Sustainable animal agriculture and crisis mitigation in
livestock-dependent systems in southern Africa
Proceedings of the regional conference held at Malawi Institute of Management, Lilongwe,
Malawi 30 October to 1 November 2000.
Edited by:
J.W. Banda, M.G.G. Chagunda, L.A. Kamwanja, R.K.D. Phoya and A.C.L. Safalaoh
March 2001
Bunda College of Agriculture, P.O. Box 219, Lilongwe, Malawi
USAID, P.O. Box 30045, Lilongwe, Malawi
GTZ, Private Bag 387, Lilongwe, Malawi
Table of Contents
Abstract
Foreword
Opening address
Preface
Session 1: Livestock management and improvement
Characterisation and conservation plan of the Sudanese Kenana cattle breed
Livestock production improvement at small scale household level in a developing country set
up: The fate of indigenous goat gene pools in Malawi
Characterisation of small scale cattle farming in two districts of the Free State Province of
South Africa
Genetic evaluation of grade, appendix and pedigree cow classes in Holstein, Jersey and
crossbred dairy breeds in Zimbabwe
Estimation of body composition in tropical sheep raised under seasonal feed supply
conditions: Prediction models
Fodder yield and nutritive value of leucaena, acaecia and calliandra at three cutting regimes
and two locations in Malawi
Effect of draft work on production and reproduction of Malawi Zebu cows
A study of the flock dynamics, marketing, and socio-economic roles of small ruminants in
crop–livestock farming systems in Malawi
Prospects of genetic improvement programmes in populations: A case of dairy production in
Malawi 
Relationship between annual rainfall oscillations and mohair production between 1935 and
1996
On-farm and on-station evaluation of growth and reproductive characteristics of pigeons in
Malawi
Strengths and weaknesses of smallholder poultry improvement programmes in Malawi
Sheep responses to different stocking rates and seasonal variation under rangeland
conditions of Lesotho
Forage responses to different stocking rates and seasonal variation under rangeland
conditions of Lesotho
Roasted sorghum as an alternative energy source in pig feeding
The effect of a community-based animal health service programme on livestock mortality,
offtake and husbandry applications — A field study in northern Malawi
Session 2: Policy analysis, socio-economics and sustainable integrated
animal health
Socio-economic and biological issues to consider in smallholder poultry development and
research in Southern Africa in the new millennium
Analysis of socio-economic factors influencing egg production in Maseru District, Lesotho
Epidemiology of parasite infestation in goats belonging to resource limited communities in
eastern Namibia
Session 3: Crisis mitigation in livestock dependant systems
Crisis mitigation in livestock dependant systems: The role and limitations of current initiatives
for livestock systems in South Africa
Crisis mitigation in livestock dependent systems: Concern, universal experiences and
challenges in promotion of livestock production in Dedza District
Session 4: Product enhancement and human resources development
Promotion of higher education: Is there a role for the traditional agricultural university in
development?
The framework for regional collaboration in animal agriculture: The SADC animal agriculture
network (SAARNET)
The challenges of training animal scientists in the SADC: Experiences from Malawi
Capacity building for sustainable use of animal genetic resources
Session 5: Short communications
Keeping quality of cow and goat milk under different temperatures in different seasons in
Malawi
Prospects for the 2000/2001 rainfall season in Malawi
The effect of variety and geographical location on the mineral content of sorghum and finger
millet in Kenya and Uganda
Milk yield, physico-chemical properties and composition of milk from indigenous Malawi goats
and their Saanen half-breeds
Session 6: Abstracts
Genetic diversity in indigenous cattle breeds of East Africa
Carcass composition of tethered, supplemented Tanzanian goats and prediction models for
carcass weight and composition
Post-weaning performance of local Malawian pigs fed soyabean-based diets containing
different levels of raw, untreated sunflower seed meal
Using carbon isotope ratios (∂13C) in determining diet changes in animals: the case of the
Lake Malawi bathyclarias nyassensis (pisces: claridae)
Characterisation of the local Zambian chicken — A case study of Chikondwa area in Chongwe
The prevalence, incidence and effects of trypanosomosis on reproductive performance of
cattle in north-eastern Zimbabwe
Constraints to rural poultry production and diversity in Lilongwe and Mzuzu agricultural
development divisions, Malawi
Village chicken production in Zimbabwe — Constraints and potential
Some socio-economic factors constraining farm animal production in Malawi
Use of maize and crotalaria leaves as supplements for tethered sheep and goats
Abstract
These proceedings contain papers, abstracts and posters of the Regional Conference on
Sustainable Animal Agriculture and Crisis Mitigation in Livestock-dependent Systems in
Southern Africa that took place at Malawi Institute of Management (MIM) in Lilongwe, Malawi,
from 30th October to 1st November 2000. A keynote paper opens each theme and is followed
by scientific contributions, short communications, and abstracts. Posters are published as
short communications and/or abstracts. These proceedings are organised as follows:
Session 1    Livestock Management and Improvement (16 papers)
Session 2    Policy Analysis, Socio-economics and Sustainable Integrated Animal Health (4
papers)
Session 3    Crisis Mitigation in Livestock-dependent Systems - Held as a mini workshop within
the conference (2 papers)
Session 4    Product Enhancement and Human Resources Development (4 papers)
Session 5    Short Communications (4 papers)
Session 6    Abstracts (10 abstracts)
Foreword
The numerous constraints to animal agriculture in Southern African Development Community
(SADC), including further limitations exerted by man-made and natural disasters or crises
necessitated the organisation of the "Regional Conference on Sustainable Animal Agriculture
and Crisis Mitigation in Livestock-dependent Systems in Southern Africa". The goal was to
map out strategies that would be used by all stakeholders in animal agriculture to improve
production to levels that will enhance and sustain food security and improve living standards
as well as mitigate crises in the region.
Previous attempts to consolidate information have been limited by lack of such a forum to
bring together scientists, farmers, processors, traders, development and extension agents,
decision and policy makers and all stakeholders from the government, NGO and private
sectors.
These proceedings publish papers, posters and abstracts based on research, development
and review work in animal agriculture as well as crisis mitigation aspects. The proceedings
review most of this work conducted not only in SADC countries, but also in other countries in
order to provide lessons for livestock improvement and crisis mitigation. The proceedings also
contain recommendations on strategies for present and future development in animal
agriculture and crisis mitigation as well as on enhancing/strengthening of collaboration, co-
operation, human resource development and exchange of information among the various
stakeholders.
Some recommendations require immediate attention and institutional reform, while others
require development of research and outreach projects and programmes that will assist in
improvement of the animal industry and mitigation of crises. The development and
implementation of such projects and programmes will depend on the understanding of the
rationale to change, and willingness and co-operation of all concerned stakeholders. The
recommendations are made on the presumption of good will, commitment and continued
support of donors and governments.
Those interested in the detailed summaries of the conference can request special copies of
the Conference Report and the report on The Mini Crisis Mitigation Workshop held during the
conference.
Opening address
Professor Gearge Y. Kanyama-Phiri
Principal, Bunda College of Agriculture
It is my humble duty as Principal of Bunda College of Agriculture, University of Malawi, to
present these proceedings to all participants and stakeholders with the hope that it will provide
consolidated information and recommendations required in the animal agriculture sector. I
would like to take this opportunity to express my sincere gratitude to the donors for their
support and contribution to the conference. I hope they will continue supporting activities,
projects and programmes that will emanate from the conference recommendations.
These proceedings are a result of the first ever international conference on animal agriculture.
As such I would like to thank all individuals or committees involved directly or indirectly in the
organisation of the conference and putting together of this document.
Preface
Domesticated livestock play a dominant role in most SADC countries through contribution to
the Gross Domestic Product (GDP) and agricultural output, nutrition, income and nutrient
recycling. Despite this, inadequate nutrition, animal diseases and parasites, poor performance
of indigenous animal genetic resources, low production systems, adverse socio-economic and
policy environment, low levels of processing, inadequately trained human resources, and
weak institutional linkages, have constrained improved productivity, particularly in the
smallholder crop-livestock systems. These have to be addressed in a holistic approach.
Further limitations to animal agriculture has been posed by man-made and natural crises such
as persistent droughts and floods, diseases and pests, food and feed shortages, and market
distortions, to list a few. Current early warning systems do not address or integrate these
livestock issues. Therefore there are no strategies or coping mechanisms to mitigate against
such crises. There is a need to address these issues especially in the new millennium.
It is against this background that this conference was organised. It drew together 74 national,
regional and international scientists, farmers, processors, traders, development and extension
agents, decision and policy makers, and all stakeholders from the government, NGO and
private sectors to map out strategies to improve livestock production to levels that will
enhance and sustain food security and improve living standards as well as mitigate crises in
Southern Africa. The objectives are to:
review research, development, extension, policy and human resource work in animal
agriculture
identify strategies for present and future development of animal agriculture in Southern
Africa
suggest crisis indicators and coping mechanisms or contingency plans for crisis
mitigation in livestock systems
enhance/strengthen collaboration, co-operation, human resource development and
exchange of information between universities, animal industries and relevant
stakeholders regionally.
The conference proceedings reflect the oral paper and poster presentations at the conference
held at Malawi Institute of Management, Lilongwe, Malawi from 30th October to 1st November
2000. The conference drew together over 80 national, regional and international scientists,
farmers, processors, traders, development and extension agents, decision and policy makers
and all stakeholders from the government, NGO and private sectors. The participants
presented more than 40 papers and 18 posters that covered the following themes: livestock
production management, policy analysis, socio-economics and sustainable integrated animal
health, crisis mitigation in livestock dependent systems, and product enhancement and human
resources development.
For the session on "Crisis Mitigation in Livestock-dependent Farming Systems", the strategies,
coping mechanisms and clear quantitative bases were determined in order to cope with
persistent crises in Southern Africa.
The conference was organised by Bunda College of Agriculture with funding from USAID,
GTZ, SACCAR and the NORAD Project at Bunda College.
The participants of the conference were drawn from the Republic of South Africa, Lesotho,
Namibia, Zimbabwe, Zambia, Tanzania, Kenya, Uganda, Ethiopia, Sudan, the United
Kingdom, Norway, Germany, New Zealand and the host country Malawi. The discussions
were extended to include formation of national animal science associations.
The conference was officially opened by the Minister of Agriculture and Irrigation Development
(MoAI), Honourable Leonard Mangulama, MP. The directors of the major funding
agencies/donors of USAID, Mr. Kiertsak Toh and of GTZ, Dr. R. Pollvogt gave remarks.
Session 1: Livestock management and improvement
Characterisation and conservation plan of the Sudanese
Kenana cattle breed
I. A. Yousif
Department of Genetics and Animal Husbandry, Faculty of Animal Production, 
University of Khartoum, P.O. Box 32, Khartoum North, Sudan
Abstract
Introduction
History of the Kenana cattle breed
Description
Population
Systems of production
Performance
Current effort for genetic improvement of Kenana
Genetic improvement and utilisation
References
Abstract
The Kenana cattle breed originated from Bos indicus or the humped cattle. Other evidence
suggests that Kenana cattle resulted from interbreeding the Nilotic Sana cattle with the short-
horn Zebu during tribal migration before recorded history. The name Kenana came from the
owners who are mainly nomadic and semi-nomadic tribes. The light blue-grey coat is the
characteristic colour of Kenana. The calves are frequently born with brown-red coats that tend
to change to permanent grey at 3–6 months of age. The horns are short and they are
relatively short in females. At maturity (five years) the average body weight of males and
females range from 300– 500 kg and 250–350 kg, respectively. The total population of
Kenana has been estimated at 3,000,000 head which represents 12% of the whole cattle
population in Sudan. The homeland of Kenana is the delta between the White Nile and the
Blue Nile; however they are also found along the western and eastern banks of those rivers as
far as Northern Kordofan State. This habitat could be described as semi-desert and desert
scrub.
The main system of production is the range system which includes the pastoral and semi
pastoral types. The nature of nomadism is relatively mild in Kenana; this is because Kenana
homeland is partially extended to an irrigated area of Elgezira Scheme. The Kenana breed
has been classified among the highest milk producing Northern Sudan Zebu Cattle. The
average milk production per lactation varies between 1400 and 2100 kg and the maximum
production is attained at 6–7 years of age. Umbanein Research Station was established in the
year 1957 with the primary objective of improving the Kenana breed for milk and beef
production through a continuous selective breeding program. Even though Kenana cattle are
described as relatively low production animals (milk and meat), however they are well adapted
to the harsh climatic and environmental conditions.
Attempts at genetic improvements of the Kenana breed have not progressed considerably.
Plans for further genetic improvement could be dependent on the establishment of a selection
strategy involving a multi-nucleus herd stage to solve the problem of the unavailability of an
organised record system. This will lead to the development of a well-established central
nucleus herd station, where intensive selection can be practised and will ultimately result in
selection of superior cows and bulls.
 Introduction
Sudan is one of the largest countries in Africa and the Near East with an area of
approximately one million square miles. The country possesses a large livestock wealth and
ranks first in Africa in total livestock numbers. There are approximately 22.5 million head of
cattle in Sudan, and about 98% of them are owned by pastorals. This large population is
mainly concentrated in three major regions, namely Western Sudan – the homeland for
Baggara cattle, mid-Sudan – the homeland for Kenana and Butana Breeds, and Southern
Sudan where the Nilotic breed is dominant.
 History of the Kenana cattle breed
The Kenana breed originated from Bos indicus or humped cattle. It has been suggested by
Payne (1970) that the origin of Zebu breeds of cattle appeared to be Western Asia and not the
Indian subcontinent. Zebu cattle were imported into Africa during recorded history through the
migratory people, thus representing the ancestor of the vast majority of cattle breeds found
today (Payne 1964). The migration of short-horn cattle owners (including the North-Sudan
Zebu) took place along the Southern Mediterranean and southwards along the Nile into Upper
Egypt and Sudan, but not across what is now the Sahara. Other evidence suggested that
Kenana cattle have resulted from interbreeding the Nilotic Sanga cattle with the short-horn
Zebu during tribal migration (Rouse 1970). Sanga cattle are a cross between the hamitic and
the long-horn Bos indicus. The name Kenana came from the owners who are mainly nomadic
and semi-nomadic tribes which include the Kenana, the Rufa Elhoy and the Beni Moharib.
There are a number of related types of North Sudan short-horn Zebu among which Kenana
synonymous to Blue Nile, Fung and Rufa'ai can be distinguished. Mason and Maule (1960)
described Kenana as a sub-type of the Northern Sudan short-horn Zebu. It is a representative
of the nomadic cattle of Northern Sudan.
Description
Figures 1 to 6 show animals of the Kenana breed of different sex and age. The characteristic
colour of Kenana is light blue-grey, with gradation from nearly white to steel grey. Shading to
nearly black is common on the head, neck, hump, hindquarters and legs. Black points are also
visible on the muzzles, horns, tail tips and eyes. The individual hairs are black at the base and
white (or occasionally red) at the tip. Darker coat colours and darker areas are due to the
hairs having a broader black band. Calves are frequently born with brown-red coats that tend
to change to permanent grey at 3–6 months of age. The hair with the red tips in calves may
apply to those on the poll only or to the whole body. The head is long and coffin-shaped with a
thin face. The horns are short, they seldom exceed 30–35 cm, and they are relatively shorter
in males than in females. In a cross section, this horn appears oval in shape. Even though
polledness is exceptional, animals with loose horns are very common. Horns grow from the flat
poll in an outward and upward direction. The hump is cervico-thoracic in position; it is large in
males and tends to hang over at the rear, but is less developed in females. The dewlap is
large and prominent in males. Because of the better frame and relatively good milk
production, Kenana cows provide their calves with a high level of nutrition up to the weaning
age. At maturity (about 5 years) the average body weights of males and females as reported
in the literature, range between 300–500 kg and 250–350 kg, respectively depending on the
managerial conditions (Saeed et al. 1987).
Figure 1. A typical mature Kenana bull.
Figure 2. Kenana cow with loose horns.
Figure 3. Kenana calf less than six months old.
Figure 4. A typical mature Kenana cow.
  
Figure 5. Polled Kenana bull.
Figure 6. Kenana calf three months old.
Population
The total population of Kenana has been estimated at 3,000,000 head, which represents
about 12 percent of the whole cattle population in Sudan. The homeland of Kenana is the delta
between the White Nile and Blue Nile. It is also spread along the western and eastern banks
of those rivers in a triangular area bounded by the cities of Sinnar, Singa, Roseries and Kosti
(Figure 7). The breed is also found in Northern Kordofan State and as far as the area of the
Abbassiya in the eastern part of Nuba Mountain. The general description of this habitat is a
semi-arid desert scrub with rainfall ranging between 336 and 457 mm. Under traditional
management, the sizes and compositions of the herds are influenced by a number of factors
such as seasonal availability of water and feed, market prices and infectious diseases.
Figure 7. Map of Sudan showing the main breeding area of the Kenana breed (dark triangle).
Generally herds under the nomadic system tend to include a high proportion of bulls than
those required for breeding purposes. It is also expected that a large sized herd will contain
more bulls at different ages than those with small sizes. For breeding purposes, owners select
the large sized bulls to serve the cows in a ratio of approximately 1:25.
Generally there is no specific breeding season; however, most of the cows conceive during
the rainy season when pasture is rich and abundant. As a result of nomadic migration, the
Kenana breed has been exposed to a degree of interbreeding with Western Sudan Zebu and
probably the nilotics, in addition to the artificial crossbreeding with exotic breeds, but still it is
believed that this process has not yet seriously endangered the original breed.
 Systems of production
The traditional grazing system is considered to be the most common upon which more than
80% of livestock owners are dependent. The range system is the main production system; it
includes the pastoral and semi-pastoral types. The pasture region in the country extends in a
wide belt between the northern desert and the southern forest for a distance of 1000 miles
from east to west and 500 miles from north to south. Annual rainfall varies from 10 inches in
the north to 35 inches in the south. Cattle owners who are mostly nomads or semi nomads
seasonally migrate with their herds and gather around the areas rich with water and grass.
Milk is sold to the nearest towns to generate cash income. Seasonal cheese factories may be
established by merchants in association with cattle owner settlements.
Nomadism of Kenana is less pronounced compared to Baggara cattle in Western Sudan,
which pass through a very long migratory process. This is mainly related to the fact that the
Kenana homeland is partially extended to an irrigated area of Elgezira scheme where a variety
of crops are grown. Therefore the post harvest crop by-products provide a better chance for
settlement during this period of time.
More than 80% of the milk sold to big cities and towns in Sudan is still provided by the
traditional sector. Most of the dairy farms include small or medium-sized herds (20–50 milking
cows), which are mainly Kenana, and its crosses with exotics. This system of production could
be defined as semi-intensive. Kenana cattle are not considered as a major source for beef in
the country.
 Performance
The Kenana breed has been classified among the highest milk producing Northern Sudan
Zebu. According to several reports (Alim 1960; El Amin 1969; and Saeed et al. 1987), the
average milk production per location varies between 1400 and 2100 kg (Table 1). The
maximum production is attained at 6–7 years of age. This is fairly comparable with other
important milk producing Zebu breeds in Africa. Results from Nisheisheiba Research Station
showed that milk fat and solid not fat (SNF) vary with lactation season in a range of 4.2–5.4%
and 8.7–9.0% respectively (unpublished data).
The Kenana bull is commonly characterised by a fairly large body size. In general, data on
reproductive and productive performance of the Kenana breed under nomadism is very
scanty; therefore the results, which were obtained from the research stations, are considered
as the closest an approximation. The average birth and weaning (120 days) weights of
Kenana at Umbanein Research Station was found to be 23.0 kg and 52.0 kg respectively
(Saeed et al. 1987), whereas the dressing percentage was 55%. Traditionally Kenana is used
for milk, meat and hide production. Bulls are also used for draft power in the migration season.
Under traditional management the seasonality is very marked, with more than 60% of all
calving taking place in a 3-month period related to the conception in the previous rainy season.
The age at first calving that was reported at different research stations ranges between 38.4
months at Gezira Farm (Alim 1960) and 50 months at Umbanein Station (Saeed et al. 1987).
The age at first calving is expected to be relatively higher under the nomadic system, though it
may not exceed the reported range at the research stations. The range for the number of
services per conception at different seasons was reported to be 1.4–1.8 (Saeed et al. 1987).
The calving interval at Umbanein Station was found to be 485 days which is longer than that
reported at Gezira Farm (368–405 days). The reproductive life defined as the number of
calves born per cow, is reported to be in the range of 4–5.4 calves.
Calf mortality data at Umbanien Station showed that it was highest during the neonatal period
of up to one week of age. From one week to weaning at 120 days, the mortality rate is
considerably lower; but again it rises after weaning. The overall mortality rate to one year was
found to be 16.6% (Saeed et al. 1987). 75.6% of all animals born survived up to ten years of
age.
Table 1. Performance of Kenana Cattle (location data from Kenana cattle herds).
Lactation Lactation Calving Annual  
Herd and source of data Number yield (kg) Length (days) interval (days) yield (kg)
 
Gezira Research Farm 539 1613 224 395 1491
 
Khartoum University 1537 1860 294 428 1568
Farm (El Amin 1969)  
Um Banein Research 1263 1497 251 479 1141
Station (1987) Source
Saeed et al.
1987.
Current effort for genetic improvement  of Kenana
The governmental effort to conserve and genetically improve the indigenous cattle breeds has
been going on since the early 1940s, which consequently led to the establishment of a
number of research centres in the areas of livestock concentration. Umbanein Research
Station was established in 1957, with the primary objective of improving milk and beef
production through selection within the Kenana breed. In the early 1960s, a multi-purpose
research centre was established at Kuku district in Khartoum North province. It includes
several animal production units, among which the A. I. centre is considered as an imminent
progress in animal breeding. At the same time the research effort which was carried on at
Khartoum and Gezira University farms indicated the probability of achieving some genetic
progress in milk production of Kenana cattle (Alim 1962) . However, because small herd sizes
and financial problems presented a genetic progress, a major goal of these research centres
is to conserve small pure-bred nucleus herds of Kenana. On the other hand the introduction of
foreign temperate breeds to improve the milk yield of indigenous cattle (Kenana and Butana) is
considered despite the fact that there is no clearly defined national crossbreeding policy.
 Genetic improvement and utilisation
The plans and suggestions for genetic improvement of the indigenous cattle breeds in many
tropical countries are more or less similar. Improvement could be achieved through two
parallel schemes; these are within population selection and introduction of improved exotic
germplasm. For practical reasons however, much more emphasis should be given to the
former scheme, whereas the latter should be implemented under certain restrictions.
The decentralised multi-nucleus herd scheme is suggested to deal with the problem of
unavailable organised record systems as a result of continuous migration under nomadism.
The selection criterion of individual cows for these nucleus herds could be based on direct
assessment of their milk yields. It is obvious that concurrency using this procedure is
questionable; however, it is useful to study the performance of Kenana. The substantial
variation in milk production of the breed is mainly attributable to the wide range of
environmental differences and the genetic variability within the breed. Furthermore, genetic
parameters which are essential for selection planning could also be estimated during this first
stage of the scheme.
In the next stage and after having established records, individuals should be selected on the
basis of their own and relative information in order to develop central nucleus herds on
station. This station must be well established and provided with means that are necessary for
intensive selection procedures. Since the primary objective is selection for milk production, it
would be worthwhile to include economical aspects of the dairy breed in the index of selection
according to their appropriate weights. Due to this effort it would be expected that few but
superior cows and bulls will be produced and through artificial insemination (AI) techniques the
semen could be disseminated to farmers. However, because of ritualistic beliefs and
traditional attitudes of the nomad's application of AI procedures in the nomadic herds may not
be welcomed. This program is designed to last for as many generations as possible to
maximise the genetic improvement for milk production. In addition, selection for both milk and
growth rate in the Kenana breed would be feasible.
In Sudan, importation of exotic breeds was initiated between 1925 and 1927 when Short Horn
cattle from the U.K and Holstein Friesian from the USA were introduced to the country.
Regarding the indispensable role of the indigenous cattle for the vast majority of the nomadic
tribes, any improvement programme which may jeopardise the existence of indigenous
species is not at all acceptable. However it should be noted that the application of a
scientifically well-planned crossing scheme under necessary restrictions should not be
considered a risky process. Moreover this kind of crossing programme may be best suited in
the areas where semi-intensive systems of production are prevalent; therefore relatively better
management would be provided. Even though the Kenana breed has not yet been exposed to
severe and risky crossbreeding efforts, it is believed that the initiation of an early conservation
scheme is strongly recommended. Among the foreign breeds imported to Sudan, Holstein
Friesian and their crosses with Kenana and Butana have shown the best performance
indicating they are comparatively better adaptable to the local environment (Fadl-Moula 1994).
 References
Alim K.A. 1960. Reproductive rate and milk yield of Kenana cattle in Sudan. Journal of
Agricultural Science (Cambodia) 55:183–158.
Alim K.A. 1962. Body reactions of Kenana cattle under natural and artificial shade in the
Sudan. Tropical Agriculture (Trinidad) 39:13–16.
El-Amin F. 1969. Environmental and genetic factors influencing reproduction and milk yield of
Sudanese indigenous dairy cattle. MVSc thesis, University of Khartoum, Khartoum, Sudan.
Fadl-El-Moula A.A.1994. Factors affecting reproductive and productive performance of
crossbred dairy cattle in Sudan. MSc thesis, University of Khartoum, Khartoum, Sudan.
Mason I. L. and Maule J.P. 1960. The indigenous livestock of eastern and southern Africa.
Tech. Comm. 14. Commonwealth Bureau of Animal Breeding and Genetics. CBA, Farnham
Royal, U.K.
Payne W.J.A. 1964. The origin of domestic cattle in Africa. Empire J. Exp. Agric. 32:97–113.
Payne W.J.A. 1970. Cattle production in the tropics. Vol. 1: Breeds and Breeding. Longman,
London, UK. .
Rouse J.E. 1970. World cattle: Cattle of Africa and Asia. University of Oklahoma Press,
Oklahoma City, USA, .
Saeed A.M, Ward F.A., Light, D., Durkin, J.W. and Wlisom R.T. 1987. Characterisation of
Kenana cattle at Um Banein, Sudan. ILCA Research Report 16. Addis Ababa, Ethiopia.
Livestock production improvement at small scale household
level in a developing country set up: The fate of indigenous
goat gene pools in Malawi
A. J.D. Ambali,1 J. W. Banda2 and L.A. Kamwanja2
1. Molecular Biology and Ecology Research Unit Department of Biology, University of Malawi,
Chancellor College, Box 280, Zomba, Malawi
2. Bunda College of Agriculture, Lilongwe, Malawi 
Abstract
Introduction
Genetic improvement of the local Malawi goat
Reflection on the MGLDP's goat improvement programme
Status of the indigenous goat gene pool in Salima district
The way forward
Acknowledgements
References
 Abstract
A goat breed improvement programme was initiated at Lifidzi Goat Breeding Centre in Salima
District in the mid 1980s with the objective of improving meat production and animal
population at the household level through crossbreeding. An exotic Boer breed was introduced
from South Africa and crossbred with the local Malawi goat. Three main breeding strategies
emerged in the area covered, namely stud breeder programme, buck centre programme and
uncontrolled breeder mating system. While the first two programmes were controlled at the
beginning, the last method was not controlled once the buck had been purchased from a stud
breeder. This study was carried out to determine the genetic impacts of the programme with
the aim of making recommendations for sustainable programmes that would increase
production without compromising conservation of indigenous genetic resources. Between
1985 and 1998, goat populations have increased in Salima District from less than 82,210 to
149,930 due to deliberate selection for prolific indigenous females into the breeding
programme. Growth rates of the hybrid goats improved and the offspring easily attained
weaning weights of at least 20 kg in 12 months. The number of hybrids in the communal
grazing areas visited, proliferated with more than 40% of the animals being Boer and local
crosses. The improvement programme was however implemented without proper mechanisms
for conserving the indigenous gene pools. In a setup where animals are subjected to
communal grazing and uncontrolled mating, it has been impossible to maintain pure breeds of
indigenous goats. Farmers have recently started to resent the hybrids due to low disease
resistance and yet chances of re-establishing pure populations of local breeds seem to be low.
The farmers and the animals have been abandoned the introduced Boers have succumbed to
diseases due to lack of veterinary services. The Malawi case therefore serves as a typical
example of the genetic improvement programmes in developing countries where increased
production through importation of exotic breeds is being carried out without proper planning for
sustainable conservation of indigenous gene pools.
 Introduction
Ruminant livestock contribute substantially to the animal protein supply of Malawi where a total
ruminant population of 2,319,000 was estimated in 1998 (MoAI 1998). This comprised
618,800 cattle, 1,597,500 goats and 102,700 sheep. Of the cattle population, 95.2% is owned
by small-scale rural farmers and the rest is shared by large-scale commercial operators in
estates and peri-urban areas (MoAI 1998). About 97% of the herd comprise the Malawi Zebu,
which is a naturalised indigenous breed in the country comprising several phenotypes with
varied origin (Butterworth and McNitt 1994; Jere and Msiska 2000). A similar trend is also
observed in the other two ruminant species where 98% and 95% of goats and sheep,
respectively, are owned by small-scale rural farmers. The sheep and goats constitute the
small ruminant livestock of the country, of which more than 90% are indigenous breeds raised
by small-scale farmers at an average of six animals per household distributed in the low-lying
areas (Jere and Msiska 2000).
It is observed in Figure 1 that during the period 1987 and 1996 the sheep population declined
and did not reach the 1983–85 population levels. The decline in sheep population was due to
high demand for mutton/lamb in the urban areas and parasite infestation (Chikagwa-Malunga,
unpublished). The goat population however picked up from the 1987-drop and maintained
annual population levels of more than 82,000 animals. In the early 1990s the flock growth rate
was estimated at 1.8% annually, contributing about 3,700–4,400 tonnes of meat per annum
which represented about 11% of the total livestock meat consumed in the country (Banda et
al. 1993). Banda (1992) observed that the rural households in Malawi readily accept goat and
sheep milk, only that it is not available due to low production. This paper appraises the genetic
improvement efforts that have been carried out on goats in the country, with the most
emphasis on the Malawi German Livestock Development Project.
 Source:
MoAI (1998).
Figure 1. Goat and sheep population in Malawi, 1983–96.
 
Goats are usually tethered and herded during different times of the year. Tethering is practised
during the rainy season when most arable land is subjected to intensive cultivation and little
area is available for grazing. Tethering limits the amount of feed available to the animals and
as a result they lose weight and kid birth weight also reduces. However, this does not
significantly affect reproductive performance (Banda et al. 1993). Herding is practised during
the dry season when crops have been harvested. The alternative system involves tethering
during the morning hours and herding in the afternoon. Those who feed their animals in
communal grazing areas tend to expose their goats to an extensive system where goats
browse, graze and scavenge.
Mortality tends to be high in kids primarily due to poor management, which include
overcrowding, poor nutrition, parasites and diseases. Effects of these have been discussed in
Banda et al. (1993).
The local goats in Malawi are naturalised indigenous stock that originated from Asia and
spread southwards from Egypt and Ethiopia to East Africa and then Malawi (Epstein 1971).
There are speculations that they could also be some stock that came in from the south by the
Zulu people (Banda et al. 1993). These collections have adapted to harsh climates and low
management regimes over the years. Although the local goats are adapted, they generally
have been described as low producers compared to exotic breeds, hence efforts have been
carried out to improve their performance using genetics and improved management
approaches.
 Genetic improvement of the local Malawi goat
The Government of Malawi has carried out a two-phase genetic improvement strategy for
goats in the country. The first was through introduction of an exotic breed, Boer goat, at
Mikolongwe Veterinary Station and Bunda College of Agriculture for crossbreeding and
upgrading of the indigenous goat to improve the final slaughter weights.
The F1 hybrids grew faster than the local goat progenies (Figure 2). Average weight gains of
40g/day and 23g/day were observed in the crossbred progenies and pure local progenies,
respectively. Under village conditions in Lilongwe district, Banda et al. (1993) reported results
where pre-weaning body weights were higher in the local progeny than in the crossbred (L ×
B) with daily weight gains of 110.0 and 90.5 g/day, respectively. Mortality was also lower in the
local progenies than in the crossbred with 16.7 and 37.5% mortality, respectively. Banda
(1992) argues further that overall, the biological productivity rate is higher in the local progeny
than in the crosses, which probably suggests that the local gene pool was more efficient than
the exotic breeds and their crosses under village conditions.
 
Source: Modified from Zerfas (1992).
Figure 2. Growth comparison between pure local (L) and crossed local × Boer (1/2L × 1/2B)
kids.
The second genetic improvement programme was carried out from 1983 to 1993 by the
Malawi Germany Livestock Development Programme (MGLDP) with the objectives of
producing genetically improved local bucks and does for smallholder farmers; producing pure-
bred Boer bucks for small holder crossbreeding activities; carrying out breed evaluation trials;
and packaging goat management innovations for farmers (Schmidt 1992).
The programme was phased out in 1993 and currently there are three sets of breeding
strategies in the project areas: namely the Stud Breeder Programme, the Buck Centre
Programme, and the Uncontrolled Breeder Mating System. The first two programmes were
controlled during the project period. The last strategy was not controlled once the buck had
been purchased from a stud breeder. In the stud breeder breeding strategy, 30 farmers were
provided with 10 hybrid animals (7/8B × 1/8L) comprising nine females and one male for rapid
multiplication at the farmstead. The progenies from the stud breeders were sold to other
farmers who wanted to cross their local goats with the hybrids. The male goats provided in the
stud breeder package were supposed to be exchanged among the stud breeders after every
year, hence there were 30 bucks to be exchanged among the stud breeders.
The buck centre strategy involved farmers with local goats bringing their female goats to buck
centres for mating with a Boer buck. Bucks in excess of 100 were placed in the villages in
Salima District while the local males were castrated. This resulted in the proliferation of hybrids
in the villages.
The uncontrolled breeder mating system was the long existing system practised by the farmers
themselves whereby mating was not controlled in the communal grazing areas or in the
kholas. The males in this case would be local and/or hybrids in areas where the crosses were
available. Male crosses bought from stud breeders mated with the local females in communal
grazing areas.
Results from the improvement programme generally suggest that the hybrids performed better
than local goats as summarised below.
The local goat and hybrid progeny performance over 360 days is plotted in Figure 3. The local
goat progeny had the lowest weight up to 300 days. Backcrossing the progenies to Boer
yielded an improvement in growth to the effect that they easily attained 20 kg in 360 days.
Similar observations were made by Schmidt (1992) who reported the highest growth rates in
crossed progenies than in the local goat progenies at the age of 12 months. A high weight
variation was however reported in the crossed progenies with changes in environmental
conditions (Figure 4), implying that the local goats were relatively more stable to adverse
environmental regimes than the hybrids.
Source: Zerfas (1992).
Figure 3. Weight of local and hybrid goat progenies at different ages.
 
Parturition intervals are plotted in Figure 5. Kidding interval was lowest among the hybrid
progenies than in the local goats. The decline in kidding interval with parity was however
higher in the local goats than in the hybrids.
Source: Schmidt ( 1992).
Figure 5. Parturition interval among the various goat progenies.
 Reflection on the MGLDP's goat improvement programme
The programme might have achieved its objectives whereby it demonstrated that genetic
improvement through cross breeding improved the performance in goats for the traits of
interest. There are however several factors in the programme which were not addressed and
hence could affect the validity of the findings.
The local goat that was compared to the hybrids and the Boer in this programme is not
genetically known. There are several phenotypes of goats in Salima district let alone the
Malawi nation as a whole. The MGLDP did not genetically characterise the local goats though
the Boer that was imported into the country was a well characterised breed. The data by
Ayoade and Kamwanja (1985) in Figure 6 suggests that the local goat performed better than
the hybrid in Lilongwe. It can therefore be argued that there may be several types of local
goats in Malawi hence, if the various types had been identified and compared to the crosses
the results could probably be different from what was obtained by the Project. A hypothesis
can be made that not all local Malawi goat types were inferior in performance to the hybrids.
This is yet to be addressed.
Source: Ayoade and Kamwanja ( 1985).
Figure 6. Pre-weaning body weight of local and hybrid progenies at household level
in Lilongwe district.
If indeed there were no genetic differences between the local goats in Salima and those in
Lilongwe, there might have been significant genotype × environmental interaction. The Project
did not explore the genotype × environment interaction hence the hypothesis that the
superiority of hybrid goats over local goat progenies throughout the country was not tested.
Salima district was the only locality that did not represent the broad environmental conditions
under which goats are raised in Malawi.
The local goats used in the breeding programme were not subjected to any prior selection
before they were crossed with the Boer goats or when they were being compared to the
hybrids. The data on parturition interval suggests that there was high variation among the local
goats, which could be exploited through deliberate selection. The Boer was a highly selected
breed while the local goats were not selected.
The environmental conditions under which the goats were raised within Salima were not
uniform. Zerfas (1992) indicates that there was a tendency among the farmers who raised
hybrid goats to invest relatively more in goat enterprise than those who raised local goats. In
fact the criteria for choosing farmers who raised hybrids were such that the poor farmers were
left out. Indeed even among the farmers, there were likely to have been environmental
differences between them.
There was conflict of interest between farmers' goals and those of the Project. Small-scale
farmers are primarily interested in numbers of goats and such traits as hardness and twinning
ability; growth rate is a secondary trait. The Project was interested in improving growth rate.
Farmers' interests could have been met by first selecting for the traits among the local goats
and then growth rate would be improved by either selecting among the local goats or
crossbreeding with the Boer. Unfortunately this oversight has had a negative impact on the
farmer acceptability of the hybrids especially now that veterinary services have been phased
out. In 1995, when farmers were offered to buy the bucks that were used for buck centre
breeding, they declined because the bucks were not hardy enough to survive in the local
conditions; they easily succumbed to diseases and parasites.
The statistical analyses carried out were also questionable especially regarding the sample
sizes. Banda et al. (1993) cautions on the conclusions that could be drawn out of the Ayoade
and Kamwanja (1985) study where the local goat outperformed the local × Boer progeny due
to the small sample sizes. The data obtained by the MGLDP has similar problems where there
were large sample sizes for local goats but the sample sizes for the hybrids were low. Indeed
Schmidt (1992) expressed similar concern. The statistical implications of small samples for
experiments of this nature could be enormous.
 Status of the indigenous goat gene pool in Salima district
Although the hybrids might have performed better than local goats, their poor survival under
low management regimes renders them unsuitable for the small-scale farmers in Salima and
probably Malawi as a whole. The uncontrolled mating in the communal grazing areas has
unfortunately led to high introgression of the Boer gene pool into the salima goat population.
Hybrids of other exotic breeds like Sanean and Damascus were also observed during the
study. Among the areas visited, there was a proliferation of Sanean × local hybrids in the
Ngozi area where a German researcher introduced the breed and abandoned it after trials.
Dispersive genetic effects observed in 13 mixed breeds of north European cattle over 40
years suggest that Nordic breeds have lost between 1–11% of their genetic variation.
Although the native breeds have retained a reasonably high genetic diversity, their genes
contribute less to the genetic variation of the popular production breeds than was previously
the case (Kantanen et al. 1999). It is not known at present how much of the genetic diversity in
the Salima is contributed by the indigenous breed types. It could probably be low.
Poor record keeping, uncontrolled mating and lack of genetic conservation protocols might
have led to decreased genetic purity of the Malawi local goat. In addition these unwitting cross
breeding programmes may result in outbreeding depression where the local co-adaptive gene
complex breaks down after the F1 generation. For instance in intertidal copepods (Tigriopus
californicus ), Edmund (1999) observed that first generation hybrids showed some heterosis
which was not related to the geographic or genetic distance separating the parents. However,
mean fitness decreased and its variance increased in subsequent generations when the F1
hybrids mated among themselves. This hybrid breakdown problem in the later generations
increased with the distance between the original parents. Genetic interpretation of these
patterns suggests that both the beneficial effects of dominance and the detrimental effects of
breaking up coadaptation are magnified by increasing evolutionary distance between
populations. It is not known at present how genetically distant the parents (i.e. Boer, Sanean,
Damascus etc.) are that were used to cross breed with the Malawi goat. This has gone
unchecked for several years.
 The way forward
The Malawi case therefore serves as a typical example of the genetic improvement
programmes in developing countries where increased production through importation of exotic
breeds is being carried out without proper planning for sustainable conservation of indigenous
gene pools. The continuous loss of the indigenous gene pool in Salima district can be arrested
by discouraging farmers from continuous exchange of the hybrid goats. Non-governmental
organisations like World Vision International have been involved in the transfer of hybrid goats
to other districts like Dowa, Mchinji and Karonga. They should also be discouraged from such
practices. There is a need for genetic characterisation of the Malawi local goat so that the
various types that are available, if any, can be documented, improved and conserved as such.
Genetic improvement should start with exploiting additive genetic variation through selection.
The traits of importance should be those in which farmers are interested.
The local goat gene pool that was collected by the MGLDP was lost due to lack of
maintenance of in situ conservation facilities after the project was phased out. It is
recommended that the government and quasi-government institutions should complement
farmers' efforts of conserving indigenous gene pools by establishing in situ conservation
facilities for goat genetic resources and providing continued financial support for such
facilities.
 Acknowledgements
We are sincerely grateful to the Lake Malawi Ecology Project for providing resources for
carrying out this study. Additional funding was provided by the research grants of Dr J.W.
Banda and Prof L.A. Kamwanja. We thank the Salima ADD staff, especially Messrs. Nyoni,
Njoka and Nkochi for providing some of the information included in the report and all the
farmers we interviewed during the study.
 References
Ayoade J.A. and Kamwanja L.A. 1985. A note on the pre-weaning body weight changes and
mortality in Malawi local and Malawi local × Boer kids under village management system.
Paper presented at the 5th Veterinary Association Conference, September 1985, Salima,
Malawi.
Banda J.W. 1992. Acceptability of goat's, sheep's and cow's milk in Malawi. Journal of
Consumer Studies and Home Economics 16:129–138.
Banda J.W., Ayoade J.A., Karua S.K. and Kamwanja L.A. 1993. The local Malawi goat. World
Animal Review 74/75(1–2):49–57.
Butterworth M.H. and McNitt J.I. 1994. The Malawi Zebu. World Animal Review 49:6–
12.Edmund S. 1999. Heterosis and outbreeding depression in interpopulation crosses
spanning a wide range of divergence. Evolution 53:1757–1768.
Epstein H. 1971. The origin of the domesticated animals of Africa. Volume 2. African
Publications, New York, USA, .
Jere J. and Msiska H.D.C. 2000. Current status on the management and utilisation of farm
animal genetic resources in Malawi. Paper presented at farm animal genetic resources
workshop, Natural Resources College, Lilongwe, Malawi. March 2000.
Kantanen J., Olsaker I., Adalsteinsson S., Sandberg K., Eythorsdottir E., Pirhonen K., and
Holm L.E. 1999. Temporal changes in genetic variation of North European cattle breeds.
Animal Genetics 30:16–27.
MoAI (Ministry of Agriculture and Irrigation). 1998. National Livestock Development Master
Plan. Agricultural Sector Programme Support, Livestock component, Malawi. Draft Final
Report. Department of Animal Health and Industry, Lilongwe, Malawi.
Schmidt U. 1992. Goat performance at Lifidzi breeding station. Paper presented at the
workshop on small ruminant activities in Malawi and the contribution of MGLDP, Lilongwe,
November 1992.
Zerfas H.P. 1992. Results of MGLDPs field activities in Salima RDP. Paper presented at the
workshop on small ruminant activities in Malawi and the contribution of MGLDP, Lilongwe,
November 1992.
Characterisation of small scale cattle farming in two districts
of the Free State Province of South Africa
L. E. Moorosi., L.M. J. Schwalbach. and J.P. C. Greyling
Department of Animal Science, University of the Orange Free State, P.O. Box 339,
Bloemfontein 9300, South Africa
Abstract
Introduction
Materials and methods
Results and discussion
Conclusion
References
 Abstract
This study was conducted in two districts in the southern part of the Free State. A total of 510
small scale cattle farmers from Thaba Nchu (255 farmers) and Botshabelo (265 farmers) were
individually interviewed and responded to a questionnaire designed to assess the farming
characteristics and management practices in both areas. The most important farming
characteristics and management practices as well as the productivity and the sustainability of
these farming systems were evaluated and the most important production constraints were
identified.
Results showed that most farmers in both areas own less than 10 head of cattle (65.8% in
Thaba Nchu and 86.5.4% in Botshabelo). The total cattle herd structure in both areas
reflected major imbalances, with breeding females making up the largest proportion of the total
herd accounting for 60.4% and 67.1% of the total herds in Thaba Nchu and Botshabelo,
respectively. The proportion of heifers and calves in both areas were very low, 10.4% and
7.7% for heifers and 26.9% and 24.2% for calves in Thaba Nchu and Botshabelo,
respectively.
The proportion of bulls and oxen in the total number of animals kept in both areas showed
major differences in the two areas. Thaba Nchu and Botshabelo were farmed with 2% and
8.1% bulls and 10.7% and 0.6% oxen, respectively. The bull to cow ratio was found to be
extremely low in Thaba Nchu (1.46%). In contrast, Botshabelo had far more bulls than
needed. However very few farmers own mature bulls (9.6% for both areas) and the use of
communal bulls is a very common practice.
Most farmers keep dual-purpose (meat and milk) crossbred cattle. Brown Swiss type animals
are the preferred cattle breeds in these areas, where almost all farmers milk some of their
cows (94.5% and 96.6% in Thaba Nchu and Botshabelo, respectively) exclusively for home
consumption.
Cattle management practices in these two areas are considered very unproductive, with low
support services available for farmers. The nutritional management is characterised by
communal grazing, overstocking, and degradation of the natural resource base — the veld.
Estimated calving rates are low (54% and 41% for Thaba Nchu and Botshabelo respectively),
and are much lower than perceived by the respondent farmers. The estimated intercalving
periods are extremely long (676 and 890 days for Thaba Nchu and Botshabelo respectively).
The mean total milk production per farmer per day was recorded as 4.15 litres and 3.46 litres
in Thaba Nchu and Botshabelo, respectively.
The adoption of basic disease control measures such as vaccination and dipping against
external parasites by small-scale cattle farmers differed considerably between the two areas
studied, the differences being due to the availability and accessibility to extension services
and veterinary products. In Thaba Nchu 95.3% and 96.5% of the cattle farmers vaccinate and
dip their animals, respectively, in contrast with only 5.4% and 4.2% in Botshabelo. The major
production constraints identified were poor management practices and inefficient farming
support services. The sustainability of these small-scale productive systems for the long term
is not guaranteed under the present farming conditions.
 Introduction
Small scale cattle production is extremely important for the economy of developing countries in
general, and to Southern African countries in particular. Cattle farming is an important part of
the African culture and a means of capital investment or wealth accumulation (Kalunda 1996).
Besides its direct benefits, cattle can sustain the household by providing meat and milk. The
value of cattle in providing rural transportation, draft power for cultivation, and manure for crop
and fuel production are functions not to be underestimated (FAO 1997). Although small-scale
cattle production (mainly for subsistence purposes) is practised in most Sub-Saharan regions,
its productivity is considered to be very low and in most cases insufficient to ensure food
security and adequate financial returns.
Traditional small-scale cattle production systems in South Africa have not received adequate
attention regarding land rights, extension support services, access to credit and market
opportunities. These constraints have however been recognised by the National Department
of Agriculture (1998), that has reported poverty in the rural areas of South Africa to be
associated with poor agricultural policies.
Very little is known regarding the characteristics of these small-scale farming systems.
Therefore, to facilitate the policy makers to introduce appropriate policies and support services
to assist these farmers and to fulfil the present government objectives, characterisation of
small-scale farming systems in South Africa should be prioritised in the agricultural research
programmes.
The objective of this paper was thus to characterise small scale cattle farming systems in the
Thaba Nchu and Botshabelo areas, to evaluate their sustainability, and to identify some
constraints limiting their productivity. This can be useful for policy makers and extension
services to improve the productivity and sustainability of these farming systems.
 Materials and methods
This study was conducted over a period of 20 months in two districts of Bloemfontein, namely
Thaba Nchu and Botshabelo. Twenty-one of the 40 agricultural villages known as 'trusts' in
Thaba Nchu were randomly selected. Botshabelo, situated 10 km from Thaba Nchu, is divided
into 17 general blocks or sections of which ten were selected at random. A questionnaire with
28 questions was designed to characterise the farming systems used in these areas. During
the survey 255 and 265 cattle owners were individually interviewed in Thaba Nchu and
Botshabelo, respectively. This brings the total number of cattle owned by the farmers
interviewed to 2694 and 1907 in Thaba Nchu and Botshabelo, respectively.
 Results and discussion
According to the results obtained during the survey, the majority of the farmers in both areas
own less than 10 head of cattle, with the higher frequency distribution of 5 –10 head of cattle
per farmer. Similar results were reported by Dreyer et al. (1999) in the same districts of the
Free State Province. The mean herd size in Thaba Nchu was 10.8 cattle, which is significantly
higher (p<0.01) than the mean herd size in Botshabelo (7.2 animals per cattle farmer). The
significant difference in mean herd size between the districts of Thaba Nchu and Botshabelo,
and the fact that in general, the cattle herds in Thaba Nchu were larger than those in
Botshabelo, may be explained by the historical and socio-economic differences experienced in
the two areas. Thaba Nchu was established as an agricultural area for small-scale black
farmers for a much longer period of time. Farmers in this area also have a longer tradition of
cattle farming, a relatively larger grazing area, and better support services. On the other hand,
Botshabelo was more recently established and developed from an illegal squatter camp for
unemployed people who owned farms in areas designated for white people and others that
were evicted from white farms.
The herd structure and the percentage of farmers farming with different cattle classes (Tables
1 and 2) clearly show that breeding females make up the largest group of the total herd in
both areas (accounting for 60.4% and 67.1% of the total herds in Thaba Nchu and
Botshabelo, respectively).
Table 1. Distribution of herd size per cattle farmer.
 Thaba Nchu Botshabelo
Herd size distribution Number of farmers (%) Number of farmers (%)
<5 67 (26.2%) 77 (29.1%)
5–10 101 (39.6%) 152 (57.4%)
11–20 66 (25.9%) 31 (11.6%)
21–50 18 (7.1%) 5 (1.9%)
>50 3 (1.2%) –
Mean herd size ± SD 10.8 ± 11.5 a 7.2 ± 4.8 b
Total number of farmers 255 265
a, b - Rows with different superscripts differ significantly (p<0.01).
Table 2. Cattle herd structures in the Thaba Nchu and Botshabelo areas.
 Thaba Nchu Botshabelo
Herd Structure (distribution) Number (%) Number (%)
Old Cows (>10 years ) 1316 (48.8%) 890 (46.7%)
Young Cows (3–10 Years ) 32 (1.2%) 242 (12.7%)
Heifers (1–2 Years ) 279 (10.4%) 146 (7.7%)
Calves (<1Year ) 725 (26.9%) 461 (24.2%)
Young Bulls (2–5 Years ) 44 (1.6%) 29 (1.5%)
Mature Bulls (>5 Years 10 (0.4%) 126 (6.6%)
Oxen 288 (10.7%) 13 (0.6%)
Total number of animals 2694 1907
Total LSU equivalent 1970 1503
These herd sizes and structures are in agreement with those reported by most authors
investigating similar small-scale farming systems in Southern Africa. Swanepoel and De
Lange (1993) and Muchena et al. (1997), found herd size to be a critical factor determining
herd production efficiency. Nthakheni (1993), postulated that the smaller the herd, the less the
chances were of making a living out of livestock farming. Sieff (1999) reported breeding
females compose 70% of the cattle herds in Tanzania, while in the Southern part of
Mozambique, this group represented 54% of the total herd size (Rocha et al. 1991). In this
study an extremely high percentage of farmers (99%) keep old cows (>10 years).
This clearly reflects the objectives of keeping cattle for milk production and reproduction. The
same tendency was reported by Rocha et al. (1991) and Nthakheni (1993). These old cows,
although less productive, still produce milk and calves, which the farmers prefer from the
production system. On the other hand, productivity is not the main issue for these farmers, as
cattle numbers and not necessarily the quality or productivity of the animals reflects the wealth
of the traditional small-scale African farmer (Magadlela and Kadzere 1996). The proportion of
young cows and heifers from the total cattle herds and the percentage of farmers with young
cows and heifers in their herds in both areas studied are very low. This can limit both
productive (milk and weaner calves) and reproduction (calving percentage) rates in these
herds. The low percentage of calves maintained in these herds (26.9% and 24.2% in Thaba
Nchu and Botshabelo, respectively), confirm the low reproductive rates of these herds.
Furthermore the herd structures, particularly regarding the female breeding classes, put at risk
the long-term sustainability of these cattle farming systems.
The shortage of mature bulls is clearly demonstrated by the low percentages of these animals
in the Thaba Nchu (0.4%) and Botshabelo (6.6%) herds with only 2.4% and 7.2% of the
farmers having mature bulls in their herds, respectively. These tendencies are also reported in
Mozambique by Rocha et al. (1991) and in Tanzania by Sieff (1999), who have reported 6.1%
and 3.5% bulls in the cattle herds, respectively. The fact that fewer farmers own bulls in this
survey could be attributed to the communal grazing practice which allows bulls of some
farmers to run together with the cows of the whole village. These bulls are usually referred to
as 'communal bulls' and many farmers do not see the need of owning their own bulls,
especially if the neighbours own a bull. This practice, although economically justifiable, puts an
extra load on the bulls. In many cases the bull to cow ratio is not considered. This problem is
aggravated by the fact that most of these communal bulls are usually not tested for fertility and
venereal diseases or vaccinated against common diseases. The low number of bulls and
eventual fertility problems may affect the whole herd without being detected, limiting total herd
productivity.
In Thaba Nchu oxen constitute 10.7% of the cattle herd, which is much higher than the 0.6%
for the herds in Botshabelo. These differences are also seen as the percentage of farmers with
oxen in their herds, (42.3% in Thaba Nchu and 4.2% in Botshabelo). A high percentage of
oxen in the herd structure is indicative of the use of animal traction practices such as
ploughing and transportation of water and agricultural produce (Blench 1987). A possible
reason for low number of oxen in the herd structure of Botshabelo may be the fact that there is
very little arable land available.
The estimated conception rate (CR) and inter-calving period (ICP), based on the herd
structure, number of potential breeding females and the number of calves, are set out in Table
3. According to calving rate and ICP estimations, the calving rate of 54% for Thaba Nchu and
41% for Botshabelo is much lower than that perceived by the local cattle farmers. The
intercalving periods of 676 and 890 days for Thaba Nchu and Botshabelo respectively, is
much longer than perceived by the farmers. These estimations do not take into consideration
the mortality rates, as no records are available. However, even if these records were available,
it is unlikely that the mortality rates could account for such a high difference in the perceived
and estimated ICPs and CRs. The calving rates estimated in this study are in agreement with
the results of Dreyer et al. (1999), who have estimated a calving rate of 45.5% for Thaba Nchu
and Botshabelo combined. The low reproductive rates in Thaba Nchu and Botshabelo can be
explained by a combination of management-related factors such as poor nutrition and
unbalanced herd structures.
Table 3. Estimated calving rates and intercalving period for small-scale farmer in Thaba Nchu
and Botshabelo.
Areas Potential breeding Number of calves Estimated calving Estimated ICP
females rate (%) (days)
Thaba Nchu 1384 725 54% 676 days
Botshabelo 1132 461 41% 890 days
Milk production is one of the main reasons why small-scale farmers in the Thaba Nchu and
Botshabelo districts keep cattle. The high percentage of mature cows, as well as the high
percentage of dual-purpose and crossbred predominantly containing dairy genes, are strong
indications that as in most of the Southern African regions, milk production is a major reason
for cattle farming (Dugmore et al. 1996). Milk production is exclusively for home consumption.
The total daily milk production per farmer in the areas surveyed was generally low (Table 4).
Farmers were unable to relate to the individual daily milk production of their animals, as no
records and measurements are taken regarding milk production.
Table 4. Mean milk production (± SD) for small-scale cattle farmers in Thaba Nchu and
Botshabelo.
Areas Farmers milking cows Farmers not milking cows Mean total milk
production/farmer/
Day
Thaba Nchu 241(94.5%) 14(5.5%) 4.15 ± 1.3a
Botshabelo 256(96.6%) 9(3.4%) 3.46 ± 2.7b
ab columns with different superscripts differ significantly (p<0.01).
Application of disease control measures by cattle farmers in the Thaba Nchu and Botshabelo
areas as shown in Table 5, differ considerably. The poor adoption of basic disease control
measures such as vaccination against diseases and dipping against external parasites in
Botshabelo can be explained by lack of basic farming infrastructures such as crush pens and
dipping tanks in this area. Although extension services still operate from Thaba Nchu, farmers
and extension staff are most of the time faced with transport problems.
Table 5. Average application of health control measures in both cattle herds.
Area Number (%) of farmers adopting
 Vaccination Dipping
Thaba Nchu 243(95.3%) 246(96.5%)
Botshabelo 14(5.3%) 11(4.2%)
Farmers in Thaba Nchu have enjoyed free extension programmes offered over many years by
AGRICOR. Most of these farmers have realised the advantages of disease control measures
and are still implementing programmes even after the government stopped these free
services. These farmers now buy their own vaccine and dipping remedies. The main reasons
for low implementation levels of disease control measures are general lack of funds by most
governments, and the inability to support and maintain an efficient operational extension
service (Bembridge 1984; Nthakheni 1993; Tambi et al. 1999).
Farmers reported that cattle are sold whenever cash is needed. This usually happens around
January every year, when school fees for their children are needed. The price is usually
determined by the condition of the animal, although due to lack of a marketing infrastructure
such as transport, farmers feel that they are forced to submit to prices offered by the buyer.
The proportion of farmers buying cattle is very low. Only 5.9% in Thaba Nchu and 2.6% in
Botshabelo were reported as buying cattle. The reason is due to lack of access to credit to buy
more animals. Farmers usually depend on their existing animals to reproduce and increase
their herd size.
 Conclusion
The general management and farming practices adopted by small-scale cattle farmers in
Thaba Nchu and Botshabelo are still done in a traditional and unproductive manner. The
general farming characteristics and management practices of small-scale cattle farming
activities in Thaba Nchu and Botshabelo districts of the Free State Province are characterised
by communal grazing, overstocking of the veld, small and unbalanced herds, uncontrolled
breeding and crossbreeding, poor reproductive performance, natural weaning, low milk
production, and lack of basic disease control practices. In general, these characteristics are no
different from those seen in other regions of Southern Africa. There are however some
particularities in Thaba Nchu and Botshabelo, which are determined by the farming conditions
and the historical background of these two farming areas. Farmers from Thaba Nchu are in
general more experienced, have a better farming infrastructure, and have more access to
farming support services.
The main reasons for cattle farming are milk production for home consumption and it is a
means of capital savings. The low productivity and the progressive degradation of the basic
resource—the veld— seriously put at risk the long-term sustainability of these farming
systems.
 Key words: Small Scale Cattle Farmers, Farming Characteristics, Management Practices,
Sustainability
 References
Bembridge T.J. 1984. A systems approach study of agricultural development problems in
Transkei. PhD thesis, University of Stellenbosch, South Africa.
Blench R.M. 1987. Social determinant of animal traction in central Nigeria. Report submitted
to Agricultural Research Unit, The World Bank, Washington, DC, and International Livestock
Centre for Africa, Kaduna, Nigeria. 82 pp.
Dugmore T.G., Oosthuizen, P.A. and Du Toit J.F., 1996. Strategies for dairy production in
rural areas. Proceedings  2nd All Africa Conference on Animal Agriculture, 1–4 April 1996.
Pretoria, South Africa.
Dreyer K., Fourie L. J. and Kok D.J. 1999. Assessment of cattle owner's perceptions and
expectations and identification of constraints on production in a peri-urban resource poor
environment. Onderstepoort. Journal of Veterinary Research 66:95–102
FAO (Food and Agriculture Organization of the United Nations). 1997. Agriculture, food
security and nutrition. Agriculture, food and nutrition for Africa. FAO, Rome, Italy, pp. 5–46.
Kalunda M. 1996. The role of livestock in the socio-economic development of Uganda.
Proceedings  2nd All Africa Conference on Animal Agriculture, 1–4 April, Pretoria, South
Africa. pp 1–5.
Magadlela M.A. and Kadzere C.T. 1996. Dualism of cattle production in Southern Africa.
Proceedings 2nd All Africa Conference on Animal Agriculture, 1–4 April 1996. Pretoria, South
Africa.
Muchena M., Presse J., Thirtle C. and Townsend R.F. 1997. Herd size and efficiency on
mixed crop and livestock farms: Case studies of Chiweshe and Gokwe, Zimbabwe. Agrekon
Vol. 36, No.1:46–64.
Nthakheni D.N. 1993. Productivity measures and dynamics of cattle herds of small scale
producer in Venda. MSc Theses, University of Pretoria, South Africa.
Rocha A., Starkey P. and Dionision A.C. 1991. Cattle production and utilisation in small holder
farming systems in Southern Mozambique. Agricultural Systems 37:53–75.
Sieff, D.F. 1999. The effects of wealth on livestock dynamics among the Datoga pastoralists of
Tanzania. Agricultural Systems 59:1–25.
Swanepoel, F.J.C. and De Lange, A.O. 1993. Critical determinants for successful small scale
livestock production. Livestock Farmer Support Programme Workshop, Venda Agricultural
Training Centre, 22–23 March 1993. Venda, South Africa.
Tambi N.E., Mukhebi W.A., Maina W.O. and Solomon H.M. 1999. Profit analysis of livestock
producers' demand for private veterinary services in the high potential agricultural areas of
Kenya. Agricultural Systems 59:163–179.
Genetic evaluation of grade, appendix and pedigree cow
classes in Holstein, Jersey and crossbred dairy breeds in
Zimbabwe
S.M. Makuza, V. Muchenje and S. Chiyanike 
Department of Animal Science, University of Zimbabwe
P.O. Box MP 167 Mount Pleasant, Harare, Zimbabwe
Abstract
Introduction
Objectives
Materials and methods
Source of data
Data editing
The animal model
Results and discussions
Conclusions
Recommendation
Acknowledgements
References
 Abstract
Genetic parameters of milk yield of pedigree, appendix and grade Holstein, Jersey and
crossbred cows were estimated using the Average Information Restricted Maximum Likelihood
(AIREML) animal model. The objectives were to determine the best cow class, in terms of milk
production, and determine whether the cow classes are different in the three major dairy
breeds and whether there are any genetic differences among the three cow classes namely
pedigree, appendix and grade.
The records were obtained from the Zimbabwe Dairy Services Association (ZDSA) for total
lactation from 1979 to 1994. The fixed effects in the animal models were herd-year, month of
calving, age at calving and parity. Random effects were residual error, permanent
environmental effects of the cow and the individual animal or cow effects. The cow classes
were analysed separately within each breed group because of limitations of computer
memory.
Appendix cows had the highest milk yields, followed by pedigree cows, with grade cows
producing the least. Heritabilities (h2) for milk yield in Holsteins were 0.23, 0.24 and 0.23 for
pedigree, appendix and grade cows, respectively. Similar values for Jerseys pedigree,
appendix and grade cow classes were 0.36, 0.36 and 0.25, respectively. Heritabilities for milk
yield in crossbred cows were 0.28 and 0.33 for appendix and grade cows, respectively.
Repeatabilities (rp) were similar and averaged 0.42 for all breeds and all cow classes. Major
differences among the cow classes across the breeds were observed in the additive, residual
and phenotypic variances.
The variances observed suggest adequate genetic variation in the Zimbabwean dairy breeds
cow classes for an effective direct selection programme. This genetically improves milk yield
and its positively correlated traits.
 Introduction
Zimbabwean dairy cattle can be divided into three classes, which are pedigree, appendix and
grade. Registered classes include pedigree and appendix cows for which the Zimbabwe Herd
Book (ZHB) will issue a certificate and confirms the breed stated on the certificate. A grade
daughter mated to a registered sire will produce appendix A offspring. By continued use of
registered sires, offspring of succeeding generations go from appendices A to B to C and
finally to pedigree, thus there is an open herdbook on registered Holsteins, Jerseys, crossbred
cows and other dairy breeds. Grades are those not included in the herdbook. The registration
by ZHB is done under the appendix scheme, which is an upgrading process where non-
registered or grade animals can achieve full pedigree status in three generations. However, it
should be emphasised that this registration is not based on performance but on identification
of pedigree and parentage. It has been reported that the pedigree has the highest milk yields,
followed by the appendix and the grade cow with the least milk yields (Makuza 1988; Trigg
1989). However, Muchenje (1996) found that the appendix cows had the highest milk yields,
followed by the pedigree then the grade cow class. Makuza (1995) reported heritabilities and
repeatabilities for milk yield of 0.19 and 0.45 in Holsteins, 0.27 and 0.45 in Jerseys, and 0.26
and 0.40 in crossbred cows, respectively. He also reported additive standard deviations for
milk yield of 488, 571 and 658 kg in Holsteins, Jerseys and crossbred cows, respectively. The
corresponding phenotypic standard deviations were 1120, 1104 and 1288 kg, respectively.
Reported repeatabilities for milk yield in the literature range from 0.35 to 0.60 (Mao 1984).
Similarly the range of heritabilities for milk production is from 0.15 to 0.40 (Mao 1984). Banga
(1992) reported a heritability estimate of 0.54+ 0.16 for milk yield in Zimbabwean Jersey cows.
This is a little bit too high possibly due to the method of estimation which was Harvey's least
squares mixed models. The estimates vary due to different measurements, samples,
populations, models or estimation procedures. Literature genetic parameter estimates of
different breeds also differ due to differences in population structures among breeds. Small
heritabilities (less than 0.10) mean that regardless of genetic evaluations and selection
methods used, the genetic gain would be relatively small.
Due to the contrasting findings by the different Zimbabwean researchers on the performance
of breed cow classes, there was a need to evaluate the genetic and phenotypic performance
of Holstein, Jersey and crossbred cow classes. The Holstein, Jersey and crossbred cows are
the three major dairy breed groups in Zimbabwe (Makuza 1995). The hypothesis tested in this
study was that the pedigree class produces the highest milk yield in all the three dairy breeds
of Zimbabwe.
 Objectives
The objectives of the study were to:
establish the best class of cow in three Zimbabwean dairy breeds
determine whether there are genetic differences (heritabilities, repeatabilities and
standard deviations) among cow classes of Holsteins, Jerseys and crossbred cows
compare genetic estimates from pedigree, appendix and grade daughters across
breeds.
 Materials and methods
 Source of data
Total lactation milk yield records for multiple lactations were available from 1979 to 1994
calvings of Holstein, Jersey and crossbred commercial dairy cattle (Table 3). The records
were provided by the Zimbabwe Dairy Services Association (ZDSA). The field records
consisted of parentage, production and reproduction data.
 Data editing
To remove outliers and incorrect data, records with age at calving outside the range of 16 to
200 months were deleted. Days in milk considered ranged between 250 and 300 days with a
calving interval between 300 and 600 days. Lactations 1 to 12 were evaluated. Lactations
greater than 4 were combined with lactation 4 into one subclass. Milk yield was restricted
between 1000 to 15000 kg per lactation. Records with cow classes incorrectly recorded apart
from 1 (pedigree), 2 (grade) or 3(appendix) were deleted. The data also had to be consistent
in birth dates, calving dates, year of calving and region of the country. The records were also
deleted for missing sire and dam identification or duplication or misidentification of parentage.
Each sire was supposed to have at least 2 daughters for AIREML to run.
 The Animal model
The animal model used in the analyses was as follows:
Yijklmno = : + HYi + Mj + Pk + Agel
+ Animaln + PEn(m) + Eijklmno
where:
Yijklmno = response variable milk yield;
: = overall mean common to all observations;
HYi = ith combined effect of herd and year of calving on milk yield;
Mj = jth effect of month of calving on milk yield;
Pk = kth effect of parity on milk yield;
Agel = lth effect of age at calving on milk yield;
Animaln = nth individual random animal effect on milk yield;
PEn(m) = nth random permanent environmental effect of the mth animal on milk yield;
Eijklmno = random residual error.
The inverse of the matrix of additive genetic relationships was added to improve linkage
between sires and improve the accuracy and precision of the estimates. The average
information restricted maximum likelihood (AIREML) FORTRAN algorithm of Gilmour (1995)
was used for analyses. Sparse matrix techniques were employed to calculate those elements
of the inverse of the coefficient matrix required for the first derivatives of the likelihood.
Residuals and fitted values for random effects can be used to derive additional right-hand
sides for which the mixed model equations can be repeatedly solved in turn to yield an
average of the observed and expected second derivatives of the likelihood function.
This Newton method, using average information, generally converges in less than 10
iterations. The Restricted Maximum Likelihood algorithm (REML), using average information,
is about five times faster than a derivative-free algorithm, using the simplex method, which in
turn is about three times faster than an expectation - maximisation algorithm (Johnson and
Thompson 1995). The REML animal model accounts for losses in degrees of freedom of fitting
fixed effects and selection bias. The cow classes were analysed separately within each breed.
Each dairy breed was therefore also analysed separately. Grade and appendix Jersey cows
and grade crossbred cows were analysed as single records. We could not run multiple records
due to the small sample sizes resulting in a non full rank matrix meaning that it had
dependencies in the factors fitted in the models. This could have resulted in convergence
reaching a local rather than a global maximum.
The values reported therefore correspond to a generalised inverse approach and are therefore
less reliable as they are not unique solutions to the mixed model equations.
 Results and discussions
In the preliminary analyses with SAS's PROC GLM (SAS 1994) for the three cow classes and
in all dairy breed groups, fixed effects of herd-year, month of calving, parity and age at calving
affected milk yields (P<0.001). These results are in general agreement with those reported in
the literature (Makuza 1995; Muchenje 1996; Chiyanike 1997).
Table 1 shows the percentages of cow classes by breed group. The majority of cow classes
are grade, which comprise 65% in Holsteins, 45% in Jerseys and 81% in crossbred cows. The
proportion of pedigree cows, which are close to 100% pure-bred exotic blood, is small. They
comprise 11% and 22% in Holstein and Jerseys respectively. As is expected there are no
pedigree Crossbred cows.
Table 1. Percentages of cow classes by breed group.
 Breed
Cow Class Holstein Jersey Crossbred
Pedigree 11% 22% 0%
Appendix 24% 33% 19%
Grade 65% 45% 81%
The percentages of cow classes by breed group and agroecological region are given in Table
2. For Mashonaland, the distribution of cow classes by breed is similar to that of Table 1.
However, for Midlands and Manicaland, the percentage of grade cows by breed group is
higher than those in Table 1. This indicates that these two regions have the majority of
unregistered cows in Zimbabwe. Almost 71% of Jersey cows in Matabeleland are pedigree
cows. As this is the driest region of the country, this finding is rather surprising and
unexpected. Again, as was expected, there are no pedigree crossbred cows in all the four milk
recording scheme regions of Zimbabwe. Herds with some of each type of cow may have
represented management systems that were attempting to increase the proportion of
registered cows i.e. pedigree and appendix. Grade cows in these herds would have faced
lower probabilities of survival than appendix or pedigree cows or even comparable cows in all
grade herds of Holstein, Jersey and Crossbred cow classes. This could be true for
Mashonaland with the majority of cows for the three breed groups and shows a deviation of
proportions from those of other regions (Table 2).
Table 2. Percentages of cow classes by breed and agroecological region.
  Breed
Region Cow Class Holstein Jersey Crossbred
Matebeleland Pedigree 6 71 –
Appendix 22 9 3
Grade 71 20 97
Midlands Pedigree 2 19 –
Appendix 13 28 2
Grade 85 53 98
Mashonaland Pedigree 17 10 –
Appendix 28 40 21
Grade 55 50 79
Manicaland Pedigree 1 4 –
Appendix 19 19 13
Grade 80 77 87
Table 3 shows the raw means and standard deviations for various traits of the Holstein,
Jersey and Crossbred cow classes. Milk yield, lactation number, days dry and calving intervals
in all the three breed groups were highest in the appendix cows which were also the oldest
with an average age at calving of 59 months compared to 54 months for grade and 50 months
for pedigree. Fat percentages tended to be slightly higher in grade cows for all the three breed
groups. Days in milk were similar for all cow classes in all the breeds. Grade and pedigree
cows left herds at younger ages. Grade and pedigree cows averaged 2.7 lactations across
breeds versus 3.0 lactations for appendix cows.
Table 3. Raw means and SD for various traits of Holstein, Jersey and Crossbred cow
classes.
Class of Trait Breed
cow Holstein Jersey Crossbred
Mean SD Mean SD Mean SD
Pedigree Milk yield (kg) 6212 1667 3628 1198   
Fat % 3.55 0.37 4.68 0.52
Days in milk 292 19 286 28
Age at calving (mo) 46 22 54 31
Calving interval (d) 251 198 263 184
Days dry 62 44 58 48
Lactation number 2.4 1.5 2.9 1.8
n 8184 1925
Appendix Milk yield (kg) 6415 1750 4252 1067 4895 1042
Fat % 3.53 0.40 4.45 0.48 3.81 0.40
Days in milk 291 17 290 17 293 15
Age at calving (mo) 54 25 62 32 60 23
Calving interval (d) 285 182 308 166 285 178
Days dry 65 40 62 41 51 33
Lactation number 2.8 1.6 3.4 1.8 2.90 1.6
n 17841 2846 970
Grade Milk yield (kg) 5496 1682 3770 1099 4375 1202
Fat % 3.57 0.49 4.50 0.52 3.90 0.48
Days in milk 290 18 289 20 288 20
Age at calving (mo) 51 24 56 31 55 25
Calving interval (d) 251 193 255 187 245 188
Days dry 58 43 53 44 49 41
Lactation number 2.5 1.6 2.9 1.8 2.6 1.6
n 48791 3965 4223
Heritabilities and repeatabilities by class of cow and breed are given in Table 4. Generally,
heritabilities and repeatabilities were higher in Crossbreds followed by Jerseys. Higher results
were previously reported in Zimbabwean Jersey cows (Banga 1992). The genetic parameters
(h2 and rp) were slightly higher in appendix cows but those for grade and pedigree cows were
similar. As has been explained in the materials and methods section, heritabilities for Jersey
pedigree and appendix cows and Crossbred appendix cows had some permanent
environmental components in them because they were run as single records. The programme
could not separate permanent environmental effects from additive effects. The fact that
heritabilities and repeatabilities are higher in Jerseys and Crossbreds than in Holsteins could
also be due to sampling variation due to the small sample sizes compared to Holstein cow
classes. There could also be differences in the breed structures of these breeds. However,
generally these estimates are in agreement to the findings of Makuza and McDaniel (1995),
working with the same data. The fact that Crossbreds had higher heritabilities is expected
because they are a mixture of different breeds, some indigenous, resulting in more additive,
dominance and epistasis (genetic) variation compared to the pure-breds.
Table 4. Heritabilities, repeatabilities for milk yield in cow classes of three major breeds of
Zimbabwean dairy cows
 Breed group
 Holstein Jersey Crossbred cows
Cow class n h2 rp n h2 rp n h2 rp
Pedigree 5971 0.23 0.40 1025 0.361 0.38 – – –
Appendix 9635 0.24 0.43 1040 0.361 0.38 655 0.28 0.52
grade 8658 0.23 0.42 1646 0.25 0.45 2393 0.331 0.34
1Run as single records and used a generalised inverse approach due to a non full rank
matrix.
Standard deviations by cow class and by breed are in Table 5. In general, standard deviations
(Фa, Фpe, Фe and Фp) were highest in Holsteins followed by Crossbreds, with Jerseys having
the lowest. The same trend was reported by Makuza and McDaniel(1995).
Table 5. Additive a permanent environmental ( pe ) standard deviations (kg) for milk yield
in cow classes of three major Zimbabwean dairy cows.
 Breed group
Cow Holstein Jersey Crossbred
Class
 a pe e p a pe e p a pe e p
Pedigree 540 470 880 1134 408 –1 540 677 – – – –
Appendix 584 518 894 1186 419 –1 554 695 404 371 530 763
Grade 503 442 794 1039 328 293 488 657 425 –1 605 739
1 Run as single records therefore could not separate permanent environmental effects from
additive effects.
Appendix cows had higher standard deviations across breed groups followed by pedigree
cows. Grade cows had the lowest standard deviations.
These findings suggest that in a large population of any breed, including both registered
animals and grades resulting from several generations of use of pure-bred sires, there will
inevitably be some animals in the lower strata of the breed structure which are genetically
superior to some in elite herds.
Even if selection programmes in elite herds are highly effective and directed to goals of
commercial importance, this will occur due to genetic recombination or heterotic effects and
the fact that phenotypes are imperfect indicators of genotypes.
The fact that the data analysed covered the whole country of Zimbabwe, might indicate that
appendix cows have the right breed composition for general adaptation to the Zimbabwean
environment. Appendix cows are referred to as appendix for at least three generations under
the grading up appendix scheme. Usually the first crossbred progeny shows maximum
heterosis and this will envisage maximum milk production by appendix A cows.
Our study has indicated basic genetic and phenotypic differences in field data for grade,
appendix and pedigree milk production.
 Conclusions
Appendix cows outperformed pedigree and grade cows in milk production in all three
breed groups.
Genetic parameters (h2, rp and variances) were highest in appendix cows in Holsteins,
Jerseys and Crossbreds. Pedigree and grade cows had similar genetic parameters in
the three breed groups.
Genetic parameters (h2, rp and variances) were highest in Crossbreds followed by
Jerseys.
There is adequate genetic variation in Zimbabwean dairy breed cow classes for an
effective direct selection programme for milk yield.
Milk yield in Zimbabwe is as heritable and repeatable as in the temperate and tropical
regions of the world.
 Recommendation
There is a need to determine what genotype proportions of Pedigree and Grade cows achieve
the maximum milk production in Zimbabwe.
 Acknowledgements
We would like to thank the Zimbabwe Dairy Services Association (ZDSA) for providing the
records as input to this study. We also send our gratitude to Mrs V. Choruma for typing this
manuscript.
 References
Banga C. 1992. Genetic parameters for milk production traits in Jersey cattle. Zimbabwe
Journal of Agricultural Research 30:45–48.
Chiyanike S. 1997. Genetic evaluation of Grade, Appendix and Pedigree Holstein cows in
Zimbabwe. BSc honours thesis, Department of Animal Science, University of Zimbabwe.
Gilmour A. 1995. Average Information REML Manual. NSW Agriculture, Australia
Johnson D.L. and Thompson R .1995. Restricted maximum likelihood estimation of variance
components for univariate animal models using sparse matrix techniques and average
information. Journal of Dairy Science 78:449–456.
Makuza S.M. 1988. Influence of imported sires on production of dairy herds in Zimbabwe. MSc
thesis, Michigan State University, East Lansing, Michigan, USA.
Makuza S.M. 1995. Studies on the genetics of dairy cattle in Zimbabwe and North Carolina.
PhD dissertation, North Carolina State University, Raleigh, North Carolina, USA
Makuza S.M. and McDaniel B.T. 1995. Genetic and phenotypic parameters for production
traits in Zimbabwean dairy breeds. Journal of the Zimbabwe Society for Animal Production
7:151–156.
Mao I.L. 1984. Variation in dairy cattle population: causes and consequences. Proceedings of
the National Invitational Workshop on Genetic Improvement of Dairy Cattle. Milwaukee,
Wisconsin, USA pp.25–43.
Muchenje V. 1996. Development of age-month adjustment factors for lactation milk yield of
Zimbabwean Holstein dairy cattle. MSc thesis, Department of Animal Science, University of
Zimbabwe.
SAS (Statistical Analysis System). User's Guide; Statistics, Version 6.10 Edition. 1994. SAS
Inst., Inc., Cary, NC.
Trigg C. 1989. Holstein, sire and cow evaluation for milk production traits in Zimbabwe. MPhil
dissertation, Department of Animal Science, University of Zimbabwe. Zimbabwe.
Estimation of body composition in tropical sheep raised under
seasonal feed supply conditions: Prediction models
E. Negussie, O. J. Rottmann, F. Pirchner and J. E. O. Rege
International Livestock Research Institute (ILRI), P. O. Box 5689, Addis Ababa, Ethiopia.
Institute of Animal Breeding, Technical University of Munich, D-85350 Freising-
Weihenstephan, Germany
Abstract
Introduction
Materials and methods
Management of experimental lambs
Experimental layout
Infusion of D2O, sample collection and body measurements
Application of D2O dilution technique
Live animal measurements
Slaughter and carcass dissection
Chemical analyses
Statistical analyses
Results and discussion
Prediction of body components from body weight measurements
Prediction of body components from BW and TBW
Prediction of body components from BW and D2O space (DS)
Contributions of age and live animal measurements to the accuracy of prediction models
Acknowledgements
References
 Abstract
Prediction models were developed from isotope dilution space (D2O) and live animal
measurements (heart girth, height at withers, body length and tail volume measurements) to
estimate chemical body components of indigenous tropical fat-tailed sheep breeds in vivo. A
STEPWISE multiple regression procedure of SAS was used to assess the predictive power of
combinations of variables and models which minimise the predicted residual sum of squares.
With regard to the accuracy and robustness of prediction, models containing body weight as
the only predictor variable resulted in less accurate estimates of body components, especially
that of body fat and energy contents. However, the use of isotope dilution space (as an index
of Total Body Water) along with body weight measurements showed significant improvements
in R² and accuracy of prediction equations. Testing the predictive ability of models containing
live animal measures only, the result obtained showed that, despite a small reduction in
accuracy, indices of live animal measures gave comparable estimation of body components
with models containing isotope dilution space. Therefore, considering the cost of D2O and its
applicability in the field, indications are that the use of models containing indices of live animal
measurements only (in various combinations) is promising for field applications, to provide
longitudinal measures of change in body composition of tropical fat-tailed sheep.
Key Words: Sheep, Body Composition, D2O Space, Prediction Equations.
 Introduction
In the tropics, marked seasonality in the quantity and quality of feeds available for livestock
affects greatly the growth and development of animals. Therefore, some understanding of
changes in the body composition of animals is essential not only for understanding growth and
development processes, but also for planning appropriate mating and feeding strategies so as
to prevent adverse effects of marked seasonality in feed supply. Moreover, since faster
genetic gain in the growth rate and hence improved meat production potential of livestock can
be enhanced by the ability to accurately and easily measure sequential changes in body
composition of animals in vivo, the development of accurate prediction models for use on live
animals is vital. In this regard, however, there are only a few reports on indigenous African
sheep breeds. Particularly for the indigenous Ethiopian fat-tailed sheep breeds, such
information is lacking. The main objective of this study was, therefore, to develop prediction
equations for an in vivo estimation of the chemical body composition of indigenous Ethiopian
fat-tailed sheep breeds from D2O dilution space and live animal measurements.
 Materials and methods
 Management of experimental lambs
This study was conducted at the ILRI (International Livestock Research Institute) Debre
Berhan Research Station in Ethiopia. In this study, the two indigenous Ethiopian fat-tailed
sheep breeds (the Menz and Horro) were used. Lambs were born from ewes which were
oestrous synchronised in order to obtain contemporary groups of study lambs. All lambs were
reared on their dams until weaning at 3 months of age. Except for limited supplementation
made during parts of the dry season, lambs were raised entirely on natural pasture. At the
station, all lambs were routinely checked for any health problems and therapeutic treatments
were given to sick animals.
 Experimental layout
This study included two groups of experimental lambs. The first group comprised 52 lambs (27
Menz and 25 Horro) of both sexes born in the wet season while the second group included
sixty-lambs (30 Menz and 30 Horro) of both sexes born in the dry season. The whole
experimental period for each group was divided into five growth phases (i.e. 1st, 2nd, 3rd, 4th,
5th phases). The division of growth phases was based on the stage of maturity of lambs with
an interval of three months between each consecutive growth phase (1, 3, 6, 9 and 12
months). When lambs attain the prescribed stages of growth, randomly selected lambs of both
sexes per breed were dosed with a deuterium oxide (D2O) solution to determine D2O space
and thereby estimate Total Body Water (TBW) from blood samples collected at equilibrium of
dilution. Soon after the collection of the last blood samples, D2O dosed lambs were
slaughtered, dissected and tissue samples were collected from each of the slaughtered lambs
and used for the analyses of the chemical body components and the calorific value of the
body.
 Infusion of D2O, sample collection and body measurements
 Application of D2O dilution technique
Deuterium oxide (99.8%, Deutro GmbH, Germany) was injected at a rate of 0.5 gram per
kilogram of body weight using disposable plastic syringes (15ml, fitted with 21 gauge, 38mm
needles). During each injection, syringes were weighed full and empty to the nearest 0.01g.
The actual dose of the solution injected was calculated as the difference between the weight of
the full syringe before, and empty syringe after the injection. At equilibrium of dilution 10 ml of
blood samples were collected from each of the D2O dosed lambs and stored deep-frozen. The
vacuum sublimation technique was used to extract D2O containing blood water from deep-
frozen samples. The concentration of D2O in the sublimate was read at the Institut für
Nutztierwissenschaften der Humboldt Universität Berlin, in Germany, using an Infrared
Spectrophotometer (16 PC FT-IR attached to digital DECPc LPv 433dx PC) at a wave number
2700 cm-1 to 2400 cm-1 against water. Finally, the D2O dilution space (kg) was calculated
from the actual amount of D2O dosed and the concentration of D2O (g/kg) in vacuum
sublimates of the collected blood samples.
 Live animal measurements
Live animal measurements were taken from each experimental lamb at the end of each
growth phase prior to D2O dosing. Measurements taken included: body weight (BW), heart
girth (HG), height at withers (HW) and tail volume (TV). The tail volume was estimated using
the Archimedes' principle by submersion of the tail into a 10-1itre cylinder completely filled with
water.
 Slaughter and carcass dissection
Soon after the last blood samples were collected at equilibrium of dilution, D2O dosed lambs
of both sexes per breed were slaughtered and dissected. Tissue samples were collected from
each of the slaughtered lambs at all stages of slaughter and dissection processes. Collected
samples were put into properly labelled plastic bags and stored deep-frozen at –18°C until
ground for laboratory chemical analyses.
 Chemical analyses
Except for blood samples, all collected tissue samples were minced, sub-sampled, freeze-
dried and ground to 3 mm sieve size. Ground samples were then thoroughly mixed and
duplicate samples of about 70g were taken from each component and stored deep-frozen in
polyethylene bags until chemical analysis. Samples from each of the slaughtered and
dissected lambs were analysed for body chemical components (moisture, ether extract, crude
protein and ash) and gross energy (GE) content.
 Statistical analyses
All statistical analyses were made using the Statistical Analysis System (SAS). Simple and
multiple regression equations were developed to predict body components from all
combinations of independent variables (D2O space, age, BW, EBW, and various indices of
live animal measurements) using the SAS REG procedures. Outlier, collinearity, and influence
diagnostics were examined and models which minimise predicted residual sum of squares
were selected and fitted to the data. The STEPWISE procedures of SAS (SAS 1990) were
used to assess the contribution of each additional variable to the accuracy of prediction
models, and variables were retained in the best models only when judged a significant
predictor (after adjusting for other variables when the model contained more than one
variable).
 Results and discussion
 Prediction of body components from body weight measurements
Prediction equations expressed in kilograms for estimating body chemical components from
BW are presented in Table 1. Equations developed from the pooled data produce higher
coefficients of determination (R2) and generally smaller model standard error than any of the
age groups used individually. Moreover, such calculated models resulted in higher accuracy of
prediction when expressed in kilograms than as a percentage of soft tissue.
Table 1. Equations relating body chemical components (kg) and energy (Mcal) to body weight
(BW, kg) and empty body weight (EBW, kg) in Menz and Horro sheep
Body
Components Breed  a + β1BW  a + β1EBW
   a β1 R² RMSE  a β1 R² RMSE
 Horro  0.0591 0.1043 0.912 0.120  –0.0475 0.1530 0.985 0.049
Protein Menz  0.0319 0.1056 0.929 0.105  –0.0232 0.1507 0.983 0.050
            
