Disentangling the response of species diversity, forest structure, and environmental drivers to aboveground biomass in the tropical forests of Western Ghats, India.

Tropical forests are crucial to the global carbon cycle, but a significant knowledge gap in the precise distribution patterns of forest aboveground biomass (AGB) hinders our ability to formulate effective conservation efforts. A key unresolved issue is the lack of understanding of how forest AGB interacts with biotic and abiotic factors on large spatial scale. To address this, we used Structural Equation Modeling to disentangle the direct and indirect effects of environmental, anthropogenic, structural diversity species diversity and edaphic factors on AGB of trees, lianas and regenerating communities using the data from 96 1-ha plots in the central Western Ghats biodiversity hotspot, India. We hypothesized that the effect of structural attributes overrides AGB distribution, with relative contributions varying among plant communities. The landscape-level mean AGB was 245.12 ± 19.74 Mg ha^-1, with SEM explaining 68-85 % of variations across the three vegetation communities. Structural diversity emerged as the primary mediator of the positive effects of taxonomic diversity on AGB in the regeneration community, whereas canopy cover and stem density linked diversity to AGB in adult tree and liana communities. Further, AGB showed a positive association with soil organic carbon in adult tree and regeneration communities, underscoring the significance of belowground resource availability on AGB. The results indicate that structural features were consistently the strongest AGB predictors at all levels of data aggregation, indicating the predominant role of niche complementarity and efficient space utilization in driving AGB, albeit differently across the plant communities. Our study emphasizes the importance of maintaining high structural features and managing taxonomic diversity while promoting soil fertility and minimizing disturbances to support AGB in tropical forests. We recommend testing the effects of predictor variables on biomass of vegetation communities independently to better understand the ecological principles of forest functioning.