ENMeval 2.0: Redesigned for customizable and reproducible modeling of species’ niches and distributions

Abstract

1. Quantitative evaluations to optimize complexity have become standard for avoiding overfitting of ecological niche models (ENMs) that estimate species' potential geographic distributions. ENMeval was the first R package to make such evaluations (often termed model tuning) widely accessible for the Maxent algorithm.

Species distribution modeling to inform transboundary species conservation and management under climate change: promise and pitfalls

“…The regularization multiplier limits the complexity of the model to generate a less localized prediction (Phillips and Dudík 2008): the default value of 1 tends to allow for more complexity and tends to lead to overfit models. Higher regularization values penalize complexity, so the best practice is to try a range of regularization values and then choose an optimal model for the species based on a set of evaluation metrics (e.g., see Kass et al 2021b). Similarly, Maxent's default settings allow for multiple feature classes in same model, based on the number of occurrence records, which can also lead to model overfitting and overcomplexity depending on the particular biological system.…”

“…The new opensource version release of Maxent (Phillips et al 2017), as well as other open-source tools that facilitate best practices in model tuning and parameterization like Wallace (Kass et al 2018), are changing this landscape to lower entry barriers into robust application of SDM for conservation. For example, Wallace implements two state-of-the art R packages spThin (Aiello-Lammens et al 2015) and ENMEval (Kass et al 2021b) that facilitate some of the best practices outlined above to avoid pitfalls, and in a user-friendly graphical user interface (GUI) environment. The application guides modelers through a complete analysis, from the acquisition of data to choosing and evaluating optimal models, to visualizing model predictions on an interactive map, thus bundling complex workflows into a single, streamlined interface.…”

Driven to the edge: Species distribution modeling of a Clawed Salamander (Hynobiidae: Onychodactylus koreanus) predicts range shifts and drastic decrease of suitable habitats in response to climate change

“…under two cross-validation schemes: 10-fold cross-validation and spatial block cross-validation. While the former method is commonly used to generate and test SDMs, we also tested the latter method as its use has been recommended when model transfer to different environmental conditions is the goal (Muscarella et al, 2014;Kass et al, 2021). We used 10,000 target group background points to generate candidate models under each cross-validation method.…”

“…First, we selected the model with a combination of feature classes and regularization multiplier that had the lowest AICc value, describing model fit and complexity (Warren & Seifert, 2011;Muscarella et al, 2014). Second, we selected the model that had the lowest 10% omission rate as well as the highest test AUC (sequential selection method; Kass et al, 2021). For both crossvalidation methods, the optimal parameters based on AICc were the LQHPT features combined with a regularization multiplier of 1.5.…”

“…According to these criteria, we selected the model parameters with the lowest AICc (LQHPT 1.5) to calibrate the final current habitat suitability model. We conducted the model tuning procedure using R packages dismo (Hijmans et al, 2017) and ENMeval (Muscarella et al, 2014;Kass et al, 2021).…”