Modelling Green Gram Production in Kenya Under the Current and Future Climates

Abstract

Green gram is one of the legumes considered suitable for cultivation in the Arid and Semi-Arid Lands (ASALs) of Kenya. However, the area that is currently suitable remains small due to inadequate knowledge on the variation of climatic elements in space and time in the ASALs. The changing climate may have an effect on the areas presently suitable for green gram production. This study purposed to model the suitability of green grams in Kenya under the current and projected future climates. The CORDEX RCA4 models' ability to simulate the observed rainfall and temperature from Climate Research Unit (CRU) datasets were assessed using statistical measures of bias and normalised root mean square error (NRMSE). The bias in rainfall was reduced by using an ensemble of the models adjusted using the scaling method. The temporal analysis of temperature and rainfall were assessed using the Mann Kendall test to determine whether there was an increasing or decreasing trend in the datasets. Mapping for different levels of green gram suitability in Kenya was done through the use of a weighted overlay of climate, soil, and topography parameters. The APSIM model was calibrated for four varieties of green gram, namely Biashara, Tosha, N26, and KS20 varieties to evaluate the impact of climate change on green gram yield, biomass and days to maturity in a highly suitable region. Although the CORDEX models and their ensemble did not replicate the spatial and temporal variability of rainfall during the MAM and OND season very well, the models and their ensemble captured the temperature pattern well. The rainfall ensemble, despite performing better than the individual CORDEX models, still showed notable biases, necessitating bias adjustment before further use in green gram crop modelling. The bias-corrected ensemble of rainfall and the ensemble of temperature were then used to study the space and time variability of rainfall under baseline (1971 to 2000) and future RCP 4.5 and RCP 8.5 scenarios (2021 to 2050) and their effect on green gram production. The temporal trend of rainfall has been increasing at most stations under the baseline scenario and the trend is projected to continue under the RCP 4.5 and 8.5 scenarios for the MAM and OND seasons with statistical significance for some stations at a P-Value of 0.05. The temporal trend of maximum temperature during the MAM season has been increasing and statistical significance is noted at most stations under the baseline, RCP 4.5, and 8.5 scenarios at a P-Value of 0.05. The temporal trend of minimum temperature shows that minimum temperature has been increasing at all stations under the baseline, RCP 4.5 and 8.5 scenarios

for the MAM and OND seasons with statistical significance at most stations at a P-Value of 0.05. The increase in temperature is attributed to global warming due to a rise in the level of greenhouse gases. Most of Kenya was found moderately suitable for green gram production during the MAM and OND seasons under the baseline, RCP 4.5 and RCP 8.5 scenarios. During the MAM season, the area currently highly suitable for green gram production (67842.6 km2) is projected to increase slightly to 68600.4 km2 (1.1%) under the RCP 4.5 scenario and reduce to 61307.8 km2 (-9.6%) under the RCP 8.5 scenario. This decrease could be attributed to unfavourable temperature and rainfall above the threshold suitable for green gram production. During the OND season the area currently highly suitable (49633.4 km2) will increase under both RCP 4.5 (22.2%) and RCP 8.5 (58.5%) scenarios. This increase is attributed to good rainfall and temperature conditions in the future which are favourable for green gram production. The calibrated green gram model captured the observed yield, biomass and days to maturity of the four varieties of green gram shown by a coefficient of determination (CoD) which ranged between 87.0% and 99.0%; bias values which ranged between 1.3 and 25.3 and levels of NRMSE which ranged between 4.7% and 45.5%. During the MAM and OND seasons, a decline in yield, biomass, and days to maturity is expected under both the RCP 4.5 and RCP 8.5 scenarios. The increase in rainfall amount under both the RCP 4.5 and RCP 8.5 scenarios will translate to a lower yield and increased biomass. The increase in temperature will result in reduction of the days to maturity for green gram in Kitui County. The maps of green gram suitability indicate that the area suitable for green gram production will increase in the future. There is, however, a net decrease in yield of the four green gram varieties modelled. Kenya, currently, only produces 460kg/ha of green grams. The study found that despite the decrease in yield, potential production under the future climate scenarios was still above 460kg/ha. There is, thus, potential to expand on the current production of green grams. Therefore, despite the decrease in the future, green gram is still a lucrative crop since farmers still stand to increase their current production. Policymakers can refer to the developed green gram suitability maps under past and future climate scenarios, to determine how suitable their region will be for green gram production. Policymakers should also make use of the four green gram varieties developed under the APSIM model to mitigate against the possible impacts of climate change on green gram

yield. Given that the government aims to revive farming in the ASALs by promoting climate smart agriculture through planting drought resistance crops, there is need to develop green gram varieties which are more tolerant to the expected increase in rainfall and temperature to increase yield and in turn benefit farmers, the society and the country at large.