Background: Globally, shallow-water coral reef biodiversity is at risk from a variety of threats, some of which may attenuate with depth. Mesophotic coral ecosystems (MCEs), occurring from 30 to 40 m and deeper in tropical locations, have been subject to a surge of research this century. Though a number of valuable narrative reviews exist, a systematic quantitative synthesis of published MCE studies is lacking. We conducted a systematic review to collate mesophotic research, including studies from the twentieth century to the present. We highlight current biases in research effort, regarding locations and subject matter, and suggest where more attention may be particularly valuable. Following a notable number of studies considering the potential for mesophotic reefs to act as refuges, it is important to know how comprehensive these sources of recruits and organisms capable of moving to shallow water reefs may be. Methods: We search seven sources of bibliographic data with two search strings, as well as personal libraries. Articles were included if they contained species presence data from both shallower and deeper than 30 m depth on tropical coral reefs. Studies were critically appraised based on the number of species identified and balanced sampling effort with depth. Maximum and minimum depths per species were extracted from each study, along with study region and taxon. We quantified the degree of community overlap between shallow tropical reefs (< 30 m) and reefs surveyed at the same locations below 30 m. Proportions of shallow species, across all studied taxa, observed deeper than 30 m were used to generate log odds ratios and passed to a mixed-effects model. Study location and taxon were included as effect modifiers. Funnel plots, regression tests, fail safe numbers, and analysis of a high validity subgroup contributed to sensitivity analyses and tests of bias. Results: Across all studies synthesised we found two-thirds of shallow species were present on mesophotic reefs. Further analysis by taxon and broad locations show that this pattern is influenced geographically and taxonomically. Community overlap was estimated as low as 26% and as high as 97% for some cases. Conclusions: There is clear support for the hypothesis that protecting mesophotic reefs will also help to conserve shallow water species. At the same time, it is important to note that this study does not address mesophotic-specialist communities, or the ecological forces which would permit refuge dynamics. As we limit our analysis to species only present above 30 m it is also possible diversity found exclusively deeper than 30 m warrants protection in its own right. Further research into relatively ignored taxa and geographic regions will help improve the design of protected areas in future.

No abstract