Environmental factors driving the distribution of the tropical coral Pavona varians: Predictions under a climate change scenario

Abstract

Climate change causes shifts in the geographical distribution boundaries of many organisms. Modelling techniques predict the distribution of species by relating climatic and physical factors with species' presence records, including potential extinction areas and new potential areas of colonization, under predicted climatic scenarios. In this study, we initially explored which environmental variables most influenced the distribution of Pavona varians, a hermatypic coral from the equatorial Indian and the Pacific Ocean, which is categorized as ‘Least Concern’ by the UICN. The most influential variables were the minimum and maximum sea surface temperature, the diffuse water attenuation and the cloud cover. These variables were used to predict habitat suitability of P. varians under a current and a future (A1B IPPC) scenario using MaxEnt. Despite P. varians is an opportunistic species, with a well‐known resistance to environmental stress, the model predicted a massive decline in the suitability of its habitat in all areas by the year 2100. The information obtained by this study can be used to support the conservation decision making process to improve the preservation of the species.

Predictions Based on Different Climate Change Scenarios: The Habitat of Typical Locust Species Is Shrinking in Kazakhstan and Xinjiang, China

“…To avoid model overfitting, we first reduced spatial autocorrelation by spatially filtering occurrence data within the study area using the SDMtoolbox tool “Spatially Sparse Occurrence Data” to eliminate spatial clustering of locations for model calibration and evaluation [ 41 ]. Spatial filtering was used to minimize omission errors and commission errors, and ultimately, 109 occurrence records for the three locust species were used for potential locust distribution predictions based on historical and future climate conditions ( Figure 1 A).…”

“…Variables with correlations greater than 0.75 were removed based on variable contribution value analysis. VIFs were used for assessments of the covariance between variables [ 41 , 42 ]; variables with VIF < 10 were retained, and ultimately, nine environmental variables were used for modeling, including seven bioclimatic variables ( Table 1 ).…”

The potential habitat of desert locusts is contracting: predictions under climate change scenarios

“…First, based on Spearman’s correlation coefficient and jackknife analysis, the 23 initially selected environmental variables were rescreened. When Spearman’s |r| between a pair of variables was >0.75, we retained the variable that had a higher percentage contribution to the MaxEnt model for subsequent analysis ( Wang et al, 2019 ; Rodriguez et al, 2020 ). Then, a multicollinearity test was performed on the selected environmental variables with VIF, and the VIF values of all the environmental variables were less than 5.…”

“…Climate change, especially global warming, affects the geographical distribution of species on earth ( Parmesan & Yohe, 2003 ; Noce, Caporaso & Santini, 2019 ; Rodriguez et al, 2020 ). The life cycle of desert locusts usually needs to go through three stages (eggs, nymphs (hoppers), and adults), and the time it takes to transition from one stage to another is highly dependent on weather patterns ( Rainey et al, 1979 ; Symmons & Cressman, 2001 ; Cressman & Stefanski, 2016 ).…”

“…As a machine learning algorithm, the maximum entropy (MaxEnt) model is widely used in species distribution modeling. It quantifies the relationship between species and their environment based on the existence of species records, thereby generating potential habitat suitability maps and allowing inferences under future climate scenarios ( Phillips et al, 2006 ; Phillips & Dudík, 2008 ; Elith et al, 2011 ; Fois et al, 2018 ; Rodriguez et al, 2020 ). Compared with other statistical methods, MaxEnt provides higher predictive performance when species-absence data are not available ( Elith et al, 2011 ; Merow, Smith & Silander, 2013 ).…”