Modelling Yield Response and Water Use to Salinity and Water Relations of Six Pepper Varieties

Abstract

Better understanding of crop yield response under salinity and water deficit conditions is essential to meet food need under the circumstance of population growth and climate extremities. It has been well known that plant species response differently under stress conditions. Recent studies show that these different responses occur not only among species but also in different varieties within the same species. The aims of the study are to examine and to compare yield, yield response factors (ky), salinity thresholds, biomasses, and water productivity responses of six varieties of pepper plant (Sürmeli-Hot, Yalova, BT016-Hot, BT 016, BT Ünsal, BT Demok) under salinity conditions. In another experiment under the same conditions (location, time, growth media etc.), water deficit was applied to two of these six varieties (BT Ünsal and BT Demok) separately, and their responses to salinity and water deficit conditions were compared. The experiment was carried out in containers. The amount of irrigation water was determined manually by weighing each container. Water deficit treatments were consisted of meeting 120, 100, 70 and 50% of soil water depleted from field capacity. Water salinity levels were 0.25 (control), 2, 4 and 6 dS m-1. There was no difference in yield under non-stress and excessive stress conditions, but the yield difference was as high as 38.9% under moderate stress conditions. Varietal differences were also observed for water productivity. Salinity threshold values vary between 0.89 and 1.83 dS m-1. Yield response factor (ky) were high for all varieties under salinity. Comparing the ky values obtained under water deficit and salinity experiments, sensitivity to salinity induced water stress was found higher than that of applied water deficit itself. Using salinity (Model 1) and water deficit (Model 2) data set of two varieties, two models were created plotting relative yield and water potentials (osmotic potential + matric potential) and compered their predications statistically. Statistically better predictions were obtained from Model 2.