Proceedings of the Second Workshop of the Crop-animal Systems Research Network (CASREN) held in Kunming, People's Republic of China 1- 7 May 2000 International Livestock Research InstituteI I INTERNATIONAL lI\I?STOC.. ARCH INSTITUTE IMPROVING THE CONTRIBUTION OF LIVESTOCK TO CROP-ANIMAL SYSTEMS IN RAINFED AREAS IN SOUTHEAST ASIA Proceedings of the Second Workshop of the Crop-animal Systems Research Network (CASREN) Kunming, People's Republic of China 1- 7 May 2000 Edited by c. Devendra and A.S. Frio INTERNATIONAL. International livestock Research Institute LMSTOO< RESfARCH P.O. Box 30709 Nairobi Ken y a Au g ust 2000INSTITUT E ' , , ISBN 92-9146-082-6 Oevendra C. and Frio A.S. (eds). 2000. Improving the contribution of livestock to crop- animal systems in rainfed areas in Southeast Asia. Proceedings of the Second Workshop of the Crop-animal Systems Research Network (CASREN) held in Kunming, People's Republic of China, 1-7 May 2000. ILRI (International Livestock Research Institute), Nairobi, Kenya. 327 pp. ILRI INT?RNATI~AL lMST00< ~~MO. INSTITUT E T ABLE OF CONTENTS PREFACE v VISUMMARY SESSION I. Ecoregional research on crop-animal systems Crop-animal research perspectives, H. Li-Pun 3 Livestock to promote rural income diversification and growth in Vietnam: finding appropriate policy options M.A.Jabbar 10 GIS application for site characterization and definition of recommendation domains L.A.Lapar 17 Improved feed production and utilization to address year-round feeding systems: guidelines C.DevendraandD.Pezo 44 Design and ex-ante analysis for technical interventions D.Pezo 57 Methodologies for analysis of household survey data D. PezoandC.Leon-Velarde 85 Discussions on Resource Paper Presentations. 102 SESSION II. Country Presentations Household survey results and implications for future research in China z. Kaidian, C. Wenyang, S. Honglu and L. Jianping. ..107 Crop-animal systems research in Philippine rainfed areas: the Philippine BMS E. C. Villar, E.F. Lanting, D.C. Cardenas, NE. dela Cruz, A.M. Rebugio, A.M.P. Alo andV:M.C.DeRoma 121 Crop-animal systems research in Thailand M. Wanapat, A.Polthanee and C. Wachirapakorn 187 Household survey results and implications for further research at Cilawu, Indonesia A. Djajanegara, A. Priyanti, B. Risdiono, andD.Lubis 223 Benchmark site characterization in Dong Tam, Dong Phu, Binh Phuoc Province, Vietnam K. M. Luc, L .P. Dung, L. v: Ly, andP. V:Biem 245 iii Discussions on Country Presentations 279 281 286 289 292 293 SESSION III. Proposed Interventions China. , , ..., Philippines. ., .., , , .., ., ., , , .., , .. Thailand, , , , ., , ..., ..., , , .., , Indonesia."..,.,..",.."..."" Vietnam. , , Discussions on Proposed Interventions. . 295SESSION IV. Final Discussions ANNEXES 299 303 Workshop programme Listofparticipants Questionnaire for detailed characterisation of benchmark sites (BMS) in CASREN countries. iv PREFACE The Second Workshop of the Crop-Animal Systems Research Network (CASREN) project embodies continuing activities on increasing the contribution of livestock to improVing productivity of crop-animal systems in Southeast Asia. This workshop was funded by the Asian Development Bank (ADB) and was held at the Yunnan Beef and Cattle Research Centre, Kunming, China, between 1-7 May 2000. The project's member countries, China, Indonesia, Philippines, Thailand and Vietnam participated in the workshop. The workshop was opened by Mr. Lee Da Ping, Director of the Yunnan Animal Husbandry Bureau, China. Dr. H. Li-Pun, Project Manager of the Crop-Livestock Systems in the rainfed lowland regions of the Southeast Asia initiative, introduced the objectives of the meeting and emphasised the importance and opportunities for research and development in the CASREN project. The programme for the meeting and list of participants are reflected in Appendices 1 and 2. Four sessions were involved. Session one focused on ecoregional research on crop-animal systems and provided background information relevant to the project. They included presentations on crop-animal research perspectives; policy options for the livestock sector; GIS application for site characterisation and definition of research domains; meth~ologies to address all year round feeding systems; ex-ante analyses for technical options; and methodologies for data analyses. Session two was devoted to presentations on the results of the household surveys, and based on these, session three discussed the proposed interventions by individual countries. The latter two sessions involved considerable discussion in ensuring the relevance and implementation of the proposed interventions. Dr. Patrick Safran of ADB also made useful observations and suggestions on the survey findings. Final discussions during the open forum enabled broader discussions as well on any residual matters concerning the work plans and budgets. In the fourth session, a very useful field trip over one and a half days was made to the Bix.i Xiang, the BMS site in Nanjin County to appreciate and discuss progress in the research activities ILRI wishes to thank ADB for their interest and support, and to our Chinese hosts in the Yunnan Animal Husbandry Bureau, Beef and Cattle and Pasture Research Centre in Xiashao, and the many officials in Nanjian County for the excellent arrangements and kind hospitality. We also acknowledge with thanks, the participation and valuable contribution of members of the Steering Committee, national coordinators, the ADB and others in the continuing activities and success of the project. C.DEVENDRA' Project Coordinator vPreface IMPROVING THE CONTRIBUTION OF LIVESTOCK TO CROP-ANIMAl SYSTEMS IN RAINFED AREAS IN SOUTHEAST ASIA SUMMARY 1) Resource papers introduced various aspects of ecoregional research in crop-animal systems to include the research perspectives, policy options for promoting rural income diversification in the livestock sector, GIS applications, improved feed production and utilisation, design and ex ante analysis for technical interventions and methodologies for analysis of household survey data. These were followed by country presentations that dealt in detail on the results of characterisation of the systems in the extensive household surveys, their significance and relevance, methodologies used and the lessons learnt. 2) The survey results across the five benchmark sites (BMS) indicated the animals made a significant contribution to food production, food security, and income generation. A variety of crop-animal interactions were identified. Animal manure was consistently used as fertiliser for crop production. Major constraints to production included feed resources and nutrition, animal health, small size of farms, and access to credit and services. The results enabled wide discussions that focused on the interventions identified, their implementation and anticipated results. The detailed questionnaire that was developed and used appropriate to crop-animal systems in all five countries is given in Annex C. 3) The importance of analyzing the large body of infonnation gathered to enable more complete intetpretation of the results and cross-country comparisons of crop- animal systems was emphasised. ILRI will provide support to completing this task, which will be important in the determination of needs-based research and recommendation domains . 4) In China, the importance of the work and the relevance of the results were made more useful through a field visit to the BMS in Bixi Xiang in Nanjian County, where participants were able to see research activities that respond to overcoming constraints and the needs of poor farmers. 5) A separate meeting was held with the National Coordinators to discuss project- related issues, sharing of the available funds, and the importance of ensuring good results concerning the role and contribution of animals. There was also discussion on the new project proposal development, which will build on the results of the present project and focus especially on poverty reduction. vi Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia 6) It was decided that a training course on ex-ante assessment be convened in an appropriate location to strengthen research capacity in NARS associated with the project. 7) The participants emphasised the need for more information exchange that is appropriate to crop-animal systems. 8) It was decided that in year 200 1, the third workshop would be merged with the proposed international conference to discuss and disseminate the results. The venue remains to be finalised, but will either be Indonesia or Thailand. viiSummary SESSION 1 ECOREGIONAL RESEARCH ON CROP-ANIMAL SYSTEMS CROP-ANIMAL RESEARCH PERSPECTIVES H. Li-Pun International Livestock Research Institute Addis Ababa, Ethiopia INTRODUCTION . Importance of crop animal-systems Holistic approach needed given complexities: bio-physical and socio- economic Linking the production to consumption chain: policy and technology research The ecoregional perspective: building synergies CASREN: An ecoregional project . Role of Crop-animal Systems in Rural Development . . . . Asset building for the poor: savings, buffering against inflation and climatic change Cash flow Nutrition, health and cognitive capacity: micronutrients Environment: reduce use of fertilisers, e.g., US$ 404-697 million in N and US$124-183 million in p in irrigated areas in Asia SYSTEMS RESEARCH Why? Problems of small farmers are complex due to: . . . . . Restrictions to capital and inputs Risk aversion Access to markets and services Diversity ofactivities: crops, animals, household, off-farm Interactions Cultural preferences and practices Lessons Learned Benefits: .Understanding complexities .Development of improved component technologies: improved feeding, management, health 3Crop-animal research perspectives J tp . Technology adoption: for example, the fodder bank technology developed by ILRI in West Africa Complementarity of systems and component research Capacity building. Organisation Specialisation Intensification - ,~ Area-wide crop livestock integration ~ Intensification J Nutrient surplus .. -- Animal systems development pathways (from Steinfield, de Haan, and Blackburn, 1996). 4 Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia SYSTEMS RESEARCH METHODOLOGY DESIGN l~ I DATA BASE I--. 'L ON-FARM . TECHNOLOGY CROP-ANIMAL RESEARCH PERSPECTIVES: METHODOLOGY Steps 1. 2. 3. Characterisation .Regional: GIS maps, integration of geographic and socio-economic data .Selection of benchmark sites .Description: socio-economic and bio-physical .Analysis of constraints: on-fann and off-fann Design .Analysis ex-ante: constraints and opportunities to improve the system .Discussion of results with stakeholders .Refmement Testing .Research on component technology and improved systems .Selection of cooperating fanners .Agreement on interventions .Implementation .Monitoring and evaluation .Feedback Crop-animal research perspectives 5 4. Validation .Large-scale testing Diffilsion and transfer of technology5. ECONOMIC ::;;ti. \' SYSTEMS fi~1 s y N T H E S I S COUNTRIES socIO- POLITICAL ,'1 l"i;~1 SYSTEMSECOREGION ~ COUNTRY , / / / , LOCATION (BMS) , / / AGRICUL TURAL SYSTEMS A N A L y S I S / / BREEDS ANIMAL SYSTEMS NUTRITION HEALTH .. Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia6 ECOREGIONALRESEARCH Adopted in 1993 by the CGIAR as a mechanism for: .Increasing research on the conservation and management of natural resources, linking increases in agricultural productivity with sustainable use of resources .Promote linkages of the CGIAR with NARS partners What is ecoregional research? . . Ecoregion: an ecological area defined within geographic boundaries (TAC, 1992) It deals with a region, not a fann, nor a continent It bridges the gap between basic, applied, and bio-physical and socio-eco- nomIc sciences It rectifies the erroneous assumption that environment is an independent vari- able It allows the systematic study of changes in land use and agricultural systems (Rabbinge, 1995) Differences between farming systems research (FSR) and ecoregional research (ECOR) Characterisation .Exogenous factors: markets, services ~ FSR: at the start of the project ~ ECOR: periodically .Unit of analysis ~ FSR: farm ~ ECOR: watershed (farm and off-farm factors, differences in scale} .Participation ~ FSR: mainly scientists and farmers ~ ECOR: scientists, community members, decision-makers .Design ~ FSR: sometimes modeling, mainly bio-economic criteria, technological interventions ~ ECOR: modeling, bio-economic, social and environmental criteria, technology and policy interventions Crop-animal research perspectives 7 ILRI's ecoregional coverage O Rainfall .0-400 ~ 401 -800 D 801 -1,200 O 1,201 -1,600 ~.>1,600 Caribbean Central America-,"-" Andean Ecoregion South America 8 Improving the contribution of livestock to crop-animal systems in rainfed areas in Southe8St Asia Table 1. Alternatives for cattle fattening based on liacchu (Myriophylion elatinoides) and Totora (Cirpus totora) forage in Carata, Puno, Peru (PISA, 1991). Weight Gain (g/head/day)Alternative Housing + fresh forage Housing + dry forage No housing + fresh forage 470 1050 323 Table 2. Technological improvements in pasture-animal management. Percent Increment over Traditional System Forage Fed Production Results Alfalfa + Dactylis + Barley 7.8 tons DM/ha/year 1.5 tons crude protein/ha/year 254 88 Barley (silage) 3.1-4.8 tons DM/ha 72 9 LWG/day (sheep) 380 180 202 Aquatic forage Llacchu and Totora Traditional Shelter + pre-dried Water lentil* 0.301 kg LWG/day 0.910 kg.LWG/day (cattle) 100 9 LWG/day (sheep) 280 Sweet potato (DLP3548 and Helena varieties)** Cutting frequency of 45 days (6 cuts) 90 days (3 cuts) 135 days (2 cuts) 85 Yield in 270 days: 15.5 t DM/ha 11.2 t DM/ha 6.8 t DM/ha 75% digestibility + income***Silage (30% Sweet potato + 70% maize) * Mixture of 25% water lentil and 75 % of barley grain and oat hays ** DLP3548 is a dual-purpose variety that produces tubers and vines for feed whereas Helena only produces foliage; planting density of these varieties was 50,000 plants/ha with fertilisation *** Increase in income through sale of green corn and sweet potato tubers at harvest 9Crop-animal research perspectives PROGRESS IN GLOBALISA TION Southeast Asia Improvement of crop-animal systems in rainfed area Umbrella project .Sustainable control of endoparasites .Economic importance of foot and mouth disease: OIE-ILRI .Improvement of feed resources: CSIRO-NARS-ILRI .Trade and animal productivity: USDA-ILRI South Asia Improvement of crop residue utilisation: ICRISA T -ILRI Impravement of crop-animal systems in mountain areas: ICIMOD- ILRI . . OUTLOOK . . . . . Ecoregional research in evolution Pragmatism: learning by doing it Need to strengthen efforts in this area ~ Need for impact ~ Increasing interests of stakeholders, environmentalists, governments and donors CGIAR system review CASREN: an opportunity to demonstrate impact REFERENCES Rabbinge R. 1995. Ecoregional approaches, why, what and how. In: Bourna I., Kuyvenhoven A., Bouman B.A.M, Luyten I.C. and Zandstra H.G. (eds). Ecoregional Approaches for Sustainable Land Use and Food Production. Kluwer Academic Publishers, Dordrecht, The Netherlands. pp.3-11. TAC (Technical Advisory Committee) -CGIAR (Consultative Group for Interna- tional Agricultural Research) 1992. Review of CGIAR Priorities and strate- gies. TAC Secretariat, Food and Agricuhure Organisation, Rome, Italy. Proyecto de Investigacion de Sistemas Agropecuarios Andinos (PISA). 1992. Informe Final (1985-1992). Proyecto colaborativo INIA-CIID. INIA, Puno, Peru, 417 pp. Steinfeld H., de Haan C. and Blackbum H. 1996. Livestock-environment interactions: issues and options. F AO and Commission of the European Communities. Brus- sels, Belgium. 56 pp. Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia10 LIVESTOCK TO PROMOTE RURAL INCOME DIVERSIFICA TION AND GROWTH IN VIETNAM : FINDING APPROPRIATE POLICY OPTIONS' M. A. Jabbar International Livestock Research Institute PO Box 5689, Addis Ababa, Ethiopia BACKGROUND The agricultural sector in Vietnam, particularly the rice sub-sector, has experienced remarkable growth during the last decade. However, there is mounting evidence that the gap between rural and urban income is increasing and the prospects for strong and sustainable growth of non-farm rural income are not very encouraging (Nguyen Van Bich et al. 1998). With 80% of the population living in rural areas, these trends risk raising social tensions. These trends and the recent economic crisis in Asia induced the government of Vietnam to recognise the key role of agriculture and rural economy for the promotion of industrialisation and modernisation and its contribution to increasing employment and income of the rural population (Phan Van Khai, 1998). The rural-urban income gap, the limited growth prospects in rice cultivation, and changing pattern of demand both in Vietnam and in world markets suggest the need for rural income diversification as a key element of a strategy of rural development. Such a strategy aims at achieving higher and more stable rural incomes, reducing the incentives for migration from rural to urban areas, promoting farming systems to be more sustainable in the long run, and alleviating rural poverty, especially among ethnic groups in mountainous and hilly areas. Within this context of rural income diversification, livestock development offers the opportunity of meeting several of the objectives of rural development. Rural income diversification, however, is a complex process that involves the entire rural economy and society (Timmer, 1993 ). It entails broadening the income sources ofrural households and their active participation in the decisions affecting their livelihood at the community level. The process involves not only new cropping patterns and animal breeds, but also new marketing and agro-food-based industrial activities, and rural institutions affecting credit and local organisations. Eventually, the process of structural transformation of agriculture will lead to the exit of a significant proportion of the rural work force from agriculture, though not necessarily from rural areas. Thus, rural income diversification encompasses agricultural diversification, rural industrialisation, and participation at the local level. Within this broad framework,~ I Derived extensively from the "Research proposal on policy options for using livestock to promote rural income diversification and growth in Vietnam," submitted by IFPRI and ILRI to the DANIDA. livestock to promote rural income diversification and growth in Vietnam; finding appropriate policy options 11 the particular role of livestock and livestock-based enterprises, and appropriate policy and technology options to exploit that role need to be identified. RATIONALE OF THE STUDY Livestock contributes about 20% of total agricultural gross domestic product (GDP). In contrast to other agricultural sub-sectors such as rice, sugar, and fertiliser , the livestock and meat processing sector does not seem to have immediately recognizable policy distortions. Both domestic and international trade in livestock products and inputs are generally not affected by interventions such as tariffs, quota, and price controls. At the same time, in spite of the generally acknowledged importance of this sector for agricultural and rural development, there has not been a clearly specified set of priorities and policies comparable to that of rice and sugar. The policies related to the development of the livestock sector in Vietnam do not seem to have matured beyond a generic indication of targets for production and meat consumption. What then is the rationale behind the proposal for a livestock sector study? The argument for livestock development rests on the following pillars. First, agricultural diversification requires shifting to higher value production per hectare and per unit labour than in the case of rice. Livestock production, particularly in the context of an agrarian structure characterised by very small farm size, offers the opportunity to capture higher value added. Second, the prospects for increasing domestic and international demand for livestock products, particularly pork and poultry, appear quite good (Delgado et al. 1999). Third, livestock products are processing-intensive and subject to economies of scale in processing. The development of the livestock sector may create a lot of productive employment in the rural areas, but it needs appropriate policy interventions to ensure that its benefits are broad-based and contribute to rural income growth and poverty reduction of smallholder farmers. The research interest in this sector is further motivated by the emerging evidence that the development of an efficient livestock and meat processing sector may accelerate agricultural growth in Vietnam, promote rural industrialisation, contribute to the development of those regions where most of the poor live, and exploit the potential for high-value exports (Goletti and Rich, 1998). These claims are based on limited knowledge on the sector, a situation highlighted by the lack of adequate database and updated information related to the production-consumption chain. Moreover, the improvement of the livestock and meat sector in Vietnam today is a challenge for a rural development strategy that contributes to broad-based growth, poverty alleviation, and the creation of an environment conducive to active private sector participation in the rural economy. The challenge consists in combining measures that have a public good character Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia12 (such as improvement in veterinary and sanitary standards, research and extension) with improvements in the implementation of trade policies and access to credit, thus facilitating the emergence of an efficient private sector and the improvement of the income and employment of poor rural households. OBJECTIVES The general objective of this study is to contribute to the formulation of a strategy for the development of the livestock sector as a key element of rural development in Vietnam. The specific objectives of the study are: I) To describe the structure of the livestock system including the key characteristics of the production and marketing system, the meat processing industry, the structure of trade, and the pattern of meat consumption; 2) To identify the incentives and constraints to the development of a modern livestock system, including investment in and adoption of new breeds, use ofhigh quality feed, availability and cost of veterinary services, upgrading of meat processing facilities, production of lean meat, and quality control; 3) To study the determinants of livestock supply and feed demand in the context of a diversified agricultural system; 4) To study the characteristics of meat demand in Vietnam, particularly in the context of urban areas; 5) To identify the constraints and the opportunities of meat exports and the comparative advantage of meat production in Vietnam vis a vis other countries in Southeast and East Asia; 6) To identify the barriers to entry of the poor and women into livestock production for the market and the policies and institutions that facilitate small and medium enterprises to expand their commercial operations; 7) To identify and ana lyse alternative policy options for the development of the livestock sector and their impact on income and employment in rural areas. 8) To enhance national capacity in policy analysis. METHODOLOGY The study uses a sub-sector approach where the integration of feed use, adoption of new breeds in livestock production, marketing channels, meat processing and exports, consumption patterns in rural and urban areas, and research and extension are analysed on the basis of extensive field survey data. The primary focus of the study is on pigs and poultry as these subsectors represent over 90% of total meat production, and they are present in all agro- ecological areas of the country. The cattle and buffalo sub sector will also be given livestock to promote rural income diversification and growth in Vietnam: finding appropriate policy options 13 attention, as it plays a major role in the farming system by providing draft power and manure, and demand for cattle meat is rising, particularly in the urban areas. Small ruminants are of minor importance in Vietnam, so they will not be given particular attention except in cases where they are raised with other species already included in the sample. The surveys have been conducted in different regions including the Red River Delta, the North Central Coast, the North East and South. After a careful assessment of the secondary data, the following survey sample plan was prepared and implemented: .Pigs and poultry producers -1,800 in 18 provinces and 72 communes; 60% pig producers, 40% poultry. Also, 40% are household enterprises and 60% business enterprises, both small and large .Cattle and buffaloes -400 producers in 10 provinces and 20 communes; small 30%, medium 50% and large 20% .Meat processors/slaughterhouses for poultry and pigs -162 in 9 provinces; 50% processors, 50% slaughterhouses; both small and large .Meat processors and slaughterhouses for cattle and buffaloes -30 in 10 provInces .Feed industry/mills -162 .Veterinary service at province, district and commune levels -180 samples total .Feed traders -total 162 including assemblers, wholesalers and retailers .Pigs and poultry traders -total 324 including assemblers, wholesalers and retailers .Cattle and buffalo traders -80 including wholesalers and retailers .Consumers in urban areas -Hanoi, 200 households, Ho Chi Minh City, 300 households, each classified as poor, medium, rich Econometric and modeling techniques will be used, in addition to descriptive statistics, for analysis. The econometric work will study barriers to entry of the poor and small and medium enterprises into commercial livestock production. The modeling work will involve a spatial equilibrium model where the building blocks ofdemand, supply, and feed use are estimated from econometric work and analysed in the context of crop-livestock linkages, linkages with the rest of the rural economy, different agro-ecological regions, and domestic and international trade. The model will be used to study the linkages between livestock and crop production and consumption, and the effect of alternative policies on prices, income, production, consumption, trade and food security with particular emphasis on the impact on smallholder farmers' income and the poor non-farming population in rural areas. The training and capacity building includes different components such as technical training in survey methods, statistical and economic analysis, and conceptualisation of research issues and policy options. It is being implemented through training courses and the active participation of local staff and research collaborators during the overall course of the study. 14 Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia WORK PLAN The study is being implemented in four phases over the course of 13 months. In phase I, an initial workshop with main stakeholders was held in J une, 1999 in Hanoi, where the objectives and the methodology of the study were discussed and the key policy and research issues were identified. This was followed by identification and training of the survey team, the design of the survey sample and survey instruments, and their pre-testing. In phase 2, data were collected and computerised over a period of five months. Staff of different departments of MARD and other partners (see list below) were seconded to the project to undertake the surveys and do computerisation work. In phase 3, data are currently being analysed and a draft report will be prepared. Training courses using collected data will be organised for national staff in different regions to train them in data management and policy analysis. In phase 4, the results of the study will be presented to a national workshop and to regional seminars. Based on comments and discussions at the workshop, the draft report will be revised and a fmal report will be submitted. EXPECTED RESUL TS The main result expected out of this project is a contribution to the formulation of a livestock development strategy as part of the rural development strategy based on solid research and effective capacity building. More specifically, the study expects to contribute to several aspects including: 1) Database of indicators on the livestock sector for use by the government; 2) Analysis of policy options with impact assessment on regional and distribution effects on income, production, consumption, prices, trade, food security, and poverty; 3) Conceptualisation of linkages between livestock development strategy and rural development strategy; 4) Capacity building in policy analysis; 5) Dissemination of results and contribution to policy fonnulation; and 6) Identification of priorities for livestock development policy. COLLABORATION The success of a study in Vietnam depends on the contribution of several key organisations involved in research, policy, and extension. The collaborative effort includes both national and international experts. By exploiting the comparative advantage of all of these organisations, the study aims at the successful achievement of its objectives. Livestock to promote rural income diversification and growth in Vietnam: finding appropriate policy options 15 At the international level, the study is being coordinated by IFPRI with contribution from ILRI. The Vietnamese organisations that are part of this collaborative effort include: 1) The Department of Agricultural Policy and Rural Development at the Ministry of Agriculture and Rural Development (MARD), 2) The Department of Extension also at MARD, 3) The Animal Husbandry Research Institute in Hanoi, 4) The Binh Thang Animal Husbandry Research and Development Center belonging to the Institute of Agricultural Science of South Vietnam, and 5) The Department of Veterinary, Veterinary Research Institute The project is being guided by a Steering Committee with representatives from the partners of the project. REFERENCES Delgado C. L., Rosegrant M.W., Steinfeld H., Ehui S. and Coubois C. 1999. The growing place of livestock products in world food in the twenty-first century. Markets and Structural Studies Division Discussion Paper No.28, International Food Policy Research Institute, Washington, D.C., U.S.A. Nguyen Van Bich, Nguyen Xuan Nguyen, Chu Tiem Quang, Tran Huu Quang. 1998. Differences in economic development in urban and rural areas. Viet Nam Socio-economic Development, No.13, Spring 1998 Goletti F. and Rich K. 1998. Policy simulations for agricultural diversification . Strengthening Capacity Buildingfor the Renewal of Rural Development in Viet Nam. Final Report submitted to UNDP under project VIEI9610081 A/ 01/99. International Food Policy Research Institute, Washington, D.C., U.S.A. Phan Van Khai. 1998. Socio-economic development in 1998 and major orientations for socio-economic development in 1999. Government's report presented by Prime Minister at the 4th Session of the lOth National Assembly on October 28th 1998, in Viet Nam 1998-1999. The Gioi Publishers, Hanoi, 1999, Viet Nam. Timrner P. 1993. Agricultural diversification in Asia: lessons from the 1980s and issues for the 1990s. In: Agricultural diversification in monsoon Asia. Report ofa study meeting 18-23 January, 1993. Tokyo, Japan: Asian Productivity Organisation . GIS APPLICA TIONS FOR SITE CHARACTERISA TION AND DEFINITION OF RECOMMENDATION DOMAINS L. A. Lapar International Livestock Research Institute Los Baiios, Laguna, Philippines INTRODUCTION The site-specific nature of technologies for crops and animals is largely a function of their reliance on natural inputs such as sunlight, soils, precipitation, etc. Moreover, given the context in which these technologies are designed and proposed for adoption at the plotJhousehold/farm level, macro-based interventions to micro-oriented problems have been shown to be ineffective. The fact that centrally planned, universal solutions to rural development problems do not work has long been recognised in agricultural research (Pingali et al. 1987, MacIntire et al. 1992). The site-specificity issue also implies the need for extrapolating such site- specific results to a larger recommendation domain and the attendant technicalities involved. This requires detailed, spatially disaggregated data on important variables for analysis. Specifically, for technology targeting in crop-animal systems, this suggests the need for disaggregated information on agro-ecological endowment, physical infrastructure, and socioeconomic factors at various levels of resolution, as well as the tools to analyse this type of information. Computer-based Geographic Information Systems (GIS) are recent tools that have become available for data acquisition, analysis, and display of data within some common spatial referencing system (Nix, 1987). GIS, together with systems analysis and simulations, now offer exciting opportunities to demarcate homologous agro-ecological zones at mega-, macro-, meso-, as well as micro-level depending on the objective and data availability. This has facilitated the rapid advancement of spatial analysis applications in research on natural resource management, land use planning, technology targeting, impact assessment, extrapolation of research results, and policy, among others . The use of GIS applications in crop-animal systems research is becoming widely accepted among researchers and development practitioners. Crop-animal systems research being systems-oriented lends itself well to GIS applications. This paper presents some GIS applications that have been utilised, as well as those that are considered worthwhile doing, in crop-animal systems research. 61S applications for site characterisation and definition of recommendation domains 17 GEOGRAPHIC INFORMA TION SYSTEMS This section presents a brief overview of the defmition, structure, components, and functions of the GIS. What is a GIS? A Geographic Inrormation System (GIS) is a geo-referenced database system that contains both feature (spectral, spatial, temporal) and thematic data, with emphasis on information extraction rather than on data storage. A GIS is a powerful set of tools for collecting, storing, retrieving, transforming, and displaying spatial data from the real world for a particular set of purpose. GIS is also a computer technology consisting of hardware and software that are used to produce, organise, and analyse information. The distinguishing characteristic of GIS from other information storage and retrieval systems is the use of formal location of features in coordinate space as the fundamental referencing principle and as important variables in quantitative analysis. GIS can ana lyse and explore any relationships between these variables. As a computer system, GIS is able to handle maps with legends. It is also useful in extracting information from a single map especially when the description of data accompanying the map is very complicated. One major function of a GIS is to present likely scenarios under different assumptions. Hence, it provides the necessary information to the concerned agents to respond accordingly to these scenarios based on biophysical criteria, socio- political considerations, and cost effectiveness. It can also compare different maps and is, thus, an ideal tool for integrating data from different sources . Structure and components of GIS GIS is basically computer-based, which is the main source of its awesome capability and accuracy in handling, synthesising, interpreting, and displaying the desired output. The graphic-interactive workstation plays the central role in a GIS. It serves to control, to correct, and to store the data. In addition, workstations are available for digitising, data processing, and data output. There are four major components that make up a GIS, namely, datal information, computer hardware (e.g., the machines), software (e.g., the computer application modules), and the users (people). Data are numerical representations or other symbolic surrogates that characterise attributes of people, organisations, objective, events, or concepts. It consists of numbers, letters, words, names, and other symbols that may be stored in a computer system. Improving the contribution of livestock to crop-an/inal systems in rainfed areas in Southeast Asia18 Information is data that have been processed into a form that is meaningful to the user and have real perceived value in current or prospective decisions. There are two types of data, namely, spatial and non-spatial. Spatial data is information that shows the geo-referenced point of component of the land or area and comprises both a measurement and/or description of an attribute or characteristic and the spatial location where such is applied. There are two forms of spatial data -vector data and raster data. Vector data represents spatial data by three main geographical entities such as a point (e.g., the position of a village), line or arc (e.g., road, river), and polygon (e.g., soil series, agro-ecological zone). Raster data, on the other hand, represents spatial data in the form of a set of cells located by coordinates, where each cell is independently addressed with the value of an attribute. Non-spatial data or attribute data are data in table form, showing the characteristics or properties of spatial data. It relates the qualitative and quantitative characteristics to the described object in so-called alphanumeric form. Up to now, this relationship could only be realised using vector graphics, but methods for relating raster data to non-spatial attribute data are now under development (Buhman, 1992). The non-spatial attribute data is stored in a data bank. The data bank can be an integral component of the GIS, but may also be connected to it by an external interface. The advantage of an integrated data bank is faster access, but an external data bank provides a better chance for connections with other applications. Examples of non-spatial data include soil database, crop database, and climatic database. The GIS hardware consists of four or five basic elements: .the central processing unit (CPU), representing the computer itself, which is linked to a disk drive storage unit, for storing data and program; .the input device (e.g., digitiser, video scanner to convert data from maps and documents into digital form and send them to the computer); .the output device (e.g., printer and plotter to reproduce the results of data processing); .the visual display unit (VDU), the monitor, or terminal of the computer for user interaction and for instantaneous viewing of spatial data and result of the map analysis; and .an external tape drive for storing data and programs and for communication with other data systems.1 The software includes the subsystems of application programs, as follows: .Data input and verification, covering all aspects of transforming data captured in the form of existing maps, field observations, and sensors (e.g., satellite data) into a compatible digital form; .Data storage and database management, i.e., the structure and organisation of data such as position, linkages (topology), and attributes of geographical elements and the way they must be handled in the 1 This is optional BIS applications for site characterisation and definition of recommendation domains 19 computer; .Data output and presentation, dealing with the manner by which the results of the analyses are displayed and reported to the users such as quality maps, tables, figures, and the ephemeral image displayed on the VDU; .Data transformation, dealing with such aspects as scale manipulation, fitting data to new projection systems, area-, parameter-, or distance- calculations. It involves both spatial and non-spatial data transformation. .Query input (interaction with users), which facilitates improved acceptability of GIS because of the development of user-friendly or menu-driven application programmes Users (people) are the major infrastructure needed for an efficient GIS. Skilled personnel and managers are required to operationalise the system. Together with the appropriate hardware and software, the overall organisation of resources influences to a great extent the maximum utilisation of this set of technology for handling complex database with extreme efficiency and accuracy. GIS processing The main purpose of a GIS is to manage spatial information for decision- making. The principle of GIS processing involves three major processes, namely, data input, data manipulation, and data output. A summary of the processes of mapping and analysis for GIS is shown in Figurel. "" f-. I Data Output I .. -.. . ~ Figure 1. The major processes of GIS 2 The more popular GIS software include ArcView, ArcInfo, ERDAS, IDRISIW, SPAN, MapInfo, and MICROBRIAN, to name a few. Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia20 Data input and data entry cover all aspects of transformation of data captured in the form of existing maps, field observations, and sensors into compatible digital form. For example, spatial data such as a soil map or other existing maps are inputted using a digitiser or a scanner, while non-spatial data like soil characteristics, crop characteristics, or socioeconomic data are inputted manually using the keyboard. Data storage and management concerns the way in which the data about positions, linkages (topology), and attributes of geographical element (point, line, and areas representing objects on the earth's surface) are structured and organised, both with respect to the way they must be handled in the computer and how they are perceived by the users of the system. Several operations commonly required in manipulating attribute data may be undertaken such as the addition of new data sets to the database, insertion of new data and retrieval of old ones from existing data sets, retrieval of data from existing data sets, updating and/or transformation of data from existing data sets, deletion of data from existing data sets, and removal of data sets from the database. The great advantage of digital storage of spatial data is the various possibilities for graphic presentation in various forms for interpretation and analysis. Illustrative graphic and alphanumeric information presentation is possible as a continuous information and data flow. The rapid updating enables instantaneous recording of changes in every desired detail in the data bank and allows obtaining an actualised map instantly. One of the most important functions of digital storage is the possibility to overlay or superimpose different levels. Different thematic maps can be combined and supplemented by any explaining non-spatial attribute data, thereby providing further opportunities for thematic analysis. Data manipulation and analysis involves the creation of composite variables through processing activities directed both on spatial an'"-.c ? (/J '- . ~:;'...JE~ "C ~ m "C-~.'.,-.E-E ., .~ "C ('1 '" ~ a. ~ .. ,".ca.~,"l:'~ O 2.c "0 ... --~ .,"O,g," ~m .,U'".,'"- , - ~~-~~.~,"m?,","~~3:~~~-= '" r~ 11 I .~ .(;:;1 ? fn~a.~-0.E~.c~~ ...fnCI: E (/) ~ -(VJJ",-"'~~.~ om a.~ ...: "' O~Nu:. (V , .- ~.!.IDa.~E"'-o"'CD"!:...,fnN..,.-~ .- E~ ..E ~(V!;'~.c .-.c ~..,fn~~ ;: c: ~ ",(VCDfn .., E E .'" E . mmC:CD~~.=~ CD Ilj '... r ,- .' "i)1 ~ .. , ..9'" . -~~a , .. ~ - I. ~," z~ .,.. , , .1 0- !~--.!- 0::~, !I)NwB~g- ~~ > i~E.:. E~~i c.~" ~-"g.c .:-tt.: " ~"..E.-~0~~ ~~~g.~ 8 ~ ~e~.co~ ~~ c ~ -c ~~~~=~~~~E.e.oe.oE~~~~8~~8~~IIIIII~ ~ B g, g, 'f. 'f. ~ ~ e e ~ ~ s; s; ~ ~ Q: ~ ..(/) ~ ~~i ! ~ oc~ II ? "";,, Oc-" 1:.1:.. oQ:1: ,~,.!! .::~" -"1-Oc.. .a } - e",E~.!! a e 1"l!i~ :!!ni "="e.:: E-" c..c~" ~&c .CGs; c~G "~"AE.!!.a.a"1: ~~~OC~~~ .=G...Ce S ~~~~~~~~~~.CG..o...o~~~~~U~~~Zi~lllli D, - , ... .,. .. .. ~ . ~"t>.a z~ , '~...'v.r.1 '(l 0.- !*--L Q:~ IIINW<{...Q)~-la....5~.;;cQ)"C...clac.e~~0)Q)-~~ir: 32 ... ., ., "' "' Q. Q. r. r. .., .., -E -E ~ ~ .c .c .0 .0 ~ ~ -., -., 11: 11: ~ Q)., (/)- E~., (/)Eoot "'", -11: 18~U -18t'1 Il! -a. J: (/) J: ~ -a. ~ ~ !0 ~ o ...ooto "' -J. -..> ~ -J. -'" ~ . ..,- "'> 0..,- ",o~ 0.., -Q.," -..,"'.., Q...,~ 0--.' U ~-Il)--.' --Q)., -~U o--".~~"' e~U 0 18 -~ C.11: > 18 ~ '" --18 oot -- c.., .Q?(/) C.., .Q c.., >..Ee", ~~->E>..Ee", ~E>..e", ~ ., >~--~ '"m~-- .,- c. ..,..'"-.,- c. ., c. ~'"..'o ..,- ",.'~.'..'o ..,"'~..,c ~ "O E--- E-E ., --"0 ., "O E-~ -E-.., ., - E-~ -~uc. -"0 - "0 ~ ~ ~ ~ .c -Q. c ~ ~ ~ ~ "0 ~ ~ ~ .., ~ ~ .c --'" .c -0 :: ~ -~ ~ .c --"' .c ~ ~ ~ -- ~ ~ .Q E -~ .Q 0 ~ .Q .., ~ "0 .Q E ~ .Q ~ ~ .a E ~ -~ ~ ~ c. ~ 0 .:0 ~ ~ ~ -~ ~ ~ c. ~ -~ ~ ~ ~ 1:~.'.Ce.'~~"'~1:'""'~e"""1:"'.C~~EEE-EE'"EE~EEE-EE'"EE~-5 ~ ~ ~"'S ~ ~"'S ~ ~-5 ~ ~ ~"'S ~ ~-5 ~ ~ ?~~~~~~~~~?~~~~~~?~~~jllllilllj~llllljl D-\-". ... .r.~ .. '~~I 9' ~ . ;. . ~~...J. -.~1 a , ...\: ~., z~) ~,., Q-~.' '~ --1}, ~ 0.-;;;:(/) 5;::!5--1- Q:Q; \I)Nw?...Q)a-la....5>....UicQ)"C...clac.E~';cozo.-Q)-~ai.i: 1'---~ (.\~ ""' ,-,..'- "... , Guc:GLoG ~NC?)I:00 CD ID -- ~oqoN"O N :N.q:GG CD > oqo I I I O .N~N~.gCDooCDaloqo~oqoG O .g oo :Nzz ~I.IIID....' -. r\ .I~t~~ ~,\.:*~ .* .. ..~ ~~i~J'i (( , -..~, ~~ f,t1 ~- "' .-.v., ( ~} z~& ... A1 (-.~ ---.r -.. .~ r .. ,! , 'r; Q.~'.;>Lx:!! >....;:cQ)"C...c(Qc.E~-"Cc(Q>....(j(QQ.(QCJac.>--(QCJcQ)Q)~...Q).cQ)CJcQ)-Q)=:s>....(j(QQ.(QCJac.>--(QCJ...Q)z.-.-Q)-~aii: (TJ \D ~ \IJ-E \IJ-E \D-E (U~(/) \D.C(/)\D\D~(/) .L'C-~-C~~'C:,(n?'~",,0~,(/)O(/)C~S(/)C~-(nC~c: ? Q) .-C O (I) C C O Q) \DO~a.?~~a.:30~a.~~(1)~ ,-Q)~""(U~Q)~ ~Q)a.o:t""(UQ)C.q- Q)C.q- .-c. N q- E a. N q- ~ c. N q- ? q- 0 ...: .-c: q- 0 ...: I:; q- 0 ...: (V)ro... (V)ro...~(V)ro... -r--It)~r--1t) r--1t) \D ~r--~r--r \D E I ~..-1~~..-' N i I N (J) i I N - .-0 It) q- 0 .It) q- 0 ~ ~ ro q-(V)ror--","(V)roO "'C r-.: Or--r-.:~Or--r-.:O i ~ ..-Oo:t.-It)Oq Z ~ .- []II ~[] []: ::.:: D G) (V) N :,: ~~:: -J O ~ ..::: ,...: ....1/ . " .9-. ";: , ... ., 11 ({ : I;?;.wf ~~ ~. '- ~, a , ~ ~. z~"..:7 Q ..'. ~~...~ .vb ( -.I 0-:;::(I) 5(3-.J. ~~(j> , >...u;cQ)"0co.~la"SQ.OQ.claE~.cco>...u;cQ)"0iiE.cla-0>la~Q)>ON.-Q)~~~iL ~ ,(1) -33-L cr: ~ z< ~ .2~ "'i E .2IV "'i... 0... 0... InQ) UQ) Q. In ~ U O... InQ) ~ 0 Q)- IV~ In Q) > .~IV "ii ~ P) .- Q)- ~ = ii: 61S applications for site cflaracterisation and definition of recommendation domains 35 -vc.a.-J .iu~ .c.?o:cu ...',..'....';.:.;,:.,.Qii:.~.~oO~~.oiit~millJ III"0O.~Q)Q.O~...-OCOIII.~IVQ.EouIVIIIQ).(jQ)Q.III"iE.cIV-OCO~"U)OQ.Eou"0iIIIQ)..IV.cIIIQ)>~IV"i)~~.-Q)..~CI ii: .~ . . - c: . UJ c: -.- c: E ~ :J c: '- -I E c: :J O 'C It: Z ~ID - , .' .~. ~ ~ I. z . G- , (/) -- ~:[ a:: 9 , C/) ... c IU C .E ~ '=" c o c ~ c IU C/) ... C IU C .E ~ ~ C/) ... .c ~ i& E .c IU i& ... 0 ... - O 41 ~ IU .c C/) 41 > .~ IU "ii a: Ln .- Q) ~ ~ = ii: 0- ~ tU t:~o fII~ tU""Q.0..tUfII54)u ~ > -.w ~ tU~~- 4)t:tU::I/It:.Q~u...at:>l/ItU>..-O :J .-~ tU ~ O :J -g ~uU)0..U)u0..U)~ 4)a~-J .~ ..~10~r-(V)OO \ .00 " O O Gi'c ~ -."Q.IU - '5 O Q.4) -? Q. C IU,gIU IU Q. ..a> - CCIU 4) Q. 4) .-:J rn 4) 4) Q.Ern ~ Gi' ~ ~ rn C rn IU -5:JGi'C4) u G) .v C --C ..."Q. ~IU:O;IUc: ~ Q. 1U.. Q. u,g~m m a L-alU4) :J G) :J -IU C aD rn C rnma~ >- -.-> :J :J O C a. c-mrnErnrn L- -.III - 0 C -111 --C... U L-C8mE;L-~ ~ o~ uO4)8c: G) u~ 4)- -u,g m u -,g u 0 ->- -4) .L: ~ >- 4) 0 4) Q. u o ...~ u .-,-urni:-- ~ L-111 Oi: -- -c: I -rnQ.,- '> .~m-GI :J E m - .c~.!!:Jt:-"Ct: m '. U,grnc:"Q.:t:mm! . ~ >-Gl~ :>1U~C.v . .COC~ Q. .e.~== Q.GU rno =rn.v --' :J "C .c .c >~~~. ~ .5 D.. f- ~,(J; a - ~.I .: .~(0 z~ "\~~.-0 Q.j .~-.~., ..1 .Jr (1., -U)5<9"--1- 0:~1 '(l cnQ)~"0.UjQ)-Q.O-uocnQ)Q.~...cQ)-Q):t::sOcnQ)-ca~cnQ)>.;;ca"'ii~CD.-Q)-~C)~ ~~ ~ 0 I()~ Definition of recommendation domains GIS analysis is also used to delineate recommendation domains for research and possible interventions. For this particular study, recommendation domains (RDs) were defmed based on a very general set of criteria. This set of criteria included the following: target AEZs of humid/subhumid zones and high animal and human population densities. The rationale for the choice of this particular set of criteria is to identify areas with high concentrations of animal and human popu- lation because of the underlying notion that the potential impact of interventions will be relatively large in these areas. 4 The process of defming the recommendation domains given the scale of data available in the existing database involved overlaying the following maps: AEZ, administrative boundaries, animal density, and human population density (Figure 17). Further refinement was undertaken by disaggregating animal density into ruminant density and non-ruminant density. The output of the GIS analysis is a list of provinces that satisfy the criteria. Specifically, two sets of recommendation domains are initially identified through this preliminary analysis using GIS (Figures 18 and 19), namely RDI and RD2. RD I comprises the areas with high animal densities, while RD2 consists of areas with high ruminant densities. These two separate RDs indicate where specific interventions of any kind on animals and/or ruminants are likely to generate the most impact. 5 A caveat is that the RDs will most likely be different once addi- tional criteria are included in the analysis. For example, if specific interventions will be introduced, then a more stringent set of criteria will need to be used to accommodate the more specific requirements of the analysis. It is proposed that this will be done using information obtained from the BMS surveys and once the results of constraints and problem diagnosis have been finalised. Targeting of technology and policy interventions The information available in the GIS can be used to formulate more location- specific recommendations on interventions in contrast with very macro-oriented recommendations that cover very broad and clearly heterogeneous areas resulting in ineffective, low-impact initiatives. For example, an improved characterisation of the BMS afforded by the GIS analysis would allow better targeting of appropri- ate interventions, e.g., feeding strategies and nutrient flow management, in the BMS, and that could also be extrapolated to areas with similar characteristics and 4 Once defInite interventions are identified based on the results of the BMS surveys, more specific criteria will be imposed to delineate and refme the defInition of recommendation domains. s This is largely driven by the underlying assumption that large numbers (of animal and human population) will generate more impact. 61S applications for site characterisation and definition of recommendation domains 39 hence generate a much larger impact. Examples of the use of this GIS application include the studies by Karim (1992) on technology targeting in Bangladesh and the study by Godilano and Agustin (1990) on extrapolation of agricultural tech- nologies in Thailand. While this particular application has not yet been done for the ongoing crop- animal systems research, it is proposed that this be undertaken using the informa- tion from the BMS household/farm survey and be complemented by the results of the rapid rural appraisal activities for problem and constraints diagnosis. .Human Population Density ~ Animal Density .Administrative Boundaries .AEZ ~ Recommendation Domains 17. Delineation of recommendation domains. 40 Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia Impact assessment of technology and policy interventions As GIS facilitates targeting of interventions, it can also be used to assess the impact of such interventions. Specifically, the GIS analysis that is appropriate for this kind of research is the what if.. ? application type. Such an analysis is used to assess the potential impact of changes in, say, land use type characteristics be- cause of the adoption of certain technological interventions, or the potential im- pact of an increase in animal productivity arising from the use of an alternative feeding strategy and/or improved management. Differences between the potential productivity estimates using the baseline information and the predicted data can provide measures of impact. An example is the study of the impact of genetic enhancement of sorghum and millet residues on the productivity of ruminants by Kristjanson and Zerbini (1999). Of particular importance are not just changes in the production levels but also in the spatial distribution of production, both of which are obtainable from the GIS analysis. This is an application that is worth pursuing for the present study and utilizing the information that is obtained from the household survey in the BMS and comple- mented by the existing information available in the geo-referenced database that has been developed. Figure 18. Recommendation domain 1 (includes 32 provincesl. 61S applications for site characterisation and definition of recommendation domains 41 Figure 19. Recommendation domain 2 (includes 49 provincesl. THE NEXT STEPS Thematic analysis will be continued using BMS survey data. This will comple- ment the thematic analysis that has already been done at the administrative level, i.e., province level, across the five countries under study. Criteria to be used to define recommendation domains for specific technology in- terventions will be drawn up once problem diagnosis (from RRA and BMS survey) has revealed what the major constraints are. This activity will be followed by a second round of map overlays (using BMS survey data) to define the recommendation do- mains based on this new set of criteria to generate technology-specific RDs . Options to pursue technology targeting and impact assessment will need to be dis- cussed. 42 Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia REFERENCES Buhmann S. 1992. Geographic information system. In: Bossel, H. and Bruenig, E. (eds), Natural Resource Systems Analysis: Ecological and Socioeconomic Sys- tems and Sensitivity Analysis for the Conservation and Management of Forest Ecosystems and Natural Resources in Southeast Asia. German Foundation for International Development, Feldafing, Germany. pp. 50-57. Godilano S. and Agustin P. Jr. 1990. Geographic information systems: potential tool for the extrapolation of agricultural technologies. In: Ecosystems Analysis for Extrapolation of Agricultural Technologies: Proceedings of the Planning Work- shop, Cagayan. Philippines. 22-25 May 1990. IRRI (International Rice Research Institute), Manila, Philippines. pp. 96-119. Harder I. 1991. Back to office report -Regional Workshop on AEZ Methodology and Applications. APO Misc. Document. FAO (Food and Agriculture Organi- sation), Rome. Karim Z. 1991. A document on cropping systems-based fertiliser recommendations by AEZ in Bangladesh. Bangladesh Agricultural Research Center, Dhaka, Bangla- desh. Kristjanson P. M. and Zerbini, E. 1999. Genetic enhancement of sorghum and millet residues fed to ruminants: an ex ante assessment of returns to research. ILRI Impact Assessment Series 3. ILRI (International Livestock Research Institute), Nairobi, Kenya. 52 pp. MacIntire J., Bourzat, D., and Pingali, P. 1992. Crop-livestock interactions in Sub- Saharan Africa. World Bank, Washington, D. C. Nix H. 1987. The role of crop modeling, minimum data sets, and geographical infor- mation systems in the transfer of agricultural technology. In: Bunting, A. H. (ed), Agricultural Environments: Characterization, Classification, and Mapping. CAB International, U. K. pp. 113-117. Pingali P ., Bigot, y ., and Binswanger, H. P. 1987. Agricultural mechanization and the evolution of farming systems in Africa. Johns Hopkins University Press, Baltimore. Pongnak W. 1995. The application of geographic information system for potential productivity evaluation of lowland rice area in Chachoengsao province, Thai- land. Ph. D. Thesis, University of the Philippines at Los Baiios, Laguna, Philippines. Wood S. 1994. AEZ/GIS applications. In: AEZ in Asia: Proceedings of the Regional Workshop on Agroecological Zones, Methodology and Applications. F AO (Food and Agriculture Organisation), Rome, Italy. pp. 203-227.~ 43GIS applications for site characterisation and definition of recommendation domains IMPROVED FEED PRODUCTION AND UTILISA TION TO ADDRESS YEAR-ROUND FEEDING SYSTEMS: GUIDELINES c. Devendra and D. Pezo International Livestock Research Institute, Los Bafios, Laguna, Philippines The following are some ideas and suggestions to initiate research to address year-round feeding systems, which is one of the major objectives of the Crop- Animal Systems Research Network (CASREN) project: 1 Participatory approaches to understand 1.1 Cropping systems and calendar months 1.2 Types of feeds produced and used 1.3 Types of feeding systems 1.4 Traditional systems (rationale, use, practices, etc.) 1.5 Importance of feeds and their benefits 1.6 Critical feed and water shortages 1.7 C(J)nservation practices 1.8 Supplementation if any 1.9 Medicinal feed plants used if any 1.10 Farmers' perceptions on how to solve nutritional shortages 1.11 Role of men and women This section provides a background on prevailing circumstances at the farm level. It involves the totality of available livestock, both ruminants and non- ruminants. The analysis and assessment need to be made through PRA efforts, including group discussions, going further from the household survey in the characterisation exercise, focussing specially on feed resources. The group discussions can be held at least at three levels: (1) Key informants (government planners, cooperative chiefs, farmers' leaders, NGO representatives), (2) Farmers groups (heterogeneous mix of men and women), and (3) EMS officials and extension personnel (Province/District/Village levels). About three initial meetings are involved. The discussions with groups of farmers are especially important and serve the following purposes: (1) fully understand the activities, (2) identify individuals for subsequent in-depth interviews and monitoring, (3) identify men and women farmers for separate interviews, and (4) enable the selection of farmers who are representative of the EMS, for further work. Key elements in the selection of the individuals to be involved in the Farmers Focus Group discussions include the following: (1) willingness to participate, (2) 44 Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia enterprising, with animal ownership, (3) responsiveness to innovations and improvements, (4) have a mix of animals (ruminants and non-ruminants), and (5) have some knowledge and understanding of feeds and farming systems. The separate group discussions and interviews with men and women (and children) will provide information on responsibilities, time spent, choice of animal species, decision-making, etc. Gender issues are involved, and can provide more insights especially in the link between women and children and small animals. More importantly, this will provide valuable information on the extent of socio- economic contributions of animals, as well as their impacts on the stability of farm households. Critical feed shortages should address, when these occur, the severity, implications, mitigating practices and mechanisms to cope it. For example, is this the time when supplements (farm produced or purchased) are put to best use, and with what effect? Also, conservation types and practices, as well as the best use of conserved feeds need to be understood. Are supplementary energy or protein or both sources used, and why? Likewise, critical water shortages, their implications and coping mechanisms need to be understood. Very little is known about medicinal plants used as feeds for one or more reason, as well as for medicinal reasons, e.g., use as anthelminthics. Since they are part of traditional feeding systems, a search for more knowledge that includes diversity, use, nutritive value and active ingredients will be very useful. Finally, farmers' perceptions on how to deal with, and solve feed shortages also needs better understanding in order to improve the situation with appropriate interventions. 