Adoption of East Coast Fever Vaccine among Smallholder Dairy Farmers in Kenya: The Case of North Rift Kenya

ABSTRACT

Diseases are a major constraint limiting dairy cattle production, with East Coast Fever (ECF) ranking among those of high economic importance. The high cost of ECF control and treatment led to the development of the ECF vaccine, which is envisaged to be safer, cheaper and effective. The objective of this study was to identify factors influencing the uptake of this technology. One hundred and eighty one (181) randomly sampled smallholder dairy farmers from North Rift Kenya, where the vaccine had been introduced, were used to estimate the determinants of uptake of East Coast Fever (ECF) vaccine in these areas. Results showed that the vaccine was likely to be adopted by relatively well-off households whose main source of income was off-farm employment, had relatively large herd sizes and could afford to practise on-farm tick control. Farmers with higher levels of education and advanced in age were also likely to adopt the vaccine. The paper derives important policy implications for enhanced ECF vaccine uptake in smallholder dairy systems in Kenya.

KEYWORDS:

• East Coast Fever vaccine
• dairy cattle
• adoption
• Kenya

Introduction

Cattle are an integral part of farming in Sub-Saharan Africa. In Kenya, development of the dairy subsector has been the focus of national development strategies, as evidenced by the Agricultural Sector Development Support Programme (ASDSP) (Government of Kenya, 2011), Vision 2030 (Government of Kenya, 2007), the National Livestock Policy (Government of Kenya, 2008) and Feed the Future (FtF) Programme (USAID, 2010). Milk is the most economically important livestock product with an estimated value of KES 257.811 billion in 2009 and an estimated 70% of the total gross value of livestock's contribution to the agricultural sector (Behnke and Muthami, 2011). Diseases are a major constraint limiting dairy cattle production, with East Coast Fever (ECF) ranking among those of high economic importance (Mbogo, 1998).

The ECF is caused by the Theileria parva parasite and transmitted by the Rhipicephalus appendiculatus tick. The high cost of ECF control and treatment increases production costs to the farmer. This has led to the development of immunization methods envisaged to be safer, cheaper and more effective (Mukhebi and Perry, 1993). One such vaccine is the ‘Muguga cocktail’ live vaccine.

Muguga cocktail has mainly been used within pastoral cattle production systems and proven effective, but its administration in smallholder dairy cattle systems is not well documented. Thus, an initiative to immunize dairy cattle was implemented in 2010–2011 in North Rift Kenya by VetAID (a local non-governmental organization) with technical support from the Department of Veterinary Services (DVS), the International Livestock Research Institute (ILRI) and the Kenya Agricultural Research Institute (KARI). The main aim of the project was to promote the use of the vaccine and evaluate its effectiveness under smallholder dairy production systems.

VetAID conducted immunizations among willing households in the target areas, with impressive results on the performance of the vaccine (Turasha, 2011, Wesonga et al., 2013). However, information on factors influencing adoption of the vaccine by the smallholders remains scanty. In an attempt to fill this knowledge gap, this case study paper analyses the determinants of ECF vaccine adoption in North Rift Kenya where promotion of the vaccine was conducted. It hypothesizes that the adoption of vaccine is influenced by, among other factors, socioeconomic, demographic and dairy cattle enterprise management characteristics of households.

It is expected that an understanding of the importance of these factors will guide targeting of farmers and regions, and prioritization of ECF vaccine awareness campaigns for enhanced uptake, even after withdrawal of funded programmes.

Methodology

The study area

The study was conducted in Uasin Gishu (Ziwa and Sirikwa sites), Kericho (Kipkelion site) and Nakuru (Olenguruone site) counties located in the North Rift Kenyan region. The three counties are similar in many aspects including agro-ecological conditions, farming systems, population densities, cultural and economic activities. They were selected for this study as a follow-up to a pilot project implemented by VetAID-Kenya and financed by the Food and Agriculture Organization of the United Nations (FAO-Kenya), which focused on immunization of dairy cattle using the Muguga Cocktail ECF vaccine from December 2010. Immunizations were conducted in smallholder dairy production systems in the region.

Data collection procedures

The study utilized primary data collected in a household survey conducted between August and September 2013. The criterion for selection of households was participation in the ECF vaccine pilot delivery project.

