Using airborne laser scanning to model potential abundance and assemblages of forest passerines

详细信息    查看全文

• 作者：Jö ; rg Mü ; ller ; Christoph Moning ; Claus Bä ; ssler ; Marco Heurich ; Rol ; Br ; l
• 关键词：Aerial photography ; Breeding birds ; LiDAR ; National park “ ; Bavarian Forest” ; Temperate montane forest ; Species-abundance models
• 刊名：Basic and Applied Ecology
• 出版年：2009
• 出版时间：October 2009
• 年：2009
• 卷：10
• 期：7
• 页码：671-681
• 全文大小：375 K

文摘

Modelling and forecasting of the distribution and abundance of organisms using environmental variables is a major focus of applied ecological research. High-resolution airborne laser scanning is a recently developed remote-sensing method that provides data that can be used as surrogates for the vertical structure of the vegetation. These data can be used for modelling the occurrence and abundance of species or species assemblages. Until now, few studies evaluated the potential of these data for use in such models, or compared the suitability of data obtained by airborne systems with data gained by alternative methods. To fill part of this gap, we used forest passerine bird species to evaluate airborne laser scanning data for statistical modelling of potential bird abundances and composition of assemblages. Birds were counted in a mixed montane forest, on 223 1-ha plots along four transects. In the same period, these areas were scanned using Light Detection And Ranging (LiDAR) to characterise canopy structure. Additionally, we used visual interpretations of aerial photographs and field measurements on the same plots to derive habitat variables for comparison. We found clear correlations between the LiDAR variables and the other two variable sets using canonical correlation analysis. With a few exceptions, predictive power of the LiDAR data set for modelling abundances of single species, with up to 40 % explained variance, was superior to that of the other two data sets. Models agreed with existing ecological knowledge for these species. For modelling of species composition with redundancy analysis, LiDAR was also superior to the other two data sets with more than 20 % unique contribution to the explained variance. Our results clearly showed that LiDAR provides valuable data for describing and modelling single species as well as assemblages of forest organisms.