2 Quantitative and qualitative issues in feed availability and use throughout the year 2.1 Types available (from cropping systems and other sources) 2.2 Methods for estimating availability of forages (grasses, shrubs and tree leaves), crop residue, agro-industrial by-products (AIBP), and non- conventional feed resources (NCFR) 2.3 Distribution throughout the year (on monthly basis) 2.4 Assessment of nutritive value where necessary (i.e., no data exists) 2.5 Interventions/opportunities 2.6 Protocols for number of farms involved The methods for assessing the availability of forages is given in Appendix I, and crop residues, AIBP and NCFR are given in Appendix II. In using these, field assessments will be necessary for individual feeds in applying the methods. These will serve the purpose ofverifying individual coefficients used as well as improving on them, or providing new figures for feeds where these are not available. The quantitative measures have to be done throughout the year. Assessment of nutritive Improved feed production and utilisation to address year-round feeding systems: guidelines 45 value needs to be undertaken only where there is no information on particular feeds. In general, tables of feed composition exist in most countries, which are adequate for most practical purposes. It is suggested that between 4-6 farms be used in each benchmark site (BMS), for the detailed assessment of the year-round situation. The farms should be representative of the BMS, should be broadly similar, and involve the more knowledgeable farmers who were identified earlier. 3. Adequacy and deficits (Availability of nutrients vs. Requirements by species) 3.1 Methods for assessment 3.2 Use of secondary data and information to assess different models to predict performance 3.3 Repeatability (two years of research is proposed) There are two parts to this. One is a macro picture of nutritional adequacy and deficits at the BMS level, based on theoretical calculations using secondary information, as well as the one derived from the household survey. The data to be used include the area under crops, animal populations, animal and crop production data, feed resources available, sold and purchased. Tables of nutrient requirements for individual species in developing countries, such as the one prepared by Kearl ( 1982) will be very helpful for this purpose. This will give an initial picture of the situation. The second part of the work is a more detailed year-round verification of actual nutrient availability in each of the 4- 6 farms, for an in-depth assessment of the earlier findings. Calculations will have to be done for each of these farms. To get a realistic picture, it is suggested that this work be done for two years. The overall efforts will also contribute to the development of models that can be further adapted. 4. Intake and grazing studies (On-station) 4.1 Methods e.g., markers 4.2 Radioisotopes (feeds/nutrient transfers) The methodologies for these are found in Appendix III. It is an important section, which relates to extensive grazing and use of stubble. Also, it will provide valuable information on nutrient transfers and recycling, use, and effects. In 4.2, much more basic work is required and will need backstopping, possibly with other partners, such as the IAEA. 46 Improving the contnoution of livestock to crop-animal systems in rainfed areas In Southeast Asia 5 Feed production opportunities (examples) 5.1 Introduction of legumes into cropping systems e.g., siratro, mungbean 5.2 Multipurpose trees e.g., leucaena, gliricidia 5.3 Treatment of straw/other crop residues 5.4 Cassava/ sweet potato production 5.5 Multi-nutrient blocks These are tentative examples of possible interventions that are potentially useful. There could be others. Individual interventions will need to be assessed and then tested to identify those that are more appropriate for individual EMS. The impact of these interventions will need to be studied in depth. 6. Definition of strategies 6.1 6.2 6.3 6.4 6.5 Understanding of adequacy Strategic supplementation Sustainable production systems Optimum productivity Comparisons across countries/regional strategies The definition of specific strategies, diffusion and recommendations will be the final output of the on-farm activities. It is hoped that we can do these in about three years of concerted research, or even earlier . REFERENCES * Oevendra C. 1976. The utilization of agro-industrial by-products in Asia and the Far East. In: New Feed Resources. FAO Animal Production and Health Paper No.4. pp. 111-124. Oevendra C. and Raghavan G. V. 1978. Agricultural by-products in South East Asia: availability, utilization and potential value. World Review of Animal Production 14: 11-28. Kearl L.C. 1982. Nutrient requirements of ruminants in developing countries. International Feedstuffs Institute. Utah Agric. Exp. Sta. Utah State University . Logan, Utah, U.S.A. 381 pp. *Identified in the text and in Appendices I, II and III. Improved feed production and utilisation to address year-round feedinq systems: Ilulllelines 47 Lascano C.E. 1992. Methodology for measurement of consumption under grazing conditions. In: Ruminant Nutrition Research: Methodological Guidelines (Eds. Ruiz M.E. and Ruiz S.E.). Inter-American Institute for Cooperation in Agriculture, Latin American Network for Animal Production Systems Research. San Jose, Costa Rica. pp. 163-172. Lascano C.E., Borel R., Quiroz R., Zorrilla J, Chaves C. and Wernli C. 1992. Recommendations on the methodology for measuring consumption and in vivo digestibility. In: Ruminant Nutrition Research: Methodological Guidelines (Eds. Ruiz M. E. and Ruiz S. E.). Inter-American Institute for Cooperation in Agriculture, Latin American Network for Animal Production Systems Research. San Jose, Costa Rica. pp. 173-182. Mendoza P. and LascanoC. 1985. Mediciones en la pastura en ensayos de pastoreo. In: Evaluacion de Pasturas con Animales: Alternativas Metodologicas (Eds. Lascano C. and Pizarro E.). Memorias de una Reunion de Trabajo, Lima, Peru, 1-5 Octubre, 1984. CIA T (Centro Intemacional de Agricultura Tropical), Cali, Colombia. pp. 143-165. Pezo D., Vohnout K, Leon-Velarde C. and Camargo A. 1977. Prediction ofpasture intake based on chemical fractions and digestion parameters. Turrialba (Costa Rica) 27: 157-162. Romero F., Montenegro J., Chana C., Pezo D. and Borel R. 1993. Cercas vivas y bancos de protein a de Erythrina berteroana manejados para la produccion de biomasa comestiible en el tropico humedo de Costa Rica. In: Erythrina in the New and Old Worlds (Eds. Westley S B and Powell M H). Nitrogen Fixing Tree Research Reports. Special Issue 1993. Nitrogen Fixing Tree Association, Paia, Hawaii. pp.205-210. Shaw N .H., 't Mannetje L., Jones R.M. and Jones R.J. 1976. Pasture measurements. In: Tropical Pasture Research: Principles and Methods (Eds. Shaw N H and Bryan W W). Commonwealth Bureau of Pastures and Field Crops. Bulletin No.51. Commonwealth Agricultural Bureaux, Hurley, England. pp. 235-250. 't Mannetje L. 1978. Measuring quantity of grassland vegetation. In: Measurement of Grassland Vegetation and Animal Production (Ed. 't Mannetje L). Commonwealth Bureau of Pastures and Field Crops. Bulletin No.52. Commonwealth Agricultural Bureau, Hurley, England. pp.63-90. 48 Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia APPENDIX I METHODS FOR ESTIMATING THE AVAILABILITY OF FEEDS 1 Estimation of the Availability of Grasses, Herbaceous Legumes and Crop Stubbles The methods to be applied for the estimation of forage availability will be defined based on practical, statistical and cost considerations (Mendoza and Lascano, 1985), but the appropriateness of the technique to be applied is a function of: .Type of forages present, i.e., in terms of species composition, uniformity, and density .Method of utilisation .Size and shape of the pasture/cropland area .Precision required for the estimates .Availability of labour, equipment, as well as facilities for processing samples and data The methods used for the estimation of forage availability can be classified as destructive and non-destructive. Destructive methods Destructive methods are those in which dry matter forage yield is determined by cutting/harvesting part or the total forage area, weighing and then samples are taken for dry matter determination (Shaw, et al., 1976). (i) Short Grasses and Herbaceous legumes For short, prostrate and decumbent species (e.g., Brachiaria, Axonopus, Calopogonium, Arachis ), growing in waysides/roadsides or under tree crops, yield is estimated based on samples taken using quadrats, which vary in size between 0.25-0.5 m2. Hand shears, machete, or even hedge trimmers are used to harvest the forage contained within the quadrats, but the important point is to maintain a uniform height of cutting, being better as close to the ground level as possible, to reduce errors due to uneven cutting heights. The number of samples required varies depending on the size and heterogeneity of the plot under evaluation. In general, for heterogeneous pastures ( a very common feature in pastures selectively defoliated by animals, specially in continuous or long-occupation grazing) it is preferable to use smaller quadrats and taking as many samples as possible, rather than using large quadrats with few samples. Improved feed production and utilisation to address year-round feeding systems: guidelines 49 (ii) Fodder Species For fodder grasses (e.g., Napier, Guinea grass, Paspalum atratum, Setaria) grown in borderlines, or even in a compact pasture but planted in rows, which do not spread covering the space in between, it is possible to use linear sampling units (e.g., a 0.50 or 1.0m section of the row). The same could apply to crop stubble plots. However, if data has to be expressed on an area basis (e.g., kg DM/ m2), special attention should be put to estimate the area of the piece of land supporting the forage sampled. For example, if the distance between rows is 0.5m, the yield obtained from harvesting a 0.5m long row section corresponds to 0.25 m2 (0.5 x 0.5m), since we assume that a portion of land 0.25m wide from each side of the row is supporting the plants harvested. Regardless of the harvesting devices and procedures used, samples must be taken at predetermined random positions, allocating sufficient positions to ensure that no site is sampled more often than once a year, in order to minimise cumulative effects of sampling on subsequent growth (Shaw et al., 1976). Non-destructive methods Non-destructive techniques are based on the estimation of some attributes of the vegetation that can be used to predict forage yield (Shaw et al., 1976). Although these methods are not highly accurate, the number of measurements that could be made helps to improve the precision of yield estimates (Mendoza and Lascano, 1985). There are three main types of non-destructive methods: .Visual estimates .Height and density measurements .Capacitance measurements (i) Visual estimates Visual estimates range from direct estimates of the yield in sample quadrats (e.g., g/quadrat, kg/ha), to the double sampling techniques in which visual estimates are complemented with the yield measurements obtained after cutting some quadrats, and used to give a calibration for the observers ('t Mannetje, 1978). The comparative yield method of Haydock and Shaw (1975) is the most commonly used double sampling method for estimating forage availability in heterogeneous pastures. In this, a set of fixed quadrats representing a scale of yields is first chosen (% points is the most common number of levels), and then random quadrats are rated in relation to the scale. Once all visual samples have been evaluated, the quadrats used to define the scale are harvested, dried and weighted. A calibration curve (linear regression) is established with the harvested quadrats, and the average of all visual estimates is used to predict yield, based on the 50 Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia calibration curve. The basis of the method is that it is easier to estimate the yield of a sample as being at some point on a visual scale, than to estimate the actual weight. In fact, in a study made with experienced observers, their yield estimates ranged between 4 to 8% of the actual mean derived from cutting all quadrats (Shaw et al., 1976). It is suggested that this method is particularly appropriate for the measurement of forage availability in grazing situations and crop stubble (e.g., rice, maize). (ii) Height and density measurements Height and density measurements have been used as indirect estimators of forage availability. Again, this involves some cutting to develop calibration curves, which are "site specific," as they vary greatly with species, seasons and other factors. Good precision of estimates is obtained only when more homogeneous pastures are evaluated. (iii) Capacitance measurements The capacitance methods measure the amount of water available in forage tissues, and not dry matter yield, but the latter is estimated through a calibration equation. Unfortunately these equations tend to be not only site specific, but errors arise from environmental factors (e.g., dew, soil moisture), as well as from the variation in water content within green vegetation and between these and dead material (litter) present in the pasture (Mendoza and Lascano, 1985). 2. Measuring foliage availability from shrubs and trees In the case of shrubs and trees, a large proportion of the biomass is non- edible; especially when these are pruned occasionally, but non-edible fractions could be valuable to farmers for example as fuel. From a nutritional standpoint it is useful to know in individual situations, what proportion of the biomass is useful for the animals. There is no convention on what is considered edible and non-edible, thus a final decision on it should be made based on field observations. In the case of Gliricidia, Erythrina, Calliandra and other tree species, leaf and green stems are considered edible (Romero et al, 1993 ), whereas in Leucaena leucocephala this is represented by leaf plus stems less than 6 mm in diameter (Shaw et al. , 1976). In the case of trees and shrubs, it is common to use a single plant as the smallest sampling unit, and estimate total availability based on the density of trees (in sparse or compact arrangements) or the number per unit of longitude (in the case of fence lines). When trees are frequently pruned, it is possible to estimate availability through harvesting, weighing and sampling the total biomass of a given Improved feed production and utilisation to address year-round feeding systems; guidelines 51 number of trees, but this becomes non-practical in the case of trees pruned occasionally. Under those conditions it is recommended to count the number of branches, and sample few of them after stratifying the whole plant in at least three portions, i.e., basal, medium and top. 52 Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia APPENDIX II CROP RESIDUES, AGRO-INDUSTRIAL BY -PRODUCTS AND NON-CONVENTIONAL FEED RESOURCES I. Major By-Product Feeds from Trees and Field Crops, with Approximate Extraction Rates in Asia and the Pacific Approximate Extraction Rate (%) Crop By-product Feed Tree crops Cocoa (Theobroma cacao) 5-10 70 35-40 2 12 2 55-60 55 Coconut ( Cocus nucifera L. ) Oil palm (Elaeis guineensis) Cocoa bean waste Cocoa pod husks Coconut meal Oil palm sludge (dry)* Palm press fibre Palm kernel cake Rubber seed meal Sago refuse Rubber (Hevea brasiliensis) Sago (Metroxylon sago ) 2. Field Crops Castor (Ricinus communis L.) Coffee (Coffea arabica) Cotton (Gossypium spp.) Maize (Zea mays) 45-50 70 40-45 8-10 16-18 10 4-5 10 15-17 100** Castor meal Coffee hulls and coffee pulp Cotton seed meal Maize bran Maize germ meal Maize stover Broken rice Rice bran Rice husk Rice straw Rice ( Oryza sativa) Sugar cane (Saccharum officinarum ) Bagasse Green tops Molasses 12- 15 15- 20 3-4 55 -59 10 100** Tapioca waste Wheat bran Wheat straw Cassava (Manihot esculenta Crantz) Wheat (Triticum aestivum L.) * Now referred to as palm oil mill effluent (POME) ** Implies equivalent weight to the yield of grains Source: Devendra (1976) Improved feed production and utilisation to address year-round feeding systems: guidelines 53 II. Minor By-Product Feeds from Various Sources, with Approximate Extraction Rates in Asia and the Pacific Crop By-product Feed Approximate Extraction Rate (%) Cassava leaves 6-8 70 -73 57 53 -57 70 -80 55 -58 55 -60 40 -42 67 -70 15- 40 50 -55 60- 62 45 -50 60- 62 60- 80 40 -42 35 -40 60 Cassava (Mahinot esculenta Crantz) Dhupa (Veteria indica) Groundnut (Arachis hypogaea) 70- 75 24- 35 30 -35 60-65 Guar ( Cyampsis psonaloides DC) Kakan (Salvadoza oleoides) Karaj (Pohogomia pinnata) Kakum (Garcinia indicachois) Kusum (Schleichara oleosa) Mahura (Madhuka indica) Mango (Mangifera indica) Nahor (Mesuaferrcalinnn.) Neem (Azadirachta indica) Oak (Obercus dilatata) Pineapple (Annonas comosus) Pisa (Actinedaphne hooberi) Sal (Shorea robusta-Gaerth) Sesame (Sesammum indicum L.) Soyabean ( Glycine soya) Sweet potatoes (Ipomoea batatas) Tamarind (Tamarindus indica) Dhupa meal Groundnut vines (stems + leaves) Groundnut meal Guar meal Kakan meal Kajah meal Kakum meal Kusum meal Mahura meal Mango kernel Nahor meal Neem meal Oak meal Pineapple waste Pisa meal Sal seed meal Sesame cake Soybean Sweet potato vines (stems + leaves) Tamarind seed hulls Tamarind seed kernels Source: Devendra and Raghavan (1978) III. Straw Grain Ratios 1:1. 1:2 1. Millet, rice and wheat 2. Maize and sorghum These are average values. The ratios are likely to vary due to crop variety , agronomic manage- ment and location, and will therefore need to be verified. For example in India, grain:straw ratios range from 1:1.3 to 1:3.0. 54 Improving the contnDution of livestock to crop-animal systems in rainfed areas in Southeast Asia APPENDIX III ESTIMATION OF INTAKE OF FORAGES 1 Pen fed In cut and carry systems, forage harvesting and distribution to animals is part of management. Therefore, weighing forages as fed, and taking samples to determine dry matter content are the only additional steps needed to estimate the amount of forage on offer, and the same can be applied to the refusals. Accordingly, forage consumption and efficiency of utilisation, can be estimated based on the forage on offer and the residues left by animal, based on the following formulas: Intake, kg DM = DM offered, kg -Residual DM, kg (1) (2) However, as the number of animals and weight change along the year, these data need to be recorded and used to express intake on a common weight basis ( e.g., kg DM as% BW). If that is the case, the intake obtained through equation (1) could be adjusted using the following formula: (3) ~BW. I where: Intake (kg DM) = Total intake by the group ofanimals, as estimated in (I) BWj = Sum of the body weights ofal1 animals fed 2. Grazing The estimation of intake under grazing/browsing conditions is more complicated, and in controlled experiments it is usually estimated indirectly based on fecal production and forage digestibility, as indicated in equation (4). (4) Improved feed production and utilisation to address year-round feeding systems: guidelines 55 In most studies, fecal production is estimated using chromium sesquioxide (Cr2OJ) in capsules, Cr2OJ stained paper with it, using Ytterbium (Yb), or using forage fiber treated with either Cr2OJ or Yb. However this procedure is not practical under farm conditions, as the marker needs to be administered for at least 14 days -7 days adaptation and 7 days for collection of feces (Lascano, 1992). U se of chromogens, lignin (Pezo et al., 1977), indigestible neutral (I-NDF) or acid detergent fibers (I-ADF) (Lascano, 1992), has been suggested to estimate digestibility, the other component of equation (4). Also the short-term in vitro digestibility of extrusa samples have been proposed (Pezo et al., 1977). All these options may be appropriate for on-station research, but results are tedious and/or impractical under farm conditions. An alternative to be applied under farm conditions is the estimation of intake as the difference between the DM available before and after grazing (Lascano et al., 1992). The availability of forage can be estimated using quadrats or the double sampling techniques as described before. Under experimental conditions, this method has been recommended for relatively homogeneous pastures, managed under a high grazing pressure, which results in the use of more than 50% of the available forage in a short period (less than 5 days). Longer occupation periods may result in significant errors due to forage growth occurring during the time pastures are under grazing. The latter could be corrected either by measuring growth in cages that prevent animal defoliation, or assuming that the growth rate during the grazing period is similar to the one observed during the resting period. Although this alternative could be also applied to pastures managed under continuous grazing and in communal areas, estimates tend to be less accurate under those conditions. To estimate intake on a body-weight basis (e.g., kg DM as % BW), it is necessary to establish and maintain grazing records. These should include the following information for each animal grazing in the pasture: .animal identification .date on entering the pasture .date when it was removed from the pasture .number of grazing days .estimated live-weight of the animal (kg) The stocking rate (SR) expressed as kg BW -days can be estimated using the summation of the data collected for individual animals. In turn, intake could be estimated as the quotient between total forage disappearance during grazing time, divided by the stocking rate expressed in kg BW -days, as indicated in equation (5): kg DM available before grazing -kg DM available after grazing= x 100 (5) DM Intake, kg SR, in kg BW -days 56 Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia DESIGN AND EX-ANTE ANAL YSIS FOR TECHNOLOGICAL INTERVENTIONS D. Pezo International Livestock Research Institute Los Bados, Laguna, Philippines INTRODUCTION The design of technological interventions and their ex ante evaluation is one of the most critical steps in farmillg systems research, since it is an initial effort to find solutions to those problems prioritised by fanners. Therefore, the credibility of the research team is going to be scnrtinised by them. The design of technological options is the connection between Farmillg Systems and Components Research (Borel et al., 1985). Options should be based 00 the results ofprevious investigations available in the literature and/or in institutional databases (Figure 1), as well as on the indigenous knowledge identified in the Characterisation phase. During the design phase, the research team could also identify new needs for component research. The design of technological options is the bridge between Characterisation and On-Fann Experimentation. The Characterisation process should result in the description of the main agro-climatic and socio-economic features of the site, the farmillg systems practiced in it, as well as a clear understanding of the fanners' conditions, expectations and motivations (Figure 1). Also, the main Recommendation Domains should be identified and characterised, with their constraints and opportunities for intervention defined. All these serve as a basis for the design and ex-ante evaluation of potential technological options, in terms of: .Biological and technical feasibility .Ecological soundness .Economic viability .Social and cultural acceptability DESIGN OF TECHNOLOGICAL INTERVENTIONS Criteria for Designing Technological Interventions Some critical aspects to be considered for the design of alternatives are the following: 1. Does it answer basic needs and priorities of the majority of farmers in a given recommendation domain ? Most researchers have been trained under a discipline- oriented and "technology-push" approach for doing research. Therefore, it is 57Design and ex-ante analysis for technological interventions co U) .~ E .~ Q) .. ...Q)U)...>u(/)10 ..10.cU I/) ~Q).-u:: ~ .c ;:) 10 O :: I/) 10 Q) >a:~'t- O ~ U!"OE 0) 0);;=:0 .+:- 0 c ~ 0) o..~ f/)c f/) o Q) .+:- E~'- u 10 0) .1. 0.XW ~ >at'-o0.!?-.0 .0 10 .." C - .c .i6 (.) > Q) ozU)0 ~~ t ~ r . ~0')0')...N.0-.5O'CCcaG)'C-ca"'Q):::-C'0G)..JE0--'CG)..Q,ca'Cca.cU-caG)f/I~E~-IC0.5f/IQ,G)..f/I"'iV.~cG)~C"G)(/)... t ~ z ~ii: frequently difficult to become adjusted to the "demand-pul1" research approach required for solving real problems of farmers (Borel et al., 1985; Sagar and Farrington, 1988). In other words, the research team must check whether the proposed interventions respond to the priorities and needs stated by the majority of farmers within a given recommendation domain, or are biased by their professional interests. It does not mean that the problems perceived by farmers are the only ones that exist. In many cases they may not identify the problem roots, but the research team can orient farmers on how to do so, and help them in the prioritisation. 2. What is/are the objectivefunction(s) offarmers? The proposed options could be targeted to different objectives, such as to increase productivity and/or income, prevent capital losses (e.g., animal deaths), enhance sustainability of the system through conservation or even improvement of the water availability and soil nutrient status. However, the research team must be clear about what objective( s ) farmers place their priorities on. In other words, the information gathered during characterisation regarding farmers' interests and expectations, as well as the conditions under which farming is practiced and potentials for improvement is a key to determining the objective(s) to target. These criteria need to be considered when defining potential options for ex-ante assessment. 3. What is the level ofintervention? The hierarchical level (e.g., system, subsystem, component) to be focused by the intervention is a function of the project's and farmers' goals and objectives, but the research team also needs to determine where the greatest payoffs will be obtained (Quiroz et al., 1990). For example, if the problem tree analysis developed during the Characterisation phase identified poor live weight gains of growing animals due to a low nitrogen content of the basal diet (e.g., rice straw), supplementation with a source of fermentable nitrogen (an intervention at the component level) may alleviate that problem. However, if there is another problem that may hinder a good response to that intervention ( e.g., high parasite infestation), then the intervention should consider the interaction between nitrogen supplementation and parasite control, and in that case the intervention is at a subsystem level. There are situations where the proposed intervention may involve more than one subsystem; therefore it is at a system hierarchical level. For example, the proposed intervention may require the introduction of a new crop component into the system, and this may compete for land and labour with other crops already present, as well as for labour with the animal component. 4. How to manage limitingjactors ? The research team can deal with limiting factors by tackling them either directly or indirectly through the use of the system's flexibility. Usually it is easier and cheaper to use the second approach, but frequently it takes time to get the expected results. In some cases, the use of the 59Oeslgn and ex-ante analysis for technological interventions direct approach could be a way to temporarily solve the problem faced by fanners, even at a greater cost, but they must be aware that the research team is also working with them in the preparation of a more sustainable option. Using the problem described in the previous paragraph, a direct way to overcome the deficiency of fennentable nitrogen in rice straw-based diets is the use ofurea-molasses blocks. This option could be effective, but the fanners will depend on purchasing external inputs. A mid-tenn and more sustainable alternative could be the use of leguminous trees, but they need to be planted, and will take time before they are ready for harvesting and used as protein supplements. 5. Are there indigenous techn%gies we can bui/d on ? Indigenous knowledge embodies a capacity to interpret biological processes in the local environment. However, it should be seen as a dynamic process of experimentation to be fostered rather than a static source of knowledge to be "mined" for occasional gems (Farrington and Martin, 1987). Therefore, it is important to check the inventory of indigenous technologies prepared during the Characterisation phase before designing any technological options. In case one or more are identified as possible options to attack the problem, then researchers should try to define its (their) rationale, and look for synergies with the formal scientific knowledge, which they could bring in based on their own experiences and/or literature search. Focus on simp/e component interventions or "techn%gica/ packages?" It is clear that most of the time there are synergistic effects between different components of a "technological package", but very seldom do farmers adopt the whole (Pezo, et al., 1993). For example, it is known that improved grasses respond to fertilisers better than the native ecotypes, and it is under this management that the former exhibits its yield potential. However, in most cases farmers tend to take the new forage, but will manage it as the other grasses they have in the farm. Hence, designing a technological option based on improved forages needs to consider response differences at various levels of inputs and/or management (sensitivity test). 6. What variables to measure? The criteria and/or variables used by farmers to rate technological options may not be the same as those applied by researchers. Therefore, when planning ex-ante evaluation of technological options, researchers need to identify and include in the protocols the criteria used by farmers to determine if an option is better than others, as well as the units of measurement commonly used by them. In some cases these aspects may have not been determined during characterisation, so that additional effort to collect such information is needed. For example, an experiment conducted by Ash by et al. (1987) showed that the ranking of bean genotypes made by farmers and researchers differed, because the criteria used by both groups and the relative importance given to each criteria were not the same. The same happened within the farmers group, when men and women made the evaluation separately. 60 Improving the contribution of livestock to crop-animal systems in rainfed areas ID Southeast Asia Elements Required for the Description of Technological Options More than one technological option for the solution of each problem should be identified, and each needs to be thoroughly described and documented for ex-ante evaluation. Inputs from all research team members, as well as cooperating farmers, must be considered for the identification of the alternative( s ) to be tested, but the research team should take the final decision on which are implemented at the farm level. For the analysis of pros and cons of each alternative, a complete set of information regarding needs and expected results has to be prepared. In this respect, Le6n- Velarde and Quiroz (1994) suggest the following: .Identification of the proposed technological option to be tested .Rationale for the proposed technological option .Actual farmers' practice which will be used as a control treatment. This information is usually obtained during Characterisation, but may need to be complemented through specific interviews with farmers, specially if details were not identified .Description of the technological option(s) to be tested, stressing on the difference(s) with actual farmers' practice .Expected results to be obtained after the application of the proposed intervention. It could be higher yields, lower use of imported inputs, more ecologically friendly practices, less demand in labour, reduction in costs, etc. The criteria chosen need quantification, in order to use them as a reference for the analysis of the technical viability and economic feasibility of the alternative( s) under consideration (Figure I ) .Socioeconomic requirements and implications of the proposed technological option( s ). Although the research team could start a working list in this respect, it will be enriched by farmers' contributions in participatory planning sessions. The criteria identified will help to analyze the potential acceptapility by farmers (Figure I) .Estimate of the time (e.g., months, cropping seasons or years) required to obtain the proposed results .References that support the proposed technological option(s) Box 1 shows an example ofhow a technological option can be described using the above-mentioned guidelines and similar descriptions should be prepared for the other options under consideration. Before submitting options for consideration by a group of smallholders, especially if they are illiterate, the alternatives should be simplified, and presented preferably using drawings, which will facilitate the expression of his/her points of view by farmers. 61Oesign and ex-ante analysis for technological interventions Box 1 .Example of how a technological option is described Option: Use of urea-treated rice straw for dual-purpose cows Rationale: The poor crude protein content ofrice straw ?4.0%) limits digestibility and intake, as well as milk production (saleable and suckled by calves). Cows loose weight even when supplemented with 1.5 kg of concentrate, because of mobilisation of body reserves to produce milk. Actual Farmers. Practice: Untreated rice straw fed ad /ib to dual-purpose cows, supplemented with 1.5 kg of commercial concentrate at milking time Technological Option: Urea-treated (6%) rice straw fed ad /ib, keeping constant the amount of concentrate (Note: All details and needs for urea treatment should be described here, as this is not commonly practised by the goal farmer population). Expected Results: 60% increase in rice straw intake; saleable milk production will increase by 1.0 kg (from 2.5 to 3.5); cows will maintain weight instead of losing 250 g/ day; and calf gains will improve from 150 to 250 g/day. Additional cost and income: $0.11 and 0.23/cow/day, respectively (Note: Details on the costs for urea treatment should be included). Special Requirements: For efficient urea treatment two enclosures with at least two lateral walls are needed, as well as thick plastic to cover the straw and prevent escape of ammonia. Each batch should be enough to feed animals for 2 or 3 weeks (time required for the other batch to be ready). Water has to be added with urea (30 liters/lOO kg of straw). It is desirable to have a sprayer for better distribution of urea. Urea is a toxic comnound. needs to be managed with care, and kept away from children and animals to prevent poisoning. Simple training is needed for farmers to apply this technological option. No indigenous technology to support the proposed option. Estimated Time to obtain Results: Results are observed almost immediately after urea- treated straw is fed. Reference: Chenost M. and Kayouli C. 1997. Roughage Utilisation in Warm C/imates. F AO Anima/ Production and Hea/th Paper No. J 35. F AO (Food and Agriculture Organisation of the United Nations), Rome, Italy. 226 pp. EX-ANTE ASSESSMENT OF TECHNOLOGICAL INTERVENTIONS Estimation of Potential Biological Responses Before starting the implementation of on- farm trials, every effort should be made to estimate which will be the biological and economic impact for each of the technological options under consideration as alternatives to solve a given problem. There are different ways to do it, but all of them require data either available in the literature or provided by key informants. However, in most instances some assumptions are also needed. Estimates of biological response could be obtained through the use of quantitative formats or simulation models (Leon- V elarde and Quiroz, 1994). Use of quantitative formats Quantitative formats are simple models comprising a logical series of arithmetic operations performed either with pocket calculators or utilising any spreadsheet computer program. The advantage of the latter is that once formulas are input, new estimates can be easily calculated with different sets of data, and a sort of test of sensitivity could be performed. The same can be done with calculators, but it will take more time, and depending on how complicated are the arithmetic operations involved, it could become a tedious exercise. Regardless of the device to be used, it is helpful to build a flowchart that describes the logical order for calculations, as well as the assumptions and formulas needed. Figure 2 illustrates the logic used for estimating the amount of milk produced or the live-weight gain/loss obtained when using different feed resources. On the other hand, in Box 2 are included the equations used by Kristjanson et al (1999) to estimate the potential impact of genetic enhancement of sorghum and millet, following the flowchart depicted in Figure 2. Based on both pieces of information, it is possible to construct a spreadsheet template to estimate the potential animal output for diets containing different qualities of forages, which will serve as a basis for technological options to be proposed. These flowchart and spreadsheet could be more complex considering more than one feed at a time, and the potential interactions among them, but the general procedure (definition of assumptions and equations) remains the same. The same type of quantitative formats could be used for agronomic management interventions, and/or combinations of these with the utilisation of crop residues for animal feeding, but the principles illustrated in the simple example shown in Figure 2 also applies to those situations. Box 2. Equations used to estimate milk production and/or live weight change based on digestible organic matter (DOM) intake OMI = 42.78 + 2.3039*DOM -0.0175 DOM2 -1.8872*NoM2 + 0.2242 * NOM*DOM Where: OMI = organic matter intake DOM = digestible organic matter N = nitrogen OM = organic matter Maintenance requirements: 32 9 DOM per kg BW 075 Milk requirements: 1 kg of Milk = 5.09 MJ ME Live weight gain (L WG) requirements: 1 kg L W change = 48.9 MJ ME Where BW = body weight MI ME = megajoules of metabolizable energy I kg digestible OM = 15.8 MI ME 63Oeslgn and ex-ante analysis for technological interventions ""t", c0"CQ)tnIU.cCO.~CJ;,"CO~c.~.E~OtntnO:::c.caC)....cC).Qj . ~~ ... Q) IU > CJ = C -.-0 Q) C.?0 IU ...CIU .- E ~ .-Q)tn ... Q) IU Q) E .c CJ ~ c 0 IU -C)t: O IU Q) .c-CJ.c~ .~ 0 Q) -C)u. .-"C NQ)~;,C)u: Use of simulation models A mathematical model is an abstract representation of a real system through the use of equations, and simulation refers to the use of those equations to predict the performance of such system under a given set of conditions. Simulation models could be deterministic if for each condition only one value is obtained, and stochastic if probabilities are associated to the solutions given by the model. In other words, different values could be obtained each time the mathematical model is run (Leon- Velarde and Quiroz, 1994). They could also be dynamic or static, depending on whether the time factor is considered or not in the model. The simple model illustrated in Figure 2 as a basis for the spreadsheet calculations is deterministic since probabilities are not associated to any element of the system, and static since changes in time are not involved in the calculations. For example, to get an estimate ofhow animal performance varies with changes in feed quality along a given season, then the model has to be run each time with a new set of inputs, as determined by the variations with time. Simulation is a powerful tool to assist in the evaluation of potential responses (biological and/or economic) to interventions in the system, assuming that the model used is an appropriate abstract representation of the reality where changes are supposed to occur. With the capability of personal computers to do complex calculations and iterations in a few seconds, the use of simulation models is more accessible to users. However, before utilising these, the user has to be clear that the model has been validated under conditions similar to the ones he/she wants to apply. Crop-animal systems are complex, not only in terms of their structure but also because of the interactions among components. For that reason, a modular approach for the construction of simulation models has been proposed, utilising existing crop and livestock models as building blocks, and efforts have been put on developing valid and robust interfaces among them (Thome, 1998). Some validation of these models has been performed considering the systems practised in Africa (Thome, 1998) and Latin America (Herrero, 1998), but there are opportunities for adapting those to the crop-animal systems practised by smallholders in Southeast Asia. Figure 3 shows the general structure of the model proposed by Herrero (1998) for the simulation of crop-animal systems using a modular approach. The model has modules (each is a separate model routine) for the animal, the herd, pastures, soils and crops. The animal module consists of a dynamic model of digestion linked to a nutrient requirement system. The outputs of the model are animal intake (forages and supplements), animal outputs (milk, live-weight gain), and excretions. The animal module is linked to DYNAFEED, a database of dynamic nutritional information on tropical feeds, including data on the contents of crude protein and different carbohydrate fractions, their digestibility and rate of degradation, as well as energy values. On the other hand, the herd module is able to simulate the effects of changes in management practices on the dynamics of the herd, its production and economic performance. 65Design and ex-ante analysis for technological interventions Figure 3. Modular structure of a simulation model for crop-animal systems (adapted from Herrero, 1998). The pasture module is driven by environmental variables, and allows the prediction of pasture biomass availability and botanical composition throughout the year, under different management practices (e.g., stocking rate, fertiliser application, grazing management). Up to now, the pasture module has only been validated for Kikuyu (Pennisetum clandestiunum ), signal (Brachiaria decumbens ) grasses, and native grasses. The crop module uses the CERES simulation models for different crops, which form part of the Decision Support System for Agrotechnology Transfer (DSSA T). The soil module is based on the soil organic matter CEN11JRY model, which is used to simulate the effects of different resource management strategies on nutrient cycling, as well as the availability of nitrogen for crops and pastures. Determining Economic Feasibility There are different procedures for estimating the economic feasibility of technological interventions in a system, but the decision on which procedure is the best to use will depend on the availability of information, the type of technological option proposed, and the characteristics of the systems intervened. To perform an analysis of the economic feasibility of any option, it is necessary to have an estimate of the biophysical r~sponse to be obtained when the alternative is applied, and this could be obtained through any of the procedures described above. Also, it is important to define clearly the units and/or indexes to be used. Some of the units of measurement and indexes commonly used for the type of economic analyses frequently applied for the ex-ante evaluation of technological interventions are defined in Table I. The same can also be used to evaluate the performance of actual systems. Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia66 Table 1. Units of measurement and indices used for economic analyses Physical Measurements Total Area (ha): Area of land used for productive purposes. It allows comparisons between two or more farms with similar conditions Cropping Area (ha): Area ofland used for crop production. It allows comparisons between farms that have crops with similar intensities, since it is used for the estimation of crop productivity Pasture Area: Area of land covered by grasses or other forage species. Its estimation is sometimes difficult when tethering is practised in alleys, roads, and canal bunds. Under those circumstances, the estimations of animal productivity on pasture basis are misleading Animal Inventory: Refers to the total number of animals available in the farm. As different animal species, sexes, sizes and physiological stages are represented in the farm, for comparisons among farms it is usually recommended to use a common unit, especially in the case of ruminants. That unit is frequently called Animal Unit (AU), and the equivalencies used are: = 1.25, 1.0, 0.7, 0.6, 0.5, 0.2, 0.2 and 0.2 AU, for bulls, cows, young bulls, >2-year old heifers, <2-year old heifers, calves, mature sheep and mature goats, respectively Capital Investment: Refers to the value of land, infrastructure, machinery and animals held in the farm Total Labour: Represents the household members and hired people who work at the farm on temporary or full-time basis. Total Production: Represents the amount (in kilograms, liters, tons) ofproducts obtained in a farm Production Measurements Yield per unit of land: Refers to the total production obtained per unit of land used to produce it, and is expressed in kilograms or tons per hectare ( or the units commonly used in the site) Production per animal unit: Refers to the amount of milk, meat, eggs that an animal produces on a given period of time (kg/day, kg/lactation or season) Labour Measurements Labour-day: Refers to the number of hours per day that a person is supposed to devote to productive activities. Eight hours is a common reference value, but frequently in agriculture it is equivalent to 5 hours of field work. However, the definition should be based on what the labour laws indicate. 67Oeslgn and ex.ante analysis for technological interventions Table 1. Units of measurement and indices used for economic analyses (continuation) Man-day Equivalent: It is a common unit to account for the work done by different gender groups. The reference value is the labour performed by an adult male during a labour-day, and equivalencies of 0.75 and 0.5 have been proposed for women and 6-14 year-old children, but these estimates may change with sites and type of activities Measurements of Income Gross Income (GI): For a specific farm activity, GI is equivalent to the Total Value of Production (TVP), which is estimated as a product of the yield (TP) times the price per product unit ($/unit). For the farm, GI is the summation of the gross income obtained in each farm enterprise (e.g., rice, mungbean, coconuts, goats) (GI = TVP = TP x $/unit) Gross Margin (GM): Is the difference between Gross Income (GI) and Total Variable Costs (TVC) GM = GI -TVC Net Income (NI): Is the difference between Gross Income (GI) and Total Costs (TC). The latter is estimated as the addition of Variable Costs (VC) plus Fixed Costs (FC) NI = GI -(VC + FC) Net Family Income: Is the result of adding the Net Farm Income (NI) plus the one derived from off-farm activities (OFI) (NFI = NI + OFI) Indices of Return to the Productive Factors Cost-Benefit Ratio (CBR): Is an expression of the gross benefit or loss per monetary unit invested. It is estimated as a quotient between Gross Income (GI) divided by the Total Cost (TC). If this is greater than 1.0, the activity under analysis is profitable, if it is equal to 1.0 that is the break-even point and the activity is not profitable, but if the ratio is less than 1.0 then the activity is losing money (CBR = Gl!fC) Profitability of the Investment (PI, %): Represents the net benefit or loss per unit of money invested. It is expressed in percentage, and estimated as the quotient between the Net Income (NI) divided by the Total Costs (TC), and multiplied by lOO (PI, % = 100 x (NI/TC) Return to the Cash Invested in Inputs (RCII): Represents the gain or loss of money per unit of cash invested in buying inputs. It is estimated by subtracting from the Gross Income (GI), the amount of money used to buy inputs (e.g., fertilisers, concentrates, pesticides) (CII), and the amount spent in labour (LC); and the difference divided by the amount invested in inputs (CII) (RCII = (GI -CII -LC)/CII Return to Labour (RL): Represents the gain or loss of money per unit of labour (LU) used in a given activity or the farm as a whole. It is estimated by subtracting total costs (TC), except for those corresponding to labour (LC), from the Gross Income (GI), and dividing it by the number of man-days used in the activity (LU). Therefore, the formula is: RL = (GI -TC + LC)/ LU Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia68 Partial budget techniques A budget is an ordered representation of expected income and expenses. It can be a total or complete budget when it considers the farm as a whole, or even an enterprise within the farm, but considering all costs and income. A partial-budget analysis is the tabulation of expected gains (benefits) and losses (costs) due to a relatively minor change in farming practices, such as changing the basal diet, using a new supplement, vaccination, or drenching (T. Soedjana, personal communication). Partial-budget analysis is not suitable for answering questions in which several factors determine the contribution of a treatment. Other conditions required for a partial-budget analysis to be a useful choice are the following: .Inputs and outputs are measurable and easy to price .Yields vary little among farms .Profitability is the major concern .Fixed costs do not change Some of the advantages of the partial-budget analysis are: .Simplicity .Easy to learn, and can be readily taught to extension workers and farmers .Examines only net changes in costs and benefits .Effective for assessing the economic viability of single-intervention technologies .Requires less data than whole-farm budgeting .A firm conclusion about adaptability of new technology can be drawn without going into more details, as needed for whole-farm analysis On the other hand, some of the disadvantages attributed to the partial-budget analysis are: .Frequently neglects farmers' resource limitations .Technologies are often ana lysed without realising the effects on farmers' resource base .Lack of understanding of farmers' objectives .Time analysis of farm activities and labour distribution are not considered .Assumes that the input-output relationship changes linearly with the scale of operation The following steps should be followed to carry out partial-budget analysis to evaluate technological options in the context of crop-animal systems managed by smallholders: .Define clearly the farmers' objectives for keeping animals, growing crops or both. This information is available from the characterisation phase. .Identify the changes associated to the new technology, and determine which factor(s) remains the same and which change. The detailed description of the technological option, as illustrated in Box I, is a good 69Oesign and ex-ante analysis for technological interventions basis for defining which elements change and which remains the same. In Box 3 are listed those elements that are specifically needed for the implementation of the treated rice-straw option, therefore their costs need to be accounted for . .Quantify the amount of inputs (e.g., supplements, agrochemicals) and other resources (e.g., labour, land) associated to each alternative under study (those who are not constant across alternatives). Determine costs per unit of those inputs and/or resources that vary with the alternatives under evaluation. .Estimate the expected production outputs ( e.g., live-weight gain, excreta production), using any of the procedures for the estimation of biological responses previously described, and determine the price per unit of product. .List all non-cash considerations that are relevant to the farmers' choice. Information on this respect should also come from the Characterisation Phase. .With the information generated in the previous three steps, construct a table including columns for gains and losses, and each column should be divided into two categories: added returns and reduced costs for the "gains" column, and added costs and reduced returns for the "losses" column. All this information is relevant for the comparison of the different alternative practices or technologies under testing. .Interpret results considering whether the best profit criterion has been used, and whether farmers' interest to maximise the benefit of particular resources (i.e. land) is relevant for the analysis. Table 2 illustrates the way this type of analysis is carried out to evaluate the effect of nutritional interventions on the performance of cattle, in a system where cattle is not only important as beef producers, but also for manure. Gross-margin analysis Gross-margin analysis is similar to partial-budget ana1ysis in many ways, especially because both calculate total gains over variable costs, but the difference is that in the former all variable costs are considered (Table 1 ). Gross-margin is usually expressed in the units of a common resource -for example, the gross-margin of the animal component can be estimated on a per head basis, as a measure of the efficiency of animal production activities in pen- fed animals, or per hectare of land in the case of grazing systems. The gross margin can be computed for a given component of the system (e.g., gross margin for animals, gross margin for crops) or the total margin for the entire operation, in which case it is the result of the addition of the gross margins obtained for different components. Improving the contribution of livestock to crop-aminal systems in ralilfed areas in Southeast Asia70 Table 2. Structure of a partial budget analysis for nutritional interventions in a farming system where live weight and manure are important outputs of the cattle component IT. Soedjana. personal communicationl. Gain-loss components Quantity Price WKg MKg FKg L hours/days $w/kg $m/kg $f/kg $/hour or day Additional weight of cattle Manure sold Additional feed required Additional labour Gains (a) Losses (L: Added returns ($) Weight gain RW=(wkg x $w/kg) Manure RM=(rnkg x $m/kg) Added costs ($) Feed CF=(fkg x $f/kg) Labour CL=($) Reduced costs (none) Reduced returns (none) Total G = RW+RM Total L = CF+CL Difference (0- L) = Total a -Total L The advantage of the gross-margin analysis is that it can be easily used to rank different technological options, the best rank being the one with the highest value. However, there is a tendency to conclude that the farm profit can always be increased by expanding a given activity that has higher gross margin per unit, at the expense of those that have lower returns, and ignoring farmers' perspectives, objectives, and/or competition among components for available resources. 71Oesign and ex-ante analysis for technological interventions Cost-benefit analysis Cost-benefit analysis is another alternative for the evaluation of the feasibility of technological options. The procedure is similar as the one described for the Gross Margin Analysis, but the difference is that for this purpose all costs (variable and fixed) are considered. Also cash and non-cash costs and benefits are taken into account. As described in Table I, the Cost-Benefit Ratio (CBR) is estimated as the quotient of Gross Income (GI) divided by Total Costs (TC). Thus, the most attractive alternative according to this criterion wiIl be the one with a higher value. Caution has to be taken when using this procedure because total costs (TC) include all variable (VC) and fixed costs (FC), and many times some (speciaIly fixed costs) are left out or are not estimated properly. For example, a fixed cost could be the use ofland or corral facilities, but frequently there are difficulties estimating their costs (i.e., opportunity cost for land use, depreciation costs for buildings and equipment). Some could leave these costs out of the analysis, invalidating the estimates. If that is the case, the estimation of the Cost-Benefit Ratio is not valid. Under such circumstances, it would be better to use the partial budget technique. Sensitivity analysis Sensitivity analysis is an option through which the researcher can determine how much change an economic indicator exhibits when yield responses, prices of outputs and/or costs of inputs vary .It can be applied to either one of the previously mentioned indices, or even to one in which returns to a given production factor are estimated ( e.g., return to labour, return to land, return to the cash invested in inputs). It is a means of dealing with uncertainty about future events and values (Frio and Bhasayavan, 1988). This can be done by varying one element at a time or a combination of elements, and determining the effect of those changes on the Gross Margin, the Cost-Benefit Ratio, or any other index. As indicated before, sensitivity analyses can be easily performed using spreadsheet templates. Cash flow analysis A cash flow analysis is a procedure that allows identifying variations in cash expenses or income along the year, a crop cycle, a season or any other unit of time. It helps to find out whether a given technological option requires excessive use of cash at a given time, compared to the farmers' present practice, but also if the net cash flow from the alternative under evaluation will be able to finance successive production requirements (Frio and Bhasayavan, 1988). This analysis is particularly critical for those crop-animal activities practised by farmers with limited resources, where cash inflow and outflow are seasonal or periodical (e.g., when animals or crops are sold), and have limited access to credit or are reluctant to use it. Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia72 Cash flow analysis should be perfonned using a more holistic approach. In other words, it should be done not only for the activity or enterprise directly affected by the intervention, since usually there are simultaneous peak demands for cash in more than one enterprise at the fann level. For example, if the intervention is oriented to changes in the rice system, and has special cash demand during the planting season, when other crops are also demanding, the fanner may not adopt the alternative just because applying it will limit his/her possibilities of planting other crops. Moreover, the additional cash demand may also coincide with other requirements in the household ( e.g., start of school year) that can not be ignored. On the other hand, an animal production activity which assures fanners additional income just in time for the planting season when there is need for cash could be preferred over another that results in even higher net income that can be achieved later in the year. Labour profile analysis A labour profile analysis refers to the estimation of labour requirements and availability in the farm, along a given period of time (Frio and Bhasayavan, 1988). To elaborate these profiles, a detailed sequence of tasks to be done and the labour (amount and type) required for each task must be defined. As in the case of the cash flow analysis, even when a partial intervention in the system is under evaluation, it can not be analysed in isolation of the rest of activities in the farm. This analysis can not ignore gender issues, because different household members have their own responsibilities in the household and farm. Also, certain activities are preferentially performed by a given member of the family, or could be culturally restricted to a given gender group, which may have competitive demands of labour for other purposes at the time the proposed technological option needs to be implemented. FARMERS' PARTICIPATION IN EX-ANTE EVALUATION OF TECHNOLOGICAL INTERVENTIONS The partnership between farmers and researchers initiated in the characterisation phase continues during planning of on-farm research, which involves the design and ex-ante evaluation of technological interventions, the definition of how to implement trials at the farm level, which parameters are measured, and other decisions related to on-farm participatory research. However, the role of researchers in the planning of participatory research is more active than facilitating farmers' contributions. When researchers meet farmers to discuss different technological alternatives proposed to tackle the problems prioritised by farmers, they have already done a primary screening, and discarded some that are not technically viable (Figure I ), at least based on the procedures previously described. For the consultation, researchers should provide Oesign and ex-ante analysis for technological interventions 73 farmers complete and neutral information for them to contribute their points of view on the options under analysis. Farmers are expected to help with their knowledge and local expertise for the selection of options to be tested, but their observations may also serve to adjust either the way treatments are applied or measurements are taken (CIA T - FSP,1995). Moreover, farmers can tell the research team if the procedures proposed for testing technology are considering not only the local priorities in terms of costs, biophysical conditions and availability of inputs, but also their social conditions and cultural beliefs. According to Cubillos et al (1988), some criteria that farmers consider when analysing the feasibility of a given technological alternative are: l) Compatibility. One criterion used by a farmer in analysing the feasibility of a given alternative is how easily a technological option can be integrated in the system he/she manages. Although a farmer may recognise that a proposed alternative can contribute to improve significantly the outputs of his/her system, or any other criteria considered relevant, it is less likely that he/she will favour an intervention that represents an important disturbance in the system. 