Date Published: May 21, 2019
Publisher: Public Library of Science
Author(s): Umberto Diego Rodrigues de Oliveira, Paula Braga Gomes, Ralf Tarciso Silva Cordeiro, Gislaine Vanessa de Lima, Carlos Daniel Pérez, Shahid Farooq.
Climate and environmental conditions are determinant for coral distribution and their very existence. When changes in such conditions occur, their effects on distribution can be predicted through species distribution models, anticipating suitable habitats for the subsistence of species. Mussismilia harttii is one of the most endangered Brazilian endemic reef-building corals, and in increasing risk of extinction. Herein, species distribution models were used to determine the present and future potential habitats for M. harttii. Estimations were made through the maximum entropy approach, predicting suitable habitat losses and gains by the end of the 21st century. For this purpose, species records published in the last 20 years and current and future environmental variables were correlated. The best models were chosen according to the Akaike information criterion (AIC) and evaluated through the partial ROC (AUCratio), a new approach which uses independent occurrence data. Both approaches showed that the models performed satisfactorily in predicting potential habitat areas for the species. Future projections were made using the International Panel on Climate Change (IPCC) scenarios for 2100, with different levels of greenhouse gas emission. Representative Concentration Pathways (RCPs) were used to model the Future Potential Habitat (FPH) of M. harttii in two different scenarios: stabilization of emissions (RCP 4.5) and increase of emissions (RCP 8.5). According to the results, shallow waters to the south of the study area concentrate most of the current potential habitats for the species. However, in future scenarios, there was a loss of suitable areas in relation to the Current Potential Habitat (RCP 4.5 46% and RCP 8.5 59%), whereas there is a southward shift of the suitable areas. In all scenarios of FPH, the temperature was the variable with the greatest contribution to the models (> 35%), followed by the current velocity (> 33%) and bathymetry (>29%). In contrast, there is an increase of deep (50–75 m) suitable areas FPH scenarios, mainly in the southern portion of its distribution, at Abrolhos Bank (off Espirito Santo State). These deeper sites might serve as refugia for the species in global warming scenarios. Coral communities at such depths would be less susceptible to impacts of climate change on temperature and salinity. However, the deep sea is not free from human impacts and measures to protect deeper ecosystems should be prioritized in environmental policies for Brazilian marine conservation, especially the Abrolhos Bank, due to its importance for M. harttii.
Coral reefs are one of the most ecologically valuable ecosystems on earth  providing a number of ecosystem services , such as shelter for associated fishes  and crustaceans [4, 5, 6, 7], also serving as substrate for coralline algae [8, 9]. Stable water conditions are determinant for the maintenance of living corals on reefs . However, effects of climate changes put at least 50% of shallow-water species in critical risk of extinction in the next 20 years [11, 12].
The variables used to model CPH (set 12) were,: in decreasing order of contribution: long-term minimum mean current velocity (42.7%); bathymetry (31.9%) and maximum temperature (25.4%) (S3 Appendix).