Breeding scheme based on community-based participatory analysis of local breeding practices, objectives and constraints for goats around Dire Dawa, Ethiopia

G Gebreyesus, A Haile* and T Dessie**

ILRI, PO Box 5689, Addis Ababa, Ethiopia
g.gebreyesus@cgiar.org
* ICARDA, PO Box 5466, Aleppo, Syria;
** ILRI, PO Box 5689, Addis Ababa, Ethiopia

Abstract

This study was conducted in the rural kebeles around Dire Dawa for designing a simple, yet, feasible breeding scheme in the context of community-based management of animal genetic resources. Range of participatory rural appraisal tools, including focal group discussions and participatory mappings, were employed to study the local community’s Indigenous knowledge and practices in managing the goat gene pool. The breeding objective and local trait preferences were defined in a participatory manner through own-flock ranking experiments.

The community generally practices selective pure breeding where by the own flock and flocks in the neighbourhood were the units of selection for bucks. There are social regulations in the community against sale of breeding does outside the community while encouraging communal use of outstanding breeding males. Goats are kept for multifaceted purposes ranging from products like milk, meat and live-sale to functions in socio-cultural, financial and ritual state of affairs. The breeding objective is to ensure improved milk production, through increased daily yield per doe and increased fertility per flock, and increased net income per flock, through increased number of marketable animals. Traditional criteria such as conformation, behaviour and adaptation were as important as most “production” traits in selecting breeding animals. The breeding goal traits considered were, accordingly, milk production, conformation and reproductive traits. Based on these findings, village breeding schemes, where-by flocks and breeding groups in a village are taken as focal points, is recommended as way forward in genetic improvement. The framework for a feasible implementation of such genetic improvement scheme is outlined based on the rationale of utilizing available social regulations, indigenous knowledge and traditional systems of breeding as well as future market prospects.

Keywords: indigenous knowledge, village breeding schemes

Introduction

Large, and possibly expanding, parts of the globe can be used for food production only by livestock that are adapted to local conditions (FAO 2009). Pastoral livestock production systems have long evolved to provide, perhaps the only, means to utilize such extremely harsh environmental conditioned areas as deserts (Savero and Nikolas 2010). Despite the fact that the pastoralist enterprise is supporting considerable proportion of the country’s rural population, it has not regrettably attracted the attention of research and development.

The fact that there is no documented national level genetic improvement strategy for the pastoralist system shows the stumpy attention to the system (IBC 2004). Nevertheless, there are some legitimate sources of difficulty in going forward with genetic improvement in the pastoralist production system. Apart from the obvious absence of infrastructures, high mobility and harsh environmental conditions, the highly communal value and socio-cultural associations the pastoralists have towards their livestock render the conventional research and development approaches irrelevant in this context (Koehler 2005).

In developing nations an enabling farm animal genetic resource conservation policy could be successful by placing high priority on a community-based participatory approach and focusing on food security and poverty alleviation (Wollny 2003). However, information on sheep and goat breeding objectives targeting the needs and perceptions of farmers and design of community or village-based breeding schemes is virtually absent (Solomon 2008).

Designing effective genetic improvement strategy for the pastoralist system stipulates adequate prior knowledge of the different social underpinnings, indigenous knowledge and traditional systems and institutions pertaining to livestock breeding. Communities have different sets of cultural and social values by which to judge, appraise and decide on breeding animal selection (Solkner et al 1998). Nevertheless, such sociological factors are so far ignored or considered less important than they really are (Solkner et al 1998). A coherent and comprehensive breeding program suited to the existing production systems (decentralized breeding program) is required to guide stakeholders in the sustainable management of animal genetic resources. Baker and Rege (1994) stress the fact that defining objectives in comprehensive economic terms (i.e. returns minus costs) is difficult enough in temperate agriculture and much more difficult in the tropics.

The basic objective of this study was to explore the potential socio-cultural determinants for implementing breeding schemes, to define local breeding objectives and traits preferences, and probe for indigenous knowledge, traditional systems and institutes that can favourably be utilized towards successful breeding scheme.

