Climate, soil and crop yield relationships in Cascade County, Montana

Abstract

The Productivity Index (PI) model estimates the productivity effects of erosion by the simulated removal of surface soil and consideration of available water-holding capacity, bulk density and pH. Although it has performed well in the US Corn Belt and elsewhere, further testing is required to demonstrate its applicability in other semi-arid environments. Evaluation of model performance in four fields in Hill and Jefferson counties, Montana, revealed a weak relationship between PI and small grain yield, possibly due to local conditions. Therefore, soils and crop data were extracted from the USDA-SCS SOILS-5 and Montana Agricultural Potentials System (MAPS) databases, as well as the county soil survey, to evaluate PI model performance and indicate appropriate changes in its design in Cascade County, Montana. Results indicate that model performance can be improved with the addition of factors to account for water balance, slope, growing degree days and calcium carbonate content. Regression of barley, spring wheat and winter wheat yield data against PI values from the original model accounted for 34, 31 and 31 per cent of the variability in yields of these three crops, respectively. R 2 increased an average of 77 per cent and accounted for 54, 59 and 58 per cent of the variations in yields when the four new factors were added to the model. These results have encouraged further efforts to develop a modified version of the PI model that uses computerized databases for county-scale assessments of semi-arid environments.