 Horro  0.0062 0.0199 0.876 0.027  -0.0055 0.0281 0.877 0.027
Ash Menz  0.0025 0.0200 0.927 0.020  –0.0037 0.0279 0.942 0.018
            
 Horro  –0.0132 0.0978 0.656 0.263  –-0.2110 0.1558 0.834 0.182
Fat Menz  –0.0863 0.1005 0.791 0.188  –0.1748 0.1484 0.897 0.132
            
 Horro  0.1978 1.4532 0.775 2.913  –2.1787 2.2430 0.927 1.657
Energy Menz  –0.6147 1.4846 0.865 2.131  –1.7177 2.1632 0.956 1.126
(Mcal)
Table 2. Equations relating body chemical components (kg) and energy (Mcal) to body weight
(BW, kg) and total body water (TBW, kg) in Horro and Menz lambs
Body
Components Model  Horro  Menz
   a β1 β2 R² RMSE  a β1 β2 R² RMSE
Protein a +β1  0.0591 0.1043 – 0.912 0.118  0.0319 0.1056 -- 0.929 0.105
BW
 a +  0.0451 0.3580 -0.3509 0.965 0.070  0.0604 0.3573 – 0.965 0.070
β1BW + 0.3515
β2TBW
              
Ash a +  0.0062 0.0199 – 0.875 0.027  0.0025 0.0200 – 0.927 0.020
β1BW
 a +  0.0055 0.0319 -0.0166 0.875 0.001  0.0048 0.0402 – 0.934 0.017
β1BW + 0.0282
β2TBW
              
Fat a +  - 0.0978 – 0.656 0.263  - 0.1005 – 0.790 0.188
β1BW 0.0132 0.0863
 a +  – 0.7951 –0.9647 0.983 0.054  – 0.7033 – 0.984 0.044
β1BW + 0.0517 0.0182 0.8420
β2TBW
              
Energy a +  0.1978 1.4532 -- 0.775 2.913  – 1.4846 – 0.865 2.131
(Mcal) β1BW 0.6147
 a +  – 9.1703 – 0.990 0.612  0.1592 8.3301 – 0.929 0.564
β1BW + 0.2283 10.6766 9.5619
β2TBW
BW, Body weight; TBW, Total body water; Coefficient of determination; RMSE, Root mean
squares of error; Mcal, Mega calories; EBW, Empty bodyweight
Results presented in Table 1 shows that body protein was predicted relatively accurately from
a model containing BW alone with an R2 of 0.912 for the Horro and 0.929 for the Menz sheep,
respectively. The prediction of body fat and body energy content was, however, less accurate
with an R2 of 0.656 and 0.775 in the Horro, and 0.791 and 0.865 in the Menz sheep,
respectively. In this regard, Velazco et al. (1997) also working with urea dilution and dairy
cows at different ages, reported more accurate predictions of body water and protein from
body weight than for body fat and ash, agreeing with the results reported by Swartz et al.
(1991). This is also in line with the results of the present study in that for BW and ash, greater
than 90% of the variation was explained by body weight alone in contrast to 60 to 80% of the
variation for body fat and energy contents. Therefore, at times when accurate information on
body reserve and energy status of tropical sheep is required, body weight alone can not be a
suitable index to obtain reliable estimates of these components on live animals.
Similar equations calculated with EBW in the model are also presented in Table 1. As
compared to BW, the use of EBW as a predictor variable showed a significant improvement in
both R2 and root mean squares of error (RMSE) of prediction equations for all the different
body components. Particularly, body fat and body energy contents were predicted more
accurately from the model containing EBW rather than BW as a predictor variable. The
inclusion of EBW to the model resulted in an increase of 10.6 and 17.8% and 9.1 and 15.2%
in the R2 of prediction equations for body fat and energy in the Menz and Horro sheep,
respectively. Although different researchers have shown that live weight preceding slaughter is
the most important variable for predicting body components, there are mixed findings with
respect to its value in sheep. Kempster et al. (1982) reported that live weight at evaluation is
normally the most effective predictor of carcass composition as it gives the most precise
prediction when used alone and, also, the highest level of precision when scan measurements
were used along with it in multiple regression equations. On the contrary, Sykes (1974) in
sheep, and Johnson and Farrell (1988) in poultry, reported that body weight per se provides a
less accurate prediction of body components and especially that of body fat and energy.
Thompson et al. (1977) and Scents et al. (1982) also indicated that live weight offers much
less as a predictor of body composition because of the many factors affecting the rate and
onset of fattening in meat animals. The results of this and other studies also showed that the
use of EBW instead of BW increased the accuracy of prediction models. The main reason for
this improvement may be related to the exclusion of the large variability associated with gut fill
of ruminants, which normally accounts for about 13.8 to 17.3% (Devendra 1980) of the live
weight in small ruminants, depending on such factors as age, diet, breed or species of animal.
 Prediction of body components from BW and TBW
The close relationship between TBW and body chemical components has long been noted
and its use with other variables in prediction models gives an accurate estimation of body
composition in animals. But in this study, the inclusion of TBW as a predictor variable along
with body weight resulted in a very little or no improvement in the R2 and accuracy with which
both body protein and ash components are predicted. On the contrary, however, the inclusion
of TBW into the model resulted in a highly significant improvement of the accuracy of
prediction of body fat and body energy content. When TBW was added to the model
containing body weight alone, the R2 was increased by 32.7 and 21.5% and by 19.4 and
12.5% for body fat and body energy in the Horro and Menz sheep, respectively (Table 2).
Such use of TBW for routine applications is, however, limited by lack of methods for its direct
measurement on live animals.
 Prediction of body components from BW and D2O space (DS)
Since, there are no direct means of quantifying TBW on live animals, its use in prediction
models has been restricted to controlled scientific studies. Instead, therefore, the dilution
space of hydrogen isotopes has been used for an in vivo estimation to provide an index of
TBW in prediction models for accurate estimation of body composition in pigs (Houseman et
al, 1973; Shields et al. 1983; Farrel and Cornelius 1984), sheep (Searie 1970b; Donnelly and
Freer 1974) and cattle (Crabtree et al. 1974; Chigaru and Topps 1981). Thus accurate
estimation of TBW from DS in sheep (Wright and Russel 1984 ;Rozeboom et al. 1994) and the
relative ease of obtaining BW measurements has led to the use of multiple regression
equations with these variables as best predictors of the main body chemical components
(Panaretto 1963; Keenan et al. 1969; Searie 1970a) in live animals.
In this study the predictive power of DS (as an index of TBW) when used along with BW in
prediction models was investigated (Table 3). The results show that in both breeds the
addition of DS to the model containing BW had no significant effect on the accuracy of
prediction of body protein and ash components. This may be due to the fact that compared to
BW, DS is not a particularly good predictor of body protein and ash and hence its contribution
in multiple regression equations with live weight is not significant. The most significant
improvement due to the addition of DS to the model containing BW was, however, observed in
the prediction of body fat and energy contents in both breeds. In the Horro sheep, the R2 was
increased from 65.5 to 87.3% and from 77.5% to 91.6% with corresponding reduction of the
RMSE from 0.263 to 0.161 and from 2.913 to 1.799 for fat and body energy contents,
respectively (Table 3).
These results are in general agreement with literature findings. Numerous experiments with
young growing animals have shown that body fat can be predicted precisely from live weight
and either tritiated water or deuterium oxide space (Panaretto 1963; Donnelly and Freer 1974;
Trigg 1974; Searie 1970a, b; Robelin 1977). Working with three different isoltopic methods on
sheep and with tritium and D2O concurrently, Trigg et al. (1978) and Johnson and Farrell
(1988) also reported that multiple regression equations including either deuterium or tritium
space with live weight predicted body fat content with significantly greater precision than live
weight alone.
Table 3. Models for estimation of body chemical components (kg) and energy (Mcal) from BW
(kg) and DS (kg) in Horro and Menz lambs
Body
Components Model  Horro Menz
   A β1 β2 R² RMSE A β1 β2 R² RMSE
Protein a +  0.0591 0.1043 – 0.912 0.120 0.0319 0.1056 – 0.929 0.105
β1BW
 a +  0.0998 0.2921 –0.2555 0.945 0.094 0.0546 0.1592 –0.0745 0.934 0.102
β1BW
+
β2DS
             
Ash a +  0.0062 0.0199 – 0.875 0.027 0.0025 0.0200 – 0.927 0.020
β1BW
 a +  – – – – – 0.0042 0.0240 –0.0055 0.928 0.020
β1BW
+
β2DS
             
Fat a +  –0132 0.0978 – 0.656 0.263 –0.0863 0.1005 – 0.790 0.188
β1BW
 a +  0.1029 0.6339 –0.7297 0.873 0.161 –0.0186 0.2601 –0.2221 0.835 0.168
β1BW
+
β2DS
             
Energy a +  0.1975 1.4532 – 0.775 2.913 –0.6147 1.4846 – 0.865 2.131
(Mcal) β1BW
 a +  1.4749 7.3469 –8.0230 0.916 1.799 0.1244 3.2264 –2.4253 0.891 1.926
β1BW
+
β2DS
Table 4. Models for estimation of body chemical components (kg) and energy (Mcal) from BW,
DS and age in Menz lambs
Body
Components Model Constants R² RMSE
  A β1 β2 β3   
 a +β1BW 0.0319 0.1056 – – 0.929 0.104
 a +β1BW + 0.0546 0.1592 –0.0745 – 0.934 0.100
β2DS
Protein a + β1BW + 0.0268 0.1701 –0.0760 –0.0104 0.938 0.100
β2DS + β3T
        
 a +β1BW 0.0025 0.0200 – – 0.927 0.044
 a + β1BW + –0.0001 0.0209 –0.0010 – 0.928 0.031
β2DS
Ash a + β1BW + 0.0014 0.0250 –0.0056 –0.0010 0.929 0.031
β2DS + β3T
        
 a +β1BW –0.0863 0.1005 – – 0.790 0.187
 a + β1BW –0.0186 0.2601 –0.2221 – 0.835 0.167
+β2DS
Fat a + β1BW + –0.0796 0.2839 –0.2254 –0.0227 0.852 0.158
β2DS + β3T
        
 a + β1BW –0.6147 1.4846 – – 0.865 2.130
 a + β1BW + 0.1244 3.2264 –2.4253 – 0.891 1.926
β2DS
Energy a + β1BW + –0.5817 3.5022 –2.4633 –0.2635 0.903 1.834
(Mcal) β2DS + β3T
BW, Body weight; DS, D2O space; T, Age; R2, Coefficient of determination; RMSE, Root
mean squares of error; Mcal, Mega calories.
On the contrary, however, working with 40 mature ewes in late pregnancy, Lodge and Heaney
(1973) found no increase in accuracy from inclusion of D2O space in prediction models
agreeing with the results reported by Velazco et al. (1997) for dairy cows.
 Contributions of age and live animal measurements to the accuracy
of prediction models
The inadequacies of published equations support the suggestion of Seebeck (1968) that live
weight and dilution spaces (D2O space or TOH space) alone are insufficient to specify body
composition of animals in vivo. In most cases, however, it is also not clear what additional
predictors, measurable on the live animal, could be usefully added to them to increase the
accuracy and robustness of prediction equations. It might be expected that an index of
skeletal development would merit inclusion, as this would differentiate between a fat young
animal and a thin adult of the same live weight. Age is one such index and equations
calculated for the prediction of body components from BW, DS and age in various
combinations are presented in Table 4. The result shows that except for a very little increase
in the accuracy of prediction of body fat and energy contents, the inclusion of age as an
additional variable to the model containing BW and DS did not result in significant
improvements of the accuracy with which the other body components are predicted. But
Hofmeyer et al. (1971) and Benjamin et al. (1993) found that body components as well as
body energy contents are predicted more accurately from multiple regression equations that
included body weight and age as predictor variables in addition to isotope dilution space. On
the contrary, however, Viljoen and Coetzee (1988) and Gizaw and Abegaz (1995) reported
that the addition of age as an additional independent variable to live mass in the prediction of
body components did not show any significant improvement. In general, in this study the fact
that body weight and D2O space are associated with a considerably greater proportion of the
variability in body constituents than in age (when both are used in the same model) might
explain the reason why age did not have a significant effect, and accounts for only 0.3 to 0.5%
of the total variability.
The contribution of each of the different live animal measures to the predictive power of a
model containing BW and DS are presented in Table 5a and 5b. The result in general shows
that the addition of indices of live animal measures (HW, BL, HG and TV) to the model
containing BW and DS resulted in very little improvement of the accuracy of prediction. For
both breeds, the increase in the R2 and the reduction in RMSE ranged from 0.5 to 4.5% and
from 0.05 to 0.50 for body fat and energy content, respectively. This is, however, contrary to
the results of Seebeck (1968) and Donnelly and Freer (1974) who have strongly suggested
that indices of skeletal size are of great value in prediction of body composition in domestic
animals. Literature reports regarding the added advantage of including such measurements in
prediction models are also conflicting. Wright and Russel (1984) working on mature cows used
indices of skeletal size in addition to BW and DS and found no significant advantage in
accuracy of predictions, especially for body protein, which is in line with the results of the
present study. The reason for this lack of significant contribution of indices of live animal
measurements (when used along with DS) in this and other studies, may be due to the fact
that variation within breeds in live animal measurements is too small to allow it to improve
upon the level of prediction given by live weight and D2O space alone. However, in situations
where meaningful live weight measurements cannot be obtained such as in gestation, etc.,
indices of live animal measures could still be used to improve upon the accuracy of prediction
of body components.
Table 5. Accuracy of prediction models for estimation of body components (kg) and energy
(Mcal) from various indices of body composition, including BW, DS, T and linear body
measurements (TV, BL, HG and HW) in Horro and Menz lambs
HORRO MENZ
Models and independent
variables R2 RMSE Models and independent
variables R2 RMSE
Protein   Protein   
a +β1BW 0.898 0.118 a + β1BW 0.909 0.118
a +β1BW + β2DS 0.930 0.100 a +β1BW + β2T 0.947 0.089
a + β1BW + β2DS + β3HW 0.944 0.089 a + β1BW + β2T + β3DS 0.949 0.089
a + β1BW + β2DS + β3HW + 0.961 0.077 a + β1BW + β2T + β3DS + 0.951 0.089
β4HG β4TV
   a +β1BW + β2T + β3DS 0.956 0.083
+β_4TV+ β5HW
   a + β1BW + β2T + β3DS + 0.958 0.083
β4TV+β5HW + β6BL
      
Ash   Ash   
a + β1BW 0.965 0.044 a + β1BW 0.975 0.031
a + β1BW + β2HW 0.970 0.031 a + β1BW + β2T 0.983 0.001
a + β1BW + β2HW + β3HG 0.976 0.001 a + β1BW +β2T +β3DS   
a + β1BW +β2HW + β3HG 0.980 0.001    
+β4TV
a +β1BW + β2HW + β3HG 0.982 0.001    
+β4TV +β5T
      
Table 5b Continued   Fat   
Fat
a + β1BW 0.437 0.296 a + β1BW 0.695 0.230
a +β1BW + β2DS 0.800 0.178 a + β1BW +β2DS 0.791 0.194
a + β1BW + β2DS + β3TV 0.820 0.176 a + β1BW + β2DS + β3T 0.833 0.176
a + β1BW +β_2DS + β3TV + 0.836 0.170    
β4T
      
Energy (Mcal)    Energy (Mcal)   
a + β1BW 0.639 3.212  a + β1BW 0.805 2.559
a +β1BW + β2DS 0.863 2.026  a + β1BW +β2DS 0.880 2.039
a + β1BW + β2DS + β3HW 0.878 1.962  a + β1BW + β2DS +β3T 0.904 1.857
a + β1BW + β2DS + β3HW 0.890 1.915  a + β1BW + β2DS + β3T + 0.906 1.873
+β4BL β4TV
a + β1BW + β2DS + β3HW + 0.895 1.919  a + β1BW + β2DS + β3T + 0.918 1.778
β4BL+ β5TV β4TV+ β5HW
a + β1BW + β2DS + β3HW 0.901 1.920     
+β4BL+ β5TV + β6T
DS, D2O space (kg); BW, Body weight (kg); T, Age (months); TV, Tail volume: BL, Body
length; HG, Heart girth, HW, Height at withers; R2, coefficient of determination;
RMSE, Root mean squares of error, Mcal, Mega calories.
Table 6a. Accuracy of prediction models for estimation of body components (kg) and energy
(Mcal) from the body weight (BW) and various indices of live animal measurements (TV, BL,
HG and HW) in Horro and Menz lambs
HORRO MENZ
Models and independent R2 RMSE Models and independent R2 RMSE
variables variables
Protein   Protein   
a + β1BW 0.898 0.118 a + β1BW 0.909 0.118
a + β1BW + β2HW 0.907 0.114 a + β1BW + β2T 0.947 0.089
a + βBW + β2HW + β3HG 0.931 0.100 a + β1BW + β2T + β3TV 0.949 0.089
a + β1BW + β2HW + β3HG + a +β1BW + β2T + β3TV + 0.955 0.083
β4T 0.935 0.100 β4HW
a + β1BW + β2HW + β3HG + a + β1BW + β2T + β3TV +
β4T + β5TV 0.938 0.100 β4HW+ β5BL 0.957 0.083
   a + β1BW + β2T + β3TV
+β4HW+ β5BL + β6HG 0.959 0.083
      
Ash   Ash   
a + β1BW 0.965 0.031 a +1BW 0.975 0.031
a + β1BW +β2HW 0.970 0.031 a +1BW + _2T 0.983 0.001
a + β1BW + β2HW + β3HG 0.976 0.031    
a + β1BW + β2HW + β3HG    
+β4TV 0.980 0.001
a + β1BW + β2HW + β3HG +    
β4T +β5T 0.982 0.001
      
Fat   Fat   
a + β1BW 0.437 0.296 a + β1BW 0.695 0.230
a +β1BW + β2TV 0.493 0.288 a + β1BW + β2T 0.791 0.194
a +β1BW +β2TV +β3T 0.600 0.262 a + β1BW + β2T + β3TV 0.797 0.194
a + β1BW +β2T + β3TV + a + β1BW + β2T + β3TV +
β4BL 0.622 0.260 β4HG 0.824 0.184
a + β1BW + β2T +β3TV + a + β1BW + β2T + β3TV +
β4BL + β5HW 0.637 0.264 β4HG+ β5BL 0.828 0.184
 a + β1BW +β2T + β3TV +
  β4HG+ β5BL + β6HW 0.833 0.187
      
Energy (Mcal)   Energy (Mcal)   
a + β1BW 0.639 3.212 a + β1BW 0.805 2.559
a +β1BW +β2TV 0.676 3.119 a +β1BW + β2T 0.880 2.039
a + β1BW +β2TV +β3T 0.739 2.872 a + β1BW +β2T + β3TV 0.886 2.029
a + β1BW + β2T + β3T + a + β1BW + β2T +β3TV +
β4BL 0.746 2.909 β4HG 0.903 1.900
a +β1BW + β2T + β3TV + a + β1BW + β2T + β3TV
β4BL+ β5HW 0.760 2.911 +β4HG + β5BL 0.906 1.904
a + β1BW + β2T +β3TV + a + β1BW +β2T + β3TV +
β4HG+ β5HW+ β6HG 0.768 2.952 β4HG + β5BL+ β6HW 0.911 1.893
BW, Body weight (kg); T, Age (months); TV, Tail volume; BL, Body length; HG, Heart girth;
HW, Height at withers; R2, Coefficient of determination; RMSE, Root mean squares of error;
Mcal, Mega calories.
Considering the cost, technical and/or analytical difficulty and the ease with which
measurements of the different predictors are obtained under field applications, the predictive
power of models containing live animal measures only was tested (Table 6). The result
showed that except for a slightly reduced accuracy of prediction of body fat and energy
contents, the use of indices of live animal measures alone in prediction models gave estimates
of body components which are comparable to estimates from models containing isotope
dilution space. Therefore, the general indications are that in view of the economics of D2O and
its applicability, the use of properly measured body indices alone in prediction models will have
great promise for field applications involving tropical fat-tailed sheep breeds.
 Acknowledgements
The authors would like to thank the German Academic Exchange Services (DAAD) and the
International Livestock Research Institute (ILRI) for technically, materially and financially
supporting this study as part of the PhD research project of the principal author.
 References
Benjamin R. W., Koenig, R. and Becker, K. 1993. Body composition of young sheep and goats
determined by the tritium dilution technique. Journal of Agricultural Science, (Cambodia) 121:
399–408.
Chigaru P. R. N. and Topps, J. H. 1981. The composition of body weight changes in underfed
lactating beef cows. Animal Production 32:95.
Crabtree R. M., Houseman R. A. and Kay M. 1974. The estimation of body composition in
beef cattle by deuterium oxide dilution. Proceedings of the Nutrition Society 33, 74A–75A
(abs).
Devendra C. 1980. Goat production in the Asian region: current status, available genetic
resources and potential prospects. International Goat and Sheep Research 1: 55–78.
Donnelly J. R. and Freer M. 1974. Prediction of body composition in live sheep. Australian
Journal of Agricultural Research 25: 825–834.
Farrel C. L. and Cornelius S. G. 1984. Estimation of body composition of pigs. Journal of
Animal Science 58: 903.
Gizaw S. and Abegaz S. 1995. Changes in carcass characteristics of Horro rams with
increasing age and body weight under different feeding regimes. In: "Proceedings of the 7th
Annual Conference of the Ethiopian Society of Animal Production (ESAP) on "Livestock
production and the environment-implications for sustainable livestock production and the
environment - implications for sustainable livelihoods". 27–29 April, 1995.  Addis Ababa,
Ethiopia . pp. 233–239.
Hofmeyr H. S., Oliver I., Kroon F. E., and VanRensburg W. J. J. 1971. Mededeling oor
verwantskappe tussen tritiumruimte en liggaamsamestelling by 3 skaaprasse. South African
Journal of Animal Science 1: 13.
Houseman R. A., McDonald I., and Peenie K. 1973. The measurement of total body water in
living pigs by deuterium oxide dilution and its relation to body composition. British Journal of
Nutrition 30: 149–156.
Johnson R. J. and Farrel D. J. 1988. The prediction of body composition in poultry by
estimation in vivo of total body water with tritiated water and deuterium oxide. British Journal
of Nutrition 59: 109–124.
Keenan D. M., McManus W. R., and Freer M. 1969. Changes in the body composition and
efficiency of mature sheep during loss and regain of live-weight. Journal of Agricultural
Science, (Cambodia) 72: 505–508.
Kempster A. J., Cuthbertson A. and Harrington G. 1982. Carcass evaluation in livestock
breeding, production and marketing. Granada, London.
Lodge G. A., and Heaney D. P. 1973. Energy cost of pregnancy in single and twin bearing
ewes. Canadian Journal of Animal Science 53: 479–489.
Panaretto B. A. 1963. Body composition in vivo. III. The composition of living ruminants and its
relation to the tritiated water spaces. Australian Journal of Agricultural Research 14: 944.
Robelin, J. 1977. In vivo prediction of lamb body composition using heavy water space.
Annales de Biologie Animale, Biochemie, Biophysique 17: 95–105.
Rozeboom D. W., Pettigrew J. E., Moser R. L., Cornelius S. G. and Kandelgy S. M. E. 1994.
In vivo estimation of body composition of mature gilts using live weight, back fat thickness and
deuterium oxide. Journal of Animal Science 72: 355–366.
SAS (Statistical Analysis System). 1990. SAS Institute Inc., Cary. North Carolina.
Searie T. W. 1970a. Body composition in lambs and young sheep and its prediction in vivo
from tritiated water space and body weight. Journal of Agricultural Science (Cambodia). 74:
357–362.
Searie T. W. 1970b. Prediction of body composition of sheep from trititated water space and
body weight: Tests of published equations. Journal of Agricultural Science (Cambodia). 75:
497–500.
Seebeck R. M. 1968. Developmental studies of body composition. Animal Breeding. Abstract
36: 167–181.
Scents A. E., Walters L. E. and Whiteman J. B. 1982. Performance and carcass
characteristics of ram lambs slaughtered at different weights. Journal of Animal Science 55:
1360.
Shields, R. G., Jr., Mahan, D. C., and Graham, P. L. (1983). Efficiency of deuterium oxide to
estimate body composition of growing swine. Journal of Animal Science 57: 66.
Swartz L. A., Heinrichs A. J., Varga G. A. and Muller L. D. 1991. Effects of varying dietary
undegradable protein on dry matter intake, growth and carcass composition of Holstein calves.
Journal of Dairy Science 74: 3884.
Sykes A. R. 1974. The prediction of body composition of hill sheep from body weight, red cell
volume and tritiated water space. Journal of Agricultural Science (Cambodia) 269–275.
Thompson J. M., Pattie W. A. and Butterfield R. M. 1977. An evaluation of the "Sancogram" as
an ultrasonic aid in assessing carcass composition of live sheep. Australian Journal of
Experimental Agriculture and Animal Husbandry 17: 251.
Trigg T. E. 1974. Body composition of ruminants in relation to plane of nutrition and
production. PhD thesis, University of Aberdeen. Scotland.
Trigg T. E., Domingo E. A. and Topps J. H. 1978. A comparison of three different isotopic
methods for measuring body components of sheep. Journal of Science Fd. Agric. 29: 1007–
1016.
Velazco J., Morril J. L., Kropf D. H., Brandt R. T. Jr., Harmon D. L., Preston R. L., and
Clarenburg R. 1997. The urea dilution for estimating of carcass composition of Holstein steers
at 3, 6, 9, and 12 months of age. Journal of Animal Science 75: 139–147.
Viljoen J. and Coetzee S. E. 1988. The in vivo prediction of body composition in Boer goat
does by means of the tritiated water space technique. South African Journal of Animal
Science 18: 63–67.
Wright I. A. and Russel A. J. F. 1984. Estimation in vivo of the chemical composition of the
bodies of mature cows. Animal Production 38: 33–44.

Fodder yield and nutritive value of leucaena, acaecia and
calliandra at three cutting regimes and two locations in Malawi
G.Y. Kanyama Phiri, H.D.C. Msiska, B.H. Dzowela, 
W.M.M. Ngwira, M.J. Njewa and L.L. Kalanje
Introduction
Materials and methods
Location of the study
Experimental design and treatments
Data collection
Chemical analysis
Data analysis
Results and discussion
Fodder matter yields at Bunda
Nutritive value
Fodder yields of accessions at Bvumbwe (1998/99 season)
Conclusion
References
 Introduction
Multipurpose trees and shrubs (MPTS) constitute a rich and critical source of supplementary
fodder and protein in the diet of ruminants especially during the dry season (Chen Halim and
Chin 1992; Devendra 1992). In Malawi, leucaena (leucaena leucocephala (ham) de wit) has
been naturalised and is widely adapted in many parts of the country (Savory and Thomas
1971; Savery and Breen 1979). It used to rank high among the tree legumes used as fodder
for protein supplementation in beef and dairy diets. However, continued use of this
multipurpose tree legume has been threatened by the leucaena psyllid pest called Iteteropsylla
cubana which is capable of massively defoliating and killing the tree (Austin et al. 1997; Bray,
Palmer and Ibrahim 1997; Mugundi and Nair 1997). At the University of Hawaii, Beldt and
Nampopeth (1992) have researched on psyllid resistant varieties. Their results have shown
variety K636 to be resistant to the psyllid pest.
The psyllid problem has also forced researchers to explore other alternative species of
leucaena as well as species other than leucaena leucocephala which would be suitable
candidates for supplementary feeding. Leucaena tricendra, leucaena esculinta, leucaena
dwersifolia and leucaena pallida have demonstrated remarkable ability to resist psyllid damage
(Bray and Woodrope 1998; Dzowela 1994; Kavachi and Shirima 1997; Akyeampong 1998).
Fodder species have also been evaluated for adaptability; for instance, calliandra (Calliandra
calothyrsus (Meiss) L. and acacia (Acacia angristissina (Mill) Kuutze have been shown to be
suitably adapted to the tropics (Chamberlain 1998; Akyeampong 1998; Lesueur et al. 1996
and Paterson et al. 1996) Lesueur reported that Calliandra callothysus was successfully
introduced in the highlands of Burundi, Kenya and Rwanda.
Multipurpose tree species have also been evaluated for dry matter yields and results have
revealed remarkable variations. Bray et al. (1997) reported lower dry matter yields from
Leucaena leucouphala than from Calliandra callothynsus, Acacia anguistissima, Gliricidia
selsium, Leucaena dillensifolia and Leucaena pollida. This disparity in dry matter yield
between Leucaena leucocephala and the other species was also related to the psyllid attack
on Leucaena leucocephala.
Acacia angunstissima is believed to have originated from Belize on the Caribbean Coast. It is
resistant to the psyllid pest and produces large quantities of leaf biomass which is rich in
protein. Its leaves also form a rich source of mulch and litter for soil moisture conservation
and nutrient cycling (Dzowela 1994). Calliandra callothysuis is yet another multipurpose tree
which originated in Central America and is a rich source of protein. In a comparative
performance trial Akyeampong (1998) reported higher leaf production (4.5 tonnes per hectare)
from Calliandra calothyrsus than from Leucanena pallida (1.2 tonnes per hectare). Wood
biomass yields were 5.0 and 1.1 tonnes per hectare, respectively.
Mkgundi, Kettler and Nair (1997) reported higher nitrogen and carbon contents in Calliandra
calothyrsus foliage than in Cordia africana and Grevillea rubusta. Fresh foliage of Calliandra
calothyrsus was reported to be more digestible than the dried foliage (Paterson 1996). The
digestibility of dried foliage was 40% while that of fresh foliage was 60%. Apparently, the low
digestibility of dried foliage was attributed to high values of condensed tannins. This suggests
that digestibility of some species may depend on the form in which there are presented.
Acacia anguistissima has been shown to respond well to cutting by producing leafier and more
digestible material, but it tends to defoliate its leaves under infrequent cutting especially during
the dry season (Dzowela 1994). Under frequent cutting, its leaves contain larger amounts of
nitrogen than those of Sesbania sesban (L.) Merr, Cajanus Cajan, Gliricidia sepium,
Calliandra callothyrsus and Flemingia macrophyne, (Dzowela 1994). Calliandra callothyrsus
coppices well following cutting (Tomaneng 1990).
Dry matter yield can also vary with cutting time and variety of the multipurpose tree species. In
Malawi, Savery and Breen (1979) reported higher dry matter and nitrogen content from
Leucaena leucocephala C.V. Cumingham than from either CV Peru or CV Hawaiian giant
CK8. Frequent cutting has been reported to produce a high proportion of shorter and thinner
branches which are relatively high in crude protein and palatability. The only problem is high
concentrations of numosine at the tips of branches especially the new growth (Adeneye 1991).
This could bring about variations in the nutritive value of leaf and stem fractions (Karachi
1998).
The overall objective of this study was to determine the yield and nutritive value of Leucaena,
Acacia and Calliandra fodder at three cutting heights and two locations in Malawi. Specifically
the study sought to evaluate the effect of cutting height on:
growth rate of the leucaena, acacia and calliandra species and accessions
leafy and woody biomass yields of leucaena, acacia, and calliandra species and
accessions
crude protein Neutral and acid detergent fibre contents of leucaena, acacia and
calliandra species and accessions
leaf to stem ratio of leucaena, acacia and calliandra species and accessions
 Materials and methods
 Location of the study
The study was conducted at Bunda College of Agriculture (1158 m above sea level, 14o18` S,
33o 76' E with 1100mm average annual rainfall) and Bvumbwe Agricultural Research Station
(1350 m above sea level, 15o 55' S, 350 o 05' E with an annual rainfall averaging 1100 mm).
The study spanned a period of two years.
 Experimental design and treatments
There were a total of 8 accessions of multipurpose trees (MPTS) of which four were leucaena
species, namely, Leucaena esculenta (OFI 52/87) Leucaena tricandra (OFI 37/94) and
Leucaena diversifolia (OFI 135/98), three accessions of Acacia anguistissima (OFI 68/92, OFI
65/92 and OFI 70/93), and one accession of Calliandra calothyrsus (OFI 12/91). Seedlings of
these accessions were sown in the nursery and raised as bare - rooted seedlings. These
seedlings were transplanted in rows 18 m long, 0.5 m apart. Each row was spaced 1.5 m
apart. The trees were then allowed to establish for twelve months. Thereafter, cutting height
treatments of 0.25, 0.50 and 0.75m were imposed on the plants within the next plot of 9 m
long, discarding 4.5m from either side of the row.
The eight accessions and three cutting height treatments served as main plots and sub-plots,
respectively. Thus, the treatments were factorially combined in a split plot design with four
replicates per treatment combination.
 Data collection
The growth of trees was monitored at monthly intervals within the net plot. The trees were
filled after every four months. The filled trees were hand stripped of the leafy biomass which
was weighed fresh. A sub sample of the leafy biomass was weighed fresh then oven dried and
weighed to determine dry matter yield per hectare. Fifteen centimetre sections were cut from
all base, tip and middle sections of the stems. These were weighed fresh and oven dried at
105oC for 48 hours. The remainder of the stems from the net plot were weighed fresh and
oven dry weight was determined by extrapolation.
 Chemical analysis
Leafy biomass samples were oven dried at 70oC and ground to pass through a 1 mm sieve
and analysed for nitrogen using a Kjeldahl digestion method (Anderson and Ingram 1989).
Neutral and Acid were determined using the procedure as described by Goering and Van
Soest (1970).
 Data analysis
All data were subjected to analysis of variance by General Linear Models (GLM) procedures
using a Statistical Analysis System (SAS) package. Significant differences among the
treatment means were separated by fishers Least Significant Difference (LSD).
 Results and discussion
Multipurpose tree growth performance at Bunda (Table 1) indicates that Leucaena tricandra
and Acacia anguistissima (OFI 68/92) exhibited the fastest growth rate followed by Leucaena
esculenta and Acacia anguistissima (OFI 70/93). Lower growth rates were observed in
Calliandra calothyrsus and Leucaena diversifolia.
Table 1. Growth rate of leucaena, acacia calliandra species and accessions at three cutting
heights at Bunda College
Species Accession Growth rate (cm/day) Mean
Leucaena OF1 52/87 1.89 1.42 1.93 1.95
Esculanta
Leucaena OF1 53/88 2.48 2.45 1.40 2.11
Tricandra
Leucaena OF1 137/94 1.27 1.50 1.82 1.53
Pallida
Leucaena OF1 135/98 1.32 1.01 1.66 1.33
Dullrsifolia
Acacia OF1 68/92 1.87 2.68 1.92 2.16
Anguistissima
Acacia OF1 65/92 1.77 1.47 1.29 1.51
Anguistissima
Acacia OF1 70/93 1.78 1.88 1.81 1.82
anguistissima
Calliandra OF1 12/91 1.77 1.29 0.80 1.29
callothyrsus
Growth rates of Leucaena pallida and Acacia anguistissima (OFI 65/92) were intermediate.
Growth rates of the multipurpose trees during the dry season (Table 2) followed a similar
pattern except that in most cases the dry season growth rate was one third of that in the wet
season, apparently a reflection of declining moisture content normally experienced during the
dry season. Nevertheless, if selection was based on growth rate performance, then Leucaena
tricandra and Acacia anguistissima (OFI 68/94) would be primary candidates for further
agronomic and nutritive evaluation.
Table 2. Growth rate of leucaena, acacia and Calliandra species and accessions at three
cutting heights at Bunda (Dry season)
Species Accession Growth rate (cm/day) Mean
Leucaena
esculanta OF1 52/87 0.66 0.60 0.49 0.58
Leucaena
tricandra OF1 53.88 1.04 0.74 0.89 0.89
Leucaena
pallida OF1 137/94 0.49 0.82 0.61 0.64
Leucaena
paliideia OF1 135/98 0.57 0.57 0.58 0.57
Acacia
anguistissima OF1 68/92 0.42 1.15 0.62 0.73
Acacia
anguistissima OF1 65/92 0.80 0.75 0.62 0.72
Acacia anguistissima OF1 70/93 0.92 0.83 0.95 0.90
Calliandra callothyrsus OF1 12/91 0.44 0.45 0.23 0.37
The results presented in Table 3 suggest that a low cutting height of 0.25m stimulated
branching in Leucaena pallida (92 primary branches per plant) and Acacia anguistissima (OFI
70/93;86 primary branches per plant) when averaged across the cutting regimes. However, it
was observed that Acacia anguistissima accessions OFI 65/92 and 70/93 exhibited branching
performance. Two lessons could be drawn from this table. First, branching behaviour could be
species specific. Secondly, branching could be dependent on height of cutting. From these
results it is arguable that the Acacia anguistissima accessions OFI 65/92 and OFI 70/93 have
the potential for higher dry matter yields assuming that dry matter production is positively
correlated to the number of primary branches per tree. A cutting height of 0.25m appeared to
the most suitable cutting height in so far as branching was concerned and could therefore be
adopted as a suitable cutting height.
Table 3. Number of primary branches per plant for leucaena, acacia and calliandra species
and accessions at three cutting heights at Bunda
Species 0.25cm 0.50cm 0.75cm Mean
Leucaena esculanta OF1
52/87 52c 64bc 69b 67
Leucaena tricandra OF1 53.88 85 a 74b 61bc 73
Leucaena pallida OF1 137/94 92 a 73b 57c 74
Leucaena paliideia OF1
135/98 82ab 78b 66b 75
Acacia anguistissima OF1
68/92 54c 68b 82ab 68
Acacia anguistissima OF1
65/92 78b 79b 80ab 79
Acacia anguistissima OF1
70/93 86a 81ab 71b 79
Calliandra callothyrsus OF1
12/91 67b 81ab 59c 69
Mean             72
LSD 0.05      58.91
cv %           41.97
     abcMeans in rows or columns bearing the same superscripts are not significantly different
 Fodder matter yields at Bunda
Leaf biomass yields were highest for Leucaena esculenta (2139 kg ha -1) Leucaena tricandra
and Leucaena pollida (2020 kg ha -1) when they were both at 0.75m (Table 4).
Table 4. Fodder yield of leucaena, acacia and calliandra species and accessions at three
cutting heights at Bunda during the 1998/99 season.
Species Cutting height Leaf yield Stem yield Total yield Leaf:stem ratio
(cm) (kgha-1) (kgha-1) (kgha-1)
Leucaena 0.25 1793 3001 4794 0.60
esculanta 0.50 1180 2387 3567 0.49
(OF1 52/87) 0.75 2139 2066 4205 1.04
Leucaena 0.25 1293 2509 3802 0.52
tricandra 0.50 1833 1833 3666 1.00
(OF1 53/88) 0.75 1322 2530 3852 0.52
Leucaena 0.25 1255 2462 3717 0.51
pallida 0.50 1480 2688 4168 0.55
(OF1 137/94) 0.75 2020 2382 4407 0.85
Leucaena 0.25 1011 2218 3229 0.46
diversifolia 0.50 1974 2215 4189 0.89
(OF1 135/98) 0.75 1540 2747 4287 0.56
Acacia 0.25 1074 2273 3347 0.47
anguistissima 0.50 1709 2916 4625 0.59
(OF1 68/92) 0.75 1700 2908 4608 0.58
Acacia 0.25 1159 2377 3526 0.49
anguistissima 0.50 1467 2649 4096 0.55
(OF1 65/92) 0.75 1270 2478 3748 0.51
Acacia 0.25 1205 2414 3619 0.50
anguistissima 0.50 1531 2739 4270 0.56
(OF1 70/93) 0.75 1679 2886 4565 0.58
Calliandra 0.25 1554 2751 4305 0.56
callothyrsus 0.50 1287 2495 3782 0.52
(OF1 12/91) 0.75 1293 2472 3765 0.52
At this height of cutting, their leaf to stem ratios were 1.04 and 0.89, respectively. Highest dry
matter yields of stems were recorded in Leucaena esculenta (OFI 52/87) and Acacia
anguistissima (OFI 68/92). When cut at 0.25m, leucaena and esculenta produced 3001 kg ha -
1. When cut at 0.5 and 0.75m, stem biomass yields for Acacia anguistissima (OFI 68/92) were
2916 and 2908kg ha-1, respectively. The same accessions gave the highest total yields. From
these results it could be argued that Leucaena esculenta and Acacia anguistissima OFI 68/92)
are the best candidates for fodder and fuel wood, both of which are required by small holder
farmers.
 Nutritive value
Nutritive value of the 8 accessions were generally comparable. Table 5 shows phosphorous
content was higher in Acacia anguistissima (OFI 70/93) (0.5% P) Calliandra calothyrsus (0.5%
P) than the rest which averaged 0.4% P. Crude Protein contents for Leucaena diversifolia
(21.3 cp) Acacia anguistissima (OFI 65/92; 21.3%) and Acacia anguistissima (OFI 68/92;
20.6%cp) were significantly higher than those of Leucaena tricandra (18.1% cp) and calliandra
callothyrsus (19.4% cp). There was insufficient evidence to indicate that neutral and acid
detergent fibre contents were significantly different. This suggests that the digestibilities of
these multipurpose trees are similar.
Table 5. Nutritive value of leucaena, acacia and calliandra species and accessions at Bunda
(average cutting heights)
Species %N %P %CP %NDF %ADF
Leucaena 3.3 0.4 20.6 36.8 20.2
esculanta (OF1
52/87)      
Leucaena 2.9 0.4 18.1 36.8 19.1
tricandra(OF1
53/88)      
Leucaena pallida 3.2 0.4 20.0 37.8 20.6
(OF1 137/94)
Leucaena 3.4 0.5 21.3 36.9 20.5
diversifolia (OF1      
135/98)
3.3 0.3 20.6 34.5 21.1
Acacia
anguistissima (OF1      
68/92)
3.4 0.4 21.3 31.3 21.2
Acacia
anguistissima (OF1      
65/92)
3.2 0.5 20.0 35.8 21.1
Acacia
anguistissima (OF1      
70/93)
3.2 0.5 19.4 35.1 20.3
Calliandra
callothyrsus (OF1
12/91)
Mean - - 20.2 35.6 20.5
SE±Mean - - 0.44 0.51 0.28
cv - - 16.44 11.80 13.96
 Fodder yields of accessions at Bvumbwe (1998/99 season)
At Bvumbwe Agricultural Research Station, highest dry matter yields were obtained when the
trees were cut at 0.75m. Table 6 shows for instance, Acacia anguistissima accessions OFI
68/92, OFI 65/92 and OFI 70/93 were 2391, 2742 and 3442 kgha-1, respectively. Except for
Acacia anguistissima (OFI 68/92) the accessions with the highest dry matter yields did not
necessarily have the highest leaf to stem ratio. It would appear, however, that the ecology of
Bvumbwe Agricultural Research Station favoured Acacia anguistissima in terms of dry matter
yields. A similar trend occurred in the 1999/2000 growing season (Table 7).
Table 6. Fodder yield of leucaena, acacia and calliandra species and accessions at three
cutting heights at Bvumbwe during the 1998/99 season
Species Cutting height Leaf yield Stem yield (kgha- Total yield (kgha-1) Leaf:stem ratio
(cm) (kgha-1) 1)
Leucaena 0.25 432 2160 2992 0.36
collinsii 0.50 603 1921 2524 0.36
(OF1 51/88) 0.75 904 2208 3112 0.42
Leucaena 0.25 1193 2520 3713 0.47
tricandra 0.50 1198 1904 3102 0.58
(OF1 53/88) 0.75 1503 2313 3816 0.67
Leucaena 0.25 957 2380 3337 0.41
pallida 0.50 735 1560 2295 0.46
(OF1 137/94) 0.75 1337 2412 3749 0.55
Acacia 0.25 1675 2408  0.71
anguistissima 0.50 2289 4366 4083 0.52
(OF1 68/92) 0.75 2391 3335 6655 0.73
Acacia 0.25 1516 2525 5726 0.60
anguistissima 0.50 2119 3500 4041 0.64
(OF1 65/92) 0.75 2741 4687 5691 0.58
Acacia 0.25 2992 5028 7428 0.50
anguistissima 0.50 2985 4734 8020 0.60
(OF1 70/93) 0.75 3442 5690 7219 0.64
Calliandra 0.25 1137 1663 9132 0.69
callothyrsus 0.50 1228 1649 2800 0.72
(OF1 12/91) 0.75 944 1628 2877 0.60
    2512
Mean (overall) 1739 3029 4692 0.59  
SE± (interaction) 307 490 771 0.08
cv (%) 32 29 30 24
 
Table 7. Fodder yield of leucaena, acacia and calliandra species and accessions at three
heights at during the 1999/2000 season.
Species Cutting Leaf yield Stem yield Total yield Leaf:stem ratio
height (cm) (kgha-1) (kgha-1) (kgha-1)
Leucaena 0.25 1414 3073 4487 0.50
collinsii 0.50 1959 4526 6485 0.41
(OF1 51/88) 0.75 1985 3827 5012 0.51
Leucaena 0.25 2874 4746 7621 0.60
tricandra 0.50 1748 3239 4987 0.59
(OF1 53/88) 0.75 4098 5803 9880 0.70
Leucaena 0.25 2338 5135 7472 0.45
pallida 0.50 3397 6655 10,051 0.50
(OF1 137/94) 0.75 3805 6979 10,783 0.49
Acacia 0.25 3591 6530 10,121 0.58
anguistissima 0.50 4632 7351 11,983 0.65
(OF1 68/92) 0.75 4442 7273 11,715 0.64
Acacia 0.25 4486 7870 12,356 0.58
anguistissima 0.50 5268 8989 14,257 0.63
(OF1 65/92) 0.75 6131 10,190 16,321 0.60
Acacia 0.25 4837 11,343 15,864 0.54
anguistissima 0.50 7796 11,315 20,736 0.58
(OF1 70/93) 0.75 7587 10,589 18,176 0.68
Calliandra 0.25 2642 4551 7093 0.61
callothyrsus 0.50 2978 4104 7081 0.77
(OF1 12/91) 0.75 3075 4735 7810 0.71
Overall mean 4160 7114 11,338 0.63  
SE± (interaction) 972 1743 2617 0.07
cv (%) 50 53 50 23
 Conclusion
Under the conditions of this study, it can be concluded that growth, yield and quality of fodder
from leucaena, acacia and calliandra are influenced by cutting regimes and location growth
rate and number of plots per tree were highest at a cutting height of 0.25m. On the contrary,
dry matter yields were highest at a cutting height of 0.75 m regardless of location. Based on
location, Acacia anguistissima accessions gave the highest fodder yields at Bvumbwe. On the
other hand, the new introduction of leucaena species appeared to give higher yields than the
acacia and calliandra accessions.
Finally it is concluded that the 8 accessions offer an attractive alternative to leucaena
leucocephala. Further studies should screen leucaena leucocephala against these 8
accessions to confirm the results.
 References
Adeneye J.A. 1991. Mimosine content in various fractions of Leucaena leucocephala grown in
Western Nigeria. Journal of Animal Science and Technology 33: 349–353.
Akyeampong E. 1998. Comparative performance of Leucaena pallida and Calliandra
callothyrsus on an acid soil in a sub humid environment in Burundi. International Tree Crops
Journal 9(3): 227–232.
Anderson J.M. and Ingram J.S.L. 1993. Tropical Soil Biology and Fertility: A Handbook of
Methods. Second Edition CAB International. Wallingford, U.K.
Bray R.A.F. and Woodroffe T.D. 1998. Resistance of some Leucaena species to the Leucaena
psyllid. Tropical grasslands 22(1):1–10.
Chamberlain J.R. 1998. Isozymic variation in Calliandra callothyrsus: Its implications for
species delineation and conservation. American Journal of Botany 85(1): 37–47.
Chen C.P., Halim R.A. and Chin F.Y. 1992. Fodder trees and fodder shrubs in range and
farming systems of the Asian and Pacific Region: Legume trees and other fodder trees as
protein sources for livestock. Animal Production and Health 102:11–22.
Devendra C. 1992. Fodder trees and fodder shrubs in range and farming systems in Tropical
Humid Africa: Nutritional potential of fodder trees and shrubs as protein sources in ruminant
nutrition. Animal Production and Health 102:95–108.
Dzowela B.H. 1994. Acacia anguistissima: A Central American Tree that is going places.
Agroforestry Today 6(3):13–14.
Goering H.K. and Van Soest P.J. 1970. Forage and Fiber analysis. United States Department
of Agriculture (USDA) Agriculture Handbook No. 379. USDA, Washington DC, USA.
Karachi M. 1998. Variation in the nutrition value of leaf and stem fractions of nineteen
Leucaena lines. Animal Feed Science and Technology 70(3): 304–314.
Karachi M. and Shirima D.S. 1997. Agronomic evaluation of eleven Leucaena lines at Tambi,
Tabora, Western Tanzania. Agroforestry Systems 35(2): 155–165.
Kettler J.S. 1997. Fallow enrichment of traditional slash/mulch system in Southern Costa Rica.
Comparisons of biomass production and crop yields. Agroforestry Systems 35: 165–176.
Lesueur D., Tassin J., Evilorae M.P., Sarrailh J.M. and Peltier R. 1996. Improving Calliandra
growth. Agroforestry Today. 8(3): 12–13.
Savory R. and Breen D.A. 1979. Production of Leucaena leucocephala in Malawi.
UNDP/FE/MW/75/020. Working Paper numbers 15–22.
Savory R and Thomas D. 1977. Pasture Handbook for Malawi. Bunda College of Agriculture,
Lilongwe, Malawi.
Paterson R.T., Kariuki I.W. and Roothaert R.L. 1996. Calliandra for cows in Kenya.
Agroforestry Today 8(4): 20–21.
Tomaneng A.A. 1990. Calliandra callothyrsus: Observation on coppicing characteristics.
Agroforestry Today 2(2):15.
Van Den Beldt R.J. and Napompeth B. 1992. Leucaena psyllid comes to Africa. Agroforestry
Today 4(4): 11–12.

Effect of draft work on production and reproduction of Malawi
Zebu cows
S.K. Chikagwa-Malunga, W.F. Kumwenda and F.K. Nyondo
Chitedze Agric, Res. Station, P. O. Box 158, Lilongwe
Abstract
Introduction
Objectives
Materials and methods
Results and discussion
Work performance
Cow weight changes
Calf weight development and milk yield
Reproductive performance
Conclusion
References
Abstract
A study was carried out to evaluate the effect of draft on the performance of Malawi Zebu
cows in order to supplement work oxen, which are in short supply. A total of 24 cows were
randomly allocated to one of the following three treatments: a control group in which cows
were not exposed to any type of work, a light work group where cows pulled a cart loaded with
300kg weight, or a cultivated light soils (sandy soils) and heavy work group where cows pulled
a cart loaded with 600kg weight or cultivated heavy soils (clay soils). Work had no effect
(P>0.05) on the milk yield of cows. Similarly there were no significant differences (P>0.05) on
birth weight and weaning weight of their calves. The results indicate that Malawi Zebu cows
can be used for draft work and calf growth provided the animals are adequately
supplemented. Farmers are encouraged to use cows as a multipurpose animal for cow meat,
milk and traction/draft power.
Introduction
The Agriculture Policy of the Malawi Government is to increase smallholder production of both
food and cash crops through improved yields. In trying to achieve this, the government is
encouraging and assisting smallholder farmers to adopt improved farm technologies including
farm mechanisation in the form of animal draft power. Mechanising cropping operations using
draft animals reduces timeliness of operations thereby reducing bottlenecks at peak periods of
labour during land preparation, weeding and harvesting.
Animal traction fits into the existing smallholder-farming system and allows mixed cropping.
The proportion of farmers using draft animal power (DAP) is only 13% (Moald 1990) and the
estimated total population draft animals is 6,971 (Kumwenda and Kunkwezu 1987). A
shortage of work animals (steers) limits the use of draft animals by smallholder farmers who
are willing to have work animals. This constraint is partly due to competition between use of
steers for draft and beef production. Therefore, use of Malawi Zebu cows is now encouraged
in order to increase the number of draft animals, to allow more steers to be sold to other
farmers for fattening thereby increasing meat supplies for human consumption as well as
production of calves and milk, thus, making cows efficient utilisers of scarce feed resources.
Cows are 10 –15% lighter than males (Reh and Horst 1985) but females may be able to
perform the same work as males especially if they work in sandy soils. Use of cows as draft
animals would enhance the changing of pairs of work animals and result in healthier animals
being used. Farmers with more animals may hire them out to farmers with few or no animals
and this will increase the income obtained from the farmer's surplus animals.
Work on draught animal power (DAP) has been carried out for five years in order to support a
viable introduction and development of a multipurpose draft option for the smallholder farmers
in Malawi. From the little information available, it appears that the extent to which cows could
be used for draught largely depends on the provision of adequate supplementary feed
supplies, proper especially during the dry season. Studies on the effect of work on fertility and
productivity of cows suggest that draft use may reduce consumption rate and milk yield if poor
quality or insufficient feed is provided (Jabbar, 1993; Agyemang et al., 1983). However, data
from the International Livestock Centre for Africa (ILCA) shows no significant differences in
terms of productive and reproductive performance between draft and non draft cows which
were both fed adequately (Agyemang et al. 1983). The provision of adequate feed in the dry
season can be a problem since during this period both the quality and quantity of feed is very
low. Availability of concentrates is sometimes a problem and their costs may hinder their use
as supplementary feed for draft animals by most of the smallholder farmers. In such cases,
feed stuffs, which are locally available, for example, tree fodder mixed with crop residues, can
be important sources of feed for working animals.
The work by Chamdimba (1991) shows that Malawi Zebu cows can be used for draft work and
that such a practice would impose minimal adverse effect on milk production and calf rearing.
However, the effect of work on the fertility of Malawi Zebu cows was not studied in detail.
Many factors were also omitted from the study such as differences in field conditions, like soil
condition and temperature. It is also not known how much energy was expended by cows in
specific groups.
Although Reh and Horst (1985) and Agyemang et al. (1983) indicated that cows could be
used for draft without any adverse effect on fertility and milk production, especially if the
requirement for draft is light and for short periods, the minimum work for draft cows is still not
clear. Malawi Zebu cows traditionally are utilised over long periods of the day, as they are
used to travelling long distances to provide transport for smallholder farmers. Therefore, there
is a need to come up with the minimum work for draft cows as well as feed packages for
recommendations to smallholder farmers.
Draft cows may be used during early or late pregnancy and this may lead to abortions or
stillbirth even with minimum use at these times. Agyemang, et al. (1983) did not observe this
probably because cows worked for 3 hours per day only. It is therefore important that further
investigations should be carried out to determine when the cow should be worked depending
on the condition of the animal.
Objectives
To evaluate the potential of Malawi Zebu cows for draft work in order to supplement
work oxen, which are currently in short supply.
To evaluate the effect of work on body weight changes, fertility and milk production of
Malawi Zebu cows.
To investigate factors affecting work output of cows as draft animals.
To investigate the effect of supplementary feeding on live weight and work output of
Malawi Zebu cows.
Materials and methods
This research which was initiated by Bunda College and continued by Chitedze Research
Station was carried out at Mbawa experimental station for five years. The experiment involved
a total number of 24 cows, which were randomly allocated to three treatments comprising of 8
animals each. Treatment I involved a control group in which cows were not exposed to any
type of work. In treatment II (light work group), cows pulled a cart loaded with 300kg weight or
cultivated light soils (sandy soils). Treatment III was the heavy work group and cows pulled a
cart loaded with 600kg weight or cultivated heavy soils (clay soils). Each pair of cows was
worked three times a week for three hours per day. The cows were working throughout the
year alternating between pulling a load, ploughing or ridging. The route for carting was fixed at
a stretch of 11km.
Milking was done everyday in the morning and daily milk yields were recorded. Cows were
weighed once a month and calves were weighed at birth and at two week intervals up to
weaning (7 months). All data were analysed using statistics version 3.0 Analytical Package
Programme. Data, which was subjected to analysis of variance (ANOVA), included plough,
ridge and cart rate, body weights, deaths and weaning weights, milk yields and conception
rate.
Results and discussion
Work performance
The effect of each treatment on working area, time and speed are presented in Table 1.
Treatment had no effect on working area, time and speed as the differences between the two
treatment groups were small (P>0.05). However, cows working on heavy soil tend to cultivate
a bigger area than cows on light soils.
The results on working time are consistent from 1990 to 1994 although the results reported by
Jere and Zimba (1994) were somewhat lower than those reported by Chamdimba (1991)
which ranged between 0.19 and 0.20 ha per day. One of the causes would have been the
level of supplementation, which was generally reduced to maize bran only i.e. 1.3kg per
animal per day while Chamdimba fed 5kg Rhodes grass per animal per day in addition to
maize bran. In this study the cows had a much lower work rate than that of Malawi Zebu oxen.
Several factors might have influenced the speed of work of these cows and these include
weather, soil condition and depth of ploughing. Hot days and very wet soils, especially for the
heavy clay soils, reduced work speed greatly. Thus, animals slowed down in their work in hot
weather conditions because they got tired easily. Dry soil conditions, especially in clay loam
soils, increased the draft power requirement and hence lowered working rates. Ploughing
operations had higher draft requirements than ridging. However, cows developed adequate
power for ploughing provided they were adequately supplemented for work. In all operations
the heavy work group had a higher draft power output than the light weight group (Table 1)
indicating that the performance of the heavy weight group was better than that of the light
weight work group.
Table 1. Performance of cows in light and heavy work
Operation Year Light work Heavy work
Ridging area (ha) 1991–93 1.03 ± 0.06 1.28 ± 0.08
 1993–94 0.09 ± 0.06 0.12 ± 0.12
 Mean 0.56 0.70
Ridging time (hr.) 1991–93 2.22 ± 0.06 2.27 ± 0.06
 1993–94 2.55 ± 0.44 2.10 ± 0.32
 Mean 2.39 2.19
Ridging speed 1991–93 0.46 ± 0.43 0.56 ± 0.59
 1993–94 0.03 ± 0.03 0.05 ± 0.04
 Mean 0.25 0.31
Ploughing area (ha) 1991–-93 0.82 ± 0.02 0.93 ± 0.03
 1993–94 0.05 ± 0.07 0.05 ± 0.05
 Mean 0.44 0.49
Ploughing time (hr.) 1991–93 2.26 ± 0.03 2.23 ± 0.03
 1993–94 2.29 ± 0.40 2.21 ± 0.34
 Mean 2.28 2.22
Ploughing speed (ha/hr) 1991–93 0.36 ± 0.38 0.42 ± 0.21
 1993–94 0.02 ± 0.02 0.02 ± 0.02
 Mean 0.19 0.22
Carting distance (km) 1991–93 10.4 ± 0.28 10.4 ± 0.10
 1993–94 8.68 ± 2.26 8.07 ± 2.14
 Mean 9.54 9.24
Carting time (hr) 1991–93 2.8 ± 5.10 3.0 ± 0.01
 1993–94 3.05 ± 0.06 2.98 ± 0.26
 Mean 2.93 2.99
Carting speed (km/hr) 1991–93 3.7 ± 0.01 3.5 ± 0.03
 1993–94 2.87 ± 0.73 2.77 ± 0.61
 Mean 3.29 3.14
 Cow weight changes
Table 2 shows the average live weight changes of cows for each treatment group from March
1991 to March 1994. There were no differences in live weights among the three treatments
groups (P>0.05). In general, all cows in the different treatment groups lost weight between
1991 and 1992 but gained weight in 1993 although the difference in live weight changes were
not significant. These results indicate that work did not have an adverse effect on live weight
of the cows. This was probably due to adequate supplementation. However, Chamdimba
(1991) observed that the period of calving (September–October and November–December)
had an effect on cow live weight. September–October calvers had a higher weight loss due to
slightly heavier birth weights of their calves (18.25kg) and the weight loss thereafter was an
effect of the deteriorating feed quality and quantity towards the end of dry season. Those that
calved in November–December experienced a lower weight loss as an immediate effect of
calving due to slightly lower birth weights of their calves (17.11kg). These late calvers did not
experience a long period of weight loss after calving as quantity and quality of feed improved
within 3 to 6 months after calving, providing sufficient nutrients for cow and calf growth as well
as for increasing milk production. It is also important to observe that in some years (1993–94)
cows that were not exposed to any work were the lightest, although this was not significant
there may have been additional factors affecting the weight of these cows apart from draft
work.
Table 2. Average live weight of cows for light and heavy work
Year Control Light work Heavy work Mean
1991 293.5 ± 6.1 283.3 ± 8.6 298.0 ± 8.0 291.6 ± 7.28
1992 283.3 ± 3.44 264.7 ± 3.44 273.6 ± 5.4 273.8 ± 2.05
1993 286.8 ± 87.0 304.7 ± 35.2 297.5 ± 85.6 296.6 ± 85.65
1994 271.1 ± 105.3 310.7 ± 32.9 298.0 ± 19.13 293.0 ± 66.54
 Calf weight development and milk yield
Birth weights, weaning weights and average milk yields for 1991, 1992 and 1993 are indicated
in Table 3. Birth weights between the treatment groups were not significantly different
(P>0.05). Similarly, there were no differences in weaning weights and 200-day adjusted
weaning weights although the heavy work group calves were heaviest followed by the non-
working group and then the light work for 1991, 1993 and 1994.
Table 3. Productive performance and milk yield of cows for each treatment group
Year Parameter Control Light work Heavy work
1991 Birth wt (kg) 16.75 ± 1.55 17.88 ± 2.55 18.41 ± 0.69
 Calf ADG (g) 343.5 ± 52.7 292.6 ± 57.8 361.3 ± 48.4
 Milk yield (kg/d) 1.780 ± 0.38 1.433 ± 0.28 1.682 ± 0.55
1992 Birth wt (kg) 16.7 ± 2.51 17.4 ± 5.77 18.2 ± 2.07
 Weaning wt(kg) 77.2 ± 32.55 74.4 ± 31.57 89.2 ± 11.07
 200 days(kg) 72.3 ± 29.77 67.1 ± 28.33 83.5 ± 7.11
 Milk yield(kg/d) 1.5 ± 0.55 1.2 ± 0.42 1.4 ± 0.46
1993 Birth wt (kg) 14.9 ± 3.42 15.4 ± 6.53 15.7 ± 11.71
 Weaning wt (kg) 64.8 62.9 72.2
 200 day wt (kg) 63.2 ± 19.96 59.6 ± 26.56 71.1 ± 46.74
 Milk yield kg/d 1.61 ± 0.97 1.92 ± 0.66 1.94 ± 0.74
Chamdimba (1991) observed that the average daily gain (ADG) was highest for the calves in
the heavy work group followed by those in the non-working group while the light work group
had the lowest ADG. He also observed that calves that were born in November/December
were slightly lighter in birth weight than those born in September/October (Table 4). Because
of poor pasture conditions, their dams went through immediately before they calved and there
was a limited supply of nutrients for the growing foetus. The severity of poor feeding increased
as the dry season progressed. For these early calvers, grass was insufficient during the time
that calf growth was totally dependent on milk supplied by the dam. Therefore,
September/October calves delayed rumen development because they could not consume the
feed, which had high crude fibre (CF).
On the other hand, calves born in November/December showed an overall faster growth than
those born in September/October because of improved pasture conditions with the rains,
which started soon after they were born. Work of the dam had a significant disadvantage on
growth of their calves only in the light work group.
Milk yield was not different between the treatment groups. Thus, work did not have an effect
on milk production in cows kept under the proper feeding management. This is in agreement
with the results of Zerbini (1993). The period of calving had an influence on milk yield
(Chamdimba 1991). The milk yield of animals that calved in September/October rose for a
short time and then declined abruptly with inadequate feed availability until week 10, when it
rose again. The rise in milk yield from the 10th week onwards took place after the first rains
had started when pasture was plentiful. Those that calved in November/December were able
to increase their milk yield (Table 4) consistently, as pasture became available before the body
reserves were depleted. Malawi Zebu mean daily milk yield was 1.63 litres for the non-working
group; 1.52 litres for the light work group and 1.67 litres for the heavy work group from 1991 to
1993. Malawi Zebu can produce an average daily milk yield of 1.591 with a mean lactation of
261 days in its first lactation (Kasowanjete 1981). The national average is estimated at 1.7
litres/day; (range 2.21–1.2 litres) (Bunda College of Agriculture and Chitedze Research Station
1988).
Table 4. Effect of period of calving on birth weight, milk yield and calf average daily gain
Parameter Period of calving
 September/October November/December
Birth weight (kg) 18.25 ± 2 .02 17.11 ± 1.17
Milk yield (kg /day 1.484 ± 269 1.716 ± 0.491
Calf ADG (g) 351 ± 60 346.8 ± 60.1
 Reproductive performance
Cows were served every year between December and March, and the same bull was used in
all the three treatment groups. The effect of work on conception rate is shown in Table 5.
Conception rates were 100% in 1990 and 1992 in all the three treatment groups indicating that
work did not have any adverse effect on conception in those years. No negative effect of work
was observed on the calving performance of the working cows. None aborted, gave stillbirths
or showed any sign of difficulty in calving. Conception rates varied from year to year in all the
three treatments. It is also important to note that in 1991, no animal conceived for the light
work group. This was probably because the Malawi Zebu cow is a low producer with a long
calving interval, which can go beyond 12 months. The period of calving had an effect on cow
body weight development. September/October calvers did not only experience a higher weight
loss at the time of calving but that weight thereafter continued to decrease until about
December, when the animals started to regain weight. Those calving in November/December
did not lose weight due to calving but were able to regain weight shortly after calving.
Table 5. Effect of work on conception rates
Year Control Light work Heavy work
1990 100 (8) 100 (8) 100 (8)
1991 12.5 (1) 0 (0) 0 (0)
1992 100 (8) 100 (8) 100 (8)
1993 50 (4) 62.5 (5) 12.5 (1)
1994 62.5 (5) 75.0 (6) 62.5 (5)
Figures in brackets indicate number of animals, which conceived.
 Conclusion
The results of this 5-year on-station study indicate that Malawi Zebu cows can be used for
draft work and calf growth. Cows can do heavy work without significant negative effect on
other productive performance, like milk production for calf rearing and for subsistence
consumption. Work did not disadvantage the general performance of their calves when
compared to non-working counterparts. However, effects of work on fertility is still not clear as
conception rates tended to fluctuate as the years progressed in all the treatment groups,
including the non-working cows.
In addition, work rate achieved by the treatment animals was much lower than that reported in
the literature. Therefore, further investigations should be carried in these areas so as to
substantiate the current findings especially on conception rate and work output of Malawi Zebu
cows before they are recommended for smallholder use. Future studies should also look at
parameters like feed intake of the animals, amount of energy expended when working, amount
of time spent working, soil condition and ambient temperature. It is therefore recommended
that future research on draft animals should be carried out by the Farm Machinery Commodity
with the Livestock Commodity playing a supporting role.
 References
Agyemang K., Asteatke A., Anderson F.M. and Youlbeab W. M. 1983. Early indicating of
effects of work on reproduction and reproductive performance of dairy cows in the Ethiopian
highlands. ILCA (Mimeograph).
Bunda College of Agriculture and Chitedze Research Station. 1988. Beef/dairy production
systems in Malawi, Annual Technical Report. Lilongwe, Malawi.
Chamdimba E.A.T. 1991. The effect of draft work on some productive characteristics of the
Malawi Zebu (Bos indicus) heifers and their calves. MSc thesis, Bunda College of Agriculture,
University of Malawi. Malawi.
Jabbar M.A. 1993. Research on cow traction in Africa; some lessons to be learned. World
Animal Review/Revue Mondiale de Zootechnie 74/75: 1–2.
Jere A.J. and Zimba A.W.C. 1994. Effect of draft work on some production characteristics of
Malawi Zebu cows. Technical Report. Chitedze Research Station, Lilongwe, Malawi.
Kasowanjete M.B.B. 1981. Breeds of cattle and breeding programmes in Malawi. Presented at
the FAO/SIDA workshop on breeding and feeding of cattle for milk production. December
1981. Lilongwe, Malawi.
Kumwenda W. and Kunkwezu E. 1987. Research on use of draft animal power and
mechanisation in Malawi: Proceedings of a workshop on animal tracking and agricultural
mechanisation research in SADC countries. Maputo, Mozambique, 8–12 August 1987.
Unpublished data.
MoALD (Ministry of Agriculture and Livestock Development). 1990. A guide to agriculture
production in Malawi (1989–1990). Ministry of Agriculture and Livestock Development, Malawi.
Reh I. and Horst P. 1985. Beef production from draft cows in small-scale farming systems.
Quarterly Journal of International Agriculture. Vol. 24 no. 1:38–47.
Zerbini E. 1993. Cow tracking - What about milk and calves? Newsletter 10 (1). ILCA
(International Livestock Centre for Africa), Addis Ababa, Ethiopia.

