Resiliences to water deficit in a phenotyping platform and in the field: How related are they in maize?

Resilience to soil water deficit is a priority of many maize breeding programmes. Its genetic analysis requires estimators that characterise each genotype in a reproducible way. We have tested and compared three methods for that, namely (i) the ability of hybrids to maintain leaf growth in a range of soil water potentials in a phenotyping platform, (ii) a direct estimator of resilience of seed number to water deficit in a network of field experiments and (iii) classical methods involving tolerance indices and variance analysis. A drought index was obtained by averaging soil water potential, measured with tensiometers, for the phenological phase during which seed number is determined (evaluated for each individual hybrid in each site). It closely correlated with seed number in each of the 19 hybrids that were analysed over 14 environmental situations in France, Hungary and Chile. The slope of the regression line between drought index and seed number, established for each hybrid, was taken as an estimate of the resilience to soil water deficit of that hybrid. Resilience estimated in this way varied 2-fold in the set of studied hybrids, and correlated with resilience of leaf growth to soil water deficit in short-term experiments in the phenotyping platform. In contrast, the resilience estimated via the interaction genotype x watering treatment resulted non-significant because of large differences in drought indices between genotypes in a given watering treatment. We therefore propose that direct estimation of resilience to water deficit is feasible in the field with a minimum amount of environmental measurements. This provides estimations that are consistent with those measured in a phenotyping platform. The combination of both estimates provides insight into the mechanisms associated with resilience of each hybrid, potentially contributing to accelerate the genetic gain. (C) 2011 Elsevier B.V. All rights reserved.