Remotely sensed habitat variables are poor surrogates for functional traits of rocky reef fish assemblages

Conservation planning often relies on measures such as species richness and abundance to prioritize areas for protection. Nonetheless, alternative metrics such as functional traits have recently been shown to be useful complementary measures for detecting biological change. Timely conservation planning often precludes the collection of such detailed biological data relying instead on remotely-sensed habitat mapping as a surrogate for diversity. While there is evidence that habitat maps may predict taxonomic species richness and diversity in some coastal ecosystems, it is unknown whether similar strong relationships exist for functional traits and functional multimetrics. We compared the performance of physical habitat structural complexity obtained from high definition swath mapping in explaining variation in traditional taxonomic metrics as well as functional traits (e.g., maximum length, trophic level, gregariousness) and functional multimetrics (e.g., functional richness, dispersion) of fish assemblages. Reef complexity measures were good surrogates for fish species richness and abundance but not for functional traits or multimetrics, except functional richness at the scale of 1 m. Remotely sensed habitat maps may not be a good surrogate for predicting functional traits and multimetrics of fish assemblages, and must be used with caution when maximizing such aspects of assemblages is a priority for conservation planning.