Environmental heterogeneity and multidimensional filtering decouple insect richness and abundance in a montane wet-island

Environmental heterogeneity is a fundamental driver of terrestrial insect diversity, yet the synergistic effects of its multiple dimensions remain insufficiently quantified. Focusing on a representative “wet-island effect” region in the Central Asian mountains (the Tianshan Mountains, Northwest China), which features exceptionally cool and humid microclimates compared to the surrounding arid matrix, this study investigated how multidimensional environmental gradients regulate insect community diversity and structure. Insect communities and environmental variables were surveyed across 20 horizontal habitat plots representing distinct habitat types. Significant differences were detected in insect diversity metrics and functional group composition among habitats. Correlation analysis, Gephi network visualization, and principal component analysis (explaining 94.99% of total variance) identified three dominant environmental gradients: a soil salinity-nutrient gradient (PC1), a vegetation productivity-microclimate gradient (PC2), and a soil hydrothermal gradient (PC3). Species richness and individual abundance exhibited pronounced spatial decoupling and responded to distinct environmental pathways. Specifically, we observed that the environmental gradients maximizing species diversity fundamentally differed from those supporting high population densities. Consequently, this decoupling highlights that conservation strategies focusing solely on habitat heterogeneity may fail to protect both biodiversity and biomass simultaneously. Bayesian structural equation modeling revealed niche differentiation in their drivers. Abundance was primarily promoted by the direct effects of PC1 and PC2 and indirectly via increased plant cover, indicating an energy-resource-mediated pathway. In contrast, species richness was regulated by more complex direct effects, responding negatively to PC1 but positively to PC2 and PC3, with minimal mediation by plant cover. This divergence in the drivers of abundance and richness underscores the need to explicitly incorporate multiple dimensions of environmental heterogeneity in biodiversity assessment and management. The proposed “multi-axis environmental heterogeneity-community assembly” framework elucidates insect community responses to habitat variation and provides a theoretical basis for predicting insect diversity dynamics in arid montane ecosystems under global environmental change. • Insect community assembly in an oasis is driven by the type, not the amount, of environmental heterogeneity. • A fundamental decoupling is found between drivers of species richness and individual abundance. • Three environmental gradients act as antagonistic filters on community structure and stability. • Network analyses reveal richness and abundance are driven by semi-independent modules of factors. • We propose a multi-dimensional “heterogeneity template” framework to predict biodiversity change.