An integrated model-approach to the effect of water management on crop yield

Abstract

Quantifying the effect of drainage on crop yield is of essential importance in agricultural management. In this article a model is described with which this effect can be computed. For both arable land and grassland the factors acting in spring, summer and autumn are dealt with separately.Arable land. In spring sowing date is the main factor affecting the crop yield. Sowing date depends on the tillage conditions of the soil toplayer. By means of an existing model, the course in time of the soil water tension of the upper layer is simulated in connection with rainfall, evaporation, drain depth and drain intensity data. Using specific criteria on minimum soil water tension for tillage operations, the dates and number of workable days can be established from the model output. The expected yield depression is then derived, using an experimental relationship between yield depression and number of days of sowing delay.During the growing season the yield directly depends on the magnitude of the actual evapotranspiration. This value can be computed by means of a known evapotranspiration model for various drought frequencies, groundwater table depths in spring, drain intensities and amounts of water supplied. The yield can be obtained from the relationship between yield and relative evapotranspiration. Combining this yield with the yield depression obtained by means of the workability model gives the actual yield.In autumn crop yield is influenced by the working conditions during harvest. Via the workability model, the dates and the number of days available for harvesting are determined. Yields are derived from an experimental relationship between yield depression and number of days of earlier harvesting. An example is given for summer cereals growing on a heavy sandy loam soil under meteorological conditions prevailing in The Netherlands.Grassland. The effect of shallow groundwater table depths in winter and spring on the yield of the first and second cut can be determined with the workability model in an identical manner to that given for arable land. Because of lack of data, a slightly different approach was followed in this paper. With the workability model the course of groundwater table depth during winter and spring can be simulated and the mean depth determined. From the relation between yield depression and mean groundwater table depth over the period November through May the yield depression can be found. Combining this with the yield obtained with the evapotranspiration model gives the actual yield. An example representative for The Netherlands is given for grass on peat soil.