Exploring models to predict the establishment of the leaf-feeding beetle Zygogramma bicolorata (Coleoptera: Chrysomelidae) for the management of Parthenium hysterophorus (Asteraceae: Heliantheae) in India. (August 2019)

• Record Type: Journal Article
• Authors: Gharde, Yogita; Sushilkumar, ; Sharma, A.R.
• Is Part Of: Crop protection. Volume 122(2019)
• Page Start: 57
• Page End: 62
• Publication Date: 2019-08
• Subjects: Logistic model -- Machine learning approach -- Modelling -- Rainfall -- Relative humidity -- Temperature; Plants, Protection of -- Periodicals; 632.05
• DOI: 10.1016/j.cropro.2019.04.014
• Languages: English
• ISSNs: 0261-2194

Abstract

Abstract: Parthenium hysterophorus L. is considered globally as one of the most troublesome weeds, which has invaded non-cropped as well as cropped lands in many tropical and sub-tropical countries including India. Biological control using the Mexican beetle ( Zygogramma bicolorata Pallister) among other agents, has been considered most successful in India. In the present study, models were developed using climatic indices (minimum temperature, maximum temperature, rainfall and relative humidity) as independent variables to determine the suitable sites for the establishment of Z. bicolorata . Regression approach for model building revealed 79.6% concordance, 20.0% discordance, and 0.4% tie between predicted probabilities and observed responses whereas, decision tree using J48 algorithm classified 81.7% instances correctly which helped to frame the rules for predicting the establishment of Z. bicolorata . Based on these rules, it was concluded that a site experiencing climate with indices values of average minimum temperature 24.2–26.2 °C and rainfall 191.2–257.3 mm during July to October would be highly suitable for setting up the Z. bicolorata population in the region. These models might be useful to decide the most suitable sites for release and establishment of Z. bicolorata in India as well as in other parts of the world with similar climatic conditions. Highlights: Parthenium hysterophorus L. is one of the most troublesome weeds globally. Biological control of P. hysterophorus using Zygogramma bicolorata Pallister has been considered most successful in India. Regression approach and decision tree using J48 algorithm were used to model the establishment of beetle using climatic variables. . Minimum temperature and its interaction with rainfall and relative humidity were found to be important climatic variables These models might be useful for the selection of most suitable sites for release and establishment of Z. bicolorata.