Comparing species distribution models: a case study of four deep sea urchin species

Abstract

There is an increasing demand for biodiversity mapping to address new challenges in the management of marine ecosystems. Species distribution models are a key tool in supplying part of this information. However, the use of these models in the marine environment is still developing and the reasons for the underlying use of different methodological approaches are not always clear. In this work, we compared four different statistical techniques: the ecological niche factor analysis (ENFA), the MAXimun ENTropy algorithm (MAXENT), general additive Models (GAMs), and Random Forest. ENFA and MAXENT were applied using presence-only data whereas GAM and Random Forest used presenceabsence data. As a case study, we used four deep sea urchin species: Centrostephanus longispinus, Coelopleurus floridanus, Stylocidaris affinis, and Cidaris cidaris. The distribution of the studied sea urchins showed strong bathymetric segregation. Depth was the most important variable, followed by reflectivity and slope. The correlations between the predictive outputs of the models were similar between GAM, Random Forest and MAXENT, and lower for ENFA. Models using presence/absence data showed the highest scores in the four species, significantly outperforming ENFA in most of the cases, although differences with MAXENT were significant in only one species.

Species distribution models of the Spotted Wing Drosophila (Drosophila suzukii, Diptera: Drosophilidae) in its native and invasive range reveal an ecological niche shift

“…We modelled the potential distribution of D. suzukii in three separate models using (a) occurrences in the native range (SDM native ), (b) occurrences in the European invasive range (SDM Europe ), and (c) all present occurrences (SDM global ). For this purpose, we used the software Maximum Entropy Modelling ( Maxent v.3.4.1; Phillips et al., ; http://biodiversityinformatics.amnh.org/open_source/maxent/) because of its superior accuracy with presence‐only records compared to many other SDMs (González‐Irusta et al., ). In addition, Maxent produces response curves of each environmental predictor, which is essential in interpreting and comparing model performances.…”

The present is the key to the past: linking regime shifts in kelp beds to the distribution of deep‐living sea urchins

“…In contrast, González‐Irusta et al. () found that depth, bottom type, and slope were the most important predictors of suitable habitat for other sea urchins ( Centrostephanus longispinus , Coelopleurus floridanus , Stylocidaris affinis , Cidaris cidaris ) at 158–1,663 m depth on a seamount near the Canary Islands, where BPI, curvature and aspect were not important predictors.…”

Assessment of the effect of climate changes in the Late Pleistocene and Holocene on niche conservatism of an arvicolid specialist

“…The SDMs were built with three robust modelling algorithms: Generalized linear model (GLM), RF (R library randomForest 73 ) and Maximum Entropy (Maxent 74 ) (R library dismo 75 ). We specifically included GLM because of its flexibility to control all the factors involved in the model such as interactions or variable fitting, while Maxent and RF were used for their good predictive performance 7 , 76 . All models were built using current occurrence data of the species plus a background dataset (with 20% of the points in the global territory considered) for Maxent and weighted background for the GLM algorithm, whereas the RF and GLM were developed with a set of pseudoabsences.…”