Regional-scale patterns and predictors of species richness and abundance across twelve major tropical inter-reef taxa

Species richness and abundance are biodiversity metrics widely used to describe and estimate changes in biodiversity. Studies of marine species richness and abundance typically focus on one, or just a few, taxa. Consequently, it is currently not possible to understand the performance of predictors of species richness and abundance across marine taxa. Using a taxonomically comprehensive dataset of twelve major taxa of flora and fauna from eight phyla sampled from the inter-reef seabed region of the Great Barrier Reef, Australia, we used boosted regression trees to test the performance of fourteen environmental and spatial predictors of species richness and abundance. Sediment composition predicted richness best for all taxa: gravel contributed up to 39% relative influence for one group and all taxa had low richness in muddy habitats. Sea surface temperature, seabed current shear stress, depth and latitude were also influential predictors for species richness for eight groups. Sediment was frequently an influential predictor for abundance also, while distance to domain (reef/coast) and longitude were relatively influential for six taxa. Within-site richness was correlated between nearly all pairs of taxa, as was within-site abundance, however ρ values were low. Overall, model performance was high, explaining up to 62% deviance of species richness, and 38% of abundance. Typically, deviance explained was greater for richness than abundance and may indicate that some drivers of species richness operate independently of any effects on species richness mediated by their effect on abundance. Deviance explained differed most between richness and abundance for bryozoans (23.3% difference) and soft corals (15.2% difference). While sediments were consistently the best predictors across all taxa, the inconsistent influence of all other predictors across taxonomic groups, as well as the low correlation of richness and abundance across taxonomic groups, cautions against predicting regional patterns of species richness and abundance from few taxa.

Document type: