Livestock Research for Rural Development 27 (4) 2015 Guide for preparation of papers LRRD Newsletter

Citation of this paper

Ex-ante perceptions and knowledge of artificial insemination among pastoralists in Kenya

D N Khainga, G Obare1 and A Murage2

Department of Agricultural Economics and Business Management, Egerton University, P.O. Box 536-20115, Egerton, Kenya,
dkhainga@gmail.com
1 Department of Agricultural Economics and Business Management, Egerton University, P.O. Box 536-20115, Egerton, Kenya;
2 International Centre for Insect Physiology and Ecology, P.O. Box 30772-00100 Nairobi, Kenya

Abstract

We measured the preferences for Artificial Insemination (AI) for the stocking of Sahiwal cattle by Maasai pastoralists. Multistage sampling was employed in collecting cross-sectional data of 384 respondents using structured questionnaires. An ordered probit regression model was adapted and data analysed using stata.

We found group membership, access to extension, AI awareness, production system, county of residence, years of education, household size and herd size to have a positive significant effect while experience and age had a negative significant effect on the perception of farmers towards AI preference. Moreover, pastoralists’ perceptions towards AI uptake were influenced by its affordability, accessibility, success rate and its calf tolerance in arid and semi-arid environment. To ease demand for Sahiwal bulls from National Sahiwal Stud, we recommend extensive farmer education through existing agricultural extension networks to positively influence farmer perceptions towards AI technology for its rapid uptake.

Keywords: arid and semi-arid lands, ordered probit, preference, production systems, Sahiwal


Introduction

In sub-Saharan Africa (SSA), dairy cattle produce about 70% of total national milk output with an estimated annual per capita milk consumption ranging from 19 kg in rural areas to 125 kg in urban areas (Muriuki 2011). Dairy production in Arid and Semi-Arid Lands (ASALs) has been perpetually low and neglected by government agencies over a long period of time. The interaction between pastoralists and commercial ranches remain active given the exchange of breeding material and where necessary semen is obtained from the Sahiwal Cattle Breeders Society (FAO 2010). The farmers’ choice of a particular breed to keep is determined by a number of factors not limited to characteristics of the production systems, infrastructural limitations and changes in environmental conditions with respect to prevelence of diseases and cattle feeds (Ouma et al 2004). Introduction of new breeds to pastoral areas has been in progress, however, there is need to match these genetic resources to the existing production systems and the distribution of such advanced germplasm should be viable with the use of available technologies and infrastructure (FAO 2010).

Over the years, Sahiwal breed known for high milk production, endurance for hot climate of tropics, resistance to tropical diseases, low cost of maintenance and higher feed conversion efficiency has been used in the upgrading of the well adapted Small East African Zebus (SEAZs) (Ilatsia et al 2007; Ilatsia et al 2011a). Trade-offs between the sahiwal and the SEAZs include how to tolerate drought and diseases as compared to indigenous breeds (Roessler et al 2010, Ilatsia et al 2011b). These need to be considered when determining the degree of cross breeding in pastoral areas in a bid to safeguard their wealth and production goals.

Major breeding aims of livestock farmers include high productivity in terms of milk and beef, improved fertility and adaptability to local production environment (Roessler et al 2010). High milk productivity per cow makes economic sense for the households to rear cattle which is evidenced in large number of animals per herd, a characteristic of the Kenyan Maasai community. Cattle traits considered important by breeders include milk production per lactation, reproductive efficiency, growth potential, adaptability, udder conformation and temperament which are traits that the Sahiwal is credited for (Muhuyi et al 1999). The introduction of new breeding techniques such as AI is vital to structural change in pastoralist production in Kenya. The need to denuclearise and disseminate Sahiwal genetic material through such technologies including crossbreeding in ASALs is core given the aforementioned benefits. Hence, the objective of this study was to determine factors that influence pastoral farmer’s perceptions and preference for AI as an alternative breeding technology to the natural service.


