.. ^ -_ -. . -_ .
, -' ..- ._-- .
:''. -: -*--**.^ "4 ^S
~E~~ ~ -.-~
,w --W " :. :
-- -- .4. ~
- r *- 44
c ...r. -,i -r
-~ ....-.. I
L -. '-: : -
~~'A WWI4. ~ t
d .0, 4
-ti.m J" ; A;
A five year report of the Small Ruminant
Collaborative Research Support Program.
Edited by R.D. Blond
Development Design Associates
A Joint Venture Between
the United States Agency for
Agricultural Research Institutes
in the United States and Overseas
Design by Jan Conroy, production by Repro Graphics, University of California, Davis, California 95616, USA.
Printed by UC Printing Department, Berkeley, California.
In accordance with applicable Federal laws and University policy, the University of California does not discriminate in any of its policies, procedures, or
practices on the basis of race, color, national origin, sex, sexual orientation, age, or handicap. Inquiries regarding the University's equal opportunity
policies may be directed to the Vice Chancellor of Academic Affairs-Affirmative Action Officer and Title IX Coordinator, 521 Mrak Hall, (916)
752-2070. Speech and hearing impaired persons may dial 752-6TTY for assistance.
TThe origins of the Small Ruminant Collaborative Research Support Program (SR-CRSP) lie in the
passage by Congress in 1975 of the Famine Prevention and Freedom from Hunger Act. The provisions
of Title XII in the Act were to:
". .. strengthen the capabilities of US land grant... universities in program-related agricultural
institution development and research .. improve their participation in the US government's
international efforts to apply more effective agricultural sciences to the goal of increasing world
food production and in general should supply increased and longer term support to the
application of science to solving food and nutrition problems of the developing countries."
In this simple but eloquent language, the key elements that have been instrumental in molding the design,
goals, and administration of the SR-CRSP were finally established. Specifically:
* To strengthen the capacities of US universities as opposed to conducting research exclusively overseas.
* To conduct institutional development and research as opposed to direct handouts in the form of
* To supply increasedand long-term support as opposed to small, short-term, stop-and-go projects that have
characterized so much of the past technical assistance to developing countries.
Implementation of these high ideals could have been attempted in any number of ways. One that was chosen
was the Collaborative Research Support Program (CRSP) mode, whereby several US institutions
representing several disciplines would be engaged in work at several overseas locations. The Small Ruminant
CRSP was the first CRSP implemented and has to a degree pioneered this new technical assistance mode. The
focus of the work has been limited to research, training, and institution building which are the areas of
endeavor in which universities excel. With no precedent to follow or template in place, it is neither surprising
that the design, structure, and administration of the SR-CRSP were the subject of considerable debate at its
inception, nor that the SR-CRSP continues to be a dynamic program of flexible design. Also, because an
inherent aspect of the new program concept was to allow universities and host countries to design their own
programs collaboratively, the technical work, especially overseas, has actually been in progress for only three
and one half years in most cases, and even less in others. The first eighteen months were devoted almost
exclusively to program development which involved such tasks as identifying overseas worksites, counterpart
institutions, and host country scientists with whom to develop collaborative workplans.
In retrospect, it is likely that one of the hidden accomplishments of the SR-CRSP which has resulted in a
continuing spirit of mutual cooperation is the time that was taken in the beginning to develop this program in a
collaborative mode, involving USAID, US institutions, and host country agencies as equal partners in this
Currently, there are seven CRSPs in place, all with a superficially different form and structure and yet
possessing some common features. All have several US participating institutions and many overseas locations.
All have a management entity, technical committee, a board, and an external evaluation panel to provide
constant quality control. All are multidisciplinary in their approach and all are joint ventures between USAID
(the primary source of funds), US institutions (with their matching contributions), and the overseas host
There is no doubt that for every CRSP, a hundred possible designs could be proffered. None would be
perfect, but all might succeed equally well. The essential element in the inter-disciplinary program is a balance
between the necessary expression of each scientist's individualism and the collective need for integration. In
this respect, CRSPs, like all other social structures, possess the universal characteristics of holarchic order
which, as Koestler asserts, "... is possessed of two opposite tendencies or potentials: an integrative tendency to
function as parts of a larger whole, and a self-assertive tendency to preserve individual autonomy." This
applies within the CRSP itself and to the CRSP as part of the larger body of researchers worldwide. An
example within the SR-CRSP will suffice to illustrate the point. There are a number of principal
investigators/scientists at various institutions across the US who work in collaboration with scientists/
colleagues overseas. It is imperative that each scientist operate with a degree of "self-assertiveness" or
independence so that they may excel in their specialty. However, it is also necessary that the "integrative
tendency" be strongly felt so that scientists can function as "part of a larger whole," to solve problems with a
wider scope than encompassed by their own subject area. These two forces are opposite, and assertion of too
much of one leads to as many problems as an excess of the other. At times, the SR-CRSP has heard strident
voices from within calling for more independence and from without the voices call for more integration. What,
in fact, we have strived to achieve is a balance between the two in order to preserve the atmosphere of"creative
tension" in which critical inquiry flourishes. The reader may judge whether the accomplishments over such a
short span of time represent acceptable progress.
Progress toward what?-the aggrandizement of a handful of US or overseas scientists through the
publication of scientific papers? Certainly these are by-products of the quest for knowledge, but they are not
the primary product of the SR-CRSP. The primary goals are institutions that will endure, trained scientists
capable of solving problems as they arise at the local level, and new knowledge that will lead to improvement in
the welfare of those with limited resources.
Research is a very long-term investment. It gives the impression of being costly but experience has shown
that the returns on capital invested are very high indeed. Research in the "international" sphere is doubly
expensive and requires even more patience before benefits are realized than that required of domestic research
programs. But the pay-off, particularly in human terms, is also very high. Not to engage in the adaptive
research that allows technological advances to benefit the less-developed countries with their burgeoning
populations and unprecedented food supply problems, is to condemn millions to the ranks of the
malnourished. Some will argue that research, both basic and applied, is a long way removed from the
immediate needs of the smallholders. Not so. Millions of smallholders and their families are now benefiting
from basic research that was considered exotic ten and twenty years ago, but is now recognized as the
foundation of the "green revolution."
These successes were the outcome of a long period of sustained research that depended on well-established
institutions in wealthier nations and on the long-term endeavors of the international research centers overseas.
The successes have only highlighted the pressing need for effective agricultural research systems in the
developing countries, so that new technology can be adapted for local use. Few would question the view that to
build such research capability requires, inherently, a long-term effort. This is a view which is embodied in the
language of Title XII and the SR-CRSP Grant and which recognizes the critical importance of the long-term
The outward manifestations of our efforts to accomplish these program goals have been an enormous
emphasis on institution building, especially training of overseas colleagues at all levels, and of continued
support on their return. Also, we have made a heavy investment in high calibre scientists placed overseas who
have worked with an emphasis on field research in close proximity to the smallholder. It has been a
cooperative effort and testimony to the willingness of the Agency, the participating institutions, and the host
countries to work in a spirit of close collaboration as equal partners. From every section of USAID, be it the
Contracting Office, Science and Technology, the Regional Bureaus, and particularly the Overseas Missions,
the SR-CRSP has received outstanding cooperation. From each of the US participating institutions, be it
from Board Members, Principal Investigators, or co-workers, the SR-CRSP has received commitment of the
highest calibre. And from the host countries and their collaborating institutions, be it from ministerial offices,
university departments, research agencies, or the remote experiment station, the program has been sustained
by enthusiasm, the provision of facilities, and a dedicated staff.
This volume appears at the end of the first five years of USAID Grant No. AID/DSAN/XII/G-0049 and
summarizes work carried out overseas and in the United States under the aegis of the Small Ruminant
Collaborative Research Support Program. It represents the report of the University of California, Davis, the
Grantee, to USAID. Research aspects of this report closely follow the five year reports of Principal
Investigators which have been edited for this publication.
We will not single out any individual or institution for special acknowledgement. Any attempt to do so
would fall short of a just recognition of the efforts of so many. The program discussed in this volume is
dedicated to the fulfillment of the hopes embodied in Title XII. It is our hope that the SR-CRSP may continue
to strive for improvement in the welfare of the poor, and that it may do so in a spirit of mutual cooperation.
David W. Robinson
University of California, Davis
Contributions by the Principal Investigators and Site Coordinators from whose reports this volume was
developed are gratefully acknowledged. Also acknowledged are those who provided photographs from which
to illustrate the Report. The assistance given by the staff of the Management Entity in the preparation is also
most gratefully acknowledged: David Robinson and William Weir who have managed this SR-CRSP over
the past five years and whose initiatives set the stage for this report; to Richard Waters and Marcella Pieratt
who provided fiscal data; and to Michele Lipner who undertook managing the production of this report as
well as providing throughout, invaluable editorial assistance.
TABLE OF CONTENTS
Foreword ..................................................................................... iii
Editor's Acknowledgements ........................................................................ .v
EXECUTIVE SUMMARY .................................. ................... ................. ix
PROGRAM BACKGROUND ........................................................................
Economic Analysis of Small Ruminant Production and Marketing Systems. A.J. De Boer .................... 19
Improving Reproductive Performance of Small Ruminants. W.C. Foote ................................. 25
Goat and Sheep Nutrition and Feeding Systems Research. W.L. Johnson ................................. 30
Range Research for Increasing Small Ruminant Production. J.C. Malechek ............................... 37
Sociological Analysis of Small Ruminant Production Systems. M.F. Nolan ................................. 44
Small Ruminant Flock/Herd Health Program in Smallholder Systems. H.J. Olander ...................... 47
Breeding and Management of Sheep and Goats. J.M. Shelton .......................................... 50
Introduction .................................................................. .......... 56
Genetic Improvement of Sheep and Goats for Smallholder Production. G.E. Bradford ...................... 58
Economic Analysis of Improved Small Ruminant Production Systems. A.J. De Boer ....................... 63
Nutrition and Feed Systems Research. W.L. Johnson .................................................70
Sociological Analysis of Small Ruminant Production Systems. M.F. Nolan ............................... 77
A Small Ruminant Production Model on a Programmable Calculator. N. Thomas ......................... 82
Introduction ............ ..................................... ................................. 85
Genetic Improvement of Sheep and Goats for Smallholder Production Systems. G.E. Bradford ................ 87
Development and Breeding of Genetically Improved Goats. T.C. Cartwright ............................... 89
Systems Analysis and Synthesis of Livestock Herds. T.C. Cartwright .................................. ... 92
Economic Issues in a Dual-Purpose Goat Production System. A.J. De Boer ............................... 96
Dual-Purpose Goat Production Systems for Smallholder Agriculturalists. H.A. Fitzhugh.................... 101
Animal Health Constraints on Small Ruminant Performance. T.C. McGuire .............................. 113
Sociological Analysis of Small Ruminant Production Systems. M.F. Nolan .............................. 117
Introduction ....................... ..................... ............................ .......... 121
Prolificacy and Productivity of Moroccan Breeds of Sheep and Their Crosses. G.E. Bradford ................. 123
Nutrition and Feeding Systems Research. W.L. Johnson ............................................. 126
Sociological Analysis of Small Ruminant Production Systems. M.F. Nolan ............................ 129
Rangeland Research for Increasing Small Ruminant Production. J.T. O'Rourke ............................ 131
Introduction ..................................................................... .........137
Evaluation and Genetic Improvement of Sheep and Goats
in Extensive Management Systems. R.L. Blackwell ................................... ............. 139
Improving Small Ruminant Nutrition, Management, and Production Through
Management of Native Range and Improved Forages. F.C. Bryant ........................ .......... 144
Economic Aspects to Increased Productivity of Small Ruminants. A.J. De Boer ............................ 152
An Investigation of Small Ruminant Health Problems. J.C. DeMartini ................................... 158
Improving Reproductive Performance of Small Ruminants. W.C. Foote .................................. 164
Sociological Analysis of Small Ruminant Production Systems. M.F. Nolan .............................. 168
Goat Research Project in Northern Peru. B. Quijandria ............................................... 173
BASIC SUPPORTING RESEARCH
Introduction ............................................................................. 180
Breeding Research. R.L. Blackwell ............................................................... 181
Breeding Research. G.E. Bradford ................................................................ 183
Range Management Research. F.C. Bryant................... ................. ................... ... 185
Systems Analysis Research. T.C. Cartwright ............................................................ 187
Agricultural Economics Research. A.J. De Boer .................................................... 192
Animal Health Research. J.C. DeMartini ..........................................................194
Production Systems Research. H.A. Fitzhugh...................................................... 196
Reproduction Research. W.C. Foote .......................................................... 197
Small Ruminant Nutrition Research. W.L. Johnson..................................................200
Animal Health Research. T.C. McGuire .................. ................................... .... 206
Range Management Research. J.C. Malechek...................................................... 208
Animal Health Research. H.J. Olander ......................................................... 210
Breeding and Management Research. J.M. Shelton ..................................................211
Training Accomplishments .................................................................... 217
Organization Strengthening ........................ ..........................................219
PROJECT DEVELOPMENT AND PUBLIC SERVICE ....................................................220
Summary of Program Expenditures from USAID Grant Funds .......................... 223
Summary of Program Expenditures from Matching Funds and Contributions ............... 223
Summary of Sub-Grant Program Expenditures of USAID Grant Funds by Category .........224
Summary of Centrally Administered Funds .......................................... 224
Expenditures of Centrally Administered Funds by Function ............................. 225
Program Administration Personnel
Board of Institutional Representatives ................ ............................... 226
Technical Committee ........................................................... 227
External Evaluation Panel ..................................................... 227
Current AID Washington Staff ....................................................227
Host Country Counterparts ....................................................... 228
Overseas Site Coordinators ...................................................... 228
Management Entity Office and Overseas Support Staff ................................. 228
SR-CRSP Research Participants
Indonesia ..................................................................... 232
Kenya ................. ..................................................... 234
Morocco ................. .................................................. 236
Peru ......................................................................... 238
Basic Supporting Research ............................... .................... 242
Overseas Institutions Participating in On-Site SR-CRSP Activities ........................ 247
Chronology of Important SR-CRSP Events..........................................250
Acronyms ................. .................................................... 251
T he Small Ruminant Collaborative Research Support
Program (SR-CRSP) is a joint effort of the US Agency
for International Dexelopment. US universities, and go\-
ernment agencies in fi e developing countries. It is supported
b, a Title XII grant of $15 million dollars over a fixe-year
period and by contributions of o\er 30. of this amount from
participating institutions. The program is administered as a
grant to the Uni ersity of California. Dax is. which, through a
management entity, provides subgrants to US participants.
and maintains fiscal accountability .
Currently the program is engaged in research and training
in fi\e countries: Brazil. Indonesia. Kenya, Morocco, and
Peru. Supporting research of a more basic or technically
complex nature is carried out in the US.
US institutions responsible for the following projects are
presently participating in the program:
lnil\ersity of California.
Texas A&M Unixersity
Texas Tech University
Utah State UnixersitN
Colorado State Uni\ersity
North Carolina State
Montana State tlniversit\
University of Missouri
Forages and Nutrition
Production and Feed
The range of projects deployed in each overseas site x aries
trom eight in Peru to four in Morocco and is based on the
overall research oblecti\es set for each site, the major
constraints to improved production, and the needs of host
country institutions for support.
The SR-CRSP research programs are all multi-dis-
ciplinar\: arming degrees of integration between disciplines
reflect, in large measure, the demands of the program in each
At its inception, the SR-CRSP undertook the planning
and development of its ow\n research and training agendas,
and selection of overseas \orksites and collaborators-a
task which took o\er a year to complete. The longest running
country research program has been in place for only three
and one half years. Funding is available for an additional
program year. through September 1984. and is contemplated
for an additional three years. This \will enable training and
overseas research to generate sufficient progress and
momentum such that further research, after the conclusion
of the SR-CRSP. can become sell-sustaining under the
auspices of host country governments.
The semi-arid interior of Northeast Brazil. the sertao, is an
area encompassing about one million square kilometers and
inhabited b. over 12 million people, most of whom rely on
smallholding agriculture for subsistence. The area is plagued
by the unreliability of its rainfall, which together with the
long hot summer severely restricts plant growth and crop
}ields. As a survival strategy. farmers practice mixed crop-
li\estock farming, raising sheep and goats which depend
principally on the native bush, the aatwinga. for forage.
Although 90. of the nation' goats and 30'- of its sheep are
raised in the region, their productivity is extremely low,
averaging only one fifth of that found in other areas of Brazil.
The SR-CRSP. in collaboration %\ith EMBRAPA and
the National Center for Goat Research (CNPC). initiated a
research effort in early 1980 to improve the productivity of
small ruminants in the screao, with the prospect of raising the
incomes and standards of living of the area's limited resource
The approach adopted bi the SR-CRSP in Brazil has
been to field teams of scientists to conduct research with
Brazilian collaborators in the major areas constraining small
ruminant production \while Brazilian scientists undergo
training overseas in a wide variety of subject matter areas.
Research projects in Animal Breeding and Management.
Reproductive Physiolog,. Animal Health. Forage and
Nutrition. Range Management. Economics and Sociolog
are currently in place. Each project pursues its research
objectives independently but cooperates closely w here issues
necessitate an interdisciplinary approach.
The Range Alanagement Proie r, recognizing the oxer-
whelming constraint of dry-season feed deficits, has focused
research on achieving a better balance between \wet-season
surplus of feed and dry-season shortages. while at the same
time tring to minimize the impact ol recurrent drought on
land and vegetation Accomplishments include a satellite
sunres of local range resources, studies of preferred diets of
sheep and goats in the caaiinra. and development of
techniques for manipulation of caatinga vegetation to
prolong the critical availability of green forage into the dry
The Nutrton adiil Feeding Systemns Project is also
tackling the problem of seasonal scarcity of herbage available
from the native range. The project is emphasizing research on
the use of crop residues and preserved forages as feeds to
supplement range resources when these are deficient at the
most critical stages of the animal production cycle: post-
weaning. breeding, late gestation, and lactation.
The .-nimal Health Protect is investigating methods of
economically controlling the major diseases that were found
to limit small ruminant product\ it in the area: intestinal
parasitism. caseous lymphadenitis. pneumonia, and mastitis.
Research is directed at disease control by improved animal
management practices rather than bi costly accination or
The Manageumen and Breeding Project has addressed the
question of characterizing existing breeds and tv pes of sheep
and goats in the sertao, \ith a view' to choosing stock for
initiating selection and breeding work. The intention here is
to develop a breed or tipe \with a genetic potential
appropriate to the biological en% ironment and management
The Reproducition Prolet ha.- oriented its research
towards improving the overall productivity of flocks and
herds by developing procedures to relax constraints on
reproductive performance. Management practices which
ha\e been recommended for impro ing reproductive output
include optimal age for \weaning. optimal condition of the
female for first pregnancy. restriction of the breeding season,
reduction of the parturition internal, and a technique for
diagnosis of pregnancy\ to facilitate special arrangements for
the gravid female.
The Economics Project has developed a \ hole farm linear
programming model based on descriptive suneys of farms in
Ceara State. The model pro\ ides a means for testing the
economic impact of technical interventions in sheep or goat
production systems proposed by other SR-CRSP projects.
as well as revealing how\ production responds to market
The Rural Sociology Proiect has pursued goals closely
related but largelN complementary to the Economics Project.
Developing a comprehensive nie\w to total agricultural
production s stems. the project has sought to understand the
production decision Iramework of farmers with special
reference to the decision rules governing small ruminant
production. In addition, the project is providing a service to
the biological projects by e\ aluating the likely social impact
of various ne\\ technologies and providing criteria to guide
development of new techniques w which have the most
likelihood of adoption.
SR-CRSP research in Indonesia, developed in collabor-
ation with the government's Agency for Agricultural
Research and Development. is directed towards improve-
ments in sheep and goat production in two distinct
agroecologies: coastal lowland and interior highlands. In
both areas w here crop production is intensive. products ity
of small ruminants on smallholdings is low. The lour SR-
CRSP projects and their counterparts have worked cooper-
atiiel\ to alles late some of these problems here they most
matter--at the villagee le\el.
The Breediing Proitec has focused on developing from the
nati e thin-tailed sheep. a type \ hose prolificacy is less erratic
and matches its production ens ironment more closely. This
will reduce neonatal losses and improve flock performance.
A major cause of lo\w reproductive performance has been
identified and remedial steps taken to change management
The Nutrition and Feed Syilems Proiect in its sun ey of
current small ruminant nutrition in \illages, had identified
digestible energy, not protein, as the limiting factor. Feeding
trials are in progress to determine sheep and goat responses
to various feed supplements available to villagers.
The Etonomhic Projcel has participated with the other
SR-CRSP projects in the village sure s. to de elop a picture
of the economic roles and importance of small ruminants.
Research has also been carried out to determine mechanisms
of price formation and marketing, and on analyzing
product ity constraints.
The Sociologt Project. after completion of the initial
survey, has undertaken studies to understand the farmers'
decision framework as it affects small ruminants. The nature
and extent of farmers' contacts \with external institutions
such as credit, extension, and markets, has also been studied
and will he important in designing an acceptable impro\e-
ment package for smallholders.
The work of the SR-CRSP in Kenya is unique among the
Program's five worksites. In the intensive farming region of
Western Kenya, an attempt is being made to develop an
entire production technology package for meat and milk
goats for a situation in which dual-purpose goats are not
traditionallI raised. Working towards this goal has de-
manded increasing integration between component projects,
and training Kenyan scientists in interdisciplinary ap-
proaches as well as in traditional subject areas. This work is
conducted in collaboration with the Ministry of Livestock
Successful development of such an economically viable
technology package depends on its ability to fit the existing
farming system and cause minimum dislocations. The
understanding necessary to achieve this is the outcome of a
comprehensive survey of economic aspects, cultural norms,
and farm resources, conducted jointly by the Economics,
Sociology, and Production Systems Projects.
The Economics Project has completed a linear pro-
gramming model of the target farm system, which optimizes
economic benefits under different patterns of resource use
subject to biological and social constraints specified by other
The Sociology Project. after the completion ofthe suree,
has focused on questions of division of labor and labor
availability for animal care and management on socio-
cultural constraints to goat production and on the magnitude
and types of external services and supports that would be
needed by farmers adopting a dual-purpose goat production
The Production Systems Project is charged with devel-
oping the technology packages from components supplied
by the other projects, and conducting farm-trials necessary to
validate new production systems. Pending completion of
research on package components, the project has focused on
the survey and on pursuing research on topics not otherwise
covered, such as acceptability of dairy goat products. This
project is also responsible for feed production research
focused on identification of feed production strategies to
meet the needs of a dual-purpose goat production system.
The fHealh Prolect completed a survey to identify the
most important disease problems likely to face dairy goats in
the target area, and is pursuing research on methods to
control parasites and contagious caprine pleuropneumonia.
The project's work on caprine encephalitis arthritis (C.-\EI
enabled the presence and extent of this important disease to
he established and eradicated from Kenya.
The Breecdin Proieu is seeking to develop. through
genetic selection and cross-breeding, a dual-purpose goat
appropriate to the conditions in Western Kenya. The
performance of existing cross-breeds in Kenya is being
evaluated under target area conditions, and attempts are
being made to develop a suitable cross using a hard> local
breed and semen from exotic dairm breeds.
The Si Ycnt .-Anal'sis Proiet has developed a mathe-
matical model which simulates the performance of sheep and
goats in Kenya and shows the influence of factors such as
nutrition, disease stress, and management on product\ ity.
Small ruminant production has been a traditional activity\
on the extenstie rangelands ol the Middle and High Atlas
regions of Morocco, and sheep and goat products are in high
demand. In recent decades, the equilibrium established
through traditional range resource use has been upset by an
increased demand for small ruminant products from an
expanding human population, and by reduction in the range
pasture by encroachment of \\heat cultivation. The SR-
CRSP is collaborating \ith the Institut Agronomique et
Veterinaire. Hassan II in research to solve the problem of
improved small ruminant production and stabilization of
range resources and in training Moroccan staff to continue
The Rane lae Mnaguetn Proic(i is investigating tio ti pes
of range resources. sagebrush-grasslands and oak w\ood-
lands, to determine feasible range management practices that
\till provide for the needs of livestock producers and at the
same time protect and develop the producti\its of the
The Foratc andm ,Ntrttioin Proiect has confronted the
problem of seasonal forage deficits and is looking lor \wa s in
which crop residues can be used as feed for small ruminants.
either to supplement range resources or as a preserved feed
for use in \\inter months.
The Breedint, Proitci is investigating the desirability and
methods for adjusting the genetic potential of traditional
sheep breeds \which have outstanding hardiness but lo\w
prolificacy. To make full use of feed supplies from irrigated
agriculture and the shift from a transhumant to semi-
sedentar> production s stem. a different type of sheep needs
to be developed with higher reproductive performance.
The Rural So( olog i Proici is pursuing research to ad ise
the biological projects on the social and organizational
aspects of production. OI particular interest, in each of the
different ecologies, are the producers' decision frameworks
and the rules that go\crn use of communal grazing lands.
The major research locus of the SR-CRSP in Peru is
improvement of sheep and alpaca productivity in Andean
peasant communities. With conditions rarely suitable lor
adequate crop production, more than 350.0i10 families are
dependent on small ruminants for blood and income.
Unfortunatel.. product ity is not high and pressure from an
expanding population threatens degradation of the range.
The research is conducted in collaboration with INIPA
and several universities including La Molina. IVITA.Cuico.
Puno, Arequipa. and Lambaeque.
The Soctology ProIelI has devoted its resources to
studying the social dimensions of sheep and alpaca
production in agro-pastoral communities in the Andes and
characterizing the goat production systems in Northwest
Peru. These understandings help in guiding the directions of
biological research so that proposed interentions are
The Ec.ononucs Proictc has focused its ellorts on
descriptive aspects of small ruminant production s\stens in
Peru, on the mechanisms of price formulation of small
ruminant products, and on the models needed to carry out
bioeconomic analysis of proposed improvement intenen-
The Raniee .Manaienenti amn Foraw,. Proie t has in' esti-
gated ways to mitigate effects of seasonal forage deficits on
livestock productivity\ and to develop grazing systems to
optimize animal production and maintenance of range
resources. Stocking rate guidelines have been established and
research is continuing on impro ed forages, determination of
seasonal nutrient delfiiencies, and supplementation sched-
The Animal Heahul Proiec research priorities ere ba.ed
on an assessment of the importance of particular diseases and
the current lack of eflecti\e control methods. Continuing
investigations include: developing control measures for
chronic respirator\ disease in sheep and neonatal enteritis in
alpaca; disease-related infertilit\ in sheep: and heavy \ metal
toxicit\ in sheep liv ing near mining and smelling operations.
The Breeding Pro/ca has completed preliminarN studies
directed at identilfing native or improved breeds ot sheep
and alpaca with high genetic potential for improved meat
and liber production under the en\ ironmental and manage-
ment conditions of both smallholder and large cooperative
producers in the Andes through progens and performance
testing. Breeding and selection of improved stock is the
longer term goal.
The Rep1itdlctioni Projeit has pursued research on
reproducti e factors in sheep and goats which limit flock or
herd product\ it\. The project has generated basic data on
reproductive performance in local breeds and, based on this.
recommended management practices to reliee reproduct\e
The Incerracdl Goat Projecu has been funded out of
contingency funds with seed money for two years. This
project. under the direct control of the overseas SR-CRSP
Site Coordinator, and using resident staff. emphasizes an
explicit systems approach to develop improvements to local
goat production in the Northwestern desert. Surxeys of
existing conditions ha'e now been completed and major
constraints identified. Future xork on the project is being
sponsored hb a grant from the Canadian IDRC.
Basic Supporting Research
Nearly all projects haxe found it necessary to conduct a
portion ot their SR-CRSP research in the US. Such research
is a part of the v orkplan developed for each overseas site, but
is conducted in the US either because it is of a more
fundamental nature-and therefore has universal appli-
cabilit. -or because of availability of facilities and equip-
ment not accessible in the host country. Computer facilities
are a case in point, and ha\e been used in the US for
developing production systems and biological models
(Economics and Systems Analysis projects) as well as in data
processing Ie.g.. Breeding Project). Some analytical work has
required use of specialized techniques and equipment
a ailable in the US (Nutrition and Animal Health projects)
and the Breeding projects have made use of specialized flocks
and herds established for experimental purposes.
Developed in close collaboration with host country
institutions, the Program has devoted a large effort to
training host country graduate students and technical staff,
as well as US graduate students for work overseas.
The SR-CRSP has provided support for 56 overseas
students pursuing advanced degrees at US universities.
T\%ents of these students are in doctoral programs. Forty
students from our host countries are being supported for
training to Bachelors and Masters degrees at home
uni ersities. The Program has also sponsored 22 US students
conducting research on SR-CRSP related topics as part of
the requirements for their MS and PhD degrees.
Ad anced technical training has been provided for nine
overseas professional staff who spent a total of 31 months in
More than a dozen short training courses lasting over 15
weeksinall ha e been conducted in host countries, with over
350 local staff participating.
On-the-job training is a continuing and important part of
the collaborative research effort. Though not quantified, the
results are evident in the improved quality of work produced
b\ all staff.
Though never a formally planned part of the program
strategy the SR-CRSP has materially improved the research
capabilities of participating institutions in ways other than
scientific skills training.
SR-C RS P resources and act i it ies for example have often
been the catalyst that has inspired investment from other
sources to pro ide essential long-term research facilities.
Recognition of the importance of sound administration as
an ingredient of any research system has led to efforts to
provide training and experience for host country adminis-
trators. In some cases host country\ administrative staff have
taken o\er responsibilities pre iously held by expatriates in
residence at overseas sites.
A most important accomplishment has been the linkages
that the Program has helped develop between institutions
and between subject disciplines. Applied research and
problem solving demand contributions from x various subject
areas which are frequently segregated for historical reasons
or administrative convenience. Developing operational
linkages \within the context of the Program's collaborative
research has, in several cases, provided a basis for extending
this collaboration to other problems of national importance.
Several opportunities have arisen for furthering the
principal objectives of the Program in ways apart from the
main strategic approach. Three that deserve special mention
were funded by the Program to the point when outside
funding could take over.
Caprine Arthritis Encephilitis (CAE), introduced into
Kenya in the mid 1970's from imported exotic goats, went
unnoticed until SR-CRSP scientists spotted its symptoms,
launched diagnostic tests, and on confirmation, worked with
the Kenyan government to develop measures for its
eradication-a feat which has now been accomplished.
The Program provided seed money for an Integrated Goat
Production Project in Northwest Peru. With its initial
surveys now complete and its future research agenda
established, the project was able to successful] appl. for
funds from the Canadian IDRC to support a further three
The SR-CRSP has taken several initiatives in trying to
stimulate research on increasing prolificacy in sheep in
developing countries. Two international workshops have
been partly sponsored by the Program, and a project
proposal commissioned on using prolific sheep research as a
vehicle for building livestock research capacity.
(r ., ,
The Role of Agricultural Research in Developing Countries
In most developing countries, sustained economic growth and food security for expanding populations
require a steady increase in agricultural production. Increases in agricultural output can no longer rely as they
did in the past on expansion of farming into new lands. Instead, new technologies must be applied which can
increase the productivity of existing land or bring otherwise marginal land into use.
The view that new technology increases agricultural growth has been demonstrated by the experience in
Western Europe and the United States during the last century, but only recently has this been demonstrated
for food crops in tropical countries. The increases in yields per unit of land of the improved rice varieties and
hybrid corn in developing regions testify to the cost effectiveness of agricultural research. Recognition of this
has been a major factor responsible for the substantial increase in the agricultural research effort over the last
twenty years, directed towards increasing food supplies in developing countries.
One consequence of this perspective has been the emergence of a global agricultural research network.
International agricultural research centers, such as IRRI and CIMMYT, which concentrate on commodities
of global importance or on the agriculture of large agro-climatic regions; inter-country and regional research
programs in developing countries; and national research organizations in both low and high income countries,
are now in place.
The drive towards rapid industrialization, adopted by many low income countries as a means to economic
growth, obscured the relatively high rates of return to investment in new agricultural technology. However,
"trickle down" theories of economic development and economic development fuelled by industrialization are
somewhat discredited. The need to feed growing populations is now seen as paramount, and the efforts of
millions of small farmers are vital to achieve it. In this new context, it has been possible recognize and accept
the high cost-effectiveness of investment in agricultural research.
Notwithstanding the highly uncertain outcome of research and the often inefficient use of money through
inadequate project design and management, it has been suggested that there are few avenues for investment of
public funds which can be expected so consistently to yield returns as high as those from investments in
agricultural research programs. The increased funding for agricultural research in the last decade is a reflection
of these expectations.
The development of the international agricultural research centers (IRRI, CYMMYT, ICRISAT, IITA,
etc.) and the establishment, in 1971, of the Consultative Group on International Agricultural Research
(CGIAR) provided a vision of hope and incentives for governments to enlarge research expenditures in
The international research centers, by their very nature and mandates, are required to devote their main
effort to tackling problems of broad applicability and to developing the technological foundation upon which
location specific technologies can be established. It is also clear that each country has specific research needs
that can best be met by research carried out in the environment in which the new technology is to be used.
Unfortunately, it is a common experience that national research programs in developing countries are often
weak in conducting effective research to adapt the new technology that is becoming available. Efforts to
remedy the weakness have been gaining momentum over the last decade as the economic importance of their
potential contribution has been recognized. International agencies and research centers, as well as the
development agencies in high income countries, have collaborated with the governments and research
establishments of developing nations to improve local research systems.
Research institutions and universities from the more technically developed countries have become
increasingly involved in efforts to improve the research systems of low income nations. It has been suggested
that help from these types of institutions has particular merit in that the task they are asked to do overseas is
very similar to what they do best at home: research, training, and extension.
In the United States, the special contributions that universities can provide has long been recognized by
USAID, which has called upon university expertise extensively over the last three decades.
The Concept of the CRSP
In December 1975, the US Congress approved an amendment to the Foreign Assistance Act of 1961.
Included in the amendment was Title XII, "Famine Prevention and Freedom from Hunger."A prime purpose
of the new title was to provide the means by which US universities could make their expertise in science and
technology more available to low income countries in their search for technical solutions to food and nutrition
problems. Title XII provides for a strengthened university capacity and greater participation in overseas
development activities. One component of Title XII is "program support for long-term collaborative
university research, in developing countries themselves, to the maximum extent possible, on food production,
distribution, storage, marketing and consumption." This provision is being implemented by Collaborative
Research Support Programs (CRSPs). The CRSP format has been designed to tackle broadly focused but
interrelated multi-disciplinary research needs, and draws upon the capabilities of several institutions rather
than any single one.
Funding for the CRSPs was to be provided by the United States Agency for International Development
(USAID). A Board of International Food and Agricultural Development (BIFAD) was established to initiate
implementation of the Act and to assist in the administration of the Title XII program. A Joint Research
Committee (JRC) was appointed by BIFAD to help in conceptualizing, planning, and implementing CRSPs,
a task which included help in identifying research priorities as well as in program selection, development, and
By mid-1977, BIFAD had conducted an extensive process for identifying research areas, and recommended
three to be used in testing the CRSP concept and procedures. Only two of these topics (Small Ruminants and
Sorghum/ Millet) were approved at the program selection stage, and became the first two CRSPs to be
funded. Subsequently, CRSPs in the areas of Beans and Cowpeas, Human Nutrition, Pond Dynamics, and
Tropical Soils have been developed.
The Origins of the Small Ruminant CRSP
There are approximately 400 million sheep and over 300 million goats in the developing countries of Asia,
the Near East, Africa, and Latin America, representing over half of the total world population of these
ruminants. Most of these sheep and goats are owned by small pastoralists and farmers with very limited
resources. Generally, the various breeds used and their management are traditional; little effort has been made
to upgrade them by incorporating modern technology, and the productivity per animal is low. However,
because of their large numbers, small ruminants make a significant contribution to the economy and food
supply in these countries. Their importance is often not seen in official statistics on commodities mainly
because, being smallholder livestock, they are utilized by the farming family itself, or within the village. But
there is little doubt of their widespread importance to the small farmer: they meet his needs and match his
The low initial and maintenance costs of small ruminant livestock places them within reach of the small
farmer. Because the investment is low, the risks of loss are acceptable. Feed costs are minimal since the animals
are grazed on marginal lands unsuitable for cropping or fed crop residues. The cost of labor for supplying feed
or grazing supervision is again minimal, since children can be given these tasks. The products from sheep and
goats are convenient for local use and require no infrastructure, markets, or external technologies for their
utilization. Their meat and milk outputs come in small quantities which can be readily consumed; their skins
and fiber help support cottage industry and local demand.
Improving the performance of sheep and goats under smallholder management offers a direct route to
improving the diets and living standards of more than 100 million people living in some of the poorest and least
hospitable areas of the world. Much of the technology for improving small ruminant performance is already in
existence in the more developed countries, but a major research initiative is required to adapt this technology
to farmers' needs and the environments found in developing regions. By far the greatest part of this research
has to be conducted in the country where the technology will eventually be used. Rarely, however, do the
respective national research systems have the capability or resources for achieving this unaided.
The Goals and Organization of the Small Ruminant CRSP
The long-range goal of the Small Ruminant CRSP is to increase the food supply and incomes of
smallholders raising sheep and goats. The major route for accomplishing this goal is by developing new
technology to improve animal productivity. Research of this type, aimed at improving production, has to
address many components of the production system. There is no single constraint to productivity, but usually
several, which interact in a dynamic and complex manner. In the SR-CRSP, this is reflected in the number of
scientific disciplines which are addressing the problem simultaneously: breeding and genetics, physiology and
nutrition, health, range management, systems analysis, economics, sociology, and farm management. At the
beginning of the program, thirteen US institutions were collaborating in seventeen research projects in.these
disciplines. Those institutions and the areas of research were:
* University of California, Davis Ohio State University
-Breeding and Genetics -Forages and Nutrition
-Animal Health Texas Tech University
* Washington State University -Range Management
-Animal Health Utah State University
* Colorado State University -Range Management
-Animal Health -Female Reproductive Physiology
* Montana State University Tuskegee Institute
-Breeding and Genetics -Intensive Farm Management
* California State Polytechnic University, Pomona University of Missouri
-Male Reproductive Physiology -Rural Sociology
* Texas A&M University Winrock International
-Systems Analysis -Economics
-Breeding and Genetics -Goat Production Systems
* North Carolina State University
-By-Products and Nutrition
The efforts of each institution are directed by a Principal Investigator (PI) who was initially responsible for
drafting the research proposal. Typically, each PI works on the SR-CRSP in a part-time capacity and remains
at home base, making working visits to overseas sites. The proportion of a PI's time spent on the SR-CRSP
varies considerably, depending on such factors as number of overseas sites and how the research project is
organized. On average, PIs contribute about 25% of their time to the program.
Although the Small Ruminant CRSP comprises a number of closely interrelated projects which seek to
contribute to a common set of objectives, each project has its own distinct objectives, and has considerable
autonomy in designing and implementing its own activities. Apart from the way in which projects are
interrelated through overlapping objectives, much of the cohesive strength of the program is provided by its
organization and internal management.
The University of California, Davis was selected by the 13 collaborating US institutions to serve as the legal
grantee for the program grant and to provide the Management Entity (ME).
The Management Entity, administered by a Program Director, provides leadership in the integration of
research and training activities of the component projects, and is responsible for the overall research
performance and fiscal management for the program. Within UCD, it is responsible to the Dean of the
Though the role of the ME is carefully defined in the Grant document and the Board By-laws, practical
realities indicate that its major tasks are to:
* Receive on behalf of the SR-CRSP, the funds committed by AID and assume accountability for their use.
* Provide funds to the participating institutions for SR-CRSP activities, and ensure compliance with the
Terms of the Grant.
* Provide a focal point for the interaction among the Technical Committee (TC), the Board of Institutional
Representatives (BIR), and the External Evaluation Panel (EEP) within the SR-CRSP and AID, JCARD
and BIFAD outside the SR-CRSP.
* Execute the decisions of the TC and BIR and seek ways to implement the advice of the EEP.