Materials and methods

The study area

This study was carried out in the rural kebeles around Dire Dawa located between 9^o27’-9^o49’ north latitude and 41^o38’- 42^o19’ east longitude in the Eastern part of Ethiopia (DDAC, 2002a). The north eastern part of Dire Dawa is relatively sparsely populated lowland exhibiting pastoral and agro-pastoral system and the southeastern part of the administration comprises of the escarpment with mixed farming system (DDAC, 2002b). The study sites include Jeldesa, Goladeg and mudianeno kebeles found in the north-eastern and north-western parts of the Dire Dawa Administrative council. Goat and camel are the dominant livestock species in the study sites kept under pastoral and agro-pastoral production systems.  In the study sites, the Short-eared Somali goat breed is preferred and predominantly kept by the Issa pastoralists.  

Sampling framework

Using purposive sampling, three kebeles were selected for the study. Accordingly, two kebeles (Mudianeno and Goladeg) were selected to represent the pastoral system while one kebele (Jeldesa) was selected to represent the agro-pastoral production system. Villages within kebeles were selected randomly.

For administration of semi-structured questionnaire, respondent households were first stratified into flock size categories (small, medium and large) and then respondents within each category were selected randomly. For this particular study, households owning less than twenty goats were grouped as small flock sized (N=41), while those with goat holding starting from twenty to thirty five were categorized as medium flock size class (N=49). Meanwhile, households with more than thirty five goats were grouped within large flock-size category (N=36). This categorization was based on prior information from the Dire Dawa Agricultural Bureau pertaining to livestock holding data in the three kebeles.

Data collection procedures

Range of PRA tools and semi-structured questionnaire were employed to collect the data. Discussions were made with a focal-group established at each kebele. Members of the focal groups included individuals communally known to have high quality breeding animals, people believed to be knowledgeable about past and present social and economic status of the area, community elders and story tellers.

Semi-structured questionnaire was used to collect data on livestock holding, husbandry system and production constraints, purposes of keeping goats, household labour divisions and breeding management systems. Owner-recall approach was also included in the questionnaire to collect data on reproductive performances of goats.

Own flock ranking experiment (implicit ranking) was used to study breeding objectives and trait preferences. Participants of the own flock ranking were asked to rank their first, second, third and least preferred male and female goats within their own flock giving reasons for each preference rank. The respondent could mention as much reasons as possible for each rank but only the first mentioned reason, under each rank, was taken as the “revealed” trait preference for that rank. The own flock rankings allowed easy and unguided ranking by participants with a concrete presence of the animals bearing the attributes, as opposed to an “abstract” ranking against predetermined list of traits.

Matrix ranking was also carried out to rank among the available grazing species in the area according to different criteria.

Seasonal calendars were also prepared in participation with the locals, to define local ecological seasons, seasonal availability of grazing species, labour division and migration patterns. These will be helpful in the planning stage of a possible breeding program to determine seasonal variation in flock dynamics and resource availability and suitable timing for the program.

Informal enquiries, to randomly accessed local informants, and talks with local agricultural officers were made as a method of triangulation in order to cross check reports on similar issues.

Participatory observation, staying within the community, was also made to identify important goat traits and their expressions in the system’s context as part of the participatory definition of the breeding objective.

Data Analysis

The SPSS statistical software (SPSS for window, release 15.0, 2006) was used to analyse the qualitative data from the semi-structured questionnaire. F test was carried out as appropriate, following analysis of variance (ANOVA), to assess statistical significances.

Indexes were calculated for all ranking data according to a formula:  Index = sum of ranks(3 for rank 1 + 2 for rank 2 + 1 for rank 3) given for an individual reason (attribute) divided by the sum of ranks(3 for rank 1 + 2 for rank 2 + 1 for rank 3) for overall reasons. For the matrix ranking activity, a rank score was calculated for each top five ranked grazing species based on the proportion of ranks each received: 1 for highest ranked, 5 for lowest ranked species according to each criteria.

Results and discussion

Production system and major constraints in the area

The average reported livestock possessions are presented in Table 1. There was no significant difference (P>0.05) between the kebeles in goat, sheep and donkey holdings. However, respondents in Goladeg had significantly higher (P<0.05) number of cattle than their counterparts at Jeldessa and Mudianeno.