Materials and methods

The study area

Narok and Kajiado counties were selected because they both have high number of pastoralists and high concentration of Sahiwal breed and its crosses. Narok County is within a high agricultural potential area with a relatively high population of both pure pastoralists and agro-pastoralists. Narok County covers an area of 17,944 km² (GOK 2010),with mean annual rainfall ranging from 500 to 1800 mm, temperature ranges from 8°C to 28°C in the North and West. The southern part has a semi-arid climate (Jaetzold et al 2005). The climate is suitable for crop-livestock farming. Narok South is largely semi-arid with pastoralism as the major occupation of the residents. Narok County is home for approximately 770,000 cattle heads, out of which 5000 are pure Sahiwal cattle while 69,000 are crosses of Sahiwal and the local Zebu (MOLFD 2006). The County has a total human population of 850,920 with a population density of 47 persons per km² and 12 % living below the poverty line (KNBS 2010).

Kajiado County lies on a land mass of 21,901 km² characterised by semi-arid to arid tropical environment,conditions that favour pastoral livestock production. The SEAZ is the predominant cattle breed, followed by Sahiwal and their crosses with SEAZ, and unimproved Boran (MOLFD, 2006). The County is estimated to be home for 440,000 cattle population of which 39,000 are pure Sahiwal while approximately 130,000 are crosses of the Sahiwal breed and SEAZ (MOLFD 2006). Kajiado falls under the research mandate area of the NSS where Sahiwal breeding activities have actively been promoted, hence the relatively high concentration of Sahiwal genetic resources. The County has a total human population of 687,312 with a population density of 31 persons per km² and 11.6 % living below poverty line (KNBS 2010).

Survey design and data collection

The survey was carried out in two phases. The first phase involved stakeholder group discussions through workshops held at the Kenya Agricultural Livestock Research Organization (KALRO) to obtain background and situational information about farmers’ perceptions on introduction of assisted reproductive technologies in arid and semi-arid areas given their cultural beliefs. The information gathered was incorporated in a structured questionnaire, and then pre-tested for both qualitative and quantitative aspects which were then employed in collection of data in the second part of the survey.

Each County formed a sampling stratum based on dominant production system practiced by the farmers. Kajiado and Narok Counties are majorly pastoral and agro- pastoral zones, respectively with a few ranches scattered within each. The locations in each division as per the 2009 population census formed the sampling frame from which 45 sub-locations were sampled (GOK 2010). In the second stage, simple random sampling was performed with assistance from community animal health officers and district livestock production officers. A total of 384 households comprising 178 and 206 from Kajiado and Narok counties respectively were sampled. The survey was conducted from December 2012 to January 2013. Household heads responsible for decision-making, especially in relation to livestock production were interviewed. The questionnaire covered household characteristics, socioeconomic and institutional factors. Data for addressing this specific objective were elicited by asking respondents to rank the preference for AI services based on a range of factors.

Table 1. Definition of variables
Variable Description
AI preference Preference ranking for AI on a scale of 1-3; least preferred, preferred and most preferred respectively.
Youth-aged Household head aged between 18 and 35 years old (0 = no, 1 = yes)
Middle-aged Household head aged between 36 and 50 years old (0 = no, 1 = yes)
Elderly-aged Household head aged above 50 years old (0 = no, 1 = yes)
Education Education of household head (years of schooling)
Household size Number of household members
Land size Land owned by household head (acres)
Herd size Number of cattle owned by household head
Off-farm activity Household head has Off-farm income generating activities (0 = no, 1 = yes)
Experience Experience in keeping livestock (years)
AI awareness Household head is aware of artificial insemination (0 = no, 1 = yes)
Group membership Household member belongs to local group (0 = no, 1 = yes)
Credit access Household member accessed credit in the past 5 years (0 = no, 1 = yes)
Extension services Household member accessed extension services in the past 5 years (0 = no, 1 = yes)
Distance to market Distance to livestock markets (kilometres)
Agro-pastoralism Household practice agro-pastoralist production system (0 = no, 1 = yes)
Pure pastoralism Household practice pure pastoralist production system (0 = no, 1 = yes)
Narok County Household resides in Narok county (0 = no, 1 = yes)
Kajiado County Household resides in Kajiado county (0 = no, 1 = yes)
Nomadism Household is living nomadic lifestyle (0 = no, 1 = yes)
AI affordability Household head is willing to accept AI based on its affordability (0 = no, 1 = yes)
AI accessibility Household head is willing to accept AI based on its accessibility (0 = no, 1 = yes)
AI success rate Household head is willing to accept AI based on its success rate (0 = no, 1 = yes)
AI calf tolerance Household head is willing to accept based on AI’s calf tolerance (0 = no, 1 = yes)
Attribute index Household attribute index is greater than the mean (0 = no, 1 = yes)
Empirical model and specification

