Ecological niche modeling in Maxent: the importance of model complexity and the performance of model selection criteria

TLDR

Maxent is a widely used tool for species distribution modeling, but guidance on setting its L1 regularization level is scarce and the consequences of mis‑specified model complexity remain largely unknown. This study demonstrates how information‑criterion approaches can be used to set Maxent regularization and compares these models to those selected by other common criteria. We assess model performance by generating occurrence data from a known true Maxent model and evaluating several metrics of quality and transferability. Our results show that models that are too complex or too simple reduce the ability to infer habitat quality, variable importance, and transferability, while information criteria offer significant advantages over the methods commonly used in the literature.

Abstract

Maxent, one of the most commonly used methods for inferring species distributions and environmental tolerances from occurrence data, allows users to fit models of arbitrary complexity. Model complexity is typically constrained via a process known as L1 regularization, but at present little guidance is available for setting the appropriate level of regularization, and the effects of inappropriately complex or simple models are largely unknown. In this study, we demonstrate the use of information criterion approaches to setting regularization in Maxent, and we compare models selected using information criteria to models selected using other criteria that are common in the literature. We evaluate model performance using occurrence data generated from a known "true" initial Maxent model, using several different metrics for model quality and transferability. We demonstrate that models that are inappropriately complex or inappropriately simple show reduced ability to infer habitat quality, reduced ability to infer the relative importance of variables in constraining species' distributions, and reduced transferability to other time periods. We also demonstrate that information criteria may offer significant advantages over the methods commonly used in the literature.