2) Divisibility. Those technological options that require the interaction of several elements to express their potential will be less likely chosen by farmers. For example, if the option is a crop genotype which requires irrigation, high levels of fertilisers and a very complicated pest control program to express its maximum potential will be probably discarded by most farmers, even though it may have higher yield potential. On the contrary , farmers will prefer a genotype that is adapted to the main constraints prevailing in their farm (e.g., tolerance to drought, resistance to the most common pests), even ifit is not as productive. 3) Complexity. Simple technological options are more manageable and applicable than the complex ones. Therefore, farmers will tend to choose the simple ones, and will even sacrifice potential differences in output. 4) Risk. Less-endowed farmers tend to be more resistant to those technological options that represent higher risks. The same may not be true for those farmers with less limitation in resources availability. 5) Profitability. In commercial agriculture, the retribution to invested capital is a must, but not necessary for all smallholders. Although they would like to have a more profitable system, in many cases they can not afford those options, or may have other priorities, such as the reduction of costs in cash, or use less of a given production factor in one subsystem, to allocate to another subsystem. Getting Feedback from Farmers on the Acceptability of Technological Options The consultation with farmers can be done on a one-to-one basis or as a group. However, for the latter, special care has to be taken in how groups are composed. Groups should be the most uniform as possible in terms of gender, availability of resources and other criteria. One-on-one consultations The one-on-one consultation recognises that farmers may react differently when approached individually than as a group. If this option is selected, special care has to be taken on who is interviewed. There is a tendency to consider for interviews only the head of the household, but those members of the household who manage the subsystem affected by the intervention should be the ones approached; otherwise, the research team will not be able to appreciate the real implications of the proposed alternatives. The interview should be conducted in the farm or house of the interviewee, with not many members of the research team present, but one of them must be the technician who knows the farmer for a longer time. All this will contribute to prevent inhibitions of those farmers or family members who tend to be shy with strangers. To start the meeting, one of the members of the research team explains in a very simple manner the purpose of the consultation, and how the data provided by the farmer during the characterisation phase was used for the formulation of the technological options under consideration (Cubillos et al., 1988). A recognition visit to the farmer's field and animals, demonstrating interest on the resources they have and the practices they apply is a good way 'to break the ice' for the consultation. Also, asking farmers to help draw a map of their property is a valuable means to establish rapport with him/her , and in fact it may serve later to identify potential sites where the alternatives could be implemented, in case these involve crops or pastures. This will serve as a concrete basis for discussion, since it will allow the farmer and other members ofhis/her household to identify how feasible is a given option under their farm conditions. The use of graphs and carefully prepared analyses of the different options will facilitate retrieving valuable information from the farmers. The research team must be clear on all details of the proposed alternatives, and prepared to register the evaluation and comments made by farmers. Although researchers can take some notes during the discussion with farmers, it is important to have the discussion quite open and smooth as possible. One possibility is having at least two team members working at the same time in the interview, one conducting the discussions, and the other taking notes (Ash by, 1992). However, to be successfui in getting the valuable contribution from farmers, the research team must assume an open-minded attitude, and demonstrate genuine interest on the farmers' comments and observations. 75Oeslgn and ex.ante analysis for technological interventions Group consultation Group consultations have the same purpose and structure as the one-on-one consultation, but is conducted with a group of farmers or household members. It requires a better preparation of audiovisual aids to allow all of them to become aware of the requirements and expected results for each technological intervention that is proposed. When the research team works with groups, frequently there is a risk that only few express their points of view, but this can be managed politely by the facilitator. Also, when there are two team members in the discussions, one of them should identify the ones not participating, and he/she can ask separately the points of view of those who did not participate in the discussions. It is recommended that the facilitator be a researcher or technician well known by the group; better, a person who has been in contact with the farmers for a longer time. He/she must be clear that his/her role is to encourage farmers to react to a given set of alternatives, and not to teach them how to do things. The feedback from farmers will be very poor if they are not convinced that the research team is trying to learn from this interaction, and looking for genuine participation of farmers in identifying potential alternatives to improve their systems. It must be remembered that through this consultation process the research team is not only looking for the farmers' reactions on the feasibility of each of the technological options to be proposed, but also for new ideas on how to implement these. Consequently, records regarding both aspects should be kept. Regardless of the option applied (i.e., one-on-one or group discussions), attention should be paid to time management in order to avoid long sessions which may negatively affect the participants' contribution. Also, to obtain better results, it is desirable to concentrate on just one set of options, selecting those that tackle the problem identified with the highest priority .In following sessions, with the same group or others, different options could be discussed, if needed. TECHNIQUES FOR RANKING AND SCORING ALTERNATIVES The main purpose of having farmers as evaluators is to discover if the criteria applied by them to select a given option differ significantly from the ones used by researchers. Therefore, farmers should have the opportunity to express even concepts that are unexpected by researchers, and explain their rationale (Ash by, 1992). Those concepts and their analysis are probably the most important contributions farmers could do to the research program. Moreover, this step is necessary considering that farmers will also be involved in the actual evaluation of technical options in their own farms. The use ofranking and scoring techniques is an adequate option to get that type of feedback from farmers, since those are analytical instruments designed to manage information regarding the preferences and perceptions of farmers with respect to problems and opportunities in the systems they manage. These also help to determine Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia76 why different people have diverse criteria for making decisions and judgements about choices of technologies or any other things relevant to the farmer (T. Soedjana, personal communication). Among the techniques used for this purpose are the following: .Preference ranking .Pair-wise ranking .Matrix ranking Preference ranking Preference ranking is a technique in which each individual option is compared against the others, until they are ranked from the most to the least preferred. To compute the assessments made by different farmers, they are requested to assign scores of 5 (most favorite) to I (least favorite) for the different options under evaluation, and once all are registered, the "average" ranking is computed. Table 3 shows an example of how data is tabulated and computations are made for a hypothetical example in which problems associated to a given technological option are ranked. The same procedure is used to rank different technological options using the same scoring criteria, but in this case, it is desirable to ask each respondent to speak aloud about the reason why he/she ranked each option in the way they do (CIA T -FSP , 1995). Table 3. Application of the preference ranking technique for the evaluation of a given technological option in terms of the problems associated to it. Respondent Problem Score Rank B c DA E F Not enough land 4 3 5 34 4 23 3 Scarcity of financial 4 3 5 4 5 4 25 2 Demands labour in a critical time 3 4 4 1 3 3 18 4 Too much risk in the implementation 5 5 3 5 4 5 27 1 Requires high-quality animals to payoff 1 2 1 3 1 1 9 5 77Oesign and ex-ante analysis for technological interventions Pair-wise ranking The pair-wise ranking method produces a matrix ranking and scoring by comparing each item against the other, and the respondent will define if a given option is better or worse than the one it is compared with. It is recommended not to have more than six options, otherwise the method becomes tedious (Ash by, 1992), and the number of comparisons increases very fast as treatments increase. For example, for four treatments there are six possible comparisons, but these increase to 10 when five treatments are considered, and to 15 with six treatments. Each option is identified and annotated in a card, and two of these are shown to the interviewed person at the same time. He/she is asked which option he/she prefers, and frequently, the reasons for their preference are also asked. These procedures are repeated until aIl possible combinations have been used, and the results are put into a table. Table 4 shows an example of pair-wise ranking, including information on how computations are made to rank options, considering a total of five interviews, and comparing five treatments. Each cell is inputted the number of interviewees who preferred the option listed in the row and compared to the one listed in the column. For example, when S I was compared to S2, three informants said S2 was better than S I, and the other two said the opposite. Then in the cell defined by column S2 and row S I, there is a 3, whereas in the cell defined by column SI and row S2, 2 is written in parenthesis. Also, all five informants said that option S2 was better that S3; hence, S is recorded in the cell defined by column S3 and row S2, and zero (0) in the cell corresponding to column S2, row S3. Table 4. Summary of hypothetical results obtained by applying pair-wise ranking to different land use options. Land-use Option* Score RankTreatment S2 S3 84 SSSI SI 13 23 4 3 3 5 S2 16(2) 5 4 4 S3 9 3(I) (0) 4 4 S4 8 4(2) (I) (I) 5(I) (I) SS 4(2) (0) .81 = upland rice, 82 = strip cropping; 83 = alley cropping; 84 = hillside; 8S = conventional practices To compute the total score assigned to S I, add all the values in the corresponding row. For example, the score for SI equals 3+4+3+3 = 13, and for S4, 2+1+1+4 = 8. After doing all computations, identify the option that received the highest score. This is ranked as number I. In the case of the example, S2 is the most preferred option, with 16 points, followed by option S 1 with 13 points. Although the table does not include the reasons, these could be also analysed, in order to identify why each option was ranked in a given position. Matrix ranking Matrix ranking is a technique in which farmers compare items or technological 'options against selected criteria used to judge them. It is desirable that farmers also participate in the identification of the most relevant criteria for evaluation. For this purpose it is highly recommended to avoid having too many options and criteria for evaluation, since due to fatigue, responses of informants may become mechanical (Ash by, 1992). In matrix ranking, scores are assigned based on the relative suitability of the option to each criterion. The highest score (or the first rank) is given to the option which fits the best according to the criterion under consideration, and the lowest to the one most poorly suited or absolutely non-suitable. Special care needs to be taken when formulating questions regarding the criteria used. For example, if the criterion refers to demand of cash (assuming it is a scarce resource ), the question should emphasise "~ demand for cash." Then the highest score will be obtained by the option which requires the least; otherwise, the scoring may be misleading. Once all criteria have been considered, and all informants interviewed, the data is summarised adding the scores obtained by each option, and the option preferred by evaluators is the one with the highest score. In Table 5, there is an example applying this tool for the evaluation of five dry- season feeding strategies for ruminants, using five criteria previously identified by evaluators. Alternatives were ranked based on all criteria, but using one at a time. As the computation is made by simply adding the scores obtained using each criterion, the same weight or importance is given to all, and this may not be the rationale when decisions are made. One option to overcome this problem is to arrange the scoring in a way that the researcher can identify not only the option evaluators prefer, but also the most relevant criterion for them. For that purpose, evaluators are asked to distribute a total score ( e.g., 100) among all cells resulting from the combination of options and criteria. Although it is a little more difficult procedure, it has been successfully applied with farmers in different developing countries, frequently using a simple way to allocate scores (e.g., small stones, grains) in a two-way table. Table 6 shows the same example as in Table 5, but after giving evaluators the opportunity to consider a relative weight for each criterion. Notice that it was possible to identify the criterion to which evaluators gave more importance (i.e., effectiveness 79Design and ex.ante analysis for technological interventions to prevent live-weight losses was the one with the highest score). Also, the ranking of options changed due to the different weight evaluators gave to all criteria. When each criterion was considered independently, feeding only crop residues was the option selected (Table 5), whereas the use of these plus tree foliages was the preferred alternative when the relative importance of each criterion was also taken into account (Table 6). ACKNOWLEDGEMENTS This paper is based on the drafts of two chapters of the Training Manual entitled "Research Approaches and Methodologies for Improving Crop-Animal Systems in Southeast Asia". Co-authors of those chapters are Drs. Tjeppy Soedjana and Carlos Leon- V elarde. Special thanks to both of them for allowing the use of the drafts, and their examples as used in this paper . Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia80 Oesign and ex-ante analysis for -cG)..G)......:c...oco.~fO~ca>G)G).c-..0..."C0.c-G)EC)c:.v:cfO..><.~-fOEG).c-C)c.> . -IIJc.-c.CfO fOC G) E .::' ~ fO .. "C .. G) O C....-IIJfO G) -.-.cC)0 G)- IIJ fO -..--~ IIJ IIJG) C) ..CG):C.c G) -G)O C G) 0 -IIJC.fOE G) fO IIJ >< > W .. ."C It)G):cfO~ "' ] ~ ~:gu"'-~~:;::,d Q. 0 e Zu fII -0 U p..gtIS UfII-.bUNI~-~ 0, ..0;::, ..U f/)~ ~ ~ +?~:::- 00 ~ -Ql~ ~ ~!0.0 I.--.::. ~ ==;0.~ .g U + ..8 .Q -=~-=- I.l: - -~;3 = "' ~O~~ .l: u + .E .9 -;3~N = - ?.I:~'I:u "'~"'"'..9-fo?~I~=~>~c.o-"'"'~~>0.=U~~ - N ~ M ~ ~~.0u'-'~.8]u~u..~1u"1:)~o..J II"\ r'\ ~ N - .s"'u.~.>.~u.uu-so-s...~~ o E..J:Ju .u Q Si;8 ? -.~0 ... d:gO .u O Ir\ N M "'t" i=eQ.'g.4)]0.~~.Uu,~ .a g.~~ ~ -.u~~0 0/') - -.t ~ ~8:~-;u]~]O':=:gcu-0~=80Q IrI N ..,. M - ~u(/)3~ -N 0- 0.- '0- 0\ ~~u>< '"'" N ~ 11"1 81 REFERENCES Ash by J.A. 1992. Manual para la Evaluacion de Tecnologia con Productores. CIAT (Centro Internacional de Agricultura Tropical). Proyecto IPRA. Cali, Colombia. 101 pp. Ash by J.A, Quiros C.A. and Rivera Y.M. 1987. Farmer Participation in On-Farm Varietal Trials. Overseas Development Institute. Agricultural Administration (Research and Extension) Network. Discussion Paper No.22. London, UK. 30 pp. Borel R., Ruiz M. E., Pezo D. and Ruiz A. 1985. A methodological approach for the development and evaluation of animal production systems alternatives for the small farmer. In: Report II Workshop on Tropical Animal Production Systems. (Eds. Li Pun H. H. and Zandstra H.). IDRC Manuscript Report No. 62e. pp.36-77. CIAT-FSP. 1995. TrainingModulesforFarmerParticipatoryResearch. CIAT(Centro Internacional de Agricultura Tropical). Forages for Smallholders Project. Los Bafios, Philippines. Cubillos G., Ruano S. and Vargas H. 1988. La confrontacion de la tecnologia en el proceso de generacion y transferencia. In: Informe VIII Reunion General de RISPAL. Ciudad de Guatemala 17-21 Octubre 1988. (Eds. Ruiz M E and Vargas H). IlCA (Instituto Interamericano de Cooperacion para la Agricultura). San Jose, Costa Rica. pp. 377-388. F arrington J. and Martin A. 1987. F armer Participatory Research: A Reviewof Concepts and Practices. Overseas Development Institute. Agricultural Administration (Research and Extension) Network Discussion Paper No.19. London, UK. 88pp. Frio A.L. and Bhasayavan N. 1988. Target area and research site selection and description for crop-animal systems research. In: Proceedings of the Crop- Animal Systems Research Workshop held in Serdang (Malaysia), August 15- 19, 1988. IRRI (International Rice Research Institute). Los Bafios, Philippines. pp. 473-514. Herrero M. 1998. Integrating methodologies for analyzing crop-livestock production systems. In: Proceedings of the Workshop on Ecoregional Research at ILRI. Addis Ababa 5-8 October 1998 (Eds. Thornton P K and Odero A N). ILRI (International Livestock Research Institute). Nairobi, Kenya. pp.87-95. Kristjanson P.M., Zerbini E., Rao K.P.C., Kiresur V. and Hofs P. 1999. Genetic Enhancement of Sorghum and Millet Residues Fed to Ruminants: Ex Ante Assessment of Returns to Research. ILRI Impact Assessment Series 3. ILRI (International Livestock Research Institute). Nairobi, Kenya. 44 pp. Leon-Velarde C. and Quiroz R. 1994. Antilisis de Sistemas Agropecuarios: Uso de Metodos Bio-matemtiticos. CIRNMA (Centro de Investigacion en Recursos Naturales y Medio Ambiente). La Paz, Bolivia. 238 pp. 83Design and ex-ante analysis for technologicallilterventions Pezo D., Villegas L.A. and Romero F. 1993. Modulos lecheros ITCO/CATIE: Una experiencia de adaptacion tecnologica a nivel de finca. In: Investigacion con Pasturas en Fincas. Memorias de una Reunion de Trabajo realizada en Palmira, Colombia, 27-29 Agosto de 1990. CIAT (Centro Internacional de Agricultura Tropical), Cali, Colombia. Documento de Trabajo No.124. pp.203-214. Quiroz R., Mamamni G. and Arce B. 1990. Aplicacion del Enfoque/Analisis de Sistemas en la Investigacion Pecuaria. In: II Seminario Taller Enfoque y Ancilisis de Sistemas Agropecuarios Andinos. Proyecto de Investigacion de Sistemas Agropecurios Andinos (INIAA-PISA). Puno, Peril. Serie Didactica. Material de Ensefianza No.4. pp.120-127. Sagar D. and Farrington J. 1988. Participatory Approaches to Technology Generation: From the Development ofMethodology to Wider-Scale Implementation. Overseas .Development Institute. Agricultural Administration (Research and Extension) Network. Network Paper No.2. London, UK. 50 pp. Thorne P. 1998. Developing integrated models for application in crop livestock systems. Proceedings of the Workshop on Ecoregional Research at ILRl. Addis Ababa 5- 8 October 1998 (Eds. Thomton P K and Odero A N). ILRI (International Livestock Research Institute). Nairobi, Kenya. pp.105-121. Improving the contrib/Jtion of livestock to crop-animal systems in rainfed areas in So/Jtheast Asia84 METHODOLOGIES FOR ANAL YSIS OF HOUSEHOLD SURVEY DATA D. Pezo International Livestock Research Institute Los Baiios, Laguna, Philippines C. Leon-Velarde International Livestock Research Institute and Centro Internacional de la Papa, Lima, Peru INTRODUCTION Characterisation of a benchmark site is a research activity, therefore the scientific method has to be applied. Within this context, Rey et al. (1993) suggested the following general steps for the characterisation of market-oriented dairy systems in Sub-Saharan Africa. These are also applicable to the crop-animal systems practised in Southeast Asia: .Fonnulate questions and hypotheses to be answered or tested in relation to perceived problems in the systems under study .Determine the analytical methods to be used for answering the questions fonnulated and to test the proposed hypotheses .Define the data needs to respond to questions and test the proposed hypotheses .Identify the sources and the most appropriate data collection procedures for the different sources .Design the survey protocol, including the sampling design, the field instrument to be used to gather the information and the analytical procedures to be applied. As indicated earlier, this step is required only if primary data is deemed necessary . .Analyse the data collected to test the proposed hypotheses .Draw conclusions based on the hypotheses tested A logical sequence of steps for the choice of analytical methods, data and data sources, methods of data collection, and sampling procedures is shown in Figure I (Jabbar et al., 1997). Most of these steps are discussed in detail in following sections, but the emphasis of this paper is on the methodologies used to analyse household survey data. A household survey is a part of the analytical methods that could be applied for the characterisation of farms within a benchmark site (Lapar, 1999). Its use is recommended only if primary data is deemed necessary to obtain baseline information on the prevailing crop-animal systems, and when quantitative data is needed for making valid inferences on a set of hypothesis formulated (Rey et al., 1993). The method of analysis to be applied is a functioo of the research objectives, the type of data collected and the hypothesis formulated. While it is highly recommended 85MethtX/ologies for analysis of household survey data that statistical procedures to be used for the analysis of data be defined before starting to collect data in the field, the opposite is frequently practised. Statisticians are often approached for advice on which analytical procedures could be applied to a set of data already obtained without defined hypotheses to test. STEP1 j a nS::~a~I: :~dt~~able? I.n~e,.ble ;' bJe? . ~ .:!' STEP 2 -~ STEP3 ST EP 4 Figure 1. Steps in the characterisation of crop-animal systems (Jabbar et al. 1997). 86 Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia Yes ~ s TE p 5 Determine method of data collection (Dq s YE p 6T1 For primary data, design survey including sample design Will survey & secondary data be adequate for analysis and interpretation? NO YES Characterise crop- animal systemReassess OS, MOC Figure 1. Steps in the characterisation of crop-animal systems (continuation). Methodologies for analysis of household survey data 87 REDUCING THE NUMBER OF VARIABLES Household surveys usually result in the collection of too many variables. This eventually limits the capability of the research team to visualise, interpret and under- stand the multiple relationships existing among variables within households and among households. This problem is exacerbated when more components are present in the farms under study and more diverse backgrounds are represented in the research team. The reduction of the number of variables in the database is a must, and different options could be used. A priori, the research team can determine whether a given attribute is relevant and/or reliable information can be obtained. If the answer is NO, then that variable should not be included in the questionnaire used for the survey. Once data collection has been completed, the research team can also review the list of variables and check for their relevance and/or reliability .Other options to reduce the number of variables a posteriori are: .To use statistical procedures which may result in building indexes which summarise relevant variables and discard those that provide limited information (e.g., Principal Component Analysis) .To build indexes with a specific meaning, incorporating several variables that have some degree of association (not necessarily statistical association). STATISTICAL ANALYSIS FOR DETERMINING BASELINE STATUS The determination of the baseline status entails quantification of descriptors that characterise a population at a given point in time. Therefore, simple analytical tools can be used, but the appropriateness of the method will be a function of the type of data collected. Some attributes are categorical { e.g., type of feeds used to supplement the basal diet) or even of binary nature { e.g., provide supplements or not, pregnant or non- pregnant, preferred or non-preferred), and others are continuous {e.g., performance indicators such as milk production, live-weight gain). In the case of categorical data, results can be expressed either as frequencies or percentages, and this can be applied to the whole sample or to separate sub-sets of the sample. An example of this is on the use of rice straw for animal feeding and is shown in Table I. If the variables are continuous, then measures of central tendency { e.g., mean, mode, median), and dispersion {e.g., standard deviation, standard error, coeffi- cient of variation, range) for the whole sample or by sub-sets could be appropriate {Steel and Torrie, 1980). An example of baseline productive performance parameters for the cattle component in a sample if dairy farms is shown in Table 2. 88 Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia Table 1. Baseline information on the use of rice straw for animal feeding in a sample of farms covered by a household survey. Criteria of Classification Frequency % Total% District 43 14 57 75.4 24.6 100.0 33.1 10.8 43.8 District I Yes No Sub-Total 63 10 73 86.3 13.7 100 48.5 7.7 56.2 District 2 Yes No Sub-Total TOTAL 130 100 Table 2. Baseline performance parameters for the cattle component in a sample of farms covered by a household survey. Number of Animals Mean:f: S.D.Parameter Range Calving interval, days 127 355.1% 32.1 276 -694 Lactation length, days 102 218.0:i: 96.0 50 -574 Length of the dry period, days 72 148.4:j: 87.2 II -455 Milk yield, kg per lactation 102 1355.9:r. 622.4 191 -3254 Milk yield, kg/cow/day 102 5.16% 1.55 1.65- 10.4 Methodologies for analysis of household survey data 89 TEST OF HYPOTHESES Observations made during preliminary visits to the benchmark site, information obtained through PRA activities (e.g., focus group discussions, interviews with key informants), or even the review and analysis of secondary information can serve as basis for defining hypotheses to be tested with the information generated by the house- hold survey. Therefore, the questionnaire used for the survey should include questions that respond to those hypotheses. However, the complexity of crop-animal systems in terms of components and interactions may lead to a very long list of hypotheses. This could become even longer due to the multi-disciplinary character of the research team, as each member may have different perceptions of the problems and their causes, based on his/her professional background. Therefore, consensus is needed for the definition and prioritisation of researchable questions and hypotheses to be tested in the household survey. It is also important to have in mind that hypotheses need to be defined at the outset of the project. This avoids the costs of changing research plans after they are already under way, or repeating activities to obtain information that is missing (Mullins et al., 1994). Some examples of hypotheses that can be formulated when studying crop-animal systems, are the following: Animal Component . . . . Cut and carry is practiced by the majority of farmers Land allocation to forage production is positively related to farm size and secu- rity of land tenure The extent of use of crop residues is negatively correlated to the length of the growing period The use of tree foliages as feed is not a common practice Ruminants and monogastric animals are complementary, since both do not com- pete for feed resources Native animal breeds are preferred by smallholders because of their adequate reproductive performance even under feed scarcity conditions Endo-parasite infestation is greater in exotic breeds than in native animals Utilisation of veterinary services for curative interventions is greater when im- proved breeds are used . . Crop-Animal Interactions . The main purpose of raising cattle is for draft, but the use of draft animals is declining Farmers do not efficiently use the excreta produced by their animals Farmers do not like dwarf sorghum genotypes, although they produce high grain yields, because of their poor stover production Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia90 . The number of cattle/buffaloes maintained in the farm is directly associated to the availability of crop residues Farmers raising animals with higher genetic potential have a larger area de- voted to the production of improved forages Supply, processing and marketing . . Middlemen are the most common means of cattle trading Farmers with better access to the market practice more intensive feeding sys- tems Lack of knowledge and facilities for processing/handling milk on-farm is the major constraint to increasing milk production Socioeconomic aspects . . . . Farmers give preference to crops when allocating labour The relative contribution of animals to the total farm gross income is higher in those farms better endowed in capital and land resources. Smallholders consider specialisation undesirable because it is perceived as too risky Household security and the nutritional status of children are enhanced by in- corporating milking animals into crop-animal smallholder systems Access to credit is a major constraint to increasing the animal component in the smallholdercrop-animal systems Smallholders are not maximizing profit generated by their activities on the farm, rather, they maximise utility or satisfaction of all farm family members . Gender issues Women are responsible for forage collection and distribution to the animals, therefore they are more open to improved forage technology issues Supplemental income associated to increases in animal production is dispro- portionately retained by the men of the household Women and children are the ones managing animal species, but very seldom do they have access to livestock extension services Women with young children tend to emphasise back-yard crop and animal production activities Households with older children at home and more labour upon which to draw may take on more labour-demanding activities Temporary or permanent migration may leave a high proportion of female headed-households, with less available labour and more limited access to re- sources for production Methodologies for analysis of household survey data 91 In the case of the household survey carried out by the CASREN project, intrinsic in the proposed methodology is an additional set of hypotheses to test, comparing the two villages sampled. As one village will serve as a "control" for project activities, and the other will have technological interventions applied to it, then the "non-written" hypothesis is that both are similar in baseline attributes. On the other hand, the research te;im could decide to test hypotheses based on the classes identified through the survey. For example, if in a given benchmark site there is more than one ethnic group, it could be of interest to define if this characteristic influences the practices applied at the farm level, the use of family labour or any other attribute. Also, it is possible to find farms that raise exclusively monogastrics, rumi- nants or both; and within the latter two, some may have cattle, buffaloes and small ruminants, and some hypothesis could be established based on the type of animals maintained in the farm. In general, the procedure used for testing hypotheses will be a function of: .Type of variable (categorical or continuous) .Number of classes (two or more than two) .Nature of the hypothesis Test of hypotheses with categorical data Many of the questions included in a household survey refer to a qualitative rather than a quantitative characteristic (e.g.,number of farmers using rice straw, number of households using hired labour), which after computation result in enumeration data, expressed as a discrete variable. To test hypotheses related to this type of attributes involving only two classes (e.g., two villages) it is suggested to use either the Normal Approximation Analysis or the Chi Square (X2) Criterion for a 2 x 2 or Fourfold Contingency Table (Steel and Torrie, 1980), as illustrated in Box I. The same procedure could be applied to a greater number of classes, but making independent comparisons each with one degree of freedom, or even grouping some of them, using the principle of the additivity of Chi Square(x2). Details on how to compute this can be found in any statistics textbook. Test of hypotheses with continuous data The analytical option to be used for the analysis of continuous data could be either a simple t-test in case of only two classes, or the Analysis ofVariance with an F- test, in case more than two classes are considered. The latter is also used when more complex treatment structures are evaluated ( e.g., comparisons among systems within sites). Least Square procedures are frequent]y used for statistica] ana]ysis ofhouseho]d survey data, but in many situations the data avai]ab]e may ]ead to biases in the estimation of parameters (Draper and Smith, 1981) or the tests of hypotheses (Leon- Ve]arde and Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia92 Box 1. Application of X2 for the analysis of categorical data with two classes Using the data shown in Table I, the research team wants to know ifboth districts differ in the proportion of farmers that uses rice straw for feeding cattle. X2=(nlln22 -nI2n21)2n.. = [(43)(l0)-(63)(14)]2(130) (106)(24)(57)(73)nl.n2.n3.n4 X2 = 2.2302 Inference: The estimated x:2 is smaller than tabular r with I df and a = 0.05 (3.84), but greater than x:2 with a = 0.25, therefore although villages were not different regarding the use of rice straw for cattle feeding at 5% probability, but they tend to be different at 25% probability . and Quiroz, 1994). Because of these reasons, it is highly recommended to look for the advice of a statistician during the Planning stage for the analysis and interpretation of data. The problems most commonly faced when analyzing data using Least Square procedures are the following: I. Lack of normality -One of the basic assumptions for the Analysis of Variance is that data is normally distributed, with mean equal to zero and variance equal to one. However, in many cases, the observations collected have a skewed distribution and the standard deviation greater ~ than the mean is not uncommon. This affects the test of significance of .the parameters, as well as the estimation of confidence intervals for the parameter estimates. Under these circumstances, it is recommended to use data transformations based on the analysis of the residuals. 2. Heterogeneity of variances -Another assumption in the Analysis of Variance is that all observations have a common variance, and the Ordinary Least Square procedure applied gives the same weight to each observation. Again, the solution to this problem is transformation of data based on the distribution of residuals. Methodologies for analysis of household survey data 93 3 4 Co-linearity -Many attributes used for characterisation could be correlated, and this can contribute to increase in the variance, and may change the magnitude and even the sign of a given parameter estimate. A means to detect if variables are correlated is to estimate the Correlation Matrix, and look for the association between each pair of variables, but some could not be detected by this procedure. For this reason it is recommended to run a Principal Components Analysis to develop new variables that integrate those that are correlated (Le6n- Velarde and Quiroz, 1994). This aspect is discussed later in this paper. Correlated errors -This problem is mostly faced in the case of dynamic surveys where data is repeatedly collected in the same units. Under these circumstances, variables tend to present correlation between their residual values, which in turn diminishes the precision of estimates, and could even invalidate any test of significance of the estimated parameters. This problem could be overcome by using Time Series Analysis and Generalised Least Squares (GLS) procedures. Multiple and partial regression and correlation analysis Household surveys include a wide range of variables, some of them could be correlated, and some could be used for the prediction of others. With the use of computer packages it is relatively easy to run multiple correlation analysis, and this is frequently a procedure applied when household survey data is analyzed. However, it must be clear that this analysis determines the degree of empirical association among a set of variables, therefore a high correlation value between two variables does not mean that there is a causality relationship between them (Gomez and Gomez, 1984). Multiple regression is also a procedure frequently used to analyze household survey data, but the research team needs to identify which could be adequate dependent and independent variables, and rationalise the basis for running the analyses. Stepwise procedures for Multiple Regression Analysis are commonly used for this purpose (Rawlings, 1988). The advantage of applying such techniques is that these could help to reduce the number of independent variables used for the prediction of a given dependent variable, dropping from the equation those that make a small contribution to the Coefficient of Determination (R2). However, it is necessary to re-run the analysis after dropping variables because the equation contains "partial regression coefficients" whose magnitude varies depending on the presence of other terms in the regression equation. Multivariate analyses Multivariate analyses are a tool for the classification and typification of households in the benchmark site, based on their commonalties regarding the set of variables included Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia94 in the questionnaire. The following procedures used in multivariate analysis are briefly discussed in this paper: Principal Components, Cluster and Discriminant Analyses. Principal component analysis is a procedure used to present a set of variables in terms of a smaller, more manageable set of linear combinations of the variables, which retains as much of the information of the original set as possible. It is also a tool for controlling co-linearity problems among different variables; which is in fact, a frequent problem in household survey data sets. Mathematically, principal components are linear combinations of the original variables (XI' X2' X3' Xn)' which can be represented as: YI= al X1+a2X2+ +an Xn Each principal component has a maximum variance and satisfies the condition of n Ea~ =I In other words, the sum of the coefficients equals one ( 1 ), and all principal components are orthogonal among them. In short, the correlation between each pair of principal components is zero. Although "p" principal components are required to reproduce the total variability, often a small number of them ("k" principal components) can account for much of this variability .If so, there is almost as much information in the "k" components as there is in the original "p" variables. The "k" principal components can then replace the initial "p" variables, and the original data set, consisting of"n" measurements on "p" variables, is reduced to one consisting of"n" measurements of"k" principal components (Johnson and Wichem, 1982). The latter are the ones used for subsequent analysis, such as Cluster Analysis. Tables 3 and 4 show an example of the results obtained after running Principal Component Analysis to a household survey data set resulting from a research project involving dual-purpose cattle production systems (Anderson and Santos, 1997). In Table 3 are shown the Eigen values for the Principal Components identified, as well as the relative contribution of each to account for the total variability .Notice that the first two principal components explain 99.86% of the total variability, whereas the contribution of the others is negligible; therefore, it is advisable to concentrate the attention just on the first two principal components. 95Methodologies for analysis of household survey data Table 3. Eigen values of the covariance matrix for a set of livestock farms surveyed (Anderson and Santos, 1997). Principal Component Eigen Value Difference Proportion Cumulative PCl eG~ PC3 PC4 PC5 PC6 PC7 PC8 2234551 609914 3999 67 1624637 605915 2933 66 0 0 0 0 0.7845 0;2141 0.0013 0.0001 0.0000 0.0000 0.0000 0.0000 0.7845 0.9986 0.9999 1.0000 1.0000 1.0000 1.0000 1.0000 o o o Table 4 lists the variables included in the first five principal components, and highlight those that make the greater contribution to each principal component. Based on the list, the most important attributes to discriminate among farms are stocking rate, use of concentrates, animal records and seasonal mating (included in PCl), as well as the total farm area and the one in pastures, the use of technical assistance and off-farm work (included in PC2). As indicated before, new variables can be estimated based on the coefficient assigned to the elements of each principal component. Cluster analysis Clustering is the grouping of objects based on their similarities. In the case of household surveys, clustering of farms is a mechanism to identify and characterise groups of households that are similar among them, but quite different from those belonging to other clusters. Each group of farms within a cluster may constitute a recommendation domain, so that a specific technological recommendation can fit properly to all members of a cluster. Variables that are largely the same for all households have little clustering power, whereas those manifesting differences from one household to other are more likely to induce strong distribution. Cluster groupings depend on the selection of variables. Cluster groups could be poorly defined when a relevant variable is ignored, or by including a non-relevant one. Also the scale of the variables is an important factor that could influence the cluster composition (i.e., variables with a larger scale will dominate the conformation of clusters). Therefore, it is recommended to standardise all variables before running cluster analysis to give a uniform importance to each variable. Standardisation is possible through subtracting their means and dividing by their standard error. 96 Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia Table 4. Description of the first five Eigen vectors related to each principal component in a set of livestock farms (Anderson and Santos. 1997). PCSVariable PCl PC2 PC3 PC4 0.16699 -0.39645 0.18731 0.15736 0.26451 0.18459 0.41134 -0.16480 0.19577 -0.48440 -0.29480 0.49254 -0.30850 0.10454 -0.17892 0.20298 0.27510 -0.20863 -0.00169 0.29756 Total land, ha Forage Area, ha Stocking rate, heads/ha Use of concentrates Individual animal records Seasonal mating Rotational grazing Uses irrigation Uses family labour Receive technical assistance Off-farm work Occasionally hire labour -0.17869. 0.29751 0.02124 0.01567 0.14434 0.20403 -0.04248 0.04918 -0.43801 0.364~7 0.15534 0;57927 -0.33960 -0.36491 0.00142 0.24220 -0.01442 0.07172 0.15 0.24 0.22 0.07284 -0.16019 0.20407 0.28331 There are different procedures (algorithms ) used to measure the distance between classes, which is the basis for defining clusters, but based on previous experiences it is suggested that the Ward procedure (SAS, 1998) be used. This estimates the variance within each cluster, and tries to minimise it. One important decision in cluster analysis is how to define the adequate number of groups (clusters). The distance among groups becomes greater as the size of clusters increases (Fig. 2), but also increases the variability within clusters. For each variable it is possible to estimate the Sum of Squares within groups (W k)' as well as the Total Sum of Squares (T k)' then the multiple correlation coefficient (R2) can be estimated as: n tWk n- tTk R2 = 1 - 97Methodologies for analysis of household survey data -0.25204 -0.14446 0.15269 -0.18224 0.27196 0.21721 0.07029 371 952 685 Figure 2. Average distance between the last two conglomerates formed. as a function of the number of conglomerates formed. The multiple correlation coefficient tends to decrease with the reduction on the number of clusters; therefore, the researcher needs to make a decision on how much reduction in R2 is he/she prepared to accept. In the case of the example shown in Figure 3, it seems reasonable to accept from 3 to 5 clusters, as there is an important decline in R2, when only two clusters are considered. When cluster analysis is run, the output includes the grouping of units in the different clusters, as well as the statistics for the different attributes used for clustering. It is advisable to run analyses of variance for the different variables, using clusters as a class variable, in order to detect potential differences among groups. 9 8 7 6 5 4 Number of clusters 3 2 Figure 3. Changes in R2 as number of clusters varies. An example of the results of the application of cluster analysis to the information gathered in a household survey involving 100 farms practicing crop-animal systems in the Red River Basin of Vietnam is shown in Table 5. Notice that six clusters were identified, two grouping high-income households, with and without special activities (e.g., fruits, garlic production, cattle production, fish farming, and income from off- 98 Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia farm labour). Another two clusters were composed of medium-income households, with and without special activities, and had no cattle production activities. A fifth cluster grouped alllow-income farms, which do not have special activities. Finally, a sixth cluster covered those farms with the highest income and was strongly influenced by commercial swine operation. Considering the components of the farming systems represented by the different clusters, as well as the level of incQme, and eventually their availability of resources, it is clear that the same technological options not necessarily apply to all farms (they correspond to different recommendation domains). A more detailed analysis of other attributes, as well as farmers' expectations and concerns within each cluster will give more hints on the type of interventions to be proposed. Another important decision the research group should take is which cluster(s) to target in their research efforts. Table 5. Clusters identified in a household survey involving crop-animal systems practiced by smallholders in the Red River Basin of Vietnam (Chan-Dung. 1993). Farm Size Number of Farms Total Income Rice Income Income from Pigs Special ActivitiesCluster 3600 Fruits, garlic, cattle, fish- fanning, income from off-fann labour 3686 1174 780I. High Income, Special activities 13 731 3830 None13 3685 1313High Income, Basic Activities 521 3340 Onions, income from off-farm labour 17 3421 11043. Medium Income, Special Activities 2870 None861 4974. Medium Income, Basic Activities 22 2551 None669 379 237033 18755. Low Income 1093 975 34602 4782 Intensive pig production 6. Exceptions 99Methodologies for analysis of household survey data Discriminant analysis Once clusters have been defined with a sample of units, other units (i.e., farms) not used for the analysis can be assigned to one of the cluster groups, based on the similarities they have with respect to attributes that characterise each cluster. As each unit may have attributes in common with more than one cluster, then it is possible through mathematical procedures (i.e., Mahalonobis distance estimates) to elucidate to which group each unit is more likely to belong, with a probability factor associated to this decision. The greater the probability level, the unit is a more typical representative of the cluster. Table 6 shows an example of the outputs of discriminant analysis for a set of farms. Notice that Farm No.2 has been assigned to Cluster 5, and this is the most typical among the ones listed, as it has the highest probability to pertain to that cluster . Table 6. Example of the output of discriminant analysis classifying farms within clusters. Fann Number Assigned to Cluster Number Probability of belonging to a given cluster 1 2 3 4 5 1 2 3 4 5 2 0.11 0.05 0.22 0.22 0.15 0.42 0.01 0.10 0.11 0.40 0.17 0.12 0.53 0.13 4 0.17 0.14 0.05 0.48n 100 Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia REFERENCES Anderson S. and Santos J. 1997. Monitoreo de sistemas de doble proposito a nivel de la vaca, el hato y el productor. In: Conceptos y Metodologias de Investigacion en Fincas con Sistemas de Produccion de Doble Proposito (Eds. Lascano C E and Holmann F). CIA T (Centro Internacional de Agricultiura Tropical). Consorcio Tropileche. pp. 191-212. Chan-Dung L.T. 1993. Analysis of Activities of Farm Households in a Commune in Vietnam using Multivariate Statistical Techniques. M.S. Thesis (Statistics). UPLB, Los Baiios, Philippines. 206 pp. Gomez K.A. and Gomez A.A. 1984. Statistical procedures for agricultural research. 2nd Ed. IRRI (International Rice Research Institute). Wiley, New York, U.S.A. 680 pp. Jabbar M. A., Tambi E. and Mullins G. 1997. A Methodology for Characterizing Dairy Market Systems. ILRI (International Livestock Research Institute). Market Oriented Smallholder Dairy Research Working Document 3. Nairobi, Kenya. 62pp. Johnson R.A. and Winchern D. W .1982. Applied multivariate statistical analysis. Prentice Hall, Englewood Cliff, New Jersey, USA. 594 pp. Lapar M.L. 1999. Crop-animal systems research in Southeast Asia: data collection for benchmark site characterization. In: Proc. Planning Worhhop on the Crop-Animal Systems Project. Held at IRRI, Los Baiios, Philippines on 1-4 June 1999. ILRI (International Livestock Research Institute), Los Baiios, Philippines. pp. 16-19. Leon-Velarde C. and Quiroz R. 1994. Anillisis de SisJemas Agropecuarios; Uso de Metodos Bio-matemilticos. CIRNMA (Centro de Investigacion en Recursos Naturales y medio Ambiente). La Paz, Bolivia. 238 pp. Mullins G., Rey B., Nokoe S. and Shapiro B. 1994. A research methodology for characterizing dairy product consumption systems. ILCA (International Livestock Centre for Africa). Market Oriented Smallholder Dairy Research Working Document 2. Addis Ababa, Ethiopia. 40 pp. Rey B., Thorpe W ., Smith J., Shapiro B., Osuji P ., Mullins G. and Agyemang K. 1993. Improvement of dairy production to satisfy the growing consumer demand in SubSaharan Africa; A Conceptual Frameworkfor Research. ILCA (International Livestock Centre for Africa). Market Oriented Smallholder Dairy Research Working Document I. Addis Ababa, Ethiopia. 13 pp. Rawlings J.O. 1988. Applied Regression Analysis; A Research Tool. Wadsworth and Brooke-Cole Statistical/Probability Series. Davis, CA. U.S.A. 553 pp. SAS. 1988. SAS/ST A T'M User's Guide; Release 6.03 Edition. SAS Institute Inc. Cary , North Carolina, USA. 1028 pp. Steel R.G.D. and Torrie J .H. 1980. Principles and procedures of statistics: a biometrical approach. McGraw Hill, New York, USA. 633 pp. DISCUSSIONS ON RESOURCE PAPER PRESENTATIONS Hugo Li-Pun's presentation emphasised the positive aspects of the on-going project: .The surveys .Participatory approaches .The identification of common constraints markets, services (constraints outside the farm) nutrition (constraints within the farm) .Diversity -can be an advantage, e.g., exposure to a range of different situations and find alternatives subsistence-oriented systems where crops are not important and livestock can be important (China) systems where crops may be important and livestock plays a complementary role. The challenges include: .Regional analysis representativity of the BMS integration, i.e., how information that have been generated can be integrated .National analysis secondary information cluster analysis (within sample variations, for targeting interventions) interpretation, critical for their implications on interventions gender variable needs much better interpretation, e.g., women may playa major role in some species like the small species while the males do in larger species .Contribution to poverty reduction Pigs and poultry Small ruminants .Ex ante assessment ILRI has to play an important role in providing support to national scientists to undertake these activities Need for practical training on ex ante assessment, e.g., hands-on training .Capacity building -an important element and we need to organise activities along this area Discussions centered on some of these aspects and approaches, transfer of technology, farming systems research and ecoregional research, and impact assessment, and the need to build convincing arguments and documenting cases concerning the role and contribution of animals. 102 Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia Mohammad Jabbar's presentation on policy options evoked comments from Vietnam about its relevance to livestock development and marketing issues concerning smallholder production, because taste preference of consumers have changed due to the changing economic conditions. Questions were also raised about contract farming but evidently not many linkages exist in the product market. However, linkages exist in the feed markets, e.g. advance payment arrangements, differential pricing. The intense competition among feed sellers to get the largest market share induces them to provide incentives to attract more feed buyers. It was suggested that it would be more realistic to include two to three countries in the ASEAN region to get quick answers, instead of just focusing on one country . On Lucy Lapar's presentation on the GIS applications, there was a suggestion that in respect of crop residues, it will be useful to show whether there have been substitutions in the relative share of their availability over time. In the paper jointly authored by C. Devendra and D. Pezo, emphasis was made about the overreaching importance of feed production and utilisation technologies, which was the main thrust in the present project. It was also mentioned that modeling and simulation methods in Vietnam are very weak. Policy should be put in place to drive the process of technology development improvement. Concerning the paper on ex-ante assessment by D. Pezo, it was suggested that applying the several methods for the economical evaluation of technological interventions will take too much time, and these should be implemented as soon as possible after finishing the household survey. In response, it was indicated that in most cases, partial budget analysis is a good choice. However, in many situations, it would be desirable to complement it with cash flow and labour profile analysis. On the paper on household survey data analysis that was also presented by D . Pezo, there was a query on the use of SAS for data analysis and whether it was the only option. It was suggested that any statistical package containing the procedures described in the presentation could be used. Social sciences people prefer the use of Statistical Package for the Social Sciences (SPSS). 103Oiscussions on resource paper presentations SESSION 2 COUNTRY PRESENT A TIONS HOUSEHOLD SURVEY RESULTS AND IMPLICATIONS FOR FUTURE RESEARCH IN CHINA z. Kaidian Yunnan Beef Cattle and Pasture Research Center C. Wenyang, S. Honglu Animal Husbandry Bureau ofNanjian County L. Jianping Chinese Academy of Agriculture Science ABSTRACT The results of household surveys in four villages in Bixi Xiang in Nanjian county are given, involving a total of 182 households. The average number of persons per household is 4.7. The land area for cropping is very low (0.06 haper capita). The main crops are maize, wheat, potatoes, beans and barley. All households kept beef cattle, buffaloes, goats, pigs and chickens. The productivity of animals is low and is related to a low level of nutrition and low reproductive rate. This is further exacerbated by high mortality rates of 16-41% in piglets and chicks. A variety of crop-animal interactions exist and all farmers used all the manure produced as fertiliser. The proposed interventions include feed production and improved nutrition and animal health, and will involve both on-station and on-farm work. INTRODUCTION Following the selection of the BMS and using a questionnaire provided by ILRI, a survey of 100 households was completed to characterise the site in detail. The sections below present the results. THE SITE Location The 100 households surveyed are distributed within four villages of Bixi Xiang, which is located at the southern part of Nanjian county (Figure I ). The Xiang headquarters is about 40 kilometres from the county town and about 400 kilometres from Kunming. It is located between 100005' -100020' E, 24051' -24058' N. The total area is 124.88 km2 with a population of 17,368. 107Household survey results and Implications for future research in China The four villages surveyed are along the national road going through the Xiang, the distance between the first one and the last one is about 25 kilometres. The altitude of the lowest village is about 1,200 metres and that of the highest village is about 1,800 metres. To Kunming, 401 km ! ITo Cali 161 km .."-- To county town, 41 km /,- ~ ~ '""-.~~ To Leqiu Xiang, y 18 km To Lancang Xiang, 20 km To Yongcui Xiang ? 