Environmental correlates of Whinchat Saxicola rubetra breeding territory retention in a declining upland population.

Capsule: In a 461 km² unenclosed upland landscape in south-west England, long-term breeding Whinchat Saxicola rubetra persistence is more likely in areas characterized by a mixture of Bracken Pteridium aquilinum, ericaceous vegetation, and very low tree densities, situated in steep-sided valleys at mid-altitudes, further from agriculturally improved grassland or arable. Aims: To establish the correlates of Whinchat breeding territory retention between 1979 and 2015–2022 in a declining upland population. Methods: Historical Whinchat territories (identified in 1979) were revisited to assess range occupancy, producing a lost (n = 104) and retained (n = 60) sample. Territory retention probability was modelled at local (100 × 100 m) and broad (500 × 500 m) scales against remotely sensed data and field habitat data measured in 2022. Results: At the local scale, territory retention probability was greater in steeper valleys, further from arable or agriculturally improved grassland, peaked at five trees per ha in a quadratic response, and where Bracken cover was greater. Bracken cover effects were enhanced when ericaceous vegetation was also present. At the broad scale, retention probability was again greater in steeper valleys, with greater Bracken cover, and at very low tree densities in a quadratic response (but the latter was not important at the bottom of steep-sided valleys). Retention also peaked at around 350–400 m elevation in a quadratic response at the broad scale. Conclusions: In this population in south-west England, steep-sided valleys at 350–400 m, with a light scattering of trees, situated further from intensive enclosed farmland are more likely to retain Whinchats long term. Within these areas, a mixed Bracken and ericaceous field-layer should be encouraged, and blanket afforestation avoided, although lower densities of native trees appear to be more tolerated. [ABSTRACT FROM AUTHOR]