Abstract
At 10 locations in Oregon and Washington, tree mortality resulted in dry-matter transfer of 1.5-4.5 Mg .cntdot. ha-1 .cntdot. yr-1 of boles and branches to the forest floor and 0.3-1.3 Mg .cntdot. ha-1 .cntdot. yr-1 of large-diameter roots directly to the mineral soil. The first value is about the same as that reported for leaf fall in similar stands; the second value generally is smaller than that reported for fine root turnover. Results are based on measurements by the U.S. Forest Service spanning 16-46 yr and areas as large as 42 ha. Values based on intervals < 10 yr were highly variable and potentially misleading. At an old-growth Douglas-fir [Pseudotsuga menziesii] stand in Washington, fallen boles accounted for 81 Mg/ha, standing dead for 54 Mg/ha. Density of fallen boles averaged from 0.14-0.27 g/cm3 depending on decay state. Values were lower than some previously reported because the sample included small-diameter fallen boles that tend to decay rapidly, and density was measured with techniques that minimized compaction and shrinkage. The decay rate at the old-growth stand, calculated indrectly by dividing bole mortality (megagrams per hectare per year) by the amount (meagagrams per hectare) of fallen and standing dead woody material, was 0.028 yr-1. This rate, 3-5 times those previously calculated directly from change in density alone, was almost identical to values calculated elsewhere from change in both volume and density. Decay rates based on change in density alone include only respired and leached material and exclude the large amount of material lost in fragmentation. This study shows the value of permanent plots, undisturbed by salvage logging, for retrospective studies of decomposition, nutrient cycling and productivity.