Lightning strikes thousands of trees every day and ca. 500 million hectares of tropical forest exist in regions with high lightning frequency. Anecdotal evidence and model-based estimates suggest that lightning primarily affects canopy trees and could influence forest dynamics. However, existing information regarding lightning-caused tree mortality in tropical forests is sparse and mainly limited to fortuitous observations.
Over the past three years, we located lightning strikes by triangulation from video images generated by tower-mounted cameras on Barro Colorado Island, Panama (BCI). We used these data in combination with bootstrapping methods and satellite-based lightning frequency data to calculate rates of lightning-caused tree mortality. We contrasted the results with background mortality rates in the BCI 50 ha forest dynamics plot and confirmed the lightning-mortality estimates using real rates of lightning mortality in this plot during 2016. Finally, we estimated long-term mortality rates and the effects of predicted increases in lightning frequency.
The immediate effects of lightning (within 3-10 months) are responsible for 34% (95%CI = 9%) of annual large canopy tree mortality (trees > 60 cm in diameter) and reduce average canopy residence time for large trees by 33% (ca. 30 years). Lightning-caused mortality rates were similar in the 50ha plot during 2016 (49% of annual mortality). Although 1 in 4 trees affected by lightning died within the 3 months of the strike, lightning-related tree deaths continue to occur for years. Using a conservative 40% total death rate for lightning-struck trees, we estimate that 56% (95%CI = 9.5%) of large tropical tree mortality is caused by lightning. Given that lightning frequency is predicted to increase with climatic change, large tree mortality rates could increase by as much as 23% in the coming decades. Such increased mortality is likely to have important effects on forest structure and ecosystem function.