Explainable artificial intelligence enhances the ecological interpretability of black‐box species distribution models

Abstract

Species distribution models (SDMs) are widely used in ecology, biogeography and conservation biology to estimate relationships between environmental variables and species occurrence data and make predictions of how their distributions vary in space and time. During the past two decades, the field has increasingly made use of machine learning approaches for constructing and validating SDMs. Model accuracy has steadily increased as a result, but the interpretability of the fitted models, for example the relative importance of predictor variables or their causal effects on focal species, has not always kept pace. Here we draw attention to an emerging subdiscipline of artificial intelligence, explainable AI (xAI), as a toolbox for better interpreting SDMs. xAI aims at deciphering the behavior of complex statistical or machine learning models (e.g. neural networks, random forests, boosted regression trees), and can produce more transparent and understandable SDM predictions. We describe the rationale behind xAI and provide a list of tools that can be used to help ecological modelers better understand complex model behavior at different scales. As an example, we perform a reproducible SDM analysis in R on the African elephant and showcase some xAI tools such as local interpretable model‐agnostic explanation (LIME) to help interpret local‐scale behavior of the model. We conclude with what we see as the benefits and caveats of these techniques and advocate for their use to improve the interpretability of machine learning SDMs.

Urban Heat Island and Its Regional Impacts Using Remotely Sensed Thermal Data—A Review of Recent Developments and Methodology

“…Deep learning methods are a powerful complement to classical machine learning tools and other analysis strategies, and have been used in a number of applications in UHII and remotely sensed image analyses [241]. The explainable artificial intelligence in UHII and UHIRIP modeling has become more and more important [242]. Interpretable machine learning methods either target a direct understanding of the model architecture (i.e., model-based interpretability) or interpret the model by analyzing the model behavior (post hoc interpretability) [242].…”

“…The explainable artificial intelligence in UHII and UHIRIP modeling has become more and more important [242]. Interpretable machine learning methods either target a direct understanding of the model architecture (i.e., model-based interpretability) or interpret the model by analyzing the model behavior (post hoc interpretability) [242].…”

Challenges and opportunities of species distribution modelling of terrestrial arthropod predators

“…Once this relationship is determined, the model is used to characterize the ecological niche of a given species by projecting a probability surface into a geographical space to represent its potential range of distribution (Guisan et al 2017). These models can be construed using a wide range of algorithms, from simple logistic regression up to sophisticated techniques based on machine learning (Elith et al 2011, Ryo et al 2020 and other artificial intelligence methods (Cardoso et al 2020a) .…”

Challenges and opportunities of species distribution modelling of terrestrial arthropod predators

“…Natural history is indeed entering its next-generation phase (Anderson et al, 2021;Jarić et al, 2020;Tosa et al, 2021), one characterized by increasingly available data (not only distribution data but also species traits and phylogenies) that can be routinely integrated in our modelling exercises. This is made possible by a parallel development of new methods, ranging from computationally fast multispecies modelling platforms (Pichler & Hartig, 2021) to flexible techniques able to account for traits (phenotypic plasticity) and genetic data in making predictions (Brewer et al, 2016;Bush et al, 2016;Garzón et al, 2019), along with tools to ease model interpretability (Ryo et al, 2021). As entomology is entering a nextgeneration phase too (Høye et al, 2021;Liu, Clarke, et al, 2020;Liu, Blackburn, et al, 2020), in all likelihood these advances will soon cascade to positively affect our understanding of the distribution of less studied arthropod groups.…”