A Critical Appraisal of Grasshopper Forecast Maps in Saskatchewan, 1936-1958

Abstract

Final Grasshopper Forecast Maps issued annually in Saskatchewan are based on data obtained from two surveys—an adult survey undertaken in summer and an egg survey undertaken in the fall. Using the rural municipality as the basic unit of study the accuracy of the maps during the period 1936-58 has been tested against the following sets of data: 1) nymphal surveys, (1945-47 only); 2) adult surveys; and 3) crop damage reports. The forecast maps achieved their greatest degree of accuracy (82%) when grasshopper populations remained at low levels and decreased slightly from year to year. They were less reliable when the severity of the outbreaks increased and at the highest levels the degree of accuracy was no greater than that which could be expected to arise by chance. When the grasshopper population increased from year to year the forecasts underestimated the severity of the outbreak and when the population declined they tended to overestimate it. The Preliminary Forecast Maps, which are based on data obtained during the adult survey only, differ little from the Final Forecast Maps but have been slightly more successful in predicting the more severe outbreaks.