Weather based insect pest forewarning models in castor

Survey and monitoring of castor insect pests was carried out both in the farmers’ fields and in the research station to study the influence of weather parameters on the incidence of insect pests in castor for six years i.e. from 2006 to 2011 during Kharif season at two villages and research farm of Regional Agricultural Research Station, Palem. The data were pooled and analyzed statistically with weather parameters using correlation and regression techniques. Correlation studies showed significant and negative relation of leafhopper and semilooper damage with rainfall both in the farmer's fields and in the research station, while Spodoptera litura damage was negatively influenced by evening relative humidity. The multiple regression equations were developed for castor insect pests using linear regression and regression with forwarded selection models. By using the multiple linear regression models, biotic stresses viz., leafhopper, semilooper, Spodoptera and capsule borer were predicted to an extent of 62, 49, 46 and 68%, respectively in the farmers fields, while in research station the prediction rate is 67, 57, 66 and 46%, respectively.