Development of a Wheat Model for Land Evaluation
Simulation is a new approach to rural land evaluation. Its major advantage over more conventional methods is that it lends itself to a consideration of a large number of alternatives. If properly used, it can also meaningfully synthesize large amounts of varied information to provide reliable predictions. The prime objective of the study was to examine the possibility of using computer simulation techniques as a means of evaluating the biological (wheat) productivity of various land units in the prairie provinces of Canada. The literature on crop production modelling was discussed at length since there exists much controversy regarding why and how such models should be developed. To ensure general applicability of the approach, the information needed to develop and test the model, WHTMOD, had to be readily available or obtainable. For model development, existing data on hard red and soft spring wheat production, moisture use and climate were used. Using a method similar to de Wit (1958), 'm' values were calculated for the hard and soft wheats. Different sites and years from those used in model development were used to test WHTMOD. Data required for model testing included daily temperatures and precipitation, tabulated astrometeorological data, soil moisture at seeding, readily obtainable soil physical properties, the 'm' value and seeding date. WHTMOD simulated phenology from seeding to harvest using Robertson's (1968) biometeorological time-scale. A moisture budget was kept to assist in modelling transpiration and evaporation. Top and root growth were simulated on a daily basis using de Wit's (1958) equation modified to take account of soil nutrient status and the nutrient-water interaction. After heading, top growth was partitioned between grain and straw. Yield predictions were made on a soil series basis as a larger unit would not be sufficiently homogenous in its crop yields. However, a method is suggested to predict yields for a tract of land of any specified size. The yield predictions were generally encouraging, although there was a tendency to underpredict yields, apparently due to underprediction of moisture use. Tests also showed considerable variation in the nutrient factors for identical sites in different years. This was due to different moisture conditions between test years. It was suggested that tables of nutrient factors be calculated for a range of moisture and nutrient conditions to simplify future use of WHTMOD.