High resolution LiDAR and environmental variables enable fine scale predictions of hollow-bearing tree abundance across a complex temperate forest landscape

Hollow-bearing trees (HBTs) provide a critical habitat resource for hollow-dependent animals, many of which are threatened due to a range of natural and human disturbances. Developing strategic conservation plans for HBT-dependant fauna requires an understanding of the spatial distribution of HBTs across multiple scales. Historically, landscape-scale estimates of HBT abundance have been based on extrapolations from much smaller areas where they have been directly observed. Remote-sensing technologies that provide high-resolution data at large spatial scales have the potential to improve the quality and coverage of HBT estimates across complex forested landscapes. In this study we used high-resolution, LiDAR-based metrics of forest structure and topography, coupled with detailed ground-based assessments of HBT abundance and environmental covariates, to develop predictive models of HBT abundance across ∼340,000 ha of temperate forest in southeastern Australia. Our analyses estimated a mean density of 4.4 (95 % CI: 3.7 – 5.1) HBT per ha across the study region. The abundance of different HBT form classes varied substantially by forest tenure and vegetation type. Our analyses identified complex relationships between the abundance of HBTs and key environmental and forest structural predictors that reflect how fine-scale variability in environmental conditions, disturbance history, and forest management practices within complex landscapes influence HBT abundance and occurrence. High-resolution, spatially explicit remote-sensing data can be used to provide rigorous empirical estimates of the distribution and abundance of HBTs in complex forested landscapes. These are critical to informing conservation and forest management planning to ensure long-term continuity of HBTs within these landscapes. • Remote-sensing data provide rigorous empirical estimates hollow bearing trees (HBTs). • A mean density of 4.4 (3.7 – 5.1) HBT per ha was predicted. • Abundance of HBTs varied by environmental and forest structural factors. • Spatial predictions can inform conservation planning within the study landscape.