SMALL RUMINANT BREED
PRODUCTIVITY IN AFRICA
PROCEEDINGS OF A SEMINAR
HELD AT ILCA, ADDIS ABABA, ETHIOPIA
IN OCTOBER 1982
Edited by
Ruth M. Gatenby
and J.CM. Trail
 
DECEMBER 1982
INTERNATIONAL LIVESTOCK CENTRE FOR AFRICA
P.O. BOX 5689, ADDIS ABABA, ETHIOPIA

SMALL RUMINANT BREED
PRODUCTIVITY IN AFRICA
PROCEEDINGS OF A SEMINAR
HELD AT ILCA, ADDIS ABABA, ETHIOPIA
IN OCTOBER 1982
Edited by
Ruth M. Gatenby
and J.CM. Trail
DECEMBER 1982
INTERNATIONAL LIVESTOCK CENTRE FOR AFRICA
P.O. BOX 5689, ADDIS ABABA, ETHIOPIA
 
ABSTRACT
This report on small ruminant breed productivity in Africa is
based on a seminar held at ILCA, Addis Ababa, Ethiopia in October 1982.
Basic considerations of breeding plans and production objectives are
discussed. Current research approaches in Africa, India and Australia
are outlined and methods of collection and analysis of production data
are described. Some results of research on the productivity of small
ruminants in Africa are reported.
KEY WORDS
Sheep, goats, Africa, tropics, breed evaluation, productivity, animal
nutrition, disease.
PREFACE
Small ruminants are a major source of livelihood in many areas of
Africa. Unlike cattle, they are owned by even the poorer sectors of the
community, and attempts to increase the productivity of small ruminants
are an important route to improving the standard of living of the rural poor
and the landless peasants. In addition, small ruminants can be an important
step in increasing the productivity of small-holder agropastoralist systems
and thus producing more grain in a continent desperately short of food for
human consumption: with increased livestock production the small farmer
has a surplus for sale, with livestock sales he has a cash income, and with
cash he buys the inputs needed to increase grain production.
The objectives of this seminar, held at the International Livestock
Centre for Africa (ILCA), Addis Ababa in October 1982, were to outline
how research on small ruminants in the tropics and especially on breed
productivity could contribute towards increased output.
The seminar was structured to cover aspects of four broad areas:
basic considerations of production and market objectives, and breeding
plans relevant to these; current research approaches in sub-Saharan
Africa and relevant information from other tropical areas; requirements,
collection, analysis and interpretation of production data; and indications
of current research findings in Africa.
Eight speakers from Africa, Australia and the UK presented
papers and 31 additional workers from 13 countries of Africa took part
in the discussions. These discussions focussed finally on the possible
contributions that ILCA might make towards increasing the impact of
results from small ruminant research operations in Africa.
Grateful acknowledgement is made to the ILCA staff who assisted
in the organisation and conduct of the seminar, in particular
Dr. E. Mukasa-Mugerwa; and Mrs. G. Maloba who typed the report.

CONTENTS
CONTRIBUTORS
Page
GENERAL CONSIDERATIONS
Basic considerations of breeding plans: Helen Newton Turner 1
Production objectives and market forces: Addis Anteneh 7
CURRENT RESEARCH
Research on small ruminants in sub-Saharan Africa:
Ruth M. Gatenby 13
Research on small ruminants in India: Ruth M. Gatenby 21
Trends in sheep production in Australia: Helen Newton Turner 25
METHODS OF COLLECTION AND ANALYSIS OF FIELD DATA
Techniques for field evaluation: Helen Newton Turner 31
Evaluation of breed productivity in Africa and ILCA resources
for data analysis: J. CM. Trail and J. Durkin 37
Handling and analysis of production data: J. CM. Trail
and J. Durkin 41
RESULTS OF RESEARCH IN THE FIELD
Husbandry, nutrition and productivity of goats and sheep in
tropical Africa: R.T. Wilson 61
Seasonal breeding effects on productivity: Christie P. Peacock 77
Disease as a constraint to productivity: S. Mack 81
CONCLUSION 85
REFERENCES 87
PARTICIPANTS 91
INDEX 95
CONTRIBUTORS
Mr. Addis Anteneh: Agricultural Economist in ILCA, Addis Ababa,
Ethiopia.
Mr. Jeff Durkin: Computer Manager in the ILCA Biometrics and
Computer Unit, Addis Ababa, Ethiopia.
Dr. Ruth M. Gatenby: Lecturer in Tropical Animal Production at the
Centre for Tropical Veterinary Medicine,
University of Edinburgh, Scotland
Mr. Simon Mack: Animal Scientist in the ILCA Humid Zone
Programme, Ibadan, Nigeria.
Miss Christie P. Peacock: Animal Scientist in the ILCA Arid Zones
Programme, Nairobi, Kenya.
Dr. John CM. Trail: Animal Geneticist in the ILCA Livestock
Productivity and Trypanotolerance Unit,
Nairobi, Kenya.
Dr. Helen Newton Turner: Honorary Research Fellow at the C.S.I.R.O.
Division of Animal Genetics, Australia.
Mr. R. Trevor Wilson: Animal Scientist in the ILCA Arid Zone
Programme, Bamako, Mali.
GENERAL CONSIDERATIO NS

BASIC CONSIDERATIONS OF BREEDING PLANS
Helen Newton Turner
Genetic considerations are part of an improvement package for small
ruminants. The whole package includes
1. Definition of breeding objectives,
2. Choice of selection criteria, and decisions on measurement techniques,
3. Decisions on the techniques for genetic improvement,
4. Establishment or improvement of marketing systems, to ensure that
increased production gives extra income as an incentive to improvement,
5. Improved management (including veterinary care) to ensure realization of
genetic potential,
6. Development or improvement of an extension service, which will involve
training extension officers to assist producers in implementing improve
ment plans.
These aspects are not necessarily in order of sequence of application;
some are preferably concurrent. Only the first 3, i. e. the genetic aspects, will
be discussed in this section.
Breeding Objectives
The main objective is to maximize the output of the system per unit input.
The output may be the economic return, or amount of product (food, fibre, skins,
etc. ) and its quality. The inputs are food, land, investment, labour, veterinary
care etc., many items of which are on a per head basis, and so it is convenient to
measure production too, per animal. The quality of animal products must be
defined in relation to the requirements of the end user or target market. This
market may be export, the internal market or home consumption. These objectives
have been called "breeding objectives" but they apply whether improvement is
sought by breeding or improved management.
The quantity of meat produced depends on the number and weight of surplus
animals at age of sale. The importance of reproduction rate must be stressed in
relation to surplus animals for sale for meat. Meat quality is poorly defined in
most less developed countries. Urban markets in general require lean meat, but
the fat tail is a delicacy in some countries, and there may be specific flavour
requirements in some areas. Skins are a valuable by-product of the meat industry,
but there is no clear definition of skin quality.
There are basically two types of wool: apparel and carpet. The quality of
wool depends on average fibre diameter (fine for apparel, coarse for carpet wool),
percent clean yield (including freedom from vegetable contamination), staple
length, percent medullated fibres, kemp (shed fibres) and colour (white fetches a
higher price on most commercial markets, but pigmented fleeces are sometimes
preferred in cottage industries). Most small ruminants in Africa produce hair
which is used locally, and there are no clear definitions of quality, apart perhaps
from staple length and colour.
For milk the most important measure of productivity is yield per year.
Quality is less important although goat milk may be desired because of its flavour,
and sheep milk because of its high solid content.
Selection criteria and measurement techniques
Selection will be involved at some stage, whatever technique is chosen for
genetic improvement. Where small ruminants are kept primarily for meat produc
tion, selection will be on the number and weight of offspring weaned per female per
year. The number of offspring born can be increased by decreasing the number of
females which fail to lamb (or kid), by increasing litter size, or by increasing the
frequency of parturition. Females which fail to produce offspring after consecutive
opportunities should be culled, but the heritability of such failure is low, and such
culling may not have a marked effect on the reproductive performance of the whole
flock. Litter size, on the other hand, will respond to selection. The acceptability
of twins depends on the environment (particularly nutrition) and management
system. There is considerable interest around the world in increasing the fre
quency of parturition, and there have been reports of success. There is little
information so far, however, of the effect of more frequent parturition on total life
time productivity. With the year-round reproduction which occurs in Africa, there
is scope for research on the heritability of the parturition interval, and the effect
of decreased parturition interval on lamb survival and growth.
For wooled sheep, selection is based on weight of fleece, but considering
also fibre diameter and percent medullation. A long staple is required for most
markets; in general, selection for increased fleece weight will also result in a
longer staple, but it might be necessary to pay particular attention to staple length
for indigenous Africa sheep.
Selection of animals for milk is in terms of quantity of milk produced per
year, which is a function of quantity of milk per lactation, lactation length and
parturition interval.
Techniques for genetic improvement
Techniques for genetic improvement include selecting between breeds (or
strains), selecting within breeds and crossing. Crossing includes grading-up,
developing new breeds and exploiting heterosis. It should be noted that even if
crossing is chosen as a technique, selection is usually involved as well.
Selection between breeds. Many international meetings over the last few years
have stressed the need to look more closely at indigenous (also called local or
native) breeds, because of their adaptation to their environments, instead of
assuming that importation of a so-called "improved" breed is a short cut to in
creased production. Selection between breeds will therefore mean comparing
indigenous breeds, as well as comparing an indigenous with an exotic or its cross,
as in the past. Selection between breeds must be based on performance data
collected from groups run in the same environment. Comparisons in the past have
usually been on experiment stations; there is a great need for them to be made in
the environment where the animals will run. Techniques for field evaluation will
be discussed at a later session. Comparisons MUST be carried through to give
realistic estimates of lifetime production including not only measures of products
such as body, fleece and milk weights but also reproduction, mortality and mor
bidity rates.
Selection within a breed. Selection is usually done within a flock, i.e. among
animals of the same age which have been run together. Genetic progress through
selection (selection response) depends on heritability, selection differential and
generation interval.
Heritability is fortunately moderately high for many of the characteristics
with which we are concerned. Body weights later than at weaning and fleece weight
can be selected on a single measurement for both sexes. For sex-limited charac
ters such as weight of offspring weaned or milk production, selection of males
must be on performance of female relatives. The most useful female relative is the
the dam; selection on records of half-sibs or progeny lengthens the generation
interval and so lowers the annual rate of genetic progress.
Selection on number (or weight) of offspring weaned per female at a single
parturition per year should preferably be based on more than one record.
Selection of both ewes and rams can be based on the dam's lifetime record using
the formula for heritability of a repeated record, namely:
Dam's ranking coefficient = J h P - P
D D
1 + (k-l)t
o
Where h is heritability of number (or weight) of offspring weaned per ewe mated;
t is repeatability; k is number of matings; Pj) is number (or weight) of offspring
weaned per mating for all matings of the dam under consideration; and Prj is
average number (or weight) of offspring weaned per mating for females in the
same flock, up to and including the dam's present age. This formula is most
readily used in flocks where full records are kept. Where they are not, it is still
possible to devise a system of recording a female's performance at each
parturition (at least for numbers of offspring) by means of ear-notches, eartags
or tattoo-marks. The formula was devised for once-a-year lambing. Its advant
age is that younger dams are placed on the same footing as their elders. It could
probably be extended to cover year-round breeding, if estimates of the heritability
of the interval between parturitions are available.
In areas where twins are acceptable, females which twinned consistently
could be marked, while another mark could be placed on females which consistently
had a short interval between parturitions. Where twins are not desired, a shorter
interval might be used as the selection criterion, though, as already mentioned,
more information is needed about the effect of shortening the interval on lifetime
production.
The selection differential depends on the accuracy of selection (i. e. measure
ment rather than eye appraisal), the proportion of animals saved and the variability
of the character concerned. When flocks are small, the proportion of animals saved
will be high, and the selection differential lowered, particularly for rams;
progress through selection will thus be very slow. Further, if rams
are always chosen from within the flock, inbreeding levels become high. Ways of
overcoming this problem include establishment of a central sire- breeding nucleus
of larger size, either by government or private ownership, establishment of
co-operative breeding schemes, and the use of ram circles and ram exchange.
A central nucleus should contain at least 300 breeding females, preferably
more. In it selection on measured performance would be practised, so that res
ponse to selection would be predicted. Males from the nucleus would be distributed
to small flocks. Whatever rate of genetic progress was achieved in the nucleus
would also be achieved in the flocks which consistently drew their males from it.
This progress could be 1-2% per year for the suggested selection criteria. If this
seems small, remember that it is cumulative, so that after 10 years production
would be 10-20% higher. Since indigenous breeds have become adapted to various
harsh environments throughout Africa, it would be logical to choose among the
breeds within each region, then establish a sire-breeding nucleus for each. The
question of the population of commercial flocks which could be influenced by a
nucleus of a given size will be discussed later.
Cooperative breeding schemes have become popular in New Zealand and
Australia for both sheep and cattle. They involve the establishment of a ram-
breeding nucleus through the cooperation of a number of flocks, each of which con
tributes its best ewes to the nucleus each year, receiving rams in return. Suppose
a central nucleus of 300 breeding females could be established by a contribution
of 5 females from each of 30 cooperating flocks in each of 2 years. These flocks
might consist of 25 breeding females each. The annual contribution thereafter
would depend on the length of life of females in the nucleus and on its reproduction
rate; calculations have shown that genetic gains in the nucleus are maximized if
half its replacement females come from within itself, half from the cooperators.
Annual contributions after the first two years would thus be of the order of 1-2
ewes from each cooperator. The nucleus would be able to supply sires to smaller
flock as well as to its cooperators. Details of such schemes could be worked out
if they were thought feasible. Selection in the nucleus would be based on measured
performance, but the contributed females could be selected visually. Record
keeping would be required only in the nucleus.
Ram circles, as operated in Scandinavia and Ireland, require backing from
extensive record-keeping and data analysis in all flocks involved. In the absence
of such infra- structure, they are probably not adaptable to African conditions at
present.
Ram exchange is more feasible. If several small flocks are normally
herded together so that their general environment can be regarded as the same,
then the females' reproductive performance can be recorded provided they are
identified by ear-notch, eartag or tatoo. Males can be selected from the combined
flock on a combination of their own weight (age adjusted) and their dam's
reproductive performance. This would help to increase the selection differential,
and if rams were rotated among similar flocks, inbreeding would be lowered.
Crossing. Crossing may be between indigenous breeds, or between indigenous
breeds and exotics. Very often in the past it has been assumed that exotic breeds
from temperate areas, with high production in their homelands, will automatically
raise productivity when crossed with local breeds in tropical areas, and very often
the introductions have failed. Crossing may have a place, but is not necessarily a
rapid highway to success, and introductions should be carefully evaluated in the
field before widespread crossing is undertaken. Such evaluation would include
comparison with indigenous breeds.
Grading up involves replacement of one breed by another, with continual
back-crossing, while development of a new breed means crossing of two (or more)
breeds, then at some stage mating inter- se, with selection. Again evaluation in
the field is needed to assess the new type of animal.
Exploitation of heterosis is a technique widely used in Europe and North
America, particularly with cattle. It is useful if a cross between two (or more)
breeds has higher production than the superior parent. This is sometimes the
case, and the technique is used in Australia for lamb production. Merino ewes
are crossed with Border Leicester or Dorset rams, the crossbred ewe progeny
then being mated to a meat breed sire such as Dorset or Suffolk. The crossbred
ewe has a higher lambing percentage than either parent, and the terminal sire
gives fast growth rate to the lamb. Use of such a system means there must be a
continual supply of the parent breeds. Estimates of heterosis for sheep and goats
vary widely, and more data are required before such schemes could be recom
mended for small ruminants under African conditions. There could be a place
for stratification of the industry, with crossing of two indigenous breeds to develop
a system similar to the Merino x British breed crossing used in Australia, but
research is needed.
Gene distribution. It is instructive to see how long it takes to distribute
superior genes once a source is established. If males are being distributed from
a selected central nucleus of an indigenous breed, any flock drawing those males
will make the same rate of genetic progress as the nucleus. The number of
males available annually will depend on the size of the nucleus and its reproduc
tion and mortality rates. A nucleus of 300 breeding females might have 80 males
available for distribution annually, and a nucleus of 500 breeding females, 130.
If the average life of a male in the field is 3 years, then males can be distributed
to 3 sets of flocks in rotation. In this way the numbers of females reached can be
calculated as (average male life) x (males available annually) x (male: female ratio).
Some examples are in Table 1. If artificial insemination could be used, the
numbers would of course be greater.
Table 1. Numbers of females mated by sires from central nucleus
Size of central nucleus Male : female ratio
(females) 1:10 1:20 1:30
300
500
2,400
3,900
4,800
7,600
7,200
11, 700
If exotic or synthetic breeds are being distributed, the first decision to be
taken is the level of the new distributed genes required in the target flocks.
Assumptions can be made about reproduction rates and death rates in the village
flocks and, using these, calculations can be made of the length of time taken for
the required level of genes to be reached in a village flock. At that point, males
from within the target flocks can be used to maintain the level, though with some
rotation of rams to avoid inbreeding. When a given set of flocks becomes self-
maintaining, distribution of males can be switched to another set. Calculations
along these lines have shown that, 10 years after establishment, a nucleus of 500
breeding females could raise 4, 500 females in village flocks to 50% level of intro
duced genes, provided males with 100% of the genes were distributed. Again, with
AI this number would be greater. These calculations have been made to show that
evenifafirstcross between an exotic breed and an indigenous shows a great
superiority over the indigenous, distribution of genes through a population takes a
long time.
PRODUCTION OBJECTIVES AND MARKET FORCES
Addis Anteneh
Small is beautiful.' "... Smallholder, Juan Batista, maintained a herd of
38 mature females of mixed Criollo-Nubian breeding. The goats were exposed for
breeding year-round so that 3-5 females were always in lactation. The herd
grazed freely on unfenced pastures and brush lands in and around Batista's house.
The next morning the lactating females were driven from door-to- door and milked
to each customer's order. This was the example of a small entrepreneur adopting
to a definite but limited, market niche. Feed costs for the herd were essentially
nil" (Winrock, 1977).
Production from small ruminants
Throughout the world, sheep and goats are raised for several purposes.
The first things that come to mind are meat, milk and fibre. In many developed
countries, the production of sheep meat is a specialized undertaking where large
flocks are kept and managed under commercial ranching conditions. Milk produc
tion from dairy goats is an established industry in western Europe. Wool produc
tion is a major activity in Australia, New Zealand, the United Kingdom and the
USSR. Sheep and goat skins are major non-food products accounting for almost
20% of world production of hides and skins. Sheep and goat production in developed
countries is mainly a commercial undertaking emphasizing these products.
In the developing countries too, farmers raise sheep and goats for meat,
milk and fibre production. Africa excluding South Africa has about 22% of the
world's sheep and goats, and produces about 14% of total sheep and goat meat.
About 13% of sheep and goat skins are produced in developing Africa. In some
countries such as Ethiopia hides and skins are major export items (17% of value
of total merchandise trade in 1979). Goat and sheep milk production in Africa
accounts for approximately 14% of world production (FAO 1981a). In some
African countries sheep and goat milk is a major source of milk supply for con
sumption. Gall (1975) reported that as a percentage of total milk production sheep
and goat milk accounted for 57% in both Morocco and Niger, 56% in Mauritania
and over 30% in Mali, Algeria and the Sudan. Developing countries in Africa are
not major producers of wool or goat hair. Excluding South Africa again, Africa
produces about 37, 000 tonnes of greasy wool per year (FAO, 1981a); this
represents only about 1% of world production.
So meat and milk production from sheep and goats are major objectives
in Africa as in many developed regions. There are, however, two important
differences to note between production objectives in the majority of the producers
under these two categories. Firstly, in developed countries the objectives are
commercial whereas most production in the developing countries takes place on
holdings mainly geared towards subsistence food production. Secondly, sheep and
goat keeping in Africa serves other purposes which, though difficult to quantify may
be as important as food production per se.
For the small farmer or agropastoralist in Africa, sheep and goats as well
as cattle provide security of continued food supply in times of crop failure; their
meat or milk is directly consumed or the animals are sold for cash to purchase
grain or are exchanged for grain products. Sheep and goats have special advantages
over cattle in this and other respects. For the small subsistence farmer or
pastoralist who considers his livestock as a kind of investment or "money in the
bank", sheep and goats represent a current account or working capital in relation
to large stock which can be considered as equity investment (McDowell and
Hildebrand, 1980).
Skins are an important byproduct to meat production. Goat skins of the Red
Sokoto in West Africa and Bati goat skins in Ethiopia fetch premium prices in world
market, but there is no evidence of these or other similar breeds being produced
for their skins as a primary production objective. Sheep skins are used for clothing
in the highland area of Ethiopia, and skins with their hair kept intact serve as
sitting rugs in many rural households in Ethiopia. Goat skins are specially suitable
for use as water carriers and grain sacks. In some parts of Africa manure from
tethered sheep and goats is collected and spread on crop fields as fertilizer (ILCA,
1979a).
In most developing countries, these production objectives are inseparable
from meat and milk production. The general picture given by production statistics
indicates that in Africa sheep are preferred for meat and goats for milk, and both
meat and milk are mainly used for local consumption.
Individual objective
The subsistence producer keeps livestock to meet his family's immediate
requirements for food, fuel and fertiliser, to give him cash for off-farm expendi
ture, and he aims to minimize production and investment risk. Small ruminants
therefore have several advantages. They provide meat and milk in quantities
suitable for immediate family consumption, they are easier than cattle to buy and
sell so that they are better suited to meeting relatively small and frequent cash
requirements, and their low individual value means low initial investment and low
risk of total flock loss from individual deaths. Small ruminants can thrive in
extremely dry areas and in tsetse-infested areas where large stock cannot survive.
There are also certain disadvantages of small ruminants. Particularly in
extensive systems sheep and goats are more susceptible than large stock to pre-
dation and theft. In cultivated areas, sheep and goats may damage crops; a
survey in the derived savanna zone of Nigeria (ILCA 1979c) showed that 95% of the
people who did not keep small ruminants gave crop damage as the reason.
National objectives -
National production objectives for agriculture in sub-Saharan Africa, as in
other developing countries, include increasing food production, increasing employ
ment, increasing foreign exchange, more equitable income and wealth distribution,
and more efficient utilization or conservation of resources. In sub-Saharan Africa
livestock products (meat, milk, eggs, wool, hides and skins) account for approxi
mately $5, 000 million or 17% of agricultural GDP (Jahnke, 1982), and of this about
17% is produced by small ruminants. Export of live animals accounts for about 1%
of the total stock of small ruminants compared with only about 0. 6% for cattle (FAO,
198la; b). Small ruminants are particularly important in the economy of arid
zones where the milk production from sheep and goats is 220 kg per livestock unit,
more than twice that from cows. In areas of high population density where farms
are small and a farmer's fields may be scattered, small ruminants are more
easy to keep than large livestock.
Livestock provide opportunities for landless families to obtain both employ
ment and income (McDowell and Hildebrand, 1980). Small ruminants are particu
larly important because of their low initial cost. Goats are frequently condemned
as destroyers of vegetation and a major cause of desertification. Gall (1975) and
Rice (1976) conclude that it is mismanagement which destroys vegetation, but
nevertheless one of the factors limiting their productivity may be the negative
attitude of planners towards development projects including small ruminants.
Public policy can be a serious constraint to the productivity of small
ruminants. For instance genetic improvement programmes to introduce exotic
breeds may be undertaken without giving attention to exploiting the potential for
production of the indigenous breeds. As an example, the growth rates of indigenous
Malaysian goats (Devendra, 1980) can be increased substantially by improved feed
ing, so that the edible weight of carcasses at slaughter increases by 37%. If even
50% of the small ruminants slaughtered annually in sub-Saharan Africa were to
achieve half this weight increase, then the increased meat production would reduce
the meat deficit by about 5%.
Prospects for increased supply of meat from small ruminants
Based on recent trends, the demand for meat from both ruminants and non-
ruminants in sub-Saharan Africa is projected to increase at 4. 2% per year until
2000, whereas supply will increase at only 2.4%,and supply from ruminants will in
crease at a mere 1.2% per year (de Montgolfier-Kouevi and Vlavonou, 1981). The
rate of growth of sheep and goat meat production will be lower than that of beef,
indicating the need for a greater effort to increase supply from the former. The
possibilities for increasing production by expanding low input- low output extensive
systems are limited because of the decreasing availability of grazing land in both
small holder and pastoral areas. Also, most small holders tend to regard sheep
and goats as secondary to crop production and as a means of raising occasional
cash to meet immediate family requirements. On the other hand, small holders
have been known to exploit favourable market conditions to maximize this cash in
come. Many small farmers sell animals at times of peak demand (such as religious
festivals), some fatten sheep or goats to take advantage of consumer preference,
and some farmers are part-time sheep and goat traders.
10
Such situations could be exploited to increase marketed supply from small
holders. However, sustained growth in meat supply from small holders could be
achieved only if the animals could be integrated into the small holder's overall pro
duction activities to achieve higher farm productivity. One way of making small
farmers respond to increased demands and higher prices by producing more, is to
demonstrate that they could achieve increased crop production by using their cash
income from livestock to buy inputs such as fertilizer. Small ruminants for meat
appear to have special advantages in making an important contribution to this, if the
appropriate technical solutions and economic incentives could be provided.
Marketing
Small ruminants may be sold directly to a consumer, sold to a trader, sold
in an auction or sold to a marketing organisation. In many parts of Africa most
farmers sell to a trader or middleman and animals may pass through the hands of
many traders before reaching the consumer. The middleman is much maligned; he
is accused of giving a low price to the farmer and making large profits. In fact,
unless there are limits to the number of traders (either by law or by tradition) then
the trader must earn his income. He provides a selling service to the farmer who
may not be able to go to market, and he may move animals over long distances from
areas of supply to areas of demand. However, the market is generally unregulated,
peasant farmers have only limited access to market information and there is little
incentive for export.
Marketing cooperatives, such as operate in Botswana and Ethiopia can help
to bypass the middleman. In countries such as Kenya and Zimbabwe there are
commercial organisations which have grading schemes and buy meat at advertised
prices. However, even in these countries the majority of meat from small rumi
nants does not pass through these official channels. To accommodate small rumi
nants in slaughter-houses special lines are required, and organisations are under
standably reluctant to provide for sheep and goats when the supply is not assured.
