RESUMO
Finger millet has gained considerable attention worldwide due to its nutritional and health benefits. Being a rainfed crop of semiarid and arid regions, drought is one of the major constraints to its yield stabilisation. To address this, a set of 38 accessions of finger millet were evaluated in both field and mini-lysimeters under both well-watered (WW) and water-stressed (WS) conditions. The objectives of the study were to identify the range of variations for yield components, water-use (WU) and transpiration efficiency (TE) and to examine the potential of the mini-lysimeter system in assessing the genotypic performance in the field conditions. Approximately 2-fold variations in shoot biomass and ~9-fold variations in grain yield under WS conditions were observed. Reproductive growth was more sensitive to WS than the vegetative growth. Our results indicate that in addition to yield potential under WW conditions, WU followed by TE were the other two major contributors toward shoot biomass, whereas, HI followed by TE were the major contributors toward grain yield under WS. The close association between the yield components recorded in the field and in mini-lysimeters suggests that the lysimetric system has the great potential to reflect the genotypic performance under field conditions. Regression analyses suggest that HI explained almost all the variations in grain yield under WW conditions, whereas under WS treatment, next to HI, both TE and WU had also contributed significantly to grain yield. The absence of interrelationship between WU and TE suggests that both these components contribute independently toward the yield components under WW or WS conditions. The accessions with higher shoot biomass and grain yield extract much more water during the post-anthesis stages than the poor performers under WS. Results also suggests that higher WU contributed more towards shoot biomass and higher TE contributed more towards grain yield by improving the harvest index.
