Supraspecific Ecological Niche Models as a Tool for Predicting Burrowing Crayfish Habitat

ABSTRACT Sparse distributional data, taxonomic uncertainty, and cryptic diversity present significant barriers to building reliable species‐level ecological niche models (ENMs). Supraspecific ENMS hold potential to address these challenges by combining data across multiple species to capture shared habitat associations. This study investigated the application of supraspecific ENMs to predict the distribution of primary burrowing crayfish, a freshwater faunal group with taxonomic ambiguity and limited historical data. We developed four Maxent ENMs for three of the most widespread burrowing crayfish species in the United States— Creaserinus fodiens (Cottle 1863), Procambarus gracilis (Bundy 1876), and Lacunicambarus ludovicianus (Faxon 1884)—and a combined supraspecific dataset using spatially filtered historical records of all three species. We conducted field validation across 96 semi‐randomly selected sites in Missouri, United States, stratified by habitat suitability quartiles. Model performance was assessed using a suite of threshold dependent and independent metrics to evaluate discrimination and calibration. Modelling above the species level successfully identified suitable habitat for burrowing crayfish as a group. Field validation confirmed the model's predictive ability, with crayfish detected at 35 locations, including 11 sites where Lacunicambarus species not used in model training were found. Our Maxent models performed significantly better than expected by chance and showed that broadly suitable burrowing crayfish habitat consisted of shallow water tables, low elevation, low average maximum daily temperatures, and moderate annual precipitation. Our findings suggest that burrowing crayfish species may exhibit shared habitat associations, making them strong candidates for supraspecific modelling approaches, especially in regions with limited historical records or lineages with unresolved taxonomy and cryptic diversity. Further, by using species from three different genera, we demonstrated that supraspecific ENMs based on ecological traits, rather than phylogeny, offer a promising tool for extending distributional knowledge and informing conservation planning for data‐poor taxa.