Indexed on: 01 Jul '91Published on: 01 Jul '91Published in: Irrigation Science
The utility of water production models as irrigation management tools is dependent upon their accuracy. Development of precise water production models requires a thorough understanding of how water and other factors interact to affect plant growth and yield. The objective of this experiment was to identify significant environmental variables which control water production function (transpiration vs. yield) variability between harvests and seasons for alfalfa (Medicago sativa L.) over a seven year (1981–1987) period in northwestern New Mexico. A single line-source design was used to supply a continuous gradient of irrigation (I) to the crop, and transpiration (T) was calculated as the difference between evapotranspiration, as estimated by the water balance method, and modeled soil water evaporation at each I level. Yield per cutting was found to be a function of T, growing degree-day accumulation, average daily solar radiation, year and harvest number within year. A multiple regression equation formulated with these variables explained 82% of the yield variability. Average yield per cut in 1981 at 50 mm of T was l Mg ha-1 and in 1985 at the same level of T was 2 Mg ha-1 based on the regression model. Yield per cut at any given level of T, as estimated by the coefficients of this equation reached a maximum at year 5.7 and a minimum in year 1. Within a season, yield per unit T was generally greatest at cut 1 and lowest at cut 2. Total seasonal yield was found to be a function of T and year which explained 90% of yield variability. Yield varied from 0.83 Mg ha-1 to 18.1 Mg ha-1 and T varied from 186 mm to 1298 mm.