Regional mapping of vegetation structure for biodiversity monitoring using airborne lidar data

- Regional mapping of vegetation structure for biodiversity monitoring has rarely been done
- Cluster analysis based on lidar-derived height and cover metrics reflected natural grouping of vegetation structure
- Structure classes were distributed differently between natural subregions
- Human disturbance created unique patterns of vegetation structure

Vegetation structure is identified as an important biodiversity indicator providing the physical environment that generates, supports, and maintains forest biodiversity. Airborne lidar systems (light detection and ranging) have the capacity to accurately measure three-dimensional vegetation structure, and have been widely used in wildlife habitat mapping and species distribution modeling. Large-area structural inventories using lidar-derived variables that characterize generic habitat structure have rarely been done, yet would be helpful for guiding biodiversity monitoring and conservation assessments of species at regional levels. This study provides a novel approach for processing regional-scale lidar data into categorical classes representing natural groupings of habitat structure. We applied cluster analysis on six lidar-derived habitat-related variables to classify vegetation structure into eight classes for the forested areas of ten natural subregions in boreal and foothill forests in Alberta, Canada. Structure classes were compared across different natural subregions and under anthropogenic/non-anthropogenic disturbance regimes. We found that the Lower Foothills Natural Subregions had the most complex vegetation structure, and wildfire was the most prevalent disturbance agent for all classes except for the rarest class (i.e. stands with high standard deviations of height and low canopy cover) which was more heavily altered by timber harvesting. This data product provides continuous, regional mapping of vegetation structure directly measured from lidar, with a spatial resolution (30 m) relatively finer than what was provided by polygon-based forest inventories. This vegetation structure classification and its associated spatial distribution address the fundamental issue of habitat structure in biodiversity monitoring. It can serve as a base layer used together with species and land cover data for forest resources planning, species distribution and animal movement modeling, as well as prioritization of conservation efforts on critical habitat structures.