| Source of technology and funding | IFAD-supported research carried out by the International Livestock Research Institute (ILRI) |
| Expected Benefit: | Improved efficiency in delivery of General: Understanding farmer perceptions of risk so as to enhance the adoption of livestock disease risk management strategies, including vaccines for East Coast fever in Kenya |
| Crops and enterprise: | Livestock-keeping |
| Agro-ecological zones: | All |
| Target region and countries: | Developing countries |
| Keywords: | Understanding farmers’ perceptions, management styles, risk management, animal disease prevention, East Coast Fever, technology adoption, Participatory Learning Appraisal |
Often, farmers do not adopt potentially effective and cost-beneficial technology and information packages. One contributing factor is that these packages are not tailored to the perceived and actual needs of farmers. An important step in improving technology and information packages so as to maximize their uptake and impact is to understand farmer perceptions of agricultural constraints and the benefits of different technologies in solving these constraints. For livestock disease control, farmer perceptions of disease risk and the inherent benefits and risks associated with available disease control options will be important criteria in adoption decisions.
In this note, a participatory approach to the assessment of farmer perceptions of the risk of livestock disease and the different methods for disease control is presented. This approach is applied to the study of the risks of East Coast fever and the control strategies, including vaccines, in four smallholder mixed farming villages in Kenya.
Farmers manage risk through a continuous adaptive process, whereby decisions are made based on perceptions of the external environment, resources and the farmers’ own attitudes and preferences. Hardaker, Huirne and Anderson (1997) have characterized the process by which farmers arrive at risk management decisions and practices. Briefly, the farmers acquire knowledge of their own context. Then the risks are identified, analysed and assessed. After assessment of the risks, if action is deemed worthwhile, the most suitable option for avoiding, preventing or managing the risks is selected. This process is continuously monitored and reviewed. Figure 1 is a schematic representation of this process.
Figure 1. Phases in the risk management decision cycle
In
the risk management process, farmers consider and respond to a combination
of external and internal factors, such as market access and the resources
available to the farm household. It is not necessarily only the factors
affecting risk, but also the farmers’ perceptions of these, that
are crucial in decision-making. And, for disease risk management, it is
not only the farmers’ perceptions of disease risk that are important,
but also their perception of potential additional risks associated with
available disease control strategies.
Farmers consider many and varied risks in their day-to-day decision-making. In considering risk, one must also understand the overall agro-ecological context, the production systems of the farmers and the main household types within these broader systems. Farming risks vary greatly among systems, from low- rainfall, extensive, dryland systems to high-rainfall, densely populated highland systems. The contexts which condition household risk perceptions within production systems are also variable. One important criterion is household resources. Resource-poor farmers have less capacity to manage risk. Farmers as individuals also differ in their goals and attitudes towards risk.
Many farmers apply multiple livelihood strategies. These involve general household risks and more specific agricultural and livestock risks, including disease risk. To consider the perceptions of any particular risk, such as disease risk, one should place the risk perceptions within the broader context of the household, which has different, sometimes competing risks.
In any understanding of farmer perceptions of risk management options,
some key questions are:
a. Of which risk management options are farmers aware?
b. Which among these do they use?
c. How do they choose between different options or combinations of options?
Because it is the farmers’ perceptions of risk management options, rather than merely the actual factors influencing risk, that influence adoption, participatory methods that can both synthesize and identify the diversity of perceptions can be valuable for understanding current and potential risk management options.
Participatory learning appraisal
There are many participatory appraisal methods available. For the assessment of farmers’ perceptions of risk, we considered the main methodological requirements to be a systematic coverage of issues for cross-site comparisons, an ability to explore and identify the main common risk perceptions and their range, as well as farm management types, and a relatively rapid procedure for both farmers and researchers. The method chosen to meet these criteria was participatory learning appraisal (PLA). The PLA process encourages initial input from many farmers so as to capture the diversity and complexity of perceptions and practices. Then interactive group discussions are used to examine in more detail and to seek consensus and an emphasis within the group on the correction of initial misconceptions (Chambers, 1997). In a PLA, unlike the more well known participatory rapid appraisal, the focus is not so much on the stimulation of community participation, but on the enhancement of information exchanges and analysis of the farmers’ situation with the farmers themselves through contextual understanding and learning.
Similar to participatory rapid appraisals, PLAs should be a structured process and require facilitation. There are a number of key steps that require attention in the four phases – preparation, implementation, analysis and reporting – because they can greatly affect the value of the study:
a. selection of the villages: the criteria and sampling method are critical;
b. choice of participatory tools and techniques, including language: to
secure mutual learning by researchers and farmers;
c. facilitator training: to acquaint facilitators with the research objective(s),
choose the strategy for facilitation and recording, rehearse facilitation
methods and develop information checklists to ensure all topics are covered
and allow for cross-site comparisons;
d. choice of key informants: to introduce the researchers and facilitators
to the community;
e. choice of time and place of event: to increase attendance in the sessions;
f. formation of focus groups: to guarantee information quality; focus
group discussions with 8-10 people who are similar as regards the topics
of discussion give the most reliable information;
g. synthesis and the use of the results: the information that has been
generated during a PLA needs to be verified.
Pla application to farmer perceptions of the risk and control of east coast fever
The risk of East Coast fever (ECF) and the control strategies for ECF can vary greatly depending on climatic zone and production system. Three regions, namely, the central highlands, the western highlands and the Lake Victoria basin, had been selected for studies on ECF epidemiology and thus represented a good opportunity to examine the biologically measured risk of ECF in light of the perceived risk and the risk management practices adopted. Four districts (Kiambu and Maragua in the central highlands, Uasin Gishu in the western highlands and Kakamega in the lake basin) were selected. One administrative division in each district was randomly selected using a geographic information system. Villages within the division were listed, and, from among those that had smallholder farmers, one village was randomly selected. The four study villages thus selected are listed in Table 1, along with some of their characteristics.
Prior to the village visits, a three-day facilitator-training workshop was held. The PLA team comprised two facilitators, two researchers and the local animal health assistant. The PLA study was conducted over five days in each village. Two facilitators were present for all community and focus group meetings. At any given moment, one focused on encouraging responses, while the second listened and recorded. Thus, one needed to have cheerleader qualities to stimulate an information flow, while the other needed to have associative qualities to stimulate analysis among the farmers. They alternated in facilitating and recording roles, depending on the requirements of the process. Both had training in livestock-related subjects and spoke the local languages. The first two days comprised community-wide activities. On the first day, a transect walk was taken, and logistical arrangements were made for later meetings. On the second day, villagers described their natural resources, farming systems, institutions, main economic activities and household needs using participatory tools such as village maps and Venn diagramming. Farmers also classified farm types in their village based on differences in farm activities, livestock practices and other criteria. The underlying criteria used to differentiate categories were discussed. Although the four study villages differed markedly in farm and cattle management practices, the criteria by which farmers classified themselves into farm management groupings were very similar. In each village, two main criteria, with two levels each, emerged, thereby determining four management categories (Table 2). Farmers were then asked to assign themselves to one of the categories for focus group formation.
Table 2. Participant criteria types among focus groups
| Village | FG 1 | FG 2 | FG 3 | FG 4 |
| Waguthu | W1 good resources, not committed |
W2 few resources, committed |
W3* (*key informant) good resources, committed |
W4 few resources, not committed |
| Emurumba | E1 free-grazing, zebu cattle |
E2 zero-grazing, grade cattle |
E3 road-side grazing, zebu cattle |
E4 semi-zero-grazing, cross-bred cattle |
| Seiyo A | S1 good resources, modern knowledge |
S2 few resources, modern knowledge |
S3 few resources, traditional practices only |
S4 good resources, traditional practices only |
| Munyaka | M1 men, free-grazing, zebu cattle |
M2 women, zero-grazing, cross-bred cattle |
M3 men, zero-grazing, cross-bred cattle |
M4 women, free-grazing, zebu cattle |
Focus group sessions were spread over three days. To begin, matrix scoring (Chambers, 1997) was used to assess the relative importance of different farm enterprises. The groups first listed the main household needs (food, income, etc.) and ranked the six to eight most important farm enterprises. Then the participants were given a fixed number of beans and were asked to weigh the relative importance of the different enterprises, the different household needs and the ability of the different enterprises to meet these needs.
Farmer perceptions of disease risk and their knowledge of cattle diseases and control options were then discussed. The diseases were listed, and the farmers asked about the symptoms, mode of transmission, prevention and treatment of the diseases. The farmers then ranked the diseases based on perceived importance. After doing the ranking, the farmers were asked to explain their ranking criteria. If ECF had not been mentioned, the facilitators specifically probed for any knowledge of ECF at this stage.
