Extension services as key determining factor for adoption of minimum tillage practice in Kenya: A plot level analysis
Abstract
There is a consensus that conservation agricultural practices increase crop productivity and save resources. However, even after a decade of active promotion of conservation agriculture in sub-Saharan Africa, the adoption of such practices as minimum tillage is low. This article finds out the determining factors that influence the adoption decision of smallholder farmers of Kenya about minimum tillage practice. The study is based on survey data from a randomly collected sample of 300 households in the maize-dominant farming system of eastern Kenya. A plot-level analysis consisting of 437 plots is carried out using the Probit regression to estimate the adoption decision model. Access to extension information is found to be a key determinant of adoption decisions for minimum tillage. Household labor availability, land allocated for maize crop, and plot-level characteristics such as soil type, plot size, and intercropping practice followed on the plot are also found to have influenced farmers' decision to adopt minimum tillage practice. Extension services being a vital element in the efforts to promote minimum tillage practice, more effective implementation of promotion programs need to be undertaken by the Government in Kenya in particular and sub-Saharan Africa in general.
1 INTRODUCTION
Conservation agriculture (CA) which is an integrated agricultural technological innovation has been widely promoted as a resource and environmentally conscious agricultural practice (Erenstein et al., 2012). Such agricultural practices hold significant relevance for the resource-poor countries such as Kenya where food security is threatened.CA employs three main principles in crop agriculture; continuous minimal mechanical disturbance of the soils, use of permanent organic soil cover, and diversified crop rotation or intercropping (FAO, 2008; Hobbs, 2007). Continuous mechanical tillage (hoe or mouldboard) has been observed to reduce soil organic matter causing weakening of soil structure and making the soils prone to; soil erosion, loss of fertility, and environmental degradation (FAO, 2008; Richardson & King, 1995). On the other hand, minimum disturbance of the soil (Minimum tillage) and mulching have been shown to improve the soil structure while at the same time reducing surface runoff hence conserving the soil (Arshad et al., 1999; Richardson & King, 1995). The current study is based on the farmers' perception and adoption decision of the minimum tillage aspect of CA in Kenya. Minimum tillage practice which results in improved soil fertility leads to improvement in crop yields; especially cereals. At the same time, it also cuts down the costs associated with labor mainly for land preparation and weeding thereby increasing farmers' overall net income (Ahmad et al., 2014; Erenstein et al., 2008; Gathala et al., 2014; Jaleta et al., 2016; Lahmar, 2010). An IFAD/FAO joint study has found out that minimum tillage practice can save labor hours as much as 75% (in the hand labor/hoe system). In the draught animal powers category, the minimum tillage system (with knife roller and direct seed drill) can reduce labor requirement by 80%. Ngoma (2018) shows that adopting minimum tillage in Zambia is associated with an average yield gain for several crops such as maize, groundnut, sunflower, soybean, and cotton. Deines et al. (2019) applied a machine-learning causal inference approach to satellite-derived datasets of tillage practices and crop yields spanning the US Corn Belt from 2005 to 2017. They found an average of 3.3% and 0.74% yield increase for maize and soybeans, respectively, for fields with long-term conservation tillage.
Despite the well-documented benefits of minimum tillage, adoption rates in the developing countries have been low. A major reason for the unenthusiastic adoption of minimum tillage can be linked to the concern over the practical effect of it for smallholder farmers in the context of diverse, small-holder farms and farming systems of resource-poor countries in sub-Saharan Africa (Giller et al., 2009). Several studies have examined the adoption rate of minimum tillage and the determinants of adoption. Derpsch and Friedrich (2009) reported that over 80% of cultivated world land under zero tillage was found to be in both North and south America, compared to 2.3% in Asia, 1.1% in Europe, and 0.3% in Africa. Ahmad et al. (2014) observed that despite the efforts by the Government to promote the adoption of minimum tillage in Pakistan, the response still was not impressive as only 13% of the farms were actively practicing the same. The same authors also established that farmers practicing resource-saving technologies were able to increase profits, though a third of the farmers reported a decrease in yields and recommended that further refinement on the technologies need to be done to ensure the farmers optimally benefit from it (Ahmad et al., 2014). Knowler and Bradshaw (2007) reviewed past studies on the factors influencing farmer decisions on the adoption or dis-adoption of minimum tillage and established that there were few if any universal variables that regularly explained the adoption. However, extension services, delayed onset of rainfall, oxen numbers owned, and wealth index are some of the key factors that have been found to influence the adoption of minimum tillage (Arsalan et al., 2014; Pannell et al., 2014; Jena & Majhi, 2021).