* Maintain liaison with Regional Sub-Programs through the Directors of the overseas collaborating
institutions and Site Coordinators and service them through provision of memoranda of understanding
* Generate the documents of the SR-CRSP including minutes ofthe TC and Board, reports of the EEP, the
Integrated Program Plan, annual reports, the budget and the fiscal reports and provide these to AID and
In addition to staffing the program office at UCD, the ME also provides each overseas worksite with a site
coordinator, whose principal function is to provide logistical support for PIs and other participating scientists
in work conducted at the foreign worksite. Depending on specific needs and opportunities, site coordinators
also serve as communication linkages between scientists.
The work of the Management Entity is supported by three committees: the Technical Committee, the Board
of Institutional Representatives, and the External Evaluation Panel.
The Technical Committee is comprised of all US Principal Investigators of each component research
project, and advises the Program Director on:
* development of plans for the research and training programs and technical services including the addition,
modification, or deletion of component projects and program elements;
* evaluation and recommendation of foreign worksites;
* development of staff and facilities at foreign worksites and planning their utilization;
* development of the annual budget plan for allocation of funds for component projects and work in foreign
* development of policies on publication and dissemination of research results, including joint publications;
* preparation of reports.
The Board of Institutional Representatives comprises one representative from the administration of each
participating US institution. The functions of the Board are to:
* provide liaison between institutional administrations and the ME;
* advise the ME on general program policy and objectives, taking into account changing technical
requirements of the program and the recommendations of the External Evaluation Panel;
* assess the content and balance of research in the SR-CRSP and the adequacy of funding;
* review cost sharing by the participating institutions and make recommendations as needed;
* review the general expenditure pattern of the SR-CRSP and approve the annual budget plan for allocation
of funds to component projects and work in foreign sites;
* approve the addition or deletion of component projects and program elements and changes in program
* review the progress and accomplishments of the SR-CRSP including research and training elements and
The External Evaluation Panel, a five member panel comprised of eminent scientists from non-participating
institutions, acts in an advisory capacity and is responsible for periodic review and evaluation of program
activities to assess whether particular project goals and program objectives are being accomplished.
Specifically, this requires:
annual reviews of the projects and the program;
recommendation of changes in program objectives; and
recommendation of changes in component projects or program elements.
SR-CRSP Overseas Worksites
The research product of the SR-CRSP is ultimately directed toward benefiting the low-resource producers
in developing countries, such as smallholder farmers and nomadic herdsmen.
The problems unique to their situation make research overseas not only appropriate, but essential if
meaningful progress is to be made in improving small ruminant productivity under these conditions. Because
the overseas research component of the SR-CRSP was considered the cornerstone of the project, great care
was taken to select appropriate overseas worksites.
Sites were chosen which are representative of the various ecozones and production systems encountered in
the tropics so that the applicability of SR-CRSP findings can extend beyond the borders of any nation in
which the research is conducted and be useful in other areas of similar climate and topography. Additionally,
the countries in which the sites are located have established agricultural institutions staffed by scientists,
trained personnel, and students with whom the SR-CRSP investigators have an opportunity to collaborate.
These institutions are critical in providing the extension links which are the basis for adopting SR-CRSP
findings by farmers.
Identification of overseas worksites, selection of host country collaborating institutions, development of
collaborative research programs, and staff selection all took place during the first year of program
implementation. There were initial exploratory visits to 21 countries before Brazil, Indonesia, Kenya,
Morocco, and Peru were recommended as overseas SR-CRSP worksites.
In May 1979, these recommendations were accepted by the BIR and the Program Director began a series of
administrative overseas site visits, the purpose of which was to discuss:
* which institute should collaborate with the SR-CRSP in the selected countries;
* which scientists within each of the selected institutions should collaborate with US counterparts;
* the content and nature of the MOU between the overseas institution and UCD representing the SR-CRSP
* the time schedule for the scientist-to-scientist contacts required for planning and initiating the research
* the preparation of a workplan for each overseas site.
Based on these visits and the discussions between scientists in 1979 and early 1980, an overseas worksite
matrix (Table 1) was developed for SR-CRSP participating institutions.
Table 1. Participating US Institutions, Study Areas, and Overseas Locations as Arranged Early 1980
Study Institution Peru Brazil Indonesia Morocco Kenya
Range Texas Tech X
Management Utah X X
Nutrition Ohio X X
Nutrition North Carolina X X
Health UC Davis X X
Washington X X
Breeding & UC Davis X X
Genetics Montana X X
Texas A&M X
Management/ Tuskegee X
Production Winrock X
Reproduction Utah/Cal Poly X X
Economics Winrock X X X X X
Rural Sociology Missouri X X X X X
Systems Texas A&M X X X X X
Agreement between the University of California, Davis, and each of the overseas institutions participating
in the SR-CRSP could not be negotiated in the same manner as the highly complex agreements between UCD
and the individual US institutions involved in the Program. Many of the mandatory management
responsibilities required by AID in accordance with Federal regulations are defined and transferred by these
agreements. The use of such a model with each of the overseas institutions would inevitably take an inordinate
amount of time to negotiate and arrangement through Government-to-Government agreement would take
even longer. A model recommended by AID, overseas missions, BIFAD staff, JRC, and ME alike which has
proven to be a valuable rapid implementation tool has been the development of a relatively simple
Memorandum of Understanding between UCD and the official overseas collaborating institution. This
broadly defines the scope of work and the anticipated contribution from US and host country institutions.
While the same template was used for each of the collaborating institutions, each one was also modified to
match local requirements and idiosyncracies. Research activities did not commence until the MOU's with each
country were in place.
Table 2 shows the countries and the primary institution with whom the SR-CRSP is collaborating, the dates
MOUs were signed, and the type of ecology that is being focused on in each situation.
Table 2. Overseas Research Locations of the SR-CRSP
Country Ecology of Worksite
Brazil Semi-Arid Lowlands
Peru Semi-Arid High
Indonesia Humid Tropics
Kenya Sub-Humid Tropical
Morocco Semi-Arid Savannah
Extensive Empresa Brasileira
de Pesquisa Agro-
Extensive Instituto Nacional
de Investigation y
Intensive Agency for Agricul-
tural Research and
Intensive Ministry of Live-
Extensive Hassan II University
Period of Program
Existence at 9/30/83
March 1980 3 years, 6 months
March 1980 3 years, 6 months
3 years, 7 months
3 years, 7 months
1 year, 8 months
Funding for the Program
Funds for the SR-CRSP have been provided by USAID in the form of a grant, which requires participating
US institutions to contribute a matching amount of not less than 25 percent. AID's total commitment up to
September 30, 1983, is approximately $15.6 million. The cost sharing contribution by US institutions, taken at
25%, amounts to $3.9 million but in fact, after the first five years, matching has been $5.5 million. The overseas
collaborating institutions have shown their own commitment to the program by contributions, mostly in kind,
which amount to the equivalent of $1 million per annum.
The Funding Process
UCD places AID grant funds in a master account and establishes separate accounts for each subproject.
The amount of the annual budget for each subproject, as approved by the BIR, is transferred to that account
and administered under a written agreement between UCD and the participating institution. The institution is
authorized to submit expense reports on a monthly basis for reimbursement, and these are approved by the
ME before submission to UCD fiscal authorities for payment.
After the first year of funding of subprojects, the ME developed a set of guidelines for subsequent annual
budget preparations by PIs. Budget detail is submitted to the ME for the ensuing year along with the annual
report of prior year activities and expenditures.
The guidelines issued by the ME include target figures to guide the PI in budget preparation. These figures
are based on prior expenditure.patterns and the annual subproject plan approved by the TC.
The ME reviews all subproject budgets and makes adjustments found to be necessary. These budgets are
made available to the TC for comment. The ME considers any observations made by the TC and submits a
final budget proposal to the Board for approval. The Board may make further adjustments within the limits of
The ME is responsible for executing the budget as approved by the Board. On occasion, the Board has
approved special funds for expenditure through the ME. For example, funds were set aside for site
development, for developing a program in Morocco, and for a contingency fund. In the case of the
contingency fund, the Program Director is pre-authorized to utilize this fund for items up to $5,000. Larger
amounts can be expended only with prior Board approval.
Except for site coordinator support, all host country expenditures flow through subproject budgets.
Research Objectives of the Small Ruminant CRSP
Accomplishment of the long range goal of increasing the efficiency of production of milk, meat and fiber by
sheep and goats depends on satisfactory resolution of the constraints that presently limit productivity. The
main problem areas that each project is tackling are indicated below, together with the more intermediate
The genetic potential of traditional breeds is generally the outcome of a selection process which reflects the
impact of environmental, technological, economic and cultural constraints. It represents a strong bias towards
survival of the animal under adverse conditions; productivity is only a secondary consideration, and is
therefore usually low. With the availability of modern inputs, such as health protection, the genetic base of the
stock can be profitably shifted towards greater productivity. Genetic and breeding projects therefore have the
* Develop baseline data on the performance of indigenous small ruminants in order to initiate selection
procedures for upgrading the genetic potential of native stock. The survival, fertility, and performance of
local and local/exotic crosses will be compared in order to identify superior types, well adapted to local
* Characterize the production norms for indigenous and introduced breeds under local environmental
* Estimate the inheritance of important production traits, the genetic and phenotypic correlations among
them, and genotype x environment interactions influencing them.
* Evaluate the potential for genetic improvement through selection and crossbreeding of native stock.
* Design coordinated large scale selection, breeding, and genetic improvement programs which can be
sustained after the SR-CRSP terminates.
The key to improving livestock productivity is to optimize reproductive performance. Commonly in
traditional production systems, performance falls well below potential. Factors which may contribute to this
include delayed first estrus and parturition, low fertility and prolificacy, high embryo and postnatal losses,
short reproductive life of breeding females, and maintenance of non-productive males and females in the
flocks. The lack of knowledge of the basic reproductive processes of such indigenous animals as the alpaca,
hair sheep, and highland wool sheep further hamper efforts to improve the reproductive performance of these
animals. Accordingly, the aims of the reproductive physiology projects are to:
* Determine the male and female reproductive parameters for local goats and sheep, identify the
reproductive processes limiting productivity, and establish estimates of reproductive potential for the
* Establish recommendations for management practices which will optimize reproductive rates.
Nutrition and Feeding
Provision of year-round feed is an important step in improving small ruminant productivity. In most
situations, the inadequacy of year-round feed supply is the primary limitation to productivity. When feed
supplies fall to the level of maintenance rations and below, feedstuff and labor are non-productive inputs:
animals, rather than gaining weight, are more likely losing it. There is little information on the intake,
digestibility, and nutritive value of many common tropical forages and crop by-products, which could be
utilized for developing a sustained feedbase.
This lack of knowledge about how various levels and combinations of common and exotic feeds influence
potential animal productivity leaves unclear what constitutes optimum feeding regimens for animals at
different stages of their production cycle. As there is currently no organized system for the efficient
incorporation of these feedstuffs into rations, they are fed in combinations which do not necessarily make the
best use of the available resources.
Research objectives for these projects are therefore to:
* Characterize the nutritional and economic value of available forage, by-product, and native range
* Determine the nutritional requirements for indigenous goats and hair sheep in different stages of their
* Develop guidelines for formulation of balanced maximum profit rations which use the most advantageous
combination of natural and cultivated forages at any given time of the year.
* Establish recommendations for herd and flock grazing practices on the range which will help optimize
animal productivity and range conservation.
* Establish recommendations for mineral, protein, vitamin, and energy supplementation practices which will
help optimize reproductive rates, disease and parasite resistance, growth rates, feed deficiency, and carcass
grade at market age.
Traditional grazing practices tend to be based on use of maximum land area, with minimum management
of the rangeland itself. Today, with diminished grazing areas and greater populations of man and livestock,
this approach leads to overgrazing and permanent damage to rangeland. These projects seek to establish
improved grazing practices which provide livestock with more nutritious year-round grazing while at the same
time maintaining the quality of the range. In striving towards these goals, the immediate objectives of the range
management projects are to:
* Characterize range sites and evaluate existing plant communities in relation to their ecological potential.
* Determine the seasonal variation in availability and nutritive value of range forage plants.
* Evaluate the nutritional requirements, intake, and diet preferences of the grazing livestock at different
stages of their productive cycle.
* Determine proper species and numbers of grazing animals to assure optimal long-term productivity and
stability of the range.
* Establish recommendations for herd and flock grazing practices which will help optimize animal
productivity and range stability.
A wide spectrum of health problems exists among the livestock in developing countries. Ailments which
range from poorly characterized endemic parasite burdens to highly specific vector-borne diseases act in
concert with other production system constraints and exert a considerable influence on overall animal
performance. Poor health depletes the animals' resistance to environmental stress, compounds the debilitating
effects of inadequate nutrition, and hampers reproductive performance, all of which cause reduced productive
output and profit for small ruminant producers.
Though there are often no cures for many tropical diseases afflicting domestic livestock, their incidence can
be reduced enormously by prophylactic measures and improved husbandry. The research objectives of the
animal health projects are therefore to:
* Characterize the prevalence and impact of parasitic and infectious diseases in local herds and flocks,
including seasonal, nutritional, management, and genetic effects.
* Establish practical guidelines for prevention and control of major diseases using improved diagnostic
techniques and treatments for infections and internal parasites, in addition to an improved herd health
The many biological parameters that govern the productivity of livestock interact amongst themselves in a
very complex manner. This complexity makes it impossible for any one scientific discipline to predict the
repercussions of changes introduced into the production systems.
The Systems Analysis Project is therefore seeking to integrate the interaction of the many factors
contributing to the function of the production system, through a workplan which has as its objectives:
* Develop a computer simulation model of individual animal and flock performance based on data derived
from nutritional, reproductive, and health studies.
* Study the dynamics of traditional and improved production systems, using the simulation technique,
identifying the key factors for influencing productivity and productive efficiency, pointing out knowledge
gaps, and establishing research priorities.
The social environment exerts a pervasive influence in determining overall patterns of sheep and goat
husbandry. The decision making process of the farming family is set within this matrix of culture and tradition.
If improved husbandry practices are to be adopted by smallholder society, they must offer benefits without
concomitant social disruption. To help chart this course, the Rural Sociology Project has developed a research
program which aims to:
* Establish, through extensive surveys, the key socio-cultural constraints to small ruminant production.
* Develop an understanding of the process of smallholder decision making, including the role and influence
of women and children in sheep and goat production, and the role of livestock in the social fabric of the
* Evaluate prospective and implemented production system interventions, with a view to understanding the
trade-off between increased productivity and social disruption.
The economic benefits that accrue from livestock rearing must be of major importance in the decision
making process of the smallholder farmer. However elegant the solutions to problems of biological and
technical constraints to production, they are not likely to have much practical meaning unless they appear
economically sound to the smallholder. The gap in our understanding of existing production and marketing
practices, the economic rationale underlying these activities, and the economic value of proposed
interventions, all present major problems in proposing sound interventions. A further problem is the general
lack of data collection and processing facilities which are needed to assess the progress being made at the
smallholder level, as new technology is adopted. As such, the general objectives of the Economics Project are
* Characterize the existing production systems, including input/output relationships, the role of risk in
decision making, and the overall economic rationality and efficiency.
* Study the constraints in transportation, processing, pricing, and storage systems as they relate to producer
incentive and market efficiency, and how these may need to be modified to accommodate improved
* Study the availability of key inputs for the implementation of new recommended practices, including
physical inputs, technical assistance, and credit.
* Assess the likely and actual impacts of proposed and implemented new production practices, especially as
they concern nutrition, employment, income distribution, and local village economics.
In Kenya, an attempt is being made to develop and put in place a complete goat production system in a
situation where small ruminants are almost entirely absent from the smallholder farming regime.
To be acceptable to smallscale farmers, the dual-purpose (meat and milk) goat production system must be
based on low cost, low risk technology and be minimally competitive with cropping activities for land, labor,
and capital resources. Even in developed temperate countries, relatively little research has been conducted on
dairy goat systems, and what is known is generally not suited to production conditions of tropical developing
Potential advantages of small, short gestation, litter bearing dairy ruminants for small-scale farming
systems in the humid tropics are readily apparent. However, this apparent potential remains essentially
untested. It is necessary to characterize existing farming systems in order to determine if and how this
component can be introduced. The research problem, therefore, will be to develop an animal of the
appropriate genotype, develop a health-nutrition-management package appropriate to the small farm
resource base, and to ensure that this component is economically and sociologically acceptable.
The major research objectives of this project that are needed to meet this challenge are to:
* Characterize existing farming systems in locations selected as having potential for dual-purpose goat
* Analyze the farming systems to identify key constraints and propose alternative systems in which these
constraints are relaxed.
* Synthesize and evaluate new technology-management packages initially through on-station research and
subsequently through on-farm research.
* Assist in extending the new proven technology-management package to farmers in the target areas.
Inter-Relationships Between Subprojects
With the SR-CRSP as the pioneer of this particular program mode, it is not surprising that the grant
agreement with USAID specifies the research objectives for the overall program in only the broadest terms.
This has allowed a wide latitude of interpretation of the program in designing its operational components, a
feature that proved to have particular value when worksites and overseas collaborating institutions were
subsequently selected. It meant that each country program could be designed to take account of the specific
prevailing circumstances. What emerged, as a consequence of this flexibility, was a series of country programs
ranging conceptually from technical assistance to reduce a major production constraint to development of an
entire technology production package. Certainly the reason for this spectrum of approaches lies in the different
circumstances of each country worksite, such as the state of the existing livestock production system, the
institutional setting, and government policies. But there was a further factor. The SR-CRSP, breaking new
ground with the collaborative research mode, had little precedent to guide it in choosing implementation
strategies that would meet the needs of diverse participants and institutions and targeted beneficiaries.
Conceptual considerations offered no real help in providing such guidelines. It is true, for example, that an
"integrated approach" has an appealing ring to it, but it is considered a difficult exercise to bring off
successfully, and too often is nothing more than window dressing. Equally, the "single factor approach," with
its strict focus on tackling a single constraint with little provision for other dimensions of understanding, leaves
much to be desired.
The willingness of the SR-CRSP to accept the conceptual spread of the strategies for its overseas projects
provides an opportunity for assessing the relative effectiveness of differing program strategies operating under
the same management structure. It is convenient here to point out the slightly different approaches of the
research activities in the five country worksites.
The objective of the program in Kenya is the development of a "technology package" intended to offer
smallholders in intensively cropped areas of Western Kenya the entire technology required for dual-purpose
goat production. In a situation where the extent of goat production is minimal and modern inputs almost
non-existent, this requires not only new technologies, but ones that are consistent with the social setting and
economic realities facing farmers. The development and testing of such a production package clearly demands
very close working collaboration of PIs in different disciplines and a significant management effort by all
involved has been required to achieve this. It has meant, for example, that research agendas of subprojects
have had to be inter-related and research activities coordinated; in the interests of developing an integrated
package, it has required surrendering some autonomy of individual research projects. This integrated mode of
operation is one which is also in harmony with the needs of the Kenyan collaborating institution, the MLD,
who wishes to develop the capability for this type of research approach.
In the four remaining worksites, Brazil, Peru, Indonesia, and Morocco, research is targeted at farming
systems in which small ruminants already occupy a position of some importance. In these locations, research is
aimed at reducing constraints in existing production systems to improve their poor productivity. The exact
definition of the SR-CRSP program in these circumstances depends on a combination of effects that arise
principally from specification of the key constraints, and the scope of research already being carried out on
related topics by the collaborating institution. In Brazil and Peru, the problems that face small ruminant
production are researched best by a SR-CRSP program that fields a broad spectrum of disciplines. However,
the inter-disciplinary relationships are considerably less integrated than is the case in Kenya. This is partly in
the nature of the problems being tackled, but in part also a result of institutional circumstances. Where the
collaborating institutions are either strong and centralized, or operate in a fragmented fashion, a collaborative
program with a highly integrated approach is either unnecessary or inadvisable.
Further away from the whole system approach with its full range of disciplines and mutual
interdependencies, is the situation found in the Indonesian and Moroccan worksites. In each, the activities of
the SR-CRSP serve mainly to complement the continuing work of the collaborating agencies, rather than to
spearhead a broad attack on a fresh problem. In Morocco, for example, where the collaborating agency is
strong in reproductive physiology and health care and wishes to integrate results as part of its own program,
the SR-CRSP has only four projects on site.
Institution Building Objectives
A major goal of the SR-CRSP is to strengthen the capability of participating institutions to conduct
research on small ruminant production systems.
Typically, the performance of national research systems in developing countries is hampered by shortages of
trained manpower and deficient organizational structures. The latter are frequently manifested as a
fragmentation of effort between institutions, weak linkages between research and extension, and the
generation of research output that has little practical value to farmers.
Although the program was not planned to make formal improvements in the organizational structures of
host country research establishments, it was always foreseen that there would be beneficial spin-offs in this
area. The types of strengthening that have arisen in this way have been derived from:
opportunities to work together with other scientists, which adds to host country scientists' scope of
experience and understanding;
exposures to the team approach to problem solving and the organizational cooperation and coordination
essential to this research mode;
complementarity of resources that strengthens host institutions' abilities to realize their own objectives; and
availability of improved technology and information.
It is obviously difficult to quantify how these benefits improve a research system; they are generally regarded
as important ingredients, but their value must rely on the subjective judgements of the participants themselves.
The program has taken a much more deliberate approach to making investments in the training of young
scientists. The goals here have been to:
* give promising students from developing countries the opportunity of training for a degree at a US
university in an area related to small ruminant production;
* offer students from the US the chance for post-graduate training in a developing country, within the
context of a live problem-oriented program;
* provide host country scientists with the means to upgrade their professional skills, using mainly overseas
visits and in-country workshops; and
* provide technical staff with training in specific skills needed for the program's research agenda.
Current Status of the Program
The Small Ruminant CRSP was the first program to be initiated in the CRSP format. Planners and
directors did not have the advantages of any model to guide them in their choice of strategies for program
implementation. There was little experience to suggest how best to manage and coordinate 13 US and 5
overseas institutions in order to generate effective research results. The existence of 17 semi-autonomous
subprojects operating under a management entity, the scope of whose authority was yet to be expressed in
operational terms, and the undefined functional relationships between the various management bodies in the
program, all left much to be worked out during the initial years of project implementation.
The fact that the program has successfully negotiated these, and many like pitfalls, attests more to the
sincerity and commitment of all those involved, than to the quality of the original program design.
Inevitably as the program has evolved, changes have been made to accommodate the many unforeseen
circumstances that a long term endeavor faces. Some of these changes have been responses to budget
constraints, some to allow more effective management and others to modify research directions to take
account of new circumstances.
As it stands today, the overall program retains basically the main characteristics as originally conceived. A
few of the major changes and comments on the present status of key issues are given briefly below.
Almost inevitably there have been changes in the original pattern of participating institutions and their
worksites. Original project proposals were drawn up before worksites had been chosen. When this was
completed, the magnitude of the tasks facing projects in various worksites became clearer, and, to make best
use of fixed funding, several projects subsequently consolidated their activities to encompass fewer worksites.
The Texas A&M Systems Analysis Project, for example, which had originally proposed activities in all five
countries, has consolidated, and its major thrust is now in Kenya, with some lesser activities in Peru, and none
at all in other sites.
Difficulties in finding suitable host country counterparts have led other projects to shelve their activities at
certain sites. The situation of the UCD Health Project in Indonesia is a case in point.
Two institutions, Ohio State University and Tuskegee Institute, withdrew entirely from the program in
September 1982. The essential components of their research agendas were incorporated by other participating
PIs, and in this report, their accomplishments have been included with the reports of these investigators.
A somewhat different situation arose in the case of the California State Polytechnic University, Pomona
project on reproductive physiology. Given the budgetary constraints that the program was beginning to feel,
and given the similarities of the Utah State University project on reproduction, it was decided that one of these
institutions should play the lead role, and the other operate as a sub-contractor to it. The continuing research
contribution of Cal Poly, Pomona, is now subsumed under the Utah State University Project, and is not
reported separately in this volume.
Table 3 summarizes the current position of US participating institutions and the locations of their overseas
Table 3. Participating US Institutions, Study Areas, and Overseas Locations, Mid-1983
Study Institution Peru Brazil Indonesia Morocco Kenya
Range Texas Tech X
Management Utah X X
Nutrition/ North Carolina X X X
Forages Texas Tech X
Health UC Davis X
Breeding & UC Davis X X
Genetics Montana X
Texas A&M X X
Management/ Winrock X
Production Texas A&M X
Reproduction Utah X X
Economics Winrock X X X X
Rural Sociology Missouri X X X X X
Systems Texas A&M X
Although the SR-CRSP has an MOU with only one primary institution in each overseas country, other
host country institutions have been drawn into the program. Each subproject, with the concurrence of the
primary overseas institution, has made arrangements to work with other national institutions where this best
meets the needs of the particular research plan.
A summary of participating host country institutions is given in Appendix IV.
Research and Training
From the outset, the SR-CRSP was envisioned as a long-term program. It was anticipated that, subject to
satisfactory performance, the initial 5-year grant would be extended. Part of the rationale for this lies in the
inherently long-term nature of research on livestock production systems. It is simply not feasible to complete
some major aspects of this research (e.g., breeding) in short-term projects. But the desirability of a long-term
program is also due to the extra benefits that technical assistance programs derive from longer term
relationships between the people and institutions involved.
Though this report covers the first five year grant period, it is in no sense to be construed as a terminal
report. Every project in each of the different worksites is continuing both its research and training activities.
The direction of future research and a sense of what still needs to be accomplished, is given more fully in the
reports of the Principal Investigators.
A notable trend in the continuing work is the increasing responsibility being undertaken by host country
scientists and administrators. Much of this represents the benefits from the program's heavy investment in
training of overseas personnel in technical matters as well as in the management of research programs.
In several worksites, research projects pursued initially along disciplinary lines are increasingly finding the
need for closer conceptual and working relationships with other SR-CRSP subprojects. This need for greater
integration is a natural consequence of convergence towards a common goal. Even so, there is a resulting
demand for a planned and orchestrated process for managing integration of functions.
This need has presented a challenge to participating scientists, because for the most part, they are educated
in specialized disciplines, rather than in the management and systematic integration of technical, economic,
and social sciences towards a common goal. But the challenge has not been limited to the scientists; it is also a
matter which confronts institutions. Many research institutions are neither structured nor organized to engage
in research that requires the balanced input of a diversity of disciplines. Progress towards this is being made
through the activities of the program, but changes of this nature are intrinsically protracted.
The first five year grant from AID for the Small Ruminant CRSP ended on September 30, 1983. An
extension of one year has been granted, during the early part of which AID will review the program to
determine whether funding for an additional three years would be justified. Research agendas for all projects
continue to be planned in anticipation of this extension.
E eight states of Northeastern Brazil form a natural B R A Z I L
ecological region separated by climate and vegetation *Salvador
from the rest of the country. In this region, there are
essentially two ecoclimatic zones: a dry inland region called s
the sertao and a narrow coastal strip with a humid climate.
The semi-arid sertao, the site of SR-CRSP research, is an
area encompassing approximately one million square
kilometers, with a population in excess of 12 million people. A Worksites
The climate is characterized by distinct wet and dry periods, Q Northeast drought region 2S-
each lasting approximately 6 months, and relatively little d t
seasonal variation in temperature. The entire annual o oo
precipitation occurs from January to June with a rainfall KILOMETERS
ranging from 300-900 mm but evaporation is high, frequently 55ow 50oW 45oW 40oW 35W
as much as 66% of precipitation. Compounding the effects of I I
a water deficit, is the unreliability of rainfall; minor droughts Small Ruminant CRSP worksites in the semi-arid region of Northeast
are common, and major droughts periodic-there have Brazil.
already been six catastrophic droughts this century.
The typical vegetation of the sertao, known as caatinga, is a secondary cover which follows centuries of
felling and over-grazing, and consists of a mixture of small trees and bushes with some herbaceous ground
cover. In some places the caatinga has been cleared for grazing; in others, cotton has been planted. Apart from
this and small patches of subsistence crops (maize, pulses, manioc, vegetables, fruit), there is little cultivation
outside the areas of irrigation.
Both crop and livestock production provide the rural people of the sertao with food and agricultural
commodities for home use and cash income. Major agricultural produce includes corn, cotton, beans, rice and
cassava; the first three of which are the primary crops of the semi-arid Northeast. Sheep and goats are the
predominant livestock species although there are some cattle confined primarily to the wetter regions closer to
the coast. The 5 million sheep in the sertao are the smooth hair type rather than the traditional wool type
commonly found in the more temperate areas of Southern Brazil, and represent about 30% of Brazil's total
Over 90% (6 million) of the country's goat population is found in the arid and semi-arid areas of the interior.
Their proportion of the total livestock population increases as one moves away from the coast and the rainfall
Sheep and goats in Northeast Brazil provide a major source of animal protein and cash income. They are
raised primarily to produce meat for home consumption, but the animals themselves, and their hides and
skins, are also important generators of revenue. Despite the fact that goats appear to be better adapted and
more suited to the rigorous semi-arid environment, the farmers in Northeastern Brazil seem to prefer keeping
sheep to goats, perhaps because sheep are more easily herded and confined and their meat is more popular
than that of goats. Although sheep are apparently never milked, goats are, and most of the milk is used on the
farm. Except for the use of goats for dairy production, sheep and goats serve a similar function. It is not known
to what extent they are competitive for the same limited resource base and if each species is being raised in the
numbers and proportions most advantageous for the optimum management and utilization of the caatinga.
Any program for improving livestock production in the semi-arid interior of Northeast Brazil must address
the major constraint to animal productivity: the pronounced seasonal vegetation pattern and the drastic
decline of available dry season feed on the range. The prolonged period of nutritional inadequacy causes
livestock severe nutritional stress, particularly during the post-weaning, late gestation, and early lactation
stages of their productive cycle. They have great difficulty maintaining acceptable rates of growth and
reproduction, indeed even surviving during the dry season.
This problem is reflected in their small size, low yield, and depressed offtake, which is estimated at less than
20% of the average in many other areas of the country. There is currently a lack of knowledge of the caatinga 's
ability to support livestock, of the management of this resource, and of the sheep and goats which depend
upon it for sustenance. Without more information about factors relating to the caatinga, it would be difficult
to devise management schemes to mitigate the problem of the dry season feed deficit and enhance animal
A number of long-term programs to raise the standard of living of Northeast Brazil's rural and urban sectors
are currently being undertaken by the Brazilian government. A high priority has been placed on improving the
performance of small ruminants as a means of improving the nutritional level and earning capacity of the
people of the sertao, especially the most deprived and impoverished. In addition, improved sheep and goat
production is expected to aid the economies of both the rural and urban sectors by increasing the supply of
livestock products (hides and skins) upon which Northeast Brazil's regional industries depend. The urban
sector is also expected to benefit by the increased meat supply available to city dwellers and the alleviation of
migration of rural people into the already overcrowded urban areas.
The government agency, Empresa Brasileira de Pesquisa Agropecuaria (EMBRAPA), is responsible for
coordinating agricultural development and improvement programs throughout Brazil. They have recognized
that currently existing technological improvements for sheep and goat husbandry, which for the most part
were developed in the temperate regions of Europe, North America, Australia and New Zealand, cannot be
transferred directly to Northeast Brazil. Only technologies appropriate for and adapted to semi-arid tropical
environments can be effectively implemented and achieve positive results. Research must be conducted to
develop these new technologies in situ so they are responsive to the actual conditions in which they are
intended to function.
At the time of initiation of the SR-CRSP, EMBRAPA was a new organization in the unique position of
being able to develop an original research program to meet this challenge. In response to the dynamic and
complex manner in which physical, biological, economic, and social factors interact in agricultural systems,
EMBRAPA's leadership has emphasized the formation of multi-disciplinary teams to focus on the entire
production system and not on isolated components or constraints. This approach was already central to the
Small Ruminant CRSP research philosophy.
In September and October 1979, SR-CRSP scientists travelled to Brazil to confer with EMBRAPA
administrative and technical personnel at two of EMBRAPA's National Research Centers in Northeast
Brazil: Centro Nacional Pesquisa en Caprinos, in Sobral, Ceara (CNPC) and Centro de Pesquisa
Agropecuaria no Tropico Semi-Arido, in Petrolina, Pernambuco (CPATSA). Located at these centers are the
Brazilian counterpart investigators with whom the SR-CRSP researchers are currently collaborating.
The collaborative research program on the production and socio-economic problems related to goat and
hair sheep enterprises in Northeast Brazil consists of projects in seven subject matter areas, each directed
jointly by US and Brazilian investigators. Details of institutions collaborating on each project are provided in
These projects, Animal Breeding and Management, Range Management, Nutrition, Reproduction,
Animal Health, Economics and Sociology, have cooperated closely to examine all aspects of the small
ruminant production system in the caatinga in an integrated and collective manner. The reports of their
research and training accomplishments which follow are the outcome of the first three years' endeavors.
Economic Analysis of
A.J. De Boer
he major long-term goal of the Economics Project was
to set up and institutionalize a research approach that
was analytically appropriate to the agro-ecological condi-
tions of this region, flexible enough to incorporate the wide
range of production systems found in the sertao, and that
could be understood and used by biological scientists within
the National Goat Research Center (CNPC). The rapid and
accurate assessment of technology generated by the cooper-
ating research stations and the training of Brazilian personnel
in the various components of this research approach are all
key elements in the success of the project.
Other goals which are critical to project success are
developing reliable data bases for research and policy
purposes, improving design and evaluation of on-farm
testing of technology, and answering some important
questions about the nature and magnitude of off-farm
constraints which limit small ruminant productivity.
The following research agenda was developed to achieve
* Conduct descriptive (diagnostic) surveys to characterize
small ruminant production systems in the sertao including
farm profiles, farmers' goals and objectives, family
consumption requirements, land-use patterns, on-farm
levels of animal productivity, grazing patterns, levels of
technology applied to cattle and small ruminants, and
* Design a long-term farm monitoring survey to obtain
production traits of sheep and goats on participating
farms under a range of agro-economic conditions using
the survey information.
* Design experiments at the CNPC to help overcome
production constraints related to nutrition, range man-
agement, animal production, health, and breeding.
* Carry out economic analysis of research trials conducted
at the CNPC stations, at cooperating research stations
and on-farm trials designed to test the suitability of
specific technologies under actual farming conditions.
* Study the marketing and price formation procedures for
small ruminants and their products in Northeast Brazil
and identify policies to improve the performance of the
overall marketing system.
* Construct and use a whole-farm linear programming
model as a device to synthesize survey information,
experimental results, farmer objectives, and alternative
management and investment strategies into simple
measures of economic costs and benefits.
* Train Brazilian staff in all steps of the research process
and in the requirements for interdisciplinary research.
The descriptive and diagnostic stages began in April 1980
and were initially limited to the main small ruminant
production areas of Ceara State. The objectives of the survey
were to characterize the major farming systems under which
sheep and goats are produced, analyze the importance of
cattle and cropping enterprises vis a vis small ruminants,
describe interactions between enterprises, estimate farm
resources and their patterns of use, and describe the
production practices applied to small ruminants.
The 1976 crop/livestock Census of Ceara State indicated
that seven of the twenty-two micro regions in Ceara
contained over 50% of the sheep and goats. Within each of
the five micro regions selected, two counties were selected for
the economics survey. An average of fifteen questionnaires
were filled out per county. One county was subsequently
discarded. Studies were conducted on 127 mixed crop/live-
stock farms with large populations of cattle. Data was
analyzed and published in Brazil. Table 1 gives the
distribution of agricultural activities by the surveyed farmers.
Seven groups of small ruminant owners were distinguished:
landowners, manager-sharecroppers, sharecroppers, land
settlement scheme farmers (parceiros), cash renters, com-
munal owners, and employees or permanent workers. In
general, landowners (64% of the small ruminant owners
sampled) have large holdings which are jointly utilized by
themselves, sharecroppers, and permanent or temporary
wage laborers. The average size of goat herds was 116 head
and 142 head for sheep flocks. Most specialization occurred
in sheep (39%) rather than goats (3%). Fifty-eight percent
had both species (Table 2). Levels of management varied but
most producers had corrals, full-time herdsmen, used studs,
and castrated; controlled breeding was rare, however.
The baseline data was used to separate farms into large
and small categories based on median farm size within the
county sample. The sixty-two farms in the small grouping
had an average of 172 ha, the 65 large farms averaged 1,165
ha while the sample mean was 680 ha. Linear regression was
used to test hypotheses about land-livestock relationships.
The major factors hypothesized as influencing total animal
units (AU) per farm were farm size, total man equivalents of
labor available per farm, and percentage of farm perimeter
area fenced. The first two variables were highly significant
but total variation in AU's explained by the equation was
small (Gutierrez et al., 1981).
A subset of these 127 farms was chosen to participate in the
long-term farm monitoring survey. A total of four farms (two
large, two small) were selected in each of eight counties.
Questionnaire construction started in June 1980, and the first
round of surveys started in August-September 1980. A total
of eight rounds have thus far been completed. Two studies
have been completed using this data: the first examined
Table 1. Distribution of Major Agricultural Activities by Survey Farmers
Farms with Farms with Farms with
Crops, Sheep Cattle, Sheep Crops, Cattle,
County and Goats and Goats Sheep and Goats
Granja 87 87 80
Sobral 94 88 81
Crateus 93 85 85
Independencia 100 93 93
Taua 94 94 87
Parambu 100 100 100
Quixada 100 100 100
Quixeramobim 100 89 89
Morada Nova 94 94 87
Average % 96 92 88
Source: Gutierrez et al. (1981), p. 14.
capital structure and income generation on the thirty-two was used as the delimiter. Small herd/flock holders realized
sample farms while the second defined general marketing 40% of their net farm income from small ruminants while
practices and utilization of small ruminants, large herd/flock owners had only 19% of net income from
Analysis of capital structure concentrated on capital small ruminants and 50% from cattle. Conversely, small
assets, income generation, variable costs, and income/ capital landholders had 71% of their net income from cattle and 10%
ratios between the small farm and large farm groupings. from small ruminants while large landholders had income
Results were much different between the large and small almost equally divided between cattle, small ruminants and
groups depending upon whether herd/flock size or farm size crops (Tables 3 and 4). Overall, there was good corre-
Table 2. Characteristics of Small Ruminant Populations on Sample Farms
Average Small Ruminant
Population Per Farm (Head) Producers
Producers Producers with
Sheep and with only with only Sheep and
County Goats Sheep Goatsa Goats (%) Sheep (%) Goats (%)
Granja 163 149 271 13 13 74
Sobral 60 115 145 0 56 44
Crateus 49 172 194 0 54 46
Independencia 213b 173 233 0 71 29
Taua 262 218 448 0 13 87
Parambu 90 49 118 0 14 79
Quixada 94 165 209 0 47 53
Quixeramobim 57 133 161 0 44 56
Morada Nova 58 108 121 6 40 54
Average 116 142 211 3 39 58
a The average sheep and goat population does not equal the average of the two preceding columns since not all farms had both
b A very small number of farms had goats.
Table 3. Average Net Farm Incomes by Activity and by Farm Type in 1980 (US$)*
Small Large Small Large
Herd/ Herd/ Land- Land-
Flock Flock Holder Holder
All Farms Farms Farms Farms Farms
Activity ($) (%) ($) (%) ($) (%) ($) (%) ($) (%)
Cattle 880 42 516 28 1110 50 1064 71 823 36
Sheep/goats 544 26 736 40 422 19 150 10 686 30
Crops 663 32 589 32 689 31 284 19 777 34
Total 2,087 100 1,841 100 2,221 100 1,498 100 2,286 100
Small herd/flock farms = 13
Large herd/flock farms = 20
Small landholder farms = 9
Large landholder farms = 24
*Official exchange rate, December 1980 was CR 64 = $1US
Source: Periodical Survey, 1980.
spondence between levels of capital invested in cattle,
expenses on cattle and income derived from cattle. This did
not hold for small ruminants, however, reflecting their
subsistence orientation as a ready source of food and a
species on which very low levels of capital investments or
current inputs are used. The ratio or income generated to
capital invested is thus very favorable for small ruminants in
this region (Gutierrez et al., 1982).