For each disease and specifically for ECF, the farmers were asked to list the disease control and prevention measures that were known and available to them. Control methods that were not used and the reasons for this were then explored. The main approach used was a laddering technique (Malhotra, 1996). First, the group was asked to estimate what proportion of their group would not implement the listed control options. Then specific reasons for non-implementation were listed, and the proportion of non-adopters for each reason was estimated through an iterative process. Ranking techniques were also used to assess general risk management perceptions and the relative ranking of livestock diseases within these. The non-adoption of specific dairy technologies, such as zero-grazing and grade or exotic cattle, was also explored. Such practices are associated with ECF risk. (Exotic cattle are more susceptible to ECF, and zero-grazing reduces exposure to ticks.)
Farmer perceptions of the ecf risk management components
Farmer perceptions relative to the risk management decision cycle (Figure 1) were as follows.
1. Farmer perceptions of the risk context
The perceptions of the farming context (Venn diagram, listing of enterprises and household needs) did not differ greatly among people from the same village except initially in Munyaka. Munyaka is in a relatively newly settled area in which the communities do not have a long history of shared experiences and values.
Risks can be classified as production, market, institutional, personal, or financial (Hardaker, Huirne and Anderson, 1997). The two villages with extensive dairy systems (Emurumba and Munyaka) did not perceive market risks. In Waguthu, personal risk (illness) was not perceived as a farming risk. Livestock disease risk (a production risk) was identified by all focus groups, but assessed (ranked) differently between the extensive and intensive dairy systems.
2. Risk identification, analysis and assessment
Risks were ranked relative to the role of farm enterprises in meeting basic needs. In Waguthu and Munyaka (both Kikuyu villages), all needs were summarized succinctly as either food, or cash needs. In Emurumba (Luhya village) and Seiyo A (Kalenjin village), social needs were added. In three villages, Waguthu, Emurumba and Seiyo A, the cattle enterprise was considered most important in meeting basic needs. In Munyaka, cattle and maize farming shared the highest ranking. Table 3 shows the relative rank of the cattle enterprise.
Table 3 - Relative ranking of cattle in meeting
household needs
(0 = not important; 1 = all-important)
| Focus group village | 1 | 2 | 3 | 4 | Average |
| Waguthu | 0.36 | 0.32 | 0.50a | 0.38 | 0.39 |
| Emurumba | 0.47b | 0.27 | 0.26 | 0.36 | 0.34 |
| Seiyo A | 0.25 | 0.25 | 0.24 | 0.28 | 0.26 |
| Munyaka | 0.23 | 0.19 | 0.25 | 0.23 | 0.23 |
a This rank was provided by a key informant
as a focus group for this category could not be formed.
b This group decided a priori to allocate half their beans
to cattle and made very few adjustments to this.
Since cattle were the most important farm enterprise, cattle diseases
ranked high as risks, although not always the highest. In Waguthu, only
the resource-poor, committed group ranked disease as their greatest threat.
In Seiyo A, drought, hunger and the illness of family members were considered
the greatest risk. In this village, livestock diseases were perceived
as constraints rather than risks. In Emurumba, all groups except the free-
grazing zebu group ranked livestock diseases the highest. Across villages,
all focus groups that owned exotic breeds or crosses feared ECF the most
among all livestock diseases. ECF was identified as a serious risk and
analysed as a concrete problem in all but the two zebu-owning focus groups.
An interesting distinction was made between a risk and a constraint. In
Emurumba, livestock diseases are considered the major risk and are a risk
primarily because farmers have little knowledge of how to prevent them.
In Seiyo A, livestock disease was considered only a constraint because
farmers had the knowledge and ability to control them.
Table 4. Rankings of main diseases by the focus groups
Beginning with 1, rankings go from the most important to the least important
disease. FMD = foot and mouth disease. LSD = lumpy skin disease. For this
ranking, W3, W4 and S3 had less than five participants. In W1 and W2,
other diseases were mentioned, but not ranked.