Selection of the sites was based on key informant interviews conducted with VetAID, Kenya Dairy Farmers Federation (KDFF) and Heifer International. Within each site, a cooling/chilling plant was the focal point during the ECF vaccine pilot delivery project where vaccinations were conducted among members of the milk cooling/chilling plants. With the guidance of personnel in each cooling/chilling plant, sampling frames consisting of members whose cattle had been vaccinated were developed, and those who had not participated in the ECF vaccination programme. From these, a total of 181 households were randomly selected. The sampled households were interviewed by trained enumerators using a semi-structured questionnaire which had been pretested. In addition to socioeconomic and demographic characteristics, data on cattle production practices and ECF vaccination characteristics were also collected.

The empirical model

The econometric model developed in this study is based on the random utility theorem (Gujarati, 2003) which postulates that consumers (in this case, farmers) will choose or adopt a technology which can maximize their utility. The decision to adopt a technology or not is a binary decision which can be represented as a qualitative variable whose range is actually limited since it can only take on two values: adopt or not adopt. An adopter in this study is defined as any farmer who had vaccinated cattle (at least one) against ECF at the time of the study. Thus adoption at the farm level describes the realization of a farmer's decision to apply a new technology in the production process (Rogers, 1995).

An empirical model (equation (1)) was derived for econometric analysis and analysed using probit regression: (1)

Description of variables, rationale and hypotheses

The explanatory variables (independent variables) used in this study include demographic characteristics, socioeconomic and dairy cattle enterprise characteristics. Their definitions and hypotheses are presented in Table 1. The dependent variable is VACC_ADOPT (adoption of ECF vaccine technology). It is coded 1 if a household has vaccinated against ECF, and 0 if otherwise.

Table 1. Definition of explanatory variables

Variable	Definition	Hypothesis
GENDER	Sex of the household head (1 = Male, 0 = Female)	+
EDUCATION	Education level of the household head in years	+
OCCUPATION	Main occupation of the household head (1 = Farming 0 = Off-farm)	−
AGE	Age of the household head in years	+
MKTD_MILK	Whether the household markets more milk than consumed and calf fed (1 = Yes 0 = No)	+
TICK_CONT	Method of tick control (1 = Spraying 0 = Dipping)	+/−
EXOTC_BREED	Breed of cattle (1 = Exotic/crossbred 0 = Indigenous)	+
CATTLE_NUM	Number of cattle kept	+
EXP_ECF	Whether the household has experienced ECF (1 = Yes 0 = No)	−
GRAZE_SYST	Whether the cattle management system is zero grazing (1 = Yes 0 = No)	+

The rationale for the selection of the variables was based on a literature review of similar agricultural technology adoption studies. Demographic factors are represented by the variables AGE and GENDER, which are hypothesized to have a positive influence on the adoption of the ECF vaccine just like in other adoption studies (Adesina and Chianu, 2002; Baltenweck and Staal, 2000).

Among the socio-economic variables, the education level of the household head (EDUCATION) is likely to influence the adoption decision positively, as demonstrated in dairy cattle studies (Baltenweck and Staal, 2000; Staal et al., 2002). The main occupation (OCCUPATION) of the household head is hypothesized to negatively influence the adoption decision, because farmers with an off-farm income source are more willing and able to pay for the vaccine.

Dairy cattle enterprise characteristics include breed of cattle (EXOTC_BREED) and size of herd (CATTLE_NUM). It is expected that farmers with relatively large cattle herds and with exotic breeds would be willing to make additional investments to improve the productivity of their dairy cattle enterprises. It is also hypothesized that those producing milk and selling a higher proportion than the quantity consumed at the household level (MKTD_MILK) would be willing to take measures to reduce production costs and ensure minimal disruptions to the income flow obtained from the enterprise. Experience with ECF incidences in the past (EXP_ECF) are hypothesized to expose farmers to the high cost of treatment of ECF or death of the infected animal thus prompting the farmer to ensure other cattle in the herd are protected. The grazing system (GRAZE_SYS) and method of tick control (TICK_CONT) are proxies for the farmer's willingness and ability to invest into the dairy cattle enterprise.

Data management

Data coding, entry and cleaning was done using SPSS Version 19 software. Descriptive statistics, which included frequencies, means and cross-tabulations, were generated using SPSS and used to characterize households’ socioeconomic and demographic characteristics and practices. The econometric probit model was analysed using LIMDEP statistical software.

