Population growth curves in three species of Diabrotica in the region of Irapuato, Guanajuato, Mexico, were estimated using multivariate multiple regression techniques through a third-degree polynomial function defined by:
where T is the average monthly standard temperature, PP is the average monthly accumulated rainfall, and Y1, Y2, and Y3 are the numbers of Diabrotica undecimpunctata (Harold), Diabrotica balteata (LeConte), and Diabrotica virgifera (Krysan & Smith) adults, respectively. The Roy criterion was used in the multivariate analysis of variance (MANOVA) of the multivariate multiple regression, which statistical test value was 95.2 with a probability for its corresponding approximation F of 8.548e-05. Consequently, the estimated polynomials for the population growth curves were adequate. Multivariate association measures Wilks (Λ; = 0.9984) and Roy (θ = 0.9896) were calculated, allowing estimation of growth curves as prediction tools. These models of polynomial growth curves were compared to each other through the statistic W = 23171.78, which yielded a probability value P = 0**, concluding that the growth curves were different. The critical points for temperature and rainfall (T, PP), at which the number of individuals of each species reached its maximum, calculated through standard techniques of mathematical programming were (14.0562, 4.5) for D. undecimpunctata, (13.7000, 168.0) for D. balteata, and (14.7020, 4.5) for D. virgifera.