Niche-based prediction of establishment of biocontrol agents: An example with Gratiana boliviana and tropical soda apple

Document Type

Article

Publication Date

4-1-2012

Abstract

Successful establishment of a biological control agent is a prerequisite for effective reduction of an invasive weed. Niche-based species distribution models can generate valuable information about the potential spread of a biological control agent and help to predict its distribution. The Maximum Entropy Species Distribution Model was used in our study to predict distributions of the leaf beetle Gratiana boliviana Spaeth (Coleoptera: Chrysomelidae) and its target weed, tropical soda apple (TSA), Solanum viarum Dunal (Solanaceae). The specific objectives of this study were to (1) assess the climatic suitability for establishment of this insect across the invasive range of the weed by overlapping predicted current distributions and, (2) examine the niche-related restriction in distribution of the insect by visualising the predictive niche in multivariate space. Accuracy of predicted distributions was tested using binomial tests and area under the curve scores. The results of statistical tests confirmed that the predictions were significantly better than random. The predictions indicated that the potential distribution of G. boliviana in the USA will be more restricted that of its host plant. Consequently, the beetle's ability to inflict damage to TSA will be geographically limited. Niche visualisation, using a PCA-based analysis, provided evidence of the niche imposed restriction on the distribution of G. boliviana. Overall, this study proposes a new approach for understanding the spatial limitations in establishment of biological control agents, allowing researchers to establish more realistic expectations of success. © 2012 Copyright Taylor and Francis Group, LLC.

Publication Source (Journal or Book title)

Biocontrol Science and Technology

First Page

447

Last Page

461

This document is currently not available here.

Share

COinS