Results and discussion

Farming was the main occupation for the majority (72.8%) of the households while 21.1% were salaried workers and 6.1% were self-employed off-farm. The majority (92.2%) of the sampled households were male-headed with the mean age of the household heads being 50 years. Ninety five percent of the household heads had at least acquired basic education with the mean number of years of formal education being 10 years. This indicates that most of the households in the research area had acquired the basic primary education level. The majority of the household heads (85%) were married and living together with their spouses while the rest comprised those married but with one of the spouses living away, single and widowed.

The results of the probit model are presented in Table 2. The model correctly predicted 84% of the responses and the Chi-square value (53.96) (P<0.001). The psuedo R² was 26%, which is within the range allowed with the kind of data being analysed in this study (Mbata, 1997; Greene, 2003). The other model statistics are presented at the bottom of Table 2.

Table 2. Probit results – factors influencing adoption of ECF vaccine

Variable	Coefficient	Standard error	t-ratio
CONSTANT	−2.046**	0.814	−2.513
GENDER	−0.368	0.443	−0.832
EDUCATION	0.065**	0.031	2.102
OCCUPATION	−0.562**	0.270	−2.082
AGE	0.026***	0.009	2.816
MKTD_MILK	−0.057	0.230	−0.250
TICK_CONT	0.720***	0.252	2.853
EXOTC_BREED	0.056	0.262	0.215
CATTLE_NUM	0.087***	0.022	3.997
EXP_ECF	−0.217	0.229	−0.948
GRAZE_SYST	0.424	0.543	0.781

Log likelihood function −92.35

Restricted log likelihood −119.33

Chi-squared = 53.96***

Pseudo R² = 0.259

No. of observations = 181

** and *** = significant at 5% and 1% levels, respectively

The probit model results showed that except for GENDER and MKTD_MILK, all the other variables had the hypothesized signs. Thus, our hypotheses that male farmers and those who market higher quantities of milk than used by the household (consumed and fed to calves) were more likely to vaccinate their cattle against ECF, could not be proved by this study.

Among the household head's characteristics, EDUCATION and AGE of the household head emerged as key variables that significantly and positively influenced the probability of adoption of the ECF vaccine. The positive coefficients of these factors indicate that farmers who were more educated were more likely to understand the benefits of the vaccine, and hence vaccinated their cattle against ECF. Those more advanced in age were likely to have more experience in cattle rearing, and hence understood the likelihood of cattle getting ECF and the associated negative effects, thus they vaccinated against the disease. OCCUPATION of the household head had a significant, albeit negative, influence on the farmer's likelihood to vaccinate against ECF. This factor indicates that farmers whose main occupation was off-farm employment were more likely to vaccinate their cattle against ECF compared to full-time farmers. This relationship may be attributed to the higher purchasing power of off-farm employees, hence their ability to meet the cost of vaccination.

As hypothesized, TICK_CONT (method used in control of ticks in a herd) had a positive influence on adoption of the vaccine. Farmers spraying cattle as a tick control practice were more likely to adopt the vaccine. Spraying as a tick control method was indicative of a relatively well-off farmer, implied by the ability to purchase equipment and acaricides, and employ labour required for spraying at the farm level. CATTLE_NUM had a positive significant influence on the adoption of the vaccine. This factor indicates that households with larger herd sizes had a higher propensity to vaccinate against ECF than those with smaller ones. This may be attributed to the ease of selling cattle or cattle products (e.g. milk) to meet the cost of vaccination for other cattle in the herd.

Conclusions and recommendations

This study made an attempt to estimate empirically factors that influence farmers’ uptake of the ECF vaccine. The stutdy identified characteristics of farmers likely to adopt this technology. The factors generated can therefore assist in identifying and targeting farmers who are likely to adopt the vaccine in the future. To some extent, these results could also be used for prioritizing ECF vaccine technology awareness campaigns.

The study shows that the likelihood of adopting the vaccine is positively linked to off-farm occupation, relatively large cattle herds and on-farm tick control methods (spraying). These factors are indicators of relatively well-off farmers. Thus, to enhance adoption, farmers with these characteristics can be targeted since they are more likely to adopt the technology than their resource-poor counterparts. In addition, to ensure that their resource-poor counterparts are not marginalized in adopting the technology, lobbying for subsidies targeting them would accelerate their adoption thus enhancing equitable access to benefits of the technology.