Comparison of transient state models that include salinity and matric stress effects on plant yield

Transient-state models that account for continually changing salinity and matric stress on crop yields have been developed by several research groups. The objective of this research was to compare the simulated yields of forage corn obtained from a common set of soil and water conditions for ENVIRO-GRO, HYDRUS, SALTMED, SWAP and UNSATCHEM. The physical and hydraulic properties of Panoche clay loam were used. The amounts of water applied weekly, based on the climatic conditions in the San Joaquin Valley of California, ranged from 0.9 to 1.3 times (Kirr) the potential evapotranspiration (PET) of corn. The salinity of the applied water (ECiw) ranged from 0.5 to 6 dS/m which brackets the threshold soil-water salinity of forage corn (Zea mays L) of 3.6 dS/m. The model simulations were run for sufficient back-to-back crop seasons to establish transient matric and osmotic conditions within the root zone that did not change from one crop season to the next, a quasi steady-state condition. SALTMED simulated lower relative yields (RY) than the other models for all combinations of Kirr and ECiw. For the other models, RY values were similar (within about 7% or less) for ECiw <= 3 dS/m for all Kirr values. Plots of RY versus ECiw for HYDRUS, SWAP, and UNSATCHEM approximately paralleled each other except that UNSATCHEM produced higher values. ENVIRO-GRO produced the highest RY where Kirr >= 1.1 and ECiw <= 2.0 dS/m but decreased more rapidly for greater ECiw. ENVIRO-GRO has plant-based compensation which allows water uptake to meet PET as long as any portion of the root zone is not exposed to matric or osmotic stress that exceed threshold levels. This compensation factor produced higher RY at the lower ECiw values. More rapid decrease in RY with increasing values of ECiw simulated by ENVIRO-GRO is attributed to the assumption that the osmotic and matric stresses are additive, whereas the others assume that they were multiplicative. All the models except UNSATCHEM assume a constant relationship between EC and salt concentration in solution. UNSATCHEM takes into account the effects of the ionic composition and ion concentration on osmotic potential, resulting in higher RY values obtained with this model. Since the chemical composition of irrigation waters are all unique, this aspect of UNSATCHEM poses an important capability in the assessment of osmotic effects on crop yields. We conclude the models provide a valuable resource to assess the utility of moderately saline irrigation waters, for a broad range of transient conditions which include variable crops, precipitation, irrigation water management, and irrigation water salinity. We also highly recommend their use to assess the results obtained in experiments that focus on the responses of crop growth and yield to transient changes in soil water content and salinity. (C) 2011 Elsevier B.V. All rights reserved.