Long-term recovery dynamics following hurricane-related wind disturbance in a southern Appalachian forest

Wind disturbance affects thousands of km2 annually in eastern temperate forests, yet few studies address longterm recovery. Here, I assess changes in forest structure and composition before (pre-Opal), immediately after (Y0), and 21 years after (Y21) Hurricane Opal in 0.166-1.08 ha gaps created by microbursts, and undisturbed controls. In gaps, an average of 24.0% of trees (41.1% BA) were windthrown and 1.0% of trees (1.1% BA) died standing during Opal; scarlet and black oak were disproportionately windthrown. Subsequent windthrow or standing tree mortality rates did not differ between treatments. By Y21 an average of 32.2% (49.5% BA) of trees were windthrown in gaps and 3.2% (2.4% BA) in controls; 7.2% (7.6% BA) died standing in gaps and 12.0% (11.7% BA) in controls. Pit depth and windthrown rootmass area decreased by 63% by Y21. Total snag density differed over time (23.8-38.8/ha in controls; 6.7-19.7/ha in gaps); "new" snag (e.g., died after Opal) density did not differ between treatments. Disproportionately more scarlet oak, shortleaf pine, and hickory but fewer yellowpoplar, red maple and sourwood died standing. By Y21 live tree density and (marginally) basal area recovered to pre-Opal levels in gaps but remained lower than controls. In gaps, hurricane-related mortality of trees 38.1-50.7 cm dbh was heaviest, and density of this size-class remained lower by Y21; density of trees 50.8-63.4 cm was also lower in gaps in Y21. Trees grew faster in gaps than controls but stand-level gains in average dbh were smaller in gaps due to a greater increase in small trees (ingrowth) by Y21. Initial hurricane-related mortality changed relative importance (average of relative abundance and relative BA) of species in gaps; differences in mortality and growth rates among species contributed to subsequent shifts by Y21, especially in gaps. Post-Opal gains in shade-intolerant yellow-poplar importance were small but significant by Y21; shade-tolerant generalist species such as red maple and sourwood also increased, whereas scarlet and black oak decreased. Cumulatively, small changes by multiple species resulted in decreased (-19.9%) importance of the oak-hickory group and a corresponding increase in "other"; this trend was also evident in controls at a much smaller scale (-5.6%). Results indicate that a wind-related "pulse" of heavy mortality unevenly distributed among species - in this case scarlet and black oak - can alter stand structure and relative importance of species for decades, with accelerated loss of the oak-hickory group and replacement largely by shade-tolerant generalist species.