35 km Figure 1 Location of the BMS. General Profile of the Four Villages Amiao There are 35 households in this village; all the people are ofHan nationality .The population is 162 and 87 of this are working labourers. By the end of last year, there were 71 cattle (38 cows) and 142 pigs (21 sows) in the village. The total area is 250 hectares and the total cropping land is 20 hectares. The maize grain yield is 52,240 kg. The village has some native pasture and an artificial insemination site (AI) nearby. The altitude of this village is around 1,200 meters. The topography is mainly rainfed upland, with a little rice paddy. Crop residues and agriculture by-products are fed to animals and manure is collected and applied to cropping and vegetable areas. Huilongshan There are 51 households in this village and all the people are ofHan nationality . The population is 320; 197 people are working labourers. In October 1999, there were 36 cattle (12 cows) and 350 pigs (75 sows) in the village. The total area is 450 hectares, with a total cropland area of 35 hectares. The maize grain yield is 958,510 kg. The village does not have any grazing land because of reforestation work but has an AI site nearby. This village is closer to the Xiang headquarters, so its socio-economic situation is better compared with the other three villages. 108 Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia The area is a rainfed upland site. Crop residues and agriculture by-products are fed to animals, and manure is collected and applied to cropping areas. A few households use animal excreta to produce bio-gas for lighting and cooking. Yanzitou There are 63 households in the village and all the people are also ofHan nationality . The population is 275, 177 of which are working labourers. In October 1999, there were 78 cattle (45 cows), 278 pigs (75 sows), and 68 goats in the village. The total area is 57 hectares and the total cropland is about 20 hectares. The maize grain yield is 79,398 kg. The village has some grazing land. The village is a bit far away from the AI site. The topography is purely rainfed upland. .Crop residues and agriculture by-products are fed to animals. Manure is collected and applied to the cropping area. A few households also use animal excreta to produce bio-gas for lighting and cooking. Guoditang There are 33 households in the village. All the residents are Yi, a minority nationality .The population is 146 with 87 people who are working labourers. In October 1999, there were 52 cattle (25 cows), 187 pigs (55 sows), and 78 goats in the village. The total area is 3 krn2, the total cropping land is about 11 hectares. The maize grain yield is 46,906 kg. The village has some grazing land but is a bit far away from the AI site. The topography is purely rainfed upland. Crop residues and agriculture by-products are fed to animals and manure is collected and applied to the cropping area. A few households also use animal excreta for producing biogas for lighting and cooking. CONDUCT OF THE SURVEY The questionnaire prepared and provided by ILRI was first translated into Chinese. A team consisting of members from CAAS, Yunnan Agriculture University, Animal Husbandry Bureau ofNanjian county, Bixi Xiang Animal Husbandry and Veterinary station and Yunnan Beef Cattle and Pasture Research Center was established with Mr . Zhao Kaidian as the team leader. The survey was divided into three stages. Stage One Team members convened and went through the questionnaire twice to determine what information was easy, difficult or impossible to get. Then, some discussion was held on what score should be given to some qualitative indicators such as soil fertility, Household survey results and implications for future research in China 109 feed stuff quality, etc. to make the results more realistic based on the same score standards and disciplinary knowledge. Stage Two A simulated interview was carried out with the team leader acting as the interviewer . One team member acted as a farmer while the others played an audience role. Afterward, the team had some more discussion. Then, one farmer was invited to Bixi Xiang Animal Husbandry and Veterinary station for interview, with the team leader as the interviewer while the other team members served as audience. Stage Three The team members were divided into groups. Each group consisted of two members with one of them from Bixi Xiang Animal Husbandry and Veterinary station. The groups went to the first village in the morning to interview three households and came back to the station for a discussion session and exchange of information. During the discussion, most members felt that the two-person-group system did not work well as some information were missed out with one member asking and the other, recording. It was also observed that the method was inefficient in terms of time. As a result, team members worked individually and on most occasions, three or four interviews were going on at the same time in one household. For some collective information for each village, village accountants and village heads were interviewed as key informants. During late October and early November, the farmers are not very busy. When there are some visitors coming to someone's house, the neighbours just automatically come to him. Taking these opportunities, the team also conducted some group discussions. RESULTS General Household Characteristics Among the 100 households surveyed, the average number of people per household is 4.7. The range is one to eight. When clustered into four groups, group one ( 1-3 people) takes up 12%, group two (4-5 people) 61 %, group three (6- 7 people) 24%, and group four (over 8 people) 3%. For the age structure, 8.6 % is under the age of six, 11.2% is between 6 -16, 67.5% is between 16- 60, and 12.7% is over 60 years old. The age structure shows that the working labourers in these villages 'are more than enough and there was quite a high proportion of aged people (over 60) who still do a lot in helping the households. 110 Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia The education level is relatively good. Apart from the school attendance, most people also received some training on cropping and raising animals. For training in cropping, 74 households were trained in plastic sheet covering and high-nutrients-bag transplanting of maize and tobacco; the rest did not receive any training. For raising animals, 49 households have received some training and 51 households did not. Not much difference existed in the education level between Ban and Yi nationalities. Among the respondents, illiteracy rate was low (only 3%). Seven percent (7%) attended elementary school grades 1- 3, 57% finished elementary school (grade 6), and 30% finished secondary education. Crop Production The main crops in the BMS are maize, wheat, potatoes, beans and barley; only one village produces a little rice. Two crops are grown each year and farmers maintain a small patch specially planted to vegetables. The cropping patterns are as follows: maize with/without soybean or- kidney bean r::> wheat or barley The crop production patterns are summarised in Table I Table 1. Crop production. Second growing seasonFirst growing season Quantity (kg) Average yield (kg/ha) Parcel Quantity (kg) Average yield (kg/ha) Area (ha} Area (ha) 53,657 36,680 22,820 9,665 4,669.8 4,506.1 3,841.7 4,956.4 10.38 31,385 3,023.6 3,447.3 2,934.5 2,900.4 Parcel I Parcel 2 Parcel 3 Parcel 4 11.49 7.31 5.57 2.26 25,200 16,345 6,555 8.14 5.94 1.95 119,822 25.52 79,485Total* 27.52 *Total crop production: first growing season + second growing season :; 119,822 + 79,845:; 199,307 kg Household survey results and implications for future research in China 111 Animal Inventory The animal numbers in this area are relatively stable except those for pigs and chickens. The main purpose ofkeeping beef cattle is for drafting, so, there is not much slaughtering and selling. People do like eating beef, but there is not enough supply and the marketing system is not functioning well. Among the total 137 head, there were only 60 cows (Table 2). The reproduction is low. Within the last 12 month period, 60 cows gave birth to only 27 calves. There may be a need to study this aspect . Table 2. Number of beef cattle. Item Adult Male Adult Female Young Animal Calves 39 27 n.a. 24 1 13 O 8 60 n.a. 12 2 20 0 39 31 o.a. 12 2 O O 8 29 27 5 o o At start Current Born Sold Died Bought Slaughtered n.a. = Not applicable. The purpose of keeping beef buffaloes is the same as that ofbeef cattle. Because of the topography and physical environment, buffalo is not a main species in the area, and the total number is decreasing (Table 3). Table 3. Number of buffaloes. Item Adult Male Adult Female Young Animal Calves 4 3 o.a. 2 O 1 O 11 7 n.a. 4 O O O 1 2 O 1 3 O O O o.a. O 0 0 O At start Current Born Sold Died Bought Slaughtered n.a. = Not applicable. 112 Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia Goats are one of the main ruminant species in the county as a whole, but their distribution is uneven. Some villages keep more, some villages keep less. Most villages belonged to the latter. There are only four households keeping goats. One important finding in the survey is that the mortality is too high. The cause(s) of this high death rate is unknown -whether it is due to poor nutrition, disease or poor management. The buying and selling of goats is more active than in cattle and buffaloes (Table 4). Table 4. Number of goats. Item Adult Male Adult Female Young Animals Kids 7 7 n.a. O O O O 51 27 n.a. 10 15 2 0 11 6 n.a. O 5 O O 21 32 O 12 O O At start Current Born Sold Died Buy in Slaughtered n.a. = Not applicable. Within the four villages surveyed, the dominant species is the pigs, which is also the main source for home meat consumption and cash income. Among the lOO households, 98 households kept pigs and the rest did not. Sixty five (65) households slaughtered pigs for home consumption and more than 80% of households sold piglets and/or fattened pigs for cash income. The average herd size is 8.1 head (ranging from 1 to over 40). Fifty (50) households keep sows for producing piglets for sale. The mortality is very high, especially among the piglets and young animals (Table 5). Table 5. Number of pigs. Adult Female Young Animals PigletsItem Adult Male 84 76 n.a. 10 3 6 3 394 235 n.a. 232 22 130 53 143 451 805 406 129 59 0 38 32 D.a. 22 O 19 II At start Current Born Sold Died Buy in Slaughtered n.a. = Not applicable. 113Household survey results and implications for future research in China Raising chickens is a popular enterprise using a special breed in the area. The flow of chickens (born, buying and selling, gift and slaughter) is very active {Table 6). The mortality of chickens is extremely high, over 40%. The causes of this are poor facilities and inadequate veterinary service. People normally consume or use eggs produced on their farm for reproduction. Table 6. Number of chickens. Item Number 2,078 1,813 3,118 202 62 953 124 1,270 1,320 At start Current Born Bought in Gift-in Sold Gift-out Died Slaughtered Switching of species and the enterprise is common but there are quite a few households that have either started or stopped raising beef cattle and buffaloes (Table 7). The reasons for this are complex. Table 7. Switching of species and enterprises. Number of Households Species Start Present Pigs Beef cattle Buffaloes Goats Sheep Chickens II 6 2 I I 11 8 O 1 114 Improving the contribution of livestock to crop-anlinal systems in rainfed areas in Southeast Asia Animal Management and Productivity In Bixi Xiang, the proportion of beef cows to bulls is low. Among the total 147 head, there are only 60 females that can be bred. This may relate to the fact that some people keep cattle as draft animals and that the native cattle is small and the female cannot fulfil the task of a draft animal. The age at first parturition varies from 24 to 48 months. For weaning weight and age, people normally do not wean their calves at any fixed age. They just leave the calf with its mother and let it naturally weaned. The estimated figures that farmers gave us during the surveyare from 6 to 12 months for weaning age and 70 (native cattle) to 180 kg (maybe the second generation of crossbreeding) for weaning weight. -The marketing age and weights are also not fixed, only one farmer says that he can get one calf every year and he weans the calf at six months old and sells it soon after at a weight around 180 kg. Most farmers sell their cattle as culls or for some other reasons. There are two main types of cattle management. One type is cut-and-carry and pen-fed, no grazing. The other is a combination of grazing, cut-and-carry and pen-fed. With the latter, people graze their cattle during the day, with cut-and-carry and supplementary feeding (including crop residues, maize and salt) during the night. The average amount of maize supplemented is 0.805 kg, ranging from 0.1 to 3 kg The management of buffaloes is the same as that of cattle. The animal health practice for cattle and buffaloes is about the same. No vaccination is given and only a small proportion is deworming. Pigs are the main animal species in the BMS and 98% of households raise them. The average performance is summarised in Table 8. Table 8. Cattle and pig production. Cattle Pigs Item ModeRange Average Mode Range Average 24-48 40-10 33.3 96.7 36 8-36 6-25 14.6 12.9 12 10 1-14 1.5-2 6-180 2-18 7.6 8 5-12 120-300 7.3 ~ Age at first month parturition (mth) Weaning weight (kg) Litter size of last parturition Weaning age (mth) Selling weight (kg) Selling age (mth) 115Household survey results and implications for future research in China The management of pigs is variable., For piglets, about 5% of the farmers graze their pigs with cattle or goats, and 95% have their pigs pen-fed. People buy commercial feeds to feed their pigs at 1-3 months of age and also at the finishing stage. Most people use kitchen refuse to feed their pigs. Among the 55 households keeping breeding sows, 24 use AI and 31 use natural service including introduced and native breeds. Ninety one (91) households vaccinate their pigs, one does not. The management of chickens is more or less similar to that of pigs (Table 9). Table 9. Percentage of village households practising pen-fed and scavenging systems for pigs and chickens. Pigs Chickens Village Dry Season Wet Season ScaVen- ging Dry Season Wet Season Scaven- gmg Pen- fed Pen- fed Scaven- ging Pen- fed Scaven- ging Pen- fed 4.2 20 95.8 80 8 20 92 80 45.8 58.3 54.2 41.7 40 54.2 60 45.8 o 0 6.1 100 loo 93.9 o 0 1.1 100 loo 92.9 50 96.6 62.1 50 4.3 37.9 33.3 86.4 52.6 66.7 13.6 47.4 Amiao Yanzitou Huilong- shan Guoditang Average Economics of Animal Enterprises For daily management of the animals, the role ofhousehold members is not fixed. More commonly, it is determined by specific activities of the household on that day. Men tend to do heavy work while women, children and aged people tend to do easy work (Table 10). Generally speaking, the input and output of the animal industry in the area are low and this is probably related to the low efficiency of production (Tables 11-13). Table 10. Number of persons per family performing various farm chores. ChickensChore* Cattle Pigs Grazing Male Female Child Feeding.. Male Female Child Vaccination Male Female Marketing Male Female Cleaning manure Male Female Elders 44 46 3 D.a. D.a. D.a. n.a. n.a. n.a. 18 40 14 84 55 41 5 79 20 49 49 ~ 73 26 12 86 65 27 70 18 7 70 18 7 n.a. = Not applicable; -= No data .Column totals less than 100 are due to some families who did not feed their animals at the time of the survey ..Total is more than 100 and is accounted for by some people who performed two jobs, e.g. cattle grazing and feeding Table 11. Cost of production per head excluding feeds, (Yuan) Expense Item Animal Raised Medicine Vaccine Drench De- worming AI Salt Ropes 648 451 962.7 ,769.5 0733.5 4,583.5 1248 105 825.8 588 35 393 74 59.5 430 n.a. o.a. n.a.n.a. D.a. Cattle Pigs Chickens D.a. = Not applicable 117Household survey results and implications for future research in China Table 12. Marketing and consumption of pigs and chickens 71 1352 8220 1.8 115.5 2688 29310 1429 6.26 15.35 Butcher Middleman Meats produced on fann* Number of animals slaughtered Slaughter weight (kg/head) Amount consumed (kg) Amount sold (kg) Selling price per kg (yuan) Major buyer Weekly 2.1 10.5 Butcher/ market D.a, D.a, D.a, D.a, Meat purchased** How frequent/y purchased Amount purchased (kg) Price per kg (yuan) Source ofmeat purchased *No meat was produced on farm from cattle, buffaloes, goats and sheep ** No meat from cattle, buffalo, goat and sheep meat was purchased n..a. = Not applicable Table 13. Eggs produced on farm Item Quantity 401 627 561 321 0.5 596 0.42 Number of laying birds Number of eggs produced last month Number consumed Number sold* Selling price (yuan) Eggs purchased last week Price paid (yuan) .Major buyer is the consumer Landholding The land in China belongs to the government and collectives and the farmers who manage the land do not own it. The cropping land is very limited, just 0.058 hectare per capita. Each household has several pieces of various sizes. The distance from the homestead is not very far, but transportation is very difficult. Soil fertility is generally low and people use manure and chemical fertilisers (Table 14). Table 14. land use. Area (ha) Distance from home (kID) Fertility Parcel Range RangeAverage Range Average Average Parcel I Parcel 2 Parcel3 Parcel 4 0.21 0.077 0.067 0.058 0.01-0.27 0.01-0.2 0.02-0.33 0.03-0.13 0.5 0.51 0.53 0.44 0.01 -2 0.01 -1.5 0.01 -2 0.01- 1.5 2.78 2.83 2.88 2.86 2-3 2-3 2-3 2-3 Crop-animal Interactions Cropping and animal raising are closely interrelated. The survey results of crop- animal interactions are shown in Figure 2. All farmers who raise animals use 100% of the manure as fertiliser. Eighty (80) households use maize straw as feed, one household uses all the straw for other purposes. Other households use 40 -80% of the maize straw as feed and 20 -60% for other purposes. For wheat straw and bean straw, 100% is used as feed and nothing is burnt. A few households use manure to make biogas for cooking and light. Figure 2. Crop-animal interactions. 119Household survey results and implications for future research in China IMPLICATIONS OF THE SURVEY RESULTS The survey results indicate the following: . . . There are surplus working labourers in the BMS and people can participate to increase productivity; The natural resources are very limited, including cropping land, grazing land and feeds. The area of cropping land per capita is low, only 0.058 ha/person. This means that people have to concentrate more effort and input per unit area; The productivity ofanimal is low. The reasons may be due to: 1) Low nutrition level. Most farmers just feed crop residues to their cattle and buffaloes, or with a small amount of maize supplement, especially during the dry season. This type of feeding system cannot meet the nutritional requirements of animals. It is even worse with calves, as indicated by the weaning weight estimates. Some households do not have enough straws to feed their animals. 2) Low reproduction. In beef cattle, during the 12 month period surveyed, 60 head of breeding cows only gave birth to 27 calves (including both from AI and natural service). The reproduction rate is just 45%. This may also, to some extent, relate to low nutrition level and longer calving interval. " The mortality rates are too high and animal health procedures are not carried out properly, even though we have a veterinary station at the Bixi Xiang level. For chickens, the high death rate may relate to poor warm-keep facility for use immediately after hatching. For cattle, goats and buffaloes, the vaccination and drenching service is very poor and this also may relate to slow growth of young animals. The mortality rate ofpiglets is as high as 16% and that for chicks is 40.7%. All these cause huge losses to the farmers and contribute to big wastage of natural resources. . EXPERIENCES AND LESSONS LEARNT One suggestion or observation from the questionnaire of this survey is that, to quite a few questions, the answer should be the combination of several possibilities. For example, in taking manure out of the shed, people often do it with all able members of the household, including male, female and aged people. The biggest lesson that was learnt from this survey is that the death rates of animals are much higher than what was estimated before. Some farmers are good at animal- keeping but others are not. CROP-ANIMAL SYSTEMS RESEARCH IN THE PHILIPPINES E. C. Villar, E. L. F. Lanting, D. C. Cardenas, A. P. Alo, V. M. C. de Roma Philippine Council for Agriculture, Forestry and Natural Resources Research and Development, Los Balios, Laguna, Philippines N. E. dela Cruz Central Luzon State University, Mulioz, Nueva Ecija A. M. Rebugio Pangasinan State University, Lingayen, Pangasinan ABSTRACT Two adjacent sites were chosen that were typical of the prevalent lowland rainfed- based crop-livestock farming systems in the Philippines. These were Don Montano in Umingan, Pangasinan, chosen as the focal village where technological interventions will be introduced and Sta. Maria, also in Umingan, as the control site. The selection was based on the agricultural potential of the area, importance of livestock in terms of its contribution to GDP, income share, productivity gaps, resource degradation, apparent poverty, and accessibility. Various participatory rural appraisal techniques were employed to select the sites from the region down to the village level. A benchmark survey was then conducted using the ILRI-developed questionnaire. To validate the survey, an in-depth inquiry on matters related to farming was made using story mapping and focus group discussion (FGD). The average household size was six and that of landholding was 1.3 ha. Multiple cropping systems were common involving rice, corn, mungbeanand vegetables. Farmers raised buffaloes, goats, pigs, chickens and ducks. About 32% of the farmers used crop residues as feeds, and between 50-64.7% collected the manure for use as fertiliser. A number of bio-physical, animal and crop problems, and socio-economic issues were identified as major constraints to production that contributed to low farm productivity .Proposed interventions were developed. INTRODUCTION In the process of selecting the benchmark sites (BMS), a fann household survey was conducted with the following objectives: (I) to generate baseline data on the selected sites for future impact assessment studies, and (2) to identify local fanning constraints that could provide valuable direction/infonnation on the appropriate intended technology-based interventions for the area. Crop-animal systems research in the Philippines 121 BMS Selection Process The selection of the Philippine BMS has been an arduous process primarily because the Team exercised caution in selecting the site representative of the larger rainfed- farming domain in the Philippines. Extra care has been exerted, as the Team's experiences have shown that the wrong choice of project site, institutional linkages and farmer participants can spell the difference between project success and failure. Hence keen attention was given to ensure that majority of the elements in the selection criteria were met. In selecting the sites, the Team went through a series of consultations with key officials of the Department of Agriculture (DA), local government units and village leaders. In most of these visits, ILRI scientists were with the Philippine Team and they were witnesses to the difficult process of selection. Through rapid rural appraisal (RRA) these interviews were validated until finally Don Montano in Umingan, Pangasinan was chosen as the focal village where technological interventions will be introduced. To serve as control site, Sta. Maria, another village in Umingan, Pangasinan was chosen. Figure 1 gives the relative location of the chosen site vis-8-vis the larger research domains. The results of this RRA and the initial series of consultations were presented during the steering committee meeting held at IRRI, Philippines on June 2, 1999. After this, the survey instrument for the BMS characterisation was prepared in Englishjointly by the Philippine Team and ILRI scientists. The interview schedule consisted of the farming characteristics, livestock inventory , cropping pattern and productivity, animal health practices, feed resource utilisation, land holding, labour utilisation, and problems on crop-livestock farming, among others. Initially, the instrument was pre-tested among farmers in Barangay (village) Mayuro, Rosario, Batangas, again with a group ofILRI scientists. After the pre-testing, it was finalised and farmed out to all collaborators ofCASREN. In the meantime, the finalised questionnaire was translated into the local dialect, l1ocano, and once again pre-tested among l1ocano crop-animal farmers to determine ease of question comprehension, acceptability and reactions of farmers to such an interview. A one-day training of the interviewers was conducted to familiarise them on the needed information, definition of terms and how questions should be asked to generate the desired information. METHODOLOGY From the initial RRA conducted, it was learned that the focal site has 329 households, 90% of whom are involved in farming. Of the 90%, half are also into 122 Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia Figure 1. Location of the focal village relative to the larger research domains. Crop-animal systems research in the Philippines 123 livestock production. Using Slovin's formula' (Dulay, 1997) for determining sample size and pegging the margin of sampling error at 10%, the Philippine Team arrived at a sample size of 60 households both for the target and the control sites. This represented roughly 18% of the Don Montano population. Having computed the objective sample size, the Team then proceeded to gather the names of all the crop-animal farmers in each village. This was done through the assistance of the Municipal Agriculture Office. From the list, the Team randomly selected the 60 respondents to be interviewed. Prior to data collection, the survey coordinator sought clearance from the municipal mayor and the concerned village chiefs. The survey was then formally conducted in November -December 1999 and periodic validation were made from time to time. When the interviews were completed, data were encoded and entered into the computer. Descriptive statistics such as frequency counts, percentages, and means were used to characterise Don Montano and Sta. Maria. After consolidating the results of the survey, the Philippine Team validated the picture and made an in-depth inquiry on matters related to farming. These were done through a participatory rural appraisal technique called story mapping. This is a participatory diagnosis and planning method that centres on talking, acting and learning with the target beneficiaries. Using this technique, participants were clustered into two, the farmers' group and the change agents' cluster. The first group was composed of farmer leaders and women farmers. The second cluster consisted of the municipal agricultural officer; the agricultural technician assigned to the village, the livestock inspectors of the town, a feed and drug supplier, trader/middlemen, a stud breeder and the village officials. The Team met each of the two groups separately. On the first day, the farmers' group was allowed to draw their community map and relate their farming situation, practices, and realities. They were also made to plot their cropping calendar. After this, the focus group discussion (FGD) and modelling techniques were employed to focus the discussion on specific solutions to farming problems. On the second day, the Team asked the change agents' cluster to (I) validate the community map drawn by the farmers; (2) story tell their role in the farming set- up; and (3) identify what they perceived to be the farmers' problems and solutions. Through this FGD, the survey data were validated and additional in-depth information were generated. These were elicited from the farmers through informal discussions. These outcomes are included and blended in the succeeding discussions on the characteristics of the focal site. For this report, a thorough description of the focal site, Don Montano, is presented and the control village will just be used for comparison as required in the CASREN outline. I Slovin's formula for determining sample size: N where n = sample size n = ~ N = population e = margin of error that can be tolerated (10%) CHARACTERISATION OF DON MONTANO: THE BENCHMARK SITE The village now known as Don Montano used to be a wide tract of land owned by a Spanish haciendero (Spanish term for rich landed farmer) named Don Montano Castillo. The latter donated part ofhis land to the municipal government, constructed schools and sponsored civic activities. Hence the village was named after him. With the implementation of the Agrarian Reform Law, the hacienda was subdivided into smaller parcels of land and distributed to tenants and workers. The farmers ofDon Montano today are the beneficiaries of the Land Reform Law that subjected the village to parcelling and distribution. .The following characterisation of the focal site was gathered from the Team's field audit as well as from the municipal records. Succeeding characterisations however , were based on the results of the survey and the FGD. Biophysical Characteristics Don Montano is one of the 58 agricultural villages in the Municipality ofUmingan in Eastern Pangasinan. It is characterised by farmlands situated at the foot of the partly denuded Caraballo Mountain. The familiar agricultural landscape boasts of strips upon strips of rice, onion, corn and peanut, alternately planted in the wet and dry seasons. Land use The total land area is about 297 hectares which is generally flat with a slope of 0-3%. The land area is used as follows: 45 ha ...214.21 ha 6.0 ha 30.5 ha ...0.1878 ha Irrigated. Rainfed Pastureland Residential Community plaza, school and church. . Climatic characteristics The climate in Don Montano belongs to Type 1, characterised by distinct wet and dry seasons (wet from May to October, and dry from November to April). During the dry period, farmers use water from shallow tube wells to augment water needed for crop production. The temperature in Don Montano averages 26.5?C. Crop-animal systems research in the Philippines 125 Soil type Just like the rest of the municipalities, the soil ofDon Montano is sandy loam soil, fit for raising lowland rice, corn, mungbean, sweet potato and some vegetables. Accessibility Don Montano is quite accessible by any form ofland transportation. The habitual highway traversing the whole strip of the village and the whole ofUmingan is concrete where regular means of public transport ply. The proximity ofDon Montano to key points in terms of economic activity, support services, market, source of farm inputs, is presented in Figure 2. La Union RFU 130 km - Baguio, 75 km Urdaneta UPang Livestock 18 km ~Market ~45km Uminganmunicipality w 3km .."'Lingayen Don Montano: 97 km ~ The BMS MAO3km San Jose City 21 km CLSU, PhilRice, PCC, BPRE 35 km Manila, 201 km Figure 2. Location of Don Montano relative to critical locations in the North. 126 Improving the contribution of livestock to crop-aminal systems in rainfed areas in Southeast Asia Socio-economic characteristics There are 329 households in Don Montano consisting of a population of 1,738 persons or a household size of 5-6 members. Ninety percent of the residents are farmers with an average land holding of 1.5 ha. The regular crops grown in Don Montano are rice, corn, onion, peanut, mungbean and vegetables and the animal raised are beef cattle, goat, carabao, poultry, and swine. While farming is the major source of income, other non-farm income is derived from serving as hired labourers, carpenters, PUV drivers, housemaids, security guards, and overseas contract workers. Nine percent of the potential workforce is unemployed. RESUL TS OF THE SURVEY AND FGD Respondents' Profile Based on survey results, 42 of the 60 farmer-respondents (70%) in Don Montano are male, out of which 97.6% are considered head of the family (Table I). Similarly, Table 1. Gender. status and household size of respondents. Don Montano (n=60) Sta. Maria (n=60) Item No. % No. % 42 18 60 70.0 30.0 100.0 43 17 60 71.7 28.3 100.0 Gender Male Female Sub-total Status Male Head of family Son Sub-total Female Wife Daughter Sub-total Household Size 0-1 2-3 4-5 5-6 >7 Average 41 I 42 97.6 2.4 100.0 41 2 43 95.4 4.6 100.0 18 O 18 100.0 0.0 100.0 16 1 17 94.1 5.9 100.0 O 12 19 22 7 0.0 20.0 31.7 36.7 11.7 10 19 16 14 6 6 Crop.animal systems research in the PhilIppines 127 1.7 16.7 31.7 26.7 23.3 43 of the 60 Sta. Maria farmer-respondents (71.6% ) are male, out of which 95.4% are heads of their respective households. These results are not surprising, as Filipino households, particularly those in rural areas, remain patriarchal in nature. The average household size is 6 members per household in both Don Montano and Sta. Maria. For both villages, most respondents are between 41-60 years old, with the average age being 48 years in Don Montano arid 50 iri Sta. Maria (Table 2). In terms of educational attainment, all respondents in Don Montanohave undergone formal education, while a number of respondents in Sta. Maria (8.3%) were not privileged enough to have done so. Majority of the respondents in Don Montano (46. 7% ) was able to at least reach or finish secondary education. In Sta. Maria, most of the respondents (43.3%) were able to reach the primary level of education. During the FGD, it was learned that there were various training and seminars on crop and livestock production sponsored by the Department of Agriculture as well as the veterinary drug and feed companies. However, survey results showed that for both villages, an alarming number of respondents was unable to participate in such training, either for crops (Don Montano, 81.7% and Sta. Maria, 100% ) or for livestock (Don Montano, 85% and Sta. Maria 96.7%). This may be attributed to the fact that the training courses were improperly timed such that farmers had difficulty attending such activities; albeit they were aware of such offerings, they were preoccupied with their on-farm activities. Crop Production Farmers are practising multiple cropping systems as evidenced by the several crops being planted both during the wet and dry seasons (Tables 3-4). These crops include rice, corn, mungbean, and vegetables during the wet season and onion, rice, peanut, tomato, corn, beans and vegetables during the dry season. Rice remains to be the major crop during the wet season while the other crops are planted only in small portions of their farm. On the other hand, onion is the major crop during the dry season, obviously because of limited water availability during the dry season. The main cropping pattern followed is rice + corn + vegetable-onion + rice + peanut. Cropping starts as early as April or May when the first monsoon rain begins with corn as the crop being planted. This practice according to them is to maximise utilisation of resources such as land and water. The amount of the first rainfall is enough to facilitate germination and early vegetative growth of the corn. Corn, which serves as catch crop, is purposely used as food for the family and as additional source of family income. This happens whenever rainfall can sustain the growth of the crop; otherwise when water is not enough, it serves as feed for livestock. Seedbed preparation for lowland rice usually starts early June and the crop is harvested early October in preparation for the onion seedbed. On the other hand, vegetables, peanut and mungbean are planted only in smaller areas and are intended for home consumption. When the farm produce exceeds their needs, this serves as additional income for the family. Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia128 ~ I E ~ 8~ ..,. u ~.. ~ ~ o 8 Eo u ~ ;" In vu ;;1 "' 00'D"'.u NoO~.O NO", "' > tU ""I",,", -.t-=o-~ .. .."rIO.'.' -..;"!o~~ ..,o..,~o =1 ~ -"' > ..M U -,.;"'0M ~ ->': ~ -MMU M O""' -0-* "' * ., c o c " ~ O c " 0: 0 c " I: O " " I: O I: 0 0 Q 0 ~ >.".",0tU U,,"!o- -0 >= 0--"' "" d >'" oooU VI O "' 00 OOOO.U 00 ~ > OON~ ~ >~ o--u N ~ "" o ~ !::! N- ~ ... Q ~ C) 0 0 ~ "' VI 0 ... "' N "" 0 C> ... "' .,., c:)c:)c:) ~~o 000 ~oo 000 000 The average rice yield was 3.32 tons for the dry season and 3.31 tons for the wet season. These are quite low considering the potential yield of the varieties they are using. The low yield is due to the fact that most of them are not knowledgeable on what and how much fertiliser should be applied. Moreover, the inadequate budget for rice farming adds to this low yield. For onion, the average yield is 1.3 ton/ha, which is very far below the average yield obtained by nearby farmers in San Jose and Mufioz, Nueva Ecija. It should be noted that the rate offertiliser application at an average of3 bags per hectare was very low. This was validated during the FGD and one of their reasons was lack of capital. On the other hand, the average yield of corn is 3.26 ton/ha, also considered low because all of them are not applying fertiliser. In general, yields obtained from all the crops are low. They perceived that this is because of inadequate funds necessary for the purchase of important farm inputs such as good quality seeds and fertilisers. Likewise, they perceived that low yield is correlated with the low fertility level of the soil as well as poor water- holding capacity . During the FGD, farmers requested for assistance in coming up with sound fertiliser and water management to increase crop production. Increase in their crop production will certainly help increase family income and ultimately improve the purchasing capacity particularly for inputs necessary for the improvement of their livestock production. Cropping calendar The following is the cropping calendar as related by the fanner-participants at the FGD. I Crop Ian Feb Mar Apr May Jun Jul SepAug Oct Nov Dec Rice Direct TPR Onion Corn YC WC .~ . Peanut .~ Mung- .~ bean Legend: Planting Harvesting ~ 132 Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia Animal Ownership Animal holdings In both Sta. Maria and Don Montano, the common animals being raised are buffaloes, cattle, goats, pigs, chickens and ducks (Tables 5-6). Buffaloes are commonly raised for draft and work purposes. Their importance in farm work persists even as farm mechanisation gains popularity with the onset of such farm machinery as "kuliglig" for soil cultivation. Farmers prefer buffaloes since they can be used in cultivating areas near paddy dikes, that farm machinery cannot reach. Moreover they require less maintenance costs and can easily be converted to cash during "times of emergency. More than half of the respondents in both villages raise buffaloes (Don Montano, 55.9% and Sta. Maria, 63.4%). On the average, a farmer owns 2 head of buffaloes in both Don Montano and Sta. Maria. Backyard cattle fattening is also largely practised in both agricultural villages, as in the whole ofPangasinan. In Don Montano, 67.8% of the respondents raise cattle; in Sta. Maria, 43.3% own cattle. On the average, each cattle raiser in Don Montano and Sta. Maria owns 3 and 2 head, respectively. Generally, the farmer-respondents prefer "native cattle" to imported breeds. The so-called "native breed" is evidently of Brahman bloodline considering the phenotype raised. As explained during the FGD, theyattributed this preference to the regularity (yearly) and ease in parturition, and its adaptability to local conditions. But still they look forward to raising the fast growing breeds considering the ultimate price these will command at the Urdaneta Livestock Auction Market. The farmer respondents in both villages also raise goats as they too command a good price in times of necessity .Likewise, they serve as source of meat for their own consumption, while having minimal input requirements. In Don Montano, 52.5% of respondents raise goats while in Sta. Maria, 41.6% keep the species. The average animal holdings in the sites are almost similar, with Don Montano having four goats per farmer and Sta. Maria, three. Goats are commonly tethered in vacant or fallowed areas in both villages. Respondents commonly raise monogastrics in their backyards. Swine raising, although capital intensive due to commercial feeds utilisation, is also practised in Don Montano (44.1%) and Sta. Maria (41.6%). On the average, a farmer in Don Montano fattens nine pigs while in Sta. Maria, a farmer averages only two head of pigs. The high average in Don Montano is attributed to the big animal holding of one particular farmer whose operation is already classified as semi-commercial in scale. The rest of the farmers, however, have an animal holding of only four pigs per household. Native chicken raising is vastly practised in both Don Montano (89%) and Sta. Maria (75%). This may be due to the fact that native chickens require minimal inputs and management as they are left to subsist and range freely. On the average, a farmer raises 20 chickens in Don Montario and 12 head in Sta. Maria. Ducks, on the other Crop.animal systems research in the Philippines 133 Table 5. Types of animals raised by respondents. Don Montano (n=59) Sta. Maria (n=60) Animal Average no. of animal holding Average no. of animal holding ~ Response No. Percent Response No. Percent 33 55.9 51.5 48.5 12.1 21.2 2 38 17 21 5 9 63.4 44.7 2 17 16 4 7 Buffaloes Adult Male Adult Female 2 Young Calves 13.1 23.71 1 40 10 33 4 19 67.8 25.0 82.5 10.0 47.5 3 2 2 2 26 9 20 0 8 43.3 34.6 76.9 0.0 30.7 2 2 2 O Cattle Adult Male Adult Female Young Calves Goats 31 II 24 2 15 52.5 35.5 77.4 6.4 48.4 4 2 2 2 3 25 5 21 O 9 41.6 3 20.0 84.0 0.0 Adult Male Adult Female Young Kids 2 O 236.0 26 7 18 8 13 44.1 26.9 69.2 30.7 60.0 9 2 3 14 5 25 5 19 5 2 41.6 20.0 76.0 20.0 8.0 2 3 2 1 Pigs Adult Male Adult Female Young Piglets Chickens All ages 53 89.0 20 45 75.0 12 Ducks All ages 25 42.4 7 5 8.3 22 Table 6. Number of animals raised by respondents. Don Montano Sta. Maria Animal Number Percent Number Percent }9 25 4 9 57 33.4 43.9 7.0 15.7 100.0 17 Buffaloes Adult Male Adult Female Young Calves .Total 27 5 10 59 45.7 8.5 17.0 Cattle Adult Male Adult Female Young Calves Total 13 66 8 35 122 10.6 54.1 6.6 28.7 100.0 15 35 O 13 63 23.8 55.6 0.0 20.6 100.0 Goats Adult Male Adult Female Young Kids Total 18 58 3 38 117 15.4 49.6 2.5 32.5 100.0 10 43 0 20 73 58.9 0.0 27.4 Pigs Adult Male Adult Female Young Piglets Total 14 46 112 67 239 5.9 19.3 46.8 28.0 100.0 16 28 II 7 62 25.8 45.2 17.7 11.3 100.0 Chickens All ages ,048 551 Ducks All ages 168 109 Crop-{Jnim{J{ systems research in the PhIlippines 135 hand, are not as popular. In Don Montano, 42.4% of respondents raise ducks while a measly 8.3% of respondents practise duck raising in Sta. Maria. On the average, the number of ducks raised by a farmer in Don Montano and Sta. Maria is seven and 22 head respectively. A more detailed inventory of animals for Don Montano and Sta. Maria is shown in Table 7. Shifting of animal species The reasons for raising farm animals as component of the total farm system are somehow influenced by the socio-psychological factors (aspirations/goals and perceptions), economics, intricacy in management and availability of resources (Table 8). Within the past five years, only 15.0% and 23.3% of the respondents in Don Montano and Sta. Maria, respectively, raised new species of livestock in addition to what they have been raising before. Of the respondents who raised new animals, the most common species raised are cattle in Don Montano ( 40% ), buffaloes and ducks in Sta. Maria (57.1 % and 7.1 % respectively) and goats, pigs and chickens in both villages (Don Montano, 11.1 %, 44.4% and 11.1% respectively; Sta. Maria, 28.6%, 21.4% and 7.1 % respectively). The most common reasons for starting new species are: source of income (Don Montano, 50% and Sta. Maria, 43.4%); ease in raising (Don Montano, 28.6% and Sta. Maria 21.1%); low input requirement (Don Montano, 21.4% and Sta. Maria,26.3%). On the other hand, Table 9 shows that there were also farmers who stopped raising certain species of animals within the past five years (Don Montano, 41.7% and Sta. Maria,33.3%). It was evident in both villages that pig raising was the predominant activity stopped (Don Montano, 60% and Sta. Maria, 31.0%). This could be attributed to the high cost of production inputs such as the commercial feeds and stocks required in pig raising. Other reasons cited by farmers were difficulty in raising particular species, lack of time for management, need to sell the animals, lack of feeds and feeding area and mortality . Animal preference Based on the results shown in Tables 10 and 11, fanner-respondents in both villages prefer raising ruminants such as beef cattle (Don Montano, 48.3% and Sta. Maria, 46.6% ) and water buffaloes (Don Montano, 31.7% and Sta. Maria, 41.7% ). The reasons for these preferences are ease in raising the species, use in draft/work purposes, low cost of inputs and good source of income in times of necessity .Some fanners also indicated preference for raising pigs, as this is also a good source of income. However, the high input requirement needed for purchase of stocks and commercial feeds may have some bearing on their decisions. 136 Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia .~~~CQ...(/)"CcCQ0cCQ...co~coc.5>..0...cQ)>.5"'iOE.2CQ'0-0>..CQEE~(/)r-.Q):cCQ~ ] 1# ~~~ . ~~u "'-.;0~ "'00:0: ~ I "'~u=0 E~ oz oz ~ oz ~ oz ~0 oz ~ oz ~0 Oa.onNOOOOon oO\~o-N- -\oV\NNONO NOr-- -"'I -N - -- Q\\I'IIOOOOIOOr--Q\OONO -\1'1- ~- ONI-'-00 0 \ON001-'- O'Ir--NOMO 0000 11'10 000 II'IIDO11'1 "'cxi.';dddd...:~.';...: QO-NQOOO\I0~~~N~~oOoo-oioN~-oi'..;N- ~",r-.r-.o",0~--oo..':~~o~oN~..':o"' - O\onoOO'-tO~'-t~NMNooo...;oNM-oo'-t -N .0 O\r--r o~o,..,r ,.,.., oOooooooa\.';oo.0 N r--NO\oOOO'ot\O-'otM-=oooNocxi-=oo\0 N O~ 0.,..0 0 N .,.. ~Nr--NN N N O\-\0.,..~~0\~ ~. ll"\N-r-. 11"\ 0-""N ~\C NO\O\NOONNon- ~O\f'--NOO..'.Ot"\OIn lOr--,..,r--N ~oooooooooo o~ r ,. 0 COr--\0 OOr--ON,",","," -0\ CONI/") I/") r--N N I/") O\N-oOV\ 00\0000\0\0...,...,...,0oci...;...;ooNN..;...;..;oN- - I'1-NOOI'100\-1'10""'-.-iOO.'.;0.-i.-i...:O o-..,.ooO\o\Ooo~or..:..j...:oo...:o",-or..:ot"\ "'C\C\OOOOC\N...'O..tNNoooo'...:\Ooo-0 N- C\-I'\OO'..O~t"I~t"I;-ooooor-.:N~oI'\ N r-.",C\OOOONOOOOOr-:~OOOOO~..;r-:O~ N OONO\O~~0...' 1-- oooor--",ooO\ "' 0 0 C:- C:- C:- C:- s s s s ~ ~ ~ ~ ~ ~ ~ ~ ~ :- -~ :- ~ -., ~ O~ ~- ~ ~ ~ ~- ~ ~ c: '- ,5 ~ c ~ ~ 5 ~ c ~ ~ .- I3 bo .-C ---bo '- C --- c:~ ~~~...~... ~~~.05 ~~~...~... ~~~ o 'E E ~ ~ .: ~ ~ ~ ~ ~ ~ ~ ~ '~ .'t E ~ ~ ': ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ c c ~ $ ~ ~ ~ .~ c ~ $ ~ ~ ~ c c ~ $ ~ ~ ~ .-c ,,9 $ ~ c:~~~~~~~~~~~~u .~~~~~~~~~~~~u 0 .0 0 rn 0\0 \0 OM -0 Table 8. Respondents who raised new species (not raised before) within the last 5 years. Don Montano (n=60) Sta. Maria (n=60) Item No. % No. % Raised new species Yes No 9 51 15.0 85.0 23.3 76.7 14 46 Animal newly raised Cattle Buffaloes Goats Pigs Chickens Ducks (0=9) (n=14) 4 O 1 4 44.4 0.0 II.I 44.4 II.I 0.0 7.1 57.1 28.6 21.4 7.1 7.1 8 4 3 1. 0 (n=14) (n=19) 7 50.0 28.6 21.4 0.0 9 43.4 21.1 26.3 5.2 4 4 53 O Reason for raising new species Source ofincome Ease in raising Low input requirement Use as draft animal Improving the contribution of livestock to crop-anlinal systems in rainfed areas in Southeast Asia138 Table 9. Respondents who stopped raising animals within the last 5 years, Don Montano (n=60) Sta. Maria (n=60) Item No. % No. % Stopped raising species Yes No 25 35 41.7 58.3 20 40 6 3 3 15 4 24.0 12.0 12.0 60.0 16.0 12.0 5 8 3 9 2 2 17.2 27.6 10.3 31.0 6.9 6.9 Animal species Cattle Buffaloes Goats Pigs Chickens Ducks 3 8 2 3 3 1 8 2 27.6 6.9 3.5 10.3 10.3 3.5 27.6 6.9 3.5 8 6 1 2 2 O 7 3 O 21.4 3.6 7.1 7.1 0.0 25.0 10.7 0.0 Reason for stopping Had to sell the animals Mortality Given away Lack of time for management Difficulty of raising Animal is too old Lack offinancial inputs Lack offeeds andfeeding area No reason stated Crop-animal systems research in the Philippines 139 Table 10. Species of animals preferred by respondents. Don Montano (n=60) Sta. Maria (n=60) Animal No. % No. % First choice Cattle Buffaloes Goats Pigs Chickens Ducks No answer 29 19 3 7 O 48 31 5 11 O I I 28 25 1 6 O O O 46.7 41.7 1.7 10.0 0.0 0.0 0.0 Second choice Cattle Buffaloes Goats Pigs Chickens Duck No answer 20 14 9 8 5 I 3 33.3 23.3 15.0 13.3 8.3 1.7 5.0 14 19 10 9 7 I 0 23.3 31.7 16.7 15.0 11.7 1.7 0.0 Third choice Cattle Buffaloes Goats Pigs Chickens Ducks No answer 2 6 14 14 16 O 8 3.3 10.0 23.3 26.7 0.0 13.3 13.3 2 4 20 16 9 3 6 3.3 6.7 33.3 26.7 15.0 10.0 10.0 140 Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia .3 .7 .0 .7 .0 .7 .7 Table 11. Reasons for animal preference of respondents. Don Montano (n=169) Sta. Maria (0=175) Reason No. % No. % 53 36 40 17 4 31.4 21.3 23.7 10.0 2.4 51 35 46 22 4 29.1 20.0 26.3 12.6 2.3 Easy to raise Source of income Draft purposes Own consumption Breeding purposes Low cost of inputs/use of crop residues 18 10.7 0.6 15 2 8.6 1.1No answer These results confonn with the findings in the RRA and FGD. During the FGD, the fanners voiced out that they preferred raising beef cattle because of its high price in the market, while buffaloes aside from being an "emergency" source of income, are also used for draft purposes. Moreover, the fanners indicated their preference for raising pigs, given the financial support/assistance, since they can easily be marketed aside from serving as additional source of income. ANIMAL MANAGEMENT PRACTICES: FEEDS AND FEEDING SYSTEMS FOR RUMINANTS DURING THE DRY SEASON Feeding systems The dominant feeding system for ruminants in the dry season is tethering. Nineteen out of 40 farmers (47.5%) tether their cattle and 12 out of31 (36.3%) do so with their buffaloes. Other feeding systems being practiced are stall-feeding, stall-feeding + free-grazing, stall-feeding + tethering, free-grazing and free-grazing + tethering. However, the degree of use of each feeding system varies with the species being raised (Table 12). Crop-animal systems research in the PhilIppines 141 Types of basal feeds Farmers use almost the same type ofbasal feed for cattle and buffaloes. The most used basal feeds are grasses + crop residues, grasses alone, crop residues alone, grasses + crop residues + tree forages, and grasses + tree forages, in that order . Goat raisers use different types ofbasal feed from those raising cattle and buffaloes. They use mostly grasses (54.8%) as basal feed for their goats. The second most used basal feed for goats are tree forages (29.0% ) while grasses + crop residues + tree forages (6.45%) and grasses + crop residues (3.2%) are not extensively used by farmers raising goats. During the RRA and FGD, it was learned that tree legumes such as ipil-ipil (Leucaena), kakawate (Gliricidia) and acacia are seldom utilised as feeds, although these species particularly, giant Leucaena trees, abound in the area. Common reason given for non-use was that these legumes cause diarrhea in the animals. These resources are however used as firewood by the residents. Source of feeds Farmers have different sources offeed for each ruminant species during the season. For cattle, farmers use crop residues from their own farm (37.5% ) and from their own pasture + neighbors' pasture (35.0%). Buffaloes raisers get their feeds mostly from their own pasture supplemented by those from their neighbor's (54.6%). On the other hand, 58.l% of farmers raisinggoats have their own pasture/grazing area while 25.8% use their own pasture + neighbors , as their main sources of feed during the dry season. Frequency of feeding Table 12 also shows that for large ruminants (cattle and buffaloes) fanners in Don Montano practice almost similar feeding frequency. More than half of the respondents feed their cattle and buffaloes daily. A good number of them also feed their cattle and buffaloes thrice a day (45.0% and 39.4%, respectively). Goat raisers feed their goats differently. Some fanners feed their goats twice a day (32.3%) while others give feeds daily (29.0%) or thrice a day (29.0%). Improving the contribution of livestock to crop-animal systems in ralilfed areas in Southeast Asia142 Table 12. Feeding system and types of basal feed for ruminants during the dry season. Cattle (n=40) Buffaloes (0=33) Goats (0=31) No. % Item No. % No. % 19 O 3 6 47.5 0.0 7.5 15.0 12 2 6 36.3 6.1 18.2 3.0 19 O 9 0 61.3 0.0 29.0 0.0 Feeding system Tethering Free grazing Stall-fed Tethering & free-grazing Tethering, free-grazing & stall-fed Tethering & stall-fed Free grazing & stall fed No answer 4 2 5 10.0 5.0 12.5 2.5 5 7 O O 15.2 21.2 0.0 0.0 3.2 0.0 0.0 6.5 o 0 2 Type of basal feed Grasses Tree forages Crop residues Grasses & crop res. Grasses & tree forages Grasses, crop residues & tree forages No answer 9 O 5 23 1 1 22.5 0.0 12.5 57.5 2.5 2.5 9 O 5 18 O 1 27.3 0.0 15.2 54.6 0.0 3.0 17 9 O 1 O 2 54.8 29.0 0.0 3.2 0.0 6.5 r 2.5 0 0.0 0 0.00 Adequacy of feeding Most farmers find the feeds they give their animals to be adequate. Out of 40 farmers raising cattle, 97.5% indicated that their feeds are adequate. All farmers with buffaloes also think they have adequate feed and the same is true with those raising goats (93.6%). Types and amount of supplements The survey revealed that supplementation for ruminants even during the dry season is seldom practiced. Crop-animal systems research in the Philippines 143 Table 13. Source of feed, frequency and adequacy of feeding, and supplements given to ruminants during the dry season. Cattle (n=40) Buffaloes (0=33) Goats (n=31) No. % Item No. % No. % 2 15 8 5.0 37.5 20.0 6 3 6 18.2 9.1 18.2 18 O 2 58.1 0.0 6.4 Source of feed Own pasture/grazing area Crop residuesfrom ownfarm Community pastures/road -side grazing Own pasture & neighbor's pasture No answer 14 35.0 18 54.6 8 25.8 1 2.5 0 0.0 3 9.7 20 I 18 I 50.0 2.5 45.0 2.5 18 2 13 O 54.5 6.1 39.4 0.0 9 10 9 3 29.0 32.3 29.0 9.7 Frequency of feeding Daily Twice a day Thrice a day No answer Adequacy of feeding Adequate Inadequate 39 1 97.5 2.5 33 O 100.0 0.0 29 2 93.6 6.5 Supplements type and amount BMS Rice Bran Consumix Vetracin None I. O O O 39 2.5 0.0 0.0 0.0 97.5 I. O O O 32 3.0 0.0 0.0 0.0 96.9 o 0 0.0 0.0 3.3 3.3 93.6 1** 29 * Average quantity reported was 2 kg per head per day. **Average quantity reported was 0.25 kg per head per day. 144 Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia ANIMAL MANAGEMENT PRACTICES: FEEDS AND FEEDING SYSTEMS FOR RUMINANTS DURING THE WET SEASON Feeding systems During the wet months, most farmers in Don Montano tether and stall-feed their cattle and buffaloes, while their goats are mostly tethered (Table 14). For cattle, the other feeding system being used are tethering + free-grazing (22.5%) and tethering + free-grazing + stall-feeding (10.0%). A greater percentage of buffaloes (30.3%) are also tethered. Some 15.2% are tethered at times, freely grazed at other times or stall- fed. Feeding systems for goats are tethering (87.1%), tethering + stall-feeding (3.2%) and free-grazing + stall-feeding (3.2%). Types of basal feeds Majority of Don Montano farmers use grasses + crop residues as the basal feed for their cattle (70.0%) and buffaloes (69.7%) during the wet season. Grasses are the second most used basal feed for large ruminant animals. For goats, the type of basal feed is mostly grasses (48.4%) and grasses + tree forages (35.5%). Source of feeds During the wet months, cattle raisers get crop residues from their own farm (35.0% ), own pasture + neighbor's pasture (32.5%), community pastures/road side grazing (22.5%), own pasture/grazing area (5.0%) and own pasture + contour hedgerows (2.5%). For buffaloes, the sources of feeds are the farmers' own pasture + neighbor's pasture/ road side grazing (18.2%) and crop residues from own farm (6.1%). Majority of the goat raisers ( 67.7% ) have their own pasture/grazing area as their source of feed in the wet season. Others have their own pasture + neighbor's pasture ( 16.1% ) and community pasture/road side grazing (9.7% ) as their sources of feeds for goats. Frequency of feeding Almost 50% of the fanner-respondents feed their cattle daily during the wet season while 42.5% of them feed their animals thrice a day (Table 15). For buffaloes, an almost similar feeding frequency is employed by fanners. On the other hand, 11 out of 31 respondents (65 .5% ) feed their goats twice a day while others are feeding their goats daily (29.0%) or thrice a day (29.0%). Crop-animal systems research in the Philippines 145 Table 14. Feeding system, types of basal feed, and source of feeds for rumi- nants during the wet season. Cattle (n=40) Buffaloes (n=33) Goats (n=31) No. % Item No. % No. % 12 30.0 30.3 3.0 3.03 15.2 10 27 O O 1 87. 0. 0. 3. Feeding system Tethering Stall-fed Tethering & free-grazing Tethering, free-grazing & stall-fed Tethering & stall-fed Free grazing & stall fed No answer 9 4 22.5 10.0 1 5 14 O 1 35.0 0.0 2.5 16 0 0 48.5 0.0 0.0 1 3.2 2 6.5 Types of basal feed Grasses Crop residues Tree forages Grasses & crop res. Grasses & tree forages Grasses, crop residues & tree forages No answer 6 3 1 28 O 1 15 7 2 70 0 2 7 O 21.2 0.0 3.0 69.7 3.0 3.0 48.4 0.0 3.2 3.2 35.5 3.2 15 O 23 1 1 1 11 1 1 2.5 0 0.0 2 6.5 2 14 9 5.0 35.0 22.5 7 2 6 21.2 6.06 18.2 21 O 3 67.7 0.0 9.7 13 1 32.5 2.5 18 O 54.6 0.0 5 O 16.13 0.00 Source of feed Own pasture/grazmg area Crop residues from own farm Community pastures/road side grazmg Own & neighbor's pastures Own pasture & contour hedgerows No answer 1 2.5 0 0.0 2 6.45 Improving the contflDution of livestock to crop-animal systems in rainfed areas in Southeast Asia146 1 0 00 1 .0 .5 .5 .0 .0 .5 Table 15. Feeding system, types of basal feed, and source of feeds for rumi- nants during the wet season. Cattle (n=40) Buffaloes (n=33) Goats (n=31) No. % Item No. % No. % 19 3 17 1 47.5 7.5 42.5 2.5 7 2 14 O 51.5 6.0 42.4 0.00 9 11 9 2 29.03 65.48 29.03 6.45 FrequencyofFeeding Daily Twice a day Thrice a day No answer Adequacy of feeding Adequate Inadequate 39 1 97.5 2.5 33 O 100.0 0.0 29 2 93.55 6.45 Supplements type and amount BMS Rice Bran Consumix Vetracin None 2.5 0.0 0.0 0.0 97.5 1 O O O 32 3.0 0.0 0.0 0.0 97.0 o o 0.00 0.00 3.23 3.23 93.6 o o 0 39 1 29 * Average quantity reported was 2 kg per head per day. ** Average quantity reported was 0.25 kg per head per day. Adequacy of feeding Majority of the farmer-respondents (97.5% for cattle, 100% for buffaloes and 93.6% for goats) indicated that their feeds are adequate during the wet season. Types and amounts of supplements Farmers ofDon Montano seldom practise supplementation of any kind during the wet season. Crop-animal systems research in the Philippines 147 ANIMAL MANAGEMENT PRACTICES: FEEDS AND FEEDING SYSTEMS FOR NON-RUMINANTS DURING THE DRY SEASON Feeding system Majority (92.3%) of the farmer-respondents ofDon Montano indicated that their pigs are pen-fed (Table 16). Only 7.7% let their pigs scavenge for their feeds. On the other hand, farmers raising native chickens emp]oy three systems of feeding: scavenging, pen-feeding and the combination of the two systems. However, most (75.5%) of them just let their native chickens scavenge for feeds. The same is true in duck raising. Scavenging is the most (72.0%) practiced feeding system. Type of basal feeding Commercial feeds (61.5%) are the most used basal feed for pigs by farmers- respondents of Don Montano. Only few of them use either grains (19.2%) or the combination of grains and commercial feeds (19.2%). For native chickens, most farmers rely on kitchen refuse (30.2%). Others feed their native chickens with the combination of kitchen refuse + grains + commercial feeds (18.9%), grains (17.0%), kitchen refuse + grains (15.1 %), grains + commercial feeds (5.7%) or commercial feeds (3.8%). Ducks are mostly fed kitchen refuse (32.0%). Source of feeds The main source of feed for pigs in Don Montano is the Umingan market (65 .4% ). For native chickens, the dominant sources of feed are home supply (39.6%), home mixed (24.5%) and home supply + home mixed (24.5%) and home supply + home mixed (20.8%). Most duck raisers have home supply (72.0%) as their source of feeds. A few rely on home mixed feeds (20.0%). Frequency of feeding Table 17 shows that farmers feed their pigs daily (46.2%), thrice a day (38.4%) or twice a day (15.4%). Native chickens are fed thrice a day (49.1%), daily (32.1%) or twice a day (1.3%). Ducks are usually fed twice a day (64.0%). Adequacy of feeding All the farmer-respondents indicated that their pigs are adequately fed. For native chickens, 90.5% indicated adequacy of feeding. On the other hand, 92% of the farmers raising ducks think their ducks are fed adequately. 