A study of the flock dynamics, marketing, and socio-economic
roles of small ruminants in crop–livestock farming systems in
Malawi
J. W. Banda, S. C. F. Chapa, A. K. Nyasulu, W. Muyaya and C. S. Mataya
Bunda College of Agriculture, Lilongwe, Malawi
Abstract
Introduction
Materials and methods
Results and discussion
Productive and reproductive coefficients
Marketing
Socio-economic contribution
Conclusions
References
Abstract
The study was aimed at establishing the flock dynamics, supply patterns and the socio-
economic importance of small ruminants, especially goats, to household income and food
security. The performance of 337 goats from 43 households owning goats was monitored for
some productive and reproductive parameters. Questionnaire surveys of the producers and
sellers (stock agents and or middlemen) were conducted in four locations covering a total of
722 records of sellers and 651 buyers between January 1996 and January 1998. Data
collected included supply of goats and sheep, characteristics of animals, prices and forms
sold, constraints in buying and selling goats and sheep, and suggestions for future
improvements.
The average flock size was 8. The annual parturition rate was 83% with a prolificacy of 1.3
kids per female per year. Overall mortality rate was 34% with a risk of 28.8%. Mortality rate
was higher in rearing males (42%) than in breeding stock (15%) or suckling stock (27%)
showing that rearing kids are at high risk of dying before 12 months of age. The overall offtake
rate was 21% with a risk of 19%. It increased from young to old stock, and was highest in
rearing and breeding stock.
Of the total goat supplied, producers themselves sold 44.8% while middlemen sold 55.2%
(P<0.01). For sheep, only 28.1% were supplied by the producers while middlemen supplied
71.9%. Middlemen supplied more animals to the markets irrespective of season. Their animals
were heavier and they obtained significantly (P<0.001) more cash for their animals. The profit
margins were 29–39% for producers and 33–58% for middlemen (P < 0.05). The margins for
sheep were 24% for producers and 35–44% for the middlemen. As the age of the animals
increased, the live weights, and selling and buying prices increased (P <0.05). However, profit
margins significantly decreased after animals attained a third pair of permanent incisors. The
results show how much the middlemen contribute to the offtake and income of the farmers.
The proportion of profit goats produced was 77.4% of the capital invested, which is greater
than 46–52% interest charged by the lending commercial banks in Malawi. The contribution of
small ruminants to total gross returns of the farms was 30.3%.
It is suggested that mortality, especially of rearing stock, and offtake of females should be
reduced in order to sustain goat numbers and make goat production profitable and
sustainable. Improvements in the marketing of goats could be achieved through improved farm
productivity and social as well as physical marketing facilities.
Introduction
The goat, sheep and cattle populations in Malawi are 1,600,000, 100,000 and 600,000,
respectively (NLDMP 1998). Most of these animals are kept by smallholder farmers who own
less than 1 to 1.5 ha of land. For survival, these farmers practice integrated farming based on
crop/livestock production. The small pieces of land have encouraged many households to
keep small ruminants as opposed to cattle. This has led to an increase in the goat population
from 800,000 in 1990 (NLDMP 1998) to 1,600,000 in 1998 and a decrease in cattle population
from about 1 million in 1990 to 600,000 in 1998. However, the productive and reproductive
coefficients that contribute to the changes in goat populations have not systematically been
determined.
The small land holdings make it difficult for the households to produce enough food for the
family members to last the whole year. Consequently, there are regularly recurring periods
every year when the staple food (maize) runs out from September (World Bank 1995). This
period coincides with the time of greatest workloads in crop fields. During such times, selling
of goats to earn cash to purchase food is one of the major strategies farmers take to make
their households survive. Due to these sales, the contribution to the socio-economic and
cultural objectives of the rural community is quite large, but is quantitatively unknown.
Despite the recognised contributions of goats to smallholder farms, such roles and
contributions have not fully been studied and quantified. The knowledge of the system in
which goats interact with other farm enterprises is essential for identifying opportunities to
shape the future of small ruminant production in Malawi.
The objective of this study was, therefore, to characterise and quantify some of the productive
and reproductive coefficients and the marketing and socio-economic roles of small ruminants
with a view to influencing policy formulation that may lead to economically sustainable
production in crop/livestock systems.
Materials and methods
Data were collected during implementation of the EU- Science and Technology in
Development Programme 3 (EU-STD3) Research Project on small ruminant production. The
study was carried out in four locations of Mitundu, Nanjiri, Kamphata and Nkhoma in Central
Malawi (crop/livestock production area). Tobacco and maize are the major crops grown and
goat production is one of the major livestock activities households are engaged in. Very few
households keep sheep.
The methodology included initially a cross-sectional baseline socio-economic household
survey. The household survey was carried out to provide information on the family size and
age structure, household activities other than livestock production, land use, intra-household
division of labour, as well as ownership patterns and production goals. It included livestock
production systems and constraints in general. This survey covered 100 households in all four
locations and 43 flocks (households) were selected randomly based on the outcome of the
100 farms covered for the monitoring phase as well as during the marketing surveys.
The monitoring phase involved monitoring flocks of goats that were tagged for identification.
All animal related production parameters regarding events such as births, deaths, sales,
purchases, slaughters, gifts and thefts, as well as factors influencing these events, were
recorded on structured forms by trained enumerators, who visited each household once or
twice a month over 12 months. Weights and ages of all the goats were recorded during the
visits. From the data collected, parturition rates, offtake rates, mortality rates, survival abilities
and their associated risks including those due to sales and slaughters were derived. Incidence
and risk rates were calculated for all classes of animals covered.
At the same time that monitoring of the flocks took place, the relative profitability and income
contribution of the goats were studied on 40 out of the 43 farms. Simple budgeting and gross
margins analyses were used at household level and the results were then pooled over 40
households.
Detailed marketing studies were conducted between January 1996 and January 1998.
Structured questionnaires were used in both phases covering four major markets of Mitundu,
Nanjiri, Kamphata and Nkhoma where monitoring studies were conducted. The interviews
included sellers (producers/middlemen) and buyers (consumers). Information on sellers
included their background profile, supply of goats/sheep and the characteristics of animals
supplied, prices and forms sold, constraints in selling animals or buying animals and what the
government should do to assist them. Information on buyers (consumers) included their
background profile, demand consumption patterns, expenditures on the various foods, meat
and market preferences, and problems faced in buying goat/sheep meat.
All data were analysed using SPSS/PC+ software by generating frequencies and other
descriptive statistical parameters. Where appropriate cross tabulations were conducted. One-
way analysis of variance was carried out where appropriate.
Results and discussion
Productive and reproductive coefficients
The average flock size was 8 (range: 2–21). Over 56% of the flocks had less than 8 goats per
flock. Breeding and rearing females (before first parturition) comprised 82% of the flocks. This
implies that farmers are conscious of keeping only those animals which are productive for
increasing the herd. The flock size reported here is similar to that reported by Rischkowsky
and Steinbach (1997) and Banda et al. (1998) for other parts of Central Malawi. The annual
parturition rate was 82.6 with a prolificacy of 1.3 kids/parturition. This rate is lower than that
obtained by Banda (1992) of 1.47 for local goats. Rischkowsky and Steinbach (1997) obtained
even higher values of 2.0 for local Malawi goats.
Table 1. Some productive and reproductive coefficients of the flocks studied
Number of flocks 43
Total number of parturitions 200
Number of first parturitions 29
Annual parturition rate (%) 82.6
Number of kids born per parturition 1.3
First parturition before 18 months of age (%) 79
First parturitions between 18 and 24 months (%) 21
Males surviving to 6 months of age (%) 88.2
Females surviving to 6 months of age (%) 88.2
Table 2 shows some productive and reproductive parameters by class of goats in the villages
studied. The large proportion of the flocks was composed of breeding females. This shows
that farmers are consciously keeping the females for breeding. The number of breeding and
rearing males was kept to a minimum because most of these animals are sold before they
reach 18 months. The highest offtake risk of about 88% was observed in males above the age
of rearing.
Table 2. Some productive and reproductive parameters by class of goats in Central Malawi.
 No. anim. Flock Mortality Offtake Theft
 in flock structure* (%) (%) Lost(%)
Suckl. Stock(0-6mo) 92 27.3 27 27 0
Rear. males(6-18mo) 17 5.0 42 25 0
Breed. males(>18mo) 2 0.6 – 3 0
Rear.fem.6-1st Kidd) 87 25.8 – 20 3
Breed.fem.(>1st Kidd) 139 41.4 – 43 6
Rear.+Breed fem. 226 67.1 15 24 9
Total 337     
* Add to more than 100% due to the combination of animal classes.
The overall mortality rate was 34.0% with a risk of 28.8%. The mortality rate was highest for
rearing males, followed by suckling stock. These values are higher than those reported by
Rischkowsky and Steinbach (1997), although the methodology of calculation was different.
Mortality rates were higher in rearing males (42%) than in either suckling stock (27%) or
rearing plus breeding stock (15%), particularly starting from June, soon after harvest. The
trend was also similar for suckling stock.
The overall offtake rate was calculated to be 21.1% with an offtake risk of 19.0%. The offtake
rate increased from young stock to old stock. The risks of offtake were highest in rearing and
breeding stock. The major reasons for the offtake were mainly the need for cash, if the
animals were sold or slaughtered for ceremonies and family consumption.
 Marketing
Male animals were mostly sold and this agrees with results of Kusina, Dzudar, Chikura and
Sibanda (2000) in Zimbabwe. Since most males were sold, the season of sale was monitored.
It was observed that most sales took place from September to March during the study period.
Normally, most slaughters and sales of goats take place from September to March each year
(Figure 1). During this period yearly food shortages recur in households. Hence as a family
survival measure, selling of goats or slaughter are the main ways for obtaining cash for food,
which runs out by September (World Bank 1995). It is also the period when slaughters are
most predominant due to hunger, and when the social and ritual dances take place. It is
common to see goats being slaughtered.
Figure 1: Percent of sales and slaughters of rearing males in Central Malawi.
Of the goats sold at all farms, 47% of the goats were sold directly at the market or to other
farmers by the producers themselves while 53% were sold by the middlemen who purchased
the goats from the farmers. The prices of goats sold by the producers and those purchased
and sold by the middlemen were recorded. The summaries are shown in Table
Table 3. The average sale prices and marketing margins of goats by the producer in 1997/98
(US$1=MK 26.00)
 Producer/own Middleman Sign
% of total  
sold 45 55.2 **
Av. weight  
buying (kg) 22.8 ± 7.0 23.9 ± 6.0 NS
Av. buying
price (MK) 256 ± 68 285 ± 9 NS
Av. selling
price (MK) 298 ± 107 377 ± 142 ***
Av. Margin
(MK) 72 ± 87 92 ± 83 ***
% Margins  29 ± 31 33 ± 28 *
Av. sell  
price dead
(MK) 356 ± 160 449 ± 148  
% final  
margin 39 ± 51 58 ± 82 NS
NS = Not significant; * = P<0.05; ** = P<0.01; ***= P<0.001
Producers make 29–39% profit margins while middlemen make 33–58%. The contribution of
small ruminants to the producer (owner) and even to the middleman is quite immense. The
producers who also sell their own animals walk an average of 8 km only while middlemen
walk an average of 12 km. There were no significant differences in distance sellers moved
with regard to sheep marketing. Irrespective of type of supplier, the mode of transport was
predominantly either on foot (40.2%) or on bicycle (57.1%). The differences in profit margins
and goat/sheep selling prices could be attributable to the distances moved, and not to mode of
transport. The observations in the present study are not in agreement with the results reported
by Lovelace, Vaz, Mikuka, Simukoko and Chitambo (1997) of 38% for farmer, 32% for stock
agent and only 16% by market trader. The low prices obtained by small ruminant producers in
the villages suggest that farmers have less bargaining power than the middleman who took
33–58% of the final price as profit for transportation, market fee, market storage and for the
labour (time).
Table 4. The average sale prices and marketing margins of sheep by the producer in 1997/98
(US$1=MK 26.00)
 Producer Middleman Sign
% of total sold 28.1 71.9 **
Av. weight buying (kg) 22 ± 6.5 24 ± 6.2 NS
Av. buying price (mk) 281 ± 70 279 ± 95 NS
Av. selling price (mk) 300 ± 133 377 ± 146 NS
Av. Margin (mk) 57 ± 81 98 ± 72 NS
% Margins 24 ± 36 35 ± 22 NS
Av. sell price dead (mk) 664 ± 122 495 ± 117 NS
% final margin – 44 ± 22 –
NS = Not significant; ** = P<0.01
The market prices and marketing margins for sheep are higher than for goats due to the long
distance covered in sheep movements (over 26 km) compared to goat movements (10 km),
but also due to the scarcity of sheep compared to demand for sheep. Sheep in Malawi could
therefore be considered as a luxury if purchased by Malawians. More effort in multiplication of
sheep and investment in marketing infrastructure could lower marketing costs of sheep to that
closer to goats.
Table 5. Characteristics and prices of goats marketed related to age of the animals
Age by dentition (pairs)
 0 1 2 3 4 Broken
teeth
Average Number 1.0 1.3 1.4 1.4 1.6 1.0
Sold
Liveweight (kg) 14.4a 17.5b 22.3c 24.3cd 26.6d 28.4e
Buying price 198a 219b 304c 291c 313cd 315d
Selling price 212a 300b 370c 320b 367c 402d
% Margin 15.0a 64.6c 46.5b 23.9a 20.0a 40.6b
Price of dressed 296 345 461 363 446 245
goats
% Final Margin – 32.0 48.4 31.2 31.5 36.1
There are clear differences (P<0.01) in prices related to age and weight of the animals. As
age advanced, live weights and hence buying and selling prices significantly increased. This
trend was also reported by Lovelace et al. (1997). However, the margins obtained decreased
significantly after the animals reached 3 pairs of teeth. A similar trend was observed in sheep.
It is therefore important that farmers be advised to sell goats when they have reached this
stage. The best option for farmers to get the best price is to sell at the right age and weight
which could be achieved through improved nutrition. This will improve the income and relative
profitability of the enterprise. This is particularly important in this mixed crop/livestock system
of Central Malawi where only a few goats (1–4) may be culled or selected for marketing.
At the other end of the chain, buyers of meat or goats for consumption (consumers) were
interviewed on the problems they face in buying goats or goat meat. Table 5 indicates the high
prices of meats, poor handling or poor hygiene during slaughter and selling (with no
refrigerators) and non-availability of goat/sheep (especially sheep) meat, ranked in this order,
irrespective of the type of consumer. The high cost of goat meat discourages consumers from
purchasing them in large amounts.
Table 6. Problems with buying goats, sheep and their meat
 Rural Urban Total
 No % No % No %
Meat not available 27 19.7 3 2.2 30 21.9
Prices very high 33 24.1 11 8.0 44 32.1
Long distance to market 8 5.8 1 0.7 9 6.6
No standard/cheating 4 2.9 0 0.0 4 2.9
Handling/hygiene poor 30 21.9 4 2.9 34 24.8
Others 10 7.3 6 4.3 16 11.7
Total 112 81.8 25 18.2 137 100.0
 Socio-economic contribution
Results on the profitability of the goat enterprises as well as their income contribution to the
rural households are presented in Table 6.
Table 7. Simple budget for goat production enterprise in the 40 villages in central Malawi
Variables Total value (MK)
A. Revenue 2425.00
B Operating costs 115.98
C: Fixed cost value 322.90
D: Interest on capital invested (opp. cost) 49.00
E: Total cost of enterprise (B+C+D) 487.88
F: Net profit of enterprise (A-E) 1937.12
G: Net profit per animal 387.42
H: Percentage return on capital 77.4
On average, the small ruminant production enterprise earned a household a total net profit of
MK1937.12 (US$49.18) for an average flock of 5 per family. An animal, therefore, made a net
profit of MK387.42 (US$9.68). The proportion of this is 77.4% of the capital invested which is
greater than 46–52% interest charged by the lending commercial banks in Malawi. This very
high contribution is due to very low inputs injected into small ruminant production compared to
the output. Panin (1997) found similar results. This suggests that small ruminants are
profitable under this system of management.
Table 8. Summary of income analysis in the in Lilongwe, Central Malawi
Income source Total income (MK) % Contribution
A:Net crop income 925.45 11.6
B:Net livestock income 5800.00  
C:Net income: small ruminants 2450.00 30.3
D:Net income: cattle 2850.00 35.6
E:Net income: others livestock 500.00 6.2
F:Net farm income 5775.00  
G:Off farm income 1310.00 16.4
H:Total net household income 8010.45 100.00
The average net household income was MK8010.00 (US$200). About 30% of this was from
small ruminants. Compared to crops which contributed 11.6% to the total net farm income,
livestock contributed 72%. Cattle alone accounted for 35.6%. Although the households were
predominantly goat owners, the contribution of small ruminants surpassed that of crop
production by 17%. This provides ample evidence for the farmers to reallocate resources to
small ruminants. The conclusions may, however, vary depending on the farm enterprise
combination.
Therefore, more technologies and development initiatives should be implemented in order to
reduce poverty and contribute to nutritional and incomes of the rural households through goat
production. These results could be used for decision making on resource allocation by farmers
themselves, policy makers and other institutions involved in small ruminant production.
 Conclusions
The study revealed that the Malawi goats in the villages are more prolific, but have a high
mortality rate in rearing stages of males (42%) and that 27% of the suckling stock risk dying
before 12 months of age. The offtake rate was 21% with a risk of 19%. It increased from
young to old stock, and was highest in rearing and breeding stock. These observations are
very high and pose a problem in sustaining flock numbers and making the enterprise
competitive in the crop-livestock farming systems. It is suggested that in order to sustain goat
numbers and make goat production profitable and sustainable, mortality, especially of rearing
stock and offtake of females should be reduced. In addition, improvements in management
should be made to avoid the factors which cause mortality and reduced survival of the goats.
For the marketing system to be sustainable and benefit all the producers, middlemen and
consumers, the producer has to make goats available that are in good condition and age as
well as weight. This calls for improved productivity at the farm level by increasing the
reproduction/breeding and nutrition, management and disease control. This will automatically
reduce meat prices. Middlemen and sellers themselves must endeavour to maintain high
handling and hygienic standards, if consumers are to be satisfied. The sellers and/or
middlemen could form groups and contribute to a revolving fund from sales to improve the
marketing facilities, other than waiting for funds from the government.
 References
Banda J.W. 1992. Genotypic and seasonal influences on milk yield and milk composition of
sheep and goats in Malawi. PhD thesis, Justus-Liebig-University of Giessen, Germany.
Banda J.W., Mtukuso A.P. and Jere J.A. 1998. Small ruminants in smallholder crop/livestock
systems in Malawi Identification and prioritisation of constraints. UNISWA Research Journal of
Agriculture, Science and Technology. 2(1): 180–183.
Kusina N.T., Dzudar L., Chikura S. and Sibanda S. 2000. Goat marketing systems in a crop-
livestock farming system in Wedza Communal Area and consumer preferences for goat meat
in some areas in eastern Zimbabwe. FVSc/DGIC/ITM Workshop on Health, Production and
Management of Small Ruminants. 14–16 August, 2000, Harare, Zimbabwe.
Lovelace C.E.A., Vaz Y., Mukuka J., Simukoko H. and Chitambo H. 1997. The marketing chain
for goats from farms in the Zambezi Valley in Zambia. Technical Report, EU-STD3
Project/Zambia, Lusaka, Zambia.
NLDMP (National Livestock Development Master Plan). 1998. Customary land livestock
census. National Livestock Development Master Plan. Department of Animal Health and
Industry (DAHI), Ministry of Agriculture and Irrigation (MoAI), Lilongwe, Malawi.
Panin A. 1997. Profitability and household income contribution of small ruminants to small-
scale farmers in Botswana. Small Ruminant Research 25: 9–15.
Rischkowsky B. and Steinbach J. 1997. Goat's milk for smallholders in Malawi - A way of
improving the nutrition of the rural population. Animal Research and Development 46: 55–62.
World Bank. 1995. Malawi Food Security Report. Report No. 8151 - MAI, 27 June, 1995.
Prospects of genetic improvement programmes in small
populations: A case of dairy production in Malawi
Mizeck G. G.Chagunda,1, 2 , Clemens B. A. Wollny2 and Erich W. Bruns1
Department of Animal Science, Bunda College of Agriculture, University of Malawi, P.O. Box
219, Lilongwe, Malawi Institute of Animal Breeding and Genetics, Georg-August-University of
Göttingen, Germany.
Abstract
Introduction
Materials and methods
Results and discussion
Productive and reproductive coefficients
Marketing
Socio-economic contribution
Conclusions
References
Abstract
Genetic improvement is inherently a long-term process in which the progress in the future is
built upon improvement in the past (Gibson and Wilton 1998). To this end, setting up of well-
defined breeding goals, organising a well co-ordinated recording system, defining elaborate
breeding strategies, and establishing selection programmes for improving economically
important variables including adaptation traits, are a prerequisite. Genetic evaluation of
breeding strategies to generate information that could be utilised in making breeding decisions
is essential. It is important to set up systems of getting reliable data on the phenotypic and
economic performance of the breeds, improving infrastructure and technical assistance,
developing and installing selection programmes for raising productivity, and optimising
production systems.
Lack of adequately defined breeding objectives and insufficient description of the animal
populations as regards to their phenotypic and genetic performance ability in production,
reproduction and adaptation, are some of the constraints in livestock development in Malawi.
Although some on-station performance testing has been carried out in the past, they have
been limited and discontinuous due to organisational and financial problems. Regular
performance recording under field conditions is rare, but essential for improving and
preserving local populations, which is a main component in the crop-livestock production
systems.
Since Nicholas and Smith’s paper (1983) on nucleus herds, the idea has been intensely
discussed and reviewed by different authors. To date it is known that nucleus schemes, more
so those supported by multiple ovulation and embryo transfer (MOET), can lead to 10 to 20 %
extra annual gain (Lohuis 1998). However, different versions of nucleus schemes need to be
adjusted and modified in order to adapt to prevailing farming conditions and production
systems. The main aspects of recording structure, sampling of farms and family groups, and
performance of small animal populations, with respect to biometrical problems are discussed.
The major benefits and optional parameters of nucleus scheme are reviewed. An open
dispersed nucleus scheme is proposed for the Malawian dairy production system.
(Key words: genetic improvement, dairy, small populations)
Introduction
Genetic improvement is inherently a long-term process in which the progress in the future is
built upon improvement in the past (Gibson and Wilton 1998). To this end, setting up of well-
defined breeding goals, organising a well co-ordinated recording system, defining elaborate
breeding strategies, and establishing selection programmes for improving economically
important variables including adaptation traits, are a prerequisite. Genetic evaluation of
breeding strategies to generate information that could be utilised in making breeding decisions
is essential. It is important to set up systems of getting reliable data on the phenotypic and
economic performance of the breeds, improving infrastructure and technical assistance,
developing and installing selection programmes for raising productivity, and optimising
production systems.
Currently, milk production in Malawi is performed on smallholder farms on the one hand, and
large scale dairy farms on the other. Among others, the major differentiating features are the
holding size, the genotype of cattle raised, and the level of management. Recent information
(Malawi Government 1997) indicates that there are about 3600 smallholder farmers who use
over 6000 Holstein Friesian x Malawi Zebu cows, and about 1700 smallholder farmers who
use an unknown number of Malawi Zebu cattle for commercial milk production in the peri-
urban setting. Although the smallholder farms play an important role in milk production in
Malawi, genetic evaluation studies are currently difficult if not impossible to carry out because
of the unavailability of systematically kept records (Chagunda 1996). In addition to the
smallholder farmers, there are 15 private large scale dairy farms accounting for about 2200
milking cows. The predominant genotype on the large scale dairy farms is the Holstein
Friesian although some of these farms also have Aryshire and Jersey cattle.
The total milk production from both the large scale and the smallholder sub-sectors is
estimated to be 34 million kg per year (Banda and Mwenifumbo 1998). Using the figures of
total milk available on the Malawi market and the country’s human population of about 10.5
million, Banda and Mwenifumbo (1998) estimated the average milk consumption of 3.2 kg per
capita. Official estimates from FAO (1994) put Malawi’s per capita consumption at 6.9 kg. Both
these estimates indicate low milk consumption in Malawi. The average milk consumption for
Africa is 15 kg per capita. This pronounces the need for improving milk production and the
consequent milk consumption in Malawi. This paper reviews some aspects of recording
structure, sampling of farms and family groups, and performance of small animal populations,
with respect to biometrical problems. The major benefits and optional parameters of nucleus
schemes are highlighted and prospects of applying some are discussed and reviewed.
Constraints in genetic improvement programmes
Genetic improvement in the dairy sector in Malawi is hampered by several constraints. Among
them are the following:
1. Lack of adequately defined breeding objectives and insufficient description of the animal
populations as regards to their phenotypic and genetic performance ability in production,
reproduction and adaptation, are some of the constraints in livestock development in
Malawi (Wollny 1995).
2. Although some on-station performance testing has been carried out in the past, they
have been limited and discontinuous due to organisational and financial problems.
Regular performance recording under field conditions is rare, but essential for improving
and preserving local populations, which are a main component in the crop-livestock
production systems.
3. The existence of genotype by environment interaction on some large-scale dairy farms
in Malawi with respect to Canadian imported semen highlights an important feature to be
considered in assessing the genetic merit of Holstein Friesian cows in Malawi
(Chagunda 2000). The estimate of genetic correlation for milk yield of Holstein Friesian
cows in Malawi and Canada was low and the sires ranked significantly different,
suggesting a substantial genotype by environment interaction.
4. Small livestock population sizes, small herd sizes and unreliable animal identification
(Wollny 1995). Although some pedigree records go up to the grandparent generation,
survey records indicate that there is a substantial number of animals with missing
records on parent or grandparent identification (Chagunda 2000). This limits the
estimation of the additive genetic variance, a component used in determining heritability.
Towards a national selection and breeding programme
So far it is clear that environmental factors of nutrition and management play an important role
in milk production in Malawi (Mwale et al. 1999, Chagunda 2000). These factors are normally
masked within the herd and year of calving (Schmidt and Van Vleck 1974). The high milk
production potential of Holstein Friesian cows may not have been fully expressed due to the
environmental factors. This indicates that considerable improvement could be achieved by
improving the production environment. The same applies to the relatively long calving
intervals, which were observed in the study. A need is therefore apparent to identify the most
effective management and breeding practices to circumvent the environmental constraints on
genetic expression of Holstein Friesians in Malawi.
Since importation of foreign genetic material involves the use of foreign exchange which is
extremely scarce, and the existence of genotype by environment interaction slows down the
rate of genetic progress (Mpofu et al. 1993a), there is a need therefore to revise the breeding
strategy to find the most suitable bulls for the production system in Malawi.
As Smith (1988) pointed out, if genotype-environment interactions are important across
countries, then an independent domestic genetic improvement programme would be needed
and would be justified. He, however, cautioned that what developing countries should avoid is
the unnecessary duplication of expensive breeding efforts being carried out in other countries
from which they could benefit at little cost.
For the smallholder and communal sector in southern African countries, Wollny (1995)
suggested an open nucleus scheme approach. This approach could be adapted for the large-
scale dairy farms. However, the more appropriate breeding programme for the large scale
dairy farms would be the dispersed open nucleus scheme as represented in Figure 1. By
definition, in a dispersed open nucleus scheme, the animals are not physically located in one
place and there is allowance of inflow of foreign germplasm into the system (Lohuis 1998). As
compared to the centralised closed nucleus scheme, the dispersed open nucleus scheme has,
on the one hand, the following advantages:
a. There is less threat from high inbreeding levels since the nucleus has no size limitations
and the scheme is still open to some percentage of foreign germplasm.
b. The nuclear environment automatically reflects the production environment hence a very
reduced (if any) rate of within-country genotype by environment interaction.
c. There is increased farmer participation since more farmers have a direct impact on the
breeding programme.
d. Farmers still own and control top germplasm.
e. There is less demand for additional investment since there is high utilisation of the existing
infrastructure.
On the other hand the centralised closed nucleus scheme has the following advantages:
a. There is more flexibility and options such as collection ore more trait information, protection
of best cows and heifers, and development of ground-parent lines.
b. There is more accurate female rankings in that bias is reduced and also the best females
compete head to head.
c. Technologies like artificial insemination embryo transfer, and in vitro fertilisation are more
effectively used.
Setting up a nucleus breeding scheme in Malawi would mean all the effective breeding and
selection activities would be confined within the nucleus scheme, which would be the source
of male stock in the system. These also proposes some restructuring and adjustment of some
of the current functions of the National Artificial Insemination Scheme and also ensure that the
major constraints that retard the efficiency in the delivery of the artificial insemination services
in Malawi (Chagunda et. al. 1998) are alleviated.
On the other hand, the effective population size should be large enough to sustain continued
genetic improvement over many generations (Smith 1988). An effective population of 72 per
generation would ensure a sustainable programme. Nucleus stocks would be open
continuously to the introduction of higher genetic merit from any source to increase genetic
variation and reduce inbreeding levels. Currently, the strategy in the large scale farms is
based on 100% continuous semen importation. In the nucleus scheme a lower percentage of
imported semen is proposed. In a simulation study, Mpofu et al. (1993b) evaluated breeding
strategies for commercial dairy cattle in Zimbabwe from a genetic and also an economic
viewpoint. In that study, the authors compared local programmes based on progeny testing in
a closed population, progeny testing combined with semen importation to sire 30% of the
cows, progeny testing with foreign sires as sires of bulls, and a closed nucleus scheme using
embryo transfer started from elite imported cows. Further strategies, which were evaluated,
were based on continuous importation of semen for 30, 50, or 100% of the cows and semen
from elite foreign bulls used on elite local cows and, finally, bulls from elite imported embryos.
The results of their study showed that continual semen importation for 30% of the population
was better in the initial period than using progeny testing schemes initially. Since the rate of
imports depends on the immigration rate and the initial difference in the genetic mean between
the exporting and the importing populations, need is there to determine the actual level of
importation suitable for the Malawian production system.
The optimal family size in a half-sib analysis can be calculated as, n = 4/h2. If heritability (h2)
for milk yield is assumed to be 0.25 the optimal number of half-sibs (progeny) per sire should
be 16. Assuming progeny of 20 sires are available then the standard error of the estimated
heritability will be (32 h2 /T)½  whereby T is the total number of cows recorded, in this case
being 320. This gives the standard error of the heritability estimate of 0.16. This indicates that
with half-sib analysis, the accuracy of the heritability estimate improves with the increase in
number of the half-sibs groups assuming the same number of sires. With the current situation
where there is no united recording system and sourcing of sires in Malawi, this number can
only increase if a nucleus breeding scheme would be in place. This is because nucleus
breeding encourages utilisation of the same elite bulls and also proper record keeping.
Phenotypic herd comparison is a sound management tool for identifying sub-optimal
management procedures. The number of animals per farm that would warrant such a
comparison is calculated as follows: n = k (c/d)2  where, k = constant depending on the
significance level and the probability of detecting the difference (@ 20); c = coefficient of
variation (%); d = important difference between the group means to be detected ( % of
the mean).Table 1 gives an example of such a calculation with known parameters from a
previous study on a Holstein Friesian herd in Malawi (Chagunda 1996).
Table 1 Estimates of number of animals required for Holstein Friesian herd average
comparison in Malawi
Number of flocks 43
Total number of parturitions 200
Number of first parturitions 29
Annual parturition rate (%) 82.6
Number of kids born per parturition 1.3
First parturition before 18 months of age (%) 79
First parturitions between 18 and 24 months (%) 21
Males surviving to 6 months of age (%) 88.2
Females surviving to 6 months of age (%) 88.2
* Add to more than 100% due to the combination of animal classes.
If the between-herds difference for milk yield is to be detected and declared significant at 5%
then up to 659 cows per farm are necessary. For age at first calving and calving interval, 104
and 356 cows are required, respectively. The combination of production and reproductive
traits like calving interval by applying the basic principles of selection index (Hazel 1943)
would need, therefore, to be supported and applied in the nucleus programme (Chagunda et
al. 1998).
This programme would only function well with the full participation of the dairy farmers (both
large-scale and smallholder) and all institutions involved in dairy production. This would help
pool efforts, existing facilities and technical know-how together for the effective and efficient
establishment of the programme.
 Conclusion
Setting up of well-defined national breeding goals, organising a well co-ordinated recording
system in the large scale dairy and some smallholder farms, defining more elaborate breeding
strategies, and the application of a properly constructed selection index are a prerequisite to
dairy improvement in Malawi. It is important to set up systems of getting reliable data on the
phenotypic and economic performance of the breeds, improving infrastructure and technical
assistance, developing and installing selection programmes for raising productivity, and
optimising production systems.
 References
Banda J. W. and Mwenifumbo C. 1998. Dairy production and marketing. Working Paper.
National Livestock Development Master Plan. Ministry of Agriculture and Irrigation. Lilongwe,
Malawi.
Chagunda M. G. G. 1996. Evaluation of the Malawian artificial insemination program for small
and large scale dairy farms. MSc thesis, University of Malawi, Zomba, Malawi.
Chagunda M. G. G., Wollny C. B. A., Bruns E. and Kamwanja L. A. 1998. Evaluation of the
artificial insemination program for small scale dairy farms in Malawi.
Chagunda M. G. G. 2000. Genetic evaluation of Holstein Friesian cattle on large scale dairy
farms in Malawi. PhD thesis, Institute of Animal Breeding and Genetics, Georg-August-
University of Göttingen, Germany. Cuvilier Press, Göttingen. 114 pp.
FAO (Food and Agriculture Organization of the United Nations). 1994. Yearly Handbook,
Rome, Italy. 235 pp.
Gibson J. P. and Wilton J. W. 1998. Defining Multiple-Trait Objectives for Sustainable Genetic
Improvement. Journal of Animal Science 76: 2303–2307.
Hazel L. N. 1943. The genetic basis for constructing selection indices. Genetics 28: 476 – 490.
Lohuis M. 1998. Establishment and use of nucleus herd schemes for genetic improvement in
dairy cattle. Paper presented at Congress CAAB/CETA Convention, Saint-Hyacinthe, Quebec,
August 30–September 2, 1998. Available at: http://www.aps.uoguelph.ca/lohuism/
CAAB/CAAB.html. Accessed Dec. 2, 1999.
Malawi Government. 1997. Annual Livestock Census. Department of Animal Health and
Industry, Ministry of Agriculture and Livestock Development. Lilongwe, Malawi.
Mpofu N., Smith C. and Burnside E.B. 1993a. Breeding Strategies for genetic improvement of
dairy cattle in Zimbabwe. 1 Genetic evaluation. Journal of Dairy Science Handbook76: 1163–
1172.
Mpofu N., Smith C., Van Vuuren W. and Burnside E. B. 1993b. Breeding Strategies for
genetic improvement of dairy cattle in Zimbabwe. 2 Economic evaluation. Journal of Dairy
Science Handbook 76:1173–1181.
Mwale S.E., Wollny C., Banda J.W., Chagunda M.G.G. and Bruns E. 1999. Evaluation of
Malawi Zebu and its crosses on smallholder dairy farms in Mzuzu, Northern Malawi. Paper
presented at Tropentag 99, Humboldt University of Berlin, 14–15 October, 1999.
Nicholas F.W. and Smith C. 1983. Increased rates of genetic change in dairy cattle by embryo
transfer and splitting. Animal Production 36: 341–353.
Schmidt G. H. and Van Vleck L.D. 1974. Principles of Dairy Science. W. H. Freeman and
Company. San Francisco, California, USA. 288 pp.
Smith C. 1988. Genetic improvement of livestock using nucleus breeding units. World Animal
Review 65: 2–10.
Wollny C. 1995.
Breed improvement and future breeding strategies in livestock in Southern Africa.Proceedings
of the International Symposium of Animal Production through Breeding and Genetics. Dzama,
K., Ngurume F.N., and Bhebhe, E. (Ed), Harare, Zimbabwe, 10th – 11th May 1995, University
of Zimbabwe, Harare, Zimbabwe. pp 61–70.
Wollny C. B. A., Namwaza A. G, Makamba T. S. W. and Küttner K. 1998. Zum Stand der
Rinderzucht in Malawi. Archiv fur Tierzuct. Dummerstorf 41(1/2):33–44.
Relationship between annual rainfall oscillations and mohair
production between 1935 and 1996
Ng'ambi, J.W. and L. Khit_ane
National University of Lesotho, P.O. Roma 180, Lesotho
Abstract
Introduction
Methodology
Site
Data collection
Data analysis
Results
Discussion
Conclusion
References
 Abstract
Angora goats in Lesotho depend mainly on communal rangeland pastures which depend on
rainfall for their growth. The main objectives of this study were to determine trends in annual
rainfall oscillations and mohair production, and calculate the relationship between annual
rainfall oscillations and mohair production in Lesotho between 1935 and 1996.
An exponential regression equation of the form LnY = LnA + bx was used to estimate trends.
Between 1935 and 1965 Angora goat numbers, mohair production and mohair yield per goat
increased at annual rates of 1.1, 1.9 and 1.1%, respectively (P < 0.01). During years of
independence (1966 to 1996) annual goat numbers remained largely stagnant at around one
million animals (P > 0.05). However, Mohair yield per goat declined at an annual rate of 1.2%
to around 0.85kg in 1996 (P<0.01). Similarly, mohair production declined at an annual rate of
1.2% to 970,000kg in 1996 (P<0.01).
A long-term (1935 to 1996) annual rainfall mean of 700mm was calculated for Lesotho. There
were recurrent wet years (rainfall above long-term mean) and dry years (rainfall below long-
term mean). However, no clear alternate rainfall oscillations of wet and dry years were
observed. It was thus difficult to predict years of drought (rainfall below long-term mean) from
alternate annual rainfall oscillations. A positive but non-significant (P > 0.05) relationship
(r=0.261) between annual mohair yield per goat (kg) and rainfall (mm) was observed, possibly
indicating that there were other major factors that affected mohair yield per goat.
Keywords: Rainfall oscillations; mohair production trends; relationships.
 Introduction
The agricultural sector is the largest single sector in the economy of Lesotho. It contributes to
income and employment for more than 85% of the population (Eckert and Nobe 1982; Bureau
of Statistics 1987). In 1990, agriculture contributed 23.9% of the Gross National Product.
Seventy five percent of this contribution came from the livestock sector (Phororo, Kidman and
Ponzoni 1993).
Angora goats have played an important role in the economy of Lesotho for a long time. They
provide an important source of cash income through the sale of their mohair, their meat or the
animals themselves. In the food system, Angora goats provide meat which is a high quality
protein food for Basotho people (Phororo 1979; Lawry 1986; Hunter 1987). Mohair sales
contribute between 14 and 20% of the total income generated by the livestock sector in
Lesotho (Hunter 1987). In 1990, more than R12 million was realised from mohair sales
(Phororo, Kidman and Ponzoni 1993). However, most studies indicate that annual mohair
production, mohair yield per goat and Angora goat numbers have declined or remained low
and stagnant over the past 30 years (Uys 1977; Wyeth et al. 1983; Hunter 1987; Belete et al.
1994). These studies listed poor nutrition as being the most important cause of low annual
mohair production, mohair yield per goat and Angora goat numbers.
Communal rangelands play an important role in the livestock production system in Lesotho
(Phororo 1979). Thus, Angora goats depend mainly on natural pastures in these rangelands
for their nutritional requirements. These natural pastures, in turn, depend on rainfall for their
growth. Reduced amounts of rainfall or droughts have a negative effect on pasture growth,
resulting in less food for the animals and hence low mohair yield per goat (Hunter 1987). Thus,
each time there is drought, farmers in Lesotho are caught unaware and hence lose a lot in
terms of a decline in annual mohair yield per animal and deaths of animals. Knowledge of
rainfall oscillations and the relationship between rainfall and mohair production in Lesotho
would help farmers prepare themselves for drought years. However, such information has not
been generated in Lesotho.
The objectives of this study were:
To determine trends in production and interactions between annual Angora goat
numbers, mohair production and mohair yield per goat in Lesotho between 1935 and
1996.
To determine the annual rainfall oscillations in Lesotho between 1935 and 1996 and
estimate their possible impact on mohair production.
 Methodology
 Site
The study was carried out in Lesotho. Lesotho is a semi-temperate country with cold winter
and warm summer months. The temperature may range from –20 to 15.9 °C in winter and
from 24 to 32 °C in summer. The average annual rainfall is between 600 and 1775 mm, with
most of the rain falling in summer (Bureau of Statistics 1987). Owing to its mountainous
character, much of Lesotho is well suited for Angora goat and sheep production (Hunter
1987).
 Data collection
Secondary data on annual rainfall, Angora goat numbers, mohair production and mohair yield
per goat in Lesotho were used in this study. Major sources of this data were government
reports, annual statistical bulletins, published research articles and books. The data collected
covered the period between 1935 and 1996. However, trends in the above factors were
calculated for the colonial period (1935 to 1965) and years of independence (1966 to 1996).
These are distinct periods in terms of government policies, management practices and
livestock production systems (Hunter 1987).
 Data analysis
An exponential regression equation of the form LnY = LnA + bx was used to estimate trends in
annual Angora goat numbers, mohair production and mohair yield per goat in Lesotho (Steel
and Torrie 1960). In this equation 'LnA' is the intercept, 'b' is the slope and 'x' is the time
indexed to 1, 2, up to 31. Correlation analysis was used to estimate the relationships (r)
between the above production parameters.
Mean annual rainfall in Lesotho between 1935 and 1996 was calculated. Annual rainfall data
was then calculated as deviations above or below the 62-year mean. This data was plotted to
observe variations above and below the long-term mean. A ten-year weighted moving
average was used as a simple smoothing and generalising device (Dyer and Tyson 1977;
Eckert and Nobe 1982). Patterns of wet (annual rainfall above long-term mean) and dry
(annual rainfall below long-term mean) years were observed.
 Results
Between 1935 and 1965 Angora goat population in Lesotho increased (P<0.01) from 501,000
to 1,018,000 at an annual rate of 1% (Table 1). Mohair production and mohair yield per goat
increased at annual rates of 1.9 and 1.1%, respectively (P < 0.01). There was some significant
(P<0.01) indication that during that period, growth in mohair production depended on annual
growth in both Angora goat numbers and mohair yield per goat (Table 2).
During years of independence (1966 to 1996) annual goat numbers remained, largely,
stagnant (P > 0.05) (Table 3). However, during this period mohair yield per goat declined at an
annual rate of 1.2% to a value of around 0.85kg (P < 0.01). The end result was a decline in
mohair production at an annual rate of 1.2% to 970,000kg in 1996 (P<0.05). There was some
relationship between the decline in annual mohair production and the decline in both annual
goat numbers (r=0.658) and mohair yield per goat (r=0.453) (Table 4).
Figure 1 presents a graphical display of the annual rainfall results. Annual rainfall is shown as
deviation above or below the 62-year mean of 700 (+113) mm. The data show that there were
dry (rainfall below long-term mean) and wet (rainfall above long-term mean) years. However,
no clear alternate oscillations of wet and dry years were observed. A positive but non-
significant (P>0.05) relationship between annual rainfall and annual mohair yield per goat was
observed (Table 5).
Table 1. Regression model for annual Angora goat numbers, mohair production and mohair
yield per goat in Lesotho between 1935 and 1965
Variable Parameter, Y = Aebx Growth rates (%)
Angora goat numbers (' 000) A = 588  
 b = 0.010 1.0
 r = 0.619  
 P = 0.000  
Mohair production (' 000kg) A = 427  
 b = 0.019 1.9
 r = 0.614  
 P = 0.000  
Mohair yield/goat (kg) A = 0.662  
 b = 0.011 1.1
 r = 0.525  
 P = 0.002  
r = Correlation co-efficient
P = Probability value
A = Intercept
b = Slope or growth rate
 
Figure 1. Annual rainfall deviations from long-term mean of 700 mm in Lesotho between 1935
and 1996.
Table 2. Relationship between annual Angora goat numbers and mohair production and
between annual mohair yield per goat and mohair production in Lesotho between 1935 and
1965
Variables Relationship
Annual goat number and mohair production r = 0.732
 p = 0.01
Annual mohair yield/goat and mohair production r = 0.695
 p = 0.01
r = Correlation co-efficient
P = Probability value
Table 3. Regression model for annual Angora goat numbers, mohair production and mohair
yield per goat in Lesotho between 1966 and 1996.
Variable Parameter, Y = Aebx Growth rates (%)
Angora goat numbers (' 000) A = 950  
 b = 0.005 –0.5
 r = 0.257  
 P = 0.163  
Mohair production (' 000kg) A = 836  
 b = –0.012 –1.2
 r = 0.394  
 P = 0.030  
Mohair yield/goat (kg) A = 1.089  
 b = –0.012 –1.2
 r = 0.498  
 P = 0.004  
r = Correlation co-efficient
P = Probability value
A = Intercept
b = Slope or growth rate
Table 4. Relationship between annual Angora goat numbers and mohair production and
between annual mohair yield per goat and mohair production in Lesotho between 1966 and
1996
Variables Relationship
Annual goat number and mohair production r = 0.658
 p = 0.01
Annual mohair yield per goat and mohair production r = 0.453
 p = 0.05
r = Correlation co-efficient
P = Probability value
Table 5. Relationship between annual mohair yield per goat (kg) and annual rainfall (mm) in
Lesotho for the period 1935 to 1996
Variables Relationship
Annual mohair yield and annual rainfall r = 0.261
 p = 0.381
r = Correlation co-efficient
P = Probability value
 Discussion
Annual mohair production can be influenced by the number of goats shorn and mohair yield
per goat (Drummond 1985; Hunter 1987). Between 1935 and 1965 annual goat numbers and
mohair yield per goat increased. The overall result was that there was an increase in annual
mohair production. Thus, annual mohair production depended on growth in both goat numbers
and mohair yield per goat. The growth in goat numbers may have been possible during that
time because the communal rangelands were still able to sustain more animals (Hunter 1987).
However, the increase in goat numbers and other livestock species resulted in overstocking
and degradation of the communal rangelands (Phororo 1979; Hunter 1987). This, in turn, led
to the communal rangelands being unable to sustain good levels of livestock production
(Hunter 1987). Thus, between 1966 and 1996 annual goat numbers remained largely stagnant
at around one million (P > 0.05) while mohair yield per goat declined at an annual rate of 1.2%
(P < 0.05). Indeed, it could be argued that the future prospects for increasing mohair
production in Lesotho may depend, mainly, on increasing the annual mohair yield per goat
rather than on increasing goat numbers. Unfortunately, annual mohair yield per goat has been
declining over the past 30 years. The present low annual mohair yield mean of about 0.85kg
per goat may be compared with a mean annual yield of about 4.0kg mohair per goat in South
Africa (South African Wool and Mohair Board 1992). Thus, improvements of around 370% are
possible.
The major factors limiting mohair yield per goat in Lesotho have been identified as poor
nutrition, poor management practices, too many old goats and poor genetic mohair production
potential among the goat population (Uys 1977; Wyeth et al. 1983). Previous Lesotho
governments responded to the problems of low mohair yield per goat with a variety of policies
and remedial programmes (Uys 1977; Wyeth et al. 1983; Hunter 1987). These included
compulsory rotational grazing in the communal rangelands, exchange of old goats with young
ones imported from South Africa, introduction of grazing fees, routine dosing and dipping
against parasites, formation of associations, etc. However, it seems that past policies and
remedial programmes did not have a positive impact on annual growth of mohair yield per
goat during the 1966 to 1996 period. It is therefore suggested that in future, attention should
be placed on increasing mohair production through improvements of mohair yield per goat.
Initially, an attempt should be made to quantify the major factors that limit growth in mohair
yield per goat. Then, policies and programmes putting more emphasis on those factors having
higher limiting effects can be instituted.
In the present study recorded data show that there were wet years (rainfall above long-term
mean) and dry years (rainfall below long-term mean). However, there were no clear alternate
10-year oscillations of wet and dry periods as suggested by Eckert and Nobe (1982). In that
study, which involved data from lowland districts of Lesotho, there was some pattern of
alternate 10-year periods of below or above long-term average annual rainfall. However, even
in that study, there were a number of deviant years within wet or dry spells.
It is also clear from the present study that drought (annual rainfall below long-term mean) was
a recurrent feature in Lesotho during the period under study. This is similar to the findings of
Eckert and Nobe (1982). There is, thus, a need for Angora goat farmers in Lesotho to have
resources and knowledge so that they can prepare for droughts and adjust their management
strategies accordingly. Droughts should be seen and treated as part of the whole livestock
production industry and not as something which takes the country by surprise every now and
then. The aim should be to minimise the impact of droughts on livestock production. This will
ensure the industry's sustainability.
Angora goats in Lesotho depend mainly on natural pastures for their nutritional requirements.
These natural pastures depend on rainfall for their moisture requirements and hence their dry
matter production (Phororo 1979; De Waal 1994). However, the relationship between annual
rainfall and annual mohair yield per goat in Lesotho was very poor. Such poor relationships
may indicate that there are other major factors determining annual mohair yield per goat
(Hutchings and Stewart 1953; Eckert and Nobe 1982; De Waal 1994). These include the
rainfall distribution within the year and any intervention by the farmers in terms of feed
supplementation to the animals (Eckert and Nobe 1982; De Waal 1994).
 Conclusion
The communal rangelands of Lesotho are not able to sustain the present Angora goat
numbers, mohair yield per goat and mohair production. It is recommended that in future,
emphasis should be on increasing mohair production through improvement of mohair yield per
goat.
A long-term annual rainfall mean of 700mm was calculated. There were annual rainfall
oscillations of wet (rainfall above long-term mean) and dry (rainfall below long-term mean)
years between 1935 and 1996. Drought (annual rainfall below long-term mean) was a
recurrent feature. It is, thus, important for Angora goat farmers to always have necessary
resources and management skills for drought preparedness. However, the magnitude and
extent of the impacts of such climatic changes on dry matter production of natural pastures
and hence mohair production were not predictable.
 References
Belete A., Ng'ambi J.W., Makoanyane M.P. and Morojele E.M. 1994. Growth trends of wool
and mohair production and export in Lesotho since independence in 1966. University of
Swaziland Journal of Agriculture 3:86–93.
Bureau of Statistics. 1987. Statistical Yearbook. Bureau of Statistics, Maseru, Lesotho.
De Waal H.O. 1994. The effects of long-term variation in rainfall and dry matter production of
veld on the financial position of a beef weaner enterprise. South African Journal of Animal
Science 24:113–118.
Drummond S.B. 1985. Angora goats the northern way. Stoney Lonestone Farm, Michigan,
USA.
Dyer T.G.T. and Tyson P.D. 1977. Estimating above and below normal rainfall periods over
South Africa, 1972–2000. Journal of American Meteorology 16:5–11.
Eckert J.B. and Nobe K.C. 1982. Toward the year 2000: Strategies for Lesotho's agriculture.
Research Report No.10. Ministry of Agriculture, Maseru, Lesotho.
Hunter J.P. 1987. The economics of wool and mohair production and marketing in Lesotho.
Institute of Southern African Studies Research Report No.16, Roma, Lesotho.
Hutchings S.S. and Stewart G. 1953. Increasing forage yields and sheep production on
intermountain winter ranges. U.S. Department of Agriculture Circular, U.S.
Lawry S.W. 1986. Livestock and range management in Sehlabathebe: A study of commercial
resources management. A Research Report, Ministry of Agriculture, Maseru, Lesotho.
Phororo D.R. 1979. Livestock Farming in Lesotho and Pasture Utilisation and Suggested
National Policy. Ministry of Agriculture and Co-operatives and Marketing, Maseru, Lesotho.
Phororo D.R., Kidman P.R. and Ponzoni R.W. 1993. Developing strategic plan for the wool
and mohair industry in Lesotho. UNDP/FAO Technical Mission Report. Maseru, Lesotho.
South African Wool and Mohair Board 1992. Statistical Analysis of Mohair Clip. South African
Wool and Mohair Board, Port Elizabeth, South Africa.
Steel R.G. and Torrie J.H. 1960. Principles and Procedures of Statistics. McGraw-Hill Book
Company, Washington D.C., USA.
Uys D.S. 1977. The Lesotho Mohair Industry: History and Evaluation 1970 and Five More
Years. South African Wool and Mohair Board, Port Elizabeth, South Africa.
Wyeth P.W., Moletsane N. and Motsamai M. 1983. Trends in mohair sales, and reflections
concerning economic incentives for mohair producers. A Research Report. Ministry of
Agriculture, Maseru, Lesotho.
On farm and on station evaluation of growth and reproductive
characteristics of pigeons in Malawi
T.N. Gondwe, F.C. Chilera, R.K.D. Phoya and J.B. Jonazi
Department of Animal Science, Bunda College of Agriculture, University of Malawi, P.O. Box
219, Lilongwe, Malawi
Abstract
Introduction
Methodology
Results and discussion
On-farm production characteristics
On-station production characteristics
Reproduction characteristics in pigeons
Conclusion
Acknowledgement
References
 Abstract
A field study was conducted to evaluate diversity and production performance of pigeons in
Lilongwe and Mzimba Districts of Lilongwe and Mzuzu Agricultural Development Divisions,
respectively. The districts are in two distinct agro-ecological zones. On-farm results initiated an
on-station study at Bunda College of Agriculture to evaluate growth and reproductive
characteristics of pigeons using the village production system. Sixty young pigeons were
collected in pairs from rural areas of Lilongwe and Mzimba districts and these formed an initial
stock for evaluation. These were housed in traditional pigeon type houses and released in the
morning for roaming around. Minimal supplementation with maize grits was provided in the
morning and afternoon, while water was provided in clay pots for ad libitum consumption.
The on-station study showed that pigeons had fast growth rates during the first 10 weeks of
age. Live weights reached 0.3 kg at 15 weeks of age, and thereafter, live weights stabilised or
declined. Sexual dimorphism was not clearly distinct in live weights. These results confirmed
field study observations in which highest live weights were attained when pigeons were
squabs (0.33 kg at less than 10 weeks) and they became lighter afterwards. Most people in
rural areas traditionally consume squabs, leaving adults for breeding. Both field and on-station
findings showed no significant differences in live weights due to location or origin of pigeons.
The on-farm study showed that pigeons are prolific, producing a pair of squabs almost every
month. In the villages, farmers reported that pigeons took 13 weeks to start laying but the on-
station study observed that pigeons started laying after 30 weeks of age.
Pigeon growth and reproductive characteristics have been studied and the results support the
potential of integrating pigeons in other poultry production in order to diversify and improve
food security among rural farmers in Malawi, hence sustaining human livelihood.
Key words: Pigeons, diversity, integration, and potential
 Introduction
Pigeon production has remained neglected in Malawi but has the potential to be promoted or
integrated with other poultry species in the rural areas. This neglect is basically due to the fact
that chickens are the only poultry species that have received considerable attention in Malawi
(GoM 1998). Pigeons are however, widely owned and ranked second to chickens in
prevalence in the villages (Gondwe et al. 1999). Pigeons are mainly kept for consumption with
minimal care. Their existence and diversity show the potential to contribute meat to human
diets. Promoting their production would be a means of contributing animal protein to the diets
of the majority of the Malawi human population that stay in rural areas. This would directly
contribute to increased food security for people who are below the poverty line. Such efforts
could not be attained without due attention to their production, health and reproduction. There
is a need to maintain and improve the diversity in pigeons as the species form part of the rural
poultry integrated production system. In the absence of relevant information in Malawi, the
productive and reproductive performance and potential for pigeons remains unknown. This
calls for the need to study their characteristics and determine technical parameters pertaining
to traits of importance. Studies were therefore initiated to determine the production and
reproduction performance of different phenotypes of pigeons that exist in Malawi. Some of the
results are presented in this paper.
 Methodology
A field study was conducted as part of a survey on biodiversity in poultry in the rural areas of
Lilongwe and Mzimba Districts of Lilongwe and Mzuzu Agricultural Development Divisions
(ADD), respectively, in 1998. The two districts are in two distinct agro-ecological zones of
Malawi with different ethnic tribes. Different types of pigeons were observed and weighed in
the morning, with an age determined from estimates by the farmer. Most reproductive
characteristics were recorded from farmers' observations. All data were recorded on formatted
sheets by enumerators. Kitchen scales were used to determine weights of pigeons.
On-station studies were initiated after compiling the on-farm results and these studies took
place at Bunda College of Agriculture, Students' Livestock Farm. Sixty young pigeons were
purchased in pairs from the villages of Lilongwe and Mzimba as initial stock. These were kept
in specially constructed traditional pigeon type houses (kholas). Pigeons were released during
the day for scavenging with chickens and ducks. Minimal supplementation in the form of maize
grits was provided in the morning and afternoon, just like what was observed in the villages.
Water was provided in special clay pots placed on a raised pole.
Initial live weights and physical characteristics were collected on the arrival of pigeons. They
were then weighed fortnightly using a kitchen scale. Pigeons were individually identified using
numbered wing bands, hence individual data recording was possible.
Quantitative data collected were analysed for means and their standard deviations in
Statistical Analysis System (SAS 1994).
 Results and discussion
 On-farm production characteristics
Table 1 shows flock sizes and mean live weights of adult pigeons (over 20 weeks old)
determined in Lilongwe and Mzimba. There were equal numbers of male and female pigeons
in the flocks. This was basically due to the fact that pigeons are hatched in pairs and these
grow and reproduce. Flock sizes were higher in Lilongwe Agricultural Development Division
(ADD) than in Mzuzu ADD. Mature live weights showed the potential for growth performance
in pigeons. There were no significant differences in weights between sexes or between ADDs.
Within the species though, variations was large, indicated by the large standard deviations.
The components of within species variation could however not be split during this study, but
show the potential for improving the species.
Table 1. Mean live weights of pigeons determined in villages by sex (means (g) and standard
deviation)
Parameter Sex  Lilongwe ADD Mzuzu ADD
  n Mean SD n Mean SD
Flock size per household Female  15.6 9.1  11.0 8.5
 Male  15.3 9.2  11.2 8.8
Live weights Female 195 290 100 335 310 200
 Male 183 330 100 336 310 100
Figure 1 shows the trend in growth of pigeons in the two ADDs by age group. There was no
specific pattern of growth in pigeons. Young pigeons (< 10 weeks old) and growers were
heavier than old ones. Farmers indicated that they consume squabs (young pigeons) and
leave old ones for breeding. Probably farmers take advantage of the large size of squabs
which are more tender to consume.
 Figure 1.
Live weight trend of pigeons by age group.
Note:- Age range in weeks are: squabs (<10), growers (11–20), grower adults (21– 30),
mature (31–52) and old (>52).
 On-station production characteristics
The on-station initial results showed that pigeons grow fast during their first 10 weeks of life.
Most pigeons reached 0.3 kg live weights by week 15 with males weighing significantly
(p<0.05) more than females (Table 2). This is unlike the findings observed on-farm where
sexual dimorphism was not significant.
Table 2. Live weights and weight gain for pigeons raised on station at Bunda (means in g and
standard deviation) by sex and origin.
Parameter Sex  Lilongwe Mzimba
  n Mean SD n Mean SD
        
