Habitat and water quality variables as predictors of community composition in an Indonesian coral reef: a multi-taxon study in the Spermonde Archipelago.

Assemblages of corals, sponges, foraminifera, sediment bacteria and sediment archaea were assessed at two depths in the Spermonde Archipelago. Our goal was to assess to what extent variation in composition could be explained by habitat and water quality variables. The habitat variables consisted of depth, substrate type and scleractinian coral cover while water quality variables were derived from ocean color satellite imagery, including the colored dissolved organic matter index (CDOM), chlorophyll-a (Chlor-a) and remote sensing reflectance at 645n m (Rrs_645). Together, habitat and water quality variables explained from 31% (sediment bacteria) to 80% (forams) of the variation in composition. The variation in composition of corals, sponges, forams and sediment archaea was primarily related to habitat variables, while the variation in composition of sediment bacteria was primarily related to water quality variables. Habitat and water quality variables explained similar amounts of variation in the composition of corals and sediment bacteria. CDOM (sponges, sediment bacteria and sediment archaea), Chlor-a (corals and forams) and Rrs_645 (sponges and forams) proved significant predictors of variation in composition for the studied taxa. In addition to water quality variables, all taxa responded to a range of habitat variables including depth and the percentage cover of various benthic life forms including coral cover variables, rubble and sand. Sand cover was the most important habitat variable for corals, sponges, sediment bacteria and sediment archaea. Coral life forms including the cover of branching and tabular corals were important habitat variables for sponges and forams. These results show marked differences in how various taxa respond to variation in habitat and water quality in the Spermonde Archipelago. Moreover, our results indicate that variables estimated from ocean color satellite imagery proved to be better predictors of variation in marine community composition than commonly-used proxies such as the distance offshore or distance to the nearest river.