Inter-African trade is encouraged by the OAU, yet political decisions limit
the trade between east and west Africa. For instance, Nigeria has import restric
tions, and the Kenyan government discourages the export of live animals because it
employs few people compared with the canning and skinning industries. In addition,
governments may be reluctant to encourage the sale of meat in urban areas and for
export because this results in higher meat prices and consequently lower meat
consumption by the rural population and the poorer sector of the community.
Accessible markets with short supply; two examples
East African sheep and goats for the Middle East. Public policy may be
partly responsible for failure to take full advantage of quite favourable trading
conditions even where markets are ready to absorb sheep and goats. A case in
point is sheep and goat exports from some East African countries to the Arabian
Peninsula and the Gulf States. According to a report prepared by AOAD and
FAO (1979), Arab countries on the Red Sea and in the Gulf imported from East
11
African countries live cattle, sheep and goats equivalent to the 56, 000 tonnes
carcass weight. This accounted for slightly over 20% of the total fresh meat con
sumption in the importing countries. Of the imported animals Somalia's share was
about 75%, Sudan's share was about 20% with only 5% being taken by Ethiopia.
About 60% of the live animals exported were sheep and goats or approximately 1. 6
million for Somalia, 400, 000 for the Sudan and about 100, 000 for Ethiopia. These
numbers are inversely proportional to the small ruminant populations of these
countries: Ethiopia with approximately 40 million, Sudan with 30 million and
Somalia with about 25 million sheep and goats (FAO, 1981a). Even if one were to
assume that about one-third of Somalia exports were unofficially transferred from
Ethiopian flocks, the Somali export offtake rate is still about 4 times that of
Ethiopia. While government interventions in price setting have an effect on both
supply and demand for live sheep and goat exports to the importing countries (such
as Sudan's export duty and taxes which accounted for 21% of export value as against
4% for Somalia), the very poor performance for Ethiopia and the relatively poor
performance for the Sudan must be attributed to internal price structures and poor
marketing support services.
What is even more interesting is the share distant producers such as India
and Australia take in the live animal export trade of the "doorstep" markets for
East African livestock. Despite very real consumer preference for fresh meat
from East African fat-tailed or fat-rumped sheep and the enormous advantage East
African countries enjoy in cheaper sea freight costs (6-10 times lower than the
Sidney-Jeddah rates for live sheep), Australia's share of the large Saudi Arabian
import market was approximately 15% in 1978 (AOAD and FAO, 1979), Ethiopia's
share was only 1% while that of the Sudan was approximately equal to Australia's.
The higher unsubsidized prices for fresh meat from African live sheep and goats
prevailing in the markets (20-30% higher than Australian prices) are believed to be
reflection of consumer preference and do not seem to act as a deterrent to the
growth of future demand in these markets. A 100% increase is projected in the
consumption of fresh meat from African stock as opposed to a 30% increase in the
supply from Indian and Australian sources by 1983/1984.
Of the total requirement of 113, 000 tonnes carcass weight equivalent from
East African sources, Somalia and Sudan are expected to supply about 80% while
about 22, 000 tonnes will have to come from other East African sources (AOAD
and FAO, 1979). If Ethiopia were to capture the market for about 5, 000 tonnes
carcass weight equivalent of fresh sheep meat, it would mean foreign exchange
earnings of about US$ 20-25 million. This is about 5 times the total value of live
animals and animal products (excluding hides and skins) exported in 1979 and over
30% of that of hides and skins. The export of about 400, 000 sheep and goats from
the 24, 000 exported in 1980 is a phenomenal increase. However, even under the
present relatively unrealiable supply pattern from the Ethiopian smallholder and
pastoralist, a little more agressive marketing strategy and internal reorgani
zation should make this possible.
The East African and Ethiopian situation is used for illustrative purposes.
The "fallout" from such externally induced trading activity in terms of the cash
benefits to the smallholder or pastoralist should not be underestimated.
12
Not enough white wool: the Ethiopian carpet industry. Only about 10% (3500
- 4000 tonnes) of the wool produced in sub-Saharan Africa is exported, which
implies that 90% of the domestic production of greasy wool is traded domestically.
In value terms, exports of greasy wool in 1980 represented only about 3% of the
value of live sheep and goats plus fresh sheep meat exports (FAO, 1981b). This
would still be smaller if the value of skins were included as part of the export of
livestock products. Import figures for greasy wool (FAO, 1981b) are reported for
only Ethiopia and Kenya with 79 tonnes and 24 tonnes worth about US$ 71, 000 and
US$ 75, 000 respectively.
There is scant information as to what use domestically produced wool is put
in many parts of the region. Some information on production from indigenous flocks
and use of carpet wool has been available for the central highlands of Ethiopia
(Cossins and Bekele, 1974; Technical Committee, 1975; Weiner, 1972; 1975;
Livestock and Meat Board, 1974). Coarse wool is used to make hand-woven carpets
and blankets. Making runner carpets is a thriving activity in western Wollo, and
also uses wool purchased from farmers in eastern Gondar and northern Shoa.
Cossins and Bekele (1974) estimated a monthly demand of 5 to 7 tonnes in only one
locality renowned for carpet making.
Small-scale "industrial" production of woollen carpets is concentrated in
Addis Ababa. Most of the carpet wool required by these handicraft centres is
obtained from rural production in northern Shoa and western Wollo. The Livestock
and Meat Board (1974) reported that the major carpet making centres in Addis
Ababa had an estimated annual demand of 65-70 tonnes in the form of coarse wool,
cleaned yarn from Debre Birhan Wool Factory and scoured wool from some
tanneries as well crossbred wool from Debre Birhan sheep station. The industry
has been characterized by irregularity and overall shortage of supply (Technical
Committee, 1975). The current situation is no better. The Handicraft and Small-
scale Industries Development Agency (HASIDA) has confirmed that this problem
still persists particularly for white carpet wool and is further exacerbated by the
high prices being asked by trader suppliers. Prices have about doubled in the last
3 to 4 years from Birr 3. 50 - 4. 00 per kg to Birr 7. 00 for white wool. Prices
paid to farmers are estimated to be quite low, less than half the prices charged by
traders supplying the pile carpet making industry. Farmers give shortage of
sheep types producing white wool as the main reason for the inadequate production
and supply although the low prices paid to them must clearly be a disincentive.
HASIDA considers the lack of farmer organization and support services as part of
the problem.
Hand-made carpets are in great demand as export items from Ethiopia. If
supply of the raw material could be better assured there is an indication that
carpet making could be a robust activity with backward linkages to increasing the
cash income of small off-the-road farmers in the remote parts of western Wollo,
northern Shoa and estern Gondar.
Again, the Ethiopian situation was discussed for illustrative purposes.
Although for Africa as a whole the role of wool in generating cash income may be
limited, there are possibilities for increasing producer cash income from wool in
areas where there is an already established and unsatisfied market for the product.
CURRENT RESEARCH

13
RESEARCH ON SMALL RUMINANTS IN SUB-SAHARAN AFRICA
Ruth M. Gatenby
Until the last decade small ruminants in Africa have received relatively
little attention from research workers. Recently more attention has been paid
to small ruminants as their advantages are becoming understood - their ability to
produce meat and milk even in hostile environments, their resistance to many
diseases of cattle, and the ease of marketing of their products.
Research on small ruminants can be classified into basic and applied
research. In the basic approach an in-depth study is carried out in a narrow field.
In applied research a complete system is analysed, and much of the field work
carried out by ILCA falls into this category. The final objective of both approaches
is to improve the system of production. For successful outcome, a flow of ideas
between the scientists working on both approaches is essential; this flow takes the
form of scientific literature, conferences and scientific visits. Applied research
can generally be undertaken only in the locality of the system whereas basic
research may best be done in a place with appropriate facilities away from the
small ruminants. One of the advantages of ILCA and other research centres in
Africa is that they provide facilities for basic research as well as being ideally
suited to applied research. Thus the two approaches can progress in parallel,
with good contact between scientists undertaking each type of research.
The research process
All types of agricultural research are undertaken with the objective of
improving the standard of farming. Research can rarely be put directly into
practice in the field but innovations "discovered" by research workers must be put
into appropriate packages, a process called development, and then distributed to
farmers via extension workers. This whole process is shown in Figure 1. In
addition to the primary flow of information downwards from research to farming,
there are also secondary flows upwards between all levels. In other words,
research, development and extension all need feedback to function properly.
In putting research into practice all 4 steps shown in Figure 1 are equally
important. Too often extension receives scant attention. To be successful, an
extension or advisory worker must be accepted by the farmers, but he must also
be respected by the scientists. In most less developed countries the social gap
between scientist and peasant farmer is very wide; progress cannot take place
unless the extension worker can bridge this gap. Thus extension is often the
weakest link in the chain, and recognition of the importance of extension by agri
cultural scientists is essential.
14
Research
Development
Extension
i
 
Farming
Figure 1. The flow of information in agricultural research. Arrows indicate
exchange of ideas.
The first step in applied research is the understanding of the present
system. This may be called "static" research, and involves finding out how the
system operates, what are the outputs and inputs, and how the system under study
interrelates with other systems. The second step is to see how the system responds
to a change in circumstances; perhaps another input is added (a better source of
feed, veterinary care or a new breed of animals), one input may be reduced (less
land), or there may be more demand for an output (better marketing facilities).
These responses may be called the "dynamic" characteristics of the systems, and
once they are known attention can be given to solving specific problems that limit
the functioning of the system. At the problem- solving stage there is likely to be
the greatest input from basic research.
In the development of new systems care must be taken to ensure that they
are appropriate and stable; present traditional systems have evolved slowly and
have been tested ecologically over centuries. Even though we may want to change
systems rapidly we cannot afford to give unsound schemes for the advisory staff to
propagate because this will merely ruin the livelihood of the farmers and give rise
to distrust which will make future extension even more difficult.
Objectives of research
The objective of research is to increase the productivity of sheep and
goats. This increase in production most often takes the form of an increased
quantity, but it may also be an improved quality of product or alternatively a new
product may be introduced. To a small farmer the security value of owning small
ruminants may be as important as their tangible production; small ruminants are
a low risk investment which keeps its value. In a drought sheep and goats can be
eaten whereas even if a farmer had cash there is no guarantee that he could buy
food, and after the drought small ruminants have the capacity to multiply again.
15
The national objective for research on small ruminants may be to increase
their production of exportable products and thus reduce the balance of payments
deficit. From the point of view of an international development agency the aim may
be to improve the standard of living of the poorest sector of the community. These
objectives are all subtlely different, and each one must be considered.
In addition it must be remembered that small ruminants are only part of the
ecosystem. It would be foolish to think of increasing the productivity of sheep and
goats to the detriment of say crop production and cattle production. It is the output
of small ruminants per unit input that must be considered, bearing in mind to
measure productivity over a long period to ensure that the resource base is not
diminished.
By the time present research findings are put into practice, say in 2000,
the economic and social climate may have changed. Almost certainly there will be
little or no cereal available for ruminants because of the demand of the increased
human population for food. The area of grazing land is diminishing as cropping
spreads to marginal areas, and the production per hectare of grazing land is said
to be falling too as a result of overgrazing. However better uses of some food
resources such as straw and industrial and vegetable by-products should have
developed. A better understanding of animal requirements should mean that better
nutrition is given to those animals which will use it most efficiently. All persons
involved with animal production are more likely to have received more education
than their present-day counterparts and will be more aware of management
practices to minimise disease problems and to make better use of available feeds.
There will probably De a swing to production for sale rather than home consump
tion as urban and export markets become more important.
The state of research on small ruminants in Africa
A considerable quantity of knowledge about small niminants in Africa is
being collected by research workers in both national and international institutions.
Our understanding of production systems is obviously far from complete but we
should know enough to be able to proceed from static research to dynamic research
in some areas. There is a tendency for research workers to become so involved
in elucidating points of detail that the objective of research is forgotten, and the
time when attention should be given to solving specific problems which limit
production may be long passed. Even when the research approach has progressed
beyond the static approach, and basic information about the system may continue
to be collected as a by-product.
Let us consider briefly some research findings in Africa. Africa has a
greater meat/milk consumption ratio than any other continent and 27% of this
meat comes from sheep and goats. However meat is still a luxury for the
majority of the population and is eaten only on special occasions. In a continent
with primitive storage facilities the small individual size of sheep and goat is an
advantage because meat from them comes in family-size packs. Any animal can
be eaten for meat. In a system designed to produce meat as the main product
16
most meat is produced from young males and culled females. However other
classes of animal are eaten if the need arises, and sick animals may be slaughtered
at point of death.
The growth curve of small ruminants is sigmoid with the point of inflexion
occurring at a few weeks of age. The maximum liveweight attained (i. e. adult
liveweight) is determined primarily by genotype, whereas the rate of weight gain is
more under the influence of environmental factors. Typical growth rates from 0 to
8 months of sheep under traditional husbandry in semi-arid areas are between 50
and 100 g/day (Wilson, 1980). As the animals get older so their rate of liveweight
gain decreases under traditional systems. It is however possible to increase the
growth rate of animals destined for slaughter by improved nutrition, a process
known as finishing. Brinckman (1981) reports liveweight gains of up to 250 g/day
for sheep in Africa. In an experiment in the Sudan, El Mag and Mukhtar (1975)
fattened 1 year old rams for 70 days on different levels of locally- available con
centrate. Liveweight gain, killing out percentage and food conversion ratio all
improved with increasing proportion of concentrate in the diet, and they concluded
that if finishing is to take place it should be done quickly. In practice the avail
ability of concentrate food and infrastructure prevent finishing being widespread in
Africa.
Meat quality is not well-defined in Africa, except where carcass grading
takes place in an organisatioL such as the Kenya Meat Commission. Some objec
tive assessment may be made on the basis of fat content (the fat tails of sheep are
a delicacy in parts of East Africa) and flavour. People in different parts of the
world have very different flavour preferences: Arabs like meat from entire males,
whereas people in other areas prefer meat from castrates or females. In Yemen,
meat from East Africa is much preferred compared with meat from Australia or
New Zealand which is said to have an unpleasant flavour. Whether this is due to
the origin of the meat or its subsequent freezing is unclear. Tenderness, a
quality prized in developed countries, is of little importance in most of Africa
where cooking methods have adapted to deal with tough meat.
Milk for human consumption is a relatively minor product from small
ruminants in most of Africa. Goats are more frequently used for milk production
than sheep which are milked in parts of Somalia, Sudan and the arid regions of
West Africa. There is little literature on quantitative aspects of milk production
but yields are probably less than 1 kg/day. Goat milk is said to be more easily
digestible than cow milk, and sheep milk has a much higher content of total solids
than cow milk - a factor which gives it a premium price. Most communities boil
milk before use thus minimizing the risk of diseases such as brucellosis.
Controlled fermentation produces yoghurt which may be dried and stored in oil.
Cheese too keeps for several months and ghee which contains little or no water
may be stored for years in sealed containers.
The advantages of milk production over meat production are that milk is
produced every day and the efficiency of utilization of nutrients to produce milk
is high. Factors limiting the adoption of milk production by small ruminants
17
include the established milk supply from cows whereas the milk yield of meat
breeds of small ruminants is low and they have milk let-down problems. In
addition lactose intolerance is widespread in many parts of Africa so that the
demand for fresh milk is low.
Wool sheep are found mostly in the drier and cooler areas of Africa. Wool
is obtained when the sheep are shown or slaughtered, and a specialized marketing
structure or local industry is needed. Problems occur if wool contains much
vegetable matter, and environmentally associated conditions such as blowfly strike,
canary colouration and lumpy wool disease can be serious. Mohair and cashmere
production from goats is negligible except in some highland areas. On the other
hand sheep and goat hair is used throughout Africa in the production of felts, carpets,
ropes and bags. Skins are a useful by-product of meat production. The production
of good quality leather depends both on the treatment of the animals and the treat
ment of skins. Sheep and goats each produce about 30 m of intestines which with
proper treatment can be used for surgical thread, sports' racquets, etc. Manure
is another important product. Much more research attention should be given to
these "minor" products of small ruminants which are so important in the economy
of the small farmer.
In summary, considerable effort is being put into studying the use of small
ruminants for meat production in Africa., Enough knowledge has been gathered to
allow effort to be diverted into developments in response to particular needs such
as finishing. In contrast, relatively little attention has been given to other products
such as milk, skins and manure, and a strong case can be made for diverting more
research effort here.
As a broad generalization the major factors limiting the productivity of
small ruminants in Africa appear to be poor nutrition in the semi-arid areas and
disease in more humid areas. Drinking water may appear to be more critical than
nutrition in some dry areas, and water and food are obviously closely related.
Marketing too is a big problem and socio-economic limitations may override tech
nical problems. The characteristics of the breeds of animal available in Africa
are not primary limiting factors except where nutritional, disease and marketing
problems have been eased.
Nutrition
Poor nutrition is identified as one of the most serious factors limiting small
ruminant productivity throughout Africa but especially in arid and semi-arid areas.
Even when there appears to be an adequate quantity of vegetation available its
nutrient content in terms of metablisable energy, digestible protein and minerals
may be poor0 In semi-arid areas the main problem occurs in the dry season and
strategies to ameliorate the problem can be divided into those which attempt to
improve nutrition in the dry season and these which attempt to improve nutrition
throughout the year.
18
Possibilities for the dry season include:
limit numbers of animals. For communal grazing areas this depends on a
good social structure in the community. Restrictions imposed by govern
ments are rarely successful. Widespread veterinary treatment can
aggravate the problem, but improvement of marketing opportunities to
increase offtake should be encouraged.
grow species of grass and legumes which retain their nutritive value into
the dry season. Stylosanthes is successfully used in this way in tropical
Australia.
grow shrubs and trees. Many palatable trees (such as Prosopis spp. and
Acacia spp. ) retain their leaves into the dry season and so can be lopped
for fodder.
conserve fodder as hay or silage. These are not widespread practices.
To make good silage a watertight and airtight silo is required. Hay making
relies on sub-drying, and if hay is made towards the end of the wet season
its quality can be reduced by leaching of nutrients. Both practices demand
labour when it is also needed for harvesting crops. So in reality grass is
left as "standing hay".
use crop and industrial byproducts. Considerable quantities of cereal
straw, husk, oil- seed residues, vegetable waste and cotton coffee, cocoa,
groundnut and sugar residues are produced in less developed countries.
Some residues are well-utilized, some are beginning to be utilized but a
lot are wasted. Their feeding value must be appreciated and practical
problems such as transport and method of feeding must be tackled.
use supplements. Where one specific nutrient is lacking a supplement can
have a dramatic effect on productivity. For instance urea, minerals and
molasses gi re non-protein nitrogen, specific minerals and energy,
respectively.
bring in fodder from other areas. This is likely to be impracticable for
most of Africa but is practised when droughts occur in the semi- arid parts
of Australia for instance. A
export animals to other areas. Traditional movements of animals out of
arid areas in the dry season are known as nomadism or transhumance.
Trucking of animals introduces enormous problems where basic facilities
such as roads are poor. In addition, animals moved to a more humid area
are susceptible to disease and may not thrive.
Better feed supply throughout the year may be achieved through growing
species of grass with a higher nutritive value; growing legumes; growing fodder
crops; controlled grazing; the use of fertilizer and the integration of small
ruminants with plantations of coconuts, oil palm, rubber, tea and trees for wood
production. The success of these practices depends on their economics. Besides
investigating their technical feasibility, research workers must devise packages
which are suitable for different parts of Africa.
19
Diseases
Diseases of small ruminants seriously limit their productivity especially in
humid and sub-humid zones. The effects of disease interact with malnutrition in
that an undernourished animal is less resistant to disease, yet disease control is
more justifiable in economic terms when animals are we11-nourished. The serious
ness of diseases can be measured according to the loss in production they cause,
the cost of their control or the effect they would have if no preventive measures
were taken.
Internal parasites are a major cause of low productivity of small
ruminants. Helminths (Haemonchus^ Ostevtagia, TrichostrongyluSj Bunostomwn)
can cause clinical symptoms such as diarrhoea, but more important lead to reduced
growth rates and fertility. The effect of helminths is most serious towards the end
of the wet season. Fascioliasis occurs throughout Africa as is seen from slaughter
house records of live infections. Its life cycle depends on a snail and the disease
is probably picked up by animals at watering places or in marshy areas.
Dictyocauliasis or parasitic bronchitis causes unthrifitness of young animals,
particularly in highland regions.
Peste des petits ruminants (PPR) is endemic in West Africa. It is a
disease which mimics rinderpest (to which small ruminants are not susceptible)
causing fever, mucosal erosions, diarrhoea and often death. Ticks cause serious
economic loss by skin damage resulting in poor hide quality. Theileriosis in
sheep and goats is usually the benign form, but heartwater is a problem in the
southern half of Africa. Other diseases constraining the productivity of village
flocks are pasteurellosis causing pneumonia, contagious caprine pleuropneumonia
(CCPP), sarcoptic mange, hydatid cysts on lungs and liver, anthrax, brucellosis,
foot and mouth disease, rift valley fever, bluetongue, trypanosomiasis, sheep and
goat pox and clostridial diseases. Details of these diseases are given by BVA
(1976) andlLCA (1979a).
Breeding
Attempts to improve the productivity of sheep and goats earlier this
century in Africa concentrated on the importation of exotic breeds, and indigenous
sheep were "improved" by crossbreeding. This approach was unsatisfactory
except in a few areas such as the highlands of Kenya. One of the main reasons
for failure was the poor quality food available for sheep and goats. Disease, too,
caused problems with imported stock. In particular, importation of animals into
the tsetse belt resulted in high mortality. Some environments in Africa are much
more harsh than others and there may seem to be a place for stratification of the
sheep and goat industries. Stratification depends on moving breeding females
from a poor environment to a better one where they are mated with faster growing
sires to give offspring for meat production. Stratification may have several tiers.
At present the small ruminant industry is not sufficiently structured for such an
exploitation, and there appears to be little possibility of exploiting hybrid vigour.
However, it is possible that such a structure may be developed in the future.
20
The small size of most sheep and goats in Africa has sometimes been
attributed to inbreeding. The amount of inbreeding in a flock depends on how much
movement of animals there is between flocks. Inbreeding is minimised by
communal grazing, by a deliberate policy to swap animals (swapping breeding
animals takes place in Nigeria), by transfer of animals between relatives and
friends, and by financial transactions involving small ruminants. It therefore
seems unlikely that inbreeding is a major cause of small adult size; adaptation to
a harsh environment and poor nutrition are probably more important causes.
Why research fails
Millions of dollars of research fundings are put into Africa each year, yet
the improvement in productivity seems very small. There are numerous reasons
why research fails and some of these (the ones that the research worker can do
something about) are:
the objectives of research are poorly defined. Too much emphasis is put
on "improving SHEEP for MEAT production" or "improving GOATS for
MILK production" without appreciating wider issues. All the time the
farmer and his motives must be considered.
individual research workers or research teams become so involved in
their specialized fields that they do not notice when it is time to take the
acquired knowledge and put it to use.
a research worker becomes isolated from other research workers so that
he may lose sight of his research objectives and his motivation may be
low. Poor communication between scientists working the same field can
result in duplication of research and waste of effort. Scientific meetings
reduce this isolation, and visits of research staff to universities and other
institutions where research goals are discussed are valuable.
research scientists are not integrated into the local community. The
locals think the scientists are different and may resent their higher salaries.
Scientists on the other hand may think the locals are ignorant.
scientists squabble with one another and the goal of the research team
turns from successful completion of the project to individual glory. Some
critical discussion within a group is needed to stimulate ideas, but
feelings of isolation, insecurity and non- appreciation are not constructive.
It is largely the duty of the director of research or project leader to see
his staff remain in an emotional environment where they can function
satisfactorily as scientists.
research reports are unsatisfactory. They may be badly written, fail to
report important details and get wound up in irrelevant facts. Failures
are rarely reported, so that no-one learns from them. Even when
satisfactory reports are produced, their distribution is poor.
21
RESEARCH ON SMALL RUMINANTS IN INDIA
Ruth M. Gatenby
Small ruminants systems in India are in many ways similar to those in
Africa; technical aspects are broadly similar but the sociological conditions are
rather different. The following is a brief look at some aspects of small ruminants
research in India which may be of benefit to research workers in Africa.
Some statistics of India and Africa are given in Table 2. The area of India
is approximately one-tenth that of Africa. Within India there is a wide range of
climatic zones. The lowland areas range from desert to humid zones, with
temperate conditions in the northern states. The human population in India is
more than that in Africa so that the population density is about 15 times greater.
Thus the food production per unit area must be considerably higher to support the
population. The number of cattle in India is similar to that in Africa, and there
are 3 times as many small ruminants in Africa than in India. Nevertheless the
density of small ruminants in India is more than 3 times as great as in Africa.
Table 2. Area and populations of India and Africa. Source: FAO (1981a).
India Africa
Area (ha) 3.0 x 10°
Human population 7.0 x 108
1.8 x 108Cattle population
0.4 x 108Sheep population
0.7 x 108Goat population
Number of sheep and goats
per hectare 0.37
30 x 10
4.8 x 108
1.7 x 108
1.8 x 10®
1.5 x 10
0.11
Agricultural Research in India includes research coordinated by the
Indian Council of Agricultural Research (I.C.A.R.), research in universities
and research undertaken by international organizations. Veterinary and agricul
tural research is being carried out in universities throughout the country.
International organizations such as the World Bank have agricultural schemes
and there are many charities operating in India. However there appears to be
little unilateral foreign aid related to production of sheep and goats. The
I.C.A.R. coordinates a large proportion of research on small ruminants. It
operates three all India coordinated research projects on breeding sheep for fin
wool, breeding sheep for mutton and breeding goats for meat.
22
The Central Sheep and Wool Research Institute (C.S. W.R.I.) was set up in
1962 by I.C. A. R. with the help of U.N. D. P. It is situated near Malpura on the
edge of the Thar Desert in Rajasthan, one of the most important sheep rearing
areas. It is 26°N, annual rainfall is 500 mm and temperatures range from 4 to
430C. A considerable effort at C.S. W.R.I, is being put into the development of
crossbred sheep. Crossing of local carpet wool breeds (Chokla and Nali) with exotic
fine wool breeds (Rambouillet and Soviet Merino) has produced sheep which yield a
2.5 kg fleece with a fibre diameter of about 23 jum which is suitable for processing
on worsted systems. Similarly crossing extremely coarse wool breeds (Malpura,
Sonadi) with fine wool breeds has given a superior carpet wool. Crossbreds for
mutton production have been developed by crossing local breeds with Suffolk and
Dorset rams.
The division of nutrition has been studying the nutritive value of local
grasses and fodder trees as well as improved feedstuffs. Supplementation with
molasses, sugarbeet pulp and urea is being investigated. An interesting project is
underway to evaluate the relative economics of sheep and goats on degraded range-
land. Technical problems of artificial insemination are being studied in the
division of physiology. Climatic physiology studies include determining the water
requirements of different breeds of sheep and the causes of canary colouration i.e.
yellowing of wool. The health section concentrates on diseases observed in
research flocks, and these include pneumonia, Johne's disease, sheep pox,
enterotoxaemia, parasitic gastroenteritis, facial mange and abortion.
The division of wool science at CSWRI has been evaluating local and cross
bred wools, evolving an appropriate grading system and developing methods of
processing. Finally there is an extension division which disseminates research
fundings to the farmers by radio, leaflets, farmers' meetings and individual
contact.
The success of CSWRI in introducing crossbred sheep into the semi-arid
areas is limited because the farmers are unwilling to adopt the innovation. The
price paid by the wool traders for the better quality wool is no higher than for the
local wool. Any increased fleece weight and growth rate are offset by higher
mortality of the crossbred rams and their progeny. Until the harsh conditions of
Rajasthan can be ameliorated there seems little point in trying to improve produc
tivity by introducing exotic blood. At present breed comparisons are made on the
research station where the sheep are zero-grazed with good quality forage and
lambs are allowed to suckle a goat if they are not receiving enough milk from their
own dams. Outside the research station the sheep are managed on a transhumant
system and for most of the year exist on a poor supply of low quality forage.
Unless comparisons of breeds are made under realistic conditions field perform
ance is unlikely to be improved by breeding.
A much more satisfactory approach to improvement by breeding is in
operation at Palmaner in Andra Pradesh. The Animal Husbandry Department buy
local rams which they have selected for weight (at a given age). These rams are
distributed to farmers some distance away and give superior growth rates in the
progency. This practice has been operating in many states since about 1960.
23
The limiting factor is the number of rams which the Department are able to select.
Care must be taken to ensure that the rams are of an acceptable appearance in the
area in which they are distributed; for instance farmers refuse to use a ram of a
colour not normally found in their locality.
In the south of India in Tamil Nadu are the Niigiris hills. The climate here
is pleasantly cool and the local Niigiris breed of sheep was developed by the
European settlers. It produces a fine-wool fleece of about 1 kg/year, whereas the
other breeds in Tanil Nadu are hair breeds. At the Sandynallah sheep breeding
station in the Niigiris hills a breeding project is underway to cross the Nilgiri
sheep with Merinos to evolve a new, larger breed of sheep with a fleece weight of
2.5 kg. The "improved" sheep are meeting with little success in the area because
the majority of land in the hills is devoted to intensive crop and dairy production.
The sheep must survive on steep hillsides, field boundaries and on crop residues.
It is not economical for the farmers to devote resources to their sheep which
operate on a low input - low output basis.
There is little local demand for fine wool (it is sent up to the north of India)
yet the crossbred animals are demanding in terms of nutrition if mortality is not to
be very high. There are only 15, 000 sheep in the area (and 2000 of these are on
the research farm) so that the effort put into breed development seems misplaced.
The government of India is aware of the importance of small ruminants to
the landless peasant. The banks give a loan to the peasant to allow him to buy 15
to 20 ewes and one ram. An agricultural adviser helps in the purchase and sees
that the animals are insured, dewormed and vaccinated. The loan is repaid with a
low level of interest over 2 years. This scheme began several years ago with the
purchase of cows. It has now been extended to sheep and ducks. Goats are not
included in most areas because of their alleged destruction of crops and trees.
Despite the efforts being made to increase productivity of sheep the CSWRI
reported in 1981 that "sheep development activities taken up in different states of
the country in the last few years have not made much impact except in Jammu and
Kashmir." The reasons for lack of progress in the tropical and sub-tropical
areas in India are many. Primarily, the objective of research appears to be
poorly defined. Considerable effort has been put into the development of breeds
for wool production without due consideration of the other uses of sheep (meat,
milk, hides, manure, investment) and without realizing that considerably greater
inputs into the system are required if an exotic breed is to be introduced. The
single-minded breeding approach must be replaced by an interdisciplinary
approach giving attention particularly to problems of nutrition and marketing.
Scientists studying small ruminants in India tend to be geographically
isolated with respect to other areas in India and abroad. Within India there are
different languages and cultures as well as strong caste divisions. The scientists
have no economic incentive to develop packages that are satisfactory, and there
is a temptation to stagnation. Despite good extension activities, farmers are
reluctant to take up the new breeds proposed by the scientists because there is
24
little or no benefit in doing so. Improvements in productivity are likely to come
from more appropriate schemes such as ram selection and distribution, and govern
ment grants allowing motivated people to enter the sheep production industry.
25
TRENDS IN SHEEP PRODUCTION IN AUSTRALIA
Helen Newton Turner
Australia has just over 137 million sheep, all woolled, of which 75% are
Merino and 22% have some Merino blood. A further 3% are other pure breeds,
mainly British breeds used in lamb production. Australian Merinos are of
several genetically distinct strains, ranging from fine-wool (19 microns or less in
average fibre diameter) through medium-wool (average 21 microns) to strong (24
microns or more). Medium-wool strains are in the majority. The main product
is still wool, though meat is increasing in importance.
Management is with year-round open-air grazing on fenced pastures;
there is no shepherding, but dogs are widely used in mustering and handling.
Rainfall in general follows the coast, decreasing towards the inland. In the north
it is monsoonal (summer), in the south-east it is less seasonal while in the south
and south-west the climate is Mediterranean, with dry summers. There has
been considerable improvement of pastures with sown introduced species of
grasses and fertilizer application (mainly super-phosphate) where the rainfall is
adequate (usually at least 500 mm). Supplementary feeding is usually confined to
rams on studs, but sometimes to ewes at mating time, and to all sheep in times
of severe drought.
Sheep are not run in the humid tropics in Australia, and not on the eastern
coast. The Great Dividing Range runs down the east coast; sheep run on the
tablelands up to about the Tropic of Capricorn, and then in an inland area following
approximately the 400 to 600 mm rainfall isohyets in Queensland to the edge of the
monsoon belt. They are run throughout New South Wales west of the Range, in
all Victoria, in eastern South Australia up to about the 125 mm rainfall isohyet,
and in Western Australia in the south-western corner.
Strains and breeds differ in the various regions. Medium and strong wool
Merinos run in the northern areas and in the drier areas further south. Fine-
wools are on the tablelands of NSW and in Victoria and Tasmania. Dual-purpose
breeds (wool and meat) such as the Corriedale, and crossbreds are run in the
cooler, higher rainfall areas. The lamb industry is supplied mainly by the dual-
purpose breeds and by a system in which Merinos are crossed with Border
Leicester or Dorset rams, the crossbred ewe progeny then being mated to
terminal sires, most frequently of the Dorset or Suffolk bred. Some Merinos are
sold as meat. Productivity increases from the north to the south of Australia.
In the northermost sheep areas clean fleece weights are 1.4 - 2. 3 kg per head
and lamb-marking percentages (lambs present at the end of lambing) 20-50.
These figures increase to 2.7 - 3.2 kg (or more) and 80 - 90 percent (or more)
in the south.
26
Australia has a ban on the export of Merinos, instituted in 1929. It was
lifted briefly in 1970, 1972 and again in the early 1980's. The lifting was only
partial, sale of 300 rams being permitted annually provided they were purchased
at specified stud ram sales. Various trade unions objected to the sales, and
refused transport,but with government help the rams were moved. An enquiry was
promised, however, and early in 1982 a team visited various of the purchasing
countries to assess whether the sales were likely to affect the Australian sheep
industry. The report of this team was favourable to continuing to permit limited
sales, and even to extending them to flock rams (as distinct from the higher priced
stud rams).
Genetically, the Merino industry has been structured in the past on a 3-tier
system. The top tier consists of a few closed "parent" studs, and the second of
"daughter" studs drawing rams from only one parent, and "general" studs from a
number of parents. The third tier consists of commercial flocks, drawing rams
from any part of the stud system. Traditionally, selection of breeding stock has
been mainly on visual appraisal. This system certainly developed the many
strains, and increased fleece weight markedly from the middle of last century till
well into the 20th. After that a plateau seemed to be reached, and for the last 30
years scientists and research workers have been urging that further progress
could be made if measurements were included in the selection process. Some
stud-breeders adopted measurement, but some still have not. Since the message
from animal-breeders was also that flocks would make the same rate of genetic
progress as the studs from which they drew their rams, dissatisfaction among
some flock-owners has led to changes in the traditional system. Visually- selected
studs still exist, but side-by- side with other systems, discussed later.
Marketing systems
Wool. Not only were sheep selected mainly by eye, but wool was graded for sale
mainly by eye, being placed into some 2, 660 categories. Number of crimps per
inch was used to assess average fibre diameter, the most important processing
characteristic, while visually assessed staple length, percent clean yield and
colour were also used to determine the grade. From the 1940' s evidence from
research laboratories mounted to show that crimp was a very unreliable guide to
diameter. The challenge to wool by synthetics led to a realization that wool could
no longer depend on unreliable grading; it had to be "true to specification" if it
were to compete with accurately graded synthetic fibres.
This conclusion led in the 1960rs to the formation of the Objective Measure
ment Policy Committee (OMPC), specially funded by the Commonwealth (Australian)
Government and with representatives from all branches of the industry, from
Universities, Government Departments and from CSIRO. The body was charged
with investigating how to institute sale of wool on measurement. The technical
work involved covered determining first what characteristics should be measured,
deciding how to measure them and how to sample the wool in the whole wool-
selling chain. This technical work was in the hands of the Australian Objective
Measurement Project (AOMP).
27
The wool-selling chain started with classing of wool into a multitude of
lines in the shearing shed as soon as it came off the sheep's back. From there
the wool passed to brokers' stores, where it was often classed again, and before
sale there was considerable physical movement of bales within the store, and
opening of them for buyers' inspection.
Reports issued by OMPC in 1973 led to considerable reduction in the
number of classes used for grading wool. It was shown that much of the classing
was a waste of effort; lines were frequently put together again when they reached
the mill. A new schedule, called "Objective Clip Preparation" has been prepared
to replace traditional classing, and is being increasingly used for classing in the
shearing shed.
The OMPC report also listed the characteristics of importance in process
ing (for Australian wools) and worked out sampling and measurement techniques.
All bales of wool submitted in the various categories (called "lines") by each
owner are "core- sampled" with an instrument like a large cork-borer. The
samples extracted are put through machines which measure average fibre diameter
and percent clean yield, as well as amount and type of vegetable contamination.
Grab samples are also taken from each bale and combined to form a display
sample for the line, to be shown at the sale, with the measurements.
In Australia's last wool-selling season, about 96% of auctioned wool was
sold in this way on measured sample, thus presenting a product more likely to be
"true to label". Work is in progress to develop techniques for quick measurement
of staple length, staple strength and colour, the ultimate aim being to sell by speci
fication only, thereby saving even more labour in handling. There were great ob
jections at first from classers in shearing sheds and from buyers, who felt them
selves threatened, but the objections have been overcome and the current system
is working well, backed by measurement laboratories established in all states.
About 80% of wool is sold by auction, but auctioning is to a certain extent
controlled by the Australian Wool Corporation (AWC), a growers' organization but
with some government representation. The AWC operates a flexible reserve price
system for wool and intervenes to protect the grower. Levies are charged on every
bale of wool sold, one being used to fund research and promotion, and the other to
fund the reserve price scheme. Because of the profitable nature of this scheme,
the latter levy is eventually refunded to the grower. Research is both on produc
tion and on processing; techniques for shrink-proofing, moth-proofing, permanent
pleating and drip-drying wool have all been developed. The AWC is linked to the
International Wool Secretariat (IWS), whose members are Australia, New Zealand,
South Africa and more recently, Uruguay. In addition to research laboratories in
the member countries, IWS has a large wool processing research establishment in
Yorkshire, England. It is also responsible for promotion, carrying the results of
processing research into mills in various countries.
Meat, Although the main product of Australian sheep is still wool, there is an
increased interest in meat. This has led to attempts to introduce objective grad
ing schemes for meat, which are so far not successful. The work is being hand
led by the Australian Meat and Livestock Commission. Some States, notably West
28
Australia, have their own grading schemes, but one for the whole country is not
yet in operation. Some progress has been made in uniform grades for live animals,
but these are not yet objective. Market prices for livestock and wool are given
daily over the radio during selling seasons, as well as being published in the press.
There is a growing trade in live sheep for Middle East markets from
western and southern ports. Ships carrying up to 100, 000 sheep call at these
ports, and on-board losses have been reduced to 2-3 percent. There have been
union protests from abattoir workers who feel threatened, and some wharves have
been picketed, but so far no shipments have been completely stopped.
Changes in breeding objectives and industry structure
Increased interest in meat has led to an increased interest in raising
reproduction rate. There has been active as well as passive selection against
twins in the Merino industry - active in that one of a pair would often be deliberately
killed, passive in that a smaller twin would be discriminated against at selection if
its birth type was unknown. In Australia, once-a-year lambing is the general rule.
Research has shown that increased twinning rate is the best avenue for increased
lambing percentage, and there has been response to such selection in the Merino.
In 1954 CSIRO established two Merino flocks, one based on ewes which had
twinned twice at 5 and 6 years of age, one on ewes which had singles in the same
years. When some response to selection was evident, the fact was publicised
through extension journals, and a letter was received from a family who had been
selecting for multiple births for some years, and who offered a ram born in a set
of 5. The family was the Seears Brothers, and their property was named
Booroola. So in 1959 the now famous Booroola flock was added to the CSIRO
experiment, founded on two quintuplet rams, donated in successive years, and
twelve purchased ewes born as triplets or quadruplets.
The Seears had raised their lambing percentage to 170-180 by selecting
only on the ewe side. The lambing percentage of the CSIRO flock has gone up to
220 as a result of selection on both sexes. This makes it the only fine-wool flock
with a reproduction rate as high as this. Recent evidence, presented by research
workers Drs. Piper and Bindon (now in charge of it) has shown that its high
fecundity is due to a single gene. Crossing experiments are being conducted in
many places in Australia and New Zealand, and Booroola rams are in great
demand in the sheep industries of both countries.
Animal breeding research and extension workers in Australia have
advocated for many years that more attention be paid to increasing reproduction
rate, and that measurement of performance be used in selection. Increased
interest in meat has led to more attention being paid to reproduction rate, while
sale of wool on measurement has led to a break-through in the attitude to selection
on measurement. Some studs, indeed,have used measurement for many years,
others have not. As mentioned earlier, many flock-owners have become
dissatisfied with lack of information about progress in the studs on whom they
29
depend. Two movements have followed which have made some changes in the
structure of the industry. Some large flock-owners have decided to breed their
own rams by setting aside a portion of their flock as a ram-breeding nucleus, and
selecting on measurement within it. Others have formed cooperative breeding
schemes.
The system in a cooperative scheme is that members send some of their
best ewes to a central nucleus, which selects on measurement and produces rams
for the whole group. The system may have two or three tiers, the central tier
consisting of multiplying units for the nucleus. Ewes move up one tier, while
rams move down one tier. Annual rates of genetic progress can be increased by
16 percent, because of the larger numbers drawn upon and the increased selection
differential. Cooperative schemes developed first in New Zealand, but have been
taken up in Australia, the largest being one founded very early by Jim Shepherd
in Western Australia. His nucleus now influences about 2 million ewes.
There still are studs as in the former tiered system, and some of them
still select on traditional lines. Many, however, are now using measurement, and
May 1982 saw the formation of Association of Performance Recording Merino
Breeders, on the initiative of a pioneer in measurement, Jim Maple Brown, of
Goulburn, who even uses his own computer to select his rams. There are per
formance recording schemes, backed by computers, run by various State Depart
ments of Agriculture and Universities to aid breeders, though none is a compre
hensive as the Sheeplan scheme in New Zealand.
Other developments
Irrigation is used in Australia, but only to a minor extent to grow pastures
for sheep. Mineral deficiencies have been found in some areas (e.g. cobalt,
copper, selenium etc). Some, such as cobalt, are countered by the administration
of cobalt "bullets", which are slowly absorbed from the stomach.
The cost-price squeeze has been strongly felt in Australia, and has led to
interest in labour-saving devices which increase the number of sheep which can be
maintained by one man. Shearing has for a long time been done by contract, with
teams moving through the country; other operations formerly done by on-farm
labour are now often done by contract, such as crutching (removal of breech wool),
and mulesing (removal of skin on the breech to render it less susceptible to fly-
strike). A great deal of research effort has gone into investigating techniques for
removing wool biologically or mechanically. Biological shearing uses a chemical
which will reduce the diameter of fibres to the point where they can be readily
pushed off while leaving enough staple for protection. Automatic mechanical
shearing involves the use of robots. Neither technique is as yet completely succes
sful. There has also been a move towards "easy care" sheep with fewer skin
folds so they are less susceptible to flystrike; with more open faces so that no
clipping ("wigging") is required; and which will lamb without assistance.
30
Goat production
Interest in goats has increased over the last few years in Australia. There
are unknown thousands of feral goats in the inland areas, many brought in originally
for milk production and rendered redundant with the advent of refrigerated trains
and trucks. For many years these goats have been harvested for meat, much of it
exported, but they are now being caught and re- domesticated, mainly for meat and
fibre.
Observations have been made on growth rates and kidding percentages, both
of which make the goat an attractive proposition for meat production, and some
farms are running them. A mohair industry was started some years ago with
small-scale flocks. When larger -scale enterprises were wanted, shortage of
Angoras was a problem. This has been overcome by using captured feral does as
the recipients of fertilized ova from super-ovulated Angoras and the industry is
expanding.
Some of the bush goats were observed to be carrying cashmere, the most
expensive of all animal fibres. A main source of raw material used to be China,
but as the Chinese are now processing their own cashmere, manufacturing
countries such as Britain and Japan are interested in another source. A few farms
have been started in Australia with recaptured goats. One problem is harvesting
the cashmere. In countries where the fibre is traditionally grown, labour is avail
able to comb out the down, but the cost of this in Australia is prohibitive. A shorn
fleece can be sold, though at a lower price. In the meantime, attempts are being
made to develop a mechanical combing device.
•
METHODS OF COLLECTION AND ANALYSIS OF
FIELD DATA

31
TECHNIQUES FOR FIELD EVALUATION
Helen Newton Turner
Because animals on experiment stations are much more under control than
those in the field, most performance data have been collected on such stations.
Performance on experiment stations is not necessarily the same as that in the
field, however, and information is needed on productivity under field conditions.
This is especially true if breeds are to be compared, particularly if one is an
exotic or new synthetic breed, or the cross of one such with a local breed.
Evaluation is defined here as the comparison of two (or more) groups of
genetically different animals; these may be two indigenous breeds, an indigenous
breed and an exotic (or new synthetic), or an indigenous and its cross with an
exotic or synthetic. More than two groups might sometimes be involved and plans
can be extended.
Before any plan is begun, it is desirable to decide on the characteristics
to be recorded. Many organizations, including the Society for the Advancement of
Breeding Researches in Asia and Oceania (SABRAO) are in the process of estab
lishing data banks for performance records and some coordination among such
banks would be an advantage. A form for the documentation of breeds is given by
SABRAO (1980).
Gathering useful information on the performance of animals in the field is
not easy. Some of the problems encountered include obtaining a representative
sample of households and animals, identification of animals, deciding who owns
the animals, finding the animals for sampling, getting suitable enumerators and
training them, deducing events that happened when the enumerator was absent,
maintaining the owners' interest, maintaining the enumerators' interest and lack
of continuinty of persons collecting, analysing and interpreting data.
One essential feature of evaluation is that the groups under comparison
should be under the same environment and management. If we say this means the
conditions under which the animals are to run, the question that immediately
arises is whether those conditions should be the current ones, or any improved
set which might be envisaged for the future. The question cannot be answered
categorically. If any improvement in conditions is likely to be very long term,
then current conditions are indicated. If short-term improvement is possible, it
would be logical to make the comparison under the improved conditions.
The method of field evaluation that appears to be most feasible in Africa
at present is the collection of information on the performance of flocks of the
different breeds in the field. Comparison of the different breeds will give a
picture of their potential under realistic conditions. Between-flock variation may
be high, but if the number of flocks is high enough, meaningful comparisons can
32
be made. A more serious criticism of this method is that is not a controlled
experiment and differences between flocks are unlikely to be due solely to the
difference in genotype between them, but may be partly a result of different man
agement. For instance, suppose a new breed is being introduced into an area.
The farmers who have the new breed will be those who have the resources or the
imagination to try something new, and together with the new breed they are also
likely to have superior management.
If adequate resources are available for field evaluation and the farmers'
cooperation is assured, then the evaluation can be carried out according to an
experimental plan. Two such plans were drawn up for consideration by SABRAO;
they are termed within-flock and grouped-flock comparisons. Within-flock com
parison requires a number of flocks each containing not less than 100 breeding
females; it is thus possible only in highland range systems and in pastoral systems,
The grouped-flock comparison requires a large number of paired flocks of any
size and thus is suitable for small-holder systems.
Within-flock comparison
For comparison of indigenous breeds (LI and L2) at least four flocks each
of 100 breeding females are required, two of LI and two of L2. Half the females
are interchanged between flock 1 and 2, half between flock 3 and 4, keeping the
same age distribution in each half. This gives 4 flocks, each half LI and half L2.
LI females are mated with LI males, and L2 females with L2 males. Preferably
controlled mating seasons are used, so that the progeny under comparison are of
the same age. Mating can be effected by shepherding separately, by hand-mating
or AI following teasing; probably separate shepherding is the most likely
method. There is no need for individual pedigrees, but offspring should be
identified as LI or L2 by coloured eartag, ear-notch or tattoo and then reared
together, except at mating time. This procedure is continued until 3 lots of
progeny have been born, putting LI females into the LI half and L2 into the L2 half
as they reach breeding age. Female and male replacements are chosen at random.
The comparison of the LI and L2 progeny includes survival rate from birth to
weaning, weaning weight, 6 month and 12 month weights, annual death rate, annual
fleece weight (if woolled sheep), percent offspring born to progeny, percent of
these offspring surviving to weaning, weaning weight of these offspring and milk
per lactation (if applicable).
For comparison of an indigenous breed (L) with an exotic or new breed (D)
the procedure is as before, but with each of four flocks composed of half local
breed and half D. Since the D breed will come from an experiment station, the
displaced L sheep can be held and mated on the experiment station until the com
parison is completed.
For comparison of an indigenous breed (L) with its cross with an exotic or
new breed (D), only males of D will be used. The procedure is complex and the
matings over 4 years are shown in Table 3.
33
Table 3. Matings and observations for within-flock comparison of distributed (D)
and indigenous (local-L) rams
Year Season
Mating Matings
Group Rams Ewes
Progeny
Year
born
Observations
Comparisons
Hams (I) and L) Progeny of D and L
Mating
.Lambing
Weaning
IM.
1.1.
I)L
LL
:»
L;nnb.s lH>ni/cwe joined
Limits weiined/laml> born
Weaning weigiit
Fleece weight and quality
IjUmliH si red
Survival rales anil
|>roductivily at weaning
Mating
Lambing
Shearing
Weaning D
L
Pre- mating D
L
LL
1)1.
LI.
DL
LL
DI,
LL
Ham count from previous \ Survival rates
y ea r
Lambs bom/ewe joined
"
}
Lambs sired
Numl>er surviving
Fleece weight and quality
(if there are 2 shearings
per year)
Lambs weaned/lamb born
Weaning weight
Fleece weight and quality
Number surviving
Body weight
Fleece weight and quality
Survival rates and
productivity at about
10 - 12 m.
Survival rales and
productivity at weaning
Survival rates and
productivity at about
16 - 18 m
Mating General / L L
flock ( D \ L
> same rams
D d) fDL(l)
(LL(l)
Lambing D
Shearing D
? L
Weaning D
Pre-mating D
L
D.DL
D.LL
DL
LL
D.DL
D.LL
DL
LL
Ram count - no other
observations
3 1 Lambs born/ewe joined
3 /
2 \ Number surviving
2 f Fleece weight and quality
( (if there are 2 shearings
I per year)
3 > Lambs weaned/lamb born
3 f Body weight
2 J Number surviving
2 ) Body weight
Fleece weight and quality
Survival rates
Reproductive performance
Survival rate and
productivity at about
10 - 12 m
Reproductive performance
Survival rate and
productivity at atout
16 - 18 m.
Mating General
flock
Lambing D
Shearing
1
Weaning D
L
L
same rams
DL(1)
DL(2)
LL(1)
LL(2)
D.DL
D.LL
D.DL
D.LL
I }
1 )
Ram count - no other
observations
Lambs bom/ewe joined
No observations
Lambs weaned/lamb born
Survival rates
^Reproductive performance
) Reproductive performance
Body weight
This schedule gives, within individual flocks:
(i) Comparison of survival rates over 4 years, and number of lambs sired over 2 years, for D and L rams
(ii) Comparison of the productivity of the progeny bom in the first two years, for:
Survival rate birth to weaning and weaning to first mating (assumed to be at lj years)
Body weights at weaning and first mating
Fleece weight and quality at weaning and first mating (and at an intermediate shearing if there are 2 shearings a year)
Reproductive performance, over 2 years for profjeny born in Year 1, and one year for progeny born in Year 2.
Note.
There are never more than 2 mating groups in a flock. In Years 3 and 4 the important comparison is for the reproductive performance of the
DL and LL progeny, which must be mated to the same rams. It is suggested that these might be D rams, and to make the numbers up to half
of the total breeding flock some of the L ewes be added to them. The remainder of the L ewes can be maded to L rams, and a longer
comparison of survival rates of D and L rams \w>uld be obtained. An alternative choice would be to have only one mating group in Years 3
and 4, all ewes (L, DL and LLO being mated to either D or L rams. Information on a comparison of ram survival rates in those years would
then be lost.
Notation: D - distributed rams;
L - indigenous (local) rams and base ewes
DL - progeny from D rams and L ewes; DL(1) progeny born in Year 1
LL - progeny from L rams and L ewes; LL(1) progeny born in Year 1
34
Grouped-flock comparison
Since it is not always practicable to find flocks of 100 breeding females an
alternative is to have a number of small flocks. When two genetic groups are
under comparison, the flocks should be paired as far as possible for environmental
and management conditions. There should be at least 16 small flocks, 8 of which
will run one of the genetic groups under comparison and 8 the other. The same
observations should be made as for the within-flock comparison.
Comparison of two indigenous breeds (LI and L2) requires 8 flocks to run
LI, and 8 L2. In each case the animals in the flocks should be representative of
the breed concerned. It would again be preferable to have controlled matings, so
that the animals compared are of the same age and progeny should be identified in
case of mixing for grazing.
Comparison of an indigenous breed (L) with an exotic or new breed (D) is
not so easy to manage, as it involves replacing all L animals in one set of flocks
with D. If this can be done, the procedure is as for comparison of indigenous
breeds. A plan for the comparison of an indigenous breed (L) with its cross with an
exotic breed (D) is shown in Table 4.
35
Table 4. Matings and observations for grouped-flock comparison of distributed
(D) and indigenous (local-L) rams
Year Season
Mating Flock Mating
Group Number Bams Ewes
Progeny
Year
born
Observations
Comparisons
Bams (D and L) Progeny of D and L
Mating D 1-8 D L
L 9-16 L L
Lambing D 1-8 DL 1
L 9-16 DL 1
Weaning D 1-8 DL 1
L 9-16 LL 1
Mating D 1-8 D L
L 9-16 I. L
Lambing D 1-8 DL 2
L 9-16 LL 2
Shearing D 1 - 8 DL 1
•}
L 9-16 LL 1
Lambs born/ewe joined ) Lambs sired
Lambs weaned/lamb born
Weaning weight
Fleece weight and quality
Survival rates and
productivity at weaning
3 Mating D
L
D
L
Lambing D
L
D
L
1 1 Shearing D
Weaning D
L
D
L
Pre-mating D
L
Ram count from previous
year
Lambs born/ewe joined
Number surviving
Fleece weight and quality
(if there are 2 shearings
per year)
Survival rates
Lambs sired
Survival rates and
productivity at about
10- 12 m
Weaning D 1-8 DL 2 \ Lambs weaned/lambs born Survival rates and
L 9-16 LL 2 \ Weaning weight productivity at weaning
' Fleece weight and quality
Pre-mating D 1 - 8 DL 1 \ Number surviving Survival rates and
L 9-16 LL 1 ; Body weight productivity at about
16 - 18 m) Fleece weight and quality
1 - 4
5-8
9-12
13 - 16
1-4
5-8
9-12
13 - 1G
1 - 8
9-16
1-4
5-8
9-12
13 - 16
1 - 8
9-16
{ L
(DL(1)
\DL(1)
I LL(1)
1 LL(1)
DL
D.DL
LL
L.DL
DL
D. LL
LL
L. LL,
DL
LL
D.DL
L.DL
D.LL
L.LL
DL
LL
. Ram count from previous
year
Lambs bom/ewe joined
Number surviving
Fleece weight and quality
(if there are 2 shearings
per year)
Lambs weaned/lamb born
Weaning weight
Number surviving
Body weight
Fleece weight and quality
Survival rates
Lambs sired:
(1-4)+ (9-12)
versus
(5-8)+ (13-16)
Reproductive performance
(DL) v (LL)
(1-4) + (5-8)
V
(9-12)+ (13-16)
Survival rates and
productivity at about
10 - 12 m
Reproductive performance
(as for lambs born)
Survival rates and
productivity at about
16 - 18 m
4 Mating L
DL(1)
DL(2)
L
DL(1)
DL(2)
L
LL(1)
LL(2)
L
LL(1)
LL(2)
Ram count from previous
year
Survival rates
Lamhing D 1 - 4
I. 5-8
1)
L
9-12
Li - 1*
Shearing 1) 1-4
f
L
!t - 12
5 - 8
Li - Hi
Weaning 1)
1.
1 - 1
r, - 8
Ii
I.
!) - 12
l :\ - l <;
DL
D. DL
LL
L.UJL
DL
D. LL
LL
L.LL
D.DL
L.DL
D. LL
I.. 1,1.
Lambs born/ewe joined Lambs sired:
(1-4) + (9-12)
versus
(5-8) + (13-16)
No observations
l„iml>s wcuicd/himli
Weaning weight
Reproductive
performance
(DL) v (LL)
(1-4)+ (5-8)
(9-12) + (13-16)
Reproductive
performance
(as for lambs born)
36
37
EVALUATION OF BREED PRODUCTIVITY IN AFRICA AND ILCA
RESOURCES FOR DATA ANALYSIS
J. CM. Trail and J. Durkin
In planning for increased productivity of animals in tropical environments,
breeders need information on the performance of major animal types in different
ecological zones, production systems, management levels, etc. Logical decisions
on selection between breeds require both (a) comparisons made in the same
environment, and (b) information on a sufficient number of performance traits to
allow an acceptable index of overall productivity to be constructed. An assess
ment of past research work which satisfies these criteria can save on future in
puts. The first part of this paper comprises an examination, using cattle as an
example, of the quantity of past research work and its value in determining the
comparative performance of breeds in Africa. Secondly there is a description of
the present ILCA resources available for analysis and interpretation of field data.
Quantity and value of previous animal productivity research in Africa
Previous research on the productivity of cattle breeds in Africa has been
assessed and is described here. There has been no similar assessment for sheep
and goats although the procedure would be parallel to that for cattle. In a
bibliography (Trail, 1981) covering performance aspects of indigenous, exotic and
crossbred cattle in Africa south of the Sahara, approximately 550 items are listed,
of which 50 are review or descriptive papers. The remaining 500 contain objective
original data on some aspect(s) of reproductive performance, growth, viability or
milk production. This bibliography was built up over a three-year period as a
by-product of the studies of production systems being carried out in various zones
of Africa by the International Livestock Centre for Africa. It. covers the 30-year
period, 1949 to 1978. For the purpose of evaluation, each of the 500 items was
classified by year of publication, type of publication, language, ecological zone,
cattle group, and performance trait(s) measured.
The 500 references are distributed throughout the five ecological zones as
shown in Table 5. There appears to be some positive correlation between cattle
numbers and publications produced, although the very arid zone, where conditions
are extremely harsh and the temperate highland zone where conditions are
relatively favourable, are both the subject of few publications relative to their
cattle populations and the tsetse infested zone has been the subject of a consider
ably larger proportion of publications.
The number of publications has increased sharply over the past three
decades: 9% were produced from 1949 to 1958, 38% from 1959 to 1968 and 53%
from 1969 to 1978. Fifty-eitht percent of the publications were produced in
English, 38% in French, 2% in German and 2% in Portuguese.
38
Table 5. Percentages of cattle and bibliography references in the five
ecological zones
Ecological zone (rainfall)
Percentage of Percentage of
cattle items in
population bibliography
6 2
22 22
44 51
22 9
6 16
Very arid ( < 400 mm)
Arid to semi-arid (400 - 600 mm)
Semi-arid to humid ( > 600 mm)
Temperate highland
Humid, tsetse infested
With reference to types of publication, 52% of the items listed are articles
in scientific journals, 17% appear in bulletins or university or FAO publications,
15% appear in research station reports, 9% are papers presented at regional
meetings, 4% are theses, and 2% are papers presented at international meetings.
(Annual reports covering the same topic over several years were counted only
once).
Eighty-one percent of the items contain information on indigenous breeds,
30% on crossbreeds and 14% on exotic breeds. With reference to performance
traits, 72% of the publications contain information on growth, 41% on reproductive
performance, 33% on milk production and 18% on viability.
Among publications on indigenous, crossbred or exotic cattle types, the
percentage of publications containing information on the four main performance
traits are shown in Table 6. Reports on milk production thus represent a much
larger proportion of all publications on exotic breeds than on indigenous breeds,
with crossbreeds in an intermediate position. The distribution of reports on growth
runs in the opposite direction, while reports on reproduction and viability are
distributed evenly among all three cattle types.
Table 6. Distribution of performance traits by cattle types
Cattle types
Percentage of items covering:
Milk Growth Reproduction Viability
Indigenous 27 78 41 19
Crossbred 37 65 41 20
Exotic 48 57 41 17
39
To sum up the results of the overall analysis of this bibliography, it shows
that only about 20% of the references contain information on three or more perform
ance characters sufficient to allow characterisation of breed types through a pro
ductivity index. For example one simple productivity index used is 'weight of calf
plus liveweight equivalent of milk produced per unit weight of cow maintained per
year', this index being extended to cover more traits if information is available.
Additionally only 20% of the references contain comparative information on two or
more breed types. When these two necessary attributes are put together, only 5%
or 25 of the reports provide sufficient data to allow breed comparisons on the basis
of a productivity index. This illustrates that the majority of research work does
not give useful information about the comparative performance of cattle breeds in
Africa. If a similar analysis was performed for small ruminants it is likely that a
similar (or worse) picture would emerge: i. e. breed comparisons must depend on
a small amount of relevant information.
Present ILCA resources for analysis of current animal productivity
research
In 1980, ILCA purchased a Hewlett-Packard 3000 series III computer
system. The system consisted of the Computer Processor, two 300 line-per
minute printers, two 120 million- character on-line disks, a nine-track magnetic
tape unit, a 300 card-per minute reader, and 12 interactive terminals. An extra
5 terminals have now been added because the need for data entry and analysis has
increased.
ILCA cannot afford to develop large computer software, that is computer
packages comprising sophisticated programs. Programmers today are very
expensive relative to the cost of computer hardware. As a result the ILCA
Computer Unit has sought out and purchased software already written for the
HP 3000 III. This software is then supported by larger organisations with many
programmers and it is repeatedly being enhanced. For Research, ILCA has
acquired SPSS (Statistical Package for the Social Sciences) and BMDP (Biomedical
Computer Programs) from McMaster University, Hamilton, Canada; LINDO
(Linear Programming Package) from Chicago and "Harvey's Least Squares" from
Ohio State University.
While the design of such complete packages are beyond ILCA's resources,
the programming staff are frequently writing smaller programs to edit, verify
and construct data files for use by the larger software packages.
The computer system is used predominantly for analysis of scientific
research data, mainly agricultural production statistics. SPSS and BMDP are
used extensively in ILCA and they allow the scientist to tabulate and statistically
analyse his data quickly and efficiently. SPSS developed by Nie, Hull, Jenkins,
Steinbrenner and Bent (1975), has been updated to its present version SPSS 9. 1.
It allows many difficult, time-consuming and generally unrewarding tasks to be
undertaken using a variety of single purpose computer programs, each with it own
idiosyncratic control-card syntaxes and input data formats. The researcher thus
40
spends less time as a data-preparation clerk and more time as a scientist
analysing substantive results.
One very simple SPSS procedure is tabulation of data to show frequencies so
that errors can be isolated very quickly and corrections can be made prior to the
larger more expensive Least Squares Analysis. In SPSS each variable and values
in certain variables are labelled so that data verification can be performed very
easily and the 'breakdown' procedure gives a provisional estimate of the means for
different traits. LINDO is used by researchers to analyse resource allocations
within the constraints of various farming systems. Harvey's Least Squares
(Harvey, 1977) is used to analyse animal productivity data and is especially valua
ble if genetic parameters have to be estimated.
The unit provides statistical analysis for staff from all ILCA field pro
grammes and has started to analyse data from national research institutes. In the
past twelve months, work has been completed for the Institute of Agricultural
Research in Ethiopia and the Centre de Recherche Zootechnique in Kolda, Senegal.
It is hoped that over the next twelve months assistance with data analysis will be
given to more research institutes.
In an effort to streamline data collection from ILCA's field programmes,
micro-computers will be installed within programmes. These computers will
enable scientists to more easily verify their data and undertake initial analysis in
the field. Eventually it is hoped that a complete data collection network can be
set up using the micro-computers. In all interaction with computers, the coop
eration between scientist and programmer is of utmost importance. The scientist
must ensure the biological correctness of his results, and the programmer must
perform analyses by the simplest, fastest techniques.
41
HANDLING AND ANALYSIS OF PRODUCTION DATA
J. CM. Trail and J. Durkin
Computer handling and analysis of smallstock productivity data involves a
number of basic steps:
1. Extraction of relevant records from field data sheets and preparation
of coding sheets
2. Entry of data into computer file
3. Validation of data, necessitating examination and correction
4. Calculation of parameters, construction of logical sub-classes for
environmental effects, and preparation of files for analysis
5. Statistical analyses
6. Interpretation
7. Peripheral studies linking statistical outputs with biological effects.
In order to demonstrate these steps, records of birth and weaning weights
of 88 lambs on 2 farms were analysed. These lambs were born during a three
month lambing season and their individual sires were recorded. This small sub
set was treated as if it was a large block of legitimate data and all the individual
steps in an analysis to determine environmental and genetic parameters were
illustrated.
The SPSS package was used to validate the data, into which deliberate
errors had been introduced; a simple program was written and used to construct
an adjusted weaning weight and a pre-weaning daily liveweight gain; and Harvey's
Least-Squares and Maximum Likelihood Program (LSML 76) was used to evaluate
all possible environmental influences, and obtain least biased estimates of genetic
parameters.
The environmental effects of station, month of birth, dam parity number,
sex, and type of birth (single or twin) on the three traits birth weight, weaning
weight adjusted to 4 months and daily liveweight gain from birth to weaning, were
estimated. In the case of genetic parameters, the heritabilities of, and the gene
tic correlations between the three traits could be estimated through a paternal
half- sib analysis.
1. Extraction of records and preparation of coding sheets
Table 7 illustrates a typical lambing record sheet with records added daily
as iambings occur. Each lamb was assigned a unique number at birth, its dam
and sire numbers, sex and type of birth noted, and birth weight was recorded.
42
Table 7. Field lambing record sheet.
Date of
lambing
Ewe
No.
Lamb
No.
Sire Sex
Single/
twin
Birth
Commentsweight
(kg)
A-|L-?o *C4S 5lt i£ <? T — died o& birfc.
X- IL-8.o M3 til xs $ T 2/7
JL-/L-*o iio C(Z 31 * T XI
2-(L- So iio Siq. 31 * 7 - cMi-d od" brtt*
i-/l-«T6 -UT fr/S 2.S <? •S X?
S-ll -5o Aqo 8"f6 2.T c? S 2..*
^-/1-Jd Z3l 5"ll SU. $ £ *M.
Table 8 illustrates a weaning weight record sheet, produced once per
month in this example, so that all lambs at weaning were between 4 and 5 months
of age. The lamb number, its date of birth, sex, dam number, and actual weight
were recorded.
Table 8. Weaning weight record sheet.
Weaning
weight
(kgs)
Date of
birth
Weaning
date
Lamb No. Sex Dam No.
xJ\l\eo *
o*
MS
Xlo
Xl°l
/ /Mcuj H?/
Sli JL.I H-l to
± jl^jto
j//J-/*o
i* jujto
IJL.Z.
Ill5/£
5/6 3 mo
x±\5/1 /is M
Table 9 lays out the coding sheets produced from the two field records,
containing the information required for data checking and actual analyses. The
station number, sire number, month of birth, type of birth, sex, and parturition
43
number are used in identifying the environmental and genetic classes for the
analysis. The date of birth, date of weaning and weaning weight are used in the
construction of adjusted weaning weight and daily liveweight gain, which together
with birth weight comprise the three traits for analysis.
Table 9. Coding sheet
s LAMB
. NO
DAM
NO
SIRE
NO
DATE OF
BIRTH
T
B
s BIRTh p WEANING
DATE
W W '
T E
X
WT.
Uxio
N jtqXlO
A
2 M 6 8 10 12 M 16 18 2C 22° 24 26 28 SoL 32 34 36 3S
S I 3> Z I o 3 t ck I z & 0 2 2 Z 1 2 0 1 o s 6 ! I
S i < Z I 1 Z $ a j, i z s o ( I 2s 2 0 I 0 S 5 1 2*1
s ! b .I k 0 Z r1 c J I 2 8 0 ( ( 2 % 2 0 ( 0 S 8 i ZT
* i 7 2 i i s" if 0 u i 2 2 o I 2 Z q- i 0 1 0 s 5 1 I ?
& i 8 Z u M- 4 2 0 s i I 10 o 2 2 Z s t 0 1 o <{- $ i ££ -
« i 1 Zu i 4 S 0 b i 2 S 0 ( i i i 2 o / 0 5 f 2 o
e 2 0 Z I 3 i> q- 0 ") i 2h $ c 1 f 1 i i fil 1 c s S I 2*
S Z i Z4 2 Z s 0 s i 1 5 0 1 i 2 s 2 0 ( oa rS j Z /
$ z3> 2 S" o 2\ °i o s 1 1 ? 0 / I Z <s 1 0 ! oa % i Z (
s 2 4. 2 ,.x ^2 0 J 1 O * 0 Z 1 Z u 1 0 ( o I s i Z/4. 1s I S Z J *) <U 2 0°/ < 0 ? 0 1 2 o 2 °l i 0 I dil 2 6 J— . . ~i
2. Entry of data into computer file
Data entry would be either through punch cards or directly into core.
Table 10 is a listing of the records entered; note from the 12th lamb onwards,
lamb numbers and dam numbers have been omitted.
3. Validation of data
For checking and identification of errors, the SPSS package was used to
produce relevant frequencies, cross tabulations and breakdowns. Table 11 lays
out the basic information required by the SPSS package covering the variable list
and labels (e.g. ST station number, LAMB lamb number, etc); the input format
(e.g. fixed, field 1 = col 1, field 2 = cols 2-5 etc); the value labels (e.g. in sex,
1 = male, 2 = female, etc); and the first task name, frequencies of all relevant
parameters (e. g. station, month of birth and year of birth in Table 12 and type of
birth, sex and dam parity number in Table 13).
Tables 12 and 13 indicate the outputs from the frequency checks on station
number, month and type of birth, sex, and dam parity. Obvious errors are that
one lamb appears under a station code 3, one was born in August, one was born
in 1979, one has 0 for type of birth, one has 0 for sex and one has 0 for dam parity
number.
44
Table 10. Listing of records entered from coding sheet
31 5 233 46 2126322 2 7 2 0 ' 5 31 22
31 5 23 > 2532126011 28 2 0 " 5311 29
51 » 24 3 2 9 1 t801 1 >8 2 0 1 5 31 2>
31 7 271 54 ^u^fc.)12 24 1 c . 5i1* 1l
31 5 244 42 5117922 2 5 1 0 * 4311 33
6 31 9 241 45 5126331 51 2 c 531 2 J
7 32 J 2 3 3 54 71283 1 51 1 r . 531 2 i
3 521 242 25 311K01 28 2 0 4^1 :\
9 323 252 27 3118011923 1 C 431 21
1 J i2w 25: 42 2108021 24 c 3311 2;
11 32 5 237 4 2 9133012 50 ~t 0 ' 331 26
12 a J 4510126312 2 6 1 01 5?1 2 4
M j 3 2 511118012 26 1 0 631' 25
1* j j 2511128021 29 2 c . 5 30 25
15 j J 2911118012 26 1
0 • 431 >23
16 j 3 3111128021 29 2 0 ' 531 Z*
17 j 3 5413118022 29 2 0 ' 4 51 27
M j 3 231411801 I 53 1 Ci 4 51 2r>
19 3 j 2914118011 33 2 0 'I 431 2 5
20 3 3 5414118011 32 2 0 431 2<
21 J 3 2 31710802 2 22 2 c 331 3D
22 J 3 2517128012 23 1 0 ' 5.M 72
23 J J 2917106022 22 2 c 3*1 3j
24 3 3 3117128012 23 1 c I 331 32
25 J J 4217108011 52 2 0 *I 381 3 )
26 0 3 451712801 I 29 1 c I 531 25
27 J 3 4523118012 23 2 0 I 431 27
2a 3 3 4221108012 2t 1 0 I 331 3 4
29 3 3 5421108022 25 1 0 1 331 23
33 3 3 2322108011 31 1 c 1 331 33
31 3 3 2922106011 31 1 0 1 331 3J
32 3 3 2523128011 28 1 G 1 531 35
33 3 3 3123128011 28 1 0 1 531 33
34 3 3 5423108021 30 1 0 1 331 35
35 3 3 2324108012 25 1 0 1 331 34
36 3 3 2924108012 25 1 0 1 331 34
37 J 3 4224108011 27 1 0 1 331 136
33 3 3 2325108011 26 2 0 1 331 37
39 3 3 2925108011 26 2 0 1 331 137
40 j 0 4526118021 29 2 c 1 431 175
*1 3 3 5429108011 29 2 0 1 331 135
42 2 3 3 95 1128311 53 2 0 1 581 33
43 2 3 3 97 1118011 27 1 c 1 331 29
44 2 3 0 95 3128012 30 1 0 I 581 17
45 2 3 3 87 4118012 28 1 0 1 431 21
46 2 3 D 89 6128022 29 2 0 1 531 19
47 2 3 3 95 7108011 29 2 0 1 331 33
43 2 3 3 89 8128021 31 2 0 1 531 22
49 2 3 3 9o 3128012 36 1 0 1 531 32
5J 2 3 3 93 3128011 31 2 c 1 581 73
51 2 3 3 87 9118011 22 1 0 1 431 24
52 2 3 3 95 9128011 31 1 c 1 531 25
53 2 0 3 8711108012 32 1 0 1 331 25
54 2 3 3 8911128012 52 2 0 1 531 123
55 2 3 J 9611128011 33 1 0 1 5 31 1 3
56 2 3 3 9311118022 25 1 0 1 431 134
57 2 3 ] 8914128011 30 1 0 1 5 31 31
53 2 i 3 9514118012 33 2 0 I 4311 23
59 2 3 3 9714108021 26 1 0 I 381 37
63 2 3 3 9314128012 28 1 0 'I 531 13
61 2 3 3 9515118021 31 2 0 4811 23
62 2 3 3 9915118012 33 2 0 1 431 19
63 2 3 3 9316128011 29 1 0 531 23
64 2 3 3 8717108021 28 2 0 331 31
65 2 3 3 9617118022 31 2 0 t 431 22
63 2 3 3 9917118011 32 2 0 I 431 21
67 2 3 3 9513118022 27 1 0 I 431 26
63 2 3 3 9e13118011 34 2 c I 4311 22
69 2 3 3 9720108012 26 1 0 1 381 23
70 2 3 3 8721108012 27 2 0 I 331 33
71 2 3 3 9921108022 25 1 0 I 331 24
72 2 3 3 9522108021 28 1 0 I 331 23
73 2 3 3 9622108022 27 1 G 331 25
74 2 3 3 9722108011 32 2 0 '1 331 23
75 2 3 J 9822118012 23 2 c I 431 23
76 2 3 3 8923128012 50 1 c 5 31 33
77 2 3 3 9o23108021 25 1 0 'I 3 31 27
73 2 3 3 8724108011 28 1 0 'I 3 31 35
79 2 3 3 9"24106021 33 1 0 'I 381 26
83 2 3 ] 9525108022 28 2 c I 3311 29
81 2 3 3 892711801 1 30 2 0 1 431' 1 ?
62 2 0 ) 9727108022 26 2 0 4 3 311 19
63 2 3 3 9927108012 29 2 c 3311 23
84 2 j 3 9523138011 28 1 0 ' 33r 27
85 2 3 3 9=23108311 31 2 0 '1 331 33
85 2 3 j 9323118011 33 2 c I 431 23
87 2 3 3 9530136012 28 1 0 'I 331 32
83 i 3 3 25 2126022 27 2 0 'I 531 22
45
Table 11. Basic information required by SPSS package.
SPSS BATCH SYSTEM
SPSS FOR HE hP/3030 SPR ING-RELE AS z, VERSI3N 9.1/
DE=AULT SPACE ALLOCATION..
WORKSPACE £2752 3YTES
TRANSPACE 324s 3YTES
ALL3WS FOR.. 32 TRANSFORMATIONS
133 RECOOE VALUES ♦ LAG VARIABLES
523 IF/COMPUTE OPERATIONS
1 RUN NAME
2 «=IL E NAME
J IN°UT MEDIUM.
4 'AGESIZE
5 VARIAbLE LIST
INPUT FORMAT
ILCA 1982 OATA ANA.YS1S SMALL RUMINANTS
COURSE
OISC(*IN)
NOEJECT
ST/LAM? /DA «N, SIREN/ OB /MB/Y3/TB/SEX/BHT/PARITY/
QK/MW/YW/W>JT
FIXE0(F1.0,3F4. 0/3=2. 0/2F1.0/F3.1/=2.0/4X/
3F2.3/F3.1)
ACCOROING T3 YOUR INPUT FORMAT/ VARIABLES ARE TO 3E READ AS FOLLOWS
VARIA3LS F3RMAT REC3RD COLUMNS
ST F 1. '.! 1 - 1
LAMS F 4. ~} 2- 5
DAMN F 4. 2> 6- 9
S I R 2 N F 4. .J 10- 13
C3 F 2. J 14- 15
Xi F 2 . 0 16- 17
Y 1 F Z . j 18- 19
T1J F 1 . j 20- 30
s;.x F 1 . J 21- 21
t4l F
■;
1 22- 24
PARITY F 2. 0 25- 2o
DW F 2 . ; 31 - 52
f « F 1' . j 33- 5*
Y* F 2. o 33- 36
W -T F 3. 1 37- 39
THE INPUT FORMAT PROVIDES FOR 15 VARIABLES. 15 WlL. 3E REAO
IT PROVIDES F03 1 R2C3RJS ('CARDS') PER CASE. » MAXIMUM 3F 39 'COLUMNS' ARE USED ON A R2C0RJ.
1 J
1 I
1 2
1 5
1 4
15
1 i
1 7
1 1
1?
*J
21
22
22
24 VALUE l.A;EL!
25
2i
27
2o
2V TASK NAME
3J FREQUENCIES
31
ST/STATI3N/
LAMB/LAM3 NUMB.:R/
DAMN/DAM NJM3ER/
SIREN /SIRE N U M 3 E R /
OB /CAY Oc ilRTH/
M?, MONTH OF 3IRTH/
YS/YcA* 3F BIRTH/
TB/TYPi OF BIRTH/
SEX/SEX /
8WT/3IRTH •EIGHT/
PARITY/OAM PARITY DUMBER/
OW/WSA <ING DAY/
MW /WEARING M3NTH/
Vk/W£ANING Y:A3/
WWT/WEANIN3 .VEIGMT/
SEX(1)MAlE(2)FEMAL;/
TB(1)SINGLF(2)TWINC3)TRIPL£TS/
"IB/M^(1)JANUARY(2)CE3RUARY(3)MARCH(4;APRIL
C5;M4Y(6;j'JN;<7)JU-Y(S)A'UGUST(9)SE°TEM3ER
<10)3CT03ER<11 )NDVEM3ER(12)DECEM3ER/
FREQUENCIES ALL RELEVANT PARAMETERS
GENERAL = S T/ SI RE M/ 06 / M3/ YB/T 3/ SE X, * A R I T Y,
Dw/*W/YW
GItH WORKSPACE ALLOWS FOR 1250 VALUES AND 437 LA3F.LS PER VARIABLE C0R 'FREQUENCIES'
AFTER READING 38 C-ScS FRCM SU3FILE COJRSE END OF OATA WAS ENCOUNTERED ON LOGICAL UMT P 8
46
Table 12. SPSS frequencies for station, month of birth and year of birth.
IL:A 1^82 JATA ANALYSIS SfcALL RUMINANTS
=RcQJENCIES ALL RELEVANT PARAMETERS
FltE COURSE (CREATION OATE = 12/13/82)
ST STATION
CATEGORY LABEL
RELATIVE ADJUSTED CUM
ABSOLUTE FREQ FREQ FREQ
CODE FREQ (PCT) (PCT) (PCT)
1 . 41 1,6.6 46.6 4 6.6
1. 16 52.3 52.3 98.9
3. 1 1.1 1.1 100.0
TOTAL 3? 100.0 100.0
VALID CASES 83 MISSING CASES
MB MONTH OF EIRTH
CATEGORY LABEL
AUJUST
0CT03ER
NOVEMBER
DECEMBER
VALID CASES £3 MISSING CASES
TB YEAR OF rjIRTH
RELATIVE ADJUSTED CUM
ABSOLUTE FREQ FREQ FREQ
CODE FREQ (PCT) (PCT) (PCT)
3. 1 1.1 1.1 1.1
1D. 35 39.3 39.8 40.9
11. 25 28.4 28.4 69.3
12. 27 30.7 30.7 100.0
TOTAL 38 100.0 130.0
CATEGORY LABEL CODE
ABSOLUTE
CR£Q
RELATIVE
FREQ
(PCT)
ADJUSTED
FREQ
(PCT)
CUM
FREQ
(PCT)
7*. 1 1.1 1.1 1.1
BO. i7 98.9 98.9 100.0
TOTAL °8 100. 0 100.0
VALID CASES l.> MISSING CASES 0
47
Table 13. SPSS frequencies for type of birth, sex and dam parity number.
RELATIVE ADJUSTED CUM
ABSOLUTE FR5C FRSQ FREJ
CODE FRcC (PCT) (PCT) (PCT)
2. 1 1.1 1.1 1.1
1. 50 67.0 67.0 68.2
2. 27 30.7 30.7 93.9
J. 1 1.1 1.1 103.0
rOTAL 38 130. 0 100.0
RELATIVE ADJUSTED CUM
ABSOLUTE FR:Q FREQ FREQ
CODE FREQ (PCT) (PCT) (°CT)
3. 1 1.1 1 .1 1 .1
1. 47 53.4 53.4 54.5
2. 40 45.5 45.5 100.0
0T4L <.t 100.0 100.0
RELATIVE ICJUSTcO CUM
ABSOLUTE FREC CPEC FREQ
ODE FR:C (PCT) (PCT) (PCT)
0. 1 1.1 1.1 1.1
1 . 45 51.1 51.1 52.3
? . 41 46.6 4e.6 93.9
3. 1 1.1 1 .1 100.0
il:a 1?82 oata ANALYSIS SMALL RUMINANTS
=R£QJENCIES ALL RELEVANT P AR AMETERS
FIlE COURSE (CREATION DATE = 12/13/82)
TB TYPE Op ilRTrt
CATEGORY LABEL
SINGLE
TWIN
TRIPLETS
VALID CASES 83 MISSING CASES
SEX SEX
CATEGORY LABEL
MALE
FEMALE
VALID CASES 83 HISSING CASES
PARITY DAM PARITY N'JHiER
CATEGORY LABEL
TOTaL .iE 130. C 100.0
VALID CASES Hi MISSING CASES
48
The programme lists the identification numbers of the anomolous types,
for reference back to the coding sheets and correction. Acceptable ranges for
parameters can also be indicated, so that, for example, lambs with birth weights
below 1 kg and over 5 kg might be listed.
Table 14 lists a crosstabulation requested between sire number and month
of birth of lamb. In all genetic analysis it is vital to check that there is no con
founding of time with use of individual sires, or apparent genetic effects may
actually be time period effects. If interaction effects are to be estimated it is
essential to identify any subcells with missing values before commencing the main
analyses, and crosstabulations are a useful method.
Table 15 lists an example of a breakdown, in this case birth weight by
sire number, both to give an indication of the possible importance of sire effects
and pinpoint any gross data errors (e. g. sire 29).
After adjustment of all errors, a rerun of SPSS confirms that all
corrections have actually been carried out. Table 16 indicates the corrected
picture of station frequencies from Table 12.
Table 16. SPSS frequencies for station, after data correction.
ILSA 1^82 DATA ANALYSIS SMALL RUMINANTS
FREQUENCIES ALL RELEVANT PARAMETERS
FIuE CDURSE (CREATION DATE = 12/13/82)
ST STATION
CATEGORY LA3cL CODE
1 .
ABSOLUTE
FRiC
u2
*6
VALID CASES b 5
TOTAL 68
MISSING CASES
RELATIVE ADJUSTED
eR=C
c p ; t >
h7.7
32.3
130.0
FREQ
(PCT;
47.7
52.3
100.0
CUM
F R E ,w
(PCT)
U7.7
100.0
Table 17 is a listing of the corrected records and should be compared
with the original listing in Table 10.
Sorting of a file is very often required before analyses, and Table 18 lists
the records after use of a standard sort program to arrange their sequence on
sire number within station. Thus the order is progeny of sires 23 to 54 within
station 1, followed by progeny of sires 87 to 99 within station 2.
49
Table 14. SPSS crosstabulation between month of birth and sire number.
IL:A 1 >82 JATA ANA.YSIS SMALL <?U^I:)ANTS
CrIECK C^LLS F3R I NT lR AC Tl 3KS AK3 SI^ES
FI.E COlHSc <C«E1TI0N DATE = 12/13/82)
* » * * * * * «
io.'jh of ?:rth
R0SSTA3JLATION
6T SIREN
* » * » *
SIRE "JlMeHR
3CT03ER
S 1 1 E N
CCJNT I
ROM PCT I
COL PCT 1
TOT PCT I 23.1 .5.1 2?. I 51.1 42.1 45.1 46.1 54! i J i 1 J J 1
6. I C . j I 0 1 0 1 CI 0 1 0 1 1
I .0 1 .0 1 .0 1 .0 1 .0 1 .0 1 .0 I 100.0
I .01 .01 .01 .01 .01 .01 . 0 I 14 . 3
I .CI . j I .01 .01 .01 .01 .01 1.1
10. I 41 Jl 41 01 51 01 01 3
I 1 1 . 4 I . 0 I 1 1 . 1 I .0 I 14.3 I .3 I- .0 I 3.6
I iU.O I . :' I 50.0 I .0 I S3. 3 I .0 1 .0 1 42.3
I ",.5 1 .0 1 4.5 1 .0 1 5.7 1 .0 1 .0 1 3.4
11.1 11 2 1 3 1 J I 11 2 1 0 1 2
I 4.0 I 3.0 1 1?.0 I .0 I 4.0 I 8.0 I .G I 9.0
I 20.0 I 23.6 I 37.5 I .0 I U.7 I 40.0 I .0 I 23.6
I 1.1 I 2.3 I 3.4 I .0 I 1.1 I 2.3 I .0 I 2.3
12. I 0 1 5 1 11 3 1 0 1 3 1 11 1
I .0 I 13. 5 I 3.7 I 11.1 I .0 I 11.1 I 3.7 I 3.7
I .0 I 71.* I 12.5 I 1O0.O I .0 I 60.0 I 100.0 I 14.3
1 .0 I 5.7 I 1.1 I i.<- I .0 I 3.4 I 1.1 I 1.1
-t r r I I T I I
COL JMN
TOTAL
5
5.7
3
.'.4
6
6.8 5.?
1
1.1
7
3.0
*B
oc rojE^
I 87.1 39.1 95.1 96.1 V7.I 9e.I 99.
8. I 01 01 L'l 01 01 01 0
I .01 .01 .01 .31 .01 .0 1 .0
I .01 .01 .CI .01 .0 1 .01 .0
I .0 1 .0 1 .0 1 .11 .0 1 .0 1 .0
10. I 41 01 51 31 4 1 0 1 3
I 11. * I .0 I U.3 I 3.6 I 11.4 I .0 I 8.6
I 66.7 I .0 I '.5.5 I *2. •> I 30.0 I .0 I 60.0
I 4.5 I .0 I 5.7 I 3.4 I 4.5 I .0 I 3.4
11. 1 2 1 II 3 1 2 1 11 3 1 2
I 8.0 I 4.0 I U'.O I 3.0 I 4.0 T 12.0 I £.0
I 33.3 I 16.7 I J7.3 I 23. r> I 2C.0 I 50. 0 I 40.0
I 2.3 I 1.1 I 3.4 I 2.3 I 1.1 I 3.4 I 2.3
12. I 0 1 5 1 II 2 1 0 1 5 1 0
i . o : 1 s . 5 i 11.1 i 7.i i . l : 11.1 i .o
I .0 i £3.3 I ^7.3 I L3.3 I .0 I 50.0 I .0
I .01 5.71 .'.41 ? . 3 I .01 3.41 .0
COL JM I
TUTAL
11
12.5
5
5.7
ROM
TOTAL
1
1. 1
35
3". 3
25
23.4
27
30.7
83
100. 0
50
a
a
CD
u
•i-H
CQ
CD
'S
•8
■a
CD
U
Ph
CQ
CD
z a
X 1/1
S. j:
»- j o
o z o
uj »;
J ^ 4
UJ ^1 UJ
i/> < a.
'Ti *-■ CO
UJ Z
uj 3 r;
Z
O r
1- LU
a ex
o *:
l-l o
t- f^.. o n t- f^. '
fiOf^r'..J.ooa m o >t o co n «i
o »» r\i r^ <> r>* «~t m n ^ st it ^ k. *—
rjor^T-ooooT-ooT-OT-T-
.ooo*inr*.oora^ MT- ro ai o /i
O O U"> N. K.i O O t- O Ki o <-t T N. .n
ocr^.OMoor^oroMoo.oo
.J. O "^ C fl o C u.\ u* "O O o 4. *— t
NNa.OMOMONOO>T-S JO O
r«JNro(\»rJojryNf\jMf\(rorgojM
OOOOOOOOOOOOi
-J X
«l 1—
z •-<
«. i
<t t-
I- 3
< ID
"\J o
uj e «* ►-
otk > 3
m o a.
3 u. uj X.
cr v> a O
UJ z o o
ot a o x
OC iU li.
UJ H « *."
*U ZZZZZZZZ.ZZZ.ZJeZZ
UJUJJJUJUJUJUJUJiJJUJUJUJUJIUUJ
££ Qc.3£a£Q££fO£0£afJ&£a£A£££m££tf
O MMMMWIHl-IMl-ll-IMMMt-tlH
51
Table 17. Listing of corrected records.
1 1 0 0 25 2128022 17 2 C ' 531 2 2
i 1 0 0 31 2128022 17 2 C ' 5M1 22
3 1 0 0 2 5 3128 311 1 5 2 0 ' 531 29
4 1 0 0 H 3126011 2 8 2 C ' 5 31 25
5 1 0 0 54 4128012 24 1 0 531 1 3
6 1 0 0 42 5118U22 25 1 c 1 431 33
7 1 0 0 45 5128011 51 2 0 I 531 2 3
1 1 0 0 54 7128011 51 1 0 * 5 31 23
v I 0 0 25 311S011 28 2 c 1 4 31 21
U 1 0 0 29 3116011 28 1 0 1 431 21
11 1 p 0 4 2 3ll)6'o21 24 1 G 3*31 Z4
12 t J 0 4 2 9108012 50 2 P 1 3 31 25
15 1 0 0 4513126012 2b 1 c 1 531 24
14 1 0 0 2511118312 2 6 1 0 '1 431 23
15 1 0 Q 2511128021 29 c c 1 531 25
1s 1 0 0 2 911118312 26 1 c 1 4<>1 2?
17 1 3 0 31111 28021 29 2 0 5 31 25
13 1 3 J 541 3118022 29 2 0 1 4 31 25
1? 1 J 3 2314118011 53 1 0 1 431 26
2j 1 j j 2914116011 33 2 0 1 4 31 25
21 1 0 3 5 41411631 I 52 2 0 1 431 23
22 1 3 J 2317106022 22 2 c 1 331 3J
23 1 0 3 2517128J12 2 7 1 0 1 5 31 32
24 1 3 j 2917108022 22 2 c 1 331 33
25 1 0 3 3117128312
•> t
1 0 1 5 31 32
26 1 J j 4217106011 52 2 0 1 3 31 32
27 1 3 j 4517128011 29 1 c 1 581 26
28 1 3 ] 4 520116312 23 2 0 1 431 27
29 1 3 3 42211 J8012 26 1 0 1 331 34
3J 1 3 3 54211 38022 25 1 c 1 331 33
31 1 3 3 2322106011 51 1 0 1 331 33
32 1 3 3 2922108011 31 1 C 'I 331 33
33 1 3 j 2523128011 26 1 0 1 531 33
34 1 3 3 31231 26011 28 1 0 1 581 33
35 1 3 J 5..23108021 50 1 0 1 331 33
3a 1 3 0 232^106012 25 1 0 '1 3 31 34
37 1 3 3 2924136012 25 1 0 1 331 34
33 1 3 3 4224138311 27 1 c 1 3*1 26
39 1 3 3 2325108311 26 2 c 1 331 37
40 1 3 3 2925108011 26 2 c 1 3i1 37
41 1 3 3 4526116021 29 2 0 1 431 35
42 1 3 J 5429108011 29 2 c 1 381 35
43 2 3 3 95 1128011 53 2 0 1 531 33
4* 2 3 J 97 1118011 27 1 0 1 331 2?
45 2 J J 95 312SQ12 30 1 0 1 531 1 7
45 2 3 J 87 -.1 16012 26 1 0 1 431 21
47 2 3 3 89 6128022 29 2 0 1 531 19
43 2 3 3 95 7108011 29 2 0 1 331 25
49 2 0 3 89 3128321 51 2 0 1 531 22
50 2 3 3 96 3128312 5c 1 e 1 531 32
51 2 3 3 93 3128011 51 2 0 1 531 33
52 2 0 3 87 9116011 22 1 G 1 431 24
53 2 3' 3 95 V1 26011 31 1 c 1 531 25
54 2 3 3 67111 38U12 52 1 G 1 331 25
55 2 0 J 8 911128312 5 2 2 c 1 531 23
56 2 3 3 9511128011 53 1 o 1 531 1 3
57 2 0 0 9111118022 25 1 G 1 431 134
53 2 0 3 8914126311 30 1 0 1 531 131
59 2 3 3 9514118012 53 2 G 1 431 23
6J 2 3 0 9714108321 26 1 0 'I 3 31 37
61 2 3 3 93141 28012 28 1 0 1 531 1 3
62 2 3 j 9515116021 31 2 0 'I 4 31 25
63 2 3 3 9915118012 33 2 0 I 4 31 19
64 2 3 0 9315128011 29 1 0 I 531 25
65 2 3 J 671710602 1 2S 2 0 ' 331 71
66 2 3 ] 9617113022 31 2 0 'I 431 22
67 2 3 3 99171U011 5 2 2 0 431 21
63 2 3 3 9513118322 27 1 0 * 431 2b
69 2 0 J 9o13118311 5 4 2 c 1 4 51 21
70 2 0 3 9 72 0138012 26 1 c 3?1 L 1
71 2 3 J 8721108312 27 2 0 ' 331 3 3
72 2 3 J 9921U6322 2'. 1 0 331 2 i
73 2 J j 9522136^21 7 P. 1 r 3*1 13
74 2 3 J 9622108022 27 1 c 3i1 Z 5
75 2 3 J 9722106011 32 2 0 * 3 31 2 5
76 2 3 3 9322118u12 23 2 0 ' 431 23
77 2 3 J 8923128012 30 1 0 1 531 3 i
73 2 3 3 9623108021 2? 1 0 * 3 31 27
79 2 3 J 372410601 I 2i 1 0 ' 331 35
&0 2 3 3 9^24108021 5? 1 0 ' 3 311 25
81 2 0 J 9525138022 2E L. 0 ' 331 29
82 2 3 3 8*27118011 JO 2 0 1 4J11 1«
83 2 3 J 9727108322 26 2 t 1 3 311 19
84 2 3 j 9927106312 2C 2 C 1 3811 2 5
85 2 3 3 95281 J0 011 2 fc 1 0 ' 3811 27
86 2 3 3 9628108011 51 0 ' 3811 3?
87 2 3 ) «3281 18311 5 3 L. C ' 48V 29
83 2 3 J 953Q1QS01 I It 1 G ' 381 1 2
52
Table 18. Listing of corrected records sorted by sire nun
1 J 3 23251 06011 lb 0 1 3 31 137
2 0 3 2324ldS012 J 5 1 C 1 3 3113*
5 3 3 2 52 2108011 51 1 C 1 3 <1133
•♦ 3 3 2 31710802 2 22 2 0 1 3-1133
5 j 3 2314116011 53 1 0 1 431123
6 3 3 2511118012 26 1 I 1 4*1123
7 3 3 2 5 3116 311 2 3 2 0 1 431121
3 J 3 2 5 3128 311
N q
2 0 1 531129
9 J 3 2523128C11 - C 1 C 1 531 133
10. J 3 2517126012 "3 1 0 1 531132
11 3 3 2511128021 29 2 0 1 5*1123
12 3 3 2 5 2126 32 2 -'7
1
0 1 53112?
13 3 3 2?2 5108011 It 2 0 1 331137
14 3 3 2 9 2410:012 '5 1 0 . 331134
15 j 3 2^22108011 51 1 0 ' 33113d
1o 0 3 291713632: 22 2 I 1 331133
17 J 3 2 91411801! 53 2 0 1 431126
13 J 3 291 11.1 SOU lb 1 0 1 4 31125
1 9 J 3 29 3118 311 2 6 1 0 * 4 31121
2J : 3 2-> 51 26 31 I £6 2 0 « 531123
21 j 3 3123128311
i -
1 C 1 531133
22 j 3 3117126012 JT 1 0 1 531132
23 3 3 3111128021 2 9 2 0 " 531126
24 3 3 31 2128022 27 2 0 1 531122
25 3 3 4224136011 27 1 C 1 331136
23 J 3 4221108012 26 1 0 1 33113*
27 j 3 4217U8011 32 2 0 ' 331132
23 3 3 42 91J6012 30 2 0 ' 331126
2? 3 3 42 3108021 24 1 0 ' 3S1124
30 J 3 42 5118322 25 1 c 4 31133
31 0 J 45261 1 E021 29 2 ' 0 ' 431135
32 J 3 4520118012 23 2 0 431127
33 0 3 4517128011 29 1 c 5 3112 6
34 3 3 4513128012 28 1 0 1 531124
35 3 3 45 5128311 31 2 0 531123
36 3 3 5429108011 29 2 0 331135
37 3 3 5423108021 30 1 0 ' 331133
33 3 3 5421106022 25 1 0 1 331133
39 3 3 5414118011 32 2 0 431123
40 3 3 5413118022 29 2 0 431123
41 3 3 54 7128011 31 1 0 ' 531123
42 3 3 54 4128012 24 1 0 ' 581118
43 3 3 8724108011 28 1 0 331136
44 3 3 8721108012 27 2 c 331133
45 3 3 8717108021 26 2 0 ' 331131
4a 3 3 8711106012 52 1 0 ' 331125
47 3 3 8 7 9116011 22 1 0 ' 431124
43 3 3 87 -.116012 28 1 0 431121
49 J 3 8927116011 3 0 2 0 431119
50 J 3 8923128012 30 1 c 531133
51 3 3 891*128011 30 1 G ' 5 31131
SI j 0 8911128012 *2 2 0 5 31123
53 3 3 8 9 312 8 021 31 2 0 ' 531122
54 3 3 89 6128022 29 2 0 I 531119
55 3 3 95 7108011 29 2 0 331133
5b 3 3 9530138012 28 1 0 381132
57 3 3 9523138011 28 1 c 331127
53 3 3 9525108022 2 6 2 0 1 331129
59 3 3 9522108021 28 1 0 1 331123
60 J 3 9513118022 27 1 0 1 431126
01 J 3 9515116321 51 2 0 1 431123
62 J 3 9514118012 33 2 0 1 431123
63 3 3 95 9128011 51 1 0 1 531125
64 3 3 95 3128012 30 1 c 1 531117
65 2 3 3 95 1128311 33 2 0 1 531133
66 2 3 3 962310e011 31 2 0 1 331133
67 2 3 3 9623108021 25 1 0 1 3 31127
63 2 3 3 9a22108022 27 1 0 1 381123
69 I 3 3 9613116011 54 2 0 1 4 31122
70 2 3 3 9617116022 31 2 0 1 431122
71 2 3 3 9611128011 33 1 0 1 531113
72 2 3 3 95 3128012 36 1 0 1 531132
73 2 3 0 97 1116011 27 1 0 1 331129
74 2 J 3 97201 u601 2 26 1 0 1 331128
75 2 3 3 9722108311 52 2 0 1 331123
7o 2 3 3 9727108322 2 6 2 0 1 331119
77 2 j 3 9714108021 26 1 0 1 331137
73 2 3 3 9 51111802 2 25 1 c 1 431134
79 2 3 3 9323116311 33 2 0 1 431123
8 J 2 3 3 9322118012 23 2 0 1 431123
81 2 3 3 9 315128011 29 1 . 0 1 531123
82 2 3 3 9 31*12 8u12 2? 1 0 1 531113
83 2 3 3 93 3126011 31 2 c 1 531133
64 > : 3 9 927136012 29 2 c 1 331123
85 .> j 3 9924106021 ^3 1 0 1 331126
8b 2 j 3 9 921106 022 25 1 0 1 331124
87 2 j 3 9 9171 16 011 32 2 c 1 431121
83 2 j 3 9915116012 33 2 ft 1 431119
53
4. Calculation of parameters
Table 19 lays out a simple program to calculate for each lamb the weaning
weight adjusted to 120 days and the daily liveweight gain from birth to weaning.
Lines 19 to 29 of Table 19 show the use of a function NDAY to calculate the number
of days between the lamb's birth date and weaning date; the subtraction of birth
weight from weaning weight to give total growth; division by the number of days to
give the daily liveweight gain; conversion to grams; and multiplication of daily
liveweight gain by 120 and addition of birth weight to give adjusted 4 month weaning
weight.
Table 20 is a listing of the sorted records with the adjusted weaning weight
and daily liveweight gain indicated in the last two fields, each of 3 digits (e. g. the
first lamb has an adjusted weaning weight of 130 kg x 10 and a daily liveweight gain
of 87 gm).
5. Statistical analyses
The general model these data fit is:-
yijkl = u + *i + b« + Fk + eijkl
where a^ is the station set of fixed effects
b« are the sire set of random nested effects
F^ are all the other fixed sets of effects and their interactions (month,
parity, sex, birth type, station x month). Interactions of random and
fixed effects are assumed to be unimportant.
Table 21 indicates the parameter cards required for the analysis (Harvey,
1977).
The distribution of classes and subclass numbers and the overall means and
standard deviations for the three traits are listed by the program (Table 22), and
the combined least-squares analyses of variance are laid out (Table 23).
6. Interpretation
Taking birthweight as an example, Table 23 indicates that parity of dam and
sex (P < . 01) and station, month and birth type (P < .05) had statistically significant
effects, while sires within station and station x month interaction had no significant
effects.
Table 24 lists the constants, least squares means and standard errors.
Taking birthweight as the example, lambs from station 2 were 0. 20 kg or 7%
heavier than those from station 1; lambs born in December were 0. 24 or 9% heavier
than those born in October, etc. The environmental effects on the three traits can
thus be quantified and compared.
54
Table 19. Program to calculate adjusted weaning weight and daily liveweight gain.
