Predicting Habitat Changes and Vulnerability of Climate-sensitive Insects under SSP Scenarios

Abstract

Listen

Translate article

Climate change significantly impacts the distribution and habitat suitability of insects, particularly those highly sensitive to environmental fluctuations. This study evaluated the habitat changes of 12 climate-sensitive insect species in South Korea under shared socioeconomic pathways (SSP) scenarios, SSP2-4.5 and SSP5-8.5, using random forest (RF) models. Bioclimatic variables, including annual mean temperature (BIO1) and annual precipitation (BIO12), were identified as key contributors to habitat suitability changes. The model demonstrated high predictive accuracy, with receiver operating characteristic (ROC) values exceeding 0.8 for five species, such as <i>Papilio helenus</i> and <i>Argynnis hyperbius</i>, while six species, including <i>Sympetrum pedemontanum elatum</i>, exhibited lower predictability due to data distribution challenges. The results revealed that SSP2-4.5 allowed more stable or expanding habitats for certain species, such as <i>Argynnis hyperbius</i> and <i>Lampides boeticus</i>, where habitat areas significantly increased by 2070. In contrast, SSP5-8.5 showed drastic habitat reductions for most species, including <i>Camponotus kiusiuensis</i> and <i>Sympetrum pedemontanum elatum</i>, with some habitats shrinking by over 90% by 2090. The study underscores the importance of climate variables, with temperature and precipitation consistently influencing habitat changes across species. This research provides critical insights into the ecological risks posed by climate change and emphasizes the necessity of mitigation strategies. While some species demonstrate adaptive potential under moderate scenarios, others face severe vulnerabilities under extreme climate conditions. These findings offer valuable guidance for biodiversity conservation and policy-making, highlighting the need for integrated approaches that account for non-climatic factors such as land-use changes.