Literature review indicates extensive use of either probit or logit models in analyzing factors that influence farmers’ perceptions for a given technology or a programme. Unlike recent study by Makohha et al (2008) that employed binary choice models for dependent variable, this study adapts ordered probit because ordered dependent variable informs us about the level of preference for AI among various pastoralists across the study area. The respondent revealed his/her perception by ranking AI services on a Likert scale of 1 to 3, with 1 denoting least preferred; 2, preferred; and 3, most preferred, respectively. Despite attractiveness of ordered probit in analyzing categorical data, it fails to account for protest attitudes of respondents and choice task complexities which may influence consistency of results (DeShazo and Fermo 2002). The above challenges were addressed through participatory approach adopted in data collection in which choice tasks were simplified and farmers’ concerns addressed by researchers from KALRO during the survey.

The ordered probit is related to the latent class of models. We adapt the approach by Long (1997), where we consider variable Y which denotes preference rank given to AI by farmer i and takes on j values which are naturally ordered on the Likert scale. However, these observed values are assumed to be derived from some unobservable latent variable Yi* such that:

where Xi refers to the observable individual specific factors, β is a vector of parameters to be estimated and ei is the stochastic-disturbance term with normal distribution (Greene 2003). The observed choice outcomes Yi are assumed to be related to the latent variable Yi* as:

where µi unknown threshold parameter for outcome i that separate the adjacent boundary values and is estimated together with the β’s. The estimated µi, (where i=0, 1, 2) follows the order µ0 <µ1 <µ2.

The probability that the case falls into each category j, using the estimated µi parameters as threshold limits is given as:

where λ represents the cumulative density function of.The β’s can be estimated by computing the marginal effects using maximum likelihood functions defined by Greene (2003). Thus:

The estimated marginal effects indicate the change in the likelihood that a farmer would “prefer” or “most prefer” (as opposed to least preference) AI as a result of a unit change in the specifc explanatory variable. An ordered probit regression was fitted for AI technology to obtain estimates of the coefficients and marginal effects.

The estimated marginal effects indicate the change in the likelihood that a farmer would “prefer” or “most prefer” (as opposed to least preference) AI as a result of a unit change in the specifc explanatory variable. An ordered probit regression was fitted for AI technology to obtain estimates of the coefficients and marginal effects.


Results and discussion

Descriptive results

Pastoral production is a livestock production system characterized with large herds reared on huge chucks of arid and semi-arid areas. Pastoralism involves movement of farmers with animals over long distances in search of pasture and water. This exposes the farmer to risk of losing the animals due to death especially during dry seasons. Table 2 presents a descriptive summary of key variables on farmer and farm characteristics used in this study.

Table 2. Descriptive statistics and frequencies for farmers and farm characteristics by County
Variable Whole sample Counties Tests P
Kajiado Narok t-test χ²
Education 6.63 (6.18) 6.29 (0.49) 6.59 (0.47) -0.437 0.662
Household size 11.3 (6.39) 10.4 (0.48) 12 (0.52) -2.20 0.029
Land size (acres) 94.1 (55.3) 146.7(54.4) 47.3 (30.2) 4.20 0.000
Herd size 110 (12.7) 90.7 (8.17) 133 (11.34) -2.97 0.003
Experience 23.8 (12.1) 25.6 (0.98) 23.7 (0.91) 1.41 0.161
Distance to market 9.64 (8.90) 8.97 (0.69) 10.1 (0.75) -1.10 0.271
AI awareness 262 130 132 6.99 0.008
Group membership 182 81 101 0.377 0.539
Credit access 228 101 127 1.06 0.304
Extension services 105 33 72 13.3 0.000
Off farm activity 49 20 29 0.627 0.429
Pure pastoralism 191 74 117 9.81 0.002
Nomadism 52 33 19 7.80 0.005
Youth (18-35 years) 57 24 33 0.424 0.515
Middle aged farmers (36-50) 224 108 116 1.40 0.237
Elderly farmers (above 50) 45 18 27 0.760 0.383
Note: Standard errors in parenthesis
Source: Survey data, 2013.