Examination of general marketing and consumption
practices revealed that sheep and goats are disposed of
through home consumption and commercial sales, the latter
comprising surplus males (1-1/2 years old), cull females, and
cull male breeding stock. In Ceara, there are two markets for
live animals, rural and urban. The rural market consists of
sales to local farmers and farm laborers as well as sales to the
small towns scattered throughout the rural areas. Rural
markets are the main outlet for sheep and goats because
mutton/chevon is less expensive and because of the
convenience factor related to small carcass size and regular
availability. During the periodical survey, data on sales,
home consumption, animal prices and disposition of animals
and animal products was collected for three periods covering
December 1979 to March 1981. Table 5 gives some summary
data from these farms. The results indicate that on-farm
consumption of sheep and goats is common, with twenty-
eight of the twenty-nine sheep producers consuming an
average of sixteen sheep per family over the 16-month
period, and eighteen of the nineteen goat producers
consuming an average of nine goats over the same period.
Sales were about twice the level of home consumption for
each species. Offtake over the 16-month period was 43% for
sheep flocks (16 consumed + 27 sold) per 99 head and 41% for
the goat herds (9 + 17) per 64 head.
On a yearly basis, offtake averaged 24%. About 40% of
this offtake was home consumption and 60% was sold off the
farm. The offtake for sheep was marginally higher than for
goats (25% vs. 21%), but percentages of total herd home/
consumed were identical. Therefore, home-consumption is
in direct proportion to flock/ herd size. This is because larger
farms have larger flocks/herds; they also have larger labor
forces to feed or to whom home-slaughtered meat can be
Selling strategies favored sales directly at the farm.
Farmers cited more flexibility and better bargaining power as
well as the cost and convenience of taking animals to market.
Of the remainder, 24% were sold to local markets, and 2%
cited other outlets. Purchasers were local middlemen (70%),
other producers (24%), and urban middlemen (6%). Sales
were "by head," not by weight, and an average price for a lot
of animals is arrived at. Almost all sales were on a cash basis
(Gutierrez and De Boer, 1982).
Work is continuing at Winrock International and in Brazil
on statistical analysis of the complex data set on animal
performance which is being collected by the periodic survey.
In addition, the Animal Health Project has participated in
this survey and the Range Management component will
collaborate on some research on these farms in 1983/84.
The above information was synthesized into a formal
economic model of the whole-farm production unit. The
survey results clearly indicated that technology packages
aimed specifically at small ruminants could not be analyzed
in isolation because of economic and biological interactions
with the other important farm enterprises (cropping and
cattle). This modeling was carried out at Winrock and
Purdue during 1982. Substantial inputs into the model were
also made by the Utah State Range Project.
The bioeconomic model was developed using a linear
programming approach to analyze sheep and goat produc-
tion at the whole-farm level. It has four major components:
crops, animal feed production, livestock production, and
home consumption. The model consists of 690 activities and
Table 4. Components of Gross Income by Activity
A. By totalfarm income
Cattle Goats Crops All
Com ponent .......................... % ----..................----
Sales 46 12 13 71
Farm consumption 7 9 16
Seeds 1 1
Animal feed 3 3
Payment in kind 1 8 9
Total 46 20 34 100
B. By activity gross income
Cattle Sheep & Goats Crops
Component ---. % -- -.............................
Sales 100 60 38
consumption -35 26
Animal feed -- 9
Payment in kind 5 24
Total 100 100 100
Source: Periodical Survey, 1980.
288 rows. The time unit used was a calendar year divided into
four quarters to capture the seasonal effects on productivity.
The values for some of the 1,995 parameters were obtained
from the baseline and periodical surveys, but many of the
biological parameters were derived from experimental data.
The objective function of the model maximizes net cash
income subject to certain constraints on resource endow-
ment, home consumption requirements, capital and credit
limitations. The model determines the inventories of cattle,
sheep, and goats by quarter. The model also determines the
optimal livestock activities (breeding, growing, fattening) as
well as the optimal age for sale and purchase. For crops, it
determines planted areas as well as the production system
(monoculture or mixed cropping). To illustrate the model
performance, an average .traditional farm was analyzed,
some technical changes tested and, finally, two weather
conditions simulated to estimate the influence of weather on
production patterns. This study resulted in a PhD thesis
submitted to Purdue University (Gutierrez, 1983).
The model generated optimal farm plans which were
found to be consistent with actual input and output levels
observed on the "typical" farm types modeled. Initial results
gave cattle as the only livestock species because it was not
possible to segregate vegetation species according to live-
stock preferences. The higher prices per kg for cattle and
poor reproductive levels of traditional sheep and goat
flocks/herds resulted in cattle utilizing all available feed
nutrients. After consultations with other researchers, avail-
able biomass was portioned and sheep and goats entered the
optimal model solutions at levels close to those actually
The optimal small ruminant activity shown by the model
was purchasing and fattening young stock rather than the
breeding activity which is predominant in the sertao. When
reproductive performance was increased from the net
weaning rate of 70% assumed in the model to a 100% rate, the
breeding activity entered the optimal solution and net cash
income increased by 3%. The model was further used to
simulate the impact of good and poor weather conditions on
net farm income and levels of sheep and goat production.
Under good weather conditions, income increased to 185%
(average situation = 100%) while under poor weather, income
was 53% of the average level. If the income effects generated
with weather variability are compared with the income
effects generated by the technological improvement cited
above (a 42% increase in weaning percentage leading to a 3%
increase in net farm income), this indicates the tremendous
influence of weather on feed production and farm income
relative to an improvement of small ruminant herd/flock
The other major area of research complemented the
survey work on sheep and goat marketing methods and price
formation procedures. In the first study (de Souza, 1981), we
hypothesized that weather conditions, prices, a time trend,
and inventories all influenced the annual supply of sheep and
goats. Statistical supply equations were estimated and the
coefficients indicated that producers' annual adjustments of
inventories were of a similar magnitude for both sheep and
goats. The sales equations are formulated as adjustments of
producer inventories which are themselves functions of
prices. Again, the estimated coefficients between the sheep
and goat supply equations are similar. However, in neither
equation is the current (i.e., January 1) stock of animals a
statistically significant predictor of sales during the year. The
time trend variable is not significant either. By far the most
important influence of current year sales is the previous year's
inventory of animals, which is a function of the previous
year's weather conditions and prices.
In this type of animal supply situation, the highly variable
moisture supply conditions lead to major fluctuations in
animal inventories and sales. Adjustments of these in
response to prices are less important, and a critical economic
factor is the ability of producers to adjust their production
conditions so they can avoid selling when prices are
depressed, such as during drought periods (three consecutive
years of much below normal rainfall is a general rule of
thumb in this region).
Table 5. Sheep and Goat Populations, Animal Disposals and Product Prices for 33 Sample Farms, Ceara State, 1980-81
N* Range Mean sd
Total farm area (ha) 33 10-800 265 236
Sheep population, May 1980 29 21-260 99 59
Goat population, May 1980 19 2-129 64 40
Total sheep and goat population, May 1980 33 38-273 124 68
Dec. 1979-March 1981:
Sheep 28 1-45 16 12
Goats 18 1-19 9 6
Number sold, Dec. 1979-March 1981:
Sheep 28 1-155 27 34
Goats 16 1-44 17 9
Prices of meat sold from farm ($/kg):
May 1980 16 1.00-2.39 1.68 0.40
Sept. 1980 21 1.20-2.06 1.51 0.27
Dec. 1980 17 1.57-2.19 1.75 0.20
March 1981 21
Prices of skins sold from farms ($/piece)
Sheep: May 1980 15 1.00-3.60 2.95 0.40
Sept. 1980 18 1.54-2.57 1.84 0.26
Dec. 1980 13 1.72-2.82 2.22 0.28
March 1981 22
Goats: May 1980 14 3.00-3.60 3.13 0.30
Sept. 1980 15 1.72-2.57 2.04 0.27
Dec. 1980 10 1.88-2.66 2.21 0.25
March 1981 10 2.66-3.44 3.07 0.17
*n refers to the number of farms in each category of data. For example, 29 farms had sheep, 19 had goats, and 28 farms
slaughtered one or more sheep for home consumption during the 16 month period. The means and standard deviations (sd) for
each category are computed using the n appropriate to that category, not the n=33 for the entire sample.
The second study examined the international market for
Brazilian sheep and goat skins. Increased export of skins,
leather and finished leather products is an important
objective of the Brazilian government and this study
attempted to determine what factors influenced exports.
Secondary data from 1969 through 1975 were utilized. The
price elasticity of supply for goat skin exports was 0.30 which
shows less than a proportional export response to a change in
export price. The quantity exported during the current year
was highly influenced by quantity produced the preceding
year. The price of goat skins in the Brazilian market was not a
significant variable influencing exports.
The price elasticity of sheep pelts for export was -2.18
which does not concur with economic theory. Other
variables influencing sheepskin exports included sheepskin
production during the preceding year and increased domestic
incomes which reduced supplies for export (de Moraes,
Significance of Research Findings
It is important to note that the research results to date are
derived exclusively from data collected in Ceara State, much
of which is not typical of the extensive small ruminant
producing areas in parts of other northeastern states. In
general, cattle and cropping were important competing
enterprises. Small herd/flock holders gained relatively more
income from small ruminants than holders of large numbers
of ruminants. Large landholders had income coming about
equally from small ruminants, cattle, and crops. This is
significant for the research direction in that owners of smaller
landholdings and smaller herds/flocks may be more
receptive to innovations since their small ruminants are more
important economic assets. Goats, in particular, have many
characteristics of a subsistence enterprise: the surveys
identified low levels of inputs used, almost no supplemental
feeding practiced during the dry season, high levels of home
consumption, and little selective breeding being practiced.
Technical improvement will have to go hand in hand with
commercialization of the industry and programs to educate
producers about the potential economic benefits of small
ruminants in general and goats in particular.
The whole-farm linear programming model was found to
be an adequate representation of existing farm practices and
is a flexible analytical tool which can be used to test the
economic impact, as measured by net farm cash income, of
specific practices (e.g., supplemental feeding, forage crop
production, planting improved pastures, worming), or sets of
practices (e.g., pasture rotation on different types of pastures,
cross-breeding combined with a culling regime). The model
also indicated that climatic variation will be a much larger
influence on net farm cash income than even substantial
improvements in herd/flock productivity. The impact of
strategies for better managing drought may help overcome
these large fluctuations in animal numbers and farm income.
The research on marketing and price formation identified
no obvious constraints within the traditional system given
the extensive nature of production, long distances between
producers and markets, small numbers marketing per farm
and low offtake rates. The price seemed most directly related
to animal size. The forthcoming analysis of three years' data
from the periodical survey will allow more accurate estimates
of factors determining the on-farm prices received for small
Implications for Future Research
Economic Analysis of Research Results
The framework is now in place to train Brazilian
researchers in the use of the linear programming model and
to use it for the routine analysis of proposed small ruminant
interventions. For single experiments, simple budgeting
procedures will continue to be used. The focus will now be on
forming interdisciplinary teams in the major research areas
who can develop specific technology packages which we can
then proceed to analyze. A major component of the linear
programming model deals with feed utilization and the
Range Management Project in particular will work closely
with us to improve this aspect of the model and to develop
specific range management and feeding regimes that can be
subjected to model analysis. An important feature of the
model is the ability to trace the impact of a technology over a
sequence of weather conditions; the stability of model
predictions over a range of possible precipitation regimes can
thus be assured.
The ongoing operation of the 32-farm periodic survey not
only continues to provide data on production characteristics
of sheep and goats, but serves also as a field laboratory for the
animal health team to assess the farm level incidence of
specific animal health problems. It will further be used in a
joint project by Range Management and Economics to
identify vegetation types and relate these to animal per-
formance, herd/flock numbers, and farm economic status. It
is anticipated that these farms will be increasingly used for
on-farm experimentation and the economic analysis of these
results will be made easier because of the long-term nature of
monitoring their performance.
One constraint to increased production of small ruminants
in the Northeast may be lack of favorable prices and market
demand for small ruminant meat. Beef is much more
expensive in retail markets in the Northeast and if the long-
term demand prospects for small ruminant meat are
unfavorable, this could have serious implications for small
ruminant improvement programs. Therefore, a major study
is needed to examine consumer characteristics and the long-
term consumption and price prospects for small ruminant
meat. Several institutions in the Northeast have an interest in
this work and the necessary data so there are good prospects
that this research can be carried out before the conclusion of
P r P srA
Goat shelters such as this have been subsidized bv the Brazilian government
with the hope that they would increase survival and reduceparasitism in the
livestock. Photo: W. C. Weir.
The emphasis of the Economics Project has been on
developing a network of Brazilian social scientists with
experience and expertise in the analysis of small ruminant
problems. The project was handicapped during the first 2/2
years by not having an EMBRAPA economist assigned as a
Co-PI at CNPC/Sobral, but a significant investment was
nevertheless made in training. Two computer software
packages were purchased and installed at the Federal
University of Ceara's computer center. Training in the use of
these packages was provided by Nestor Gutierrez, and these
were subsequently used by two graduate students in the
Department of Agricultural Economics at that University
who were partially supported by SR-CRSP funds. On-the-
job training was provided to both students, and they
participated in the periodic surveys.
Three advanced degrees have been received under full or
partial SR-CRSP support. Titles of theses are indicated in
the publication section of this report.
Gutierrez-Aleman, Nestor. 1983. Characterization and Linear Program-
ming Analysis of Sheep and Goat Production in the Sertao Zone of
Northeast Brazil. PhD Thesis, Purdue University.
Gutierrez-Aleman, N., A.J. De Boer, and E.W. Kehrberg. 1983. Modelo
Matematico para Valuar Tecnologias Pecuarias a Nivel do Toda la
Finca en el Nordeste de Brazil (Mathematical Model to Evaluate
Livestock Technology at Whole-Farm Level in Northeast Brazil). To be
Presented at IX ALPA Meetings, Santiago, Chile.
Gutierrez-Aleman, N., A.J. De Boer, and J. Ubiraci-Alves. 1983. Character-
isticas Economicas de los Systemas de Produccion de Pequenos
Rumiantes en el Sertao de Ceara, Brasil (Economic Characteristics of
Small Ruminant Production Systems in the Sertao Zone of Ceara,
Brazil). To be Presented at IX ALPA Meetings, Santiago, Chile.
Gutierrez, N. and A.J. De Boer. 1982. Marketing and Price Formation for
Meat Goats, Hair Sheep, and their Products in Ceara State, Northeast
Brazil. Proc. III Intl. Conf. on Goat Prod. and Dis. Tucson, AZ. pp.
Gutierrez, Nestor F., A. John De Boer, and Enrique Ospina. 1982. An
Econometric Model of the Colombian Beef Sector: 1950-1970.
Canadian J. Agr. Econ. 30(1):61-70.
Gutierrez-A, Nestor, A. John De Boer, e Jose Ubiraci-Alves. 1982.
Interacoes de Recursos e Caracteristicas Economicas dos Criadores de
Ovinos e Caprinos No Sertao do Ceara, Nordeste do Brazil: Resultados
Preliminares. EMBRAPA-CNPC Boletim de Pesquisa Numero 3.
Gutierrez, N., A.J. De Boer, and V. Vieira. 1982. Capital Structure and
Farm Incomes for a Sample of Sheep and Goat Producers in Northeast
Brazil. Proc. III Intl. Conf. on Goat Prod. and Dis. Tucson, AZ. p. 512.
Gutierrez, Nestor, John De Boer, y Enrique Ospina. 1981. Un Modelo
Econometrico para la Ganaderia de Carne en Colombia: 1950-1970.
Revista de Planeacion y Desarrollo. 13(3):91-102.
Gutierrez-A, Nestor, A.J. De Boer, and Jose Ubiraci. 1981. O Papel do
Economista, Agricola Dentro de uma Equipe Multidiciplinar, no Caso
de Caprinos-Cultores no Sertao do Ceara. Presented at the 19th
Brazilian Congress of Rural Economy and Sociology. Olinda,
Gutierrez-A, N., J. De Boer, and Jose Ubiraci. 1981. Some Resource
Interactions and Economic Characteristics of Sheep and Goat
Producers in the Sertao of Ceara State, Northeast Brazil: Preliminary
Results. Presented at First National Symposium on Tropical Sheep and
Goats. Fortaleza, Brazil.
De Moraes Eloy da Costa, Odorico. 1982. Oferta de Exportacao de Peles de
Caprino e Ovino do Brazil (Export Supply of Sheep and Goat Skins in
Brazil). MS Thesis, Federal University of Ceara.
De Souza-Neto, Jose. 1981. Analise Econometrica de Oferta de Caprinos e
Ovinos do Estado do Ceara (Econometric Analysis of Sheep and Goat
Supply in Ceara State). MS Thesis, Federal University of Ceara.
Utah State University
The lack of knowledge on reproductive performance of
sheep and goats in Northeast Brazil is recognized as a
constraint to any program intended to improve small
ruminant productivity. A primary objective of the Utah State
Reproduction Project has been to help remedy this situation
with particular emphasis on researching those aspects with
the highest potential payoff. This goal synchronized well with
the activities of the newly established National Goat
Research Center (CNPC) and also provided support for
development of their research agenda in this field.
During this first phase of the SR-CRSP, the Reproduc-
tion Project has worked effectively with scientists from the
Center to generate basic information on reproduction. More
recently, plans have been developed and approved by the
CNPC to expand the influence of the Reproduction Project
to other units of the EMBRAPA system in Northeast Brazil.
Specific research objectives of the SR-CRSP Reproduc-
tion Project have been:
* To determine reproductive potentialities of native and
exotic genotypes of goats and sheep in Northeast Brazil.
* To measure the influence of environmental factors on the
reproductive performance of relevant genotypes of goats
* To develop and apply management practices that are
compatible with the resources, objectives and under-
standing of the producers so as to improve reproductive
performance of their flocks.
These objectives have been pursued in close collaboration
with the CNPC and other appropriate EMBRAPA units so
that the other objective of the project, building research
capacity, could be accomplished through formal and
informal training and contact between scientists.
Knowledge of the age at which kids reach sexual maturity
is important to herd management because it defines when
male and female kids must be separated or the males
castrated to avoid unplanned matings. A study was initiated
in the first year (1981) with Moxoto kids. It was found that
the separation of the penis from the sheath occurred when the
kids were an average of 134.5 days old and 12.2 kg of live
body weight, and the first ejaculation via artificial vagina
with the presence of sperm cells occurred when they were
139.5 days old and weighed 13.3 kg. The semen character-
istics (volume, sperm concentration, and individual motility)
were mostly stabilized at 174 days of age. This is the age at
which the Moxoto males reach sexual maturity or zootech-
nical puberty (Figure 1).
L Divided Scrotum
f Non-Divided Scrotum
= Head Abnormalities
= Detached Heads
= Cytoplasmic Droplet
= Cytoplasmic Droplet
= Tail Abnormalities
HA DH PCD DCD TA
Figure Frequency of abnormal sperm cells in ejaculatedsemen ofMoxoto
bucks related to scrotal morphology.
Three breeds of goat (Anglo Nubian, Bhuj and SRD) were
studied in the second year. In this study, the separation of the
penis from the sheath and the presence of sperm in the semen
via electroejaculation occurred at about the same age.
Physical examination of the goats established a morpho-
logical difference in the degree of scrotal division bipartitee).
To see whether this had influence on semen quality and hence
potential significance for breeding management, the fol-
lowing study was conducted. The morphological differences
were scored where 0 refers to no division and 5 refers to
complete division. Semen from animals with divided and
non-divided scrotum was taken to measure the difference due
to scrotum morphology. The information obtained showed
no difference in ejaculated volume, sperm cells concen-
tration, and individual motility. But the same semen samples
tested following thermal stress showed that the animals with
divided scrotum have a significantly higher proportion of live
sperm after incubation at 30 minutes (P < 0.05), 60 minutes
(P < 0.01), 90 minutes (P< 0.05) and 120 minutes (P< 0.01)
than the semen from animals with non-divided scrotum. The
proportion of pathological sperm cells (Figure 1) was greater
(P< 0.01) in semen from animals with non-divided scrotum
(32.9%) compared with semen from animals with divided
In female goats, the age (363.6 days) and weight (12.6 kg)
at puberty were similar between breeds (Caninde, Marota,
Moxoto and Repartida) and relationships between the
weight at birth and weaning and age at puberty were
negative; however, a significant difference occurred (P< 0.05)
for age and weight at puberty between single and twin born
doe kids. The first detected ovulation was one in all breeds
studied. Based on estimates from the presence of corpus
albicans at laparotomy, ovulation had occurred previously
without estrus in approximately 40% of the does.
The estrous cycle and period duration were studied during
one complete year with thirty 18-24 month old SRD does.
The length of estrous cycle was 21.2 days and of the estrous
period 56.8 hours. The estrous cycle duration agrees with
information published elsewhere, but the estrous period was
longer than the published data indicated for different
environmental conditions. These observations need to be
reconfirmed. This information is important if there is to be
planned management of mating.
Three breeds of female goats (Moxoto, Marota and SRD)
were studied during two years in two different nutrition-
management conditions (native pasture and confinement), to
measure the frequency of estrus and ovulation, ovulation
rate, estrous cycle length, and the number of estrous periods
per female per month throughout the year and per year. This
study will be used as a partial requirement for a PhD thesis of
a host country scientist at Utah State. Preliminary results
showed that the SRD breed type goats have the highest
ovulation rate (1.68) in both management conditions
followed by Marota(1.43) and Moxoto (1.36). The ovulation
rate of goats on native pasture was higher (1.59) than in
confinement (1.45) for all breeds combined. The frequency of
estrus was evenly distributed throughout the year, indicating
that the goats of Northeast Brazil are continuous breeders.
On the average, the goats showed 1.06-1.19 estrous
periods/month, 3.50-3.92% of does showing estrus daily and
12.72-14.28 estrous periods/year. The goats failed to adapt
well to confinement and consumed inadequate roughage
(elephant grass) and concentrate to gain in body weight and
demonstrate reproductive levels comparable to those on
native pasture. Part of the difference may have been due to
the poor quality of elephant grass fed, particularly during the
earlier part of the experiment.
A study was conducted with 203 does to measure and
demonstrate the usefulness of ultrasound equipment in
pregnancy diagnosis. The study in goats was conducted at an
average of 109.5 days of pregnancy. The diagnosis was
accurate in 93.1% of the cases, 84.3% in case of pregnant, and
100.0% in the case of non-pregnant does, accuracies which
are very similar to those experienced in dealing with sheep.
A collaborative study in goat reproduction was under-
taken between the CNPC, Universidad Federal Rural de
Pernambuco, Recife and the Reproduction Project. In this
study, the influence of feeding on reproduction and
production performance of producer-owned native SRD
goats were measured during two reproductive phases: late
pregnancy and early lactation. The results showed that
heavier birth and weaning weights were recorded in kids born
from does supplemented with 300 g/ day/ doe of 16% CP and
75% TDN concentrate 45 days before kidding compared
with kids from the non-supplemented control group does
and does supplemented 45 days post partum. The kids from a
group of does supplemented for 90 days (45 days before and
45 days after kidding) showed better performance than the
control group as well as those given 45 days supplementation
before kidding. In terms of kid mortality, no difference was
found. Higher abortion rates were recorded for the control
group (16%) and does supplemented during lactation (12%)
than does supplemented 45 days before kidding (8%).
A study is underway to estimate reproduction and
production performance and to identify the constraints of
goat production systems under traditional conditions. The
study uses 4,000 goats in 17 private farms located in five
municipalities in Northeast Brazil.
The data being collected are: flock structure (numbers,
age, sex, color and breed type), frequency or presence of
congenital defects (polledness, prognatism, supplementary
teats, intersex, etc.), frequency of kidding throughout the
year, kidding percentage, kidding intervals, kidding rate, age
at first kidding, sex ratio, mortality rate (by season, sex, type
of birth, weight), off-take rate (sales, home consumption),
interval between parturitions and first post partum estrus,
whole flock body liveweights (at least three per year), body
weight of offspring at birth, weaning and one year of age,
body weight of dams at parturition and weaning, and growth
rate of the kids.
Herd composition at the initiation of the study gave 51.5,
19.0, and 6.5% of adult, yearlings, and kid females,
respectively. In males, the proportion was 4.04, 13.3, and
5.66% for adults, yearlings, and kids, respectively. The low
proportion of kids shown in this data indicates a high
mortality rate or low fertility and/or prolificacy in female
goats in the flock. Concerning the frequency of defects,
35.2% out of 2,770 animals have supplementary teats and
12.24% are polled. Sixteen color patterns and combinations
were found in goat flocks of this project. The proportions
varied from 0.29% for light brown/dark brown and 17.40%
for black, followed by 17.30% for light brown, 15.7% for
brown, 4.99% for brown/white, etc.
The kidding in Northeast Brazil occur throughout the
year, showing that the goats are continuous breeders.
However, a peak of 30.4% of the parturitions occur during
July-September and 21.1% occur in May. Seventy-three
percent of the kidding were singles and 26.8% twins or
triplets. The month or season of the year seems to have no
influence on prolificacy, showing that type of birth is evenly
distributed throughout the year. When a limited breeding
period was imposed on the goats at a private farm, 100% of
the does were bred with 78.2% fertility at kidding, compared
to 88.4% fertility in the control farm. No abortion was
observed in does in the limited breeding season and 10%
abortion was recorded in the control farm. In the farm with a
limited breeding period, only 4.2% of kids died to weaning
age and in the control farm 27.6% of kids were lost.
Marota, Moxoto, and SRD breed type offemale goats are shown confined
during a study on reproductive capability in which they are compared with
goats grazed in native range pasture. Photo: W. C. Foote.
During 1980 and 1981, five Brazilian Somali rams were
used to study the semen characteristics as they varied
seasonally and individually. The data show that ram semen is
fertile throughout the year with no significant differences
between wet and dry season, except for individual motility,
which was shown to be better during the dry season. When
the information was classified by quarterly periods, however,
the semen characteristics showed statistical differences for
ejaculated volume, mass motility, individual motility, concen-
tration, and linear dimensions of the testicle. This obser-
vation could be important in planning seasonal mating.
A three year puberty study was conducted using 112 ewe
lambs from three breeds (Morada Nova, Santa Ines and
Somalis), grazed on native pasture. Breed differences (P<
0.01) were found for weight and age at puberty. The Santa
Ines breed was heavier (27.7 kg) followed by Morada Nova
(22.5 kg) and Somali (18.6 kg). The Santa Ines breed reached
puberty at 274.5 days of age, the Morada Nova at 277.5 days,
and the Somali at 321.6 days. The ovulation rate was 1.34 for
the Morada Nova, 1.31 for the Somalis and 1.24 for the
Santa Ines. The ovarian response was determined by
laparotomy performed 40 hours after heat was detected. The
data available also shows that 77.3% of the ewe lambs
ovulated before the first standing estrus at puberty.
The mean estrous cycle and period length for these same
ewes was 18.2 days and 31.3 hours, respectively. No
difference was found between breeds for estrous duration.
The length of the estrous cycle differed for the Morada Nova
and Somalis breeds (P < 0.05) but neither differed from the
Santa Ines. In total, 90.14% of the estrous cycles were from
14 to 19 days in length, 1.8% were less than 14 and 8.1% were
greater than 19 days.
The interval between parturition and the first post partum
estrus in 279 sheep was studied during 1980-1982, with the
same three breeds. The Morada Nova (65.04 days) and
Somalis (61.90 days) did not differ statistically (P< 0.05) but
both were significantly shorter (P < 0.05) than the Santa Ines
(76.96 days). Overall, post partum interval for sheep was
64.96 days. Ewes producing single lambs had a shorter post
partum interval (63.67 days) than those producing multiples
(72.26 days; P < 0.05).
An experiment, initiated in July 1982, is underway with 72
ewe lambs divided in two nutrition-management groups (fed
in confinement and grazed in native pasture) and three
breeds (Morada Nova, Santa Ines and Somali). In confine-
ment, the ewe lambs receive elephant grass ad libitum plus
1% of their body weight per day in concentrate containing
16% crude protein and 75-80% of TDN. The feed
consumption is measured or estimated in both groups to
discover the feed availability in native pasture related to the
season of the year and to compare it to the feed consumed in
confinement. Consumption on native pasture is estimated in
cooperation with range scientists at the Center.
138 166 194 222 250 278
Figure 2. Growth age and weight at puberty of ewe lambs (Morada Nova,
Santa Ines, Brazilian Somali)fed in native range pasture and confinement.
Preliminary results demonstrate a significant effect of
nutrition-management on the onset of puberty. Preliminary
comparisons were made between the two nutrition-man-
agement groups when the lambs were approximately 8.5
months of age. For the three breeds combined, 69% reached
puberty at 234 days of age and 25.9 kg body weight in
confinement compared to 25% at 265.6 days of age and 23.2
kg of weight for ewes on native pasture (Figure 2). Ewe lambs
from the three breeds, fed in confinement, reached puberty at
different ages and weights: Santa Ines at 249 days and 32 kg,
Morada Nova at 225 days and 23 kg, and Somalis at 229 days
and 22 kg. The ovulation rate was 1.14 for confined ewe
lambs and 1.04 for lambs fed on native pasture. This
experiment will be carried out for 30 more months to find the
frequency of estrus and ovulation, ovulation rate, and
estrous cycle length throughout the year.
The use of ultrasound equipment for pregnancy diagnosis
was evaluated using 112 ewes from the three breeds used in
the other study reported. The results showed that at 92 days
of pregnancy, 92% of the diagnoses were accurate (100% in
the case of ewes that were pregnant and 50.1% for non-
pregnant ewes). The use of this method which is easy to
perform on the farm is being encouraged in the management
of small ruminants, especially for extension technicians in
Northeast Brazil. Earlier diagnosis of pregnancy is an
important tool in flock management as it allows separation
of pregnant ewes for preferential treatment, and for
additional matings of ewes not yet pregnant.
Significance of Research Findings
Although the goat and sheep industry in Northeast Brazil
provides only a subsistence standard of living, it plays a vital
role in the economy of the area. The low level of income is, to
a large extent, due to environmental constraints and to lack
of resources required for improved production. But efforts to
improve production have been hampered by the paucity of
information on reproduction, production capabilities and
management practices needed to realize potential capa-
bilities. The results obtained from the Reproduction Project
demonstrate that reproductive performance could be im-
proved by the use of important management practices:
* Under traditional management, sheep and goats are not
weaned, nor do producers know when and how to do so.
Our results show that weaning, or at least separation or
castration of the male kids must be done by 140 days to
avoid undesirable mating.
* Puberty in female goats and sheep takes place around 335
days of age and 16.6 kg of body weight. Pregnancy at this
stage should be prevented since this body weight is
insufficient to maintain a successful pregnancy to term
due to inadequate feeding and growth.
* The ovulation rate at puberty in native breeds of goats is
usually low (1.0) and does not permit making selections at
this age for higher ovulation rate. In sheep, breed
differences appear to be evident at puberty, thus allowing
for selection at this age. Our results show that the Morada
Nova has a higher ovulation rate than the Somali and
Santa Ines, implying that the Morada Nova, with better
management, will be more productive than the other two
* Information on the incidence of estrus and ovulation
indicates that in both sheep and goats, reproduction is
possible year round if breeding season, nutrition, and
health are properly managed. This has been proven in the
private flock study where producers have traditionally
kidded year round. However, production was signifi-
cantly improved by restricting the breeding season.
* According to our data, two parturitions per year are not
feasible in sheep but with proper management three
parturitions in two years could be easily achieved in sheep
and is possible in goats. This would significantly reduce
the present parturition interval.
Establishing the research project in one of the seventeen privately owned
flocks of goats in Northeastern Brazil, near Taua. Photo: W.C. Foote.
* Results on reproduction in the ram and buck indicate that
they can be used for breeding throughout the year.
The accuracy of ultrasonic pregnancy diagnosis in sheep
and goats enables producers to introduce this as a
management practice to identify non-pregnant animals
and avoid their maintenance in the breeding flock.
As a result of our research, reproductive capabilities of
sheep and goats are much better understood and some
management practices have been introduced at the producer
level and information is now available to develop others.
Implications for Future Research
The objective of establishing the most relevant infor-
mation on reproductive capabilities of sheep and goats has
been largely achieved. However, extension of these results to
local producers is still in its beginning stages. Major efforts in
the future will involve:
Continuation of programs to develop, apply, and test
reproduction-related management practices at the pro-
ducer level. The program initiated by the Reproduction
Project, working directly with producers, is proving to be
very successful. Its results have direct application to the
low income producer as well as strengthening extension
programs and contributing to developing practices with a
wider application. It will still be necessary, however, to
identify additional basic information that is needed for
improved management and reproduction. This will be
pursued through continuing research with present CNPC
Initiation of the Reproduction and Related Management
Network Organization in Northeast Brazil will provide a
means for research collaboration on constraints related to
reproductive performance. Physical and personnel re-
sources are limited and this network organization will
provide a way to pool these to solve local and general
regional problems in an economical manner. The
Reproduction Project has accepted the invitation by the
CNPC to participate in the development of the planned
dairy goat program and expects to invest the required
effort to help make it successful.
* Arrangements have already been made by the Center with
support from the Reproduction Project and the SR-
CRSP Management Entity to establish a Radioisotope
facility and initiate a Radio-Immuno-Assay (RIA)
program in reproduction. A new IICA/OAS consultant
has been provided to the Center to replace Dr. Simon
Riera and will continue the work in reproduction and help
establish the RIA program. Other programs for facilities
development will be supported by the Reproduction
Project as they are identified and as funds permit.
Training, always one of the most important objectives of'
the SR-CRSP Reproduction Project, has been conducted
from the start to provide both degree opportunities and a
wide variety of non-degree training for technicians, as well as
junior and senior scientists.
In Brazil, a shortcourse entitled "Management of Repro-
duction" was presented for researchers from the EMBRAPA
system units, northeast universities, and extension service
personnel. The training covered aspects in experimental
design, data collection, organization of data for analysis,
analysis and interpretation of results, etc., as they related to
reproduction and management for reproduction. Nineteen
Semen collection from Brazilian Somali ram for studies on fertility and
seasonal variation in semen quality. Photo: W.C. Foote.
persons completed the shortcourse and received certificates
Training in specific techniques and procedures required
for research have also been provided to scientists and
technicians. These include updating techniques for semen
research in the male, and laparotomy, laparoscopy and
pregnancy diagnosis procedures for research in the female. In
addition, training has been provided on data collection,
processing and interpretation.
Training has also been undertaken at Utah State to
provide graduate degree programs for overseas personnel.
Mr. Aurino A. Simplicio from the EMBRAPA/CNPC,
started his formal training for a PhD in Reproduction in
August 1982 and is expected to complete this by summer
1984. A major part of his research for his dissertation was
conducted in Brazil under the direction of the SR-CRSP
Reproduction Project and the CNPC. Mr. Jose Ubiraci
Alves is now conducting his research with the project at the
private farms with goats under a similar arrangement.
Information from this study will serve as thesis material for
his MS degree. He expects to begin formal training at Utah
State in the fall of 1984.
Riera, G.S. 1982. Reproductive Efficiency and Management in Goats. Proc.
III Intl. Conf. on Goat Prod. and Dis. Tucson, AZ. p. 162.
Riera, G.S., A.A. Simplicio, and W.C. Foote. 1982. The Accuracy of
Pregnancy Diagnosis in Goats Using Ultrasound. Proc. III Intl. Conf.
on Goat Prod. and Dis. Tucson, AZ. p. 497.
Riera, G.S., A.A. Simplicio, and W.C. Foote. 1981. Diagnostico de Prenhez
em Ovelhas das Racas Morada Nova, Santa Ines e Somalis. Simposio
Nacional de Reproducao Animal, 4. Belo Horizonte, MG. p. 47.
Simplicio, A.A., E.A.P. Figueiredo, G.S. Riera, and F.A.M. Lima. 1982.
Reproductive and Productive Performance of the Undefined (SRD)
Genotype of Goats Under the Traditional Management System of
Northeast Brazil. Proc. III Intl. Conf. on Goat Prod. and Dis. Tucson,
AZ. p. 349.
Simplicio, A.A., G.S. Riera, E.A. Nelson, and K.P. Pant. 1982. Seasonal
Variation in the Semen and Testis Characteristics of Brasilian Somali
Rams in the Hot Semi-Arid Climate of Tropical Northeast Brazil. J.
Repro. & Fert. 66:735.
Simplicio, A.A., G.S. Riera, and J.F. Nunes. 1982. Estrous Cycle and
Period Evaluation in Undefined Breed Type (SRD) for Goats in
Northeast Brazil. Proc. III Intl. Conf. on Goat Prod. and Dis. Tucson,
AZ. p. 310.
Simplicio, A.A., G.S. Riera, W.C. Foote, and J.F. Nunes. 1981. Puberdade
em Femeas Capri nas Nativas no Nordeste do Brasil. Simposio Nacional
de Reproducao Animal, 4. Belo Horizonte, MG. p. 34.
Simplicio, A.A., G.S. Riera, E.A. Nelson, and J.F. Nunes. 1981. Puberdade
em Machos Caprinos da Raca Moxoto. Simposio Nacional de
Reproducao Animal, 4. Belo Horizonte, MG. p. 13.
Simplicio, A.A., G.S. Riera, and J.F. Nunes. 1981. Ciclo Estral e Estro de
Ovelhas das Racas Morada Nova, Santa Ines e Somalis. Simposio
Nacional de Reproducao Animal, 4. Belo Horizonte, MG. p. 30.
Simplicio, A.A., G.S. Riera, J.F. Nunes, and W.C. Foote. 1981. Idade, Peso
e Atividade Ovariana a Puberdade em Borregas das Racas Morada
Nova, Santa Ines e Somalis. Simposio Nacional de Reproducao
Animal, 4. Belo Horizonte, MG. p. 43.
Simplicio, A.A., G.S. Riera, H.E. Bezerra, and J.F. Nunes. 1980. Ciclo
Estral e Distribuicao de Estro Sem Raca Definida. Congress Brasileiro
de Medicine Veterinaria, 17. Fortaleza.
Goat and Sheep
North Carolina State University
ne of the most serious limiting factors to small
ruminant productivity in Northeast Brazil is the
seasonality of herbage available from the native caatinga
vegetation. During the extended 6-8 month dry season,
growth and reproduction rates suffer tremendously without
supplemental feed. However, if feed supplements must be
purchased from off-farm sources, their use is unlikely to
prove economical. The challenge therefore is to develop
feeding systems from resources available on the farm which
can correct the gross nutrient deficiencies of the dry season
and contribute to a more profitable production enterprise.
The design of improved dry season feeding systems relies on
* Nutrient availability from the caatinga, measured in terms
of daily animal intake.
* Nutritive value of available feeds such as crop residues
and by-products, cultivated forages, and native
* The production response of animals to increments of
The Nutrition and Feeding Systems Project is directed
toward the second and third of these research areas. Our
major objectives over the long term are to:
* Determine the nutritional value of local feedstuffs which
could be available for dry season supplementation of
goats and sheep.
* Develop practical guidelines for dry season supple-
Complementary to these two major objectives are three
additional areas of activity: forage preservation and storage;
correlation of feed composition to animal performance; and
verification of the nutrient requirements and relative
efficiencies of goats and sheep of the breeds commonly
encountered in the target production systems.
Throughout the project, major attention is focused on
animals at their most critical stages of the production cycle:
post-weaning growth, breeding, late gestation, and during
After project planning visits by the Principal Investigator
in 1979, collaborative research was launched in June 1980.
During the 1980 dry season, crop residue samples were
collected from farms in three states of the Northeast, and
feeding trials were started with weanling lambs and kids at
the Sobral station of the National Goat Research Center
(CNPC). A set of rustic, but highly serviceable, pens were
built at the CNPC for confinement feeding trials. These pens
have been used throughout the 1980, 1981, and 1982 dry
seasons for growth trials with Northeast Brazil native SRD
goats, and Morada Nova and Santa Ines sheep.
Starting in 1981 and continuing through 1982, an animal
nutrition analytical laboratory was established at the CNPC.
