Multiplicative Regression Models Application in Identification of Effective Factors on Integrated Pest Management in Khuzestan Province

The intensive use of pesticides has detrimental effects on the environment. Such challenges show that integrated pest management (IPM) plans should be extended and applied. In this regard, the investigation of effective factors on the number of applied IPM operations by farmers can help to extend IPM. Furthermore, these kinds of studies can be an appropriate guidance for programmers in an administration of sustainable agriculture. For this, in this study, we investigated the effective factors on the number of applied IPM operations in Khuzestan province to reduce pesticides environmental risk. The required data were collected from 180 farmers of Khuzestan province in 2014. According to dependent variable identity, i.e. it is discrete and count data, the multiplicative regression models were used. The dispersion parameters for negative binomial I, II and generalized negative binomial was about zero, while this parameter for generalized Poisson model was -0.429. In addition, the statistics in LR test for negative binomial models and generalized Poisson were zero and 44.71, respectively, that showed our model is under-dispersion. Consequently, the generalized Poisson model was most suitable. With attention to estimated coefficient, farmer's experience, education, knowledge, confronting with pesticides environmental risk and farmer's revenue have positive and significant effects on IPM, while pest severity has negative impacts. The incident rate ratios results showed knowledge level with 1.099 and pest severity with 0.896 were most and least effective factors, respectively. Therefore, poster, visit of farms, which show the effect of chemical poisons, and adoption of persuasive policies to use biologic fight and other IPM operations are necessary.