Late mid-summer dry-spells and planting dates of maize in the western maize-growing region of South Africa.

Abstract

South Africa experiences a climate characterised by arid conditions and significant variability in annual rainfall. This exerts a severe impact on water resources, agriculture, and the socio-economic sector of the region. Drought is a natural occurrence known for sustained below-average precipitation over a specified period, and it is a well-documented aspect of a region's climate variability. Existing research has predominantly focused on drought occurrence, impact, intensity, and duration, primarily on seasonal and annual scales. This study deviates by emphasising the importance of recognising mid-summer droughts in the agricultural sector, where the timing and severity of dry-spell periods hold greater significance than overall seasonal rainfall deficits. Mid-summer droughts, occurring typically from December to January in the summer rainfall regions, coincide with a shift in atmospheric circulation from baroclinic to barotropic. This shift is dominated by tropical systems and is particularly critical as the timing aligns with the water stress-sensitive flowering stage of summer crops like maize and sorghum. Given the potential detrimental effect on crop yield, especially maize, there is a need to investigate and understand the patterns associated with these events. By delving into the specifics of mid-summer droughts, this research aimed to contribute valuable insights to aid and mitigate the adverse impacts on agricultural productivity in South Africa. However, in this study, the late-midsummer period, which spans from mid-January to the end of February is of particular importance. This is because maize is planted later in the western region of maize production than it is in the temperate and cooler eastern maize production regions. Therefore, dry-spells, which occur around the mid-January to the end of February period, normally correspond with maize's flowering cycle in the western maize growing regions of South Africa. However, it is worth noting that due to shifting planting dates, the flowering stage may occur later and remain unaffected by the phenomenon. During this sensitive stage of the growth cycle, even a few days without rain could result in lower crop production. The Markov chain model analysed the probability of initial and conditional dry-spell pentads. In addition, the Mann-Kendall monotonic trend test and Sen's slope estimator were performed to check the direction and magnitude of dry-spell trends. It consistently revealed the impact of dry-spells throughout the mid-January to late February period, the most critical phase in the flowering of maize for the western maize-producing region. Although most districts showed a downward trend in the occurrences of dry-spells, this was not statistically significant. Sen's slope estimator further supported a reduction in the magnitude of dry spells over the study period. This research thus provides important information to farmers by enhancing their understanding of shifting risk profiles associated with dryspells. Furthermore, utilising AquaCrop simulation simulations, the study explored the influence of planting dates on maize yield in the western maize growing region. The latest simulated planting date of 20 December was found to result in the best maize yields across most Rainfall Districts. However, certain districts with low dry-spell frequencies experienced lower average yields compared to mediumfrequency seasons, potentially linked to heavy rainfall distribution during the flowering stage of maize. The analysis of the rainfall data showed that during some low frequency dry-spell seasons, extreme rainfall events occurred. These extreme rainfall events may have led to waterlogging issues, resulting in a reduction in crop yield below the typical maize production levels. In conclusion, the AquaCrop model provided reliable and realistic outputs, with crop yields aligning well with observed rainfall patterns. A two-way ANOVA test indicated that there is a statistical significance between maize yield and planting dates in most of the Rainfall Districts, with the planting date of 20 December simulating the best maize yields compared to the other planting dates in most of the Rainfall Districts. Thus, the period after the 20th of December has higher crop yield. This is because crops planted around this date will flower outside the late midsummer period, and as a result, they will not be affected by the dry spells. It is recommended that future research should focus on examining the impacts of dry spells on different maize varieties and assessing the effects of climate change on dry spell occurrence.