Education levels in the study area were low with resulting statistics indicating an average of 6 years (grade 6) of schooling while majority never attended school at all. Pastoralists still hold the view that a huge household size is a sign of wealth and the summary statistic corroborates this with a mean size of 11 members. Unlike in most Counties in Kenya, the Maasai community still own huge tracks of land parcels with a mean holding of 146.7 and 47.3 hectares in Kajiado and Narok counties, respectively. It is worth noting that most of these tracks of land are held in groups. Large herd sizes (mean of 113 heads of cattle) among pastoralists is considered as a social security besides being the only source of livelihood, especially in Kajiado County. Respondents had many years of experience in keeping livestock with a mean of 24.58 years and a significant deviation of 12.09 years. This implies that one starts keeping livestock at youthful age and assumes full ownership immediately upon getting married according to the Maasai culture. Given the vastness of the ASALs, pastoralists cover long distances (mean of 9.64 km) in both counties to reach the nearest market centres which couple as livestock markets as well. Most farmers were aware of AI services as an alternative breeding method to the bull. A total of 130 and 132 of the sampled farmers in Kajiado and Narok Counties respectively were aware but had not used AI services before.

Most farmers, 55.8% in both counties belonged to either a production or a marketing group. It’s within these groups that they are able to constructively share production knowledge on livestock feeds and general animal health. Most of the sampled farmers, 31% and 39% from Kajiado and Narok counties respectively, had access to credit from their informal groups in the last 12 months. Survey results further indicate that 15% of sampled farmers had other income generating activities besides livestock and crop farming. Only 10.1% of sampled farmers in Kajiado were nomads and 5.83% in Narok County. Nomads are less likely to adopt new innovations in livestock production because of their movements.

Pastoralist Farmers’ Preference for Artificial Insemination

Pastoralists had different perceptions of AI (Table 3). Approximately 68.4% of farmers least preferred AI while 23.62% and 7.98% revealed their preferences as preferred and most preferred respectively within the counties sampled. Comparison of farmers’ preference for AI by county showed that more farmers in Narok County preferred AI (9.66%) than their counterparts in Kajiado County (6.00%). A significantly high proportion (81.33%) of farmers in Kajiado County (81.33%) had the least preference for AI compared to 57.39% in Narok County. In contrast, 32.95% of farmers in Narok County preferred AI compared to 12.67% in Kajiado County. This difference can be explained by the differences in nature of lifestyles. Pastoralists move large herds for long distances and thus may neither have time for close observation for heat detection in cows or response for artificial insemination. The infrastructure and cost of maintaining frozen semen is exorbitant. This is unlike agro-pastoralists in Narok who live on the farm with their animals and might access the breeders.

Table 3. Pastoralist farmers’ perception of AI
Preference level Whole sample (n=326) Narok County (n=176) Kajiado County (n=150)
Least preferred 223 (68.4) 101 (57.4) 122 (81.3)
Preferred 77 (23.6) 58 (33.0) 19 (12.7)
Most Preferred 26 (7.98) 17 (9.66) 9 (6.00)
Note: Percentages are in parenthesis
Source: Survey data, 2013
Factors that influence pastoralists’ perception of artificial insemination

Most pastoralists (66.9%) preferred AI based on its affordability across the two counties. There is insignificant difference in perceptions of pastoralists about the cost of using this technology across the two counties with 65.3% and 68.7% of all farmers in Narok and Kajiado acknowledging it as the best cheap alternative breeding method as shown in table 4. The high percentages could be attributed to the inadequate supply and high demand of Sahiwal bulls that influence farmers to be willing to pay for alternative breeding methods apart from the bull to enhance their Sahiwal production.

Based on the ability of AI calf to tolerate arid and semi- arid climate, 25.8% of the sampled population preferred to use AI. Disaggregated results by County mirror the same low levels with a significant difference in levels of preference across the two Counties. A high proportion of pastoralists in Narok County (30.7%) approve of AI calves survival ability compared to their counterparts in Kajiado County (20%).

