Ecosystem-Level Biodiversity Metrics

Abstract

Measuring biodiversity at the ecosystem level requires an integrated framework that captures taxonomic, functional, and phylogenetic dimensions of diversity across spatial scales -- from alpha diversity within local communities to gamma diversity at landscape and regional extents -- while remaining operationally feasible for national biodiversity monitoring programmes. This study systematically evaluated 24 candidate ecosystem-level biodiversity metrics against four criteria: ecological validity, sensitivity to anthropogenic change, scalability via remote sensing, and alignment with Essential Biodiversity Variables (EBVs) and the Kunming-Montreal Global Biodiversity Framework (GBF) reporting targets. Field validation was conducted across 312 ecosystem plots spanning six biome types in 14 countries over three years (2020-2023), measuring ground-truth biodiversity against satellite-derived spectral diversity indices, UAV-derived canopy structure metrics, and acoustic diversity indices from passive monitoring arrays. Spectral diversity (coefficient of variation of NDVI, CV-NDVI; spectral heterogeneity index, SHI) showed the strongest correlation with ground-truth species richness (r = 0.79, p < 0.001) and functional diversity (r = 0.74, p < 0.001) across terrestrial ecosystems. Acoustic diversity index (ADI) correlated strongly with vertebrate community completeness (r = 0.71, p < 0.001). A composite Ecosystem Biodiversity Score (EBS) integrating five complementary metrics explained 84.7% of variance in ground-truth multi-taxon biodiversity across all biome types, substantially outperforming any single metric. The EBS framework is proposed as a standardised monitoring tool for GBF Target 21 implementation, enabling scalable national biodiversity reporting without requiring exhaustive species-level field surveys.