EFFECT OF CROP ROTATION AND FERTILIZATION ON THE QUANTITATIVE RELATIONSHIP BETWEEN SPRING WHEAT YIELD AND MOISTURE USE IN SOUTHWESTERN SASKATCHEWAN

Abstract

The effects of crop rotation and fertilization on the quantitative relationship between spring wheat (Triticum aestivum L.) yields (y), available spring soil moisture (AvSpSM), and growing season precipitation (GSP) were determined from an experiment conducted on a loam soil in the Brown soil zone of southwestern Saskatchewan for the period 1967-1984. Moisture used (MU) was defined as GSP + (SpSM-Harvest SM in 0- to 120-cm depth), and AvSpSM as volumetric moisture content at time of planting minus the lowest soil moisture measured at harvest. For discussion of moisture use characteristics GSP was defined as precipitation received from 1 May to 31 Aug.; for predictive equations GSP was more effective if GSP was taken as 1 May-31 July. When the 18-yr data were analyzed, the relation of y to MU was curvilinear. If data for 1 or 2 yr (1970 and 1983) with skewed distribution of GSP were omitted, the relationship between y (kg ha-1) and MU (mm) for well-fertilized fallow- and stubble-seeded wheat combined was y = -544 + 10.07 MU (R2 = 0.63 ***) (where GSP = 1 May to 31 July). This equation performed well when it was used to estimate yields of fallow- and stubble-seeded wheat in 1985 and 1986. Stubble-seeded wheat required 68 mm of moisture to produce the first kilogram per hectare of grain; fallow-seeded wheat required about 46 mm. There was no effect of fertilizer on these lower threshold values, but the yield increase per millimetre of MU tended to be greater for the better fertilized rotations. The main difference in our y vs. MU relationships compared to those reported for the period 1925-1950 was that the lower threshold level of MU for grain production decreased from about 140 mm to the values cited above; this has resulted in substantially greater moisture use efficiency in recent years likely due to better, more timely crop management and the improved cereal varieties now used. When y was related to AvSpSM and GSP (1 May to 31 July) analyzed as separate variables, the relationship was only improved in terms of coefficient of determination (R2) value when a quadratic model was used. However, the latter did not predict the 1985 and 1986 yields as well as the y vs. MU relationship did. For fallow-seeded wheat, the relative effect of GSP on yield variability was about 5.4 times as great as that of AvSpSM; for stubble-seeded wheat it was only 1.5 times as great. GSP was equally important in affecting yields of wheat grown on fallow or stubble. The grain-filling period was confirmed as the most important for the occurrence of precipitation for both fallow- and stubble-seeded wheat, but precipitation at or near seedling time was almost as important for stubble-seeded wheat since this ensures the establishment of an adequate plant stand.