Towards integrated soil fertiuty management: quantitative Analysis of tradeoffs between risk and returns for small scale Farmers in Kiambu district, Kenya
Food insecurity in Kenya has been attributed to a number of factors, among them, fragile ecosystems and a steady decline in the growth of the agricultural sector, population pressure and slow growth in other sectors of the economy. This has led to the gradual but persistent degradation of the natural resource base. Through efforts have been made to develop and disseminate soil nutrient replenishment technology packages, adoption still remains low even in the high potential Central highlands of Kenya. An example of such interventions is the multi-institutional, multi-disciplinary project by the Kenya Agricultural Research Institute, the Tropical Soil Biology and Fertility Institute, the University of Nairobi and the Ministry of Agriculture that sought to address the perennial problem of declining soil fertility in the Central highlands of Kenya. In a participatory process, the researchers in this multi-institutional project facilitated the farmers through an exercise where they classified themselves into 3 soil fertility management classes based on their management ability. Soil fertility class 1 are those who were deemed to have good SFM practices (high manure and fertilizer input, hybrid seed, terrace maintenance) while class 3 are characterized by low input (including low soil fertility replenishment) and output. To understand adoption based on this project activities and enrich future efforts in dissemination of soil fertility based options, this risk assessment study submitted that the risk attitude of the target farmers is a major factor influencing adoption. Further, the study submitted that farmers adopt technologies/techniques based on their available resources and past experience. Using the SFM classes, the main objective of this risk assessment study was to quantitatively measure the trade-offs between risks and returns for selected cropping systems amongst the 3 SFM farmers' classes in Githunguri division, Kiambu district of Central Kenya. The Target-MOTAD model was used to analyse for risk using farmers' input-output details. The model is a two-attribute risk and return linear programming model that entails a constraint on income deviations from a target level of income. The model results were used to develop risk frontiers for various income and variance combinations for each class. The coefficient of variation, derived from dividing the mean by the standard deviation, was also used as a measure of risk. Input, output and risk efficiency results were distinctly different for the three classes. The farm annual incomes at the time of data collection averaged KES.114,877, KES.35,656 and KES.18,153 respectively for the 3 SFM classes from respective average farm sizes of 1.04, 0.76 and 0.52ha. Out of the six priority enterprises selected, maize, maize/beans, potatoes, bananas, tomato and dairy/Napier production, only tomato and dairy enterprises entered the risk efficient plans. Based on the specified resource availability and obligations for the farmers interviewed, the maximum attainable income levels (at zero-risk) was KES.278,551, KES.79,036 and KES.47,618 respectively. Coefficient of variation (CV) showed higher risk levels as one moved from farmers practice, through various risk levels, to optimal results for classes 1 and 3. Deriving risk efficient frontiers using research trial results, the income as well as risk levels were much higher than farmers' current income and risk levels. Lowest risk levels were depicted by CV of 51%, 22.6% and 45% for the 3 classes. Only class 2 farmers activities had higher CV compared to 3 of the plans in the normative plan. Risk is important to the smallholder resource-poor farmers of the study area and their active participation in formulating research agendas is inevitable. The study concluded that in Integrated SFM research in Kiambu district, what is more urgently needed may not be blanket new technologies but a research focus to improve productivity and enterprise selection in face of the varied levels of resource availability. Whereas class 1 farmers may be able to take higher risks, most class 3 farmers might not be able to increase productivity sufficiently based on their current resource levels. Class 2 farmers, who form 50% of the area population, would be an effective target group for researcher knowledge in order to increase efficiency in resource use. In terms of enterprises, researchers should move towards including high value crops in their analysis, away from the traditional use of maize in experimental designs. This will help get better and affordable options for increasing income using sustainable soil fertility replenishment options.