Fit to flutter: Species-specific models for alpine butterflies

Modeling the habitat use and distribution of rare insects presents unique challenges, especially for alpine butterfly species with small populations and restricted ranges. Our research focuses on two such species: the White Mountain fritillary (Boloria chariclea montinus) and the White Mountain arctic (Oeneis melissa semidea), both endemic to New Hampshire’s Presidential Range. Although they inhabit the same region, their differences in habitat specificity and data availability require distinct modeling approaches to accurately assess their ecological needs and vulnerabilities.
For the White Mountain fritillary, a species with limited occurrence records, we used an Ensemble of Small Models (ESMs) through the R package ‘biomod2’. This method prevents overfitting while incorporating climate, topographic, and other abiotic variables to identify potential shifts in distribution under future climate scenarios.
In contrast, the White Mountain arctic, which has more documented sightings but a narrower range, was analyzed using a Resource Selection Probability Function (RSPF) with the ‘ResourceSelection’ R package. This fine-scale analysis compared available versus used habitat, revealing a broader used habitat type than previously reported.
These findings underscore the need for species-specific modeling in conservation entomology. Our models suggest the fritillary faces substantial habitat loss, and the arctic may be utilizing a wider range of alpine habitat types. By tailoring analytical methods to each species' data and ecology, this research offers practical tools for managing rare Lepidoptera in rapidly changing environments and supports targeted conservation strategies, including listing decisions and habitat restoration efforts.