Table 4. Factors that influence pastoralists’ perception towards AI
Determinants Whole sample Narok County Kajiado County χ² P
Affordability 218(66.9) 115(65.3) 103(68.7) 0.404 0.525
Calf tolerance 84(25.8) 54(30.7) 30(20.0) 4.830 0.028
Accessibility 76(23.3) 50(28.4) 26(17.3) 5.556 0.018
Success rate 77(23.6) 42(23.9) 35(23.3) 0.013 0.911
Note: Percentages are in parenthesis
Source: Survey data, 2013

The accessibility of a technology to its intended recipients is critical in achieving its purpose. Among the pastoralists sampled, 23.3% expressed their interest in using AI based on their ability to access it. Kajiado County registered the lowest percentage of farmers (17.3%) who had access to AI compared to those in Narok County (28.4%). The sparse nature of settlements in Kajiado County explains low adoption rates of assisted reproductive technologies in livestock production compared to Narok County. These statistics confirmed the views of most farmers who participated in focused group discussions that it is not easy to access AI in the ASALs. An insignificantly low percentage of pastoralists were interested in adopting AI based on their knowledge of success rate (23.6%). This can be attributed to lack of relevant information about AI among pastoralists, which explains why there are low adoption rates in ASALs compared to highland dairy farmers.

Ex-ante determinants of Artificial insemination adoption in arid and semi- arid areas of Kenya.

Results of ordered probity regression are presented in Table 5. The Likelihood Ratio (LR) Chi-Square test for the goodness of fit shows that at least one of the covariates in the model is not equal to zero thus the model provides good fit for the data. The chi statistic (χ²) is highly significant (p < 0.0000). The test for multicollinearity resulted in a mean variance inflation factor (VIF) of 1.27. This means that there was no multicollinearity among the variables based on VIF, since it was less than the threshold of 10, as explained by Maddala (1997)

A summary view of the model coefficients in the first column indicate that group membership, access to extension, AI awareness, agro-pastoralism as a production system, county of residence, nomadic lifestyle, years of education, household size and herd size have a positive significant effect while experience and age had a negative significant effect on the perception of farmers towards AI adoption. Access to credit, distance to local market, land size and off-farm income were insignificant. The marginal effects presented in the 2nd, 3rd and 4th columns refer to a small change in the dependent variable due to a marginal change in the explanatory variable, ceteris paribus. We discuss in detail the model results below.

Membership to a group was found to have significant positive influence on farmer’s perception towards AI. Group membership increased the probability of preferring AI by 9.4% and ranking AI as the most preferred by 1.6 %. This could be attributed to the fact that group members benefit from established social capital that enhances sharing of production information and knowledge. Most groups that were found to exist were mainly engaged in livestock production and marketing of animals among the men while women formed milk selling groups. Members shared knowledge within the groups and could invite livestock experts to teach them on better methods of production, which was common in Narok County. These results suggest that groups provide a better avenue through which interventions targeting farmers could be disseminated which confirms findings by Mignouna et al (2010) on adoption of new maize and production efficiency in western Kenya.

Table 5. Result for ordered logit regression for revealed preference of AI by pastoralist farmers
AI preference Coefficients Marginal effects
(least preferred)
Marginal effects
(preferred)
Marginal effects
(most preferred)
Group membership 0.358(0.054) - 0.109 (0.046) 0.094(0.048) 0.016(0.085)
Extension 0.327(0.066) - 0.105(0.074) 0.089(0.071) 0.017(0.073)
AI awareness 0.369(0.072) - 0.123(0.087) 0.102(0.079) 0.021(0.170)
Agro-pastoralism 0.517(0.003) - 0.165(0.004) 0.139(0.004) 0.027(0.031)
Narok county 0.538(0.005) - 0.165(0.004) 0.140(0.004) 0.024(0.028)
Nomadism 0.406(0.087) - 0.138(0.109) 0.113(0.096) 0.025(0.207)
Education 0.421(0.000) - 0.131(0.000) 0.112(0.000) 0.019(0.001)
Access credit 0.188(0.348) -0.059(0.332) 0.049(0.337) 0.008(0.330)
Household size 0.051(0.000) - 0.016(0.000) 0.014(0.001) 0.002(0.010)
Herd size 0.001(0.039) -0.0004(0.039) 0.0003(0.042) 0.0001(0.072)
Experience -0.025(0.002) 0.008(0.002) -0.007(0.003) -0.001(0.017)
Youth-aged -0.742(0.013) 0.190(0.002) -0.168(0.002) -0.022(0.013)
Middle-aged -0.508(0.029) 0.167(0.036) - 0.138(0.032) - 0.029(0.104)
Distance 0.004(0.627) - 0.001(0.628) 0.001(0.628) 0.0002(0.631)
Attribute index 1.94(0.001) - 0.602(0.001) 0.515(0.001) 0.087(0.011)
Land size 0.00005(0.727) - 0.00001(0.727) 0.000001(0.727) 0.00002(0.727)
Off-farm income -0.121(0.611) 0.037(0.600) -0.032(0.603) - 0.005(0.583)
/cut-off 1 2.92 0.614
/cut-off 2 4.31 0.646
Number of observations 296
Prob > χ² 0.0000
Note: P values in parenthesis
Source: Survey data, 2013

