Biodiversity is inherently multidimensional, combining information on the numbers, abundances, taxonomy, and traits of species. Quantifying biodiversity with finite sets of metrics is an ongoing challenge, relying on essential variables with high complementarity, accessibility, and ecological importance. Here, we explore the large-scale multivariate structure of coral reef biodiversity, identifying metrics that represent major dimensions of variation in reef composition and function. Focusing on two large-scale datasets from The Great Barrier Reef and Coral Sea, we conduct a range of dimensionality-reduction analyses to explore how 18 widely accessible reef biodiversity metrics are related and the multivariate information represented by each metric. We identify six variables that describe major components of variation in reef biodiversity. These are (1) hard coral cover, (2) fish biomass, (3) structural complexity, (4) coral functional composition, (5) fish functional composition, and (6) algal composition. These complementary metrics accurately predicted the multivariate structure of long-lists of biodiversity metrics ( n = 14, Mantel r = 0.76), and captured large-scale information on species richness, size structure, shelter volume, non-coral sessile groups, and coral-fish interactions. Functional composition estimated via the abundances of 5–10 major functional groups was effective in predicting multi-trait functional diversity. While these six biodiversity dimensions capture a small component of coral reef species richness, they represent large components of reef biomass, habitat structure, and ultimately ecosystem functions and services. Moreover, these metrics are optimized to have high complementarity and accessibility, making them appropriate for use in a range of biodiversity conservation practices with diverse goals. • Measuring biodiversity with finite sets of metrics is a key challenge in conservation. • We identify complementary and accessible ‘dimensions’ of coral reef biodiversity. • Six metrics predict complex multivariate information across 14 biodiversity variables. • They capture further variation in trait diversity, size structure and habitat quality. • Consistent metric predictions were found in two reef regions spanning >9000 transects.
McWilliam et al. (Wed,) studied this question.