1 SCONTROL USLINIT/FIL5=1-i
2 »INTS&tR*4
5 PROGRAM A3DWWT
4 C
5 C T3 CALCULATE ACJUST5D W5ANING WEIGHT (120 OAYS)
6 C AND DAILY GR0«T,H (GRAMS) ilRTN TO WEARING
? C
3 COMON IN(23),I01 (3>,ID2(3)
9 1 F3RMATU1, 314, 312, 211, 13, 12, 14, 312, 13, 312, 13)
13 IN0«1=0
11 13 CONTINUE
12 RcAD(2,1,fNO=999)(I>1(K),K = 1,20)
13 03 13 K=1,3
14 L=K*4
15 M = K + 12
1o I01(K)=IN(L)
17 I02(K)=IN(M)
13 13 C3NTINUE
19 I3S = NOAY(ID1,ID2)
20 A= IDS
21 B= IN(10)
22 C = IN(16)
23 0= (C-b;/a
24 G=100*G
25 IN(18) = IFIX(G)
2s t = D*120
27 E = E ♦ E
23 IN(17) = IFIX(E)
29 WRITE(1,2>(IN(K),«=1,18>
30 2 F0RMATU1, 314, 312, 211, 13, 12, 14, 312, 13, 213, 212, 213)
31 GO TO 13
32 999 CONTINUE
33 STOP
34 END
35 INTEGER FUNCTI 3N NO A Y ( 4, 3)
36 INTEGER 4 ( 3) /NOD (1 3; / ANS / Y1 , Y2/8 (3)
37 OATA NOD/365/334/306/275/245/214/134/153/122/92/61,31,0/
33 IF(A(2).SE.1.AND.A(2).Lc.12.AN0.3(2).GE.1.AN3.8(2).LE.12.
39 1 ANO.A(1).GT.0.ANO. 8(1). GT. 0)50 TO 1
40 100 CONTINUE
41 NDAY=-N00(1)
42 RETURN
43 1 CONTINUE
44 Y1=A(3)
45 . Y2=B(3)
4d IF(Y1. LT.100.AND.Y2.LT. 130) GO T3 101
47 IF(Y1.GE.100.AND.Y2.GE.130)G0 TO 102
43 IF(Y1.GE.100) CO T3 103
49 Y1=(Y2/10G)*10J + Y1
50 SO TO 102
51 103 CONTINUE
52 Y2=(Y1/10G)*100+Y2
53 GO TO 1G2
54 101 :0NTIN'JE
55 Y1 = Y1+1900
56 Y2= Y2 +1900
57 102 CONTINUE
53 11= A(2)
59 12=B(2)
60 LEAP1=0
61 LE4P2=0
62 IF(M1 .LE.2.AND.((M0D(Yl,^).Ea.3. ANO. 100 ( Y1 ,1 30) . NE . 0)
63 1 .03.10D(Y1/40C'). :Q.0))LEAP1 = 1
6f IF(M>.LE.2.AND. ( (K3D ( Y2/ 4) . E 3. 3 . ANO. 10D(Y2/1 30).NE.O)
65 1 .03.MOO(Y2,400). EC.0))LE4P2=1
6b C
67 C
63 C
69 1T1= NOO(.11>-'403( 11*1)
70 IT2= IT1 ♦ LE4P1
71 JT1 = N0C(H2) -NO 3 (12*1)
72 JT2=JT1*LE»P2
73 Ir((Afl).GT.IT1.ANO.Kl.JE.2).OR.(A(1).GT.IT2.AN0.11.Ej.2)
74 1 .3R.(3(1).GT.jT1.ANi).M2.N5.2).0R. (B (1 ) . GT . JT2 . AND.M2. EQ. 2 )
75 2 > GO TO I00
75 C
77 ANS = NOO(M) - A(1) -N0C(*2) + 6(1) +L EAP1 -Lc AP2
7 3 IF(Y1.LE.Y2) GO T3 2
79 ANS=-4NS
80 M*Y1
81 Y1=Y2
82 Y2=M1
8 3 2 C3NTINUE
84 Y1=T1*1
85 . I-(Y1.GT.Y2) GO T3 4
8 6 DO 3 M1=Y1/Y2
87 ANS-ANS+365
£3 Ir(M03(Ml,4'i0).:C.C.OR.(M0D(P1,4).Ej.0.ANO.M0D('4l,130).
89 1 N.2.0)) ANS=4NS + 1
9 J J C3NTINU-
91 4 CONTINUe
92 NOA»=IAdS(ANS)
9J RiTJPN
9* E JO
55
Table 20. Listing of records with adjusted weaning weight and daily liveweight
gain in last two fields.
1 1 j 3 23^5 I'Jt'31 1 26 c C 1 331 371 30 57
i 1 j J 2 32* 106 31 i 25 1 0 1 3 31 341 27 35
i 1 J 3 2322 106 011 31 1 0 I 3 i1 1331 22 73
<t 1 J 3 2 517 0 8 0 2 i 22 2 0 ' 331 I2j1 18 83
3 1 J : 2214 I1Cj1 1 33 1 0 1 -.31 251 13 67
6 1 : j 2511 116 j12 36 1 c 1 431 1231 38 6'
7 1 ] 3 2 3 i 118011 i o 2 0 1 431 1211 35 64
i 1 3 j 25 3 I26u1 I 28 2 0 1 5 31 291 39 67
i 1 J j 2523 12631 1 28 1 c 1 531 1351 30 85
1J 1 J 3 2517 I2t012 22 1 0 1 5 31
•» >.
19 83
11 1 : J 2511 I 28021 29 2 c 1 531 26' 11 63
1 2 1 j j 25 2 126322 27 2 t 1 531 1221 33 63
1 i 1 j J 2<25 108 311 26 2 0 1 331 1371 30 87
H 1 ] j 2*2* 1 J601 2 25 1 o 1 231 13*1 27 85
15 1 j J 2^22 1 06011 31 1 0 1 331 1331 22 76
16 1 j J 2°17 106022 22 2 c 1 3 31 1331 18 83
1 7 1 j j 2<>14 1 16011 33 2 0 1 431 I2c1 12 67
1 i 1 j j 2911 116012 26 1 c I 431 1231 38 63
1 J 1 ) j 29 3 116011 2 8 1 0 I 4i1 211 35 ©4
20 1 j 3 2^> 3 126011 26 2 0 1 5 31 291 J9 67
21 1 j 3 3123 1 28011 2 8 1 0 I 531 331 30 85
22 1 3 3 3117 1 2601 2 23 1 c 1 531 1321 19 83
23 1 0 3 3111 I 2602 I 29 2 0 1 531 1261 11 63
2* 1 J J 31 2 I 28022 27 2 c 1 5 31 221 33 63
25 1 J j 4224 136011 27 1 0 1 331 3a1 29 85
2a 1 J 3 4221 1 38012 26 1 c 1 331 13*1 24 82
27 1 j 3 4217 1 J601 I 32 2 c 1 331 321 20 7*
23 1 3 3 42 9 136012 30 2 c 1 381 261 10 6 7
29 1 3 ) 42 3 106021 24 1 0 1 331 241 37 69
3J 1 J 3 42 5 118022 25 1 0 1 431 1231 13 75
31 1 3 3 4526 116021 29 2 0 1 431 1351 29 84
32 1 J 3 4520 11601 2 23 2 0 1 451 I271 17 73
33 1 3 J 4 517 128 011 29 1 0 1 531 1 2 6" 15 71
3* 1 J ) 4510 128012 26 1 c 1 531 1241 39 67
35 1 3 J 45 5 126011 31 2 0 I 531 1231 33 63
2a 1 J 3 5*29 108 01 1 29 2 0 1 2 31 1331 33 8i
37 1 j 3 5*23 1 06'j21 50 1 0 1 2 31 1351 25 79
33 1 3 3 5421 1 38022 25 1 G 1 331 1331 21 83
39 1 3 3 5*1* 118011 32 2 0 1 431 1231 15 69
40 1 3 3 5413 1602 2 29 2 0 1 431 1231 10 67
41 1 J 3 5., 7 1 26011 31 1 0 1 531 231 11 66
-.2 1 J 3 5* * 1 2 1 J 1 2 2* 1 0 1 5*1 11 31 JO 6 5
43 2 J J 872* 06011 it 1 0 1 331 3o1 29 s;
44 2 J J 872I J801 2 27 2 0 1 3 31 331 24 83
45 2 3 3 6717 I J6021 23 : c I 331 311 19 76
46 > J 3 8711 08012 52 1 0 ' 331 261 12 66
47 2 3 j 17 9 11601 I 22 1 c 4 31 241 37 71
Hi 2 3 J e7 4 116312 26 1 c I 431 211 33 62
49 2 0 3 6927 116011 30 2 G 1 431 191 15 71
50 2 3 3 6923 26012 30 1 G ' 5?1 351 25 7?
51 2 j j 691* 2801 1 50 1 C ' 5^1 311 17 75
52 2 j 3 6*ir 26012 32 2 0 ' 531 231 13 63
53 2 J 3 g0 v 26021 31 2 0 ' 531 221 3e 63
54 2 3 3 go 5. 2802 2 29 2 0 ' 5?1 191 32 61
55 I 0 3 95 7* 06011 29 2 G ' 3 31 331 15 71
55 6 3 j 9533 06312 26 1 0 ' 331 3 I' 5C 8 5
57 i 3 3 952 3' 3601 1 2i 1 0 ' 331 27' 23 79
53 2 D 3 C525' 3 6 0 2 2 28 2 0 ' 3 31 291 23 7)
5v 2 3 3 9522' 06021 26 1 0 ' 3311 231 20 73
t J 2 j j 951 3' 1602 2 27 1 0 ' 431 261 15 7i
61 2 3 3 9515' 16 021 51 I 0 ' 4 31 231 11 67
0 2 2 .1 3 9514' 1601 2 33 2 C 1 431* 231 3? 65
63 2 3 J 95 r 2 6 011 31 1 0 1 5311 251 39 65
b* 1 J J 95 5' 2801 i 30 1 0 ' 531 171 30 5 3
65 I 3 J 95 r 28011 33 2 0 1 5811 331 12 65
6o 2 3 J 962 3' 0601 1 31 2 G ' 331' 331 26 79
67 I J 3 9 ',2 3 06021 25 1 0 ' 2 31 271 19 7)
6J I J J 9622' J8022 27 1 0 ' 331 261 16 73
6V 2 J 3 9613 118311 3* 2 0 ' 4?1 221 1? 63
70 I 3 J 9617' 18022 31 2 0 ' 4^1 221 11 67
7 1 1 j J »a1 1' 2 b J 1 1 53 1 c 5S1 1 31 J*. e> J
71 2 J 3 96 3 28012 36 1 G I 5 31 3 2' 16 65
73 2 J 3 97 1 18011 27 1 0 ' 321 2*1 29 85
7h 2 J J 9 72 3 0801 1 26 1 0 ' 331 231 18 77
73 2 3 J 97 2 2 06 311 32 2 0 1 3*1 231 16 73
75 2 J 3 9727 10802 2 2o 2 0 1 7S1 191 15 7i
77 > 3 3 971'4 3801 1 26 1 0 ' 331 371 22 83
73 j 3 J 9 31 1 18022 25 1 C ' 431 341 17 77
7> 2 J 3 9 32 3 116011 53 2 0 ' 431 231 24 7 6
8 J : J j 9 'I c 2 116012 23 I G 1 431 231 19 83
81 2 3 2 9 115 128011 29 1 0 '1 531 231 11 6'
82 2 j 3 9 Ai 2 8 312 28 1 G 531 1 31 n 65
83 2 j J 9" i 2 801 1 51 2 G 531 731 13 63
0.4 2 J j W27 I 3tOl2 29 r 0 1 3 31 231 24 7J
83 ? J J 9 >24 )Z02 1 53 1 0 1 351 251 20 7 2
33 2 3 j 9921 1 0 1 3 2 2 .> C A G I 3 61 2 4* 15 75
87 2 j j 9V17 18 011 52 2 0 1 431 211 11 6S
il 2 j } 9yl 5 1801 2 5 2 2 c • 431 1 91 18 6 2
56
Table 21. Parameter cards required for Harveys Least Squares Analysis.
1 01 06 01 003903 0005 0000Q1 00 3001 0001 200403 32 3201
2 SIR/STOG300003011213
3 01STAT 02010100010002
4 02M0MTH 330216001000113012
5 03 PARITY 0201 2600010032
6 04SEX 02012100010002
7 J53THTrP0201 200001 0002
3 0102
9 0002 02301 G000230000BTHWT
10 300304301 0001 150Q00WWT
11 0003 04 30 0001 0OO000OG«OWTH
12 0202 3100 37030104 0000 310000010000 2004 030 2 3202
13 SIR/ST0125750101021213
14 31STAT 32010100010002
15 02MOMTH 030216001000113012
16 03P A RITY 0201 26000100 32
17 34SEX 32012100010002
13 353Tr1TYP 0201 2000 01 0002
19 0102
20 0002 02301 000 02 S0000BTHWT
21 0003 0400100011 50000WWT
2 2 0003 04 300001 00 0OO00GROWTH
Table 25 lists the variance and covariance component estimates and uses
these in calculation of the estimates of heritabilities and genetic correlations.
Thus for birth weight, in this paternal half-sib analysis, the heritability = 4 x
V among sires t (V among sires + V within sires) = 0.317. Similarly the genetic
correlation between birthweight and weaning weight = (Cov. birth wt. weaning wt)
•/tJ V birth wt. V weaning wt = -1.202
Certainly many more data would be required before seriously attempting
to estimate such parameters.
7. Peripheral studies linking statistical outputs with biological effects
In this data subset, the effect of month of birth on weaning weight, for
example, might be correlated with the rainfall over the 4 month period following
each individual month of birth. This linking would usually be more relevant to
analyses of mortality data, for example, when the statistical identification of
contrasting subclasses might indicate that detailed evaluation of post-mortem
reports, etc. should be attempted.
57
Table 22. Distribution of class and subclass numbers
DISTRIBUTION OF CLASS ANO SUBCLASS NUMBERS FOR P*03L6M NO.
IDENTIFICATION NO.
T3TAL
STAT
STAT
M3NTH
MONTH
M3NTH
PARITY
PARITY
SEX
SiX
BTHTYP
BTHTYP
STAT
STAT
STAT
STAT
STAT
STAT
10
11
12
1
2
1
2
MONTH
mONTi
MONTI
MCNT.1
^ONT.1
*CNT1
1C
11
12
10
11
12
88
42
46
35
25
23
44
42
43
*3
61
27
1S
11
15
19
H
13
OVERALL MEANS AND STANDARD DEVIATIONS Or RHM
dTHWT
ww r
GR3WTH
«EAN =
MEAN =
MEAN =
2.H4545
11.58*09
72.75000
S.D.=
S.D.=
S.D.-
.31143
.82725
7.774h8
Table 23. Combined least squares analysis of variance
COMSINEO LEAST-SJUARES ANALYSIS OF VARIANCE
SThWT
SOURCE O.F. SU* OF SQUARES MEAN SQUARES F PROS ERROR LINE
STAT 1 . 1422J3 .342203 9.199 .0104 SIR/ST
SIR/ST 12 1 .'9 J6 2.7 .091552 1 .484 .1521 RE MNOR
MONTH : 417917 .203959 3.383 .0397 RcMNDR
PARITY i ..3o852 .♦33852 7.083 .0097 REMNDR
SEX 1 354611 .354611 13.357 .0004 REMNCR
yT.lTYP 1 344417 .34,417 5.585 .0210 REMND3
STAT X MONTH 2 15 H7C .079235 1.285 .2334 REMNDR
REMAINDER *>7 * 13.'0 32 .061672
SOURCE O.F. sui ~>f s juak :s MEAN SQUARES F PROS ERROR LINE
STAT 1 .124la1 .12*1o1 .212 .6536 SIR/ST
SIR/ST 12 • 7. j2>479 .565373 1.325 .2257 REMNCR
MONTH 2 22.922158 11.461079 25.396 .0000 REMNDR
PARITY 1 1.16*655 1.164655 2.631 .1095 R E M N C R
SEX 1 t . -379 J6 1 .43 7906 3.249 .0760 R ; M '1 0 R
BTHTYP 1 1.152837 2.152807 4 • 36 * .0 30 REM.JCR
STAT X MONTH 2 .6C5231 .302641 .384 .5082 R : M <j C R
REMAINDER 67 ^..>«31>7
G.iO-rfT'.i
.443535
SOURCE O.F. SUM OF SwUAR:S MEAN SQUARES c PR03 =RRCR LINE
STAT 104.826483 104.826433 1.672 .2203 SIR/ST
SIR/ST 12 752.274139 62. 589512 1.731 .0795 R z M N D R
MONTH 2089.540811 1044.770406 2-5.347 .0000 REMNDR
PARITY 211.763133 211.763133 5.347 .0183 R z V N D R
SEX 7.571679 7.571679 .20? .6490 RiMNDR
9T«TYP 49.848020 49.343020 1.37a .2449 R 1 f N 0 R
STAT X MONTH 2 83.276743 41.633371 1.150 .3229 R?MND<!
REMAINDER 67 2426.548850 35.217147
STable2T.Listingofconstants,le stsqu rmeansa dsta d rdrrors
AANOARO ERROROF
LSMEAN
.03703156 .05078116 .04935903 .05846421 .05771742 .05629383 .04653952 .04622796 .04552783 .04640553 .04048233 .05526735 .03924319 .103453973 .09357244 .07539236 .07643204 .09238103 .09331971 .12792372 .12346707 .15277649 .15072494
.17M1449
.12009024
.1189M74
.11698372
.1194330"!
.13232337 .14398411 .23297323 .22069945
.24M8029
.19545572
.1989M75
.24175735 .97750466 1.33595141
1.3367C463
1.46833027 1.45138224
1.6524C391
1.19432765 1.13539787 1.16937691 1.139466410 1.05479559 1.39425486 2.24510987 2.12678414 2.354102247 1.89669747 1.93053557 2.34630737
PROBLEMN .2
LEAA-1wUARE1
MEAN
2.320e7122
2.72149735 2.92024538 2.59466377 2.33363640 2.93431348 2.74631364
2.395M830
2.92227835 2.71943458 2.39467341 2.74706902 2.60323?13 2.79317538
2.76802811
2.78603941 2.97409772 3.10059810 11.41557358 11.45372896 11.37741820 12.363137 4 11.27323154
1T.63033136
11.53731118 11.29393118 11.54703471 11.28406245 11.60003778 11.23105938 12.50803806 11.17486433 13.67131450 12.21336762 11.33159874 13.58223823 71.55767290 72.66631129 70.44901052 83.58411218 73.14722595 63.^4163058
73.19924M2
69.91o13159
71.35945494 71.27589037 72.44554666 73.56979915 82.61363136 69.78007772 65.60529678 73.55459299 70.51437419 62.27806439
AANDARO ERROROF CONAANT
.03703156
.03338127
.03415959 .049473*9 .04135681 .05358343
.02831427
.02801425 .02723590 .02723590
.03U2998 .03U2993
.04829702 .04092537 .05142327 .04829702 .04092537 .05142327 .09301371 .03409459 .01158547 .13274715
.11079T11
.14115733
.075T4693
.07504693 .107296182 .07296182 .03366150 .05366150 .12938213
.1T963432
.13775693 .12933213 .13963432 .15775693 .97750466 .87819443 .93432264 1.19902754
1.0T221235
1.29862410 .67837794 .67837794 .60001589
.6o001589
.75680567 .75680567 1.17039657 .99175701 1.24615591 1.17039657 .99175701 1.24615591
-10UAR61MEAN1ANDANOARO€RROF
CONAANT EAIMTE
2.32387122 -.09937336 .09937336 -.12620745 .01276518
.1134M26
-.07455758 .07455758 .10133684 -.10138694 .37380220 -.07380220 .00799822 .05391274 -.06691076 -.00799622 -.05391254 .06691076 11.41557358 .03315538 -.03315538
.94761M6
-.16234204 -.78527222 .12174000
-.121740T0
.13151113 -.13151113
.18451M0 -.18451M0
.10666434 -.11652279 .00985775 -.10666434 .11652279 -.00985775 71.55767290 1.10366238 -1.10866238 9.02643927
-1.41T44695
-7.61599232 1.64157131 -1.64157131 .30173233 -.30173203 .88787376 -.88787376 .92085680 -1.47581061 .55495381 -.92085680 1.47581061 -.55495381
EFFECTIVE
NO.
71.3 36.9 43.3 21.2 21.3 15.3
36.?
37.7 39.3 37.4 54.3 24.2 9.1 10.2 3.3 13.2 12.7 3.6 71.3
36.?
43.3 21.2 21.3 15.3 36.V 37.7 39. 37.4 54.3 24.2 9.1 13.2 3.3 13.2
U.7
3.6 71.3 36.9 43.3 21.2 21.3 15.3 36.9 37.7 39.3 37.4 54.3 24.2 9.1 10.2 8.3 13.2
U.7
3.6
MO. OB1.
85 42 4s 35
27
23 46
M
43 4T 61 27 16 1 15
U
14 13 83
M U 11
25 23 44 42 43 43 61 27 16 11 15 19
1*
15
U
42 46
11
25 23 46
M
43
U 6)
27 16 11 15
1?
14
U
LeAA
U
11
U
10 11
U
10 11
U
10 11 12 10 11 12 110 11
U
LIAIMGOFCON ANTT,
ENDENTVARIA3LE
1 1 1 2 2 2 1 1 1 2 2 2 1 1 1 2 2 2
1 2 1U 1
U
1 2 1 2 1 I
MONTH MONTH MONTH MONTH MONTH MONTH
1 2 10 11
U
1 2 1 2 1 2
MONTH MONTH MONTH MONTH MONTH MONTH
2 1U 11 12 2 2 2
MONTH MONTH MONTH MONTH MONTH MONTH
INDEP
MONTH MONTH MONTH
PARITY PARITY 9THTYP 3TMTYP
X X X X X X
MONTH MONTH MONTH
PARITY PARITY BTHTYP 8THTYP
X X X X X X
MONTH MONTH MONTH
PARITY PARITY 3THTYP 3THTYP
X X X X X X
CODE
Ml)
AT AT TEX TEX AT AT AT AT AT AT
U
AT AT 1EX TEX AT AT AT AT AT AT
U
AT AT 1EX 1EX AT AT AT AT AT AT
ROW 1 c 0 3 4 G 5 0 6 10 7 C 8 9 0 0 0 G 1 Z 0 3 4 0 5 0 6 0 7 10 8 9 0 10 G G 1 2 0 3 4 0 5 G 6 G 7 0 8 9 10 0 0 G
RHM
NAME aum
3THWT 3THWT 3THWT TUkT dTHWT BUNT dTHWT 8THWT tTTHwT JHKT bthwt BTHWT BTHWT 8THWT BTHWT BTHWT 8THWT
HUT wwr wwr wwr wwr wwr wwr wwr wwr wwr wwr wwr wwr wwr wwr wwr wwr wwr
6RUWTH GROWTH GROWTH GROWTH GROWTH GRUWTH GROWTH GROWTH GROWTH GROWTH GROWTH GROWTH GROWTH GROWTH GROWTH GROWTH GROWTH GROWTH
Table25.Varianceandcov iancecompone tsfesti ationofg ne cparam te s.
11ORCPMOYCOM^ONEST1 1.109362635.09155 24. 053075 -1.31315123.0 418310139? 52 -13.29153914.5 429243-.73735?0
7.03647909.53637526.0275 C71
6?.021055755.7517546.35 60858
752.27413S9362.639511534.70221450
0H:NOTtPICENVIRONMENTLNEGVARIANCEORCOC04POHEKT1
RhM
ETHwT
WWT
GROWTH
WwT
GROWTH GROWTH
RHM
6THWT BThwT BTHWT
HWT W4l
GROWTH
COL 1 U 3 3 3
ROW 1 1 1 2 l 3
U103
2 I I I I I
AANOAR!
•37>>
1.h35
.357 .144ERRORT 1.231
HfcRlTAUILITY ORGENETICR
.317 -1.202 .215 1.026-1.Oil
RhM
BTHWT
GROWTH
WWT
GROWTH
h*T
RHK
3T-IWT iTHWT sTHhT
rfWT HUT
COL 1 2 3 2
ROW 1 1 1 2 2
E 2 ? 2 2 2
AMONGWI HIN
n0053T7'53.0616^212
1.0301.30.'fAlillz-.3 5 9931
-•H0'!"2"JJ.17«J5-.56280739
4.7022146035.21714702
VARIANCENOCOVARI EMPONENTEAIMT 1FR MIN ECTANALY1 1
KFORRANDOMEFFECT1COMP NENT(1IR/1T>»5.6298DEER mOFE 301=12.
11,CPIS/ACP,VARIANCENOCOVARIANCCOMPONE T1
NEGTIVEVARIANCECOMPONENTE IM TE11ETZEROFORH 1COMPUTTION
E1flHTE1OFHRITAEILIT1,G:NETIC,PHcNOTYPlC,ENVIRONME TLCORRELTIO 1 CORRELTIONCORRELTION
33500W3535OWTH00 5.,1 W1.10 0
-:;;3-.0500.1707359 .02754071 .35560853
.958.9430.3556JS59
cn CD
60
RESULTS OF RESEARCH IN THE FIELD

61
HUSBANDRY, NUTRITION AND PRODUCTIVITY OF GOATS AND
SHEEP IN TROPICAL AFRICA
R.T. Wilson
Tropical Africa has one- sixth of the total world flock of sheep and one-
third of all goats. There are 0. 71 goats and sheep per inhabitant in tropical
Africa but their distribution is uneven, tending to the drier areas (Table 26).
Within the semi-arid and humid zones the number of goats and sheep per head of
human population varies from as low as 0. 07 in Sierra Leone and 0. 13 in the
Congo and Zaire to as high as 6. 0 in Djibouti and 5.4 in Mauritania. In the
countries in which ILCA has zonal programmes the figures are 2. 13 for Mali,
1.65 for Ethiopia, 0. 73 for Kenya and 0.47 for Nigeria; production from small
ruminants is important in all these countries. Total meat production from small
ruminants in Africa is 1.15 million tonnes (16% of world production), total milk
production is 1. 99 million tonnes (14% of world total) and total skin production is
211 000 tonnes (15% of world production). The total milk production from goats in
Africa is about 3 times that from sheep.
Sheep and goats contribute about 17% of the total ruminant biomass in
Africa (Table 26). This percentage varies from 9. 3% in wet tropical Africa (in
cluding Tanzania and Zambia) to 35% in the Mediterranean Littoral. There are
slightly more sheep than goats, and as sheep are generally bigger than goats they
contribute more to meat production, although the value of goats is increased by
their better milk production.
Table 26. The regional importance of sheep and goats in Africa. Numbers and
areas adapted from FAO (1981a). Biomass calculated from mean
population weights (cattle 206 kg, camel 307 kg, sheep 30 kg, goat
18 kg, buffalo excluded).
