Comparison of traditional and ET-based irrigation scheduling of surface-irrigated cotton in the arid southwestern USA

Abstract

The use of irrigation scheduling tools to produce cotton under-surface irrigation in the arid southwestern USA is minimal. In the State of Arizona, where traditional irrigation scheduling is the norm, producers use an average of 1460mm annually to grow a cotton crop. The purpose of this paper was to determine whether or not the use of ET-based irrigation scheduling methods could improve lint yield and irrigation water use productivity over traditional cotton border irrigation scheduling practices in the region. A field study with four irrigation scheduling treatments replicated in 4 blocks was conducted for two cotton seasons (2009 and 2011) in 16, 12-m×168-m cotton borders at the Maricopa Agricultural Center (MAC), in Arizona, USA. Remotely-sensed vegetation indices (VI) were used to estimate basal crop coefficients (Kcb) at 40, 4-m×8-m zones within borders for two treatments, denoted as VI_A and VI_B, whereas a single Kcb curve was applied to all zones in borders for a third treatment (FAO). Daily ETc for these three treatments was estimated using FAO-56 dual crop coefficient procedures with local weather data and irrigation scheduling for the three treatments were based on soil water balance predictions of soil water depletion (SWD). For the VI_A and FAO treatments, irrigations were given when predicted SWD of all 160 zones in the treatment averaged 45% of total available water (TAW). For the VI_B treatment, irrigations were given when 5% of the 160 zones in the treatment were predicted to be at 65% SWD. A fourth treatment (MAC) represented the traditional irrigation scheduling treatment and was scheduled solely by the MAC farm irrigation manager using only experience as a guide. The study showed that the lint yields attained under the MAC farm manager’s irrigation scheduling equaled or exceeded the yields for the three ET-based irrigation scheduling treatments. Although the MAC irrigation scheduling resulted in somewhat higher irrigation input than for the other treatments, the MAC treatment maintained or exceeded the irrigation water productivity attained for other treatments that had lower irrigation inputs. A major conclusion of the study was that present-day irrigation water use for cotton in surface-irrigated fields could be substantially reduced. When compared to Arizona state cotton averages, any of the four treatments presented in the study could potentially offer methods to significantly reduce cotton irrigation water use while maintaining or increasing current lint yields levels.