Unique Canopy Structural Signatures of Moderate Disturbance Derived from Terrestrial LiDAR

Moderate disturbances alter forest structure and function due in part to how they reorganize forest canopies, resulting in resource redistribution (e.g. light, water, nutrients). There are multiple types of moderate disturbance including low intensity fire, pathogens, ice storms, wind-throw, age-related senescence, and insect damage—each resulting in a gradient of partial defoliation, gap generation, or species loss, with mixed effects on various ecosystem functions. It would be expected that different disturbance types would differ in how they reshape canopy structure. However, the resulting canopy reorganization following moderate disturbance remains poorly defined. Canopy structural complexity (CSC) measures derived using terrestrial laser scanning (TLS) describe structural aspects of forests canopies (e.g. canopy height, arrangement, and variability) and potentially offer a means of quantifying the magnitude and directionality of canopy reorganization. We hypothesized that moderate forest disturbances modify canopy structure in one of four ways based on the directionality and magnitude of change. These groups include top-down disturbances (ice storm damage) that primarily reduce leaf area index (LAI) in the upper canopy strata, bottom-up disturbances (ground fires) that reduce subcanopy LAI, random disturbances (wind-throw) which lack directionality, and targeted disturbances (beech bark disease) that affect specific species or areas of the forest, often resulting deep gaps across canopy strata.

To test this hypothesis, we surveyed sites across the eastern United States that were subjected to an array of disturbance types, including: a ground-fire in Tennessee; a hemlock woolly-adelgid outbreak in Massachusetts; chronic acid-deposition in West Virginia; ice storm damage in New Hampshire; and beech bark disease and age-related senescence in Michigan. At each site, we used terrestrial LiDAR to derive canopy structural complexity (CSC) metrics summarizing canopy height, arrangement, heterogeneity, openness, variability, and density. A machine learning approach identified which CSC metrics most significantly distinguished disturbed from non-disturbed areas for each disturbance. Outputs from these models were then run through an ordination analysis to distinguish specific clusters of structural change. Results show, for example, that following ground fires in the Great Smoky Mountains, vegetation area index (VAI) decreases overall, but also the distribution of VAI within the canopy shifts upward, indicating a loss of subcanopy. Conversely, results from ice storm damage in New Hampshire indicate a change in the outer canopy height of the forest. In total, our results indicate that CSC measures can provide unique structural signatures of moderate disturbance-types, with potential to inform how various disturbances affect ecosystem functioning.