Small Diameter and Decreased Radial Growth Control Pinus palustris Tropical Cyclone Damage in Georgia, USA

Traditional correlative analyses often struggle to parse the confounding drivers of storm damage in forested systems. To address this limitation, this study utilized a causal inference framework to separate broad observational trends from underlying biological mechanisms. We applied a non-parametric causal forest algorithm to evaluate how diameter at breast height (DBH) and DBH growth rates influenced tree-level damage probabilities in even-aged longleaf pine ( Pinus palustris Mill.) plantations following Hurricane Helene (2024). The data were from two contrasting sites, an old-field and a cutover plantation, in Berrien County, Georgia, USA, ranging from 5 to 24 cm DBH. Pre-tropical cyclone correlative analysis showed that the old-field site possessed greater aboveground productivity, like increased DBH, height, volume, and basal area. Post-cyclone analysis revealed that while aboveground trends remained, the old-field site suffered a 17% reduction in stand density. While a simple correlative assessment would likely attribute this decline entirely to the vulnerability of larger trees, the causal forest model parsed a more nuanced storm damage response. Specifically, smaller trees (< 12 cm DBH; 1.69 to 8.06 m height) demonstrated high sensitivity to moderate stem damage, while larger, slower-growing individuals (> 12 cm DBH; 10.09 to 12.17 m height) were more susceptible to severe stem failure. These results highlight proactive density management - specifically tailored to maintain increased individual tree growth rates - as a critical tool for building wind resilience in southern pine plantations. Additionally, causal inference applied to site-intensive inventory data provided a precise, data-centered understanding of storm-induced mortality dynamics.