For the conservation agriculture practices like minimum tillage to be effective in Sub Saharan Africa, Giller et al. (2009) proposed that the identification of situations where these practices can offer major benefits to the farmers need to be researched. Given this backdrop, the current study is aimed at finding out the key factors that influence farmers' decision to adopt minimum tillage in Kenya. The major contribution of this study is that these key factors will guide the policymakers as to which regions and farmers to target for the promotion of minimum tillage adoption. This article is structured as follows. Section 2 outlines the sampling technique, survey site, and survey instrument. It also discusses the method of estimation. Section 3 presents the results of the study and discusses their implication. Finally, the last section concludes with some policy recommendations.
2 MATERIAL AND METHODS
2.1 Sampling
Kenya is divided into 47 counties from which Embu county has been selected for this study. Embu in the south of the eastern province is a county with high agricultural potential where over half the population is classified as poor. Maize is the major crop in the county as it is in the whole country. Other crops such as coffee, beans, tea, and vegetables are also grown in the county. The high-altitude regions are cropped with coffee and tea. We, therefore, restricted the survey to the major maize growing regions of the county. A random selection of 16 villages was done from all villages from the maize growing regions of the county. Factoring in the population size of the villages sample size was chosen proportionally from each village, the sampling rate is 20%. Finally, an exhaustive list of total household names from each selected village was collected from which 300 households are randomly selected. Further, we also had the replacements for each village which was randomly chosen after the main sample. The questionnaire was precise yet detailed regarding the crop plantation, soil conservation, crop rotation, crop residue retention, minimum tillage, marketing, market access, extension service access, credit access, condition of village roads, shocks and risks, household consumption, food security, and other household characteristics. The survey took place between June and July 2013.
From the sample, about 80% of the sample households were male-headed while 20% were female-headed. Household heads had an average age of 55.6 and 7.3 years of schooling. On average, each household had about 4 family members and owned a total of 1 ha of farmland. From the total sampled households, the average number of maize plots was 1.4 per household with an average area of 0.26 ha.
2.2 Conceptual framework
Since rainfall in Kenya is bimodal there are two cropping seasons and the majority of the farmers plant maize in both seasons. Crop rotation is rare given the high preference for maize in the country. Data on all the parameters have been collected for both seasons and the analysis is done on both seasons. This can show whether the farmers' behavior and the outcome are consistent for both the seasons or they are different. It is also necessary to treat the seasons differently since rainfall, temperature, seed and other input uses are different between the seasons. The analysis done in this article is at the plot level since each plot is different in terms of soil type, plot slope, and soil fertility. These plot-level characteristics are included in the analysis.
The probit model is used in the study to model the adoption decision of the respondents. The decision to adopt minimum tillage in a particular plot depends on the farmer's perception regarding the suitability of the practice in the plot as well as the profitability prospects. Whether to adopt minimum tillage or not in a plot is a binary decision variable and the Probit model estimates the probability of this event conditional on certain explanatory variables. These explanatory variables are divided into three categories—household characteristics, farm characteristics, plot characteristics, and institutional factors. The household characteristics are age, education, and gender of the household head, household size, available household labor, and gender of the plot manager. The last variable is important since gender is touted as an influential factor that determines the input use and farming practice in a plot. Hence, the questionnaire asked the gender of the household member managing the plot.
The farm characteristics are the total size of the farmland owned by the household and area under maize production in the particular season. The plot characteristics are the size of the plot, plot ownership, distance of the plot from the homestead, soil fertility, slope of the plot, and soil type. The institutional factors are distance to the main market, distance to the agricultural extension office, information received on maize variety, and minimum tillage practice.
3 RESULTS
3.1 Comparative analysis between adopters and non-adopters of minimum tillage
The mean and standard deviation of household characteristics, assets, and distance to key input markets are compared between adopters and non-adopters of minimum tillage and presented in Table 1. There seems to be a significant difference between the levels of education of households adopting minimum tillage and of those without it. The adopters have a comparatively higher education level than the non-adopters. This implies that higher educated household heads chose to adopt new technologies by a readiness to take the risk that is associated with it. Regarding land ownership, there is a statistical difference between the two groups with the minimum tillage adopters possessing higher land size.
