Fine-scale habitat selection in a grassland mosaic: electivity-based vegetation-community preferences of the endangered Hungarian meadow viper ( <i>Vipera ursinii rakosiensis</i> )

Abstract Quantifying habitat preference of a taxa is a key step in conservation practices, particularly for rare and strictly protected species where management decisions must be based on robust evidence. We compared three widely used indices, Strauss’ linear index (L), Ivlev’s electivity index (E), and Vanderploeg &amp; Scavia’s relativised electivity (E*), using the Hungarian meadow viper (Vipera ursinii rakosiensis), a rare grassland specialist reptile of the Carpathian Basin, Europe, as a model species. Presence data from 2010–2023 in the Kiskunság region, Hungary, were combined with pseudo-absence points generated proportionally to survey effort within occupied grid cells. Habitat availability was derived from high-resolution vegetation community maps, and a spatially balanced resampling approach (1,000 iterations) was applied to reduce bias and estimate uncertainty in selection estimates. All three indices consistently identified closed sand steppes, mesotrophic wet meadows, uncharacteristic mesic meadows and tall-herb communities, uncharacteristic dry and semi-dry grasslands, and open sand steppes as preferred habitats, while several other habitat types were consistently avoided. Differences among indices were generally small but occasionally altered the ranking of preferred habitat types, demonstrating that metric choice can influence ecological interpretation. Our findings refine the understanding of habitat preferences for one of Europe’s most threatened reptile and highlight the importance of method selection in guiding conservation priorities for rare species.