Distribution modelling of Acrosternum spp. (Hemiptera: Pentatomidae) in south of Iran

Abstract

Species distribution models (SDM) are increasingly used in the regional biodiversity assessments, pest management strategies, conservation biology, ecology, and evolution. In the present study, the maximum entropy model was used to predict the potential distribution of three hemipteran stink bugs, namely Acrosternum arabicum, A. breviceps, and A. millierei in Kerman province, south of Iran, using the presence records of the species sampled during 2012–2014 alongside seven environmental predictors. Besides, having described the climatic profile of the species, we explored the contribution percentage of the bioclimatic variables. The accuracy and performance of distribution models were also evaluated by the area under the receiver operating characteristic curve (AUC). According to Jackknife, the annual precipitation, the precipitation of the wettest month, and the precipitation of the coldest quarter were regarded as the most important predictors for A. arabicum distribution model. The maximum temperature of warmest month, the precipitation of the wettest month, and the precipitation of driest quarter for A. breviceps; and also for A. millierei temperature seasonality, the precipitation of coldest quarter, and the precipitation of wettest month were the most effective variables on species distribution. The AUC values, based on training data, were respectively 0.83 for A. arabicum, 0.89 for A. breviceps, and 0.83 for A. millierei. The suitable distribution sites and the most effective bioclimatic variables could be used in a more practical management program for three stink bugs. The MaxEnt algorithm had a good predictability based on the AUC values for the species under study.