An overall minimum number of 8 and maximum of 160 goats were reported per household giving an overall mean of 32.8. The figure is higher than the case reported by Farm Africa (1996) which suggested a mean flock size of 23 for the Small-eared goat population in general in Ogaden and around Dire Dawa. This may be explained by a suspicion that pooling the data from the two regions, by the former study, may have lowered the overall mean vis-à-vis the actual mean for Dire Dawa.

The proportion of castrated males was 3.8% of the whole population while male and female kids less than a year of age made up to 19.9% and 25% of the whole flock, respectively. In terms of proportion for pooled sex groups, females constituted 67.4%, while intact and castrate males constituted 28.8% and 3.8 % of the whole flock, respectively. Out of the total kids with less than six months of age, 52.5% were female and 47.7% of them were males. This may be referred to speculate that the natural birth is slightly skewed towards the female sex. This is in agreement with the report of Devendra and McLeroy (1982) for the tropics and subtropics.

The results of matrix ranking showed that five species of pasture and shrubs were highly important as sources of feed (Table 2). Grazing land ownership was totally communal in all the cases.

Feed shortage was the most frequently mentioned production constraint in the study area. Water shortage was the second frequently mentioned constraint at Jeldesa and Mudianeno while disease was ranked second most important constraint at Goladeg. Recurrent droughts exacerbate feed and water shortage problems in the area. Disease and predators were also mentioned as important constraints threatening goat production in all the three Kebeles while theft was mentioned exclusively at Mudianeno.

Relatively longer dry season and weak rain during the wet seasons render the grazing lands discontinuous feed sources able to support only part of the year. Reported responses to feed shortage include migration, cutting and providing leafs and branches from trees and sale of livestock. Migration is reported to be the ultimate means of drought coping up specifically in the longer dry season (Jilal). The type of migration may involve taking all flock and family on board (50.4%) as well as flock splitting (49.6%). There were differences between the flock size categories in the type of migration (Table 3). Households with small flock size tend to take the whole flock on-board migration while majority of the households with large flock sizes practiced split flock migration. This suggests that flock size must receive due consideration in designing village breeding schemes in pastoral areas as such schemes out to be “movement-friendly”.

Household labour division

Majority of the respondents reported a fairly equal trend between adult male and female family members in making decisions regarding selection and breeding of goats (Table 4). Since house wives are primarily responsible for majority of the routine husbandry and hence with better knowledge about the flock, it looks apparent that they are well placed in making breeding decisions. These findings stress the fact that ensuring active participation of women will be indispensable for the success of a village breeding program in the Area.

Reproductive performances of the goat population

According to summary of respondents’ view, mean age at first service of male goats was 14.6 months whereas the value for female goats was 14.9 months (Table 5). This was in agreement with the result of Tolera and Abebe (2007) who reported mean age at first service of 14.9 + 6.6 months for female goats in southern Ethiopia. A twinning rate of 5.8% was computed for the population. The finding was higher than corresponding values computed for the Nubian, Afar, Abergelle and the Long-eared Somali breeds by FARM Africa (1996). Nevertheless, the computed value is not compatible with the high preference for twinning expressed by livestock owners in the study area and calls for strong selection in its favour. 

Range of goat products and functions prioritized in the area

Results of the rankings have shown that goats are expected to provide multifaceted benefits apart from the quantifiable products (Table 6). Milk, income, meat and skin were the reported tangible benefits. Meanwhile, tradition, social status, saving, social payments, ritual and manure were among the intangible benefits respondents keep their goat flock for and reasons for their wish to expand it.

Milk has received larger proportion of the overall ranking in all the kebeles. This is in agreement with its significant contribution towards the traditional staple diet in the community. Income from sale of live-goats was the second highest ranked benefit. This goes smoothly with the reported higher contribution of live-goat sale to the household cash income.

Local breeding strategies and mechanisms

In this study, majority (97.6%) of respondents reported to select buck for breeding whether from own flock or other sources. Meanwhile, 38.7% of respondents also reported selection of breeding females. Stages of selection for breeding males were early; (birth-6months old), (13.6%), young; (6 months-1year old) (50.8%), adult; (more than year old) (34.7%) and multi stage (0.8%).  Selection criteria reported were mother (ancestral) history, own performance (appraisal) and some other traditional systems.

