Goatfishes and cross-validation – new publication

In an era of ongoing biodiversity crisis, it is critical to map biodiversity patterns in space and time for better informing conservation and management actions. Species distribution models (SDMs) are widely applied in various types of such biodiversity assessments. Cross-validation represents a prevalent approach to assess the discrimination capacity of a target SDM algorithm and determine its optimal parameters. Several alternative cross-validation methods exist; however, the influence of choosing a specific cross-validation method on SDM results remains unresolved. Scientists from China, Japan, Italy, and Hungary tested the performance of random versus spatial cross-validation methods for SDM using goatfishes as a case study. Ákos Bede-Fazekas, assistant professor in our department, is a co-author of the paper that summarized the results and is published in the D1-ranked scientific paper ‘Ecography’. Hongwei Huang and Zhixin Zhang from the Chinese Academy of Sciences and their co-authors found that the random versus spatial cross-validation methods resulted in different optimal model parameterization. While no significant difference existed in predictive performance between the random and spatial cross-validation methods, the two cross-validation methods yielded different projected spatial distribution and redistribution patterns of goatfishes under climate change exposure. The findings highlight that the choice of cross-validation method is an overlooked source of uncertainty in SDM studies and researchers should not rely on model performance to select the proper cross-validation approaches.