Initial weight (4.0 weeks) Female 6 219.2b 24.0 21 261.9b 34.1
 Male 7 246.4a 44.8 21 286.9a 24.5
Weight at 15 weeks Female 6 287.5b 20.9 17 283.8b 27.9
 Male 5 310.0a 28.5 20 305.0a 23.8
Live weight gain Female 6 68.3 23.2 17 23.5 36.9
 Male 5 56.0 62.1 20 18.8 26.8
a,b Means between sex within a column with different superscripts differ significantly (P<0.05)
Weight gains at 15 weeks were significantly (P<0.05) different between pigeons of Lilongwe
and Mzimba. Pigeons from Lilongwe had significantly (P<0.05) higher gains than the ones
from Mzimba. The trend in sex was ,however, similar between pigeons from the two areas.
Initial live weights (from day old to 4 weeks) were not taken in these birds. The lower weight
gains in pigeons from Mzimba may imply that adult weight is reached at earlier age than those
from Lilongwe. More data that is being collected will verify the findings.
The trend in growth pattern on-station (Figure 2) follows that observed on-farm in Lilongwe
and Mzuzu ADDs (Figure 1). The two figures show that large weights were attained during the
grower stage (between 5 and 10 weeks of age) in pigeons. Thereafter, live weights stabilised
or declined. This probably supports the fact that farmers consume squabs. It can be
determined therefore that pigeons should be 
Figure 2. Growth trend in pigeons raised on - station at Bunda.
consumed at the ages of 5 to 10 weeks to take advantage of large weights (more meat). With
this, adult pigeons will be left for breeding. Raising pigeons in integration with chickens could
best complement each other in terms of supplying meat at different times. Chickens reach
table size at 22 weeks of age (Gondwe et al. 2000), while faster returns are obtained from
pigeons.
 Reproduction characteristics in pigeons
Table 3 shows some estimated reproduction traits in pigeons obtained on-farm in Lilongwe
and Mzimba. Farmers reported that pigeons start laying eggs within three months. Two eggs
were laid per clutch that usually hatched. This observation suggests that pigeons are a highly
inbred species. Pigeons were noted to be prolific, producing a pair of pigeons almost every
month. The estimated values obtained on-farm did not agree with those determined on-station.
Pigeons took 30 weeks to start laying on-station. More will be verified during subsequent
studies as on-farm data is still inadequate to analyse reproduction.
Table 3. Means and standard deviations for reproduction traits in pigeons, on-farm
Trait Lilongwe Mzimba
 n Mean SD n Mean SD
Age to point of lay (weeks) 56 11.9 4.9 83 13.0 5.0
Number of eggs per clutch 58 2.0 0.0 87 2.0 0.0
Number of clutches per year 56 11.7 4.7 87 8.6 3.2
n = number of respondents
 Conclusion
Results are still inconclusive but have generated initial information needed for further research
in pigeons. There is potential in promoting pigeons to be raised in integration with other poultry
species in rural areas.
 Acknowledgement
Malawi - Norway NORAD project and USAID / MEMP are acknowledged for financial support
to the studies in indigenous poultry at Bunda College of Agriculture.
 References
GoM (Government of Malawi). 1998. National Livestock Development Master Plan, Final
Report. Department of Animal Health and Industry. Ministry of Agriculture and Irrigation,
Lilongwe, Malawi.
Gondwe T.N.P., Ambali A.D.J., Chilera F.C., Lwesya H. and Wollny C. 1999. Rural poultry
biodiversity in Lilongwe and Mzuzu Agricultural Development Divisions, Malawi. Malawi
Journal of Science and Technology 5: 17–25.
Gondwe, T.N.P., Chilera, F.C., Ambali, A.D.J., Lwesya, H. and Wollny, C. and Phoya, R.K.D.
(2000). Characteristics and potentials of rural poultry in Lilongwe and Mzuzu Agricultural
Development Divisions (ADD), Malawi. Paper presented at the 6th University Research
Conference held at Boadzulu Holiday Resort, Mangochi from 21st to 24th February, 2000
SAS (1994). Statistical Analysis System. Users' Manual. Version 6.0 Release 6.12.
SAS Institute, Inc., Cary, NC, USA.
Strengths and weaknesses of smallholder poultry
improvement programmes in Malawi
A. C. L. Safalaoh
Department of Animal Science, University of Malawi, Bunda College of Agriculture, P. O. Box
219, Lilongwe, Malawi.
Email: AndyS@chirunga.sdnp.org.mw
Abstract
Introduction
The smallholder commercial poultry project
The smallholder poultry improvement programme
Program operations
Choice of breed for improvement
Marketing and distribution of birds
Support services
Project risks
Impact of the SPIP
Program evaluation
Suggestions for improvement
The Malawi smallholder poultry model
Project components
Current project status
Conclusion
References
 Abstract
This paper gives a review of three poultry improvement programmes in Malawi viz.: the
Smallholder Commercial Poultry Project (SCPP), the Smallholder Poultry Improvement
Programme (SPIP), and the Malawi Smallholder Poultry Model Project (MSPMP). The SCPP
was aimed at promoting commercial egg and broiler production in peri-urban areas. The
project provided farmers with a ready market for their products. Due to management and
funding problems, the project has since gone into oblivion. On the other hand, the SPIP was
aimed at improving egg and meat production of indigenous chickens through crossbreeding
using the Black Australorp. The programme is hampered by a myriad of problems such as ill-
defined programme objectives and implementation strategies, inadequate availability of
improved stock, lack of appropriate poultry husbandry skills, lack of cheap quality feed,
prevalence of diseases and lack of programme evaluation. The MSPM Project is aimed at
enhancing employment opportunities and income generating activities of smallholder farmers
through poultry production or related activities. With proper planning, monitoring, evaluation
and modification, the MSPM Project has the potential to be used as a viable and relevant
poultry production model in Malawi. This paper focuses on the activities of the three poultry
programmes in terms of their strengths, weaknesses and makes suggestions for improvement.
 Introduction
Poultry are the most widely kept class of livestock in Malawi in both rural and peri-urban areas
(Safalaoh 1992; MOAI 1999). The majority of poultry kept are chickens followed by doves,
ducks, and guinea fowls (Haule and Jere 2000; Safalaoh 1997). The trend in commercial
poultry production has been unsteady due to an unreliable supply of feed, an inadequate
supply of day old chicks and the prevalence of diseases. The MOAI (1999) indicated that
Malawi had a poultry population of about 8.6 million free-range chickens, 1.6 million broiler
chickens and about 0.2 million layers kept under intensive production. Realising the
importance of poultry production, the Malawi Government has embarked on a number of
poultry improvement programmes geared at improving productivity of subsistence and
smallholder commercial poultry production. This paper discusses the Smallholder Commercial
Poultry Project (SCPP), the Malawi Smallholder Poultry Improvement programme (SPIP), and
the Malawi Smallholder Poultry Model Project (MSPMP).
 The smallholder commercial poultry project
The Smallholder Commercial Poultry Project (SCPP) was implemented in the early 80s as a
pilot project to promote commercial poultry production among small-scale farmers in Malawi.
Under this scheme, farmers were provided with a starter pack that comprised day-old chicks,
feed, drugs and equipment (cages, feeders, drinkers) on loan. Farmers were required to sell
their products (eggs and broiler meat) using the government-owned Blantyre, Lilongwe and
Mzuzu Egg Marketing Centres. Farmers therefore had a ready market for their products. Due
to lack of funds, poor management and poor loan recovery mechanisms, the project was
closed. Sadly, there is no documentation regarding what caused the failure of the project and
no efforts have been made to evaluate the project. It is paradoxical to find that most egg
producers in Malawi have problems marketing their eggs yet the now defunct SCPP provided
farmers with a ready market for their products. There is a seemingly urgent need to revisit the
whole poultry products marketing system so that an appropriate market structure can be
established. Formation of poultry farmer co-operatives and/or clubs should be explored and
the Poultry Industry Association of Malawi should take a leading role in this regard.
 The smallholder poultry improvement programme
The SPIP started in the 1950s. The main objective of the programme is to improve productivity
of indigenous chickens (IC) in terms of egg and meat production through crossbreeding with
the Black Australorp (BA). Safalaoh et al. (1996) reported that IC attained 8-week and 20-
week live weights of only 615 g and 2100 g, respectively. The need to improve the productivity
of IC cannot be overemphasised. Studies carried out in Malawi (Safalaoh et al 1996; Mjojo
1983) have shown that crossbreeding the IC with the BA bird can improve 8-week bird weight
and feed efficiency by 16 and 6%, respectively.
 Program operations
 Choice of breed for improvement
The Black Australorp (BA) was chosen as the exotic breed for crossing with the IC. The BA
was chosen based on its dual-purpose nature and its ability to adapt well to tropical
environments such as Malawi. On the other hand, no concerted efforts were made to identify
and rigorously select an IC type that would optimally combine with the BA in order to produce
an ideal crossbreed for the selected egg and meat production traits. For instance, the naked
nake IC has been reported to have a higher meat yield than the dwarf village chicken
(Safalaoh 1997).
Apart from the egg and meat production, other traits such as feed conversion, productive life,
fecundity, quality of eggs (fertilisation and hatchability), growth rate, feed conversion, carcass
yield and quality could have also been evaluated. Once identified, these IC genotypes would
then be conserved so that they do not become extinct.
 Marketing and distribution of birds
Until recently BA birds were sold to farmers at 6 weeks of age at a cost of MK80.00 (MK80.00
= ~US$1.00). This was a deliberate decision made to relieve rural smallholder farmers of
important and critical early life operations such as brooding, feeding and vaccinations. In order
to increase access of birds to smallholder farmers of lower income levels, farmers can now
purchase either day-old chicks or fertilised eggs at K15.00 and K5.00, respectively. Farmers
place the purchased fertilised eggs under broody IC hen for natural incubation. Some
smallholder farmers under the GTZ-supported Integrated Food Security Programme (IFSP) in
Mulanje, Southern Malawi, are already practising this method which is referred to as the BA
Multiplier Programme (Safalaoh 1998).
When birds are ready for sale, farmers are informed through radio advertisements or
announcements made using public speaker vehicles that go around the villages. However,
due to funding problems, such promotional activities are sporadic and done on an ad hoc
basis. In order to decentralise production and distribution of the BA birds, the SPIP
established breeding and multiplication centres at Mikolongwe, Bwemba and Choma in the
southern, central and northern regions respectively.
While birds were initially delivered to each of the distribution centres (in the villages) using
government transport, farmers are now required to provide their own transport for purchase
and delivery of birds to their homesteads. This arrangement has resulted in many farmers not
buying the birds due to lack of transport. Consequently, not all birds produced at the
multiplication and breeding centres are sold at six weeks of age as planned. The price of
MK80.00 per bird has also proved to be prohibitive for the majority of rural farmers (the target
clientele of the programme) most of who live below the poverty line. These factors limit further
growth in demand of BA birds. Failure to sell birds on time has also been costly for the SPIP.
Keeping birds for longer than six weeks results in increased production costs for feed, disease
control and water bills among others. A case has been reported where Bwemba Multiplication
and Breeding Centre had about 5,206 nine-week-old birds that were consuming 650 kg per
day costing the project US$112.94 (Saini, personal communication). With proper planning
such unnecessary costs and expenses could be easily avoided.
 Support services
Provision of services such as vaccines and extension services were an integral part of the
SPIP during inception. However, in an attempt to promote self-reliance and programme
sustainability, farmers are now required to purchase their own drugs and vaccines. Not many
farmers can afford to purchase these vaccines and drugs. Disease prevention is therefore an
exception under smallholder conditions. This has resulted in the prevalence of outbreaks of
diseases such as New Castle Disease (Christiansen 1986). Adequate and reliable extension
services are mainly constrained by a low extension worker to farmer ratio.
 Project risks
The SPIP has two major risks. The first is the abrupt change birds face when they are moved
from breeding centres to the villages. The change from an intensive deep litter system of
management to a free range system is in itself stressful in terms of feeding (nutrition), housing
and general bird welfare. Under such conditions, the performance of the BA birds is heavily
compromised, as the birds may need some time to adjust. Secondly, the birds are exposed to
an environment where outbreaks of diseases such as New Castle Disease come in
horrendous and epidemic proportions.
 Impact of the SPIP
The Impact of the SPIP leaves a lot to be desired. Safalaoh (1998) reported that only 5.3% of
the rural population in Mulanje keep BA birds while approximately 97.3% keep indigenous
chickens. Gondwe (1994) has reported similar observations. Low availability of BA birds has
been attributed to erratic and biased distribution, high mortality rates, unaffordable bird prices
and lack of good quality feed (Safalaoh 1992, Gondwe 1994). There is virtually no evidence of
planned breeding programmes in the rural areas of Malawi. The breeding strategy used is
indiscriminate where chickens themselves are responsible for selection, mating, incubation
and brooding. Under such circumstances, no clear crossbreeding programme can be
implemented with any hope of success.
Another often-overlooked fact is that most farmers in the rural areas prefer to keep hens to
cockerels. However, BA hens tend not to sit on their eggs (Safalaoh 1992) hence no eggs are
hatched. For crossbreeding purposes, the male BA cocks should therefore be introduced in
order to cross with local hens, which naturally sit on their eggs. Beneficial effects of
crossbreeding the BA with the IC include improved weight gain and improved feed efficiency
(Safalaoh et al 1996).
 Program evaluation
Albeit being in existence for more than 40 years, the SPIP has never been evaluated.
Programme evaluation is important for beneficiary impact assessment and modification of the
programme, where necessary, in order to achieve the desired goals/objectives. Additionally,
timely programme evaluation would identify strengths and weaknesses that would be used to
draw up recommendations that would be used to strengthen the SPIP.
 Suggestions for improvement
Several options exist for improving the SPIP. Accessibility to BA birds can be improved by
encouraging farmers to purchase fertilised eggs to be incubated using IC hens. However,
farmers need to have proper poultry feeding, general management and disease control skills
so that bird survival can be improved. Vaccination campaigns of diseases such as New Castle
Disease should be strengthened and drugs for all-important diseases made available at
affordable rates through reasonable cost recovery mechanisms to ensure sustainability.
Mechanisms for reducing the price of six-week old birds should also be sought so that a larger
proportion of the target clientele can afford to buy the birds. Finally, there is an urgent need for
the Malawi Government to evaluate the programme so that the SPIP can be put on the right
course.
 The Malawi smallholder poultry model
Poultry production at the smallholder level is generally characterised by low productivity
(Safalaoh 1997). However, the importance of poultry as a source of income to rural landless
and economically marginalized households in Malawi has recently received tremendous
recognition by both government and non-governmental organisations. With appropriate
improvements to the existing poultry production systems it is envisaged that small-scale
farmers in the rural areas of Malawi can greatly benefit from keeping poultry. In view of this,
the Government of Malawi, through the DANIDA Agricultural Sector Programme Support
Phase II (DASPS II) - Livestock Component, initiated the Malawi Smallholder Poultry Model
(MSPM) Project. This project is an adaptation of the Bangladesh Poultry Model. This project is
aimed at development of a poultry production model to enhance employment opportunities
and income generating activities of smallholder farmers in Malawi (Haule and Jere 2000).
It is envisaged that development of the MSPM would invariably work as a poverty reduction
strategy thereby improving the economic status of the poorest of the poor households with
priority given to female-headed households. A Participatory Rural Appraisal was conducted
before commencement of the project in order to generate information regarding perceptions
and constraints of smallholder farming communities in relation to livestock production with
emphasis on poultry. The project is being implemented in two sites of Lilongwe West Rural
Development Project and Kasungu Rural Development Project.
 Project components
The MSPM is designed as an integrated system that provides all the prerequisite
inputs/supplies and services that are required to maintain and sustain a viable semi-
scavenging poultry production system. The model is designed to synchronise all related
poultry activities hence their interdependency. The poultry model has the following
components:
The Poultry Worker: The poultry worker is responsible for provision of services such as
vaccination of birds and training farmers in managerial skills.
The Feed seller: The function of the feed seller is to supply feed to chicken growers.
The Chicken rearer: The function of the chick rearer is to raise birds up to eight weeks of age
for sale to the key rearer.
The Key rearer: This is the main target of the model. The key rearer keeps local broody hens
and exotic ones. In Malawi, the Hyline is being used for production of table eggs.
Pullet rearer: The pullet rearer raises birds up to 18 weeks of age for sale to the Model
breeders.
Model Breeders: These are supposed to maintain breeding stock for production of chicks for
use by chick and key rearers.
Egg seller: These collect and sell eggs from the key rearers.
Mini hatcherers: These hatch day old chicks which are distributed to chick and key rearers.
 Current project status
Mini hatcherers: In the initial stages of the MSPM Project, the use of mini hatcherers has been
temporarily removed. This has been done due to lack of appropriate incubation equipment for
use at the smallholder level. The use of rice husk incubators is currently being tested at
Chitedze Agricultural Research Station. Considering the fact that rice husks are not readily
available in the pilot project sites, there is a need to look for alternative materials that could be
used as incubation materials instead of rice husks. In the absence of incubators, fertilised
eggs are purchased from commercial hatcheries and sold to farmers. The farmers then place
these eggs on broody IC hens for natural incubation. However, one of the problems with this
approach is that there is no assurance that all the eggs bought from the hatcheries are fertile.
Another apparent problem with regard to sustainability is that the project buys the fertilised
eggs on behalf of the farmers. It is important that farmers themselves should be involved in the
procurement of these eggs so that they should get used to the process, thereby
simultaneously promoting an element of project ownership.
Egg sellers: The main purpose for keeping hens is the egg. As such, the expectation of most
farmers is that they will produce and sell the eggs themselves. However, the MSPM is
designed in such a way that the eggs should be sold by a farmer other than the producer - key
rearer. Preliminary observations have shown that the key rearer would rather sell the eggs
himself and not through the egg seller. The need to thoroughly analyse and evaluate various
components of the MSPM becomes apparent. For instance, could the key rearer and egg
seller components be combined? The views and feelings of the farmers should be part and
parcel of the model before any success is anticipated.
Model breeders: Currently, farmers are using hatched eggs from commercial breeders. Use of
commercial hatcheries as a source of fertilised eggs may not be sustainable. The project
should establish its own breeding stock so that farmers and mini hatcherers can easily access
the fertilised eggs. Initial indications are that the Hyline breed is performing well under village
conditions using the semi scavenging system of production. The performance of the Hyline
breed should be studied thoroughly before it can be recommended for use at the village level
in Malawi. There is also a need to assess how crossbreeds from the local chickens with the
BA or other exotic breeds can perform under conditions as stipulated in the MSPM.
The Feed seller: Feed is a very important component of the MSPM as it is one of the major
constraints to poultry production in Malawi (Safalaoh 1992). Currently, farmers are using feed
purchased from commercial millers. This feed is generally expensive and in the absence of a
sound economic base, the sustainability of such an approach is questionable. Another flaw of
the approach is that the project is responsible for the purchase and delivery of feed to the
farmers. There is a need to evaluate the project in these early stages to assess if the farmers,
without project intervention, can effectively manage and sustain the project. It might be
prudent to consider teaching feed sellers how to mix their own feed on the farm. This would
ensure that feed is always available on the farm. The farmers should also be taught to grow
crops such as soybeans that form a large proportion of poultry feeds.
It is therefore of paramount importance that the MSPM Project should be accorded all the
financial and technical support so that it can accomplish its laid out objectives. There is a
need for an ongoing monitoring and evaluation system so that timely project modifications can
be made and implemented in order to suit prevailing conditions and farmers' aspirations.
 Conclusion
Poultry production is, and will continue to be, an integral part of life in Malawi. For any poultry
improvement programme to succeed, there is a need to take into consideration the prevailing
situation of the farmers and their aspirations and ambitions. Proper planning and constant
monitoring and evaluation are prerequisites. There is a need to review the Smallholder
Commercial Poultry Project concept in line with the newly introduced Malawi Smallholder
Poultry Model. The failures associated with the SCPP may be applied to the MSPM and
appropriate corrective steps taken accordingly. Revival of egg marketing centres would
provide farmers with a ready market for their products. Components of the MSPM should be
thoroughly examined in terms of their applicability and suitability at the local level. More
importantly, the Malawi Government should always evaluate its projects before embarking on
new ones. Under normal situations, it would have been expected that the Smallholder
Commercial Poultry Project and Smallholder Poultry Improvement Programme could have
been evaluated before commencement of the Malawi Smallholder Poultry Model Project.
 References
Christiansen N.H. 1986. Disease and mortality in poultry in the Shire Highlands of Southern
Malawi. World's Poultry Science Journal 42:243–247.
Gondwe T.N. 1994. Prevalence of black Australorp chickens in extension planning areas 11
and 12 of Lilongwe ADD. BSc thesis Project Report, Bunda College of Agriculture, University
of Malawi, Lilongwe, Malawi.
Haule A.P. and J. Jere 2000. Report on Participatory Situation Analysis of Livestock
Production Smallholder Farmers in Lilongwe West and Kasungu Rural Development Project.
Department of Animal Health and Industry. DASPS II - Livestock Component.
MOAI (Ministry of Agriculture and Irrigation). 1999. Final Report, National Livestock
Development Master Plan. Department of Animal Health and Industry.
Mjojo D.P.K. 1983. Comparison of Reproductive and Productive Performance of Black
Australorp and Local Chickens, Degree Projects, Bunda College of Agriculture, University of
Malawi.
Safalaoh A.C. L. 1992. A Review of Poultry Production in Malawi: Constraints and Possible
Solutions. 3:15–22. Malawi National Commission for UNESCO. Malawi.
Safalaoh A., Kadigi H. J. S. and Phoya R. K. 1996. Growth performance of the Malawi local
chicken, the Black Australorp Chicken and their crosses at twenty weeks of age. Uniswa Res.
Journal of Agriculture Science & Technology 1(1): 61–66
Safalaoh A.C.L. 1997. Characteristics of indigenous chickens of Malawi. AGRI 22: 61–69.
Safalaoh A.C.L. 1998. Poultry Survey Consultancy Report. GTZ Integrated Food Security
Programme (IFSP), Mulanje, Malawi.
Sheep responses to different stocking rates and seasonal
variation under rangeland conditions of Lesotho
L. M. Mpiti- Shakhane1, G. Y. Kanyama- Phiri1, J. W. Banda1 and W. O. Odenya2
1 Bunda College of Agriculture, Lilongwe, Malawi
2 National University of Lesotho, Roma, Lesotho.
Abstract
Introduction
Materials and methods
Experimental design
Ewe handling and management
Data analysis
Results and discussion
Live weights of Merino ewes
Fleece production and quality
Birth weight, growth rate and mortality of lambs
Conclusion
Acknowledgements
References
 Abstract
A study was conducted at Lekubane Rangeland Research Station in Lesotho to determine
sheep responses to stocking rates over a one-year period. The experimental design consisted
of three stocking rates namely, light (3.5 ewes ha-1 year-1), moderate (5.6 ewes ha-1 year-1)
and heavy (14 ewes ha-1 year-1) in a Completely Randomised Block Design (CRBD) with
three replicates per treatment. Data collected from ewes included live weights, grease fleece
weight, fibre diameter and crimp frequency. Data collected from lambs included birth weight,
weekly weights, mortality and weaning weights. Weight changes of both ewes and lambs were
calculated.
Ewes in the heavy stocking rate had significantly (p<0.05) lower final liveweights (36.0 kg)
than ewes in the moderate (39.0 kg) and light (40.0 kg) stocking rates. Heavy stocking rate
had significantly (p<0.05) lower fleece weight (2.2 ± 0.1 kg) than moderate and light stocking
rates (2.8 ± 0.1 kg and 2.9 ± 0.02 kg, respectively). Ewes in the stocking rates of 3.5, 5.6 and
14 ewes ha-1 year-1 had fibre diameters of 23.3± 0.8 μm, 24.5 ± 0.7μm and 23.1 ± 0.8 μm,
respectively; staple lengths of 66.4 ± 1.1 mm, 74.9 ± 1.0 mm and 57.7± 1.1 mm, respectively
and crimp frequencies (25mm) of 9.89 ± 0.6, 11.13 ± 0.6 and 12.4 ± 0.6, respectively.
Birth weights of lambs in the light-stocking rate were significantly (p<0.05) higher (3.24 ± 0.1
kg) than in the heavy stocking rate (2.82 ± 0.2 kg). The birth weights of lambs in the moderate
stocking rate (5.6 ewes ha-1) did not differ significantly (3.06± 0.3kg) from those in the heavy
(14 ewes ha-1) or in the light stocking rates (5.6 ewes ha-1). The lambs in the heavy and
moderate stocking rates had significantly (p<0.05) higher mortality (64% and 58%) than those
in the light stocking rate (40%). There were no significant differences in the average weight
gain per lamb between moderate and light stocking rates However, the heavy stocking rate
registered a significantly lower (0.059 Kg) average weight gain than the moderate (0.088 Kg)
or lighter (0.069 Kg) stoking rates. The weaning weights for light, moderate and heavy
stocking rates, were 10.4 kg, 13.3 kg and 9.0 kg, respectively. The average weight gain per
hectare increased significantly (p<0.05) from 20.2 kg through 41.4 kg to 69.4 kg for light,
moderate and heavy stocking rate, respectively.
Under the conditions of this study, it is concluded that the heavy stocking rate significantly
reduced ewe and lamb performances.
 Introduction
The livestock industry in Lesotho has long been a vital component of the economic and social
structure of the country. Basotho (people of Lesotho) have long invested in cattle and have
extracted many valuable products from them including meat, milk, manure and draft power.
Basotho rapidly adopted Angora goats and Merino sheep, which generate cash income
through sales of wool and mohair as main exports for the country. Of the total income
generated by the livestock sector (75%), wool and mohair sales contribute between 14 and
20% (Hunter 1987; Phororo, Ponzoni and Kidman 1993). Cattle, sheep and goats have also
become important in ceremonies and social obligations.
While there is evidence to suggest that livestock remains one of the best investments
available to individual Basotho, there is a cause for concern about the future of the livestock
industry. There is a low-marketed offtake of livestock and livestock products in Lesotho.
Angora goat population and mohair production per goat have been declining at annual rates
of 0.5% and 1.8%, respectively (Bureau of Statistics 1999). The Merino sheep population and
wool production have also been declining at annual rates of 0.2% and 1.5%, respectively
(Bureau of Statistics 1999). This progressive decline over the years is indicative of the
reduced capacity of rangelands in Lesotho to sustain livestock productivity (Buzzard 1995). It
is obvious that at the current stocking rates and range conditions of Lesotho, there will be a
continuous low-marketed offtake of both animals and animal products. Further, with an
increase in annual population growth of 2.6% and an increase in demand per capita for
livestock products, communal rangelands of Lesotho should receive careful management to
ensure their functional and sustainable productivity and thus improve production of livestock
grazing on these rangelands.
The purpose of the study was to establish an appropriate stocking rate, which would lead to
an increase in sheep productivity whilst ensuring conservation of forage resources.
Specifically this study was designed to:
Determine liveweight changes of Merino ewes under different stocking rates;
Determine wool production (grease fleece weight) and wool quality (fibre diameter,
staple length and crimp frequency) of Merino ewes under different stocking rates;
Determine birth weights and mortality of lambs under different stocking rates;
Monitor growth rate of lambs from birth to weaning.
 Materials and methods
A study was conducted at Lekubane Rangelands Research Station in Lesotho from February
1999 to February 2000. The study area was located at 29.5º South Latitude and 27.9º East
Longitude. The elevation ranges from 2275 m above sea level at substation headquarters to
2618 m above sea level at the highest mountain peak. The area is characterised by extremely
low and high temperatures ranging from –20oC in winter to 32oC in summer. Mean annual
precipitation ranges from 800 mm to 1000 mm (Bureau of Statistics 1999). Soils are deep dark
brown mollisols, moderately coarse textured and fairly well drained. The station has two
slopes: the east and west facing slopes. The west facing slopes of the station are dominated
by Themeda triandra and Aristida bipartita and the east facing slopes are dominated by
Festuca caprina and Aristida bipartita. The station has been fenced and stocked at a
moderate rate of 5.6 ewes ha-1 on a year round basis. The condition of the camps is
estimated to be good while most of the surrounding rangeland outside the station is estimated
to be in poor condition based on rainfall distribution (Figure 1).
 Experimental design
The study area had a total of 11.25 hectares of rangeland pasture. This rangeland pasture
was subdivided by barbed wire fencing into nine paddocks. Three paddocks of 2.00 hectares
were stocked at 3.5 Merino ewes ha-1 year-1 while three paddocks of 1.25 hectares were
stocked at 5.6 Merino ewes ha-1 year-1 and the remaining three paddocks of 0.5 hectares
were stocked at 14 Merino ewes ha-1 year-1, representing light, moderate and heavy stocking
rates, respectively. Sixty-three Merino ewes were involved in the study and were divided into
nine groups of equivalent weight. These groups of ewes were randomly allocated to each
paddock. The study was conducted over four seasons, namely, autumn (February to April),
winter (May to July), spring (August to October) and summer (November to January).
The experimental design consisted of three stocking rates laid out in a Randomised Complete
Block Design (RCBD) with three replicates per treatment. The "moderate" stocking rate (5.6
merino ewes ha-1 year-1) was arrived at using the Standard Animal Unit (one mature cow of
equivalent weight of 450 kg with a daily dry matter intake requirement of 12 kg) (Tainton
1981).
 Ewe handling and management
Before the initiation of the trials, all ewes were ear-tagged with plastic tags for ease of
identification. Thereafter, the ewes were de-wormed monthly against internal parasites, using
Valbazine at the rate of 2.5 ml/ 10 kg live weight. Each ewe was sprayed with chlorofenvinfos
every two months against external parasites such as lice and fleas. The ewes were taken to
pasture every morning at 0830 hours and were allowed to graze in their respective paddocks
for 8 hours during the day and then were penned at night. The animals were allowed free
access to coarse salt and water.
Grease fleece for each ewe was weighed during the shearing season in October. Fleece
samples were then taken to Lesotho Products and Marketing Services (LPMS) to be graded
by staple length, fibre diameter and crimp frequency.
Mating of ewes was done between May 5th and June 7th 1999. Lambing was from late
October to late November 1999. Lambs were weighed immediately after birth. Thereafter,
lambs were weighed weekly to monitor their growth rates up to weaning at 12 weeks of age.
Mortality of lambs was also recorded from birth to weaning. It was assumed that the liveweight
gain from birth to twelve weeks of age was linear and constant. Thus average daily gain
(ADG) was calculated by deducting the birth weight from the weaning weight, divided by the
number of days to weaning as follows:
ADG = Weaning weight – birth weight
             Number of days to weaning
 Data analysis
Data on dam live weights, fleece production and quality, and lamb performance were
subjected to Analysis of Variance by the General Linear Model (GLM) Procedure using
Statistical Analysis Computer Package (SAS Institute 1994). The means were separated and
compared using the Least Square Means Procedure.
The model used for analysis of stocking rate effect on liveweight changes, grease fleece
weight, fibre diameter, staple length and crimp frequency of Merino ewes, birth weight, daily
weight gain and weaning weight of Merino lambs was:
Yij = µ + Ri + Sj + b (Wi) + Eij
Where
Yij = liveweight changes, grease fleece weight, fleece diameter, staple length, crimp frequency
of Merino ewes, birth weight, average weight gains and weaning weight of Merino lambs
µ = Overall mean
R th
i = i  effect of the replicate where i =1..., 3
S th
j = j  effect of the stocking rate where j = 1..., 3
b = Partial linear effect of initial live weight on animal performance
WI = Initial live weight (a covariate of subsequent animal performance)
Eij = Error component
The percentage mortality of lambs from birth to weaning age was calculated and compared for
different treatments using the Chi- Square (x2) test.
 Results and discussion
 Live weights of Merino ewes
The live weights of Merino ewes at the beginning and at the end of the study are shown in
Table 1. There were no significant differences in the average live weights of Merino ewes
attributed to the stocking rate at the beginning of the study. This was expected since at the
initiation of the study, the ewes were allocated to stocking rate based on equivalent weights.
However, there were clear indications that these live weights followed seasonal trends. There
was a consistent pattern of weight loss as the season advanced from autumn to winter. This
declining trend occurred across all the stocking rates. The reduction may be ascribed to
reduced forage intake and diet quality caused by the little or lack of rainfall (Figure 1) and very
low temperature (–20oc) during winter months, which resulted in little or no growth of
vegetation. The weight loss observed during spring might also be attributed to lambing, which
coincided with harvesting of fleece. This was more pronounced in the heavy stocking rate than
in the light and medium stocking rates.
Table 1: Least square means of live weights (kg ± se) of Merino ewes in response to different
stocking rates
AVERAGE LIVEWEIGHTS  
Stocking rate Start weight Final weight αAdj. R2
Ewes ha-1 kg ewe-1 kg ewe-1
 Mean ± se mean ± se  
3.5 35.8 4.8 40.0a 4.11 0.93
5.6 36.1 5.0 39.0a 3.75 0.89
14 34.7 4.9 36.0b 4.10 0.93
a, b Means in a column with different superscripts are significantly different (p<0.05)
se = Standard error
αAdj. R2 : Adjusted R2
  Figure 1: Rainfall distribution in Lekubane Rangeland Research Station during January 1999
to February 2000
Final live weights were significantly lower (p<0.05) in the heavy stocking rate treatment than in
the light and moderate stocking rates (Table 1). These observations are similar to results
obtained by Nolan (1972) and Sharrow, Krueger and Thetford (1991) who applied similar
stocking intensities. Nolan (1972) observed a decrease of 8.8 kg mean peak live weight ewe-
1, if stocking rate was increased from 10 to 20 ewes per ha-1. This compares well with this
study, in which a decrease of live weight of 4 kg ewe-1 was observed, if stocking rate
increased from 3.5 to 14 ewes ha-1.
 Fleece production and quality
Least square means for wool production and quality of Merino ewes in response to stocking
rate are presented in Table 2. The average fleece weight per sheep ranged between 2.2 and
2.9 kg. This is in agreement with the results of Ngambi and Belete (1994) who reported the
performance to be between 2.4 and 3.5 kg in Lesotho. Their work, however, never considered
differences in stocking rates.
Table 2: Least square means of wool production and quality of Merino ewes in response to
different stocking rates
Rate Fleece production Fleece quality
 Grease weight Diameter (µm) Length (mm) Crimp (25mm)  
(kg ewe-1) (kg ha-1) **
Ewe ha-1 mean ± se Mean ± se mean ± se mean ± se mean ± se
3.5 2.9 a  0.02 10.2 a  0.03 23.3  0.8 66.4 b 1.1 9.9  0.6
5.6 2.8 a   0.1 15.7 b  0. 1 24.5  0.7 74.0 a 1.0 11.1  0.6
14 2.2 b  0.1 30.8 c  0.1 23.1  0.8 57.7 c 1.1 12.2  0.6
a, b, c Means in a column with different superscripts are significantly different (p<0.05)
** (p<0.001)
se = standard error
Stocking rate had significant effects on fleece production (Table 2). The light and medium
stocking rates had significantly (p<0.05) higher grease fleece weight per ewe than the heavy
stocking rate. These results are supported by the work of White and McConchie (1976) and
Langlands et al. (1984), which indicated that wool production per sheep was reduced when
ewes were stocked at the high rate. Stocking rates of 3.5, 5.6 and 14 ewes ha-1 produced
10.2, 15.7 and 30.8 kg of wool ha-1, respectively.
Fleece lengths differed significantly across the stocking rates. Ewes on the heavy stocking
rate had significantly (p<0.05) shorter fleece length than those on the light and moderate
stocking rates. Similar observations were made by White and McConchie (1976) and
Langlands et al. (1984) who reported that increased stocking rate resulted in a decline in
staple length and fibre diameter and an increase in staple crimp frequency (waviness of the
fleece). However, fibre diameter and staple crimp frequency found in this study did not differ
significantly across the stocking rates. This could be explained by the fact that the stocking
rates used in this study were not significantly reliable to show the effects on fleece quality.
 Birth weight, growth rate and mortality of lambs
Birth weight, daily weight gains, mortality and weaning weights of Merino lambs are shown in
Table 3. The birth weight of lambs in the light stocking rate was significantly (p<0.05) higher
than the birth weight of lambs in the heavy stocking rate. On the other hand, birth weight of
lambs in the moderate stocking rate was not significantly different from the birth weight in
either heavy or light stocking rate. It was however, observed that the lambs in the heavy and
moderate stocking rates registered significantly (p<0.05) higher mortality than those in the light
stocking rate. These results are comparable to previous reports, which indicated that lamb
birth weight decreased as stocking rate increased (Nolan 1972).
Table 3: Least square means of birth weights, growth rate and mortality of lambs response to
different stocking rates
-1 2  
Ewe ha Birth weight Weight gains Weaning Mortality(%) (x  = 7.20 )
(kg) kg lamb-1 kg ha-1 weight (kg) (p < 0.05)
 mean ± se mean ± se mean ± se mean ± se  
3.5 3.24 a 0.1 0.069 a 0.003 20.2 c 0.004 10.4 b 0.1 40
5.6 3.06 a b 0.3 0.088 a 0.001 41.4 b 0.002 13.3 a 0.1 58
14 2.82 b 0.2 0.059 b 0.004 69.4 a 0.01 9.0 b 0.2 64
a, b, c Means in a column with different superscripts are significantly different (p<0.05)
X2 = Chi- square
Stocking rate had no significant effect on daily weight gains of lambs. There was however, a
tendency for daily weights to decrease as stocking rate increased. A similar tendency was
observed by Sharrow et al. (1991).
Lamb production per hectare significantly (p<0.05) increased with increasing stocking rate
(Table 3 and Figure 2). This increase occurred in spite of the declines in the birth weights,
weight gains per lamb and percentage lamb crop. Similar results have been reported by
Timberlake (1994), MacLeod and McIntrye (1997), McCollum III, Gillen, Karges and Hodges
(1999) who noted that increasing the stocking rate from 14 to 20 ewe ha-1 reduced average
birth weight and average weight gain per lamb, but increased average weight gains per
hectare. The medium stocking rate had significantly higher weaning weight than the light and
heavy stocking rates. These results are in agreement with the report of Sharrow et al. (1991),
which indicated that average weaning weight of lambs during the study was significantly
(p<0.05) higher at 7.4 ewes ha-1 than at the 12.4 ewe ha-1 stocking rate.
Figure 2: Weight gains of Merino lambs in response to stocking rate
 Conclusion
The results of this study indicated that the heavy stocking rate significantly (p<0.05) reduced
the birth weight of lambs. The study further demonstrated that the lambs in the heavy and
moderate stocking rates had significantly (p<0.05) higher mortality than their counterparts in
the light-stocking rate. While the lamb production per hectare significantly increased with
increasing stocking rate, the average daily gain per lamb remained similar across the stocking
rates. It can therefore, be concluded that the "moderate" (optimum) stocking rate required in
this study to confirm linearity between gain per lamb and stocking rate was not achieved.
Since the results of this one-year study were inconclusive, it is recommended that, the study
should be repeated. The moderate stocking rate of 5.6 ewes ha-1 year-1 currently used at the
research station should be adjusted upwards until optimal stocking rates that will permit
efficient forage utilisation and sheep productivity are realised. It is further recommended that in
the new study, the stocking rate of 5.6 ewes ha-1 year-1 should be included as a control.
 Acknowledgements
The authors would like to acknowledge Deutsche Gesellschaft fuer Technische
Zusammenarbeit (GTZ) for funding the study and the National University of Lesotho for
providing field and laboratory facilities.
This study would not have materialised without the assistance of the Research Division,
Conservation Division, Livestock and Range Division. The Research Division assisted with
workers to fence the study site, transport to the research site on weekly basis, and the study
site and herdsmen. The Conservation Division surveyed the research site and came up with
the map demarcating the paddocks. The Livestock and Range Division provided transport
during the collection of vegetation data and provided research technicians who assisted in the
data collection and identification of grass species.
The smallholder farmer is gratefully acknowledged for providing Merino ewes used in this
study. The Lesotho Products and Marketing Services is equally acknowledged for grading of
fleece samples.
 References
Bureau of Statistics. 1999. Annual Wool Statistical Bulletin. Bureau of Statistics, Maseru,
Lesotho.
Buzzard R.F. 1995. An introduction to ecological monitoring. Lesotho community natural
resource management project. Range Management Division. pp121–142.
Hunter J.P. 1987. The economics of wool and mohair production and marketing in Lesotho.
ISAS Research No. 16, Roma, Lesotho. pp 15–56.
Langlands J. P., Donald G. E. and Paull D.R. 1984. Effects of different stocking intensities in
early life on the productivity of merino ewes grazed as adults at two stocking rates. Australian
Journal of Experimental Agricultural and Animal Husbandry. CSIRO Division of Animal
Production, Pastoral Research Laboratory, Armidale, NSW 2350, Australia. 24: 124, 34–36.
MacLeod N. D. and McIntyre S. 1997. Stocking rate impacts on the production and economic
performance of steers grazing black speargrass pastures. Rangeland Journal 19 (2): 174–89.
McCollum III, F. T., Gillen, R. L. Karges, B.R. and Hodges, K. M. 1999. Stocker cattle
response to grazing management in tall grass prairie. Journal of Range Management 52:1–7.
Ngambi J. W. and Belete A. 1994. Mohair production trends in Lesotho between 1935 and
1990. An International Workshop on Research and Development of Small Ruminants held in
1994 in Southern Africa, Zimbabwe. pp 1–7.
Nolan T. 1972. Fat lamb production in the west of Ireland. 2. Effects of stocking rates on Lamb
growth rate and on production of lamb carcass meat and wool per hectare. Irish Journal of
Agricultural Research 11:47–55.
Phororo D.R., Ponzoin R.W., Kidman P.R. 1993. Developing a strategic plan for the wool and
mohair industry in Lesotho. UNDP/FAO. Technical Mission Report, Maseru, Lesotho. pp 1–5
SAS (Statistical Analysis System). 1994. User Guide: Statistics, version 6. 10th Edition. 1994.
SAS Institute, Inc., Cary, NC. USA.
Sharrow S. H., Krueger W. C. and Thetford F.O. 1991. Forage standing crop and animal diets
under rotational and continuous grazing. Journal of Range Management 36: 447– 450.
Tainton N. M. 1981. Veld and pasture management in South Africa: Grazing management.
Shuter and Shooter, Pietermarizburg in association with University of Natural Press,
Pietermaritzburg, Republic of South Africa. pp 322–354.
Timberlake J. R. 1994. An analysis of a stocking rate trial at Chobela, Southern Mozambique.
African Journal of Range and Forage Science. 11:51–53
White D. H. and McConchie B. J. 1976. Effect of stocking rate on fleece measurements and
their relationships in merino sheep. Australian Journal of Agricultural Research. S. S.
Cameron Lab., Department of Agriculture, Victoria, Werribee, Australia. 27: 163–174.
Forage responses to different stocking rates and seasonal
variation under rangeland conditions of Lesotho
L. M. Mpiti- Shakhane1, G. Y. Kanyama- Phiri2, W. O. Odenya1 and J. W. Banda2
1 National University of Lesotho, Department of Animal Science, P. O. Box 180, Lesotho.
2Bunda College of Agriculture, Lilongwe, P. O. Box 219, Malawi.
Abstract
Introduction
Materials and methods
Experimental design
Measurement of vegetation composition, biomass production and forage quality
Data analysis
Results and discussions
Establishment of botanical composition
Biomass production
Crude protein content
Conclusion
Acknowledgements
References
 Abstract
A study was conducted at Lekubane Rangeland Research Station in Lesotho to determine
forage responses to stocking rates and season over a one-year period. The experimental
design consisted of three stocking rates namely, light (3.5 ewes ha-1 year-1), moderate (5.6
ewes ha-1 year-1) and heavy (14 ewes ha-1 year-1) and four seasons (autumn, winter, spring
and summer) combined factorially in a Completely Randomised Block Design (CRBD) with
three replicates per treatment. Data collected included biomass production, botanical
composition and crude protein content.
Total biomass production of forages at the end of the study was significantly (p<0.05) lower
(3191 kg ha-1) in the heavy stocking rate than in the moderate (3821 kg ha-1) and light (3834
kg ha-1) stocking rates. Seasonal variation significantly (p<0.05) altered crude protein (CP)
content of the grass species as evidenced by a decline in crude protein from 5.6 % CP in
summer and 7.8 % CP in autumn to 4.2 % CP in winter and 4.3 % CP in spring.
The results of this study suggest that an appropriate stocking rate that would lead to an
increase in sheep productivity while ensuring conservation of forage resources was not
achieved. Under the conditions of this study, it would be concluded that season exerted
greater impact on sward productivity than stocking rate.
 Introduction
Lesotho is a small mountainous country covering an area of 30,555 km2. Its population of 1.7
million is growing at an annual rate of 2.6% (Bureau of Statistics 1999). Owing to its
mountainous character, 75 percent of the land is non-arable, subject to physical limitations
such as low and erratic precipitation, poor soils, rough topography and cold temperatures
making it unsuitable for cultivation, and is therefore best put to use as a rangeland.
Based on carrying capacity estimates reported by the Range Management Division (RMD
1988), the rangelands were estimated to be overstocked by about 45 percent, an increase of
15 percent from ten years earlier. This has led to a severe reduction of available forage
despite the fact that rangelands of Lesotho have grass with a high regeneration capacity. The
current estimation of rangeland carrying capacity indicates that stocking rates in the country
are exceeded by 75 percent (RMD 1998), an increase of 30% since 1988.
It is obvious that at the current stocking rates and range conditions of Lesotho, there will be
continuous low-marketed offtake of both animals and animal products. Further, with the
increase in annual population growth of 2.6% and an increase in per capita demand for
livestock products, communal rangelands of Lesotho should receive careful management to
ensure their functional and sustainable productivity and thus improve production of livestock
grazing on these rangelands. The purpose of the study was to establish an appropriate
stocking rate, which would lead to an increase in sheep productivity whilst ensuring
conservation of forage resources at Lekubane Rangeland Research station in Lesotho.
The purpose of this study was to:
establish botanical composition under different stocking rates.
determine biomass production of species under different stocking rates.
determine quality of species (crude protein) under different stocking rates.
 Materials and methods
A study was conducted at Lekubane Rangelands Research Station in Lesotho from February
1999 to February 2000. It is a mountainous area located at 29.5º South Latitude and 27.9º
East Longitude. The elevation ranges from 2275 m at substation headquarters to 2618 m at
the highest mountain peak. The area is characterised by extremely low and high temperatures
ranging from –20oC in winter to 32oC in summer. Mean annual precipitation ranges from 800
mm to 1000 mm (Bureau of Statistics 1999). Soils are deep dark brown mollisols, moderately
coarse textured and fairly well drained. The station has two slopes; the east and west facing
slopes. The west facing slopes of the station are dominated by Themeda triandra and Aristida
bipartita and the east facing slopes are dominated by Festuca caprina and Aristida bipartita.
The station has been fenced and stocked at a "moderate" rate of 5.6 sheep ha-1 on a year
round basis. The condition of the camps is estimated to be good while most of the surrounding
rangeland outside the station is estimated to be in poor condition based on rainfall distribution
(Figure 1).
 Experimental design
The study area had a total of 11.25 hectares of rangeland pasture. This rangeland pasture
was subdivided by barbed wire fencing into nine paddocks. Three paddocks of 2.0 hectares
were stocked at 3.5 Merino ewes ha-1 year-1 ,while three paddocks of 1.25 hectares were
stocked at 5.6 Merino ewes ha-1 year-1, and the remaining three paddocks of 0.5 hectares
were stocked at 14 Merino ewes ha-1 year-1, representing light, moderate and heavy stocking
rates, respectively. Sixty-three Merino ewes were involved in the study and were divided into
nine groups of equivalent weight. These groups of ewes were randomly allocated to each
paddock. The study was conducted over four seasons, namely, autumn (February to April),
winter (May to July), spring (August to October) and summer (November to January).
The three stocking rates and four seasons were combined factorially in a Randomised
Complete Block Design (RCBD) with three replicates per treatment. The "moderate" stocking
rate (5.6 merino ewes ha-1 year-1) was arrived at using the Standard Animal Unit (one mature
cow of equivalent weight of 450 kg and a daily dry matter intake requirement of 12 kg (Tainton
1981).
 Measurement of vegetation composition, biomass production and
forage quality
The Metric Belt Transect (Schmutz et al. 1982) was used for measurement of vegetation
composition and biomass production (Figure 1).
Figure 1. Rainfall distribution in Lekubane Rangeland Research Station during January 1999
to February 2000
 Crude protein was analysed using spectrophotometric method (Weatherburn 1987). Data
were collected during the first month of every season. Thus, for the autumn, winter, spring and
summer, data were collected in February, May, August and November, respectively.
 Data analysis
Data were subjected to Analysis of Variance by the General Linear Model (GLM) Procedure
using a Statistical Analysis Computer Package (SAS Institute 1994). The means were
separated and compared using the Least Square Means Procedure.
The model used for the analysis of the effect of stocking rate and season on botanical
composition, biomass production and crude protein was:
Yijk = µ + Ri +Sj +(R*S)ij +Tk +(S*T)jk +(R*T)ik +Eijk
where
Yijk = Observation for botanical composition, biomass production, and crude protein
µ = Overall mean
Ri =
th
 i  effect of the replicate where i =1..., 3
S t
j =
h
 j  effect of the stocking rate where j = 1..., 3
(R*S)ij = th
ij  interaction effect between replicate and stocking rate (Error a)
T t
k =
h
 k  effect of season where k = 1..., 4
(S*T) t
jk = h
 jk  interaction effect between stocking rate and season
(R*T)ik = th
 ik  interaction effect between replicate and season
Eijk = Error component (Error b)
 Results and discussions
 Establishment of botanical composition
Table 1 presents least square means for botanical composition of grass species in response
to stocking rate. At least nine species of grasses and forbs were identified in the area. This
shows some biodiversity within the 11.25 ha study area. This biodiversity should serve as an
impetus for sustainability (efficient resource use) especially in Lesotho rangelands where
weather patterns are common.
Table 1. Least square means of botanical composition (% ± standard error) of grass species
in response to stocking rate
Species 3.5 ewes ha-1 5.6 ewes ha-1 14 ewes ha-1 Ecological status
mean ± se mean ± se mean ± se
Andropogon 16.6a      1.8 11.4b    1.8 10.9b   1.8 Decreaser
appendiculatus
Themeda triandra 18.3      1.9 21.7       1.9 23.1     1.9 Decreaser
Aristida bipartita 19.9ab   1.4 22.6a      1.4 18.2b     1.4 Increaser
Festuca caprina 6.5        1.3 5.5          1.3 5.3       1.3 Variable
Elionurus muticus 7.5        1.7 9.3        1.7 8.4       1.7 Increaser
Eragrostis capensis 6.0b        1.0 7.2 b      1.0 11.2a     1.0 Variable
Eragrostis chloromelas 10.0      1.7 10.1      1.7 6.7         1.7 Increaser
Harpochloa falx 2.4        0.4 2.1        0.4 2.4       0.4 Increaser
Heteropogon contortus 2.5b      1.9 8.3a        1.9 4.8ab    1.9 Variable
Other species 1.9a      0.3 1.0b      0.3 2.9b      0.3 –
Forbs 2.4b        0.5 2.5b      0.5 4.6a     0.5 –
Total 100.0 100.0 100.0  
a, b Means in a row with different superscripts are significantly different (p<0.05)
se = standard error
Andropogon appendiculatus significantly decreased as stocking rate increased. The light-
stocking rate had significantly (P<0.05) higher composition of this species than the moderate
and heavy stocking rates. This was expected because this species is a Decreaser. It is
palatable and has a high grazing value. It is, therefore, more preferred by the animals than the
other species categories and decreases under continued grazing (Tainton 1981). Normally
one would have expected Themeda triandra to decrease with increased stocking rate. On the
contrary, the composition of Themeda triandra was not significantly reduced by increased
stocking rate. This anomaly could be explained by the differential heights of the species.
Sheep generally prefer short grasses and forbs to tall grasses. In this study, Themeda triandra
was taller than Andropogon appendiculatus hence the preference over the latter. Botanical
composition of Aristida bipartita was significantly (p<0.05) higher in the moderate than in the
heavy stocked pasture. However, the botanical composition of this species in the light stocked
pasture was not significantly different from that in the moderate or heavy stocked pasture.
Aristida bipartita is an Increaser, extremely unpalatable and is consequently of little grazing
value. The high composition of this species is an indication that the study site was previously
under poor grazing management and incorrect fire management (Tainton 1981). The
composition of Eragrostis capensis was significantly (P<0.001) higher in the heavy stocking
rate than in the light and medium stocking rates. The composition of Heteropogon contortus
increased significantly (p<0.05) as stocking rate was increased from 3.5 to 5.6 ewes ha-1
although it was not significantly different from the composition in the heavy stocking rate. The
ecological status of Eragrostis capensis and Heteropogon contortus was variable, and
therefore the composition of these two species tended to vary under different stocking rates.
Botanical composition of Festuca caprina, Elionurus muticus, Eragrostis chloromelas,
Harpochloa falx and other species (Merxmullera disticha, Eragrostis racemosa, Pennisetum
spacelatum, Trachypogon specatus Eragrostis gummiflua, Tristachya leucothrix, Poa annua
and Cymbopogon excavatus) remained similar across the stocking rates. Forbs were
significantly higher (p<0.05) in the heavy stocking rate than in the light and moderate stocking
rates. Sharrow, Krueger and Thetford (1991); Hassan and Krueger (1980); Du Toit and
Aucamp (1985); Van Oudtshoorn et al. (1992) reported that continuous grazing allows
livestock a considerable amount of dietary selectivity. This selectivity, while allowing animals
to choose a high quality diet, often results in over-utilisation of the most preferred plants
allowing less preferred plants to become rank. The lack of response by the vegetation in terms
of changes in botanical composition to stocking rate in this study, could be an indication of
selectivity under-utilisation of the pasture. Further, most botanical changes take place between
2 to 10 years (Jones et al. 1995).
Figure 2. Tenth-metre- square quadrant for measuring vegetation along a Metric Belt
Transect
 Biomass production
The standing biomass production of individual grass species in the enclosure plots is
presented in Figure 3. This figure indicates that generally, all species registered greater
biomass in 2000 than in 1999 with the highest biomass production recorded for Themeda
triandra and Eragrostis gummiflua. This may be attributable to residual plant material
experienced in 2000 from 1999. The lowest production was recorded for Poa annua in 1999.
This is probably because the species is a poor competitor in mixed pastures (Van Oudtshoorn
et al. 1992).
 
Figure 3. Standing biomass of individual grass species in the enclosure plots during 1999 and
2000.
Key:
T.t = Themeda triandra
A.a = Andropogon appendiculatus
A.b = Aristida bipartita
E.h = Eragrostis chloromelas
E.c = Eragrostis capensis
E.g = Eragrostis gummiflua
H.c = Heteropogon contortus
T.p = Trachypogon spicatus
P.a = Poa annua
The total biomass production of grass species in the enclosures was significantly (p<0.001)
higher than that in the grazed plots (Table 2). When averaged across the species, the total
biomass production of forage at the end of the grazing trial was significantly (p<0.05) lowered
by the heavy stocking rate (Figure 4). 
Figure 4. Total pasture biomass in response to stocking rate and seasonal variation
Thus the biomass production in the heavy stocking rate was 643 kg ha-1 lower than that in the
light stocking rate and 630 kg ha-1 lower than that in the moderate stocking rate. The lowered
herbage production could be due to reduced plant vigour. There were significant (p<0.05)
differences in the forage produced per ewe per grazing season between stocking rates of 3.5,
5.6 and 14 ewes ha-1 (1,917, 1,274 and 456 kg of forage produced per ewe per grazing
season, respectively).
Table 2. Least square means of biomass production (kg ha-1) in the enclosure plots and
under different stocking rates over a period of one year
Stocking rate Month  
Ewe ha-1 February May August November February Total
1999 1999 1999 1999 1999 2000
Enclosure       
3.5 4918 3731 4136 3479 2905 4292a
5.6 4865 4074 4108 3422 2638 3834b
14 4349 3656 3548 2219 2181 b
3821
      3191c
a, b, c Means with different superscripts are significantly different (p<0.05)
These observations are supported by the work of Nolan (1972) in Ireland who applied stocking
intensities of 10, 15 and 20 ewe ha-1 (1,079, 726 and 464 kg of forage produced per ewe per
grazing season, respectively) to a mixed pasture of perennial rye grass (Lolium perenne L.)
and white clover (Trifolium repens L.). Further, Hassan and Krueger (1980) concluded that
stocking density should be based on at least 650 kg forage per ewe (9.9 ewe ha-1) if forage is
to remain vigorous. It is therefore clear that in this study the moderate stocking rate should be
between 5.6 and 14 ewes ha-1. Cook (1970) reported that herbage production on most ranges
could be substantially increased by switching from heavy to moderate or light grazing
intensities particularly for grassland ranges. Du Toit and Aucamp (1985) observed light
grazing to be an economically effective means of improving shortgrass prairie ranges. This
suggests that there is an optimum stocking rate, which should be recognised if maximum
pasture productivity is to be sustained. This optimum stocking rate is the best compromise for
the welfare of both plants and the animals.
When averaged across the stocking rate treatments and the species, it was found that a 44%
utilisation level could maintain both the animal performance and vegetation performance over
a one-year period of study. This was comparable with the study of Vallentine (1990) who
found that a 32% utilisation level gave a high rate of recovery for deteriorated black gram
ranges. Similarly, Martin and Cable (1994) found that an average utilisation level of 40%
maintained the perennial grasses over a 10-year period of study.
 Crude protein content
Table 3 shows the crude protein (CP) content (%) of individual grass species in the enclosures
and under three different stocking rates. Crude protein content of individual grass species in
the enclosure plots recovered after one year was relatively higher than that from the grazed
plots. The low crude protein content of the species in the grazed plots could be attributed to
the effects of grazing which include loss of vigour through depletion of nutrients and
carbohydrate reserves. Generally, crude protein content of grass species did not differ across
the stocking rate. This could be attributed to the fact that the rangeland was under-stocked
and the vegetation was under-grazed, consequently the protein content remained similar in all
stocking rate treatments.
Table 3. Least square means of Crude protein content (% ± se) of individual grass species in
response to stocking rate effect
Species 3.5 ewes ha-1 5.6 ewes ha-1 14 ewes ha-1 Enclosure
mean ± se mean ± se mean ± se mean ± se
Andropogon 8.3bc     0.5 9.3 b     0.5 7.6c    0.5 12.6 a     0.4
appendiculatus
Themeda triandra 7.8 b      0.5 7.3 b      0.5 7.4 b     0.5 11.5 a    1.5
Aristida bipartita 8.1 b      0.4 7.4bc    0.4 6.6 c    0.4 10.3 a     0.9
Festuca caprina 6.5a      1.0 6.4 a     1.0 4.7a     1.0 -
Elionurus muticus 4.6ab     0.8 6.5a      0.8 4.5b     0.8 -
Eragrostis capensis 7.0 b       0.5 5.6c      0.5 5.8c     0.5 10.4 a     1.0
 b       b  b  a     
Eragrostis chloromelas 7.3  0.8 6.8      0.8 7.3     0.8 13.8  1.7
Harpochloa falx 5.7a      0.7 5.6a      0.7 5.0a     0.7 -
Heteropogon contortus 4.2 b     1.0 4.2 b     1.0 5.6 b     1.0 10.9 a     1.1
Other species 2.0 b      0.2 1.6 b       0.2 1.0 c     0.2 15.1 a       1.7
a, b, c Means in a row with different superscripts are significantly different (p<0.05)
se = standard error
Table 4 shows the crude protein content of individual grass species in response to seasonal
variation. 
Table 4. Least square means of crude protein content (% ± se) of individual grass species in
response to seasonal variation
Season February May August November February
Species Mean ±se mean ±se mean ±se Mean ±se mean ±se
Andropogon 10.6a    0.6 6.9b     0.6 5.1c    0.6 8.3b      0.6 11.1a      0.6
appendiculatus.
Themeda triandra 9.4a       0.6 5.6b      0.6 5.1b     0.6 7.9a     0.6 9.7a        0.6
Aristida bipartita 9.2a     0.5 5.2b     0.5 5.0b     0.5 8.1a       0.5 9.3a        0.5
Festuca caprina 8.5a     1.2 5.0ab    1.2 4.3b    1.2 6.4ab     1.2 5.0ab      1.2
Elionurus muticus 5.8ab     1. 3.5b     1.0 3.9ab   1.0 6.2ab     1.0 6.6a         1.0
Eragrostis capensis 9.0 a      0.6 4.6b     0.6 5.3b      0.6 3.4b       0.6 8.4a        0.6
Eragrostis chloromelas 10.0a     1.7 5.8b     1.1 5.6b    1.1 6.9b       1.1 7.3ab      1.1
Harpochloa falx 8.3a       0.9 2.9c      0.9 5.0bc    0.9 3.6c       0.9 7.5ab      0.9
Heteropogon contortus 5.8b       1.3 1.5c     1.3 2.4bc   1.3 3.5bc     1.3 10.2a    1.3
Other species 1.8a      0.2 1.3abc  0.2 0.8c     0.2 1.7ab     0.2 1.8a      0.2
Average 7.8 4.2 4.3 5.6 7.7
a, b, c Means in a row with different superscripts are significantly differe
The table shows that season had a highly significant (p<0.001) effect on the crude protein of
species regardless of the stocking rate. There was a consistent decrease in crude protein
content as the season advanced from autumn through spring. Thereafter, there was a pickup
as the season advanced from spring to summer and from summer to autumn (Figures 5, 6a,
6b and 6c). 
 Figure
5. Crude protein content (%) of grass species under different stocking rates
MONTHS
Figure 6a. Crude protein content (%) of individual grass species under stocking rate of 3.5
Ewes ha-1
Key:
A.a = Andropogon appendiculatus
A.b = Aristita bipartita
F.c = Festuca caprina
E.m = Elionurus muticus
E.c = Eragrostis chloromelas
H.f = Harpochloa falx
H.c = Heteropogon contortus
T.t = Themeda triandra
Ot = Other species
Figure 6b. Crude protein content (%) of individual grass species under stocking rate of 5.6
Ewes ha-1
Key:
A.a = Andropogon appendiculatus
A.b = Aristita bipartita
F.c = Festuca caprina
E.m = Elionurus muticus
E.c = Eragrostis chloromelas
H.f = Harpochloa falx
H.c = Heteropogon contortus
T.t = Themeda triandra
Ot=Other species
Figure 6c. Crude protein content (%) of individual grass species under stocking rate of 14
Ewes ha-1
Key:
A.a = Andropogon appendiculatus
A.b = Aristita bipartita
F.c = Festuca caprina
E.m = Elionurus muticus
E.c = Eragrostis chloromelas
H.f = Harpochloa falx
H.c = Heteropogon contortus
T.t = Themeda triandra
Ot = Other species
This significant decline in crude protein in winter and spring could be attributed to the
extremely low temperatures (–20°C) experienced in winter and relatively little or no
precipitation from May to September. In agreement with the present study, Stoddart and Smith
(1975), Tainton (1981), Heady (1982) and Limbach (1985) reported that forage quality
declines with grass maturation and translocation of nutrients to reserve pools. Stocking rate
becomes manifest during late winter (Black 1978; Ash and Stafford 1996). Similarly, Tainton
(1981) reported that the changes in crude protein and crude fibre are consistent with rainfall
causing a flush of growth and so increasing the proportion of young leaf and twigs high in
protein and low in fibre.
Table 4. Least square means of crude protein content (% ± se) of individual grass species in
response to seasonal variation
Season February May August November February
Species Mean ±se mean ±se mean ±se Mean ±se mean ±se
Andropogon 10.6a    0.6 6.9b     0.6 5.1c    0.6 8.3b      0.6 11.1a      0.6
appendiculatus.
Themeda triandra 9.4a       0.6 5.6b      0.6 5.1b     0.6 7.9a     0.6 9.7a        0.6
Aristida bipartita 9.2a     0.5 5.2b     0.5 5.0b     0.5 8.1a       0.5 9.3a        0.5
Festuca caprina 8.5a     1.2 5.0ab    1.2 4.3b    1.2 6.4ab     1.2 5.0ab      1.2
Elionurus muticus 5.8ab     1. 3.5b     1.0 3.9ab   1.0 6.2ab     1.0 6.6a         1.0
Eragrostis capensis 9.0 a      0.6 4.6b     0.6 5.3b      0.6 3.4b       0.6 8.4a        0.6
Eragrostis chloromelas 10.0a     1.7 5.8b     1.1 5.6b    1.1 6.9b       1.1 7.3ab      1.1
Harpochloa falx 8.3a       0.9 2.9c      0.9 5.0bc    0.9 3.6c       0.9 7.5ab      0.9
Heteropogon contortus 5.8b       1.3 1.5c     1.3 2.4bc   1.3 3.5bc     1.3 10.2a    1.3
Other species 1.8a      0.2 1.3abc  0.2 0.8c     0.2 1.7ab     0.2 1.8a      0.2
Average 7.8 4.2 4.3 5.6 7.7
a, b, c Means in a row with different superscripts are significantly different (p<0.05)
 Conclusion
The study demonstrated that botanical composition of individual grass species was not
significantly affected by the stocking rate. Thus, it is possible that productive native pastures
dominated by perennial grasses can be maintained at these stocking rates without damaging
the range.
The study has clearly demonstrated that biomass production decreased significantly (p<0.05)
with increased stocking rate, with lower biomass being observed under heavy stocking rates
than under moderate and light stocking rates. This could be attributed to the effects of grazing,
which include reduced plant vigour resulting from depletion of carbohydrate reserves and
reduction of photosynthetic tissue.
The stocking rates (3.5 ewes ha-1 to 14 ewes ha-1) applied in this study could not show
significant effects in the nutritive value of grass species. Seasonal variation, however, exerted
great impact in the crude protein of grass species, which significantly (p<0.05) declined in the
winter and spring than in the summer and autumn.
Under rangeland conditions of the one-year study, it would therefore appear that an
appropriate stocking rate that would lead to an increase in ewe productivity while ensuring
conservation of forage resources was not obtained. The implication could be that the stocking
rate of 5.6 ewes ha-1 year-1 currently practised in Lekubane Rangelands Research Station in
Lesotho is not optimal and could not elicit significant effect in forage utilisation.
Since the results of this one-year study were inconclusive, it is recommended that the study
should be repeated. The "moderate" stocking rate of 5.6 ewes ha-1 year-1 currently used at the
research station should be adjusted upwards until optimal stocking rates that permit efficient
forage utilisation are realised. It is further recommended that in the new study, the stocking
rate of 5.6 ewes ha-1 year-1 should be included as a control.
 Acknowledgements
The authors would like to acknowledge Deutsche Gesellschaft fuer Technische
Zusammenarbeit (GTZ) for funding the study and the National University of Lesotho for
providing field and laboratory facilities.
A study such as this would not have materialised without the assistance of the Research
Division, Conservation Division, the Department of Livestock Services, and Range
Management Division. The Research Division assisted with workers to fence the study site,
transport to the research site on weekly basis, the study site and herdsmen. The Conservation
Division surveyed the research site and came up with the map demarcating the paddocks. The
Department of Livestock Services, Range Management Division provided transport during the
collection of vegetation data and provided research technicians who assisted in the data
collection and identification of grass species.
The smallholder farmer is gratefully acknowledged for providing Merino ewes used in this
study.
 References
Ash A.J. and Stafford D.M. 1996. Evaluating stocking rate impacts in rangelands. Animals
don't practice what we preach. CSIRO Division of Wildlife and ecology, Alice Spring NT 0871.
Range Journal 18 : 216–43.
Black W. J. M. 1978. Winter grazing of pasture by sheep. 1. Some effects of sheep stocking
density on performance of pasture including sward recovery, botanical composition and
animal performance assessments, Irish Journal of Agricultural Research 14: 275–281.
Bureau of Statistics. 1999. Annual wool statistical bulletin. Bureau of Statistics, Maseru,
Lesotho. pp1–21.
Cook W.C. 1970. Energy budget of range and range livestock, Colorado State University,
Colorado, USA. Expt. Sta. Bull. TB. 109.
Du Toit P. F. and Aucamp A. J. 1985. Effect of continuous grazing in the Dohne sourveld on
species composition and basal cover. J. Grassl. Soc. South Africa 2 : 41– 45.
Hassan B. and Krueger W. C. 1980. The impact of grazing pressure and season of grazing on
carbohydrate reserves of perennial ryegrass. Journal of Range Management 33: 200–210.
Heady H.F. 1982. Rangeland wildlife management in the tropics. Longman Group Limited.
USA pp 5–24.
Jones R. M., Jones R. J. and McDonald C. K. 1995. Some advantages of long - term grazing
trials, with particular references to changes in botanical composition. Australian Journal of
Experimental Agriculture 35: 1029–1038.
Limbach, W.E. 1985. Introduction to range management . Report to the farm policy committee
of Lesotho Agricultural College. Maseru, Lesotho. pp 25–50.
Martin S. C. and Cable D. R. 1994. Influence of stocking rate on forage production. Journal of
Range Management 35: 174–176.
Nolan T. 1972. Fat lamb production in the west of Ireland. 2. Effects of stocking rates on Lamb
growth rate and on production of lamb carcass meat and wool per hectare. Irish Journal of
Agricultural Research 11:47–55.
Range Management Division. 1988. Lesotho National Rangeland Inventory. Ministry of
Agriculture and Marketing. Maseru, Lesotho. pp 2–45.
Range Management Division. 1998. Lesotho National Rangeland Inventory. Ministry of
Agriculture and Marketing. Maseru, Lesotho. p 33–35.
SAS (Statistical Analysis System). 1994. User Guide: Statistics, version 6. 10th Edition.. SAS
Institute, Inc., Cary, NC. USA.
Sharrow, S. H., Krueger, W. C. and Thetford, F.O. 1991. Forage standing crop and animal
diets under rotational and continuous grazing. Journal of Range Management 36: 447– 450.
Schmutz, E. M., Reese M. E., Freeman, B. N., Weaver, L. C. 1982. Metric belt transect
system for measuring basal plant cover, composition and production of plants. Rangelands 4:
162–164.
Stoddart L.A. and Smith A.D. 1975. Range management. McGraw - Hill, New York. pp. 33–79.
Tainton N. M. 1981. Veld and Pasture Management in South Africa. Shuter and Shooter,
University of Natal Press. Pietermaritzburg, South Africa. pp 325– 334.
Vallentine J. F. 1990. Grazing Management. Academic Press Inc. N. Y. pp. 553–560.
Van Oudtshoorn F.P., Trollope W.S.W., Scotney D.M. and Mcphee P. J. 1992. Guide to
grasses of South Africa. National Book Printers, Capetown ,South Africa. pp. 257– 278.
Weatherburn N. W. 1987. Phenol- hypochloride reaction for the determination of ammonia.
Analytical Chemistry 23: 971–979.
Roasted sorghum as an alternative energy source in pig
feeding
O. Mwale1, R.K.D. Phoya2, T. Mujahid2, J.P. Mtimuni2
1Ministry of Agriculture, Food and Fisheries, P.O. Box 21916, Kitwe, Zambia
2Bunda College of Agriculture, University of Malawi
Abstract
Introduction
Materials and methods
Determination of tannins
Rapid tannin analysis: semi-quantitative method
Follin-Ciocalteu assay technique
Laboratory analysis on feed ingredients
Feed formulation
Housing and management of animals
The feeding trial
Phase I (starter period)
Phase II (grower period)
Data collection
Kleiber ratio
Statistical analysis
Results
Tannin analysis
Starter phase performance
Starter phase performance (ADG)
Effect of initial weight on overall performance
Effect of starter ration on overall performance
Overall performance over the 18 week period
Time taken to reach target weight
Effect of sex on the overall performance
Cost of production of the pigs
Discussion
Tannin analysis
Roasting as a method of reducing tannin content
Starter phase performance
Overall performance over the 18 week period
Average daily gain and the Kleiber ratio
Time taken to reach maturity
Cost of production of the pigs
References
 Abstract
Eight sorghum varieties (Masotongo, Kapile, Kalombo, Kawaladzuwa, Shabala, Misinde, Pilira
l and Pilira 2) were analysed to establish their tannin status and subsequently the higher
tannin variety was used to replace maize diets in a growth trial with pigs.
Tannin (proanthocyanidin) concentrations were determined using the Folin-Ciocalteu assay
technique with Catechin as the standard. The tannin concentrations were as follows: Misindi
(1.193mg/ml), Kapile (0.396mg/ml), Kalombo (0.062mg/ml), Kwaladzuwa (0.284mg/ml),
Shabalala (0.535mg/ml), Masotongo (0.640mg/ml), Pilira-1 (0.469mg/ml) and Pilira-2
(0.448mg/ml). The variety Misinde was tested to be higher in tannins ranging from 0.29–0.99%
catechin equivalents compared to 0.10–0.25% for the rest, and hence was used in feeding
trials. Roasting Misinde on a charcoal flame reduced its tannin level to 0.10–0.25% catechin
equivalents representing a 37% decrease.
The performance of Large-White and local Malawian pigs fed on maize based- diet (Ration1),
raw-sorghum based diet (Ration 2), 50% maize and 50% roasted sorghum- based diet
(Ration 3) and 100% roasted sorghum-based diet (Ration 4) as a major energy source was
investigated. Average daily gains (ADG) and the Kleiber ratio were determined during the
starter phase. For Large-Whites fed on Ration 1, 2, 3 and 4 ADG in kg were not significantly
different at 0.443, 0.525,0.453 and 0.491 and Kleiber ratios 0.483, 0.582,0.491 and 0.535,
respectively. For local Malawian pigs also fed on the above diet in the same order, the ADG in
kg were 0.192, 0.344, 0.126 and 0.359 and the Kleiber ratios were 0.349,0.525,0.466 and
0.497, respectively, significantly lower for Ration 1. In the grower phase, the ADG (in kg) for
Large-Whites on Rations 1 and 2 were significantly higher (P<0.05) at 0.431 and 0.459
compared to 0.364 and 0.379 for those on Rations 3 and 4. In the grower period the Large-
Whites had Kleiber ratios of 0.361,0.374 0.355 and 0.349 while local pigs were at 0.267,
0.327, 0.283 and 0.306 for Rations 1,2,3 and 4 respectively.
The target weight was 55kg for Large-Whites and 45kg for local Malawian pigs. Sorghum-
based rations, both starter and grower, were cheaper to compound. Although roasting of the
sorghum (variety Misinde) reduces tannin levels, it increases costs of the ration and does not
contribute to any significant difference in growth performance. It is therefore concluded that
low tannin sorghum can replace maize as a major source of energy without affecting the
average daily gain or carcass characteristics of Large-Whites or local pigs.
Key words: Low-tannin sorghum diet, Large-Whites and local pigs, growth performance
 Introduction
With an increasing human population, especially in the Southern African Developing Countries
(SADC), the meat requirements have to be increasingly met from diverse forms of production,
which would also contribute towards food security, poverty alleviation and income generation.
The greatest constraint to livestock production in non-ruminants is the cost of feedstuff.
Hence, efforts should be concentrated on using often unconventional feeds as alternate
sources of energy and proteins. In spite of its favoured statistics, maize production alone does
not meet the consumption levels of the Malawian population. Ironically, the same is also
utilised in livestock feed formulations— a scenario which exerts pressure and competition
between human population and livestock industry. Sorghum, on the other hand, a drought
tolerant crop, is mostly grown in the low rainfall areas of the Lower Shire as a staple food.
Brown sorghum varieties with a higher tannin content are not likely to be consumed by human
beings because of an unpleasant astringent taste. The land pressure resulting from an
increasing human population also calls for production systems that are not only intensive and
less demanding on space, but also give quick returns. Hence, pigs qualify when compared to
other smaller stock, because of their fecundity and growth potential. This research therefore
was aimed at reducing the dependence on maize as an energy source in intensive pig
production systems by suggesting the use of a treated high tannin sorghum variety as an
alternative.
The objectives of the study were to:
investigate the effect of replacing maize with a roasted high tannin (RHTS) variety and
raw high tannin sorghum (UHTS) on the performance of both local Malawian pigs and
Large-White breeds with respect to bodyweight changes, feed utilisation, efficiency,
average daily gain (ADG), and time taken to reach the targeted marketing weight of 55
kg for the Large-whites and 45 kg for the local pigs
compare the cost of raising piglets up to marketing weight based on rations of RHTS,
UHTS, maize and their mixtures as energy sources.
 Materials and methods
Eight varieties of sorghum were obtained from Chikwawa district in the Lower Shire Valley of
Southern Malawi, an area that is 400 metres above sea-level, and receives an annual rainfall
of 400–800 mm. Apparently, the low rainfall pattern has prompted the production of sorghum
and millet, which are relatively more drought resistant than maize and many other cereals.
The varieties collected were Kapile, Shabalala, Masoton'go, Kalombo, Misinde, Kawaladzuwa,
Pilira 1 and Pilira 2 of which the last two are newly introduced hybrids while the others are
local varieties.
All eight varieties were subjected to tannin analysis using two methods which included the
colorimetric Folin-Ciocalteu assay and a semi-quantitative rapid analysis technique.
Misinde was identified as the variety with the highest tannin content. Subsequently, this variety
was purchased in bulk for the experiment. Maize grain was obtained from the Field Research
Section of the Crop Science Department at Bunda College of Agriculture while soyabeans
were bought from farmers in surrounding villages.
 Determination of tannins
 Rapid tannin analysis: semi-quantitative method
The method is as presented in the International Crop Research Institute for the Semi-Arid
Tropics (ICRISAT) Technical Manual (1977). The rationale in this tannin analysis method is
that it provides a rapid and convenient visual estimation of the quantity of polyphenols in
sorghum grain, without the use of instrumentation and with the minimum of glassware. This is
a subjective method based on the reduction of ferric ions to ferrous ions by tannins and other
polyphenols, followed by the formation of a coloured ferricyanide-ferrous complex commonly
known as Prussian Blue (Price and Butler 1977). The intensity of the colour formed enables
the tannin content to be determined using a set of standards as reference.
 Follin-Ciocalteu assay technique
This method gives a more accurate measure of tannin content. Readings for absorbance as
estimates of tannin concentration were taken off a spectrophotometer and the tannin
concentration was calculated.
After identifying Misinde as the variety high in tannins, a sample was roasted on a charcoal
flame in a large tray and later analysed for tannins using the two methods outlined. While it
was not possible to measure the temperature of the fire, the roasting was done on a flame hot
enough not to cause charring of the sorghum grain. Once the fire was prepared, soyabeans
were first roasted so that the intensity of the heat subsided.
 Laboratory analysis on feed ingredients
Samples of each ingredient were taken and ground to pass through a 1 mm sieve at the small
animal unit and analysis was done in the Animal Nutrition Research Laboratory. Results are
given in Table 1.
Table 1. Chemical composition of the major ingredients utilised in the rations
Ingredient Dry Matter Ash Crude Protein (%) Ether Extract (%) Gross Energy
(%) (%) (kcal/kg)
Maize 9.02 1.5 8.8 3.9 4436
Sorghum 91.3 1.6 8.9 3.0 3865
Soyabeans 90.6 4.5 36.1 26.3 6172
Maize bran 90.2 3.9 11.8 8.6 4698
Kcal/kg = kilo calories per kilogram
The following proximate analysis was done:
Crude Protein (CP) which was determined by the macro Kjeldal procedure (AOAC 1984)
Dry Matter (DM) was determined in an oven (AOAC 1984)
Ether Extract (EE) was determined using the Soxhlet apparatus (AOAC 1984)
Gross Energy (GE) was done by use of adiabatic bomb calorimeter (Goering and Van
Soest 1970)
Total ash was determined by burning the samples in a muffle furnace (AOAC 1984).
 Feed formulation
Rations were formulated using a linear programming computer package BLP88 (1987) to meet
the nutritional requirements of the pigs at each stage of growth. Inclusion levels of both
sorghum and maize in the rations compounded are shown in Table 2. There were four (4)
starter and grower rations prepared with maize and sorghum as the varying ingredients as
shown in Tables 3 and 4, respectively.
Table 2. Inclusion levels of maize and sorghum as major energy sources in the rations**
Treatments Starter Grower
Ration 1 100% maize 100% maize
Ration 2 100% raw sorghum 100% raw sorghum
Ration 3 50% maize and 50% maize and 50%
50%roasted sorghum roasted sorghum 
Ration4 100% roasted sorghum 100% roasted sorghum
Ration  I =Maize based ration
Ration 2= Raw sorghum-based ration
Ration 3 = 50% maize and 50% roasted sorghum
Ration 4 = Roasted sorghum based
**All other ingredients were the same for the rations
Table 3: Composition of starter experimental diets (%)
Ration l Ration 2 Ration 3 Ration 4  
Ingredient     
Maize 54.52 – 27.26 –
Sorghum – 56.00 27.26 56.00
Maize bran 15.00 15.00 15.00 15.00
Soyabeans 22.92 21.40 22.90 21.40
Fishmeal 4.00 4.00 4.00 4.00
Salt 0.50 0.50 0.50 0.50
Monocalcium Phosphate 0.51 0.50 0.50 0.50
Pig Vitamin Premix 0.30 0.30 0.30 0.30
Dolomitic lime 1.25 1.30 1.30 1.30
DL-Methionine 0.50 0.50 0.50 0.50
Lysine 0.50 0.50 0.50 0.50
Total 100.00 100.00 100.00 100.00
*Cost/kg MK9.58 MK9.25 MK9.44 MK9.25
*Based on prices of May 1998
Table 4: Composition of grower experimental diets (%)
 Ration 1 Ration 2 Ration 3 Ration 4
Ingredient     
Maize 63.00 – 31.50 –
Sorghum – 63.80 31.50 63.80
Maize bran 19.38 18.48 19.38 18.48
Soyabeans 10.80 10.90 10.80 10.90
Fishmeal 4.00 4.00 4.00 4.00
Salt 0.50 0.50 0.50 0.50
Monocalcium Phosphate 0.35 0.35 0.35 0.35
Pig vitamin premix 0.30 0.30 0.30 0.30
Dolomitic lime 1.07 1.07 1.07 1.07
DL-Methionine 0.30 0.30 0.30 0.30
Lysine 0.30 0.30 0.30 0.30
Total 100.00 100.00 100.00 100.00
*Cost/kg MK8.15 MK7.84 MK7.99 MK7.84
*Based on prices of May 1998
 Housing and management of animals
A total of 64 piglets were raised from the breeding stock available at the Bunda College of
Agriculture. There were 32 piglets of the Large-White breed and an equal number of the local
Malawian breed whose average age was 15 weeks + 3 days for the former and 14 weeks + 4
days for the latter. At weaning, all the piglets were given Dectomax, a long-acting injectable,
broad-spectrum paraciticide solution which is recommended for pigs, cattle and sheep. This
was injected at the recommended rate of 300ug/kg or 1ml/33kg body weight. This particular
drug is effective against round worms and lungworms. It is also very effective against
sarcoptic mange mites and lice on pigs.
Pig husbandry practices such as tooth clipping, tail docking and male castration were carried
out. The animals were randomly allocated to a total of 16 pens, each with 4 piglets of the
same breed in it. A completely randomised block design was used. The breeds were used as
blocks. The initial average weight (taken when introducing grower ration) was 10.9 kg for the
Large-Whites and 6.2 kg for the local pigs.
 The feeding trial
Initial weights of the piglets were taken immediately before they were put through their
respective treatments and before feed was offered. The first phase of the feeding trial was the
Starter period and the second was the Grower period.
 Phase I (starter period)
The piglets were given the starter rations as shown in Table 3 and this was fed until the
animals had attained an average weight of 25kg/pen for the Large-Whites and 20kg/pen for
the local pigs before changing to grower diets. The duration of the starter period was 4 weeks
for the Large-Whites and 6 weeks for the local Malawian pigs. All treatments were formulated
in such a way that they were isonitrogenous and isocalorific. For each breed, there were two
pens (each with 4 pigs) for each ration. Over a period of one week they were given 1 kg of
feed per animal per day which was reduced on each subsequent day depending on the
incidence of feed refusals the previous day. This was then changed to 2kg per day per animal
in the grower period when the animals were weighing on average, 30 kg/pen for the Large-
Whites and 25kg/pen for the locals. The meals were given daily in two equal portions
depending on the number of animals in each pen, at 08.00 hrs and at 15.00 hrs.
 Phase II (grower period)
The composition of the grower rations was as shown in Table 4. The animals were weighed
weekly until they reached the targeted weights of 45 kg for the local Malawian pigs and 55 kg
for the Large-Whites.
 Data collection
Parameters of interest were Average Daily Gain (ADG), the time taken to reach maturity and
the Kleiber ratio for growth performance. Feed utilisation efficiency would not be determined
directly due to the absence of individual feeding pens for the pigs. Instead, the Kleiber ratio
was used to approximate feed conversion efficiency.
 Kleiber ratio
The Kleiber ratio was devised by Max Kleiber and is related to the feed conversion ratio as
measured in phase-D tests done with cattle (Bergh 1992). Phase-D tests are performed in
group-fed animals where individual feed consumption cannot be determined. The Kleiber ratio
is calculated as follows:
Kleiber ratio = Average daily gain/Metabolic weight
where metabolic weight is taken as the final mass of the animal during the period of
observation to the power of 0.75.
This was calculated at the end of the starter period for each of the breeds, and also at the end
of the experiment.
 Statistical analysis
Initial, weekly and final weights of the pigs were recorded. Feed consumption was calculated
from the feed offered and feed refused data. Weight gain was determined from the liveweight
records.
Data collected was analysed using Statistical Analysis System (SAS 1989). The performance
parameters were Average Daily Gain, Time taken to maturity and the Kleiber ratio as an
estimator of feed conversion efficiency.
 Results
 Tannin analysis
The results of the tannin analysis expressed as Catechin concentration using the Folin-
Ciocalteu method are shown in Table 5. The variety Misinde was found to have the highest
tannin level at 1.193mg/ml which after roasting dropped to a lower tannin level of 0.752mg/ml,
representing a 37% drop in the tannin level. The classification of the different sorghum
varieties by the rapid analysis method also showed that Misinde was in a higher tannin group
compared to the other varieties as given in Table 6.
Table 5. Tannin content of sorghum varieties by Folin-Ciocalteu method
Sorghum variety  Absorbance Catechin
(nanometers) Concentration
(mgm-1)
Kapile 0.060 0.396
Masotong'o 0.097 0.640
Kalombo 0.062 0.409
Misinde 0.293 1.193
Kawaladzuwa 0.043 0.284
Pilira 1 0.071 0.469
Pilira 2 0.068 0.448
Misinde (roasted) 0.114 0.752
Shabalala 0.081 0.535
Figures are means of four readings taken at wavelength of 675 nanometers
Table 6. Tannin content: semi-quantitative method
Sorghum variety Sample colour Range of catechin Tannin level equivalents (CE)%
Kapile light green 0.10–0.25 low
Masoton'go light green " low
Kalombo light green " low
Misinde blue-green 0.29–0.99 intermediate
Kawaladzuwa light green 0.10–0.25 low
Pilira l light green " low
Pilira 2 light green " low
Misinde (roasted) light green " low
Shabalala light green 0.10–0.25 low
Classification adapted from Price and Butler (1977)
 Starter phase performance
The parameters of interest were the Average Daily Gain (ADG) and Kleiber ratio (Klb1) at the
end of the starter period.
 Starter phase performance (ADG)
The comparison of average daily gain (ADG) and Kleiber ratio as affected by ration for the
Large-White pigs is presented in Table 7. There was no significant difference in performance
among the four treatments administered. A graphical comparison of the different levels of daily
gain for the Large-Whites on the different rations is given in Figure 1.
Table 7. Effect of starter rations on performance of Large-White and local pigs
Large-white pigs Local pigs
lsm sem lsm sem
ADG 0.443 0.041 0.192ab 0.065
Ration l   
Klb1 0.483 0.044 0.349b 0.048
ADG 0.525 0.040 0.344a 0.065
Ration 2   
Klb1 0.582 0.044 0.525a 0.044
ADG 0.453 0.040 0. 126ab 0.079
Ration 3   
Klbl 0.491 0.046 0.466a 0.062
a
ADG 0.491 0.040 0.359 0.070
Ration 4   
Klbl 0.535 0.045 0.497a 0.070
r2 values ADG = 0.59, Klb 1 = 0.71
abc=lsmeans for each parameter per breed with the same superscript do not differ significantly
(P<0.05)
ADG = Average daily gain (kg/day), Klbl= Kleiber ratio
Ration 1 = maize-based ration
Ration2 = raw sorghum-based ration
Ration3 = 50% maize and 50% roasted sorghum
Ration4 = roasted sorghum-based ration
Ism = least square means, sem = standard error mean
Figure 1. Performance of the Large-White pigs in the starter phase.
For the local Malawian pigs on similar diets, the average daily gains were not significantly
different for the pigs fed Ration1, Ration 2 and Ration 4. The rate of gain for the pigs on
Ration 4 was, however, significantly different (P<0.05) from that recorded for Ration3. Table 7
gives a summary of the effects of different rations on the performance of the local Malawian
pigs regarding ADG and the Kleiber ratio. A graphical comparison of the different levels of
daily gain for the rations is given in Figure 2.
The Large-White pigs performed at significantly higher rates than the Malawian pigs. It should
be mentioned that eight out of thirty two local Malawian pigs died in the course of the
experiment, which was attributed to their poor health status due to early weaning.
 Effect of initial weight on overall performance
The initial weights of the experimental animals were divided into three weight ranges which
were 3–7.5 kg (Inwt1), 8–12.5 kg (Inwt2) and 13–19 kg (Inwt3). The overall effect of the initial
weight on the performance of the pigs in general, was such that the heavier pigs gained at a
significantly higher (P <0.05) rate of 0.641 kg/day compared to 0.464 kg/day for the pigs in the
weight range labelled Inwt2, and 0.273 kg/day for the lighter (Inwt1) pigs. The Kleiber ratio
was also larger for the pigs in the heavier weight range (Inwt3) at 0.768 as opposed to 0.371
(Inwt2) and 0.320 (Inwt1). This represented an estimated better feed utilisation efficiency for
the heavier pigs than for their lighter contemporaries. Summarised results of the overall effect
of initial weight are shown in Table 8.
Table 8. Effect of initial weight on the general performance of both Large-white and local
Malawian pigs in the starter period
 Inwt1 Inwt2 Inwt3
 Ism sem Ism sem Ism sem
ADG 0.273a 0.087 0.464b 0.089 0.641c 0.092
Kibl 0.320 a 0.099 0.571 b 0.107 0.768 c 0.111
a,b,c lsmeans having different superscripts in a row differ significantly (P<0.05)
ADG = average daily gain (kg/day), Kib I = Kleiber ratio
sem = standard error mean, Ism = least square mean
lnwt l = initial weight range 3.5–-7.5kg
lnwt 2 = initial weight range 8–-12.5kg
Inwt 3 = initial weight range 13 –19kg
 Effect of starter ration on overall performance
The analysis of performance of the rations in general was also conducted. There was no
significant difference in both the ADG and Kleiber ratio for all the rations administered. A
summary of the results is presented in Table 9.
Table 9. Effect of starter ration on the general performance of both large-white and local
Malawian pigs.
 Ration 1 Ration2 Ration 3 Ration4 r2
 lsm sem lsm sem lsm sem lsm sem  
ADG 0.324 0.034 0.416 0.038 0.395 0.048 0.400 0.037 0.89
Klbl 0.454 0.042 0.565 0.046 0.537 0.059 0.547 0.044 0.83
ADG = average daily gain (kg/day), Klb I = Kleiber ratio
Ism = least square mean, sem = standard error mean
Ration I = maize based ration
Ration 2 = Raw sorghum-based ration
Ration 3 = 50% maize and 50% roasted sorghum-based ration
Ration 4 = Roasted sorghum-based ration
 Overall performance over the 18-week period
The overall performance of the Large-White pigs and local Malawian pigs as indicated by the
average daily gain (ADG), Kleiber ratio (KR) and the time taken to maturity (Tmat) for each
treatment is given in Table 10. While the ADG for the Large-White pigs on Ration 2 was not
significantly different from the pigs on Ration1, it was significantly different (P <0.05) from
those on Ration 3 and Ration 4, respectively. There was no significant difference in ADG for
the local pigs in all the rations fed. The comparisons of growth for the Large-Whites and for
the local pigs on different treatments are given in Figures 4 and 5 respectively.
When a comparison of ADG between the two breeds on similar rations was made, it was
found that there were significant differences (P<0.05) in daily gain with the Large-Whites
performing better than the local pigs. A summary of the results is presented in Table 10.
Table 10: Overall performance of the experimental pigs by treatment as indicated by Average
daily gain (ADG) kg/day, Kleiber ratio (KR) and time taken to reach target weight (Tmat) in
weeks
Treatment Ration 1 Ration 2 Ration 3 Ration 4 r2
 lsm sem lsm sem lsm sem lsm sem  
ADG 0.431ab 0.029 0.459a 0.029 0.364 b 0.031 0.372b 0.031 0.61
 (0.250) (0.030) (0.389) (0.029) (0.316) (0.030) (0.335) (0.032)  
KR 0.361ac 0.015 0.374a 0.015 0.355be 0.015 0.349e 0.015 0.47
 (0.267) (0.016) (0.327) (0.015) (0.283) (0.019) (0.306) (0.017)  
Tmat 15.31 0.564 13.81 0.559 14.71 0.749 15.25 0.632 0.76
 (13.83) (0.874) (14.12) (0.549) (11.39) (0. (13.71) (0.726)  
892)
Figures in parentheses represent the performance of the local Malawian pigs
a,b,e lsmeans within the same row with the same superscript do not differ significantly (P<0.05)
lsm = least square means, sem = standard error means
Ration 1 = maize-based ration
Ration 2 = raw sorghum-based ration
Ration 3 = 50% maize and 50% roasted sorghum-based ration
Ration 4 = roasted-sorghum based ration
 Time taken to reach target weight
The results as shown in Table 10 indicate that there was no significant difference in the
number of weeks taken from the start of the fattening trial to attainment of the targeted weight
for both the Large-Whites (50-55 kg) and the local Malawian pigs (40-45 kg) in all the
treatments.
Table 11. Overall performance of the experimental pigs for ADG, KR and Tmat as determined
by sex
 ADG KR Tmat
 lsm sem lsm sem lsm sem
female 0.386 0.0117 0.343 0.0077 14.74 0.319
male 0.350 0.0113 0.312 0.0079 13.29 0.349
Lsmeans in a column do not differ significantly (P<0.05)
ADG Average daily gain (kg/day)
KR Kleiber ratio
Tmat Time taken to reach target weight (weeks)
 Effect of sex on the overall performance
The effect of sex on the ADG, Kleiber ratio and time taken to reach maturity are presented in
Table 11. There was no significant difference in the growth rate of the female pigs in general
at 0.368kg/day compared to 0.350kg/day for the males.
 Cost of production of the pigs
The cost of production of the pigs was mostly centred on feeding. Tables 12 and 13 indicate
the average cost of feeding both the Large-Whites and the local Malawian pigs with each of
the four rations. Ration 1 was the most expensive followed by Ration 3. The costs of Ration 2
and Ration 4 were the same.
Table12. Feeding costs in the starter period.
 Ration 1 Ration 2 Ration 3 Ration 4
No. of Pigsa 8(7)* 8 (6)* 8 (5)* 8 (7)*
Mean 1.0 1.0 1.0 1.0
consumption per
day (kg) (0.56) (0.57) (0.57) (0.53)
Length of starter 28 (42) 28(42) 28(42) 28(42)
period (days) a
Mean total 224 (164.6) 224(143.6) 224 (113.4) 224(155.8)
Consumption (kg)'
Cost/kg (MK) 9.58 9.25 9.44** 9.25**
Average cost of 2,145.92 2,072.00 2,114.56 2,072.00
starter feed (MK)a 
(1,576.88) (1,328.30) (1,070.50) (1,441.15)
Net cost of starter 3,722.80 3,400.30 3,185.06 3,513.15
rations (MK)
Live-weight gain 0.443 0.525 0.453 0.491
(kg/day)
(0.192) (0.344) (0.126) (0.359)
Feed cost per live- 9.58 9.25 9.44 9.25
weight gain (MK) a
(5.37) (5.27) (5.38) (4.90)
a figures in parentheses are for local Malawian pigs
*note the fewer local pigs on the treatments to explain lower feed cost per live- weight gain
**The cost of Rations 3 and 4 excludes the actual cost of roasting the sorghum, estimated at
MK200.00 for the starter ration
Ration 1 = maize-based ration
Ration 2 = raw sorghum-based ration
Ration 3 = 50% maize and 50% roasted sorghum-based ration
Ration 4 = roasted sorghum-based ration
Table 13: Feeding costs in the grower period
 Ration 1 Ration 2 Ration3 Ration 4
No of pigsa 8 (7)* 8 (6)* 8 (5)* 8 (7)*
Mean 2.5 (2) 2.5 (2) 2.5 (2) 2.5 (2)
consumption per
day (kg) a
Mean length of 70.9(74.9) 66.5 (52.5) 70.9 (56) 69.13 (63)
grower period
(days)a
Mean total 1,418 1,330 1,418 1,383
consumption (kg) (1,049.9) (630) (560) (882)
a
Cost/kg (MK) 8.15 7.84 7.99** 7.84**''
Mean cost of 11,556.70 10,427.20 11,329.82 10,842.72
grower ration (8,556.69) (4,939.20) (4,474.22) (6,914.88)
(MK) a
Net cost of grower 20,113.39 15,366.40 15,804.22 17,757.60
ration (MK)
Live-weight gain 0.461 0.487 0.476 0.404
(kg/day)a (0.297) (0.316) (0.336) (0.287
Feed cost per 20.38 19.61 19.99(1 5.99) 19.61 (16.32)
live-weight (16.32) (15.67) (15.67)
gain(MK)a
a Figures in parentheses are for the local Malawian pigs
* note the fewer local pigs on the treatments to explain the lower feed cost per live-weight gain
**The cost of Rations 3 and 4 excludes the actual cost of roasting the sorghum, estimated at
MK200.00 for the grower ration
Ration 1 = maize-based ration
Ration 2 = raw sorghum-based ration
Ration 3 = 50% maize and 50% roasted sorghum-based ration
Ration 4 = roasted sorghum-based ration
The grand cost of feed for both the starter and grower rations compounded was MK82,862.92
 Discussion
 Tannin analysis
The analysis for tannins of the eight sorghum varieties showed that the red and brown coated
varieties were high in tannin levels. The two methods of analysis consistently showed that
variety Misinde (a predominantly red/brown sorghum variety) was of the highest tannin as
shown in Tables 5 and 6. The high tannin in the red-coated Misinde is consistent with the
findings of Boren and Waniska (1992), who reported that a dark seed coat is indicative of
higher tannin content with the lighter colours indicating lower tannin levels. It has been
reported that tannins give sorghum a resistance to birds, insect and fungal depredation (Harris
and Burns 1973).
The classification of sorghum in the rapid tannin analysis method suggested by Price and
Butler (1977) leaves the variety Msinde in the intermediate level (0.26–0.99% catechin
equivalents), while the rest of the varieties are in the low tannin group with a range of 0. 10–
0.25% catechin equivalents. Banda-Nyirenda and Vohra, (1990) classified some sorghum
varieties as low in tannin content at 0.001% catechin equivalents to high at 1.15% catechin
equivalents, while Ibrahim et al., (1988) had some varieties ranging from the lowest at
0.00%CE to the highest at 2.00%CE.
There are other factors which affect tannin levels in sorghum or indeed any other crops such
as location of growth (McMillian et al. 1972; Guiragossian et al. 1978), stage of growth and the
part of the plant being tested. Although Misinde showed an intermediate tannin level when
grown in the Lower Shire valley, it could probably be of higher tannin content (>1.00% catechin
equivalents), or lower when cultivated in other regions of Malawi.
 Roasting as a method of reducing tannin content
The sorghum variety Misinde that was identified as being higher in tannin content was
subjected to roasting on a charcoal flame, and analysis was done before and after roasting as
shown in Tables 5 and 6. Analysis of tannins employing the Folin- Ciocalteau method of tannin
showed the concentration dropping from 1.193mg/ml to a significantly lower 0.752mg/ml
indicating a 37% depression. A similar observation was made when the semi-quantitative
method of tannin classification (Price and Butler 1977) was employed where the pre-roasted
Msinde at 0.26–0.99%CE (intermediate) was reduced to 0. 10–0.25%CE (low tannin level)
after being roasted.
Roasting reduces the tannin levels in high tannin sorghum varieties, Bangu (1992).
Deshpande et al. (1984) cited by Huisman et al. (1990) reported that tannins can be reduced
upon heat treatment.
 Starter phase performance
There were no significant differences in the average daily gain (ADG) for the Large- White pigs
fed different rations. The data agrees with the findings of Jernigan et al. (1988) who observed
that there were no significant differences in weight gain when pigs were fed rations based on
maize-soyabean, roasted sorghum-soyabean and raw sorghum-soyabean mixtures. While
digestibility was improved due to roasting, the conclusion was that there was no corresponding
improvement in average daily gain (ADG) which meant that roasting as a treatment was
almost unnecessary.
The ADG among the local Malawian pigs showed that there were no significant differences for
pigs fed raw sorghum-based ration (Ration 2) and roasted sorghum-based ration (Ration 4).
However, there were significant differences (P<0.05) in ADG when compared to the pigs fed
Ration 3 and Ration1 which were lower. Differences among treatments were mostly
circumstantial in that most of the Malawian pigs were obtained for the experiment at a time
when they were highly malnourished, small and weak resulting from early weaning. It has
been stated that weaning weight is positively related to post weaning growth (King et al. 1997),
which means that the lower weaning weights of local pigs when compared to exotic pigs
jeopardises their growth soon after weaning.
The mortality figures of eight out of thirty two local pigs for minor incidences of scouring could
be attributed to the assertion above. Simoongwe (1998), suggested that the practice of
weaning the local Malawian pigs below 5kg live-weight was a risk to their survival.
The breed differences in performance during the starter period were significantly different with
the Large-White pigs showing better daily gains compared to the local pigs. The difference in
gain between the two breeds could also have been genetic. The ADG for the Large-Whites (in
kg/day) were 0.431, 0.459, 0.364 and 0.372 for the pigs fed Rationl, Ration 2, Ration 3 and
Ration 4, respectively. The local pigs fed similar rations gained at 0.280, 0.389, 0.316 and
0.335 kg/day. It has been stated that exotic pig breeds have a higher feed conversion
efficiency which results in a better growth rate and bigger litters than the African indigenous
pigs. Haley, d'Agaro and Ellis, (1992) reported a feed conversion efficiency of 2.7 for Large-
White pigs compared to 5.3 for the local Malawian pigs (Mapemba 1996). In another report,
Simoongwe (1998), observed significant differences (P<0.05) when a comparison of the exotic
and the local Malawian pigs fed on a soyabean based ration was made. The Large-Whites
gained at 373g/day compared to 310g/day for the local pigs.
The Kleiber ratio for the Large-White pigs followed a similar pattern to the live- weight gain
where there was no significant difference in performance among all the treatments. The fact
that the results at this stage were not significant might be due to the small number of
experimental animals. Figure 2 illustrates the differences in performance of the local Malawian
pigs fed different rations in the starter period. The pigs fed the raw sorghum-based ration
(Rationl) were growing at a faster rate than the other pigs. The explanation for this would be
that raw sorghum generally has a higher protein level than maize. Therefore, heat treatment
(roasting in this case) may have affected the protein content by denaturation since there was
no strict control of the temperature at which the sorghum grain was roasted, hence the
seemingly lower growth rate exhibited by the pigs fed roasted sorghum-based ration (Ration
4). The better performance of the pigs fed Ration 2 compared to those fed the maize-based
ration (Rationl), 50%maize and 50%roasted sorghum (Ration 3), and Ration 4 is because of
the higher protein level in raw sorghum than in maize, and also the loss in value of the roasted
sorghum grain.
This suggestion agrees with Yu et al. (1996) who reported that heat treatment decreases
proteolysis. The differences in performance among the diets are graphically presented in
Figures 1 and 2 for the Large-White and local pigs, respectively. Another explanation would
be that heat treatment also reduced the levels of lysine in sorghum. Lysine and Methionine
are already known to be the most limiting amino acids in cereals such as sorghum and maize
(McDonald et al. 1994) and therefore, roasting further serves to decrease their availability.
Figure 3. Performance of the Local Malawian pigs in the grower phase.
The results in this growth trial as shown in Tables 7 and 10 indicate a superior performance in
daily gain and the Kleiber ratio for raw sorghum-based rations. This means that the relatively
higher tannin levels in raw red-coated sorghum (Misinde) were not at levels high enough to
cause nutritional problems in the pigs, hence the apparent better performance. The
classification of sorghum according to tannin content using the protocol by Price and Butler
(1977) indicates that for a sorghum cultivar to be classified high in tannin, it must have a
range of Catechin Equivalents (CE) higher than 1%. The variety Misinde fell short of this and
was instead classified as intermediate (0.26–0.99 % CE), although much higher than the other
varieties which were in the low tannin group with a range of 0. 10–0.25% CE.
The initial weight had a significant influence on the performance of both the Large- White and
local Malawian pigs. The heavier pigs in both breeds had a better average weight gain
compared to the lighter pigs as given in Table 8.
 
Figure 4. Growth of the Local Malawian pigs on their different rations
 Overall performance over the 18 week period
 Average daily gain and the Kleiber ratio
The data revealed that over the 18-week growth trial, the average daily gain of the Large-
White pigs fed maize-based (Ration1 and the raw sorghum-based (Ration 2) rations was not
significantly different. The pigs fed Ration l did not differ significantly from those on the 50%
maize, 50% roasted sorghum (Ration 3) and roasted sorghum (Ration 4) rations, while the
pigs on Ration 2 had an average daily gain that was significantly higher (P<0.05) than for the
pigs fed either Ration 4 or Ration 3. The daily gain for the local Malawian pigs was not
significantly different for all the rations, although Figure 4 indicates a slightly better
performance for the local pigs fed raw sorghum-based ration (Ration 2). The differences
appear to have been not significant mostly because of the relatively small number of local pigs
involved in the trial due to higher mortality
However, the earlier explanation concerning the relatively higher growth rate for the local
Malawian pigs fed Ration 2 in the starter period is still valid in the grower period. Raw sorghum
is reported to be higher in protein than even maize. Roasting of sorghum just assisted in
denaturing some of the amino acids which would otherwise have been available for uptake by
the pigs. The picture portrayed in Figures 2 and 4 is proof enough of the assertion that the
performance of the pigs on the Ration 2 was superior to the rest of the treatments.
The Kleiber ratio for the Large-Whites indicates that the pigs fed Ration 2 (0.374) were more
efficient feed converters together with the pigs on Ration l. The pigs on Ration 3 had a Kleiber
ratio of 0.355 which was significantly different (P<0.05) from 0.361 for the pigs on Ration l and
0.349 for those on Ration 4. The observation agrees with the assertion made earlier about the
average daily gains of the pigs on Ration 2. The results of a better feed conversion efficiency
(as estimated by the Kleiber ratio) are evident in the average daily gain. The Kleiber ratio for
the local pigs also followed a similar pattern as for the ADG.
 Time taken to reach maturity
There was no significant difference in the time taken to attain the targeted maturity of 55kg
among the Large-Whites for all treatments. Similarly, there was no significant difference in the
time taken to reach the targeted weight of 45kg among the local Malawian pig breeds fed the
different rations. The Large-Whites attained the target weight at 15.31weeks (Ration l), 13.81
weeks (Ration 2), 14.71 weeks (Ration 3) and 15.25 weeks (Ration 4). These were not
significantly different from 13.83 weeks, 14.12 weeks, 11.39 weeks and 13.71 weeks for the
local pigs fed Ration l, Ration 2, Ration 3 and Ration 4, respectively when breed comparison
was carried out. The result is interesting in that the targeted weights were different at 55kg for
the Large-Whites and 45kg for the local pigs. However, the different growth rates account for
the uniformity observed in the time taken to reach target weight. The inability of the local pigs
to reach maturity earlier, given their lighter targeted weight is testimony of their slow growth
rate. Some of the factors of major importance for commercial meat production are the weight
at which the animal produces a carcass of the desired conformation, required degree of
fatness and the efficiency of feed utilisation necessary to obtain these objectives (Payne
1990). Growth rate is important in this particular aspect because it determines the amount of
time that the animal is kept. Mapemba (1996) suggested that the longer the animal was kept,
the greater the proportion of feed that was utilised for maintenance, rather than for production
and the result was a slow annual turnover rate per unit land or housing space. Commercially
therefore, local Malawian pigs may not be so desirable until such a point that selection will
have been done to come up with strains with a better feed conversion efficiency and improved
growth rate which will also reduce the time it takes to attain mature weight given well balanced
feed.
Apart from breed differences in the time taken to reach mature weight, the uniformity within
breed also indicates that the tannins in the ration based on raw high-tannin sorghum (Ration
2) did not affect the ultimate performance of the pigs. Probably the tannin level in variety
Misinde (a red coated sorghum cultivar) was not high enough (perhaps towards the lower end
of the range 0.29–0.99% CE) to cause growth problems in both the local and Large-White pig
breeds.
The time taken to reach maturity was not significantly different as revealed by the data in
Table 11. The notion that the average daily gain differences by sex were not significant is
further seen in the time taken to reach maturity, although there is a contradiction with
established fact that female pigs tend to grow faster than male castrates, as suggested by
(Whittemore 1987).
The effect of sex on the overall performance of the experimental pigs was such that there was
no significant difference between the females and the male pigs gaining at 0.386kg/day and
0.350kg /day, respectively. It should be noted that the males used in this experiment were
castrated. Female pigs have been reported to have better growth rates than the male
castrates (Whittemore 1987).
 Cost of production of the pigs
Feed was evidently the most demanding input on the budget. The cost of raising an average
pig per day on the starter ration shows that those fed Ration l (maize-based ration) were more
expensive to feed than the rest. The data on feed cost per live- weight gain was not
statistically analysed, but a price comparison did show differences. The feeds based on raw
sorghum and roasted sorghum were the least expensive in both the starter diets at
MK925.35/100kg bag compared to MK958.11 (maize-based ration) and MK944.32 (50%
maize and 50% roasted sorghum-based ration). In the grower period, the raw sorghum and
roasted sorghum based rations were cheapest at MK784.31 compared to MK815.19 and
MK799.44 for the maize-based ration (Ration l) and Ration 2, respectively. The cost of
roasting can be estimated at MK200.00 for each growth phase which raises the cost of ration
4 to about MK1,125.35/100kg bag and Ration 3 to MK144.32/100kg bag in the starter phase.
In this experiment, the cost of charcoal at MK200.00 is merely an estimate to emphasise the
fact that roasting is an added cost to the production of a roasted sorghum-based ration. Since
soyabeans also require roasting, and the sorghum was only roasted after soyabeans to take
advantage of the lower temperatures of the charcoal, the cost related to roasting can almost
be ignored, unless where soyabeans or indeed another ingredient requiring roasting is not part
of the ration.
 References
AOAC (Association of Official Analytical Chemists). 1994. Official Methods of Analysis, 4th
Edition, 1111, North Nineteenth Street, Suit 210, Arlington, Virginia, 22209, USA.
Banda-Nyirenda D.B.G. and Vohra P. 1990. Nutritional improvement of tannin containing
sorghums (Sorghum bicolour) by sodium bicarbonate. Cereal chemistry 67: 533–537.
Bangu, N.T.A. 1992. Traditional technologies in small grain processing: roasting and malting-
The Tanzanian experience, pp57–59. In: Utilisation of sorghum and millets. Gomez M.I.,
House L.R., Rooney L.W and Dendy D.A.V. (eds). Patancheru A.P. 502 324, India:
International Crop Research Institute for the Semi-Arid Tropics (ICRISAT).
Bergh, L. 1992. The Kleiber ratio as a measure of feed conversion efficiency in phase -D tests.
National beef cattle and progeny testing scheme (1980–1992 results) Livestock Improvement
Scheme, Irene, Pretoria, Republic of South Africa.
Boren B and Waniska R.D. 1992. Sorghum seed colour as an indicator of tannin content.
Journal of Applied Poultry Research 1:117–121.
BLP88. 1987. Linear Programming with Bounded Variables, User Manual, Eastern Software
Products.
Goering H.K and Van Soest P.J. 1970. Forage Fibre Analysis. (Apparatus, Reagents,
Procedures and some Applications). Agricultural Handbook No 379, Agricultural Research
Service, United States Department of Agriculture.
Guiragossian V., Chibber B.A.K., Van Scoyoc S., Jambunathan R. Mertz E.T. and Axtell J.D.
1978. Characteristics of proteins from normal, high-lysine and high tannin sorghums. Journal
of Agricultural Food Chemistry 26:219–223.
Haley C.S., d'Agaro E. and Ellis M. 1992. Genetic components of growth and ultrasonic fat
depth traits in Meishan and large-white pigs and their reciprocal crosses. Animal Production
54:105–115.
Harris H.B. and Burns R.E. 1973. Relationships between tannin content of sorghum and pre-
harvest moulding. Agronomy Journal 65: 957–959.
Huisman J., Van der Poel A.F.B., Vestergen M.W.A. and Weerden E.J. 1990. Anti-nutritional
factors in pig nutrition. World Review of Animal Production 25 (2): 78–82.
Ibraham S., Fisher C., EI-Alaily H., Soliman H. and Anular A. 1988. Improvement of the
nutritional quality of Egyptian and Sudanese Sorghum grains by the addition of phosphates.
British Poultry Science 29: 721–728.
ICRISAT (International Crop Research Institute for the Semi-Arid Tropics.). 1977. Technical
Manual 2, Laboratory Procedures for Quality Evaluation of Sorghum and Millet. Patancheru
802–324, Andra Pradesh, India.
Jernigan J.S., Combs W.R., Walker G.E. and White C.E. (1988) Effect of roasting grain
sorghum on the performance and digestibility of nutrients in pigs. Nutrition Reports
International 38: 79–89.
King R.H., Mullan B.P., Dunshea F.R. and Dove H. 1997. The influence of piglet body weight
on milk production of sows. Livestock Production Science 47: 169–174.
Mapemba J.P. 1996. Evaluation of the potential performance of the Malawian indigenous pigs.
MSc thesis, Faculty of Agriculture, University of Malawi.
McDonald P., Edwards R.A. and Greenhalgh J.D.H. 1994. Animal Nutrition. 4th edition.
(Longman).
McMillian W.W., Wiseman B.R., Burns R.E., Harris H.B. and Greene G.L. 1972. Bird
resistance in diverse germplasm of sorghum. Agronomy Journal 64: 821–822.
Payne W.J.A. 1990. An Introduction to Animal Husbandry in the Tropics. Longman, Harlow,
UK. 4th edition. 881 pp.
Price M.L. and Butler L.G. 1977. A critical evaluation of the vanillin reactions as an assay of
tannins in sorghum grain. Journal of Agricultural and Food Chemistry 63(3): 201–207.
SAS (Statistical Analysis Systems). 1989. SAS/STAT User's Guide, SAS Institute Inc., Cary
N.C.
Simoongwe V. 1998. The performance of large-white and local Malawian pigs fed rations
based on soyabeans (Glycine Max), cow-peas (Vigna unguiculata) or pigeon peas (Cajanus
cajan). MSc thesis, Faculty of Agriculture, University of Malawi. Malawi.
Whittemore C.T. 1987. Elements of pig science. Longman Scientific and Technical, Essex,
UK.
Yu. F., Mcnabb W.C., Barry T.N. and Moughan P.J. 1996. Effect of heat treatment upon the
chemical composition of cotton-seed meal and upon the reactivity of cottonseed condensed
tannins. Journal of the Science of Food and Agriculture 72: 263–272.
The effect of a community based animal health service
programme on livestock mortality, offtake and husbandry
applications — A field study in northern Malawi
K. Hüttner 1 ; K.Leidl 2; D.U.Pfeiffer 3; F.B.D.Jere 2 and D.Kasambara 2
1 EpiCentre, Massey University, PB 11-222 Palmerston North, New Zealand 
2 Basic Animal Health Service Project/GTZ, P.O.Box 611 Mzuzu, Malawi
3 The Royal Veterinary College, London, UK
Abstract
Introduction
Materials and methods
Farm-selection
Interviews
Follow-ups
Ecological zones and climate
Calculation of rates
Offtake rates
Mortality rates
Approach to data analysis
Results
Farm locations
Herd size
Cattle movements in and out of herds
Livestock offtake
Mean monthly cattle offtake
Period-specific cattle offtake
Mean monthly offtake for small ruminants
Period-specific offtake rates for small ruminants
Livestock mortality
Calf mortality
Cattle mortality
Mortality in small ruminants
Adult chicken mortality
Livestock husbandry applications
Seasonal pattern of selected husbandry measures
Farmer's perception about the observation period
Discussion and conclusions
Acknowledgement
References
 Abstract
Livestock dynamics, offtake, mortality and husbandry measures of 96 randomly selected users
of a community-based animal health (BAHS) program were compared with 96 matched part-
users and 96 non-users, respectively. More than 5000 farm visits were conducted between
July 1997 and February 1999, of which 3724 visits were used for this evaluation.
Univariate and multivariable analyses were performed to compare farm characteristics and
livestock performance among the three user categories. The results showed that users of the
program owned larger herds of ruminants and more chickens compared to part-users and
non-users. For non-users, mean monthly mortality in calves (3.2 %), sheep and goats (2.8 %)
and chickens (4.4 %) were twice as high compared to user farms (1.4 %, 1.3 % and 2.0 %).
All livestock species monitored in this study showed a decline in numbers during the 14 month
period of evaluation. There were large movements of cattle into and out of kraals. These were
animals that were managed but not owned by respective farmers. Offtake rates varied
considerably between different species and at different periods. The adjusted annual cattle
offtakes between December 1997 and November 1998 on average were 10.5 % (users), 9.5
% (part-users) and 7.8 % (non-users). Users of the BAHS-program applied various livestock
husbandry and management measures more often than either of the other groups.
During concluding interviews in January and February 1999, BAHS-users felt significantly
more positive about the previous year in terms of livestock health and production compared to
part- and non-users. Overall, the results indicated a positive effect of the community-based
animal health service program in northern Malawi on animal health and productivity.
 Introduction
Traditional livestock production systems in developing countries have recently been the focus
of considerable research. (Tuah and Ya Nyamma 1985; Rodgers and Homewood 1986; De
Jode et al. 1992; Uza 1997; Kudi et al. 1998). Some researchers have investigated the effects
of community-based animal health or management interventions on livestock production in
these environments (Meemark 1988; Ngategize 1989; Jemal et al. 1995; Schreuder et al.
1996; Sulistiyo et al. 1998; Jones et al. 1998).
The Basic Animal Health Service (BAHS) Project in northern Malawi was launched in the late
1980's to provide a regional service to livestock owners. Self-help, demand-driven strategies
and consequent cost-recovery are essential components of the program. At present, BAHS
operates within an area of 27,000 km2. The focal point of BAHS is part-privatisation of field
veterinary services through a Drug Revolving Fund and deployment of trained farmers,
appointed by their community as village keymen (KM). These trainees are supplied with a
paramedical kit (drug box), which allows them to treat village livestock and provide general
advice on health and husbandry in their home area. Supervision of KM is done through
veterinary assistants (VA), who are also provided with a drug-box (Leidl et al. 1995). The
BAHS-Project places much of its effort into extension, which includes participatory techniques
and specific messages in form of pictorials. Improvement of existing management techniques
in livestock production has been given priority. BAHS-users currently pay an annual
membership-fee equivalent to about US$ 0.3. They are organised in village livestock-groups.
The organisational set up of the BAHS-Project has been discussed in detail by Jere (2000). All
non-members of BAHS, who call the KM or VA for treatments, are defined as part-users. They
pay a fee equivalent to about US$ 0.2 per prescription in addition to the price of the drugs. The
range of drugs offered through BAHS, and the annual treatment pattern of VA and KM were
described by Hüttner et al. (2000a).
At present, there are about 200 drug-box holders operating in Karonga and Mzuzu
Agricultural Development Divisions (ADD), with the latter being the study area (Figure 1),
involving 42 KM and 84 VA. Demand for BAHS by livestock farmers is strongest for cattle
followed by small ruminants and chickens. The study therefore concentrated on these species.
This article was part of a series of studies that lead to a comprehensive impact assessment
including an economic analysis of the BAHS-Project (Hüttner 2000). The objective of this
article was to compare livestock offtake, mortality and husbandry applications between users,
part- and non-users of the BAHS-program. It may contribute towards a critical review of the
effects of such programmes for small holder farmers in developing countries.
 Materials and methods
 Farm-selection
Eleven KM and 21 VA were randomly selected from the 126 drug-box holders that operated
within Mzuzu ADD as of June 1997. A minimum BAHS-involvement of 12 months was
required for eligibility. At each of those 32 locations, three BAHS-users were selected
randomly from the list of paid members. Another 3 part-users and 3 non-users of BAHS were
selected in each area as matched cases, emphasis being given on similar livestock ownership
to the BAHS-users. The selection was based on the knowledge of the respective KM and VA
at those locations with the final decision being made together with the BAHS-team comprising
of two veterinarians and three veterinary field supervisors.
 Interviews
An initial questionnaire was conducted at all 288 farms in order to describe the social
background, farm characteristics, livestock ownership and self-evaluation of farmers (Hüttner
et al. 2000b). Interviews were conducted between July 1997 and October 1997 by the BAHS-
team. In addition, a concluding interview was conducted at the end of the study period in
January and February 1999. The final herd inventory, current prices for livestock, and
summary remarks of farmers were recorded.
 Follow-ups
Follow-ups commenced after the initial questionnaire in July 1997 and finished in December
1998. Respective KM and VA serving the communities in selected areas conducted them.
Farms were visited at 4-week intervals and all changes in livestock numbers and husbandry
methods were recorded on a prepared farm-monitoring sheet. Animal units were defined as
following: adult cattl ≥12 months, adult chicken ≥  8 weeks, adult sheep and goats≥ 4 moths of
age.
During the monitoring period the BAHS-team had to intervene only on a few occasions. For
instance, there were difficulties during the beginning of this exercise, mainly with KM who
were not used to handling the relatively complex farm sheets. In some locations accessibility
became an issue during the rainy season. We also had to handle expectations by a number of
farmers, whereby some kind of regular bonus was anticipated in return for being involved in
this exercise. A total of 5003 visits were conducted of which 3724 were used for the
evaluation. This reduction became necessary because complete records of all 288 farms were
obtained for the period between November 1997 and December 1998 only.
 Ecological zones and climate
A diverse physical environment comprising three ecological zones characterises the study
area. The lakeshores have warm to high temperatures, are 450–600 m in elevation and have
a mean annual rainfall of 1,500 to 1,750 mm. The Central African Plain is 800 to 1,300 m in
elevation, has mild to warm temperatures and a mean annual rainfall of 1,000 to 1,500 mm.
The highlands are 1,300–2,600 m in elevation, have mild to cool temperatures and a mean
annual rainfall varying from 1,000 to 2,000 mm. Two major seasons determine farming in the
northern region. The dry season lasts from end of April to the end of October, with July/August
making up the cooler part of it. With the onset of the rains during November, temperatures
climb and it remains hot and humid until about February (Wanda 1994).
 Calculation of rates
 Offtake rates
Offtake rates were based on monthly totals of animals, which were sold, slaughtered or used
for ceremonial purposes versus the respective number of animals kept during the same
months. The denominator for 12- and 14-months offtakes was made up by the average herd-
size. Offtake records were only determined for ruminants but not for chickens.
 Mortality rates
Calculation of mortality rates was based on monthly totals of animals that died versus the
respective number of animals kept during the same months
 Approach to data analysis
Data were stored using Microsoft ACCESS 2000 (Microsoft Corporation, Redmond, USA).
GPS receivers were used to determine the precise locations of KM-houses and VA-stations
within the study area. Maps were produced using the geographical information system
software ARCVIEW for Windows version 3.1 (ESRI Inc., Redlands, USA). The effect of user-
status, visit number (month) and VA/KM-status as explanatory variables on mortality in calves,
small ruminants and chickens was examined using a generalised estimated equations (GEE)
model suitable for analysis of longitudinal data. GEE allows the mean of a population
parameter to depend on a linear predictor through a non-linear link function and allows the
response probability distribution to be any member of a range of different distributions (Zeger
and Liang 1986). Mortality as the dependent variable was expressed as a binomial variable
(events/trials). Farm visits, where the population at risk for a particular livestock species was
zero, were excluded from the analysis. The first category of the categorical explanatory
variables was used as the reference category as outlined in Tables 3, 4 and 5. The statistical
analyses were performed using NCSS 2000 (NCSS Statistical Software, Kaysville, Utah, USA)
and PROC GENMOD in SAS for Windows version 8.0 (SAS Institute, Cary, NC, USA).
 Results
 Farm locations
Figure 1 shows the location of farms included in the study.
Figure 1: Location of study farms within Mzuzu ADD (n=288)
 Herd size
The average cattle herd-size of BAHS-users of 16.2 (SD 9.5) was higher compared with part-
(13.8, SD 9.0) and non-users (11.5, SD 8.2) (Kruskal-Wallis test; Chi-square 183.6, df 2, p=
.000). BAHS-users also kept larger herds of sheep and goats (mean 10.2, SD .37) compared
with part- (8.4, SD .57) and non-users (9.6, SD .99) (Kruskal-Wallis test; Chi-square 26.4, df 2,
p= .000). Similarly, BAHS-users maintained larger chicken flocks (22.2, SD 1.9) followed by
part-users (19.1, SD 2.0) and non-users (16.9, SD 1.8) (Kruskal-Wallis test; Chi-square 48.4,
df 2, p= .000). Overall, there was a decline for all livestock species under investigation during
the study period, which is illustrated in Figures 2, 3 and 4. The average cattle population
across farmer-groups decreased by 10 % during the 14 month observation period. This trend
was more evident among non-users (13 %) and part-users (9 %) than BAHS-users (6 %).).
Figure 2: Mean cattle herd-sizes between 11/97 and 12/98 stratified by farmer-status (n=276)
The respective decline in small ruminants was 16 % of the opening stock, 25 % for non-users,
19 % for part-users and 4 % for users (Figure 3).
 Figure
3: Mean herd-sizes of small ruminants between 11/97 and 12/98 stratified by farmer-status
(n=213)
Chicken flock sizes also decreased by 13%, 9 % for non-users, 17 % for part-users and 14 %
for users (Figure 4).
 Figure
4: Mean flock-size of village chickens between 11/97 and 12/98 stratified by farmer-status
(n=288)
 Cattle movements in and out of herds
Large numbers of cattle were "brought in" or "moved away" from cattle kraals. These are
animals, which were kept but not owned by the respective study-farmers. In addition, animals
such as oxen or bulls are commonly hired out and returned later. Our monthly records allowed
distinguishing between cattle that were slaughtered, sold or bartered as true offtakes and
those moved. Figure  5 shows the percentage of cattle moved into or out of the kraal
summarised across farmer-groups, because the temporal patterns were very similar. The
calculation is based on monthly totals of cattle moved, against the annual average herd-sizes.
 Figure
5: Monthly percentage of cattle moved out of and into the kraal between 11'97 and 12'98 for
reasons other than offtakes or ceremonial purposes (n=276)
There were no significant differences (Kruskal-Wallis test; Chi-square 2.2, df 2, p= .33)
between farmer-groups. There were two peaks in May and November 1998 (almost coincident
for both curves).
The average relative percentage of cattle moved out was 67.5 % (SD 23.4) compared to 39.4
% (SD 13.7) of cattle moved into kraals. Users on average moved the largest proportion of
cattle out of their kraals (73.5 %, SD 17.2), followed by non- (68 %, SD 23.4) and part-users
(60.4 %, SD 19.1). The percentage of cattle moved into kraals was 44.1 % (SD 25.3) for
users, 37.2 % (SD 16.5) for part- and 37.0 % (SD 13.7) for non-users. The respective out/in
ratios were 1.7, 1.6 and 1.9. These ratios were used to calculate adjusted cattle offtake rates.
Out/in ratios also differed according to ecological zone. They were highest in the highlands
(2.6), followed by the lakeshore (2.2) and the plains (1.4).
 Livestock offtake
 Mean monthly cattle offtake
The mean monthly offtake rate between November 1997 and December 1998 was 1.74 % for
BAHS-users (SD 0.64), 1.67 % for part-users (SD 1.1) and 1.62 % for non-users (SD 1.0),
which is not significantly different (Kruskal-Wallis test; Chi-square 2.5, df 2, p= .28). Overall
there was a common offtake pattern across all three groups. One peak occurred in January
followed by larger peaks around May and again between September and November. Low
monthly offtake rates occurred before the onset of the rains in October/November and
between July and August.
 Period-specific cattle offtake
Cattle offtake rates for periods of 12 and 14 months per farmer-group and ecological zone
between November 1997 and December 1998 are shown in Table 1. Adjusted offtakes rates
are calculated under the assumption of equal proportions of cattle being "moved in and out" of
kraals for reasons other than offtakes. Average herd-sizes, therefore, have been multiplied
with out/in-ratios as explained earlier.
BAHS-users had the highest 12- and 14-months (adjusted and non-adjusted) offtake rates
followed by part- and non-users throughout all periods. There were fairly large variations
within the groups during different underlying periods. No significant differences were found
between groups. Looking at ecological zones (not adjusted for cattle movements) shows that
Lakeshore-farmers had the highest offtakes and farmers in the plains sold or slaughtered the
lowest number of cattle off their herds, which is significantly different for all periods under
investigation. Applying adjusted rates change this sequence, whereby farmers in the plains
had the highest offtake followed by those at the lakeshores and the highlands, which is
independent of farm-status and remains significantly different.
Table 1: Adjusted and non-adjusted cattle offtake rates for different time-periods according to
farm-status and ecological zone calculated as total offtakes against average herd-size (n=276)
non-adjusted offtake rates adjusted offtake rates
no consideration for in/out ratios including ratios as specified below
 anual offtake 14 months offtake annual offtake 14 months offtake
rates(SD) rates (SD) rates (SD) rates (SD)
status/ 12'97–11'98 11'97–12'98 11'97–12'98 12'97–11'98
zone n=3192 n=3719 n=3719 n=3192
non-users 14.6 17.1 7.69 9.0 (8.9)
(15.1) (16.8) (7.93) ratio = 1.9
part-users 15.2 17.3 9.50 10.8 (11.4)
(16.7) (18.2) (10.47) ratio = 1.6
users 18.0 20.5 10.5 12.1 (11.8)
(18.9) (20.0) (11.13) ratio = 1.7
p-value 0.47 0.37 0.45 0.32
and (1.54, df 2) (1.96. df 2) (1.51, df 2) (2.11. df 2)
χ2
statistics
Highlands 18.2 20.9 6.98 8.0 (7.7)
(17.0) (19.9) (6.54) ratio = 2.6
Lakeshore 20.5 22.9 9.29 9.5 (9.0)
(20.8) (21.5) (9.46) ratio = 2.2
Plains 11.7 13.8 9.51 11.0 (11.5)
(15.6) (16.2) (11.13) ratio = 1.4
p-value .022 .023 .019 .021
and χ2 (7.61, df 2) (7.56, df 2) (7.84, df 2) (7.93, df 2)
statistics
 Mean monthly offtake for small ruminants
Farmers sold or slaughtered sheep and goats more frequently compared to cattle. Users and
part-users followed a fairly similar pattern with peaks around January, June and November,
while non-users showed recurrent peaks more frequently. Non-users on average had a
monthly offtake rate of 5.4 % (SD 3.0) followed by part-users with 4.4 % (SD 3.3) and BAHS-
users with 4.3 % (SD 2.8), which is not significantly different (Kruskal-Wallis test; Chi-square
1.32, df 2, p = .52). Offtakes typically were low during February/March and again during
August/September.
 Period-specific offtake rates for small ruminants
Period-specific offtake rates for sheep and goats between November 1997 and December
1998 are shown in Table 2. Non-users maintained the largest offtake rates throughout the
specified period, followed by either users or part-users. The annual offtake was generally high
amongst study-farms ranging between 31 % and almost 45 % of the annual average herd-
size. No significant differences were found between groups. There were no indications of large
movements of sheep and goats due to reasons other than offtakes unlike the data for cattle.
Thus, no adjusted offtakes were calculated.
Table 2: Offtake rates of sheep and goats for different time-periods according to farm-status
and ecological zone calculated as total offtakes against average herd-size (n=213)
Annual offtake rates(SD) 14 months offtake rates
Status/zone 12'97–11'98 11'97–12'98
Non-users 44.3(38.9) 52.2(44.5)
Part-users 37.1(37.3) 40.6(37.0)
Users 36.4(33.0) 42.9(38.6)
p-value (χ2 statistics) 0.47(1.50).df 2) 0.39(1.86.df 2)
Highlands 36.6(36.3) 42.9(40.4)
Lakeshores 57.4(47.8) 58.7(48.8)
Plains 37.6(33.7) 44.0(38.3)
p-value(χ2 and statistics) 0.10(4.48.df 2) 0.33(2.22.df 2)
   
Comparing offtake rates according to ecological zones shows that farmers at the lakeshores
had the highest rates followed by farmers in the plains and in the highlands. There were no
significant differences between zones.
 Livestock mortality
The monthly pattern of mortality in calves, adult cattle, small ruminants (kids plus adults
combined) and adult chickens is shown in Figure 6.The lowest monthly mortality of all species
occurred towards the end of the dry season around September/October 1998.
 Figure
6: Mean monthly mortality rates for calves, adult cattle, sheep and goats (including kids) and
adult chickens between 11/97–12/98
 Calf mortality
Non-users on average had a mean monthly calf (< 12 months of age) mortality rate of 3.2 %
(SD 1.9) followed by part-users (1.5 %, SD 1.1) and BAHS-users (1.4 %, SD 0.9), which is not
significantly different in the univariate analysis (Kruskal-Wallis test; Chi-square 2.31, df 2, p=
.31). Fifty percent of BAHS-users (47/94) did not experience any calf mortality. Similarly, 45 %
(41/91) of part-users and 43 % (39/91) of non-users did not lose any calf. The results of
multivariable analysis of calf mortality using a GEE model are presented in Table 3.
Table 3: Final generalised estimating equation model looking at the effects of user-status, visit
number and KM/VA-status on calf mortality between Nov. 1997 and Dec. 1998 (n=264)
Standard
Parameter Level Estimate error OR Z P
Intercept – –5.3737 – – – –
Visit 1(Nov. 97) – – 1 – –
 2 (Dec. 97 0.8457 .6238 2.330 1.36 .111
 3 (Jan. 98) 1.3578 .6110 3.888 2.22 .001
 4 (Feb. 98) 1.7909 .5779 5.995 3.10 .000
 5 (Mar. 98) 1.6391 .6001 5.151 2.73 .001
 6 (Apr. 98) 1.6868 .6037 5.413 2.80 .000
 7 (May 98) 1.8398 .5791 6.295 3.18 .000
 8 (June 98) 1.7393 .5969 5.693 2.91 .000
 9 (Jul 98) 1.3171 .5728 3.733 2.30 .010
 10 (Aug.98) 1.0530 .6086 2.866 1.73 .045
 11 (Sep.98) 1.3689 .5887 3.931 2.33 .007
 12 (Oct. 98) 1.0875 .6637 2.967 1.64 .037
 13 (Nov.98) 0.9728 .6637 2.645 1.47 .067
 14 (Dec.98) 1.0851 .5777 2.960 1.88 .040
Userstatus user – – 1 – –
 Part-user 0.2651 .2095 1.304 1.27 .1001
 Non-user 0.3802 .2121 1.463 1.79 .0218
Visit number (month) had a significant effect on calf mortality. Calves being monitored at the
first (November 1997) compared to the following visits had higher odds of death for all but the
second (December), and the 13th (November 1998) visit. The respective odds ratios varied
between 1.6 (October 1998) and 3.2 (May 1998). In addition, user status was significantly
associated with calf mortality. Non-users had a 1.8 higher odds of experiencing calf mortality
compared to users of BAHS.
 Cattle mortality
Part-users on average had a mean monthly adult cattle mortality rate of 0.44 % (SD 0.77)
followed by non-users (0.39 %, SD 0.53) and BAHS-users (0.30 %, SD 0.16), which is not
significantly different (Kruskal-Wallis test; Chi-square 2.4, df 2, p= .91). Relatively higher adult
cattle fatalities occurred between February and June (calving season commences in May).
 Mortality in small ruminants
Mortality in small ruminants was assessed for kids and adults combined. Animals were not
tagged, which made it difficult to accurately distinguish between young and mature animals
especially at 4–7 months of age. Non-users had the highest mean monthly mortality rate in
sheep and goats of 2.8 % (SD 2.2) followed by part-users with 2.1 % (SD 1.8) and BAHS-
users with 1.3 % (SD 0.8), which is not significantly different (Kruskal-Wallis test; Chi-square
3.0, df 2, p= .22). A peak occurred during the rains between November and May and there
was a general decline towards the end of the dry season in October. The final GEE model for
mortality in sheep and goats is presented in Table 4.
Table 4: Final generalised estimating equation model looking at the effects of user-status, visit
number and KM/VA-status on small ruminant mortality (kids + adults combined) between Nov.
1997 and Dec. 1998 (n=212)
Parameter level estimate standard error OR Z P
Intercept – –4.4489     
Visit 1 (Nov.'97) - - 1 - -
 2 (Dec.'97) –0.8676 .4247 0.420 –2.04 .011
 3 (Jan.'98) -0.2052 .4368 0.814 –0.47 .482
 4 (Feb.'98) 0.1159 .4180 1.123 0.28 .601
 5 (Mar.'98) 0.1168 .5096 1.124 0.23 .584
 6 (Apr.'98) 0.6096 .4460 1.840 1.37 .007
 7 (May'98) 0.8441 .4561 2.326 1.85 .000
 8 (Jun.'98) 0.6071 .4079 1.835 1.49 .007
 9 (Jul.'98) 0.5760 .4561 1.779 1.26 .013
 10 (Aug.'98) –0.4500 .4833 0.638 -0.93 .141
 11 (Sep.'98) –0.1546 .4821 0.857 -0.32 .615
 12 (Oct.'98) –0.4588 .4941 0.632 -0.93 .155
 13 (Nov.'98) –0.3817 .4977 0.683 -0.77 .235
 14 (Dec.'98) –0.4734 .5254 0.623 -0.90 .129
Zone Plains – – 1 – –
 Highlands 0.4771 .2115 1.611 2.26 .000
 Lakeshore 0.2228 .4547 1.250 0.49 .369
The animals had a significantly higher probability of death between the sixth and ninth visits
(cool dry) compared to the first visit in November 1997. The respective odds ratios varied
between 1.4 and 1.9 whereas the odds of death were about half during the time of the second
visit (December 1997) compared to the first visit. Ecological zones also had a significant effect
on small ruminant mortality. The odds of death for animals kept in the highlands were 2.3
compared to those kept in the plains.
 Adult chicken mortality
Non -users showed the highest monthly adult chicken mortality rate with 4.4 % and the largest
variation (SD 7.7), followed by BAHS-users with 2.6 % (SD 2.2) and part-users with 2.0 % (SD
1.8), which is not significantly different (Kruskal-Wallis test; Chi-square 2.0, df 2, p= .98). The
highest mortality occurred between September and December 1998. The final GEE model in
table 5 shows a significant association between adult chicken mortality and visit number.
Table 5: Final generalised estimating equation model looking at the effects of user-status, visit
number and KM/VA-status on adult chicken mortality between Nov. 1997 and Dec. 1998
(n=283)
Parameter level estimate standard error OR Z P
Intercept – –4.2642 – – – –
Visit 1 (Nov.'97) – – 1 – –
 2 (Dec.'97) –0.6523 .5048 0.521 –1.29 .019
 3 (Jan.'98) 0.0943 .4510 1.099 0.21 .636
 4 (Feb.'98) 0.3121 .4501 1.366 0.69 .116
 5 (Mar.'98) 0.0829 .4768 1.086 0.17 .670
 6 (Apr.'98) 0.1488 .4285 1.160 0.35 .474
 7 (May'98) –0.5998 .4693 0.549 –1.28 .030
 8 (Jun.'98) 0.0849 .4799 1.089 0.18 .654
 9 (Jul.'98) 0.4381 .4506 1.550 0.97 .028
 10 (Aug.'98) –0.6998 .5487 0.497 –1.28 .008
 11 (Sep.'98) –1.2260 .5359 0.293 –2.29 .000
 12 (Oct.'98) 0.3001 .4767 1.350 0.63 .139
 13 (Nov.'98) –0.1829 .4505 0.833 –0.41 .451
 14 (Dec.'98) 0.7459 .5162 2.108 1.45 .000
The odds of death for these animals was lower at the second, seventh, tenth and eleventh visit
compared to the first visit and varied between 0.29 and 0.55. A higher odds of death was 2.1
for the last- and 1.6 for the ninth compared to the first visit.
 Livestock husbandry applications
Table 6 shows the results of chi-squared cross-tabulation of the frequencies of selected
husbandry measures, which were applied by the farmers.
Table 6. Cross-tabulation of frequency of use of selected husbandry measures between 11'97
and 12'98 stratified by farm status (n=276)
Percent yes
Measure non-users part-users BAHS-users p-value
n=92 n=90 n=94 (χ 2 statistics)
Cattle kraal repaired? 42.7 52.1 58.3 .09 χ2 4.75, df 2)
New cattle kraal built? 27.1 24.0 26.0 .88 (χ2 .25, df 2)
Calf house repaired? 5.2 6.3 17.7 .005(χ2 10.5, df 2)
New calf shed constructed? 14.6 15.6 22.9 .26 (χ2 2.72, df 2)
Chicken house repaired? 24.0 20.8 25.0 .77 (χ2 .50, df 2)
New chicken house built? 13.5 16.7 21.9 .03  (χ2 2.37, df 2)
Sheep/goats shed repaired? 27.1 27.1 35.4 .34(χ2 2.12, df 2)
New sheep/goat shed built? 13.5 20.8 21.9 .27(χ2 2.59, df 2)
Cattle manure applied? 86.5 84.4 87.5 .82(χ2 .41, df 2)
Chickens manure applied? 67.7 64.6 64.6 .87 (χ2 6.28, df 2)
Sheep/goat manure applied? 56.3 62.5 58.3 .67 (χ2 .80, df 2)
Drinkers for chickens in 75.0 76.0 83.3 .31 (χ2 2.32, df 2)
use?
Nests provided for hens? 78.1 79.2 79.2 .98 (χ2 .04, df 2)
Most of the husbandry measures are not significantly different between farmer-groups but their
frequencies clearly indicate a higher adoption of improved husbandry by BAHS-users for
almost all selected measures.
 Seasonal pattern of selected husbandry measures
Figure 7 presents the total number of monthly cattle-kraal repairs. The traditional cattle kraal is
a basic stake enclosure that makes regular maintenance of the wooden structures necessary.
 Figure
7: Temporal pattern of cattle-kraal-repairs as monthly totals between 11'97 and 12'98 stratified
by farm-status (n=276)
Most repair work happened around the onset and during the wet season with a peak during
November for all farmer-groups.
 Figure
8: Temporal pattern of nest provision for brooding hens as monthly totals between 11'97 and
12'98 stratified by farm-status (n=288)
Figure 8 shows the frequency of nest-provision for brooding hens in villages. The proportion of
farmers providing nests was comparably high. Forty-four non-users (SD 6.1), 46 part-users
(SD 7.2) and 49 BAHS-users (SD 5.8) on average applied this measure. It means that about
half of all selected farmers monitored their brooding hens regularly. All curves show a similar
shape over time with a decline during January and the highest levels between May and
September 1998, except for non-users, who had the highest level during the last visit.
Figure 9 shows the monthly number of farms providing drinkers to their flocks stratified by
farm-status. There were on average 37 non-users (SD 5.2), 41 part-users (SD 6.1) and 50
BAHS-users (SD 5.5) ensuring that chicken flocks had access to proper drinkers during
respective months.
 Figure
9: Temporal distribution of total number of drinker-provision for chicken-flocks as monthly
totals between 11'97 and 12'98 stratified by farm-status (n=288)
The graph shows that during the peak of the rains (December to February) drinkers were less
frequently used, which applied to all farmer-groups.
 Farmer's perception about the observation period
Table 7: Farmers judgement about their livestock production during the study period in
comparison with previous years (n=288), (χ2 167.6, df 6, p = .000)
Status bad normal better
non-users 42 26 28
part-users 38 27 31
BAHS-user 31 17 48
As part of the concluding visits in January/February 1990, farmers were asked about their
perception towards the past 14 months in terms of livestock health and production. BAHS-
users had a significantly more positive view compared to part-and non-users Table 7.
 Discussion and conclusions
All livestock species monitored during this study showed an overall decline in total numbers.
However, chicken flock-sizes as a rule show large variations within the traditional farming
system in Malawi (Ahlers 1999) and developing countries in general (Ajuyah 1999; Kitalyi
1999; Mallia 1999). For this reason, the current level of decline is not likely to lead to major
concerns about the chicken population in the study area. Nonetheless, losses in village
chickens are largely reducible by applying better management practices and also through
improved uptake of vaccine against Newcastle Disease (Ahlers et al. 1999).
The issue of a shrinking ruminant-population particularly in cattle, however, has a different
background compared to chickens. Cattle herds are traditionally held very stable. Moreover,
the natural conditions in the study area, including pasture availability still allow an extension of
cattle herds, which is different in the central and more so in the southern region of Malawi. In
addition, the exceptional population growth in Malawi has led to a strong demand for livestock
and livestock products, which is currently only met by rising imports from neighbouring
countries (The World Bank Group 1999). A Danish appraisal mission noted a decline in
national cattle numbers, which commenced during the late eighties (Malawi Report 1999). This
decline has apparently accelerated since 1992. Possible reasons cited in the same report
included the repatriation of almost 1 million refugees (including their livestock) to Mozambique,
and also an alarming rate of cattle thefts during the nineties. Whole herds were (and still are)
stolen and moved across borders. The latter, however, was not an issue on our study farms.
The Danish report mentioned stagnant productivity under traditional livestock management as
another contributing factor for a decline in the national livestock population.
Our results show, that annual cattle offtakes (adjusted and non-adjusted) are comparably high
for all ususer-groups (Table 1). Ngategize (1989) estimated the annual offtake of smallholder
cattle farmers in northern Tanzania at 5 %. Perry et al. (1984) studied traditionally managed
cattle herds in Zambia and found a mean offtake of 10 %. Rodgers and Homewood (1986)
investigated cattle herds in north-western areas in Tanzania and reported a mean annual
offtake of 8 %. The authors, however, provided no indication on whether movements of cattle
in and out of the pen for reasons other than offtake were of relevance despite the fact that the
traditional cattle farming system is very similar within the region. During the course of on-farm
visits a farmer's response to a missing cattle beast such as "has been given away" had to be
carefully verified in order to be able to distinguish whether this very animal was consumed,
hired out or was simply returned to its owner. It is likely that the issue of cattle movements in
and out of kraals for reasons other than offtakes explains the decline in cattle numbers on one
hand and the seemingly high offtake rates on the other. A possible explanation of much larger
removals than returns of cattle is likely to be the strong demand for livestock and livestock
products and/or higher requirements for cash by people that own these animals. The largest
relative percentage of animals is moved during May and around November. During May,
maize as the main staple food is harvested and dried. Grazing of animals on communal land is
less restrictive in terms of potential crop damage. We know, for instance, that people at the
lakeshores return large numbers of cattle from uninhabited mountainous bush areas during
May because of the latter. During this time loans are also commonly repaid in either cash or
alternatives such as agricultural products. April/May also marks the beginning of the main
calving season. During November, which is the planting season, demand for cash and loans is
very strong, mainly because of the need for fertiliser and seeds. Controlled grazing becomes
essential during this time. These are factors, which possibly had an impact on cattle
movements and offtakes. They might have led to a bias towards the actual cattle herd-size of
the kraal-owners during the observation period. It would be important for future BAHS-
expansion to assess whether kraal-owners direct the same degree of attention to animals
other than their own. It also remains to be investigated why there are such large differences
between adjusted and non-adjusted offtake rates according to ecological zone.
Our results suggest that BAHS-users not only had the highest cattle offtakes but also
managed more stable cattle herd-dynamics as compared to either of the other groups. Lower
mortality in young and adult stock by BAHS-users is highly likely to have contributed to that.
Data retrieved earlier from the initial questionnaire defined BAHS-users as farmers that were
already better educated and better motivated and that owned significantly more livestock
compared to part- and non-users (Hüttner et al. 2000b). Ownership of more livestock,
therefore, is certainly a driving force for a strong interest in the BAHS-program.
The end of the dry season in October/November corresponds with low livestock mortality
overall. This is because tick born diseases but also helminthias in livestock as the main health
constraints in the region are of greater importance during the rains because of the biological
cycle of the parasites. Therefore, visit number, equivalent to months and thus an expression
of season is significantly associated with mortality for all species under investigation. Calf
mortality is highest between March and May. The calving season commences in April/May and
earlier findings (Wanda 1994) showed that 30 % of all new-born calves died within the first
four weeks of life, confirming the seasonal pattern of death in young calves (Norman et al.
1997). For small ruminants, mortality during the cooler months between April and June are
significantly higher compared to November, which is irrespective of the BAHS-user status.
Helminthiasis, particularly infections caused by Haemonchus spp., and respiratory diseases
are also management constraints and are major contributors to death in sheep and goats in
Malawi (Edelsten 1992). In addition, ecological zone was significantly associated with small
ruminant mortality. Sheep and goats kept in the highlands have higher odds of death than
those kept in the plains or at the lakeshores. It is possible that religion attributes to the degree
of attention to these animals, with predominantly Islamic influence at the lakeshores.
Our mean monthly mortality rates in adult chickens between 2 % and 4.4 % are well below of
what Ahlers (1999) found on study farms within Mzuzu ADD. Her mean monthly mortality rate
came to 9.2 % in hens and 9.9 % in cocks. It is likely that the different study design with the
explicit focus on village chicken flocks involving daily visits has led to more realistic results.
This was one of the reasons why we did not attempt to evaluate chick mortality (because this
figure changes daily). The seasonal pattern of adult chicken mortality as shown in Gigure 6
matches those identified by Ahlers (1999).
Results of the initial questionnaire already provided some indication as to the disparities of
husbandry measures applied by study-farmers (Hüttner et al. 2000b). These findings are
complemented by the outcome of the current longitudinal study results.
BAHS-users apply better husbandry and management practices more frequently than either of
the other two groups (Table 6). Repair of calf-pens or complete overhaul of chicken-houses
occurred significantly more frequently amongst BAHS-users than in either of the other groups.
Farmers of all groups, though in varying proportions, clearly followed a seasonal calendar
while being engaged in all sorts of management activities.
Addressing the issue of improved husbandry, however, may not just be based on intensified
extension efforts and education. In the context of the AIDS epidemic in Sub-Saharan Africa,
lack of labour becomes increasingly important in these production systems (Anon. 1999). In
addition, natural building materials are already difficult to obtain in certain parts within the
Project-area, because forests and natural bush are rapidly disappearing due to deforestation
and land degradation. This all has to be considered when further planning is undertaken.
Despite these constraints, the results of our study show the benefits, which can be obtained by
utilising BAHS. The findings complement results of a benefit/cost analysis that was conducted
by Hüttner (2000). Losses in young and adult stock can be reduced and increased cattle
offtake in conjunction with more stable cattle herds appears likely as a result of BAHS usage.
This is supported by the assessments made by study-farmers at the end of the observation
period, whereby significantly more BAHS-users were satisfied with health and productivity of
their livestock compared to part- and non-users. However, our study farms represent
households with larger livestock numbers and thus better income compared with the general
population. Involvement of the majority of livestock holders in the project area is one of the
goals to be achieved by the BAHS-Project. It is therefore crucial to address the special needs
of the majority of poorer families that own chickens, ducks or rabbits only.
 Acknowledgement
This longitudinal study was a major effort for all parties involved. We wish to thank in particular
all keymen and veterinary assistants as well as their supervisors for their commitment. We
also greatly appreciate the patience and support our farmers have shown during the course of
follow-up visits.
 References
Ahlers C. 1999. Characteristics and constraints of the traditional poultry keeping system in
northern Malawi. Unpublished Doctorate thesis, Free University Berlin, Germany. 175 pp.
Ahlers C., Hüttner K., Pfeiffer D.U. 1999. Comparison between a live and an inactivated
vaccine against Newcastle Disease in village chickens. A field study in northern Malawi.
Tropical Animal Health & Production 31, 3, pp. 167–174.
Ajuyah A. O. 1999. Rural family poultry production in the South Pacific Region. The First
INFPD/FAO Electronic Conference on Family Poultry
Http://www.Fao.Org/Waicent/Faoinfo/Agricult/Aga/Agap/Lps/Fampo/Leadpap4.Htm
Anon. 1999. Aids in Africa, http://www.unaids.org/highband/fact/saepap98.html Globale Aids-
Situation: http://www.rki.de/INFEKT/AIDS_STD/GLOBAL/GLB_SEL.HTM
De Jode A., Reynolds L. and Matthewman R. W. 1992. Cattle production systems in the
derived savannah and southern guinea savannah regions of Oyo State, Southern Nigeria.
Tropical Animal Health & Production 24, 90–96
Edelsten, R.M., 1992. Health and husbandry of sheep and goats in Malawi. Proceedings of a
Workshop on Small Ruminants Activities in Malawi. 5th– 8th October 1992, Capital Hotel,
Lilongwe, Malawi, pp. 149–155
Hüttner K., Leidl K., Jere F.B.D. and Pfeiffer D.U., 2000a. Characteristics and Performance of
Village Animal Health Workers and Veterinary Assistants in northern Malawi. Journal of the
South African Veterinary Association. published in 9/2000.
Hüttner K., Leidl K., Pfeiffer D.U., Jere F.B.D. and Kasambara, D. 2000b. Farm and personal
characteristics of the clientele of a community-based Animal Health Service Program in
northern Malawi. Tropical Animal Health & Production, to be published in 3/2001.
Hüttner K. 2000. Impact Assessment of a community based animal health service program in
northern Malawi. MSc thesis, Massey University, Palmerston North, New Zealand, 117 pp.
http://epicentre.massey.ac.nz/documents/Thesis/Klim_Thesis.PDF
Jemal A., Justic D. and Hugh-Jones M.E. 1995. The estimated long-term impact of Tsetse
control on the size of the population of cattle in the Didessa Valley, Western Ethiopia.
Veterinary Research Communication 19: 479–485
Jere F.B.D. 2000. Set up and implementation of FIAH's drug revolving fund. In: Hüttner K.,
Leidl K. and Jere F.B.D. (Eds.). Primary Animal Health Activities - Topic Day. Proceedings of
an International Seminar, 15th–17th Mzuzu, Malawi. Basic Animal Health Project, Mzuzu,
Malawi. p 15–17. http://epicentre.massey.ac.nz/Documents/PDF/KLIM_ProceedingsII.pdf
Jones B.A., Deemer B., Leyland T.J., Mogga W. and Sterm E. 1998. Community-based
animal health services in Southern Sudan; The experience and future. In: Proceedings of the
9th International Conference of the Association of Institutions of Tropical Veterinary Medicine
(AITVM), 14–18 September 1998, Harare, Zimbabwe (in press)
Kitalyi A.J. 1999. Family poultry management systems in Africa. The First INFPD/FAO
Electronic Conference on Family Poultry, FAO, Rome, Italy.
Http://www.Fao.Org/Waicent/Faoinfo/Agricult/Aga/Agap/Lps/Fampo/Leadpap3.Htm
Kudi A.C., Umoh J.U., Eduvie L.O. and Gefu J. 1998. Relative survival of calves in 16
traditionally managed herds in Bauchi, Nigeria. Preventive Veterinary Medicine 36: 307–312.
Leidl K., Jere F.B.D., Wanda G., Hüttner K. and Stange M. 1995. Establishment of a basic
animal health service program in northern Malawi - dimensions of activities through field work
prior to implementation. Proceeding of the 8th International Conference of Institutions of
Tropical Veterinary Medicine, 25–29 September 1995, page 113
Malawi Report. 1999. Malawi National Livestock Development Masterplan. Final Report,
Lilongwe, Malawi, 114 pp
Meemark N. 1988. The development and evaluation of a village-based parasite control
program for swamp buffalo and cattle in northeast Thailand. Unpublished MSc Thesis, Massey
University. Palmerston North, New Zealand. 169 pp
Mallia J.G. 1999. Observations on family poultry units in parts of Central America and
sustainable development opportunities. The First INFPD/FAO Electronic Conference on Family
Poultry. Http://www.Fao.Org/Waicent/Faoinfo/Agricult/Aga/Agap/Lps/Fampo/Freeco13.Htm
Ngategize P.K. 1989. Economic evaluation of improved management for Zebu cattle in
northern Tanzania. Agricultural Systems 31 : 305–314
Norman N., Claes L., Soldan N., Banda M.. 1997. Observations on undipped Malawi Zebu
cattle. In: Zimmermann, W., Pfeiffer, D.U., Zessin, K.H. (Eds.): Primary Animal Health Care
Activities in Southern Africa, Proceedings of an International Seminar held in Mzuzu, Malawi,
February 26 to March 8 1996, pp. 90–107
Perry B. D., Mwanaumo B.; Schels H. F., Eicher E. and Zaman M. R., 1984. A study of health
and productivity of traditionally managed cattle in Zambia. Preventive Veterinary Medicine 2:
633–653.
Rodgers W. A. and Homewood K. M. 1986. Cattle dynamics in a pastoral community in
Ngorongoro, Tanzania, during the 1982–83 drought. Agricultural Systems 22: 33–51.
Schreuder B.E.C., Noorman N., Halimi M. and Wassink G. 1996. Livestock mortality in
Afghanistan in districts with and without a veterinary program. Tropical Animal Health &
Production 28:129–136.
Sulistiyo U., Wahyuni D. and Leksmono C.S. 1998. Village Animal Health Workers in
Minahasa District North Sulawesi, Indonesia. In: Proceedings of the 9th International
Conference of the Association of Institutions of Tropical Veterinary Medicine (AITVM), 14–18
September 1998, Harare, Zimbabwe (in press).
The World Bank Group, 1999. Country at-a-glance, website,
http://www.worldbank.org/data/countrydata/aag/mwi_aag.pdf
Tuah K. and Yaa Nyamaa D. 1985. Preliminary studies on the performance and productivity
indices of N'dama and West African Shorthorn cattle in Ghana. Tropical Animal Health &
Production, 17: 114-120.
Uza D.V. 1997. The productivity of Muturu cattle (Bos brachyceros) under ranching conditions
in the Southern Guinea Savanna of Benu State, Nigeria. Outlook on Agriculture 26:19–23
Wanda G.W. 1994. Health constraints of calves and young stock in Mzuzu Agricultural
Development Division. MSc thesis, University of Reading, Reading, UK.102 pp.
Zeger S.L. and Liang K.Y. 1986. Longitudinal data analysis for discrete and continuous
outcomes. Biometrics 42: 121–30

