Quantifying Patterns of Lightning-Caused Canopy Disturbances via Integration of Drone Imagery and Field Surveys

Abstract

Lightning is an important agent of tree mortality and gap formation. Here we quantified spatial and temporal patterns of lightning-caused canopy disturbance in a 50-ha plot in Panama using monthly drone imagery, and compared these patterns with field measurements of disturbance severity and spatial extent. Of 22 lightning strikes that we tracked, the impacts of 18 were monitored for at least 12 months (range of 17–50 months), and 67% of these 18 strikes led to canopy disturbances. The mean time for the first and last canopy disturbance to appear post-strike was 8.2 months (range: 0.8–14 months) and 14.6 months (range: 0.8–23.9 months), respectively. Canopy disturbances were generally highly irregular in shape (i.e., not circular), and clustered around the rooting point of the directly struck tree. A mean of 43% (± 19%) of the total lightning-associated canopy disturbance area was within 10 m of the rooting point, whereas only 3% (± 5%) occurred 30–40 m from this point. Drone-based measurements of canopy disturbance area and volume were good predictors of variation in ground-estimated dead biomass (r² = 0.48 and 0.46, respectively), reflecting their strong association with overstory dead biomass (r² = 0.42 and 0.41, respectively). The total drone-estimated canopy disturbance area was 49% of the ground-estimated canopy disturbance area. Thus, lightning typically causes canopy disturbances that are detectable with drone imagery despite their irregular shape, and drone-detected gap formation lags 8–15 months poststrike, potentially disconnecting drone-detected disturbances from their ultimate cause.