Abstract

With a view to understanding the basis of cultivar differences in yield under drought, a wide range of cereal cultivars representing durum wheats (Triticum turgidum L.), triticales (X Tritosecale Wittmack), barleys (Hordeum vulgare), and especially tall and dwarf bread wheats (T. aestivum L.) were studied in field experiments in north-western Mexico over three seasons. Drought was created in this rain-free environment by permanently terminating irrigation at various stages before anthesis. Control treatments were well watered throughout the growing period. Detailed measurements of plant water status, leaf area and dry matter production, anthesis date, yield components and grain yield were made. This paper presents primarily the grain yield data.

Drought levels were such that the mean yield of all cultivars under drought ranged from 37 to 86% of control yield, corresponding to irrigation cut-offs varying from 69 days before mean anthesis date to only 10 days before. In each experiment the grain yield under drought showed highly significant cultivar differences, which appeared consistent between years. Yields were adjusted for drought escape by using a correction factor which ranged from 2.9 to 8.5 g/m² per day advance in flowering, being greater in experiments with less severe drought.

The demonstration of linear relationships between cultivar yield and drought intensity, as indicated by the mean yield of some or all cultivars, prompted the consideration of cultivar yield under drought as the function of yield potential (Y_p, yield without drought), drought susceptibility index (S), and intensity of drought. The cultivar groups showing lowest S values (most droughtresistant) were tall bread wheats and barleys; dwarf bread wheats were intermediate, and durum wheats and triticales were the most susceptible. However, because dwarf wheats have a higher yield potential (Y_p) than tall bread wheats, it is suggested that, as a group, tall bread wheats would outyield dwarf wheats only under very severe drought. Also there was considerable within-group variability of S and Y_p. Cultivar S values were consistent across experiments. Yield responses of tall and dwarf bread wheat groups obtained in these experiments agreed with those seen in extensive international trials under dryland conditions.