148 Improving the contribution of livestock to crop-animal systems in ralnfed areas In Southeast Asia Table 16. Feeding system, types of basal feed and source of feed for non- ruminants during the dry season. Pigs (0=26) No. Chickens (n=53) Ducks (n=25) Item % No. % No. % 2 24 0 0 7.7 92.3 0.0 0.0 40 2 5 6 18 4 I 2 72.0 16.0 4.0 8.0 Feeding system Scavenging Pen-fed Scavenging & pen-fed No answer 5 }6 5 O O O 19.2 61.5 19.2 0.0 0.0 0.0 9 2 3 16 8 10 17.0 3.8 5.7 30.2 15.1 18.9 1 O 2 8 5 7 4.0 0.0 8.0 32.0 20.0 28.0 Types ofbasal feed Grains Commercial feeds Grains & comml feeds Kitchen refuse Kitchen refuse & grains Kitchen refuse, grains & commercialfeeds No answer 0 0.0 5 9.4 2 8.0 1 17 2 3.9 65.3 7.7 21 I O 39.6 1.9 0.0 18 O O 72.0 0.0 0.0 I. 3.9 0 0.0 0 0.0 5 O O O O O 19.2 0.0 0.0 0.0 0.0 0.0 0 0.0 1.9 24.5 20.8 1.9 9.4 0 0.0 13 11 1 5 5 O O 2 20.0 0.0 0.0 8.0 Source of feed Home supply Market place/factory Market place/factory & cooperative store Home supply & market place/factory Market place/factory & home mixed Cooperative store Home mixed Home supply & home mixed Home supply & market place No answer Crop.animal systems research in the Philippines 149 75.5 3.8 9.4 11.3 Table 17. Feeding system, types of basal feed and source of feed for non- ruminants during the dry season. Pigs (0=26) No. Chickens (n=53) Ducks(n=25) Item % No. % No. % 12 4 10 0 46.2 15.4 38.4 0.0 17 6 26 4 7 16 2 O 28.0 64.0 8.0 0.0 Frequency of feeding Daily Twice a day Thrice a day No answer AdequacyofFeeding Adequate Inadequate 26 O 100.0 0.0 48 5 90.6 9.4 23 2 92.0 8.00 4b O Id 7a lc O O lc 26.9 3.9 0.0 0.0 3.9 3.9 61.5 7.6 0.0 1.9 1.9 1.9 0.0 86.8 o o o o o 0 25 0.0 0.0 0.0 0.0 0.0 0.0 100.0 If O 46 Supplements type and amount Vetracin Rice Bran B-12 Chloromycin Feeds Kangkong/watercabbage None 16 a Average quantity -166 9 per head per day b Average quantity -1 teaspoon per gallon c Average quantity -was 2 kg per head per day d Average quantity -1 tablet e Average quantity -1 cavan (50 kg) f Average quantity -3 kg per head per day Type and amount of supplements Majority of the farmers raising pigs and native chickens do not give supplements to their animals. A few provide rice bran, Vetracin, vitamins and water cabbage. None of the farmers give supplements to their ducks. 150 Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia 32.1 11.3 49.0 7.6 ANIMAL MANAGEMENT PRACTICES: FEEDS AND FEEDING SYSTEMS FOR NON-RUMINANTS DURING THE WET SEASON Feeding system Farmers practice the same feeding system for pigs, native chickens and ducks in both the dry and wet season (Table 18). Types of basal feeds Farmers also use the same types of basal feed for non-ruminants in both the dry and wet seasons. Source of feeds Farmers' main source of feed for pigs in the wet months is the market place/ factory .Some of them also have the market place/factory + home mixed as their source of feed for their pigs. For native chickens, the usual sources of feeds are home supply (39.6%), home mixed (24.5%) and home supply + home mixed (20.8%). For ducks, farmers' main source of feed is home supply (72.0% ). The source of feed is similar for both the wet and dry seasons. Frequency of feeding Table 19 shows that fanners feed their pigs either daily (46.2%), thrice a day (38.5%) or twice a day (15.4%). This is similar to the practice during the dry seasons. Most fanners are feeding their native chickens thrice a day during the wet months. A good number of respondents also indicated that they do feed their chickens daily (33.9%) and twice a day (15.1 %). Duck raisers feed their animals twice a day (32.00%) during the wet season. Adequacy of feeding Farmers indicated a 100%,92.5% and 92.0% adequacy of feeding pigs, native chickens and ducks, respectively, in the wet season. Only 4 out of 53 farmers (7.6%) think that their native chickens are inadequately fed. On the other hand, there were only 2 out of 25 farmers who said that their ducks are not adequately fed. Crop-animal systems research in the Philippines 151 Table 18. Feeding system, types of basal feed and source of feed for non- ruminants during the wet season. Pigs (0=26) No. Chickens (n=53) Ducks (n=25) Item % No. % No. % 2 24 O O 7.7 92.3 0.0 0.0 42 2 5 4 79.2 3.8 9.4 7.6 18 4 1 2 72.0 16.0 4.0 8.0 Feeding system Scavenging Pen-fed Scavenging & pen-fed No answer 4 17 5 O O O 15.4 65.4 19.2 0.0 0.0 0.0 9 2 3 16 8 11 17.0 3.8 5.6 30.2 15.1 20.8 1 O 2 8 5 7 4.0 0.0 8.0 32.0 20.0 28.0 Types of basal feed Grains Commercial feeds Grains & comml feeds Kitchen refuse Kitchen refuse & grains Kitchen refuse. grains & commercial feeds No answer 0 0.0 4 7.6 2 8.0 I 18 I 3.9 69.2 3.9 21 1 O 39.6 1.9 0.0 18 O O 72.0 0.0 0.0 0 0.0 0 0.0 0 0.0 6 23.0 0 0.0 0 0.0 o o o o o 0.0 0.0 0.0 0.0 0.0 I 13 II I 5 1.9 24.5 20.8 1.9 7.6 O 5 0 0 2 0.0 20.0 0.0 0.0 8.0 Source of feed Home supply Market place/factory Market place/factory & cooperative store Home supply & market place/factory Market place/factory & home mixed Cooperative store Home mixed Home supply & home mixed Home supply & market place No answer 152 Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia Table 19. Frequency, adequacy of feeding and supplements given to non-rumi- nants during the wet season. Pigs (n=26) No. Chickens (n=53) Ducks(n=25) No. % Item % No. % 12 4 10 0 46.2 15.4 38.5 0.0 18 8 23 4 33.9 15.1 43.4 7.6 7 8 8 2 28.0 32.0 32.0 8.0 Frequency of feeding Daily Twice a day Thrice a day No answer AdequacyofFeeding Adequate Inadequate 26 O 100.0 0.0 49 4 92.5 7.6 23 2 92.0 8.0 4b O Ic O Id O 46 7a 26.9 3.9 0.0 0.0 0.0 3.9 65.4 7.6 0.0 1.9 0.0 1.9 0.0 87.7 o o o o o 0 25 0.0 0.0 0.0 0.0 0.0 0.0 100.0 o o o Supplements type Vetracin Rice Bran B-12 Chloromycin Feeds Kangkong/water cabbage None 17 a Average quantity -166 9 per head per day b Average quantity -1 teaspoon per gallon c Average quantity -1 tablet d Average quantity -3 kg per head per day Crop-animal systems research in the Philippines 153 Types and amounts of supplements Most farmers raising pigs and native chickens do not provide their animals with supplements during the wet season. A few of them do provide rice bran, vetracin, water cabbage and vitamins. Ducks are not given supplements at all. Animal Production and Health Practices Table 20 shows that the average age at first parturition is 3 years for cattle, 3 years for buffaloes, 1 year for goats and 8.5 months. for pigs. The average number of offsprings is a -calf for cattle and buffaloes, two kids for goat and nine piglets for pigs. The average weaning weights are 112, 88, 18 and 18 kilograms for cattle, buffaloes, goats and pigs, respectively. The average weaning ages for cattle and buffaloes, goats, and pigs are 10, 4 and 2 months, respectively. On the other hand, the average marketing ages for cattle, buffaloes, goats and pigs are 3, 4, I and 4 years, respectively. Majority of the farmer-respondents use an open type of housing for their cattle and buffaloes, and a shed type for their goats and pigs. In breeding their animals, most farmers use the natural system (Table 21 ). Only few use artificial insemination in breeding their animals. Farmers raising cattle, buffaloes and pigs usually have their animals vaccinated while those raising goats and chickens are not practicing vaccination. On the other hand, the Don Montano farmer-respondents never practice drenching. Deworming is practiced by majority of farmers raising cattle, buffaloes and pigs. For goats, the number of farmers practicing deworming is just equal to those not practicing. Table 20. Production parameters for cattle, buffaloes, goats, and pigs. Cattle (n=40) Buffaloes (n=33) Goats (0=31) Pigs (0=26) A verage age at first parturition (yrs. ) 3 3 0.7 A verage number of offsprings 2 9 A verage weaning weight (kg) 112 88 18 18 Average weaning age (mth.) 10 10 4 2 Average marketing age (yrs) 3 4 0.3 154 Improving the contribution of livestock to crop-animal systems In rainfed areas In Southeast Asia //)~CJ.;;CJ~~c....c~E~01~C~E'Cc~~~~~~~E.2<.-N~:c~1- u--0~1uZ - 8;;:;-M~ II ='6-~ - !tl-~~cz - - rIIlCOl)N.-II~6...Q)-~(0~ oz ~ oz ~ oz * ~ ~ 00.S~ ~ , J:~~~'+.- E: ?~~~~~~~E-o or) N I") -N 'rIO'rI~""~t'\'rl O\NN N r'"It--Or-:~~N-O fX)fX)V)- -00-~.'.;IOV\N- IrI-ON 0 -0\0\OMo0\ NMIr\M 11'111'111'1r--"r--"N~ - -'0!"00N .q-N.q-IO-N...;N...;IO-N -N~N r--Ir)OOr-:\Oor)1,0 N -NMN ~r---",or-..:...:~ - ~ -.0...;::~ ~ = .~ e ~ e. "' ~ .5 I.. "' -I\) ~~ .~ ~ .= ~ ~ ~ ] ~ .;:: ~ ~~~~~ 00\-M 011"111"1v)NNt"-N -ONN- \Or'1-r.-iciID\Or'1 O\Vlr-- - o~\O0\00 N N~N VNON Mt--ONr.:o0\ 000 000000 000 000000 000 000000 000 0000'0'0' *"'u "' .~ u -UU ... e "'g u "'g Q. .~ , e .~ , c~ ~ Q.~ ~ o~U;i: tIO~~;i: .~ .~ ~ ~ .5 .~ ~ ~ c~~~ .=~~~ .~ ~ U ~ ~t~~ 5t~~ ~ > 0 0\0N- V)oV)N.,)Nr-.N NO\NN r-N-~O~ION M-OO-- r--\r)OO00.';.';~~N NMN \01'"10\.., ;~ ~- "'uu.~u~ ~ Q, .~ , co~'U~.? .~ ~ ~ e ~ ~ ~ 0 !:3 ... ~t~~o ..;q)~~9q)>~00-"'~~~.g0~~.s0~~N\I')>-.&J"0q)"'.~u~Q.-0~"'~~"'~.?;.u:Eu0~0.~~.5uu>* Labour Inputs Ruminants The most common labour activities done by farmers raising ruminants and their corresponding frequencies are listed in Tables 22-23. The majority of the farm work, such as grazing and tethering of animals, cutting and carrying grasses, preparations of feeding materials and stall feeding cleaning of pens and houses, washing animals are done on a daily basis in both wet and dry seasons. These activities, as shown in Tables 24 and 25, are done mostly by family members and more specifically by the adult male or head of family, although adult female and chilf}ren sometimes help. No one among the respondents pr~ctice exchange labour , while a few resort to hiring labourers to do farm work. Non-ruminants Farmers raising monograstric animals also follow the same routine during both wet and dry seasons. As reflected in Tables 26 and 27, feeding and cleaning the animals and barns are usually done daily in both wet and dry season. However, vaccination, breeding and marketing are commonly done seasonally in both dry and wet seasons. As in the case of raising ruminants, most farm work for monogastrics is done mostly by family members, particularly the adult males (Table 28). However, the role of adult females in farm work for monogastric is much pronounced than those in ruminants. Some children also help in these farm activities. No one among the respondents practices exchange labour, while a few hire labourers to do the farm activities. 156 Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia c0C/IIUQ)C/I>-"CQ).t:...CIC.C~"CC/I...CIUC.E~-coc-oco.~CJ~"CO-c..5C/IQ).~.>.~CJIU-~o.c~-0>CJCQ)~C"Q)-LLLnNQ):cIU1- >.u=8.~~ ~0;Q >-~0~ -;=O"'tUu00~">":3u< ..~~oz ~ ~ ~ ~ *= == d = \I"\ \I"\ r"\ N N N o o o O 0 O o o o o o o O 0 O N 0\ 0\ 1--- ~ ~ r..:"' e ] 00 ~ d c .0 ~ d ...00 ~ ~ 'C .g ~ .E .s .s .g ~ ~ "Oi) O 00 00 00 ~ .~ OIJ E c. = .= .~ ~ '0 .= .~ ~ :0 a .u ~e ] ~ ~ ~ ~ u ~ ;b ~ --~ ...~ ~~UU> ~~ ~ o o - o o 01 d d M N N ~ -0-000010100 o o o o r-- N - 0- ..". ,0 .,) 1"'- \C o o o o d d o o o o d d 0\ O ,.) 0 II'\ 0\ N IC ai ~ -M 0 '0~00 0 0:o 0 0:o 0 0.o -.r ~Ir)-NN 0:o 0 oO N OC!M0 od -\1'\ N,00\ 0 Q(.) = tIO = .."" o .8 ~ 8.~ ~ Q .~ ~ 'ao o 5 ~ ~ :6 .E ~ .!S .~ .~ Q Q a = "tJ-g 8e~b 5 -Q ~ ~ 5 ~ > ~ ~ ~ 0:o 0;0 0 od 0 0:o0 0:o 0 0 O '0 O Mr-- " 0 " O M o o o o O O 0- - N - r-- 00 ~ 0 V) N ..,.N '""! 0-. "1~ O 0\M 0 N Economics of Animal Enterprises Cost of production for ruminants Table 28 shows the average cost of production per head (excluding feeds) for ruminants. Based on the results, cattle raising entails a cost ofPhilippine pesos P205 .71 per head, while buffaloes and goats require P192.06 and P40.33, respectively. Majority of the cost incurred for raising ruminants, constituted about 80% of the total cost of production (excluding feeds), involving the cost ofropes, breeding services, medicine and vaccination. Other minor costs include castration, salt and veterinary service costs. It was also evident in the results that nobody practices drenching in ruminants. Cost of production for non-ruminants Swine raising, compared to other non-ruminants, entails the highest cost, with PI 22.22/head (Table 29). The major costs for pig production come from medicine, breeding services and vaccination, which entails 92% of the total 'cost. However, it should be noted that the cost of commercial feeds, which represents 70% in swine raising, is not reflected in the results. On the other hand, poultry raising (chickens and ducks) in Don Montano require the least investment of less than P2.00/hd, as poultry species need minimal management inputs. Native chicken raising, a popular livestock enterprise in Don Montano, entails only Pl.67/hd with the majority of the cost being incurred on medicaments. The same holds true for duck raising, which involves only costs for medicine (PO.O9/hd). Meat Production, Consumption and Marketing As indicated in Table 30 and further confinned from the FGD, Don Montano animal raisers rarely slaughter their animals but sell them on liveweight basis. In fact, a comparison of the number of raisers that slaughter animals versus the total number of animal raisers showed that the ratio of the fonner to the latter was less than one third. Furthennore, it was also noted that except for goats, the proportion of the mean number of animals slaughtered annually as contrasted to the average number of animals raised per household ranges from only 50% to 75%. For the large ruminants, average slaughter weight per animal was 147-150 kilograms while for goats and swine, averages were placed at 25 kilogram and 68 kilogram per head respectively. Except for goats, which are considered as part of the native delicacy during almost all social occasions, most of the ruminants and pigs being slaughtered were sold rather than consumed at home. Average fann gate prices received for slaughtered meat ranged from p 154 per kilo for beef to p 100 per kilo for pork and goat's meat. Crop-animal systems research in the Philippines 163 Table 28. Average cost of production for ruminants in Philippine pesos (excluding feeds). Item Cattle Buffaloes Goats 26.29 3.47 13.86 0.00 8.71 56.68 9.20 79.08 8.42 0.00 205.71 24. 6. II. 0; 15. 36. 0. 77. 16. 0. 189. 6.67 0.00 1.85 0.00 3.64 1.24 1.24 25.69 0.00 0.00 40.33 Medicine Veterinary services Vaccination (labour and material) Drenching Deworming (labour and material) l3reeding Salt Ropes Castration Other costs Total cost per head Table 29. Average cost of production for non-ruminants in Philippine pesos (excluding feeds). Chickens DucksItem Pigs 21.89 2.53 15.56 7.56 0.00 74.68 0.00 0.00 0.00 0.00 122.22 1.18 0.00 0.00 0.18 0.00 0.00 0.00 0.31 0.00 0.00 1.67 0.09 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.09 Medicine Veterinary service Vaccination (labour and material) Deworming (labour and material) Drenching Breeding Salt Ropes Castration Other costs Total cost per head Improving the contnoution of livestock to crop-animal systems in rainfed areas in Southeast Asia164 86 61 61 00 98 61 63 05 61 00 96 Table 30. Meat production, marketing and consumption {ruminants and pigs). Item Cattle Buffaloes Goats Pigs 5 3 II 9 2 147 3 25 21 4 3 68 12 56 100 150 3 147 123 13 134 154 Meat Production Number of respondents practising animal slaughtering Average number of animals slaughtered (previous year) Average slaughter weight (kg) Average amount consumed (kg) Average amount sold (kg) Average selling price (Plkg) 100 2x lx ix ix 2x ix 2x 3x 1.00 1.00 126 0.85 1.00 130 0.88 0.75 103 2.40 1.59 100 Meat Consumption/Purchase Frequencyofpurchase (per week) Livestock growers Non-livestock growers Average amount purchased (kg/week) Livestock growers Non-livestock growers Average buying price (P/Kg) Further analysis of the respondents' behaviour revealed that livestock growers tend to buy meat more frequently than non-livestock raisers and in relatively greater quantity .This is probably because the retail prices of animal meat at the local market were considerably cheaper than the reported farm gate price of slaughtered meat. Unfortunately no data is available to explain this particular observation. Poultry Meat Production, Consumption and Marketing Tables 31 and 32 show the meat and egg production for poultry .It was found that most poultry raisers prefer to sell their animals on liveweight basis rather than slaughter them. Survey results further revealed that on an annual basis, poultry raisers typically slaughter 8 chickens and 4 ducks at an average weight of 1-1.1 kg. More often, poultry species particularly native chickens are generally slaughtered for home consumption or during special occasions, like when a visitor arrives. Thus, poultry raisers tend to procure their meat requirement less frequently and in much lesser quantity than the non-poultry growers. Crop-animal systems research in the Philippines 165 Table 31. Poultry meat and egg production, marketing and consumption. Item Chickens Ducks }9 8 O 2 4 1.1 .1 Meat Production Number of respondents practising animal slaughtering Average number of animals slaughtered Average slaughter weight (kg) Average amount consumed (kg) Average amount sold (kg) Average selling price (Plkg) 0 2x 4x Meat Consumption/Purchase Frequency ofpurchase (per week) livestock growers non-livestock growers Average amount purchased (kg/week) livestock growers non-livestock growers Average buying price (P/Kg) 1.57 4.5 98.46 6 25 14 O 11 9 }9 2 O Egg Production Average number oflaying birds Average number of eggs produced last month Average number consumed Average number hatched Average number sold Average selling price (P/Kg) Eggs Purchased/Consumed Average number of eggs purchased last week livestock growers non-livestock growers Average price bought (PJ 12 13 3.13 0.00 = No data 166 Improving the contribution of livestock to crop-animal systems In rainfed areas in Southeast Asia Table 32. Source of purchased meat and eggs. Cattle Buffaloes Goats Pigs Item No. % No. % No. % No. % Source of purchased meat Market No answer 40 0 100 0.0 6 27 18.2 81.8 5 26 16.1 83.9 26 O 100 0.0 Chickens Ducks No. % No. % Source of purchased eggs Market Store 51 2 96.2 3.8 o o 0.0 0.0 In terms of poultry egg production, a chicken raiser generally harvests 25 eggs per month with 6 layers. Farmers traditionally consume majority (56%) of the eggs produced, while the rest are sold outside the home. On the average, Don Montano residents generally purchase a dozen or so chicken eggs weekly from the local market at P3.13/piece. The data also shows that incubating eggs for possible hatching is not commonly practised in Don Montano. On the other hand, 19 duck eggs are reportedly harvested in a month 's time from nine ducks. This figure is also not surprising, as most farmers do not provide either commercial feeds or other supplements to their ducks. The survey also revealed that farmers do not commonly consume or purchase fresh duck eggs. Data on purchase of salted eggs and embryonated eggs or "balut", which are commonly consumed, were not captured in the survey. Land Holdings The average land area maintained by the respondents in Don Montano is 1.26 hectares (Table 33). This figure is smaller than the generated data from the municipal records, which is 1.5ha./farm family. The farm areas of Don Montano respondents are concentrated in a contiguous location on both sides of the provincial highway, about 1.19 km away from their residence. These areas serve as land for crop production and grazing areas for animals. Almost half of the farmer-respondents have a second parcel ofland, while a few maintain additional parcels situated in adjacent villages or near the mountainous areas. Majority of the respondents (63.6%) use their land for crop production while some use it as forest tree plantation ( 1.6%). Crop-animal systems research in the Phil,ppines 167 Table 33. Land holdings and land use. Number of Land Parcels )a(n=57) 2b (n = 22) 3c (n = 6) 4d(n= I) Ave %Item No. % No. % No. % No. % 55 96.5 1.8 0.0 0.0 0.0 20 90.9 4.6 0.0 0.0 4.6 4 0 0 0 2 66.6 0.0 0.0 0.0 33.3 o o o o 0.0 0.0 0.0 0.0 100 63.5 1.6 0.0 0.0 34.4 o o o o o 34 59.7 1.8 14.0 19.3 1.8 0.0 12 O 4 4 4 0 0 0 2 0 66.7 0.0 0.0 0.0 33.3 0.0 100 0.0 0.0 0.0 0.0 0.0 70.2 0.4 8.0 9.4 9.9 1.1 o o o o o 8 II 0 2 15.4 0 0.0 7.7n.a. n.a. n.a. n.a. 3 23.0 0 0.0 11.5n.a. n.a. n.a. n.a. 4 30.8 3 75.0 52.9n.a. n.3. n.a. n.a. 0 0.0 7.7 23.0 o 0 0.0 0.0 25.0 0.0 3.8 24.0 D.a. D.a. D.a. n.a. n.a. n.a. n.a. n.a. n.a. n.8. n.8. n.8.3 3 44 9 O 5.2 77.2 15.8 0.0 3 12 4 3 13.6 52.2 17.4 13.0 16.7 83.3 0.0 0.0 O I 0 O 0.0 100.0 0.0 0.0 8.9 78.2 8.3 3.3 5 O O 18 39 31.6 68.4 '16.7 83.3 0 0.0 100.0 6.3 15.0 n.a. n.a. n.a. n.a. Land Use Crop land F orest/trees Grasslands Orchards Others Ownership Status Owned Rented In Leased In Rented Out Leased Out No answer Tenurial Arrangement Share cropping Fixed rent after harvest Fixed rent before harvest Certificate of Land Transfer Mortgaged No answer Topography Upland Lowland not flooded Lowlandflooded No answer Soil ErosionlSiltation Observed Not observed 5 .Average land area is 1.25 ha; average distance from homestead is 0.925 km; average rental share is Philippine pesos 200 b Average land area is 1.07 ha; average distance from homestead is 0.82 km; Average rental share is Philippine pesos 200 c A verage land area is 1.71 ha; average distance from homestead is I km d A verage land area is I ha; average distance from homestead is 2 km n.a. = Not applicable Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia168 54.6 0.0 18.2 18.2 4.6 4.5 Majority of the land (70.2%) are owned by the fanner-respondents while the rest are either rented in, leased in, rented out or leased out (0.44%, 8.0%, 9.4% and 10.0% respectively). In tenns of tenurial arrangement, more than half follow the "fixed rent before harvest" arrangement with an average rental share ofP200.00. Most of the land is considered lowland but not flooded (78.2%) and thus soil erosion/siltation is not usually observed. Crop-animal Interactions Use of crop residues Residues of mainstay crops such as rice, onion, corn and vegetables are commonly utilised as feeds for livestock and as compost, mulch or housing materials (Table 34). Some farmers just leave the biomass in the field, while others make additional income by selling them. On the other hand, other farmers burn such residues. Almost one third of the respondents (32.2%) utilise their crop residues as feeds for their livestock. These crop residues include rice straw, corn stalk, and peanut and mungbean herbages. Rice straw are commonly collected and piled within their farm, to be utilised during lean months when there is not enough grasses available for cut and carry feeding, and when most of the farm lands are already planted to crops. On the other hand, corn stalks are usually cut and fed to their livestock. The area then doubles up as pasture area, whereas areas planted to mungbean, onion and peanut are used as pasture areas after harvesting. Also, close to 30% of the farmers use crop residues as compost materials. Other farmers utilise residues as mulch (0.56%) and as housing materials (2.2%). It was learned during the FGD that farmers who do not have ruminant animals are those who usually utilise their crop residues as such. Additionally, close to a third of the farmer- respondents practice burning of residues, while only a few (1.1%) make profit out of them. Use of animal manure The responses of farmers raising various species of animals indicate that manure from cattle (58%), carabao (64.7%) and pigs (50%) were distinctly collected (Tables 35-36). This is understandable because of the volume and the nature of the materials. Manure of animals under shed can easily be collected while those in pasture can be dried, caked and stored in pits. Majority of the farmers who collect animal manure use them as fertiliser for crops. Nobody among the respondents utilise manure as fuels or as feed for their livestock. Figure 3 shows the diagrammatic interactions of the crop and animal components in Don Montano. Crop-animal systems research in the Philippines 169 InQ)~"0'inQ)~0.O~u0c0';:.10~';:.=>~MQ):c101- -0\"'II6-~U -0\V}IIbN8 -0-onII~r'iPo9U 0>00~~< I=~ 08,=,0 In ~ 0" ~ ~ ~ og.u~u~= ~ ~=0~~ u > O u <~~?..I:~]'.I::::> ciZ ooor---o0'1o8NooNolr\lr\.fII'\ .qo N .qo ~ 0 ..,. "'I "'I -o o - V'\000- \000 00 ci ID Iri Iri 00 ci ci - 0;0;"'!r--.t'"!t'"!0;0; 0 O 1'-- O 1'-- O ~ 1'-- O 0...:.0 "oM...:.o 0000000000000.';00 -0 00 0 -- N 1'-- N 0- 1'-- 00 1'-- NoNo\ .00 . M N N MMN ~OMM ~..;~..;~\Ooci . M \O-OJ').qo NOOO\O~NNII') ~ 1"1 ~ ~ ~ ~ ~ .q-, .q-, .q-, 0 O N O "'";.~v-"' ~ ~ v - e "' .~ -0 - oo~ e- v .= .s Q. ~ -5 ~.Y. ~ e Q.~ v ~ o o ...v .s ~ e.= u~ ~~:s "' "' "' "' ~ 3 v 0 ??0~...JCfJ OOr-..'-0Mr-..OO'-0 N "1" V) V) 00 0 O r'\ O ~ r'\ r'\ 00 00 1,0 EQ)...In>InC)c.E~ca-occa...co~coOQ).c...c:2....~Inco.;;.uca~Q)....5i;E.ccaQ.o~(JMQ)~~C)ii: Table 35. Utilisation of animal manure (cattle, buffaloes, goats) Buffaloes (n=33) Goats (n=31 ) No. % Ave. % of use No % Ave. % of use Manure collection 2~ 58.1 22 64.7 3 10.3n.a. n.a. n.a. Manure utilisation As fertiliser As fuel As feed Other purposes Sold 25 O O O O 100 0.0 0.0 0.0 0.0 63.5 0.0 0.0 0.0 0.0 21 O O 8 O 95.5 0.0 0.0 36.3 0.0 67.9 0.0 0.0 53.8 0.0 3 O O 1 O 100.0 0.0 0.0 33.3 0.0 50.0 0.0 0.0 30.0 0.0 n.a. = Not applicable Table 36. Utilisation of animal manure (pigs, chickens, ducksl. Pigs (0=26) Chickens (n=53) No. Ave. % ofuse No. % Ave. % oruse Manure collection 13 9 16.9 1 4.00 n.a.50.00 n.a. n.a. Manure utilisation As fertiliser As fuel As feed Other purposes Sold 13 O O 4 O 100.0 0.0 0.0 30.8 0.0 72.0 0.0 0.0 35.0 0.0 9 O O 3 O 100.0 0.0 0.0 33.3 0.0 72.00 0.0 0.0 43.0 0.0 1 O O O O 100.0 0.0 0.0 0.0 0.0 n.a. = Not applicable 172 Improving the contnoution of livestock to crop.anlinal systems in rainfed areas in Southeast Asia 100.0 0.0 0.0 0.0 0.0 IMPLICATIONS OF THE RESULTS Identification of Recommendation Domains The particular site selected for the BMS is Umingan in eastern Pangasinan. It is one of the 45 municipalities ofPangasinan and is strategically located adjacent to trading and commercial centres, R & D institutions and other government support agencies. Umingan is more or less similar to the other Eastern Pangasinan municipalities, but it stands out, as it has the biggest proportion of agricultural land and rainfed areas. In Umingan as in the whole province, livestock is considered an integral part of the farming operation, with cattle and buffaloes as the most preferred livestock species. Moreover, most of the 11,308 households are engaged in agriculture. Just as in other eastern towns, a number of residents in Umingan are involved in non-farming activities such as carpentry, masonry, contractual work in other fields, and tricycle driving. These are done particularly during dry periods when there are few activities in the farm. These off-farm incomes however help increase the people's buying power and their capacity to buy livestock products and farm inputs. In Don Montano, 90% of the residents are also involved in farming. Just like in the other towns of the province, most families have 1.2 -1.5 ha. of farm land, most of which are divided into at least two parcels. Majority of the farmers use these parcels for crop production. During the wet season, rice is the major crop grown and for the dry season, onion is the predominant crop. From these farms, each household generates an average monthly income ofP 4,000. It was gathered that 15-20% of this comes from livestock production, which is viewed by residents as an equally important source of Income. In Don Montano as in the whole province, livestock production is an integral part of the farming system. Itis carried out by smallholder farmers with limited landholding and little or no inputs. HenCe, animal productivity is generally low. As in the wholeI province, farmers in the area share common problems that usually stem from lack of financial resources to augment and sustain crop-animal production. Based on interviews with key informants, farmers are, however, receptive to change and open to innovations that would give more tangible benefits. In essence, the chosen focal village is representative of the whole region. From the observations made during the series of site selection activities in the whole of Region 1 up to the conduct of the FGD, it can be concluded that the characteristics of the focal site and that of the rainfed areas in the municipality, province and region are similar . Major Constraints and Potential Improvements The perceptions of farmers about their farm problems and possible solutions generated during the survey agreed well with the responses gathered during the FGD (Table 37). As expressed during the survey and the FGD, Don Montano and Sta. Crop-animal systems research in the PhilIppines 173 Table 37. Problems identified, perceived solutions and interventions agreed between the Don Montano and Sta. Maria farmers and the research team. Problem Identified Perceived Solution Intervention Agreed Upon Livestock-related problem Limited grazing areas and thus, limited feed resources Introduction of feed banks and planting of alternative feed sources Establishment of modeVdemo fanns in the fanners' field Provision of seeds/planting materials of improved forage species Conduct of training courses on livestock management, alternative feed sources, and silage production Crop-related problems Limited cropping options due to decreasing soil fertility, inadequate water sources & high production inputs Introduction of other options after soil analysis by experts Biophysical assessment of soil fertility & productivity Planting of legumes to improve soil quality Conduct of training courses on fertiliser handling and management Construction of irrigation facilities c/o NIA Development of fertiliser recommendation for each crop based on results of the biophysical assessment for wet and dry season Monitoring and evaluation of crop perfonnance Introduction of legumes in farmers' cropping system Socio-economic problems Limited financial resources to improve cropping system and increase animal holding Provision of more accessible credit schemes Involvement of labour to reduce production cost Dispersal of animals through DA Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia174 Maria, just like the rest of the rainfed areas in Umingan, are currently constrained by the following factors that relate to the individual farmer, his farming operations, and his linkages (Figure 4). These problems were linked to one another to trace their root causes (Figure 5). Considering the primary data gathered from the survey and the FGD, the above problem tree analysis was set forth by the Philippine Team. This diagrammatic analy- sis spells out the root causes of low farm productivity as felt by the farmers in Don Montano and Sta. Maria, Pangasinan. From Figure 5, it can be seen that low farm productivity stems from lack of capi- tal to improve cropping systems and increase animal holdings. These problems in turn were traced to lack of accessible credit facilities to expand current operations, and the generally low income derived from agricultural operations. Moreover, the need to sell animals during emergencies adds to the problems of sustaining the limited animal hold- ings of farmers. Income from the crop-animal system was analysed to be quite low. Talks with farmers revealed that they earned an average of only P 4,000 ($100) monthly which is not enough to support all personal needs plus their farming requirements. This prob- lem on income emanates from low crop yield coupled with low market price for farm products and animals raised. Low yield was traced to problems on small farm size (which cannot be increased due to lack of financial resources), decreasing soil fertility and limited cropping op- tions. Again, the overriding factor here was lack of financial resources to add cropping options. The El Nino-La Nina weather problems, lack of projects to maximise the rainfed areas, inadequate irrigation facilities and high production inputs all add up to aggravate the problems on cropping systems. Concerning animals, low market price was traced to the availability of poor quality animals, which resulted from having malnourished animals. Although farmers would not admit having malnourished animals, their feeding calendar and practices painted during the FGD point to an expectedly poor ADG. Moreover, lack of feed sources especially during the dry season adds to the nutrition problem. Although tree legumes abound, these are seldom utilised, for the farmers were not used to the practice. They merely allowed cattle, carabaos and goats to feed on whatever grasses and weeds are within reach. The only additional input they give to the animals is water. Even rice bran, which is plentiful in the area; are not utilised as supplement for ruminants during summer, although they are given freely to swine. Animals were also seldom bought from other farms but selected from their current herd. This caused much inbreeding and growth stunting. All these problems were traced to lack of training on livestock management. Although the DA conducted many seminars, these were usually on crop management and very little on livestock. Majority of the seminars on livestock dealt with kinds offeeds and veterinary products, as these were conducted by feed and drug suppliers/companies. 175Crop.aminal systems research in the Philippines clac.inlaCIClac..la.~la~la...C/)~ClaOCla...CO~COC.5C/I...~E...la-~~-c.iQ...>.c~~....iQC/IE~:co...c..-t~...~CI~ ..;n .~ "'0) ] 1)-",0" " .. , ?!l e - "Q ~ u j,..::.g.2~ O u , ~ ~].E~~ 8: c "a ~~..2 ~~~ QOQOc c .9 .~ ~ g- ~ 011 ~ -J,... ~ .,]~.c..~ co=.6.e...~~ 1 t ... c~c.Ui~CIC~Q"ti.-=~~IU-U)'Cc~oc~-co~coCc~Q)E...~-Q).c->-.c'CQ)u~tn-c.e-tncouQ).c-'0tn.~~c~Q)~-EQ)~0...Q"Iri~~CIii: ~~ When these root problems surfaced during the FGD, the farmers together with the Team identified the following interventions as necessary . LESSONS LEARNT AND EXPERIENCES GATHERED In any endeavour, there are experiences and lessons learned that could lead the person to perform his activities better in the future. The same is true with the characterisation of the BMS. Listed herein are some of the major obstacles faced by the Philippine Team and how these can be avoided in the future: I. Although there was a prior need for the survey questionnaire to be comprehensive, the data generated was found to be too exhaustive, to the point that some data were too stratified and too detailed to be of much relevance to the researchers. Some of the detailed questioning (if really that crucial) could have been done as part of the individual studies later on. Moreover, as was discovered during the Team's encoding, much of the BMS figures were meaningless, even if compared across countries. Hence, it was felt that there was a waste not only of financial resources but more so of time. 2. While survey research has been a traditional tool offering replicability , quantifiability and generalizability , it was found that it was a weak tool in generating data in Umingan. For one, it was too long to elicit the needed precise data from the farmers who do not usually keep records of events in their lives. It was also too structured that it was obviously boring to the respondents, as it robbed them the spontaneity associated with face-to- face interaction. Moreover, the encounter with the respondents was too short and impersonal and questions were arranged in a rapid-fire fashion, that respondents rarely had enough time to recall, compute, or comprehend the impact of the questions. Consequently, the Team has to go back to the site to validate computations and figures, as the farmers were unable to identify and articulate their circumstances properly. 3. Although senior researchers maintain that surveys and interviews are potent tools for social research, the Philippine Team however felt that the participatory rural appraisal technique done, which was in the form of the storymapping- focus group discussion technique (Annex), was far better in generating the needed data for the characterisation of the BMS. Although it was done merely to valid~te the survey data, it was felt that it could have been the better way to gather data. For one, the same information was generated, pointing to the fact that indeed it can be used as an alternative to the traditional survey method. Second, it was able to generate not only~ 178 Improving the contribution of livestock to crop-animal systems In rainfed areas in Southeast Asia figures or data but in-depth explanations of farm realities as well. Third, as it involved one cluster of people at a time, there was sharing of experiences that either confirm or refute their veracity .Thus, there and then, there was validation of utterances and ultimately, there was consensus. Hence this technique was found more useful in generating the needed data and in getting farmers' consensus on possible technological interventions for the site. Therefore it is recommended for similar endeavours in the future. Crop-animal systems research in the PhIlIppines 179 REFERENCES Bureau of Agricultural Statistics. 1998. Commodity Fact Sheets. 1997 data. Bureau of Agricultural Statistics. 1998. National Statistics. Bureau of Soils. 1940. Soil Survey Report: Province of Pangasinan. Manila: Bureau of Printing. 59pp. Bureau of Soils and Ministry of Agriculture. 1982. Land Resource Inventory. (in- house publication) 199pp. Department of Agriculture. 1998. Medium Term Food Security Plan 1999-2004 for Pangasinan. Mimeographed. 6pp. Department of Agriculture. 1998. Municipal Agriculture and Fishery Development Plan. 1999-2004: Umingan. Mimeographed. 2pp. Devendra C. and Chantalakhana, C. 1992. Development of Sustainable Crop-Animal Systems in Asia. In: Proc. 6th AAAP Animal Science Congress. Vol. 1. The Animal Husbandry Association of Thailand, Kasetsart University , Bangkok, Thailand. p.21-39. National Statistics Office. 1991. Census of Agriculture: Pangasinan. Final Report, Vol.2. 100pp. Sevilla C. C. 1994. Prospects for Sustainable Beef Cattle Production in the Philippines. The Philippine Agriculturist. Vol. 77, No.2 (Apr-Jun). p 271-284. (Felix D. Maramba Professorial Chair Lecture delivered on Feb. 28, 1994. IAS, College of Agriculture, UPLB). Villar E.C. et al. 1999. Crop-Animal Systems: Country Proposal. PCARRD, Los Bafios, Laguna (paper submitted to ILRI Workshop) Improving the contribution of livestock to crop-animal systems In rainfed areas in Southeast Asia180 ANNEX ILRI-PCARRD Crop-Animal Systems Project Livestock Research Division PCARRD, Los Balios, Laguna 4030, Philippines Participatory Rural Appraisal-Focus Group Discussion: Proposed Mechanics What is PRA-FGD? Participatory Rural Appraisal with emphasis on Focus Group Discussion is a participatory planning method, which centers on talking, acting and learning with the target beneficiaries. It is a modern tool in social science research that is considered more culturally appropriate in capturing farmers and rural audiences' realities. It deviates from the classical approach that is expert-driven and which imposes a "problem analysis on the target group derived mainly from large scale data analysis, quick countryside visits and brief communication with local male elites by high profile experts and government officials" (Kuhn, 2000). Promoted by Robert Chambers, PRA evolved as a prominent method in the late '70s-'80s which enabled outsiders to gain information and insights from local people about local conditions. This method has served as a supplement if not an alternative to the traditional survey method. While survey research has been a traditional tool offering replicability, quantifiability, and generalisability, it is a weak tool for generating data for the following reasons: 1. It is too top-down, assuming too much about the subject's realities even before actual encounter; 2. It is often too long to elicit the needed precise data from people who do not usually keep records of events in their lives; 3. It is often structured that it becomes boring, robbing the subject its spontaneity and making probing a prying matter; 4. Farmers do not often appreciate the value of the data gathered from them as they do not see where or how they are used; seldom are results of such survey fed back to them; S. The encounter is short and questions are asked in a forced, rapid-fire technique that subjects rarely have enough time to recall, compute or comprehend the impact of the questions. Consequently, they are unable to identify and articulate their circumstances properly; and 6. Respondents of survey researches have expectations of what is required of them; hence, they slant their answers to the perceived "proper response". 181Crop-animal systems research in the Philippines Although researchers maintain that surveys and interviews remain as potent tools for social research, they must however be used appropriately and in tandem with other tools. Cognizant of the limitations of survey research especially among farming communities, it is therefore imperative to adopt a new technique that involves participatory diagnosis of the farmers' realities. This involves combining the quantitative and qualitative research methods as well as restructuring the traditional top-down power relationship between researcher and the researched towards a more democratic participatory nature. Storymapping: An Alternative Technique A variety of techniques and tools have been developed to carry out PRAs. The World Bank listed some of the most common techniques in extension (Kuhn, 2000): 1. Semi structured interviewing 2. Focus Group Discussion 3. Preference Ranking 4. Mapping and Modeling 5. Seasonal and historical diagramming In the field of communication research, a technique known as "story mapping" has been used to incorporate the salient features of the abovementioned tools. This technique relies heavily on visual tools that aid overcome communication difficulties between outsiders, facilitators and local people. This serves as springboard for storytelling, discussion and conscientisation. In this technique, participants are grouped in small clusters and made to draw and story tell their farming situation, experiences and realities. In story mapping, both researcher and participants are involved (Barroga, 1991 ). The participants, who cannot see the totality of their experiences, offer bits and pieces of information, which become the maps. The researcher helps connect the maps in the story to form a total picture of the participants' life experience. After the story map is painted, the researcher uses FGD and modeling to focus the discussion on specific solutions to farming problems that may either be perceived or not by the participants. Using the spontaneous probing technique or semi-structured interview, the researcher poses farming problems and possible R&D interventions to gather pertinent hard data (e.g. production data). Sharing of experiences, thoughts and beliefs is done to confirm or refute relationships in their story maps. Although limited in its representativeness and generalisability due to the small sample involved story mapping technique is nevertheless high in internal validity . Proposed Methodology Research Design In generating data about farmers' realities, the research shall employ a triangulated approach. This will involve the story mapping technique, wherein clusters of people will be made to story-tell. Results will be triangulated with the existing socio- demographic data initially gathered from interviews with the MAO, the survey data gathered from farmers, as well as with the site appraisal conducted by the Team. In essence, it will utilise more of the qualitative research techniques. Sampling Procedure. Small groups of 6-10 people per cluster shall be involved in each story mapping session. To.ensure representation, clustering will be done, with farmer leaders from contiguous farming areas constituting each cluster. Through the assistance of the Barangay Captain in the focal site, the participants for each cluster shall be named. They shall constitute the following clusters: .Farmers' Cluster-farmer leaders who are aware of problems on crop- livestock integration and who are articulate and analytical enough will be involved. They shall be stratified by gender . .Development Workers' Cluster-village leaders from Don Montano, the Agricultural Technician and Veterinarian assigned to the Barangay, traders/buyers of animal stocks, feed suppliers, and breeders in the area shall be invited to participate; Since a story map is not complete without researchers to help the participants reconstruct farm realities, the study shall involve a team of experts consisting of the following: .Communication specialist as facilitator .Farm experts (i.e. animal, crop, soil, marketing experts) .Videographer/Cameraman Research Focus. The research shall focus where the preliminary rural appraisal and the BMS survey left off, i.e. to discover the causes oflow income or low productivity in the area*. In short, it shall revolve around the question, "What can be done to increase productivity?" Research Procedure 1 Introduce purpose of study. State clearly the objectives of the activity: .Explain interest in studying farming practices in Don Montano .Explain need to learn from them farming factors, which they think can improve farm productivity. 183Crop-animal systems research in the PhIlippines 2. 3 .Also explain focus of the Project which is on livestock, hence explain need to generate their perceptions on livestock fanning-what animals they want to raise and why, their perceived potentials of and problems with crop-animal fanning, etc. .Explain need for the entire proceedings to be videotaped for documentation purposes. To pave the way for better rapport, ask each participant and researcher to introduce himself, citing what he does for a living, etc. Divide farmers' group by gender. For the Development Workers ' group, segregate into business group (i.e. traders, feed suppliers, breeders/supplier of stocks) and government group (MAG, A T , Vet, and Barangay officials). Then ask each group to do the following: Questions to Validate Survey Data and Probe On Possible Interventions ( in Pilipino ) Activity Draw a community map of Don Montano, complete with all the physical resources and farm elements-roads, bridges, mount- ains, pasture areas and farms. .Indicate crops and animals raised, labour force, machines used and where farm inputs are bought (fertiliser, seed and feed sources). .Indicate where their produce are marketed (i.e. marketing channels including nearest market, transport system used to market goods, etc.) .Ask them to show all inputs, events, and structures in their production cycle that contributed to their income. Ask them to name where these incomes are spent, connecting each element to the expenditures. Allow each group to choose their leader and have him/her present the story map. For clarity, ask the 2 184 Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia Questions to Validate Survey Data and Probe On Possible Interventions ( in Pilipino ) Activity participants the relationship of each element. Put arrows to connect them. Together with facilitator, farmers are then asked to connect elements together until all the elements relate to form a complete picture of the farming situation. 3. Facilitator re-presents diagram to participants to allow them to see their situation in a holistic manner. Each of the farm experts is then allowed to ask questions to gather pertinent hard data. For example. they ask the farmers how large their farm is, how many farm family labour they have, how many animals are raised for breeding. how many for work, etc. Anu-ano ang tanim ng magsasaka dito sa Don Montano? Anu-ano ang a/agang hayop? Saan bumibi/i ng aa/agaang hayop. ferti/iser. seeds. magkano. saan ginagamit ? Ano ang ipinakakain sa hayop? Saan kumukuha? How sustainab/e ang supp/y? Saan nagbebenta ng ani? Sino ang bumibi/i? Ano ang pinagsasakyan ng paninda? Ano ang pinagkakagastusan ng karamihan sa taga-Don Montano? 4 At some point, fanners may mention problems. Probe them deeper until you can trace in their story maps the root cause of such problems. Ask them, "How can you rise from this problem?" Point back to each problem-relationship and ask how it can be manipulated to benefit them. Probe gender differences in terms of labour allocation, preference for livestock species, access to and control over resources and benefit ( employ preference ranking here). 5 If the participants do not mention problems, pose farming problems and possible R&D interventions. Try manipulating the relationships in their story maps or introduce another element. This process is added to determine what other interventions will be amenable to Questions to Validate Survey Data and Probe On Possible Interventions ( in Pilipino ) Activity them. Never impose an intervention if there is resistance; or try to determine and if possible, solve the cause of the resistance. 6 Facilitator then asks second question: "In your opinion how do you think can you make your situation better?" After some discussion, Team members focus on livestock aspects and pose farming problems and possible R&D interventions by manipulating relationships in the diagrams or by introducing another element. The facilitator then point back to each problem relationship and ask farmers how it can be manipulated to benefit them. This process is added to determine what other interventions will be amenable to them. Kung mabibigyan ng pagkakataon na mahatiran ng impormasyon tungkol sa agrikultura. ano ang gustong "mode .. (preference ranking) -training. polyeto. komiks. programa sa radyo. etc. 7. Validate the final picture and ask participants how many of their co- farmers do they think will have the same opinion. Allow sharing of experiences to confirm or refute relationships in the story maps. Analysis of Story Maps. Using the videotape, Don Montano's realities can be analysed by element. The story maps can be synthesised into a Farmer's Decision Tracking Grid. This grid traces actions, events, resources, and decisions made by the farmers. This grid will also explain why productivity is low. However, even without the grid, the Team can summarise what farmers in Don Montano want and how they feel things ought to be done. In a real participatory research, this information should serve as the stepping-stone for a more detailed planning activity, i.e. participatory planning of the interventions at hand. Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia186 CROP-ANIMAL SYSTEMS RESEARCH IN THAILAND M. Wanapat, A. Polthanee and C. Wachirapakorn Faculty of Agriculture, Khon Kaen University, Khon Kaen, 40002, Thailand ABSTRACT The household survey results are presented with reference to Amphur Muang in Mahasarakham province in Northeast Thailand. Rice is the predominant crop but cassava and sugarcane are also widely grown. Dairy production is of emerging importance and studies focused in all 38 dairy cattle farms and 100 non-dairy farmers serve to provide comparisons. On the average, each dairy farmer had 9.7 head of dairy cattle, and 40% of them also raised chickens, ducks and beef cattle. Milk was sold to the cooperatives. A variety of crop-animal interactions were apparent, and manure was used as fertiliser . Major constraints identified were nutritional inadequacy, dependence on commercial feeds at high cost, low reproduction rate and inadequate training. Proposed interventions include intercropping cassava with cowpea and lucaena, cowpea with rice, assessment of feed resources and availability , concentrates and supplementation, all of which address all-year round feeding systems. INTRODUCTION Crop-animal production systems in Thailand are commonly practised by smallholder farmers throughout the four regions of the country (northern, northeastern, western and north). Dairy farming for smallholders has been currently promoted and encouraged as a means to produce milk supply, increase income and improve standards of living (Chantalakhana, 1994). Many factors, especially feeds and feeding, contribute to the efficiency and success of dairy production as well as other associated dimensions (Wanapat, 1999a). The objectives of this study in the benchmark site (EMS) at Mahasarakham are focused on essential characterisation and other factors that can be used to improve existing dairy-crop production systems. BRIEF DESCRIPTION OF THE BMS (MAHASARAKHAM) Location The study was conducted at Amphur Muang district of Mahasarakham province in Thailand. Mahasarakham province is situated in the middle of the Northeastern region (Figure I). It lies roughly between longitude l5?25'N and latitude IO2?50'E, and is Crop-animal systems research in Thailand 187 188 Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia Meteorology and hydrology The study area is in the Asian monsoon zone, which is characterised by a rainy season from May to October, and a dry season from November to April. Annual rainfall is about 1,147 mm, of which about 90% falls in the rainy season, and the remainder in the dry season without any effect on crop growth. Annual mean monthly maximum temperature varies from 30.9?C in November to 38.3?C in April, and annual mean monthly minimum temperature varies from 15.7?C in December to 24.4?C in April. Annual evaporation amounts to approximately 1,630 mm. Water resources The Chi river flows across the province extending 60 km in length. In addition, there are 12 water creeks scattered in the province and the water level would normally be low or dried up in the dry season. In Amphur Muang district, there are 11 water reservoirs, 11 weirs, 71 swamp and ponds and 522 underground water wells. Socio-economic Aspects Land use and distribution The distribution of agricultural land use systems of Amphur Muang district and the province are shown in Table 1. Approximately 63% of land is devoted to rice followed by idle areas (14%) and grazing area (7%). Land use and distribution of Amphur Muang and Mahasarakham Province.Table 1 A.Muang Province % of Total Ha % of TotalHa 27,425 602 2,203 390 63 2 5 329, 24, 10, 4, 49, 20, 71 5 2 Rice Upland crop Fruit trees Vegetables Grazing area Household Others 10 5 3,120 4,069 5,944 7 9 14 21,815 5 189Crop-animal systems research in Thailand 388 024 897 148 462 785 Crop Production Crop production in farms with dairy cattle was the main activity (Tables 2 and 3). Monocropping with rice was the main crop particularly in the wet season. Some farmers also grew other crops such as cassava and corn. These were similar in both the dairy group and the non-dairy group. The first cropping season was mainly rice while cassava and other crops were also grown as upland crops. The second cropping season was corn and cassava but the extent was low as compared to the whole group. Farmers in the dairy group hired labour for crop production. Manure was the favoured fertiliser over inorganic fertiliser. In the non-dairy group, however, farmers favoured the use of inorganic fertiliser because of lack of availability of manure and the resulting high crop production from using inorganic fertilisers. Farmers in the non-dairy group also hired off-farm labour at a similar level to the dairy group. Table 2. Characteristics of first cropping season. Crop grown Item Dairy group Non-dairy group Rice Cassava Rice Cassava 2.3 2,618.6 440.1 480,504.7 0.8 5,500.0 2.7 3,491.6 394.7 1,728.7 2.4 0.8 8,~ 2 1,]200.0 0.5 1 102.9 7.5 124.4 10.1 4 4.3 D.a 4.6 16.4 115.3 3 11.8 39.3 7 7.5 19.8 Area planted (ha} Quantity produced (kg} Inorganic fertiliser (kg} Animal manure (kg Insecticides (kg/li} Herbicides (kg/li} Seeds (kg} Number of labourers hired Number of days hired labourers worked Number of tractor hours hired Number of draft animals hired Number of days animals worked 14 7 = No data; n.a. = Not applicable. 