This is author version post-print archived in the official Institutional Repository of ICRISAT www.icrisat.orgTerminal drought tolerance implies that plants have enough water to fill grains. Water saving traits, measured in tolerant and sensitive cowpea lines, showed that tolerant lines have developed several constitutive mechanisms, closely related to one another, which reduces the rate of water use and delay drought effects. This opens the possibility to decipher their genetic basis towards the development of drought tolerant cowpea cultivars. conditions and restricted TR more than sensitive lines under high VPD. Under WS conditions, transpiration declined at lower FTSW in tolerant than in sensitive lines. Tolerant lines also maintained higher TR and CTD under severe stress than sensitive lines. TE was higher in tolerant than in sensitive genotypes under WS conditions. Significant and close relationships were found between TR and TE, CTD, and FTSW in both environments under different water regime conditions. In sum, traits that condition how genotypes manage limited water resources discriminated tolerant and sensitive lines. Our interpretation is that a lower canopy conductance limits plant growth and plant water use, and allows tolerant lines to behave like non-stressed plants until the soil is drier and maintains a higher transpiration under severe stress. A lower TR at high VPD leads to higher transpiration efficiency.

Drought stress is a major constraint on cowpea productivity, since the crop is grown under warm conditions on sandy soils having low water-holding capacity. For enhanced performance of crops facing terminal drought stress, like cowpea, water-saving strategies are crucial. In this work, the growth and transpiration rate (TR) of 40 cowpea genotypes with contrasting response to terminal drought were measured under well-watered conditions across different vapour pressure deficits (VPD) to investigate whether tolerant and sensitive genotypes differ in their control of leaf water loss. A method is presented to indirectly assess TR through canopy temperature (CT) and the index of canopy conductance (Ig). Overall, plants developed larger leaf area under low than under high VPD, and there was a consistent trend of lower plant biomass in tolerant genotypes. Substantial differences were recorded among genotypes in TR response to VPD, with tolerant genotypes having significantly lower TR than sensitive ones, especially at times with the highest VPD. Genotypes differed in TR response to increasing VPD, with some tolerant genotypes exhibiting a clear VPD breakpoint at about 2.25 kPa, above which there was very little increase in TR. In contrast, sensitive genotypes presented a linear increase in TR as VPD increased, and the same pattern was found in some tolerant lines, but with a smaller slope. CT, estimated with thermal imagery, correlated well with TR and Ig and could therefore be used as proxy for TR. These results indicate that control of water loss discriminated between tolerant and sensitive genotypes and may, therefore, be a reliable indicator of terminal drought stress tolerance. The water-saving characteristics of some genotypes are hypothesised to leave more soil water for pod filling, which is crucial for terminal drought adaptation.

Five cowpea genotypes, Gorom local (Go), KVX61‐1 (KV), Mouride (Mo), Bambey 21 (B21) and TN88‐63 (TN), differing in their susceptibility to water stress, were studied under glasshouse and field conditions, to determine their physiological, biochemical and agronomic responses to water deficit at flowering stage. Effect of water deficit on leaf water potential (ψl), canopy temperature, gaseous exchange, leaf proline content, total protein and starch contents, maximal quantum yield (ϕp0) and yield components was examined. Water deficit significantly increased the canopy temperature and the proline content of the five genotypes while ψl, gaseous exchanges, ϕp0 and starch content decreased significantly. Yield components, with the exception of seed number per pod, of the five genotypes, were also significantly affected. Under glasshouse and field conditions, the results showed that stomatal closure is the common strategy used by the five genotypes to avoid dehydration. Go, Mo and TN tolerated water stress better than B21 and KV. Furthermore, Go and Mo recovered more rapidly after rewatering than B21 and KV. These latter genotypes are revealed to be sensitive with low recovery capacity. The results suggest that the maintenance of net photosynthesis and solute accumulation seem to be traits conferring water stress tolerance in Go, Mo and TN. These traits and recovery capacity could be valuable selection criteria for higher yields under water deficit conditions.

Traditional agroforestry parkland systems in Burkina Faso are under threat due to human pressure and climate variability and change, requiring a better understanding for planning of adaptation. Field experiments were conducted in three climatic zones to assess Sorghum bicolor (L.) Moench (Sorghum) biomass, grain yield and harvest index in parklands under different rainfall pattern and compared to simulations of sorghum biomass and grain yield with the Water, Nutrient and Light Capture in Agroforestry Systems (WaNuLCAS) model for calibration and parametrisation. For planning adaptation, the model was then used to evaluate the effects of different management options under current and future climates on sorghum biomass and grain yield. Management options studied included tree densities, tree leaf pruning, mulching and changes in tree root patterns affecting hydraulic redistribution. The results revealed that sorghum biomass and grain yield was more negatively affected by Parkia biglobosa (Jacq.) Benth. (néré) compared to Vitellaria paradoxa C. F Gaertn (karité) and Adansonia digitata L. (baobab), the three main tree species of the agroforestry parkland system. Sorghum biomass and grain yield in different influence zones (sub-canopy, outside edge of canopy, open field) was affected by the amount of precipitation but also by tree canopy density, the latter depending itself on the ecological zone. The harvest index (grain as part of total biomass) was highest under the tree canopy and in the zone furthest from the tree, an effect that according to the model reflects relative absence of stress factors in the later part of the growing season. While simulating the effects of different management options under current and future climates still requires further empirical corroboration and model improvement, the options of tree canopy pruning to reduce shading while maintaining tree root functions probably is key to parkland adaptation to a changing climate.

To address tree-soil-crop interactions in the Sahel, we examined the growth-limiting factors (water, light and mineral nutrients) of Sorghum bicolor growing under trees in agroforestry parklands of Burkina Faso. Growth and yields of sorghum were measured after (1) pruning crowns of Vitellaria paradoxa and Parkia biglobosa trees, and (2) applying mineral fertilizers (nitrogen and/or phosphorus) and supplemental irrigation during normal wet cropping seasons in 2007 and 2008. Irrigation treatments led to non-significant 29% and 23% gains in grain and dry matter yields (from control values of 455 and 1,140 kg ha -1 ), respectively. The fertilizer showed variable but in general significant increases in grain and straw yields and more consistently in the height of sorghum plants. The crown pruning increased the values of these variables much more strongly, by 520% and 348% (from control, no-pruning values of 282 and 612 kg ha -1 ), respectively. The growth and production of S. bicolor were also[56% higher under V. paradoxa than under P. biglobosa. The same trends were observed in both cropping seasons, although rainfall was much heavier in 2008 than in 2007, and the mean sorghum grain yield was approximately twice as high in 2008. The results clearly indicate that competition for light limits sorghum growth more than competition for other resources in the studied system, suggesting that parkland management should aim at either increasing light availability (by reducing tree density or pruning) or growing shade-tolerant crops under the trees. However, use of a poorly soluble phosphorus source during the first year, modest amount of water applied through the supplemental irrigation (48 mm) and the wetness of the rainy season in 2008 (which led to abandonment of the irrigation treatments and floods in the experimental plots) may have masked possible effects of the applied fertilizers and irrigation. Therefore, more prolonged analyses of the effects of fertilizers and deficit irrigation are required before robust recommendations can be made to farmers.