| Village | Waguthu | Emurumba | Seiyo A | Munyaka | ||||||||||||
| Focus group disease | W 1 |
W 2 |
W 3 |
W 4 |
E 1 |
E 2 |
E 3 |
E 4 |
S 1 |
S 2 |
S 3 |
S 4 |
M 1 |
M 2 |
M 3 |
M 4 |
| ECF | 1 | 1 | 1 | 1 | 1 | 4 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | |
| FMD | 6 | 3 | 5 | 2 | 1 | 2 | 2 | 2 | 2 | 6 | 3 | 3 | 3 | |||
| Mastitis | 4 | 4 | 5 | 3 | 3 | 2 | 5 | 3 | 4 | 7 | 6 | |||||
| Pneumonia | 2 | 2 | 4 | 5 | 3 | 4 | 4 | |||||||||
| LSD | 3 | 6 | 5 | 3 | 8 | 6 | 5 | 8 | 4 | |||||||
| Helminthosis | 6 | 4 | 3 | 3 | 6 | 6 | ||||||||||
| Anaplasmosis | 3 | 3 | 3 | 4 | 5 | |||||||||||
| Eye infection | 6 | 5 | 7 | 4 | 4 | 5 | ||||||||||
| Black quarter | 2 | 4 | 4 | 3 | ||||||||||||
| Diarrhoea | 1 | 2 | 5 | |||||||||||||
| Ticks | 1 | 1 | ||||||||||||||
| Anthrax | 2 | 2 | ||||||||||||||
| Milk fever | 5 | 5 | ||||||||||||||
| Footrot | 4 | |||||||||||||||
| Ear infection | 6 | |||||||||||||||
| Rinderpest | 7 | |||||||||||||||
Identify risk management options
In the 16 focus groups, a total of 29 disease measures were mentioned, including ‘having faith’ and ‘doing nothing’. The most common measures are listed in Table 5. Ten of the 29 were measures specifically to control ticks and prevent tick-borne diseases. Although the groups were asked to list preventive measures only, six curative measures were listed by groups that had relatively poor knowledge of diseases. Two preventive measures were common to almost all focus groups: vaccination and tick control. Vaccination was mentioned in 15 of the 16 focus groups. It was not listed by the S2 focus group as that group only discussed preventive measures for ECF. All groups, except one zebu-owning group in Emurumba, mentioned tick control by spraying or dipping.
| Waguthu | Emurumba | Seiyo A | Munyaka | All | |||||||||||||
| Measure | 1 | 2 | 3 | 4 | 1 | 2 | 3 | 4 | 1 | 2 | 3 | 4 | 1 | 2 | 3 | 4 | |
| Vaccination | x | x | x | x | x | x | x | x | x | x | x | x | x | x | x | 15 | |
| Spraying | x | x | x | x | x | x | x | x | x | x | x | x | 12 | ||||
| Hygiene | x | x | x | x | x | x | x | x | 8 | ||||||||
| Deworming | x | x | x | x | x | x | x | 7 | |||||||||
| Dipping | x | x | x | x | x | x | x | 7 | |||||||||
| Quarantine | x | x | x | x | x | x | x | 7 | |||||||||
| Good feeding | x | x | x | x | x | 5 | |||||||||||
| Good housing | x | x | x | x | x | 5 | |||||||||||
| Call vet early | x | x | x | x | x | 5 | |||||||||||
Choice of risk management options
The main issue evaluated was the non-adoption of known disease control
options. Farmers in Waguthu, Emurumba and Seiyo A estimated that the proportions
of non-adopters of vaccination were 20%, 30% and 15%, respectively. The
non-adoption of tick control was considered rare and was estimated at
10%, 5% and 0%, respectively.
Focus groups formed according to resource criteria proved a good forum
for discussions of the reasons for non-adoption. General reasons and their
relative rankings were (in order of importance): (a) lack of money; (b)
poor information/knowledge; (c) farmer carelessness; (d) low disease risk;
(e) additional labour, plus distance to obtain control measure; and (f)
the safety and efficacy of the prevention measure. Interestingly, farmers
estimated high adoption proportions for tick control, but also had many
reasons for non-adoption. Most farmers apply or have applied tick control
at some time, but the frequency varies considerably.
Similar to other participatory methods, this method obtained qualitative/semi-quantitative insights on farmer perspectives of risk. While good facilitation and careful attention to detail can enhance information quality, it is very valuable if participatory risk perceptions can be validated along with other risk information. Farmer perceptions were consistent with the estimates of biological risk for the different areas and farm types. A useful extension to the method used would be to discuss with farmers how their perceptions compare to those obtained through disease surveys.
The PLA approach provided additional useful insights that will assist with shaping ECF control strategies. One example: for certain farm types and in certain areas, farmers lack information on ECF and ECF control. A second example: farmer adoption of other dairy technologies, particularly exotic cattle, greatly influenced their perceptions of ECF risk and the importance of controlling ECF.