The capacity to conduct fiber, protein, gross energy, organic
matter and in vitro digestibility (artificial rumen) deter-
minations on a routine basis is now fully developed. The
resident senior investigator has organized the laboratory, set
up a data recording system, and trained a group of four
Research activities were extended in 1982 to the EMEPA
experimental farm, Fazenda Pendencia, in Soledade,
Paraiba. The emphasis there has been on three breeds of
dairy goats: SRD, Anglo Nubian, and German Pardo.
Sorghum hay, buffel grass hay, and cultivated legumes have
been tested for feeding lactating does.
Another major development in 1982 was the estabhsh-
ment of a functional metabolism unit at the CNPC in Sobral.
Space was allotted in an unused greenhouse, metabolism
crates were constructed by the CNPC carpenter after a first
model was perfected, and the first digestibility trial was
conducted in October. Several subsequent trials have already
been conducted but await sample analysis before results can
In 1983, nutrition research has been extended to producer
goat herds in a trial designed to test the effects of copper and
zinc supplementation on the doe's reproductive performance
and on the weaning height of her offspring. Plans have been
made for the fall of 1983 to initiate energy and protein
supplementation trials with SRD does, and to expand the
emphasis on semi-confinement systems to growing lambs
and kids; and to intensify the work with mature does of the
Some of the more important information generated
during 1980-82 is summarized below.
Chemical Composition of Crop Residues
The main cash crop in the Northeast is cotton. Maize and
beans are also planted for home consumption or sale. Certain
parts of the residue from these crops were sampled from
farmers' fields and analyzed for chemical composition with
results as shown in Table 1. Bean and cotton leaves were the
highest in crude protein and lowest in total cell wall fiber of
the materials sampled. However, the supply of these
particular crop residues is not expected to be great. The stem,
leaves and husks of the maize plant are similarly poor in
nutritive value but could provide some protein and energy as
a basal dietary component (Oliveira et al., 1982a).
Table 1. Chemical Composition (as a percent of DM) of Plant Parts From Crop Residues in Northeast Brazil
No. of Crude Total
Residue Samples Protein Cell Wall Cellulose
Husk 18 6.1 54 23 V.4
Samples were collected from farms in three states, early in the dry season of 1980.
Source: Oliveira et al., 1982a.
Table 2. Performance of Lambs Consuming Maize Residue Rations During 70 Days in Confinement During the Dry Season in
51% maize stovera
51% maize cobsa
25.5% cobs, 25.5% husksa
51% maize stovera
5% mata pastoc
10% mata pastoc
15% mata pastoc
Caatinga grazing, no supplement
51% maize stovera
14% mata pasta
28% mata pastd
14% cunhad 10 116 107
28% cunhad 10 102 91
Caatinga grazing, plus
200 g/day maize grain 10 42
a Plus 28% cottonseed cake, 20% maize grain and 1% iodized salt.
b Slaughtered at the beginning of the trial.
c Plus 51% maize stover, 20% maize grain, 1% iodized salt and 23, 18 or 13% cottonseed cake.
d In these rations half or all of the cottonseed cake was replaced by mata past (Cassiasericea) or cunha (Clitoria ternatea), fed as
Source: Arruda et al., 1981, and Oliveira et al., 1982b.
Acceptability and Digestibility of Maize Crop Residues
Chopped, green napier grass (Pennisetum purpureum)
and the wholeplant stover, husks, or cobs from maize crop
residues were offered ad libitum to 18 SRD kids. Relative
acceptability of these four materials was measured by pairs of
feeds in the six possible combinations. The average
palatability index was napier grass 65, stover 56, husks 45
and cobs 34% of the total feed offered. Mean daily dry matter
(DM) intakes on the best treatments (napier and husks, and
napier plus stover) were about 60 g per kg of metabolic size
(body weight075), which is insufficient to support an
adequate growth rate for kids during the stage of weaning to
maturity (Johnson et al., 1982).
Arruda (1983) conducted a digestibility trial on maize
stover rations with Morada Nova wethers. Rations con-
taining 60, 70 and 80% wholeplant maize stover were 59, 56,
and 55% in digestibility of total dry matter, at an intake level
of 72 g per kg metabolic size, or about 3% of actual body
weight. Extrapolating these digestibility coefficients to a
100% maize residue would indicate 47 5% DM digestibility
for the maize residue alone. Digestibility of the crude fiber
and non-crude fiber fractions of DM were almost identical,
indicating the inappropriateness of the crude fiber deter-
mination as an index of indigestibility.
Growth of Lambs on Maize Residue Rations
Results of trials for three years are summarized in Table 2.
In the 1980 trial, weight gains and carcass yields were not
significantly different when different parts of the maize plant
were used as the roughage base. An economic analysis with
then-current prices indicated that the dry season confinement
feeding of Santa Ines wether lambs was profitable (Arruda et
al., 1981). However, it was felt that the use of cottonseed
cake, which must be purchased outside the farm, should be
minimized. Therefore, in the following year the ground, dried
foliage of the native legume, "mata past" (Cassia sericea),
was introduced as a partial substitute for the cottonseed cake.
Average gains were slightly lower (99 g/day) than in the first
year (142 g/ day), but were not influenced by the level of mata
past; carcass yields were also lower. In 1981, six lambs
maintained for 70 days under caatinga grazing with no
supplement lost an average of 11 g/day during the 70 day
trial (Oliveira et al., 1982b).
In the third year (Oliveira, 1983), 50 and 100% of the
cottonseed cake was replaced by mata past or by a second
legume hay called "cunha." Again, intakes and rates of gain
were quite satisfactory with partial substitution (Table 2).
Both daily intake and gains were depressed with total
cottonseed cake replacement, only slightly with cunha but
severely with mata past. This was possibly due to a higher
total cell wall fiber content of this ration. Control lambs
grazing caatinga forage, supplemented with 200 g of maize
grain per day, gained an average of 42 g/day.
Morada Nova ewe lambs at the CNPC, Sobral. These lambs made
satisfactory growth during the dry season while kept in confinement andfed
a ration based on about 50% chopped maize stover, a common crop residue
of the region. Photo: W.L. Johnson.
Growth of SRD Kids on Maize Residue Rations
Rations which had given satisfactory gains with lambs in
1980 were repeated in 1981 with SRD kids (Barros et al.,
1982a). Results with goats were much less satisfactory than
for sheep (Table 3). In the first trial, gains averaged 19 g/day
when maize cobs were the only roughage. Gains were 38
g/day when either stover or husks, or maize cobs supple-
mented with napier grass, were fed. Confinement feeding
with such low gains is unlikely to be economical. However,
results for goats that grazed the caatinga with no supplement
were much more catastrophic-three died of starvation and
the remaining three lost an average of 25 g/day.
Since low daily intake seemed to be one of the limiting
factors, the 1982 trial was designed to test ways of increasing
intake, either by reducing the indigestible fiber concen-
tration, by adding molasses to improve palatability, or both
(Robb et al., 1983a). Adding molasses to the 50% maize
stover ration improved intake over the previous year but did
not influence rate of gain or carcass yield (Table 3). A ration
with only 30% maize stover and no molasses was equally
unsatisfactory. Those diets in which the stover was reduced
to 40 or 30% and molasses was added resulted in increased
rates of gain (but not DM intake or carcass yield) compared
to the 50% stover/5% molasses or 30% stover/0% molasses
treatments. The higher rate of gain was a reflection of higher
DM digestibility of the total ration (Table 3).
The best results of all were obtained for the group which
consumed 30% stover and 5% molasses, and had daily access
to caatinga grazing. This observation has led to the planning
of additional trials to further test semi-confinement systems
for young goats during the dry season.
Energy Supplementation Effect on Age and Weight at
Puberty, Morada Nova Ewe Lambs
During the 1980 dry season, twenty-four lambs averaging
140 days old and 12.4 kg liveweight were randomly assigned
to eight pens; two pens were than assigned to each of four
energy levels. All pens were offered 105 g/animal/day of
cottonseed meal plus free choice ground maize stover (70%)
and cob (30%) mixture. The four energy levels were attained
by offering either 50,200,350, or 500 g/ animal/ day of maize
grain. Age at first estrus was 326, 290, 305 and 278 days for
the four energy groups, indicating a response in precocity at
higher energy intake levels. The average weight at first estrus
was 20.5 kg and was not influenced by energy intake level.
Rates of gain were approximately 39, 52, 56, and 62 g/day. It
was concluded that the two highest energy levels could not be
economically justified and that dry season supplementation
of Morada Nova ewe lambs should be adjusted to attain
average gains of about 40-50 g/day (Barros et al., 1982b).
Supplemental Mineral Requirements
Correction of daily intake levels of energy and protein to
sustain a respectable rate of gain through the dry season may
result in certain mineral elements becoming marginal or
deficient. Given the high cost of commercial mineral
mixtures, it is desirable to know exactly which minerals are
limiting and how much of each is required. Barros and
Spears have determined from foliage analysis that the
caatinga vegetation may provide insufficient zinc and
copper. A long term experiment was initiated in 1983 to test
this possibility. Barros and Figueiredo (1983) have published
preliminary data from twelve female SRD kids grazing in the
caatinga, indicating no growth response from the addition of
bone meal to an iodized salt supplement. Average gains were
66 g/day over an eight-month period which included all of
the 1981 dry season. Given the importance of phosphorus to
normal animal growth and reproduction, it would be
advisable to further test its supplemental response. Barros
(1983) has suggested possible species differences in this
regard, reporting an average intake of 5 g/day of an iodized
salt/bone meal mixture (l:1) by 39 SRD goats, while 119
sheep consumed an average of 31 g/day. No immediate
explanation is available for such a large difference in mineral
Feeding Systems for Lactating Dairy Goats
An experiment was conducted at the EMEPA farm in
Paraiba with Anglo Nubian and German Pardo lactating
goats, in which sorghum hay was shown to be a superior
forage to buffel grass hay in terms of total milk yield and yield
of milk fat (Table 4). A daily supplement of 400 g/ animal of
ground maize grain resulted in an average of 325 g/day of
additional milk, compared with non-supplemented animals.
The German Pardo does yielded more milk (1,232 g/day)
than the Anglo Nubians (908 g) but with lower fat (3.0%
versus 4.6% for the Anglo Nubians) (Robb et al., 1983b). It
was concluded that buffel grass is inferior to sorghum when
fed as hays, probably due to a higher content of indigestible
cell wall and consequent lower intake by animals. Further
trials with dairy goats which are planned to start in Paraiba in
1983 have been designed to test the intake and digestibility of
buffel grass under grazing.
Significance of Research Findings
We are learning that instituting major improvements in
the feeding system for goats and sheep in the semi-arid
Northeast is not a simple task. This should come as no
surprise; if easy solutions were available, they would have
been applied long ago.
The goats and hair sheep of this region have adapted to
hardship. They are survivors. The physiological adaptations
that make them survivors are not necessarily compatible with
an expectation of high rates of response to improved levels of
nutrition. However, hardy as these animals have shown
Table 3. Performance of SRD Kids Consuming Maize Residue Rations in Confinement During the Dry Season in Northeast
g g/W75 % %
1981 (98 days)
51% maize stovera 6 37 56 42.9
51% maize cobsa 6 19 48 41.4
51% maize husksa 6 39 62 45.9
Napier grass, cobsb 6 38 52 44.3
Caatinga grazing, no supplement 6f -25 40.8
Reference animals 6 39.3
1982 (70 days)
50% stover, 5% molassesd 128 37 85 42.3 53
40% stover, 5% molassesd 11h 48 82 41.6 58
30% stover, 5% molassesd h11I 61 74 42.4 70
30% stover, no molassesd l1h 29 65 43.1 66
30% stover, 5% molasses,
caatinga grazingd e 5 72 74 54.4
"Plus 28% cottonseed cake, 20% maize grain and 1% iodized salt.
b Chopped green napier grass adlibitum; plus cobs 15%, cottonseed cake 49%, maize grain 35%, iodized salt 1%, restricted to 2%
of liveweight per day.
c Slaughtered at the beginning of the trial.
d The 1982 rations contained 1% iodized salt plus cottonseed cake and maize grain at the following levels: 23.5 and 20.5, 21.5 and
32.5, 19 and 45, 18 and 51, and 19 and 45.
e 12 hours grazing, 12 hours confinement per day.
f Three of the 6 original animals died from starvation; data are means of the remaining 3.
g Carcass data are from 6 animals.
h Carcass data are from 5 animals.
Source: Barros et al., 1982a, and Robb et al., 1983.
Table 4. Performance of Dairy Goats Fed Two Forages with
Two Levels of Energy Supplementation at EMEPA's
Pendencia Experiment Station, Soledade, Paraiba (1982)
Treatment Yield Fat
1. Buffel hay, no grain 814 3.7
2. Buffel hay, 400 g maize grain 1099 3.6
3. Sorghum hay, no grain 1000 4.0
4. Sorghum hay, 400 g maize grain 1365 3.7
a All animals received 250 g/day of cottonseed cake.
Source: Robb et al., 1983b.
themselves to be, they do suffer from poor, zero, or negative
growth rates, reproductive failure, and even death when the
dry season reaches its extreme. But they have shown
themselves responsive, at least during the growth stage, to
supplemental energy, protein and minerals.
The problem is recalcitrant: when supplementation is most
needed, the feed materials with which to provide it are the
least available. An abnormally poor year in terms of grazing
resources will also be poor for the crops, whose residues and
by-products might be fed to small ruminants. Also, the
farmer may opt to give his limited feed supplies to his cattle
and let the goats continue to fend for themselves. If our
research findings are to be applied, a new frame of mind must
be developed, with the idea that farm profits can be improved
if the small ruminant stock can get through the dry season
with positive rather than negative productivity.
We have shown some ways to do this. Preserving some
type of feed supply for the most severe months of the dry
season (October-January) is the first requisite. This can be
sorghum hay or a cultivated legume such as cunha, if land
can be set aside during the rainy season to grow these crops.
Alternatively, it could be the native legume mata past,
harvested as hay at the beginning of the dry season. Chopped
maize stover can also be fed, but this material, more than the
others, will require additional high quality energy and
protein feeds to accompany it.
It has been conclusively demonstrated that Morada Nova
and Santa Ines lambs can reach an acceptable slaughter
weight during the dry season, either with total confinement
feeding or, perhaps preferably, with judicious supplemen-
tation of the caatinga resource as it dwindles in supply. For
lambs destined to join the reproductive flock, their continued
growth toward maturity can be assured by similar but more
modest supplementation programs. If both of these practices
can be adopted, the flock will be smaller and more productive
as the new rainy season commences, greatly enhancing its
efficiency. With time, similar strategies for goats will be
The evidence is mounting that tropical grasses such as
napier and buffel are not high quality feeds for the goats and
sheep of Northeast Brazil. These highly lignified grass species
require a longer rumen residence time than these small beasts
can afford. Therefore, intake levels and rates of digestibility
are not sufficient to provide the nutrients for optimum
growth, lactation and reproduction.
We are also learning not to expect the native goats and hair
sheep to respond similarly to dietary interventions. While
they might resemble each other more than either of them
resemble their distant cousins, the European breeds, there are
still important differences. The ability of Morada Nova and
Santa Ines lambs to grow at a rate exceeding 90 g/day when
consuming a diet consisting of 50% maize stover, while SRD
goats gain barely one-third that rate, is a case in point.
Implications for Future Research
As the evaluation of animal feed resources in Northeast
Brazil continues, more attention should be given to those
attributes which will assure a contribution to sheep and goat
nutrition during the prolonged dry season. Laboratory
screening should include crude protein, total cell wall fiber
(NDF), and in vitro digestibility as a minimum. Promising
feeds should then be subjected to in vivo digestibility and
voluntary intake determinations and incorporated into
balanced rations for testing intake and animal performance.
In the near future, promising forage legumes, crop by-
products, and young lesser-lignified grasses which might be
preserved as silage, will be evaluated in the above manner.
As strategies for dry season supplementation continue to
be developed and tested, a closer collaboration with the
Range Management Project will be established. Simplified
techniques are needed for a quick evaluation of nutrient
availability from the caatinga so that supplementation
strategies can be timely, economical, and effective.
We also hope to devote more attention to the adult
breeding herd or flock. As soon as land and animal resources
allow, we plan to initiate dry season supplementation work
with the pregnant doe and ewe.
A final objective for the mid-term future is to catalyze a
more intensive outreach effort emphasizing dry season
feeding strategies. This, of necessity, will be a multi-
disciplinary approach with reproduction, range manage-
ment, and economic aspects included. One possible model
for such a project would be to work through selected
EMBRATER agencies to reach clusters of pilot farms in
several states of the Northeast. The flow of information in
such a project must be continual and two-way, through a
network of producers, extension personnel, and state and
national research staff. If kept flexible and pursued with
energy and diligence, such an effort could have a rapid
impact on productivity at the farm level. The alternative is to
relegate our research findings to the scientific journals and
the occasional seminar or technical bulletin, with only a
random chance of affecting rural welfare.
Coatinga which has been cleared by cutting and burning is being compared
as a source of nutrients for sheep and goats to brush (in the background)
which has not been disturbed for about 25 years. Photo: W. C. Weir.
Training of Brazilians has occurred at four levels: thesis
research support for post-graduate degree candidates, in-
service training of young scientists, in-service training of
technicians, and outreach to personnel of other institutions
by conducting workshops. Direct scholarship support for
graduate training has not been offered, partly because there
have been no identified candidates, but mainly because
EMBRAPA normally has had ample scholarship funds of its
Thesis research advice and financial support was preferred
to Francisco de Assis-Arruda, who has now completed his
Master's degree program at the Federal University of Ceara
in Fortaleza. Similar support has been arranged for Nelson
Barros who intends to apply for admission in 1984 or 1985 to
graduate school at North Carolina State University, with Dr.
Jerry Spears as his advisor. Barros and Spears have already
initiated a collaborative project in Sobral which is expected
to yield data for the PhD thesis. English tutoring for Mr.
Barros has also been supported.
The in-service training activity is especially important for
such a young institute as the CNPC and for its relatively
young cadre of scientists. This type of training is continual
and depends on the frequency and quality of communication
between collaborators. The measurement of success is
somewhat subjective, but there is no doubt about its
In-service training of technicians is easier to document. Dr.
Robb has trained four persons in various aspects of the
nutrition laboratory operation; it is now a smoothly
operating facility, with greater limitations from physical
(electricity and water) than human factors. He has also
helped train animal technicians at Sobral and Fazenda
Pendencia in conducting feeding and metabolism trials.
A two-part workshop was held in Brazil during July-
August 1982. Part I (three days) was devoted to a discussion
of new information on goat and sheep nutrition, in seminar
format. About 20 persons from the CNPC staff and state
research agencies attended. Part II (five days) was a hands-on
shortcourse in laboratory analytical procedures, supervised
by Dr. Robb and completed by ten persons.
Arruda, F. 1983. Consumo e Digestibilidade de Racoes para Ovinos cor
Diferentes Niveis de Restolho de Milho (Intake and Digestibility of
Rations for Sheep, with Different Levels of Maize Crop Residue).
Federal University of Ceara, Fortaleza.
Arruda, F., E. Oliveira, N. Barros, W.L. Johnson, and A. Azevedo. 1981.
Restolho de Cultura do Milho para Ovinos da Raca Santa Ines
Mantidos em Confinamento (Maize Crop Residue for Santa Ines Sheep
in Confinement). Anais, Congreso de Zootecnia, Goiania GO, Brazil.
Barros, N. 1983. Consumo de Mistura de Mineral por Caprinos e Ovinos,
no Estado de Ceara(Intake of a Mineral Mixture by Goats and Sheep in
the State of Ceara). EMBRAPA National Goat Research Center Tech.
Comm. No. 11, Sobral CE, Brazil.
Barros, N. and E. Figueiredo. 1983. Efeito da Suplementacao Mineral no
Peso e Idade a Puberdade de Femeas Caprinas SRD (Effect of Mineral
Supplementation on Weight and Age of Puberty of SRD Female Kids).
EMBRAPA National Goat Research Center Tech. Comm. No. 12,
Sobral CE, Brazil.
Barros, N., W.L. Johnson, E.R. de Oliveira, and T.W. Robb. 1982a.
Desempenho de Caprinos sem Raca Definida em Confinamento no
Nordeste do Brazil (Performance of SRD Goats Fed in Confinement in
Northeast Brazil). Annals, 19th Ann. Mtg., Brazilian Society of
Zootecnia, Piracicaba. pp. 274-275.
Barros, N., W.L. Johnson, E. Oliveira, T.W. Robb, A. Simplicio, and S.
Riera. 1982b. Effect of Level of Energy Supplementation on Age and
Weight at Puberty of Purebred Morada Nova Ewe Lambs in Northeast
Brazil. Joint Mtg. of the American and Canadian Soc. of Ani. Sci.,
Abstracts. p. 405.
Johnson, W.L. 1981. Efficiency of Digestion in Goats and Sheep. Invited
Paper, First National Symposium on Tropical Goats and Sheep,
Fortaleza. SR-CRSP Technical Report Series No. 8.
Johnson, W.L., N. Barros, E. Oliveira, and B. Haryanto. 1982. Relative
Acceptability of Napier Grass and Maize Crop Residues by Brazilian
SRD Goats. Proc. Ill Intl. Conf. on Goat Prod. and Dis. Tucson, AZ.
Oliveira, E., N. Barros, T.W. Robb, and W.L. Johnson. 1982a. Utilizacao de
Restolhos de Cultura na Alimentacao de Caprinos e Ovinos (Utilization
of Crop Residues for Feeding Goats and Sheep). EMBRAPA National
Goat Research Center Tech. Circ. No. 4, Sobral CE, Brazil.
Oliveira, E., N. Barros, T.W. Robb, W.L. Johnson, and L. Vale. 1982b.
Partial Substitution of Cottonseed Meal by "Mato Pasto" (Cassia
sericia) Hay in Rations for Lambs in Northeast Brazil. Joint Mtg. of the
American and Canadian Soc. of Ani. Sci., Abstracts. pp. 447-448.
Oliveira, E., N. Barros, T.W. Robb, W.L. Johnson, and K. Pant. 1983.
Substituicao do Farelo de Algodao por Feno de Leguminosas em
Racoes para Ovinos Santa Ines em Confinamento (Legume Hays
Replacing Cottonseed Cake in Rations for Confined Santa Ines Lambs).
Annals, 20th Ann. Mtg., Brazilian Society of Zootecnia, Pelotas RS.
Robb, T.W., N. Barros, W.L. Johnson, and E. Oliveira. 1983a. Effect of
Ration Fiber Level, With or Without Molasses, for Goats Maintained in
Confinement and Semiconfinement in Northeast Brazil. Amer. Soc. of
Ani. Sci. Ann. Mtg. Abstracts. Pullman, WA.
Robb, T.W., W. Sousa, C. Zometa, E. Oliveira, W.L. Johnson, and P. Leite.
1983b. Feno de Capim Buffel (Cenchrus ciliaris) e Sorgo (Sorghum
vulgare) como Fontes Volumosas na Alimentacao de Cabras Leiteiras
(Buffel Grass and Sorghum Hays for Dairy Goats). Annals, 20th Ann.
Mtg., Brazilian Society of Zootecnia, Pelotas RS.
Range Research For
Utah State University
A large proportion of the world's poor live on lands
classified as arid or semi-arid, where scarce or erratic
precipitation and marginal soils make crop production a
risky venture, or entirely preclude it. These people depend
largely on animal agriculture for their sustenance, and the
type of husbandry practiced is, almost without exception,
extensive grazing of native rangelands. Such is the case for a
large region of Northeastern Brazil, known locally as the
sertao. Characterized by scant and unreliable rainfall, and
vegetation patterns which reflect this, the crux of the problem
for animal production in the sertao, like so many range areas,
is providing adequate nutrition. Such concerns as genetic
improvement and herd health pale in the light of animals
starving on a recurring basis.
Improvement of the forage resource is the obvious short-
term rationale for our research in the sertao. This includes
two directions: achieving a better balance between the wet-
season surplus and the dry-season deficit on an annual basis,
and on the long-term basis, minimizing the impacts of
Another dimension to the long-term consequences of land
and animal management is the inherent stability and
sustained productive capacity of the land. In their attempt to
face current demands of growing human populations and
withering economics, agriculturists in developing countries
often feel compelled to over-graze or over-exploit the land.
On a world-wide basis, this has led to an unprecedented
degradation of the land resource over the past two decades.
The special term desertificationn" has been coined to describe
The sertao is not exempt from this problem. However,
before steps can be taken to change the trend of events, the
ecological characteristics of the system must be understood.
This includes not only knowing about the specifics of sheep
and goats, but also about the climatic, edaphic, vegetational,
and sociological aspects of the region and its people. It is
these additional components that separate range science as a
discipline from the more traditional animal science dis-
Given this perspective, we are approaching the problem of
sheep and goat production in the sertao region of North-
eastern Brazil with four general objectives or problem areas
*Ecological assessment of the range forage resource: Soils,
vegetation, and climate of the sertao are highly diverse
and climate varies profoundly from season-to-season and
year-to-year. To understand how these factors affect
forage and animal production and to enable local
experiment station results to be extended to the broader
region, ecological studies are necessary.
* Plant-animal relationships: Typical of all tropical eco-
systems, the vegetation of the sertao, the caatinga,
contains a wide diversity of plant species. These species
vary widely in their palatability and nutritional value to
sheep and goats, and any particular species may vary
markedly over the year. These studies are aimed at
understanding the nutritional relationships of animals to
their erratic and diverse environment.
* Rangeland improvement and brush management: One of
the few so-called "management" alternatives available to
smallholders of the sertao is manipulation of the woody
caatinga vegetation. Local wisdom holds that clearing
improves forage conditions. While this may be true in the
near-term, the long-range consequences of land clearing
are not well known and may be highly undesirable, both
from the standpoint of animal production and land
* Grazing management: With information in-hand from
the first three objectives, we will ideally be able to design
and recommend grazing management practices and land
treatments that will optimize the relationship between
effective animal production and sustained forage produc-
C. 'E- I~L
Esophageal fistulated goats and sheep are used to determine what the
animals eat to survive through the dry season in the caatinga. Photo: W.C.
Ecological Inventory and Assessment
SR-CRSP research on sheep and goat production is not
an end in itself; the program is designed to address the
management and production constraints of the smallholder
in Northeast Brazil. Therefore, understanding the entire
production system is a key element in the SR-CRSP/
EMBRAPA association. To this end, production character-
istics of caatinga must be known before extrapolation of
grazing research results is attempted. How much of the
caatinga in Northeast Brazil, or even in the Acarau Valley, is
similar to the native vegetation at the CNPC fazenda, for
which we know details of goat diet preferences and
nutritional levels and response to clearing? How many
different kinds of caatinga are there? Is vegetation response
to clearing and burning practices always the same? The
present objective is directed toward answering questions such
as these. The studies are being undertaken by Utah State
researchers Joao Queiroz and B.E. Norton.
The ecological survey of range resources is divided into
two phases. The first, now complete, consisted of intensive
on-the-ground site descriptions in an area of 1,400 km2 in the
Acarau Valley. Three types of information were collected:
soil characteristics, including evidence of erosion; vegetation
present on the site, plus evidence of dynamic features such as
successional trends; and the way the area is being utilized by
the local landowners. A survey of this kind requires expertise
in soil taxonomy, ability to recognize plant species, and an
appreciation of mechanisms responsible for ecological
change. The vegetation currently present on a site may not be
a reliable indicator of soil type or site potential. The output
from the ground survey is a classification of types of caatinga,
a thorough description of each type, and a map showing the
distribution of the types. Preliminary data analysis indicates
approximately ten distinct types of caatinga in the area
studied; cluster analysis and ordination techniques have yet
to be applied in the data to give this classification an objective
It would appear from work completed thus far that there
are very different types of native caatinga within the relatively
small study area, with contrasting vegetation mixtures and
productivities. They reinforce the importance of having an
accurate picture of the resource base before trying to
extrapolate site-specific research data to a general area.
The ecological data indicate an apparent incongruity for a
semi-arid region: vegetation characteristics are strongly
influenced by the incidence of excessive water in the soil
profile during the wet season, which may override differences
in soil parent material. Hence, soil drainage features and
topographic location are key discriminatory criteria in a
range site classification system. The small elevational
increase of 15 cm in an otherwise poorly drained site is
sufficient to permit the establishment of a patch of
Marmeleiro (Croton hemiargereus), where otherwise no
perennial would grow.
Different sites produce different kinds of regrowth
following clearing and burning, with different subsequent
changes in the vegetation as the woody species regain
An early successional phase which follows cutting and burning consists of a
sparse, patchy community of Mufumbo and Catingueira (as shown above)
with poor productivity. Photo: J. Malechek.
dominance. The intensity of cutting and burning on one site
will determine the dominant regrowth species; excessive and
indiscriminant treatment, for example, will encourage
dominance by Mufumbo (Combretum leprosum) which has
no forage value for livestock.
Brazil does not have the trained manpower to carry out
such an intensive ground survey over the entire Northeast
region, nor would it be an economically wise enterprise given
the capability of existing remote sensing technology. The
second phase of the range resource survey will focus on the
analysis of radiation from half hectare units or "pixels"
recorded by LANDSAT monitoring satellites, relating the
radiation data to the ground survey. By integrating ground
survey data and the satellite data, each type of caatinga can
be represented by a unique spectral signature. The advantage
of LANDSAT technology is that once the spectral signature
has been defined, it is a simple computer step to print a map
of all the pixels which match the signature and calculate the
total area of the type. A comparison of the computer-
generated resource map with ground sites to check for
accuracy can provide feedback to refine the signatures. The
resulting resource survey, limited to a small area of the
Northeast, is intended as a trial of the methodology with the
hope that procedures, if successful, will be adopted by a
Brazilian agency concerned with resource management in the
Diets and Nutrition of Animals on Uncleared Caatinga
By far, the most common animal management practice in
the sertao is grazing of mixed flocks of sheep and goats on
native, uncleared caatinga rangeland. Before changes in such
management can be recommended, or before advanced
research can be planned, it is critical to understand this
traditional form of grazing management. Responses of
animals and forage supplies to the highly erratic wet-
season/dry-season precipitation regime are of particular
importance. This information is important in explaining
nutritional conditions and in management of the range for
forage species that are palatable to livestock.
To gain insight into these questions, Utah State researcher
Jim Pfister stocked an area of typical caatinga rangeland
with a mixed flock of sheep and goats and measured forage
availability, animal weight responses, forage intake, and
nutritive and botanical characteristics of diets on a seasonal
basis. Animals with a surgically prepared esophageal fistula
provided representative samples of grazed diets, by season,
for laboratory analyses.
Green grasses, for example, and the herbaceous annual
forb Bamburral were found to be important foods to both
sheep and goats during the wet season; sabia', a tree species,
was selected by goats but not by sheep. As the dry season
began, the vine Jitirana was avidly selected by both sheep and
goats and continued to be eaten even into the late dry season.
Fallen leaves from woody species became more important as
the dry season progressed and herbaceous species were no
longer available. These species included Catinguera,
Mororo', and Marmeleiro. Interestingly, Pau Branco, the
dominant tree of the area, was never an important forage
From a nutritional standpoint, diets were of higher quality
than expected. Crude protein in diets (Figure 1), an
important nutritive index, was observed to remain high
throughout the year. Values never fell below about 12%, even
near the end of the dry season. The National Academy of
Science's recommended level of crude protein for goats in
late gestation is about 13 percent. Subsequent lab analysis of
some of the species that were highly selected by grazing
animals (Table 1), confirmed that even fallen leaves retain
relatively high crude protein levels into December, the usual
end of the dry season.
These findings, although preliminary, indicate that protein
may not be as great a limitation as anticipated on native
caatinga range where a mixture of herbaceous and tree
species are available. They also offer a preliminary guide for
further work on selective clearing of trees that are not eaten
by small ruminants. Work is in progress on further
Related research is being done by Pfister on theoretical
aspects of feeding behavior in sheep and goats. This is an
attempt to explain the reported greater survivability of goats
over sheep during periods of extreme nutritional stress such
as that imposed by extended drought. The approach involves
measurement of the amount of time spent by sheep and goats
at a feeding station (the hypothetical semi-circle of forage
within reach of a feeding animal at a particular position).
During dry season months when both species were
consuming fallen leaves, no species differences existed;
however, as the wet season progressed and goats began
feeding from the tops of tall (> 2m) annual forbs and from
overhead branches, their feeding station times tripled.
Foraging time is an important component of ruminant
feeding strategies and the bipedal stance often assumed by
goats seems to be costly in time. The nutritional pay-off of
this behavior will be determined.
MAY JUNE JULY
AUG SEPT OCT DEC
Figure 1. Seasonal trends in dietary crude protein for sheep and goats feeding on native caatinga range.
JAN FEB APR
CRUDE PROTEIN (%)
= 95% C.I.
. 1 \ \ ~~ ~~ \~ ~\
Table 1. Dry Season Crude Protein Values (%) of Important
Sept. Oct. Dec.
Mororo 21.2 -
Sabia 19.7 -
Marmeleiro 19.3 15.9
Dried leaves on stand-
ing dead Jitirana 15.6 -
Mororo 5.2 6.1 -
Sabia 10.5 11.1
Catingueira -11.4 11.1
Marmeleiro 12.6 9.5
Land clearing serves two apparent purposes: it provides
saleable wood products (firewood, fenceposts) and is thought
to increase forage production by annual grasses and forbs
through removal of competition by trees and shrubs.
However, the widespread and common practice raises
numerous issues including questions such as:
* What is the extent of the increase in forage yields due to
clearing and how many years does the effect last?
* How are animal diets, nutrition, and animal production
* Do different seasons and methods of clearing change the
* What are the long-term effects on soil productivity,
erosion, and ecosystem stability?
Utah State researchers Robert Kirmse and Fred Provenza
began an initial study of these questions in 1981. They have
compared forage yields, forage composition, and animal
diets and nutrition on paired cleared and uncleared sites.
Additionally, rates of forage decomposition are being
measured to determine how grazing the flush of herbaceous
vegetation produced after clearing might be deferred in order
to serve as a standing forage reserve for the late dry season
when traditional feed supplies are nearing exhaustion.
In related research, Utah State researcher Linda Hardesty
is studying the response of selected important woody species
to various times and methods of cutting. This research
should indicate how best to control, by cutting, a particular
species that presents a noxious brush problem.
We have also observed that some tree species cut at certain
seasons tend to retain leaves on the re-sprouted new twigs
and branches well into the dry season, whereas the uncut
parent trees drop their leaves in the regular dry-season
deciduous pattern. This raises the possibility of using
selective cutting of certain species to prolong, in effect, the
green period of available forage. However, we are only now
beginning our investigations into the acceptability and
nutritional value on these re-sprouts.
Initial results indicate an unquestionable "release" of
herbaceous ground-level annual vegetation from compe-
tition by over-story trees during the first growing season after
clearing. Figure 2 vividly demonstrates this effect, but it
should be pointed out that much of the increase in
production is in the form of coarse stemmy material that is
UC C UC C UC C UCC UC C
J F M A M J J
UC C U C UC C UC C
S 0 N D
Figure 2. Monthly standing crop of herbaceous annual vegetation on cleared and uncleared caatinga ranges. Much of the yield on cleared land is in theform of
plant stems (1982).
" "``' '` '"' "~" "` ~ ""'' ""' ~
not consumed by either sheep or goats. Fallen leaf litter is
abundantly available throughout the dry season on un-
cleared land while the remaining material from herbaceous
species on cleared land is in a declining phase throughout the
dry season. This raises the prospect of selective clearing or
strip clearing to provide a boost for production of annual
herbaceous species during the wet season, but would also
retain important tree species that yield nutritive and
abundant forage for use in the late dry season. Fallen leaves
from woody species also provide a valuable ground cover
that is presumably important in erosion prevention when the
first intense rains occur.
Initial results on decomposition studies show a marked
difference between rates for leaves from trees, grasses, and
forbs. Forb leaves decay rapidly, tree leaves very slowly, and
grass leaves are intermediate. The potential nutritional
consequences of this to sheep and goats is still unknown, but
is being investigated.
Significance of Research Findings
Initial computer runs of the newly-completed farm-
systems model constructed by Nestor Gutierrez of the
Winrock Economics Project point to the significance of
range related research for the sertao. Model output indicated
differences in farm income between "average" years and
"poor" years amounting to 85%. The major contributor to
this difference was corresponding variation in forage supply.
To put this in perspective, increasing small ruminant natality
rate from 70% to 100% (through model manipulation)
increased farm income by only about 3 percent.
Thus, if guiding principles of systems analysis and
computer modeling are followed, the importance of the
range forage resource must be emphasized, both in practical
management terms and in allocation of scarce research
resources to high priority problems.
The research of the Range Management Project is
highlighting the management activity which has the highest
potential for affecting forage supplies of the Northeast:
caatinga clearing and manipulation. For example, our initial
study of forage production and animal nutrition on cleared
land indicates a 6-fold increase in yields of herbaceous annual
species during the first year after clearing. However, if a
major decline occurs 3-5 years after clearing (as we are
hypothesizing, based on results from other semi-arid tropical
regions), a very close examination of marginal costs and
marginal benefits is called for. A less radical type of treatment
is already designed for possible research where well planned
strip clearing would be substituted for the traditional "clear-
cut" approach. This would theoretically give the short-term
benefit of increased herbaceous annual forage production
while maintaining the nutritionally important and pre-
dictable tree and shrub species that apparently modulate the
vast swings between wet and dry seasons and normal years
and drought years.
The practical appeal of this research approach, aside from
being biologically important as revealed by modeling results,
is that we are dealing with modification of an existing local
technology. As mentioned earlier, land clearing is an integral
part of smallholder activities in the sertao, and is practiced
using local labor during seasons when farmers are not
otherwise occupied. It is even subsidized indirectly through
certain government incentive programs.
Bearing in mind the considerable current and long-range
economical and ecological impact that land clearing can
have, it is vitally important to know as much about it as
possible, with all of its ramifications for animal nutrition and
production before making recommendations to change
traditional smallholder sheep and goat production practices.
Researchers emptying feces from the collection bag on a sheep. Forage
intakee an e determined indirectly bv knowing the daily quantity of feces
and the digestibility of the diet consumed Photo: J. Malechek.
Future Research Directions
With only just more than two years of field data and
related lab work completed, we do not yet know enough
about land productivity capabilities, seasonal and yearly
forage supplies, land responses to clearing, and animal
nutritional characteristics to make major changes in the
overall direction of our research. Range research is slow in
yielding definitive results because of the profound effects of
yearly climatic variations. For example, we are already
seeing a 5-fold difference in forage yields between 1981, a
"poor" year (519 mm precipitation), and 1982, an "average"
year (715 mm precipitation). With such major differences
between two successive years that were not extreme in total
precipitation levels but showed contrasting patterns of
rainfall distribution, it is obvious that predictions and
recommendations would be no more than speculation,
unless backed by a run of several years of data that embraced
a wider span of variability. Thus, our first priority is to
continue research now in progress so that we can witness
more of the extremes that are characteristic of the Northeast.
Another major priority will be better integration with the
Winrock Economics Project and the North Carolina State
Nutrition Project. In the first case, the Winrock production
model will be the focus of collaboration. At present, the
model does not consider ecological variability in land
production characteristics. Although this component can
easily be added to give the model greater regional predictive
power, additional field work will be required to establish a
quantifiable relationship between soil types and vegetation
on the one hand and composition and productivity of farm
livestock herds on the other hand. We will undertake this
work in 1983-1984.
Another relationship to the Winrock Economics Project
will be established through joint analysis and evaluation of
data from on-going grazing trials being conducted by
collaborating Brazilian scientists. These trials are being
conducted in two areas in Ceara State; at Taua, under the
supervision of Dr. Joao Ambrosio of the Federal University
of Ceara, and at Quixada, under the supervision of the state
research agency EPACE.
Dr. Ambrosio will be spending a sabbatical year at Utah .
State in 1984 and will have the opportunity of collaborating
with Winrock and Utah State scientists on his research. Plans
are also underway to bring one of the EPACE scientists to
Utah for graduate training at the PhD level. He would
evaluate grazing trial data from the EPACE experiments as
part of his thesis research, under the joint supervision of Utah
State and Winrock scientists.