Session 2: Policy analysis, socio-economics and sustainable
integrated animal health
Socio-economic and biological issues to consider in
smallholder poultry development and research in Southern
Africa in the new millennium
V. Muchenje1*, M. M. Manzini1, S. Sibanda2 and S. M. Makuza2
Correspondence to: 1* Zimbabwe Open University, Bulawayo Region, P. O. Box 3550
Bulawayo, Zimbabwe, Tel: 263-9-68663 / Fax 263-9-70790, e-mail: vmuchenje@hotmail.com
2 University of Zimbabwe, Department of Animal Science, P. O. Box MP 167, Mt. Pleasant,
Harare, Zimbabwe
Abstract
Introduction
Importance and uses of poultry
Poultry productivity in the smallholder farming sector
Non-adoption of poultry production technology by farmers
Possible changes to improve productivity of indigenous flocks
Feed resources and requirements
Breeding in extensive production systems
Markets
Other issues to consider in smallholder poultry research and development
Conservation of genetic resources in poultry
Computer simulation modelling in farming systems and scientific research
Conceptual structure
Model building and validation
The future
References
 Abstract
Poultry provide major opportunities for increased protein production and incomes for
smallholder farmers because of the following factors: small generation interval, the high rate of
productivity, the ease with which poultry products can be supplied to different areas, the ease
with which poultry products can be sold due to their relatively low economic values, minimal
association of poultry with religious taboos, and the complementary role poultry play in relation
to other crop-livestock activities. Despite the potential of poultry production to attain rapid
growth, constraints such as diseases, pests, predators, poor housing, poor nutrition, lack of
knowledge about poultry production, lack of genetic improvement in commercially important
traits in indigenous breeds, lack of capital and poor accessibility to markets can negatively
affect poultry production in the smallholder farming sector. To boost income generation and
nutrition for the smallholder farmers and their families these constraints should be addressed
by carrying out on-farm research involving farmers, especially women who normally carry out
the day to day activities of the poultry sub-system. Since smallholder farming systems are bio-
economically complex involving several kinds of resources and input/output flows, it would not
be advisable to study poultry production only without considering other crop-livestock
components of the farming system. Farming systems research (FSR) has emphasised
traditional cropping systems, with little attention to livestock sub-systems, let alone poultry.
The reason could be that the economic importance of poultry is not adequately appreciated by
researchers and decision-makers because poultry products in the smallholder farming sector
only passes through non-formal marketing channels. For the smallholder farmers to adopt new
technologies meant to improve poultry production, the technologies should be compatible with
local socio-cultural, political, economic and institutional conditions. This can only be done by
carrying out more holistic FSR which go beyond biological production in crop and livestock
sub-systems to include utilisation, marketing outputs and off-farm activities. Research
methods should be more participatory than extractive.
This paper therefore highlights smallholder poultry research, development and production
potential, constraints and challenges for the Southern African poultry researchers in the new
millennium. The term 'smallholder farmers' refers to farmers in communal, resettlement and
small-scale commercial areas.
 Introduction
Although poultry may be of low economic value in smallholder farming systems, it is an
important source of protein in both rural and urban areas. Due to recurrent droughts in
Southern Africa and decreasing grazing land, which tend to result in the reduction of the
national cattle herds, poultry has the potential to become a major source of protein and a
ready source of income in the smallholder farming sector. In Zimbabwe demand for day-old
chicks increased from 10 million in 1988 to 24 million, while that of eggs increased from 11.5
million to 20 million in 1995 (Government of Zimbabwe 1995). While the 1963 to 1975 period
saw an annual growth rate of livestock production in developing countries of 2.4 %, the annual
growth rate for poultry production was much higher at 6.5 % (Hrabovszky 1981).
In most developing countries poultry production systems vary from being extensive to
intensive. In Zimbabwe there are the large-scale poultry production and smallholder poultry
production systems. Large scale poultry production units are characterised by huge capital
investments, intensive management, mechanisation and specialisation. The smallholder
poultry sector can be intensive characterised by exotic breeds of broilers and layers, semi-
intensive or extensive dominated by indigenous poultry breeds that are not specifically
classified. The indigenous poultry breeds in Zimbabwe are estimated to be between 15 and 30
million (Lambrou 1993). Although they are characterised by low production, indigenous birds
are tolerant to diseases and are adapted to poor nutrition. The free-range poultry production
system can best be described as a low input-low output system where the birds are rarely
given any extra feed besides what they get from the surroundings. This system represents a
part of a balanced farming system where little or no inputs are involved. The indigenous birds
are also an important reservoir of genomes that may be used in the future to improve hybrid
birds.
Of the 36 million day-old chicks produced annually in Zimbabwe, 50 % are purchased by the
smallholder producers (Government of Zimbabwe 1995), but little information is available on
what happens in this sector. This can be attributed to lack of support and development of rural
poultry production by the government and development agencies. Some of the constraints to
poultry production are diseases, pests, predators, poor housing, poor nutrition, and lack of
genetic improvement in commercially important traits in indigenous breeds; lack of capital and
poor accessibility to markets (Faranisi 1995; Tadelle 1996; Government of Zimbabwe 1995;
Muchenje and Sibanda 1997). Improved management of these birds through, for example,
supplementing indigenous birds with energy, protein and calcium can increase poultry
productivity in extensive systems (Tadelle 1996).
Despite the constraints listed above, there are several factors that make expansion of poultry
production in the smallholder farming sector attractive. Poultry meat and egg production
enterprises grow very rapidly, not only because of the biological, physical and economic
potential, but also because well-run poultry enterprises provide high levels of protein output
per unit of feed (Hrabovszky 1981). Unlike pigs, poultry are associated with few religious or
social taboos. Supply of poultry products in poor countries is technically possible because the
birds easily adapt to most areas of the world, have a low economic value, have a small
generation interval and a high rate of productivity (Smith 1990). Meat from poultry can be
produced within eight weeks and eggs within 18 weeks of the first chick being hatched. In
addition, poultry play an important complementary role to other crop-livestock activities.
Poultry production is a dynamic sub-sector within livestock production systems and
improvements in management and technology in smallholder poultry production systems will
ultimately improve household income and nutrition. This can be achieved through participatory
research in smallholder poultry production. Such research activities would result in poultry
production models clearly describing the input-output relations in, for example, extensive
poultry production systems and interrelationships between poultry production and other crop-
livestock sub-systems.
 Importance and uses of poultry
Poultry are an important source of ready income and protein in rural areas. Rural poultry also
integrate very well and in a sustainable way into other farming activities (Tadelle 1996).
Furthermore, rural poultry play important cultural and social roles in rural communities. The
main objectives of rural poultry are producing eggs for hatching, sale and home consumption,
and birds for sale, healing ceremonies, traditional offerings, replacement, home consumption
and for gifts to visitors and relatives (Mutisi and Kusina 1996; Tadelle 1996; Muchenje and
Sibanda 1997). Money from the sale of the birds is used to buy immediate household
requirements such as food, dairy feeds and to pay school fees, among other purposes. By
eating leftovers from the kitchen and insects such as cockroaches, birds perform a valuable
sanitary function in villages. Poultry manure can be used as field manure or as feed
supplement for ruminants.
Poultry can be kept for socio-cultural or financial purposes. Under such circumstances
introducing management changes such as regular watering and feeding, cleaning the fowl-
runs and taking care of the young chicks can be complex. When designing programmes to
expand poultry production it is important to consider issues related to gender and labour-
sharing within households. Some studies have shown that poultry keeping in most developing
countries is the responsibility of women (Tadelle 1996; Muchenje and Sibanda 1997).
Understanding the roles of each household member is important in sustainable poultry
production.
 Poultry productivity in the smallholder farming sector
Free-range poultry production levels are generally low and most flock sizes average 10 to 20
birds; two to three clutches per year, with an average of 10 to 15 eggs per clutch. The average
annual egg production is between 30 and 60 small sized eggs (Smith 1990; Tadelle 1996;
Muchenje and Sibanda 1997). The low production levels can also be attributed to high chick
mortality levels (Tadelle 1996), diseases, pests, predation, poor housing and poor
management (Muchenje and Sibanda 1997). According to Tadelle (1996), chick mortality,
which may be over 60 % between hatching and end of brooding, represents a major loss in
scavenging village chicken production systems. This may be due to a combination of poor
housing, competition for the same feed resource base with stronger and more vigorous
members of the flock, low protein and energy of available feed, the low hatching weight of
chicks, predation and diseases.
In addition to keeping indigenous chickens, some farmers in rural areas of Zimbabwe rear
broilers and layers. In most cases, broilers are reared and targeted for the December, April
and August public holidays when the market is lucrative. The average broiler flock size in
Zimbabwe is 25 birds, although some farmers can buy as many as 100 chicks at a time
(Muchenje and Sibanda 1997). It is important to study the different production systems in the
smallholder sector to have an insight into the problems and come up with programmes to
alleviate such problems.
Besides chicken production, other poultry species such as turkey, pigeons and ducks are kept
in the smallholder sector. These are mainly for family consumption. The average turkey flock
size in Nharira-Lancashire and Sanyati Smallholder Farming Areas of Zimbabwe was 2–5
birds, and most farmers were interested in expanding their flocks (Muchenje and Sibanda
1997). In the same areas pigeon flock sizes ranged from 2 to 52 birds. Little research work
has been conducted on the production systems of these minor poultry species in most
developing countries (Faure et al. 1996; Kusina and Muchenje 1997).
 Non-adoption of poultry production technology by farmers
Despite considerable research and extension efforts in agricultural systems, smallholder
farmers continue to shun most available technologies, particularly those that relate to
livestock. This may be attributed to inappropriate research methodology, lack of inter-
disciplinarity within research teams, and weak farmer-research-extension interaction (Francis
and Mutisi 1997). Some of the factors resulting in slow progress in smallholder farmers
adopting new technologies in poultry production are highlighted in Table 1.
Table 1. Comparison of characteristics of smallholder crop and poultry production systems.
Factor Crops Poultry Implication for on-farm poultry
research
Mobility Stationary Mobile Difficult to measure and control
non-experimental variables. High
chances of missing data.
Life cycle All units Units seldom Difficult to find comparable units.
synchronisation synchronised synchronised
Multiple units Only grain or tuber Many:- meat, Difficult to measure and value
and residue manure, cultural treatment effect
Non-market inputs Few Many Difficult to value input/output
or outputs
PRODUCER Impersonal Personal taboos Difficult to cull or impose
ATTITUDES treatments or production "deemed"
harmful to animals
Observation units Many Few Large statistical variability:-
experimental design should allow
for loss of experimental unit
Variability of Lower HIGHER Experimental design should allow
observations for loss of experimental unit
Adapted from: Francis and Mutisi (1997).
Farmers are worried to expose their animals to risk through research or trial of new
technologies. Therefore farmers usually prefer that their weakest and/or oldest animals be
used in trials, hoping that they would be improved (Francis and Mutisi 1997). This results in
researchers having difficulties in coming up with proper experimental designs. Furthermore,
there is no synchrony in the life cycles of the birds, making it difficult to obtain enough birds
which meet the specific criteria demanded by researchers, for example the same breed, laying
stage, sex and age. The implication on on-farm poultry research is that usually there are
insufficient replications. Management of poultry by farmers varies considerably. This is mainly
due to different physical and biological site characteristics, and different socio-economic
conditions among farms. Such variations result in wide statistical variability, most on-farm trials
being associated with large coefficients of variation and hence difficulties in interpreting
results, and no statistically significant differences between treatments when such effects may
exist. Incidences of missing data may be high since farmers may fail to avail their birds when
measurements are taken. The complex nature of smallholder farm production systems makes
it difficult to account for all the possible interactions and environmental effects. There is
therefore a need to explore alternative methods of evaluating such trials. Analyses such as
nested designs and incomplete block designs need to be considered in such circumstances.
Multiple non-cash outputs of poultry, such as manure, traditional purposes, home consumption
and status compound the difficulty in analysing poultry enterprises. Since poultry products
consumed by the farming family only passes through non-formal marketing channels,
researchers and decision-makers do not adequately appreciate the economic importance of
such products. Researchers also usually regard the years when crop failure is experienced as
unrepresentative and often stop data collection. This results in recording of peak cash output
from poultry being missed. This calls for supplementing conventional economic analyses with
the farmers' own assessment of the importance of poultry outputs in any research and
development effort. Predetermined agendas and duplication of on-station methodologies
should be avoided. The search for quantifiable data should be done without ignoring farmers'
on-farm trial results based on their experience over the years. On-farm research should be
participatory rather than being extractive.
While both biological and socio-economic aspects of smallholder farm production are
important, there is lack of inter-disciplinarity in most on-farm research activities. Research
teams should be fully integrated and inter-disciplinary involving both biological and social
scientists (Hanyani-Mlambo 1997; Francis and Mutisi 1997). Studies should address the
integrated nature of livestock and crop production subsystems and off-farm activities. There is
also a need to link research and extension activities.
In most on-farm studies there is a weak farmer-research-extension interaction. Participation of
resource-poor farmers is low yet technological recommendations tend to be general.
Extensionists should also participate in research as opposed to only assisting in transferring
technology from researchers to farmers (Francis and Mutisi 1997). There should be joint
planning and review meetings, workshops and conferences between researchers,
extensionists and farmers. Consideration of farmers' knowledge/technology and language use
are pivotal in adopting any new technologies. Researchers should therefore view farmers as
consultants, collaborators or colleagues. Participatory research should be introduced/improved
in agricultural colleges' and universities' syllabi. Holistic farming systems research (FSR) that
go beyond biological production in crop and livestock subsystems to include utilisation and
marketing of outputs, and off-farm activities are needed.
Some impressive advances have already been realised through utilisation of the FSR
approach. The International Livestock Research Institute (ILRI) has been applying the FSR
approach mainly on pastoral systems (Hanyani-Mlambo 1997). In Zimbabwe the FSR Unit in
the Department of Research and Specialist Services (DR and SS) has carried out research
using the FSR approach in Chivi and Mangwende Communal Areas. The Faculty of
Agriculture, University of Zimbabwe has been using the same approach in studying crop-
livestock farming systems in several smallholder farming areas in Zimbabwe.
 Possible changes to improve productivity of indigenous flocks
According to Smith (1990), the low input-output levels are economically efficient, since with the
low output from the individual birds, the inputs are low or virtually non-existent. Small
management changes, such as regular watering, night enclosures, discouraging hens from
getting broody, vaccination against common diseases, small energy and protein supplements
and caging chicks can bring about significant improvements in the productivity of indigenous
birds (Tadelle 1996). Smith (1990) estimated that under scavenging conditions the
reproductive cycle consists of a 10-day laying phase, a 21-day incubation phase and finally a
56-day brooding period. This implies a theoretical maximum number of 4.2 clutches per hen
each year, although in reality the number is 2–3 (Mutisi and Kusina 1996; Muchenje and
Sibanda 1997). In the Eastern hills of Nepal, indigenous birds kept under semi-extensive
management conditions produced 125 eggs in 11 clutches per annum, compared to 3–4
clutches per annum produced under normal scavenging systems (Smith 1990). By improving
existing management, it is possible to increase indigenous poultry production in the
smallholder sector. Benefits from genetic improvement can only accrue if existing
management is improved.
 Feed resources and requirements
The feed resource base for rural poultry production is scavenging, and consists of household
waste, anything edible found in the environment and small amounts of grain supplements. It is
not constant. The portion that comes as a grain supplement and from the environment varies
with activities such as land preparation and sowing, harvesting, grain availability in the
household and season and the life cycles of insects and other invertebrates. According to
Tadelle (1996), protein supply may be critical, particularly during the drier months, whereas
energy supply may be critical during the rainy season. In the absence of an event that
diminishes the flock biomass, such as diseases, the village flock will normally be at the
maximum biomass that can be supported by the available feed resource base. Any additions
to the flock that increase the biomass result in increased survival pressure and selection
against the weakest members of the flock. The feed resource base is usually deficient in
protein, energy and probably calcium for layer birds.
Determination of the nutritional status of indigenous birds and an assessment of changes in
their diets throughout the year can be done by analysing crop contents and doing carcass
measurements. After slaughtering the birds, materials in the crop are combined, mixed and
physically identified (Feltwell and Fox 1978). Seeds, insects, worms and plant materials are
separated, weighed, and recorded, and the proportion of each calculated. The collected crop
contents can also be analysed for dry matter (DM), crude protein (CP), ether extract (EE),
crude fibre (CF), and ash according to AOAC (1985), for calcium (Ca) by atomic absorption
spectrophotometer, and for Phosphorus (P). The metabolisable energy (ME) levels of the
contents can also be determined by an indirect method, using the following equation
(Wiseman 1987):
True ME (Kcal/kgDM) = 3951 + 54.4EE-88.7CF-40.8Ash.
Different parts of the carcass can be separated and measured. Carcass weight, gastro-
intestinal tract (GIT), proventriculus, crop, gizzard, liver, ovary and oviduct weights can also be
measured. Lengths of parts of the body such as the GIT, caecum and circumference of the
upper leg (thigh) can also be measured. Carcass weight can be taken after the offal
comprising feathers, lower leg, heart, crop, pancreas, lungs, head and digestive and
urogenital tracts have been removed. Then the carcass dressing percentage can be
calculated.
The possibility of increasing free-range poultry production through energy and protein
supplementation, and improved housing also needs to be investigated. For the benefit of those
smallholder farmers who keep broilers and layers, the need to consider home grown feeds as
substitutes is not overemphasised. This is against a background of ever escalating feed costs.
 Breeding in extensive production systems
Practically all village producers breed and rear indigenous day-old-chicks for themselves in
order to maintain or increase the number of their flocks (Smith 1990; Lambrou 1993; Mutisi
and Kusina 1996; Tadelle 1996; Kusina and Muchenje 1997; Muchenje 1997). The breeding
of indigenous birds is not controlled because of the mixing of different flocks. The breeding
season for village birds varies according to the condition of the bird, the natural availability of
feed and the weather. According to Lambrou (1993) and Kusina and Muchenje (1997), the
breeding season for village birds in Zimbabwe is from September to April. A strategic
supplementary feeding during the breeding season will improve the indigenous bird condition
at breeding, the breeding efficiency and will increase egg production, fertility and hatchability.
There is also a need to come up with a more effective and economic ratio of males to females
in indigenous birds. Is the 1: 15 ratio generally recommended in Zimbabwe (Lambrou 1993)
the most economic and effective?
 Markets
Ways in which farmers can improve their image and offer a competitive and consistently
quality product on a regular basis to their customers need to be explored. Furthermore,
aspects such as transport, market accessibility, marketing channels and organisation need to
be addressed if farmers are to benefit from research. In rural areas poultry products are
usually sold locally. Densely populated areas such as growth points, service centres, boarding
schools and other government institutions are important markets for poultry products.
In Zimbabwe broilers are usually sold when they are at least six weeks old, although some
farmers keep the birds for more than twelve weeks, depending on season and demand.
Where people are into broiler production, indigenous birds are rarely sold and cost less than
broilers. However, in areas where broiler production is not common, indigenous birds are
usually sold. Usually, there is no defined selling age for the indigenous birds. They are sold
depending on supply and family financial demands, for example, when school fees are
needed. In a survey carried out in Nharira-Lancashire Smallholder Farming Area of Zimbabwe
(Muchenje and Sibanda 1997), layers were reported to start producing eggs at between 24 to
28 weeks and the eggs were being sold at the same price, irrespective of their size. The
market for broilers and eggs in all areas was said to be favourable during holidays such as
Christmas and Easter. The market for the less common poultry species, such as turkeys,
pigeons and ducks, is normally not clearly defined.
 Other issues to consider in smallholder poultry research and
development
 Conservation of genetic resources in poultry
Conservation of domestic animal diversity (CDAD) encompasses characterisation,
identification, monitoring and utilisation to ensure management for best short term use and
longer term ready availability (Moyo 1995). There is a need to conserve "adaptive" traits as
opposed to "commercial" traits. The conserved genetic resources will be future sources of
unique genes and will be useful when environmental concerns necessitate change in
production systems. Indigenous birds in most African countries are not adequately
characterised to understand the existing diversity to facilitate development of rational
utilisation and conservation strategies. Biotechnology techniques such as gene mapping could
be used to characterise indigenous birds. It should, however, be noted that several Southern
African countries face a lack of infrastructure for breeding purposes and have relatively small
defined populations (Wollny 1995a; Wollny 1995b). There is a need to consider
characterisation of indigenous poultry breeds if genetic conservation for economic, scientific,
cultural and social, development and sustainability (Wollny 1995b) is to succeed.
 Computer simulation modelling in farming systems and scientific research
Computer simulation models are useful tools, which can be used in studying farming systems.
They can be used to describe the behaviour of business, economic, social, biological, physical
or chemical systems over periods of time. The tremendous revolution in computer technology
has resulted in growth in simulation modelling and this has eased the integration of
agricultural sciences.
Computer simulation models mimic real life situations. They can be used to make the same
decisions that would be made if it were feasible to experiment with the system itself. Since a
system is a collection of entities acting and interacting together towards accomplishment of a
logical end, computer simulation models can play an important role in systems analysis. A
model for semi-scavenging poultry rearing has been developed in Bangladesh and has been
applied with considerable success to the benefit of rural, poor women (Rahman et al.,
undated). The model included the indigenous breed (desi) and its crosses with exotic breeds.
Some of the traits, which were considered in the model, were age of first egg, eggs/hen/year,
mortality rate, and mortality due to predator.
A computer simulation model requires formalisation of current knowledge in the form of
equations, solutions of the equations and interpretation of the results. Models have been used
in studying metabolism. They integrate a variety of nutrient inputs for their outputs as
products, heat or other metabolic consequences and satisfy an objective of intake models-
that of integrating a wide variety of factors known to influence intake. The interaction between
physical and metabolic pathways affecting feed intake have therefore been successfully
integrated by modelling.
A bio-economic model developed in Brazil has been used to help in identifying priority areas
for applied research, to explore various combinations of herd potential and level of feeding,
the behaviour of different herd sizes, and the potential returns from specific technical
innovations (Brockington et al. 1983). Another model was developed to assess the risks and
returns from establishing improved pasture in an extensive beef breeding enterprise, for use
by farmers and extensionists (Beck et al. 1982).
 Conceptual structure
The main components of the input-output system of free-range poultry production systems
and poultry-other-crop-livestock systems can be incorporated into the models. Concepts and
quantitative values from literature and field studies can be used to develop the models.
Quantitative data may be used to describe the relationships and define default values, and the
qualitative data can help in coming up with reasonable assumptions for the model.
 Model building and validation
The discipline of mathematics, which provides appropriate structure with which to represent
the biological and socio-economic concepts, can be used to interrelate the component parts.
The processes within the systems are normally represented by sets of equations. Laws of
probability guard stochastic models. They involve use of random variables, for example,
inheritance. This means that a spread of results can be obtained from the same initial values
and parameter values. Deterministic models give definitive predictions, without any probability
distribution. An independent variable causes consistent changes in the dependent variable.
Deterministic models can therefore be used for determination of nutritional requirements of
free range chickens, and egg output for both free range chickens and intensively produced
layers. They can also be used to predict the flock biomass of birds the scavenging feed
resource base can support. Stochastic models can be used for calculating aspects such as
chick mortality rates. Mathematical models can also be used to simulate the behaviour of the
poultry-other-crop-livestock systems free-range poultry production systems. Models that
predict the effect of flock size on scavenging feed resource base, weights of chickens,
mortality; output for sale/consumption and flock composition may be incorporated.
The effects of changes in fixed parameters on model performance need to be examined. This
is done to test the correctness of the internal workings of the system and the model
assumptions. Statistical comparison of observed values from an independent data set and
model predictions can be done using regression/correlation and the _2-test for goodness of fit.
The model must be able to simulate reality if it is going to be useful. Aspects such as inputs
and outputs need to be recorded in the field. Such field results are tested against model
predictions using regression/correlation and the _2-test for goodness of fit. Recommendations,
as regards the use of the model and the possibility of redesigning it, can then be made.
 The future
Poultry are an important component of farming systems in smallholder farming areas.
Research efforts in poultry should be targeted at aspects ranging from feeding, breeding,
disease control and husbandry to obtain a better understanding of poultry production systems
in relation to other crop-livestock systems and the changes in socio-economic conditions in
smallholder set-ups. Biological factors are not the only factors considered by farmers when
adopting technologies to suit their local farming circumstances. Instead, socio-economic
factors are also important to the farmer and, therefore, should not be ignored. Both objectively
measured variables and subjectively measured variables should be studied to produce more
meaningful research results.
Detailed studies of most of the issues raised above would contribute to improved poultry
production and contribute towards improving family income and nutrition. Qualitative data can
be gathered in a once off/single visit. Reliability of such data depends on the ability of farmers
to recall past events. Smallholder farmers do not usually keep farm records! More reliable data
are obtained through monitoring studies and discussions with farmers. All the different aspects
of poultry covered in this paper should be incorporated into a model together, rather than
individually, in order to identify combinations that optimise smallholder poultry production and
improve smallholder household's income and nutrition.
 References
Association of Official Analytical Chemists (AOAC). 1985. Official Methods of Analysis. 12th
ed. Washington, D.C., 957p.
Beck A. C., Harrison L., and Johnson J. H. 1982. Using simulation to assess the risks and
returns from pasture improvement for beef production in agriculturally underdeveloped
regions. Agricultural Systems 8: 55–71.
Brockington N. R., Gonzalez C. A., Veil J. M., Vera, R. R., Texeira N. M., and Assi A. G. 1983.
A bio-economic modelling project for small production systems in south-east Brasil.
Agricultural Systems 12: 37–60.
Faranisi A. T. 1995. Poultry Breeding in Zimbabwe. In Proceedings of the International
Symposium on Livestock Production through Animal Breeding and Genetics, 10 to 11 May,
1995, (Eds.) Dzama, K., Ngwerume, F. and Bhebhe, E., Sheraton Harare, Zimbabwe.
Faure J. M., Melin J. M. and Mantovani C. 1996. Welfare of guinea fowl and game birds.
Proceedings, Fourth European Symposium on Poultry Welfare. Published by Universities
Federation for Animal welfare. (Eds.) Savory, C. J. and Hughes, B. O. London, UK.
Feltwell R. and Fox S. 1978. Practical Poultry Feeding, Faber and Faber Limited, Great
Britain.
Francis J. and Mutisi C. 1997. Review of Smallholder Crop-Livestock Research in Zimbabwe:
Lessons for the future. Proceedings of the Workshop on Crop-livestock farming systems
research methodologies training, 1–12 December 1997, (Ed.) Sibanda, S, Faculty of
Agriculture, University of Zimbabwe, Harare.
Government of Zimbabwe. 1995. Zimbabwe's Agricultural Policy Framework 1995–2020.
Ministry of Agriculture, Harare, Zimbabwe.
Hanyani-Mlambo B. T. 1997. Technical Issues to consider: Socio-Economic Aspects.
Proceedings of the Workshop on Crop-livestock farming systems research methodologies
training, 1–12 December 1997, (Ed.) Sibanda, S., Faculty of Agriculture, University of
Zimbabwe, Harare.
Hrabovszky J. P. 1981. Livestock Development: Toward 2000 with special reference to
developing countries. World Animal Review, 40:2–16.
Kusina J. and Muchenje V. 1997. Micro-Project Feasibility Study in Small Animal Breeding in
Angwa-Neshangwe-Gonono, Guruve, Zimbabwe. CIRAD-IFRA Report. Unpublished.
Lambrou L. 1993. Indigenous Poultry in Zimbabwe. Farming World, 19 (3): 11–12.
Ministry of Agriculture, Food and Fisheries (MAFF). 1985. Small-scale guinea-fowl production,
London. UK.
Moyo S. 1995. Conservation strategies for endangered livestock species. Proceedings of the
international symposium on livestock production through animal breeding and genetics, 10 to
11 May, 1995, (Eds.) Dzama, K., Ngwerume, F. N. and Bhebhe, E, Sheraton Hotel, Harare,
Zimbabwe.
Muchenje V. 1997. Potential for Improved Poultry Production in the Smallholder Sector. Paper
presented at a Workshop on Commercial Smallholder Goat and Poultry Production, 15
October 1997, Mukamba, Wedza, Zimbabwe.
Muchenje V. and Sibanda S. 1997. Informal Survey Report on Poultry Production Systems in
Nharira-Lancashire and Sanyati Farming Areas. Zimbabwe. Unpublished.
Mutisi C. and Kusina N. T. 1996. Draft Report-Role of livestock in Sanyati Communal Area.
Zimbabwe. Unpublished.
Rahman M., Sorensen P., Jensen H.A., Dolberg F. (undated). Exotic hens under semi-
scavenging conditions in Bangladesh. Bangladesh. Unpublished.
Smith J. A. 1990. Poultry. Centre for Tropical Veterinary Medicine. University of Edinburgh.
Tadelle D. 1996. Studies on village poultry production systems in the Central Highlands of
Ethiopia. MSc Thesis, Swedish University of Agricultural Sciences, Uppsala, Sweden.
Wiseman J. 1987. Feeding of Non-ruminant Livestock. Butterworth and Co Ltd. pp 9–15.
Wollny C.BA. 1995a. Breed improvement and future breeding strategies for livestock in
Southern Africa. Proceedings of the international symposium on livestock production through
animal breeding and genetics, 10 to 11 May, 1995, (Eds.) Dzama, K., Ngwerume, F.N. and
Bhebhe, E, Sheraton Hotel, Harare, Zimbabwe.
Wollny C.B.A. 1995b. Conserving the biodiversity of Southern African livestock resources.
Keynote address presented at the Annual Symposium of Developing Areas Branch of the
South African Society of Animal Science (SAGAS), 6 to 19 October 1995, Loskopdam,
Groblersdal, Republic of South Africa.
Analysis of socio-economic factors influencing egg
production in Maseru District. Lesotho
F.O.C. Nwonwu
Department of Agricultural Economics and Rural Sociology
National University of Lesotho, Lesotho
Abstract
Introduction
Materials and methods
Surveys
Sample selection
Data
Analytical models
Test of hypotheses
Cobb Douglas production function
Results and discussion
Status of egg production in Lesotho
Age and marital status of farmers
Number of birds
Technology in use
Regression results
Linear models
The Cobb Douglas production models
Elasticities
Correlation
Conclusion
Recommendations
References
 Abstract
This study investigated the socio-economic factors that affect egg production in Maseru
District, Lesotho. The nature of the egg production function was analysed and the degree of
responsiveness of output to changes in capital, technology, number of birds, and type of feed
as decision variables was estimated. A sample of 45 poultry farmers was selected and
interviewed. Both a linear and a Cobb-Douglas production function were used to predict egg
output and to determine output elasticity with respect to the independent variables. In the
Cobb Douglas function, the natural logarithm of egg output was regressed on the natural
logarithms of capital, technology, number of birds, and type of feed. Results show that the
average age of farmers is 45 years; the industry comprises of small-scale operators with an
average of 230 birds per farmer; an average investment capital of R.2000 per enterprise; and
an annual output of 6000 eggs per farmer. Of the four independent variables used as
predictors, two of them namely number of birds and type of feed are statistically significant in
predicting egg production at the 1% level while technology is significant at the 10% level with
an adjusted R2 value of 0.99. The elasticity of egg output with respect to the number of birds
(єb) is unity showing that output of eggs would change by as much as the percentage change
in the number of birds. Egg output elasticity with respect to capital (єk), technology (єt), and
type of feed (єf) stand at 0.69, 0.06, and 0.09, respectively. It is concluded that with such a
high responsiveness of output to changes in capital, egg production in Maseru District would
be significantly increased if sufficient capital were injected into the industry.
Key Words - Egg production, Maseru, Cobb-Douglas, elasticity, natural logarithms
 Introduction
The inception of large-scale egg production in Lesotho is as recent as 1971 when the
Department of Agriculture introduced commercial poultry farming through its poultry
improvement scheme. Prior to that date, in the 1950s and 1960s, egg production in the
country was limited to non-commercial small-scale production by subsistence farmers who
stocked indigenous or range scavenging breeds of birds (GOL 1986a). Commercial egg
producers are still very few in the country and their stocks range between 1500 to 20,000 birds
(GOL 1988). The small-scale operators with laying stocks of 50 to 300 birds are fast growing
to expand the poultry industry in the country (Mafoso 1999). With the formation of the Lesotho
Poultry Association and the passage of two important laws: the Agricultural Marketing Act of
1967 and the Agricultural Marketing (Egg Control) Regulations of 1969, the commercial egg
industry began to take on the outlines of its present day structure (GOL 1986a).
It is estimated that there are currently about 300,000 egg producing layers in the poultry
industry in the country (Anon 1999). Pullets are mainly imported from South Africa. The
demand for pullets by the small-scale farmers is met purely through domestic supply by the
Lesotho Government Poultry Production plant, which has a total capacity of 75, 000 pullets
(Anon 1999). The fast growth in the industry is predicated on the high potentials in meeting the
social and economic development objectives of increased employment, higher and equitable
distribution of income, and improved nutritional standards at lower cost. The investment
opportunity and the dominant role, which the industry offers women, are seen as an added
advantage.
Recent surveys reveal that Maseru District is the highest producer and consumer of poultry
products in the country. Hence, a study of the industry situation in Maseru, which is also the
capital city of Lesotho, can be extrapolated to represent the industry situation in the country.
The marketing system for eggs is a factor, which poses a major setback to the development of
the poultry industry in Lesotho. This is because the Lesotho Agricultural Marketing Regulations
Act of 1975 established a single marketing Regulation (GOL 1991). The Act stipulates a single
marketing channel for eggs in the country. It also empowered the Lesotho Poultry Co-
operative Society to be the sole buyer and seller of eggs in the country. The poultry farmers
opposed this policy and demanded that they be allowed to operate a competitive market
structure. Their desire was further heightened by the failure of the co-operative society to pay
farmers promptly for commodities already supplied. The incentive to increase egg production
became so successful that excess supplies between the months of August through December
became a major problem of the poultry industry. The districts of Leribe, Berea and Maseru
have experienced this over supply at both producer and wholesale levels and large quantities
of eggs have spoiled as a result of the glut in the market (GOL 1986b).
Livestock farming is of both cultural and economic importance to Basotho people. Animals are
kept as symbols of wealth, for payment of bride price, to provide meat during celebrations
such as funerals, and for sale to earn income. Poultry production appears very conducive to
the Lesotho socio-economic and biophysical characteristics. As a small country and with
pastoralism as the main occupation of the people the land is over stocked and overgrazed.
Large ruminants such as cattle, horses and donkeys on the one part, and small ones such as
sheep and goats on the other, exert much pressure on the available land. The disproportionate
demand for grazing land vis-à-vis the available land, causes even fragile areas to be grazed
with serious threat to the environment. A diversion from land demanding agricultural
enterprises to less land- oriented investment such as poultry keeping is considered the way to
save the environment, diversify the investment portfolio, and guarantee employment and
better income distribution in the country. Projected demand for eggs beyond the year 2000
shows that demand will outstrip the supply. The excess demand is caused by the increasing
desire of Lesotho like other developing countries, to improve the nutritional status of her
people. Eggs and poultry meat are easy and affordable options for fulfilling this need.
This study analyses the social and economic factors that affect poultry production in Maseru
District in Lesotho. The extent to which the identified factors affect poultry production are
investigated. The degree of responsiveness of egg output to changes in the factors included in
the model are quantified and the policy implications interpreted. It is hoped that the study will
shed more light into strategies to adopt to reduce the over-dependency on South Africa for
meeting the deficit in egg supply in the country. Recommendations are made on how to
improve on the performance of the poultry industry to increase the supply of eggs and
enhance the nutritional status of the people.
 Materials and Methods
 Surveys
An initial survey was carried out nation-wide to scrutinise the size and distribution of poultry
farmers and the geographical and spatial spread in the country. This survey identified Maseru
District as the major producer and consumer of eggs in the country. The district was then
selected for study based on the strength of the survey findings. The district was stratified into
north, east and west for purposes of this study. A preliminary survey was conducted to
determine the geographical spread and concentration of poultry farmers within the different
strata. The preliminary survey was also used to pretest the questionnaire to validate its
comprehensiveness. Maseru North, which has similar socio-economic and cultural
characteristics as the East and West was used to pretest the questionnaire. The final survey
conducted in Maseru East and West involved house to house visit to farmers and interview
sessions during which discussions were held and specific questions were also posed to
farmers. Data from responses to the questions were collected with the help of structured
questionnaires.
 Sample selection
A multi-stage sampling method was applied in the selection of farmers. The farmers were first
identified in clusters of residential areas within the urban and peri-urban locations within the
chosen clusters. The sampling frame was then defined as the clusters of hamlets and
settlements contained in Maseru East and West from where samples were finally chosen. In
the first stage, a judgement or purposive sampling method was used to select clusters. The
basis of the selection is to ensure that farmers that face different scales of operation,
technology, infrastructure facilities and social amenities have equal and likely chance of being
selected. The diversities considered in the sample selection include population of clusters,
accessibility to social infrastructure such as power supply, access roads, potable water,
historical and cultural considerations and demonstrable interest in poultry production.
Within the selected clusters, a simple random sampling method (SRS) was used to select the
sample of farmers. Samples proportionate to the sizes of clusters were selected. A total of 45
farmers, which cut across the geographical spread and diversity in resource endowment,
socio-cultural characteristics and scale of operation were selected from the district.
 Data
Primary data was collected in the field from field observations during the preliminary surveys
and during interview sessions in the main survey. Data collection was done with the help of a
structured questionnaire and through note taking during the interviews. Secondary data was
collected from the Bureau of Statistics annual publications; Annual Reports and Marketing
Bulletins of the Ministry of Agriculture; the Central Bank Annual Reports; and publications of
the Marketing Division of the Ministry of Industries.
Types of data collected and used in the analysis include number of laying birds; type of
technology in use (deep litter and battery cage); types and cost of feeds; sources and costs of
point of lay (POL) birds; sources of heat energy (paraffin, gas, electricity, solar). Others are,
scale of operation and equipment in use; breeds of bird; management technique (whether
farmer buys day-old chicks or POL birds); price of day old chicks; output of egg; price per unit
of eggs; age; and marital status of farmer.
 Analytical models
Two production function models, a linear and a curvilinear function were adopted in the
analysis of the data. The linear and curvilinear functions were used to estimate decision
parameters and to determine which of the two groups of models best predicts egg output
given the prevailing production characteristics in the industry. In the linear equation, egg output
was assumed to be dependent on number of birds, type of feed, capital, technology, age, and
marital status of farmer. These variables were introduced in stepwise manner into the model
one after another. The decision rule is to choose the combination of variables or the single
variable that has the highest adjusted coefficient of determination (R2
adj). The underlying
assumption is that the introduction of more variables into the model will reduce the standard
error of estimate and therefore improve the value of R2
adj. For each model introduced, the
value of the R2
adj is observed and the model with the highest R2
adj is chosen as the best
predictor for egg production. The functional form of the linear model is first assumed to be a
simple relation in which output of eggs depends only on one of the variables say the number
of birds. This is represented functionally as follows:
Q = f (N,є)     (1)
The multiple form of the relationship assumes that output of eggs is dependent on all the
independent variables as shown in Equation 2.
Q = f ( N, F, K, T, M, A, є)     (2)
Where:
N = number of birds
F = type of feed
K = capital
T = technology
M = marital status of farmers
A = age of farmer
є = stochastic factor or error term due to variables not included in the model.
The prediction equations for the simple form of the model is expressed generally in the form
shown in Equations 3.
Y =β0 +β1X1 +є          (3)
The prediction equation for the multiple form of the model is shown in Equation 4.
Y = β0 +β1X1 +β2X2 +β3X3 + β4X4 + β5X5 +β6X6 +є         (4)
 Test of Hypotheses
The single and multiple regression models are used to test different hypotheses based on a
priori assumptions. The single regression analysis tests the null hypothesis H0 against an
alternative one H1, both of which state as follows:
H0:β1 = 0
H1: β1 ≠ 0
The null hypothesis states that the mean for number of birds is zero. In other words, it states
that the number of birds which a farmer has does not significantly contribute to the amount of
eggs produced in the farm. The alternative hypothesis states that the number of birds in the
farm has an effect on the amount of eggs produced.
The multiple regression model tests a null hypothesis H0 against an alternative hypothesis H1
both of which state as follows:
H0:βi's = 0
H1:βi's ≠ 0
(i = 1,2,..,6)
The null hypothesis for the multiple model states that none of the six independent variables
has effect on the output of eggs. The alternative hypothesis states that at least one of the
factors contributes to output of eggs.
 Cobb Douglas production function
The Cobb Douglas Production Function is a typical case of a production function that is
homogeneous of the degree one (Henderson and Quandt 1980). In other words, if we vary the
inputs by some amount, the output will vary by the same amount as the input. Since it is
believed that most production functions are curvilinear, the single independent variable
relationship is expressed as in Equation 5.
Q = AXα        (5)
(0<α<1)
In this multiple variable non-linear model, the marital status and age of farmer were dropped
and the output of eggs was then estimated with the number of birds, capital, technology and
type of feed. The price of eggs was also not included in this estimation because a constant
wholesale price for eggs was assumed in the analysis.
The multiple equation states that output of egg is determined by four factors represented by
X1, X2, X3 and X4 as shown in Equation 6.
Q = AX α_X β_ ψ 
2 X δ  
1 3 X4     (6)
(α+β+δ+ψ =1)
The estimation of the parameters α , β, δ, ψ is done by transforming the curvilinear function to
a linear one. This is done by taking the natural logs of the variables and regressing the natural
log of output Q on the natural logs of the constant A and those of the variables X1, X2, X3, and
X4. The coefficients α, β, δ, and ψ of the independent variables X1, X2, X3, and X4 represent
the elasticities of egg output with respect to the independent variables.
The linear form of Equation 6 following the transformation is expressed as in Equation 7.
lnQ = lnA + αlnX1 + βlnX2 + δlnX3, + ψlnX4       (7)
 Results and discussion
 Status of egg production in Lesotho
Maseru District is the largest producer of eggs among the 10 districts in Lesotho. The district
alone accounts for 34 per cent of the total egg production in the country (see Figure 1). The
other districts are Berea, Leribe, Botha-Bothe, Mafeteng and Mohale's-Hoek, Qacha's Nek,
Mokhotlong Thaba-Tseka and Quathing. The high percentage share of Maseru in egg
production and in the poultry industry at large stresses the importance of the district in the
industry and the relevance of this study. The district also occupies a dominant position in the
consumption of eggs and poultry meat relative to other districts in the country.
 Age and marital status of farmers
Results show that the age of farmers ranges between 28 to 60 years with an average age of
45 years. This result agrees with the findings of GOL (1987) which indicated that 70 per cent
of egg producers in Lesotho are over 40 years of age. The age class distribution for the
sample of 45 farmers is shown in Figure 2. With an age class interval of 10 years, the figure
shows a normal distribution with majority or 17 farmers falling within middle age group of 41–
50 years. The number of farmers who fall within the younger age bracket of 31–40 and those
that fall within the older age group of 51–60 are equal at 14. Only one farmer falls below the
age of 30 years and there is no farmer above 60 years of age. It is shown that the majority of
the farmers or 89 per cent are married or have been married with 30 of them married and 10
widowed. Only 5 farmers or a mere 11 per cent are single.
 Number of birds
The study shows that egg production in Maseru District is predominantly on a small-scale as
demonstrated by the number of birds per farmer. Results show that the number of birds per
farmer ranges from as little as 25 birds to a maximum of 1300 birds with a mean of 230 birds.
The size distribution of farms with respect to the number of birds is shown in Figure 3. The
figure shows that majority of the farmers, equivalent to about 44 per cent of the total, have
less than or equal to 100 birds. Only one farmer has more than 1000 layers in the industry.
Considering that commercial poultry farming attracts as many as tens of thousands of birds
per farmer, this result clearly shows the small-scale and subsistence nature of egg production
in Maseru District. As already noted, the country depends on South Africa for meeting the
deficit in the supply of eggs at the moment.
 Technology in use
About half of the farmers or 22 of them, representing 49 per cent of the sample, use the
battery cage system of production while the remaining 23 or 51 per cent use the deep litter
system. According to the farmers, the deep litter system is much cheaper to set up as it
requires relatively much less initial capital outlay. However, the system has its drawbacks
which include much greater disposition of birds to disease and pest attacks; it is labour
intensive to manage especially in feeding, egg collection, and cleaning of the floor and
changing of the litter. Productivity is relatively lower partly because of the pest and disease
menace but also due to the vices of birds especially with pecking one another.
The battery cage apart from its relatively high initial capital demand, better reduces all the
disadvantages associated with the deep litter such as high incidence of diseases, pecking, and
dirtiness. It also minimises labour demand and supervision time. It achieves relatively higher
productivity per unit of resource committed into production. The system also minimises the
wastage of inputs such as feeds and water through contamination and trampling which is
common with the deep litter system.
Paraffin is the principal source of energy for heating with 28 or 62 per cent of the farmers
reporting its use. Electricity is second to paraffin with 15 farmers or 33 per cent of the sample
farmers reporting the use of this device for heating. Only one farmer reported using gas for
heating the poultry house. Heating is a major contributor to cost of production considering that
Lesotho goes through some severe four months of winter every year. Paraffin is relatively
cheap, hence the majority of the farmers opt for its use. However, it has the disadvantage of
being relatively less effective in heat supply compared to electricity or gas. It also has the
disadvantage of being highly inflammable with relatively higher chances of causing fire
outbreaks especially in the deep litter system. Electricity is reported to be far more effective for
heating but more costly to use.
Two breeds of bird namely the Lohmann and Hyline Brown are commonly kept by farmers.
The two breeds were selected after some initial countrywide screening of breeds for egg
productivity, egg size, shell quality, interior quality, liveability and thriftiness in feed
consumption. Four types of feed are commonly used by farmers in the egg production industry
in the district. They are the Makhulo, Nelko, Epol and Moreson. Of the four, the most widely
used is the Makhulo with about 56 per cent of the farmers reporting its use. The least used is
the Nelko with only 9 per cent of the farmers reporting its use.
 Regression results
 Linear models
The results obtained with the single and multiple linear models are presented in Table 1.
Table 1. Estimated Parameters for Different Models Used in Predicting Egg Output with
Assumption of Linear Relationship
    
Single regression R2 R2
adj Significance (F) Coefficient (β) Significance (t)
Independent variable 0.772 0.776 0.000***   
Capital   Constant –7.272  
  Capital 3.296 0.000***
Multiple regression 1 0.992 0.991 0.000***   
Independent Variables   Constant 153.597 0.821NS
Age, marital status,   Age –15.762 0.228NS
 Number of birds   Number of birds 29.227 0.000***
  Marital status –4.005 0.979NS
Multiple regression 2 0.992 0.992 0.000***   
Independent Variables   Constant –1032.803 0.237NS
  Age –4.615 0..734NS
Age, marital status   Number of birds 28.88 0.000***
Number of birds   Marital status 127.719 0.429NS
Type of feed   Type of feed 224.518 0.043**  
NS = not significant at 10%; ** = significant at 5%; *** = significant at 1%
The results show that capital is a major limiting factor and contributor to egg production. The
importance of capital is illustrated from the single regression equation, which has capital as
the explanatory variable. It illustrates that of all the variables assumed a priori to affect egg
production in Maseru District, capital is the single significant factor in the prediction of egg
output.
In the three-variable multiple regression equation, only the number of birds among the three
variables shows a highly significant effect on egg production at the 1% level. The age and
marital status of farmer show no statistical significance. However, the highly significant F-ratio
and the high value of R2
adj indicate some goodness of fit. The significance of the F-ratio
demonstrates that all the variables included in the model contribute significantly to the
prediction of egg output. The four-variable model further demonstrates the utility of the multiple
regression approach to statistical analysis. The model shows further improvement of the R2
adj
from 0.991 in the three-variable model to 0.992 as a result of the addition of the variable type
of feed. Overall, the multiple regression model improved the quality of the prediction as
demonstrated by its improvement of the R2
adj from 0.766 in the single regression model to
0.992 in the four-variable model. This result contradicts the view of Hair (1967) who reported
that single regression models are better than multiple regression models in prediction. It is
further noted that there is no model that used all the six independent variables chosen a priori
to predict output of eggs.
 The Cobb Douglas production models
Results obtained with the curvilinear functions are shown in Table 2
Table 2. Estimated Parameters for Different Models Used in Predicting Egg Output with
Assumption of Curvilinear Relationship
    
Single regression R2 R2
adj Significance (F) Coefficient Significance (t)
(β)
Independent Variable 0.987 0.986 0.000***   
Natural Log of   Constant  2.937 0.000***
Number of Birds   Number of birds 1.058 0.000***
 
Multiple Regression 1 0.990 0.989 0.000***   
Independent   Constant 2.977 0.000***
Variables
Natural Logs of   Number of birds 1.041 0.000***
Number of Birds,  
Type of Feed   Type of feed  0.09 0.001***
Multiple Regression 2 0.806 0.797 0.000***   
Independent   Constant 3.484 0.000***
Variables
   Capital 0.690 0.000***
Natural logs of   Technology 0.06 0.77NS
Capital and
Technology
Multiple Regression 3 0.991  0.990 0.000***   
Independent   Constant 2.874 0.000***
Variables
Natural logs of   Capital 0.019 0.491NS
Capital
Number of Birds   Number of birds 1.03 0.000***
Technology   Technology 0.088 0.071*
Type of Feed   Type of Feed 0.09 0.003***
The curvilinear form generated four possible equations. They include a single regression
equation with the number of birds as the only independent variable. The other three comprise
of two models with two of the four independent variables alternating in their inclusion while
one model has all four independent variables included as shown in Table 2.
In all four models, the number of birds has proved to be highly and statistically significant at
the 1% level in predicting the output of eggs in the poultry industry. Type of feed also shows
high statistical significance at the 1% level in the prediction of output of eggs. While technology
is not significant in the two-variable model it is fairly significant at the 10% level in the four-
variable model. Capital shows high statistical significance in the two-variable model and no
statistical significance in the four-variable model.
All four models show goodness of fit as demonstrated by highly significant F-ratios at the 1%
level. The R2 and R2
adj values for the models are very high and further support the fact that
the variables included in the models contribute significantly to the prediction of output of eggs.
Only the model with capital and technology has relatively low R2 and R2
adj values of 0.806
and 0.797 respectively. This result shows that in relative terms, the two-variable model with
capital and technology are less accurate or efficient than number of birds and type of feed in
predicting output of eggs in Maseru District. From the stepwise regression analysis, it was
seen that when the number of birds and type of feed were introduced in the model, technology
and capital were the variables rejected from entering the model.
Since the battery cage and deep litter systems of egg production depict technology in this
analysis, the result demonstrates that the use of either method in egg production does not
significantly affect the amount of eggs produced in Maseru District. This result is supported by
the findings of the Planning Division of the Ministry of Agriculture, Co-operatives and
Marketing, which reported that the operating cost of a 600-bird battery cage unit is basically
the same as that of a 600-layer deep litter unit (GOL 1987).
That capital shows no statistical significance in the four-variable model portrays the dominance
of small-scale producers in the poultry industry in Maseru District. With holdings of as little as
25 birds, an average of 230 birds per farmer, and only one farmer having over 1000 birds, egg
production in the district is far from being capital intensive.
Since the four variables have shown statistical significance in both the single and multiple
regression models, we then reject the two null hypotheses and accept their corresponding
alternative hypotheses. We therefore conclude that at least one of the variables in the models
contributes significantly to the prediction of output of eggs in Maseru District.
 Elasticities
In the Cobb Douglas production function, the coefficients of the variables represent the powers
of the curvilinear equation as shown in Equation 6. The elasticities with respect to the
independent variables in the models for the four models analysed are shown in Table 3.
Table 3. Elasticities for the Independent Variables in the Curvilinear Models
Coefficients Model 1 Model 2 Model 3 Model 4
α єb = 1.058 єb =1.041 єk =0.69 єk = 0.69
β єf  = 0.09 єt  = 0.06 єb = 1.041
δ єt  = 0.06
ψ єf  = 0.09
α, β, δ, and ψ represent the powers of the first, second, third and fourth variables
respectively in the Cobb Douglas models.
єb = elasticity for number of birds
єk = elasticity for capital
єf  = elasticity for type of feed
єt  = elasticity for type of technology
The results show that in all four models, the number of birds has unitary elasticity. This
indicates that egg production will increase by as much as the number of birds is increased. In
other words, a hundred percent increase in the number of birds will increase the output of
eggs by hundred percent. Put differently, the production function exhibits constant returns to
scale with respect to number of birds. Capital shows a high positive elasticity with respect to
egg output of 0.69. This is an indication that capital is a limiting factor that can substantially
boost egg production if it is sufficiently injected into the industry. Type of feed and technology
have positive but small values of elasticity, which signify that a percentage change in them will
increase the output of eggs by less than proportionate amount. In other words, output of eggs
is highly inelastic to changes in the technology and type of feed.
 Correlation
The correlation matrix for the dependent and independent variables is presented in Table 4.
The matrix shows that egg output is highly and positively correlated with capital and number of
birds so is capital with number of birds. The correlation coefficients of the latter two are also
highly significant at the 1% level. Technology is negatively but significantly correlated with egg
output, capital and number of birds.
Table 4. Correlation Matrix of the Dependent and Independent Variables
       
