Water use and nutritional water productivity of selected major and underutilised grain legumes.

Doctor of Philosophy in Crop Science. University of KwaZulu-Natal, Pietermaritzburg 2018.Grain legumes have potential to alleviate the prevalence of food and nutrition security in water scarce areas. There is need to promote underutilised grain legumes to diversify crop production and build resilience. This requires knowledge on their water use (ET), environmental adaptation and nutritional content (NC) in comparison to major legumes. The study benchmarked underutilised grain legumes [bambara groundnut (Vigna subterranea) and cowpea (Vigna unguiculata)] to major grain legumes [groundnut (Arachis hypogaea) and dry bean (Phaseolus vulgaris)] with respect to ET, water productivity (WP), NC and nutritional water productivity (NWP). Field experiments were conducted during the 2015/16 and 2016/17 summer seasons under varying water regimes [optimum irrigation (OI), deficit irrigation (DI) and rainfed (RF)] and environmental conditions (Ukulinga, Fountainhill and Umbumbulu) in KwaZulu-Natal, South Africa. Data collected included stomatal conductance, leaf area index, timing of key phenological stages and yield. Water use was calculated as a residual of the soil water balance. Water productivity was calculated as the quotient of grain yield and ET. Grain was analysed for protein, fat, Ca, Fe and Zn. Yield, ET and NC were used to compute NWP. Results from the field trials were used calibrate and test the performance of AquaCrop model for groundnut and dry bean. Under varying water regimes, crops adapted to limited soil water through stomatal regulation and reduction in canopy size and duration. Yield, yield components and WP varied significantly (P < 0.05) among crop species. During 2015/16, groundnut had the highest yield and WP under DI (10 540 kg ha-1 and 0.99 kg m-3, respectively). During 2016/17, the highest yield and WP were observed in dry bean under DI (2 911 kg ha-1 and 0.75 kg m-3, respectively). For both seasons, dry bean had the lowest ET across all water treatments (143 – 268 mm). Dry bean and groundnut out–performed bambara groundnut with respect to yield, harvest index and WP. Yield varied significantly (P < 0.05) across environments and seasons. Cowpea was the most stable species. Results of NWP were significant among crops (P < 0.05). Yield instability caused fluctuations in NWP. Groundnut had the highest NWPfat (46 – 406 g m-3). Groundnut and dry bean had the highest NWPprotein (29 – 314 g m-3). For NWPFe, Zn and Ca, dry bean and cowpea were more productive. Overall, the AquaCrop model was successfully calibrated for groundnut and dry beans. Model testing showed AquaCrop’s potential for simulating growth, yield and ET of groundnut and dry bean under semi-arid conditions. Underutilised grain legumes need to undergo crop improvement for successful promotion. There is need to improve adaptation of grain legumes to different environments and resilience to extreme weather events. Future studies should consider benchmarking more underutilised grain legumes to major grain legumes