The Changing Forest for the Birds: Using Remote Sensing Time Series to Examine Bird Habitat, Occurrence and Niche

Abstract

Time series of airborne laser scanning (ALS) structural variables and Landsat-derived spectral variables can characterize environmental attributes where species occur, providing insight into their habitat, occurrence, and niche, and differences across species, space and time. This study explored how these data can contribute to understanding the Eurasian Blue Tit (Cyanistes caeruleus; hereafter ‘Blue Tit’), Common Chaffinch (Fringilla coelebs; hereafter ‘Chaffinch’), Common Chiffchaff (Phylloscopus collybita, hereafter ‘Chiffchaff’) and Willow Warbler (Phylloscopus trochilus) habitat, occurrence and niche in woodlands in the United Kingdom. Blue Tit and Chaffinch are woodland generalists, while Chiffchaff is an early successional woodland specialist and Willow Warbler is a mature woodland specislist. First, three ecologically important vegetation strata were characterized using ALS to develop species distribution models (SDMs) and habitat suitability indices for each bird species. Generalist species, Blue Tit and Chaffinch, were spatially and temporally widespread, while specialists, like Willow Warbler and Chiffchaff, were restricted as habitat suitability varied with changes in forest structure. Second, pixel- and object-level fusion of Landsat and ALS data was used to predict structural attributes, and to model and predict species occurrence. While Landsat predicted structural attributes, the errors negatively impacted occurrence and prediction models. Landsat SDMs were best for modelling Blue Tit and Chaffinch, while object-level fusion SDMs were best for Chiffchaff and Willow Warbler. ALS data yielded the highest prediction accuracy for Chaffinch and Chiffchaff, whereas ALS, Landsat and object-level fusion predicted Blue Tit and Willow Warbler occurrence with comparably high accuracy. Third, whether a species’ niche was conserved or divergent across successional time was assessed by characterizing the niche of each species using ALS at two seral stages. While differences along the margins of niche space were observed, similarity and equivalency statistics support conservatism, except for niche expansion and stability metrics for Chiffchaff. Overall, differences in habitat and niche reflect species’ preferences, generality or specificity, and flexibility across environmental attributes. This research demonstrated that remote sensing time series effectively capture ecologically meaningful environmental attributes of woodland, thereby offering (i) insights into Blue Tit, Chaffinch, Chiffchaff and Willow Warbler ecology, and (ii) methodological guidance for modelling habitat, occurrence, and niche.