Abstract

The models for forewarning about the infestation of green leafhopper Nephotettix virescens Dist (Cicadellidae, Hemiptera), plant hopper Cofana spectra Dist (Delphacidae, Hemiptera), C. yasumatsui Young (Kolla mimica, Hemiptera), rice gundhi bug Leptocoriza acuta Thunberg (Alydidae, Hemiptera) and yellow stem borer Scirpophaga incertulas Walker (Pyralidae, Lepidoptera) in rice growing season (July to November) was studied through light trap collection over fifteen years (1985-1999). Maximum population of N. virescens Dist (Cicadellidae, Hemiptera), C. yasumatsui Young (Kolla mimica, Hemiptera) and L. acuta Thunberg (Alydidae, Hemiptera) were recorded in the third week of October all the years. C. spectra Dist (Delphacidae, Hemiptera) had maximum population in the second and third weeks of October during the aforesaid period. Maximum population of S. incertulas Walker (Pyralidae, Lepidoptera) was recorded in the month of September in all the years. After making a transformation on the response variable that is, population of insects, the cubic polynomial model was fitted with week as explanatory variable and it described the dynamics of the populations of all considered insects during the weeks. The values of multiple correlations for N. virescens Dist (Cicadellidae, Hemiptera), C. spectra Dist (Delphacidae, Hemiptera), C. yasumatsui Young (Kolla mimica, Hemiptera), L. acuta Thunberg (Alydidae, Hemiptera) and S. incertulas Walker (Pyralidae, Lepidoptera) were in the order of 0.964, 0.947, 0.971, 0.881 and 0.949, respectively. We also include meteorological factors in the model and it provides the dynamics of the populations of all the above mentioned insects for forecasting.

Key words: Meteorological factors, transformation, regression analysis, rice insect pests.