Genetic engineering to improve plant performance under drought: physiological evaluation of achievements, limitations, and possibilities

Fully drought-resistant crop plants would be beneficial, but selection breeding has not produced them. Genetic modification of species by introduction of very many genes is claimed, predominantly, to have given drought resistance. This review analyses the physiological responses of genetically modified (GM) plants to water deficits, the mechanisms, and the consequences. The GM literature neglects physiology and is unspecific in definitions, which are considered here, together with methods of assessment and the type of drought resistance resulting. Experiments in soil with cessation of watering demonstrate drought resistance in GM plants as later stress development than in wild-type (WT) plants. This is caused by slower total water loss from the GM plants which have (or may havemorphology is often poorly defined) smaller total leaf area (LA) and/or decreased stomatal conductance (g(s)), associated with thicker laminae (denser mesophyll and smaller cells). Non-linear soil water characteristics result in extreme stress symptoms in WT before GM plants. Then, WT and GM plants are rewatered: faster and better recovery of GM plants is taken to show their greater drought resistance. Mechanisms targeted in genetic modification are then, incorrectly, considered responsible for the drought resistance. However, this is not valid as the initial conditions in WT and GM plants are not comparable. GM plants exhibit a form of odrought resistance' for which the term odelayed stress onset' is introduced. Claims that specific alterations to metabolism give drought resistance [for which the term oconstitutive metabolic dehydration tolerance' (CMDT) is suggested] are not critically demonstrated, and experimental tests are suggested. Small LA and g(s) may not decrease productivity in well-watered plants under laboratory conditions but may in the field. Optimization of GM traits to environment has not been analysed critically and is required in field trials, for example of recently released oilseed rape and maize which show odrought resistance', probably due to delayed stress onset. Current evidence is that GM plants may not be better able to cope with drought than selection-bred cultivars.