Climate in relation to deterioration of stored grain : A multivariate study

A hypothesis elucidating the basic relationship between abiotic climatic variates and the deterioration of stored grain by biotic agents was developed for the 41 crop districts in the Prairie Provinces of Canada. Principal component (7 variates measured) and canonical correlation (6 variates measured) analyses were applied to data from each of the 3 crop years (1967-1969), for each of the 3 principal component analyses and for each of the 3 canonical correlation analyses. The data consisted of crop district means of infestation reports from 2522 managers of grain elevators and also of representative meteorological observations from each crop district. The major variates measured were the number of reports of: hot spots, infestations of a) fungus beetles and mites, and b) Cryptolestes ferrugineus (Stephens), and grain fumigants. The climatic variates consisted of mean temperature, precipitation, and number of days above 20° C (68° F) for July, August and September. Three major components, which together explained 64-80% of the variation, indicated broadly similar trends in all analyses. The first component indicated that lower mean temperatures and larger numbers of cool days increased the incidence of fungus-induced hot spots in wet harvested grain. The second component indicated the extent of arthropod infestations: warmer weather increased C. ferrugineus infestation, whereas cooler weather increased infestation by mites and secondary insect species. The third component was a measure of precipitation. Canonical correlation analyses based on 1967 and 1969 data yielded canonical roots that were statistically insignificant at the 1% level. Analyses of 3 external abiotic and 3 biotic response variates based on 1968 data showed that only the first pair of canonical variates was significant, the degree of predictability (R _c² ) being 56% (P<0.01). The analyses revealed that lower temperatures, a greater number of cool days, and higher precipitation increased the number of hot spots in 1968 confirming the findings of the principal component analyses. The pattern of hot spot development in relation to climatic characteristics was similar in all crop districts for all years examined. There was a linear relationship between climatic and biotic variates for all these crop districts but the degree of relationship varied geographically. The first and second principal component values for each crop district were ranked and plotted on a climatic map of the Prairie Provinces. The ranking showed that the incidence of hot spots and arthropod infestations was highest in the Sub-boreal and Humid southeastern regions, whereas the incidence was lowest in the Dry Belt region. C. ferrugineus was most common in the Humid southeastern region.