Closer relationships will also be pursued with the North
Carolina State Nutrition Project investigators, in their efforts
to evaluate nutritional characteristics of individual native
tree and shrub species and to understand the mineral Brazilian workers construct goat-proof fences for grazing comparisons
tree and shrub species and to understand the mineral between undisturbed vegetation and the regrowth after clearing and
nutrition of sheep and goats. burning. Photo: W.C. Weir.
Table 2. Degree-Related Training as a Part of the Utah State University Range Project in Brazil
Degree Sought Type of Location and
Name Nationality or Obtained Support Area of Study Duration of Study Present Position
James A. Pfister USA PhD F1 Range Animal Brazil (1.5 yrs) Grad. Research
Nutrition Assistant, USU
Robert Kirmse USA PhD F Range Ecology Brazil (3 yrs) Grad. Research
Joao S. de Queiroz Brazil PhD F Range Ecology Brazil (1 yr) Grad. Research
Linda Hardesty USA MS F Range Ecology Utah (I yr) Research Associate,
PhD F Brazil (2 yrs) USU
Roberto Cesar Brazil MS P2 Range Animal Utah (I yr) Grad. Research
Magalhaes Mesquita Nutrition Assistant, USU
F designates full financial support (assistantship stipend plus research operating costs) from SR-CRSP funds.
2 P designates only partial support (usually research costs) from SR-CRSP funds.
Table 3. Non-Degree Training as a Part of the Utah State University Range Project
Name or Group Instructed Number Instructed, if Group Type of Training Location Length of Training
Brazilian scientists from 25 Shortcourse on Brazil 3 days
EMBRAPA and affiliated methods of
research organizations measuring caatinga
Brazilian scientists from 12 Shortcourse on Brazil 2 days
EMBRAPA and related techniques of
research organizations animal diet analysis
Sandra Mara, graduate Diet analysis; Brazil 9 months
student, Federal vegetation sampling
University of Ceara' procedures
Vanceslau Alves Costa, Care and husbandry Brazil 18 months
Field worker/animal of fistulated animals;
caretaker sampling procedures
Edith Oliver, BS, Vegetation sampling Brazil 4 months
Texas A&M Univ. methods
Daniel Hardesty -Ecological sampling Brazil 9 months
The range management group has adopted the position
that trained scientists are the most valuable and enduring
"products" of a research project for developing nations.
Thus, we have chosen to invest project dollars in advanced,
highly qualified graduate students, rather than in placing
PhD-level scientists overseas.
Our emphasis has been on US nationals as well as
nationals from Brazil and other developing countries. We
anticipate that the US students will find careers at American
universities and other institutions engaged in international
agriculture and resource management. International exper-
ience gained in these graduate programs, including language
training, will enable them to step directly into their
professional positions with only minimal adjustment. Like-
wise, we expect our students from developing countries to
return to their respective countries to initiate their own
research and educational programs, adopting techniques,
philosophies, and approaches gained in association with
their graduate work.
We have also engaged in shortcourse training for
professionals and informal on-the-job training for several
Brazilian personnel. Tables 2 and 3 present a summary of our
Hardesty, L.H., and T.W. Box. 1982. Developing Women for International
Service: A Case Study. Title XII Conf. on Women in Intl. Dev.
Kirmse, R.D., J.A. Pfister, L.V. Vale, and J.S. de Queiroz. 1983. Woody
Plants of the Northern Ceara' Caatinga. SR-CRSP Technical Report
Series No. 14. 49 p.
Kirmse, R.D. and F.D. Provenza. 1982. Herbage Responses to Clearcutting
Caatinga Vegetation in Northeastern Brazil. First Brazilian Symposium
on the Semi-Arid Tropics. Recife, Pernambuco, Brazil.
Malechek, J.C. 1982. Grazing Management of Goats in Extensive
Rangeland Production Systems. Proc. III Intl. Conf. on Goat Prod. and
Dis. Tucson, AZ. pp. 404-408.
Malecheck, J.C. and F.D. Provenza. 1983. Nutrition and Feeding Behavior
of Goats on Rangelands. World Ani. Rev. 41: (In Press).
Malechek, J.C. and F.D. Provenza. 1981. Feeding Behavior and Nutrition
of Goats on Rangelands. Proc. Intl. Symp. on Nutrition and Systems of
Goat Feeding. Tours, France. pp. 411-428.
Pfister, J.A., J.C. Malechek, and E.A. Lopes. 1982. Dry Season Diets of
Goats and Sheep Grazing Native Caatinga Range. First Brazilian
Symposium on the Semi-Arid Tropics. Recife, Pernambuco, Brazil.
Pfister, J.A. and J.C. Malechek. 1983. Diet Selection of Goats and Sheep
Grazing Caatinga Range in Brazil. Abstracts of Papers, 36th Ann. Mtg.,
Society for Range Mgmt., Albuquerque, NM.
Pfister, J.A., J.S. de Queiroz, R.D. Kirmse, and J.C. Malechek. 1983.
Rangelands and Small Ruminant Production in Ceara' State, North-
eastern Brazil. Rangelands 5: (In Press).
of Small Ruminant
University of Missouri, Columbia
The goals of the Sociology Project in Brazil were
twofold: to develop an understanding of the role of
small ruminant production within the total production
system and identify the likely constraints to increasing
production; and, to demonstrate to EMBRAPA the
contribution sociologists can make to an integrated multi-
disciplinary program designed to increase small ruminant
Northeast Brazil has a much more complex agricultural
system than one might guess at first glance. Although it is
classified as a semi-arid area, crop production is nonetheless
important and animal production has to be integrated with it
on most farms. In addition, there are substantial year-to-year
variations in rainfall which indicates that producers have
developed an ability to adapt to changing conditions. The
Sociology Project has tried to develop an understanding of
the dynamics which influence shifts between crop and animal
production, and, within the animal area, the shifts between
large and small ruminant production and how climatic
variation influences both of these. Superimposed on this is a
complicated land tenure situation which is characterized by a
considerable amount of absentee land ownership and share
cropping arrangements. An additional goal has therefore
been to understand how the land tenure situation might
influence increased production of sheep or goats.
The first year's activities in Brazil (1981-82) focused on the
State of Ceara. The objectives of this effort were to:
* Develop a comprehensive view of the total production
system of agriculturists in the sertao of Ceara.
* Understand the production decision framework of
farmers and in particular the decision rules governing
small ruminant production within this total system.
* Develop a sociological analysis which would complement
the economic analysis done with the same sample.
The 1982-83 activities had three objectives:
* To study the role of goat production within the total
production system of small producers in northern Bahia.
This was to be done much in the same manner as was
accomplished in Ceara the previous year.
* To evaluate the possibilities and sociological conse-
quences of introducing dairy goats in the sertao of
* To examine the biological performance and sociological
responses to the introduction of new management
techniques of goats among small producers in a micro
region of Paraiba.
Given the nature of the Northeastern Brazil agricultural
system, the potentials for error in attempting to improve
small ruminant production through a uni-dimensional
strategy of biological intervention are quite high. Without a
thorough understanding of the social organization of
production and the role that small ruminants play on
"typical" farms, there is a strong possibility that proposed
interventions will either not be adopted by farmers,
accomplish nothing, or help only the relatively well-off
In Northeast Brazil, goats are generally able to withstandperiods ofdrought
better than cattle. Photo: G. Primov.
From the interviews conducted with the sample of small
ruminant producers in Ceara, a technical report was
prepared (Primov, 1982) which contained a number of
interesting insights. However, overshadowing all else is the
observation that the total production system is highly
variable and greatly influenced by rainfall. Small ruminant
production has not developed as a form of specialized
commodity production and therefore it would not be
expected that it would respond significantly to market
incentives. For most farmers, the relative importance of
small ruminants within their total production strategy tends
to vary considerably between wet and dry years. During dry
years, small ruminant production becomes much more
important since animals serve as both a subsistence and cash
crop. During wet years, cotton and cattle are seen as much
more profitable while small ruminants become more
expendable. Marketing small ruminants is largely deter-
mined by the financial needs of the household rather than
external market conditions, and increases in price and/or
demand are therefore unlikely to stimulate production.
Producers, particularly the smaller land owners, cannot
increase production without sacrificing the flexibility they
need in order to cope with alternating dry and wet years.
Since producers are confident that they can market whatever
animal they produce, there is little incentive to upgrade
animals for marketing purposes. The very poorest sector of
the rural population who are principally landless producers,
raise few animals and do not have many opportunities to
increase their livestock production. Of the smallest stratum
of landowners, 25 hectares or less, many avoid goat
production entirely because these animals require either
constant supervision or fencing. This is in spite of the fact that
the land owners are quite aware that goats survive droughts
better than sheep.
On the basis of these findings it is recommended that the
SR-CRSP in Brazil should avoid encouraging small
ruminant producers to specialize in small ruminant produc-
tion nor should they be encouraged to significantly increase
the size of their herds. Since these animals represent a vital
component in the producers' abilities to respond and adapt
to changing environmental conditions, specialization in any
commodity production, including sheep or goats, is not
recommended. The producers perceive their greatest prob-
lems as being in the area of improving herd productivity and
in health related problems. It is recommended that research
efforts be directed toward overcoming these constraints.
The second major activity which was undertaken in the
State of Bahia is still incomplete. The objective of this project
was to understand the role of goat production within the
total production system and to compare small ruminant
production in sertao Ceara and in northern Bahia. The
project also sought to identify for other projects (biological
and economics) the most important problem areas of goat
production as identified by the producers. This project
involved interviewing a random sample of about 50 small
producers in four municipios in northern Bahia. Preliminary
data indicate that although small ruminant production in
northern Bahia is more important to producers than it was in
Ceara, it is less important than expected: differences in the
production system between Bahia and Ceara are differences
of degree rather than of kind. In Bahia, water shortages are
by far the most important problem in production. Since
shortages of water also lead to shortages of forage,
inadequate nutrition is a serious issue as well. Health
problems did not appear to be as severe as they were in
The third research activity was the study of dairy goat
producers in Paraiba. This project sought to develop a
general description of the local dairy goat production
systems and to identify problem areas which might help
direct future research efforts. This project, involving inter-
views with 54 goat producers in two counties in Paraiba,
produced several significant conclusions. First, production
systems were characterized by a high degree of inter-
dependence and interaction between livestock and agri-
cultural production. While goats and goat milk were not
important staples for home consumption, goats were the
most reliable source of cash income. In general, the dominant
management techniques for goats involved minimal labor
and capital investments. The main constraints to increasing
goat milk production in this area were a lack of financial
resources on the part of the small producers, a lack of water
during most of the year, the deterioration of the caatinga
during the dry season, animal health problems (particularly
endo and ecto parasites), and inadequate extension services.
The data from this study suggested that because most
producers cannot depend on a regular cash supply, research
priorities should involve the development of technologies
which require a minimum amount of cash and fixed capital
investments. Moreover, because most producers owned only
native breeds of goats, research on these species and the use
of the native vegetation seems to be crucial at this time.
Because the technology of dairy goat production was not well
known and because the small producers lacked very basic
knowledge of techniques and the minimum capital necessary
to improve goat milk production, a much closer relationship
between researchers and extension workers will be necessary
in order to achieve any success in this area.
The objective of an additional study, done in cooperation
with the Texas A&M Management Project and completed in
April 1983, was to monitor the changes in various elements of
the small farms where improved dairy goat management
techniques, primarily supplemental feeding, were initiated.
The sociological portion of the project concentrated on
assessing the labor requirements of the recommended
practices and their impact on overall labor requirements of
the farm. Particular attention was paid to the division of
labor within the family, monitoring the dynamics of the
production system for alterations after the new practices had
been introduced, assessing the actual expenses associated
with the new management practices and comparing them
with the increased monetary value of milk and cheese output.
This study also offered the opportunity for assessing the
likelihood that these producers would continue the new
practices during the next dry season. The research was
designed as a pilot project and involved five producers who
were selected on the basis of professed interest in improving
goat milk production and who had the necessary infra-
structure to implement and continue improved production
Preliminary data indicated that even though the labor
requirements for goat production on these farms had
doubled after the project started, it had not become an undue
burden for the five families involved. All families were able to
keep up with all the other farm activities. Livestock
production in general, and goat production in particular,
were traditionally male tasks. Women were only partly
involved in the feeding and caring of the goat kids but were
the key actors in processing goat milk into cheese. Only the
largest of the five producers indicated that he would be able
to feed the concentrate ration during the next dry season.
Another producer said he would give the does and kids at
least some kind of concentrate. Two producers stated they
could not afford the ration during the dry season.
Goat milk can provide a valuable addition to the household diet; it is
produced in convenient amounts and children can provide the labor for
herding and milking. A young girl is shown milking a doe near Juazeiro,
Bahia. Photo: G. Primov.
Significance of Research Findings
All of the studies undertaken in Brazil to date point to one
overall conclusion. In spite of the fact that small ruminant
production in Northeast Brazil primarily takes place within a
grazing system that is range oriented, crop production plays a
very important role in the overall farming system. Any
attempt to improve small ruminant production without
taking into consideration the parallel cropping system and
the dynamics of the interchange between animals, crops and
climate, is probably doomed to failure. For the small and
medium sized land owners, small ruminants provide a source
of flexibility and insurance for times when rains and crops
fail. Given this system, producers are not going to be willing
to make major investments in their small ruminant enterprise
nor to specialize in small ruminant production. In the context
of the Northeast with its highly variable rainfall, special-
ization in the production of any one commodity is probably
close to committing economic suicide. Thus, it is imperative
that the Range and Nutrition projects in particular examine
the cropping aspects of the farming system in order to see
how their recommendations would fit within that context.
Future Research Directions
By September 1983, the Sociology Project will have
provided to the other programs in Brazil a good baseline
description of the significant factors involved in farmers'
decisions concerning small ruminant production in North-
east Brazil. While a number of the issues mentioned in the
research accomplishments section could be studied in
considerably more detail and across a considerably broader
cross section of the Northeastern population, taken together,
the reports that are already published or are planned
constitute a good reference for those projects still active in
Brazil. It is hoped that future work can be undertaken
particularly in terms of monitoring the results of tech-
nological/managerial interventions proposed by the bio-
logical projects. Future efforts could also profitably look at
the effectiveness of the extension service and the access
farmers have to credit.
One Brazilian student has received graduate training at the
University of Missouri in the Department of Rural
Sociology. This student, Marisa C. Neumaier, completed the
requirements for her Masters degree in Rural Sociology in
September 1983. Ms. Neumaier is a staff member of the
soybean center in the EMBRAPA structure. It is hoped that
if the appropriate administrative details can be worked out,
she will be able to maintain at least minimal contact with the
goat research in Northeast Brazil in the future.
Neumaier, Marisa C. 1983. Sociological Aspects of Goat Production in
Paraiba, Brazil. MS Thesis. Department of Rural Sociology, University
Primov, George P. (In Press). Small Ruminant Production in Bahia, Brazil.
SR-CRSP Technical Report Series.
Primov, George P. 1982. Small Ruminant Production in the Sertao of
Ceara, Brazil: A Sociological Analysis. SR-CRSP Technical Report
Series No. 15.
Small Ruminant Flock/
Herd Health Program
in Smallholder Systems
University of California, Davis
It is generally recognized that the major constraint to
livestock production in the sertao, the desert area of
Northeast Brazil, is the uncertain feed supply. The next
major constraint in obtaining maximum production is
disease. The entire gamut of etiologic agents including
bacteria, viruses, parasites, exogenous toxins, and nutritional
and metabolic imbalances are known, or more often
suspected, to cause disease in the region. While it is generally
true that maintenance of a good nutritional plane enhances
resistance to disease, it must also be recognized that disease
can lessen an animal's competitive vigor and instinct for
survival under the stressful conditions associated with limited
feed supplies and harsh climates. Maintenance of adequate
nutritional levels and reduction of the disease burden are
obviously complementary to one another in maximizing
The major goal of the Animal Health Project in Brazil is to
increase the productivity of goat producers by developing the
means for controlling diseases that are major problems to the
sertao. The initial approach was to identify the major disease
problems and to set priorities for disease research. In many
cases this has involved confirmation of the existence and
establishment of incidences and prevalence of conditions
reported or suspected to occur largely on the basis of
anecdotal information. The selection and pursuit of research
projects has been made on the basis of the. perceived
importance of the disease process and of the feasibility of
performing the research needed to reach the eventual goal of
disease control. Such work must necessarily include devel-
opment of a basic understanding of the diseases involved and
relating that understanding to conditions in the field. The
most obvious need in the pursuit of these objectives at the
Brazilian National Center for Goat Research (CNPC) has
been the development of laboratory facilities and technical
support to carry on such work.
The project has identified intestinal parasitism, caseous
lymphadenitis and pneumonia as disease processes in which
research projects can be pursued with reasonable expec-
tations of progress at the CNPC. Collaborative projects have
been established with animal health scientists at the CNPC
within the framework of the Brazilian national agricultural
research enterprise (EMBRAPA). In part to service these
projects, but more importantly to establish a continuing
research resource for the CNPC, the project has provided
support for laboratory facilities and training of technical
In addition, veterinary consultation and collaboration
have been provided to other SR-CRSP projects at the
CNPC. It is the intent, in part accomplished, to integrate the
animal health research projects with other SR-CRSP
projects by collaborating with SR-CRSP scientists where
animal health could have a meaningful effect on the results.
A senior investigator, D.E. Hansen, was stationed in
Sobral for two years with the major responsibility of
determining the nature of disease problems in the sertao. A
survey was made to determine the prevalence of various
diseases among 31 farms and the incidence of specific
diseases of small ruminants on each of these farms in the state
of Ceara, Brazil. Prevalence and incidence were recorded
during scheduled visits to 21 farms, for which the animal
health survey was the primary purpose of the contacts, and to
10 farms on which the veterinary scientist collaborated with
SR-CRSP investigators from the Economics, Nutrition and
Reproduction projects. The primary purpose was to
determine which diseases were major economic constraints
in the production of meat and hides. The findings of the
survey which ran from July 1981 to June 1982 are presented
in Table 1.
Table 1. Prevalence and Incidence of Disease in Sheep and
Goats on 31 Survey Farms in the State of Ceara in
No. of Farms % Farms
Disease Affected Affected %
Mortality of 31 100 8.5-27
(< 60 days of age)
Parasitism 19 61 -
Abortion 15 48 8.2-51.3
Photosensitivity 9 29 22-60
Pneumonia 9 29 3-43
Lymphadenitis 8 26 5-18
Mastitis 3 9.7 20-25
Diarrhea 20 65 -
Other diseases were observed but not included because of
uncertain diagnoses or too few numbers to be significant.
These included contagious ecthyma, foot rot, birth defects,
ectoparasites and other skin diseases. Plant intoxications
occur in a highly seasonal pattern.
It should be noted that Ceara is on the fringe of the major
goat producing area of Brazil. Some of the diseases which are
considered to be of little importance by producers in Ceara,
such as caseous lymphadenitis, are of major concern to
producers in areas of more intensive goat production.
An analysis of the specific management practices was
made on the 31 ranches surveyed to determine the possible
reasons for the wide variation in incidence of mortality (8.5-
27%) in newborn animals. The analysis demonstrated that
four practices which in combination had the most beneficial
effect in lowering mortality were, 1) treating naval cords at
birth, 2) regular cleaning of the corrals and houses
throughout the year, 3) separation of the dams and newborns
from the rest of the herd, and 4) cleaning and disinfecting the
corrals and houses prior to parturition.
Brucellosis, the cause of undulant fever in man, is
considered to be prevalent in Brazilian cattle. It was therefore
thought to be a likely cause of abortions in sheep and goats
and a possible hazard to human health. To determine the
likelihood of this danger, a serologic survey for evidence of
infection by Brucella abortus in sheep and goats was carried
out in more than 7,000 animals. Positive titers were found in
only 0.21% of the animals tested. This suggests that the health
hazard to either animals or humans from this disease in sheep
and goats is almost negligible under present conditions.
To determine the type of forage utilized by sheep and goats
in the sertao, range management specialists required the
surgical preparation of an esophageal fistula from which
ingesta could be collected. Dr. Hansen provided surgical
service and training for the range management technicians
and developed a new surgical fistulation technique. This
technique greatly improves the feasibility of the collection
procedures by reducing the problems of maintaining the
fistula and thus prolonging its useful life.
Several long-term projects have been established to study
the dynamics of intestinal parasitism in sheep and goats in the
sertao. A major problem being addressed is determining the
chronology of contamination, exposure, infection and onset
of clinical disease which is critical in the development of
control programs. The severe climatic and nutritional
variations in the region tend to mask the effects of parasitism
and necessitate the use of indirect methods to ascertain stages
and levels of the infection cycle. The general procedure used
is to expose groups of susceptible kids, which have been
raised free-of parasites, to infected premises for 90 days after
which they are killed and examined for evidence and levels of
parasitism developed during that interval. The findings of
multiple consecutive group exposures are then correlated to
seasonal and climatic variables.
Using a similar technique, the effects of housing goats on
slatted floors during the night on the level of parasitism has
been studied. The effect is considered to be negligible. The
same procedures will be used to complete a study on
suspected breed differences in susceptibility to parasitism.
Bacterial pneumonias have been demonstrated by gross
and histopathologic examinations to be the most common
cause of death and the major cause of condemnation at
slaughter of sheep and goats in the Sobral area. A
microbiology laboratory has been established by the Animal
Health Project at the CNPC. Dr. Mary Sawyer has been
placed in the laboratory as the Senior Investigator with the
primary responsibility of organizing a functional laboratory
and training her Brazilian replacement, Dr. Cristine Almeida,
Collecting a blood samplefrom a kidfor a serologic test. Photo: H. Olander.
who has been hired by the project for that purpose. This
laboratory will enable CNPC scientists to determine the
prevalent etiologic agents associated with the common
The laboratory will also assist in a study of the etiology of
mastitis in a dairy goat operation which is managed by the
SR-CRSP Nutrition Project in the state of Paraiba.
A major project has been established to study caseous
lymphadenitis. This disease, which is characterized by
abscessed lymph nodes, occurs throughout the world and is
prevalent in sheep and goats in the sertao. Preliminary
clinical and cultural studies have indicated an incidence of
abscesses as high as 20% in some herds near Ceara, but
organisms other than Corynebacterium pseudotuberculosis
were isolated in 50% of abscesses cultured. These studies have
indicated that lymphadenitis is less prevalent in Ceara than in
the rest of the Northeast. Field studies are now underway to
determine the level of subclinical infection and the age of
natural infection of kids in herds with a high incidence of
disease. The general procedure is to determine the age of
serologic conversion of kids indigenous to herds with a high
incidence of disease and of susceptible kids newly exposed to
those herds and to correlate this with pathologic and
bacteriologic findings. The microbiology laboratory at the
CNPC is providing the necessary laboratory support.
Significance of Research Findings
The research that is in progress is addressing the need for a
general understanding of the dynamics of disease processes in
tropical desert regions. To the extent that these studies are
successful, they will help to reach the overall objective of the
Animal Health Project which is to increase small ruminant
production through the control of disease. Application of
rational control measures requires an understanding of the
basic mechanisms by which disease agents thrive or are
suppressed in a given environmental situation. The inter-
action of disease control with other aspects of livestock
management in increasing productivity has been stated in the
project goals. However, specific interaction between the
Animal Health Project and other SR-CRSP projects has
occurred with regard to improved health of test animals.
Examples are the improved surgical technique used in range
management testing and improved sanitation measures
associated with the goat mastitis study in Paraiba which
should reduce the level of disease and its associated
production variables. Further, the studies on the dynamics of
parasitic and microbial infections which are in progress
should provide the impetus and example for changing
management practices to improve animal health and
A finding of some import regarding human health has
been the very low incidence of Brucella abortus infection in
sheep and goats in the sertao, suggesting the relative safety of
these animals as suppliers of meat and milk under present
circumstances. It also underlines the need for care in
exposing highly susceptible sheep and goat populations to
Perhaps the most important effect of the project's research
will be the establishment of a functional microbiology
laboratory to support investigations and research on
infectious diseases at the CNPC. Such a facility will provide
an important aid for recognizing, monitoring and controlling
disease processes which could directly effect the outcome of
research in other projects at the Center as well as the
productivity of the sheep and goat industry of the sertao.
Future Research Directions
The most pressing need of the Animal Health Project in
Sobral is the establishment of a permanent microbiology
laboratory to support research on infectious diseases. The
current work on pneumonia, caseous lymphadenitis and
mastitis is providing the impetus for the development of such
a capability. In the initial stages, these projects will provide a
sufficient workload for the development of facilities and
techniques and training of personnel. As the capability of the
laboratory matures, research should be pursued in the areas
of abortion and neonatal death which previous work has
indicated to be major constraints on goat and sheep
The project has attempted to provide a means for
exchange of graduate level training between the CNPC and
UC Davis with the goal of establishing a long term liaison
between senior scientists and the developing staffs of both
institutions. It has also attempted to support on-the-job
training of technical help for Brazilian counterparts. A major
endeavor in this direction involves the development of a
microbiology facility and the concurrent training of a
Brazilian microbiologist as described below. In addition,
discussions have been held with CNPC administrators
concerning the presentation of an international conference
on caseous lymphadenitis to be held in Brazil. This will add
impetus to the work on the project and will provide a
valuable learning experience for project scientists.
Graduate training programs which are in progress or have
been arranged follow.
Terezinha Padilha-Charles, DVM, MS, began her train-
ing for a PhD in parasitology at UC Davis under Dr.
Norman Baker in 1982. Her work is to be completed in 1985.
Carlos Costa, DVM, MS, is scheduled to come to UCD in
the Fall of 1983 to work towards a PhD in parasitology
under Dr. Baker.
Janete Santa Rosa, DVM, MS, has been tentatively
proposed to begin work toward a PhD in pathology at UC
Davis beginning in 1984.
Cristine Almeida, DVM, has been hired to work and train
with Mary Sawyer, PhD, UC Davis microbiologist, in
developing the microbiology laboratory at the CNPC. The
project supported an initial training period for Dr. Almeida
in an EMBRAPA laboratory in Rio de Janeiro under Dr.
Charlotta Langenegger. She will provide continuing direc-
tion for Dr. Almeida in working toward an MS degree in
Brazil. It is hoped that a PhD program will ultimately be
provided for Dr. Almeida at UC Davis.
Hansen, D., B. McGowan, and E. da Silva. 1983. The Development of a
System for Surveillance of Disease of Sheep and Goats on Small Farms
in Northeastern Brazil. SR-CRSP Technical Report Series.
Hansen, D., B. McGowan, N. East, E. da Silva, M. da Silva, J. Santa Rosa,
C. Costa, M. Sawyer, J. Sawyer, and H. Olander. 1983. The Prevalence
and Incidence of Diseases of Sheep and Goats on Selected Farms in
Northeast Brazil. (In Preparation for Publication as an EMBRAPA
Bulletin and in the Journal of International Sheep and Goat Research.)
Hansen, D., J.A. Pfister and J.C. Malechek. 1983. Surgical Establishment
and Maintenance of Esophageal Fistulae in Small Ruminants.
(Proposed for Publication as an EMBRAPA Bulletin and in the Journal
of International Sheep and Goat Research.)
Sawyer, M., M. da Silva, and J. Kleber. 1983. Control of Mastitis in Dairy
Goats. (In Preparation for an EMBRAPA Bulletin.)
Management of Sheep
Texas A&M University
The Texas A&M University component of the SR-
CRSP in Brazil was initially charged with responsi-
bility for breeding, and much of this report will relate to this
activity. However, beginning in the fall of 1982, the project
assumed responsibility for the Management component and
information on this aspect will be treated briefly.
Given the environmental and socioeconomic constraints
which characterize Northeast Brazil, the presence of adapted,
productive genotypes is an indispensable requirement for
efficient animal production. If these do not already exist, they
must either be introduced or created, based on breeding
programs within the country. The goals of the SR-CRSP
Breeding Project may therefore be stated as follows:
* To describe the existing breeds in order to facilitate
communication and reporting.
* To characterize the distinctive genotypes (or breeds) in
respect to their performance traits as a guide to
determining their utility in an overall breeding and
* To identify the environmental constraints to production
and the individual animal characteristics which contribute
to improved adaptation and productivity, with the view
that emphasizing some of these in selection might increase
efficiency of selection.
* To collect or obtain the necessary genetic parameters and
design and place in motion the indicated long-term
* To evaluate the potential gain which might be realized
from introducing selected exotic genotypes.
* To attempt to extract from the work in the US and Brazil
the maximum amount of information which may be
transferable to other areas or environments.
If each of these goals is pursued consecutively and carried
to near completion, many decades and extensive resources
would be required. To economize on time, they have been
pursued in parallel in this project and some progress on each
has been made. However, it must be realized that breeding
projects are of long-term duration and therefore expectations
of this project should be viewed in light of the short period of
time that work has been in place in the field. By contrast,
alteration of management practices offer the potential for
immediate improvement but require repetition with each
season or generation. In the near term, management, in the
broadest context, offers the only means of bringing about
quick improvement. Even in an adverse environment such as
One of the unique breeds ofgoats in Northeast Brazil is this research flock of
Moxoto goats at the CNPC, Sobral. Photo: M. Shelton.
Northeast Brazil, few problems of a technological nature are
encountered in suggesting practices which will bring about
improved productivity. However, serious socioeconomic
problems are faced when introducing new technology. Thus,
the overall goal of the management component of this
project is to ferret out those practices based on existing or
new technology which can be implemented or recommended
for implementation within the socioeconomic constraints
existing in Northeast Brazil and other LDC environments,
and which have a high probability of making a contribution
to improved productivity.
The possibilities for making contributions in this area
appear to be very good. However, as of this writing, the
Texas A&M University Management component of the SR-
CRSP Brazil program has been in place for only six months
and accomplishments should be viewed in this light. Also, the
state of Paraiba (EMEPA) has an active program to
encourage and assist producers in developing the dairy goat.
At the request of Brazilian collaborators, the management
component has placed a scientist in Paraiba to assist this
endeavor. The initial efforts have concentrated on nutrition
because this is an obvious limitation and is an area where an
early response might be obtained or where useful data may
be collected in a short period of time.
One of the first accomplishments of the project was to
provide descriptive information on the various genotypes of
sheep and goats found in the area with some preliminary
information on production conditions (Shelton and
Figueiredo, 1981). The sheep are generally described as
"Ovinos Deslanados," or without wool. However, the most
prevalent type is known as the "Crioula" which often have a
vestigal fleece cover that is not exploited commercially. In
addition to the "Crioula," the recognizable breed groups
include Morada Nova, Santa Ines, and the Brazilian Somali.
The first two are largely unique to Brazil, although they
resemble hair sheep found in adjacent regions including the
Caribbean. The Morada Nova is a small hair sheep which
has been in Brazil for many years. The Santa Ines is a larger
breed derived from crossing exotic types on the Morada
Nova and is distinctive by its size and long, pendulous ears.
The Brazilian Somali is descended from the Somali
Blackhead or Blackhead Persian which has been altered
through a long period in the tropics. In addition to the above
types, the Rabo Largo sheep, found in the more arid interior,
carries at least some infusion of fat-tail types introduced in
earlier years. The only exotic likely to be encountered in the
area is the Bergamacia, a wooled sheep of Italian origin.
The majority of the goat population is known as the SRD
(without well defined type). At least four recognizable local
types exist: the Moxoto, Repartida, Caninde, and Marota.
The most prevalent exotics are the Anglo Nubian and Bhuj
(of Indian origin), but a few individuals of the Alpine dairy
breeds (Alpine, Saanen, Toggenburg, etc.) can be found in
some areas or have been introduced on an experimental
A detailed study designed to provide comparative
information on the different breeds of sheep and goats is
being conducted by the CNPC (Sobral). Although pre-
liminary in nature, some of these data have been reported
and appear to provide a reasonable basis for conclusions.
In respect to sheep, some distinct breed differences are
evident (Figueiredo et al., 1983; Figueiredo et al., 1982b). The
Morada Nova is smaller in size with a higher lambing rate.
The Santa Ines is larger with a more rapid growth rate but a
lower level of multiple births. The Brazilian Somali is both
small and has a low-twinning rate but possesses survival
advantages under nutritional stress. At this point, it is not
clear that any one of these breeds should be selected for
exploitation at the expense of the others. Thus all three are
being included in improvement programs or plans. Given a
choice, producers will usually select the Santa Ines because of
their larger size. One project is currently underway in which
Crioula are being graded up to breeds chosen by the
producer. Most of these choose Santa Ines. This should
provide an evaluation of this breed under producer condi-
tions and also contribute to long-term improvement efforts.
Another project is concerned with improvement programs
for Morada Nova. At present, efforts are underway to
introduce new genetic material prior to initiating an
improvement program for the Somali breed.
Breed characterization studies of goats have involved the
Moxoto, Caninde, Repartida, Marota, Bhuj, and Anglo
Nubian. These have been evaluated as pure types at Sobral
(CNPC) and as crosses on SRD types at Quixada (EPACE).
Although these efforts are still underway, some preliminary
results have been reported (Shelton and Figueiredo, 1981;
Figueiredo et al., 1982c). Based on studies with the pure breeds
at Sobral, it appears that in terms of growth rate, the various
indigenous types are similar and the two exotics (Anglo
Nubian and Bhuj) support a more rapid growth rate, perhaps
reflecting their larger mature size. Reproductive data are
somewhat erratic between the breeds, due perhaps to small
numbers and to widely variable numbers between breeds.
These data do not provide a good basis for a choice between
native breeds in terms of reproductive rate. However, there is
a good basis for choice between the exotics. The Anglo
Nubian breed has been shown to be markedly superior to the
Bhuj which has higher death losses. In the crossbreeding
studies, the data have been summarized for the growth rate of
the kids, but the evaluation of the Fl females is only in the
initial phase (first mating). Growth data show no evidence of
a real difference between the indigenous or native breeds but
the two exotics sire kids with a significantly increased growth
rate (Table 1).
At this stage, it appears that the Bhuj has little to offer in
Northeast Brazil in pure form, but it may well make a
valuable contribution to crossbred populations. The Anglo
Nubian appears to be a useful animal in the region and future
breeding programs should capitalize on this breed in pure
form or in selections derived from crossbred foundations.
The number of breeds or genotypes of goats (native and
exotics) exceed the number that can be included in long-term
improvement programs with current resources. Fortunately,
the Bank of the Northeast (BNB) has provided funding for a
program throughout the Northeast designed to preserve and
improve native types of sheep and goats. Collectively, this
represents an important resource which over time can have
an important bearing on animal breeding efforts in the
Northeast. It has been a high priority goal of the SR-CRSP
Breeding Project to get this effort coordinated to maximize
the long-term gain from these regional efforts in terms of
animal improvement and generation of research informa-
tion. Since this is not an SR-CRSP funded effort and a large
number of individuals and agencies are involved, coordina-
tion cannot be forced. Only time will determine if these
efforts are successful. However, since the BNB project places
emphasis on native breeds, researchers are able to concen-
trate on work with SRD and/or mixed breed populations.
At present, the latter appears to offer more potential for
genetic improvement. This could be done while preserving
some distinctive traits of native breeds such as the color of the
The above discussion relates primarily to meat produc-
tion, the primary product obtained from sheep and goats in
Northeast Brazil. However, two other products, milk and
skins, are certainly of interest, and skins are of considerable
economic significance in Brazil. Two investigations have
been undertaken to determine factors contributing to value
of skins and these should be considered in animal breeding
efforts since Brazil, including the Northeast, has a large
industry in trading and tanning of skins. In one effort, a
number of tanneries working with sheep or goat skins were
interviewed. The finding was that size of the skin obtained at
Table 1. Least-Squares Means of Body Weights (kg) of Native and Exotic Breeds of Goats and their Crosses in Tropical
Northeast Brazil (n = 142)
Mean Body Weights (figures within parentheses
indicate one standard error)
Effects Classification At Birth At 6 Months At 12 Months
Genotypes SRD x SRD 1.86 (0.064)a 13.75 (0.842)b 20.44 (0.679)bC
SRD x Caninde 1.96 (0.059)a 11.30 (0.026)a 18.51 (0.576)a
SRD x Marota 2.01 (0.052)a 12.77 (0.009)ab 18.63 (0.554)a
SRD x Moxoto 1.99 (0.072)a 12.92 (0.897)ab 19.23 (0.692)ab
SRD x Repartida 2.01 (0.053)a 12.38 (0.815)ab 19.66 (0.513)ab
SRD x Anglo-Nubian 2.42 (0.071)b 14.35 (1.056)b 22.09 (0.811)c
SRD x Bhuj 2.48 (0.065)b 12.10 (1.093)ab 20.66 (0.805)b
Sexes Males 2.28 (0.035)m 14.62 (1.122)m 21.91 (0.381)m
Females 1.94 (0.034)n 10.97 (0.295)" 17.87 (0.354)n
Types of Birth Singles 2.29 (0.025)P 13.54 (0.641)P 20.50 (0.340)P
Multiples 1.93 (0.041)P 12.06 (0.628)q 19.28 (0.413)q
Note: Means for any characteristic with the same superscript letter do not differ significantly at the 5 percent level of probability.
slaughter and defects were the primary factors contributing
to variation in value of the skins, but there was little evidence
that breeding was a major factor except that it might
contribute to size. The defects were largely related to
environment or care in slaughter, preservation and market-
ing of the skins. The need for a larger slaughter animal to
obtain maximum skin value (at least 28 kg) would require
that all animals be maintained to advanced age. The end
result would be increased stocking rate, increased death
losses and reduced percent of the flock in the breeding age.
Since castration is not widely practiced, the surplus carryover
males would interfere with a constructive breeding program
(Figueiredo et al., 1982a).
In a further study, sheep and goat skins obtained from
slaughter of experimental animals at CNPC were processed
through a local tannery. Again, the overriding conclusion
was that defects were the primary problem and that genetic
differences were minor.
Although human food of animal origin is produced more
efficiently through milk than through meat, relatively few
goats and essentially no sheep are milked in Brazil. However
there is interest, centered in Paraiba, in increasing the use of
goats for milk production. The apparent requirements are an
adapted animal genotype with the genetic potential for a
significant level of milk production, a nutritional level
adequate to support milk production in line with the genetic
potential and a market infra-structure where commercial
exploitation is the goal. Anglo Nubian and various Alpine
dairy breeds have been imported to the region. In one study
(Rodrigues et al., 1982), it was shown that German Alpines
maintained on a good ration (Parda Alema) perform
satisfactorily as milk producers. Anglo Nubian does were
intermediate between these and native types. However, it
remains to be determined how these breeds, especially the
exotics, perform under less than optimum conditions.
In the selection experiment involving Morada Nova sheep,
two lambing seasons have been completed and the third is in
process. With this short time period, there is no evidence or
expectation of a response to selection. In 1981-1982, a total of
222 females were available for breeding. Of these, 96% mated
and 90.1% lambed with a lambing rate of 132.8. If these
lambs were marketed within the year born, this would lead to
offtake values substantially greater than that typical of
Northeast Brazil. One analysis has been completed relating a
number of characteristics of the ewes to lamb production. Of
the several variables recorded including age, condition,
weight, color, and presence of external lesions of caseous
lymphadenitis, only weight and condition of the ewe were
significantly related to lamb production. It may be significant
to note that the presence of visible lesions of caseous
lymphadenitis did not significantly affect subsequent lamb
production in the first season.
Data are currently being analyzed to provide preliminary
estimates of genetic parameters for Morada Nova sheep.
These data are being utilized for a thesis for a graduate
student at TAMU. With both these sheep and for goats, a
high priority interest is to determine the net gain from
multiple births and if this is a desirable trait under the adverse
conditions found in Northeast Brazil.
In the study involving producer flocks, one mating season
has been completed. This information provides some
descriptive data on the flocks and preliminary mating results.
More sheep (52.4%) were described as Crioula, but 43.4%
showed some evidence of having an infusion of the Santa
Ines breed. Age of the selected groups ranged from milk teeth
to four years of age. A high percentage of the sheep were
white but several other shades including spotting were
observed. This is a questionable observation in view of the
fact that white sheep have been shown to suffer some
problems in Brazil. Approximately 40% of the sheep showed
some wool cover. Continued observations on the per-
formance of the animals with various colors and degrees of
wool cover should provide information on the contribution
of these traits to adaptation or productivity. Only 5-6% of the
sheep showed identifiable genetic defects (such as progna-
thism), but this is biased as the data were not collected on a
Body weights (flock means) ranged from 28 to 42 kg at
mating. The percentage of the ewes mated ranged from 60 to
98 for the various flocks involved.
Host country scientist, Aurino Simplicio, performing laparotomy to obtain
data on ovarian function throughout the year, and how it is influenced by
nutrition and management practices. Photo: W.C. Foote.
Although preliminary in nature, the Management com-
ponent has obtained some meaningful results. One of the
major problems identified as a constraint to goat production
is fencing. The styles of fences employed in Brazil are
expensive if materials and human effort are included. Costs,
even cash costs, can sometimes be reduced by use of a solar
powered electric fence if large areas are to be fenced. It must
be recognized that electric fencing has limited applications to
the smaller producers. However, of 48 properties requesting
assistance with animal sanitation practices in the Soledade
Region of Paraiba, property sizes ranged from 20 to 2,400
hectares. Of these, 75% maintained goats but only 41%
It has been shown that goats persisting under dry caatinga
conditions will respond to non-protein nitrogen (NPN) and
energy and that NPN sources can be successfully used in
mixed rations under farm conditions (Neumaier et al., 1983).
Mixed rations for lactating goats have been successfully
tested in which Algaroba pods were used as a replacement for
The low offtake from flocks in the Northeast appears to be
due in large measure to the low percentage of breeding
females in the total population and to high death losses prior
to marketing or breeding age. Reproductive rate is not as low
as the low offtake values suggest. Research results in the US
indicate a majority of the death losses occur within the first
few days following birth. Many of the flocks in Brazil are
small and intense management at parturition reduces early
deaths. However, mortality at or around weaning and during
the dry season is a significant source of loss. In one study
(Figueiredo and Pant, 1982), 45.8% of the death losses
occurred after 60 days. The same factors that contribute to
death losses also reflects slow growth and delayed marketing
or sexual maturity. Any management practice which can
improve performance and survival in this critical period
could markedly improve production. As shown above,
supplementation with protein (even NPN) and energy will
contribute to improved gain and survival.
Significance of Research Findings
Description and characterization of the genetic resources
are reasonably complete and adequate to provide a basis for
planning future directions. In the case of sheep, it appears
desirable to work with each of the breeds, though the
Morada Nova has been identified as the breed with the
higher lambing rate. Data being collected in the selection
project will be used to determine if in fact a high lambing rate
is desirable in an adverse environment or LDC conditions
where feed inputs are often not available to capitalize on a
high lambing rate. The low offtake rate and high death rate at
or near weaning brings into question the desirability of a high
rate of multiple births. It is questionable to transfer this US
practice, where excess grain production occurs and a
favorable feed price ration normally exists, to LDC's where
neither of these are in place. For this reason, the SR-CRSP
needs to approach the question of multiple births with
caution. The Santa Ines has been shown as the breed with
increased size and more rapid growth rate, and producers are
showing a distinct preference for this breed. However, it is
not clear at this point how the breed will perform under farm
conditions. Although the Somali is a very hardy breed on a
worldwide basis, the genetic sample of this breed in Brazil is
small and perhaps not typical of the parent breed. The
indicated course of action is to consider the introduction of
new genetic material before attempting a major development
program with the Somali.
Native goat breeds show a great deal of similarity in
performance, but the Anglo Nubian has a potential use for
both meat and milk production. Improvement programs
should utilize this breed either in pure form, in systematic
crossing, or in selections based on crossbred foundations. In
respect to the dairy programs, the Alpine (Parda Alema) has
shown surprising adaptation to the Northeast when provided
with a good nutritional regime. It seems very important to
obtain data on this breed or crosses involving it under more
typical local conditions.
Implications for Future Research
In the near future, long-term selection programs will be
established for both sheep and goat species. These will run
parallel with collection and refinement of data on genetic
For the SR-CRSP in Brazil, as well as other sites, a high
priority will be placed on delineating the emphasis to be
placed on multiple births in adverse environments. Com-
bined data from more than one LDC site should provide a
Unless environmental and socioeconomic constraints can
be relaxed in the Northeast, it is extremely important that
animal breeding be carefully oriented to existing field
conditions. There is always a danger that experimental work
will be conducted under improved conditions and thus not be
applicable to producer conditions. This involves not only
climate and nutrition, but disease and parasite problems as
well. There is a disturbing tendency for researchers or
extensionists to expect a great deal of expertise on the part of
the producer in dealing with the latter problems. In the
writer's opinion, such expertise in health management is not
likely to occur unless the government provides this service to
the producer. There appears to be a need for animal breeders,
veterinarians, and animal nutritionists to attempt to estimate
likely levels of management and inputs in the future and
direct efforts or conduct selection under these conditions.
The Management component of this project has been
underway for only a very brief period. During this time,
efforts have been concentrated on nutrition because this
constituted an obvious serious need and was an area in which
early results could be expected. However, over a longer
period of time, the Management component needs to expand
efforts into more traditional aspects such as alternatives to
the large amount of time spent in confinement. There is also a
great need to combine studies involving grazing management
and supplemental feeding. Possible alternative breeding
dates for goats and sheep should be considered. The current
effort in management is heavily oriented toward assisting the
dairy goat in Paraiba. It seems desirable that SR-CRSP
efforts in this area continue to be closely integrated with
EMEPA efforts and avoid direct initiatives or responsibility
in this area.
A cooperating Brazilian family near Taua proudly display a prolificfemale goat and her triplet offspring. Photo: W.C Weir.
The collaborative work mode per se has provided informal
but valuable training for all involved. But it would be difficult
to express this in quantitative terms.
In terms of formal training, a total of six graduate students
associated with the project in Brazil have requested and been
encouraged to attend graduate school at Texas A&M. From
the outset, funding has been allocated for one graduate
student per year on the project. An added factor that has
limited the numbers attending is that students are integral
participants in the work in progress and careful sequencing is
required to schedule their absence from fieldwork assign-
ments. At present, one student, Antonio Amaury Oria
Fernandes, is attending Texas A&M as a graduate student in
Animal Breeding and is doing his thesis on SR-CRSP project
activities. Other students from EMBRAPA, including the
CNPC, have been or are presently registered as graduate
students at TAMU, but are not directly associated with this
The Brazilian Co-Principal Investigator has traveled to the
US on project funds for a period of training and orientation
including participation in the Third International Confer-
ence on Goat Production and Disease.
A one-week training short course was held in Brazil for
research personnel at CNPC and collaborating state
agencies. Approximately 25 people attended this training
course which may be repeated at a future date if the project
continues and the need exists.
Dr. C.A. Zometa has provided on-site traineeships, each
lasting three months, for two graduate technicians at
Fazenda Pendencia. Others are scheduled for the near future.
Rodrigues, A., W. Hauss de Souza, E.A.P. Figueiredo, P.R.M. Leite, and
K.P. Pant. 1982. Avaliacao da Producao Leiteira das Racas Anglo-
Nubiana, Parda Alema e Sem Raca Definida No Estado da Paraiba.
Mimeograph EMEPA. Joao Pessoa, PB. Brasil.
Shelton, M. and E.A.P. Figueiredo. 1981. Types of Sheep and Goats in
Northeast Brazil. Int. Goat and Sheep Res. 1(4):258-268.
Mason, I.L. 1980. Sheep and Goat Production in the Drought Polygon of
Northeast Brazil. World Ani. Rev. 34:23-28.
Bellaver, C. 1980. As Peles. Sobral, Ce. EMBRAPA. Centro Nacional de
Pesquisa de Caprinos. Circular Tecnica No. 3. 16 p.
Figueiredo, E.A.P., E.R. de Oliveira, C. Bellaver, and A.A. Simplicio. 1983.
Hair Sheep Performance in Brazil. In Hair Sheep of Western Africa and
the Americas. A Genetic Resource for the Tropics. H.A. Fitzhugh and
G.E. Bradford (eds). A Winrock International Study. Westview Press.
Figueiredo, E.A.P. and K.P. Pant. 1982. Evaluation of Goat Breeds in the
Tropical Northeast Brazil. II. Analysis of Age of Death of Kids. Pesq.
Agropec. Bras., Brasilia. 17:803-808.
Figueiredo, E.A.P., M. Shelton, and K.P. Pant. 1982a. Goat Skins. Proc. III
Intl. Conf. on Goat Prod. and Dis. Tucson, AZ. pp. 488-490.
Figueiredo, E.A.P., A.A. Simplicio, and K.P. Pant. 1982b. Evaluation of
Sheep Breeds for Early Growth in Tropical Northeast Brazil. Tropical
Animal Health and Production. 14:219-223.
Figueiredo, E.A.P., A.A. Simplicio, C. Bellaver, and K.P. Pant. 1982c.
Evaluation of Goat Breeds in the Tropical Northeast Brazil. 1. A Study
of Birth Related Traits of Native and Exotic Goat Breeds. Pesq.
Agropec. Bras., Brasilia. 17:643-650.
Neumaier, M.C., C.A. Zometa, A. Rodrigues, R.H. Nobre, G.P. Primov,
M.F. Nolan, and P.R.M. Leite. 1983. Biological Performance and
Sociological Responses to the Introduction of New Management
Techniques for Goats Among Small Producers in Northeast Brazil.
Abstract to be Presented: Ann. Mtgs. of the Amer. Dairy Science Assoc.
B C.r. u Bandung
^ ^ AGarut
-, CJ Au V A he activities of the SR-CRSP in
Indonesia are almost exclusively
A Worksites I N D I A N confined to Java, an island with a land
-] Uplands in West Java 0 C E A N 100 200 area similar to that of California but
I I I carrying in excess of 80 million people.
105E 110E KILOMETERS Situated in the humid tropics and with
highly fertile volcanic soils, it is not
Location of Small Ruminant CRSP worksites in Indonesia. surprising that crops form the back-
bone of the Javanese rural economy,
and that livestock contribute less than 10% of the value of agricultural production. Despite this, however, the 3
million sheep and 7.5 million goats fill a vital economic niche for the smallholder.
Farm sizes have been declining precipitously in Java in recent years and the current average size has fallen to
0.3 ha per farm. Because large ruminants are not economical on such small acreages there has been a large
increase in the sale of the cattle and water buffalo formerly used for draft power. As farm sizes get smaller,
there is little land available for grazing or fodder production and intensive "cut and carry" confinement or
semi-confinement husbandry systems for the smaller ruminants have become well developed. Fueling this
trend are increased crop intensities and yields which have increased the supply of crop by-products and residue
yields available for feed. Tree crops also provide a wide variety of livestock feeds, as do grasses and leaves
gleaned from hedgerows. Tree leaves and food crops which serve as livestock feed include: banana, bamboo,
cassava, corn, cucumber, hibiscus, jackfruit, and sweet potato. In addition to the widely practiced "zero
grazing" confinement system, there is some grazing encountered along roadsides and other unused areas.
Concentrates are rarely used, as land is at a premium for growing food or estate crops. Small ruminants are
primarily a cash crop on Java, raised for their meat and manure and marketed rather than consumed by the
producer. Most sheep and goats are raised by smallholders in highly intensive, crop/livestock production
systems with a herd size rarely exceeding 5 head. Increased herd sizes are found in areas where dry farming is
There are three major cultivation zones on Java, defined primarily by elevation, and each associated with a
particular cropping profile: in the lowlands, wet rice farming predominates; in the upper elevation slopelands
the revenue producing estate crops, coffee, tea, cocoa, rubber and spices are grown; the highlands are
characterized by intensive vegetable production systems. Each zone is characterized by a different livestock
production system, with emphasis on different animals and different breeds.
Sheep breed types found on Java display a continuum extending between two main types, the Thin-Tailed
hair sheep, prevalent in West Java, and Fat-Tailed wool sheep found mostly in East Java. Both types are
known to be highly fertile, but of low mature liveweight. Lamb mortality tends to be high, partly on account of
low birth weights and partly because of unpredictable multiple births.
The Kacang goat is the indigenous type and Kacang Etawah crosses are also common. The Etawah is
actually a cross between the Jamanapuri imported from India and better local Javanese goat varieties. Though
essentially a milking goat, the Etawah is rarely used for milk in Indonesia. Both the Etawah and the Kacang
type are valued almost exclusively for their meat.
The relatively poor productive performance of sheep and goats on Java suggests that there is ample room
for improvements even within the context of the smallholder system. Feed constraints are probably less a
factor in the humid environment of Java, suggesting that with modifications to breeds, health care systems and
management, production could be significantly improved by adaptation of existing technology.
More information is needed about the genetic potential, nutrient requirements and health problems of the
livestock, the nutritional value of their common feedstuffs, and the socio-economic constraints in the
production system. Until these poorly identified and understood factors are investigated, it will be difficult to
implement modifications in the system which optimize return.
The SR-CRSP is focusing its research at two locations in Java, Cirebon on the north coast and Garut in
central Java. These locations are representative of rainfed and irrigated agricultural areas respectively, and
sheep and goat production systems show a shift from highly intensive stall-fed animals in Garut to the more
relaxed semi-grazing system of Cirebon. The SR-CRSP is concentrating its efforts on research in improved
nutrition, controlled breeding with more advantage taken of prolificacy traits, as well as examining economic
and sociological constraints to the adoption of improved technology. Transport, marketing credit, land
tenure, and the integration of husbandry with the predominant garden culture all present severe problems in
developing more productive systems.
The Agency for Agricultural Research and Development (AARD) is the official Indonesian agency with
which the SR-CRSP is collaborating. Subsidiary collaborative organizations and their relationship to each of
the four SR-CRSP projects in Indonesia are indicated in Appendix IV.
of Sheep and Goats for
University of California, Davis
At the start of the SR-CRSP program in Indonesia, the
Indonesia Animal Husbandry Research Institute
(LPP, subsequently renamed BPT) had only a limited
program with small ruminants. Two stations (at Cicadas and
Cilebut) had recently been established and a small number of
animals acquired. No direct work with village livestock
owners had been carried out. As a result, performance
parameters for Indonesian sheep and goats were not well
established, particularly under the production environments
characteristic of smallholder operations. There have been
reports of exceptional prolificacy in Javanese sheep (e.g.,
Mason, 1980) but here also, quantitative data are quite
limited. Lack of such data precludes assessment of the ways
in which productivity is limited by genetic factors.
Improving genetic potential of livestock in production
traits is generally an efficient and highly cost-effective
method of improving productivity, especially in populations
not recently subjected to selection for performance. Genetic
improvement can be effected with no new inputs other than
some simple records, and in some cases, with no changes or
only modest changes in management. While gains per year
are relatively small, they are cumulative and permanent.
Also, breeding and selection programs require performance
records which often prove to be of at least as much value for
improved management as for the genetic programs which
prompt the initiation of recording.
The initial goals of the Breeding Project were therefore to
work with the BPT staff in establishing the capability for
research on small ruminant breeding in Indonesia, to identify
those traits most in need of genetic improvement, and to
assess the merits of selecting within local breeds and types
versus introducing new germ plasm through exotic breeds.
The longer term goals are to initiate genetic improvement
programs adapted to local needs and capabilities, with a view
to producing inherently more productive sheep and goats for
Indonesia. Because genetic improvement is a long-term
undertaking, it must be continued well beyond the life
expectancy of any specific aid program if it is to realize its full
potential for increasing productivity. Training of host
country personnel to enable them to carry on the work
following termination of the SR-CRSP is therefore an
integral part of this project plan.
Since its inception, the Breeding Project has directed its
attention towards the following specific objectives:
*To characterize performance of Indonesian sheep and
goats with regard to reproduction and growth rates under
experiment station and village conditions, for purposes of
identifying constraints to production and potentials for
* To investigate in detail the reported prolificacy of
Javanese Thin-Tail sheep, assess the desirable and
undesirable effects of this prolificacy on management and
productivity, and determine its genetic basis vis-a-vis that
in other prolific breeds of sheep.
* Based on results of the characterization phase, to develop
plans for reducing genetic and management constraints to
productivity of small ruminants in Indonesia. These plans
should be appropriate to local production systems and
* To train Indonesian personnel in small ruminant man-
agement and performance recording, and data analysis
* To collaborate with personnel on other projects in
acquiring a more complete information base for use in
developing improved small ruminant production prac-
tices and systems for Indonesia.
To develop an understanding of the performance variables
of Javanese village sheep and goats, identification and
recording systems have been developed for use both in
villages and on experiment stations.
Forms, printed in Indonesian, which facilitate convenient
data recording and computerization, have been developed
and personnel trained in their use both on station and in the
A senior sheep management specialist has been posted in
Bogor for a full year, and is emphasizing identification of
constraints to production under village and station condi-
tions, summarization and interpretation of data from both
sources, development of improved management plans for
research and village production, and training of junior BPT
staff for more effective management of research animals and
Five reports documenting various aspects of performance
of the Javanese sheep and goats have been published, and
others are forthcoming as more data becomes available and
Data from the project are being used for an MS thesis in
Animal Breeding in the US, and animal performance data
from this project are also being used by the SR-CRSP
Economics and Farming Systems projects in Indonesia.
The majority of the data summarized to date are on the
Javanese Thin-Tailed sheep. The data show that these sheep
are in fact quite prolific, with litters of 3 and 4 occurring with
relatively high frequency, and litters of 5 and 6 occurring on
rare occasions (Subandriyo et al., 1981; Inounu et al., 1982;
Sitorus and Subandriyo, 1982). However, they also show a
much higher incidence of single births than expected for
sheep which have 3 or more lambs at the frequencies
observed here, i.e., they are unusually variable in litter size.
Also, mortality is very high in the larger litters (Inounu et al.,
1982) and growth rates are very low (Thomas et al., 1982).
This unusual variability in litter size in Javanese Thin-Tail
(JTT) ewes was investigated further in data from 56 ewes in
the Cicadas flock. The overall mean number of lambs born in
131 lambings was 1.98, identifying the breed as above
average but not exceptional prolificacy. However, 11% of
parturitions contained 4 or more lambs, a most unusual
result for a group with a mean of 2 or less. The distribution of
litter sizes is plotted in Figure 1, along with that for a group of
UC Davis Suffolks with a similar mean litter size.
As noted in Figure 1, the standard deviation and therefore
coefficient of variation is 50% higher for the JTT ewes than
for the Suffolks. Furthermore, our subjective assessment is
that plane of nutrition for the JTT ewes was substantially
below that which would lead to a maximum litter size. Thus,
if they respond to improved nutrition with an increase in
litter size similar to that observed in other breeds, one would
expect even more large litters. Whether that would increase
the variability is not known; it is possible that it would shift
ewes from the 1 or 2 class into the 3 class without increasing
the proportion of 4's, 5's and 6's. However, the distribution of
litter sizes among individual ewes also appeared to be
different in the JTT group. The range in total number of
lambs born per ewe was 2 to 6, 3 to 10, 5 to 13, and 10 to 25
for ewes with 2, 3, 4, and 6 lambings, respectively.
This extreme variability presents problems in manage-
ment, since prepartum nutrient requirements are much
higher for ewes carrying 3 or more lambs than for those with
singles, and post lambing management requirements for
acceptable survival rates are much higher for the small lambs
resulting from large litter size. Thus, efficient utilization of
high prolificacy requires that one have ewes among which a
high proportion will reliably produce mostly 2's and 3's, with
few singles or litters of more than 3. In this situation,
intensive management for all ewes would probably be
justified. On the other hand, if intensive management (e.g.,
supplementing with concentrates during gestation and
lactation, artificial rearing of lambs above 2 per litter using
milk replacers) is not possible, then one wants ewes which
produce no more than twins; 26% of ewes with 3 to 6 lambs,
as in this sample of Javanese Thin-Tail ewes, is a definite
Controlling this variability, which is almost certainly
genetically determined to some extent, requires that we know
the mode of inheritance of litter size in these sheep. Litter size
in sheep is generally inherited as a quantitative trait of low
heritability, which means that it can be changed by selection,
though rather slowly. However, in most sheep, including
breeds such as the Finnish Landrace with substantially
higher mean litter size than the Javanese Thin-Tail,
variability in litter size is less. There are, however, a few with
variability similar to that in the JTT sheep, with the most
intensively studied of these being the Booroola Merino. In
the Booroolas, it has recently been established (Piper and
Bindon, 1982) that the extreme variability is due to
segregation of a gene with a large effect on ovulation rate.
Ovulation rate sets the upper limit to litter size, but prenatal
loss can reduce litter size to any number down to one.
Jovonese Thin Tail
w 50- 50- S.D. 0.71
1 40- 40-
z 30- 30-
M 10- 10-
S 2 3 4 6 2 3 4
Figure 1. Frequency distribution of numbers of lambs born to Javanese
Thin- Tail ewes at the Cicadas research station. Data on UC Davis Suffolk
ewes is shownfor comparison.
Since the variability in the JTT ewes is similar to that in the
Booroola, we are investigating the possibility of a'Booroola'
gene in the JTT breed. Only a relatively few daughter-dam
pairs and paternal half sister groups are available for study as
yet. These data are shown in Table 1.
While the data in Table 1 are limited, they are at least not
inconsistent with the hypothesis of a single gene for
prolificacy, which in either heterozygous or homozygous
state confers on ewes the ability to produce 3 or more lambs
per parturition (though because of prenatal mortality, it by
no means assures a large litter each time). Sire 186 is inferred
to be heterozygous for the postulated gene, while sire 185 did
not carry it.
Further evidence in support of this hypothesis is provided
by data on ovulation rate from three groups of ewes at
Cicadas. Dr. J.F. Quirke performed endoscopies at two
consecutive estrus periods, the first of which was synchro-
nized on all ewes in the flock in April/ May 1983. The results
are summarized in Table 2.
The repeatability of ovulation rate shown in Table 2 (0.81)
is remarkably high; by comparison, the value in the UC
Hopland flock of Targhees was 0.27, following the same
protocol and measured by the same person. Removal of
effects of body weight and birth year in the ovulation rate
data in Table 2 changed repeatabilities very little; the four
values were 0.76, 0.85, 0.74 and 0.79, respectively. The
repeatability was high even when ewes with only one of two
ovulations were considered: 87 of 105 such ewes had the same
number both times. There was also a remarkably high
correlation between ovulation rate and prior litter size. For
example, all 15 JTT ewes in the study which had previously
given birth to triplets or more had 3 or more ovulations at
Table 1. Litter Sizes of Daughter-Dam Pairs and Paternal Half Sister Groups of Javanese Thin-Tail Ewes in the Cicadas Flock
A. Daughters Sired by Ram 185
Dam's Litter Size
Daughter's Litter Size
36 3,4, 1 1, 1,4; 1; 1
158 3,3 2,2
163 1, 2, 2, 2, 3, 1 2, 3, 2
168 2,2 1,2
176 3,3 1, 1, 1
Average 15-2.33 17-1.65
B. Daughters Sired by Ram 186
37 2, 2, 2 2, 1
41 4, 4, 2 3, 4, 3; 2, 2
117 2,3,4,3 1; 1
158 3, 3 1, 3, 2; 3, 2, 4
164 1, 1 3,3,3,4
165 1, 1 1, 1
175 1, 1, 1,2 1,2
701 2,3 1, 1
704 2 2, 1
Unknown 2; 2, 1; 3, 1; 2, 4
least once, 11 of them both times. The ewe with 6 litters of 3 to
6 lambs had the highest ovulation rate, 5 both times.
These results provide strong evidence for the hypothesis of
a major gene affecting ovulation rate, with the similarity in
pattern in the three groups lending further support. Data on
the limited number of daughter-dam pairs in the ovulation
rate data were also consistent with this explanation.
Additional data on daughter-dam pairs from known sires are
required to prove or disprove the hypothesis, and matings of
the ewes in the ovulation rate study have been made with that
The data base established by the Breeding Project has
revealed that a major constraint to production under village
conditions appears to be the long intervals between
parturitions, in spite of the fact that experiment station data
and the performance of a few village ewes shows that these
sheep are quite capable of lambing at intervals of 8 to 9
months. Analysis of the village management system (Bell and
Inounu, 1982) indicates that the infrequent parturition
problem is due to failure to mate the ewes. This problem
appears to be due to a lack of ram availability and the failure
to develop a plan to ensure that the few breeding rams in the
villages are rotated in a systematic fashion. A ram rotation
scheme has recently been developed by Breeding Project
personnel which should remove this constraint. This scheme
will be tested in the near future in one or more cooperating
Although less information is available, effective avail-
ability of males may be a similar constraint to goat
production. This will be investigated further, and programs
to correct the problem initiated if appropriate.
Under experiment station conditions, feed supply has been
shown to be a further major constraint to production,
especially during the dry season. The Breeding Project
scientist, Mr. Bell, in consultation with irrigation engineers in
Indonesia and California, is developing an improved
irrigation plan to correct this problem.
To assist with the longer-term objectives, a general plan for
setting goals and implementing a breeding program for
multiplication centers was drafted in 1980 by the Breeding
Project PI, in consultation with BPT staff and the
management staff of the Margawati Station in Garut, West
Java. This plan has not yet been fully implemented, and the
Margawati flock had to be reduced drastically in 1982
because of the activity of Galunngung volcano. At the
request of the Dinas Peternakan in Garut, SR-CRSP
Breeding Project staff used performance records collected at
Margawati in 1980-82 to calculate indexes on all station
sheep, for use of the Margawati staff in determining which
animals to keep in these circumstances.
In summary, the work to date has established that the
Javanese breeds have an adequate or more than adequate
prolificacy potential. Improvements in management to
increase lambing frequency and survival, and in genetic
potential or nutrition (or, most probably, both) to increase
growth rates, have been identified as areas needing attention.
A study completed as part of the SR-CRSP Breeding Project
in Kenya suggests that a reduction in amount of wool cover,
in which there is substantial variability in the Javanese sheep
would also improve growth rates. In general, the per-
formance and variability of these sheep suggests high
potential for improving performance by selecting within the
local types. Their high prolificacy potential is something that
might be difficult to find in other sheep adapted to
confinement rearing in the humid tropics. All of these points
indicate that improvement within local types is likely to be
preferable to introduction of new germ plasm.
Significance of Research Findings
The identification of very long parturition intervals in
village flocks, due to failure to mate the females, represents
probably the most important finding in terms of potential for
effecting an immediate and substantial increase in produc-
tion of small ruminants in the Indonesian smallholder
confinement production system. For example, estimated
lamb production per unit of feed fed to each ewe would
increase about 58% with a 9-month interval compared to a
15-month interval, the latter representing a conservative
estimate from the 53% breeding age ewes lambing within a
12-month period reported by Bell and Inounu (1982).
Documentation of the extreme variability in litter size
indicates potential for very high production from the
Javanese Thin-Tail ewes, but also major management
problems. Inounu et al. (1982) reported mortality rates of
16.7, 18.4, 35.5, 42.9 and 60.0% for lambs born in litters of 1,
2, 3, 4, and 5, respectively. Applying these to the proportions
of ewes producing the different litter sizes reported for JTT in
Figure 1 gives a mean of 133 lambs weaned per 100 ewes
lambing. Applying the same mortality figures to the
distribution of litter sizes shown for the Suffolk ewes
included in that figure gives a mean of 149 lambs per 100
ewes, a 16% increase simply from reducing variability in litter
size, with no change in mortality rates for each litter size. For
a population of a million ewes, this would represent an
additional 160,000 lambs. The reduced variability in litter
size would reduce variation in birth weights and hence in
weaning weights, an added advantage. Solution of these
problems and utilization of this potential will depend on
determination of the mode of inheritance (which may be
known from the data on ovulation rate to be collected later
this year, or which may take considerably longer).
The finding of unusually high reproduction potential in
the Javanese sheep and goats indicates that work should
focus on improvement of these animals, rather than going to
the expense and risk of importing animals.
A very general finding is that there are areas of
management where feasible modifications can make signifi-
cant improvements in productivity of small ruminants under
Indonesian village systems. Although not specifically our
responsibility, the Breeding Project specialist currently in
residence on-site has the experience to make important
contributions in this area. For example, he has identified the
long lambing interval problem in the villages and developed a
ram rotation plan to correct this problem. Such improve-
ments in management are critically important since produc-
tivity increases resulting from genetic improvement may be
quite limited until management in Experiment Station,
Multiplication Station, and village flocks is improved. Thus,
we feel the emphasis on management is well justified from the
perspectives of both the Breeding Project and the Program as
Future Research Directions
An important aspect of the entire SR-CRSP in Indonesia,
is monitoring sheep and goat performance in four West
Javanese villages. A high priority in the Breeding Project
research is the continuation of this work with more detailed
analysis of the data in order to indicate production
constraints and the means for their relief. Based on results to
date, work on systems for distribution and use of rams as a
means of increasing frequency of parturition will be
Table 2. Ovulation Rate, Body Weight, Cycle Length and Duration of Estrus in Javanese Thin-Tail (Garut Strain), Javanese
Fat-Tail and Semarang Ewes
JTT Fat-Tail Semarang All
No. of ewes 85 48 22 155
No. in estrusi 77 48 21 146
Mean SD Mean SD Mean SD Mean SD
Body weight (kg) 24.2 5.4 23.2 4.9 21.3 4.1 23.4 5.1
First estrus 1.83 1.01 1.83 .93 1.67 .86 1.81 .96
Second estrus 2.05 1.10 1.96 1.05 1.81 .68 1.99 1.03
Repeatability .79 .88 .74 .81
Cycle length (days)2 16.6 .88 16.3 .71 16.1 .78 16.4 .83
Duration of estrus (hrs)2 33.1 6.5 34.2 7.5 35.7 7.2 34.2 7.0
'Following synchronization with progestagen impregnated pessaries.
2Total numbers of ewes recorded for cycle length and estrus duration were
134 and 76, respectively.
Ismeth Inounu, BPT, Bogor, with Javanese Thin- Tail ewe no. 162, Cicadas
flock. Demonstrating its remarkable prolificacy, this ewe gave birth to 25
lambs in 6parturitions in years (3, 3,6, 4, 5, 4)andhad5 ovulations in each
of 2 consecutive cycles in 1983. Photo: G.E. Bradford.
The Breeding Project will also continue its collaboration
with the Margawati Station (and other multiplication
stations where there is interest) in developing selection
criteria and mating plans for genetic improvement. Results of
research on inheritance of prolificacy and growth in the local
breeds will be incorporated into these plans as rapidly as
Inheritance of Prolificacy
Preliminary analysis has suggested the possibility of a
major gene for high ovulation rate/litter size segregating in
the Javanese Thin-Tail sheep. If this proves to be the case,
control of prolificacy may in fact be more difficult than if
inheritance of the high prolificacy of these sheep is of the
more usual quantitative nature. Confirmation of the
presence of a major gene would lead to the following options:
* Development of strains homozygous for the "prolificacy"
gene. These animals would be expected to show
considerable variability in litter size, but on average to
have a high litter size and hence to require and justify
continuous intensive management and good nutrition.
* Development of strains which do not carry the "prolifi-
cacy" gene. At this point we estimate such ewes might
have mostly singles and twins. This might be an adequate
level of prolificacy for most management situations, and if
not, selection on the quantitative variation which
undoubtedly exists in these animals should increase it,
though at rates of probably only 1-2% per year as in other
We believe that following either of these options (or both,
for different management levels) would represent a substan-
tial improvement over the present situation, in which there is
no basis to predict whether a ewe will have mostly singles and
twins, or mostly 2's to 4's. Very different management of the
ewes and the lambs is required for good viability and growth
of lambs in these two situations. An economic analysis of
costs and returns from the two kinds of ewes would be called
If the basis of inheritance proves to be quantitative,
selection for ewes which do not produce more than two
lambs would be of interest to determine if this would reduce
the extreme variability in litter size.
Work on growth will be carried out in close collaboration
with the Nutrition Project. Questions to be pursued include:
* Comparisons of the growth rates at different levels of
energy intake of lambs born in litters of 1, 2, 3,>4. This
would integrate well with the economic analysis suggested
* Growth rates of lambs from ewes of different degrees of
wool cover, and of lambs differing themselves in degree of
* Comparative growth rates of Thin-Tail and Fat-Tail
sheep, and of Etawah and Kacang goats.
Training has been considered an integral part of this
project from the beginning. There are two related objectives.
One is the training of host country personnel in animal
management and animal breeding research activities, so that
data collected, now and in the future, will be from
appropriately managed animals and collected in a way to
maximize its utility for the research objectives of the project.
This aspect of training will also increase the ability of the
persons involved to assist village livestock owners in solving
management problems. Training in this area has been
accomplished through participation of UC Davis personnel
in shortcourses in Indonesia, short-term training of Indo-
nesian staff in the US, and interaction between UCD staff
working in Indonesia and BPT staff.
The second objective is support for graduate training of
young Indonesian staff, with the aim of preparing them to
take over leadership of breeding research activities in
Indonesia when the SR-CRSP terminates.
The following is a summary of training activities under this
project in Indonesia to date.
* Participation of D.T. Torell in Sheep Management
Shortcourse in Bogor, June/July 1980. Mr. Torell was
Sheep Management and Research Specialist in charge of
the University of California's Sheep Research program at
the Hopland Field Station from 1951 through 1981.
* Seven-week training in sheep and goat management,
record-keeping, etc., for Mr. Subandriyo at UC Davis
and UC Hopland, January/February 1981.
* Work by Mr. Monte Bell with Ismeth Inounu, Bess
Tiesnamurti and Mrs. Endang in livestock management
and recordkeeping at Cicadas and Cilebut stations which
included daily or weekly conferences on ways of
maximizing information obtained from an experiment
station flock. Mr. Bell has also worked with the BPT staff
to perform an important training function for non-degree
staff working at both the experiment stations and in the
village monitoring program.
* Mr. Rachmat Setiadi, of the Margawati Station staff
(Dinas Peternakan, Garut), was funded for an MS
program at Institut Pertanian Bogor (IPB) from July
* Mr. Subandriyo, of the BPT staff, is being funded for an
MS program at Montana State University from March
1982-March 1984. He will spend one quarter enrolled at
the University of California, Davis.
* Funding for an MS program at IPB is being provided to
Mrs. Endang Triwulanningsih of the BPT staff,
Bell, Monte and Ismeth Inounu. 1982. Sheep Reproduction Parameters
from Sixteen Farms in Sukawargi Village, District of Garut, West Java.
Working Paper No. 8, SR-CRSP, Bogor, Indonesia.
Inounu, Ismeth, N. Thomas, and P. Sitorus. 1982. Lambing Characteristics
of Javanese Thin-Tail Sheep. Mimeo, SR-CRSP/BPT, Bogor, Indo-
Sitorus, P. and Subandriyo. 1982. Javanese Thin-Tail Sheep Characteristics
in Highland, Medium and Lowland Areas. Proc. II World Congr.
Genetics Appl. to Livestock Production, Madrid, Ed. C.L. de Cuenca.
Sitorus, P., Subandriyo, and Endang Triwulanningsih. 1982. Performance
Traits of Indonesian Goats. Proc. III Intl. Conf. on Goat Prod. and Dis.
Tucson, AZ. p. 538.
Subandriyo, P. Sitorus, J.M. Levine, and G.E. Bradford. 1981. A
Preliminary Report on Performance of Javanese Thin-Tail Sheep under
Experiment Station Conditions. Mimeo, SR-CRSP, Lembaga Peneli-
tian Peternakan, Bogor, Indonesia.
Mason, I.L. 1980. Prolific Tropical Sheep. FAO Animal Production and
Health Paper No. 17.
Piper, L.R. and B.M. Bindon. 1982. Genetic Segregation for Fecundity in
Booroola Merino Sheep. In Proc. World Congress on Sheep and Beef
Cattle Breeding, 1:395. Ed. R.A. Barton and W.C. Smith. Dunmore
Press, Palmerston N., NZ.
Thomas, N., W. Mathius, and M. Sabrani. 1982. Small Ruminant
Production in West Java. Methodology and Initial Results. Proc.
Livestock Dev. in Asia Conf., Singapore.
Economic Analysis of
A.J. De Boer
Animals are an important component of the integrated
farming systems found in many regions of Indonesia.
Although the subsistence needs of farm families are primarily
met by food crops, animal production is often the primary
means by which farmers accumulate cash, store capital,
provide inputs such as traction and manure for crop activities
and/or provide high quality food for the household. These
functions of livestock in traditional farming systems are
further enhanced by the role of animals in social and religious
ceremonies such as "Idul Adha" and "Idul Fitri."
In Indonesia, these roles are particularly important for
goats and sheep which are easy to raise, prolific, and have a
ready market. Their initial and maintenance costs are low
and they utilize marginal land and crop residues. Special
advantages of small ruminants over large ruminants (cattle,
buffalo) include higher reproduction rates, broader adapt-
ability to a range of environments, easier marketability, and
lower risk. Despite these apparent advantages, the small
ruminant population in Indonesia has been stagnant over at
least the past 20 years. In fact, the population of sheep and
goats in Java declined at 0.5% per annum over the 1967-1976
period (Sabrani et al., 1982d).
Sheep and goats are typically identified with the small-
holder. In Indonesia, there are more sheep and goatkeepers
than cattle owners or commercial chicken/pig farmers. For
example, in Central Java in 1980, there were 0.97 million
households engaged in small ruminant activities compared
with 0.62 million in large ruminants and only 9,600 farmers
in commercial chicken and pig raising (Dinas Peternakan,
Central Java, 1982). Our studies (Knipscheer et al., 1983b)
indicate that every fifth farm in Indonesia has small
Despite these indications of the importance of sheep and
goats for Indonesian smallholders, there has been little
information about the physical and economic productivity of
the small ruminant enterprise. In cropping systems research,
where the farming systems approach has been practiced, it is
known that a large gap exists between yields found under
controlled experiment station conditions and those found
under actual farming conditions. Although there is a growing
knowledge about the productivity of small ruminants on
Java under research station conditions (Obst et al., 1980) and
under controlled conditions in villages (Winantea and
Djuniarti, 1981), little is known about the actual biological
and economic productivity of sheep and goats under
Indonesian small farm conditions. The question is how much
on-farm productivity deviates from productivity rates
obtained on research stations (the "yield gap"), the sources of
this difference, and the economic implications of closing this
Initial project goals were to establish a "farming systems
research" philosophy within the collaborating institution to
allow BPT researchers to identify relevant researchable
problems and to begin long-term cooperative programs with
villages to test the results of the research. At the center of this
approach is the development of indigenous research capacity
on the economics of small ruminant production and
The long-term goals are thus (a) to establish institutional
capacity to carry out a range of studies related to constraints
on increasing income from small ruminant farming, (b) to
develop a farming systems research methodology appro-
priate to the needs of BPT and the participating farmers, and
(c) to assist with the development of national research and
development programs for small ruminants.
To achieve these goals, the following objectives have
guided the project:
* Characterize small ruminant production and marketing
systems to assist in the identification of productivity levels
and resource use.
* Develop appropriate economic methodologies for anal-
ysis of traditional and improved production systems.
* Construct agro-economic profiles of selected livestock
farming systems to identify the on-farm testing program
* Carry out policy-oriented studies to identify pricing,
marketing, and investment needs to spur technological
* Develop institutional capacity within BPT through
formal and on-the-job training, research planning, and
* Carry out studies in other areas (outside Java) of
Indonesia to identify potential and needs for small
ruminants and build up institutional expertise in BPT
Planning for the research project started in early 1980 with
the drafting of questionnaire outlines for review in Indonesia.
In the summer of 1980, a training workshop was held to
discuss research needs, research approaches, data require-
ments, and collection procedures, the formation of "village
programs," distribution of breeding stock, and research
responsibilities. During the course, questionnaires were
developed for baseline surveys (Sabrani et al., 1982b) and for
periodic surveys to monitor husbandry practices and
resource use relationships (Thomas et al., 1982). These early
surveys were rather general in scope and served to define
overall differences between upland and lowland farming
systems (Knipscheer and Soedjana, 1983; Sabrani et al.,
Sheep or goat house, West Java. Small ruminants in the intensively farmed
areas are confined in such barns close to the homestead andfed with forage
cut by hand and carried to them. Note the large overhang of the roof giving
shade and protection from torrential rain, and the large ground clearance,
making manure collection an easier task. Photo: M. Nolan.