Parameter
Region
Semi- arid Mediter Southern
Total Humid
6 9
and arid ranean
Littoral
Africa
Total area (10 kmz) 28.8 10.6 9.4 5.7 3.1
Agricultural population (10 ) 271.9 108.2 107.4 41.7 14.5
Number of goats (10") 141.1 93.5 23.2 12.6 11.8
Number of sheep (10") 162.2 78.0 14.7 35.0 34.7
Ratio of goats : sheep 0.87 1.19 1.58 0.36 0.34
Density of (goats+sheep)/km 10.5 16.2 4.0 6.1 15.0
Number of (goats+ sheep)/
person 1.11 1.59 0.35 0.83 3.21
(goats+ sheep) as % of
ruminant biomass 16.8 16.5 9.3 34.7 18.4
62
Husbandry
The term husbandry is used in a rather wide sense to include breed types
(because these have undoubtedly developed in response to local needs and have
been influenced by selection pressure), ownership patterns (because these reflect
the preferences and needs of the human population) and management (which does
exist under traditional systems in spite of a lingering feeling in some quarters that
this type of husbandry proceeds on an ad hoc basis).
Breed types. Both goats and sheep of the semi- arid zones are generally larger
than those of the more humid zones. They have for long been considered "unpro
ductive" but it is doubtful if any other type of animal - at least of those currently
domesticated - could produce as much in terms of returns for resources utilised.
Over large areas of the West African Sahel there is little differentiation of breed
or type in conditions of similar ecology. In general in the west and to some extent
extending across the Sudan, where differentiation does occur it is found along east-
west lines which follow the main ecological zones. In East Africa the situation is
more complicated, being influenced by climate and altitude, by the diverse origins
of the ethnic groups owning sheep and goats and by importations into the area.
These importations may be of considerable antiquity - from Arabia and south-west
Asia, for example - or more recently from the developed areas of Europe,
Australia and South Africa. Table 27 indicates some of the main races of goats
and sheep in the semi-arid areas and the economic justification for them.
In the humid zones there is little differentiation into breeds or types in
either goats or sheep nor indeed is there any functional division. Both species are
of the dwarf type, the extreme in goats often having a grotesque appearance.
Goats seldom weigh more than 25 kg and sheep are little bigger although some of
the intermediate type of Djallonke males in the sub-humid zone of West Africa may
weigh as much as 35 or 40 kg. These types are generally trypanotolerant. In this
zone the principal, if not the sole, reason for keeping small ruminants is for the
production of meat (although it is said to be very greasy and is not liked by the
peoples of the more arid areas) with the skins also being cooked and eaten. Milk
production is very poor.
In the Mediterranean Littoral the Merino is important in some areas as it
is in Southern Africa. In Southern Africa the Karakul sheep and Angora goat are
also important, particularly in Namibia (South-West Africa). Recent attempts to
"improve" local races with these breeds in tropical Africa, with the exception of
the Highlands, have met with almost universal failure (see, for example, Wilson,
1981). No traces of these attempts can be seen in the stock existing to-day. In
East Africa, apart from the out-and-out "European" operations, some progress
has been made towards improvement of native breeds by the introduction of the
Blackhead Persian and its derivative, the Dorper, in particular in Masai flocks.
Similarly, the Boer goat and a prolific fast-growing local breed, the Boran, are
being introduced, by the Masai themselves, into traditional flocks.
63
Table 27. Principal types of goats and sheep in semi-arid Africa and their
aptitude for production
Species Breed/type Country/Zone
Mature live- Production
weight (kg) aptitude
GOATS
Sudan Desert/Sahel Senegal to Sudan:
arid and semi-arid
35 Meat, milk
Maradi/Red Sokoto Niger/Nigeria: 30 Skin, Milk
southern semi- arid and Meat
Nubian Sudan:
riverine
50 Milk
Abyssinian Ethiopia: 30 Milk
arid/semi-arid
Small East African Kenya: 35 Milk, Meat
upland semi-arid
Mubende Uganda: 30 Skin and Meat
upland semi-arid
Boer Kenya: 45 Meat -
SHEEP
upland semi-arid crossbreeding
Black Moor/Zhagawa Mauritania/Tchad/Sudan:
arid desert fringe
35 Meat, Hair
Sudan Desert Sudan: 50 Meat, Milk
arid desert fringe (Skin)
Sahel (West African Mali/Niger/Tchad: 40 Meat (Skin)
Fellata) Sahel, northern semi-
arid
Macina Mali: 35 Wool, Meat
Niger flood plain
East African fat-tailed Ethiopia:
semi- arid
30 Milk (Meat)
Masai Kenya: 35 Fat meat
upland semi-arid (Skin)
Sahel x Forest/Nilotic Mali: 45 Meat - supple
semi-arid/sub-humid mentary fed
Blackhead Persian/ Kenya: 40 Meat -
Dorper upland semi-arid crossbreeding
64
In general under present conditions there would appear to be adequate
genetic material in the indigenous races for production to be raised to a much
higher level than the current average. Only when this level has been reached and
when the constraining ecological and economic conditions can be overcome, should
consideration be given to further "improvement" by races less well adapted to the
rigours of the zone.
Ownersh'fp* The ownership pattern is very varied and, for an outsider at least,
extremely difficult to establish and understand. The ramifications of many
African kinship systems; the extremely complicated systems of "stock friends",
loans and herd splitting; the herding out procedures involving professional
herders often of a different ethnic group all lead to a rather fluid idea of owner
ship which often involves many displacements of an animal over its lifetime.
It would nonetheless be true to say that larger numbers are owned by
individuals or families in the drier areas than in the less dry ones. In West
Africa and the Sudan this in effect means that flock size decreases from north to
south and in Ethiopia and Kenya there is a trend to smaller flocks at higher
altitudes. This trend reflects the obvious change of system from a purely pastoral
one associated with the very dry areas through an agro-pastoral one in the less
dry areas (and where the agricultural component may be assuming more import
ance) to an agricultural one in the gradation to a sub-humid climate.
What is perhaps less obvious in the ownership pattern is the gradual
change in emphasis from sheep to goats as the macro-management system moves
from nomadism to sedentary and from pastoral to agricultural. This is reflected
not so much in the size of flocks as in the numbers of owners who either have
preferences for goats over sheep or who, for other reasons, are forced to keep
goats. Goats are, of course, generally more prolific than sheep and are probably
less trouble to manage for the agriculturalists and agro-pastoralists who are
recent entrants into animal husbandry. Table 28 provides some idea of the
distribution of ownership in an agro-pastoral system composed of two sub
systems in the semi-arid zone of Mali; Table 29 indicates patterns in the humid
zone of south-west Nigeria while Table 30 shows additional data for Kenya and
Tchad.
In recent years, although there is little hard evidence to support such a
contention, it is probable that the goat population has been increasing not only in
absolute numbers but in relative terms in comparison with sheep. This is per
haps due to their higher total reproductive rate and their wider dietary range.
Although, as can be seen in the section on productivity goats are not generally as
productive as sheep when calculated on the same basis in terms of meat produc
tion (although there are exceptions), their superior milking ability undoubtedly
renders them more attractive overall, particularly in the drier areas. Trends
in total population and in numbers of families owning small ruminants are thus
likely to continue towards goats and away from sheep, at least in the traditional
sectors.
65
Table 28. Ownership patterns of sheep and goats in an agro-pastoral area
in central Mali.
Irrigated Rice Rainfed Millet
sub- system sub.-system
Goats Sheep Goats Sheep
Number of owners studied 27 16
Number owning sheep or goats 26 15 16 9
Number owning goats but not sheep 12 7
Number owning sheep but not goats 1 0
Mean flock sizea) 9.0 6.4 38.2 7.1
± s.d 6.03 13.51 27.75 14.81
Mean flock size ' 9.3 11.5 38.2 12.6
± s.d 5.87 17.0 27.75 18.27
Range in flock size 0-23 0-64 2-91 0-58
Notes: (a) of all owners i. e. irrespective of whether the holding of one species
of stock is nil
(b) of only those flocks in which animals are held, i.e. nil holdings
excluded.
Table 29. Pattern of small ruminant ownership in the humid zone of south-west
Nigeria. Source: Mosi et at, (1982).
Forest zone Derived Savanna
Zone
Percentage of farmers owning small
ruminants 73 20
Mean flock sizes
Goats only 2.8 3.7
Sheep only 2.0 -
Mixed flocks 5.1 5.3
Management. Until recently, and indeed the feeling lingers on in some quarters,
it was considered that under traditional systems of operation no management was
practised. Only a little thought shows the inherent nonsense of this tenet.
Nomadism is a sophisticated management response to a resource base which is
always seasonally and often totally deficient. Stall feeding is equally a response
to the availability of a surplus of nutrients in a particular environment and to a
demand, often very strictly confined in time and space, for a convenient quantity
of meat. These management options are the extremes of a wide range of such
66
Table 30. Livestock owning in agro-pastoral and pastoral societies in Kenya and
Tchad. Sources: Kenya - Christie Peacock (personal communication);
Tchad - Dumas (1977).
K en}7a Tchad
Gondeye-
Tchein
Masai Karapokot Zioud Salamat
Production system Pastoral Agro- Pastoral Agro- Agro-
Average holdings
• pastoral pastoral pastoral
Cattle 157.3 11.78 36.4 133.3 2.1
Sheep 44.0 5.35 43.5 2.0 1.3
Goats 83.1 13.64 45.0 46.3 4.7
which form a continuum from the almost totally unendowed very arid end of our
spectrum to the much more favourable environment at the sub-humid end of the
scale where irrigation possibilities may provide the opportunity for relatively
sophisticated interventions. In the very humid zones all management is sedentary
with animals often being stall fed or given quantities of household and crop waste
and being tied or housed at night.
Table 31 indicates the strategies ("macron-management) and tactics
(Tfmicro"-management) of traditional owners in the African semi-arid zone. With
only few exceptions there are clear trends from low to higher rainfall which are:
nomadism to stall feeding; uncontrolled or very loosely controlled ranging by day
and open camp at night to very restricted herding by day and confinement at night;
a tendency to generally smaller flock sizes as conditions improve and an increased
emphasis on goats associated with the agro-pastoral zones as already mentioned in
the section on ownership. Large scale modern management of sheep (for wool and
to a lesser extent for fat lamb) is confined to the highland areas of Kenya.
Prestige and perverse supply were once catch words used to typify the
attitudes of traditional livestock owners. Undoubtedly African pastoralists are
conservative but it is doubtful if they are any more so than their peers in Australia
or America. Their reasons for keeping stock are rarely irrational and are
perfectly in keeping with the problems encountered and the short and long term
goals of the owners. One aspect which supports this contention relates to the age
and sex structure of the flocks. Whatever the main economic objective in keeping
goats or sheep, a remarkable similarity in flock structure is apparent across the
whole of the semi-arid zone, as can be seen from Table 32. With the single
exception of the Afar of Ethiopia whose subsistence is almost entirely milk, all
the flocks have around 75% of females (and somewhere in the region of 55% of the
total flock breeding females in excess of twelve months of age). In a sample of
flocks belonging to four different ethnic groups in Mali, covering the whole range
Gi
Table31.Ecologyandmanagem ntofgoatsshe piemi- ridAfrica
"Micro"management
Climaticregime
(rainfall)
mm
Arid
Semi-arid Highlands
"Macro"
management
Examples
Sheep
Goats
Sizeofflock/
herdinggroup
Day
Night
Day
Night
Sheep
(200)Nomadic
Mauritania/Moor4,
Ethiopia/Afar
Sudan/Kab bish
Opencamp Penned
t*
Opencamp
Penned
10
Mali/TwaregLoosef ckOpencampo
(300)Transhumant
(T00) (T0)
(100)Sedentary
Niger/Twareg
Tchad/Zhagawa Kenya/Turkana
Ethiopia/Afar
N/
W
Penned Penned
*>/
Penned
Semi-sedentarySu an/BaggaraTi htflockPenned
Loosefl ckOpencamp
Mali/Fulani
Kenya/MasaiTightflockPenned Sudan/Dajuetc.Tightflo kP n ed Mali/BambaraT ghtf ockPenned/
Tied Stall-FeedingWestAfrica/Ti d
"Moutondecase"
Kenya/"Thenges"
Penned
Loosefl ckPenned/tied
TightflockPenned
Tied
Dryseason notherded
Cropseason
tightflock
Tied
Tied Tied
Goats
Ranching
Kenya/Large
scalefarms
Extensivepaddocks
10000T0 00 200-2T 50-1T 20-1
200-T0
20-8 5-10 0-10 1-5
T0-1000
10
30-100 20-8 10-120
U10
1-20 1-5
Table32.Someex mp sofman gem ntobjectiv srelatedtflo kstructur s(structureasp rc nofto lani ls)
Sheep
Goats
Area/EthnicGroup
Males
Females
Use
Males
Females
Breeding
55.1
T8.1
51.2
T5.5 T8.3
Total 79.8 71.7 7T.T
9T.7 TT.2
Castrates
1.2 "few" 0.0 0.0 10.3
Total 20.2 28.3 23.T 3.3 33.8
Breeding
Milk/Meat Milk/Meat Milk/Meat
Milk
Meat(Fat)/
Milk
58.T 55.9 53.7 57.7 T1.T
A.2
Total 78.1
7T.5
73.3 77.8 92.2 T8.T
Castrates
T.2
11.3 "few" 0.0 0.0 15.T
Total 22.9 25.5
2§.7
22.2 7.8 31.T
Meat/Hair Meat/Wool Meat/Milk
Meat Milk
Meat(F )
Mauritania/Moor
Mali/Fulani Tchad/"Arab"
Sudan/Baggara Ethiopia/Afar Kenya/Masai
69
of Mmacro"-management systems and including more than ten thousand animals,
the mean percentage of females was 74. 7 with a standard deviation of ± 3. 07.
Breeding females in this sample showed even less variation at 54. 3 (± 2.43)% of
the total flock. Contrary to another popular misconception, there are very few old
unproductive females in the flocks, this class of stock being usually less than 5%
in large scale flocks and rarely exceeding 10% in the small agro-pastoral ones.
The main management practice used to achieve this structure and stability is the
early culling of males which are sold or slaughtered for home consumption. Males
of reproductive age are kept, strictly speaking, in numbers in excess of those
actually required for breeding. There is, of course, the consideration of losses
from diseases and of a temporary sterility as a result of nutritional deficiencies:
when these factors are taken into account the number of males is seen to be no
more than reasonable. Where older males other than breeding animals are kept
they usually contribute directly to the flock economy in terms of wool or hair or to
specialised dietary requirements such as, for example, in the case of the Masai.
In the humid zones flock structures are even more heavily weighted to females with
as many as 80% females in the derived savanna areas and 83% in the true forest
area (Mosi, Opasina, Heywood, Carew and Valez, 1982).
It has to be said that the true pastoralists are much better at flock manage
ment, for example in terms of foraging time allowed to their animals and in control
of mating to certain desired males, than the recent entrants into the livestock
industry. These agriculturalists and agro-pastoralists appear to have much to
learn before maximum production levels are achieved. With better management
and with more efficient use of agricultural by-products and tree fodder, producti
vity could be raised considerably from these areas.
Nutrition
According to IEMVT (1980) the studies done on the nutrient requirements of
small ruminants in the tropics are for the most part fragmentary. They quote
voluntary intake for sheep as between 1. 8 and 3. 0 kg dry matter for maintenance,
3. 0 - 3. 8 kg DM for the last six weeks of pregnancy and 4.4 - 6. 0 kg DM for
lactation, all figures expressed per 100 kg liveweight for 35 kg sheep. ILCA
studies in the semi-arid zone of Mali have shown large fluctuations in DM intake
over time from about 1.6 to 3.2% of liveweight for both sheep and goats with mean
values of 2. 6%. Seasonal trends here were difficult to detect but there was low
intake of DM from October to January (i. e. in the eary dry season) with goats
eating much less than sheep in the rains. On average mature male sheep ingested
0. 70 MJ of energy per day while goats averaged 0. 53 MJ. It has also been
difficult to isolate the effects of mixed species flocks on the nutrition requirements
and intake of goats and sheep. It is possible that either one or other species
largely determines the activity and consumption patterns of both at certain times
of the year, depending on the type of food available and this could have some effect
on overall productivity. In the dry zones of Mali goats spend as much as 87% of
their time on browse while sheep spend only 34% of theirs. ILCA studies in Kenya
have shown that goats spend 56% of their time browsing compared with no time at
all by sheep. While the truism that goats are principally browsers and sheep
70
mainly grazers thus appears to be confirmed, it is necessary to be cautious in this
respect. In studies in the humid zone Carew, Mosi, Mba and Egbunike (1982)
showed that goats spent 98. 7% of their feeding time browsing but sheep also spent
92. 6% of time browsing in the forest zone while in the derived savanna zone sheep
(61. 1%) actually spent more of the feeding time, and much more of total time, in
browsing than goats (52.2%).
The conventional wisdom of the greater efficiency of goats over sheep in the
digestibility of organic matter, crude protein and fibre (Devendra and Burns, 1970)
has also been challenged recently (McDowell andWoodward, 1982). It is apparent
that a great deal of further work is required before nutritional aspects can be pro
perly evaluated and the relation of nutrition to breeding physiology and reproductive
performance probably needs special attention.
Improved nutrition greatly increases the growth rate of indigenous animals,
but the seasonality of the food supply has only minor effects on reproduction with
both goats and sheep producing young all the year round even in the semi-arid
areas with monomodal rainfall. This is not the case for cattle where more than
60% of births occur in a 10-week period related to conception in the previous rainy
season.
Productivity
Table 26 showed that goats and sheep account for almost 17% of the total
domestic ruminant biomass. This in itself is a not inconsiderable figure but none
theless gives no indication of the real contribution of this class of stock to total
animal production. On account of the higher prolificacy and the shorter generation
cycle, offtake figures are much higher than for cattle or camels. Based on
figures for biomass and offtake rates, sheep and goats can be expected, as Table
33 shows, to contribute almost 30% of total meat protein (excluding poultry and
pigs) in the semi-arid zones. This high contribution is often not acknowledged,
the trade in sheep and especially goat meat being internal to the countries con
cerned or on a much smaller scale even within the flocks themselves. That the
figure of 30% per cent is not far out is supported by the data for registered
slaughterings in the principal towns in four of the Sahelian states shown in Table
34. In 1970 the figure was almost 36% and in 1976, after a lengthy period of
drought, sheep and goats contributed overall 43. 2% to meat production on average
and more than 70% in Niger. The ability to withstand drought conditions and to
recover from them much more quickly than cattle is a not inconsiderable factor
in the production potential of goats and sheep.
In addition, as can be seen from Figure 2, in drier areas, goats in
particular, and sheep to a lesser extent contribute to human welfare by assuring
a supply of milk at the time of year when cows' milk is not available.
At the level of the individual animal some of the ways in which the
relatively high productivity is achieved are shown in Table 35. Sheep give birth
for the first time at about 15 months of age on the average: goats are generally
71
Table 33. Contribution of domestic herbivores to liveweight biomass and to
meat production in eight African semi-arid countries.
Cattle Camels
Sheep +
Goats
Total liveweight biomass (tonnes)
Offtake rate (%)
Total carcass weight available at 50%
dressing percentage (tonnes)
Species as % of total liveweight biomass
Contribution of species to total meat
availability
14 100 000
14.5
990 000
1 900 000
6.2
60 000
2 800 000
30.0
410 000
75.3 10.0 14.7
2868 4
Table 34. The contribution of sheep and goats to registered slaughterings in
four African countries before and after a drought period ('000 livestock
unit equivalents and percentages)
Country
1970 1976
Cattle ^he!P + Total ^P + Cattle *he? + Total !,hef +
Goats Goats Goats Goats
% of total % of total
Mauritania 21 4 25 16 10 3 13 23
Mali 80 23 103 22 90 38 128 30
Niger 62 79 141 56 33 89 122 73
Tchad 56 10 66 15 55 13 68 19
All 219 116 335 35 188 143 331 43
two to three weeks earlier. Parturition intervals vary from 8 to 10 months.
Although the semi-arid races are not as prolific as the small races of the more
humid zones, twin births are common in goats and are far from unusual in sheep.
In these races the number of young born per year is thus in general about 1.6 per
breeding female for goats and somewhat less for sheep. In the forest races
because of the higher rate of multiple births, rates of annual reproduction in goats
are in excess of 2. 0 and not much less than this in sheep. In all types the first
litter is smaller, as would be expected, than subsequent litters.
Growth rates in the semi-arid races are some three times more rapid than
the rates in forest types when expressed simply as grams/day. However forest
types have similar or slightly greater productivity indices calculated per unit
72
Figure 2. Complementarity .of lactations in mixed species production systems
80
60
D
/
zZ
a
—>
iS
CJ
a
i—i
co 40
£
Cb
"3
c
CJ
U
CJ
U
20
■
\ Goat!
D
A v
D
\
/
Y
• □
a
Sheep °
\
n.
\
"□.
Cattle
D
' _1 I
J FMAMJJASOND
Month
cold
dry
post
rains
cold
dry
hot dry rains
Season
CO
Table35.Sommeasur dprod ctionp ram tersforAf i ang ansh pinse -arideas
AreaandEthnicGroup
Parameter
Mali
Sedentary
Tchad "Arab"
Sudan
Baggara
Ethiopia
Afar
Kenya Masai
Uganda (station)
GoatsSheepeepGoatsShGoatsSh ep
Ageatfirstparturi ion(days)T8T70154
Parturitionin erval(days)271A
Averagelittersize1.2305
Numberofyoungp rann m1.T552
Weightofyoungat1Tdays(k )2.9T 3
Dailygain(g)to1Td ys70.9 .7
ProductivityIndex
(gyoung/kgfemale/year)TT0120
1.1207 1.130T 9.521.5
T9.3120.0
(290) 238
1.57
2.T1 12.8 71.1
T20
(3T9)
275 1.1T 1.U 18.8 99.7 530
1.10 1.15
8.8 10.3 390
1.05 1.20
15.3 82.0
TT
TU 289
1.1T
1.T7 10.T 52.1
T32
510
3TT
1.02 1.08 13.7
T9.5
UT
2975
1.30
1.10
11.TT 3
TT.772 0 T80
7T0
74
weight of breeding female (ILCA, 1979b). The withdrawal of males for slaughter at
light weights (90% of marketed and home slaughtered males in Darfur, for example,
have liveweights less than 20 kg) represents some loss of potential meat avail
ability but is certainly an efficient use of the resource base.
There are few data available on aspects of production other than meat.
Such products include milk, wool, hair and skins; and the contribution of manure
to the agricultural systems with which goats and sheep in the less dry areas are
associated. Data collection on some of these aspects may well be beyond the means
of a simple animal scientist and may require considerable cross- disciplinary inputs
from sociologists and economists as well as agronomists. These products are
certainly worthy of much more attention than they have received.
While mortality rates in small ruminants are considered to be high (up to
40% in goats and 30-35% in sheep before weaning and up to 10% in older animals)
losses due entirely to disease are difficult to categorise. Country statistics in this
subject, as for animal numbers, must be to a certain extent suspect and while
there is a voluminous and rapidly growing literature on small ruminant disease -
at least when compared with productive aspects - it would be true to say that the
real causes of most mortalities are only suspected rather than known. Except in
certain clear-cut cases death usually supervenes as a result of a complex of factors
involving nutrition, management and disease.
Potential for improved production
The semi-arid zones of Africa represent a difficult environment with
seasonal, and occasionally much longer, severe stress periods. Different kinds
of stress, for example trypanosome infections and extremely heavy parasite
burdens affect forest types. Goats and sheep are obviously fairly well adapted to
these stress conditions. "Modern" technical inputs, even where available, are
beyond the financial, and often physical, reach of the owners.
Although nothing akin to a "green revolution" can be expected, this does not
mean that improvement is not possible. In the case of liveweight gain, for
example, consider one specific and one general example. In Darfur flocks studied
in 1972-74, one male sheep, reared under exactly the same conditions as his
contemporaries, gained 265 g per day to 12 weeks of age when he weighed 26. 5 kg:
this rate of gain compares very well with the 299 g mean achieved in New Zealand
for grass-fed lambs. A general example concerns the so-called "mouton de case"
of West Africa. In Mali, the feeding of rice bran, leaves of Khaya senegalensis
and cow-pea haulm can lead to weight gains which are 50% higher than those of
similar animals reared on the open range and1 on millet and rice stubbles. It is
apparent that the genetic base is not as impoverished as many people think. In
terms of weight gain, however, it is often just the fastest growing males, which
would be the best sires, which are removed from the flocks.
75
Management has been shown to be an important factor in overall perfor
mance. A general lift in production so that the worst flocks can be raised to the
level of the current average producer, thus raising the existing output by 15-20%,
should not be too difficult to achieve. This target could be attained by encouraging
the worst owners to follow the practices of the better ones. It is just this field -
the one of management - that is most likely to be successful in raising production
levels at the least cost.
76
77
SEASONAL BREEDING EFFECTS ON PRODUCTIVITY
Christie P. Peacock
A three year study of sheep and goats production has been conducted on the
Maasai group ranch at Elangata Wuas in Kenya, and at present 3 adjacent newly-
formed ranches are being studied.
The number of breeding females per breeding male varies between 10 and
36. This ratio is lower for sheep than goats, and the highest ratios are found for
the wealthiest animal owners, i. e. those who have the greatest numbers of animals.
Most flock owners practise breeding control over their flock. If there is enough
family labour males may be herded separately from the main flock. Alternatively
an apron made of leather is attached to the males. This prevents mating with very
young females or those suckling young. The annual distribution of rainfall is bi-
modal with the main rains from March to June, and the short rains in November
and December. Flock owners try to arrange that the majority of births occur after
the short rains, and so remove the aprons at the beginning of the dry season in June
when the body condition of the females is at its best. When the animal owner
observes a ewe standing to be mated he holds her, slips the ram's apron round and
holds the female's fat tail to the side while mating takes place. Once copulation
has finished, the owner holds the female's hind legs in the air to shake down the
semen. There are some births of small ruminants all year round, but on one
ranch as many as 64% of kid births occur in November indicating successful control
over breeding. This appears to be a sensible system as there should be adequate
fodder for the lactating dams and young lambs. However, if the short rains fail
then the births take place in the middle of a very long dry season.
Reproductive performance. The mean parturition intervals observed were 306 days
for goats and 312 days for sheep (Table 36). The season of parturition significantly
affected subsequent the parturition interval, those dams which gave birth in the
short dry season had the shortest intervals, and there were large differences
between flocks.
Mortality. The causes of pre-weaning mortality and the factors affecting
mortality are shown in Table 37. The major identified causes of death are disease
for goats and predators for sheep. Animals born as twins have much higher death
rates than those born as singles, and there is a seasonal effect on mortality with
the highest death rates found for kids born in the long dry season. There is a very
large difference between the mortality rates observed between the best and the
worst flocks, and this shows the potential for improvement in produc tivity by
management.
Growth. Mean weaning weights at 150 days were 10. 6 kg for goats and 14. 1 kg for
sheep. Season had a significant effect on weaning weight. Animals born in the
short dry season (and thus reared in the long rains) had the best growth rates and
were about 1.2 kg heavier at weaning than kids and lambs bom in the other seasons.
78
Productivity , The production index calculated as the weight of young at 5 months
(in grams) per kilogram of breeding female per year depends on parturition inter
val, pre-weaning mortality, weaning weight and post-partum dam weight. Table
38 shows the effect of litter size, season of birth and flock on the productivity
index. For goats, twin births give an advantage over singles, but the reverse is
true for sheep. Season of birth has a large effect (with the lowest indices resulting
from births in the long dry season) and there are large differences between flocks.
The approach to improvement in productivity is shown in Figure 3. The
main steps are identification of the best flocks and finding out why they are the best,
identifying what disease control measures could and should be taken, seeing if any
nutritional improvements can be made, and by selecting for twinning in goats and
against twinning in sheep.
Table 36. Mean parturition intervals (days) of goats and sheep in the Maasai
pastoral system in Kenya, and the effect of season of parturition on
parturition interval. Means with different postscripts differ
significantly (p < 0.05).
Goats Sheep
Overall mean 306 312
Season of parturition
Short dry
Long rains
Long dry
Short rains
281 a 286 c
326 b 304 cd
291 a 336 e
326 b 322 de
79
Table 37. Causes of pre-weaning mortality (% deaths) and factors affecting
pre-weaning death rates (%) of small ruminants in the Maasai
pastoral system in Kenya,
Cause of death
Gtoats Sheep
Lost 9.0 15.0
Disease 31.0 17.0
Predator 8.3 25.0
Drought/malnutrition 6.9 2.0
Other 44.8 41.0
Birth type
Single 18.8 7.7
Twin 25.8 31.4
Season of birth
Short dry 23.8 17.2
Long rains 14.6 13.9
Long dry 31.4 22.4
Short rains 19.3 22.6
Flock
Best 12.4 12.2
Worst 40.0 25.8
Table 38. Production indices of small ruminants in the Maasai past:on
system in Kenya (weight of young at 5 months (g)/unit weigh
dam (kg)/year).
Birth type
Goats Sheep
Single 397 489
Twin 463 313
Season of birth
Short dry 477 460
Long rains 447 403
Long dry 383 242
Short rains 412 500
Flock
Best 541 610
Worst 345 270
80
Figure 3. Potential improvement pathways for sheep and goat flocks in
Maasai group ranches.
 
Identify best flocks,
identify causes and extend
to other flocks
Identify causes of
pre-weaning mortality
Selection against
twinning in sheep
Select for twinning
in goats
 
 
Pathological
Veterinary
Measures
Nutritional
Improve
food supply
Obtain maximum
production of young
(Parturition Interval
and Litter size)
 
Select within flocks for
fastest growing males
breed from them (growth
rate) and their dams
(milk production)
Optimise parturition
interval in goats
Seasonal breeding
in sheep
 
81
DISEASE AS A CONSTRAINT TO PRODUCTIVITY
S. Mack
An ILCA base line survey is being conducted in villages in South West
Nigeria. Two ecological zones are represented, the forest and the derived savanna
(a man-made fire climax savanna bordering the forest). Annual rainfall in the area
is about 1200 - 1800 mm, minimum temperature 18 - 24° C, maximum temperature
29 - 35°C and there is a dry season extending from mid-November through to
March. Flock sizes are small with 3-4 animals per household and the ratio of
goats to sheep is about 3:1. Very few sheep are found in the savanna zone because
many villages ban sheep, as it is sheep not goats that damage crops.
West African Dwarf sheep and goats are both classed as trypanotolerant.
Their reputed low productivity may be a result of the environment in which they
live rather than a low performance potential. The reproductive performance of
these breeds is good. In South West Nigeria the mean age at first parturition is
560 (±22) days for goats and 540 (±28) days for sheep. Mean litter size is high
(1. 6 for goats and 1. 3 for sheep) giving annual reproduction rates of 2. 1 kids/doe/
year and 1. 7 lambs/ewe/year.
During the first 18 months of the survey, disease was identified as a major
constraint. Table 39 shows the main causes of morbidity. PPR (Peste des Petits
Ruminants) or Kata is a viral disease closely related to rinderpest, human
measles and canine distemper. Farmers in the region report PPR to be cyclic
with major outbreaks every 1-3 years. Goats were found to have an unexpectedly
high incidence of mange, caused mainly by Sarcoptes scabiei. Both PPR and
mange were primarily problems with goats rather than sheep. Lice and fleas were
the common ecotoparisites in goats, and Rh-ipicephalus and Amblyomma spp. of
tick were found in sheep. Haemonchosis was the major helminth problem.
Pneumonia was a commonly reported symptom. In reality it more accurately rep
resents an umbrella syndrome, covering both primary and secondary pneumonias.
Having identified PPR as a major cause of mortality and mange as a major
pre-disposing cause, a veterinary intervention package was instigated. This
package consists of control of PPR using an annual vaccination of tissue culture
rinderpest vaccine (TCRV) and control of mange and other ectoparasites by monthly
dipping with Gammatox.
After 12 months use of this package, preliminary results show a major
reduction in reported cases of PPR in goats: from 14% to 4% in the forest and
from 18% to zero in the savanna. Dipping had a major impact on the reported in
cidence of mange from 38% to 11% for forest goats, from 61% to 3% in savanna
goats and from 13% to zero in forest sheep. Pneumonia and helminthiasis rose in
relative importance. The actual number of observed cases of illness fell from 237
to 56 for forest goats, from 357 to 61 for savanna goats and from 128 to 43 for
forest sheep.
82
Table 39. Main causes of morbidity in small ruminants in Nigeria as % of
observed cases, before and after veterinary intervention.
Disease
Goats Sheep
Forest Savanna Forest
Before After Before After Before After
PPR 13.5 3.6 17.6 0.0 0. 0 0. 0
Pneumonia 7.6 17.9 5.0 21.3 3.1 7.0
Footrot 0.4 8.9 0.6 3.3 7.0 23.2
Helminths 6.7 16.1 2.8 24.6 16.4 11.6
Ecto-parasites 8.6 1.8 2.3 0.0 26.4 0.0
Mange 37.5 10.7 60.7 3.3 13.3 0.0
Accidents 13.9 16.1 4.7 19.6 28.9 30.2
Others 11.8 25.0 6.3 27.9 4.7 27.9
Total numbers 237 56 357 61 128 43
Table 40 shows the effect of the health package on mortalities. For goats
this fell from 30% to 12% in the forest and from 40% to 13% in the savanna. The
drop in mortalities for sheep was less from 20% to 14%. This was probably because
the health package was aimed more at the goat diseases to which the sheep were
less susceptible.
Table 40. The effect of veterinary intervention on the mortality and offtake of
small ruminants in Nigeria
Mortality (%) Offtake (%)
Before After Before After
Goats in forest
Goats in savanna
Sheep in forest
29.9
40.1
20.1
12.1
13.3
13.6
25.4
20.4
32.8
21.0
17.7
28.6
Overall offtakes were slightly lower after the veterinary intervention than
before, but this difference was not significant. However, for bucks in the 0-90 day
age class, offtakes fell from 6. 5% to 0. 9% (significant at p < 0. 05). Increased
viability would be expected to reduce offtake in this age group, due to the reduction
in emergency slaughters.
83
The consequence of a fall in mortality and a static offtake was that
inventories rose (Table 41). In both sheep and goats the major increases in
numbers occurred in the post-weaning age group. For forest goats there was an
overall increase of 34%, and 18% for sheep, and in the savanna 39% for goats.
Excluded from the data in Table 41 were animals transferred into the survey for
share cropping. Farmers were quick to see the benefits of the health package and
many animals were brought into survey area from friends and relations.
Table 41. Flock inventories before and after veterinary intervention
Goats Sheep
Age group
Forest Savanna Forest
Differ
ence (%)
Differ
ence (%)
End Mto!T
ence (%)
Start End Start End Start
0-90 days
91-360 days
360
45 51 13.3
91.1
10.5
43
30
50
56
98
13.3
53.6
32.4
19
19
18
30
55
-5.3
57.9
12.2
45 96
114 126 74 49
Total 204 273 33.8 147 204 38.7 87 103 18.4
The overall benefits of the health package in terms of productivity indices
are shown in Table 42. Unfortunately, there were insufficient data for sheep to
make a comparison before and after intervention. No statistical significance has
yet been attached to the differences in productivity indices. These are preliminary
results and a detailed analysis will be carried out once data for a further 12 months
have been collated.
Table 42. The effect of veterinary intervention on the productivity indices of small
ruminants in Nigeria. Index I is total weight of offspring at 5 months/
dam/year. Index II is total weight of offspring at 5 months/unit weight
of dam (kg)/year. Index III is total weight of offspring at 5 months/unit
metabolic weight of dam (kg°*75)/year.
Weight of Index I Index II Index IH
dam (kg) (kg) (g) (g)
Goats in forest Before 15.1 7.5 500 1010
After 16.7 10.3 620 1280
Goats in savanna Before 13.9 7.2 520 1050
After 14.7 10.5 710 1430
Sheep 21.1 21.1 690 1480
84
85
CONCLUSION
At present small ruminants play an important part in the livelihood of
people throughout Africa. Their productivity is not yet well documented, but there
is no doubt that their ability to survive in harsh environments and provide a number
of products is substantial.
In an attempt to increase the productivity of small ruminants both to
improve the standard of living of the animal owners, and the contribution of small
ruminants to gross national products, constraints to production are now being
investigated in a number of situations. The main technological constraints appear
to be animal nutrition (particularly in the more arid areas) and animal disease
(particularly in the more humid areas). In addition, sociological constraints,
particularly marketing, may limit productivity and the response of the system to
technological improvements. The breed types available in tropical Africa are not
limiting factors except where other constraints have been overcome. Crossbreed
ing schemes have generally resulted in failure, with disease and nutritional pro
blems causing high mortality and poor productivity. There is room however for
selection within and possibly between, the indigenous breeds. This must take
place under realistic conditions in the field where so many problems are encoun
tered.
At the seminar delegates reported that research had been undertaken to
investigate the productivity of indigenous breeds of sheep and goats in Botswana,
Cameroun, Ethiopia, Ghana, Liberia, Nigeria, Senegal, Uganda and Zimbabwe.
A large proportion of this information has not been published, and some has not
yet been analysed. However in some countries such as Nigeria, there is already
a national programme for coordinating and publishing research on indigenous
breeds.
At present it is difficult to obtain international funding to study small
ruminants in Africa. Several proposals were put forward to remedy this situation,
including a large African conference on small ruminants, approaching deans of
agricultural and veterinary faculties encouraging them to devote more time to
teaching about small ruminants in universities, and setting up an operation similar
to the ILCA/FAO/UNEP study on trypanotolerant livestock (ILCA, 1979b). In such
an approach, relevant information on sheep and goats might be collected in conjun
ction with national organisations in a number of African countries and collated into
a report which could be presented to international funding agencies as the basis of
a proposal for future financial support. A network might then evolve on similar
lines to the trypanotolerance network and ILCA could assist in coordination of
research on small ruminants throughout Africa, to ensure the best utilisation of
research resources.
86
87
REFERENCES
AOAD and FAO (1979). The market for livestock and meat in the Arabian
Peninsula and the role of supplies from neighbouring African surplus
regions. Arab Organisation for Agricultural Development and Food and
Agriculture Organisation of the United Nations, Rome, Italy.
Brinckman, W.L. (1981). Concentrate feeding of sheep in Nigeria. In Intensive
animal production in developing countries, Eds Smith, A.J. and Gunn, R.G.,
pp. 421 - 428. British Society of Animal Production, Thames Ditton, UK.
BVA (1976). Handbook of animal diseases in the tropics, 3rd ed, Ed. Robertson,
A. British Veterinary Association, London, UK.
Carew, B.A.R., Mosi, A.K., Mba, A.U. and Egbunike, G.N. (1982). The
potential of browse plants in the nutrition of small ruminants in the humid
forest and derived savanna zones of Nigeria. In Browse in Africa, the
current state of knowledge, Proceedings of the International symposium on
browse in Africa, Ed.; Le Houerou, H. pp. 307-311. International
Livestock Centre for Africa, Addis Ababa, Ethiopia.
Cossins, N.J. and Bekele, Y. (1974). Still sleep the highlands. Livestock and
Meat Board, Addis Ababa, Ethiopia.
CSWRI (1981). Central sheep and wool research institute. CSWRI, Avikanagar,
Jaipur, Rajasthan, India.
De Montgolfier-Kouevi, C. and Vlavonou, A. (1982). Trends and prospects for
livestock and crop production in tropical Africa, Research report 1.
International Livestock Centre for Africa, Addis Ababa, Ethiopia.
Devendra, C. (1980). Potential of sheep and goats in less developed countries.
Journal of Animal Science 51* 461 " 473»
Devendra, C. and Burns, M. (1970). Goat production in the tropics. Common
wealth Agricultural Bureaux, Farnham Royal, UK.
Dumas, R. (1977). Etude sur Pelevage des petits ruminants du Tchad. Institut
d'Elevage et de Medicine Vet£rinaire des Pays Tropicaux, Maisons
Alfort, France.
El Hag, G.A. and Mukhtar, A. M.S. (1978). Varying levels of concentrates in the
rations of Sudan Desert sheep. World Review of Animal Production 14(4)
73 - 79. '
FAO (1981a). Production yearbook 1980, vol. 34. Food and Agriculture
Organisation of the United Nations, Rome, Italy.
88
FAO (1981b). Trade yearbook 1980, vol. 34. Food and Agriculture Organisation
of the United Nations, Rome, Italy.
Gall, C. (1975). Milk production from sheep and goats. World Animal Review 13,
1-8.
Harvey, W.R. (1977). User's guide for LSML76: Mixed model least squares and
maximum likelihood computer program. Monograph of Ohio State University,
Columbus, USA.
IEMVT (1980). Les petits ruminants d'Afrique central et d'Afrique l'ouest:
Synthase des connaissances actuelles. Institut d'Elevage et de Medicine
Vet6'rinaire des Pays Tropicaux, Maisons Alfort, France.
ILCA (1979a). Small ruminant production in the humid tropics, Systems study 3.
International Livestock Centre for Africa, Addis Ababa, Ethiopia.
ILCA (1979b). Trypanotoierant livestock in west and central Africa, Monograph 2.
International Livestock Centre for Africa, Addis Ababa, Ethiopia.
ILCA (1979c). Base papers for programme committee meeting, internal document.
International Livestock Centre for Africa, Addis Ababa, Ethiopia.
Jahnke, H.E. (1982). Livestock production systems and livestock development in
tropical Africa. Kieler Wissenschaftsverlag Vauk, Kiel, West Germany.
Livestock and Meat Board (1974). Survey of Ethiopian carpet industry. Livestock
and Meat Board, Addis Ababa, Ethiopia.
McDowell, R.E. and Hi1debrand, P.E. (1980). Integrated crop and animal produc
tion: making the most of resources available to small farms in developing
countries. Rockefeller Foundation, New York, USA.
McDowell, R.E. and Woodward, A. (1982). Concepts in animal adaptation:
comparative suitability of goats, sheep and cattle to tropical environments.
Proceedings of the 3rd international conference on goat production and
disease, pp. 387 - 393. Tucson, USA. ISBN 0 930848 17 9.
Mosi, A.K., Opasina, B.A., Heyward, B., Carew, B.A.R. and Valez, M. (1982).
Personal communication.
Nie, N.H., Hull, C.H., Jenkins, J.G., Steinbrenner, K. and Bent, D.H. (1975).
Statistical package for the social sciences, 2nd ed. McGraw-Hill,
New York, USA.
Rice, R.W. (1976). Extensive range conditions. In the role of sheep and goats
in agricultural development, pp. 15 - 16. Winrock International Livestock
Research and Training Center, Morrilton, Arkansas, USA.
89
SABRAO (1980). Animal genetic resources in Asia and Oceania, Proceedings of
a workshop of the Society for the Advancement of Breeding Researches in
Asia and Oceania. Tropical Agriculture Research Center, Yatabe,
Tsukuba, Ibaraki, Japan.
Technical Committee (1975). National policy on sheep research and development.
Ministry of Agriculture, Addis Ababa, Ethiopia.
Trail, J. CM. (1981). Merits and demerits of importing exotic cattle compared
with the improvement of local breeds. In Intensive animal production in
developing countries, Eds. Smith, A.J. and Gunn, R.G., pp. 191-231.
British Society of Animal Production, Thames Ditton, UK.
Weiner, G. (1972). An investigational framework to assist livestock development
in Ethiopia with special reference to breeding. Food and Agriculture
Organisation of the United Nations, Rome, Italy.
Weiner, G. (1975). The cattle and sheep breeding research programme at the
Institute of Agricultural Research, Ethiopia. Food and Agriculture
Organisation of the United Nations, Rome, Italy.
Winrock (1977). Background notes for A state of the arts study on The role of
sheep and goats in agricultural development. Winrock International
Livestock Research and Training Center, Morrilton, Arkansas, USA.
Wilson, R.T. (1980). Population and production parameters of sheep under
traditional management in semi-arid areas of Africa. Tropical Animal
Health and Production, 12, 243 - 250.
Wilson, R.T. (1981). Livestock production in central Mali: attempts to produce
raw materials of animal origin for the french textile industry during the
colonial period. Textile History, 12,. 104 - 117.
90
91
PARTICIPANTS
Ato Addis Anteneh,
ILCA,
P.O. Box 5689, -
Addis Ababa,
ETHIOPIA
Dr. P.J. Brumby
ILCA,
P.O. Box 5689,
Addis Ababa,
ETHIOPIA
Ato Adugna Kitila,
Debre Birhan Sheep Station,
c/o Ministry of Agriculture,
P.O. Box 62347,
Addis Ababa,
ETHIOPIA
Dr. Frank M. Anderson,
ILCA,
P.O. Box 568$
Addis Ababa,
ETHIOPIA
Dr. M. H. Butterworth,
ILCA,
P.O. Box 5689,
Addis Ababa
ETHIOPIA
Ato Debebe Agonafer,
Rangelands Development Project,
Ministry of Agriculture,
P.O. Box 1052,
Addis Ababa
ETHIOPIA
Dr. Assefa Adane,
Animal Health & Breeding Department,
ARDU,
P.O. Box 7,
Arssi, Assela,
ETHIOPIA
Ato Demissie Tiyo,
Bako Agricultural Research Centre,
c/o Institute of Agricultural Research,
P.O. Box 2203,
Addis Ababa,
ETHIOPIA
Dr. W. Bayer,
ILCA,
P.M.B. 2248,
Kaduna
NIGERIA
Mr. J. Durkin,
ILCA,
P.O. Box 5689,
Addis Ababa
ETHIOPIA
Dr. Beyene Kebede
Holleta Agricultural Research Station,
c/o Institute of Agricultural Research,
P.O. Box 2203,
Addis Ababa,
ETHIOPIA
Ato Binyam Akallu
Melkawerer Research Centre,
c/o Institute of Agricultural Research,
P.O. Box 2203,
Addis Ababa,
ETHIOPIA
Dr. Abdou Fall,
Centre de Recherches Zootechniques
de Kolda,
B.P. 53,
Kolda,
SENEGAL
Dr. Ruth M. Gatenby,
Centre for Tropical Veterinary
Medicine,
Easter Bush, Roslin,
Midlothian, EH25 9RG,
Scotland,
U.K.
92
Ato Gebre Tsadikan Hailu,
Holletta Agricultural Research Station,
c/o Institute of Agricultural Research,
P.O. Box 2203,
Addis Ababa,
ETHIOPIA
Ato Goshu Makonnen,
Debre Zeit College of Agriculture and
Research Station
P.O. Box 32,
Debre Zeit,
ETHIOPIA
Mr. J.L. Grant,
P.O. Box 3701
Marondera,
ZIMBABWE
Mr. C. de Haan,
ILCA,
P.O. Box 5689, ;
Addis Ababa,
ETHIOPIA
Dr. Kano Banjaw,
Department of Animal Science,
Alemaya College of Agriculture,
P.O. Box 138,
Alemaya,
Dire Dawa,
ETHIOPIA
Mr. S. Mack,
ILCA,
P.M.B. 5320,
Ibadan,
NIGERIA
Ato Mohammed Yusuf,
Animal Research Section,
Arssi Rural Development Unit,
P.O. Box 7,
Arssi, Assela,
ETHIOPIA
Dr. L. Mtenga,
Faculty of Agriculture, Forestry and
Veterinary Service,
P.O. Box 643,
Morogoro,
TANZANIA
Professor J.S. Mugerwa,
Department of Animal Science,
Makerere University,
P.O. Box 7062,
Kampala
UGANDA
Mr. S.E.C. Nyoni,
Mikolongwe Livestock Centre,
P.O. Box 5193,
Limbe,
MALAWI
Dr. J. Lambourne,
ILCA,
P.O. Box 5689,
Addis Ababa,
ETHIOPIA
Mr. S. Lebende,
Service de L'elevage,
Ministere de Developpement Rural,
B.P. 7026,
Ouagadougou,
UPPER VOLTA
Dr. M. Olayiwole,
NAPRI,
Ahmadu Bello University,
P.M.B. 1096,
Zaria,
NIGERIA
Miss Christie P. Peacock,
ILCA,
P.O. Box 46847,
Nairobi,
KENYA
Miss Louise L.
APRU,
P.M.B. 0033,
Gaborone,
BOTSWANA
Lethola, Mr. A. Rennie,
Ministry of Agriculture,
P.O. Box 9010,
Monrovia,
LIBERIA
93
Miss Jennifer G. Sandford,
ILCA,
P.O. Box 46847,
Nairobi,
KENYA
Mr. P.P. Semenye,
ILCA,
P.O. Box 46847,
Nairobi,
KENYA
Dr. B. Taiwo,
NAPRI,
Ahmadu Bello University,
P.M.B. 1096,
Zaria,
NIGERIA
Mr. H. Tawanezvi,
Matopos Research Station,
P.M.B. K5137,
Bulawayo,
ZIMBABWE
Dr. J. CM. Trail,
ILCA,
P.O. Box 46847,
Nairobi,
KENYA
Ato Tsige Yohannes Habtu,
Awassa Junior College of Agriculture,
P.O. Box 5,
Awassa, Sidamo,
ETHIOPIA
Dr. Helen Newton Turner,
CSIRO,
P.O. Box 184,
North Ryde, NSW 2113,
AUSTRALIA
Mr. R.T. Wilson
ILCA,
P.O. Box 60,
Bamako,
MALI
Mr. J. Tchoumboue,
ENSA,
P.O. Box 638,
Yaounde,
CAMEROON
94
95
INDEX
Angora 17, 30, 62
Australia 4, 5, 11, 25-30
Botswana 10
Breed evaluation 3, 31-35, 37-39
Breeding 1-6, 19-20,
Breeding control 77
Breeds - see under individual breed names
Carcass quality 1, 16
Cashmere 17, 30
Computers - see under Data handling
Cooperative breeding schemes 4, 29
Crossbreeding 5, 19, 22-23, 62
Data handling 29, 39-40, 41-59
Death rate - see under Mortality
Disease 8, 19, 22, 74, 81- 83
Economics 7-12, 22
Ethiopia 8, 10-12, 61
Exotic breeds 19, 32-35
Export 7-12, 28
Extension service 1, 13, 14, 22
Fattening 16, 17, 74
Feed conversion efficiency 16
Feeding - see under Nutrition
Field evaluation - see under Breed evaluation
Flock structure 66, 68
Fodder conservation 18
Fodder trees 18, 22
Foreign exchange 9, 11
Growth rate 9, 16, 71-74, 77
Hair 2, 17
Helminthiasis - see under Parasites
Heritability 2, 3
Heterosis 2, 5, 19
Hybrid vigour - see under Heterosis
Improvement 1, 13-20, 22-24, 74-75, 80.
Inbreeding 4, 20
india 11, 21-24.
Indigenous breeds 3, 9, 32-35, 64, 74
Karakul 62
Kenya 10, 12, 19, 66, 77-80
Killing out percentage 16
Lactation 72
Lambing percentage 25, 28
Legumes 18
Litter size 2, 4, 71-73
Liveweight gain - see under Growth rate
Maasai 62, 63, 77-80
Mali 61, 65, 74
96
Management 31, 37, 65-69, 75, 77-80
Mange 19, 22, 81-82
Manure 8, 23, 74
Marketing 7-12, 17, 26-28
Meat quality - see under Carcass quality
Merino 22, 23, 25-29
Milk 2, 7, 16, 61, 72
Mohair - see under Angora
Morbidity - see under Disease
Mortality 22, 32, 74, 79, 82
Native breeds - see under Indigenous breeds
Nigeria 8, 10,65, 81-83
Nutrition 17-18, 69-70
Parasites 19, 81-82
Parturition interval 2, 78
Peste des petits ruminants 19, 81-83
Ram exchange 5, 22
Red Sokoto 8, 63
Reproductive rate 70-73
Seasonal breeding 2, 3, 77-80
Selection differential 3, 4
Skins, 1, 7, 8, 17, 19, 61
Slaughter houses 10
Somalia 11
Statistical analysis 39-40, 53
Sudan 11, 64
Supplementary feeding 9, 18, 65, 74
Twins - see under litter size
Veterinary treatment 81-83
West African Dwarf goat 62, 81-83
West African Dwarf sheep 62, 81-83
Wool 1-2, 7, 12, 17, 22-23, 25-29
Zimbabwe 10

 Printed by ILCA, Addis Ababa, Ethiopia