Predicting Maize (Zea Mays) Yields in Eastern Province of Rwanda Using Aquacrop Model

Rwanda is affected by prolonged droughts leading to reduction in crop production and livestock production with severe food insecurity. Rain fed agriculture in Rwanda is affected by drought and variation in seasonal rainfall amounts within the last decades and climate change impacts have been reported in many areas of the country. Eastern Province is considered one of the most productive rain-fed agricultural areas in Rwanda, where it contributes more than 32 % of the cereals produced in the country. Therefore, the general objective of this work was to predict maize yields under reined agriculture by using AquaCrop model in Eastern province of Rwanda. Trend analysis was carried out on climatic parameters (1981-2016) such as the rainfall, maximum and minimum temperatures, evapotranspiration and maize yield (2002-2016). Results showed that rainfall trend was not significant over the study area, indicating non significant change in rainfall during the last decade. Minimum and maximum temperatures had significant trends (increasing) over the study area, implying that the temperature has been rising over the last decade. Maximum and minimum temperature showed a negative relationship with negative correlation coefficients of -0.38 and -0.39 with maize yield respectively. This implies that an increase in temperature beyond the optimum level (low temperature below 8°C and high temperature of above 30°C) results in a decline in maize yield and vice versa. On the other hand Correlation analysis showed a positive relationship between maize yield and rainfall (0.57 and 0.59) for both Districts (Bugesera and Nyagatare) and maize and evapotranspiration (0.29 and 0.27). This means that an increase in rainfall enhances significantly or not maize production. Monthly results of coefficient of variation values indicated an increase in climate variability, which was shown by larger season to season fluctuations, with a higher coefficient of variation implying less predictability in the climate parameters. It was observed that 25.4% and 50.7% of the variability in maize yield could be explained by these climatic parameters (maximum and minimum temperature, evapotranspiration and rainfall) and was significant at p=0.024 and 0.0103. The impacts of changes in climate parameters; temperature and annual rainfall, was determined by using comparisons between the observed data (past and present) and projected data (CNRM vi Cordex Model).Climate variability results analysis revealed significant increase in minimum temperature of 9.88% and 13.09% (1.14 °C and 1.49°C ) for Bugesera and Nyagatare by 2046, while rainfall was projected to decrease by 20.65% and 20.97% (163.59 and 174.65 mm) for Bugesera and Nyagatare respectively over Eastern province area during the same period. The outputs of the study using AquaCrop model showed that by 2050, the study area’s seasonal rainfall (September-January) will decline by 233.6 mm (10.9%), The average predicted future simulated maize yield for September-January season (2021-2050) were 1282.3kg/ha and 1316.8 kg/ha over Bugesera and Nyagatare areas respectively. Comparison with the observed maize yields for September-January season (2002 up to 2016) of 1675.5 kg/ha and 1760.9 kg/ha indicated there will be a decrease in future maize yields (23.4%) in Bugesera and (25.2%) in Nyagatare area. Seasonal temperature have significant impacts in maize yields, There were projected increases in minimum and maximum temperature by 1.20°C (4.55% and 4.54%) and 1.14°C (9.88% and 13.09%) respectively for Bugesera and Nyagatare Districts. This implies that maize yield production will be very sensitive to reduction in rainfall and increase of temperature during the season over the study area. The extreme impacts of climatic parameters occurred over the last years was depicted and its associated effects on maize yields was done using correlation and regression techniques. Man Kendall Trend analysis and coefficient of variation estimates statistics was computed in order to quantify the magnitude of climate variability and change in climatic parameters and its direction