| dc.description.abstract | Reduced agricultural production and food security continues to be pressing problems for the larger part of smallholder farmers in Sub-Saharan Africa (SSA), as well as semi-arid Eastern Kenya. The main objective of this study was to assess the potential impacts of climate change on sorghum production in the semi-arid environment of Katumani in Machakos County, Kenya. The study sought to determine the effect of current and future climate and agronomic practices on the growth, development and yield of two sorghum cultivars. It also evaluated different farm-level adaptation strategies that may improve the adaptive capacity of sorghum growing farmers in the face of climate change. The study was conducted in National Dryland Farming Research Centre at Katumani, in eastern parts of Kenya.
The variability and trends of climate parameters and sorghum yields were determined using coefficient of variation, linear regression and Mann-Kendall test. The findings show that in the study area, the MAM season is characterized by high variations in seasonal rainfall compared to the OND season, but with low variability in both seasonal and annual temperatures. Additionally, higher variability (45%) in sorghum yield was noted. The results also showed insignificant trends for the intra seasonal rainfall characteristics. However, significant decreasing trends of rainfall at seasonal and annual scales were recorded at Iveti and Machakos DO station. Positive significant trends for maximum and minimum temperatures were noted at annual scale. Sorghum trend depicted positive insignificant trends.
Field experiments were conducted to determine the effect of different cultivars, nitrogen and phosphorus levels on the growth, development and yield of sorghum. The experiments were laid out in a randomized complete block design with two replications over four seasons. The analysis of variance showed non-significant effect of nitrogen (N) and (P) on phenological parameters of sorghum, growth parameters (biomass) and yield. However, the parameters were influenced by cultivar effect. The interactive effect of N, P and cultivar in the four seasonal experiments were non-significant for all the parameters.
The Agricultural Production Systems sIMulator (APSIM) model was calibrated and evaluated for growth, development and yield response of sorghum to inorganic nitrogen fertilizer. Modified index of agreement, coefficient of determination and root mean square error were used to
determine the level of agreement between the observed values and model simulated values. Model evaluation results pointed out that APSIM has a propensity to slightly over estimate the phenology, growth and grain yield. However, the error in all the parameters was below 15%, which is judged acceptable.
Model performance of the CORDEX GCMs was assessed using modified index of agreement and root mean square error. Three models (CanESM2, GFDL-ESM2M, NorESM1-M) and the ensemble performed relatively well in simulating the observed climate variables over the study region. Thus, these models were applied in the modeling of impact of climate change on sorghum production.
The simulated impacts of climate change on sorghum production in Machakos County were assessed using percentage mean changes and coefficient of variability. Both RCP 4.5 and 8.5 show an increase in temperature and rainfall compared to the 1976-2005 baseline period. Projected mean changes on phenological parameters showed a consistent decline for both sorghum varieties during both the long and short growing season. The end century (2070-2099) showed the greatest shortening in the phenological dates. There was also slightly greater variability in the phenological dates under current climate than under climate change during the MAM season. There was slight increase in biomass for both varieties under climate change with the application of fertilizer. It has been noted that under changing climate sorghum grain yields will consistently increase for both cultivars over different time periods with up to 85.3% increase towards the end of the century (2070-2099). Yields of both cultivars are more variable under current climate than under climate change and thus they can be predicted with more confidence in the future.
APSIM model was able to predict the response of sorghum to climate change and, from the results attained, changing the plant density (increase to 88,888 and decrease to 53,333 plants ha-1) as an improved agronomic practice was proposed. Results indicate that phenological dates were increased with increase in plant population and vice versa. These dates were less variable (less than 20%) for both plant populations under both the base period and changed climate. Biomass weight for both cultivars was found to increase and decrease for the highest and lowest plant population, respectively. Higher and lower grain yield were recorded from highest and
lowest plant populations, respectively during the OND, but for MAM growing season, the reverse happened. The variations of biomass and grain yield increased with increase in plant population and vice versa | en_US |