| Minimum tillage adopters (n = 58) | Non adopters (n = 242) | ||||
|---|---|---|---|---|---|
| Variable | Mean | SD | Mean | SD | p-value |
| Age | 55.8 | 17.4 | 55.3 | 15.6 | 0.42 |
| Education level of the household head | 8.2 | 4.7 | 7.2 | 44.4 | 0.07a |
| Family size | 4 | 1.7 | 4.3 | 1.8 | 0.18 |
| Size of land owned | 1.14 | 0.58 | 0.98 | 0.57 | 0.04b |
| Distance to the main market (walking minutes) | 91.5 | 70.1 | 87.33 | 63.7 | 0.34 |
| Distance to the village market (walking minutes) | 23 | 16 | 24.5 | 16.5 | 0.28 |
| Distance to the agricultural extension office (walking minutes) | 74 | 67 | 65.8 | 52.6 | 0.18 |
| Total oxen numbers owned | 1.32 | 1.47 | 1.3 | 2.1 | 0.28 |
| Number of household members working on the farm | 2.86 | 1.47 | 3.33 | 1.54 | 0.03b |
- Note: n denotes the number of households.
- a Significant at 10%.
- b Significant at 5%.
Despite government initiatives to supply all types of agricultural machinery to the farmers, the households themselves have to go to the agricultural office or the market to purchase it. In such cases, distance to market plays a major role in economic development as nearer distance consumes less time. However, no statistically significant difference is found between the two groups in terms of all the distance variables such as distance to the main market, distance to the village market, and distance to the agricultural extension office. It is further observed that the number of household members working on the farm for minimum tillage adopters is comparatively lower than the non-adopters.
The comparison of plot characteristics between the conventional tillage and the minimum tillage plots are presented in Table 2. Out of the total 437 plots, 14% of the plots have practiced minimum tillage, a finding supported by other studies that showed a similarly low level of adoption in sub-Saharan Africa (Jaleta et al., 2016; Jena, 2019). Crop stressors are comparatively higher for the minimum tillage adopted plots (52.56%) than the conventional tillage plots (37.9%). Similarly, the practice of crop residue retention is observed to be higher for the minimum tillage plots (55.7%) compared to the conventional tillage plots (32.9%).
| Variable | Description | Conventional tillage plots (n = 376) | Minimum tillage plots (n = 61) | ||
|---|---|---|---|---|---|
| Freq | Percentage | Freq | Percentage | ||
| Experienced crop stressors | Yes = 1 | 133 | 37.27 | 33 | 52.56 |
| Crop residue retention | Yes = 1 | 124 | 32.9 | 27 | 55.7 |
| Slope of the land | Gentle slope (yes = 1) | 195 | 59.2 | 28 | 45 |
| Medium slope (yes = 1) | 148 | 39.6 | 27 | 44 | |
| Steep slope (yes = 1) | 30 | 8 | 6 | 9 | |
| Soil depth | Deep (yes = 1) | 202 | 54 | 23 | 37 |
| Medium (yes = 1) | 140 | 37 | 31 | 50 | |
| Shallow (yes = 1) | 31 | 8 | 7 | 11 | |
| Soil fertility | Good (yes = 1) | 110 | 29.4 | 20 | 32.7 |
| Medium (yes = 1) | 217 | 58.1 | 34 | 55.7 | |
| Poor (yes = 1) | 46 | 12.3 | 7 | 11.4 | |
| Soil type | Black (yes = 1) | 21 | 5 | 3 | 4.9 |
| Brown (yes = 1) | 243 | 65 | 36 | 59 | |
| Red (yes = 1) | 85 | 22 | 16 | 26 | |
| Gray (yes = 1) | 24 | 6.4 | 6 | 9.8 | |
- Note: n denotes the number of plots.
3.2 Adoption of minimum tillage
The following section discusses the Probit regression results to find out the determinants of adoption of minimum tillage. The results show that available household labor, the area under maize, size of the plot, whether or not the plot is intercropped, whether or not the maize fields had a crop stressor (pest attack), information regarding hybrid variety and soil characteristics influence an average farmer's decision to adopt minimum tillage in a particular plot.
The larger the area under maize cultivation, the higher the chances that the household may adopt minimum tillage. This can be explained by the fact that those farmers who allocate a higher proportion of land to maize would like to experiment with minimum tillage in some of their plots while practicing conventional tillage in the rest. While small farmers having less land left for maize would probably not run the risk of doing minimum tillage rather would till the land conventionally that they are used to. These findings support those of Pannell et al. (2014) who observed that larger and more fertile farms had better benefits from CA compared to smaller and less fertile farms which may motivate the farmers to adopt minimum tillage under these conditions. On the other hand, the size of the plot negatively determines minimum tillage adoption which might be because the larger plot can be easily tilled by mechanical tillage methods and hence farmers prefer to undertake conventional tillage while the smaller plot is put under minimum tillage. Furthermore, it appears that the farmers understood the role of conservation agriculture in controlling crop pests and crop disease. This is because those farmers that had encountered any crop disease or crop pest attack are more likely to have adopted minimum tillage. Farmers that received extension services, particularly information regarding hybrid varieties of maize are more likely to adopt minimum tillage in their plots. Labor availability with the household appears to have negatively affecting adoption choice. From the plot level soil characteristics variables, having brown soil has the advantage of adopting minimum tillage compared to gray soil.