Majority of overall respondents (85%) run their goats as a flock alone during grazing while only 15% of them mixed their flock with other flocks in the neighbourhood. Main sources of breeding buck, according to respondents, were own-flock, relatives, neighbours and community members. There were also differences among households within different flock size categories regarding sources of breeding bucks (Table 7). Majority of households with small flock depended on other sources than their own flock for breeding buck while the reverse progresses to hold true as flock size increased. However, 14.6% of the households within the large flock-size category also relied on other neighbours for breeding bucks. This implies that the source of breeding bucks depended, apart from flock size, on the quality of the bucks in own-flock vis-à-vis other bucks from alternative sources.  

The Issa community has a culture of mentally memorizing personal ancestral genealogy as well as genealogy of their animals. Accordingly, 98.2% of overall respondents reported to mentally memorize genealogy of every goat through the maternal line of descent. Focal group discussion members also reported that owners can recall up to more than 10 lines of maternal lineages tracing back to superior individuals. Genealogy memorization is used to select breeding animals born to a superior maternal lineage (a system known as hortice).

Method of identifying individual goats was reported to range from naming every individual (48%) to branding on body (8%), ear cutting (notching) (8%), mental memorization by colour or other physical feature (1.6%) and combination of all these (34.4%). Majority of households owning small flock were able to identify their animals just by naming (58.5%). To the contrary, majority of households with large flock sizes (44%) had to employ naming, ear cutting and branding in combination in order to identify individuals within their large flocks.

It was observed that goats are named based on physical characteristics (coat colour pattern and type, body size etc), unique behaviour or special events within the family coinciding kidding of the typical animal. Naming based on physical feature was, however, the most commonly applied system. Nevertheless, naming an individual in this method usually follow special characteristics of the maternal lineage rather than its own. 

Majority (97.6%) of respondents reported to control mating. Castration was most common means of controlling mating (Table 8). Locally known “traditional healers” were observed castrating animals through crushing the spermatic cord with a wooden material. Mean age at castration of unselected males was reported to be 1.81 years (Table 8). Castration was practiced for multiple purposes which included fattening, to fetch better price at market and modify unwanted behaviour apart from breeding control.

The breeding buck left un-castrated will serve up to 5.5 years of mean age after which it will be usually castrated or disposed through sale or slaughter. Similar practice of changing the serving bulls every four years to prevent inbreeding was reported for some pastoral societies in India by Köhler Rollefson and LIFE-NETWORK (2007). However, the average age at which the breeding buck is changed, reported in the current study, may still be considered too high to significantly prevent inbreeding.

Socio-cultural underpinnings of goat breeding in the Issa Somali community

The Issa community is a predominantly pastoralist community with strong clan system as the main social fabrics. This strong and closed clan based social identity goes beyond present day political boundaries unifying and networking members within three neighboring nations viz. Ethiopia, Djibouti and Somaliland.

The Issa community maintains a perception of special association towards the Short-eared Somali (locally known as Issa) goat breed, claiming a historic role in its development and adaptation. The Short-eared Somali goat is locally known as the Issa goat after the communal name of the pastoralists that keep it around Dire Dawa and across the Shinile zone all the way to Djibouti. Such cultural links between individual ethnic or social groups and specific breeds have been reflected in many breeds being named after ethnic groups (Rege 2001; Koehler 2003; FAO 2009). Local myths persistent in the community associate the origin of the Short-eared Somali goat breed with the communal ethno-history. Local myths about origin of breeds may not necessarily be used as official evidence of origin. However, as supported by Koehler (2005), presence of such myths and their link with the community’s ethno-history can be used as clear indications of the community’s perceptions about its roles in development of the diversity.

There are different social circumstances involving exchange animals within the Issa clan. These include Dowry (Yeredh), Groom wealth, compensation in dispute resolution (guma) and help to poorer members of one’s clan. In all these circumstances, breeding does constitute higher proportion of the gift stock and the giver is traditionally obliged to include superior of them. In fact the quality of the gift and the associated weight given by the recipient lies on the quality of the breeding does constituting the gift flock. Goat possession is also seen as prestige in which case individuals with large flocks are treated with high social respect. This justifies pastoralists’ motive to have large flocks in spite of low productivity.