The results further showed that access to extension services increased the probability of a farmer moving from a lower preference level to a higher level (p <0.1). These findings are consistent with those of Kaaya et al (2005) in Uganda and Adegbola and Gardebroek (2007) in Benin who found adoption of technologies by maize farmers depended largely on receiving production information from extension agents or from other farmers. In our study area, provision of extension services (22% in Kajiado and 41% in Narok had access to extension services) is still low which forced most farmers to seek advice from private livestock input suppliers to improve their production. In response to Government policy on extension service provision to farmers, NGOs in pastoral areas organize open field days for farmers to interact and share production experiences. Field days are taken as a form of extension, and explain the reason as to why participating farmers are more likely to adopt AI than non-participant farmers. These results complement the findings of Amudavi et al (2008) on evaluation of farmers’ field days as a dissemination tool for Push-Pull technology in western Kenya. The authors observed that the farmers’ propensity to seek new agricultural knowledge motivated farmers to attend the field days and seek extension services.

Artificial insemination awareness was found to have a positive relationship with preference. The coefficient indicates that awareness increases the probability of a farmer to move from lower rank (least preference) to preferring AI by 0.102 times. The rate of adoption of AI technology depends on the availability of information to the farmers about that technology. These findings are consistent with Murage et al (2011) who found effective and efficient dissemination pathways to be fundamental in accelerating adoption rates. The low levels of AI uptake in ASALs in this study could be attributed to lack of relevant knowledge to the farmers about its potential to increase production and the technical know-how. These results are consistent with findings of Johnson and Ruttan (1995) who established breeding technologies to be highly information intensive. To effectively use AI technology, the farmer is required to understand breeding principles, performance data, management and analysis to.

Experience in keeping livestock made farmers realise that AI was less feasible than use of bulls for breeding within the pastoral context. This means that a unit increment in years of experience in keeping livestock reduces the likelihood of moving from a lower level of preference to higher preference levels for AI (p < 0.01). This result contradicts most studies that found relative farming experience to accumulate knowledge that influence a farmer to adopt new technologies that would boost production (Odendo et al 2010; Motuma et al 2010). Most pastoralists who had kept livestock for many years held a pessimistic view about the ability of breeding using Assisted Reproductive Technologies besides the bull. The coefficients for age groups indicate negative significant relationship with AI adoption. That is, the older a pastoralist becomes the less likely that he would shift his low perception towards AI to higher preference. However, young farmers are more willing to take the risk of adopting new technologies unlike their old counterparts that are more risk averse, corroborating study findings by Howley et al (2012) on AI adoption in Ireland.

Education of a household head had a significant positive influence on the perception of a pastoralist towards AI. Given that it’s a new technology among most pastoralists, those who are educated were willing to take the risk and try, while the less and non-educated were more risk averse. These findings are consistent with most adoption studies of Genius et al (2006), Abebe et al (2013), Makokha et al (2008) and Howley et al (2012). Since a new technology is developed to improve production, education has been found to increase farmer’s ability to obtain and evaluate information about an innovation before making informed decision on its use.

There was a significant difference in the perception effect of AI across Counties. Most farmers (61.3%) in Narok County are agro-pastoralists hence movement of their animals is limited to short distances from their homesteads unless there is a prolonged dry spell. Limited movement of livestock makes it easier to spot animals on heat and call service providers to their homes to administer AI compared to nomadic pastoralists in Kajiado who moved with animals from one locality to another in search of water and pastures.

Household size was significant and positively related to preference for AI. This means that an additional member to the household would increase the likelihood of preferring AI as a breeding technology. This could be alluded to the fact that huge herd sizes symbolizes wealth in Maasai community and livestock being the major source of their livelihood, any attempts to improve and grow the herd size is most welcome. The current results are in line with findings of Mignouna et al (2010) who reported household size as a proxy to labour availability and a positive relationship with adoption of Insect Resistant maize.

