Post-windstorm natural regeneration dynamics in Italian Alps: roughness indices as a proxy for disturbance legacies.

Disturbance legacy can be defined as the amount, the availability and the distribution of resource, spatial pattern, and habitat left behind after a disturbance, and are key factor in the restoration processes, especially after large high-severity disturbances. Survived trees and green islands are a fundamental source for seed dissemination, while deadwood (e.g., stumps, logs, snags) and other structures (e.g., pit and mound morphology) create favorable microsites for regeneration establishment and survival. After stand replacing disturbance, like high-severity windstorm, disturbance legacies are responsible for most of the terrain roughness in the damaged area. The aim of this study is to test roughness indices as a proxy to infer the role of disturbance legacies on promoting natural regeneration establishment in the short term (3 years after the storm). From a drone-based point cloud we compute high resolution digital surface model (DSM) in four areas affected by a stand replacing windstorm in 2018 in eastern Italian Alps, then we calculate six different metrics for five different roughness indices. In the same areas we established 100 circular plots to measure regeneration density. Logging methods (salvage logging and no-intervention) have also been considered in the analysis. Preliminary results showed that only three indices (standard deviation of profile curvature SD_PC, standard deviation of the residual topography SD_RT, and vector dispersion VD) are significantly correlated with regeneration density, and among them, the 90^th quantile of SD_PC is the best fit. Overall, there is a positive significant correlation between roughness and regeneration density in the short term. In conclusion, the results suggest that roughness could be a good proxy for the disturbance legacies abundance and drone surveys are a powerful tool to adopt to estimate the roughness at slope scale. More detailed analysis on the threshold and influence of other parameters is needed and can be implemented to provide valuable management guidelines.