Abstract
<jats:p>Habitat availability in intensive agroecosystems fluctuates rapidly across seasons and within days, with farmland birds responding by adjusting their habitat selection. Yet selection for these species is rarely assessed across the full diel cycle or year, and links to population indices remain largely unexplored. We used year-round, high-frequency GPS tracking to provide the first diel- and season-resolved assessment of space use and habitat selection in a declining farmland indicator species, the grey partridge (Perdix perdix), and tested whether landscape-scale selection predicts breeding pair density. We tracked 26 wild partridges (10 min schedule when active) in Austria (2024–2026), integrating locations with 5 m land-use mapping and monthly Enhanced Vegetation Index (EVI). We fitted seasonal Resource Selection Functions with four-way interactions among land use, EVI, edge distance, and diel period, with individual random effects. Realized habitat use was quantified via the Selection Effect and related to site- and year-level breeding pair density. Home ranges varied strongly by season and sex, being 5–6 times larger during the covey season (winter months when groups are formed) than breeding and, during pairing (pre-breeding), 13 times larger for males than females. Selection for uncultivated, grass-dominated areas persisted across seasons, favouring greater vegetation structure and field interiors, while hedgerow selection showed strong diel-dependent effects. Contrary to the literature, tree crops were the only consistently selected crop type; cereals and other arable crops were avoided. Breeding season Selection Effects best predicted between-site and interannual variation in pair density, explaining substantially more variance than other seasons. These results identify large, structurally suitable grasslands (>60 m across), well-sited hedgerows, and appropriately managed woody crops as key features supporting partridge populations, and demonstrate that coupling year-round selection with dynamic availability can reveal drivers of population variation of farmland birds.</jats:p>