| Summary: | The hypothesis of deep reef refuges predicts that deep reef ecosystems (> 30 m depth) may act as refuges for the biota of shallow reefs and recover it after human disturbances. Although the hypothesis has been postulated almost two decades, empirical evidence is lacking that support it for diverse biological groups, including corals. Little is known about how corals respond to different types of human disorders, whose magnitude and frequency vary considerably. In this work, I used coral communities on the coast of Paraíba to test the hypothesis of deep reef refuges. As a preamble, I reviewed the literature (110 studies) to identify theoretical and geographic gaps on coral responses to ocean acidification, climate change, overfishing, pollution, and disordered tourism. I ranked the studies according to the ocean, ecoregion, type of disturbance, level of biological organization, study approach, method of data collection, depth at which data were collected and type of coral response. I have found that the studies are concentrated in the Indo-Pacific (36.3%) and the Caribbean (31.9%) and have used an observational approach (59.1%) with SCUBA diving (37.2%) to assess the impact of ocean warming 55.4%) in coral communities (58.1%), especially in shallow waters (up to 27 m). These results reveal the scarcity of information on coral responses to pollution, tourism, overfishing and acidification, particularly in deep reef ecosystems and in ecoregions outside the Indo-Pacific and Caribbean. For the test of the hypothesis of the refuge, I used a community-level approach capable of partitioning the diversity in its alpha and beta components along a gradient of 3 to 61 m. Through SCUBA diving, I found 7 shallow reefs (<30 m) and 12 deep reefs (> 30 m). Contrary to expectations, shallow reefs presented greater range diversity than the deep ones (13 vs. 7 species); highest proportion of specialist species in depth (77% vs. 57%); similar alpha diversity but higher beta diversity of rare, typical and common species; and ...
|
|---|