When a tree falls: Forest inventories illustrate how wood mechanical properties influence standing to down transitions in US forests

Background/Question/Methods Forest trees represent a major terrestrial carbon pool.  When a tree falls, increased exposure to soil moisture and saprobes may accelerate decay.  How frequently trees fall, and associated changes in forest carbon stocks, may depend on variation in the mechanical properties of their wood.  We examined relationships between of interspecific variation in wood mechanical properties and tree fall rates in eastern North American forests.  To quantify variation in wood mechanical properties, we analyzed a Forest Products Lab (FPL) database including up to eight mechanical parameters (e.g., wood density, modulus of elasticity, work to maximum load) and an index of decay resistance for 63 eastern North American woody species.  To quantify the frequency of tree falls, we modeled the probability that individual trees, both living and standing dead, disappeared  from consecutive Forest Inventory and Analysis (FIA) surveys of unmanaged forest plots.  We expected that the species-specific rates of disappearance (primarily reflecting tree falls) for living trees would be correlated with their corresponding rates of disappearance for standing dead trees,  that both rates would be higher for species with weaker wood, and that snag fall rates would also depend on wood resistance to decay.

Results/Conclusions Wood mechanical properties of eastern North American tree species were highly intercorrelated.  The first principal component axis explained over 80% of the variation in the FPL mechanical properties dataset and distinguished species with dense, mechanically resilient wood, like common persimmon  (Diospyros virginiana) from species with lighter weaker wood like white pine (Pinus strobus).  In FIA plots, fall rates of living trees and snags were weakly correlated (R-square = 0.06) and influenced by both tree size and plot location.  Controlling for these factors, species with mechanically weaker wood fell more frequently (p<0.05).  This effect was stronger for living trees than for snags, in part because decay appears to change the mechanical properties of some species more quickly than others.  For example, hickories (Carya spp.), known for their mechanically resilient wood, were unlikely to fall while living but decay quickly and are among the most likely to fall as snags.  Across all species, wood mechanical strength and decay resistance together explain 60% of the variation in snag fall rates.  These results illustrate how tree species’ wood traits have complex influences on forest ecosystem dynamics and highlight the value of integrating diverse datasets to understanding the ecosystem consequences of trait variation.