Herd size was found to be significant at 1% indicating that an increase in herd size by one animal increases the probability of a pastoralist to prefer AI by 0.0003 times. This is sensible given that AI can be done simultaneous on many animals at the same time unlike a single bull in a large herd size. The small influence of herd size on adoption of AI contradicts findings of Janssen et al (2006) who found availability of AI to be practically zero and needs an organization structure that targets livestock under intensive production system. The workers also established that the use of AI required comparatively more input into the infrastructure and education since pastoralist’s experiences with this technique is very little.


Conclusion and policy implications


Recommendations


Acknowledgement

The financial support from both East Africa Agricultural Productivity Project and African Economic Research Consortium is gratefully acknowledged.

Conflict of Interest: The authors declare that they have no conflict of interest.
Statement of human rights:
Ethical approval: For this type of study formal consent is not required.
Statement on the welfare of animals
Ethical approval: This article does not contain any studies with human participants or animals performed by any of the authors.
Informed consent: Informed consent was obtained from all individual participants in this study.


References

Abebe G K, Bijman J, Pascucci S and Omta O 2013 Adoption of improved potato varieties in Ethiopia: The role of agricultural knowledge and innovation system and smallholder farmers’ quality assessment.Agricultural Systems 122: 22-32. http://econpapers.repec.org/article/eeeagisys/v_3a122_3ay_3a2013_3ai_3ac_3ap_3a22-32.htm

Adegbola P and Gardebroek C 2007 The effect of information sources on technology adoption and modification decisions. Agricultural Economics,37(1) 55-65.

Amudavi D M, Khan Z R, Wanyama J M, Midega C A, Pittchar J, Hassanali A and Pickett J A 2009 Evaluation of farmers' field days as a dissemination tool for push-pull technology in Western Kenya. Crop Protection 28(3) 225-235.

DeShazo J R and Fermo G 2002 Designing choice sets for stated preference methods: The effects of complexity on choice consistency. Journal of Environmental Economics and Management 44: 123-143. http://econpapers.repec.org/RePEc:eee:jeeman:v:44:y:2002:i:1:p:123-143

FAO 2010 Breeding strategies for sustainable management of animal genetic resources. Food and Agricultural Organization of the United Nations (FAO): Animal Production and Health Guidelines. No. 3. Rome. http://www.fao.org/docrep/012/i1103e/i1103e.pdf

Genius M, Pantzios C J, and Tzouvelekas V 2006 Information acquisition and adoption of organic farming practices. Journal of Agricultural and Resource Economics 31(1): 93-113. http://ageconsearch.umn.edu/bitstream/10150/1/31010093.pdf

GoK 2010 Kenya Population and Housing Census of 2009. Population and household distribution by socio-economic characteristics. Nairobi, Government of Kenya.

Greene W H 2003 Econometric analysis, Fifth edition. New Jersey, Prentice Hall.

Howley P, Donoghue C O and Heanue K 2012 Factors affecting farmers’ adoption of agricultural innovations: A panel data analysis of the use of artificial insemination among dairy farmers in Ireland.Journal of Agricultural Science 4(6): 171-182. http://www.ccsenet.org/journal/index.php/jas/article/view/10687/11101

Ilatsia E D, Migose S A, Muhuyi W B and Kahi A K 2011a Sahiwal cattle in semi-arid Kenya: genetic aspects of growth and survival traits and their relationship to milk production and fertility.Tropical Animal Health and Production 43(8): 1575-1582. http://link.springer.com/article/10.1007/s11250-011-9845-x/fulltext.html

Ilatsia E D, Roessler R, Kahi A K and Valle Zrate A 2011b Breeding and conservation programmes for Sahiwal cattle genetic resources in the tropics: a review. Animal Genetic Resources, 2011, 49, 65–74. Food and Agriculture Organization of the United Nations. http://journals.cambridge.org/download.php?file=%2FAGR%2FAGR49%2FS2078633611000336a.pdf&code=6c5965233a3eb18713924f0d065b8fde

Illatsia E D, Muasya T K, Muhuyi W B and Kahi A K 2007 Milk production and reproductive performance of Sahiwal cattle in semi-arid Kenya. Tropical Science 47(3): 120-127.

