A comprehensive evaluation of predictive performance of 33 species distribution models at species and community levels

Abstract

A large array of species distribution model (SDM) approaches has been developed for explaining and predicting the occurrences of individual species or species assemblages. Given the wealth of existing models, it is unclear which models perform best for interpolation or extrapolation of existing data sets, particularly when one is concerned with species assemblages. We compared the predictive performance of 33 variants of 15 widely applied and recently emerged SDMs in the context of multispecies data, including both joint SDMs that model multiple species together, and stacked SDMs that model each species individually combining the predictions afterward. We offer a comprehensive evaluation of these SDM approaches by examining their performance in predicting withheld empirical validation data of different sizes representing five different taxonomic groups, and for prediction tasks related to both interpolation and extrapolation. We measure predictive performance by 12 measures of accuracy, discrimination power, calibration, and precision of predictions, for the biological levels of species occurrence, species richness, and community composition. Our results show large variation among the models in their predictive performance, especially for communities comprising many species that are rare. The results do not reveal any major trade‐offs among measures of model performance; the same models performed generally well in terms of accuracy, discrimination, and calibration, and for the biological levels of individual species, species richness, and community composition. In contrast, the models that gave the most precise predictions were not well calibrated, suggesting that poorly performing models can make overconfident predictions. However, none of the models performed well for all prediction tasks. As a general strategy, we therefore propose that researchers fit a small set of models showing complementary performance, and then apply a cross‐validation procedure involving separate data to establish which of these models performs best for the goal of the study.

Collinearity in ecological niche modeling: Confusions and challenges

“…Though we focused on Maxent in our study with the aim to capture a common practice in ENM literature, many other algorithms are used in ENM literature (e.g., 33; Norberg et al, ). The vulnerability to degree of predictor collinearity should vary with and depend on the mechanisms in each algorithm.…”

From topography to hydrology—The modifiable area unit problem impacts freshwater species distribution models

“…The TSS values range from −1 to 1, where values close to 1 indicate a good to very good fit and values of 0 and lower indicate model performance not better than random. The assignment of weights to the models allows to automatically select the models with the best data fit for the ensemble, without completely discarding results from all other algorithms (Norberg et al, ). The ensemble model was then used to predict the final probabilistic habitat suitability for each species across all subcatchments at a given spatial resolution.…”

Using hierarchical joint models to study reproductive interactions in plant communities

“…The HMSC model has been implemented in a Bayesian framework, has been shown to perform well in terms of parameter estimation and prediction (Tikhonov et al ), and was ranked first in a recent comparison among similar models (Norberg et al, ). In the following, we will demonstrate applications of the HMSC framework to data on flower visitation to a set of plant species at a plot level recorded during multiple censuses at each plot (Figure a).…”

“…In HMSC, the response to covariates (fixed part of linear predic- The HMSC model has been implemented in a Bayesian framework, has been shown to perform well in terms of parameter estimation and prediction (Tikhonov et al 2019), and was ranked first in a recent comparison among similar models (Norberg et al, 2019).…”