Exploring mesophotic depth refuges on coral reefs

Tropical coral reefs extend deeper than many people are aware. Though we may be familiar with zooxanthellate coral just below the surface, the deepest recorded are further than 150m down. This thesis focusses on mesophotic coral ecosystems (MCEs), beginning at 30m depth and extending to the deepest photosynthetic coral holobionts. Such understudied reefs require attention not just for their novelty. As the pressures which threaten shallow reefs attenuate with depth, MCEs may represent a refuge from anthropogenic stressors such as ocean warming. This DPhil occurred during the third global coral bleaching event, and considers whether the mesophotic zone could provide a refuge for shallow coral species into the future. Unfortunately, directly testing for deep coral refuges would require the destruction of shallow reefs and patient observation of any recovery. To side-step this process I consider a number of barriers to refuge dynamics. If MCEs are to act as a refuge for shallow species both depth zones must contain similar species pools. In chapter two I undertake a systematic review and meta-analysis to synthesise existing ecological surveys. I find that after half a century of research the mesophotic literature is still largely descriptive, likely because of the difficulties associated with accessing MCEs. I quantify the proportion of shallow species also observed deeper than 30m. The degree of community overlap varies by geographic region and with taxonomic groups, yet in some cases a substantial proportion of the shallow species pool can be found on MCEs as currently defined. Chapter three is based on primary research on the MCEs of Utila, Honduras. I consider a biological approach for recognising MCEs, as a global 30m depth limit is unlikely to capture variation in community zonation dependent on local conditions. I show that a change from shallow-specialist to depth-generalist taxa is largely coincident with current depth-based definitions, but accommodates site-specific variation in scleractinian communities. Locations with deeper maximum depth ranges for shallow species may be more likely to provide refuge from negative impacts. In addition, the recognition of mesophotic reefs as a biological assemblage rather than a depth zone will aid the design of experimental manipulations and allow logical comparisons between locations. Site-specific variation in species depth ranges, and the ability to survive both on MCEs and shallow reefs, are likely determined by the physiology of the coral holobiont. In chapter four I quantify metabolic change in the depth-generalist coral Agaricia lamarcki from two locations around Utila. Depth ranges are known to differ at these two sites. I find site-specific physiological profiles, with heterotrophy increasing with depth and a reduction in photosynthetic capacity at the site with the smallest depth range. I also find evidence to suggest greater niche conservation with depth at the site with the largest depth range. The same colonies assessed in chapter four undergo experimental manipulation in chapter five. Fragments were transplanted to a shallow or mesophotic station at a third site. Following a coral bleaching event on Utila a survival and bleaching analysis was conducted. Fragments located shallow were almost twice as likely to die, and were more likely bleached, than those at mesophotic depths. Site and depth of collection did not influence the chances of survival suggesting any reduction in bleaching thresholds for mesophotic colonies were unimportant in this case. It is important to note that depth can offer protection from other agents of mortality, such as water movement and coastal development, and this experimental design does not answer what drives differences in survival. Together these findings show that there is potential for MCEs to protect species also present on shallow reefs in the face of disturbance. Further, individuals collected from MCEs may be capable of surviving on shallow reefs when artificially moved. Substantial variation in depth ranges and metabolism, at least for Scleractinia, ensures that the likelihood of a deep reef refuge existing will always be site-specific. These findings will be best supported by further work correlating the patterns observed with local abiotic measurements, considering the fecundity of mesophotic reefs, and studying shallow water settlement and survival of mesophotic propagules in the face of different competitive regimes.