1983a; Mawi and Sabrani, 1982), relationships between
resource base and small ruminant production (Muljadi et al.,
1980; Sabrani and Muljadi, 1981), and income generated by
small ruminants (Knipscheer and Sabrani, 1982; Knipscheer
et al., 1983b). Additional survey work was initiated in a
transmigration area in Sumatra to assess research needs and
development prospects for small ruminants under a situation
of vastly different resource endowments relative to Java
(Mink, 1983). The initial surveys helped identify the
important constraints to expanding small ruminant numbers
and productivity, helped develop a research methodology for
solving these problems (Sabrani et al., 1981; Sabrani et al.,
1982c; Thomas et al., 1982), and allowed subsequent
fieldwork to focus on very specific types of problems such as
low reproduction rates (Soedjana and Knipscheer, 1982;
Knipscheer and Soedjana, 1983), the impact of animal
distribution ("dropping") schemes (Knipscheer et al., 1983b),
demand for sheep and goat meat (Sabrani, 1980), seasonal
changes in meat prices caused by Islamic festivals (Soedjana,
1982), and the efficiency of traditional marketing systems
(Sabrani and Knipscheer, 1982; Sabrani et al., 1983b). The
research approach thus changed gradually from large scale
data collection efforts to shorter, more focused inter-
disciplinary surveys which examined specific components of
the production system.
Three quite distinct locations have served as field research
sites. The first, Cirebon, is a lowland, rice-producing area
along the north coast. The second, Garut, is a typical
highland, mixed-cropping system while a third location,
Ciburuy, is intermediate in elevation between the above two
and is distinct in that it is located next to a large rubber
plantation. Table I gives the general characteristics of the
sites and Table 2 gives the average farm sizes in the sampling
strata that were selected for the farm surveys. Table 3 gives
average herd/flock composition on the survey farms. These
data show two major constraints-the very small farm size
which limits feed and cash resources for small ruminant
production and the high proportion of adult males to
breeding females which constrains herd/flock offtake rates.
Tables 4 and 5 synthesize much of the survey data from the
villages into measures of herd/flock offtake (final column,
Table 4) and the relative importance of small ruminants in
the household economy of the villages. The final column of
Table 5 illustrates that smaller farms (size strata I and II)
generate a relatively larger proportion of total income from
sheep and goats. These income figures are for small ruminant
holders only. As the percentage of farmers that keep sheep
and goats is about twice as high in the upland areas as in the
lowlands (about 60% vs. 30% based on Van Santen, 1980),
the small ruminant share in total agricultural production
should be higher in the upland regions than in the lowland
The important implication of this analysis is that small
ruminants provide a vehicle to improve the income of poor
and destitute farmers. Research in village small ruminant
production will have a relatively larger impact on landless
and subsistence farmers than on medium and larger farmers.
There is therefore a strong social justification for the further
allocation of resources for research on small ruminants.
Moreover, as meat is generally consumed by higher income
groups, the small ruminant enterprise also represents a
method of income transfer from urban high-income to rural
low-income population groups.
Another important economic component in the villages is
the practice of sharing of animals. Sharing ("gadohan") is
common in all villages in Java, as well as on Sumatra ( Mink,
1982). Animals are lent by owners to neighbors, relatives, or
friends in return for a share of the offspring. The specific
conditions regarding the number of lamb/kids to be returned
and the time of repayment differs by location.
The extent of sharing is difficult to estimate, but without
doubt shared animals constitute an important part of the
flock of landless and subsistence farmers (Sabrani et al.,
1982d). Higher sharing-in rates were found in Garut and
Cirebon. This was caused by the "dropping" of adult animals
by an external agent (BPT/SR-CRSP). In each location,
about 100 animals were dropped. As it is the objective of such
schemes to increase the average flock per farmer, it is
instructive to examine the result of this dropping scheme.
In Cirebon, sharing-in rates and sharing-out rates are
virtually equal, indicating that at least part of the received
animals are shared "forward" to other farmers. Furthermore,
the high selling rate for female adult goats (35%) indicates
that some of these animals are also sold (Knipscheer et al.,
1983b). The total village sheep flock, however, has expanded
by 35 animals indicating that only the sheep farmers have
been able to absorb the dropped animals within their flocks.
This is due to the extensive grazing system that allows for
greater flexibility in flock sizes.
The number of ewes in Garut increased by only 28 head,
while the selling rate for ewes during the recall period was as
high as 46%. This selling rate cannot be maintained over the
long term and is undoubtedly caused by farmers selling their
own stock after having received animals from the dropping
scheme. This observation is confirmed by a further analysis
of the data which showed farmers selling and sharing-out
their animals during the same two months during which they
received the animals by the dropping scheme. These
observations cast serious doubts on the effectiveness of the
distribution of animals by gadohan dropping schemes in
areas where feed supplies are already fully used. Such
schemes are often launched without adequate study and
technical backstopping. It is therefore quite likely that the
Table 1. Environment and Land Use Parameter at Three Survey Locations in West Java
Cirebon Ciburuy Garut
Elevation 04 450 600-700
Topography (m) level undulating hilly
Soil texture silt loam sandy
Rainfall (mm/yr) 1,128 4,500 2,000
Temperature (o C) 28 23 25
Cultivated land (ha) 811 1,495
Uncultivated land (ha) 269 679
Rainfed rice fields (ha) 755 127 259
Dry land (ha) 56 13 1,236
Estate (ha) 0 46 0
Communal pasture (ha) 14 16 0
Forest (ha) 0 0 433
Source: SR-CRSP/BPT Survey, 1981.
Table 2. Average Size of Land Owned and Farmed by Holders of Small Ruminants by Location (ha)
Stratum Owned Farmed N Owned Farmed N
I. Landless 0.01 0.01 19 0.01 0.01 10
II. Subsistence 0.04 0.04 30 0.06 0.11 21
III. Smallholder 0.21 0.21 2 0.16 0.21 37
IV. Medium Holder 0.34 0.55 35 0.38 0.53 71
V. Large Holder 1.12 1.92 14 1.33 1.67 6
Weighted Average 0.29 0.48 0.29 0.40
increase in livestock numbers resulting from these distribu- participants with supplemental information provided by the
tion programs is only temporary. The main cause of the village monitoring surveys. The surveys in West Java
stagnant small ruminant population is the modest produc- (Sabrani et al., 1983b) and Central Java (Sabrani and
tivity of the animals rather than their numbers. Knipscheer, 1982) have identified the marketing channels
There are good prospects, however, to raise the produc- and marketing agents. A particularly critical figure is the
tivity of the animals. The reproduction rate of small village trader who controls many of the assembly and
ruminants in Indonesia is still only about one young animal distribution functions. In addition, these individuals often
per year per adult female. On specialized (commercial) farms, produce or share-out animals and are important for
reproduction rates of two young animals per adult female are introducing improved technology. Studies will continue on
often realized (Knipscheer et al., 1983a). The sheep farmers in the potential uses of these traders in introducing change. A
Ciburuy show that such a reproduction rate is also feasible study of Islamic holiday marketing was conducted by
for smallholders where food supplies are adequate. Soedjana (1982). The influence on prices during the 1982 Idul
The other major area of research has focused on the Adha festival was not as pronounced as was hypothesized.
The other major area of research has focused on the Gearing a production system to meet the demands of these
marketing and price formation process for small ruminants Gearing a production system to meet the demands of these
and the role of specialized agents in this process. An initial movable feasts may be an unduly risky strategy.
assumption was that the Islamic holidays led to lucrative
prices for small ruminants (particularly intact male sheep)
and that a focus of the SR-CRSP should be research on
management and feeding systems that would allow pro-
ducers to meet these market needs. The methodology used
has been short, focused interviews on specific market
Table 3. Flock Composition and Three Locations in West Java (1981)
of Flock composition (Head/ Farmer)a
Surveyed AM AF YM YF K/L
Cirebon (sheep) 30 0.18 3.33 0.32 0.62 0.83
(goats) 49 0.20 1.70 0.20 0.40 0.50
Ciburuy (sheep) 33 0.38 2.55 0.64 0.88 1.22
(goats) 30 0.34 2.48 0.36 0.86 1.20
Garut (sheep) 135 0.37 2.07 0.37 0.58 0.74
aAM = adult male
AF = adult female
YM = young male
YF = young female
K/L = kid or lamb
Source: SR-CRSP/BPT Survey, 1981.
Table 4. Income per Animal Unit per Location (West Java, 1981)
AU/Farmer Offtake Income/ Farmer* Offtake
(AU) (AU) (Rupiah) Income/AU (%)
(sheep) 0.1503 0.0798 39,302 261,490 53
(goats) 0.1195 0.0682 36,547 305,833 57
(sheep) 0.1781 0.1098 53,238 298,922 62
(goats) 0.1641 0.0805 40,104 244,388 49
(sheep) 0.1709 0.0770 41,466 242,633 45
*623 Rupiah = $1US at time of survey.
Source: Derived from SR-CRSP/BPT Survey, 1981.
Significance of Research Findings
The cooperative village program has identified several
phenomena that are of interest to the biological science
projects. First, the variability of animal weight gains is much
higher than expected (Thomas et al., 1982) given the full
confinement system and apparently fairly uniform levels of
feeding and management applied. Second, poor offtake rates
seem to be caused by long lambing/kidding intervals rather
than small litter sizes (Knipscheer et al., 1983b). Breeding
strategies are now a focal point of the Breeding Project.
Third, internal parasites may be more of a problem than
anticipated, particularly in Cirebon where seasonal grazing is
practiced. Fourth, production performance achieved by
specialized commercial producers with access to protein
supplements is far greater than that achieved in the villages
(Knipscheer et al., 1983a) and gives an indication of the
potential productivity gains by a straightforward transfer of
existing technology and management.
By synthesizing secondary data on sheep and goat
holdings with results of the SR-CRSP/BPT survey and
other surveys, we now have a much more comprehensive
picture of the economic role and the economic importance of
small ruminants in Indonesia. These studies indicate that
Table 5. Estimated Share of Small Ruminant (SR) Income of Total Income for Sheep and Goat Keepers at Two Locations in
West Java (1981)
No. Farms Estimated Annual SR-Income % SR Income in
Location Surveyed Income/Farm (1980) (1980-81) Total Income
Strata I + II 44 146,693 31,688 21.6
III 2 200,906 26,938 13.4
IV + V 33 318,899 43,038 13.5
Average (79) 220,000 37,593 17.1
Ciburuy (rubber plantations)a
Strata I V 66 180,000 46,671 25.9
Strata I + II 30 95,114 21,502 22.6
III 31 158,376 35,207 22.2
IV + V 74 442,391 45,692 10.3
Average (135) 300,000 41,466 13.8
aNo estimates for total income per stratum available.
Source: Derived from SR-CRSP/BPT Survey, 1981.
about every fifth farm household in Indonesia has small
ruminants which typically provide 15-25% of gross farm
income for those households. The sharing system is
widespread and additional economic studies are needed to
pinpoint resource allocation decisions between different
forms of the sharing system since these should have an effect
on biological productivity of the system. Possibilities of
expanding small ruminant numbers through animal distribu-
tion ("dropping') schemes was examined (Knipscheer et al.,
1983b). The impact was uneven and seemed directly related
to the elasticity of roughage feed supplies. Further use of
these schemes must be directed at areas with adequate feed
supplies or where the work being carried out by the Nutrition
Project can be implemented. Otherwise, the tendency is to
replace village stocks with the dropped animals. Economic
studies of commercial systems (Knipscheer et al., 1983a)
found considerable variation in economic returns per animal
unit but these systems were, in general, highly profitable
because of the high rates of compensatory gain made by
small ruminants purchased from villages and because of a
regular and low cost protein feed supply. Surprisingly, some
commercial operations were also engaged in breeding
Studies on marketing have focused on building profiles
which describe how the system operates, who the partici-
pants are, the general marketing channels, and how prices are
formed at each stage in the process. The marketing system is
characterized as diffuse (many small producers marketing
irregularly), well organized (regular marketing agents with
good knowledge of market conditions), with facilitating
functions including the holding of animals for further
fattening. Given the diffuse nature of the market, marketing
costs are relatively high, but there appear to be no major
sources of inefficiency and the marketing agents do not
appear to be making excess profits relative to the costs and
risks involved (Sabrani and Knipscheer, 1982; Sabrani et al.,
1983b; Soedjana, 1982). These studies identified a key
marketing agent, the village trader, who is found in most
villages with significant small ruminant numbers. Given this
person's regular contact with producers and his role as
holding, fattening, and sharing-out sheep and goats, future
work must consider this individual's role as a potential
change agent, demonstrator, and provider of inputs.
Finally, the underlying approach in most of the research
reported so far has been a farming systems research
philosophy. This has allowed teams of researchers to
participate in most of these studies and has helped focus the
biological research on problems relevant to village condi-
tions. The teams are now meeting regularly with villages and
an open exchange of ideas has developed. The approach is
becoming institutionalized within both BPT/Bogor and
Implications for Future Research
Research to date has concentrated on describing and
analyzing existing production systems, identifying produc-
tivity constraints, and setting in place research to eventually
improve the system. Situations have been identified where
if. .. s
Young boys herding sheep, Cirebon. West Java. In contrast to the
confinement system found in other areas, small ruminants in this location
are grazed on what amounts to communalpastures. Often individualfamily
flocks are combined andput under the care of I or 2 children to release adult
labor for other tasks. Photo: M. Nolan.
herd/flock expansion may be feasible such as on areas of
Java where villagers have access to plantations (Knipscheer
et al., 1983b) or in transmigration areas where farm sizes are
much larger (Mink, 1983). However, most situations studied
have limited opportunities to expand numbers (Knipscheer
and Soedjana, 1983; Mawi and Sabrani, 1982) and the
research must focus on improving offtake and animal size
with existing small herd/flock systems. One thrust in the
latter area was to examine productivity levels of efficient
commercial producers (Knipscheer et al., 1983a) and the
findings of this work will be incorporated into future work.
The sources of low reproduction are now being built into the
breeding program and the regular village meetings are
seeking ways to modify current breeding practices to shorten
lambing/ kidding intervals.
The studies of traditional feeding systems are also being
used to set up village trials at BPT sub-stations.
The future emphasis of the Economics Project will be to
evaluate specific technologies that are being developed, the
economic consequences of specific practices on different
groups of producers, how commercial producers' production
practices can be transferred to villages, how the village trader
can be used as a change agent, the role of women in specific
production and marketing tasks and how we can bring them
into the research and development process, and what type of
production strategy is needed to achieve the best possible
The end result, we hope, will be a set of specific
recommendations that can be utilized by BPT personnel, the
Dinas Peternakan officials, and the Government of Indo-
nesia to improve small ruminant production. The capacity is
building up within BPT to conduct the type of adaptive
farming system research required for small ruminants which
are an integral part of Indonesian small farm systems. Future
research must continue along these lines.
The training objectives have been to first develop the
general approach to economics research that enables the
economist to analyze sheep and goats as part of a larger and
very complex farming system. Both formal and informal
training has focused on the skills needed for effective field
survey work, analysis of data that reflected local producer
conditions, use of simple analytical devices such as hand-held
calculators, quick turnaround times between surveys and
publication of results, and participation in conferences and
workshops focused on interdisciplinary research on livestock
Specific training activities promoted by the Economics
Project in Indonesia are summarized below:
* Agus Muljadi is completing an MS program at Texas
A&M University (May 1981-August 1983). The thesis
work will use the baseline survey data that Mr. Muljadi
helped collect in 1980-81. He has also taken coursework in
sheep and goat production at Texas A&M.
* Sugiyanto completed an MS program at the Institut
Pertanian Bogor (August 1980-January 1983). His
research focused on statistical analysis of factors influ-
encing sheep output in the surveyed villages. Sugiyanto
was also responsible for tabulation and analysis of
periodical data on labor use by tasks.
* M. Sabrani is undertaking a PhD program at the
Department of Agricultural Economics, Gadjah Mada
University, Jogjakarta. The SR-CRSP will provide
partial support for thesis research.
Non-Degree Technical Training
* On-the-job training has been provided to Tjeppy
Soedjana, M. Sabrani, Uka Kusnadi, and S. Mawi on
design, analysis, and write up of surveys.
* A shortcourse in data analysis using a hand-held
calculator was conducted by Dr. Richard Bernsten in
March 1982. Twelve BPT staff participated.
* M. Sabrani and A. Muljadi attended the Third Inter-
national Conference on Goat Production and Disease in
Tucson, AZ, January 1982.
* M. Sabrani attended and presented a paper at the First
Asian-Australian Animal Production Congress, Kuala
Lumpur, September 1980. He also attended the Second
Congress, Manilla, November 1982 and presented two
papers on SR-CRSP research.
* M. Sabrani attended and co-authored a paper at the
IDRC Conference on "Livestock in Asia: Issues and
Policies," Singapore, February 1982.
* M. Sabrani acted as Co-Director and H.C. Knipscheer
acted as Lecturer, in "Course on Village Data Gathering,"
held February 28-March 18, 1983. This course was
sponsored by the Australian National University, at
BPT/ Ciawi. Five staff from BPT/ Bogor attended.
* Tjeppy Soedjana was supported in the course, "Technical
Writing in English," held in January 1983.
* Support was provided for Uka Kusnadi to participate in
an English course.
* Partial support was provided for the SR-CRSP/BPT
joint training workshop held July-August 1980.
* Support was provided for Tjeppy Soedjana to attend the
joint IDRC/ADC/Winrock Training Workshop in
"Socioeconomic Research Techniques for Livestock in
Asia," Bangkok, Thailand, April 18-May 18, 1983. Partial
support was also provided for H.C. Knipscheer and A.J.
De Boer to participate as lecturers.
Knipscheer, H.C. and T.D. Soedjana. 1983. The Productivity of Small
Ruminants in West Javanese Farming Systems. SR-CRSP Technical
Report Series No. 13.
Knipscheer, H.C., T.D. Soedjana, and A.J. De Boer. 1983a. New Small
Ruminant Technology for Small Farms in Indonesia: The Impact of
Flock Size. SR-CRSP/BPT Working Paper No. 6. Bogor.
Knipscheer, H.C., M. Sabrani, A.J. De Boer, and T. Soedjana. 1983b. The
Economic Role of Sheep and Goats in Indonesia: A Case Study of West
Java. Bulletin of Indonesian Economic Studies.
Knipscheer, H.C. and M. Sabrani. 1982. The Contribution of Goat and
Sheep Enterprises to West Javanese Farming Systems at Different
Income Levels. Paper Presented to Int. Conf. of Agr. Economists,
Mawi, S. and M. Sabrani. 1982. Sheep Management Systems in Two
Different Food Cropping Patterns. Ilmu Dan Peternakan 1(1).
Mink, S. 1983. Prospects for Small Farm Goat Production in a
Transmigration Area of Indonesia: Results of a Survey. SR-CRSP
Technical Report Series No. 10.
Muljadi, Agus, M. Sabrani, Agus Suporyanto, and Djoko Pramono. 1980.
Pengaruh Sistem Pemilikan Tanah dan Ternak Terhadap Kemung-
kinan Pengembangan Domba ian Kambing di Desa Ciburuy, Bogor
(The Effect of Land and Animal Ownership Systems on the Possibility
of Sheep and Goat Development in Ciburuy, Bogor). Lembaran LPP
Sabrani, M., A. Muljadi, and A.J. De Boer. 1983a. Small Ruminant
Production on Small Farms in West Java, Indonesia: Preliminary
Results of a Baseline Survey of Upland and Lowland Farming Systems.
SR-CRSP Technical Report Series No. 9.
Sabrani, M., S. Mawi, T.D. Soedjana, and H.C. Knipscheer. 1983b. A
Profile of Sheep and Goat Markets in West Java, Indonesia. SR-CRSP
Technical Report Series No. 12.
Sabrani, M. and H.C. Knipscheer. 1982. Sheep and Goat Markets in
Central Java: A Profile. SR-CRSP/BPT. Socio-Economics Working
Paper No. 5. Bogor.
Sabrani, M., P. Sitorus, and A.P. Siregar. 1982a. Peranan Ternak
Ruminansia Kecil Sebagai Sumber Protein Hewani di Pedesaan.
Journal Litbang Pertanian 1(2):68-72.
Sabrani, M., P. Sitorus, M. Rangkuti, Subandriyo, W. Mathius, T.D.
Soedjana, and A. Semali. 1982b. Laporan Survey Baseline Ternak
Kambing dan Domba. Mimeo, SR-CRSP/BPT, Bogor.
Sabrani, M., A.P. Siregar, R.J. Petheram, and H.C. Knipscheer. 1982c. An
Introduction to Livestock Farming Systems Research for Rural
Development. Working Paper, Farming Systems Program, BPT,
Sabrani, M., T.D. Soedjana, S. Mawi, and H.C. Knipscheer. 1982d. Profil
Pemasaran Ternak Ruminansia Kecil. Mimeo, SR-CRSP/BPT Draft
Report, Bogor. 47 pp.
Sabrani, M. and Agus Muljadi. 1981. Prospek Pengembangan Kambing
Domba Dengan Referensi Terhadap Pemilikan Tanah di desa Ciburuy,
Bogor (Prospect of Small Ruminant Development with Reference to
Land Holding in Ciburuy Village, Bogor). Lembaran LPP (1):1-4.
Sabrani, M., M. Panjaitan, and A. Muljadi. 1981. Prospek Pengembangan
Kambing dan Domba Bagi Petani Kecil dan Perlunya Pendekatan
Keilmuan ter Padu. Paper Presented to Seminar on Small Ruminant
Production, BPT, Bogor.
Sabrani, M. 1980. Analysis of Regional Demand for Red Meat in Java.
Paper Presented to I Asian-Australasian Congress on Animal Prod.,
Soedjana, T.D. 1982. Sheep and Goat Markets During the Islamic Holiday
"Idul Adha"in Bandung, Indonesia. SR-CRSP/BPT Socio-Economics
Working Paper No. 7. Bogor.
Soedjana, T.D. and H.C. Knipscheer. 1982. Productivity of Sheep Under
Upland Farming Conditions in Sukawargi Village, Garut District, West
Java. SR-CRSP/ BPT Socio-Economics Working Paper No. 6. Bogor.
Sugiyanto. 1983. Pendugaan Model Fungsi Produksi Usaha Ternak
Domba di Wilayah Garut dan Cirebon (Estimation of a Production
Function Model for Sheep Farming in Garut and Cirebon Districts).
MS Thesis, Institut Pertanian Bogor, Bogor. 67 pp.
Thomas, N., W. Mathius, and M. Sabrani. 1982. Small Ruminant
Production in West Java: Methodology and Initial Results. In J.C. Fine
and R.G. Lattimore (eds.), Livestock in Asia; Issues and Policies.
Ottawa, Ont. IDRC-202e. pp. 161-166.
Dinas Peternakan (Central Java). 1982. Buku Informasi Peternakan Jawa
Tengah, Ungaran, Central Java.
Obst, J.M., T. Boyes, and T. Chaniago. 1980. Reproductive Performance of
Indonesian Sheep and Goats. Proc. Aust. Soc. Ani. Prod., Perth.
Van Santen, C.E. 1980. Report on the Pilot Farm Survey 1980 in Cikajang,
Garut District and Jalaksana, Kuningan District, West Java Province,
Indonesia. FAO Regional Agr. Office for West Java, Bandung,
Winantea, A. and Yuliani Djuniarti. 1981. The Lambing Performance of
Sheep in East Java and the Influence of Season. Mimeo, Dept. Animal
Husbandry, Brawijaya University, Malang.
Nutrition and Feed
North Carolina State University
W est Java is characterized by good soils, high year-
round rainfall, very high rural population density,
and very small, intensively managed farms. Sheep and goats
are raised in small flocks, often in total confinement near the
villager's home. Their daily ration is harvested by hand by the
head-of-household or a member of his family and carried in
baskets to the animal pens.
The first task of the Nutrition and Feeding Systems
Project has been to find out what kinds of materials are used
as feeds in the villages, how much is consumed daily by the
animals, the chemical composition of village feedstuffs, and
the productivity of the village animals as related to quality
and quantity of feed consumed. These objectives have been
addressed by means of village survey and monitoring
activities. Village workers have been trained to collect feed
samples and dry them in ovens specially designed and built to
operate under village conditions. Samples properly dried,
preserved, and stored have then been transported to Bogor
for analysis of crude protein, cell wall fiber constituents,
important minerals, and in vitro digestibility of the dry
matter and fiber.
Simultaneously, feeding trials have been conducted at the
BPT experimental farms for the purpose of testing various
diet formulations and quantifying animal response. Empha-
sis has been placed on feedstuffs known to be generally used
or thought to have potential application in the villages.
As in other worksites, we are interested in building a data
base from which to improve the predictability of animal
performance from laboratory analyses. This objective is
particularly important for the Indonesia project for two
reasons: the wide variety of feeds available and in use makes
the attainment of this objective more feasible than in sites
where the feed base is more limited; and, the constant
fluctuation of feeds employed by villagers makes it more
critical to conduct a rapid test of feed quality rather than rely
on published composition data. BPT has as one of its
objectives the development of a feed composition data bank,
to which our project is contributing information. In data
banks of this type, however, the standard deviation is of as
much interest as the mean, and if the degree of variability is
high, it is especially important to analyze each new lot of feed
that one encounters.
The issue of animal nutrient requirements and relative
efficiencies is of interest, because villagers are faced with
choices between widely different breed types-the small
native Kacang goat, the large exotic Etawah goat, the native
West Java Thin-Tail ("Priangan') sheep, and the Fat-Tail
sheep that is more common to East Java. Information about
nutrient requirements and relative breed or species efficien-
cies is appropriately generated as part of thesis projects in
conjunction with experiments that also have an immediately
applicable practical objective.
Use of agricultural land in Java. The high population density combined with
the rich volcanic soils has resulted in a highly intensive cropping system. If
small ruminant production is to be increased in areas such as this without
competing with crops, crop residues will likely comprise a larger proportion
of livestock feed. Photo: N. Thomas.
During the three-year period of this report, considerable
activity has been undertaken toward the fulfillment of the
above-stated objectives. A list of project titles, leaders, and
current status follows:
* Baseline survey of small ruminant production systems in
villages of West Java. N. Thomas, I.W. Mathius, A.
Djajanegara. Completed in 1980.
* Monitoring of small ruminant production in West
Javanese villages. I.W. Mathius, N. Thomas, H.
Pulungan, J.E. van Eys, M. Rangkuti, A. Prabowo, W.L.
Johnson, A. Djajanegara, and village staff. In progress.
* Profile study on feeds and feeding of sheep in Sukawargi
village, West Java. I.W. Mathius, J.E. van Eys. Com-
pleted in 1982.
* Intake preferences for cassava, sweet potato, banana and
napier grass foliage by Priangan sheep and Kacang
goats. B. Haryanto, N. Thomas, W.L. Johnson. Com-
pleted in 1980.
* Cassava leaves as supplement for napier grass for lambs
and kids. I.W. Mathius, A. Djajanegara, M. Rangkuti.
Completed in 1981.
* Substitution of cassava meal for commercial concentrate
in sheep rations. H. Pulungan, B. Nusa Bakti (Margawati
Station, Garut). Completed in 1982.
* Utilization of leucaena leaf meal-molasses mixtures by
sheep and goats. A. Djajanegara, M. Rangkuti. Com-
pleted in 1981.
Gliricidia foliage as a protein supplement for growing
lambs and kids fed chopped napier grass. M. Rangkuti,
I.W. Mathius. First experiment completed in 1982;
second experiment in progress.
Cassava leaf and cassava root meal as supplements to
chopped napier grass for growing lambs. H. Pulungan.
Completed in 1982.
Energy and crude protein requirements of local goats. M.
Sitorus, T. Sutardi. Completed in 1981.
Trace mineral status of Javanese Thin-Tail sheep. A.
Prabowo, S. Lebdosoekojo, J.E. van Eys. Completed in
Effects of mineral supplementation on performance,
intake and blood characteristics in Javanese Thin-Tail
sheep. A. Prabowo, I.W. Mathius, M. Rangkuti, J.E. van
Eys. Completed in 1982.
* Ground nut straw as feed for growing lambs. I.W.
Mathius, M. Rangkuti, L.P. Batubara. Completed in
* Coconut oil cake and rice bran as supplements for lambs
fed pelleted napier grass. I.W. Mathius, M. Rangkuti,
L.P. Batubara. Completed in 1982.
* Chopped cassava root as an energy supplement for
growing lambs fed napier grass. H. Pulungan. In progress.
* Effects of supplementation with corn meal or rice bran on
utilization of cassava leaf-napier grass diets by growing
sheep and goats. I.W. Mathius, M. Rangkuti, H.
Pulungan. In progress.
* Mineral supplementation of sheep fed native grass diets.
A. Prabowo, J.E. van Eys, W.L. Johnson, H. Pulungan.
* Effects of sulfur fertilization and cutting interval on in
vitro degradation of cell wall constituents of various plant
parts from tropical forages. B. Betta, J.E. van Eys, W.L.
Johnson. In progress.
* Evaluation of small silos for preservation of tropical
forages and nutritive value of ensiled grasses for sheep.
J.E. van Eys, I.W. Mathius. In progress.
* Utilization of cassava root meal as an energy supplement
for growing lambs. J.E. van Eys, M. Sitorus, H.
Pulungan. In progress.
* Laboratory evaluation of forages and data collection on
feed composition. P. Mahyuddin (BPT/Ciawi; collab-
orative activity). In progress.
* Mineral supplementation requirements and development
of mineral licks. T. Panggabean (BPT/Ciawi; collab-
orative activity). In progress.
* Evaluation of soybean curd sludge as a supplement to
native grass diets for sheep and goats. H. Pulungan, J.E.
van Eys, M. Rangkuti. In progress.
* Mineral status of sheep and goats as affected by level of
dietary energy. A. Prabowo, J.E. van Eys, M. Rangkuti,
W.L. Johnson. In progress.
* Treatment of fibrous residues with emphasis on rice
straw. M. Winugroho (BPT/Ciawi; collaborative activ-
ity). In progress.
The paragraphs that follow present a brief summary of the
highlights that have emerged from the above experiments,
particularly those already completed. The reader who wishes
more detail is referred to the publication list at the end of this
Village Baseline Surveys
In order to design the detailed Village Monitoring Survey,
a preliminary baseline survey was conducted with a larger
sample. A total of 374 families were surveyed in three
villages. There were important locational differences in type
of feeding practiced. In the highland location (Garut) only
sheep are kept, and always in confinement. In the lowland
location (Cirebon) sheep are usually allowed to graze but
goats are always kept in confinement. Across locations, the
percent of farms using certain feeds were: native grasses,
95%; banana leaves, 28%; maize tops, 28%; and cassava
leaves, 18% (Djajanegara et al., 1982).
Village Monitoring Results
Monitoring sites were selected to be representative of the
different climatic and geological strata of West Java. Garut
represents the upland site, Cirebon the coastal plains, and
Ciburuy an intermediate site with a large estate crop
(rubber). Data from the first year of monitoring are still being
analyzed. Some preliminary observations are possible,
however (Mathius et al., 1983c). Important differences in the
management of the small ruminant enterprise were observed
among sites. In Garut, animals are maintained under
complete confinement. Two-thirds of all activities related to
sheep production are carried out by the head of the family.
Women (wives) are responsible for 4% and children for 23%
of the activities. In Cirebon and Ciburuy, both goats and
sheep are found, but each farmer has only one species. Goats
are kept in complete confinement while sheep are allowed to
graze on the rice stubble (Cirebon) or in the rubber
plantation (Ciburuy). In these villages, husbands are
responsible for 45% of all activities related to small ruminant
production, wives for 15% and children for 40%. The
relationship between family size and size of animal enterprise
was poor for all three locations.
Although the number of confined animals and the average
liveweight per farm is similar among villages, the amount of
feed collected per animal or per unit of liveweight differs
considerably. This was most pronounced during the wet
season. Garut farmers offer the largest amount (277 g/kg
BW-wet season) and Cirebon farmers the smallest amount
(226 g/kg BW-dry season). Assuming an average of 25%
dry matter (DM), these amounts of feed on offer are in excess
of the animals'daily requirements (about 50 g DM/kg BW).
However, only if the feed were of highest quality would daily
intake reach the level of 50 g DM/kg BW. Casual
observation, reinforced by animal performance records,
would suggest that actual intakes are somewhat lower and
that performance is limited by the digestibility of the feed
offered, or perhaps by specific nutrient deficiencies.
Sheep and goats in Garut and Ciburuy seemed to have
more opportunity for diet selection and higher DM intakes
than in Cirebon. Sheep production under grazing in Cirebon
was depressed due to high endo-parasite infections (par-
ticularly Haemonchus spp., Trichostrongylus spp., and
Paramphistomum spp.). Observed death losses were as high
as 25%. Endo-parasite infections were of less importance for
grazing animals in Ciburuy, and not observed in confined
animals in any location.
Comparisons of liveweight gains between locations
indicated highest weaning weight and daily gain for animals
in Ciburuy, while Cirebon was lowest. Protein and mineral
content of diets on offer do not explain the observed
differences in gain. Crude protein levels were highest in
Cirebon and are higher than NRC requirements for growing
sheep and goats in all locations. This is also true for Ca and P.
The high levels of total cell fiber, generally above 55% and
highest in Cirebon, play a role in reducing intake and animal
Much of the difference in nutritive value of small ruminant
diets can be attributed to differences in botanical compo-
sition. In Ciburuy, the proportion of native grasses was
relatively constant across seasons, ranging from 70 to 85%. In
Cirebon and Garut, the contribution of native grasses ranged
between 45 and 85% with the lowest proportion occurring
during the dry season. In Cirebon, crop by-products were fed
at 20 to 35% of the diet and the rest was made up of tree
foliage or waste products from home or markets (e.g.,
jackfruit and pineapple peelings). In Garut and Ciburuy, the
remainder of the diet was made up of crop by-products such
as cassava leaves, maize tops, sweet potato vines and melon
or pumpkin leaves. Tree legumes constituted only a minor
proportion of the total ration in Garut and Ciburuy.
The total liveweight of animals carried per farm was only
slightly affected by season in spite of the fact that significantly
larger amounts of feed were available in the wet season,
especially in Garut. Much of the wet season increase in
quantity was from native grasses, which may be lower in
nutritional value. Feeding of crop by-products decreased
proportionally but remained quantitatively constant. Crop
by-products enter the feed supply as a result of harvest and
are not directly considered to be a feed resource. This attitude
may change if adequate methods of crop preservation can be
developed and if a nutritional superiority for certain crop
by-products can be demonstrated.
Results to date indicate that the most important constraint
to flock size is the amount of fodder a farmer and his family
can collect daily. If increases in animal performance are
important, improvements in feed quality rather than
quantity will be necessary.
An effect of season (dry vs. wet) on diet chemical
composition could be observed in all locations. Crude
protein levels increased during the wet season, probably due
to an increase in the proportion of more immature native
grasses. Concentration of total cell wall fiber (NDF),
calcium, and phosphorus remained relatively constant.
A major seasonal difference between locations could be
observed in growth rates of young stock. In Cirebon, growth
rates dropped dramatically at the beginning of the wet
season, which could be due to health problems or mineral
deficiencies. Micro-mineral analysis of liver samples indi-
cated significantly lower concentrations of copper and zinc in
the wet season than in the dry season.
Additional supplementation above the all-forage diet is
observed only sporadically. Occasionally, the feeding of
soybean curd sludge or rice bran may be encountered.
Mineral supplementation is limited to salt at irregular
intervals, if at all, and the purchase of any supplement is
generally considered too expensive, or farmers are not aware
of the benefits.
Intake and Digestibility of Village Feedstuffs
The evaluation of feedstuffs available to villagers is an
important line of research at the BPT station in Bogor. In a
trial with eight Priangan lambs and eight Kacang kids
(Haryanto et al., 1982), both species showed an over-
whelming preference for cassava leaves when they were
offered free choice and paired with either sweet potato vines,
banana leaves, or chopped napier grass. Significant species
differences were noted for banana leaves (preferred by goats
over sweet potato vines or napier grass, but rejected by sheep)
and napier grass (equal preference by sheep to sweet potato
vines, but rejected by goats). Half the animals were offered
200 g of concentrate per day in addition to forage. The sheep
readily ate the concentrate, reducing their forage DM intake
by 112 g/day. The goats, however, ate very little concentrate
and actually slightly increased their forage intake.
In another trial, leucaena leaf meal and molasses were fed
ad libitum to 20 wether lambs and 20 kids in proportions of
100/0, 80/20,60/40, and 40/60, as a supplement to chopped
napier grass which was also offered ad libitum. Maximum
intake of the total ration (53 g/kg W.75) and maximum DM
digestibilities (56% for sheep, 65% for goats) were obtained
with the 60/40 supplement. Goats tended to prefer the
rations with molasses, while sheep tended to prefer those
without. Also, the goats consumed a higher proportion of
napier grass when the supplement contained molasses, but
when it did not, the sheep consumed higher levels of napier
grass. Regardless of diet, the digestibility coefficients for DM
averaged eleven percentage units higher for goats than sheep.
Growth Trials with Javanese Thin-Tail Sheep and Kacang
A series of trials have been conducted with lambs or kids in
which a basic ration of napier grass has been supplemented
with fresh or wilted gliricidia foliage, wilted cassava leaves,
cassava root meal plus urea, a commercial concentrate
mixture, peanut straw, coconut oil cake, or rice bran.
Lambs (initial weight 11.6 kg) and kids (9.9 kg) were fed
gliricidia maculata either wilted (six lambs) or fresh (six
lambs and six kids), and chopped fresh napier grass. Both
forages were offered ad libitum. Excellent growth rates were
attained over a six-week period for the lambs (107 g/day) and
kids (82 g/day). Average voluntary DM intake per kg W.75
was 63 g/ day of gliricidia by sheep, 53 g of gliricidia by goats,
and 31 g of napier grass by both species. Weight gains and
DM intakes were higher but DM digestibility was lower
(P < .05) for the sheep consuming wilted gliricidia, compared
to fresh. It would appear that this tree legume has excellent
potential as an energy and protein source for small
Cassava is a common cash crop throughout West Java; its
foliage is high in crude protein content and palatable to
ruminants. Cassava root meal is a highly digestible energy
source because of its high starch content but contains very
little protein. In a series of growth trials, sheep fed wilted
cassava leaves as a supplement to adlibitum chopped napier
grass did not increase total DM intake but did consume more
crude protein, with average gains of 50 to 60 g/day. Adding
50 or 100 g/day of cassava root meal did not change the
results. Goats reacted differently from sheep in that total
intake was higher with cassava leaves but average gains were
only 23 g/day. Goats may be more sensitive to marginal
levels of hydrocyanic acid in cassava products.
When 20 wether lambs (14 kg average initial weight) and
20 male kids (19 kg) were fed adlibitum napier grass alone or
with .5, 1.0 or 1.5 kg of cassava leaves per day, the
digestibility of DM, protein, and NDF was higher (P < .05)
by goats than sheep. These levels of cassava leaf did not
influence digestibility of the total ration by either species.
Trials with cassava root meal have proven inconclusive; this
product needs further testing for both goats and sheep.
Mathius et al. (1983b) offered peanut straw at 100, 200 or
400 g/day to 12 lambs (initial weight 18 kg), in addition to ad
libitum chopped napier grass. Average gains over a 12-week
period were 23 g/day for the lowest peanut straw level (actual
intake of peanut straw was 91 g/day) and 36 g/day at the
highest level (actual intake, 297 g/day). Intake of DM and
digestible DM was higher with peanut straw, but the DM
digestibility of the total ration was lower.
