The relative importance of different causes of tree mortality in old-growth forests is poorly understood.  We studied the frequency and causes of tree mortality in old-growth conifer forests through repeat censuses of a series of 12, 1.00 ha permanent forest dynamics plots.  All study plots were located in old-growth forests (300-1200 years since stand-initiating disturbance) within Mt. Rainer National Park, WA, USA. 

Results/Conclusions

A total of 1130 tree mortalities occurred during the three decade study period.  Of the total, 457 trees (40.4% of all mortalities) died proximately from mechanical causes.  Within the mechanical class, 297 trees (65.0% of mechanical mortalities) died due to direct mechanical processes: uprooting or stem breakage by wind, snow and ice, or structural failure.  In contrast, 160 trees (35.0% of mechanical mortalities) died due to indirect mechanical causes; specifically, crushing by falling trees or tree parts (limbs, branches, boles).  Strong size discrimination of direct and indirect mechanical mortality agents was apparent: mean diameter of trees killed by direct mechanical processes was significantly greater (P<0.01) than mean diameter of trees killed by indirect mechanical processes.  Evidence for a contributing role of decay organisms (heart rot, structural root rot, hollow stem) was recorded in 33.3% of trees killed by direct mechanical processes, but was evident in only 5.0% of trees killed by crushing.  These results support the ideas that (1) asymmetrical interactions between trees in old-growth forests are not limited to competition for resources (e.g., light), but also include physical interactions; and (2) decay organisms indirectly contribute to the mortality of uninfected trees by predisposing large, “crush-capable” trees to mechanical mortality.