 Egg output Capital No.of birds Technology Type of  
feed
Egg output 1.00      
Capital 0.898*** 1.00     
No. of birds 0.993*** 0.894*** 1.00    
Technology –0.318** –0.333** –0.349*** 1.00   
Type of feed 0.335** 0.359*** 0.285** –0.047NS 1.00  
NS implies not significant at the 10% level
* implies significant at 10% level
** implies significant at 5% level
*** implies significant at 1% level
 Conclusion
This study has shown that the output of eggs in the poultry industry in Maseru District is not
affected by the marital status and the age of the farmer. The major factors that influence egg
output are found to be, the number of birds, capital, technology in use (deep litter or battery
cage), and the type of feed used. These factors proved highly and statistically significant at the
1% level in predicting output of eggs. The models containing these factors as single or multiple
variables show goodness of fit with high R2 and R2
adj values.
Farm sizes are generally small with one farm having as little as 25 birds and only one farm
has over 1000 birds with an average farm size of 230 birds. Egg output has unitary elasticity
with number of birds; it is highly elastic to changes in capital; and highly inelastic to changes in
technology and type of feed. It is concluded that given the elastic nature of egg output to
changes in capital, the output of eggs in Maseru District can be boosted by introducing
sufficient capital in the poultry industry.
 Recommendations
In the light of the findings of this study, the following recommendations are made:
Efforts should be made to provide capital to poultry farmers to boost egg production, as
capital is a seriously limiting factor in egg production in the country.
Farmers should invest more in the relatively cheaper deep litter system than the more
capital-intensive battery cage technology since there is no statistical significance in their
contribution to egg production in the district.
 References
Anon. 1999. Poultry production section policy objectives and programmes. Livestock
Department, Ministry of Agriculture, Maseru.
GOL. 1986a. The commercial egg industry of Lesotho. 1986. Ministry of Agriculture and
Marketing, Agricultural Marketing Section, Maseru, Lesotho. 31 pp.
GOL. 1986b. Current Thinking of the District Poultry Officers About Issues Relating to the
Poultry Industry. Proceedings of the First Seminar on Egg and Broiler Production and
Marketing Situation in Lesotho. Lesotho Agricultural College. Ministry of Agriculture and
Marketing, Maseru, Lesotho. 10 pp.
GOL. 1987. Lesotho Egg Study. Planning Division, Ministry of Agriculture, Co-operatives and
Marketing, Maseru, Lesotho. 38 pp.
GOL. 1988. Poultry Marketing Bulletin, Ministry of Agriculture, Livestock Department, Maseru,
Lesotho..
GOL. 1991. Poultry Marketing Bulletin, Ministry of Agriculture, Livestock Department, Maseru,
Lesotho.
Hair M.W. 1967. Use of regression equations for projecting trends in demand for paper and
board. Forest Research Report No. 18. US Forest Service, Washington, USA. 178 pp.
Henderson James M. and Richard E Quandt 1980. Microeconomic Theory. A Mathematical
Approach. Third Edition, McGraw-Hill Book Company, International Edition 420 pp.
Mafoso L. 1999. Economic analysis of factors influencing egg production in Maseru Urban.
Dissertation submitted in partial fulfilment for the award of Bachelor of Science in Agriculture.
National University of Lesotho, Roma, Lesotho
Epidemiology of parasite infestation in goats belonging to
resource limited communities in eastern Namibia
F.F. Kumba1, M. Undi1. H. Katjivena2, G. Kauta2, A. Hengua2, and F. Tjiteere2
1Department of Animal Science, Faculty of Agriculture and Natural Resources, University of
Namibia, Private Bag 13301, Windhoek, Namibia.
2Directorate of Research and Training, Ministry of Agriculture, Water and Rural Development,
Private Bag 1437, Otjiwarongo, Namibia.
E-mail: fkumba@mail.unam.na
Abstract
Introduction
Material and methods
Results and discussion
Constraints to goat production.
Epidemiology of parasitic diseases
Coprological analysis
Conclusions
Acknowledgement
References
 Abstract
The epidemiology of parasitic diseases in farm animals with special emphasis on indigenous
goats was undertaken in the communal farming areas of eastern Namibia. Monitoring surveys,
dialogue workshops, farmers' interviews and coprological analysis were the approaches
adopted in the investigation. Findings indicated that environmental conditions on the
communal farms were most favourable for the survival of free-living forms of intestinal
parasites during the summer and autumn months and most adverse during spring and early
summer. Conditions favourable to parasites existed within goat and sheep kraals all year
round due to poor management of these structures. Heavy faecal-egg-counts (FEC) were
observed in goats during all seasons of the year but to a lesser degree during the drier
periods. Major intestinal parasites found in goats were the strongyles and strongyloides.
Increased mortality, attributed by most farmers to famine, was observed among goats in spring
and early summer seasons of feed shortages. The heavy burden of gastrointestinal parasites
seems to augment the effect of famine by decreasing the natural ability of indigenous goats to
survive periods of reduced feed in-take and other stressful conditions. Based on the findings,
a rational management measure recommended by the study was to establish, at the end of
winter, a strategic de-worming programme covering all small ruminants in the communal
grazing area. This strategy, accompanied by preventive management measures such as the
relocation of kraals and the prevention of access to contaminated water points, would appear
a cost-effective intervention to eliminate the negative effect of gastrointestinal parasitism
during the critical seasons of famine.
 Introduction
Indigenous Namibian goats are known to be hardy, highly resistant to endemic diseases and
prosper relatively well under the harsh environment (Thawana and Visser 1999). Being mostly
a semi-arid country, Namibia's climate is generally characterised by frequent droughts that
give rise to serious livestock feed shortages especially in the east. This combination of factors
partly explains why most resource limited communities in the communal grazing area in the
east of the country have, in recent years, tended to reduce cattle numbers, traditionally the
preferred domestic species, in favour of goat production. Census figures of the Directorate of
Veterinary Services (1997) show that over 70% of Namibia's goat population of 1.8 million
animals are raised by small farming households in the communal areas. In a recent study,
Kumba et al. (1999) found that about 95% of households in the eastern communal area
owned goats, 90% owned cattle and 59% owned sheep. Communal farming areas in Namibia
account for 41% of the country's agricultural land and provides a means of livelihood for
47.9% of the country's 1.5 million inhabitants (Ministry of Agriculture Water and Rural
Development 1995; Putz 1996; Rothauge et al. 1999). Commercial farms cover 44% and are
owned by 0.3% of the population. Overstocking, droughts and bush encroachment have
caused severe strains on communal pastures resulting in feed shortages during certain
seasons of the year (Moyo et al. 1996). Farmers have noticed an increased mortality rate,
especially among goats, in these seasons. Katjivena et al. (1998) reported heavy FEC of
parasites in all goats examined in the eastern communal farms. An attempt by these workers
to reduce the parasitic burden in this species, by the administration of anthelmintic remedies,
proved unsuccessful. High FEC were quickly re-established in the treated animals. It was not
clear if high FEC appearing in the treated animals was a result of rapid re-infestation by new
colonies of parasites or if this was due to some hypobiotic forms unaffected by the treatment.
The likelihood of rapid re-infestation seems plausible since poor control measures appear to
promote high infective doses of gastrointestinal parasites in the grazing environment of
communal farms. De-worming can eliminate the premune status of farm animals and render
them more prone to rapid re-infestation if the environment remains contaminated (Radostits et
al. 1994). Therefore, it became important to investigate the epidemiology of intestinal worms in
goats and other farm animals in order to provide a rational basis for strategic planning of cost-
effective de-worming and other preventive management measures.
 Material and methods
Preliminary dialogue workshops (Ison and Ampt 1992; Mariner and Van't Klooster 1994) were
undertaken with farmers in 15 villages all over the communal area in order to establish the
role of parasitic diseases within the perspective of major livestock constraints.
A standard questionnaire was designed and tested (Thrusfied 1995a;1995b) before being
used to collect information about the evolution of local environmental, management and
animal health factors that influence infestation rates of livestock by intestinal parasites and the
onset of clinical manifestations they cause through the year. Forty-five farming households in
18 villages participated in the investigation. Villages were pre-selected and participating
households were randomly chosen. Factors investigated included the following (Radostits et
al. 1994):
Factors related to the micro-climate of the environment: temperatures and rainfall.
Factors related to pasture management and micro-climate of the environment: stocking
densities; duration spent by animals on same pastures; changes in the length and bulk
of pastures; distribution of vegetation on the grazing pastures; extent of water lodging on
pastures; presence of known parasite vectors; methods of watering animals; and the
practice of zero-grazing.
Factors related to the kraal: general hygiene in kraals; state of faecal accumulation in
kraals; kraaling different species of animals; kraaling young animals.
Factors related to the general nutrition and health of animals: scoring body condition of
animals; variations in pasture availability; morbidity and mortality rates.
Timing seasonal movements of livestock on the communal grazing lands.
The above parameters were later monitored by direct monthly observation over a period
lasting one year in order to determine changes in the environment, and management and
health conditions of the animals that promote the survival of free-living forms of intestinal
parasites and predispose animals to infestation and disease. This investigation was
undertaken to validate findings obtained in the dialogue workshops.
Coprological analysis was undertaken during all the months of the year to identify the main
endemic parasitic species in goats and periods of heavy parasitic burden. Six farms located in
different parts of the communal area were identified for this study. Owners of the selected
farms never de-wormed their animals. Ten goats in each farm were selected and marked.
About 10g of faeces was obtained intra-rectally from the selected goat. Faecal samples
collected on each farm were pooled in plastic bags, stored in cold boxes on ice and
dispatched to the laboratory for FEC and parasite egg identification. The lapse of time
between faecal sample collection and laboratory examination was 1 - 5 days. The floatation
McMaster technique was applied for faecal examination (Reinecke 1983). Species
identification was done by microscopic observation of eggs for specific morphological features.
Rainfall and temperature records were obtained from the local meteorological department.
 Results and discussion
 Constraints to goat production
Preliminary dialogue workshops indicated that the main livestock constraint in the eastern
communal areas of Namibia was the scarcity of pasture due to drought. Livestock diseases
were also incriminated, but farmers did not seem to perceive disease as an important
constraint comparable to famine. This is probably because, in Namibia, the existence of a
veterinary cordon fence, which isolates major disease endemic areas in the north of the
country, has, for many years, provided an effective control barrier against major epizootics.
The effect of less dramatic disease conditions, such as parasitosis, appears largely masked by
the predominant role played by famine. Losses in production, caused by parasites in livestock
farms, are usually manifested by non-specific signs that the unsuspecting farmers cannot
easily link to these etiological agents of disease (Radostits et al. 1994). Kumba et al., (1999)
observed that animals from this communal area did not compete favourably in the local market
with those from the neighbouring commercial farms even when pasture conditions were good.
 Epidemiology of parasitic diseases
Findings related to the epidemiology of parasitic diseases showed the following: Rainfall
(Figure 1) was confined to the months of summer and early autumn (January, February and
March) as previously reported (Du Pisani 1999). During this period, the study found that open
ground water points accessible to livestock abound in the grazing areas in ponds, streams,
marshes, dug-outs, ditches and dams. Environmental temperature was quite warm (Figure 1).
However, some rays do not easily reach the soil surface in many places during summer and
autumn because of significant increases in the height and bulk of the vegetation cover. The
above findings show that conditions in summer and autumn are very favourable for the
survival of free-living forms of gastrointestinal worms. Goats and other domestic livestock are
most likely to be infected during this period. In spring, on the other hand, environmental
conditions were largely unfavourable for free-living forms of parasites. Open water points had
dried-up, vegetation cover was sparse, temperatures were high and the sun's rays were very
intense.
It was also observed that all farm animals tended to spend much of the summer and autumn
near villages. The consequent increase in faecal contamination in summer pastures close to
villages appears to enhance the build up of infective doses of parasites in the environment.
The study, however, showed that farm animals are in good nutritional condition during this
period probably because grazing is plentiful. This explains why most animals did not show
clinical conditions during this season. Well-fed animals are more resistant to the negative
effect of parasitosis than debilitated malnourished animals (Radostits et al. 1994).
At the beginning of winter, summer pastures close to villages were mostly run down. Cattle
were seen to graze on the more distant pastures which were rested since they were not
utilised during summer and would consequently be expected to remain much less
contaminated. Goats and sheep still remained in the contaminated environment close to
villages for fear of jackals and other predators. Furthermore, unlike cattle, which spent the
nights on the veld, small ruminants returned to spend nights in kraals. Kraals were rarely
moved and never cleaned. Consequently, faecal accumulation in goat and sheep kraals was
often seen to reach enormous proportions. Although different livestock species were kraaled
separately, young animals often shared the same kraals with adults. During the day, when
adult animals were released into the veld, kids and lambs were retained and usually confined
and fed in kraals. These management practices increased considerably the likelihood for small
ruminants, especially the young animals, to become heavily infested by gastrointestinal
parasites.
Heavy parasitic infestation is probably an important contributing factor that explains why goats
tend to succumb more easily than cattle to the effect of reduced feed intake experienced
during spring and early summer (Figure 2). It was observed that goat mortality was nearly nil
during the summer and autumn months when grazing is abundant (Figure 2). This is a further
indication that when nutritional levels are good, indigenous goats are more resistant to
endemic pathogenic agents and adverse environmental conditions than cattle.
Information collected from the farmers during the interviews showed that increased livestock
mortality was mostly experienced in spring and early summer than at any other time of the
year. This is the season when grazing is most scarce, especially for small ruminants. Most
animals exhibited poor nutritional conditions during this season. It appears that intestinal
parasites that seem to infect animals mainly in summer and autumn, later play a big role in
increased mortality rates experienced among small ruminants during the feed gap period in
spring and early summer. However, environmental conditions observed in grazing pastures,
during this time of the year, were invariably unfavourable for free-living forms of intestinal
worms. Temperatures were very high (Figure1), humidity was low and vegetation cover was
generally absent or minimal- conditions which are generally lethal for these parasitic stages.
However, it would appear that continuous utilisation of the same pasture facilities by infected
small ruminants all year round ensures that these remain contaminated at all times.
The institution of a strategic de-worming programme for all small ruminants at the end of
winter or the beginning of spring would have a dual advantage. On one hand, this would
eliminate the important role played by gastrointestinal parasites in increasing livestock
mortality during the seasons of famine (spring and early summer). On the other hand, the
continuous contamination of pastures by carrier animals would be eliminated and the natural
sterilisation of the pasture environment by harsh weather conditions would be permitted.
 Coprological analysis
Coprological analysis showed high counts of nematode eggs during all the seasons of the
year in all the villages under study (Table 1). However, FEC appearing in summer and autumn
were much higher than at other times of the year. FEC dropped considerably during the drier
periods (winter and spring). The reasons for this phenomenon could not be immediately
established. Some authors observed increased periparturient output of nematode eggs in
infested pregnant ewes (Nuvor et al. 1997). The number of adult Haemonchus contortus
worms in sheep was seen to diminish significantly during the dry season (Mukaratirwa et al.
1997). It is possible that parasite egg output is influenced by changes that affect the
physiological state of the host animal. It has been established that during adverse
environmental conditions, gastrointestinal helminth parasites tend to develop hypobiotic forms,
a largely sexually inactive mechanism to ensure that progeny worms will survive (Urquart et al.
1996; Mukaratirwa et al. 1997).
Table 1. Parasite egg counts in six selected villages in eastern Namibia
Month Okatuhoro Okanjopo. Omatupa Oruvize Otjimati Otjiyere
Jan xx xxxxxxxxxxxxxxx x xxxxxxxx xxxxxxxxxxxxxxx xx
xxxxx o oooo x oo
Feb xxxxxxxxxxxxxx Xxxxxxxxxxxxxxx Xxxxxxxxxxxxxxx Xxxxxxxxxxxxxxx Xxxxxxxxxxxxxxx Xxxxxxxx
ooooooo Xxxxxxxxxxxxxxx Xx Xxxxxxxxxxxxxxx Xxxxxxxxxxx
Xx o Xxxxxxxxxxxxxxx o
m Xx oooooooo
Mar xxxxxxxxxxxxxx Xxxxxxxxxxxxxxx Xxxxxxxxxxxxxxx Xxxxxxxxxxxxxxx Xxxxxxxxxxxxxxx Xxxxxxxxxxxxxxx
ooooooooooo Xxxxxxxxxxxxxxx Xx Xxxx Xxxxxxxxxxxxxxx Xxxxx
Xxxxxxxxxx Xxxxxxxxxxxxxxx ooooooooooooooo
o m Xxxxxxxxxxxxxxx o
Xxxxxxxxxxxx o
Apr xxxxxxxxx o Xxxxxxxxxxxxxxx Xxx Xxxxxxxxxxxxxxx Xxxxxx x
Ooooooooooooooo o Xxxxx O
Ooooooooooooooo ooooooo
Ooooooooooo m
May xxxxxxx Xxxxx T xxxxxxxxxxxx xxxxxxxxxx Xxx oo Xx
Jun xxx xxxxxxxxxxxxxx X Xxxxxxxxxxxxxx o Xxx xxxxxxx
o oo
Jul       
Aug X xx xxxxxxxxxxxxx Xx oooooo xxxxxxxxx xx
Sep  x Xxxx o Xxxxx o Xxxxxxxx o xx
Oct xxxxxxx Xxx o Xxxxxxxxxx o xxxxxxxxx Xxxxxxxxx o xxxxxxxxxxxx
Nov xxxx Xxxxxxxxxxx Xxxxxxxxxxxxxxx Xxxxxxxxxxxxxxx xxxxxxxx x
oooooooooooooo o Xx Xxxxxxxxxxxxxxx
Ooooooo Xxxxxxxxxxxxxxx
Xxxxxx
Ooooooooooooooo
Ooooooooooooooo
Ooooooooooooooo
oooooooooo
Dec x xxxxx Xxxxxxxxxxxxxxx Xxxxxxxxxxxxx Xxxxxxxxxxxxxxx xxxxxxxxxxxxx
Xxxxxxxxxx ooooooooo Xxxxx
Oooo oo
Key:
X = 100 strongyle eggs counted
O = 100 strongyloide eggs counted
m = presence of Monezia eggs
T = presence of Toxocara eggs
The predominant species of worms seen in this study were the strongyles and the
strongyloides. Faecal culture is recommended to establish the exact species of the endemic
parasites. More serious pathogenic species like Haemonchus contortus might well have been
present as it is difficult to distinguish between most of these eggs under the microscope. Few
ascaris eggs of the Toxocara species were seen on one farm during the summer and autumn
months and eggs of the cestode, Moniezia, were also encountered in animals on the same
farm. The above observation shows how the epidemiology of gastrointestinal parasitism can
vary in farms within the same geographical location depending on differences in management
and micro-climatic factors. This is why it is not always correct to prescribe blanket control
protocols across farms even within the same district before carrying out epidemiological
investigations on each of them. No trematode eggs were reported in the study. The habitat in
the east of Namibia is unfavourable for the molluscan intermediate hosts of these parasites.
 Conclusions
Heavy losses among indigenous goats raised on communal farms in the east of Namibia have
been observed during seasons of feed shortages: spring and early summer. Although these
animals are naturally resistant to famine, heavy infestation by gastrointestinal parasites
appears to augment the effect of famine in causing these losses. Heavy parasitic burden
seems to eliminate the natural tolerance of indigenous goats to famine. This study showed
that factors that predispose animals to parasitic infestation in these communal farms included
access to contaminated water, absence of appropriate hygienic measures in kraals, and the
inability to prevent pasture contamination during periods detrimental to free-living forms of
intestinal worms. The months of spring and early summer were found to be the most
unfavourable period for the free-living forms of intestinal worms, but also the time when small
ruminants are most vulnerable to the effects of parasitosis. The main intestinal parasites of
goats seen in the area were the strongyles and strongyloides. As a rational control measure, it
is recommended that an affordable, but potent anthelmintic drug against both adult and
hypobiotic forms of strongyles and strongyloides be identified and used by all communal
farmers in strategic de-worming interventions in small ruminants to be undertaken yearly at the
end of winter or beginning of spring. This would appear a cost-effective intervention to reduce
the negative influence of parasitosis in animals during the feed gap period and to eliminate
continuous contamination of pastures. The intervention would be accompanied by the
relocation of kraals and access to faeces accumulated in the old kraals prevented by burning
or fencing them off. In addition, draining or fencing off contaminated water points within the
farms would prevent access of animals to these sources of infestation.
 Acknowledgement
This study was funded by the University of Namibia Research and Publications Committee
and the Ministry of Agriculture, Water and Rural Development. Coprological analysis was
carried out by Dr. R. Hillbert at the Regional Veterinary Laboratory in Grootfontein. Special
words of appreciation go to the communal farmers of Okakarara district for their co-operation.
 References
Du Pisani A.L. 1999. The Idiosyncrasies of Namibian Rainfall. Proceedings: 8th Congress of
the Agricultural Scientific Society of Namibia, Ehoro Lodge, 2nd – 4th November. Namibia. pp
9–14.
Ison R.L. and Ampt P.R. 1992. Rapid rural appraisal: A participatory problem formulation
method relevant to Australian agriculture. Agricultural Systems 38: 363–386.
Katjivena H., Kauta G., and Hengua A. 1998. The effects of de-worming of goats with both
registered and traditional remedies to improve production and reproduction in eastern
communal areas (Okakarara, Okamatopati, and Epikro). In: Proceedings of the second
national annual agricultural research reporting conference, Neudamm Agricultural College,
Windhoek, 8th–10th September. pp 135.
Kumba F.F., Undi M., Katjivena H., Kauta G., Hengua A. and Tjiteere F. 1999. Taping
indigenous knowledge to time strategic de-worming treatment of livestock on communal
grazing lands in the east of Namibia. Review of Southern Africa Studies: The special issue on
sustainable agriculture in Southern Africa, vol. 4, Number 2.
Mariner J.C. and Van't Klooster G.G.M. 1994. Community-based approaches to veterinary
privatization in pastoral systems. Paper presented at the Pan Africa rinderpest campaign
conference on privatization, 21st November. Nairobi, Kenya. pp 1 –15.
Ministry of Agriculture, Water and Rural Development, 1995. National Agricultural Policy.
White paper presented to parliament. MAWRD, Windhoek, Namibia.
Moyo S., O'keete, P. and Sill M. 1996. The Southern African environment: Profiles of the
SADC countries, Chapter 6: Namibia. Earthscan publications, London, UK.
Mukaratirwa S., Hove T., and Kyvsgaard N. 1997. An abattoir study of seasonal fluctuations of
adults and larval stages of haenonchus contortus in sheep in Zimbabwe. Bulletin of Animal
Health and Production in Africa 45: 165–168.
Nuvor S.V., Assoku R.K.G. and Agyei A.D. 1997. Peri-parturient rise in faecal nematode egg
counts in Djallonkex Nungua Blackhead crosses in the forest zone of Ghana. Bulletin of
Animal Health and Production in Africa 45: 169–171.
Putz J. 1996. The agricultural sector. Business guide to Namibia. An investment center
publication. Ministry of Trade and Industry, Windhoek, Namibia. pp 29–32.
Radostits O.M., Blood D.C. and Gay C.C. 1994. Diseases caused by helminth parasites. In:
Veterinary Medicine: a textbook of diseases of cattle, sheep, pigs, goats and horses, 8th
edition. London, Balliere Tindall. Pp 1223–1230.
Reinecke R.K. 1983. Diagnosis of helminthes in the live animal. In: Veterinary helminthology.
Butterworths Durban/Pretoria pp 66–69.
Rothauge A., Abate A.L., Kavenji G. and Ngbikembua M.L. 1999. What goats eat in the
highland savanna of Namibia in summer. In: Proceedings 8th congress of the Agricultural
Scientific Society of Namibia, Ehoro lodge, Windhoek, 2nd– 4th, Namibia. November, 1999. pp
66–69.
Thawana S.B., and Visser W.H. 1999. Indigenous goats: communal farmers' wealth. Spotlight
on Agriculture No 26. Ministry of Agriculture Water and Rural Development, Windhoek,
Namibia.
Thrusfield M. 1995a. Surveys in Veterinary epidemiology. Blackwell science ltd., Osney Mead,
Oxford pp 178–198.
Thrusfield, M. 1995b. The ecology of diseases. In: Veterinary epidemiology. Blackwell
Scientific Publication, UK. Pp 9 ltd., Osney Mead, Oxford.
Urquart G.M., Armour J., Duncan J.L. Dunn A.M., and Jennings F.W. 1996. Veterinary
helminthology. In: Veterinary parasitology, second edition, Oxford: Blackwell Science.
Session 3: Crisis mitigation in livestock dependant systems
Crisis mitigation in livestock dependant systems: The role and
limitations of current initiatives for livestock systems in South
Africa
Sahr. H. B . Lebbie
International Livestock Research Institute (ILRI)
P. O. Box 30709, Nairobi, Kenya
Abstract
Introduction
The SADC initiatives for drought crisis mitigation
The early warning systems (EWS)
The challenges for addressing livestock issues
Conclusions and recommendations
Acknowledgement
References
Abstract
Livestock are central to the livelihoods of livestock-dependent systems in Africa, particularly
pastoralists and agro-pastoralists. Over the years, however, climatic extremes in this sub-
region have impacted the agricultural production systems including livestock-dependent
systems and hence the livestock owners. In recognition of this problem, the SADC member
states have elaborated a Sub-Regional Action Programme (SRAP) to Combat Desertification
in Southern Africa under the umbrella of the SADC Environment and Land Management
Sector. In addition, the sub-region has, with the assistance of collaborating partners,
established early warning systems (EWS) to assist farmers anticipate and prepare for climatic
extremes such as drought. This paper discusses these initiatives focusing on their role and
limitations to serve the livestock sector and suggests opportunities to enhance their
effectiveness to support and improve the management of livestock and drought.
Keywords: livestock; climatic extremes; drought; early warning systems; ESA
Introduction
Livestock are central to the livelihoods of the livestock-dependent production systems in
Southern Africa, accounting for about 25% of the total agricultural output of the member
states. In particular, livestock are central to the well-being of the pastoralists and the agro-
pastoralists.
Climatic extremes are a critical phenomenon in the pastoral and agro-pastoral communities
largely because of the extreme dependence of these communities on the natural environment
for their livelihoods and the welfare of their animals. In most parts of the arid and semi-arid
areas of southern Africa, a common feature is unreliable rainfall and periodic droughts. This
climatic phenomenon is known to impact biodiversity, human health, the productivity of the
systems and hence the well-being of the livestock owners (Ford 1971; Lewis 1975; Hussein
1976; Swift 1976; Horowitz and Little 1987). For the Southern African Development
Community (SADC) region, drought can no longer be regarded as a catastrophe but rather
accepted as a recurrent feature of the environment. (Pereira 1968)
This and other climatic phenomena have been recognised as crucial to the development of the
sub-region at the highest political levels. In direct response to this concern, the SADC member
states have elaborated a Sub-Regional Action Programme (SRAP) to implement the UN
Convention to Combat Desertification under the umbrella of the SADC Environment and Land
Management Sector. At the same time, a network of early warning systems (EWS) has been
established in the sub-region addressing regional as well as national perspectives.
In order to contribute to these efforts, the SADC Regional Animal Agriculture Research
Network (SAARNet) has also started an initiative to undertake studies on "Crisis Mitigation in
Livestock-Dependent Systems" to understand the risks associated with climatic variability and
to provide policy options, technologies and knowledge products that will make the livestock
sector more resilient in times of drought and thus promote food security, poverty alleviation
and sustainable natural resource conservation.
 The SADC initiatives for drought crisis mitigation
The SADC Sub-Regional Programme to Combat Desertification (SADC-SRAP): Following
the SADC Council of Ministers approval of the "SADC Policy and Strategy for Environment
and Sustainable Development - Towards equity-led growth and Sustainable Development in
Southern Africa", in August 1996, the Council established the distinct Policy and Strategy Co-
ordination Unit called the SADC Environment and Land Management Sector (SADC-ELMS)
under the SADC Committee of Ministers of Environment.
The overall aim of SADC-ELMS is "to achieve sustainable utilisation of the natural resources
and effective protection of the environment" (SADC 2000). To achieve this overall goal, SADC-
ELMS will co-ordinate activities at the level of three programmes, namely the Land
Management Programme, the Environment Management Programme and the Information
Management Programme.
Within this framework, the SADC Council of Minister approved and adopted a Sub-Regional
Action Plan (SRAP) in August 1997 to combat desertification in Southern Africa as the key
operational instrument for implementing the United Nations Convection to Combat
Desertification (UNCCD) in the sub-region. One of the priority programme areas identified and
agreed upon to be implemented by SRAP is the "Strengthening of Early Warning Systems".
SADC-ELMS is evolving and has already undertaken an assessment of institutions in SADC
countries that could assist it in leading the co-ordination and implementation of activities in the
SRAP priority programme areas. Thus far the Desert Research Foundation of Namibia
(DRFN) has been selected and approved by Council as the Lead Institution for SRAP's
Capacity Building and Institutional Strengthening priority programme area. DRFN will, among
other things, establish a network of institutions to build capacity to combat desertification and
mitigate the effects of drought as well as provide leadership in the various aspect of research
on desertification and drought.
 The early warning systems (EWS)
A number of EWS have been established in SADC over the past decade or so, with focus on
the communities that are extremely vulnerable to climatic variability. Following is a brief review
and evaluation of the most successful EWS in SA so far, highlighting their strengths and
weaknesses.
The major objective of most EWS is to rapidly provide decision makers with timely and
accurate information to quickly take informed decisions on food security and environmental
management planning (Walker 1989; Wakesa 1997; Agastiva 1997). Largely, these EWS
essentially collect, process, analyse and disseminate observational data on a number of
interrelated variables and depend very little on predictive models.
The Regional Centre for Services in Surveying, Mapping and Remote Sensing
(RCSSMRS): This is a Food and Agriculture Organisation of the United Nations (FAO) funded
project. The overall objectives of the project are to strengthen EWSs for food security in east
Africa by establishing satellite remote sensing capabilities in the sub-region and preparing and
distributing monthly early warning information to users in the SADC member countries and is
based in Harare. It uses satellite data from NOAA HRPT, Meteosat Artemis and Meteosat
PDUS to generate and produce timely information on the status of agricultural and
environmental resources to assist national and regional institutions concerned with food
security and environmental monitoring. The information bulletins are distributed as both GIS-
based maps and reports. End-users surveys have shown a general satisfaction with the
quality of its products and considerable enthusiasm for EWSs to continue operation in the
region.
Famine Early Warning Systems (FEWS): These are two parallel regional initiatives operated
and financed by the USAID Africa Bureau in both the Inter-governmental Authority for
Development (IGAD) countries of eastern Africa and the SADC countries of southern Africa.
They report on price status of food commodities as well as assess weather and growing
season status. In addition they provide information on crop production outlooks. Based on
these the systems provide quantitative estimates of food aid requirements as well as
confirmed emergency food aid and pledges. They do not provide animal production estimates.
They distribute their products electronically and by post, targeting national and sub-national
level agencies and policy makers.
Drought Monitoring Centres (DMCs): These are jointly operated by the United Nations
Development Programme (UNDP) and the World Meteorological Organisation (WMO) with
one based in Harare covering southern Africa and the other based in Nairobi covering eastern
Africa. These generate quantitative data on the status of the weather and growing season on
a monthly basis using monthly and seasonal rainfall forecast information produced by the
Climate Prediction Centre (CPC) in Washington D.C., USA along with satellite-based
Normalized Difference Vegetation Index (NDVI) and locally collected agro-meteorological data.
They produce monthly bulletins which are distributed in hard copy only to national-level
technicians through courier and normal mail for informing the relevant policy makers.
 The challenges for addressing livestock issues
SADC-ELMS and EWS provide many opportunities for a holistic approach to mitigating the
effects of drought in the region. However, a number of constraints need to be addressed to
enhance the value of these initiatives:
With the exception of a few, the climate predictions or crisis indicators provided by most
EWS are point forecasts and therefore are of limited value for livestock impact forecasts
and long-range climate forecasts. The short-notice nature of these systems could
largely account for intervention failures, especially among the resource poor
stakeholders, who are usually ill prepared to effectively mobilise the necessary
resources to adequately respond to crisis situations.
Information produced by most EWS tends to be oriented to short-range action time
scales, drawing largely on locally available data (observational data) and very little use
of data external to the region (satellite information). The systems are thus oriented to
adverse shocks and hence demand ex-poste rather than ex-ante responses to shocks.
They promote coping strategies oriented to relief operations and handouts rather than
mitigation.
There is a general inability to provide reliable and complete information. The satellite
technologies currently used by the EWS are relatively new and therefore not well
understood and appreciated by ground technical staff. Equally true is that the
involvement of different players employing different methodologies to gather the same
data creates problems for the harmonisation of data to produce integrated and accurate
information and knowledge products for use in decision-making.
Most EWS do not provide the required information rapidly and on time to enable end-
users to make quick decisions to ameliorate prospective risks and exploit emerging
opportunities. This is largely due to their use of courier and normal mails rather than
electronic mail for the dissemination of the information they generate.
 Conclusions and recommendations
There is an urgent need to improve and strengthen the existing mechanisms for rapid and
timely provision of reliable and relevant information for quick decision making. The relevant
institutions in the region must therefore build capacity for use of electronic mail for information
dissemination both for those that require the primary data to generate early information and
those change agents that need to inform the communities to be affected.
There is also an urgent need to enhance the value of the existing EWS by increasing the
capacity for extended range climate forecasting and livestock impact forecasting. This can be
achieved by the use of emerging inter-seasonal and inter-annual climate forecasting
capabilities. There also exist well-established models of animal diseases and growth, range
ecology (for example at ILRI) and crop growth, that could be linked to climate forecasts to
establish climate-driven forecasts for fodder, animal disease epidemiology and animal growth.
For decades now, development resources have been increasingly used up by relief operations
necessitated by unsatisfactory anticipation and/or mitigation of drought. Enhancing the
capacity of EWS to mitigate crises, but more importantly to cue opportunistic development
interventions will break the cycle of dependence on external relief and the diversion of
development or aid resources toward development. Increasing the capacity of EWS for long-
range climatic forecasting will enhance the ability of the end-users of their forecast information
to make timely decision to take advantage of emerging opportunities and or ameliorate
prospective crises rather than depend largely on coping strategies.
As satellite technology is new, there is a need to train (in-service) the ground technical staff to
make them more effective in data collection, processing and interpretation. This will improve
the accuracy and quality of the forecast information.
EWS must ensure that the countries they serve, build the capacity for quick response to risks
shocks as well as opportunities presented by the forecast information. The current lack of
harmonisation and co-ordination of the activities of the EWS is counterproductive. Currently
different EWS use different data sets to produce climatic predictions, thus tending to differ in
the quality and accuracy of these predictors. In Kenya national EWS, for example, several
institutions are involved in collecting the same data but using different methods thus putting in
question the reliability of the results.
There is a need to bring satellite technology to those countries that do not have their own
receiving facilities by the RCSSMR supplying satellite data to those countries.
The human and economic value of the forecast information must be estimated to rationalise
the investment into EWS in the region. This will require regular end-user valuation of the
information provided by the EWS and the level to which the use of the information has
impacted food security.
 Acknowledgement
I wish to acknowledge the Conference Organising Committee for inviting me, and to SACCAR
for sponsoring my participation.
 References
Agatsiva J.L. 1997. Experiences, operational mechanisms and lessons from Early Warning
and Environmental Remote Sensing for Eastern Africa Sub-Region, paper presented to July
1997 ASARECA workshop in Nanyuki, Kenya.
Ford J. 1971. The role of trypanosomiases in African ecology. Oxford, Clarendon Press. UK.
Horowitz M.M. and Little P.D. 1987. African pastoralism and poverty: some implications for
drought and famine. In: Glantz, M.H. Drought and hunger in Africa: denying famine a future,
Cambridge, Cambridge University Press.
Hussein A.M. (ed) 1976. Rehabilitation: drought and famine in Ethiopia. African Environment
Special Report 2, London, International African Institute.
Lewis I.M. (ed) 1975. Abrar: the Somali drought. London, International African Institute. UK.
Pereira H. C. 1968. Land use and misuse. In: M. O. Collins (ed). Proceedings of Symposium
on Drought and Development, First Rhodesian Science Congress, Bulawayo, 16–-20 May,
1967. pp 9–12.
SADC (Southern African Development Community). 2000. SADC programme for environment
and sustainable development: Implementing the UN Convention to Combat Desertification. An
Information Note for the SADC Annual Consultative Conference, Mbabane, Swaziland, 9–24
February 2000 (Unpublished), SADC Environment and Land Management Sector (ELMS).
Swift J. 1976. Desertification and man in the Sahel. In: O'Keefe P., and Wisner B. (eds) Land
use and development. African Environment Special Report 5, London, International African
Institute, pp 171.
Walker P. 1989. Famine early warning systems: victims and destitution, London Earthscan
Publications. England.
Wekesa E. M. N. 1997. Experiences and lessons from the drought preparedness interventions
and recovery programme in the arid lands of Kenya. A paper presented during the
International Livestock Research Institute's Seminar on "Crisis Mitigation in Livestock
Production Systems in the Horn of Africa", held at the Muro Lodge, Nyanyuki, Kenya, 6– 11
July, 1997.
Crisis mitigation in livestock dependent systems: Concern,
universal experiences and challenges in promotion of
livestock production in Dedza District
Richard Mwanza and Jacob Mapemba
Food Security & Sustainable Livelihoods Programme
Concern Universal P.O. Box 217, Dedza, Malawi
Introduction
Crisis indicators
Inadequate extension and training on livestock management:
Scarcity of feeds due to seasonality of forages and high cost of concentrates:
High incidences of diseases and parasites:
Theft and predation:
Lack of cash to buy livestock drugs
Insufficient land for grazing:
Uncontrolled livestock breeding:
Poor livestock marketing
The impact of the crisis on the bio-diversity and farmers
Coping mechanisms deployed by farmers
Concern Universal experiences in mitigating the livestock crisis
Reducing diseases and parasites:
Development of community based livestock breeding centres
Cross breeding
Livestock credit scheme
Improved livestock extension
Local feed formulation and stocking
Livestock marketing
Local capacity building
Challenges in mitigating the livestock crisis
Sustainability
People centred livestock mitigation programme
Empowerment of livestock farmers
Participatory livestock development
Capacity building of livestock farmers and communities
Institutional development
Learning process oriented
Forging linkages and alliances
Policy issues
Conclusion
References
ANNEX: Concern Universal livestock strategies
 Introduction
The importance of livestock in sustaining the livelihoods of the rural communities cannot be
overemphasised. Despite its low contribution to the GDP (7%) and agricultural production
(12%), it is a very important sector in improving the food and livelihood security status of the
smallholder households in rural Malawi. In Dedza District, recent surveys by Concern
Universal have indicated that 64% of the households keep livestock. And 37.8 % of
households keeping livestock view livestock as the asset with the highest value. The Concern
Universal Commissioned Survey by James Banda (1999) revealed that households benefit
from livestock as follows (in order of priority):
Source of income
Source of food
Risk aversion
Source of manure
Social activities (funeral and wedding ceremonies)
Payment of court cases
Transport
The same surveys by Concern Universal in the District have revealed that the livestock sector
in the District is in a crisis situation. This paper, therefore, outlines some of the crisis
indicators, coping mechanisms deployed by farmers and the impact of the crisis on the bio-
diversity and farmers. The paper also shares experiences and challenges of Concern
Universal in mitigating the crisis.
 Crisis indicators
The Concern Universal Commissioned Surveys {Small Scale Dairy Production conducted by
Banda (1999), Factors Affecting Rabbit Production by Mapemba (1999a) and Constraints to
Livestock Production by Mapemba (1999b} reveal several crisis indicators in livestock
production which are discussed below.
 Inadequate extension and training on livestock management
There is increasing evidence that there is very little contact between front line staff and
livestock farmers, there are limited extension packages developed and disseminated to
livestock farmers and there is inadequate training of farmers in livestock management. In
Dedza District, 69.8 % of the households keeping livestock do not get any advice on livestock
management. In a survey on Small Scale Dairy Production commissioned by Concern
Universal and conducted by James Banda revealed that 89 % of the households keeping
goats have never been visited by a Veterinary Assistant. In another Concern Universal survey
on constraints to adoption of agricultural technologies by Jacob Mapemba, 88.1 % of the
households keeping livestock indicated that they have never been visited by a Veterinary
Assistant. In a survey by Jacob Mapemba on Factors Affecting Rabbit Production, 42.3 % of
the households keeping rabbits did not seek any advice from front line agricultural staff
because they did not know the staff responsible for livestock production. Forty-seven percent
of the households who sought advice said they did not receive any advice and 69.2 % of
households did not receive any training on rabbit management.
This state of affairs is evidenced by poor livestock housing, feeding, health, breeding
practices, and the fact that livestock are not multiplying in most of the villages.
 Scarcity of feeds due to seasonality of forages and high cost of
concentrates
This results in low quality of feed. The Concern Universal Survey on Factors Affecting Rabbit
Production indicated that farmers feed their rabbits with cabbages, maize bran, potato vines,
bidens pilosa and some other natural green plants/vegetables of which most are scarce during
dry season. Production of pasture crops and storage of these crops for use in the dry season
is an idea on the shelf in the University and Research Institutions and is foreign to most
farmers. Development organisations promoting livestock production in Malawi are still far from
moving the ideas from the University and research institutions shelves to the farmers' fields.
 High incidences of diseases and parasites
In a Concern Universal Survey on Factors Affecting Rabbit Production, 48.8 % of the
households indicated diseases and parasites as the major cause of deaths in rabbits
(Mapemba 1999a). Although diseases appear to be a major cause of mortality, most
households could not indicate the type of disease. A few households, however, mentioned
running noses and discharges from the eyes (Pasteurellosis) and Coccidiosis as being the
type of disease causing death in rabbits. This further justifies the inadequate training and
extension given to farmers. Red ants and fleas were mentioned as major parasites affecting
rabbit production. This survey revealed that 78 % of households have never given any kind of
treatment to sick rabbits either due to unavailability of drugs or lack of cash to buy drugs and
only 25% of the households vaccinate their chicken against New Castle Disease.
The survey on small-scale dairy production by James Banda in Dedza District revealed that
53.4 % of the households keeping goats lose their goats during the rainy season due to
worms, diarrhoea and skin diseases. The government introduced the Drug bank System
whereby the communities themselves buy drugs and Veterinary Assistants administer the
drugs to the animals. This system, however, is limited by the following factors:
Free range system that favours outbreaks of African Swine Fever and New Castle
diseases
High costs of veterinary drugs
Scarcity of drugs
 Theft and predation
There are increasing high cases of theft and predation of livestock in the communities. The
Concern Universal commissioned surveys in Dedza District have revealed that 61 % of the
households keeping livestock lose their stock through theft and 69 % of the livestock farmers
lose their stock through predation. This is evidenced by the reduced number of large scale
livestock such as cattle. Theft and predation is aggravated by the following factors:
Free range system of keeping livestock
Poor housing structures
Weak leadership at community level to police the livestock security
Poor marketing systems
Open slaughter of livestock
 Lack of cash to buy livestock drugs
Most farmers do not view livestock farming as a business and instead they depend on other
sources of income to feed their livestock and buy drugs for their livestock. The cash crop
enterprise is also crumbling and this makes the situation even more difficult for farmers to
forge ahead with livestock keeping. It is only when farmers take livestock keeping seriously as
a business that farmers can realise adequate income to maintain their livestock.
 Insufficient land for grazing
Farmers are experiencing further reductions in land as more and more marginal land is put
into crop cultivation. Most farmers in Malawi keep livestock as a hobby and as a secondary
means of livelihood. This being the case the first value for land is crop production and not
livestock keeping. Animals are therefore left to graze on very marginal lands and
dambos/wetlands.
 Uncontrolled livestock breeding
The system of rearing livestock in a free-range system contributes to uncontrolled breeding.
The Concern Universal Commissioned Survey conducted by James Banda revealed that 47 %
of the households keeping goats in Dedza District use bucks from other households during
free range grazing.
 Poor livestock marketing
There is very high evidence of limited organised markets in livestock and wide prevalence of
open and uncontrolled markets for livestock and livestock products.
 The impact of the crisis on the bio-diversity and farmers
The major impact of this crisis on bio-diversity is a declining trend in livestock production and
population and increasing food and livelihood insecurity of the rural households. Eighty-five
percent of the households in Dedza district are food insecure form October to March every
year according to Participatory Rural Appraisal Reports by Concern Universal. During the
same period, 36.3 % of the children under-five are malnourished according to Concern
Universal Nutrition Assessment Surveys. In order to maintain the status quo in sustaining their
livelihoods, rural communities resort to other means of getting income.
A Concern Universal recently Commissioned Baseline Survey by Gasheke Simons (1998/99)
has revealed that 11 % of the total households get their income from livestock. Other sources,
according to the same survey, include crop production (53.4%), natural resources (20.6%),
wage employment (7.4%), asset sales (0.01%), land rentals (0.28) and gifts and remittances
(1.52%).
 Coping mechanisms deployed by farmers
Farmers deploy several coping mechanisms in livestock production including:
Feeding animals with inferior feed
Not keeping livestock at all
Using inferior breeding stock
Not vaccinating livestock and not giving drugs to sick livestock
Using traditional drugs
Increased open slaughter of livestock
Killing diseased animals and selling meat of these animals
Overgrazing
 Concern Universal experiences in mitigating the livestock crisis
Concern Universal, recognising the importance of livestock in sustaining the food and
livelihood security of the rural households and recognising the current crisis in livestock
production, developed a three-year (May1999 to April 2002) livestock development strategy
for Dedza District. This strategy was developed in liaison with key stakeholders in the livestock
industry in Malawi.
The following are the key strategic areas to be addressed during this period (see annex for
details):
Strategies for monogastric livestock improvement on smallholder farms.
Strategies for small ruminant improvement in smallholder farms.
Strategies for large ruminant improvement on small holder farms.
In order to implement the livestock development strategy, Concern Universal increased its
budget on livestock in Dedza District and is currently supporting the following initiatives:
 Reducing diseases and parasites
This is being done through mobilising farmers to contribute money towards the costs of New
Castle Vaccine and training of farmers on livestock management.
 Development of community based livestock breeding centres
The task is being done in liaison with Land O Lakes and Bunda College of Agriculture. The
idea is to make available improved livestock breeds in the local communities to meet farmers'
demand in livestock production. This will eventually reduce the problem of scarcity for
breeding stock and hence increase the number of farmers keeping livestock and also increase
the number and species of livestock being kept by farmers. The centres would also supply the
breeding stock beyond the impact area.
 Cross breeding
The idea is to improve the genetic potential of livestock breeds. The CU programme is
supporting cross breeding of local goats with boar goats, land race/large white pigs with local
pigs, and Black Australorp chickens with local chickens.
 Livestock credit scheme
CU is supporting farmers with livestock for multiplication through an in kind livestock credit
scheme - pass a livestock scheme. The project has since July 1999 supported over 800
livestock farmers with various types of small-scale livestock including pigs, goats, rabbits and
chicken. The programme budget allows for an increase in this number to more than four times
in a year but due to a scarcity of breeding stock, the majority of farmers are not being
supported.
 Improved livestock extension
CU is supporting the development of extension materials for good livestock housing, feeding
and disease control. The development of farmer led extension and research in livestock
production in also supported. Farmer field schools in livestock production have been
developed in the impact area and are being supported by CU. Village resource centres as
learning centres and means of disseminating information have been developed by farmers
themselves and are being used by both farmers and extension staff. These allow farmers to
share their experiences and lessons in livestock production. The project is supporting farmer
to farmer learning through farmer exchange visits.
 Local feed formulation and stocking
CU is also supporting farmers in processing locally produced feed for their animals including
feed formulation.. The CU Food Security and Sustainable Livelihoods Programme also
integrates the practice of agro-forestry into livestock production with the aim of increasing
livestock feed availability.
 Livestock marketing
CU is supporting the development of livestock clubs and associations to improve marketing of
livestock and livestock products. These associations will also give farmers a voice in issues
affecting their livestock marketing.
 Local capacity building
CU is strengthening community capacity in managing livestock development issues. The
development and training of livestock committees is the key focus in this process. This also
reduces theft of livestock in the communities.
 Challenges in mitigating the livestock crisis
In order to mitigate the crisis in the livestock industry, the following challenges need to be
addressed among many others:
 Sustainability
There is a need to put in place a livestock mitigation programme that meets the needs of the
present without compromising the ability of future generations to meet their own food and
livelihood security needs. The programme should take sustainability as a process of change in
which the exploitation of resources, the direction of investments, the orientation of
technological development and institutional change are all in harmony and enhance both
current and future potential to meet farmers needs and aspirations in livestock production. The
basic understanding is that the programme can be sustainable if it is farmer/people centred.
The philosophy is to put in place a livestock mitigation programme that that will be viable and
continuously revolving from generation to generation.
 People centred livestock mitigation programme
The livestock mitigation programme that focuses on the needs and cultural practices of the
livestock farmers is what is desired. This entails a programme that recognises the farmer's
systems of production and builds on it for future development. The principle focus is to
develop a livestock mitigation programme approach that is accepted by rural people.
 Empowerment of livestock farmers
The challenge is to develop a livestock mitigation programme that will focus on the creation of
awareness and stress solidarity and mutual responsibility within local communities and groups
of farmers. Thus the programme should be concerned with individuals gaining strength,
confidence and vision to work for positive changes in their livestock crisis situation. This is a
measure of people's capacity to bring about change in the livestock industry.
The thinking is that livestock farmers must be able to have a critical understanding of their own
livestock production situation and they must organise themselves to undertake their livestock
development. The idea is to build confidence in them in undertaking their livestock
development process with a united vision.
 Participatory livestock development
The challenge is to work with a practice and approach which ensures that those who bear the
risks and costs of livestock development retain authority in its inception, planning,
implementation and evaluation. Empowerment will be demonstrated by the quality of farmers'
participation in the decisions and processes affecting their livestock production. The
philosophy is that farmers will determine the nature and extent of participation in the livestock
industry that they themselves require. Livestock mitigation will have little effect or may not
even reach its most important objectives, unless farmers participate. Participation will enhance
the efficiency, effectiveness and sustainability of the livestock development process. The use
of participatory methods such as Participatory Rural Appraisal, Participatory Learning And
Action, Learning For Transformation, Theatre for Development need to be encouraged in the
mitigation process.
 Capacity building of livestock farmers and communities
Farmers have their own ability to work and the skills and knowledge with which to produce.
Communities have social/organisational capacities such as leadership and decision-making
systems, and clan/family loyalties. They also have attitudinal/motivational capacities and
shared beliefs.
 Institutional development
A livestock mitigation programme that supports farmers' capacity to create appropriate
structures through which to participate effectively in the decisions that affect their livestock
industry need to considered. Institutional strengthening is a key factor in partnership and self-
reliance. The improvement in livestock development requires competent institutions to turn
labour, land, capacity and technology into on-going improvements in farmers' lives. The
philosophy is that strong organisations at all social levels are critical for the cost-effective
transformation of rights into outputs; on - going participation of stakeholders; the mobilisation
and regulation of local resources; the resolution and management of conflicts; effective control
in the division of benefits; the monitoring, evaluation and validation of externally suggested
change; and the transformation of livestock policies into practice.
A programme that supports local institutions that promote the rights of the farmers in a way
that is accountable to them needs to be the main focus. The basic idea is that insufficient and
weak institutions are a critical bottleneck to sustainable and more equitable livestock
development in Southern Africa.
 Learning process oriented
The livestock mitigation programme should be able to learn from its own successes and errors
in development. It should be able to embrace the error, not to externalise it. The challenge is
to create and sustain mechanisms for learning and sharing lessons. International workshops
like these are good and need to continue. There is however more need to have these
workshops at all levels - local community, district, regional and national levels. More farmers
should be allowed to participate in such workshops in future than at present.
There is a need in integrate farmer-led research and participatory research in the process of
learning.
 Forging linkages and alliances
The challenge in all this work of mitigating the crisis is to forge linkages and alliances with key
stakeholders in livestock production.
 Policy issues
While the government is creating a conducive policy environment in livestock development,
there is a need to further examine some of the policies and practically implement such
policies. Livestock marketing and slaughter of livestock are very critical issues to address in
Malawi.
 Conclusion
The importance of livestock in sustaining the livelihoods of rural communities need not be
overemphasised. The livestock industry is however in a crisis situation as evidenced by the
following indicators: in-adequate extension and training, lack of feeds, high incidences of
diseases and parasites, theft and predators, lack of cash to buy livestock drugs, insufficient
land for grazing, uncontrolled livestock breeding, and poor livestock marketing.
The major impact of this crisis on bio-diversity is a declining trend in livestock production and
population and increasing food and livelihood insecurity of the rural households. In order to
maintain the status quo in sustaining their livelihoods, rural communities resort to other means
of getting income.
Concern Universal, recognising the importance of livestock in sustaining the food and
livelihood security of the rural households, and recognising the current crisis in livestock
production, developed a three - year (May1999 to April 2002) livestock development strategy
for Dedza District. This strategy was developed in liaison with key stakeholders in the livestock
industry in Malawi.
In order to implement the livestock development strategy, Concern Universal is currently
supporting the following initiatives: reducing diseases and parasites, development of
community based livestock breeding centres, cross breeding, livestock credit scheme,
improved livestock extension, local feed formulation and stocking, livestock marketing, local
capacity building.
In order to mitigate the crisis in the livestock industry, the following challenges need to be
addressed among many others: sustainability, people centred livestock mitigation programme,
empowerment of livestock farmers, participatory livestock development, capacity building of
livestock farmers and communities, institutional development, learning process oriented,
forging linkages and alliances and policy issues.
 References
Banda J.W. 2000. Small Scale Dairy Production In Dedza District: A Survey commissioned by
Concern Universal in Dedza West. Dedza, Malawi. polycopy
Banda J.W. 1999. Livestock production in Malawi: Constraints and Options for Improvement in
Productivity of Small Ruminants. Dedza, Malawi. polycopy
Gasheke S. 2000. Baseline Survey of Household Income, Assets and Farm Production,
Dedza Sustainable Livelihood Project, Concern Universal. Dedza, Malawi. polycopy
Mapemba J. 1999a. Factors Affecting Rabbit Production and Impact On Smallholder farmers:
A Commissioned Concern Universal Case Study of Rabbit Beneficiaries. Concern Universal,
Dedza, Malawi. polycopy
Mapemba J. 1999b. Constraints To Livestock Production In Dedza District: A Concern
Universal Commissioned Case Study in Dedza West. Concern Universal, Dedza, Malawi.
polycopy
 ANNEX: Concern Universal livestock strategies
Table 1: Strategies for monogastric livestock improvement on smallholder farms
POULTRY
Diseases and parasites:
1. New castle disease/ worms
Vaccination - Farmers should be in groups so that vaccination can be carried out easily.
Farmers should be made aware of the services.
Use of Ethno veterinary medicine should be encouraged e.g. Chitedze Roots and Dema.
2. Fleas and mites
Regular cleaning of the khola.
Disinfection of the khola.
Paraffin application on affected body parts of livestock.
Assessment of the existing drug supply system and its effectiveness.
3. Poor housing
Promote standard low cost and strong khola.
4. Poor feeding
Supplementation with madeya.
Provide clean water in troughs.
Release livestock to feed in the morning in the case of a free-range system.
5. Lack of extension service
Strengthen extension in poultry through provision of training to field assistants in poultry
production techniques.
Veterinary assistants should be trained in poultry production and extension.
Develop poultry extension packages.
Encourage farmer participation in the poultry improvement programme.
6. Low genetic potential
Promote selection within breed
Promotion of cross breeding of strains that are prolific.
7. Theft and predation
Promote good and strong housing.
Facilitate the formation of strong livestock development committees.
Use of ownership certificates
Confine the movement of brooding hens.
COMMERCIAL POULTRY
1. Poor quality/ high cost and scarcity of feed
Encourage farmers to use available or alternative materials for feed formulation.
Train farmers / frontline staff in feed formulation.
Promote bulk purchasing of essential/ expensive ingredients.
Promote formation of clubs/ association.
2. Source of day old chicks
Encourage the establishment of local hatcheries where-by farmers can come up with business
plans for hatcheries for funding.
3. Incidence of diseases (NCD, Gumboro and Coccidiosis) and parasites
Follow vaccination programme.
Drugs should be accessible to farmers.
4. Poor management
Proper advice on - feeding
- housing
- Vaccination
- Cleaning and disinfection of kholas
General comments on poultry
There is a need to promote other species of poultry like guinea fowl, ducks, pigeons, turkeys.
Research in other poultry species should be encouraged.
PIGS
1. Diseases (ASF) and parasites (mites, worms, fleas) could be controlled through:
Confinement of pigs
De worming
Pig cleaning and disinfection.
2. Poor management
Confinement
Promote good housing
Promote on farm feed formulation
Promote extension and research
RABBITS
1. Diseases
Promote good housing
Use of drugs
Regular cleaning of houses
2. Feeding
Proper feeding programme.
Promote supplementation of concentrates.
3. Poor housing
Promote good housing.
4. Lack of extension services
Strengthen extension in rabbit production
Train front line staff and farmers in rabbit production techniques.
5. Marketing
Promote consumer awareness and preparation of rabbit meat.
Table 2: Strategies for small ruminants improvement on smallholder farms.
STRATEGIES FOR IMPROVING SMALL RUMINANT PRODUCTION
1. Diseases and Parasites
Design extension messages on proper feeding and housing and prevention, control of diseases
and parasites.
Promote use of ethno veterinary drugs e.g. T. Vogelli (Jerejere).
Encourage and strengthen the use of drug boxes at dip tanks.
Improve accessibility of drugs to farmers.
Conduct a diagnostic survey on prevalence of diseases and parasites and their causes to
establish an annual trend.
Document the required/ essential drugs.
2. Theft
Form and strengthen livestock development committees to assist in security during sales and
purchases of Livestock.
Facilitate exchange visits to successful livestock development committees.
Encourage construction of secure houses.
3. Grazing and feeding
Encourage herding and appropriate tethering.
Encourage integration of Agro-forestry with livestock production.
Encourage the establishment of fodder banks/ pastures.
Encourage conservation and utilisation of crop residues.
Liaise with research Institutes on available alternatives on feed technologies.
4. Destructive behaviour/ and labour
See feeding above ( No.3).
Awareness meetings on the behaviour of the small ruminant vs. husbandry practices
Use of live fences e.g. sisal.
5. Lack of technical knowledge and poor management
Conduct training for frontline staff and farmers on livestock management.
Produce and disseminate extension messages.
Organise field tours to successful livestock projects.
6. Housing
Training of farmers/ staff on proper construction of recommended low cost housing.
Organise field tours to organisations and farmers with recommended housing.
7. Lack of improved breeds
Encourage within breed selection among the local small ruminants.
Train farmers/ staff on selection within breed.
Liaise with organisations with improved breeds on supply and multiplication.
Initiate and support stud breeders and multiplier farmers.
8. Research and extension
Conduct a situational analysis on existing small ruminant systems.
Identify research needs.
Assess and evaluate livestock technologies and identify research gaps.
Facilitate formation of livestock farmers groups for transmission of extension messages.
Strengthen research and farmer linkages through demonstrations, meetings, field days etc.
9. Marketing systems
Encourage formation of associations.
Establish linkages with marketing organisations.
10. Input supply
Drugs ( see solution 1).
Group formation for credit.
Table 3:Strategies for large ruminant improvement on smallholder farms
BEEF
Form very strong livestock committees. They should sensitise people to stop slaughtering young
animals which are future stock.
Practice village level selection and multiplication within the available stock.
Evaluation and monitoring of production costs.
Encourage farmers to participate in price negotiation through organisation of auction.
Community leaders should participate in curbing theft e.g. Chikwawa livestock committees and
Lilongwe west RDP- Khongoni.
Sensitise the farmers on the benefits of raising beef animals through meetings.
Encourage strategic de-worming and vaccination of beef animals against ECF.
Improve nutritional status through good feeding.
Improve on kraals through extension.
Introduce tick control measures through application of acaricides, tick grease, T. Vogelli,
Mpanjovu.
Staff and farmer training
Develop extension leaflets, manuals to strengthen research, extension and farmer linkages.
Conduct PRA on beef production.
Encourage farmers to buy drugs in bulk.
Isolate good examples of indigenous medicines for research.
Conduct awareness meetings on utilisation of fodder, legumes and crop residues.
Conduct on farm Demonstration on the above.
Introduce Brahman and high grade bulls.
DAIRY
Village level selection and multiplication of dairy breeds.
Introduce other types of breeds e.g. Jersey , Gurnesy, Ayreshire, Sahiwal.
Proper feeding in relation to breeding.
Improve management in feeding: demonstration in feed conservation, legume utilisation, agro-
forestry ( with close supervision).
Housing- demonstration and extension materials.
Disease control strategies- vaccination, dipping and tick control.
Calf rearing: train farmers. Develop extension packages, on farm demonstrations.
Train farmers in feed formulation.
Encourage farmers to buy premixes in bulk.
Strengthen research extension and farmer linkages.
Evaluation and monitoring of production costs.
Form co-operatives to negotiate prices.
Conduct farmer training at village level.
Increase GoM and NGO initial support i.e. Training and financing.
Train farmers in AI and heat detection.
Train more breeders.
The presentations by different experts resulted in a wide range of discussion through the
working groups. Table 5 below gives a summary of the constraints limiting livestock production
in Dedza District and the possible solutions to the constraints.
Table 4: Constraints and solutions to improving livestock production
 Problems Solutions
Monogastrics Husbandry and management Exploration into
alternative feed
Diseases and parasites resources
Low genetic base
lack of extension services Promotion of on-farm
lack of participatory feed formulation
programme Training in on-farm feed
formulation and mixing
Bulk acquisition of some
ingredients
Dairy production Shortage of dairy cows to Encourage production of
potential farmers small diaries
Poor grade of existing dairy Encourage rural
cows processing
Shortage/unavailability of Introducing bulls and AI
fodder for sale Training of farmers in AI
High feed prices from Marketing/potential all
commercial factories animals have to be TB
Low milk prices tested and if free of TB
Lack of incentives to extension the milk can be
staff consumed.
High farmer to staff ratio (200 Select stud breeders to
farmers) due to big areas produce cows for other
being covered farmers
Erratic training programmes
Difficult financing terms (to
beginner)
Livestock (in general) High cost of feed Encourage on-farm
compounding of feed.
Diseases and parasites
Low prices of products Improvement on
Lack of link between research housing, management
and extension and nutrition.
Government policy on Encourage formation of
marketing associations which
Lack of good genetic base for could negotiate for
improvement better prices.
Lack of extension packages for Strengthen research -
certain species of livestock extension-farmer
Limited contact between linkages through a
livestock farmers and participatory approach.
extension workers Support and strengthen
Theft already existing rural
marketing systems.
Session 4: Product enhancement and human resources
development
Promotion of higher education: Is there a role for the
traditional agricultural university in development?
Clemens B A Wollny
SACCAR-GTZ Project, Bunda College of Agriculture, P.O. Box 219, Lilongwe, Malawi
Abstract
Introduction
Situation analyses
The potential of the agricultural and natural science universities
The role of information technology and distance learning
Conclusions
References
 Abstract
This paper analyses the status of higher education in Southern African Development
Community (SADC) with a focus on agricultural universities and recommends innovative
strategies and actions. The central role of higher education for development has been recently
re-emphasised at various international conferences and in reports. Institutions of higher
education should serve as a 'gate to global knowledge' - the key factor for development.
Knowledge becomes the critical factor in the process of globalisation. The level of higher
education has an immediate impact on economic competitiveness and democratic
development of the society. The efficiency and effectiveness of universities are still insufficient
to meet the national and regional demand for adequately trained personnel in SADC. In
agriculture, the backbone of economy in SADC countries, and in the livestock sub-sector the
situation is alarming. Land degradation, food insecurity and erosion of plant and animal
genetic resources are some of the major challenges to be faced. Universities, therefore,
should become innovators and key players in the sustainable development of the region.
Agricultural universities have the tools and experiences to provide leadership in the following
areas:
Education sector - through training of teachers in agriculture and managing, reform of
the educational sector through revision of curricula, and development of relevant
outreach and distance learning programmes
Poverty reduction strategy - through providing access to and transferring knowledge on
technical, economical and political methodologies relevant to rural development and
food production. This requires an integration of the institution itself into development
programmes within and across the agricultural sector.
Utilisation and conservation of plant and animal genetic resources - through
development of strategies, concepts and action programmes. Collaboration with the
CGIAR system would ensure state-of-the-art application of methodologies.
At present, eight universities host regional postgraduate programmes in agriculture for SADC
and they have the experiences and tools to develop interdisciplinary training and research
programmes of variable duration, to create regional technology transfer centres and to
improve networking within and between regions and build formal linkages (South-South).
Effective quality control of training and research programmes and regular curriculum reviews,
however, are considered critical. The professional implementation and reporting of projects is
essential to build up a good reputation. Cross-sectoral development programmes financed by
the international donor community offer opportunities to agricultural universities to integrate
their knowledge and capacity. The development of outreach and information technology based
distance learning programmes could improve the efficiency of institutions. In conclusion, the
institutions of higher education in agriculture should become more active and should play a
leadership role in the acquisition and implementation of programmes which are relevant to the
national and regional development objectives.
 Introduction
The central role of higher education for development has been recently re-emphasised at
various international conferences and in reports (UNESCO 2000; WORLD BANK 2000).
Institutions of higher education should serve as a `gate to global knowledge' - the key factor
for development. Knowledge becomes the critical factor in the process of globalisation. The
level of higher education has an immediate impact on economic competitiveness and
democratic development of the society (Benell 1999). The availability of high quality
agricultural and natural resources management experts with skills in policy development, and
a strong technical background and management skills is essential to achieve progress in
SADC countries, where the agricultural sector still dominates the standard of living of the
majority of the population. The major social, political and economic forces that are influencing
the development process within and across the agricultural sector are still poorly understood.
Graduate and post-graduate training is therefore critically important in improving the
knowledge base and to cope with the influx of new information and technologies from outside
the region. The design and implementation of adequate agricultural policies, the development
of entrepreneurship, the development of the private sector in several SADC countries, the
empowerment of traditional communities, and the co-ordination of donor-financed activities are
professional challenges that require highly trained and skilled personnel. This paper analyses
the status of higher education in SADC with a focus on agricultural universities and
recommends innovative strategies. In this paper the understanding of an 'agricultural
university' may include a faculty of natural science or even veterinary medicine.
 Situation analyses
Table 1 presents indicators of higher education in SADC. Major progress has been made in
primary and secondary gross enrolment of students over the last thirty years. Higher
education, which includes college and university based studies, has become significant but it
is still well below world average for SADC. The actual numbers of students and the number of
students per 100,000 inhabitants show clearly low minute availability of trained professionals
compared to the world average, not to mention western countries. Furthermore, such
quantitative figures should be carefully interpreted, as they do not imply a standardised level of
quality, as most university lectures in SADC know from their own experience. There might be
flagship universities or programmes or colleges, which produce poorly qualified students for a
number of reasons. The lowest figures of tertiary students are found for Mozambique,
Tanzania, Malawi and Angola. From the latter the 71 students per 100,000 inhabitants
corresponds to the year 1990, indicating the instability in the country. The proportion of
science based students, which include students in natural sciences, engineering and
agriculture but excludes students of the medical sciences, ranges from 4 to 46 % for actually
enrolled students. In most countries this may indicate the capacity of the existing universities
rather than the need or the original interest of the student. Many systems are still inflexible
and the undergraduate student may not be in a position to choose freely his or her field of
study (World Bank 2000). Presumably a considerable number of students pursue their post-
graduate training outside of SADC but statistics are not available for any of the countries of the
SADC region. It is understandable that expenditures of 15 (South Africa) to more than 20 %
(Malawi, Zambia) of the public currently spending money on higher education as a percentage
of total public current spending on education in 1995 leaves little room for an increase of the
budget for universities and colleges. The relative extraordinary high costs of one tertiary
student become even more impressive when expressed as expenditure per student as a
percent of GNP per capita. It ranges from 59 (South Africa) to 979 % (Malawi) for SADC
countries, where data are available (World Bank 2000). Compared with the world average of
26 % of high-income countries, it is obvious, that national efforts are tremendous just to train
one professional. The economic figures do not encourage donors to invest in university
education as a priority giving the burning issues of food security and the HIV/AIDS problem.
On the other hand, it must be argued that the agricultural sector remains the economic
backbone of the region and land degradation, erosion of plant and animal genetic resources,
food insecurity are major prevailing challenges in SADC. The efficiency and effectiveness of
universities appears to be still insufficient to meet the national and regional demand for
adequately trained professionals. The supply of computers and the access to the Internet
(Table 2) demonstrates the wide range within and between the region and the rest of the
world. The widely used Human Development Index (HDI) does not indicate the extreme gaps
in information and communication technology (Table 2).
As early as 1988 the situation prompted the German Government to support through its
implementing technical co-operation agency GTZ (Deutsche Gesellschaft fuer Technische
Zusammenarbeit) post-graduate training in agriculture in SADC. The overall goal of the project
is to improve postgraduate training in agriculture in SADC and to contribute through capacity
building to improvement of food security.
The immediate project purpose is to establish competitive MSc programmes in agriculture
relevant to SADC. The project aims follow:
To establish effective management of the MSc programmes, and to provide well trained
and motivated staff
To offer adequately structured courses with relevant content
To implement departmental research strategies
To improve the quality and facilities of teaching and research
To strengthen regional and international networking.
The Southern African Centre for Co-operation in Agricultural Research and Training
(SACCAR) in Gabarone is the executing agency. The German contribution targets agricultural
faculties of the universities of Malawi, Tanzania, Zambia and Zimbabwe. The support package
strengthens post-graduate capacity through provision of regional MSc scholarships, expatriate
assistance, supply of equipment, network support, link programmes and regional and
overseas PhD scholarships. Starting from almost non-existent capacity, the currently offered
annual intake capacity of about ten postgraduate students per faculty is impressive but still low
compared to the situation outside of SADC. In marked contrast, academic institutions in
agriculture offering courses for tropical and developing countries in Europe have a large and
growing post-graduate capacity and compete on the international market. Looking at several
institutions the income might be of higher priority than the relevance or the needs of the
programmes. Courses range from short-term residential to long-term distance learning courses
covering all kind of subjects, options and degrees. Due to the high costs and the recognised
need to train the student on relevant topics within the region the number of donor-sponsored
participants from SADC is not increasing and in some disciplines it is even sharply decreasing
(Benell 1999). SADC based universities should realise that they compete with universities in
the region and internationally for the best students and international recognition. The need to
adjust and to change to being more competitive is obvious.
 The potential of the agricultural and natural science universities
Agricultural and natural science universities have the potential to provide more and better
training utilising their tools and experiences. The SACCAR directory of agricultural and
veterinary medicine faculties lists the various institutions in the region (SACCAR 199?). The
challenge is to play a more active role and aim to gain leadership in the field of expertise. This
sounds far from reality under the given situation of many faculties where most undergraduate
students cannot even afford a textbook and only "those students, who regurgitate a credible
portion of their notes from memory achieve exam success" (World Bank 2000). Independent
and flexible thinking and learning by problem solving is often not the standard in under-
graduate and even post-graduate classes of today. It is obvious that the improvement in the
quality of teaching and research among other things deserve high priority and must go hand in
hand with additional engagement. There is an urgent need to expand the role of the
universities if they wish to play a significant role in the development of their country and the
region in the future. Today's traditional universities have the potential to contribute to
innovation and sustainable development in the region at a much higher level of impact.
Agricultural universities, for example, could achieve impact through pursuing various strategic
options such as the following:
In the education sector through training of teachers and managers in agriculture, reform
of the educational sector through revision of curricula and development of relevant
outreach and distance learning programmes.
Focus on poverty reduction programmes through providing access to and transferring
knowledge on technical, economic and political methodologies relevant to rural
development and food production. This requires an integration of the institution itself into
development programmes within and across the agricultural sector.
Utilisation and conservation of plant and animal genetic resources through development
of concepts, management and action programmes. Collaboration with the CGIAR
system would ensure state-of-the-art application of methodologies.
At present, eight universities host regional postgraduate programmes in agriculture for SADC
and they have the experiences and tools to develop interdisciplinary training and research
programmes of variable duration, to create regional technology transfer centres and to
improve networking within and between regions and build formal linkages (South-South).
Effective quality control of training and research programmes and regular curriculum reviews,
however, is considered as critical to sustainability.
At the regional level universities may submit proposals for regional training programmes to
SACCAR, which need to fulfil the following criteria (SACCAR 1993):
Relevance and importance of the subject to SADC region
The host institution must be able and willing to host and support the programme
The host institution must have adequate staff and facilities and should be a recognised
leader in their field
A record of performance in training and research must be available and continuously
updated
The host institution should be in a position to attract financial and political support at
various levels
Such criteria are presently under review. It has been recognised that the quality aspect must
be more emphasised and a quality assurance system needs to be installed. A standardised
accreditation system for the host universities and reporting system needs to be introduced as
well.
The professional implementation and reporting of projects is essential to build up a good
reputation towards the scientific community and also towards sponsors. Cross-sectoral
development programmes financed by the international donor community offer opportunities to
agricultural universities to integrate their knowledge and capacity. For example, the key
elements of German technical and financial co-operation, which are not particular different
from other international donors, are poverty reduction, equal opportunity, conservation of
natural resources and crisis prevention. It is obvious that agricultural universities could be
more involved in planning and execution of development projects and programmes. The
intensified interaction with the prevailing problems would have a positive effect on the
relevance of teaching and research for both staff and students. In addition, the institutional
ability to develop outreach programmes based on on-farm research could become attractive to
stakeholders.
 The role of information technology and distance learning
With proper support, adequate teaching material and effective networking a student can most
likely learn better at home than under crowded college conditions. The provision of higher
education will be affected by the on-going revolution of the information and communication
technology sector. It may happen, that conventional campus-based universities will soon
become outdated and obsolete. Overseas institutions can offer a great choice of interesting
and cost efficient courses and local universities may become simply agents of franchised
programmes or even left out if they resist change. Currently, the University of South Africa
offers distance education programmes and is one of the world's largest providers of distance
learning courses. In addition, private institutions offer numerous courses in different subjects.
The University of Zimbabwe is developing a distance-learning programme in agricultural
sciences. To the author's knowledge, no other university of the region offers a distance-
learning programme in agriculture or natural sciences at the undergraduate level. There is
considerable scope to replace the currently time and capacity consuming, inefficient
undergraduate teaching through IT based interactive technologies. The standard of teaching
needs to be improved in terms of quality and quantity (World Bank 2000). Learning courses for
distance programmes are easier to validate - they are simply `on file'. An update is no problem
from a technical point of view. However, the initial investment is high and the reliability of the
technology remains to be tested. A more efficient undergraduate programme may result in
improvement of research-based post-graduate programmes, which require the student's
presence on campus and in the field.
As a pilot project the Animal Science Department of the University of Malawi and the
International Livestock Research Institute (ILRI) are presently developing, with the support of
the SACCAR-GTZ project, a virtual library on CD-ROM on animal genetic resources and
related fields. This library contains a search function and serves as a tool for researchers,
trainers, managers, and policy makers and for under-graduate and postgraduate training. At a
later stage it will be available through the Internet and will have links to other web pages. It
provides opportunities for self-learning but students still need guidance and supervision. The
Swedish University of Agriculture initiated a training resources project on animal breeding,
which is currently under development by ILRI (Mpofu et al. 2000).
 Conclusions
The institutions of higher education in agriculture should become more active and should play
a leadership role in the acquisition and implementation of programmes, which are relevant to
the national and regional development objectives. The colleges and universities should leave
their `institutional ivory tower' and get actively involved in, for example, the education sector,
the national poverty reduction programmes and the management and utilisation of indigenous
genetic resources. The development of attractive training materials produced on CD could be
a first step to disseminate knowledge in a cost-efficient way. There is considerable scope for
IT based distance-learning programmes for undergraduate courses. The improvement in the
efficiency and efficacy of the existing higher education system could be significant through the
use of modern communication technology. The use of IT based training resources, the
expanding market for distance learning programmes and, finally, the virtual university
developments, have the potential to provide knowledge to all and offer opportunities for
lifelong learning.
 References
Bennell P. (1999) Strengthening post-graduate training capacity in development policy
analysis and management in Sub-Saharan Africa. Institute of Development Studies, University
of Sussex, Report prepared for the Department for International Development (DFID), UK.
Mpofu et al., 2000 i.e. Paper presented at the same conference.
SACCAR (Southern African Centre for Co-operation in Agricultural Research and Training).
1993. Procedures and criteria for the selection of SADC regional training programmes and
host institutions. SACCAR, Gaborone, Botswana
SACCAR (Southern African Centre for Co-operation in Agricultural Research and Training).
1997. Directory of agricultural and veterinary medicine faculties in SADC. SACCAR,
Gaborone, Botswana
UNESCO .2000. World education report 2000. The right to education: Towards education for
all throughout life. UNESCO Publishing, Paris, France.
World Bank. 2000. Higher education in developing countries. Peril and Promise. The Task
Force on Higher Education and Society (Ed.). World Bank. Washington D.C., USA.
The framework for regional collaboration in animal agriculture:
The SADC animal agriculture network (SAARNET)
Sahr. H. B . Lebbie
International Livestock Research Institute (ILRI)
P. O. Box 30709, Nairobi, Kenya
Abstract
Introduction
The SADC animal agriculture research network (S - AARNET)
The evolution
Vision
Mission
Objectives
The institutional framework of the network
Regional dimension of the network
Structure of the network
Stakeholders
Governance
Office and technical committees
THE S-AARNET livestock research and development strategy
Problems to be addressed
Strategy formulation
Guiding principles
Species and commodity output
Agro-ecological zones and production systems
Research and research related priorities
Nature of research
The beneficiaries
Considerations driving the research and development agenda
Research priorities formulation process
Research agenda goal and objectives
Goal
Objective
The research agenda and related activities
Genetic improvement, conservation and utilisation
Research activities
Expected products
Nutrition and feed resources development
Research activities
Expected products
Sustainable integrated animal health management
Research activities
Expected products
Natural resource management
Research activities
Expected products
Systems, policy and socio-economic analysis
Research activities
Expected products
Crisis mitigation in livestock dependent systems
Research activities
Expected products
Human resource development (capacity building)
Facilitating training for MSc and PhD in the following areas of specialisation
Short courses/workshops through collaboration with ILRI and other relevant organisations in
the following areas
Training of trainers
Mode of operation and implementation
Development and approval of research and research related activities
Management of activities
Collaborators
Monitoring and evaluation
Guidelines and format for developing project proposals
Guidelines
Format
Conclusion
References
 Abstract
SACCAR's strategy emphasises that dynamic, productive and responsive agricultural
research systems are essential to realising the food security potential of the Southern African
Development Community (SADC) region and the continuous development of technologies for
different sectors of agriculture will depend on a co-ordinated effort by countries of the SADC
region. It is against this backdrop that SACCAR has adopted networking as a strategy to
achieve strong co-ordination and collaboration among SADC member states for regional
development and advancement in agriculture. The SADC Animal Agriculture Research
Network (SAARNET) was therefore formed in 1997 as a joint effort between SACCAR and
International Livestock Research Institute (ILRI) to provide the framework for regional co-
ordination and collaboration in animal agriculture. SAARNET as a regional network aims at
promoting partnerships between different stakeholders (institutes, governments, donors,
farmers and research disciplines) to solve common problems. Partnerships will be
strengthened between stakeholders through participatory approaches (workshops, seminars,
field days and training sessions) to plan and prioritise activities, and share information,
technology, and materials. The organisation and operation of a collaborative research network
such as SAARNET is very challenging though rewarding. It is therefore critical that the
institutional framework and activities of the research network are clearly related to the needs
and goals of all its stakeholders, including the farmer. Also crucial to the success of the
network are simple but efficient mechanisms for co-ordination, information dissemination,
monitoring and evaluation, and impact assessment. Above all, a network needs sustainable
funding to be able to implement its programmes and ensure the effective participation of the
stakeholders, particularly the National Agricultural Research Institutions (NARIs). This paper
elaborates SAARNETS strategy to achieve this daunting task.
Keywords: SADC, SACCAR, Regional collaboration, Animal agriculture, SAARNET
 Introduction
The central problem facing SADC member states today, like many other regions of the
developing world, is providing sustainable food security for a human population that is
increasing by almost 3% annually while protecting the environment at the same time. In
confronting this daunting task, SADC member states agreed to work together and created the
Food, Agriculture and Natural Resources Sector to address this challenge, with the Botswana
Government responsible for the co-ordination of agricultural research and training (Natural
Resources, Fisheries and Forestry were added later).
To fulfil its mandate, the Botswana Government established SACCAR in 1984 to co-ordinate
regional research and training activities. SACCAR adopted networking as a mechanism for
accomplishing its regional mandate for research in collaboration with its partners, mainly the
IARCs. Up to 1996, the ILRI supported collaborative research networks, AFRNET, CARNET
and SRNET, worked within SADC, but largely on a bilateral basis. There were no direct links
between these networks and SACCAR, except through reciprocal representations at meetings
at institutional levels.
A number of developments have begged for the revolutionisation of this modus operandi. The
1990s saw a strong focus on regional associations by the World Bank and other major donors
as providing the best opportunity for reinvigorating and strengthening collaboration in
agricultural research compared to the traditional bilateral initiatives. An external review of the
NARS- ILRI collaborative research networks (AFRNET, CARNET and SRNET) by major
donors (USAID, GTZ, IDRC and ODA) in 1994 recommended that the commodity networks be
transformed from their predominant disciplinary and pan-African operational strategy into a
multidisciplinary livestock network with regional focus. Following the review and analysis of its
1995 priorities and long-term strategy, the Board of SACCAR and the major stakeholders
recognised in 1997, the need to give livestock research, especially animal production
research, a higher profile in future regional activities compared to the past. This review also
underscored the need to adopt networking as the mode of operation in the implementation of
all regional initiatives. Another important development worth noting is the gradual but
consistent reduced funding of international and national agricultural research, especially
livestock research, due to several reasons including misconceptions about the impact of
livestock on the environment and human health as well as economic stress in the developed
world.
It is against this backdrop that SACCAR, in consultation with its major stakeholders (including
ILRI), decided to do a fundamental restructuring of its livestock sector, including the creation of
a livestock research network known as the SADC Animal Agriculture Research Network (S-
AARNET), to ensure a balanced development towards the achievement of sustainable food
security in the region. The proposed network strategy and programmes that follow provide the
platform for the implementation of the SADC-SACCAR livestock research and development
strategy.
To ensure food security and stimulate development, the region must increase the productivity
of agriculture, including animal agriculture through science - based technologies and inputs.
This can be achieved largely through the establishment of regionally co-ordinated research
systems that are strong and sustainable. This proposition was underscored in SACCAR's
Regional Research Priorities for Crop and Livestock Production document (SACCAR 1997),
which indicates strongly that "Dynamic, productive and responsive agricultural research
systems are essential to realising the potential of the SADC region. The continuous
development of technologies for different sectors of the farming community will depend on a
co-ordinated effort by countries of the SADC region"
The new paradigm in networking, that is, the regional approach, is thus very consistent with
this regional strategy.
 The SADC animal agriculture research network
(S-AARNET)
 The evolution
ILRI (then ILCA), like SACCAR, adopted networking as a mechanism for capacity building for
research in the National Agricultural Systems (NARS) in sub-Saharan Africa. This led to the
establishment of three pan-African livestock commodity networks between 1989 and 1990:
AFRNET, CARNET and SRNET.
In 1994, an external review of the NARS- ILRI collaborative research networks (AFRNET,
CARNET and SRNET), by major donors (USAID, GTZ, IDRC and ODA) recommended that
the commodity networks be transformed from their predominant disciplinary and pan-African
operational strategy into a multidisciplinary livestock network with regional focus. In the same
year, the Special Programme for African Agricultural Research (SPAAR), the donor
community, NARS and IARC's decided to work through regional organisations including
ASARECA, CORAF and SACCAR, to reinvigorate and strengthen collaboration in agricultural
research in sub-Saharan Africa. These developments were seen as a driving force for change
in the operations and structure of the existing networks as well as providing new opportunities
for technology generation and transfer to fuel sustainable increases in market oriented
livestock production in the sub-region.
In March 1996, a meeting of the Steering Committees (composed of elected NARS scientists)
of the three pan-African collaborative commodity research networks endorsed the
recommendation to reorganise the ILRI supported networks into a multidisciplinary regional
livestock network. This network was to be affiliated to the sub-regional research organisations
(ASARECA, SACCAR and CORAF) in order to make them more relevant to regional animal
agriculture research needs and priorities.
In September 1996, the recommendation to reorganise ILRI-supported collaborative research
networks was approved by the SACCAR Board of Directors (CD).
In April 1997, at a joint SACCAR-ILRI workshop, the representatives of SADC member states
formally launched the SADC Animal Agriculture Research Network and suggested its
institutional framework and mode of operation. The stakeholders also reviewed the revised
SACCAR Livestock Research Priorities (SACCAR 1997) and a proposal on " Collaborative
Research for Livestock Development in Southern Africa" from which they developed the
networks priorities for regional research in animal agriculture. The same year, the SACCAR
Board endorsed the recommendations and approved the establishment of a Regional Steering
Committee (RSC) comprising of two nominated eminent scientists from SADC regions,
representatives of each SADC member state, ILRI, SACCAR, the SADC Livestock Sector Co-
ordinating Unit and the donor community to serve as a governing body of the regional network.
A SADC Secretariat representative could have a seat on the Committee as an observer to
advise on regional policy issues.
 Vision
The vision of SAARNET is to enhance food security and the well-being of present and future
generations in the SADC region through sustainable increase in the productivity of the animal
agriculture systems.
 Mission
The mission of SAARNET is to co-ordinate and promote regional integration of animal
agriculture related research, training and information exchange activities to promote a
sustainable increase in livestock productivity.
 Objectives
provide leadership in animal agriculture research in the region
develop and deliver research products that enhance the welfare of the people,
particularly the rural poor
build and sustain regional capacity, partnerships and alliances for effective regional
research, training and information dissemination
assist in setting priorities and mobilising resources for implementing the regional animal
agriculture research and related activities.
 The institutional framework of the network
 Regional dimension of the network
As a regional network, S-AARNET will cater for the needs, interests and expectations of the
relevant stakeholders in all SADC member states. These stakeholders include scientists in the
NARS, NGOs, farmers and other public and private sector organisations.
This broad based participatory approach will ensure the creation of a critical mass of regional
expertise that will surpass the individual capacities of member states to tackle the numerous
and variable constraints to the improvement of the productivity of animal agriculture systems.
The network will provide opportunities for a collaborative approach to research with each
partner contributing according to its comparative advantages. Another benefit of the
collaborative approach is that it provides opportunities for multi-site and multi-location trials of
which results could be extrapolated and applied broadly.
 Structure of the network
The organisational structure of S-AARNET is aimed at providing opportunities for full
participation and involvement of all stakeholders interested in research and development of
animal agriculture from both the public and private sectors.
 Stakeholders
S-AARNET is open to all individuals, institutions and organisations with an interest in animal
agriculture research and development in the SADC member countries. Members are expected
to actively participate in S-AARNET activities as well as assist SACCAR and the Regional
Steering Committee (RSC) in promoting the interest of the network at all times.
 Governance
S-AARNET programmes and activities will be articulated and managed at various levels,
namely, the SACCAR Technical Advisory Committee (TCART), the Regional Network Steering
Committee, the National Network Committees, the Co-ordination Office and Technical
Committee.
 Office and technical committees
The SACCAR Technical Advisory Committee on Agricultural Research and Training (TCART):
defines regional research policy guidelines and approves regional research priorities and
programmes for technology development and transfer. The TCART will, through the SACCAR
Secretariat and ILRI, assist the network in mobilising the required resources for the
implementation of its programs.
The Regional Steering Committee (RSC) is composed of representatives of each SADC
member state, ILRI, SACCAR, the SADC Livestock Sector Co-ordination Unit and donors, two
nominated eminent regional scientists and the Network Co-ordinator who will act as Secretary
to the RSC. A SADC Secretariat representative will have an observer status.
The RSC ensures that the network policies are followed and implemented. As an advisory
committee, the major responsibilities of the RSC are:
identification of the regional research and development priorities to be submitted to
SACCAR TCART for approval through the statutory SADC channels;
advising SACCAR on programmes and projects to be supported by the region
assisting in the development of research programmes in relation to SACCAR regional
priorities
peer review and selection of research proposals for funding
monitoring, evaluation and publicising network activities
facilitating impact assessment by individual NARS
assisting in capacity building in the NARS
participating in the recruitment and evaluation of the Co-ordinator
fostering linkages within their respective countries and with the network
facilitating linkages with other organisations implementing similar projects
creating public awareness.
The National Network Committees (NNC): In the spirit of encouraging the full participation and
involvement of all stakeholders as well as addressing their needs, interests and expectations,
each participating country will be encouraged to put in place a National Network Committee.
The membership of the Committee should include NARS scientists, farmers' representatives,
interested private entrepreneurs, and NGOs and regional and international organisations
operating in the country. The major responsibility of the committee will be to articulate
constraints to livestock production at the national level, identify research priorities and link with
the RSC to develop programmes to be implemented within the framework of S-AARNET.
The national committee will also facilitate the validation and transfer at the national level of S-
AARNET developed technologies and information.
For each country, the country representative to the RSC will serve as the link between the
National Network Committee and the RSC.
The Co-ordination Office: The Co-ordination Office will be in ILRI or SACCAR. The
implementing agents and the functions of the network co-ordinator include:
the day to day management of the network activities under the operational directives of
the RSC
monitoring and co-ordinating the implementation of the network research programme in
member countries
assisting and whenever necessary guiding national scientists in the implementation of
their protocols as well as assisting them or organising assistance in the analysis and
interpretation of experimental data emanating from their research
organising and facilitating the running of identified training courses for national scientists
and other partners
organising thematic workshops and seminars and editing their proceedings
assembling and disseminating to participating scientists and other partners database
and literature searches relevant to the network research programmes
preparing an annual programme of work as well as annual reports for consideration and
approval by the RSC and SACCAR TCART
linking participating NARS scientists with ILRI scientists whenever necessary
maintaining close contacts with donors through ILRI and SACCAR Secretariat
acting as Secretary to the Steering Committee meetings
in collaboration with the Chairman, represent the network at the relevant SACCAR
meetings and other forums and establish functional linkages with the ASARECA and
CORAF Animal Agriculture Research Networks as well as with other relevant networks
and institutions operating in the region.
Technical Committees: The RSC will identify within the region and in ILRI, key experts to
serve as members of Technical Committees (TC) in the following areas of expertise: (i) animal
breeding and genetics, (ii) forage agronomy and animal nutrition, (iii) socio-economy and
policy analysis, (iv) systems and natural resources management, and (v) animal health. Each
Technical Committee will include a minimum of three experts and will assist the RSC to review
proposals and technical reports submitted by collaborating scientists.
 THE S-AARNET livestock research and development strategy
 Problems to be addressed
Without improved agricultural productivity, the SADC countries will not meet future demands
for food security. The major opportunities for increasing livestock productivity lie in the better
integration of crop and livestock production in smallholder systems and large-scale livestock
enterprises as well as improvement of range management for large-scale beef production.
The most serious constraints to improved livestock productivity in the major ecological zones
of the region (humid, sub-humid and semi-arid) are: inadequate nutrition, low animal
productivity, animal diseases, lack of information reaching the farmers on potential
innovations, and specific agro-ecological constraints. Livestock production is also adversely
affected in a number of countries by inappropriate socio-economic and policy environments
such as price control, lack of credit and market information, monopolistic behaviour, excessive
regulations and control of livestock movements or inadequate infrastructure that limit
producers' access to inputs and appropriate technologies. These result in severe setbacks in
the development of the sector, as there is no incentive for the promotion of livestock
production and the marketing of livestock and livestock products at national and regional
levels. Furthermore, livestock marketing is often constrained by zoonotic and epizootic
diseases such as brucellosis, anthrax, Rift Valley Fever, rinderpest, foot and mouth disease
etc., which result in animals or livestock products being banned for export or local/regional
consumption.
In relation to pastoral systems, some countries have developed contingency plans for
emergency responses to disasters but the operations of the existing Early Warning Systems
(EWS) do not address the interests of livestock systems. The frequency, quality and timeliness
of information provided are still insufficient to constitute proper tools for disaster management
at national or regional levels. Again findings from vulnerability studies are rarely pieced
together to form a regional picture that could lead to the formulation of comprehensive
strategies for disaster management and contingency plans (prevention, preparedness, early
warning, mitigation, public education and training, impact assessment and need assessment,
relief and rehabilitation).
From the institutional perspective, existing post-graduate education is not preparing scientists
adequately for research on integrated crop-livestock systems and many scientists are working
in isolation, not able to keep abreast of new techniques and methodologies. Extension
services are not efficient enough in promoting the adoption of modern techniques by farmers.
There are also weak links between research, extension services and farmers. The
infrastructure available for livestock research and development is also usually not fully utilised
for lack of operating funds or relatively minor capital items.
In order to overcome the above problems, NARS must adopt strategies to mobilise human and
capital resources for implementing research projects, as well as provide continuing education
opportunities to restore and strengthen their research capacity for improved technology
generation and transfer to producers.
 Strategy formulation
This network strategic plan is consistent with the revised SACCAR regional priorities and
strategies for achieving food security in the SADC region through increases in agricultural
production, efficient and sustainable utilisation and conservation of the natural resources, and
increased economic growth. The plan is also in line with the recommendations made by
representatives of SADC member states at the workshop that launched the network and
formulated its research agenda in April 1997. Formulating a livestock research strategy to
anticipate and address the needs of the region is a task with multi-dimensions. These include
guiding principles and considerations for species, commodity outputs, agro-ecological zones,
farming systems, research and research related priorities, the focus and nature of research,
equity, poverty, the environment, sustainability and potential impact. This strategic plan is built
on these elements.
 Guiding principles
The success and impact of the activities to be undertaken by the network will depend largely
on the extent to which these activities are focused and relevant to the livestock research and
development strategy of the SADC region. Therefore the choice of activities to be undertaken
by this network will be guided largely by the following criteria:
The activity must be regional in character and consistent with SADC-SACCAR strategy
and priorities. The first objective is the generation of new knowledge and products, while
the others include training, information dissemination and partnerships.
 The activity must be one for which the network has a comparative advantage.
 The activity must be research or research related.
The activity must take cognisance of global and regional concerns on food security,
poverty reduction, environmental protection, gender sensitivity and sustainability.
 The activity must be demand-led, and target SACCAR-SADC priorities.
 The activity must be multidisciplinary and multi-institutional in character.
 The activity must have potential for regional impact.
 Species and commodity output
The strategy targets five livestock species including cattle (beef and dairy), poultry, goats,
sheep and pigs and nine commodity outputs (meat, milk, eggs, fibre, hides, skins, pelts,
manure and traction). In the long term, however, with recurrent droughts and increased
potential for increased commercialisation of cattle and the intensification of small-holder mixed
crop-livestock systems, equines (especially donkeys) are likely to assume an important role in
providing draft power in the small-holder production systems.
 Agro-ecological zones and production systems
For the purpose of addressing research priorities, the region is broadly divided into three agro-
ecological zones, namely, the semi-arid, the sub-humid and humid zones. Emphasis will be
placed on mixed crop-livestock systems in the semi-arid and sub-humid agro-ecological
zones for all species and intensive commercial production of pigs and poultry across all the
zones. Some arid zone and pastoral systems have potential for increasing production and
offtake, but the opportunity to enhance these will be to focus on natural resource management
and drought mitigation.
 Research and research related priorities
The research and research-related priority domains identified include animal genetic
improvement (including the characterisation, conservation and utilisation of animal genetic
resources, and the evaluation and exploitation of resistance/tolerance to diseases and
parasitism); improving feed supply and nutrition (with emphasis on locally available forages,
crop residues and agro-industrial by-products); integrated animal health management
(including epidemiology, vaccine development, disease control and animal health delivery
systems); natural resource management in crop-livestock systems (including systems analysis
and impact assessment, range land ecology studies, monitoring of land-use systems and their
impact on the environment); policy and socio-economic analysis on issues relating to animal
agricultural development; and strengthening regional research capacity.
 Nature of research
The network research activities will focus on applied and adaptive research in the development
and management of the production systems, the conservation and management of resources,
and selected commodity development and improvement. However, the network will undertake
basic and strategic research activities needed to support its applied and adaptive research
activities in collaboration with partners including International Agricultural Research Centres
(IARC's), Advanced Research Institutions (ARI's) and the private sector. The complex nature
of animal agriculture demands a holistic approach to research, hence the research activities
will emphasise a farming systems approach with opportunities to ensure the full participation of
the relevant stakeholders, including farmers.
 The Beneficiaries
The majority of the key players in livestock dependent production systems are resource poor,
especially women and children. The primary concern of the strategy is therefore the
generation of knowledge and products that will enhance the welfare of the resource poor.
The immediate beneficiaries will be the NARIs and their scientists as well as the collaborating
research and development institutions including the NGOs, the extension services and the
regional/international centres.
In the medium term it will be the farmers participating in the implementation of the network
strategy programmes.
In the long-term the main beneficiaries will be farmers and the national economies through
economic growth at household and national levels.
 Considerations driving the research and development agenda
The logical consideration in developing a broad based regional livestock research and
development agenda is to look at the development needs of the sector and the factors driving
those needs.
A number of major trends are known to be affecting the livestock sector and determining the
needs of the sector and therefore merit consideration. These are, inter alia:
almost all of the ruminant meat and milk as well as most of the pig meat production will,
for the foreseeable future, continue in the smallholder systems
however, most of the increase in production will come from the intensification of livestock
production in these smallholder-mixed farming systems, in which there will be a greater
interdependence of livestock and crop sub-systems
and as intensification increases, research must increasingly recognise the need for
ecosystem management
also with intensification of agriculture, the productivity of smallholder production systems
will largely depend on increased use of inputs such as traction, manure and fertilisers
with increase in urbanisation, changes in consumer preferences could lead to
specialisation and in some cases, a shift to industrial scale production as in the case of
pig and poultry production
resource management issues are perceived as the major challenges in some pastoral
systems with potential for increasing production and offtake
new science-based technologies have potential to significantly enhance animal
agriculture productivity
strengthening human and institutional capacities will create greater capacity for research
and development in the region.
 Research priorities formulation process
The research agenda largely addresses the revised SACCAR priorities and strategy. The
agenda also draws on several other sources of information including a number of studies,
workshops and regional sector reviews that have identified opportunities for increased and
sustainable livestock productivity deriving from the application of research-based technologies.
These include among others:
November 26–27 1986: SACCAR workshop on smallholder dairy, small ruminants, pig,
poultry and rabbit production in the SADC countries, Maseru, Lesotho.
July–August 1988: ILCA-commissioned feasibility study on ILCA/SACCAR livestock
research collaborative activities in SADC countries.
March 1989: SACCAR/Rockefeller Foundation/CGIAR Task Force on African
Consultations.
July 1992: ILCA workshop on the future of livestock industries in East and Southern
Africa, Kadoma, Zimbabwe.
November 1992: World Bank/SPAAR/SACCAR Task Force Report on Agricultural
Research in Southern Africa: A Framework for Action.
1995: Consultancy report on the long-term strategy for regional research priorities on
food, agriculture and natural resources in the Southern Africa Development Community
(SADC) (Vol. 3. Livestock Sector).
January 1997: Consultancy reports on the "regional research priorities for crop and
livestock production.
February 17–21, 1997: FAO/SADC/UNDP workshop to discuss "a basic strategy for the
development of farm animal genetic resources in SADC countries, Gaborone, Botswana.
March 10–14, 1997: SACCAR major stakeholders meeting to discuss and review
consultants' recommended strategies for crops, farm animals, natural resources
management and human resource development in Southern Africa, Gaborone,
Botswana.
April 21–25, 1997: SACCAR/ILRI workshop on "The Regionalisation and Reorganisation
of ILRI - NARS Collaborative Research Networks ( Cranes ), Gaborone, Botswana.
The current SACCAR strategy for agricultural research and training in SADC.
 Research agenda goal and objectives
 Goal
The main goal of the research agenda is to enhance and sustain food security and improve
the living standards of the people of Southern Africa through sustainable increases in the
productivity of the animal agricultural production systems, especially in the smallholder
systems.
 Objective
The specific objective of the research agenda is to achieve sustainable improvement in animal
agriculture productivity in the SADC region in order to decrease importation of livestock
products and generate employment and income in the smallholder farming systems.
 The research agenda and related activities
 Genetic improvement, conservation and utilisation
The objectives of the genetic research are to: identify indigenous farm animal genetic
resources with exceptional productive and adaptive characteristics to serve as a basis for
sustainable farm animal production improvement; develop breeding strategies and feasible in-
situ and ex-situ conservation methods for sustainable utilisation and conservation of identified
genotypes; build a regional database on animal genetic resources and contribute information
to the geo-referenced database developed by ILRI, the FAO and other organisations on farm
animal genetic resources and the systems in which they exist. These activities will be linked
and complementary to the FAO Global Programme for the Management of Animal Genetic
Resources, the on-going FAO/SADC initiative on Farm Animal Genetic Resources and the
ILRI Pan-African initiative on the characterisation of animal genetic resources.
 Research activities
Characterisation (phenotypic and genotypic) of the animal genetic resources and the
systems in which they exist.
Breed improvement and conservation.
Development and utilisation of disease resistant breeds/strains.
 Expected products
Animal genetic resources and their systems characterised.
Genetic relationships of known breeds established for better conservation and utilisation.
Breeding and management strategies developed.
Regional database on animal genetic resources established.
 Nutrition and feed resources development
The objectives of the feed resources research activities are to identify and characterise
available feed resources in the region; introduce and evaluate forage germplasm (grasses,
MTPs etc.); and develop appropriate feeding packages based on crop residues, forages and
agricultural and industrial by-products, especially for intensive production systems.
 Research activities
Introduction, evaluation and conservation of forage germplasm.
Feed improvement (forages, crop residues and agro-industrial by-products).
Feeding and management options for intensive management systems.
Research to exploit indigenous knowledge.
 Expected products
Adapted forages established.
Technologies for incorporating forages into smallholder systems developed.
Suitable feed resources identified for supplementary feeding.
Database on available and suitable feed resources.
 Sustainable integrated animal health management
The aims of the animal health research are to develop cost effective and affordable integrated
disease and parasite control regimes tailored for different farming systems and to maximise
productivity by reducing the impact of target diseases.
 Research activities
Epidemiology and disease control.
Development of vaccines and diagnostic tools.
Disease resistance/tolerance studies for the development and promotion of disease-
resistant/tolerant domesticated animals.
Ethnoveterinary research to exploit indigenous knowledge.
Drug efficacy and resistance studies to develop improved drug use regimes.
Economics of diseases and their control measures.
 Expected products
Community-based integrated disease management and control strategies.
Economic impact of target diseases.
Vaccines for target diseases.
 Natural resource management
The objective of this research domain is to improve the efficiency of use and conservation of
the natural resources that support the crop- livestock systems.
 Research activities
Rangeland ecology and management (technical and socio-economic).
Soil-plant-Animal Relationships.
Monitoring of land-use systems and their impact on the environment.
Research to exploited indigenous knowledge.
 Expected products
Better understanding of the interaction of livestock and environment.
Methodologies for on-farm R&D.
Strategies for rangeland management.
 Systems, policy and socio-economic analysis
The objective of this research area is to promote better formulation of regional policies for
informed decision-making.
 Research activities
Analysis of the socio-economic and policy constraints related to land tenure and land-
use, marketing (including post-harvest technology), inter-regional trade, privatisation,
and delivery of veterinary and other animal agricultural services.
Economic, social and environmental impact assessment of farm animal production
programmes and interventions.
Identification and recommendation of policy options that are supportive of sustainable
farm animal production.
 Expected products
Knowledge and understanding of the technical, social and economic factors affecting
animal agriculture.
Policy options for informed decisions.
 Crisis mitigation in livestock dependent systems
The objective of this work is to make livestock-dependent systems more resilient in times of
crisis, especially drought.
 Research activities
Identification of the major environmental and socio-economic indicators of crisis
incidents.
Identification and evaluation of existing and potential coping mechanism and
contingency plans.
Assessment of the impact of crisis on bio-diversity and the farmers.
 Expected products
Major crisis indicators identified.
Coping mechanisms and contingency plans developed.
 Human resource development (capacity building)
The objective of this activity is to strengthen NARS capacity to carry out livestock research
through training and the provision of technical and specialised information, especially within
the framework of regional research projects. The tasks/activities identified are:
 Facilitating training for MSc and PhD in the following areas of specialisation
Forage science.
Animal breeding/genetics.
Biotechnology/molecular biology.
Livestock economics.
Natural Resources Management/Environmental Services.
Social anthropology with emphasis on livestock
Animal health economics.
Animal Health.
 Short courses/workshops through collaboration with ILRI and other relevant
organisations in the following areas:
Project proposal development
On-farm research (surveys and monitoring methodologies including livestock breeds
and diseases).
Participatory Research methods including gender analysis.
Statistical techniques and tools for livestock data management
Dairy processing technologies.
Recent advances in animal nutrition research techniques
 Improvement of agricultural information management and delivery systems
(Communication).
Diagnostic methodologies for target diseases.
Reporting and presentation of research results.
Impact assessment.
 Training of trainers
Efforts will be made to identify, from among the participants in the various training
courses/workshops, potential trainers for additional preparation to enable them assume
responsibility for regional or national training.
 Mode of operation and implementation
The regional collaborative research programme will be developed and implemented within the
framework of several collaborative research programmes:
 Development and approval of research and research related
activities
A team of experts will be drawn from member states to identify and prioritise researchable
issues for each research theme identified. This is to bring focus to regional research as well as
avoid duplication of efforts
NARS scientists in SADC member states and their collaborating partners will develop
operational research proposals addressing one or more topics of the regional collaborative
research priorities identified by the regional team of experts following the guidelines and
format defined by the S-AARNET RSC (see page 18).
The proposals will be reviewed by appropriate technical committees based on set criteria
(including scientific merit, potential impact and regional relevance). The RSC will decide on the
award of the competitive research grants to successful proposals. The Co-ordinator will
provide the feedback to the participating scientists.
The successful proposals will be implemented by the investigating teams in their respective
NARS institutions with the technical backstopping of the S-AARNET Co-ordinator, the
members of the RSC and identified experts from within the region.
 Management of activities
The programmes will be managed daily by the S-AARNET Co-ordinator under the guidance of
the RSC with administrative support from ILRI and SACCAR. The Co-ordinator and the RSC
chairperson will be responsible for reporting on programme activities to SACCAR and ILRI.
Network programme funds will be kept in-trust by ILRI but the allocation of operational project
funds and their disbursement will be decided by the RSC. ILRI will be responsible for the
financial accounting to SACCAR and the donors.
 Collaborators
The potential collaborators include scientists in ILRI and other regional and international
research centres operating in the region (CIAT, CIMMYT, ICRAF, ICRISAT, IFPRI, IITA inter
alia) as well as scientists from Advanced Research Institutions (ARI's) including Universities in
Europe and the USA.
The scientific community will work in close collaboration with other stakeholders in the public
and private sectors including farmers, relevant NGOs, entrepreneurs, extension services,
government and international agencies inter alia.
Whenever appropriate, linkages will be established with other SADC networks and projects to
avoid duplication and ensure complementarities.
 Monitoring and evaluation
Overall, the achievements and successes of S-AARNET activities will be assessed in terms of
its objectives as set out earlier.
Specifically, the impact of network activities will be measured in terms of research products
and their direct or potential effect on the target groups, the extent and quality of human
resources development and the level of information exchange (including the number of
publications). Other measurable indicators at various levels (short, medium and long term) will
be indicated in the logical framework developed for operational projects.
Members of the RSC will assist the Co-ordinator in monitoring the network's activities in their
respective countries as well as at the regional level whenever requested by the Co-ordinator.
The RSC will review all progress reports submitted by collaborating scientists. Such progress
reports will follow a given format including statements on the objectives of study, work done,
achievements, constraints and budget. Attached to this report should be the future work plan
and the budget.
The progress of S-AARNET programmes will be reviewed annually by the Steering
Committee. In addition, there will be mid-term and end-of-projects external evaluation as well
as mid-term and end-of-projects workshops to discuss and share the results.
 Guidelines and format for developing project proposals
 Guidelines
Guidelines for project proposals follow:
The study must address researchable areas indicated in the regional collaborative
research agenda.
Applicants should be employees of national agricultural research institutions,
universities, relevant NGOs and private sector from SADC member's states.
All proposals to the RSC should be channelled through the designated national
representative to the network RSC.
Proposals will be evaluated based mainly on scientific merit, potential impact and
regional relevance.
Multidisciplinary and trans-boundary institutional project proposals will have added
advantage for funding.
 Format
The proposal should follow the following format:
Title.
Theme: refers to the identified regional themes.
Principal investigator(s).
Collaborators/ partners: other scientists, extension officers, NGOs, farmers.
Site(s).
Starting date and ending date.
Background and justification: description of the region; production system; problems to
be addressed; current knowledge on the topic (brief review of the literature relating to
key elements of the intended research).
Objectives :
Overall
Specific objectives
Expected output: the proposal is expected to generate products including technologies
and new information. Milestones for the expected products must be indicated.
Beneficiaries.
Material and methods: should be clearly indicated including experimental design, taking
into consideration each of the objectives of the study. Methods should be clear on what
data/information will be collected, and how and when the data will be collected.
Workplan: clearly indicate the different phases of the study. It should be an annual work
plan indicating details of the activities to be carried out.
Logical framework.
Budget: the structure of the budget should indicate, per year, the budget requested from
the network as well as the host institution's contribution by component. A summary of
that budget will indicate the following components:
Personnel
Equipment
Operational cost
Human resources development: training and information
Overheads (not more than 10% of total operational funds requested).
Budget notes: These should justify the funds requested by indicating how estimates
were derived
 Conclusion
A strategic plan is a statement of intent with a time frame. This network strategic plan is for the
next 5 years. But this is also a document to guide the stakeholders in running the network. It is
there not cast in stone. It will be reviewed as and when the majority of the stakeholders feel
that circumstances warrant a review.
For more information, please contact the co-ordinator at this address:
Prof. S. H. B. Lebbie
Regional Co-ordinator
SADC Animal Agriculture Research Network (SAARNET)
International Livestock Research Institute (ILRI)
P. O. Box 30709, Nairobi, Kenya
Phone : 254-2-630743 Ext. 3402
Fax : 254-2-632013 or 631499
E-Mail : s.lebbie@cgiar.org
web site: http://www.cgiar.org/ilri/
 References
SACCAR (The Southern African Centre for Co-operation in Agricultural Research and
Training) 1997. Regional Research Priorities for Crop and Livestock Production. African
Centre for Cooperation in Agricultural and Natural Resources Research and Training,
Gaborone, Botswana. January, 1997. pp 119.

