Enhancing wheat productivity under various water availability scenarios is critical to achieving food security. The challenge, however, is that these scenarios dictate distinct traits to enhance yields. Under terminal drought conditions, a water-saving strategy, consisting in reducing the amount of water lost via transpiration under high atmospheric vapor pressure deficit (VPD) conditions, could enhance yields by increasing the amount of stored soil moisture available during the critical seed-fill. However, under more uncertain, stochastic rainfall conditions, a more risk-taking behavior would be needed as the crop needs to maximize CO2 and nutrient uptake by aggressively transpiring in response to VPD, until the next precipitation event.

The goal of this research was to identify the physiological, genetic and molecular controls of these strategies in wheat and leverage them to quantify their impacts on yield in a location-specific manner to inform a new breeding strategy. To this end, we developed a high-throughput, gravimetric phenotyping system that enabled us to screen several wheat populations for their whole-plant transpiration rate (TR) response curves to increasing VPD, a trait that could enable the expression of both strategies dependently on the shape of this curve. In a mapping population descending from a cross between a check cultivar and a drought-tolerant line that exhibited a water-saving TR response to VPD, we identified several QTL controlling the slope of this response, among which a major locus which explained over 25% of the genetic variance. The peak region of this QTL was mapped to genes controlling root vascularization and auxin signaling. In a series of independent experiments, we found that those genes were specifically expressed in the roots, and that root auxins played a role in regulating TR response to VPD in this population by controlling root hydraulic conductivity through modulating root metaxylem size and aquaporin expression. Furthermore, we found evidence indicating that such traits may be under ‘unconscious selection’ as breeders were found to be favoring risk-taking or water-saving responses, dependently on their local precipitation.

Finally, we confirmed the benefits of such breeding strategies using crop simulation modeling. Using the Mediterranean environment of Tunisia as a case study, we identified major yield gains (up to 80%) resulting from the deployment of these traits, with potential to enhance wheat food security in this key region. Overall, this integrated approach highlights the critical importance of designing cultivars with hydraulic properties that match the local water availability regime to maximize wheat yields.