Effects of Rainfall Variability and Selected Non-Climatic Factors on Wheat Yields and Farmers’ Adaptation Strategies in Narok County, Kenya
John Maina, Kimamo
MetadataShow full item record
In recent decades, variability in climate has been experienced globally. The rainfall patterns in Kenya are becoming unpredictable just like other regions in tropical Africa. Changes in rainfall patterns occur in the onset of long rains and low average amount throughout the year. Droughts in Narok County usually lead to wheat shortages and fluctuations in agricultural production. The current study specifically examined rainfall variation patterns during the growing season (March, April and May) and effects of this variability on wheat yields in Narok County. The study was guided by four objectives: To determine variability of seasonal rainfall characteristics from year 1981 to 2016; to establish effects of these rainfall characteristics on wheat yields; determine non-climatic variables affecting wheat yields and to analyze factors affecting adaptation strategies by wheat farmers. This was done by collecting a 36-year rainfall data and wheat yields per hectare in tones from Kenya Meteorological Department and Narok County government respectively. A purposive sampling design was used to select wheat farming households in four sub-counties where wheat is grown. Proportionate allocation was used to determine the sample of farmers to be picked in each Sub-County. Random sampling technique was used to pick wheat farmers in this sample. A structured questionnaire was administered to the sampled population to collect data on adaptive strategies employed by the farmers and non-climatic variables in the region. Seventy six percent (76.0%) of the administered questionnaires were fully filled and returned. Variability of rainfall characteristics was analyzed using variability indices. Pearson correlation analyses followed by linear regression were used to determine the degree to which rainfall characteristic predicts wheat yields. The coefficient of multiple determinations for correlation (R2) was used to explain the percentage of wheat yields explained by rainfall characteristics. The individual R2 for number of rainy days, amount, cessation and onset was 0.215, 0.205, 0.029 and 0.016 respectively. This showed that the most influential rainfall characteristics on wheat yields are amount and number of rainy days. The results further showed that, while rainfall amount and number of rainy days were significant, deviation on onset and cessation dates had no significant effects on wheat yields (P > 0.05). Stepwise regression (Y = 0.727 + 0.464X1) showed that rainfall amount had the highest significant relationship with wheat yields. Pearson correlation analysis and measures of central tendency were used to assess how the non-climatic variables influence the yields. Results showed that non-climatic variables did not have a significant effect on wheat yields where the R2 for number of agricultural officers, wheat price and cost of CAN, DAP and NPK was 0.132, 0.007, 0.116, 0.001 and 0.002 respectively. A SWOT analysis was used to explain adaptation strategies by farmers. The study findings will to benefit policy makers in the field of Agriculture, academicians and other stakeholders’ sector by making decisions that will improve adaptation measures by wheat farmers.