A simple framework for estimating potential distributions of amphibious marine species and implications for conservation

Abstract

Due to their complexity, coral reefs are difficult to study especially when considering the role that the interplay between the terrestrial and marine environments has in shaping distribution of marine, terrestrial, and amphibious species. Many organisms live in remote areas of the ocean and inhabit both terrestrial and marine environments. Such amphibious lifestyle poses analytical difficulties due to broad distribution and scale of coral reefs. Ecological niche modeling is a widely used technique that allows to estimate the environmental set of conditions (niche) in which organisms can survive and reproduce. Estimating the distributions of species with complex life histories (i.e., dependent on various natural resources) at broad geographic scales is crucial, as many of these taxa are threatened (i.e., amphibians, aquatic reptiles, birds, and mammals). However, distribution estimates of such species remain challenging; thus, here we propose an approach to account for marine and terrestrial environmental domains to estimate the distribution of amphibious species. We also test whether inclusion of both environments leads to improved estimates of these species’ distributions. First we calibrated ecological niche models for marine and terrestrial domains separately, and subsequently we outlined a method to combine the marine–terrestrial potential distributions by integrating estimates of the two ecological niches into a single predictive model. Our ecological niche models produced inaccurate distribution predictions of species with amphibious life histories when only one of the environments was used in model calibration. When both aquatic and terrestrial environments were included, our models predicted narrower and more accurate potential distributions. Accounting for the dual environments involved in shaping the niches of amphibious species and their distributions is essential for studying the ecology and proposing conservation management actions for the species studied here. Models that take into account only a subset of the environmental factors are prone to overestimating species’ distributions and should be interpreted with caution.