Sale of breeding does is not practiced under normal circumstances. It is not such that an explicit social rule existed against sale of breeding does, but such act is considered as extreme weakness and is subject to negative social judgments.

Breeding bucks are also highly exchanged within the community and it is unlikely that owner of a superior buck would deny a fellow member an access to. Even though there were no explicit rules reported obliging an owner to share his bucks, a strong culture of sharing characterizes the social network not only in case of goats but also other resources.

The Breeding objective

Definition of the breeding objective was made through participatory analysis of priority problems, contribution of goats towards mitigating these problems and most limiting factors in the production system.

Deep rooted food insecurity and soaring gap between financial demand and supplies were identified as priority problems faced by the pastoralists. These problems are also threatening the pastoral enterprise in the area at a system level. Goats are significantly contributing to the household staple diet through, primarily, milk production. Significant proportion of the household cash income is also secured through sale of live-goats. Meanwhile, harsh environmental conditions including feed and water shortages, recurrent droughts, heat stress and diseases are considered persistent constraints in the area. Risk aversion is thus taken as the most feasible way forward production mode in the system and area.

The breeding objective is thus to ensure improved milk production, through increased daily yield per doe and increased fertility per flock, and increased net income per flock, through increased number of marketable animals, without undermining adaptation

Local trait preferences and the prioritized breeding-goal traits

According to Solkner et al (1998) traits to be selected for must reflect the breeding objective, must be easy to measure, heritable and not too many. Reproductive performances, including good mothering, twinning and short parturition intervals, were frequently mentioned during the flock-rankings for females (Table 9). The inherent problem in selecting for such traits is the associated low heritability values. Inclusion of reproductive traits in the breeding goal is, however, justifiable as the goal should reflect owners’ preferences. Moreover, Zewdu et al (2006) also argued in favour of including reproductive traits not only based on economics of production, but also because of its influence in genetic improvement through selection intensity.

Milk yield has received the highest preference rank in the female flock-ranking activities. It was also mentioned as one of the primary functions of goats which also makes significant portion of the staple diet. Conformation, as an aggregate of traits such as horn, colour, body condition and size, was also frequently mentioned for both sexes (Table 9 and Table 10). These traits also highly affected “marketability” of male goats in the local as well as central markets at Dire Dawa.

Body size and mother history were mentioned with relatively higher proportion of participants as reasons for ranking first preferred male goats (Table 10). However, with regard to overall index, mother history and color were the leading reasons, with an index of 0.2 each, owing to higher frequency of being mentioned as second and third reasons. Polledness, body condition and body size follow as important preference reasons, each with an index value of 0.1.

These preference patterns appear to be in line with the breeding objective. Adult male are usually kept for sale, if not selected for breeding. Price is determined by body condition of the animal, color, sex and other attributes as horn both at the central market in Dire Dawa as well as local villages markets. Use of body weight for price determination was not even observed and effect of body size on the price of the animal was not as critical as body condition, horn and color. Polled male goats were more preferred in the market than their horned counterparts for alleged association between the trait and meat quality in terms of flavor (tenderness) and also dressing percentage. Plain black colored goats were least preferred on the market among the other coat color groups while castrates were reported to fetch the highest price as compared to intact male goats.

Adaptation traits, including drought resistance and walking ability, were also mentioned by respondents. In fact, the whole system primarily depends on utilizing and optimizing these attributes. However, the difficulty in this situation is the broadness of the collectively called adaptation traits (FAO 1999) and associated intricacy in assessment of these parameters on individuals for selection. An experience from the approaches used by Rege et al (2001) was in this case followed. Rather than directly selecting for adaptation, concentrating on selection for the productive and reproductive traits in the presence of environmental stress, thus allowing animals to be selected while responding also to the stressors will be the most feasible practical way.

Accordingly, the traits considered for improvement are milk production, reproduction and conformation (aggregate of features including colour, horns, body condition and size).