Assessing the ecological application of lasergrammetric techniques to measure fine-scale vegetation structure

Revealing microhabitat requirements of an endangered specialist lizard with LiDAR

A central principle of threatened species management is the requirement for detailed understanding of species habitat requirements. Difficult terrain or cryptic behaviour can, however, make the study of habitat or microhabitat requirements difficult, calling for innovative data collection techniques. We used high-resolution terrestrial LiDAR imaging to develop three-dimensional models of log piles, quantifying the structural characteristics linked with occupancy of an endangered cryptic reptile, the western spiny-tailed skink (Egernia stokesii badia). Inhabited log piles were generally taller with smaller entrance hollows and a wider main log, had more high-hanging branches, fewer low-hanging branches, more mid- and understorey cover, and lower maximum canopy height. Significant characteristics linked with occupancy were longer log piles, an average of three logs, less canopy cover, and the presence of overhanging vegetation, likely relating to colony segregation, thermoregulatory requirements, and foraging opportunities. In addition to optimising translocation site selection, understanding microhabitat specificity of E. s. badia will help inform a range of management objectives, such as targeted monitoring and invasive predator control. There are also diverse opportunities for the application of this technology to a wide variety of future ecological studies and wildlife management initiatives pertaining to a range of cryptic, understudied taxa.

Identifying Tree-Related Microhabitats in TLS Point Clouds Using Machine Learning

Tree-related microhabitats (TreMs) play an important role in maintaining forest biodiversity and have recently received more attention in ecosystem conservation, forest management and research. However, TreMs have until now only been assessed by experts during field surveys, which are time-consuming and difficult to reproduce. In this study, we evaluate the potential of close-range terrestrial laser scanning (TLS) for semi-automated identification of different TreMs (bark, bark pockets, cavities, fungi, ivy and mosses) in dense TLS point clouds using machine learning algorithms, including deep learning. To classify the TreMs, we applied: (1) the Random Forest (RF) classifier, incorporating frequently used local geometric features and two additional self-developed orientation features, and (2) a deep Convolutional Neural Network (CNN) trained using rasterized multiview orthographic projections (MVOPs) containing top view, front view and side view of the point’s local 3D neighborhood. The results confirmed that using local geometric features is beneficial for identifying the six groups of TreMs in dense tree-stem point clouds, but the rasterized MVOPs are even more suitable. Whereas the overall accuracy of the RF was 70%, that of the deep CNN was substantially higher (83%). This study reveals that close-range TLS is promising for the semi-automated identification of TreMs for forest monitoring purposes, in particular when applying deep learning techniques.