Jaetzold R, Schmidt H, Hornetz B and Shisanya C 2005 Farm Management Handbook of Kenya – Natural conditions and farm management information. Nairobi, Government printer. http://www2.gtz.de/dokumente/bib/07-1286.pdf

Janssen-Tapken U, Kadarmideen H N and Rohr P V 2006 Cattle breeding strategies using genetic markers as a pathway for improving competitiveness of pastoral systems in Kenya. Paper contributed to the conference on Pastoralism and Poverty Reduction in East Africa, 27-28, June 2006, Nairobi, Kenya.

Johnson N L and Ruttan V W 1995 Diffusion of livestock breeding technology in the US: Observations on the relationship between technical change and industry structure. Department of Applied Economics, College of Agricultural, Food, and Environmental Sciences, University of Minnesota.

Kaaya H, Bashaasha B and Mutetikka D 2005 Determinants of utilization of artificial insemination (AI) services among Ugandan dairy farmers. Eastern Africa Journal of Rural Development, 21(1): 34-43.

KNBS 2010 Population census of 2009. Kenya National Bureau of Statistics.

Long J S 1997 Regression models for categorical and limited dependent variables: Advanced Quantitative Techniques in the Social Sciences 7. Thousand Oaks, CA. Sage Publications.

Maddala G S 1997 Limited-dependent and Qualitative Variables in econometrics. Cambridge University Press, Cambridge.

Makokha S N, Karugia, J T, Staal S J and Oluoch-Kosura W 2008 Analysis of factors influencing adoption of dairy technologies in western Kenya. Paper presented at the 2nd African Association of Agricultural Economists (AAAE) International conference on Agricultural Growth, Poverty Reduction and Millennium Development Goals in Africa, No 209-213, 20-22, August 2007, Accra, Ghana. http://ageconsearch.umn.edu/bitstream/52094/2/Makokha.pdf

Mignouna D B, Mutabazi K D S, Senkondo E M and Manyong V M 2011 Imazapyr-resistant maize technology adoption for witch weed control in western Kenya.African Crop Science Journal 19(3): 173-182. http://www.bioline.org.br/request?cs11017

MOLF 2006 Annual report. Ministry of Livestock and Fisheries Development, Division of animal production, Kajiado and Narok Districts, Nairobi, Kenya.

Motuma T, Dejene A, Wondwossen T, Roberto LR, Girma T, Wilfred M and Germano M 2010 Adoption and continued use of improved maize seeds: Case study of central Ethiopia. African Journal of Agricultural Research 5(17):2350–2358. http://repository.cimmyt.org/xmlui/bitstream/handle/10883/2186/94068.pdf

Muhuyi W B, Lokwaleput I and Ole Sinkeet S N 1999 Conservation and utilisation of the Sahiwal cattle in Kenya.Animal Genetic Resources Information, 26: 35-44. http://www.cattlenetwork.net/docs/agri/agri26_3.pdf

Murage A W, Obare G, Chianu J, Amudavi D M, Pickett J and Khan Z R 2011 Duration analysis of technology adoption effects of dissemination pathways: A case of ‘push–pull’technology for control of Striga weeds and stem borers in Western Kenya.Crop Protection30(5): 531-538.

Muriuki H G 2011 Dairy development in Kenya: Food and Agriculture organization of the United Nations. http://www.fao.org/docrep/013/al745e/al745e00.pdf

Odendo M, Obare G and Salasya B 2010 Farmers’ perceptions and knowledge of soil fertility degradation in two contrasting sites in western Kenya. Land Degradation and Development, 21: 557–564.

Ouma E, Awudu A, Drucker A and Obare G 2004 Assessment of farmer preferences for cattle traits in smallholder cattle production systems of Kenya and Ethiopia. Paper presented at the Deutscher Tropentag 2004 Conference on International Agricultural Research for Development, 5-7 October 2004, Berlin, Germany.

Roessler R, Ilatsia E D and Valle Zrate A 2011 Optimized breeding strategies for the local Sahiwal cattle in Kenya: Use values and important breeding traits. Paper presented at 5th All African Conference on Animal Agriculture, pp. 25-28, Addis Ababa, Ethiopia. http://www.esap-ethiopia.org/Publications/Proceedings/18th_Book_Abstracts.pdf


Received 14 February 2015; Accepted 13 March 2015; Published 1 April 2015

Go to top