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

By A. Norberg, N. Abrego, F. Blanchet, F. Adler, Barbara J. Anderson, Jani V. Anttila, Miguel B. Araújo, Tad A. Dallas, David B. Dunson, Jane Elith and 25 more. Ecological Monographs, 2019

493 citations recorded by the source corpus.

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.

Subfields: Ecological Modeling, Genetics, Nature and Landscape Conservation

View Paper: https://www.semanticscholar.org/paper/4ae6b2a11bce56d17202753af2383545686fa3df

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Ecological Modeling

In Genetics

In Nature and Landscape Conservation

Citations in the corpus

Defining and evaluating predictions of joint species distribution models (2020, methodology): “To our knowledge, no JSDM studies have yet included conditional or conditional marginal predictions, although they have been identified as desirable avenues of research (Blanchet et al., 2020; Norberg et al., 2019; Zurell et al., 2019).”
Machine learning and deep learning—A review for ecologists (2022, unknown): “However, these trade-offs are difficult to predict a priori, which explains the common practice of comparing different algorithms for a given task (Faisal et al., 2010; Norberg et al., 2019 ; Pichler et al., 2020 ).”
A comparison between Ensemble and MaxEnt species distribution modelling approaches for conservation: A case study with Egyptian medicinal plants (2020, methodology): “The use of species distribution models (SDMs) for this purpose has advanced the understanding of many ecological issues and is a key tool in predicting species responses to environmental change (Elith et al., 2006; Elith and Leathwick, 2009; Norberg et al., 2019; Zimmermann et al., 2010).”
Joint species distribution modelling with the r‐package Hmsc (2020, background): “In a comparison among a large number of single-species and JSDMs, HMSC ranked first in terms of predictive performance (Norberg et al., 2019).”
A standard protocol for reporting species distribution models (2020, methodology): “In principle, this protocol should work for any empirically-based biodiversity model beyond single species distribution models, including e.g. community-level models (D’Amen et al. 2017, Norberg et al. 2019, Zurell et al. 2020) and models of functional composition (Wüest et al. 2018).”