Coconut oil cake was offered to 15 lambs (initial weight 16
kg) at 100, 200 or 400 g/day in addition to ad libitum napier
grass pellets. DM intake increased at supplementation levels
up to 200 g/day; DM digestibility was highest (62%) on the
same treatment. Average gains were influenced very little by
coconut oil cake supplementation (Mathius et al., 1983a). In
a separate but identically designed growth and digestibility
trial, rice bran replaced coconut oil cake with similarly small
effects on total DM intake. When 340 g/day of rice bran was
consumed, average gains were still only 42 g/ day. The quality
and form of napier grass undoubtedly influenced the
outcome of these trials.
Energy and Protein Requirements for Growth of Kacang
Twenty-seven male Kacang kids, age eight to ten months
and with initial weights of 15 + 2 kg, were assigned to one of
nine diets for an 8-week growth study followed by seven days
of total fecal collection for digestibility estimates. At the end
of the trial, all animals were slaughtered for carcass
composition estimates. Diets consisted of chopped napier
grass, rice bran, maize meal, coconut oil cake, cassava waste,
soybean oil cake, urea and minerals to provide three levels of
energy (2.6,3.0, and 3.2 Meal DE/kg DM) and three levels of
crude protein (10.3, 13.1 and 15.3%), in a factorial design.
Increased levels of dietary energy caused faster daily gains
and greater protein retention (P < .01); DM digestibility also
increased. The fat to protein ratio of the carcass increased
with the level of dietary energy but decreased with higher
dietary crude protein. The intake level of 230 kcal of
digestible energy (DE) and 88 g of crude protein per kg of
body weight'75 resulted in the highest nitrogen retention.
Maximum DM digestibility was achieved at an intake of 196
kcal DE/kg BW.75. DM intake was relatively constant across
treatments, with a mean of 440 + 15 g/ head/day. Daily gains
were maximized at 195 kcal DE and 8.2 g CP/kg BW.75
(Table 1). Average gains were not affected by the level of
protein in the ration but energy intake had a large effect
It seems that in diets providing more than 140 kcal DE/ kg
BW75/day, increasing the level of crude protein above 10%
had little effect on the growth of local goats. However, large
improvements in daily gain might be expected from an
increase in digestible energy content of the diet up to 200 kcal
DE/kg BW75/day (Sitorus, 1983).
Concentrations of copper, iron, zinc and manganese were
measured in liver samples from 32 sheep slaughtered over an
eleven-month period in Garut (upland) and Cirebon
(lowland) villages (Prabowo et al., 1983). Iron concentration
was 298 + 26 ppm which was considered normal; no location
or seasonal effects were noted. Results for the other minerals
are given "in Table 2. Copper values are also considered
normal, but varied significantly from the wet to dry seasons.
Zinc and manganese values are considered below normal.
The locational differences for these two elements (lower
values for Garut) and the seasonal differences (lower values
in the wetter season) may be related to situations of greater
availability of feed and consequent higher metabolic activity
of the enzymes which incorporate these minerals.
A complete mineral mix was added in graduated levels to
100 g of concentrate mix and fed daily to twelve lambs. Three
additional lambs received only the concentrate (no minerals)
and three received no supplement. All lambs were offered ad
libitum chopped napier grass. The group consuming only
napier grass gained only 8 g/day. The napier grass/concen-
trate group gained about 36 g/day, while the supplemented
groups gained 69 g/day and experienced a 25% increase in
feed DM intake over the napier/concentrate group. Ten
grams of mineral supplement per animal per day was
Significance of Research Findings
Although much work remains before we have a complete
understanding of present small ruminant feeding practices in
West Javanese villages as well as the best strategy for
bringing about improvements, several preliminary con-
clusions are possible.
Crude protein levels of village diets for small ruminants
appear adequate, but digestible energy intake appears to be a
limiting factor in present village feeding systems. This could
be related to a high content of cell-wall fiber in tropical
grasses, which means that they will be more slowly digested
by goats and sheep, and that a ceiling will be imposed on total
DM intake. The addition of wilted cassava leaves has been
shown to increase dietary intake for lambs sufficiently to
allow average gains of 55 g/day. Gliricidia foliage as a
supplement for lambs will result in average gains of 90 g/ day.
Deficiencies or impaired metabolism of certain micro-
minerals may also be a problem. This is an area that needs
The local Kacang goats respond to supplementation
differently from the West Java Thin-Tail sheep. At adequate
levels of digestible energy, goats consume more low or
medium quality forage than sheep. However, sheep seem to
utilize unsupplemented low quality diets better than goats
and respond better to protein supplementation in terms of
DM utilization and weight gain.
Table 1. Average Daily Liveweight Gains (g/day) of Male Kacang Kids Fed Diets with Different Levels of Digestible Energy and
Energy Intake Intake of Crude Protein (g CP/kg BW-75/day)a
Energy Intake --------------------------
(kcal DE/kg BW75/day) 5.8 7.55 8.8 mean
142 -12 6 4 -1
170 28 30 29 29
195 67 70 73 70
aCorresponding dietary crude protein concentrations were 10%, 13% and 15.5%.
Source: Sitorus, 1983.
Table 2. Micro-Mineral Concentration (ppm) in Sheep Liver Samples from Two Locations in West Java, in the Wet and Dry
Copper Zinc Manganese
wet dry wet dry wet dry
Cirebon-lowland 236a 265b 71a 84b 7.7 8.1
Garut-upland 229a 254b 38a 44b 4.8a 5.5b
a,bValues for each element with different superscripts in the same row are significantly different (P < .05).
Energy and protein requirements of local goats for
maintenance and weight gain are in close agreement with
recently published NRC requirements.
Implications for Future Research
In reality, the research program for small ruminant
nutrition and feeding systems in Javanese villages has only
just begun. Much work remains if the project objectives are
to be fulfilled. However, initial results can be of value in
bringing to focus the necessary next steps.
In the village monitoring program, a new simplified
questionnaire will emphasize practices of immediate nutri-
tional importance. Feed sampling will continue and will be
expanded to allow quantitative estimates of nutrients
actually consumed by the animals as well as feed nutrients
placed on offer. The feed offered and the next day's feed
remaining will be weighed on selected farms at regular
intervals; samples will also be taken of the feed refused.
Techniques will be developed (fecal nitrogen index, or inert
external markers such as ytterbium) to provide rapid indirect
estimates of intake.
Feed evaluation will continue but with more rapid
performance of chemical analyses (crude protein,.important
minerals, and cell wall fiber constituents including cellulose,
hemicellulose, and lignin). Attention needs to be paid to
anti-quality factors such as tannins and cyanogenic com-
pounds. Rates of cell wall degradation will be measured in
selected important feeds, as this is the main determinant of
total dry matter intake.
More emphasis will be placed on native grasses since they
are the most abundant feed resource and probably the
cheapest in terms of labor and management input. There is,
however, likely to be an optimum ceiling beyond which
native grasses should not be offered to animals and
supplementation with higher quality crop by-products or tree
legumes such as leucaena and gliricidia will probably be
profitable in most cases. A few commonly available
industrial by-products, such as soybean whey ("ampas tahu')
need to be evaluated.
Feeding trials at the BPT farms will continue to quantify
the growth response to measured levels of energy, protein
and mineral intake. This line of work needs to be expanded
to mature ewes and does in order to measure their response in
terms of reproductive performance.
Feed preservation technology must be adapted to and
tested under village conditions. Small scale storage may be
important in improving the availability of high quality feeds
for the dry season, or for use during temporary periods when
labor supply is diverted to meet periodic high demands in rice
cultivation or other crop enterprises.
More intensified studies of the rumen physiology of native
sheep and goats are called for. Differential indices of
selectivity, rumen degradation rates, particle size reduction
by mastication, and pass-through rates may indicate that
different feeding strategies are required if both species are to
perform to their economic optimum.
In all of this work, an expanded level of collaboration
outside of BPT/ Bogor will be desirable. Collaboration with
scientists at BPT/Ciawi has begun and will accelerate. A
more intensified involvement with animal nutrition faculty
and graduate students at Gadjah Mada University and
Bogor Agricultural Institute will also be helpful.
Now that preliminary findings are available, it is time to
begin a formalized outreach activity. This should involve a
two-way dialog with researchers and teachers at other
institutions, with extension program leaders throughout
Java and, to a limited extent and on a pilot basis, directly
with farmers-as has already begun to take place in the
villages of the monitoring program. A proposal for a project
along these lines is currently being considered by the USAID
Mission in Indonesia.
The first training activity was held in August 1980 for the
young men who had been selected as village workers. This
involved classroom and field activities. The in-service
training of village staff has since been a continuous activity.
Agreement in principle has been given to eventually help
support these individuals for university training in Indonesia.
In-service training of the young research staff at BPT is a
daily activity and one which has had high priority for both
Neil Thomas and Jan van Eys, as well as for senior BPT
Support has been given for English language training to
four young scientists who are potential candidates for
graduate study abroad.
Training village staff in feed separation and identification, Tenjonegara,
near Garut, West Java. Photo: N. Thomas.
Indonesians already studying abroad include Ms. Sorta
Silitonga, who was provided a part-time assistantship
stipend while studying for her MS degree at the University of
Minnesota. Her program was successfully completed in
March 1983. Also, Mr. Budi Haryanto is studying at NCSU
under a Government of Indonesia/ World Bank fellowship.
He completed his MS in June 1983 having already started his
PhD coursework in the fall of 1982. He plans to undertake
his PhD thesis research in Indonesia.
At least three other young scientists are potential
candidates for MS study at NCSU or other US universities.
A fourth person is now working out arrangements to start a
PhD thesis project with the intention of later enrolling for the
coursework portion of her program at NCSU; a fifth person
has expressed an interest in a PhD program in the United
States. Yet another young researcher (Marudin Sitorus) has
enrolled at the Bogor Agricultural Institute for his PhD; the
Principal Investigator is serving on his committee of advisors
and will help in the design of his thesis research. Mr. Sitorus
may also spend one semester as a visitor at NCSU for
Other BPT nutritionists are being supported for graduate
training by the Australian government (Andi Djajanegara
from Bogor, Thamrin Panggabean from Ciawi).
Short-term training in the United States was supported for
Mr. Andi Djajanegara who spent a week in Dr. Van Soest's
laboratory at Cornell University and two weeks at Raleigh
looking at nutrition laboratory management procedures.
Several BPT scientists have been supported for inter-
national travel to present papers at scientific meetings in
Singapore, Malaysia, the United States and Japan.
Djajanegara, A. and M. Rangkuti. 1983. The Palatability of Leucaena Leaf
Meal for Sheep and Goats. Bulletin Lembaga Penelitian Peternakan
No. 32. In Press.
Djajanegara, A. and I.W. Mathius. 1983. The Utilization of Cassava Leaves
by Sheep and Goats. Proc., Seminar Pemanfaatan Limbah Pangan dan
Limbah Pertanian untuk Makanan Ternak. Yogyakarta.
Djajanegara, A., N. Thomas, and W.L. Johnson. 1982. Patterns in Goat and
Sheep Production in Java. Proc., Ill Intl. Conf. on Goat Prod. and Dis.
Tucson, AZ. p. 518.
Haryanto, Budi. 1983. Digestibility and Retention Time of Forage Fiber as
Affected by Level of Intake in Sheep and Goats. MS Thesis, North
Carolina State University, Raleigh.
Haryanto, B. and W.L. Johnson. 1982. Rumen Inoculum from Sheep or
Goats on Two Intake Levels of Alfalfa or Coastal Bermudagrass for In
Vitro Digestibility. Abstracts, Joint ASAS-CSAS Annual Meeting,
Haryanto, B., W.L. Johnson, and N. Thomas. 1982. Intake Preferences for
Cassava, Sweet Potato, Banana and Napier Grass Forages by
Indonesian Sheep and Goats. Proc., Ill Intl. Conf. on Goat Prod. and
Dis. Tucson, AZ. p. 279.
Mathius, W. and J.E. van Eys. 1983. Feeds and Feeding of Sheep in
Sukawargi Village, West Java; a Profile Study. Working paper No. 12.
North Carolina State University, Raleigh and Balai Penelitian Ternak,
Mathius, W., M. Rangkuti and L.P. Batubara. 1983. The Use of Peanut
Straw in Sheep Rations. Proc., Seminar Pemanfaatan Limbah Pangan
dan Limbah Pertanian untuk Makanan Ternak. Yogyakarta.
Mathius, I.W., A. Djajanegara, L.P. Batubara, and M. Rangkuti. 1983a.
Supplementation of Grass Rations with Coconut Oil Cake for Growing
Lambs. Bulletin Lembaga Penelitian Peternakan No. 32. In Press.
Mathius, .W., M. Rangkuti, and L.P. Batubara. 1983b. Ground Nut Straw
as a Feedstuff for Sheep. Bulletin Lembaga Penelitian Peternakan No.
32. In Press.
Mathius, W., H. Pulungan, and N. Thomas. 1983c. A Preliminary Analysis
of Results from the Nutrition Village Monitoring Program. Working
Paper No. 11. North Carolina State University, Raleigh and Balai
Penelitian Ternak, Bogor.
Mathius, W., J.E. van Eys, A. Djajanegara, and M. Rangkuti. 1983d.
Effects of Cassava Leaf Supplementation on the Utilization of Napier
Grass by Sheep and Goats. 5th World Conf. on Animal Prod. Tokyo.
Mathius, W., J.E. van Eys, and N. Thomas. 1982. Aspek Nilaigizi Makanan
Dalam Usaha Peternakan Domba Kambing di Jawa Barat (Nutritional
Aspects of Sheep and Goat Production in West Java). Seminar
Teknologi Peternakan untuk Menunjang Pengembangan Pedesaan.
Malang, East Java.
Prabowo, A., J.E. van Eys, W. Mathius, and Soekanto Lebdosoekojo.
1983. Trace Mineral Status and Effect of Mineral Supplementation in
Javanese Thin-Tail Sheep. 5th World Conf. on Animal Prod. Tokyo.
Prabowo, A., D. Samaih, and M. Rangkuti. 1983. The Use ofAmpas Tahu
as a Concentrate in Fattening Sheep Rations. Proc. Seminar
Pemanfaatan Limbah Pangan dan Limbah Pertanian untuk Makanan
Prabowo, A., W. Mathius, M. Rangkuti, S. Partoutomo, and N. Thomas.
1982. Some Aspects of Anthelmintic Treatment for the Improvement of
Small Ruminant Production under Village Conditions. Seminar
Teknologi Peternakan untuk Menunjang Pengembangan Pedesaan.
Sitorus-Silitonga, Siti Sorta. 1983. Evaluating Forages. Digestibility Studies
of Alfalfa and Brome Hay using Sheep. MS Thesis, University of
Minnesota, St. Paul.
Sitorus, Marudin. 1982. Kebutuhan Kambing Lokal akan Energi dan
Protein (Energy and Protein Requirements for Local Goats). MS
Thesis, Bogor Agricultural Institute.
Thomas, N. and M. Rangkuti. 1983. Small Ruminant Production in the
Small-Farm Perspective. Working Paper No. 10. North Carolina State Socioc I As
University, Raleigh and Balai Penelitian Ternak, Bogor.
Thomas, N. 1982. Constraints to Improving Small Ruminant Production Sm all Rum inant
on Multi-Enterprise Smallholdings in the Humid Tropics. Proc., III
Intl. Conf. on Goat Prod. and Dis. Tucson, AZ. p. 292. Production
Thomas, N., W. Mathius, and M. Sabrani. 1982. Small Ruminant IP Oduction System s
Production in West Java: Research Methodology and Initial Results.
Proc., Conf. on Livestock Dev. in Asia. Singapore. M.F. Nolan
University of Missouri, Columbia
In February 1980, when the formal memorandum of
understanding was signed between the SR-CRSP Man-
agement Entity and the Indonesia Ministry of Agriculture,
the Indonesian Animal Husbandry Research Institute
(subsequently called BPT) had only a limited program in
socioeconomics and no trained sociologist on its staff. Yet, to
the credit of the Institute's leadership, it was recognized that it
was important to have a viable sociology and economics
component within the Institute's structure in order to
adequately understand the human component of the animal
enterprise. Consequently, one of the first goals of the
Sociology Project in Indonesia was to strengthen the
capacity of BPT to engage in village oriented research on the
social aspects of small ruminant production in Java.
The Indonesian farming system is complex and the role
that small ruminants play within it varies. While crop
production occupies the majority of a family's resources
(labor, capital and land), small ruminants are considerably
more than a trivial or hobby activity. The over-arching goal
for the Sociology Project was to develop an understanding of
the social and cultural contexts in which small ruminant
production occurred. This is necessary if specific techno-
logical interventions are to be evaluated for their social
soundness and the likelihood of their acceptance by the
population as a whole. It will do little good for the biological
projects to propose new methods of management, or some
other intervention, if they are either unacceptable to the
population by virtue of their conflict with some dominant
cultural norm or if they actually harm rather than help the
small farmers to whom the program is directing its attention.
As a step toward reaching these goals, the following
specific objectives have formed the locus of activity of the
Sociology Project since its inception in 1980:
To describe the framework within which small ruminant
producers in West and Central Java allocate resources
and make decisions. Among those areas considered
important were labor allocation, social status consider-
ations, and division of labor within families.
Once the initial characterization was completed, to
develop more specific studies on issues which emerged
from the initial work. As the study developed, significant
questions regarding the role of women, the characteristics
of nonproducers of small ruminants, the role of the village
trader in small ruminant production and marketing, and
the implications of animal sharing arrangements were
largely unanswered by the baseline survey.
* To increase the ability of BPT to carry out sociological
analyses with its own staff resources.
Throughout, the emphasis of research has been to provide
the biological projects with social criteria by which they
might evaluate the potential value of their research.
Ultimately, the Sociology Project will have to "service" the
biological projects in the SR-CRSP to the extent of
providing them with an understanding of the implications/
feasibility/desirability of those interventions they suggest.
We are only now entering this phase.
In order to develop an initial characterization and
description of the role small ruminants occupy in the
Javanese farming system, an interdisciplinary baseline survey
was undertaken in two West Java villages. In addition to
collecting data on the social, economic, and demographic
characteristics of the sample families (in this case the samples
were drawn from sheep and/or goat producers in each
village), the sociology portion of the data collection effort
ascertained the extent and nature of farmers' contact with
various institutions (credit, markets, and extension), their
attitudes and values, particularly as they would relate to their
willingness to adopt new farm practices, and collected
information on nutritional intake and migration patterns.
These data formed the core of an MS thesis of the BPT staff
member who was trained at the University of Missouri. In
addition to this activity in West Java, a similar effort was
undertaken in Central Java through collaboration with the
Faculty of Economics at Satya Wacana University.
In both the West and Central Java locations, the villages
chosen for study were selected because they represented
different production systems-a coastal, intensive rice
production system and an upland, mixed-crop system. In the
case of the upland village in West Java, the situation was
Table 1. Descriptive Data on Sample Villages (1980)
further complicated by the fact that this particular area was
the location of the locally famous fighting sheep which were a
highly valued and prestigious animal in that area.
One hundred and forty-five farmers were chosen as a
sample from the upland site which was representative of the
region of Garut, while a lowland sample of 100 farmers
represented the region of Cirebon (see Table I for a
description of the sample villages). The samples were divided
into five groups based on farm size. The two study sites were
characterized by intensive farming and animal production
activities. Multicropping predominated in the upland com-
munities while wet rice monoculture prevailed in the lowland
village. Animal production activities in the upland site were
dominated by sheep production while villagers in the
lowland site raised both goats and sheep. Sheep were raised
confined in pens in the upland site as were goats in the
lowland villages (Table 2). While individual ownership was
the most common form of animal raising, some sharing
arrangements did exist among the lowland farmers. In both
sites, there was no significant relationship between farm size
and the number of animals owned. Men and women shared
the activities in crop production relatively equally in both
locations. However, the sexual differentiation of labor was
more pronounced for animal raising where male household
members participated more than females (Table 3). Male
household heads had the greatest role in sheep and/or goat
raising whereas women were responsible for most of the
decision making in poultry production. In both sites,
husbands played the key role in deciding what farm products
would be sold, when they would be marketed and how the
money would be used. The larger the farm, the greater the
frequency in which villagers sold farm products.
Farmers preferred to obtain credit from formal govern-
mental institutions and they were responsive to advice from
extension agents. In actual practice, however, most found
that they could not get funds from government institutions
and they relied instead on the informal money lender system.
Description VII.1 VII.2 VII.3 VII.4
1. Altitude, m 600 600 10 10
2. Area, ha 736 1,438 415 665
3. Population 7,550 5,485 4,130 6,288
4. No. of household 1,930 1,556 726 1,392
5. Density/sqkm 1,026 381 995 945
6. Cultivated area, ha:
-wet rice field 154 105 366 389
-dry field 495 741 54
7. Non-cultivated area 87 592 42 -
8. No. of sheep/goats 1,900 2,028 166 443
Table 2. Sheep and Goat Management
Raising Methods (Sheep) (Sheep) (Goats)
................. ................................. percentage ...........................-.....-- --- --
1. Grazed 6 74 57
2. Confinement 85 26 100
3. Combination 1 & 2 9 0 0
Sample Size 145 43 57
Similarly, since extension personnel seldom visited the study
sites, villagers were forced to rely on neighbors for advice
In data that were available only from the Garut (upland)
site, the main reason given for raising sheep was to acquire an
asset which could easily be turned into cash (Table 5). Farm
families rarely consumed their own animals even though they
felt it was important to serve meat.
Sheep are the highest status animals in Garut villages and
high quality sheep were priced accordingly. In the highland
villages, sheep gave status to villagers and 77% of the farmers
stated that sheep received special attention whereas only 23%
indicated that poultry was the most preferred animal. Most
farmers (77%) stated that raising high quality sheep was more
important than owning a large number of animals, regardless
of quality. There was a predisposition on the part of farmers
in the highland villages to improve management techniques
for both sheep and poultry and this was especially true for
larger farmers. However, when asked about specific changes
they had made during the past two years and their
management techniques or if they had borrowed money to
buy animals, only a handful of responses indicated that they
had done anything. One cannot be sure if this represents their
true feelings or if they simply did not have access to new and
reliable information about improved animal husbandry.
In response to questions addressing more general atti-
tudinal belief systems, there was an interesting intersection of
traditional and modern beliefs in the farmers of the Garut
villages. The vast majority of respondents felt that traditional
crop priorities were not necessarily better and to get ahead
one must take chances, yet they felt that success was more
dependent on God's will than man's effort and that the sons
of businessmen had a better chance of success than the sons
of farmers. Eighty-five percent agreed with the statement that
success in farming depended more on God than on the efforts
of man. Thus, in developing a program to modernize
agriculture in this society, efforts should recognize not only
technical issues but also the spiritual dimensions that govern
Overall, it is important to recognize that there are
considerable differences among Javanese farmers. Village
location, relative wealth, and farm size all seem to provide
some points of differentiation when one talks about devising
programs for sheep and goat producers. Interestingly
however, differences in herd sizes across categories of farm
size were negligible. Sheep and goats appeared to be a
commodity which, by the nature of the management system
(i.e., a confinement, cut and carry system of feeding for the
most part) could be kept by nearly any farmer-even those
with little or no land.
Table 3. Division of Labor: Relationship of Sex to Tasks Performed
Description Male Female Male Female
----------............... --..--.. -- ------- percentage ...................................... .....
1. Help with crops 65 62 33 29
2. Help with livestock 70 58 70 34
3. Off-farm work 39 43 42 34
Sample Size 279 281 266 221
Table 4. Extension: Usual Source of Information by Farm Size
Category Farm Size (hectares)+
Label It .02 .02-.15 .15-.30 .30-1.0 over 1.0
.......... -........................................ percentage ..................... .........................
1. Neighbor 86 90 82 81 0
2. Other friends 0 0 6 2 0
3. Village leader 0 0 6 10 0
4. P.T.D.++ 14 6 6 1 0
5. Extension worker 0 0 0 6 0
Sample Size 7 21 34 77 0
1. Neighbor 84 93 0 59 57
2. Other friends 16 0 0 7 7
3. Village leader 0 0 0 0 7
4. P.T.D." 0 3 0 25 7
5. Extension worker 0 0 0 9 22
Sample Size 19 32 0 34 14
In Garut only 3 farms were larger than 1 hectare and they were combined with the next smaller category (.30-1.0 ha). In
Cirebon, only 2 farms fell in the .15-.30 hectare category. These were combined with the next smaller category. This
accounts for the lack of case in the respective cells.
"Village staff in charge for agriculture.
Significance of Research Findings
One of the primary tasks for the Sociology Project is to
identify constraints to and consequences of the adoption of
new technology. The research conducted to date does not
point to any over arching concerns in this area although some
points should be noted. One of these would be whether the
current extension system can deliver new technology to small
ruminant producers in such a way as to significantly affect
Table 5. Reason for Raising Sheep and Poultry, Garut
production over a wide area. In addition, the credit system is
perceived as being relatively inaccessible and is not used by
most farmers. If new capital is required for the adoption of
new technology, this might prove to be a stumbling point. On
the positive side, small ruminants appear to be an important
component of the farming system, and are raised by a wide
variety of farmers-large and small alike. In the current
Javanese farming system they do not compete with crops for
land, labor or capital resources, and are complementary to
Category Farm Size (hectares)
Label It .02 .02-.15 .15-.30 over .30 Total
...................................- percentage .............................
-cash 10 10 16 9 10
-consumption 0 0 0 1 1
-saving money 80 76 57 71 69
-other reason 10 14 27 19 20
-cash 20 10 14 3 8
-consumption 10 0 21 5 7
-saving money 60 48 27 55 11
-inheritance 0 14 8 8 70
-other reason 10 28 30 29 4
Sample Size 10 21 37 76 144
intensive crop production in West and Central Java. It is
probably unreasonable to expect great nutritional improve-
ments to occur through the increased production of these
animals as most families raised them as a cash crop.
However, by providing farm families with additional cash
income, one could assume that their overall quality of life
would be enhanced.
Future Research Directions
Having established a macro view of the role of small
ruminants in the Javanese farming system, the Sociology
Project is now prepared to launch a number of more specific
studies designed to answer questions which have only
tangentially been addressed to date. Studies just begun or
planned for the near future include:
* The role of women. To date, all data collected by the
SR-CRSP in West and Central Java have been collected
by men from men. Yet the indications are that women
play more than a minimal role in the typical Javanese
farm and especially in the management of sheep and
goats. As the SR-CRSP moves into an "outreach" phase,
it will be crucial to adequately consider women's roles and
responsibilities in planning implementation programs. A
study is planned to explore this area in more depth.
* The role of the village trader. Previous research has shown
that most farm animals are sold at the farm gate to a
trader who deals directly with the farmer. This individual
acts as a middle man and in turn sells the animals he
purchases to an urban market or slaughterhouse. The role
of this individual in setting prices, encouraging farmers to
produce small ruminants, perhaps through loans of cash
or animals, and his ability to accumulate animals for
market during times of peak urban demand (e.g., religious
holidays) needs to be better understood.
A study of non-small ruminant producers. To date, all
data has been collected on farmers who raise sheep
and/or goats. Most of these farmers are probably
operating at their maximum level of animal numbers
(although the herds could undoubtedly be made more
productive). It seems that for most farms, labor is not
available to manage more than four or five animals.
However, in a number of villages, the producers of small
ruminants do not constitute a majority of the farm
families. Developing an understanding of the problems
farmers encounter in beginning to produce sheep or goats,
and why farmers are not raising them if they have the
resources to do so, is important for any program designed
to expand production.
Animal sharing arrangements. This is a very complex area
which requires more study. Our limited data suggests that
perhaps 20-25% of small ruminants are raised on some
kind of "share" basis. In addition, the Breeding Project
seems to be moving toward a strategy to increase the
genetic potential of village herds by providing improved
male animals to villages. Obviously, it is impractical to
provide every individual producer with an improved
Fighting rams, Sukawargi Village, Garut. Rams especially raised for
fighting are a very prestigious animal in this area, where exhibitions ofsheep
fighting are a major community event. This illustrates that animals are often
kept for reasons other than economic or subsistence. Photo: G.E. Bradford.
male. Thus, some form of village sharing arrangement
level will be necessary. Further research on the nature of
current sharing arrangements and the potential for such
arrangements for improving the genetic pool within a
village is obviously needed.
The Sociology Project has sponsored two graduate level
training programs for BPT staff members.
Kedi Suradisastra, a member of the BPT staff, was funded
for an MS program in Rural Sociology at the University
of Missouri, from January 1981-December 1982. Mr.
Suradisastra was awarded his degree in May 1983. He has
now returned to the BPT staff in the socio-economics
S. Mawi of the BPT socio-economics working group is
being funded for an MS program in Rural Sociology at
Institute Pertanian Bogor (IPB). This program was begun
in August 1982 and will be completed in June 1984. Mr.
Mawi will return to the BPT staff at that time.
In addition to degree training, the SR-CRSP Principal
Investigator participated in the village survey methodology
shortcourse held at BPT in June 1980. This course was
designed to acquaint junior and senior BPT staff and village
enumerators with the essentials of social science data
collection methods and analysis techniques.
Ihalauw, John. 1982. Penelitian Eksploratif Terhadap Beberapa Aspek
Sociologis-Ekonomis Sistem Produksi Kambing/Domba. Faculty of
Economics, Satya Wacana University, Salatiga.
Suradisastra, Kedi. 1983. Social Aspects of Small Ruminant Production: A
Comparative Study of West Java, Indonesia. MS Thesis, Department of
Rural Sociology, University of Missouri-Columbia.
Suradisastra, Kedi and Michael F. Nolan. 1983. Social Aspects of Small
Ruminant Production: A Comparative Study of West Java, Indonesia.
SR-CRSP Technical Report Series No. 19.
Suradisastra, Kedi and Michael F. Nolan. 1982. Small Ruminant
Production in the Humid Tropics: A Sociological Perspective. Paper
Presented at the Ann. Mtg. of the Rural Sociological Soc., San
A Small Ruminant
Production Model on a
initial survey and monitoring activities of small farms in
Java carried out by the SR-CRSP have confirmed the
small scale of sheep and goat production and the low level of
productivity (e.g., Sabrani et al., 1982; Thomas et al., 1982),
even though economic studies show the importance of this
activity to the farmers engaged in it (Knipscheer and
The combination of small scale and low productivity
creates difficulty in the assignment of priorities to an
improvement process. Small scale enterprises may show
considerable variation in their major management features,
and low productivity implies one or several constraints to
realizing the genetic potential of the animal. The need to
integrate biological and management aspects of production
in the design of an effective improvement program prompted
the development of a simple computer model as a research
tool. The objective was to provide Indonesian researchers
with a means of evaluating different strategies for increasing
animal production. A major consideration was to keep the
model sufficiently simple so that it could be developed and
operated on a programmable calculator which would allow it
to be used in the field, and would avoid any potential
difficulty in access. This approach was proven successful
under other circumstances (Thomas and Farias, 1981).
Model development was centered on those parameters of
animal production considered important in any initial
analysis and being recorded in the village monitoring
program. Inclusion of every possible variable would cause a
rapid escalation of complexity of such a model and probably
render it usable because of data limitations. Simplicity was
considered vital to the present purpose. The chosen
parameters were grouped according to their relationship with
the individual, the breed, the farm, or locality (Table 1).
The model was designed to be iterative, allowing
modifications of the values of certain parameters during
different time periods. The form of the model is described in
the flow chart in Figure 1. While the model operates at the
level of the individual animal, input data is formulated for the
population being considered, variation in the individual
being introduced through probability distributions defined
for each parameter.
Animal growth is the basic element of the model, growth
rates of study or experimental populations being derived
from the equation Y=a+bx, where Y is defined by:
Table 1. Animal Production Parameters Included in the
Litters born (if female)
Months pregnant or open
Weight at first estrus
Pre- and post-weaning mortality
Breeding flock size
LW change during period tn-tn+
LW at t
and x is defined by loge LW at t This model has fitted well
with observed populations, and was adopted because of the
need to include LW change of mature animals in order to
Figure. Flowchart of the smallfarm animal production modelfor operation on a programmatic calculator.
detect environmental effects on growth during different
periods. Under village conditions, there are generally too few
young animals to be able to derive satisfactory growth
functions for short periods.
The model was designed for, and tested on, a Hewlett
Packard 41CV programmable calculator fitted with the
extended function and extended memory modules, though it
is adaptable to other machines. In its current form, the
program requires approximately 5K of memory, with about
the same amount reserved for data manipulations and
The model allows the evaluation of animal production
over any time period for a given set of input data. Iteration is
monthly. For each month of the year, different growth
functions of the same basic form can be specified if necessary.
Most other parameters are defined as constant over time (for
a given run), though some can be modified (e.g., mortality) to
take into account dry season/wet season differences. Most
parameters are probability-linked. In its present config-
uration, 20 different farm flocks of up to 20 animals each can
be simulated at one time. On the basis of user-defined values,
the model will 1) sell market-weight males, 2) maintain a
specified size of breeding flock based on the number and age
of mature females, 3) cull the oldest females when (2) is
exceeded, 4) keep or sell young females based on consider-
ation of (2) and (3) and, 5) provide a monthly summary of
offtake and feed requirements.
The model provides a rapid means of: 1) evaluating the
data being collected from village monitoring programs such
as those being conducted in Indonesia under the SR-CRSP;
2) examining the impact on animal production of different
improvement strategies (these may have been derived from
experimental studies or may be imaginary); 3) assigning
priorities to strategies tested under (2); and 4) teaching
research workers the degree of integration between different
aspects of animal production, and allowing them a greater
appreciation of the impact of other disciplines upon their
own. In this sense, it is self-teaching.
While the model has been developed on the basis of
information derived during the Indonesia program, it is
applicable to a broader range of production conditions. It
allows a quantitative assessment of the impact of different
factors on the offtake from, and the longevity of, small-farm
breeding flocks. It is a tool that could be used by any research
worker, or even a trained extension worker, directly involved
in farm production studies.
Funding for the development of the model was provided
by the SR-CRSP Nutrition and Breeding projects and it is
anticipated that the model will be used by both programs in
Future Research Directions
The model is currently being used to examine differences
in productivity at different localities in Indonesia. It has
confirmed the low offtake realized by farmers under these
conditions, and has highlighted the feed constraint to
increased flock size.
While the model is fully operational and the principal
objective has been met, the potential of the model to meet
some of the more basic interests of the SR-CRSP research
scientists would require that it be expanded. The success of
the initial development on the HP 41CV, could be positively
continued by transferring to a desk-top microcomputer, thus
meeting the need for greater memory that increased
complexity implies. Equally, the HP 41CV is limited in
execution time and output format, and desk-top backup
would significantly improve usability. However, the main
features of simplicity and portability should not be lost. At
this point, more emphasis should be given to applying the
model to wider situations rather than to improved sophisti-
cation and a narrower set of opportunities. On-going field
studies in other SR-CRSP sites provide excellent oppor-
tunities for this.
Thomas, N., W. Mathius, and M. Sabrani. 1982. Small Ruminant
Production in West Java; Methodology and Initial Results. Invited
Paper to Conference: Livestock in Asia, Singapore. Published in
Livestock in Asia: Issues and Policies. Ottawa, Ontario, IDRC. p. 192.
Thomas, N., and J.M. Farias. 1981. Intensive Forage Production in
Northern Mexico. I. The Response of Italian Ryegrass to Nitrogen and
Irrigation. Experimental Agriculture, 17: 291.
Knipscheer, H.C. and T.D. Soedjana. 1982. The Productivity of Small
Ruminants in West Javanese Farming Systems. BPT Farming Systems
Program, Bogor, Working Paper No. 4. 20 pp.
Sabrani, M., P. Sitorus, M. Rangkuti, Subandriyo, I.W. Mathius, T.D.
Soedjana, and A. Setiadi. 1982. Laporan Survey Baseline Ternak
Kambing dan Domba, BPT Bogor. 88 pp.
KEN YA I- aseno "*Kisumu agricult
KENYA YM potentiz
uaI M *Kericho
-0 OKisi '1 Mag
S< I| ., Naiva
enya has a wide range of environments. With i o L..
K elevations ranging from sea level to 16,000 feet, its 3- o Na
ecozones include arid to semi-arid rangelands, a humid > 6,I=
equatorial coastline and environs of Lake Victoria, the
highland slopes of Mt. Kilimanjaro, and the Abedere
Mountains, the moist temperate region of the South Central Principal Small Ruminant CRSP worksites in Western Kenya.
highlands. Within this ecological diversity, there is also great
diversity in the agricultural production systems, which correspond closely with precipitation zones. Annual
rainfall ranges from 200-2,000 mm and generally increases with elevation.
Three quarters of Kenya's total land area and most of its rangeland falls into the category of low agricultural
potential, with less than 600 mm rainfall per year. In these areas transhumant pastoralism is widespread, with
herds of cattle and small ruminants. The remaining quarter of Kenya's land is in the medium and high
potential categories for agricultural production.
Over 70% of the Kenyan population relies on smallholder agriculture for its livelihood in these zones.
Livestock, although a secondary enterprise, are found on most of these holdings, with cattle on approximately
one million farms and sheep and/ or goats on .6 million farms. Small ruminants are raised in an intensive "zero
grazing" or limited grazing husbandry system and are both consumed by the producer and sold to provide
The expansion of human population in these already densely populated areas is causing farm size to shrink
and diminishing the available resources per capital. Average farm size is now 2.3 hectares and mean household
size is 7 persons.
Smallholder, subsistence level farms in the humid tropics often produce inadequate and inconsistent
amounts of high quality protein foodstuffs such as milk needed for good health, especially of children and
pregnant women. Many farms no longer have the land to provide sufficient feed for a lactating cow, however,
they may support several lactating goats. Goats, (3 to 5) in place of a cow, have multiple advantages: loss of an
individual animal has less impact on family welfare; initial investment per animal is low; mating does to kid in
different seasons allows a small, but consistent daily supply of milk throughout the year; and litters of two or
three kids at 7 to 8 month intervals markedly increase offtake of slaughter stock for family consumption or sale
to provide needed income.
Realization of these advantages for goats producing both milk and meat within a small-scale integrated
farming system requires resolution of many constraints. In the past, little effort has been directed toward the
development of a dairy goat adapted to the climatic and health stresses of the humid tropics. Furthermore, the
genetic potential of even an adapted dairy goat will not be realized without simultaneous development of
appropriate feed production, management, and marketing practices.
The advantages of a small, short gestation, litter-bearing dairy ruminant for small-scale farming systems in
the humid tropics are readily apparent. However, this apparent potential remains essentially untested. To be
acceptable to small-scale farmers, the dual-purpose goat production system must be based on low-cost,
low-risk technology and be minimally competitive with cropping activities for land, labor, and capital
resources. It is necessary to characterize existing farming systems, including those that do not now incorporate
a dual-purpose goat component in order to determine if and how this component can be introduced. The
research problem is, therefore, to develop an animal of the appropriate genotype, to develop a feed
resource-nutrition-health management package appropriate to the small farm resource base, and to ensure
that this component is economically and sociologically acceptable. Successful resolution of this researchable
problem will require the holistic, multidisciplinary approach embodied in the SR-CRSP.
SR-CRSP research activities in Kenya focus on dual-purpose goat production as a component of small
farm systems in the medium to high rainfall region of Western Kenya. Currently, seven projects administered
by four SR-CRSP institutions are working in Kenya in collaboration with the Ministry of Livestock
Development. These institutions and projects include: Breeding, Systems Analysis, Sociology, Health,
Economics, Production Systems, and Forages.
An important feature of the SR-CRSP approach is the focus on integrating economic and sociological
factors with agricultural data obtained at experiment stations and in the field, both to delineate the current
small ruminant production systems and to develop a coherent series of recommendations for improving
animal productivity and marketing efficiency.
The initial characterization phase primarily involved surveys of small-farm systems and goat health status in
1981-82; work continues with case studies of small farm families, market surveys, and on-farm evaluation of
interventions to improve feed production and goat nutrition, health and management. To date, component
research has primarily been conducted on experiment stations: breeding at 01 Magogo Station; health at
Kabete Station; and goat nutrition, management and feed production at Maseno. However, a few promising
interventions are now undergoing preliminary trials on selected farms.