The use of airborne lidar to assess avian species diversity, density, and occurrence in a pine/aspen forest

摘要

Vegetation structure is an important factor that influences wildlife-habitat selection, reproduction, and survival. However, field-based measurements of vegetation structure can be time consuming, costly, and difficult to undertake in areas that are remote and/or contain rough terrain. Light detection and ranging (lidar) is an active remote sensing technology that can quantify three-dimensional vegetation structure over large areas and thus holds promise for examining wildlife-habitat relationships. We used discrete-return airborne lidar data acquired over the Black Hills Experimental Forest in South Dakota, USA in combination with field-collected vegetation and bird data to assess the utility of lidar data in quantifying vegetation structural characteristics that relate to avian diversity, density, and occurrence. Indices of foliage height diversity calculated from lidar data were positively and significantly correlated with indices of bird species diversity, with the highest correlations observed when foliage height diversity categories contained proportionally more foliage layers near the forest floor (<5 m). In addition, lidar-derived indices of vegetation volume were significantly correlated with bird density. Using lidar-derived vegetation height data in combination with multispectral IKONOS data, we delineated five general habitat types within the study area according to the presence of prominent vegetation layers at lower levels of the forest and predominant tree type (deciduous or conifer). Habitat type delineations were tested by examining the occurrence and relative density of two bird species common to the study area that prefer lower level vegetation for foraging and nesting. Dark-eyed Juncos were significantly associated with the 0.5-2.0 m high vegetation layer in pine-dominated stands, and Warbling Vireos were significantly associated with this same layer in aspen-dominated stands. These results demonstrate that discrete-return lidar can be an effective tool to remotely quantify vegetation structural attributes important to birds, and may be enhanced when used in combination with spectral data. (C) 2008 Elsevier Inc. All rights reserved.