190 Improving the contribution of livestock to crop-animal systems in rainfed areas In Southeast Asia 0.8 177.8 ~05.8 28.6 Table 3. Characteristics of second cropping season. Crop grown Non-dairy groupDairy groupItem Rice Maize CassavaMaize Cassava 1.4 3,560 212.2 1,410 3.7 0.8 2,333 350 0.6 16,000 66.7 2.2Area planted (ha) 0.2 Quantity produced (kg) 9,100 Inorganic fertiliser (kg) 25 Animal manure (kg) 1 ,500 Seeds (kg) 1 Number of labourers hired Number of days hired labourers worked Number of tractor hours hired Number of draft animals hired Number of days animals worked - 160 173 8 168 0.25 -= No data. Major crops Most upland crops are commonly cultivated. The planting area and yields of the major crops in the Amphur Muang district and province are summarised in Table 4. Table 4. Types of major crops, planted area and yields in Mahasarakham (19971 1998). Yield (kg/ha) A. Muang Province Planted Area A.Muang ProvinceCrop %of Area %of Area HaHa 3,269 2,938 4,506 14,406 55,600 1,563 1,298 2,713 3,438 3,944 8,475 61,781 1,625 1,325 58 33 2 4 2 0.005 0.4 58 35 4 209,554 119,833 8,173 14,400 8,194 18 1,326 17,121 10,304 1,296 302 212 2 86 Glutinous rice Non-glutinous rice Second rice Cassava Sugarcane Kenaf Tobacco 0.7 0.006 0.3 The major crop in the Amphur Muang related to the study area is glutinous rice where rice cultivation occupies about 93% of total planted area. Next to rice, cassava and sugarcane are planted in large areas. However, about 2% of the study area is upland where the major crop is cassava. Sugarcane is planted on a contract basis and the area grown for sugarcane is currently expanding as there is a sugar mill existing in Mahasarakham province. Rice in the study area is usually harvested manually by farmers and dried in the same field for two to three days. Almost all the dried rice is threshed in the field by hired machines with an operator, for immediate sale of the surplus through local traders after some has been secured for domestic consumption. Sometimes rice is stored in a farmer's storehouse for timely sale. Marketing of threshed rice is conducted in the paddy field at the provincial level and sometimes at the regional level by local traders, middle men and large-scale rice millers. The marketing of rice is also carried out by agricultural cooperatives and the Bank for Agriculture and Agricultural Cooperatives (BAAC) for their member farmers at the provincial level. Farmers can select a buyer at any time but as small-scale farmers, they do not have a beneficial marketing route. They sell their produce from the field or at the farm gate for an unreasonably low price. Harvested cassava is commonly sold immediately to traders or middle men. The crop is not dried or processed in the study area. Sugarcane is generally planted on a contract farming basis and immediately loaded on a truck sent by the owner/contractor after harvesting. Other agricultural crops such as vegetables, fruits, are planted but they are generally cultivated for domestic consumption. Livestock and fishes are also raised. Animals Cattle, buffaloes, pigs, ducks and chickens are the major animals raised (Table 5). Buffaloes are raised for draft but they are decreasing in number in proportion to an increase in mechanical power cultivators and a high rate of slaughtering. Chickens and ducks are raised for domestic consumption. Mahasarakham province promotes the development of animals which is set as one of the strategies of the provincial development plan and some farmers in the study area are also raising beef cattle. Table 5. Animal population in Amphur Muang district and province IWanapat, 1999b). Animal A.Muang Province % of Province Cattle Buffaloes Ducks Chickens Pigs 18,453 11,044 24,616 123,129 5,930 136,137 90,196 284,423 1,444,267 65,251 14 12 9 9 9 192 Improving the contnoution of livestock to crop-animal systems in rainfed areas in Southeast Asia Potential Impact of the Project Low household and fann income was observed in Mahasarakham province (Table 6). Fann income in Mahasarakham is less than 41% of total household income and about 43% of average fann incomes in the whole country . Table 6 Household and farm income and expenditure (Baht/year). Item Whole kingdom N.E region Mahasarakham Household: Income Expenditure 108,088 90,855 78,015 62,926 63,723 61,210 Farm: Income Expenditure 61,817 32,006 31,191 19,386 26,029 14,637 In order to improve the standard of living of the poor, it is considered essential to promote and enhance integrated agricultural development projects in Mahasarakham province and northeastern region where poor rural people are concentrated. It is hoped that this will achieve increased income for them and an improvement in the quality of their lives in parallel with government efforts in general for establishment of job opportunities as well as securing sources of income in the regions and rural areas. Integrating dairy cattle raising into the existing farmer's farm is one of the strategies of the provincial development plan. Dairy production has the potential to generate a farm income of 120,000-360,000 Baht per year. RESEARCH METHODOLOGY Site Selection The Northeastern region covers the largest portion of the area, as compared with other regions. It contains four main agro-ecological systems: Western Hills, Korat Plateau, Mekong Provinces and Southern Hills. Korat plateau includes Changwat, Udon- thani, Kalasin, Khon Kaen, Mahasarakham, Roiet, Yasothorn and some parts of Chaiyaphum, Nakhonratchasima, Burirum, Surin and Srisaket. Mahasarakham province is promoted as an area for dairy production by the Department of Agricultural Extension. Mahasarakham is divided into 11 districts. Amphur Muang district is purposively Crop-aminal systems research in Thailand 193 selected for the study, based on dairy farming and development during the last two years. Diagram showing the process of study site selection is shown in Figure 2. Household Selection Only 38 farmers raised dairy cattle in Amphur Muang. Therefore, there was no sampling procedure for household selection. In the present study, the 40 non-dairy farms were randomly selected in Amphur Muang to understand prevailing farming systems. Based on geographical factors 'Y I Northern I ~ ~~ I South I Based on agroecological factors . "f' ... Based on provincial promotion of dairy .production Based on dairy farm during the last years production and development Figure 2. Diagram showing the process of study site selection Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia194 Planning and Preparation for Fieldwork The team members were introduced to the Rapid Rural Appraisal (RRA) methodology (Anonymous, 1985). The application of the RRA methodology and the objectives of the study were discussed, and subtopics and conceptual tools to be used in the interviews were explained and clarified. Pre-test of the subtopics was also conducted before actual household interviewing. All team members were introduced and acquainted with the EMS and all farmers before the actual interviews. Data Collection Field investigations were carried out with two household groups: dairy production (38 fanns) and non-dairy production (40 fanns). Infonnation about the households was collected according to the subtopics prepared by the ILRI team. The interview team consisted of 3-5 members. Appointments with the fanners were made beforehand and each interview lasted for about 3-4 hours. If data were incomplete, the team members had to return several times to gather the remaining infonnation. Data Analysis Both quantitative and qualitative data were obtained from the field. Most of the data were discussed in the field during the process of interviewing and during the team meetings. All data were tabulated and encoded following the procedures given by ILRI. The quantitative data were processed by the use of simple statistics. RESULTS Household Characteristics The actual survey was carried out between September 1999 to January 2000 and all data were accumulated and analysed as shown in the following tables. Household characteristics of farmers who run dairy farm and non-dairy farm exhibited similar characteristics as shown in Table 7. Most respondents were husbands who had elementary education. Male and femal~ ratios were 68 to 32 and 60:40 in the dairy and non-dairy groups, respectively. Training in livestock was evident in the dairy group while very little could be seen in the non'-dairy group. The average number of household members for both groups was similar (4.9 and 4.2 persons). Crop-animal systems research in Thailand 195 Table 7. Household characteristics of respondents. Characteristic Dairy group (0=38) Non-dairy group (n=40) Percent of respondents Respondent Husband Wife Son Daughter 55.3 26.3 13.2 5.3 50.0 30.0 10.0 10.0 Sex 68.0 32.0 60.0 40.0 38.4 38.3 48.5 45.2 Male Female Age (years) Male Female Level of education Grades 4-6 (Elementary) Grades 7-9 (Secondary) Grades /0-12 (Secondary) University 68.4 15.8 13.2 0.0 87.5 5.0 5.0 2.5 65.8 34.2 0.0 77.5 20.0 2.5 13.2 86.8 87.5 12.5 O 2 2 36 32 69 II 3 14 6 7 13 4 4 3 8 38 34 72 5 4 10 6 2 8 4 Training on crops None Short term courses Vocational school Training on livestock None Short term courses Household members by age group) Males >60 Females >60 Sub-total >60 Males 16-60 Females 16-60 Sub-total 16-60 Boys 6-15 Girls 6-15 Sub-total 6- 15 Boys <6 Girls <6 Sub-total <6 Average size of household (n) Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia196 .5 .1 .6 .9 .6 .5 .2 .2 .4 .4 .0 .4 .9 .7 .5 .2 .6 .1 .9 .9 .1 .0 .5 .4 .9 .2 Animal Ownership In the dairy group, farmers owned about 9.7 head of dairy cattle including milking cows (5), heifers (2.7) and calves (2). In addition, there were about 40% of farmers in the dairy group who also raised chickens, ducks and beef cattle while only two farmers owned buffaloes (Table 8). Farmers in the non-dairy group owned beef cattle, chickens, ducks and buffaloes in similar numbers as the farmers in the dairy group. When farmers were asked for the new species of animals which they preferred to start to raise ( except dairy cattle ), most of them answered beef cattle and buffaloes because of favourable market conditions (Table 9). Dairy cattle were the most preferred stock for raising as they provide a good source of income (Table 10). Table 8. Animal ownership of farmers Cattle Item Buffaloes Pigs Chickens Ducks Dairy Beef 100 14.5 4.4 1.6 7.9 9.8 1.0 5.0 15.5 -I 2.6 4.0 47.4 36.1 14.3 23.6 13.8 16.0 3.0 ~ 101.0 4.0 4.0 5.8 4.0 1.0 10.0 53.0 17.5.Q 1.0 9.7 13.5 71.4 15.0 2.825.3 4.0 Dairy group (n=38) Farmer-owned (%) Starting No. (hd) Born (hd) Bought (hd) Shared in (hd) Shared out (hd) Gift in (hd) Barter in (hd) Barter out (hd) Died (hd) Sold alive (hd) Slaughtered in the farm (hd) Current (hd) Non-dairy group (n=40) Farmer-owned (%) Starting No. (hd) Born (hd) Bought (hd) Shared in (hd) Died (hd) Sold alive (hd) Slaughtered in the farm (hd) Current (hd) 25.<) 7.8 1.~ 3.0 4.<) 5.1 7.5 13.0 5.0 2.0 2.0 3.0 55.0 54.4 18.5 37.5 34.9 7.0 4.0 5.0 n.a n.a n.a n.a n.a n.a n.a 21.0 21.0 20.0 5.9 1.0 3.0 1.0 2.3 20.0 10.9 11.0 21.9 n.a. n.a. 6.6 15.06.2 = No data; D.a. = Not applicable 197Crop-animal systems research in Thailand Table 9. Reasons for starting with new animal species. Reason Beef Buffaloes cattle Pigs Chickens Ducks Fish Percent of respondents Dairy group (n=38) Start with new species Yes No 73.7 26.3 59.5 40.5 10.5 89.5 34.2 65.8 15.8 84.2 15.8 84.2 14.3 13.6 0.0 0.0 Reasons for starting with species Good as assets Favourable market conditions Provide draft power Government dispersal/ promotion program Improve family diet Others 0.0 67.8 7.1 9.1 77.3 50.0 n.a. 0.0 n.a. 0.0 D.a. 33.3 D.8. 3.5 0.0 3.5 0.0 0.0 0.0 0.0 100.0 0.0 0.0 100.0 0.0 0.0 66.7 0.0 Non-dairy group (n=40) Start with new species Yes No 67.5 32.5 72.5 27.5 17.5 82.5 32.5 67.5 13.5 87.5 2.7 97.3 Reasons for starting Valued as assets Favourable market conditions Provide draft power 7.03.8 16.7 76.9 66.7 100.0 92.3 3.8 31.0 62.0 83.3 n.a. 23.1 D.a. 33.3 n.a. 0.0 n.a. D.a. =Not applicable; -= No data, Table 10. Preferences for animal species. Item Preference ThirdFirst First Second Third Percent of respondentsPercent of respondents 2.7 89.2 0.0 4.3 5.4 21.7 2.7 17 8 4 25 17 37 26 0.0 0.0 12.5 5.4 12.5 8.1 0.0 64.9 0.0 8.1 10.0 8.1 15.0 5,4 15.0 15.0 30.0 0.0 20.0 20.0 15.0 20.0 40.0 20.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 20.0 0.0 66.7 4.2 4.2 8.3 14.3 28.5 0.0 0.0 7.1 0.0 80.0 0.0 0.0 0.0 0.0 88.6 0.0 2.4 0.0 9.1 13.6 0.0 0.0 4.5 54.5 9.1 9.1 40.0 40.0 9.0 0.0 12.5 4.2 7.1 35.7 7.1 0.0 20.0 0.0 2.4 2.4 0.0 Species Beef cattle Dairy cattle Buffaloes Pigs Chickens Ducks Fish Reason for each choice Easy to manage Good source of income Produce manure Use crop residues Draft animal Good market for its products Family consumption Others Animal Management and Productivity The performance of animals raised by farmers is shown in Tables II and 12. For dairy and beef cattle, farmers took good care of them. Vaccination, drenching and deworming were regularly practised. For poultry farmers, no vaccination was given because diseases were generally not a problem. Most of the farmers used artificial insemination in dairy cattle as the main type of breeding. However, some farmers still used bulls because female animals showed poor sign of heat. In the non-dairy group, farmers preferred to use bulls in both beef cattle and buffaloes. .4 .7 .3 .0 .4 .5 .0 Table 11. Productivity, health and reproductive practices of animals among dairy farmers. CattleItem Buffaloes Pigs Chickens DucksBeef Dairy 28.1 61.4 3.2 6.4 88.6 33 3 n.a, n.a. n.a. n.a n.a. n.a. Age at first parturition (mth.) 30 Weaning weight (kg) 110 Weaning age (mth.) 10 Marketing age (mth.) 12 Marketing weight (kg) 70 Type of housing (%) Shed 100.0 Open 0.0 Type of breeding (%) Natural 50.0 AJ 50.0 Reason for not using AI (%) Poor reproduction 0.0 Others 100.0 Vaccination (%) Yes 66.7 No 33.3 Reason for no vaccination (%) Diseases are not problems 0.0 Difficult to get vaccines 0.0 Costly 0.0 Others 100.0 Drenching (%) Yes 100.0 Deworming(%) Yes 100.0 89.5 10.5 100.0 0 ~ 35.0 65.0 92.3 7.7 n.a. n.a. n.a. n.a. 100.0 0.0 13.2 86.8 0.0 100.0 90.9 3.1 0 0 64.7 5.9 100.0 n.a. n.a. 100.0 n.a. = Not applicable; -= No data. The feeding practices of animals in the dry and wet seasons are shown in Tables 13-18. Free grazing, cut and carry/stall fed were the main feeding practices for dairy cattle while tethering and free grazing wer~ the main feeding practices of beef cattle and buffaloes. There were at least three basal feeds used as ruminants' feeds. In the dry season, the main basal feed from farmers' own cropland were rice straw and native grasses as contour hedgerows. Ruzi grass (Brachiaria ruziziensis) from pasture and grazing areas was also used as a ruminant feed during the dry season, but was not sufficient due to limited regrowth. 200 Improving the contribution of livestock to crop-aminal systems in rainfed areas in Southeast Asia Table 12. Productivity, health and reproductive practices of animals among non- dairy farmers. Item Beef Buffaloes Pigs Chickens Ducks 35.2 38 12.0 ll.a n.a. 9.0 10.0 5.7 9.6 126.7 n.a n.a, n.a, n.a n.a n.a, 130.0 12.2 18.5 313.8 106.9 10.6 27.1 306.2 88.9 II.I 78.6 21.4 100.0 0.0 93.8 6.2 100.0 75.0 25.0 16.7; 0.0 0.0 8.3 25.0 50.0 25.0 33.3 8.3 0.0 0.0 33.3 ll.a ll.a ll.a ll.a. ll.a. ll.a. n.a n.a n.a n.a n.a. n.a. ~ 83.3 16.7 93.3 6.7 71.4 28.6 43.2 56.8 38.9 61.1 66.7 33.3 100.0 0.0 100.0 0.0 95.0 5.0 100.0 0.0 46.7 53.3 50.0 50.5 100.0 0.0 n.a. n.a. D.a. D.a. Age at first parturition (mth.) Number of off-springs last parturition Weaning weight (kg) Weaning age (mth.) Marketing age (mth.) Marketing weight (kg) Type of housing (%) Shed Open Type of breeding (%) Natural A/ Reason for not using AI (%) Service not available Prefer own sire Difficult to detect heat Costly Not familiar with it Others Vaccination (%) Yes No Reason for no vaccination (%) Diseases are not problems Others Drenching (%) Yes No Deworming (%) Yes No 46.7 53.3 50.0 50.0 100.0 0.0 n.a. = Not applicable; -= No data Crop-animal systems research in Thailand 201 Table 13. Feeding practices for ruminants during dry and wet seasons among dairy farmers, basal feed 1. Dry season Dairy Bccf Buffa- Cattle Cattle locs Wet season [tem Dairy Cattlc Beef Cattle Buffa- locs Percent of respondents 5.4 32.1 62.5 50.0 50.0 0.0 50.0 50.0 0.0 0.0 19.4 80.1 60.0 40.0 0.0 66.7 33.3 0.0 Main feeding practice Te/hering Free grazing Cu/ and carry/s/all fed Basal feed I Grasses Crop residues Source of feed Own pas/ure/grazing area Con/our hedgerows Own crop land (residue) O/hers crop land (residue) Neighbor's pas/lIre/grazing area Communal pas/lIre Frequency of feeding Twice a day Once a day Every o/her day Adequacy of feed Adequate Inadequate Qualityof feed Good Average Poor 18.4 81.6 100.0 0.0 75.0 25.0 94.7 5.3 100.0 0.0 100.0 0.0 10.9 40.6 45.3 3.1 0.0 0.0 20.0 0.0 ~o.o 0.0 0.0 60.0 50.0 0.0 16.7 16.7 0.0 16.7 97.0 3.0 0.0 0.0 0.0 0.0 66.7 0.0 0.0 0.0 0.0 33.3 33.3 0.0 33.3 16.7 0.0 16.7 50.0 50.0 0.0 73.0 24.3 2.7 25.0 75.0 0.0 50.0 50.0 0.0 73.5 23.5 2.9 25.0 75.0 0.0 77.8 22.2 so.o so.o 66.7 33.3 86.8 13.2 60.0 40.0 66.7 33.3 86.8 13.2 0.0 0.0 66.7 33.3 91.9 8.1 0.0 0.0 100.0 0.0 0.0 100.0 0.0 40.0 20.0 40.0 In the opinion of the farmers, the quality and quantity of feed were good and adequate, respectively. However, the farmers' great concern was the inconsistent feed supply, which was seasonal and sporadic. In the wet season, ruzi grasses were the main basal feed from the farmer's own pasture/grazing area. In addition, native grasses were also used. To a small extent, farmers also used crop residues (rice straw) as a basal feed in the dry season and during the wet season. Tree fodders were also used to a limited extent. Although farmers said that the quality of the basal feed was good, they had to supplement a high level of concentrates, either commercial or farm-mixed. However, farm-mixed concentrates tended to be more favourable than commercial concentrates mainly due to the lower price. Improving the contflDution of livestock to crop-animal systems in rainfed areas in Southeast Asia202 Fanners also used other feeds and by-products such as cassava chips, rice bran, copra meal, dried cassava leaves/hay, dried leuceana leaves and other unspecified items but the amount used was small when compared to concentrates. In the non-dairy group, beef cattle and buffaloes were mairily tethered or free-grazing on their own cropland and communal pasture. Crop residues and grasses were basal feeds during the dry and wet season, respectively. However, fanners did not supplement any concentrate to their ruminants. Feeding practices for non-ruminants in the dairy group and the non-dairy group were similar in both the dry and wet seasons (Table 19). Kitchen refuse from home/ househo\d supply was the main poultry feed. Grains were also used as poultry feed while some fanners used commercial feeds obtained from market/store. Table 14. Feeding practices for ruminants during dry and wet seasons among dairy farmers, basal feed 2. Dry season Dairy Buffa- Cattle loes Wet seasonItem Dairy Cattle Buffa- loes Percent of respondents 65.0 35.0 0.0 0.0 100.0 0.0 40.0 50.0 10.0 0.0 100.0 0.0 50.01 0.0 16.7 20.8 12.5 25.0 0.0 50.0 25.0 0.0 36.4 0.0 27.3 27.3 0.0 0.0 50.0 50.0 0.0 5?.0133.3 11.1 5.6 100.0 0.0 0.0 0.0 40.0 50.0 0.0 10.0 Basal Feed 2 Grasses Crop residues Treeforages Source of feed Own pasture/grazing area Contour hedgerows Own crop land (residue) Others crop land (residue) Communal pasture Frequency of feeding Twice a day Once a day Weekly Occasionally Adequacyoffeed Adequate Inadequate Quality of feed Good Average Poor 50.0 50.0 50.0 50.0 80.0 20.0 100.0 0.0 50.0 38.9 11.1 0.0 100.0 0.0 90.0 10.0 0.0 0.0 100.0 0.0 -= No data Crop-animal systems research in Thailand 203 Table 15. Feeding practices for ruminants during dry and wet seasons among dairy farmers, basal feed 3. Dry season Dairy Cattle Buffaloes Wet season Dairy Cattle Item Per~ent of respondents 0.0 100.0 0.0 100.0 0.0 0.0 ~ 0.0 100.0 100.0 0.0 100.0 100.0 50.0 50.0- 42.1b 57.9d 47.4" 52.6" 6.8 6.8 40.9 6.8 0.0 34.1 4.5 6.4 8.5 40.4 6.4 2.1 29.8 6.4 Basal Feed 3 Grasses Crop residues Tree forages Source of feed Own pasture/grazing area Own crop land (residue) Frequency of feeding Once a day Adequacy of feed Adequate Inadequate Quality of feed Good Supplementation Supplement I Commercial concentrates Farm-mixed concentrate Supplement 2 Rice bran Copra Dried cassava leaves/hay Cassava chips Brewer's grain Dried leucaena leaves Others a Given at 5.4 kg per head per day b Given at 4 kg per head per day c Given at 7.2 kg per head per day d Given at 4.1 kg per head per day -= No data. Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia204 Table 16. Feeding practices for ruminants during dry and wet seasons among non-dairy farmers, basal feed 1. Item Beef Cattle Buffaloes Beef Cattle Buffaloes Percent of respondents 73.3 20.0 7.7 94.4 5.6 0.0 80.0 20.0 0.0 53.3 3.3 43.3 Main feeding practice Tethering Free grazing Cut and carry/stall fed Basal feed 1 Grasses Crop residues 83.3 16.7 66.7 33.3 100.0 0.0 94.4 5.6 Source of basal feed Own pasture/grazing area Contour hedgerows Own crop land (residue) Others crop land (residue) Neighbor's pasture/grazing area Communal pasture 80.0 0.0 13.3 0.0 0.0 6.7 64.3 0.0 21.4 0.0 0.0 14.3 73.9 0.0 4.3 0.0 4.3 17.4 81.3 6.3 6.3 0.0 0.0 6.3 Frequency of feeding basal feed Twice a day Once a day 58.8 41.2 16.7 83.3 50.0 50.0 33.3 66.7 Adequacy of basal feed Adequate Inadequate 88.9 11.1 64.7 35.3 72.2 28.8 94.4 5.6 Quality of basal feed Good Average Poor 50.0 38.9 11.1 88.9 11.1 0.0 88.9 11.1 0.0 37.5 56.3 0.0 Table 17. Feeding practices for ruminants during dry and wet seasons among non-dairy farmers, basal feed 2. Dry season Wet season Item Beef Cattle Buffaloes Beef Cattle Buffaloes Percent of respondents Basal Feed 2 Grasses Crop residues Tree forages 6.7 80.0 13.3 0.0 100.0 0.0 0.0 100.0 0.0 10.0 90.0 0.0 Source of basal feed Own pasture/grazing area Contour hedgerows Own crop land (residue) Others crop land (residue) Communal pasture ~.o 7.7 92.3 0.0 0.0 100.0 12.5 0.0 87.5 ~ 38.5 46.2 7.7 1.7 25.0 75.0 0.0 0.0 30.0 70.0 0.0 0.0 Frequency of feeding basal feed Twice a day Once a day Weekly Occasionally Adequacy of basal feed Adequate Inadequate 7S.0 2S.0 75.0 25.0 90.0 10.0 Quality of basal feed Good Average Poor 50.0 41.7 8.3 41.7 50.0 8.3 80.0 20.0 0.0 sQ.o sQ.o Q.o 0.0 50.0 50.0 Supplementation Supplement I Rice bran Dried leucaena leaves Others = No data. 206 ImprovIng the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia Percent of respondents 0.0 100.0 100.0 0.0 69.2 30.8 0.0 100.0 100.0 0.0 69.2 30.8 0.0 0.0 100.0 0.0 60.0 51.4 25.7 63.6 27.3 0.0 9.1 0.0 0.0 100.0 0.0 20.0 18.0 1.0 8.0 63.6 27.3 0.0 9.1 0.0 100.0 0.0 91.4 2.9 2.9 90.9 9.1 0.0 0.0 100.0 0.0 31.0 90.9 9.1 0.0 0.0 100.0 42.9 28.6 66.7 33.3 0.0 100.0 15.0 9.0 66.7 33.3 100.0 0.0 74.3 20.0 90.9 9.1 100.0 0.0 26.0 7.0 90.9 9.1 100.0 0.0 60.6 39.4 1.03 72.7 27.3 1.0 100.0 0.0 20.0 13.0 1.03 72.7 27.3 1.0 16.7 66.7 16.7 16.7 66.7 16.7 ~ 100.0 100.0 40.0 60.0 40.0 60.0 83.3 16.7 83.3 16.7 83.3 16.7 1.0 83.3 16.7 1.0 M ah mft ~ rRtiE Scavenging Pen fed Feed type of feed Kitchen refuse Grains Crop residues Commercial feeds Source of feed Home/household supply Market place/factory/coop Others Frequencyoffeeding Twice a day Once a day Adequacyoffeed Adequate Inadequate Quality of feed Good Average Quantity of feed (kg/head/day) Feed 2 type of feed Kitchen refuse Grains Commercial feeds Source of feed Home/household supply Frequency of feeding Twice a day Once a day Adequacyoffeed Adequate Inadequate Quality of feed Good Average Quantity of feed (kg/head/day) -= No data Crop-animal systems research in Thailand 207 Table 19. Feeding practices for non-ruminants during the dry and wet seasons among non-dairy farmers. Practice Pigs Ducks Pigs Chickens Ducks Percent of respondents 0.0 100.0 IQo.o 0.0 88.9 11.1 0.0 100.0 97.3 2.7 83.3 16.7 40.0 0.0 40.0 20.0 10.8 81.1 8.1 '0.0 10.5 68.4 5.3 15.8 40.0 0.0 40.0 20.0 10.8 83.8 5.4 0.0 16.7 66.7 5.6 16.7 33.3 50.0 16.7 86.1 13.9 '0.0 66.7 33.3 0.0 40.0 40.0 20.0 81.6 18.4 0.0 66.7 33.3 0.0 100.0 0.0 0.0 83.8 13.5 2.7 73.7 26.3 0.0 100.0 0.0 0.0 81.1 16.2 2.7 88.9 11.1 0.0 100.0 0.0 89.2 10.8 89.5 10.5 100.0 0.0 97.3 2.7 97.3 2.7 100.0 0.0 2.0 94.6 5.4 3.3 97.4 5.3 1.5 100.0 0.0 97.3 2.7 2.0 100.0 0.0 1.5 0.0 0.0 66.7 33.3 0.0 60.0 30.0 10.0 0.0 16.7 66.7 16.7 0.0 33.3 33.3 33.3 9.1 45.5 36.4 9.1 0.0 22.2 44.4 33.3 33.3 36.4 16.7 0.0 36.4 ]2.5 66.7 0.0 63.6 0.0 83.3 0.0 66.7 33.3 63.6 0.0 87.5 0.0 100.0 0.0 72.7 27.3 50.0 50.0 100.0 0.0 81.8 18.2 57.1 42.9 100.0 0.0 ;12.7 ~7.3 50.0 50.0 100.0 0.0 90.9 9.1 62.5 37.5 100.0 0.0 2.0 90.9 9.1 1.0 83.3 16.7 1.5 100.0 0.0 90.9 9.1 1.0 87.5 12.5 Main feeding practice Scavenging Pen fed Type of feed 1 Kitchen refuses Grains Crop residues Commercia/ feeds Source of feed Home/househo/d supp/y Market p/acelfactory/coop Others Frequencyoffeeding Twice a day Once a day Every other day Adequacyoffeed Adequate Inadequate Quality of feed Good Average Quantity of feed 1 (kg/head/day) Feed 2 Kitchen refuse Grains Crop residue Commercia/ feeds Source of feed Home/househo/d supp/y Market p/ace factory/cooperative Mixed onfarm Frequency of feeding Twice a day Once a day Adequacy of feed Adequate Inadequate Quality of feed Good Average Quantity of feed (kg/head/day) -= No data 208 Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia Table 20. Meat production and marketing during the previous year . Dairy group Non-dairy group Item Cattle Pigs Chick- Ducks ens Cattle Pigs Chick- Ducks ens 3.4 15.0 3.1 2.7 1.6 2.0 .4 2.2 1.6 19.5 52.0 2.0 40.0 1.4 30.5 51.0 5.8 35.0 45.0 ~ Meat produced on farm Number of animals slaughtered Average slaughter weight (kg/head) Consumed in the household (kg/head) Sold (kg) Selling price per kg Major buyer (%) Middleman Consumer Other farmers Others 10.5 84.2 0.0 6.7 0.0 100.0 0.0 0.0 0.0 42.9 57.1 14.3 0.0 100.0 0.0 0.0 4.0 0.0 0.0 0.0 10.5 10.0 28.0 36.0 12.0 8.0 12.0 0.0 0.0 0.0 66.7 33.3 33.3 33.3 16.7 0.0 16.7 26.7 20.0 26.7 16.7 10.0 57.9 47.4 26.3 10.5 0.0 10.0 0.0 0.0 50.0 0.0 Frequency of meat purchase (%) Every day 0.0 Two to three times per week 21.7 Once a week 39.1 Every two weeks 4.3 Occasionally 13.0 F or special events 17.4 Amount purchased each time, kg/week 1.3 1.3 .3 .5 .0 1.0 1.5 Price per kg 120 86.2 48 50.0 128.0 82.2 51.3 Source of purchased meat (%) Butcher/market Other farmer Others 100.0 0.0 0.0 91.7 8.3 4.2 75.0 25.0 0.0 0.0 100 0.0 100.0 0.0 0.0 100.0 0.0 0.0 0.0 0.0 100.0 = No data Crop-animal systems research in Thailand 209 Economics of the Animal Enterprises Meat production in the dairy and in the non-dairy groups was mainly from poultry particularly native chicken and mainly for home consumption and for sale (Table 20). Households purchased meat (cattle, pig, chicken) from butchers or the market for their consumption (two to three times a week or once a week being the most frequent). Egg production from chickens was not sufficient for the household consumption but duck eggs produced were sold to the consumers. The values in the dairy group and in the non-dairy groups were similar (Table 21 ). Table 21. Poultry meat and egg production and marketing. Item Dairy group Chickens Ducks Non-dairy group Chickens Ducks 4 26.6 31.6 O 13.3 340 60 0 260 2 100.0 0 O 3.4 40.6 84.28 2.4 61.5 56.7 - 30 2 100.0 30 2.5 ~ - 33.8 2.1 - 14.9 2.2 Number of laying birds Egg produced last month Number consumed last month Number hatched last month Number sold last month Selling price per unit Major buyer is other farmer (%) Number of eggs purchased last week Price paid per unit -= No data Milk production from dairy cattle at peak yield was high (21 kg/hd/d) but was low at late lactation (6.2 kg/hd/d). Most of the fanners sold their milk to the cooperatives at a price of 10.8 Baht/kg. However, fanners bought some milk products such as ice cream and skim milk for their children (Table 22). The fanners expressed interest in more return income than at present. The common expression made by the fanners was" more to go out than to return". This was an indicator for potential improvement on lower inputs. Cost of production especially for the dairy cattle is shown in Table 23. Medicine was the main input accounting for 40% of the total cost of production ( excluding feeds) while vaccination was not included because it was free of charge from the government service. For poultry, the cost of production was similar as in dairy cattle but the fanners had to pay the cost of vaccination by themselves. This similar statistic was also observed in the non-dairy group. ." Labour use for livestock production is shown in Tables 24a-25b. The household' s male and female members shared all activities in the fanns. This means that household males and females in the dairy group worked closely together. In the non-dairy group, labour was concentrated on the crop enterprise and partly on the animal section, particularly pig and poultry .Household males were dominant in all activities while household females participated only in some activities. 210 Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia Table 22. Milk production and marketing of dairy cattle, dairy group, Item Response 4.29 279.6 65.0 21.6 6.2 361.8 1.3 7.0 4.3 49.9 10.8 7.9 92.1 100.0 100.0 50.0 50.0 0.2 Number of milking animals (head) A verage lactation length ( days) Length of peak milk yield period (days) Daily production per animal, peak period (kg) Daily production per animal, lean period (kg) Total milk produced last week (kg) Amount consumed (kg) Amount processed (kg) Amount given as gift (kg) Amount sold (kg) A verage price per kg sold (Baht) Buyer of fresh milk (%) Processing plant Cooperatives If processed, what are the products (%) Others (ice cream) To whom the processed product is sold (%) Consumers Milk and milk products purchases Frequency of purchase of milk products (%) More than once a week Once a week Quantity of milk products purchased recently (kg) From where milk products was purchased Store/market 3.0 Table 23. Cost of production per head excluding feeds (Baht). Cost Item Cattle Ducks Cattle Ducks 2,365.3 775.0 562.10 442.3 607.0 905.0 63.3 85.3 1,623.1 1,300.0 444.0 97.5 90.0 80.0 252.5 156.7 2, 750.0 85.0 85.0 167.0 2,870.0 523.8 96.0 32.5 100.0 ~ 400.0 300.0 Medicines Veterinary services Vaccination Drenching Deworrning AI Breeding fee Salt Ropes Other input 1 Other input 2 Crop-animal systems research in Thailand 211 Table 24a. labour use for ruminant production among dairy farmers. Item Wet season Dry season Percent of respondents 6.3 87.7 0.0 6.3 0.0 14.8 0.0 7.4 0.0 50.0 39.5 10.5 45.8 44.0 10.2 97.0 3.0 0.0 0.0 70.0 0.0 20.0 10.0 49.2 40.9 9.9 51.1 35.3 4.0 42.9 14.3 14.3 28.6 80.0 0.0 20.0 0.0 58.3 41.7 0.0 55.6 33.3 11.1 93.7 0.0 6.3 93.9 3.0 3.0 43.7 40.9 15.5 40.9 43.7 15.5 69.0 10.3 20.7 60.6 12.1 27.3 Frequency of grazing of animals Twice a day Once a day Weekly Monthly Occasionally Who does it? Household male Household female Household children Frequency of cutting/carrying of grass and other green roughage for feeds Twice a day Once a day Weekly Occasionally Who does it? Household male Household female Household children Frequency of tethering Twice a day Once a day Weekly Occasionally Who does it? Household male Household female Household children Frequency of giving supplementation Twice a day Once a day Weekly Who does it? Household male Household female Household children Frequency of collection ofanimal manure Twice a day Weekly Occasionally Who does it? Household male Household female Household children 47.6 38.1 14.3 48.6 38.9 12.5 Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia212 Table 24b. Labour use for animal production among dairy farmers Item Rumi- nants Poultry Rumi- nants Poultry Percent of respondents 20.7 30.0 10.4 137.9 100.0 0.0 0.0 0.0 91.7 0.0 0.0 8.3 23.3 0.0 26.7 50.0 100.0 0.0 0.0 0.0 91.7 0.0 0.0 8.3 51.7 41.4 100.0 0.0 0.0 39.0 41.5 19.5 50.8 41.3 7.9 100.0 0.0 0.0 39.0 41.5 19.5 100.0 0.0 0.0 0.0 100.0 0.0 0.0 0.0 100.0 100.0 0.0 0.0 0.0 100.0 0.0 0.0 0.0 100.0 46.2 41.0 12.8 ]00.0 0.0 0.0 66.7 33.3 0.0 46.8 40.3 13.0 100.0 0.0 0.0 66.7 33.3 0.0 100.0 46.3 41.5 12.2 47.5 41.3 11.3 100.0 46.5 9.3 2.3 41.9 45.5 13.6 ]00.0 100.0 60.0 40.0 60.0 40.0 0.0 100.0 14.3 85.7 0.0 100.0 14.3 85.7 Gathering, preparation of feeding materials other than green grass and roughage Twice a day Once a day Week/y Occasiona//y Who does it? Househo/d ma/e Househo/d fema/e Househo/d chi/dren Frequency of cleaning pens/barns Twice a day Once a day Occasiona//y Who does it? Househo/d ma/e Househo/d fema/e Househo/d chi/dren Frequencyofmilking Twice a day Who does it? Househo/d ma/e Househo/d fema/e Househo/d chi/dren Frequency of delivery ofmilk to collection centers Twice a day Who does it? Househo/d ma/e Househo/d fema/e Househo/d chi/dren Hired ma/e Frequency of transporting animals for marketing Occasiona//y Who does it? Househo/d ma/e Househo/dfema/e Frequencyofvaccinationl medication Month/y Occasiona//y Who does it? Househo/d ma/e Househo/dfema/e Househo/d chi/dren 100.0 0.0 0.0 69.2 23.1 7.7 100.0 0.0 0.0 69.2 23.1 7.7 -= No data Crop-animal systems research in Thailand 213 Table 25a. labour use for animal production among non-dairy farmers Wet season Dry seasonItem Rumi- nants Pigs Poultry Rumi- nants Poultry Percent of Respondent 0.0 100.0 100.0 0.0 100.0 0.0 0.0 100.0 100.0 0.0 100.0 0.0 66.7 29.6 3.7 50.0 50.0 0.0 62.5 31.3 6.3 68.0 28.0 3.6 50.0 50.0 0.0 62.5 31.3 6.3 80.0 10.0 10.0 0.0 0.0 0.0 0.0 100.0 n.a. n.a. n.a. n.a. n.a n.a n.a, n.a ~ 63.0 37.0 66.7 33.3 n.a. n.a. n.a. n.a. 100.0 100.0n.a. ll.a. 64.1 33.3 2.6 63.2 34.2 2.6 ll.a. ll.a. ll.a. n.a. n.a. n.a. 25.0 50.0 25.0 75.0 25.0 Frequency of grazing of animals/feeding Twice a day Once a day Who does it? Household male Household female Household children Frequency of cutting/carrying of grass and other green roughage for feeds Twice a day Once a day Weekly Occasionally Who does it? Household male Household female Frequency of tethering Once a day Who does it? Household male Household female Household children Frequency of gathering/ preparation of feeding materials other than grass and green roughage Twice a day Once a day Weekly Who does it? Household male Household female 66.7 33.3 66.7 33.3- -= No data; n.a. = Not applicable Improving the contribution of livestock to crop-animal systems in ralnfed areas in Southeast Asia214 Table 25b. labour use for animal production among non-dairy farmers. Wet season (%) Item Ruminants Pigs Pigs Dry season (%) 50.0 0.0 50.0 50.0 0.0 50.0 ~ 75.0 25.0 100.0 0.0 0.0 5.6 94.4 73.3 23.3 3.3 100.0 100.0 100.0 100.0 100.0 0.0 0.0 84.2 15.8 0.0 84.2 15.8 0.0 100.0100.0 100.0 0.0 0.0 75.0 18.8 6.2 Frequency of giving supplementation Twice a day Once a day Weekly Who does it? Household male Household female Frequency of collection of animal manure Twice a day Weekly Occasionally Who does it? Household male Household female Household children Frequency of cleaning pens/barns Occasionally Who does it? Household male Household female Household children Frequency of vaccination/ medication Occasionally Who does it? Household male Household female Household children -= No data 215Crop-anima/ systems research in Thai/and Fish Production Activities Farmers in the dairy group had at least two ponds. However some farmers owned three to four ponds (Table 26). Mixed species of fish were raised with fingerlings put in during May to June and harvested when needed for consumption and some were also sold. Labour used in most activities was shared by the household male and female but fish breeding was done only by the household male Table 27). Characteristics of fish production activities of the non-dairy group were the same as in the dairy group, but household males were in charge of more responsible activities than household females. Table 26. Fish production and marketing. Item Dairy group Non-dairy group 1-4 787.2 1-4 ,058.9 17.6 8.1 30.9 22.8 20.6 4.28 May-July Variable Variable 15.1 21.7 33.0 17.9 12.3 1.21 May-July Variable Variable Number of ponds Size (m2 ) Type or breed (% of respondents) Tilapia Carp Milkfish Black carp/amur Other Fingerlings put in (kg) Date put in (mth) Date to harvest (mth) Expected output (kg/m2) Landholding Most farmers in the dairy group owned three parcels ofland, located at an average of 1.4 km from their houses (Table 28). The area of each parcel was 1 ha. Ninety five percent of land was owned by the farmers. Only a few farmers rented land from others. Sixty six percent of the land was upland while 29% was unflooded lowland, and the rest (5%) was flooded lowland. Nearly 60% of the land was used for annual crops and grassland. Approximately 70% of the land were classified as good by soil fertility indicators, while 20% was average and lO% poor. As for soil erosion of land, about 50% were lightlmoderate to non-detectable but farmers observed very little severe soil erosion. In the non-dairy group farmers owned about three parcels and the area of each parcel was slightly lower than the parcels owned by farmers in the dairy group. Most of their land (89%) was used for annual crops (67%) and perennial crops (22%) with some were used for forestltrees or orchards. Topography, soil fertility indicators and soil erosion of land owned by the non-dairy group were similar to the land owned by the dairy group. Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia216 Table 27. labour inputs in fish production and marketing. Item Dairy group Non-dairy group Percent of respondents 100.0 100.0 100.0 60.0 40.0 90.0 10.0 32.0 20.0 4.0 36.0 8.0 21.4 21.4 0.0 17.1 50.0 57.5 27.5 15.0 68.4 31.6 0.0 20.0 20.0 20.0 40.0 0.0 0.0 15.4 84.6 72.7 18.2 10.1 86.7 13.3 0.0 20.0 13.3 6.7 60.0 20.0 20.0 20.0 40.0 95.5 4.5 90.9 9.1 100.0 100.0 Frequency of breeding Seasonal Who does it? Household male Transferring frequency Seasonal Who does it? Household male Household female Feeding frequency Daily Weekly Monthly Occasionally Seasonal Who does it? Household male Household female Household children Frequency of maintaining ponds Daily Monthly Occasionally Seasonal Who does it? Household male Household female Household children Harvesting frequency Daily Weekly Monthly Occasionally Who does it? Household male Household female Selling/marketing frequency Occasionally Who does it? Household male Household female 66.7 33.3 100.0 0.0 -= No data Crop-animal systems research in Thailand 217 Table 28. landholding and use. Item Dairy group Non-dairy group 1-8 1.5 1-5 1.2 36.7 9.3 31.6 4.3 16.5 1.4 67.3 4.4 3.5 2.7 22.1 84.7 2.3 1.5 1.5 9.9 98.2 0.9 0.9 0.0 0.0 36.0 18.4 0.94 - 3.24 66.4 28.9 4.7 58.4 38.0 3.6 57.8 25.0 17.0 56.6 33.7 9.7 Number of parcels Land area (ha) Land use (%) Annual crops Forest/trees Grassland Orchard Perennial crops Fallow Ownership status (%) Owned Rented in Rented out Leased out Others If rented, tenurial arrangement Share cropping (%) CLT (owned) ro/a) Distance from homestead ( kIn) Topography (%) Upland Lowland not flooded Lowland flooded Soil fertility indicators (%) Good Average Poor Observe soil erosion (%) Non-detectable Slight/moderate Severe 45.6 51.8 2.6 45.1 54.9 0.0 = No data Crop-animal Interactions Crop-animal interactions are shown in Tables 29 and 30. Farmers used manure mainly as fertiliser and partly as feed, for other purposes and for sale. While residues from crops such as rice and cassava were used as animal feeds, a small amount was left in the field. Residues/by-products from cassava were used as animal feeds to a large extent. 218 Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia Table 29. Use of manure. Dairy group Non-dairy group Item Pigs Chick- Cattle Buffa- ens loes Cattle Buffa- loes Pigs Chick- Ducks ens 97.4 2.6 7.9 45.0 45.0 5.0 50.0 32.5 87.5 3.6 44.4 53.8 44.5 41.1 100 51.4 49.6 70 30 90 10 Manure collected,(%) % used as fertiliser % used as feed % used for other purposes % sold 5.1 3.8 14.5.8 = No data Table 30. Use of crop residues. Non-dairy groupDairy group Utilisation Maize CassavaRice Cassava Rice 10060.5 88.6 71.8 39.5 1.1 4.9 3.3 1.5 100 % used as feed % used as mulch/compost/left in field % burned % used for other purposes % sold 11.4 -= No data IMPLICATIONS OF THE RESULTS Identification of Recommendation Domains The chosen site of this study was shown to have a similar topography and geology as described in terms of agro-ecological characteristics when compared to other areas in the northeast of Thailand. This site was located in the rainfed area and over 60% of the land was in the upland with the majority of the remainder being lowland, not flooded. The major crop-animal systems in the study were dairy cattle based-rice/cassava system with most of the income coming from dairy cattle. In addition, farmers in the dairy group and non-dairy groups were comparable in terms of household characteristics, labour and economics. Thus, this site where the research results were obtained, is typical and the research findings can be extrapolated. Major Constraints to the Recommendation Domains Identified Based on the analysed data from the questionnaires and from observations during farmers' interviews, the major constraints identified were as follows: I) Nutritional inadequacy both in terms of quantity and quality is the main constraint to dairy production. Although farmers said that they have enough basal feeds of good quality for dairy cattle production, they still supplemented these feeds with a high level of concentrates. In addition, farmers did not feed animals appropriately in terms of meeting animal requirements for production. Protein sources are the first limiting feed source on many farms. Farmers have to buy them at a high price from market/store/cooperatives. Thus, to improve their productivity, they need to increase the quantity and quality of the basal feeds through existing technology such as urea treatment, cassava hay and the use of leguminous feeds as well as improved pasture, etc. 2) Since farmers largely depend on commercial feeds that are expensive (5.8-6.2 Baht/kg), it is important for farmers to produce home-mixed feeds and be trained to prepare farm-made concentrates. 3) There is inadequate training to improve farmers' skills to support livestock development. Farmers have too few outlets and channels where they can sell their livestock products such as milk and beef. Often, this has resulted in overstorage and deterioration of such products. 4) Organisational management is not sound among extension workers, technical personnel and farmers to effectively collaborate on livestock programs with a view to increase production efficiency. Low reproductive performance has also been reported. 220 Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia EXPERIENCES AND LESSONS LEARNT Problems The problems encountered by the dairy farmers are quite numerous. As family income was a primary justification for the developmental and the occupational engagement of these smallholder farmers, low cost inputs are important in increasing efficiency. Farmers lack appropriate and suitable knowledge and experiences in small dairy farming management. Means of exchanging information and learning from each other are seen essential. The biological factors involve improving techniques of making urea-treated straw, the establishment of on-farm dry season high protein roughage especially cassava hay and cowpea intercropping. On-farm concentrate mixture has been a problem and should be developed and be fully exploited. Improvements Improved land use can increase the quality and quantity ofbasal feeds. Integrated, animal- based farming systems are a way to solve some of the problems. The use of animal manure as fertiliser would improve crop production, reduce use of inorganic fertiliser, and improve soil fertility and structure, etc. The use of crop-residues or by- products on the farm would improve both the quantity and quality of animal feeds. With regard to economics, if farmers can reduce their cost of inputs and increase output, they will increase their farm income and improve their standard of living. Protein sources may be increased by growing more leguminous crops like cowpea or leucaena and/or cassava (for hay) (Wanapat et al., 1997; 1999; 2000). ACKNOWLEDGEMENTS The authors and the research team members sincerely express their gratitude to all the farmers and everyone who have closely collaborated with our work and to the International Livestock Research Institute (ILRI) for the kind financial support. 221Crop-animal systems research in Thailand REFERENCES Agricultural statistics of Thailand. 1997. Ministry of Agriculture and Cooperatives, Bangkok, Thailand. Anonymous. 1985. Rapid Rural Appraisal. Rural Systems Research and Farming Systems Research Project, Khon Kaen University , Khon Kaen, Thailand. Anonymous. 1997. The situation of the economic crops in Northeast Thailand. Annual Report for 1997, The Northeast Agricultural Extension Office, Taphra, Khon Kaen, Thailand. Anonymous. 1999. Technologies recommendation for Hom Mali rice production in Mahasarakham Province. Improvement efficiency of Hom Mali Rice Project. Faculty of Agriculture, Khon Kaen University, Khon Kaen , Thailand. Chantalakhana C. 1994. Report on Status and Development of Dairy Production and Milk Products in Thailand: Research Approach and Development in the Future. Thailand Research Fund (TRF), Bangkok, 88p. Wanapat M. 1999a. Feeding of ruminants in the tropics based on local feed resources. Khon Kaen Pub Comp Ltd., Khon Kaen, Thailand, 236 pp. Wanapat M. 1999b. Improvement of crop-animal systems research in Thailand. Proc. Planning Workshop on the Crop-Animal Systems Project, (ed). C. Devendra. International Livestock Research Institute. pp81-95. Wanapat M., Pimpa 0;, Petlum A. and U.Boontao U. 1997. Cassava hay: A new strategic feeding for ruminants during the dry season. Proc. International Workshop on Local Feed Resources Based Animal Production organised by Ministry of Agriculture, Forestry and Fisheries. Kampuchea and F A0/Japan Regional Project, University of Agriculture, Ho Chi Minh, Vietnam. Wanapat M., Petlum A. and Pimpa 0. 1999. Strategic supplementation with a high- quality feed block on roughage intake, milk yield and composition and economic return in lactating dairy cows. Asian-Australasian J. Anim. Sci., 12:901-903. Wanapat M.. Puramongkon T. and Siphuak W. 2000. Feeding of cassava hay for lactating dairy cows. Asian-Australasian J. Anim. Sci., 13:478-482. 222 Improving the contribution of livestock to crop-aminal systems in rainfed areas in Southeast Asia HOUSEHOLD SURVEY RESUL TS AND IMPLICA TIONS FOR FURTHER RESEARCH A T CILA WU , INDONESIA A. Djajanegara, A. Priyanti, B. Risdiono and D. Lubis Agency for Agricultural Research and Development Ministry of Agriculture, Indonesia ABSTRACT The first step in the Crop-Animal Systems Research project activity was the selection of a benchmark site (BMS) that serves as a location of a large geographic entity for later transfer of the fin,dings. The village of Cilawu was selected as the benchmark site in Indonesia, and a household survey was conducted to characterise the benchmark site. The reasons for the conduct of the BMS characterisation include (I) to obtain a baseline information on the prevailing crop-animal systems; (2) to identify and understand the factors and constraints that influence the production systems; and (3) to identify and understand the linkages between crop and animal production. At a later stage, the potential suitability of alternative technologies and areas for further research will be identified and assessed. During the household survey at Cilawu, in- depth interviews using questionnaires were conducted among 71 farmers representing 25% of the farmers population that grow crops and raise animals. Data collected includes secondary data from available sources as well as primary data. Quantitative data collected during the household survey are presented, including proposed alternative technologies for consideration in future work. INTRODUCTION The important contribution of the rainfed areas for agriculture production in Inclonesia has long been recognised, as the majority of farmers grow food crops in these areas. In the lowland irrigated areas, rice is the major crop grown. In contrast, in the upland areas where the supply of water depends very much upon rainfall, it is not surprising that the type of agricultural crops produced heavily relies upon those crops that can thrive under relatively limited water supply, in particular, during the dry season. Most farmers in Inclonesia raise animals as an integral part of the agricultural production system. Hence, the general approach is to develop agricultural production systems for food crops and animals in the rainfed areas by promoting integrated crop- animal systems. The main focus of the Crop-animal Systems Research is to increase crop and animal production in integrated systems based on locally available resources for improving the welfare of the farmers. Household survey resuhs and implications for further research at Cilawu, Indonesia 223 The Benchmark Site The Crop-animal Systems Research area is located at Kecamatan (sub-district) Cilawu, District ofGarut, West Java, Indonesia, which is about eight kilometres east of the city of Garut. The District of Garut is located about 180 kilometres southeast of Bogor or 60 kilometres from Bandung, the capital city of West Java. The sub-district ofCilawu is bordering on the west with Sub-districts Tarogong, Margawati, Bayongbong, north with the City of Garut, on the east with the District of Tasikrnalaya and on the south with the sub-district of Banjarwangi (Box I). The sub-district of Cilawu covers around 75 square kilometres, consisting of 18 villages, 44 Dusun (Rural District), 160 Rukun Warga (District Society) and 557 Rukun Teta:ngga (Neighboring Society). Dangiang is one of the 18 villages in Ci1awu selected as location for the conduct of the study. Agro-ecology and farming systems The altitude ofCilawu reaches 950 meters above sea level (750-1235 m) and is typical of upland rainfed areas in Java. The temperature ranges between 17- 30?C and humidity between 76- 90% on the slope of Mount Cikurai. The average rainfall intensity reaches 2,423 millimetres and during the rainy season it could reach over 1 ,000 millimetres per month, generally from October to April, at which time agricultural production faces problems with sometimes too much water supply and low light intensity (Table 1, Figure 1 ). Soil drainage is categorised as medium to good, consisting of volcanic material with soil types as follows: Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia224 ~~..!!"c3...u.c...1/1'6.c~1/1...mEE..?.c..GI......mc.'i6...c.ia~GI:cm1- ;> ; "fi~ :;~ >-~~ =~ ~~< 00=< u0g-oo ~z uuO 3~ ~ 00 ~ r--r-- r-IO r-IO\o-OMV) -r-I- -00 M\O M 1011'\ ~~ OIOOOOIOOS:;OO~~ NN M . -.t 1"'1 0000-00 r--,,",I"'II"'I-.t1""1 O-.t,,",-.tO\00-,,",O\r--N ~I"'INNN,,", N ~ -1"'1 -.t -.t\O-.t-.t- r--~ r--Or--O\OOOO""'r9IOO-.t1"'1 """'"'--1"'1 -~ N- r-- ""'0\ -.t ""'N I"'IOOO?-.t::0-1"'I\0 -.t- : O '"'" \0 -r-.. r'\.N N r-.. ~ O\IrIOON,","NIrIOO\O,","N\ONO\O,","NO ~ ~ ~ ~ ~ ~ ~ ~ ~ ~O\ ,","\Or'\r'\NNN-r'\N -\0 \Or-..OOOONNr'\-1rI1rINO-IrIIrIr'\Nr'\,","OO'"'" ~Nr'\-- -r'\ ~ -N 00N- t"Ir-. "1"-.0- 10r--Nr--~- "1~ r---r'\ 0t')N~t') ~u~Q.Q.~'-'=tU.a-2uQ.=tU'3-~uQ.~~u~0~ o o o o = ~ ~ .,-li~- 00"' . 'C.(1) + a)0)0).-dla)0).-~~.!!C3..IQi6-c.ia..>:E..coE~C)IQ..~>ct.-~..~C)i! O 01 .., -~ OOO~01'"'1V\~OV\-N -'"'I -~ - Farmers grow upland rice in the rainy season and at the end of the rainy season. At the start of the dry season, secondary crops like vegetables, sweet potato, maize, potato, soybean, ginger, cabbage, red chili and onion, are genera11y grown after the rice harvest. Farmers in Cilawu who raise cattle or sheep have grown Napier grass on the benches of their land, and the grass is also found in spots along the roadside and in the backyard. Farmers, however, prefer to obtain roughage from the surroundings and often from distant places particularly during the dry season. They have to travel over 2-3 kilometres to get the roughage. Table 2 shows the major agricultural commodities grown in the area. Table 2. Major agricultural commodities grown. Commodity Area planted (Ha) Production(Kg/ha ) 3,653 1,804 1,504 358 519 41 72 12 11 57,589 5,600 3,000 6,100 1,100 12,200 12,800 10,000 10,000 20,000 7.4 kg/plant Food crops Irrigated rice Rainfed rice Maize Soybean Sweet potato Potato Chili Red onion Cabbage Banana (plants) Estate crops Tea Ginger Coconut 472.9 97 21 4,300 0.82/nut = No data Land utilisation gives priority mainly for crop production, the main crop being upland rice and vegetables. Growing rice is apparently the first option because farmers rely on rice as staple food for home consumption. Products of other crops are sold and mixed cropping systems are common in upland areas although the size of land holding is relatively small {less than 0.5 ha per household). Other areas within and outside Java are mostly similar. except for the culture. The type of agricultural commodities being planted may vary depending upon the demand and eating habits but most crops grown in the upland areas are similar. A large proportion of the agricultural land areas in Indonesia are rainfed, and the proportion of dryland areas covers 7% of the total arable land {130 million ha). The general cropping pattern is shown in Table 3. Table 3. General cropping patterns in Cilawu. 443 574 249 117 110 Cropping systems in rice fields Rice + rice + rice Rice + rice + second crop Rice + rice + vegetables Rice + second crop + vegetables Rice + vegetables + vegetables 598 150 200 178 112 57 154 81 258 7~ Cropping systems in dryland area Upland rice + corn + cassava Upland rice + corn/soybean + cassava Upland rice + upland rice/cassava/corn Corn/cassava + soybean Corn/cassava + sweet potato Corn/soybean/cassava +corn/sweet potato Upland rice + vegetables Corn/cassava + vegetables + vegetables Vegetables + vegetables Upland rice/corn + chili 67 87 38 102 198 248 Second crop + potato Ginger/chili + corn Ginger/corn + corn Sweet potato/ cassava + sweet potato/ corn Tea/upland rice/corn Tea/vegetable The topography of the research area is hilly with slopes ranging from 0 -40 %. The main soil type in the Sub-district of Cilawu is Oxisol being generally acidic to Household survey resuhs and implications for further research at Cilawu, Indonesia 227 neutral. It has high organic matter, medium to high Phosphorous (P), Calcium (Ca) and Magnesium (Mg) contents. It is low in Nitrogen (N) and Potassium (K). The general arable land can be classified into paddy-rice field, upland, estate crops, mixed garden, and forest. About 152 hectares of upland areas at Cilawu are planted with mixed crops, and 35 hectares are planted with estate crops such as vetiver grass and tea. The cropping pattern in Cilawu is determined by rainfall pattern and farmers' view of the best market. A variety of animals are raised by farmers. Raising of beef cattle is under a development program implemented by the Garut Livestock Services in which farmers were given a loan to raise one or two young steers that should be paid when the animals are sold. The animals were bought from dairy production centres in other villages as far as Bandung/Lembang. The male calves are generally sold at two to three months of age, as dairy farmers generally do not want to raise male animals. Sufficiencyofforage supply to meet the animal feed requirement becomes more important as the farmers raising the Frisian Holstein male calves for beef production did not make efforts to produce forages for their animals. The socio-economic condition in the area is a typical chronic poverty situation. However, there are no exact figures that describe the poor condition in Cilawu. In view of the housing conditions, most farmers at Cilawu could be considered as poor. Most farmers produce food crops and some of them raise cattle or sheep. It is a general situation where animals are sold when the farmer needs urgent cash, and where the services of banking systems are generally not available. METHODOLOGY Site coordinator One important aspect in an on-farm research operation is the presence of a site coordinator who is prepared to stay in the village and has a social attitude to assist the rural poor farmers. His/her education background should at least cover the basic understanding of livestock farming. The main task is to gather all the necessary information needed, but most importantly, to understand the social atmosphere that will determine the success of the operation at a later stage. The site was selected from 18 villages in Garut based on many characteristics that may influence the farming systems. The village of Dangiang was chosen as the most promising village where farmers that grow crops also raise animals (cattle or sheep). At the start, only one village was selected, while other villages will be included depending upon responses received. A dairy development scheme is underway under the program of the Garut Livestock Services in the villages of Sukatani and Sukamumi. There are 18 villages in Cilawu and a site coordinator was stationed on site to monitor activities in the village that may have an impact at a later stage. Thus, a daily report that recorded all the information and activities over the period was made. 228 Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia Sampling procedure The first step undertaken in the study was the collection of secondary data from sources around the village, i.e., Local Livestock Services, Local Food Crop Services, Balai Penyuluhan Pertanian (Agricultural Extension Services Institute), head of the village, and farmer groups. These data served as base information on the general conditions in the village. Prior to the conduct of the household survey, a general meeting was conducted with all farmers growing crops and raising livestock where the general direction of the survey was explained. This was carried out to minimise misunderstanding by farmers who may not be selected in the interview. The important reason is to clearly explain ~hat will be done and minimise expectations that will not be provided. Most farmers are expecting something out of a survey and would feel neglected if they were not included in the survey. In the conduct of a survey that involves outside people and government officials, it is generally considered that a new project will be introduced, thus, an opportunity to obtain assistance. In general, farmers are willing to support improvement schemes that will benefit them. In 1999, the population ofCilawu was 81,281 people (40,369 male and 40,912 women). The number of households in the village of Dangiang was 811,641 ofwhich or 79% were farmers (Box 2). Farmers who raise livestock account for 287 households (35.4 % of all households and 44.8% of farm households). The total land area of the village of Dangiang was 603.1 hectares and 548.1 hectares were used for agricultural purposes, of which only 26 hectares could be considered as rainfed rice fields. Farmers who are raising animals in Dangiang are organised and grouped into three farmer groups. These are Kelompok Tani Mekarsaluyu, Kelompok Tani Mekarrasa and Kelompok Tani Mekarsari. One farmer group (Kelompok Tani Mekarrasa) is not active anymore. About 25% of the farmers in the group that grows crops and raises livestock were selected at random for the household survey. Seventy-one farmers selected from farm households in Dangiang were individually interviewed in their homes. RESULTS General Conditions The general information available showed that there are 717 households in Dangiang. Land ownership data of 610 farmers revealed that 223 farmers grew rice. Of those that work in the agricultural sector, 471eased the land and 279 were farm labourers. There were 24 farmers who raise cattle, 390 raise sheep, while there was none raising goats. The number of farmers who raise chickens was 426. There were 37 households with fishponds, three households operating a home industry and 96 households in the industrial workforce outside the village. In the village, 92 people are working for the Government or the army and 60 are retired government officials. Of the workforce in Household survey results and implications for further research at Cilawu, Indonesia 229 the village, there were 32 carpenters, three bricklayers and two tailors. Seventeen people work as farm labour. The types of landholding are shown in Table 4. Table 4. Size of landholding. Size of landholding (Ha) Number of farmers <0.1 0.1 -0.5 0.6 -1.0 1.1-1.5 1.6- 2.0 3.0- 5.0 >5 Total 249 137 159 36 14 3 598 -= No data There is no banking facility available in Dangiang and there are only 26 small stalls (warung) and three small shops. Transportation within the village is provided by 11 motor bikes and there are only six cargo vehicles available. Because animals are the most ready asset for sale, farmers will offer the animals in their possession for sale. An alternative cash source might be needed to assist farmers in gaining ready cash. Improving the contribution of livestock to crop-anlinal systems in rainfed areas in Southeast Asia230 Household Characteristics Seventy one (71) respondents from eight Village Districts in Dangiang were interviewed (Box 3). Of the respondents, 69 were male (heads of the family) and only two were female. The average age of the respondents was 47.7 years. The youngest respondent was 24 years old and the oldest was 75 years old. All respondents were from West Java and belonged to the Sundanese ethnic group. Formal education of the farmers was generally at elementary school level, and six farmers have never went to school. Of those that have been to school, 24 did not finish sixth grade and 33 passed grade 6 at the elementary school. Only three farmers passed Junior High School, while three farmers finished Senior High School and another three obtained a Diploma. Around 75% of the farmers had no additional training in food crop -or animal management aspects. Only 14 farmers obtained additional training on food cropping and 17 on animal production aspects. The average number of family members in the household was 4-5 people. Most of the farmers were around 16-60 years of age. On the average, each household has one family member who is over 60 years old. Land utilisation and crop production Almost all of the land owned by fanners has a soil fertility between medium to good and according to the fanners in Dangiang, only a small portion of the land is poor. Most of the land area are planted with annual crops, mainly food crops such as rice, Household survey resuhs and implications for further research at Cilawu, Indonesia 231 corn, cassava, peanut and vegetables. There are plenty of banana trees growing in the survey area. Land is mostly owned and only some were rented. Few young fanners cultivate their parents' land. They are not paid as generally, they live together with their parents. The fannyards are mostly close to the homestead, but some fanners having more than three parcels ofland, also own fannyards in remote areas of about five or more kilometres away. There were four main food crops planted by farmers in Dangiang, namely rice (both upland and lowland), corn, cassava and peanuts. The crops are usually planted in the rainy season. Cassava is generally planted at the early rainy season, but since this crop takes nine to ten months to harvest, itoccupies land until the end of the dry season. Hence, most of the land areas are planted to cassava in the dry season. Rice Production The yield of rice crops varies among rice fields in Dangiang and many factors influence the variable rice yield, i.e., fertiliser and animal manure application, use of pesticides, seed quality and management. The average rice yield in the rainy season was 3.65 tlha, while in the second season it only reached 3.13 tlha. Inorganic fertilisers used by farmers are mainly urea and TSP applied in moderate amounts (between 150 to 300 kg/ha). In the rainy season, the average amount of inorganic fertiliser applied to the crop was 224 kg/ha while in the second season, it was 271 kg. Most farmers also used animal manure from beef cattle or sheep being raised by some farmers in the area as basal fertilizer. The average amount of manure applied to the cropland was about 1.8 tlha/year. According to the farmers, they do not have an exact guideline on determining the amount of manure needed. Insecticides were applied by less than 50% of the rice farmers in Dangiang, the main constraint being the increasing price since 1997. As an example, the price of pesticides in 1999 rose five- to tenfold compared to the price before the economic crisis. On the average, pesticide application was 1.75 kg/ha in the rainy season and 0.78 kg in the second season. None of the farmers used herbicides to control weeds in their farm and they preferred the traditional way of manual control. Corn production At the first planting season, corn yield was about 0.75 tlha while in the second season it could reach as much as 1.76 tlha. However, there was no explanation for this. Animal manure applied in corn in the first season was higher than that for rice. More than 2.3 tlha of animal manure was used in corn compared to only 1.9 tlha in rice. The amount of inorganic fertiliser applied in corn was slightly less (200 kg/ha) compared to 224 kg in rice. Only few farmers use pesticides to protect the corn plant. The average application for corn was 0.68 kg/ha in the rainy season and 0.28 kg/ha in the second season. The amount of seeds used varied between 16.7 kg to 30.31 kg/ha in the rainy season and 15.4 to 47.6 kg in. the second season. Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia232 Cassava production Cassava is an important food crop for upland farmers as almost all small-scale upland farms (even in whole West Java) are planted with it. Farmers usually plant cassava in the early rainy season and harvest 9-11 months later. Small farmers plant cassava for home consumption, while the richer farmers cultivate a special varietY of cassava for industrial purposes. The average cassava yield in Dangiang was between 5-16 t/ha in the first season and between 6-13.7 t/ha in the second season. Very little inorganic fertiliser was applied to cassava, only 98-132 kg kg/ha in the first season and 25-60 kg/ha in the second season. Animal manure used for cassava was between 4.4-6.6 t/ha in the rainy season and about 4.3 t/ha in the second season. No insecticides and herbicides were applied to cassava. The planting distance varied between parcels, depending upon the cropping pattern practised. The plant densitY ranged between 1,800 to 5,000 plants/ha. Peanut production Not many farmers in Dangiang plant peanut. The average area planted was 0.21 ha per household. The average peanut yield was about I t/ha in the rainy season and only about 150 kg/ha in the second season. Farmers applied as much as 100 kg/ha inorganic fertiliser in the rainy season and 190 kg/ha in the second season. Animal manure was applied to the crop at about 8 t/ha. Only a small amount of pesticides was used. Farmers used 7 -28 kg of seeds to plant one hectare of peanuts. Animal Production and Productivity Cattle During the previous year, the number of adult male cattle that have been raised by the farmers was 21 head, or an average of 1.75 head per farmer household. The total number of cattle shared in was II head or an average of 1.57 per farmer household that practised a sharing arrangement. The number of cattle being sold live reached 31 head or 1.94 head per farmer household. None of the animals are sold when they are already dead, as this is not allowed by custom. The sick animals that could not recover are generally sold before they die at a reasonably cheap price. The average number of animals raised by farmers was 1.6 head per farmer household. Raising of young male cattle at the start accounted for 21 head or 1.5 head per household. The number of young male cattle bought reached 19 head. Farmers are willing to raise young cattle under a sharing arrangement and 12 animals are under this arrangement or an average of 1.4 animals per farmer household. The total number of young male cattle was 35 head or an average of 1.94 animals per farmer household. The average market age of cattle was 17.1 months and these were sold at an average weight of 236 kg per animal. Raising of cattle is generally in a shed. Only six Household survey resuhs and implications for further research at Cilawu, Indonesia 233 fanners raise cattle in the open. There were only nine respondents that could provide infonnation on the type of animal raised, but it was clear that most animals raised in the village are Frisian Holstein cattle. Most animals are bred naturally. Only one fanner knew about artificial insemination. Most respondents do not vaccinate their animals for the following reasons: disease is not considered a problem, there is a lack of equipment/facilities, and vaccination is costly. These reasons might be valid because the fanners do not see a problem as long as the animals are alive and feeding could be properly done. Vaccination might not be a problem yet in the area where raising of cattle has just started and the movement of cattle in and out of the village is still relatively small. According to the Local Livestock Services officer, vaccination against anthrax is the only vaccination program being enforced. Only a few respondents drench their cattle and again the reason for not drenching the animals is that there was no problem posed by external parasites. In contrast, dewonning is more familiar to the respondents. Almost half of them treat their animals against internal parasites. The dewonner used is generally a traditional medicine that could be purchased in the shops. However, the problem with wonns could not be clearly seen. The symptoms are generally slow growth and poor appetite in affected animals. Sheep In the previous year, the number oframs raised by the respondents was 95 head or an average of3.3 head per farmer household. A relatively large number of rams were purchased over last year (132 head) where an average of 12 head per household were purchased. The number of male rams sold alive was 187 head or 8.5 animals per household on average, whereas the number of rams present at the time of the survey was 41 head or on average 2.4 head per household. The total number of ewes being raised by the respondents at the start was 120 head or an average of 3.01 head per household. In the previous year, 19 ewes were purchased (average of2.4 per household), whereas the number of ewes sold alive reached 57 head or an average of2.8 head per household. The number of ewes raised at present is 91 head, or an average of 3 head per household. Young sheep that were raised at the start were 67 head (average of 3 head per household) and 30 young sheep were bought (average of2.5 head per farmer that bought sheep). From the total number of sheep raised by farmers, 39 head of young sheep were sold (average of3.2 head per household) and the present number of young sheep reached 52 head (average of2.7 head per household). The dynamics of the animal population could not be clearly defined, as movement of animals was not recorded. The first parturition in sheep occurred at the age of 14.9 months, while young sheep are reported to be weaned at 4.6 months reaching an average weight of 8.01 kg. Sheep are usually marketed at the age of 13.4 months and average weight was reported at 29.7 kg. Poultry In the previous year, 353 head of chickens (average 7.8 chickens per farmer household) were raised by farmers: The number of chicks hatched was 575 (average of 26.1 head per farmer household) and 10 chickens were bought last year. Chickens were raised mainly for small cash income and home consumption, with the number of chickens raised as gift-in reached 13. A relatively high number of young chickens that died in the previous year were reported and the mortality reached 505 head of chickens or each farmer lost 18 head of chickens last year . The number of chickens sold alive was 98; each farmer sold an average of 5.2 head of chickens in a year. About 121 chickens were slaughtered last year with an average of 6.1 head. The number of chickens gift-in was 15 head and the total number of chickens raised by the respondents reached 354 head or 8.4 head on the average per household a year. Ducks are also raised in the village with 33 ducks recorded early last year. Around 24 have hatched, but the mortality rate was high (16 ducklings) leaving at present only 35 ducks some of which have been sold. The price of ducks is not known as raising them is considered more as a hobby. Meat and Egg Production, Marketing and Consumption In general, marketing of livestock products is in general satisfactory and market outlets and competition between traders exist. Depending upon market volume, new market sites could be established. In Indonesia, the beef market is dominated by a wet market selling second and third grade products. Only a small quantity of prime grade beef sold especially to big hotels and restaurants is available. Most of the beef cattle are sold at the local market or sent and slaughtered in Jakarta for both wet and institutional markets. For small ruminants, the main market outlets are the village collectors and the local market. In isolated areas, farmers generally have access to few village collectors. In Java, although farmers have easy access to daily or weekly markets, farmers more commonly trade through the local village collector. Marketing of large ruminants and small ruminants follows almost similar channels, however, small ruminants are easier to transport and have a relatively higher turnover rate compared to that oflarge ruminants. Market information is obtained by the farmer through the middlemen. This can be improved by developing a reliable livestock market information system. A routine monitoring would yield market price data and production projections that should cover the periodic verification of livestock numbers and production. The market of livestock in Indonesia is becoming increasingly integrated and this is visible in price trends. Correlation between Jakarta prices and those in the other cities increases as distance from Jakarta decreases, illustrating the flow of animals from the east to Jakarta. Comparison between locations in terms of marketing margins relative to Jakarta prices showed a similar relationship. In a well-integrated market, marketing margins increased Household survey results and implications for further research at Cilawu, Indonesia 235 as distance from the primary market increased. The Denpasar marketing margin as a Meat produced on farm None of the ruminants (cattle, buffaloes, goats and sheep) were slaughtered on farms to produce meat for consumption. Thus, the question on number of animals slaughtered was interpreted as number of animals sold. The average value of parameters representing meat produced on farm during the previous year is shown in Table 5. Table 5. Average value of parameters representing ruminant meat produced on farm. Cattle Buffaloes GoatsParameter Sheep 1.3 318 9,910 2.5 25.75 11,750 1 n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. Number of animals sold (head) Selling weight (kg/head) Selling price (Rp/kg live weight) Major buyer D.a. = Not applicable Meat purchase The daily meat consumption of the farmers in the vitlage was very low, as they only consume meat during special parties or an Islamic festivity in the year. The fact is that meat consumption will always relate to per capita income. This is supported by findings of a previous study where, in a breakdown by income group, urban and rural, data showed that Indonesians spend a greater share of the total income in animal products as income increases. In other words, the level of income determines expenditure of high quality goods that will increase at a faster rate. The average value of parameters representing meat purchased by household farmers during the previous year is shown in Table 6. Table 6. Average value of parameters concerning purchase of ruminant meat. Cattle Buffaloes Goats SheepParameter Frequency of purchase (times/year) Quantity purchased each time (kg) Buying price (Rp/kg) Source of purchased meat 5.7 5.57 22,445 1 2 I 18,000 I D.a. D.a. D.a. D.a. D.a. D.a. D.a. D.a. n.a. = Not applicable 236 Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia percentage of Jakarta prices was higher (43 %) than Surabaya (36%), which is c]oser to Jakarta. Chicken and egg production Kampong (village) chickens and a few ducks are raised traditionally but they are not considered as a main source of family earnings. The products of these animals are an important component in the Indonesian family consumption pattern as source of meat and eggs. Most kampong chickens and ducks are marketed through a long marketing chain. Some village collectors carrying bamboo cages for about 20-30 mature kampong chicken or ducks travel around the village, buying in cash one or more live chicken or duck. The village collectors generally travel in the afternoon to collect chickens and ducks. Early next morning, they sell the animals to big collectors from the cities. There are usually two or more big collectors, in charge of more than ten village collectors, who do trading in each collection area -every morning or twice a week in most areas. A traditional farming system with small number of chickens and ducks (on average 5-10 birds) is a common practice in most family farms in the village. The birds are allowed to scavenge in the backyard or in the garden, and are provided with limited facilities such as a simple cage and small amount of food like scraps or sometimes rice bran/corn. Between 5 to 12 eggs are left brooded by hens in each batch where chicks are raised over the following six months. With this production system, farmers get a small benefit from selling the birds when in need of cash or when consumed. The additional revenue from keeping kampong chickens or ducks is therefore unpredicted as the farm activity is considered as family savings. The average number of eggs produced on farm over the previous year is shown in Table 7. Table 7. Average values of parameters concerning egg production ChickensParameter 4.58. 13.71 7.57 5.00 n.a. 475 1 1.8 500 Number of laying birds (head) Eggs produced last month Number consumed Number hatched Number sold Selling price (Rp/piece) Major buyer Number of eggs purchased last week Price paid (Rp/piece ) n.a. = Not applicable Household survey resuhs and implications for further research at Cilawu, Indonesia 237 Cost of production inputs excluding feed costs As is generally the case, the main cost of production inputs in raising animals is feed, aside from the cost of the animal itself. For the present purpose, the cost of production inputs does not consider the feed component, since the activity to obtain feed was done by farmers during and after working in the field. The average of material inputs representing cost of production excluding feed, during the previous year is shown in Table 8. Table 8. Average cost in Rupiah of material inputs representing cost of production per head (excluding feeds). Cattle Goats Sheep ChickensMaterial inputs 10,200 n.a. n.a. 4,750 11,750 5,000 n.a. 4,000 19,000 6,000 2,000 3,000 2,000 1,900 3,450 7,250 o.a. o.a. o.a. o.a. 2,000 6,785 n.a. 2,000 n.a. n.a. n.a. n.a. n.a. n.a. n.a. Medicine Veterinary services Vaccine Drenching Deworrning Artificial Insemination Breeding fee Salt Ropes Cleaning materials Others D.a. = Not applicable; -= No data. Animal management and productivity Cattle are used for draught purposes since the village topography is very undulated with land slope varying from 5 to 40%. Large and small ruminants are kept in covered stalls in close proximity to the farmers' houses, so that the stalls can easily be cleaned and the manure can be saved or directly used for soil fertilisation. None of the farmers sell animal manure; neither do they use it for other purposes. In general, most farmers do not breed cattle. Instead, they buy young steers for fattening. The most preferred breed is the Frisian Holstein, followed by Ongole. The fattening period varies from 8 to 17 months, probably related to the age of the steer bought and how good the animals are fed as the general target is to reach at least 250 kg liveweight. Currently, due to limited facilities, animal weight is computed by using a girth length (in centimetres) and converting the measurement to kilograms, a practice guided by experience. In fact, it was found that the average marketing weight of cattle was 289.9 kg. Improving the contn1Jution of livestock to crop-animal systems in rainfed areas in Southeast Asia238 For the few farmers that breed cattle, natural mating is always applied. The main reason for not using artificial insemination was due to the almost no accessibility to livestock service from the local government in the area. However, breeding practices done by the local farmers are quite good, which in this case is reflected by the relatively good breeding parameter values. Average age at first parturition is 24 months, and the calves are weaned at four months old, at a 115-kg mean liveweight. All farmers stated that their cattle have never been vaccinated, while anthelminthic drugs ( for deworming) were given several months prior to selling of the steers. The animals were sometimes drenched with traditional herbs or commercially sold mineral mix for health and higher weight gain reasons. Farmers consider small ruminants as a living savings account. They can sell the animals when they need cash or they can sell the meat, usually to cater for special occasions or ceremonial parties. Therefore, marketing of sheep depends upon smallholders' needs rather than factors related to maximum yields. In addition, good performing rams with a nice horn shape are used for ram fighting, a traditionally common game in Garut district. In this case, selling price of the ram is much higher than the common price of sheep. The price could increase further when the ram wins the game, which could reach up to ten times the regular price, i.e., reaching Rp 6,000,000 (approximately US $ 857). Sheep are mated naturally to rams with good performance and the reproductive performance and management of sheep are fairly good (Table 9). When small ruminants are sold, the preferred marketing weight is 29.7 kg for sheep and 20 kg for goats, with an average selling age of 13.4 and 12 months, respectively. Table 9. Reproductive performance of sheep. SheepParameter Age at lst parturition (mth.) Number of offspring Weaning age (mth.) Weaning weight (kg) 15.0 1.9 4.6 8.0 Sharing systems in keeping ruminants are not common in the area. Farmers practise more share-in of cattle rather than share-out, which suggests that some cattle that are kept under sharing arrangements belong to wealthier farmers in neighbouring hamlets or villages. Like most villagers in Java, farmers feed their ruminants almost exclusively with forages. The common feeding system in the area is the cut and carry system, with some concentrates fed (approximately 2 kg/d) for fattening animals, in particular during the first and fourth quarter of the fattening period. Roughages of annual food crops (especially maize) are also fed to ruminants just after harvesting. There are no farmers Household survey resuhs and implications for further research at Cilawu, Indonesia 239 who store excess crop residues for feeding. Farmers prefer to travel long distances to obtain native grasses during times when forage feed nearby the households is scarce. Chickens (native species) are left to scavenge and they are raised as a source ofa living savings account of farmers. They are sold when small amounts of cash are needed. They are also produced for home consumption of meat and eggs. Duck raising is less common, and the ducks are kept primarily for egg production. Crop-animal Interactions Animal manure is exclusively used as soil fertiliser mainly for annual crops that are grown not too far from homestead/stalls. The most common food crop is maize, followed by sweet potato, peanut and beans. At harvest time, corn stovers, sweet potato vines and other crop wastes are fed to ruminants. During the hot, dry season (June to September), banana leaves and stubbles are valuable to meet the amount of forage and water needed. Only small amounts of rice straw are used for feed and bedding, as the rice fields are mostly located out of the study village. Table 10 shows the types of crop residues and extent of their use by farmers. Table 11. Utilisation of crop-residues of annual crops Percent Used As Annual Crop Feed Mulch Burnt 23 100 47 65 79 73 39Rice Maize Cassava Sweet Potato Peanuts Beans 38 35 21 27 = No data. IMPLICATIONS OF THE RESULTS Constraints and Recommended Domains In a crop-animal system, the main emphasis is to link crop and animal production in a system. It is apparent that crops provide residues for feeding the animals and in return the animals produce manure for fertiliser. The system is in fact already implemented by farmers in Cilawu, however, optimal use of available resources could Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia240 still be considered as not being fully exercised. A large amount of the crop residues are still wasted or burned, while animal manure (without any processing measures) is already returned to the soil to maintain the productivity of the particular crop grown by the farmers. Table 11 gives an indication of the nature and extent of the constraints faced by farmers. The major problem faced by farmers is meetirig the animal feed requirements in sufficient amounts, hence, animals are at present growing at a low productivity level. The supply of feed from present land resources does not guarantee the daily need. In terms of total land area and production of agricultural residues in relation to the number of animals raised, it would seem that the supply of feed would be sufficient. However , since land ownership is individually oriented, it is not possible to collect all the agricultural residues available, without any special effort. Farmers could obtain the material for free from other farmers when they are involved in harvesting the crop produced. The amount of crop residues that could be obtained is therefore limited by the capacity to carry the forage by the individual farmer and there is no special service to collect all the material. It is often a situation that the materials are produced far away from where the animals are. Constraints faced by farmersTable 11 Constraint Commodity Socio-economic Biological Crops Rice Capital, price of input (fertilisers) Pest, diseases, local seeds, soil fertility Price of output (harvest time) Pest, diseases, local seeds, soil fertility Small size potatoes Soil fertility Price of output; market guaranteeCorn Sweet potato Animals Cattle Capital, sharing arrangements Farmers' group Feed supply uncertainty Feed quality, diarrhea Dry season (4-5 mths.) Time taken to get feeds Feed supply, worms, Small size, growth rate CapitalSheep General Small size of fanns, price of produce Access to technology Household survey results and implications for further research at Cilawu, Indonesia 241 Processing of manure The importance of using animal manure as fertiliser is apparent in maintaining soil fertility .Over the years, the use of chemical fertilisers has been emphasised and this has created a problem with reduced soil fertility in terms of C:N ratio. A challenge is to process manure into a good fertiliser compound using rumen and soil microbes as reported (Haryanto, 1999). The processed manure has good qualities in that it induces higher crop production and after a processing period of 28 days, the manure is then ready to be used as pathogenic agents do not survive the 60- 70 ?C heat generated. The change in production goals is now being seen as an alternative for farmers to produce fermented organic manure instead of live animals, the production ofwhich is now considered as a by-product. This concept is seen as a change in vision of the animal industry from cattle production to manure production and this is expected to provide better income to the farmers. Tree legumes Introduction of tree legumes has been considered as a way to meet the animal feed requirement. The problem is that these have to be planted and the question that arose is the availability ofland resources. Many research activities have looked at various leguminous tree crops like Leucaena leucocephala. Gliricidia sepium, and Calliandra callothyrsus. While there are some limitations in the use of these tree legume crops, these limitations could be overcome quite easily. Three-strata forage production systems For a sustainable feed production system based on locally available resources that would apparently assist farmers for the production of animals, the production of forages would have to meet the demand during the rainy and dry seasons. It would probably be a challenge to introduce the Three-Strata Forage Production System developed in Bali, as the size of land holdings might not guarantee sufficient land resources needed. It might be necessary to find an alternative land area that could be used by farmers collectively to plant forage. However, this may face foreseen problems with management and care of the facility .Alternative systems relying on available resources are apparently difficult to find with the limited resources offarmers. In addition, the priority of farmers who own the land is to grow food crops for home consumption, hence, the production of forage by farmers for feeding the animals does not come into the picture. In the rainy season, forages that grow close to the vicinity of the animal are abundant and a large supply of forages is available. This has led to the practice of farmers getting forage from distant places when forage is not available nearby. Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia242 Pre-treatment of crop residues One approach to guarantee feed supply is probably to store forages when abundant and in addition provide supplements to the animal to meet the daily nutrient requirements. Forage supply for the dry season is probably a major solution expected, however, in Dangiang, it is during the rainy season that forage is scarce. All arable land areas are utili sed for food crop production while storage facilities are not available in the small land holding and limited amount of forage can be collected in one day by human labour. The maximum load that one man can carry is limited to one or two loads on the shoulder. Women and children can only carry half the amount, which they carry home when returning from work in the field. -Buying offorages to feed animals is not feasible with the low-income situation of farmers. Techniques to treat fibrous crop residues have been developed in the early 1980s involving application of physical, chemical and microbial pretreatments. Since farmers do not store crop residues and prefer to collect grass daily, the idea of pretreatment of fibrous crop residues is not encouraging. The poor farmer, with only less than a quarter of a hectare of land, cannot be burdened with the extra cost of pretreatment facilities. LESSONS LEARNT AND EXPERIENCES Over the years, many alternative approaches have been conducted with farmers to increase animal production and productivity. The results have shown that production and productivity could be increased substantially, in conditions where feeding is not a problem. The introduction of new breeds carries with it the need for better management. Under the prevailing conditions of the farmers, it has not succeeded when farmers have to run the operation themselves without any additional support. It is quite frustrating to consider that farmers do not see the animal industry as a business, but more as a living savings account. This leads to farmers seeming to be interested more in having the animals alive rather than improving growth rates. In facing financial problems, farmers would first sell their animals, meaning that they do not target animal growth for a certain time frame, except in closing the lebaran qurban, when intact male animals have a big market. The introduction of new technologies often lacks continuity as these have notbeen obtained from the farmers' point of view but more often from the researchers , point. It would be more of a challenge to start at the farmer level but the time frame setting of the research operation to achieve results may not be simple. Development achievements will take a slow pace as it depends upon acceptance and availability of resources. Household survey results and implications for further research at Cilawu, Indonesia 243 ACKNOWLEDGEMENTS The present work has the support of the International Livestock Research Institute- Nairobi, Kenya. Many have contributed to the work and appreciation is hereby extended to Dr. Kusuma Diwyanto and Dr. Tjeppy D. Soedjana for the support given. Thanks are also extended to all the survey team members including those from CRIAS, Faculty of Animal Science -IPB. Further, the hard work ofMr. Sonny Ahmad at the village of Dangiang has been very excellent in coping with the social conditions. The support of the Livestock Services and Food Crop Services ofGarut, the Balai Penyuluhan Pertanian and Farmers groups at Cilawu, especially the cooperation given during the surveys is gratefully appreciated. Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia244 BENCHMARK SITE CHARACTERISATION IN DONG TAM, DONG PHU, BINH PHUOC PROVINCE, VIETNAM K. M. luc, l. P. Dung, P. V. Bien Institute of Agricultural Science of South Vietnam, Ho Chi Minh City , Vietnam l. V.ly National Institute of Animal Husbandry, Hanoi, Vietnam ABSTRACT The household survey involved a total of 105 households in two hamlets out of 581 households in nine hamlets in Dong Tam village. Crops grown are rice, cashew, cassava and maize. All farms raised beef cattle, buffaloes, pigs, chickens and ducks, of which pigs were the most important. About 16-31% of the households stopped raising beef cattle, buffaloes and pigs due to reproduction problems. Data are given in the management functions for each species, and also feeding systems and labour use for managing animals. About 61-84% of the farmers collected manure, but only 50-51% of them used it as manure and the remaining was sold for cash. The major constraints were inadequate feeds and poor nutrition, inadequate number of beef cattle, and rampant cassava production. Proposed interventions include assessment of feed availability and use year round, intercropping tree legumes with cassava, utilisation of feeds from cassava and resultant soil fertility and economic problems. INTRODUCTION Improving productivity of crop-animal systems in the rainfed area in Vietnam is an important issue in the livelihood of the farmers in this area, recognised as the poorest part of the country .In Vietnam, 80% of the population is involved in agricultural production with a total sown area of 8.5 million ha for food crops (7.4 million ha paddy rice and I.I million ha other food crops). The irrigation capacity presently serves 3.7 million ha, 44% of which is cultivated area (General Statistics Office, 1999). These figures reveal that a large part of agricultural production in Vietnam is still under rainfed condition. The living condition of the farmers in the rainfed areas is relatively poorer than those in irrigated areas because of the low productivity of their crops and animals. Improvement of productivity of the crop-animal systems in Vietnam was defined as a general objective of the CASREN project, which was started in 1999 with five participating countries (Li-Pun, 1999a, 199b). In Vietnam, Dong Tam village was seleCted as the BMS site of the project. A general information of the site selection in Vietnam was reported in the first workshop held in Los Bafios, the Philippines in June 1999 (Luc, 1999). This paper summarises the information obtained from a household survey conducted in December 1999 on benchmark site characterisation. Benchmark site characterisation in Dong T am, Dong Phu, Binh Phuoc Province, Vietnam 245 MA TERIALS AND METHODS The procedures for the BMS characterisation are surnrnarised in Figure 1. Before implementation of the household survey, several visits to the village were made by the research team of the Institute of Agricultural Science (IAS), where rapid appraisal and focus group discussions were conducted. A pre-survey was also conducted on entire households of the village (3,081 households) using a short questionnaire with categories. Data of the pre-survey was analysed to understand the general picture of the village and to select Hamlet 4 as monitoring and Hamlet 2 as control sites. A total of105 households (50 from Hamlet 2 and 55 from Hamlet 4) were randomly sampled from 581 households of the two hamlets selected from the 9 hamlets in Dong Tam village (Figure 2). The data was analysed using the SAS program (SAS, 1999). J ..Site Visit ~r ~ ~ ~ ~ Report Hamlet 2: Control I. Hamlet 4: Monitoring -. ~ 1 .Site Visit Report Figure 1. BMS characterisation procedures. . Figure 2. Map of Dong Tam village. RESUL TS AND DISCUSSION General Characteristics of Households General infonnation on tl1e surveyed households in tl1e control hamlet (Hamlet 2) and monitoring hamlet (Hamlet 4) is presented in Table I. Most oftl1e respondents in both hamlets were tl1e husbands and majority oftl1em were 20 to 60 years old. Most of tl1em had an educational level from tl1e first to ninth grades, and had no training on crop and animal production. Among tl1e ethnic groups including Kinh (Vietnamese) Tay Nung, Khome and Stieng in tl1e two studied hamlets, tl1e Stieng people are tl1e most "backward" ones. The production practice oftl1e Tay Nung people is almost similar to Vietnamese. During tl1e survey, some assigned respondents in Hamlet 2 were not at home. Interviewers tl1en had to change respondents and interviewed Stieng people. This resulted in a higher percentage of Stieng households in Hamlet 2. Table 1. Information on the surveyed households. Percent Item Hamlet 2 Hamlet 4 58 38 4 80 18 2 92 8 44 96 4 15 7 20 22 33 18 30 24 20 20 6 84 16 O 69 29 2 Respondent Husband Wife Other Age of respondent 20-60 years old Over 60 years old Stieng (Ethnic group) Education None Grades 1-3 (Elementary) Grades 4-6 (Elementary) Grades 7-9 (Secondary) Grades 10-12 (Secondary) Training on crops None Short courses Vocational school Training on livestock None Short courses 87 13 90 10 3 51 25 21 6 47 25 22 Age distribution of respondents Over 60 yrs J 6-60 yrs 6-JSyrs Less than 6 yrs Crop Production Table 2 summarises the information on crop production. Cashew and rice production is the most popular in both hamlets, followed by cassava and maize. There were 90% households in Hamlet 2 and 89% households in Hamlet 4 growing cashew, and 82% households in Hamlet 2 and 42% households in Hamlet 4 growing rice. However, most rice production in Hamlet 2 is paddy rice, while upland rice is more popular in Hamlet 4. The productivity of rice, maize, cassava and cashew in both hamlets is relatively low, because of intercropping and low investment in inorganic fertiliser as well animal manure. This indicated that crop production in Dong Tam is still extensive, except for some high benefit perennial trees such as coffee, orchard and rubber. In cashew production, a large amount of herbicides was used for the purpose of preventing grasses from catching fire in the dry season. Animal Inventory The changes in number of animals within a year are presented in Tables 3 and 4. Animal population in both hamlets is relatively low. Hamlet 2 had 0.94 ruminant units, 0.6 pigs and 16.6 heads of poultry per household while Hamlet 4 has 1.1,0.95 and 23.6. No shared in, shared out, gift in, gift out, and barter in/barter out of animals were observed in the two hamlets, except for a small amount of chickens and ducks. Table 3. Animal numbers in Hamlet 2. Animal Starting Number* Born Bought Died Sold Alive Slaugh- Current tered Number 28 I 10 10 7 Beef cattle Adult males Adultfemales Young Calves I 12 9 o 0 3 O 2 O O o 0 2 O o o o o5 19Buffaloes Adult males Adultfemales Young Calves I 10 6 0 1 1 01; 0! 1 1 O O O 1 o o o o5 9 5 41 30 Adult males Adult females Young O 2 25 I 4 31 O 1 12 o 0 27 o 0 1 ~ 5 16 7 6 4 9 0 0 8 Chickens 2009 1197 29 941 311 855 804 Ducks 64 0 54 9 0 79 26 *The starting number was the number of animals at the beginning in December 1998. -= No data. 250 Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia Table 4. Animal numbers in Hamlet 4. Animal Starting Number* Born Bought Died Sold Alive Slaugh- tered Current Number Beef cattle Adult males Adult females Young Calves Buffaloes Adult males Adult females Young Calves 28 2 22 16 2 27 28 2 O 1 4 O 3 4 O O 3 II o o 5 0 9 2 6 2 O o o o o o o o o o 0 0 02 1 1 O 5 2 58Pigs Adult males Adult females Young Piglets Chickens Ducks O 7 92 8 1,448 226 0 O 3 145 17 793 541 o 0 6 O 1,741 213 O 116 4 158 184 3 26 28 1,225 77 35 3,115 534 31 2 928 102 *The starting number was the number of animals at the beginning in December 1998. -= No data The reproduction rate of cattle in Hamlet 4 was relatively low, indicated by five calves produced from 27 cows within a year. Almost all farmers did not slaughter cattle and buffaloes on farm for home consumption or for sale. There were a few young pigs slaughtered on farm. Chicken mortality was rather high in the two hamlets with Hamlet 2 having a higher figure than Hamlet 4. The reason for the high mortality of chickens was non-vaccination of the animals. Shifting Animal Situation Fifty six percent (56%) of the respondents in Hamlet 2 and 47% in Hamlet 4 started raising new animal species in the last five years (Table 5). These figures are relatively high because most farmers in these hamlets are new immigrants from different parts of the country .They possess different agricultural production backgrounds. Some animal species have not yet been raised before, e.g. farmers from paddy rice areas have raised buffaloes rather than cattle. Moreover, the lack of capital at the beginning years of the settlement prevented them from raising animals for quite a time. Benchmark site characterisation in Oong Tam, Oong Phu, Binh Phuoc Province, Vietnam 251 Table 5. Percent of respondents that started raising new animal species. Item Hamlet 2 Hamlet 4 56 44 47 53 30 22 16 6 24 7 20 7 22 4 O 2 11 O 4 O Started raising new species Yes No Species raised Beef cattle Buffaloes Pigs Chickens Reason for stopping Beef cattle Government program Provide draft power Valued as assets Favourable market conditions Buffaloes Provide draft power 22 7 Pigs 6 O O 6 4 2 Good as assets Can use crop residues Easy to sell when cash is needed Chickens Improve fami Iy diet 2 2 Beef cattle was the most preferred animal species for those who started raising them. Among the studied respondents, there were 30% of respondents in Hamlet 2 and 24% of respondents in Hamlet 4 that started raising beef cattle. These figures are in accordance with the practice in Dong Tam where many settler-farmers from other areas where beef cattle production is not common started raising beef cattle after moving to Dong Tam because of the availability of its feed resource. Being a government project as the reason stated for starting the raising of beef cattle elicited the highest response, 22% of the respondents in Hamlet 2 and 11% of the respondents in Hamlet 4. A high number of farmers stopped raising cattle, buffaloes and pigs for the last five years (Table 6). The percentage of respondents that stopped raising pigs was the highest (20% in Hamlet 2 and 31% in Hamlet 4). However, the relevant reason for this was not given. Apart from a common reproduction problem that made farmers stop raising beef cattle and buffaloes, some farmers sold the animals for cash for other inputs or food emergency. At this time, it was quite difficult for the farmers to recover the capital to buy another animal. A few respondents stopped raising chickens, though mortality rate was the highest in this study. 252 Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia Table 6. Percent of respondents that stopped raising animals. Item Hamlet 2 Hamlet 4 42 58 56 44 16 20 Stopped raising animals Yes No Reason for stopping Beef cattle Reproduction problem Buffaloes Reproduction problem Pigs 14 II No reason 20 31 Tables 7 and 8 show the preferences of the fanners and the reasons for their choice in raising new animal species. Beef cattle was the first choice (46% in Hamlet 2 and 52% in Hamlet 4) followed by pigs and chickens. Thirty one percent (31%) of the respondents in Hamlet 2 and 37% of the respondents in Hamlet 4 preferred to raise cattle because it is "easy to manage." On the other hand, chickens were chosen for raising mainly for home consumption (43% of the respondents in Hamlet 2 and 29% of the respondents in Hamlet 4) . Table 7. Respondents' preference for raising animal species. Species Hamlet 2 Hamlet 4 First choice Second choice Third choice First choice Second choice Third choice Percent of Respondents Beef cattle Buffaloes Pigs Chickens 46 27 11 15 10 7 62 19 20 10 15 55 52 7 32 6 13 2 35 46 8 16 50 = No data Table 8. Reasons for choice of animal species to raise. Reason Hamlet 2 Hamlet 4 Percent of Respondents Beef cattle Easy to manage Draft animal Good source ofincome 31 29 27 37 11 22 Pigs 35 34 13 18 43 21 14 Good source ofincome Easy to manage Use crop residues Chickens Family consumption Easy to manage Good source ofincome 29 29 17 Animal Production and Management Practices The productivity of cattle, buffaloes and pigs are presented in Table 9. Age at first parturition in cattle were 29 months in Hamlet 2 and 27 months in Hamlet 4. The calves in Hamlet 4 obtained a weaning weight of 103 kg at 8 months of age, and were sold to the market at 16 months of age and weighing 160 kg. Information for Hamlet 2 was not available. Table 9. Productivity of cattle. buffaloes and pigs. Cattle Buffaloes Pigs Item Hamlet 2 Hamlet 4 Hamlet 2 Hamlet 4 Hamlet 2 Hamlet 4 29 27 39 54 9.7 9.7 12.3 1.7 5.5 73.4 ~ 103 8 16 160 105 10 15 150 100 12 66 367 9 68 Age at first parturition (months) No. of offspring last parturition Weaning weight (kg/head) Weaning age (mth.) Market age (mth) Marketing weight (kg) -= No data. 254 Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia Tables 10-13 show the management practices for cattle, buffaloes and pigs in both hamlets of Dong Tam village. Most of the cattle, buffaloes and pigs were housed with pens, mated naturally and vaccinated. However no drenching and deworming for ruminants and pigs were observed in Hamlets 2 and 4. Vaccination is not commonly practised in chickens and ducks (90-100%) in both Hamlets 2 and 4. The most popular reasons cited for not drenching, deworming in ruminants and pigs, and vaccinating poultry in these hamlets were that the farmers are not familiar with these activities or the services are not available in these hamlets. Table 10. Management practices in beef cattle. Item Hamlet 2 Hamlet 4 Percent of Respondents 94 6 93 7 92 8 75 25 100 100 95 5 81 19 37 63 27 73 36 36 33 25 17 83 14 86 Type of housing Shed Open Breeding system Natura/ AI Reason for not applying AI Service not avai/ab/e Vaccination Yes No Drenching animal Yes No Reason for not drenching Parasite not a prob/em Notfami/iar with it Deworming animal Yes No Reason for not deworming Notfami/iar with it 42 44 Benchmark site characterisation in Dong T am, Dong Phu, Binh Phuoc Province, Vietnam 255 Table 11. Management practices in buffaloes. Item Hamlet 2 Hamlet 4 Percent of Respondents 83 17 20 loo 100 100 100 100 100 100 23 77 20 80 44 33 - 25 0 100 20 80 Type of housing Shed Open Breeding system Natura/ Reason for not applying AI Service not avai/ab/e Vaccination Yes Drenching animal Yes No Reason for not drenching Parasite not a prob/em Notfami/iar with it Deworrning animal Yes No Reason for not deworrning Worms are not a prob/em Notfami/iar with it 27 45 25 25 -= No data 256 Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia Table 12. Management practices in pigs. Item Hamlet 2 Hamlet 4 Percent of Respondents 100 100 100 0 71 29 91 9 91 Q 100 0 0 100 II 89 37 56 36 36 24 76 73 27 Type of housing Shed Breeding system Natural A1 Vaccination Yes No Reason for no vaccination Service not available Drenching animal Yes No Reason for not drenching Parasite is not a problem Notfamiliar with it Deworming animal Yes No Reason for not deworming Notfamiliar with it 92 17 Table 13. Vaccination practices in poultry. Item Hamlet 2 Hamlet 4 Percent of Respondents 5 95 9 91 47 30 10 0 100 0 100 Vaccination in chickens Yes No Reason for no vaccination Notfamiliar with it Se",ice not available Vaccination in ducks Yes No Reason for no vaccination Not familiar with it Se",ice not available 83 - 17 -= No data Benchmark site characterisation in Oong T am, Oong Phu, Binh Phuoc Province, Vietnam 257 Feed Resources and Feeding Systems Tables 14-18 show the feed and feeding systems for beef cattle, buffaloes, pigs and poultry, which were observed to be similar during the wet and dry seasons. Tethering is a main feeding system for cattle and buffaloes in Dong Tam (94% and 83% of the respondents applied it in cattle, and 100% and 60% of the respondents practised it in buffaloes for Hamlets 2 and 4, respectively). Animals grazed grasses twice a day while tethered in the household or in communal pastures. However, pen-fed is the most popular feeding system for pigs, where 14 to 30% of the respondents use grain and crop residues to feed the animals. Scavenging is the popular feeding system in poultry as reported by 86% and 89% of the respondents in Hamlets 2 and 4, respectively. Table 14. Feed resources and feeding systems in cattle. Hamlet 2 Hamlet 4Item Percent of Respondents Main feeding system Tethering Free grazing Type of feed Grasses Crop residues Source of feed Own pasture Contour hedgerows Communal pasture Feeding frequency Twice a day Once a day Feed adequacy Adequate Inadequate Feed quality Good Average Supplement* Rice bran 94 6 83 17 100 0 94 6 3956 17 44 76 24 67 33 78 22 100 0 42 58 4} 59 4250 *Given at 1.2 kg/head/day in Hamlet 2 and 1.4 kg/head/day in Hamlet 4. Table 15. Feed resources and feeding systems in buffaloes. Item Hamlet 2 Hamlet 4 Percent of Respondents Main feeding system Tethering Free grazing Type of feed Grasses Source of feed Own pasture Communa/ pasture Feeding frequency Twice a day Once a day Feed adequacy Adequate Inadequate Feed quality Good Average Supplement. Rice bran 100 0 60 40 100 100 43 50 60 40 58 42 50 50 86 14 100 0 15 85 25 75 33 100 .Given at 0.7 kg/head/day in Hamlet 2 and 1.8 kg/head/day in Hamlet 4. Table 16. Feed resources and feeding systems in pigs. Hamlet 2 Hamlet 4Item Percent of Respondents s 95 12 88 20 30 30 20 5 14 18 63 83 17 27 72 86 14 91 9 74 26 64 36 S8 42 O 1.7 43 48 9 0.9 Main feeding system Scavenging Pen1ed Type of feed Kitchen refuse Grain Crop residues Commercial feed Source of feed Household Market Feeding frequency Twice a day Once a day Feed adequacy Adequate Inadequate Feed quality Good Average Poor Feed quantity (kg/head/day) Benchmark site characterisation in Oong Tam, Oong Phu, Binh Phuoc Province, Vietnam 259 Table 17. Feed resources and feeding systems in chickens. Item Hamlet 2 Hamlet 4 Percent of Respondents 86 14 89 11 7 88 5 4 88 4 93 7 76 24 70 30 83 13 74 26 78 22 46 54 0 0.05 45 50 5 0.05 Main feeding system Scavenging Pen-fed Type of feed Kitchen refuse Grain Crop residues Source of feed Household Market Feeding frequency Twice a day Once a da.v Feed adequacy Adequate Inadequate Feed quality Good Average Poor Feed quantity (kg/head/day) Meat Production, Marketing and Consumption The farmers never slaughtered cattle, buffaloes and pigs for consumption or for sale at home (Table 18). There were 98% of the respondents in Hamlet 2 and 75% of the respondents in Hamlet 4 that have never bought beef. However, the figures for having never bought pork are only 6% and 2% in Hamlets 2 and 4, respectively indicating that pork is the most popular meat consumed in Dong Tam. Poultry production in Dong Tam was observed as a main source ofprotein supply for family consumption (Table 19). The number of slaughtered chickens and ducks was 16 per household per year for Hamlet 2 and 30 per household per year for Hamlet 4. Most of these slaughtered animals were consumed at home. This was indicated by 94-100% of the respondents in both hamlets who reported that they have never bought chicken meat. This trend was also observed for egg production. 260 Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia Table 18. Beef and pork purchases. Hamlet 2 Hamlet 4Item Percent of Respondents 2 98 2 25 75 0.7 Beef Frequency of purchase Occasiona//y Never Quantity purchased (kg) Where purchased Market 61100 Pork 2 11 5 17 45 4 6 0.81 16,000 12 24 20 6 34 2 2 0.77 19,700 Frequency of purchase Dai~v 2-3 times a week Once a week Every two weeks Occasional~v Special events Never Quantity purchased (kg) Price/kg (VND) Where purchased Market 73 32 Labour Input for Animal Production Tables 20-23 summarise the frequency of ruminant and non-ruminant production activities and who in the household are performing these. Fifty nine percent (59%) of households in Hamlet 2 and 33% of the households in Hamlet 4 tethered their animals twice a day. For the tethering activity, the household aduh male and female in both hamlets contributed a significant part. In general, more household males than females were involved in management of the ruminants. However, in pig and pouhry production, more females performed more farm activities than the males. For instance, 77% of the households in Hamlet 2 and 65% of households in Hamlet 4 reported that females, rather than the males, did the feeding of pigs. Couples, and parents and children performing farm chores together, is a very popular and important practice in agricuhural production in Vietnam. Benchmark site characterisation in Oong T am, Oong Phu, Binh Phuoc Province, Vietnam 261 Table 19. Poultry meat and egg production. marketing and consumption. Chickens Ducks Item Hamlet 2 Hamlet 4 Hamlet 2 Hamlet 4 16 1.3 24.4 1.4 18,000 30 1.6 41.6 2.8 21,000 2 1.4 2.3 3 1.9 5.5 0.4 100.0 0.0 0.0 20.0 20.0 60.0 100.0 100.0 1.6 25,000 25.0 75.0 ~ 3.7 18 6 12 O 4.2 41 12 17 12 1,000 0.6 II 4 6 60.0 20.0 20.0 20.0 60.0 20.0 60.0 Meat produced on fann Number of animals slaughteredlyr Slaughter weight (kg/head) Amount consumed (kg/yr) Amount sold (kg/yr) Selling price per kg (VND) ,Major buyer (% of respondents) Storekeeper Middleman Other farmer Meat purchased Frequency of purchasing f/0 of respondents) Never Amount purchased each time (kg) Price per kg (VND) Source of purchased meat f/0 of respondents) Market Other farmers Eggs produced on fann Number of laying birds Eggs produced last month Number consumed Number hatched Number sold Selling price (VND/egg) Major buyer (% ofrespondents) Market Consumers Other farmers Egg purchased Number of eggs purchased last week Price paid (VND/egg) 3 1,100 9 1,000 8 1,200 7 1,200 -= No data. 262 Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia Table 20. Frequency of management activities in ruminant production. Activity Hamlet 2 Hamlet 4 Percent of Respondents 59 38 33 60 7 80 20 0 47 47 6 33 O 33 33 17 50 33 0 7 53 20 20 O 56 22 22 19 42 8 4 23 0 40 20 7 33 Tethering Twice a da:y Once a da.v Occasionally Grazing of animals Twice a day Once a day Occasional~y Giving supplementation Twice a day Once a day Weekly Monthly Collection of animal manure Twice a day Once a day Weekly Occasionally Cleaning pens/OOms Twice a day Once a day Weekly Monthly Occasionally Transporting animals to market Occasionally 100 100 -= No data Benchmark site characterisation in Oong Tam, Oong Phu, Binh Phuoc Province, Vietnam 263 Table 21. Household members involved with the management of ruminants. Activity Hamlet 2 Hamlet 4 Percent of Respondents 33 20 7 30 3 3 28 28 22 6 6 6 35 6 O 35 6 17 25 19 31 6 6 6 67 O O 33 O 33 33 17 O 17 50 13 6 25 6 40 30 10 20 0 67 IS O IS 4 39 39 II II O Tethering Househo/d ma/e Househo/d fema/e Househo/d chi/dren Househo/d ma/e and fema/e Househo/d ma/e, female and chi/dren Househo/d fema/e and chi/dren Grazing of animals Househo/d ma/e Househo/d fema/e Househo/d chi/dren Househo/d ma/e and fema/e Househo/d ma/e, fema/e and chi/dren Househo/d fema/e and chi/dren Giving supplementation , Househo/d ma/e Househo/d fema/e Househo/d ma/e and fema/e Househo/d ma/e, female and chi/dren Househo/dfema/e and chi/dren Collection of animal manure Househo/d ma/e Househo/d fema/e Househo/d chi/dren Househo/d ma/e, fema/e and chi/dren Household fema/e and children Cleaning pens/barns Househo/d ma/e Househo/d fema/e Househo/d chi/dren Househo/d ma/e and fema/e Househo/d fema/e and chi/dren Transporting animals to market Househo/d ma/e Househo/d fema/e Hired fema/e 80 20 0 47 47 7 264 Improving the contn:/Jution of livestock to crop-animal systems in rainfed areas in Southeast Asia Table 22. Frequency of performing management activities in pig and poultry production. Activity Pigs Poultry Pigs Poultry Percent of Respondents 96 4 56 44 91 9 78 16 17 44 22 6 11 0 0 0 0 100 27 59 9 0 0 0 0 1005 100 0 79 100 100 100 Feeding Twice a day Once a d~y Cleaning the barn Twice a day Once a day Weekly Nfonth~y Occasional~y Vaccinationl medication Occasional~y Breeding Occasional~y Marketin g/t ra nspo rtin 9 Occasional~y 100 100100 100 No data. Benchmark site characterisation in Dong T am, Dong Phu, Binh Phuoc Province, Vietnam 265 Table 23. Household members involved with the management of pigs and poultry. Activity Hamlet 2 Hamlet 4 Percent of Respondents 9 77 O 9 18 53 8 18 O 65 O 26 }6 44 4 29 5 O O 5 O 9 O 2 32 53 5 11 O 80 15 0 5 0 8 67 17 11 O O 89 14 21 57 46 36 18 O ~ 100 0 0 O 33 67 Feeding Household male Household female Household children Household male and female Household male, female and children Householdfemale and children Cleaning the barn Household male Household female Household children Household male and female Household female and children Vaccination/medication Household male Household female Household male and female Hired male Breeding Household male Household female Hired male Marketing /transporting Household male Household female 33 67 25 75 50 33 33 67 No data landholding Land and landholding situations in the two studied hamlets are shown in Table 24. Most of these lands are owned by households and have moderate soil fertility with no erosion observed. Households in Hamlet 2 owned 1-6 parcels of land while those in Hamlet 4 had 1-5 parcels. On the average, each household in Hamlet 2 had 2.8 land parcels with an average area of 0.74 ha. In Hamlet 4, each household had an average 2.1 parcels of land with an average area of 1.4 ha. The highest number of parcels per household was 6 and 5, for Hamlets 2 and 4 respectively. 266 Improving the contribution of livestock to crop-animal systems In rainfed areas in Southeast Asia Forty seven percent (47%) of the parcels in Hamlet 2 and 29% in Hamlet 4 were used for annual crops. Perennial crops also occupied a large proportion of the land, 49% in Hamlet 2 and 51% in Hamlet 4. Table 24. Land and landholding situation. Item Hamlet 2 Hamlet 4 Percent of Respondents 2.8 1-6 0.74 0.66 2.1 1-5 1.41 0.82 47 1 1 49 1 1 29 I 12 51 6 I 91 1 8 54 7 38 41 37 22 54 42 4 34 49 17 36 46 18 Parcels per household Average arealparce/ (ha) Number of parce/s Average area! parce/ (ha) Distance from homestead (km) Land use (% parcels) Annua/ crops Forest/Trees Orchard Perennia/crops Fa//ow Others Ownership status (% parcels) Owned Leased out/ Rented out Others Topography (% parcels) Up/and Low/and, notflooded Low/and, flooded Soil fertility indicator (% parcels) Good Average Poor Soil erosion (% parcels) Non-detectab/e S/ight/moderate Severe 63 34 3 55 40 5 Crop-animallnteractions The interactions between crop and animals are seen in the utilisation of the animal residues for fertiliser, animal feed and for other beneficial purposes (Table 25). Fanners in both hamlets collected animal manure. Majority of the respondents in both hamlets used manure as fertiliser. Fifty percent (50%) of the respondents in Hamlet 2 and 44% of those in Hamlet 4 sold manure at US$IO per cubic meter. In Hamlet 2,44% of the respondents used the entire rice straw crop residue as mulch or compost while the rest burned all the rice straw in the field (Table 26). Only 11% of the respondents in Hamlet 2 used a very small proportion of cassava residues (2%) as animal feed. A similar trend of the crop-residue utilisation was observed in Hamlet 4, except that no fanners in Hamlet 4 used cassava residues for animal feed. Table 25. Utilisation of animal manure. Hamlet 2 Hamlet 4 Item Cattle Buffa- loes Pigs Chick- ens Cattle Buffa- locs Pigs Chick- ens 61 39 44 56 42 58 24 76 84 16 31 69 67 33 36 64 Manure collection Yes No Utilisation Asfertiliser Asfuel Asfeed Sold Other purposes 50 0 0 50 0 100 0 0 0 0 86 O 7 O 7 51 O 5 44 O 90 0 0 10 0 86 O 5 O 9 100 0 0 0 0 ~ = No data. Table 26. Utilisation of crop residues. Rice Maize Cassava Item ~ntity % Res- Quantity % Res- ~ntity % Res- (%) ponc.knts (%) ponc.knts (%) ponc.knts 2 88 77 11 56 33 100 100 44 56 100 100 73 27 Hamlet 2 Feed Mu/ch/compost//eft in fie/d Burned Hamlet 4 Mu/ch/compost//~fl infie/d Burned 100 loo 26 74 100 loo 63 37 77 85 57 43 -= No data. 268 Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia LESSONS LEARNT The survey is a method of quantifying crop-animal production phenomena to characterise the production systems and relationships between parameters of the production systems. The accuracy of the obtained parameters depends on the infonnation collected by the interviewers and sample size. In general, the assigned sample size of 50 households per site is appropriate. However, the production situation in each household did not cover all items of interest in the questionnaire. This resulted in obtaining parameters that were a good reference for the common items but were biased for some items that had a small frequency of occurrence. CONSTRAINTS TO PRODUCTION Available food resources and nutrition are the major constraints to production, especially during the severity of the dry season. Given the priority to beefproduction, the significance of better understanding of the quantity and quality to improve feeding systems for beef production is important. Associated with this, pig, and duck production are constrained by poor feeding and this will require improvement in the future. To support beefproduction, increased numbers and a breeding program to support this are required. Cassava production is important in the village, but due to a lack of knowledge and access to technology, serious soil fertility and erosion problems have been created. The introduction of legume trees into the cropping system and training of farmers to understand the advantages of improved soil fertility and also increased availability of feeds from cassava will be important. ACKNOWLEDGEMENT This work was funded by the CASREN project of the International Livestock Research Institute (ILRI). The authors would like to express their sincere thanks to the Director, Board of the Institute of Agricultural Science of South Vietnam who energetically supported us during the implementation of the work. Special thanks are also due to Dr. C. Devendra and Dr. D. Pezo from ILRI and the Leaders of Dong Tam village for their valuable help. Benchmark site characterisation in Oong T am, Oong Phu, Binh Phuoc Province, Vietnam 269 REFERENCES Li-Pun H., 1999a. Improvement of Crop-animal Systems in Rainfed Areas: Introductory Remarks. In: Proceedings of the Planning Workshop on the Crop-Animal Systems Project. (ed.) C. Devendra. ILRl, Los Bafios, Philippines, 1-4 June 1999, pp.I-2. Li-PunH., 1999b. The External Environment: Implications for ILRl's Systems Research. Proceedings of the Planning Workshop on the Crop-Animal Systems Project. (ed.) C. Devendra. ILRl, Los Bafios, Philippines, 1-4 June 1999, pp.3-8. Luc, K.M., 1999. Improvement of Crop-animal Systems in Vietnam. Proceedings of the Planning Workshop on the Crop-Animal Systems Project. (ed.) C. Devendra. ILRl, Los Bafios, Philippines, 1-4 June 1999, pp.96-108. General Statistical Office. 