The cultivation of pearl millet in India is experiencing important transformations. Here, we propose a new characterization of the pearl millet production environment using the latest available district level data (1998–2017), principal component analysis, and large-scale crop model simulations. Pearl millet cultivation environment can be divided in up to five environments (TPEs). The eastern part of the country (Rajasthan, Haryana, Uttar Pradesh) emerges as the only region where pearl millet cultivation has grown (+0.4 Kha/year), with important yield increase (+51 kg/ha/year), and potential surplus that are likely exported. Important reductions of pearl millet cultivated area in Gujarat (−4.5 Kha/year), Maharashtra and Karnataka (−4 Kha/year) are potentially due to economy-driven transition to other more profitable crops, such as cotton or maize. The potential rain increase could also accelerate this transition. With [0.15–0.61], the tested crop models reflected reasonably well the pearl millet production system in the A1 (North Radjasthan) and AE1 (South Rajastan and Haryana) TPEs covering the largest area (66%) and production share (59%), especially after the use of a new strategy for environment and management parameters calibration. Those results set the base for in silico system design and optimization in future climatic scenarios.
Transpiration restriction under high vapour pressure deficit (VPD), measured indoors with individual plants, increases water use efficiency (WUE). However, VPD is not the only factor driving transpiration, individual plants in a field rapidly become a canopy, and reports on the transpiration restriction versus WUE are scant. We analyzed the transpiration response to the evaporative demand (Penman‐Monteith reference evapotranspiration, ET ref ) and WUE in sorghum canopies outdoors. These responses showed no plateau at high ET ref in 47 genotypes. The slope of the resulting linear relationship over the whole range of ET ref showed a large genetic variability. Unexpectedly, this slope was positively correlated with WUE in experiments with high ET ref . Conversely, a (classical) negative correlation was observed under low ET ref . Genotypes with high WUE and response to ET ref allowed maximum light penetration into the canopy, via more erect leaf orientation. VPD in the canopy was also lower than in open air when the leaf area index reached 2.5‐3. We interpret that higher WUE related to a larger proportion of plant photosynthesis being contributed by lower level leaves that received light and faced lower VPD than leaves exposed to air VPD. This study opens new opportunities, agronomic and genetic, to improve WUE.