Salinity-grain yield response functions of barley cultivars assessed with a drip-injection irrigation system

Success in breeding crops for yield and other quantitative traits depends o n the use of convenient methods to accurately evaluate genotypes under fiel d conditions. We report the evaluation of a drip-injection irrigation syste m (DIS) for assessing the salt tolerance of barley genotypes. Ten barley cu ltivars were randomized within each of nine salinities imposed by a DIS in an experiment with two replications. Grain yields were regressed against so il salinity (ECe) using a sigmoidal growth response model to obtain the sta tistics Y-m, ECe50, and p, The data fitted the model well; the average corr elation coefficient was 0.89 (P < 0.001) when the observations for each cul tivar in both replications were pooled, and the average SEs were <12% of th e mean Y-m and ECe50 estimates. We concluded that the DIS is a reliable sys tem for estimating the salinity response functions of barley. The grain yie lds obtained in the control(ECe = 4 dS m(-1)) and intermediate (ECe = 9 dS m(-1)) soil salinity were highly correlated (r = 0.81, P < 0.01), indicatin g that the highest-yielding cultivars under nonsaline conditions were also most productive under intermediate saline conditions, However, yields at hi gh soil salinity (ECe = 17 dS m(-1)) were not correlated with the control y ields. For the same set of genotypes, the estimates of ECe50 obtained with the DIS and with a triple-line source system (TLS) were strongly correlated (r = 0.84, P < 0.01), even though the direct absorption of salts by the le aves, which is a feature of the TLS, had a deleterious effect on grain yiel ds. Results from the DIS trial (mean ECe50 = 13.1 dS m(-1)) suggest that th e salt tolerance of barley quoted in the literature for similar climatic co nditions (ECe50 = 18 dS m(-1)) could be overestimated by 40%.