1999. Statistical Data of Agriculture, Forestry and Fishery 1990-1998 and Forecast in the year 2000. Statistical Publishing House, Hanoi. 1999. SAS Institute. 1999. SAS User's guide: Statistics. SAS Institute Inc., Cary, NC. 270 Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia DISCUSSIONS ON COUNTRY PRESENT A TIONS China C. Chantalakhana We want to make sure that we know the real problems why we need these interventions. We want to be forward-looking; whatever interventions will be proposed, we need to ask what will happen after the project is over . There are reservations about drenching as an intervention. There is a need to get a handle on what really are the problems causing high mortality. We may need to go slow on drenching at this point. C.Devendra How to reduce disease outbreak in animals is important.z. Cungen Philippines The report indicates that most farmers are educated and have other activities. How much income is generated from other off-farm activities? H. Li-Pun When there are no agricultural activities in the fann, fanners work temporarily in other activities. Although we haven't quantified that income, we estimate it in the order ofUS$IOO/month. E. Villar T. Soedjana The cost of cattle production, not including feeding costs, seem to be high (US$200/animal). Why do farmers keep cattle? Is the activity still profitable? E. Villar Farmers keep cattle for two reasons: a) it is part of their culture to raise animals, and b) market opportunities for cattle are good as there is an auction market relatively close by. I guess the costs for cattle production need to be reviewed.P. Faylon I agree. The accuracy of some of estimates is low.E. Villar How do you feel regarding the data on household consumption of livestock products? M. Wanapat E. Villar We found data in some way messy. I guess it is because they were too disaggregated. Le Viet Ly Being the BMS in the rainfed lowlands, I would expect that the population ofbuffaloes should be high, but apparently, that is not the case in your data. Any comments? E. Villar Roads make services more accessible, and the use of mechanisation is high in the region. E. Villar P. Faylon People say that in Pangasinan crops are for food and animals are for generating cash. Do your data support this hypothesis? E. Villar The contribution of farm animals to the total farm income is 7- 15% but it could be higher. Farmers are more crop-oriented and animals are a source of cash like a piggy bank. The contribution of animals seems to be limited. How can this be increased? C.Devendra Probably, interventions should be oriented first to improve crop production. It may have implications on animal production activities and on the totality of the system. E. Villar The volume of data collected is large. I guess some refinement! reduction of data is needed. One possibility is to stratify the sample and then make analysis by stratum. It will help to target technology to different groups of farmers. M. Jabbar It seems there are no differences in the feeding systems between dry and wet season, therefore you could integrate the data ofboth seasons. P. Faylon Thailand You surveyed a total of 78 farms, 38 dairy and 40 non-dairy . Were those in different locations? P. Faylon No, all of them are in the same district, although some of the dairy farms are in some way clustered. M. Wanapat Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia272 H. Li-Pun How was dairy developed in the BMS? M. Wanapat Dairy is an option promoted by the government to give farmers an opportunity to improve their standard of living. Now, for the purposes of this project, all the dairy farmers (38) present in the BMS were surveyed. M. Jabbar Hearing your answer to H. Li-Pun, I have some concern about the representativeness of the sample. What proportion of the population of farms in the BMS was included in the sample. M. Wanapat I do not have the information with me now. M. Jabbar I understood in your presentation that non-dairy farms would be the controls for dairy farms. Is it what you are proposing? M. Wanapat No, we will work with dairy farms only. Therefore, interventions and control will be implemented in equivalent farms. Indonesia H. Li-Pun You mentioned that the BMS was chosen based on previous experiences. What were those? Which were the main outcomes? A. Djajanegara The CRSP -Small Ruminants Project worked there in the' 80s. At that time, the project provided sheep to the farmers, in a sort of rotational scheme. Once the project finished, it was difficult to monitor the animals Le Viet Ly Do fanners consider animals as a sort of savings account? A. Djajanegara I would say, more than that, as marketing is not a problem in the area of study. C.Devendra In the presentation, it was not clear which are the constraints and where the opportunities for research were. A. Djajanegara Definitely, the market is not a constraint. Credit could be, and also, availability of feed resources. M. Jabbar Are most cattle purebred Holsteins? When were these introduced to the area? Is it a new approach limited only to a small area? A. Djajanegara Purebred Holsteins are basically for fattening, and were introduced to the area about four years ago. This has been made on a loan basis. P. Faylon I understand one of the criteria you used for selecting the BMS was that there was a development project in place. Would it be possible to get higher animal responses (e.g., 800 g LWG/day) using local resources? A. Djajanegara To obtain those gains year-round will probably need the use of urea-molasses blocks, but our interest is on doing everything using lacally available feed resources. C.Devendra In the way data has been presented, it gives the impression that it is more of a development-oriented project rather than a research project. I would suggest that there are good opportunities in the BMS for doing more research oriented to a much more integrated use ofresources that are available. For this reason, it will be very important to reassess the results, constraints and define the needs- based research. Vietnam T. Soedjana We all share common goals. We need to come up with a common solution, with modifications here and there, recognizing the diversity across EMS. There is also a need to come up with the kind of research /technology that can fill the gap in the EMS. P. Faylon I suggest to use the term Vietnamese chicken instead of scavenging chicken. C. Chantalakhana On the use of bigger bulls in the village, how will this be done? General comments on country presentations by P. Safran . In the presentation from China, it was not clear what the relationships were between animals and crops in those systems, and how these can help to alleviate poverty In the presentation from Indonesia, it was not clear what can be done, and what will be the proposal for the next stage of the project Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia274 . . . . . . . In the presentation from Vietnam, the importance and role of animals in the systems under study is not quite clear The presentation form Thailand had a strong emphasis on the nutritional aspects, but it was not clear what role livestock could play for poor farmers to improve Gender issues in the livestock sector need to be emphasised There is a lot ofheterogeneity in the systems under study, among and within BMS's, therefore there is a need for summarising information in order to define a sort of benchmark status of these systems None of the speakers made a reference to livestock policy matters in their countries, and how these affect the population targeted by the project. There is need to emphasise the role of animals in improving livelihood, and draw recommendations on how the smallholder livestock sector can be improved. Most of the characterisation effort has been concentrated on the crop-animal systems as such, but not much on how extension systems are, and how these can be improved. A question all participants should ask themselves is why do we want to do research in crop-animal systems? For whom? How is CASREN going to be sustained? 275Oiscussions on country presentations SESSION 3 PROPOSED INTERVENTIONS PROPOSED INTERVENTIONS* CHINA Proposed Interventions To Overcome Constraints Rationale Fanners use crop residues as the main feed for cattle and buffalo; some even hammer-mill straws for feed to pigs and even chicken. Theoretically, these residues have very high fibre content, low digestibility and crude protein. This type of feeding obviously can not meet the requirements of animals, especially the demand of young growing animals and reproduction of breeding females, resulting in slow growth in early life and poor reproduction perfonnance. Disease control and animal health are very important in the livestock industry . The survey results show that the veterinary service at the BMS is not sufficient at present, and this includes a vaccination program and extension. For grazing animals, drenching is very important because they are eating from a big "common pot" which may help the spread of parasites. It seems to us that the fanners are not aware of this enough. It is a common knowledge that the productivity of animals is detennined by the nutrition level, health status and genetic make-up of the animals. At present, the breed improvement program of pigs and cattle is fully accepted by the fanners. Priorities From the survey results, it is clear that feed production, nutrition level and animal health need to be improved. It is proposed that we start in the near future, some demonstrations on multi-nutrient blocks, straw treatment, and regular drenching, especially for young grazing animals. In the long run, we shall try to introduce fodder trees and forage crops. In the BMS and China as a whole, output of grain is more than a people need. More and more grains are used as feed. For pigs and poultry , this is acceptable, but for the ruminants, the efficiency is low. On the other hand, many experts and the government has been pushing a campaign to change the Chinese two-component growing-system (grains and cash crops) into a three-component system of grains, cash crops and feed production. * Taken from the individual country presentations. Proposed interventions 279 On-station research For the purpose ofmulti-nutrients block making, it is necessary to do some plant and soil sample analysis in the laboratory at first. More and more people are becoming anxious about environment pollution from animal keeping, especially from the big intensive farms. For our purpose, we are more concerned about nutrients transfer and recycling. So, it may be necessary to do some grazing studies. On-farm research As mentioned before, feed resources are relatively limited and the quality of feed is poor in the EMS. So, to improve the quality, supply and utilisation efficiency shall be the priorities for on-farm research, including silage making, ammonia treatment of straw, fodder growing and multi-nutrients block feeding. The second aspect we shall focus on is regular drenching of young grazing animals. For this purpose, we need to do fecal egg counts and larvae hatching and classification if necessary . In the vaccination program, we will work more closely with the local veterinary station to try to reduce the mortality of animals. This, at least, will reduce the death rate of pigs, chicken and goats, to some extent. Anticipated Output Through the above-mentioned interventions, we can anticipate that the animal nutrition status of animals will improve to some extent. Likewise, the early growth of cattle and goat shall be able to increase by 10%, the death rates of piglets, and baby chickens and goats can be reduced to 5-10%. The overall output or productivity of the animal industry in the BMS will increase by 7 -10%. Meanwhile, this will increase the confidence of farmers to keep animals. Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia280 PHILIPPINES Proposed Interventions and Activities After the conduct of the participatory rural appraisal and the focus group discussion in February 2000, the Philippine Team started with preliminary activities. The survey, the PRA-FGD and the selection offarmer-partners were some of the activities completed to date. A Project Field Technician has also been hired to assist the Team in conducting field activities. In fact, to date, th~ forage resource inventory has been started by the project field assistant, who was tasked to identify and quantify the feed resource/ration given to or accessed by the animals every two weeks. Likewise he will also record the potential feed resources during the period of collection. Also accomplished was the identification of the six farmer-cooperators. These farmers have been briefed about the Project in their village and oriented on the roles they will take on. The following is a list of activities proposed and started for year 2000. 1. Establishment of a forage nursery An area, strategically located and very accessible to the farmer-collaborators, shall be developed into a forage nursery .Only improved forage species (grasses and legumes), which are known to have wide adaptability, shall be planted.The forage nursery shall serve as a demonstration area as well as the farmers , source of planting materials. This will address the farmers' problems on lack of planting materials and good quality forages which were raised during the FGD. A suitable area for this purpose shall be selected and seeds and vegetative planting materials of improved forages, requested from CIA T , IRRI. The farmer-collaborators and the project staff will jointly establish and maintain the nursery . 2. Introduction of feed technologies The intervention deals with the introduction of certain feed technologies that are suitable in the area and compatible with the farmers' resources. While it is indicated in the BMS survey that the animals of the farmer-respondents are adequately fed, the general performance of the animals shows otherwise. Looking at the kind of feeds being used, one can conclude that the available feeds are indeed, nutritionally limited. Coupled with unstable supply on a year-round basis, this has great impact on animal productivity . A number of feeding technologies have been tested and have shown improvement in animal performance. It is only a matter of choosing the most appropriate technology . The feeding interventions shall therefore address this concern with the end-view of mitigating the problem that limits animal productivity . 281Proposed interventions A "basket of options" will be offered to farmers. The technologies that will be included in the "basket of options" are only those that have passed the "ex-ante analysis". The feed technologies that may be included are the following: a. Use of leguminous fodder trees/shrubs as supplement b. Utilisation ofprotein-rich crop residues c. UMMB supplementation d. Use of mixtures of forages e. Use of forages with different physical forms, e.g. dried, pre-wilted, fresh Each participating farmer will select a particular feed technology that he will use in his animal for a period of two years. Productive and reproduction data shall be gathered and analysed. These data shall be compared to those not employing the feeding intervention. Likewise, the benefits accruing from the adoption of the interventions shall be assessed. 3. Soil fertility evaluation of the area This intervention shall be introduced to develop a fertiliser recommendation for each crop grown by the identified six farmer-cooperators. This shall be based primarily on soil test values. Likewise, the efficiency of each fertiliser recommendation shall be monitored and evaluated. Through soil analysis, the fertility level of farms in the barangay (village) shall be assessed to come up with a soiHertility map using the GIS software. As limited cropping option due to decreasing soil fertility has been identified as a pressing problem, it was agr~ed that biophysical assessment of the soil be done. It shall be upon the results of this study that fertiliser recommendations will be developed for each crop grown by the identified six farmers. Moreover, it will be these results that will dictate the soil fertility map. To achieve the aims of the study, the following shall be done: .Soil survey and analysis .Development offertiliser recommendation for each crop based on results of soil analysis for the wet and dry season cropping .Monitoring and evaluation of crop performance .Identification of gaps for further research .Training on fertiliser handling and management 4. Crop management monitoring and evaluation Information gathered on crop production per se needs actual validation in the field to come up with an improved cropping pattern/system relative to water availability and farmers' available resources as well as animal production. Through close monitoring, gaps can be identified for further research. This will hopefully improve the whole farming system and increase the role of livestock in income generation. 282 Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia This will be done through close monitoring of the farming activities of the six farmer-cooperators and provision of technical assistance particularly on soil and crop management. Along the way, gaps can be identified for further research geared towards improvement of the whole farming system. Moreover, these efforts can improve the role of livestock and eventually increase farm family income. 5. Participatory communication activities on crop-animal management These interventions shall centre on the needs vis-a-vis the resources and capabilities of the target beneficiaries. The identified farmer-cooperators, their households and other interested farmers in the area shall be invited to take part, from the planning to evaluation stages of a campaign that will tackle various needs on crop-animal integration.' Technology, no matter how adaptive, cannot contribute to development unless utilised. And utilisation cannot proceed unless such technologies are communicated. It is the aim of this study to bring to the attention of the intended clients the technological approaches on crop-animal management. Pre-campaign studies will be done to analyse the setting of the campaign -the organisational aspects, physical resources, communication channels available, audience preferences, baseline knowledge, attitude and practice (KAP). These will be used as springboard for planning succeeding strategies with the farmers and the initial KAP shall serve as the basis for evaluating the effect and ultimate impact of the technology on their individual competencies and sdcial, ecological and economic assets. Initially, pre-campaign studies shall be conducted to serve as basis for the planning of the communication campaign. This shall involve the following activities: .Situational analysis -determination of available communication channels In area .Audience analysis -preference for information materials and style; pre- test on KAP After these, participatory planning shall be done with the farmers to design the communication campaign in detail. The strategies to be included shall be audience- dictated and shall depend upon the their preferences vis-a-vis their needs. These may include any or all of the following: .Packaging of print support materials .Home visits/ education tours .School-on-the-air As hands-on training courses have been identified during the FGD and the survey, these shall be programmed to provide farmers with the rudiments of the various aspects of crop-livestock management. The following are the topics of the training courses, as identified during the FGD: .Various aspects of animal management .Feed resources and feeding management .Fertiliser handling and management Proposed interventions 283 Henceforth the communication campaign shall be conducted to promote the technologies deemed necessary for the successful implementation of the Project. The researchers and the farmers shall then jointly evaluate this. 6. Ex-ante analysis of proposed interventions Despite several research breakthroughs generated by the national research system, many of them still remain unutilised by their intended beneficiaries. Hence, a system to analyse the chances of the proposed interventions from being widely adopted is being proposed. An added benefit of this exercise is the reduction in wasted investments. This study aims to: I) examine the appropriateness and suitability of the proposed interventions to the target community in terms of actual needs, resource availability, social acceptability, technical feasibility, economic viability, environmental soundness, and sustainability; 2) determine the financial value/significance of the proposed interventions 3) develop practical recommendations for enhancing the adoption of the proposed interventions. The methodology to be used in this exercise is technology assessment and benefit- cost analysis 7. Performance evaluation of the ongoing intervention Despite claims of technological superiority by many technology generators, many recommendations offered to farmers fail to deliver on the expected net benefits they promised, resulting in frustration and loss of reputation for both the researchers/change agents and their clients. Thus, conducting periodic monitoring and evaluation is necessary to avoid committing serious mistakes. This study aims to: .assess the performance of the implemented interventions in terms of their achievements .identify implementation loopholes before their effects become more serious .come up with alternative recommendations for improving project performance Any evaluation activity typically begins with comparing something with another; therefore there must first be an agreement on what is acceptable as a standard indicator of achievement. In this case, there are several ways of measuring performance: 284 Improving the contnoution of livestock to crop-animal systems in rainfed areas in Southeast Asia Beforel AfterTi T2Outside Programme y ---y I 2 With/Withou. r-;;lX X I 2 Programme Achieved Achieved/ Expected Programme Planned x x*I THAILAND The following interventions are proposed: 1. Crops and cropping systems The existing cropping patterns practised by the farmers in the study area are monocropping of cassava and kenaf in the upland area and monocropping of rice in the lowland area. Therefore, the following are proposed (Figure I): .Intercropping cassava with cowpea/leucaena .Cowpea before/ after rice Cowpea is considered as a crop that is drought- and acid-tolerant. In addition, cowpea is a good protein vegetable for human consumption and the whole crop can be fed to dairy cattle as a high protein fodder. While leucaena is a fast growing tree and could be cut and used as a protein rich fodder as well, it can also be used to enrich soil nitrogen. Another essential dimension is to emphasise the use ofno-farm fertiliser for crop production especially in rice and cassava. Figure 1. Potential cropping pattern Cassava hay2. Farmers have been invited to provide 2-4 rais to grow cassava for cassava hay and some for tuber chip production. Cassava hay production is recommended to ensure high quality roughage availability and to increase milk yield and quality, especially milk fat and solids-not-fat concentrations. Most importantly, it is expected that use of concentrate would be reduced, thereby subsequently reducing cost of production. 3. Rice straw Rice straw plays an important role as a roughage for dairy production, particularly during the dry season. Treatment of rice straw with regard to treatment method, duration and efficient use to increase milk yield and quality (% fat, %SNF) will be pursued. Treatment and storage should be improved. 4. Pasture improvement Grass/legume mixtures will be estab]ished and used. Increasing both quantity and qua]ity will be aimed at through on-farm manure ferti]isation. Proper management and :uti]isation of grass at suitab]e maturity shou]d be taken. 5. Concentrate making and supplementation On-farn1 concentrate mixtures will be introduced using local resources to ensure lower cost and proper use especially in improving the yield and quality of milk. 6. High-quality feed block (HQFB) HQFB as a strategic supplement should be used to improve rumen ecology and subsequent milk yield and compositions. Other aspects especially reproduction and health could be improved. 7. Monitoring of dairy feeding and feed resources. A year-round feeding practice, feed resources availability and their uses by the dairy farmers will be monitored on a bi-weekly basis throughout the year. A prepared form will be used for each farmer to fill in with details on a daily basis. A bi-weekly visit of the team members will be regularly conducted. This will give team members a chance to talk, discuss, as well as to work with farmers on the forms which are left on the farms. Team members will work closely with the farmers. Training on feed preparation for farmers will be done. Sampling of feeds will be done and samples will be ana lysed for essential chemical compositions. Data on feeds used and milk yield will be recorded. Milk samples will also be collected at intervals for milk compositions analyses (% fat, protein, lactose, SNF, TS, MUN). Cost of production (inputs) and profit (outputs) of milk sale will be assessed. In addition, blood samples, when possible, will be taken for analyses of BUN, b-OH butyrate. 8. On-Station Support Research Results from the BMS reveal that the overall efficiency of livestock productivity particularly dairy cattle, would mainly depend on availability and quality of feed resources throughout the year. It is therefore imperative to conduct on-station research in parallel with the on-farm research to obtain necessary and useful information for further implementation. The following research activities have been initiated and are proposed. (i) Study on yield and nutritive values of cassava hay. A study will be conducted to compare unfertilised, fertilised (manure) and a variety of cassava production for hay making. An initial harvest of cassava will be done at three months and followed every two,months to harvest the regrowth. The samples will be assessed on yield per area per cut and will be analysed for chemical compositions. (ii) Study on cassava hay (CH) supplementation levels to reduce concentrate use. Various levels of cassava hay (0.1,2,3 kg/hd/d) will be supplemented with levels of concentrates (I :2, I :2+CH, I :3+CH, I :4+CH). Milk yield and compositions will be measured and profit will be assessed. Urea-treated rice straw will be used as a roughage. Rumen fluid will be sampled and analysed for pH, NHJ-N, VF A. Blood samples will be analysed for BUN . (iii) Study on using cassava hay as a protein source in HQFB. Two trials will be conducted to assess the effect of CH as a protein source in the block on rumen ecology (pH, microbes) and fermentation (NHJ-N, VFAs) and microbial protein synthesis using urinary purine derivatives in cattle and buffaloes. A second trial will be done with milking cows fed on urea-treated rice straw. Milk yield and compositions will be measured as well as rumen pH, VF A, NHJ-N Cost-profit will be assessed. Blood samples will be analysed for BUN. PIE ratio of the feeds will be investigated. (iv) Study on effects of different roughages on rumen ecology, fermentation and digestibility. Four types of roughages, untreated rice straw, urea-treated rice straw, grass hay and cassava hay, have been fed to buffaloes. Rumen samples are to be collected to be analysed for microorganisms, pH, NHJ-N, VF A. Digestibility will be determined using internal indicator (AlA). 8. Training and Workshops Training courses or workshops have been and will be continuously given to all dairy fanners initially at the University but mostly, these will be organised and conducted on-fann in small groups. Lectures and 'practical aspects as well as demonstrations will be organised. These will enable fanners to understand the introduced technologies through on-hand experiences. A fanner-to-fanner interactive atmosphere is also highly expected. Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia288 INDONESIA Proposed Interventions Background The production and productivity of crops very much depend on the nutrient available from the soil. With poor soil conditions, farmers face problems in buying chemical fertilisers, while the use of animal manure has been taken for granted. In the crop-animal systems proposed interventions, the cycle of crop residue utilisation as feed and the return of the manure to the soil will provide nutrients to the crops. The basic fact that needs to be understood in crop-animal systems is how to make better use of available resources. It is therefore important that the proposed interventions improve effective utilisation of manure as fertiliser for crop production and the use of crop residues as feed and as a means to increase the supply of better quality feed. Rationale Under the present condition, fanners have limited resources and they need support for their activities in crop and animal production. The harvest time of crops is detennined by the life cycle and seasonality of the particular crop and these apparently contribute to the irregular income of the fanner. Methodologies General conduct The proposed interventions will be conducted at the farmers' location with full participation of the farmers. The work will also serve as a demonstration plot for other farmers. To obtain individual and group responses, a regular meeting with the involved farmers is planned to be held. Processing of manure The number of farmers that will be included in the work will be composed of around five farmers' groups that live close to each other, hence, there will be five sites where the processing technique will be demonstrated. The site will be a communal land area and the facilities for processing the manure will be established. The animal manure of each participating farmer will be stored and each farmer will later be given the processed manure. It is expected that around five tons of manure will be processed on each site and each farmer will get around one ton of processed manure. A probiotic agent will be initially purchased and later on produced at the site. The dynamics of manure quality changes will be monitored until the manure is ready for application. Detailed descriptions will also be conducted on-station. The variation among manure sources in conjunction with feed given to the animals will be studied. After the processed manure is used as fertiliser, comparisons will be made on crop production and productivity with regular manure. The possible plant- manure interaction aspects will be analysed. Planting of tree legumes The tree legumes that will be plRnted are known by farmers, i.e. Leucaena leucocephala. Gliricidia sepium. Calliandra callothyrsus. The trees will be grown in a forestry land area as a buffer zone for environmental protection. The trees will be planted from planting material that has been grown earlier. The number of trees for each participating farmer is determined on the basis of number of animals. Discussions will be held with the head of the village and forestry services officers. It is hoped that the farmers would take care of the trees and harvest the leaves according to the directions gIven. Three strata forage systems The Three Strata Forage Systems will involve five farmers who have a reasonably large size of land area. The general cropping systems will not be changed, except that plants in the area will include grasses, fodder shrubs and fodder trees. The system is considered to provide continuous supply of forage to the animals. The daily harvesting time follows the directions for the respective plants in accordance with the amount needed. Fodder tree leaves will serve as a supplement. The feeding of the leaves will be conducted when the trees are at least six months old. Growth of animals will be monitored including production of leaves and socio-economic measurements. On-station research On-station research includes assessment of the interventions under well-controlled situations at the Research Institute for Animal Production in Ciawi. The research will have more detailed measurements, including laboratory analysis, processing of manure, growth of tree legumes growth and feeding aspects. Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia290 The constraints considered important for on-station research include: .Quality assessment of processed manure to serve as standard material .Assessment of the year-round feeding systems being introduced .Chemical analysis of all samples collected on farm. It is apparent that under on- farm conditions, it is difficult to achieve well-controlled and standard samples. This requires an assessment of the variability of factors found under village situations. The following are proposed: .Quality of processed manure coming from different sources .Changes in quality of processed manure with time .Evaluation of microbial types and population changes in the processed manure .Supplementary feeding of mineral additives for an all-forage ration VIETNAM The general objective is to increase the productivity of the crop-animal systems to improve the livelihood of the poor farmers in the EMS. In livestock production, beef cattle was recognised as the animal that can help poverty alleviation in Dong Tam. The development of beef cattle production in this village has been considered by the government authorities as indicated by several projects. However. the methodologies of increasing beef cattle production in this village have not yet been investigated so far. Among the research interventions for improvement ofbeef cattle production, studies on feed and feeding systems will be a main component and are expected to have a long-term effect in this village. Moreover, it is determined that to develop beef cattle production in Dong Tam, several interventions should be considered simultaneously. which can be summarised as follows: I. Studies on feed and feeding systems of ruminants in the village .Year-round feed and feeding systems for ruminants .Introduction of the Stylosanthes hamata for improving the grazing pastures .Utilisation of cassava leaf for ruminants 2. Introduction of bulls and cows to improve beef production is also important. This activity is proposed as presently there are very few working bulls which have been used for many years to produce many generations of cows in the village. Inbreeding has affected the performance of the herd. The introduction of at least one new bull to replace the current bull in the service area will give a more positive effect on beef cattle production in the village. The village authorities have also requested an increase in herd size of cows. but access to capital is limiting. The 6 month-age calves produced from the project cows will be transferred to other households for the purpose of increasing the herd size. 3 4. Development of local feed resources for backyard chicken production to generate jobs for women. Though poultry production is a minor activity in the framework of this project, it is an important source of protein for households and will support cattle production in the village. Some farmers have owned cattle but they had to sell the animals in times of food emergency. Chicken production is considered as a short-term benefit to support the long-term benefit of cattle production. Technologies to increase cassava production and soil erosion control in Dong Tam. For crops, cassava production for starch purposes is one of the important cash crops of the farmers in the village. Cassava production is often related to soil degradation and erosion. This activity is expected to help farmers understand the manner of soil degradation and erosion and methods of control by intercropping cassava with legume trees. It is expected that the introduction of legume tree intercropping with cassava will augment nutrients and serve as sources of feeds for cattle. Improving the contribution of livestock to crop-animal systems In rainfed areas in Southeast Asia292 DISCUSSIONS ON PROPOSED INTERVENTIONS C. Devendra introduced this section by emphasizing the criteria for identifying proposed interventions: .Need to focus and identify proposed interventions, and not produce a shopping list .Need for prioritisation, including short- and long-term goals and more realistic identification of what can be delivered .Ensure relevance .Potential opportunities for improving the contribution of animals to the small farm system livelihood and poverty soil fertility sustainability environment, etc. .Aiming for three to four interventions, to be realistic. Discussions emphasised that the project leaders need to decide that the interventions proposed really came out of the results of the BMS survey. The Steering Committee will have to decide to accept or not the proposed interventions, given the available time period for the project. China Comments included the need to know the real problems and why we need these interventions. Need for supporting data was also raised. Also, interventions need to be forward- looking and whatever interventions will be proposed, would also need to consider what will happen after the project is over . Reservations were expressed about drenching as an intervention. There is a need to get a handle on what really are the problems causing high mortality; we may need to go slow on drenching at this point. Philippines Comments included the observation that most of the interventions are technology transfer and development activities. These cover a broad range of activities. Also, more testing of technology is appropriate. This comment is applicable to other countries as well. Clarification was sought about whether feed adequacy referred to quantity or quality of feed since this has subsequent bearing on experimental design of interventions. It was pointed out that responses to the survey are not consistent with the validated situations in the EMS. Hence, nutritiori is still hypothesised as a problem. It was also pointed out that the presentation is based on agreed interventions with the farmers. The broadness is a function of the inability of the team to describe the interactions in the EMS. There is a need to connect the performance of the crop with the performance of the animals. Hence, there is a need for a model as a framework for proposing the right interventions, which can also link the interventions with the broad goal of poverty reduction. Additionally, it was also observed that some are development activities, e.g., feed nursery, some are research activities that can be undertaken by ILRI. Thailand Clarification was sought on the allocation of farms, i.e., 30 intervened, 8 not intervened, and within the 30, 5-6 minimum in each sub-group with one or more interventions. Specific comments were made on the issues of analysing the results from the interventions given the existing experimental design. This was going to be reexamined. A suggestion was made about the use of producing a module on cassava/dairy system. Add-on activities like treatment of rice straws need not be given priority . Additional comments included concern about the bias towards nutrition in the proposed interventions on-farm; need to target specific interventions, i.e., what is the role of nutrition in cassava/dairy system; how much of the proposed activities are KKU activities and which are CASREN activities; need to refocus the proposed interventions. For these reasons, it was important to realistically think about what are really the critical things that can be done given the constraints of available resources. Indonesia Specific comments and concerns were made about the constraints and interventions. These needed careful interpretation. There is a need to use the data in each BMS to come 1.1p with a detailed description of the crop-animal systems in the BMS and rethink the issues. It was also emphasised that there is a need to review from past experiences why existing technologies have not been adopted. What are the facilitating and constraining factors? Some of the constraints may be off-farm, some may be on-farm. There is also the need to understand the constraints to adoption and the whole process of adoption. Vietnam A comment was made that we all share common goals. We need to come up with a common solution, with modifications here and there, recognising the diversity across EMS. Also it was important to identify appropriate research and technologies that can meet the needs of the EMS. SESSION IV FINAL DISCUSSIONS In the open forum, the following additional comments were made. .For China, marketing is not an issue in research terms .CASREN will not try to solve all the problems; some will be done by NARS and others by ILRI scientists There are three issues that need to be resolved: .Need for a framework for the analysis and synthesis of the BMS data; related is the issue of backup support .Prioritisation of ongoing activities as well as those that will be done across all the sites .Indications of interactions to be used, noting that this was mostly an important NARS responsibility. Prioritisation of Interventions In order to focus on the validity of proposed interventions by country, discussions centered on the relative importance of seven broad interventions. The results of this analysis are shown in the table below, within which the proposed interventions will be implemented. Prioritisation of Interventions China philip- pines Indo- nesia Thai- land Viet- nam Intervention *** *** ** ** * *** ** *** * ** *** * ** * ** * *** * * * * *** * ** ** ** *** * *** *** ** ** ** *** ** Feeds and feeding Diseases Management Services Breeds and breeding Markets Policy *** -first priority ** -second priority * -third priority 295Final discussions ANNEXES ANNEX A. WORKSHOP PROGRAMME Crop-animal Systems Research Network (CASREN) May 1-7,2000 01 May (Monday) Arrival of participants at Kumming 02 May (Tuesday) 09:00-09: 1 S Introductory welcome remarks .Ho Li-Pun 09: 15-09:30 Greetings from Y unnan Animal Husbandry Bureau 09:30-09:40 Greetings from ADB P. Sa/ran 09:40-10:00 Introductions of participants and objectives of the workshop. C. Devendra lO:OO-lO:l5 Coffee break Session I -Ecoregional researc.. on crop-animal systems lO:l5-ll:00 I. Crop-animal research perspectives. H. Li-Pun 11 :00-11 :45 2. Policy options for development of the livestock sector livestock sector in Vietnam. ...M. Jabbar 11:45-12:00 Discussion 12:00-13:00 Lunch 13:00-13:45 3. GIS application for site characterisation and definition of research domains. L Lapar 299Workshop programme 13:45-14:00 Discussion 14:00-14:45 4. Methodologies to address year round feeding systems. C. Devendra and D. Pezo 14:45-15:00 Discussion Coffee break15:30-15:45 15:00-15:45 5. Ex-ante analysis for technical interventions. D. Pezo 15:45-16:00 Discussion 03 May (Wednesday) 08:30-09:15 6. Methodologies for data analyses D.Pezo 09:15~09:30 Discussion Session II -Country Presentations -Progress results 09:30-10:30 7. China. Z.Kaidian 10:30-11 :00 Discussion Coffee break11:00-11: IS 8. Philippines I E. Villar11:15-12:15 12:15-12:45 Discussion 12:45-13:30 Lunch 13 :30-14:30 9. Thailand I M. Wanapat 14:30-15:00 Discussion 15:00-16:00 10. Indonesia. A. Djajanegara 16:00-16:30 Discussion Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia300 16:30-16:45 Coffee break 16:45-17:45 11. Vietnam I Khieu M. Luc 17:45-18:15 Discussion 19:00-22:00 Dinner- Reception by Yunnan Beef Cattle and Pasture Research Centre 04 May (Thursday) 08:30-09:00 Observations on Country Presentations. H. Li-Pun and C. Devendra Session IV- Implementation 09:00-10: 15 Presentation of proposed interventions and work plans for each BMS by the National Coordinators China Philippines Thailand Indonesia Vietnam 09:00. 09:15. 09:30. 09;45. 10:00. 10:15-10:30 Coffee break 10:30-12:00 Discussions on individual work plans 12:00-13:30 Lunch Session V -Field Trip 13:30 Proceed to Nanjian (about 6 hours) 19:00-21 ;00 Dinner -Reception by ILRI Workshop programme 301 .9:15 .9:30 .9:45 .10:00 .10:15 OS May (Friday) 08:00-12:30 Field visit to Bixi Xiang 13:30 Return to Kunming 06 May (Saturday) 08:30- 13:00 Visit to Stone Forest 07 May (Sunday) Departure for home Note : The venue will be the Beef Cattle and Pasture Research Centre in Xiaoshao, about 40 krn Northeast of Kunming. The available housing, catering and jogging/walkingfacilities are very good. . 302 Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia ANNEX B. LIST OF WORKSHOP PARTICIPANTS INSnTUTIONS I. Patrick Bernard SAFRAN Asian Development Bank 6 ADB Avenue, Mandaluyong City 0401 Metro Manila, Philippines Tel: 6326325615 Fax: 632 636 2300 E-Mail: Dsafran@adb.org 2. Hector Hugo Li-PUN Resident Director ILRI p O Box 5689 Addis Ababa, Ethiopia Tel: 251-1-612505 Fax: 251-1-611892 E-Mail: h.li-oun@cl!iar.org 3. Canagasaby DEVENDRA 130AJalan Awan Jawa 582000 Kuala Lumpur Malaysia Tel: 603 783 9307 Fax: 603 783 9307 E-Mail: cdev@nc.iaring.m~ 4. Md. Abdul JABBAR Agricultural Economist ILRI p O Box 5689 Addis Ababa, Ethiopia Tel: 251-1-613215 Fax: 251-1-611892 E-Mail: m.iabbar@c2iar.org list of workshop participants 303 s Danilo PEZO ILRI P O Box 3127 MCPO 1271 Maka~i City, Philippines Tel: 632 845 0563 Fax: 632 845 0606 E-Mail: d.nezo@cgiar.org 6. Lucila Alicer LAP AR ILRI P O Box 3127 MCPO 1271 Makati City, Philippines Tel: 632 845 0563 Fax: 632 845 0606 E-Mail: 1.1aRar@cgiar.org STEERING COMMITTEE China 7. Zhang CUNGEN Professor arid Director Division of Livestock and Development Institute of Agricultural Economics Chinese Academy of Agricultural Sciences 30 Baishiqiao Road West Suburbs, Beijing 100081 Peoples Republic of China Tel: 86106891 9786/6891 9876 Fax: 861062187545 E-Mail: Zhangcg@ihw .com.cn Indonesia 8. Tjeppy SOEDJANA Director Center for Agricultural Library & Research Communication Agency for Agricultural Research & Development Jl.lr. H. Juanda No 20 Bogor16122 Indonesia Tel: 62 251 324 394 Fax: 62251326561 E-Mail: tjeRR~ds@indo.net.id Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia304 Philippines 9. Patricio S. FA YLON Executive Director Philippine Council for Agriculture, Forestry and Natural Resources Research and Development (PCARRD) Los Baftos, Laguna Philippines Tel: 63495360014 to 20 Fax: 63 49 536 E-Mail:Dsfavlon(a>.ultra. Dcard.dost. (!OV .Qh Thailand 10. Charan CHANT ALAKHANA Professor Department of Animal Science Kasetsart University Bangkok 10900 Thailand Tel: 66 2 579 8555 Fax: 66 2 579 8555 E-Mail: swkcrc(Q),nontri.ku.ac. th Vietnam II. Le Viet LY Director of International Projects National Institute of Animal Husbandry Thuy Phuonh, Tu Liem Hanoi, Vietnam Tel: 844 8344770 Phone: 844 8344775 E-Mail: NIAH@netnam.org. vn NA TIONAL COORDINATORS China 12. Zhao KAIDIAN Yunnan Beef Cattle and Pasture Research Center Xiaoshao, Kunming Yunnan 650212 Peoples Republic of China Tel: 0086871 739 1045 Fax: 0086871 7391020 E-Mail: vIDcwik(W.Dublic.km. vn.cn Indonesia 13 Andi DJAJANEGARA Coordinator Small Ruminant Production Systems Network in Asia Central Research Institute for Animal Science P O Box 221, Bogor 16002 West Java, Indonesia Tel: 62251313778 Fax: 62 251 322954 E-Mail: andidia@indo.net.id Philippines 14. Edwin C. VILLAR Director Livestock Research Division Philippine Council for Agriculture, Forestry & Natural Resources Research and Development (PCARRD) Los Bafios, Laguna Philippines Tel: 63495360014 to 20 Fax: 63 49 536 E-Mail: ecvil1ar(@,ultra.Dcarrd.dost.gov.nh Thailand 15. Metha W ANAP A T Professor Department of Animal Science Faculty of Agriculture Khon Kaen University Khon Kaen 40002 Thailand Tel: 6643 239749/244-474 Fax: 6643239749/244-474 E-Mail: metha@kku I.kku.ac.th Vietnam 16. Kieu Minh LUC Research Manager Department of Research Management & Foreign Relations Institute of Agricultural Sciences of South Vietnam 121 Nguyen Binh Khiem District 1, Ho Chi Minh City Vietnam Tel: 848829 1746/8228371 Fax: 848 8298371 E-Mail: iIIi-ias@hcm. vnn. vn Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia306 OTHER PARTICIPANTS 17. Li JIANPING Institute of Agricultural Economics Chinese Academy of Agricultural Sciences 30 Baishiqiao Road West Suburbs, Beijing 100081 People's Republic of China Tel: 861068919786/6891 9876 Fax: 861062187545 E-Mail: caasDra@;Dublica.b:.chinfo.net 18, WenJiKun Yunnan Beef Cattle and Pasture Research Center Xiaoshao, Kunming Yunnan 650212 People's Republic of China Tel: 0086871 739 1020 Fax: 0086871 7391020 E-Mail: vlDcwik@nublic.km. y!!.?n 19. Zhou ZiWei Yunnan Beef Cattle and Pasture Research Center Xiaoshao, Kunming Yunnan 650212 People's Republic of China Tel: 0086871 739 1107 Fax: 0086871 7391020 E-Mail: zhouzz72(ii),Dublic.km. vn.cn Wang AnKui Yunnan Beef Cattle and Pasture Research Center Xiaoshao, Kunming Yunnan 650212 People's Republic of China Tel: 0086871 739 1023 Fax: 0086871 7391020 Zhou ShiChun Animal & Husbandry Bureau ofNanjian County NanJian County .Yunnan. People's Republic of China Fax: 0086 872 852 2276 Tel. 0086 872 852 2276 307list of workshop participants ANNEX c QUESTIONNAIRE FOR DET AILED CHARACTERISA TION OF BENCHMARK SITES (BMS) IN CASREN COUNTRIES Improvements to the draft and resolutions to produce the final improved questionnaire were made during a Working Group Meeting in the Philippines in September 1999. The final product was the result of very useful discussions by a team comprising ofDrs. T. Soedjana from Indonesia, M. Jabbar, D. Pezo, L. Lapar and c. Devendra from ILRI, and E. Villar, E. Lanting, and M. Alo from the Philippines. The questionnaire was pre-tested in Batangas, Philippines and further refined. The following were important discussion points that were considered in producing the questionnaire. .BMS should be representative of the research domains .Data collection should be informative and relevant, without being exhaustive and wasteful of time, money and irrelevancy .The questionnaire should not be exhaustive, but needs to have focus .Activities and interactions undertaken in the BMS can be extrapolated elsewhere, in the face of diversity .The BMS should also provide both control and intervention opportunities to enable measure of change outputs and impacts .Feed and nutrition was an important intervention to increase the existing productivity gaps .Activities can improve and change in the face of external changes . Purpose of Characterisation To generate baseline information for future impact assessment studies To identify constraints in the systems practiced, as well as to define researchable issues 2. Methodology I) Step 2. Step 3. Define research domains. These are the rainfed areas (lowlands and uplands). EMS. This is the Province/Provincial level. Identify villages. It is proposed that two villages not adjacent to each other but sufficiently distanced be identified and surveyed. 309Ouestionnaire for detailed characterisation of benchmark sites (BMS) in CASREN countries Village 1. Control situation which will only be used to monitor the situation over time. No research will be done here. Village 2. Research and operational situation. All research activities will be done here, starting with detailed characterisation Undertake household survey to establish the baseline, and follow up with the PRA to get more detailed information for the formulation of specific interventions. Step 4. Research domain Rainfed area BMS Province/ Municipality level . / Village ""'.. Village I Village 2Chosen sites ::tSO households :tSO households Methods Characterisation. . Household survey PRA Monitoring Research Define research interventions In terventionslOpti ons On-fann activities Technology diffusion Adoption and impacts Figure 1 : Scheme of research activities 31 O Improving the contribution of livestock to crop-animal systems in rainfed areas in Southeast Asia Notes: I) The sample size within each village is :.!:50 households. This will vary according to the number of households, with suggested extremes of which are < 40 and > 100 households. One of the two villages will be used for research activities. 2) An important prerequisite before using the Household Survey Questionnaire is to have a dialogue/friendly broad discussion to provide an environment for the use of the questionnaire. 3) Within the Village 2 (Research), the size of households will be reduced for the interventions. 4) For the intervention a minimum of20% of Village 2 will need to be used, to be objective and within the limits of the budget. c. Devendra 30-9-99 Ouestionnaire for detailed characterisation of benchmark sites rBMS) in CASREN countries 311 ANNEX c QUESTIONNAIRE CROP-ANIMAL SYSTEMS RESEARCH IN RAINFED AREAS OF SOUTHEAST ASIA BMS Characterisation -Farm Household Survey Name of interviewer: Date interviewed: Country: Pro vin ce/Di s tri c tIM uni c i pa 1 i ty : v i llage/S ub- Distri cttr ambon : Household No.: Name of Respondent: Status in family (head, wife, son, daughter, etc.): Ethnic GroupSex Age Education: Training (Crop) Training (Livestock) Household sizel No. of members in the household Age group (years) Male Female Total >60 16-60 6-15 <6 Ilnclude all persons living permanently in the household and taking food from the same kitchen. Improving the contribution of livestock to crop.animal systems in rainfed areas in Southeast Asia312 ~"5..0"' ..:-0~..~ ~ ;0;- ~--~..~~ .:0--~'.' 0 c-;;E~~~u- .?..E"3~ ..J~"0 ii>- -- ..~ ..~ -~M ..~M ~ E .-E . ..~ ..= ~ ..= .. E E- ..E- D.. --D > --C> --C -.===- 0 === ~ ..." ." 0 ~ ." ." o iu .~ >- u :! >- ."SQ m r ~ 1'I .!! j ..=E . ..~-0,., c ~ ~ ~~ I ~ I.~ I~~ ~ ..EE.:!,"5 "5 ,"0 "0 0( 0( IL Ij 11. ..E:-6Ic[ Ij I'SI~ oo~:E.c~N....,j00C0...."Eig'irn..~ I; 314 Improving the contribution of livestock t ~ N e--."'u.["'out).0-5...c2dO"'~~~.=e1"'U.ug."'.c=t)i"'~- J ~u?s.du.?QO"'(Ij~~ ...; ~ rnr>C'-.~oSU>.U>~-~~tE,2,tIO.=rn.2=U,2,U>~=0>.-oS-SrnU.GUQ.rn>.~.grn.;~UQ.Q..9rn=0>.U~:I: I~ c-."'~.u8."'.s0.~s?.~.u~'Q....s=.9;~~...0.=e]"'~.u~0."'-5~~...- I ~ ~u~u~C'-.uU\.ao- ~ M .~ ~ (,) u u MQ. tI} [O'QU\]~OU\u.u ~I ~ "' .d ~ (,)~Ir\ systems in ,ainfed ~y-.0y..~Q,.c-...~.c~c..c.~-.0y=~e~ 0-;;lEmui{) Pouu.c"1 00:0:'iUo{')~~~~~ uu~ ~0;;~...uum ~";Q= "v;'~~oE'-'=,.2I'?"i! Q,~...I:'iUuCf.< c I .g.E~00].5"'e.oc.t:00"'I:::o'-0~I -.-gu:;'-"-C()u~~uI~ ---"'~oE'-'u~co~'iu~ ~-5?e"'.."'v~'-']v~"'-;;E.~...Ov0!.< """'toO~'-'-gv~E"'~E.~...0~...~ """"'I I""" < "' . II - Ic I 8.]o =' , .~ -CII '-" ~ ~ .c "' "' =' Ir B "' "' =' >, 0 "' .c 01)C O .- ~]1!- C!) I C'-.>,.c~i"':1Oc<:= sites (OMS) in N~ZII"'uC1"'8'00=.u~>~E.=1< 0'-.>,~Oc= N~Z-!!,U>-'-'UUC~-0i~.a< C'-'>,~Oc~ ---N~zi"'u>-'-'C'-.-0uEo~-0].21< C'-.>..c~0=:: Feeds and Feeding Systems (Cattle-Buffalo-Goats-Sheep) Feeding system Type of basal feedSpecie! Source Adrq.. ( A, I ) QualIty"Freq. Supplements I Type I k2/bd/d I Cattle Bun'alo I Goat Sheep Wet SeB50n Cattle Buffalo Goal Sheep Feeding system: I.. tethering; 2 -free gl11zing; 3 eut and carry Types or bllll reed: I "" grllsses; 2 -emp residues; 3 "",ree furuges; 4 legumes; 5 non-conventional feeds; 6 -- oth= (specify) Souree: I -own pasture/gl11zing area; 2 -contour hedgerows; 3 -crop residues from own fann; 4 -crop residues from other fllnns; 5 -neighbor's pa~ture~; 6- community pastllres/road~ide grazing; 7 -cooper.ltive store; 8 -market place/factory Frequency; I "" twice a duy, 2. once a day, 3 every other day, 4 oncc a week, 5 -occa!;ionally .Adequlcy: 1- Adequate; 2 --Inadeqllale bQUlIIty: 1 Good; 2 Average; 3- Poor Improving the contribution of livestock to crop.animal systems in rainfed areas in Southeast Asia316 Feeds and Feedin! Systems (Pi2S -Chicken --Du~ks) J)rv Season Feeding system Type of feed Source Freq. . Adeq. A.I Qualityb Supplements k2/hd/d Pig Chicken Ducks Wet Se~n Pig Chicken Feeding system: I -scavenging; 2 -pen-fed Types of feed: 1 -kitchen refusals: 2- grains (e.g., broken rice, corn, sorghum); 3 -f:-ommercial feeds (e.g., cassava chips, rice bran, soybean meal, fish meal, commercial mixtures) ; 4- others (specify) Source: I home supply; 2 markct placc/factory: 3 -coopemtive store: 4 -home mixed; 5 -others (spccily) .Adequacy: 1- Adequate; 2 -Inadequate hQua1lty: 1- Good: 2- Average: 3- Poor Q)u~ .1-- 318 hh,-.~.' '3 the contribution of livestock to crop-animal systems in rainfed areas in !'c -...~~"'==.->~...c.~.:...~=.-...='C'-==.-...c.e="'==u'C'=~~=.-...~...~e==-.cu='C'=...c.~~~'C'=~...~~~ QJ~.w~u Q-~== Q.~~.crlJ 01..~~ =~~CJ.-.=u ~~==~E~~... -;=(..? -=oiCj=E=.~~ a...~ 'tj~~--&~~I~e.~(0.-0""~"8z - ~bO 1 II Q) -Q) !~ ? c. l~ o Q) o 8.~ ~<~ ~ "'8ElOtI.s >,~'+-0~Q)"6i ~ ]~c=o0..Q)-0E:IZ ~-6~.Q...v"8oz "Q-0ri)~"8:Jz v>:3.00.~.~ -0.-~0-,,'outheast Asia ,-.cQ~=I..~Q,Q"'~I..=~~"'=Q.>~Q,cQ"0~"'=,Q'-'cQ~Q,6="'cQu"0c=~c..:~I..=6cQ~u="0QI..Q,>.I...;Q I~I IcII I~I I~I~~e~4)?I~ > 5 '; '; 0 ~ o <~O.O.I-- 1-- of benchmark C'-. ~ .'1 u=' -0 80. .s ~...1tU .d ~II0. - ti~-0eQ,-0~"'gl...,Q,~ I ]~~ ~=~ ~Q..-~~ ~ c~u-u .- ~ e -.J::~ ri) e0 ... ~'-a~e~.aou~1~0"~~ CJ.- ~ =~ ~u .- .c u ~u = Q --i-.~:u~~.c...~e"'~"'~c.=.c~--"10'-'~"'~e~u~.cu~.Q~-~u.-'0=.-~~~~~'-'~-=~=.e=T=====:u='0="'c."'~~...=c..5"'=.c~~ ..,..=(.I = = ~ ~ = g.~~ ... ~.c 1--~-~o.c~=e~~ '-!')0'-'~I =.c=-flJ==.=U~~ ,-.~'-'-0.c"~'Cf= CLI:?8I~ ~.c-501C'a.uU'0 "'1u.,setcI.~iu I.?-I ~:0Q)~O..:, "'.9UU"'> c 00 c .~ > .-~ .. "O~...~;=~~Q~~~ lCI~] ?.~~]do.~~.(3~> I.iI~~"'gj1.= 'Q)iI~I Q)~Q)c.~~~ -.~..u~Q."''-'~~.cQIIIrJ;i.-;~ = .-~"' .. "'"C~-u .- Q.CII U ..,. II ;;:..c==U= ~ Q~e ~ II e ~ ~ = ~'*' Q ~ ~ ~-~~ = ~ ~"Ccn~<~lIgN.~II < = ~~ ~ ~ ~ "C ~ = .!!.~~ ~ ..-~~=cnu"C =<~~ II ~ [~"~.c ~ ~=~"'==.->~~=-~-=~~=.-~="0t=~=~=.--=~~ .Q zI~ c.-.~E(/}- WIu..~Q,WI.=WI~ ~==-III.9 =I.~ - ...0:=...Q.= Q ~ e ..- ~ ~";.0,-...c.c = O Q ...e,,'-...-0I .c~ -i-.~>.~.wu~~-=~~e~ct~Q.Q-=~~~~e~u~-=u~,Q~.w~u.-'Q.5~~~~6:-~=.~e~~-=~=j,..s~=Q..5~Q,Q=~ >..~~;"~ >....=~=C"~..t.:. .. ~ ~:.. "" ;< I~ ~~~[ I~ :2:2u ~ I~u ~~ :E:2u ~"-6..t!~ ~='E~"'~ "'~c~~c 01) :E .5 S ] .~ 1.1. ~ e:o.5~u~:I: OQc.="~E~:.?"Q)C/) ;,;~-5a -c.y~Q."'-"'-~-==II\C..r~=="'~~"'IIIr) ~ ';=="'~~"'II~>.:s==8II"'I>.~~~~N:i-.;~II;.:.u=~=Q"~-~ 324 Improving the contribution of livestock to crop-anlinal systems in rainfed areas in Southeast Asia ~Of)=.-'C-Q..c'"0=(U~ .1 "a-cGI ,~ = = Gl t t ;c -~ s ~ ='0..= ..~ ~ e IJ "' =~.:c~ s ~ c .--=0 l oa e-. 0. = ;..-Q I Lo 0. Q ~ ~ ~ Q .&1 Lo o QJ Qi . u Q \0 = =-00 "0 ~ c ~ ~ .. -J = ] ~ ..=...J c.:= "' ~EQJ = c...:.- "' ,.0 ...c -41.! e ..41= ~ c c 41 = E-o ....~ 6.E'c.c.Q = ...~af) ~1::Q...~(.I=6.=?'t:~"S~ ~INI~I~I~I~I~I~I~I~ ,..:;.'T:jQ)~0"-"E-o~UII~ -~ ~~ ~ ~ ~ Q)~u..c~~O ~ ~ II .-.c r--- ~u.~ ~ t:I) 0 ~ .~ o~U=~.o~ Q) ~-.d~II~~\O9~.,r\O~C). ~.- 8E~u 0 II -~M~ ., ~ .-"'-Q~'1:j~UO~~..cQ) II Q) ~ ~ II~~Ir) ~ ='- ~ -d'o Q) ~~~ ~ ~ U U 0 .c ~ 0 ~ 5 ~ ~ o~..c '1:j II II ~ ~ -.:t~~ - ~ ~ ~ ~ ] .S = .3 '"tn]~ II ~~~ M ~ ~ u ~ ~ u X ~ 0 ~ .-Q) II II ~II -g Mr"i 0 U ~ ~N ~ ~~ ~ -~~ Q) .-t:I) 0 ,.g ] .S 9 o ~ ~ C).f'" ~QC). "O-M U Q) 0 Q ., ~ ~~ ~ ~ ~Q) ~II u ~~~ IIN~ o~~ N-d'.c ~U~ ~ ~ r/) II > 0 g.~! ~tr~ ~ O ~N..c U ;; Q " t:I) -II Q ~ ~ .- ~ 0 - "' C). ~ i,j8 ~~II ? =' u II II N <~~ --0 II "rn ~ ] >. ..~ -C).t::t:I)..cCIl eo ~ C). := - Q).C- t:I)~.;; .. ~~eo~~~Q=' ~ .t: 0 t:I).~ 0 -"' =' 0 -.- =QQ~C).:=~eo~ull 008 ~Of-Ir)f-r/)~ 'C'~~W}==.>CI-Q,CI.d~=.--=~-W}CI=~.-W}CI-Q,e(j~=~CI-==~e=CIW}~ I~ I~ "; I.. 1.5 = ~~ ~iII", ~ - O t:cud,t:,iI 4)~u 3~ 0(~ ~~c I~IIi ! '<"IrIi4)>- I-d~~"0U~~~ri)- ...QN:;: ~ ~ ..~ ~UQ0;= ~Q ~.E~O =~ ...'C "' "' "' 0 .-"' tUtUtUl,.:; fl)tU ~'0~~ ~-o ..Q Q Q Q :E ~ Q ~ "' "' "' "' 0 "' Q::>::>::>::>C/)==:> ,0 ,0 ,0 ~ ,0 U U ~ -0' 0' 0' 0 0' 0 I~ ~ut;:.5~~'5{/}9?~.?='e{/}="B::)~ 1- (Q'C4)~ Ib(,5 "';:Io.r:"'~]"':::>~" I]I~ !/)Q)!/)0~Q...-5o...tE]:.:J~ ~0rJ)~ systems in rainfed areas in Southeast Asia = ~...=.~=a.....~=Qu=Q,Q=~=Q.~..Q.GJua..GJQ.a..GJea..=~ ~"'"~~-C'.~~Q"'~.0t'.:c"C,I~~.:"'c.~~C,I~c""'-;e.=".5~~c=~C,Ic~=~"'e~:c~~c.~~""Gi~c~-=C,I=~~~c.~~"et~--.: . -"'~ ~ .: c~ ~ Q. ~ "' " ~~ ,,~~ e of benchmark sites ISMS) in CASREN countries 327