However, the age and gender of the household head did not affect the decision to adopt minimum tillage. Furthermore, the household size, distance to both the village and the main market, distance to the agricultural extension office, the total number of oxen owned appeared to have no effect on the adoption of zero tillage. These findings are similar to those of Knowler and Bradshaw (2007) who reviewed past studies on CA and established that there were few; if any universal variables that regularly explained the adoption of conservation agriculture (Table 3).
| Variable | Coefficient | Std error |
|---|---|---|
| Age of the household head (years) | −0.001 | 0.005 |
| Household member size | 0.02 | 0.07 |
| Education of household head | 0.002 | 0.02 |
| Total farm size owned | 0.06 | 0.05 |
| Plot size (ha) | −273.1 | 120.26a |
| Plot distance | −0.003 | 0.003 |
| Plot ownership | 0.25 | 0.2 |
| Intercropping done (1 = yes) | −0.4 | 0.17a |
| Maize crop stressor (1 = yes) | 0.33 | 0.17a |
| Area under maize in ha | 674.8 | 296.95a |
| Distance to the main market (walking minutes) | 0.001 | 0.001 |
| Available household labor | −0.16 | 0.09a |
| Gender of the plot manager (1 = male) | −0.16 | 0.19 |
| Good soil (1 = yes) | −0.24 | 0.28 |
| Medium soils (1 = yes) | −0.22 | 0.23 |
| Gentle slope (1 = yes) | −0.43 | 0.28 |
| Medium slope (1 = yes) | −0.26 | 0.27 |
| Black (1 = yes) | −0.66 | 0.42 |
| Brown (1 = yes) | −0.62 | 0.26a |
| Red (1 = yes) | −0.36 | 0.29 |
| Access to extension information (1 = yes) | 0.58 | 0.23a |
| Distance to agricultural extension office | 0.0005 | 0.001 |
| Hybrid maize variety (1 = yes) | −0.17 | 0.3 |
| Information on minimum tillage (1 = yes) | −0.07 | 0.17 |
| Wealth | −0.05 | 0.03 |
- Note: Number of observations 437. Wald chi2 263.35. Prob > chi2 0.000.
- a Significant at 1%.
4 CONCLUSION AND POLICY IMPLICATION
The current study probes into the decision-making of smallholder farmers of Kenya on the matter of adoption of minimum tillage which is one of the key components of conservation agriculture practices. A random sample of 300 households has been taken from the Eastern Kenyan district of Embu to carry out the household survey. Adoption of minimum tillage is found to be rather low in the sample with only 14%. It shows that farmers in sub-Saharan Africa are quite apprehensive of the suitability and profitability of minimum tillage practice even after a decade of active promotion of such practices. Using Probit regression and taking each plot as a unit of analysis, the determinants of adoption of minimum tillage are found to be—access to extension service, average area under maize, whether the maize fields had a stressor or not, available household labor, soil characteristics, whether the plot is intercropped or not and size of the plot. The factors that have positively swung the farmer household towards adopting minimum tillage are extension information, the area under maize, experience with pest attack and crop disease, and soil characteristics. Whereas, availability of household labor, plot size, and intercropping are found to have negatively associated with the adoption decision.
Although, plot characteristics and household characteristics influence the adoption decision of new agricultural practices the most important factor is the institutional aspects like information regarding particular maize variety which will succeed under the minimum tillage practice. Farmers are understandably apprehensive about these practices because these are not universally proven profitable practices yet and require support in the form of seed, irrigation, and other input provisions. More studies are required to understand the right environment for practices such as minimum tillage to succeed in sub-Saharan Africa.
Biography
Dr. Pradyot Ranjan Jena currently works as Associate Professor at National Institute of Technology Karnataka. Dr. Jena's core research area of expertise are impact evaluation, environmental impact valuation and climate change. He has worked across continents and have experience about rural livelihoods in many developing countries. He won several international awards, published over 50 research papers, and served as editor to journals.
Open Research
DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding author upon reasonable request.
