Spatially nonrandom tree mortality and ingrowth maintain equilibrium pattern in an old‐growth<i>Pseudotsuga–Tsuga</i>forest

Mortality processes in old-growth forests are generally assumed to be driven by gap-scale disturbance, with only a limited role ascribed to density-dependent mortality, but these assumptions are rarely tested with data sets incorporating repeated measurements. Using a 12-ha spatially explicit plot censused 13 years apart in an approximately 500-year-old Pseudotsuga-Tsuga forest, we demonstrate significant density-dependent mortality and spatially aggregated tree recruitment. However, the combined effect of these strongly nonrandom demographic processes was to maintain tree patterns in a state of dynamic equilibrium. Density-dependent mortality was most pronounced for the dominant late-successional species, Tsuga heterophylla. The long-lived, early-seral Pseudotsuga menziesii experienced an annual stem mortality rate of 0.84% and no new recruitment. Late-seral species Tsuga and Abies amabilis had nearly balanced demographic rates of ingrowth and mortality. The 2.34% mortality rate for Taxus brevifolia was higher than expected, notably less than ingrowth, and strongly affected by proximity to Tsuga. Large-diameter Tsuga structured both the regenerating conspecific and heterospecific cohorts with recruitment of Tsuga and Abies unlikely in neighborhoods crowded with large-diameter competitors (P < 0.001). Density-dependent competitive interactions strongly shape forest communities even five centuries after stand initiation, underscoring the dynamic nature of even equilibrial old-growth forests.