Airborne laser scanning and spectral remote sensing give a bird's eye perspective on arctic tundra breeding habitat at multiple spatial scales

Abstract

Abstract The effects of climate change are more acute in the Arctic than any other region, and as such, arctic tundra wildlife habitats are changing in ways that are not yet well understood. Remote sensing tools are capable of assessing dynamics in wildlife-habitat associations over a wide range of spatial scales and in a variety of ecosystems. As already well-established in other ecosystems, Light Detection and Ranging (LiDaR) technology has the potential to greatly expand our understanding of tundra wildlife-habitat associations because unlike the most commonly used spectral vegetation index - the Normalized Difference Vegetation Index (NDVI) - LiDaR directly quantifies three-dimensional (3-D) vegetation structure. Only recently has airborne laser scanning (ALS) technology evolved to having vertical resolution and point densities high enough to quantify small differences in vegetation structure, such as characterize arctic tundra ecosystems. Our objective was to employ ALS metrics, airborne spectral data (for NDVI), and Landsat 7 ETM + data (for snow cover) to determine the relative importance of a suite of ecological characteristics - considered at multiple spatial scales - on habitat use and reproductive success of two of the most common migratory songbird species breeding in northern Alaska. Our most important findings are that (1) combining ALS metrics, but not NDVI, with Landsat derived snow cover data provided useful information, revealing that while Gambel's white-crowned sparrows use breeding territories in areas with high canopy volume and large snow-free areas, Lapland longspurs will establish territories in areas with low canopy volume, small patches of snow-free tundra, and minimal surface wetness and; (2) while habitat characteristics were important determinants of habitat use for both species at the territory scale, those in the immediate vicinity of nest-sites were not important. Contrary to expectation, we also found that the reproductive success of both species was unaffected by variation in our hypothesized metrics of shelter and food availability (canopy volume, standard deviation of micro-terrain height) at both the nest-site or territory scales. Our study is the first arctic demonstration of how ALS yields novel insights into wildlife-habitat associations, suggesting it has great potential in other ecosystems with similarly small - yet often ecologically important - ranges in 3-D vegetation structure.