The challenges of training animal scientists in the SADC:
Experiences from Malawi
P. Mumba, J.W. Banda and C.B.A.Wollny.
University of Malawi, Bunda College of Agriculture
P.O. Box 219, Lilongwe, Malawi
Abstract
Introduction
Method of assessment
Results
Needs of BSc (agriculture) holders in Malawi
Needs of final year BSc (agriculture) students in Malawi
Needs of (supply to) employers in Malawi
Discussion
Conclusion
References
 Abstract
The Animal Science Department at Bunda College has been running a regional postgraduate
degree program at the Masters level for eight years. The candidates have been both local and
international from the Southern African Development Community (SADC) region. So far,
twenty-one local and fifteen internationals have graduated from the program. While the
employment distribution of the graduates in Malawi has varied, the results have shown that
eleven graduates are in positions not directly involved with livestock. Apart from the University,
which has six of those graduates, the Ministry of Agriculture has only two graduates who are
directly involved with livestock projects. However, for the internationals, nine graduates are in
the Ministry of Agriculture, four in the University and only two in the other fields. This then
gives an opportunity for Bunda College to train more animal scientists. However, the results of
a survey carried out in Malawi and partially in Zimbabwe, have shown that the majority of the
potential candidates would want to be trained outside Africa. The college should entice as
many qualified applicants as possible. The challenge to the University of Malawi is to have a
curriculum that addresses the needs, values and objectives of the SADC countries. However,
this may not be possible and so the alternative is to reduce knowledge transfer in favour of
''key competencies''. Sustainability in terms of funding is another challenge to the University.
The regional MSc program has to-date been funded by GTZ which is now phasing out. The
question now is ''What's next?'' The University has to find other sponsors in order to sustain
the program as regional. In addition, the government's priority is geared mostly toward crops
and little effort is put on livestock. The College must sustain the program against the
government's low priority in the livestock industry and the desire of candidates to go and study
abroad.
 Introduction
Bunda College is one of the five constituent colleges of the University of Malawi and was
opened at its present 2000 hectare site in 1967. The college has the following seven
departments: Crop Science (CS), Animal Science (AS), Agricultural Engineering (AE), Rural
Development (RD), Home Economics and Human Nutrition (HE/HN), Aquaculture (AQ) and
Language and Development Communications (LDC). Before 1987, the College had been
mounting postgraduate programs on an ad hoc basis only. After that year, the Animal Science
Department started a full time postgraduate program at the Masters level. This program
became regional, under the sponsorship of the German Technical Service (GTZ) in 1989.
Later, the other departments notably, CS and RD started their Masters programs as well. The
MSc program in Animal Science has been running for eight years now, with 2000 as the ninth
year of operation. The program was interrupted for two years between 1990 and 1992 due to
accommodation problems. The regional MSc program offers up to ten scholarships per year,
50% of who should be from the host country. In the Animal Science Department, it was
observed in 1997, that only 42.2% of the total number of students (based on a theoretical total
of 90) had enrolled over nine years, of whom 22.2% were Malawians (out of the required 50%)
and 20% were foreign students (out of the required 50%). The figure had increased to 49% in
1999, of whom 22.7% were Malawians (an increase of only 0.5%) and 26% were from the
region. In the 1998 academic year, the complete intake consisted of only regional students.
Up to 1999, the Animal Science Department at Bunda College has enrolled 61 students at
Masters level, (36% Malawian and 64% International students). Of the total enrolled, eleven
students either dropped or were deceased. A total of 36 students have successfully completed
and defended their research work. However, not all the candidates who graduated have
entered the Animal Science field. The needs of the potential candidates and those of a
particular country may vary considerably. These and other factors may pose as a challenge to
the training of animal science specialists in the SADC.
This paper examines the challenges of training animal scientists, with respect to the needs
and supply, based on the employment type of the previous graduates and the results of a
survey that was carried out in Malawi and Zimbabwe in 1997.
 Method of assessment
A database of the employment record of the MSc graduates in animal science from Bunda
College was compiled and used as such. Three questionnaires were formulated. One set was
for BSc (agriculture) holders and was administered to a total of seventy three people with only
a BSc in agriculture (those working in various government or private sectors). A second set
was for final year BSc (agriculture students and was administered to a total of seventy
students at Bunda College of agriculture and some animal science students at the University
of Zimbabwe. A third set of questionnaires was administered to twenty BSc (agriculture)
employers in Malawi. The data were all analysed by SPSS computer programs by generating
frequencies.
 Results
1. Distribution of Bunda College Animal Science MSc graduates in various fields in the SADC
region:The distribution, in various fields, of the twenty-one Malawians who have so far
graduated with a Masters degree in Animal Science is shown in Table 1. Almost half of the
graduates have gone into fields other than agriculture. The ''other fields'' are NGOs,
companies or jobs outside Malawi, which may or may not involve animal science or related
projects. People in these areas are therefore, not working full time as animal scientists. Only 3
candidates joined the Ministry of Agriculture, of whom, one is at Principal Secretary level and
is not involved in fieldwork. The other two are directly involved in animal science work.
Although six graduates joined the University, only three are in Animal Science. This means
that over eight years, only five Malawian candidates trained at Bunda College are directly
involved with animal sciences. It is interesting to note that over the same period, fifteen
international students defended in the same program (Table 2). Nine of these are in the
Ministry of Agriculture, involved with livestock, four are in the University/College and two are in
the ''other fields''.
Table 1. Bunda College Malawian Animal Science MSc graduates in the various fields.
Year defended Place of work
 Ministry of Agriculture University Other4 Deceased Total
1989 11 12 0 1 3
1991 0 0 2 0 2
1996 1 1 1 0 3
1997 1 2 5 0 8
1998 0 23 1 0 3
1999 0 0 2 0 2
Total 3(2) 6 11 1 21
1 Principal Secretary. 2Joined Agricultural economics department. 3 One joined Aquaculture
Department and one is at Polytechnic. 4 These are NGOs, companies or jobs outside the
country
Table 2: Bunda College International Animal Science MSc graduates in the various fields.
Year defended Place of work
 Min. of Agriculture University Other Deceased Total
1991 3 1 0 0 4
1996 2 1 0 0 3
1997 1 0 0 0 1
1998 1 1 1 0 3
1999 2 2 0 0 4
Total 9 4 2 0 15
 Needs of BSc (agriculture) holders in Malawi
Of the respondents covered during the recent survey, 56.2% were from the government,
28.8% from parastatals and 15.1% were from the private sector. A few answers relevant to the
study are presented in Figures 1–3.
Most respondents (61.1%) obtained a pass degree while 38.8% got a credit but none with a
distinction. This may not be surprising as most distinction materials are usually reabsorbed into
the university system. Figure 1 shows the fields in which the various respondents are working.
It can be seen that 19.2% are working in crops (of which 12.3% are in government and 6.8%
are in parastatals), 15.1% in Agricultural Economics, 20.5% in Extension, 6.8% in Animal
Science field (of which 2.7% are in government and an equal number in parastatals and 1.4%
in private sector) and 4.1% in Agricultural Engineering. However, the majority (31.5%) are in
other fields, such as Banking. Of all the respondents, only 41.1% had come across the
SACCAR-GTZ scholarship advertisement. Of those that had seen the advertisement, the
majority (51.3%) were interested in Agricultural Economics (Figure 2), 30.8% were interested
in Crop Sciences and 7.7% were interested in Agricultural Engineering. A relatively small
percentage (10.3%) was interested in Animal Sciences (of which, 5.1% were from government
and 2.6% each from parastatals and private sectors). The small number of respondents
interested in Animal Sciences may be related to the small number of people working in that
field as shown by Figure 1.
Figure 1: Respondents field of work
Figure 2: Field of interest in the SACCAR-GTZ advertisement by the respondents.
Figure 3: Choice of the Universities in the SADC by the respondents.
The proportion of respondents who have ever applied to the SACCAR-GTZ regional MSc
scholarship was only 22.5% while only 8.4% have ever applied for MSc in Animal Sciences at
the University of Malawi. The latter fits well with Figures 1-3. While most respondents (95.3%)
were aware of the existence of the Regional MSc program in Animal Science, many (35.6%)
have not applied to the program because they are not Animal Scientists while 20.3% said they
had financial problems and another 14.1% would prefer to study outside Malawi. A small
number of respondents (6.3%) said that they were satisfied with their present condition. Most
notable was the choice of Universities, in the SADC (Figure 3) the respondents would prefer to
go attend. Most respondents would prefer to attend any South African University (57.9%), or
University of Botswana (19.6%) or University of Zimbabwe (17.9%) but only a small number,
10.3%, would prefer to attend the University of Malawi.
 Needs of final year BSc (agriculture) students in Malawi
While only 26% of the students preferred work, a good number (33%) indicated they would
prefer further training. Although the choice for work was high, most of the students (57%) said
that they had no concrete prospect for work. A more important observation was that 97% of
the students indicated an interest in further training to the PhD level. Figure 4 shows that 32%
of the students preferred to study in Europe and 53% in USA. A similar survey carried out in
Zimbabwe showed that a larger proportion of the Zimbabwean students (17%) would prefer to
study in Africa compared to only 3% of their Malawian counterparts.
Figure 4: Choice of continent/country of study by the students
Figure 5: Choice of a SADC University by the students
Figure 5 presents a more interesting scenario on the choices of universities in Southern Africa.
Most students would prefer to study in a South African university (56%), University of
Zimbabwe (22%) and to some extent University of Botswana (16%) and few (7%) would want
to go to the University of Malawi. A similar survey carried out in Zimbabwe showed that the
majority (50%) would prefer to go to their own university, followed by any South African
university (42%) and none would want to go to the University of Malawi. Finally, a bigger
percentage of the students (55%) said that they had not come across the SACCAR-GTZ
regional scholarship advertisement. This was rather surprising since the advertisements are
posted on the notice boards all the time. However, of the students who have come across the
advertisements, 36% said they were interested in agricultural economics, 25% in animal
science, 19% in crop science and 14% in agricultural engineering. For further training, there
were an almost equal number of candidates who said they would pursue agricultural
economics (21%), animal science (19%) and agricultural engineering (16%).
 Needs of (supply to) employers in Malawi
The employers were from the government (50%), parastatals (34.6%) and the private sector
(15.5%). The proportion of employees in these sectors was 8.1% in the animal science field of
which, 1.6% have a BSc (agriculture) only degree. A larger proportion (30%) of the employees
were in the crops field. It was noted that 92.3% of the employers offered "on the job" training
to BSc (agriculture) graduates from the University of Malawi. This agrees with the BSc
(agriculture) holders' rating of the relevance (i.e. how practical) of the Bunda's BSc program
mentioned earlier.
The class of professionals preferred by the employers were BSc holders (42.3%) followed by
MSc (38.5%) and then PhD holders (11.5%). The preference for MSc is quite high, and for
this, 28.6% of the employers said they were willing to offer a full scholarship for MSc studies of
their staff while 19% said they would offer only a partial (50%) scholarship. The others
(52.4%) would not offer any. Although 57.9% of the employers have heard of the SACCAR-
GTZ regional MSc program in animal sciences, only 15% have sent their staff to the program.
The reasons given were mostly financial and that the animal science program did not meet
their requirements. It seems possible therefore, that the high preference for MSc holders is in
other fields and not animal science.
 Discussion
Tables 1 and 2 indicate that the Animal Science Department has trained a number of Animal
Scientists at the Masters level. Unfortunately for Malawi, most of them are not in the animal
science field. Since the Government has now launched the livestock policy through Malawi
Agricultural Sector Improvement Program (MASIP), this should be an opportunity for the
Department to train even more specialists on the assumption that the graduates will work in
that field. This is a challenge to the Department in that it has the chance to convince the
government of the lack of work force in the animal science field. Most international Animal
Science graduates from Bunda College have found jobs in the livestock field. The need for
more such specialists in the SADC countries gives an opportunity to get more candidates to
be trained. On the other hand, there is competition with the other universities like the
University of Zimbabwe which may attract more candidates.
However, data from the potential candidates, i.e., the needs of the BSc Holders and the
undergraduates show that there are two problem areas: the marketing of the MSc program is
not vigorous enough and the image of the College/University is not good. It is encouraging to
note that many respondents were aware of the regional MSc program in Animal Science.
Currently, advertisements for the regional programs are carried out in local daily newspapers
and also by SACCAR in their newsletters. They are also sent to all relevant government
ministries and organisations for information to their staff. Recently, advertisements have also
been sent to SADC countries national newspapers to be covered in those papers. Despite all
this effort, there are still not enough qualified Malawian applicants. This is not surprising since
most students who go out into the field are those that have obtained a pass degree at the
undergraduate level. With a credit pass or better as a requirement for the Masters degree, the
Department will continue to face the problem of getting more applicants. This problem is also
compounded by the desire of most Malawians to go and study abroad and not within Malawi.
This has also been observed by other researchers and they have suggested that this is due to
the declining quality and relevance of higher education in most African Universities
(Fernandez 1998; Matos 1998; World Bank 1997). The BSc holders' views of the relevance of
the Bunda (BSc) programs as they relate to their work is in agreement with this. Most
respondents (32%) said the programs are 60-100% practical, 41% said the programs are 50-
59% practical and 28% said the programs are 1–49% practical. The figure 28% is significant
enough and suggests that aspects that are more practical should be introduced into the
curricula. This means that the college should reduce knowledge transfer to students in favour
of more "key competencies". It is of interest to note that in a recent survey on the factors
affecting the performance and quality of the students (Mumba 2000) it was observed that the
students appreciated the quality of education at Bunda College. It must be recognised that
postgraduate training overseas does not usually provide opportunities to work on African
problems (Abegaz 1994). In this regard, the College must convince as many potential
candidates as possible in the SADC, to pursue their postgraduate degree at the University of
Malawi. Therefore, the College faces the challenge to come up with the curricula that is in
consonance with, and respond to the needs, values, priorities and objectives of the different
SADC countries. However, Aina (1998) has pointed out that because of the unequal economic
and political relations between nations, the needs and priorities of different countries are
bound to be different. This is where ''key competencies'' are more important than just
knowledge transfer. The sustainability of the program also depends on continued funding. So
far, the major sponsor of the regional program has been the German Technical Service (GTZ)
which is now phasing out. If funding is not available, then getting the required mass of trained
personnel in the Animal Sciences will present a great challenge.
The image issue is probably a more serious one. Reading through the comments of the
BSc(agriculture) holders, several issues were raised such as: (a) MSc programs at Bunda
College take too long (more than two years), (b) there is more theory in the curricula than in
practical aspects which should be reversed and (c) the scholarship package not enough.
If the programs take too long, who would be at fault? Obviously, it is either the students or the
faculty. Since students are supervised by faculty, the problem may then be with the faculty. If
the faculty are not very strict with the regulations or they pamper the students to the extent that
the latter relax, delays may be inevitable. For example, take thesis examinations: "What is the
dead line for submission of theses?", "How long should an external examiner keep a thesis?",
and "Under what conditions is extension of scholarship permitted?". Step by step analysis of
these questions may reveal why the MSc programs at Bunda take too long and who may be at
fault. The faculty must look into the analysis of these questions more closely to avoid
unnecessarily long programs as claimed by the respondents. This is a real challenge to the
faculty in order to improve its image. Although the Animal Science Department has greatly
improved graduating its students within 2 years, the general perception of potential candidates
on the prolonged graduation periods remains the same. On the other hand, the students
themselves may deliberately prolong their research work so that they can look for
employment. Experience has shown that once these students are employed, they do their
research slowly. The college faces the need for internal quality control.
For the past five years, the government's support for tertiary education has been steadily
declining. Although this is also true of many African countries (World Bank 1994), the situation
in Malawi seems to be getting out of hand. The current government policy is "poverty
alleviation", but it seems that this poverty alleviation policy is only geared toward crops and
not animal production. This can be seen from the number of people working in crops field as
compared to those in the livestock field (Figure 1). The main problem it seems is that the
government has put a low priority on livestock and no policy on livestock existed until recently.
Whether this will be implemented soon or not is a matter of speculation. If the Government
sees need to improve the nutritional status of its people, then there must be clear national
policies on both crops and animals since both are necessary. The University must take the
lead in livestock research, which deals with problems at the societal level and they must
convince the policy makers that this is important for the good health of the people. Unless this
done, fewer and fewer people will be in the Animal Science field and in the long run, nobody
will have any vested interest in Animal Sciences. As one employer put it, "It can only be hoped
that the Agriculture Sector Investment Plan which the Ministry of Agriculture and Irrigation has
developed, will include a strong component on livestock development". However, such
livestock development calls for the participation of the private sector as well. In order to
encourage primary producers, the University, the government, parastatals and private sectors
must come together with a good policy on livestock. In this way, more people can be trained in
animal sciences.
 Conclusion
The paper has highlighted the challenges to the training of Animal Scientists in Malawi with
regard to the needs and supply. Although the college has trained several Animal Scientists at
the postgraduate level, many of those are not directly working in this field. The few animal
scientists in the country is an opportunity for the college to train more people in the area.
However, the desire of the potential candidates to be trained overseas, has a competitive
effect which is a challenge to the institution. The University should reinforce research at the
grassroots level to address societal problems. However, unless the Government reconsiders
its priorities to include livestock, few graduates will actually work in this field. This will kill the
livestock industry. On the other hand, the University is faced with the problem of funding in
both teaching and research. In the past, the major source of funding was GTZ, which is now
phasing out. The College must sustain the program against the government's low priority in
animal sciences, dwindling funding and the desire of candidates to study abroad.
 References
Abegaz B.M. 1994. Universities in Africa: Challenges and Opportunities of International
Cooperation, p171–205. Reader for Conference on ''Challenges and Perspectives of
Universities in Africa, 1998''. Compiled by Bergs R. and Gmelin W. GTZ, DAAD, DSE.
Aina T.A. 1994. Quality and Relevance: African Universities in the 21st Century, p15–50.
Reader for Conference on ''Challenges and Perspectives of Universities in Africa, 1998''.
Compiled by Bergs R. and Gmelin W. GTZ, DAAD, DSE.
Fernandez L.1998. Universities between commitment to local needs and international
aspirations: The South African Experience, In: 'Challenges and Perspectives of Universities in
Africa. p25–31. GTZ Education report No. 82.
Matos N. 1998. Challenges and Perspectives of Universities in Africa. In: Challenges and
Perspectives of Universities in Africa. pp8-18. GTZ Education report No. 82.
Mumba P. 2000. Factors affecting the performance and quality of the students at the
University of Malawi, Bunda College of Agriculture. A report Submitted to The Deutsche
Gesellschaft Fur Technische Zusammenarbeit (GZT). 13 pp.
World Bank, 1994. Higher Education: The lessons of Experience. The World Bank,
Washington D.C., USA.105 pp.
World Bank 1997. Revitalising Universities in Africa: Strategy and Guidelines, p51–83. Reader
for Conference on Challenges and Perspectives of Universities in Africa, 1998. Compiled by
Bergs R. and Gmelin W. GTZ, DAAD, DSE.
Capacity building for sustainable use of animal genetic
resources1
Ntombizakhe Mpofu, Habib Ibrahim, Michael Smalley and Edward Rege
International Livestock Research Institute (ILRI), P O Box 5689, Addis Ababa, Ethiopia
Email: ILRI-Ethiopia@cgiar.org
Abstract
Introduction
Needs assessment activities
Animal genetic resources for sustainable agriculture
The project's main activities
Training courses
Training resources
Partnerships
Expected outcomes and project evaluation
Expected outputs
Project evaluation
References
 Abstract
Africa has a challenge to meet the rising food demands that will come about due to increased
population size and urbanisation. Such challenges can be tackled in many ways that include
improving livestock production and this entails making improvements in feeding and disease
control, using appropriate genotypes and genetic improvement of livestock. Research and
training are some of the major requirements to realisation of improvements in livestock
production. Sub-Saharan Africa has a number of colleges and universities and research
stations that do postgraduate training and research in animal breeding and genetics. It is
important that training and research at these institutions is relevant to Africa's needs so as to
produce scientists that will become trainers or take up leadership in research and
development in animal breeding including sustainable use of animal genetic resources. The
International Livestock Research Institute (ILRI) and the Swedish University of Agricultural
Sciences (SLU) have initiated a capacity building project. This project aims to address training
of researchers in animal genetic resources by introducing a new model of capacity building
which targets qualified lecturers and researchers working in sub-Saharan African universities
and research stations with a specific focus on making graduate training in sub-Saharan Africa
more relevant to the region. The paper describes the project in terms of identified needs and
expectations of stakeholders, the capacity building model developed to address these needs
and the activities that are to be undertaken to meet the project outcomes. The project
activities include a training of trainers course, building partnerships and developing computer-
based training resources for use by university lecturers and national scientists. The proposed
project evaluation procedures are also presented.
 Introduction
The increased demand for livestock products has been putting pressure on African livestock
owners to increase production. Steps taken to increase production included identifying and
using high-producing genotypes. Unfortunately for Africa, little attention was paid to improving
the genetic potential of local breeds for increased production. Instead, a view was taken that
increased production would be best realised by importing foreign high-producing breeds and
using them in purebreeding and crossbreeding systems. The introduction of these foreign
breeds was made easy by advances in biotechnology, particularly advances in artificial
insemination. This crossbreeding and/or replacement of indigenous breeds with foreign
germplasm is one of the most serious threats to indigenous populations. In addition, the
system is also not sustainable, as most foreign breeds are generally not adapted to African
production conditions. With the expected further increase in demand for meat and milk in
developing countries between now and 2020 (Delgardo et al. 1999), there is even a greater
need to develop and promote sustainable ways of livestock production. This will place
significant new demands on national capacities for research and development in developing
countries.
Even though research training is introduced at the undergraduate (BSc) level in most
universities, it is at postgraduate level (MSc and PhD) that most of the research training is
done. The number of national scientists for livestock research and development continues to
be a cause of concern (Wilson et al. 1995). In 1986, there were fewer than 1000 livestock
researchers in the national agricultural research institutes in African countries with a BSc
training or above, and just over 300 with PhD degrees. Whilst most universities in Africa now
have courses in the agricultural sciences, considerably fewer provide post-graduate training
(Pardey et al. 1997). Africa has, therefore, relied on sending its people overseas for
postgraduate training. However, postgraduate courses are increasing in universities in sub-
Saharan Africa. Scientists trained at these universities are absorbed (as staff) by local
universities and local research and development institutes.
The pace of scientific progress is increasing and this continues to affect both what can be
achieved by agricultural research and how it is to be achieved. New tools and approaches in
both molecular and quantitative genetics are supporting radical changes to what can be
achieved through animal genetics. These techniques can be applied in characterising
indigenous animal genetic resources and that information can promote their use in sustainable
breeding programmes. There is a need, therefore, to provide national scientists with the skills
and knowledge to allow these techniques to be used (FAO 1997). Another concern has been
that scientists are often locked in their disciplines and are poorly prepared for systems
research addressing the many components of sustainable animal breeding programmes
structured to meet the needs of farmers (FAO 1991). The situation, in some cases, is made
worse by the fact that most materials used in training are disciplinarily based and do not
encourage thinking and planning at the systems level. In addition, most of these materials use
examples and case studies from unrelated environments making them irrelevant thus often
confusing the concept or principle being presented. There is, therefore, a need for capacity
building on sustainable use of animal genetic resources to; help scientists build on their
knowledge on current techniques in animal breeding, to ensure that scientists address the
issues of sustainability in animal breeding, and to emphasise the use of relevant training
materials.
The approach that has been taken by ILRI and other organisations in capacity building for
research has been to hold short-term courses for the scientists. The scientists would then go
back to their research stations and use the new techniques. A new model of capacity building
has been adopted by the ILRI/SLU project. It stems from the realisation that most of Africa's
researchers are now being trained at local universities. This model introduces three concepts.
Firstly, it is based on the assumption that training of researchers and university lecturers who
teach at postgraduate level has an amplified impact than training researchers alone. This is
because lecturers, besides carrying out research themselves, also train postgraduate students
(future researchers) who will then use the skills learnt from the lecturers in their own research.
The benefits of capacity building using this model, therefore reach more researchers.
Secondly, the model encourages the use of relevant resources for training of Africa's research
scientists. Such resources can be made available at universities in Africa and overseas
universities that train African students. Thirdly, the model calls for developing partnerships with
universities and NARS. The project is demand driven as it caters for the needs of partners
who also contribute to the output. This paper describes the ILRI-SLU capacity building project
that employs this new model of capacity building.
 Needs assessment activities
The project carried out need-assessment activities to determine the number of universities in
sub-Saharan Africa that teach animal breeding and genetics at the postgraduate level, the
content of the courses taught, human resources and facilities available for running the
courses, constraints faced by universities when delivering the courses and the link between
universities and research institutes carrying out research in animal breeding and genetics.
Questionnaires were sent to fifty-three universities in sub-Saharan Africa by mail. An ILRI/SLU
team followed up by visiting six countries in Eastern and Southern Africa.
Only 23 out of 53 universities responded to the questionnaire and of these only 17 offered
postgraduate training in animal breeding and genetics. The countries visited were Kenya,
Tanzania, Uganda, Lesotho, South Africa and Zimbabwe. The findings of these two activities
were:
The number of staff with postgraduate qualifications was impressive with 74 percent
having PhDs and the rest MSc degrees. In addition, each university produces an
average of six postgraduate students per year. The starting point for the project would,
therefore, not be to give basic postgraduate training in animal breeding and genetics but
to offer a course, which would allow participants to update their skills. Most lecturers
had received little or no training in teaching methods or university teaching. The
postgraduate classes are generally small, an average of six per year, but the same
lecturers teach very large classes at the undergraduate level. Therefore, lecturers need
to be taught teaching methods for both small and large groups.
The postgraduate degree programmes vary a lot in terms of the actual courses offered
and the course content. The courses include statistics/biometry, computer courses,
biochemistry, physiology, production, genetics (basic, population, quantitative,
molecular), and research methodologies. However, little or no time is spent on
conservation of animal genetic resources. Most universities said they needed
strengthening in statistics, basic and molecular genetics, animal genetic resources
(characterisation, conservation, utilisation and management) and sustainable breeding
programmes.
Modern textbooks are often lacking. In addition, books and journals used are those
published in Europe or North America. These resources provide useful information on
principles and concepts of animal breeding and genetics. However, they are often
lacking in examples relevant to sub-Saharan Africa.
Universities had access to computers although the access was rather limited for some -
in numbers and quality of computers. Various statistical software were available.
The availability of molecular genetics laboratory facilities varied - some universities had
none yet some had more than one laboratory in one campus, e.g. one in a faculty of
agriculture and the other in a faculty of science or veterinary medicine.
Most universities do not generate data that can be used in animal breeding and as such
rely on data collected by research institutes or organisations running livestock recording
schemes. There is, therefore, a need to promote good links between research,
extension and universities. The need for collaboration is realised and is practised in
some countries.
Collaboration between universities is limited and is mainly through external examiners.
There is limited exchange of teachers and students between universities. Increased
contact between institutions and countries would reduce the isolation of staff at some
universities. The project should therefore promote collaboration.
Funding support is limited resulting in poor facilities and resources for research and
teaching, heavy teaching loads and high staff turnover.
A workshop, with participants from universities and research institutes in sub-Saharan Africa,
ILRI, SLU, Cornell University and the African Development Bank, was organised to discuss the
findings given above and to plan project activities. The workshop recommended that training
courses for lecturers and researchers to update them on recent techniques and methodologies
and topical issues in animal genetic resources should be held. The production of training
resources to supplement available books and journals was also recommended. The main
emphasis of these resources would be case studies that are relevant to sub-Saharan Africa
and written using information collected by research organisations and extension agencies,
universities and international organisations. Teachers need teaching and learning resources
that are relevant to their country or the country of origin of their students, and which provide
them with the tools to strengthen their teaching by using material that their students can relate
to their present experience and future responsibilities. The participants identified the following
modules on which the training course and training resources would be based:
 Animal genetic resources for sustainable agriculture
Improving our knowledge of tropical indigenous animal genetic resources.
How to make breeding programmes in tropical farming systems sustainable?
Quantitative methods to improve the understanding and utilisation of animal genetic
resources.
Teaching methods and communication.
They also developed the contents of these modules. A third recommendation was the building
of partnerships; among the sub-Saharan universities, universities and research organisations
within the same country, and ILRI and SLU with the universities and research institutes. The
workshop participants agreed to act as a reference group for the project. This reference group
would assist in identifying course participants, identifying scientists who would write the case
studies, and evaluating the training resources. More than one course will be held and
participants for each course will be drawn from a number of countries thus allowing links
between countries. Course participants will also be asked to evaluate the courses and training
resources.
 The project's main activities
Based on the concerns and issues identified by the project through the needs-assessment
activities, ILRI and SLU, together with partners from universities and research institutes in
sub-Saharan Africa, embarked on the project "Capacity Building for Sustainable Use of Animal
Genetic Resources in Developing Countries". The project planned three main activities:
training teachers and researchers on animal genetic resources via a series of courses,
building training resources, and developing partnerships.
 Training courses
The training courses for sub-Saharan Africa will be based on the modules agreed on at the
planning workshop. They will cover topics on: (a) techniques in animal breeding and genetics
and their use in characterisation and conservation of African animal genetic resources, (b) the
importance of indigenous breeds, sustainable breeding programs and the utilisation of
indigenous breeds in such programs, and (c) teaching and supervision of university students.
Participants will get a chance to search available databases including those that are being
developed by the project, use livestock recording and statistical packages, and computer
software used in teaching some animal breeding and genetics principles. They will also
evaluate the training resources produced by the project.
Criteria for selection of course participants were developed at the workshop. The ideal would
be to have two participants per country, one from a research institute and another from a
university. The participants should be people who are likely to work in research or teaching for
a minimum of five years after the course. It was also agreed that a maximum of 25
participants per course would be ideal, making it necessary to hold more than one course in
Africa. The first of these courses was to be a three-week course to be held in November-
December 2000 in Addis Ababa, Ethiopia. Sixteen participants would come from East and
South African countries. Facilitators for the first course would be from SLU and ILRI.
 Training resources
This project will develop training resources to be used by university teachers responsible for
teaching animal breeding and genetics to postgraduate students from sub-Saharan Africa. The
teachers may be in the universities of sub-Saharan Africa, or wherever young professional
from Africa are taught, including the international degree programmes of North America,
Europe and Australia.
The term training resources has different meanings for different users. It is often used in
reference to lists of courses and/or materials on the Internet. This project adopts a more
specific definition of the term training resources, namely: an interactive electronic knowledge
and information package that provides, in an integrated manner, resources for teachers and
trainers on specific subject areas. Because university faculty know the facts and principles of
their subject material, the resources are not an attempt to write an electronic textbook. The
resources being developed are not an attempt to promulgate a fixed and "ideal" curriculum.
Where a university is in the process of developing a new curriculum, the project's training
resources will offer a structure and content that may be helpful. In most situations, however,
we envisage university faculty using the resources to extract and adapt material that will
strengthen existing lectures and practicals, and to provide insights and resources for the
development of new lectures and practicals within existing courses.
The animal genetic training resource CDROM is based on five modules identified at the
workshop. Figure 1 shows a general structure of a training resource. For each module, an
introduction that summarises a set of basic knowledge, principles and concepts was written.
We refer to this material as the "core knowledge". It should be emphasised that the
introductions are a summary - albeit a comprehensive summary - and not an exhaustive
treatment of the subject. The information in the introductions is structured to help engage the
attention and interest of users - both faculty and students.
All the other resources complement the core knowledge in the modules. The opportunity to link
information in a web based product means that these other resources are woven into the core
knowledge, but they can also be accessed directly. The other resources include case studies,
information on distribution of breeds mentioned in the module introductions and maps, images
and video clips, problem solving exercises, bibliographies and full text documents, databases,
glossaries and dictionaries, research methodologies and software.
Figure 1: The general structure of training resource.
The case studies from sub-Saharan Africa will form the most important set of resources in the
training resource. Because they are linked to a particular fact, concept or principle in a
module, they provide users with information that helps to bring reality into what otherwise may
be purely theoretical and factual presentation, concept or principle. Each case study will be a
complete study and story and each is based on a practical situation. Case studies will present
examples where investigations worked and where they did not, and examine why some were
successful and others were not. Where appropriate, each case study will:
improve the understanding of a key fact, principle or concept
achieve that understanding by using practical information relevant to the country or
region of the users, or the people they are training
identify gaps in knowledge
provide guidance on possible research methodologies to address gaps in knowledge.
The case studies will be written by ILRI and SLU staff and scientists working for universities
and research institutes in sub-Saharan Africa. Up to date about 20 case studies have been
written.
The resources are designed to be downloaded by teachers. The teachers can use the
resources as they are or adapt them to be used in their own course work. All the resources
comply with established pedagogic principles and aim to create learning environments. At the
same time, the resources are flexible enough for the teachers to use information to suit their
own situation.
 Partnerships
ILRI developed and maintains a collaborative network of animal genetic resources scientists in
sub-Saharan Africa. This project was planned to be demand-driven hence made extensive
use of existing networks. The project team formed partnerships with research institutes and
universities in sub-Saharan Africa for planning and implementation of the project activities.
The course and training resources are produced in partnerships with colleagues from
universities and research institutes. Based on agreed work plans all partners will contribute to
the content, design and functionality, and testing of the animal genetics training resource and
the course structure.
The primary beneficiaries are university lecturers and research scientists in sub-Saharan
Africa and those teaching international agricultural programmes in universities in developed
countries. As a result of using the training resources, these people will have an impact on
present and future policy makers, socio-economists and agricultural researchers. Animal
breeding/genetic scientists will have access to better materials to strengthen the practical
relevance of their teaching, research and policy making. Secondary beneficiaries include
researchers and development agents in sub-Saharan Africa. This target audience will have a
greater and more relevant understanding of key areas of animal genetic resources in their
own countries. Faculty in developed countries who teach international agricultural courses to
graduate students from sub-Saharan Africa will also benefit from the training resource. The
final and ultimate beneficiaries are the smallholder crop-livestock farmers in developing
countries who will receive new technologies that have an impact on productivity, household
and national economic development and do so in ways that are environmentally sustainable.
 Expected outcomes and project evaluation
 Expected outputs
This project will facilitate the introduction of recent techniques/advances in animal genetics
and breeding into the graduate curricula of African universities. In addition, by introducing
university teachers to teaching aids and methods, it will help improve graduate supervision
and effectiveness of teaching of these subjects, not only at the graduate level, but also at the
undergraduate level. Indeed, this project could serve as a model through which effective
teaching is disseminated throughout entire university systems.
Training resources (both electronic and print) is possibly one of the most significant potential
outputs of this project. As has been alluded to, the resources will aim at providing technical
material which those involved in graduate training can use to make animal breeding and
genetics more interesting to students and relevant in the African context. These resources will
include case studies and numerical examples, illustrated with images, video clips, etc. and
supported by relevant bibliographic information. Moreover, because these resources will be in
electronic form, their continued future updating will be relatively easy. Such updating could
involve inclusion of links to independently developed, and continuously updated, Internet sites.
Not only will provision of these resources to university teachers ensure that the courses are
relevant and that teachers and students are exposed to recent technologies, it will also
improve access (by both teachers and students) to the information they need for their
research.
To the extent that postgraduate training provides the next generation of teachers, researchers
and policy makers, this project will, through a substantial multiplier effect, ensure that capacity
for animal genetic resources research and development is improved in Africa, and that the
level of public awareness on sustainable use of animal genetic resources and related issues is
tremendously enhanced.
 Project evaluation
The Delivery Pathway will be initially through individuals who participate in the training
courses. Two such courses are planned for Africa: one for Eastern and Southern and another
for West and Central Africa. This will be sufficient to cover the majority of universities in sub-
Saharan Africa which offer postgraduate training in animal breeding and genetics.
On return to their respective universities, course participants are expected to share the
information gained from the course with colleagues. Initially, this information will be technical
notes in print form. Recipients will be requested to make this material available to all those
involved in animal breeding/genetics training at the graduate level in their countries. Contacts
for additional information and/or comments will be provided, as will a website address from
where detailed information on the training resources could be obtained (at a later stage, the
entire product will be made available on the Internet).
The evaluation of electronic training resources will be continuous. The first step was to
develop a demonstration CDROM which was evaluated during the planning workshop. The
Beta version of the CDROM will be tested by the participants of the training courses and by the
reference group. Feedback from testing the beta version will be used to improve the
application and version I will be released at the end of 2001 for wider testing by universities in
sub-Saharan Africa.
Evaluation of the impact of the whole project will involve an assessment of users' opinion.
Users will include the following categories: those who will have attended the course; those
who will have attended the course and subsequently used the training resources (electronic or
print) in their teaching/research; those who will have used the training resources in their
teaching/research but without having attended the course; those who will have neither
attended the course nor used the training resources in their teaching/research; and graduate
students who may have been taught or been supervised by individuals in these categories.
The assessment to be conducted starting two years after the first course, will consist of two
activities: implementation of a comprehensive mail survey (using a questionnaire) to a
selected sample of individuals representative of all the above categories; and a workshop with
participants representing these categories. The mail survey and workshop will serve as
mechanisms to assess the value and impact of the project and how the course and training
resources may be improved. The information from both sources will be collected in a way that
allows for sound quantitative analysis. However, the workshop will also provide an opportunity
to discuss emerging issues more exhaustively.
During the implementation of the project, a comprehensive list of animal breeders/geneticists
in sub-Saharan Africa (in universities and research institutes) will be compiled. This will be
used as the basis for the sampling described above. At the same time, this list will facilitate a
realistic assessment of the coverage ("adoption rate") and impact of the project.
 References
Delgardo C., Rosegrant M., Steinfeld H., Ehui S. and Courbois C. 1999. Livestock to 2020:
The next food revolution. The International Food Policy Research Institute (IFPRI), Food and
Agriculture Organisation of the United Nations and the International Livestock Research
Institute. 72 pp.
FAO (Food and Agricultural Organization of the United Nations). 1991. The role of universities
in national agricultural research systems of developing countries: An overview. In: Expert
Consultation on the Role of Universities in National Agricultural Research Systems.
FAO (Food and Agricultural Organization of the United Nations). 1997. Issues and
opportunities for agricultural education and training in the 1990s and beyond. 87 pp.
Pardey P., Roseboom J. and Bientema N.M. 1997. Investments in African agricultural
research. World Development 25 (3): 409– 423.
Wilson T.R., Ehui S. and Mack S. 1995. Livestock development strategies for low income
countries. Proceedings of the Joint FAO/ILRI Roundtable on Livestock Development
Strategies for Low Income Countries. International Livestock Research Institute. Addis Ababa,
Ethiopia. 27 February to 02 March 1995.
Session 5: Short communications
Keeping quality of cow and goat milk under different
temperatures in different seasons in Malawi
P. Mumba, J.W. Banda and S.B. Msowoya
University of Malawi, Bunda College of Agriculture, P.O. Box 219, Lilongwe, Malawi.
E-mail: ansccof@sdnp.org.mw
Introduction
Objectives
Materials and methods
Results and discussion
Conclusion
 Introduction
Malawi is in the tropics and the temperatures range from 10 oC in the cold season which lasts
2–3 months to 30 oC in the hot season, which lasts up to 9 months. Milk is produced by the
smallholder farmers and to some extent, the large scale farmers. Smallholder farmers take
their milk to the milk bulking places from where the processing companies buy and take it to
the plants. With the prevailing high temperatures, the length of time it takes for the milk to be
transported from the farmer to the plant is important in connection with the keeping quality of
the milk. Determination of the minimum time the milk can stay outside the plant is therefore,
important both to the farmer and the processing company for maximum profit. Since most
peasant farmers do not own refrigerators, such information would be to their advantage.
 Objectives
The objective of this work was, therefore, to determine and compare the keeping quality of
milk from cows and goats under the prevailing temperatures and seasons.
 Materials and methods
The milk from each species was divided into two portions. One portion was heated to 65 oC for
30 minutes and allowed to cool under refrigeration temperature. The raw milk portion was put
in different volumetric flasks covered with rubber stoppers and exposed to temperature ranges
of 5–7 oC (using a refrigerator), 12–15 oC (using a charcoal cooler), 35–37 oC (using a water
bath) and to room temperature (18–20 oC (cool dry season) or 22–25 oC (hot dry season)) in
two seasons (cool dry season and hot dry season). Equal amounts of heated milk were put in
different volumetric flasks and exposed to the same ranges of temperature. The Resazurin
test was used to determine the keeping quality.
 Results and discussion
The overall average keeping quality of milk (hours) by animal species, season, processing and
temperature are shown in Table 1. There were significant differences (P<0.05) in the keeping
quality of cow (46.0 + 1.02 hours) and goat milk (35.5 +1.02 hours). Milk in the cool dry
season kept longer (P<0.05) than in the hot dry season. Heated milk also stayed longer than
unheated milk. The milk kept in the charcoal cooler stayed longer (P<0.05) than that kept at
room temperature in the cool dry season but there was no difference (P > 0.05) with that kept
at room temperature in the hot dry season.
Irrespective of the process (heated or not) goat milk under temperature ranges of 5–7 oC, 12–
15 oC, 18–20 oC and 35–37 oC kept good on average, for 88 + 2.0 , 30 + 2.0, 36 + 2.0 and 10
+ 2.0 hours in the cool dry season while in the hot dry season the averages were 76, + 2.0, 24
+ 2.0 , 14 + 2.0 and 6 + 2.0 hours, respectively (Table 2). Cow milk exposed to the same
temperature ranges kept good for 112 + 2.0, 36 + 2.0, 40 + 2.0 and 10 + 2.0 hours in the cool
dry season and 105 + 2.0, 34 + 2.0, 22 + 2.0 and 8 + 2.0 hours respectively, in the hot dry
season. For each animal species and in each season, the values were significantly different
from each other (P< 0.05).
Table 1: Keeping quality of milk (hours) separated by type of milk test.
Main Class Number of observations LS Means(Hours)
Sub-class
Overall Species 384 40.75
Cow 192 46.0 a
Goat 192 35.5 b
Season
Cool dry 192 45.3 a
Hot dry 192 36.2 b
Processing
Heated 192 51.8 a
Unheated 192 29.7 b
Temperature (oC)
5– 7 76 97.0 a
12–15 77 35.3 b
18– 20 77 32.5 b
22–25 77 28.0 c
35– 37 77 11.0 d
Means within variable groups with different superscripts differ significantly (P< 0.05).
Table 2: The keeping quality of cow and goat milk (hours) at different temperatures and
seasons
Season Animal species Temperature
  5–7 oC 12–15 oC 18–20 oC 35–37 oC
1 Goat 88+ 2.0a 36 +2.0a 30+2.0a 10 +2.0a
 Cow 112 +2.0b 40+2.0b 36+2.0b 10+2.0a
2 Goat 76+2.0a 24 +2.0a 14 +2.0*a 6+2.0a
 Cow 105+2.0b 34+2.0b 22 +2.0*b 8 +2.0a
* Room temperature of 22-25 oC
Means with the same letter in a column and at each season are not significantly different at P =
0.001
Regardless of temperature, heated cow and goat milk stayed without spoiling for up to
62+2.03 hours and 49 + 2.03 hours respectively, in the cool dry season while unheated
samples stayed for 37+2.03 hours and 33 + 2.03 hours respectively, in the same season.
Each set of values within a season, varied significantly from each other (P<0.001). In the hot
dry season, heated cow and goat milk kept good for up to 56 + 2.03 hours and 40 + 2.03
hours, respectively, while the unheated samples kept for only 29 + 2.03 hours and 20 + 2.03
hours respectively
 Conclusion
The average temperatures in Malawi range from 10 oC in the cold dry season to about 35 oC
in the hot dry season. The results therefore suggest that in cool dry season small holder
farmers should keep their goat and cow milk for not more than 30 hours and 36 hours,
respectively, if kept at room temperature. However, if the charcoal cooler is available, then the
farmer can keep his milk for a little longer time. The charcoal cooler is relatively easy to build
and smallholder farmers should be encouraged to have one. This may help reduce loss of
their milk and maximise their profit. In the hot dry season, goat and cow milk should only be
stored at room temperature for up to 14 hours and 22 hours, respectively. If left too long in the
heat, both types of milk may be spoiled within 6–8 hours. The charcoal cooler may also greatly
prolong the keeping quality of the milk during this season. However, for a longer shelf life,
milking should also be done in a clean environment to reduce contamination and early
spoilage. Special hygienic care should be taken to reduce contamination of goat milk during
milking.
Prospects for the 2000/2001 rainfall season in Malawi
Press Release Meteorological Department
Ministry of Transport and Public Works
The La Nina phenomenon in the equatorial Pacific Ocean, which dominated the past two
rainfall seasons, 1998/99 and 1999/2000, is observed to have died down. Sea surface
temperatures in the same area are currently near average and are projected to remain around
normal during the next six months, implying that the ocean-atmosphere interaction will be
normal (no El Nino and no La Nina). Hence, the 2000/2001-rainfall season in Malawi, which
falls within these six months, will not be affected by either El Nino or La Nina implying a
normal season. Typical normal seasons in recent years include 1989/90, 1978/79, 1960/61
and 1959/60. However, of these seasons, 1989/90 is the only season that precedes two
strong consecutive La Nina episodes.
Analysis of these four normal seasons in Malawi indicate the following:
The onset of the main rains was normal, relatively uniform across the country and
occurred during the second ten days of November.
The season progressed normally with the usual mid season dry spells.
The cessation of the rains was normal; towards the end of March over the southern half
of Malawi stretching into April over the northern half.
The seasonal rainfall totals were normal in most areas.
Additional national analyses utilising other indicators have also been done to describe the
prospects of Malawi's 2000/2001 rainfall as detailed below:
Inland surface heating over the country, which precedes the onset of the rainfall season,
was delayed up to August. However, from September to-date warming has been
observed. Normally, a continued warming trend often leads to favourable onset of the
rains.
The Inter-Tropical Convergence Zone (ITCZ), which is the main rain bearing system
over Malawi, is currently close to its normal position around the equator, slowly
progressing southwards.
Current atmospheric wind and pressure patterns favour the normal onset of rainfall over
the country. Pre-season rains, locally known as chizimalupsya, are being experienced in
some parts of the country.
During a normal season tropical cyclones develop over the Indian Ocean, which have
either adverse or favourable effects on Malawi rainfall depending on their position in the
Mozambican channel.
Based on the above analyses, the forecast for the 2000/2001 rainfall season in Malawi is:
normal rainfall across the country during the period October to March. However, pockets of dry
areas associated with a normal rainfall season should be expected.
The above forecast has been produced with additional inputs from Southern African Climate
Outlook Forum (SARCOF) that met in Gaborone, Botswana in September 2000 and other
international climate forecasting institutions.
The effect of variety and geographical location on the mineral
content of sorghum and finger millet in Kenya and Uganda
Winchester M.D Mvula1 and A. O. Makokha2
1University of Malawi, Bunda College of Agriculture, Box 219, Lilongwe, Malawi
2Jomo Kenyatta University of Agriculture and Technology, Box 62000, Nairobi, Kenya
Abstract
Introduction
Materials and methods
Results and discussion
Conclusions and recommendations
Acknowledgement
References
 Abstract
Sorghum S.bicolor and finger millet eleusine coracan grain are widely consumed in East
Africa. There are very many varieties of these grains and these grains differ widely in their
physical and chemical composition. This study determined the influence of variety and
geographical location on the mineral content of six varieties of sorghum grains grown within
three different geographical locations and two similar varieties of finger millet grown within two
different locations. The minerals were determined using an atomic absorption
spectrophotometer. In the sorghum grain the iron, calcium, zinc, magnesium and manganese
content ranged from 8.7 – 15.3, 3.3 – 6.7, 4.2 – 6.8, 15.5–- 16.2 and 2.9 – 4.3 mg/100g,
respectively. In finger millet grain the iron, calcium, zinc, magnesium, and manganese content
ranged from 10.3 – 14,2, 60.7–- 85.6, 5 – 5.9, 16.5 – 34.3 and 19.2–- 24.1 mg/100 g
respectively. There were significant differences due to location in iron content of all finger
millet varieties. However, the differences in mineral content for sorghum grain did not correlate
for either variety or geographical location.
 Introduction
Sorghum and finger millet grains are grown in semi areas of the tropics and sub- tropics. Many
types of sorghum grown traditionally are photosensitive. Between 1930 and 1950, local
sorghums were first collected in Kenya, Uganda and Tanzania in East Africa. The crop was
considered to be of no importance there until about 1975 (Doggett 1988). Finger millet and
sorghum are widely consumed in East, Central and Southern Africa. Sorghum and finger millet
are used in the form of "ugali", "uji" or porridge and also traditional beer. Finger millet and
sorghum are important sources of carbohydrate, protein and minerals. However, there are
major variations in chemical and physical composition. In addition, there are major variations
in mineral content among sorghum and finger millet grain world-wide. There is little
information on this variation in mineral content within East Africa. The objective of this study
therefore is to determine total minerals in sorghum and finger millet grain grown in different
locations within Kenya and Uganda and also to investigate the effect of variety and
geographical locations on the mineral content of sorghum and finger millet grain grown in
Kenya and Uganda.
 Materials and methods
Finger millet and sorghum were collected from Serere in Uganda, Katumani in West Kenya,
and Alupe in East Kenya. A pair of similar varieties was obtained from two different
geographical locals. Six varieties of sorghum were from three different geographical locations
and two varieties of finger millet from two different geographical locations. Moisture was
determined by placing samples in an oven at 105ºC for three hours and crude ash was
determined by heating the samples in a muffle furnace at 550ºC for 18 hours. Minerals were
analysed by using an atomic absorption spectrophotometer (AAS; procedures are outlined by
Lick et al. 1979; Khetarpaul and Chauhan 1990).
 Results and discussion
Table 1. Mineral content in sorghum and finger millet grain (mg/100gm).
Variety DM Zinc Mn Fe Ca mg
KARI Mtama 1 Alupe Sorghum 92.2 4.6 3.3 10.1 6.1 15.8
KARI Mtama 1 Katumani Sorghum 94.9 4.2 2.9 8.7 4.9 15.5
Seredo Alupe Sorghum 90.2 5.7 4.3 10.2 3.8 16.2
Seredo Katumani Sorghum 91.2 5.3 4.0 9.2 4.0 15.8
IS 9646 Katumani Sorghum 91.7 4.8 3.4 9.1 5.2 15.8
IS 9646 Alupe Sorghum 92.1 6.8 4.2 15.3 3.2 15.6
Pes Serere Finger millet 91.8 5.0 24.1 14.2 85.8 34.3
Pes Alupe Finger millet 89.1 5.9 19.2 10.3 60.7 16.5
Dry matter (DM) is more than 90% in all cases, therefore within acceptable moisture in all
cases.
In this study, there was no significant difference in iron content among sorghum varieties
(mean 10.4 ± 2.3). In addition, there was no significant difference in iron among geographical
locations except between IS9646 Alupe and Katumani, where there is a difference at the 5%
level. (compared the means using the analysis of variance). For finger millet, Pes Serere had
significantly higher iron than Pes Alupe. Therefore, there was no significant difference due to
geographical location. All the iron content figures are within the range from earlier work
(Doggett 1988).
KMI has significantly higher Calcium (Ca) than Serere and IS9646. There was no significant
difference between IS 9646 and Seredo in Ca content. Finger millet has significantly higher
Ca content than sorghum grain. Regarding geographical location, IS 9646 grown in Alupe has
significantly higher calcium than that grown in Katumani. In addition, finger millet, Pes Serere,
grown in Serere has significantly higher Ca content than the one grown in Alupe. KM1 grown
in either Alupe or Katumani and Seredo grown in the same two locations did not show
significant differences in Ca content.
Agricultural management of sorghum and finger millet grain at the three research stations is
similar. No fertilisers were added.
Table 2a: Differences in mineral content among Sorghum grain
Mineral Variety Location
Iron NS SD
Calcium SD SD
Zinc NS NS
Magnesium NS NS
Manganese SD NS
SD: Significant difference at 5% level
NS: No significant differences.
Table 2b: Differences in mineral content in finger millet due to geographical locations
Iron SD
Calcium SD
Zinc SD
Magnesium SD
Manganese SD
 Conclusions and recommendations
This study shows that there was no consistent effect of variety or geographical location on the
mineral content of sorghum and finger millet. More work on mineral determination needs to be
done with sorghum and finger millet at more geographical locations.
 Acknowledgement
Thanks to the Japanese International Corporation Agency (JICA) for providing the financial
assistance and Jomo Kenyatta University of Agriculture and Technology
for providing the facilities.
 References
Lick et al., 1979; Khetapaul and Chauhan, 1990, AOAC, 1984
Milk yield, physico-chemical properties and composition of
milk from indigenous Malawi goats and their Saanen half-
breeds
P.P. Mumba, Banda J.W., Nyoni C.C. and Kaliwo G.
University of Malawi, Bunda College of Agriculture, P.O. Box 219, Lilongwe, Malawi
E-mail: Anscconf@sdnp.org.mw
Objectives
Materials and methods
Results and discussion
Conclusion
References
 Objectives
The objective of this work was to compare the milk yields, some physico-chemical properties
and the chemical composition of milk from the indigenous Malawi goats and their Saanen half-
breeds.
 Materials and methods
The physico-chemical properties determined were pH and specific gravity. The constituents
analysed were total solids, fat, crude protein (N x 6.38), lactose, calcium (Ca), magnesium
(Mg), sodium (Na), potassium (K) and chloride (Cl).
 Results and discussion
The mean yields, pH and specific gravity from both local and crossbred goats are given in
Table 1. The crossbred goats produced significantly higher milk yields (p<0.01) than the local
Malawi goat. This means that crossing the Malawi local goat with the Saanens would improve
milk production by almost 300%. Both the specific gravity and the pH were not significantly
different in the local and crossbred goats (p>0.01). It has been reported that the specific
gravity of milk is a function of the fat content and the total solids in the sample. In this
experiment, both the fat and the total solids contents in the two breeds (Table 2) were not
significantly different and nor was the specific gravity. The pH values of the milk from both
breeds were slightly lower than the range of values (6.3–6.7) reported by other researchers.
Table 1. Milk yields (Kg), pH, and specific gravity of milk from indigenous Malawi (LL) and
half-bred Saanen (SL) goats during an eight-week period
Breed Milk yield pH Specific gravity
 Mean Sd Mean Sd Mean Sd
LL a 4.79 5.196 0.02 1.033 0.01
37.1
SL 102.0b 11.21 5.618 0.03 1.046 0.01
Means with different letters in a column are significantly different (p<0.01)
The concentration of the milk components of the locals and their Saanen crosses are given in
Table 2. The fat content of the local goat milk (6.72 + 0.25%) was not significantly different
from that of the crosses (6.08 + 0.08%). The values obtained were in the range of values
reported for the African dwarf goat (Jenness 1980). However, these values were higher than
those obtained elsewhere. Peacock (1996) reported that tropical goats have fat contents
ranging from 3–6%. The goats used by Braunsdorf (1966) and Dozet (1973) were of different
breeds and this could account for the observed differences.
Table 2: Average milk composition of the indigenous Malawi goat (LL) and half-bred Saanen
goats (SL)
Component Breed
 LL SL
 Mean Sd Mean Sd
Fat (%) 6.7 0.25 6.1 0.08
Crude protein (%) 5.2 0.29 4.4 0.20
Lactose (%) 4.7b 0.22 5.3a 0.25
Ca (mg/l) 74.4b 4.29 134.3a 12.51
Mg (mg/l) 58.0b 4.44 98.6 10.88
Na (mg/l) 55.9b 5.19 129.7a 16.58
K (mg/l) 82.7b 11.0 132.9a 17.36
Cl (mg/100ml) 92.1 0.54 132.9a 0.26
Total solids (% w/w) 13.8 0.72 13.5 0.42
Means with different letters in a row are significantly different (p<0.01)
The protein contents were also not significantly different in the local and cross-bred goats.
Again, the variations from other goat milks can be accounted for by the different genetic
constitutions. The lactose content in the local goat milk (4.7+ 0.22%) was significantly lower
(p<0.01) than that of the crossbred goats (5.3 + 0.25%). The observed values were higher
than those reported for other breeds such as the pure Saanens, but were closer to the values
of dwarf goats, which produce higher lactose than the other breeds.
The content of all determined minerals of the crossbred goats were in general higher (p<0.01)
than those of the local goat. All minerals in the local goat milk were much lower than the
values reported by other researchers. Most notable was the chloride content which was lower
in both the local and crossbred goats as compared to literature values. For example, French
(1970) reported that the chloride content is in the range of 157.8–181.1 mg/ml.
The total dissolved solids were not significantly different (p>0.01) in the local goat milk and in
the crossbred goats but were in the range of values reported by other researchers. Banda
(1992), who used the local Malawi goats and their Boer crosses, observed much higher total
solids values than in the current studies.
 Conclusion
Milk yield from crossbred goats was higher than that from the indigenous Malawi goats. This
means that the indigenous Malawi goat can be improved by crossing it with the exotic dairy
goat. There were no significant differences in fat, crude protein and total solids contents in milk
from indigenous and from the crosses. However, there were differences in the contents of
lactose and the trace elements, which were higher in the milk from crosses than in the
indigenous breed. The high levels of nutrients in the goat milk are indicative of the potential to
improve the diets of the rural Malawians.
 References
Banda J.W. 1992. Genotypic and seasonal influence on milk yield and milk consumption of
sheep and goats in Malawi. PhD thesis, University of Giessen. Germany.
Braunsdorf K. 1966. A report on an investigation of goat's milk. Dairy Science. Abstract.
28:1337.
Dozet N. 1973. Composition and nutritional value of goat's milk and goat's milk Products.
Dairy Science Abstract 35:3147
French M.H. 1970. Observation on the goat. FAO Agricultural Series No. 80. Food and
Agricultural Organisation (FAO). FAO, Rome, Italy. 299 pp.
Jenness R. 1980. Composition and Characteristics of goat milk: Review. Journal of Dairy
Science 63:1605–1629.
Session 6: Abstracts
Genetic diversity in indigenous cattle breeds of East Africa
P. S. Gwakisa
Faculty of Veterinary Medicine, Sokoine University of Agriculture, Box 3019, Morogoro,
Tanzania, Correspondence: Email - gwakisa@suanet.ac.tz
 Abstract
The cattle of East Africa are generally adapted to the environment in which they occur with
resistance to many parasites and to extremes of climate. These cattle are however,
characterised with low productivity. When the more productive European breeds are
introduced to Africa, they succumb to the local conditions. Treating these latter cattle is
expensive and negates many of the advantages due to their introduction. The options to
breeders are either to improve the local breeds via crossbreeding to the European cattle or to
improve the local breeds by selection. Improvement of African breeds requires their
preliminary characterisation, in which regard, evaluation of the level of genetic divergence
among them is particularly important. By knowing the level of divergence between local
breeds, appropriate choices can be made for selection, improvement and conservation of
genetic material. We have used two molecular genetic techniques to compare the extent of
genetic diversity among East African local Bos indicus breeds and compared these to a distant
West African N'dama breed and improved Bos indicus breeds in Australia. Polymorphisms
were generated by PCR analysis of random amplified polymorphic DNA (RAPD) and a variety
of microsatellite markers. Similarity indices and genetic distances indicated different extents of
divergence between the East African Zebu. Inclusion of distant West African N'dama and
Australian Bos indicus breeds in the genetic comparisons, enabled a clearer indication of the
hierarchy of genetic relationships between the East African breeds. The significance and
implications of this genetic diversity are discussed in the context of the merits carried by these
breeds.
Keywords: Genetic diversity, Cattle, DNA markers, RAPD, Microsatellites
Carcass composition of tethered, supplemented Tanzanian
goats and prediction models for carcass weight and
composition
Mtenga, L.A., Kakengi, A.M.V., Kimambo, A.E. and Mgheni, D.M.
Department of Animal Science and Production, Box 3004, Morogoro, Tanzania
 Abstract
A study was conducted to evaluate the effect of supplementing two strains of tethered small
East African goats on growth rate, killing out characteristics, carcass composition and
development of prediction equations for carcass weight using linear body weights. Sixteen
male goats of 9 to 12 months of age were used. Animals were sorted by their weight and
strain and then randomly allocated to tow treatments. The treatments were tethering with or
without supplementation. All animals were tethered for 8 hours for a period of 84 days.
Animals were slaughtered and carcass killing out percentage and composition were
determined. Pooled prediction equations for carcass weight and composition were developed.
Supplementation significantly (P<0.05) influenced the empty body weight and carcass weight
when expressed as a percentage of slaughter weight. Treatment effect was observed also for
the edible non carcass components such as gastro intestinal tract, pluck, kidney and gut fat
where it was higher for supplemented than unsupplemented goats albeit the differences were
merely numerical (P<0.05). Heart girth was an independent variable that could predict the
carcass weight of the goats better (P<0.01), r = 0.78). Of the three independent variables -
slaughter, carcass and empty body weight, - carcass weight was the best fit model for
prediction of lean (P<001, r = 0.87) and bone (P<0.01, r + 0.71) whereas empty body weight
was a better predictor of fat (P<0.05< r = 0.32) in carcass tissues. It was generally concluded
that tethering and supplementation could be a better grazing management for improvement of
local goat meat production in areas under scarcity of land. It was further concluded that the
developed carcass weight and composition models could be of use to farmers as tools that
can determine the time and price for disposing their goats.
Post-weaning performance of local Malawian pigs fed
soyabean-based diets containing different levels of raw,
untreated sunflower seed meal
R. K. D. Phoya, Y. N. Vibetti and J. B. Jonazi
 Abstract
The effects of feeding soyabean-based diets containing different levels of raw, unextracted
sunflower seed meal on the post-weaning performance of local Malawian pigs were
investigated. The treatments were 0 (T1), 10 (T2), 20 (T3) and 30% (T4) sunflower seed meal
- containing starter diets formulated to provide 3320 Kcal digestible energy/kg and 22% crude
protein for a period of 42 days. Then the pigs were changed to grower diets formulated to
provide 3040 Kcal digestible energy/kg and 18% crude protein such that animals initially on 0,
10, 20, 30% sunflower seed meal diets were put on 0, 20, 30 and 40% sunflower respectively
and only removed from the trail in the week they approached 45 kg.
Average daily gain (ADG) and the Kleiber ratio were determined during the starter period.
There were no significant (P > 0.05) differences in ADG between treatments 1, 2 and 3, but
treatment 4 was significantly different from treatment 2 at the 5% level. There were also no
significant differences in the Kleiber ratio between T1, T2, and T3, but T4 was significantly
different from T1 and T2 at 5%. ADG values in kg were 0.38, 0.40, 0.36 and 0.33; and for the
Kleiber ratio, the values: 0.034, 0.035, 0.034 and 0.032, for treatments T1, T2, T3 and T4
respectively.
During the grower period, the average daily gain values in kg were 0.43, 0.47, 0.45 and 0.40,
and for the Kleiber ration, the values were 0.031, 0.032, 0.032 and 0.030, and for treatments
T1, T2, T3 and T4 respectively. There were no significant (P>0.05) differences in the response
variables between the treatments during the growing period.
Piglets on higher sunflower seed meal-containing starter diets (T3 and T4) showed reduced
performance compared with their contemporaries on lower sunflower seed meal, although this
difference became non- significant ( P >0.05) during the growing phase when the pigs were
maintained on consistently higher sunflower seed meal diets. There were no significant
differences in the carcass characteristics between the treatments. It is suggested that in local
pigs, sunflower seed meal can be included up to 20 and 30% in starter and grower diets
respectively, without reducing average daily gain or increasing the backfat thickness in the
carcass.
Using carbon isotope ratios (∂13C) in determining diet changes
in animals: The case of the Lake Malawi bathyclarias
nyassensis (pisces: claridae)
E. Kaunda1 and T. Hecht2
1 Dept of Aquaculture and Fisheries Science, University of Malawi, Bunda College of
Agriculture: email: ekaunda@yahoo.com
2 Dept of Ichthyology and Fisheries Science, Rhodes University, Grahamstown, RSA
 Abstract
The isotopic composition of the body of an animal reflects the isotopic composition of its diet,
although the animal is on average enriched in ∂13C by about 1 % relative to the diet. We
examined isotopic composition of different size groups of Bathyclarias nyassensis ranging
from 37 to 73 cm total length (TL) and of its principal prey items (fish and zooplankton). The
∂13C values showed ontogenetic dietary shifts from "light" or more negative (approximately –
21 %) in small fish to "heavy" or "less negative" (–19.5 %) in medium size fish and to light or
more negative (–20.7 %) in large fish. The dietary ontogenetic changes were supported by two
other independent methods: morphology and gut content analysis.
Morphologically, filtering area was determined on 18 fish of varying sizes, and the area
changed when fish reached approximately 50 cm TL suggesting changes in fish food habits
and/or diet. Detailed gut analysis showed changes in diet from piscivory to predominantly
zooplanktivory at the same fish size (50 cm TL).
These results indicate that while limits of accuracy of using ∂13C generally restrict its
application to situations in which diet is derived from sources with relatively large differences in
their ∂13C values, such as terrestrial Vs aquatic organisms or C3 Vs C4 plants, the method can
be relied upon when other methods are available to validate the findings.
Characterisation of the local Zambian chicken — A case study
of Chikondwa area in Chongwe
Malwa-Mushota, N.M; Mfitilodze, M.W.; 
Wollny, C.B.A; Gondwe, T.N; Banda-Nyirenda, D
 Abstract
A study was conducted to characterise the local Zambian chicken in the Chikondwa area of
Chongwe district from February to July, 2000. A preliminary study involved using a
questionnaire administered to 78 households to determine demography of the farmers and
evaluate village chicken production, health, management and their importance in households.
Each household had an average of 6.29 ± 3.22 persons, the majority being women. The level
of education of most of these was not beyond primary school. Most farmers (50 %) kept 1 to
10 chickens, and a few (37.7%) kept between 11 and 20. More men (48%) claimed ownership
of the chickens than women in the household (41%). The majority (65.4%) of the farmers
acquired chickens from within the same village, 10.3% from the neighbouring village and only
5.1% from the market. The local chickens were kept mainly for consumption and as a source
of income for the rural people. Chickens were kept on free range with occasional
supplementation of maize grains and kitchen refuse. They were housed only at night in
kitchen houses (51.3%), traditional poultry houses (9%), battery cages (2.6%) and trees
(32.1%). The majority (97%) of the villagers did not give any health care to their chickens. The
birds reached table weight at 4.16 months and age at sexual maturity was attained at 5.72
months. Egg production reduced in the wet season (January to March). The study confirmed
that village chickens are the most common species of livestock kept by the rural Zambian.
There were variations in both colour and types of these chickens, which include types like the
frizzle, naked neck, dwarf and the spotted.
Key words: Zambian local chicken, management, and variation
The prevalence, incidence and effects of trypanosomosis on
reproductive performance of cattle in north-eastern Zimbabwe
N.T. Kusina,1a P. Muzavazi2 and P. den Bossche3
1Department of Animal Science, University of Zimbabwe, Harare, Zimbabwe
2Faculty of Veterinary Studies, University of Zimbabwe, Harare, Zimbabwe
3Regional Tsetse and Trypanosomosis Control Programme (RTTCP), PO Box A560,
Avondale, Harare, Zimbabwe 
 Abstract
A study was conducted in three adjacent zones stretching for a distance of approximately 60
km from the Mozambican border into Zimbabwe to investigate the prevalence and incidence of
trypanosomosis in cattle. The second objective was to determine the effects of the
trypanosome challenge on reproductive performance of the cattle. Each zone was 20km wide.
The area had been previously cleared of tsetse fly and any re-invasion was controlled by an
eight-kilometre wide odour-baited insecticide treated target barrier. The trypanosome
challenge was evaluated using serological and parasitological analysis of blood samples taken
from sentinel herds of approximately 60 adult Mashona cows in each zone. Blood samples
were collected at monthly intervals from March to August 1999. Results showed that the
incidence of trypanosomosis was low with an average of 3% of the cattle infected. The level of
infection decreased with increasing distance from the Mozambican border and infection was
higher during the rainy season than the dry season. Trypanosome infection significantly (P <
0.05) reduced the packed cell volume. However, the reproductive performance of the cattle
herds was not affected (P > 0.05) by the trypanosome challenge. In addition, abortion and
mortality were similar among the cows in the three zones. It is therefore concluded that the
current level of trypanosome challenge did not affect reproduction and productivity of cattle
reared in the 60km zone close to Mozambique. These results suggest that the target barrier
that was erected was effective in controlling tsetse fly from re-invading the area.
Key words: Tsetse fly, trypanosomosis, cattle, reproductive performance
Constraints to rural poultry production and diversity in
Lilongwe and Mzuzu agricultural development divisions,
Malawi
T.N.P. Gondwe1, F.C. Chilera1 and A.J.D. Ambali2
1Department of Animal Science, Bunda College of Agriculture, University of Malawi, P.O. Box
219, Lilongwe.
2Biology Department, Chancellor College, University of Malawi, P.O. Box 280, Zomba.
 Abstract
Rural poultry constitute over 80% of total poultry population in Malawi. Smallholder farmers
keep a diversity of species on free-range system. Chickens (mainly indigenous breed)
comprise 83%, pigeons 14%, ducks 2% with a small proportion of turkeys, geese and guinea
fowl. Productivity of rural poultry is generally low. In a study to evaluate biodiversity in rural
poultry, specific and general constraints were observed from a random sample of 323 farmers
who kept one form of poultry in Lilongwe (170) and Mzimba (153) Districts of Lilongwe and
Mzuzu Agricultural Development Divisions, respectively. Some of the constraints were of
threat to phenotypic and genetic diversity that exists in rural poultry. The study mainly
concentrated on chickens, ducks and pigeons whose constraints are discussed in the paper.
Health constraints comprised of high prevalence of diseases and parasites (more in chickens
and ducks than in pigeons), predators and bad weather conditions. Bad weather was
associated with poor housing structures. Traditional houses (kholas) were used for all species
in Mzimba while in Lilongwe, traditional houses were used for ducks and pigeons. Chickens
were housed in human dwelling units for fear of theft that was more in chickens than in ducks
and pigeons. These traditional houses were however, not protecting birds from night
predators, rain and cold weather conditions. There was minimal care taken in chicken houses
and duck and pigeon houses were not attended to. Human dwelling units in Lilongwe were
small and with poor ventilation, creating possible chances of disease transmission.
There was haphazard breeding in rural poultry and farmers did not keep records. Most
farmers exchanged their breed stock among themselves within the villages. Less than 40% of
breed stock was acquired from outside the villages. The breeding structure and the exchange
of breed stock are possible causes of inbreeding and genetic drift, considering the small
populations that exist in the villages.
There was poor health and extension services for poultry, which was non existent for ducks
and pigeons. Duck production and consumption was also affected by cultural and religious
taboos, and the preference of farmers of chickens to ducks. These also affected the diversity
of species among flocks. Rural poultry production is mainly for subsistence, and hence was
under the responsibility of women in households who were unfortunately more illiterate
(average of standard 3) than men (average of standard 5). The husbandry care for poultry
was therefore very minimal.
The study identified constraints in rural poultry relating to housing, diseases, management and
socio-cultural values. These constraints affected production and species diversity in rural
poultry. There is therefore a need to develop species-specific and general interventions to
improve rural poultry production and sustain genetic diversity.
Key words: constraints, production, diversity, and rural poultry
Village chicken production in Zimbabwe — Constraints and
potential
J. Kusina1
Department of Animal Science, University of Zimbabwe,
P O Box MP167, Mt. Pleasant, Harare, Zimbabwe
E- mail address: Kusina@ansci.uz.zw
 Abstract
Poultry production is widespread in Zimbabwe, comprising large-scale, small-scale
commercial and extensive poultry production units. Large-scale systems are characterised by
huge investments, intensive management and a high degree of mechanisation and
specialisation. These systems, because of the nature of the inputs required, are confined to a
few wealthy companies or individuals. Small-scale systems are found in both peri-urban and
communal areas. Exotic or improved breeds are used in commercial poultry production
systems. Extensive systems are found mainly in the communal areas, with almost every
household owning some chickens. These systems are dominated by poultry breeds that are
not classified into specific breeds, but are possibly crosses of imported and indigenous breeds
or indigenous breeds depending on the area and mainly scavenge for their daily nutrient
requirements. In Zimbabwe, village chickens are estimated to be between 15 and 30 million.
The estimate is based on about one million communal farmers, with an average of 20
birds/flock. Village chicken production has withstood the test of time and is often described as
a low input-low output production system. This makes it a very sustainable system for the
resource poor, communal area farmers. Needless to say, because of this attitude, very little
attention has been paid to this industry in terms of research and development, training and
marketing, resulting in the lack of information on the status of poultry production in the
communal areas. This lack of attention has led to many constraints in terms of improving
productivity of village chickens, not only in Zimbabwe, but also in other African and developing
countries. Despite the lack of attention and constraints, there is potential to increase village
poultry production. Village chickens or indigenous chickens (where they still exist), are an
important reservoir of germplasm that may be used in the future to improve hybrid birds to
better adapt to the environments where they are kept. Village chickens are also relatively
cheap to purchase, feed and house. In addition, they are an ideal source of high quality
protein (both meat and eggs), income and manure fertiliser for economically disadvantage
people, the majority of whom live in the rural areas of Zimbabwe and Africa. The constraints
and potential to improving village and small-scale chicken production will be reviewed in this
paper.
Key words: village poultry, production, constraints, protein, income
Some socio-economic factors constraining farm animal
production in Malawi
J.P.Simpson
University of Malawi, Bunda College of Agriculture, PO Box 219, Lilongwe, Malawi
 Abstract
By international standards, there is a fundamental lack of interest in stockmanship in Malawi.
Livestock, specifically cattle, has been traditionally viewed as wealth-on-the-hoof, a security
fund, with culling and marketing as alien concepts. Since time immemorial, land was owned
communally, not individually. Additionally, heavy and rapidly increasing demographic
pressures have created a land shortage. Matrilineal descent patterns have militated against
the development of breeding herds and flocks. A bewildering array of virulent diseases has
caused enormous livestock losses and official efforts to control or contain these ravages were
largely undone by non-compliance. Consequently, both the colonial administration and the
newly independent republic were confirmed in their views that `corn' and not `horn' was the
way forward for national prosperity. Finally, a lack of interest in measuring and recording, so
essential for successful livestock development, remains one of the differences between the
First and Third Worlds.
Use of maize and crotalaria leaves as supplements for
tethered sheep and goats
S.K Chikagwa-Malunga1, J.A Jere1 and J.W Banda2
1Chitedze Agricultural Research Station, Lilongwe, Malawi.
2Bunda College of Agriculture, P.O. Box 219, Lilongwe, Malawi.
 Abstract
Two experiments were conducted to evaluate the effect of supplementation on the
performance of local versus Dorper x local tethered sheep (average 33.0 kg wt) with lambs
(average 11.37 kg wt) in Experiment 1, and local versus Boer x local goats (average 25.8 kg
wt) with kids (average 6.07 kg wt) in Experiment 2 at Chitedze Agricultural Research Station.
In both experiments, a mixture of maize leaves and crotalaria (3:1, 500 g/animal/day) was
compared with maize leaves (500 g/animal/day) or no supplementation. Feed intake, live
weight change, mortality and milk intake were used to evaluate dam and offspring
performance. Maize leaves + crotalaria leaves (MC) had a higher N content than maize leaves
(M) alone. There was a significant difference (P<0.05) between supplementation and no
supplementation. In Experiment 1, sheep (dams and lambs) supplemented with MC gained
more (P<0.05) weight per day compared with those that received M only. However, there was
no significant difference (P>0.05) between the two breeds of sheep in terms of live weight
change. However, mortality of sheep due to worm infestation was influenced by breed. In
Experiment 2, dams given MC gained weight. The performance of kids was not affected
(P>0.05) by supplementation, but by body condition and mothering ability of the dams. Breed
had an effect (P<0.05) on the performance of both dams and kids in terms of liveweight
change and mortality. Boer crosses were more affected nutritionally than their local
counterparts. Defoliation of maize leaves did not affect maize grain yield in the field. It is
concluded that supplementation significantly improved the performance of tethered sheep and
goats, and the use of MC as a supplement was superior to M alone. Although breed had an
effect on liveweight change of lambs and goats, and mortality of sheep and goats, it had no
effect on the liveweight change of ewes.