Water stress management and characterization
Target environments can also be mimicked if water is controlled by imposing a water regime by gravity or, better, by drip irrigation. Water stress management (timing, intensity, uniformity) and characterization (soil, plant measurements) are essential issues in drought phenotyping.
Moisture availability can itself be a complicating factor when comparing genotypes in field experiments. Although plots growing the different genotypes may receive the same quantity of water, the genotypes can vary in their water use and/or access to underground water, thereby confounding measurements associated with plant water relations. Study of water profiles (either experimentally or by using simulation models) can provide very useful information. Trait evaluation should preferably be carried out under field conditions, avoiding experimental situations (growth chambers, greenhouses, pots) that differ significantly from the agricultural growing environment. The ability to access water deep in the soil profile, which is an important drought-adaptive mechanism, is eliminated as a variable in pot conditions. Furthermore, the relative humidity of the air, which has an important influence on stomatal conductance (Ben Haj Salah and Tardieu, 1997), is extremely difficult to simulate in controlled environments.
When possible, drought tolerance evaluation should be done out-of-season, under irrigated conditions. This option allows better management of water stress but needs a dry season sufficiently long to cover the whole growth cycle. The photoperiod and temperature should not differ too much from the growing season, as is the case in the dry tropics, to avoid genotype-by-season interactions and allow results obtained from the out-of-season experiments to be extrapolated to the growing season conditions.