Nitrogen transformations in fallen tree boles and mineral soil of an old-growth forest

Ecology, June, 1999 by Stephen C. Hart

Well-decayed boles in this forest (which is typical of other old-growth Douglas-fir/western hemlock forests in the PNW) cover only a small fraction of the stand area ([approximately]5%; M. E. Harmon, personal communication). Using two published estimates of the mass of class 4 and 5 boles in similar forest stands within the H. J. Andrews Experimental Forest (Sollins et al. 1987, Means et al, 1992) and assuming that my measured field rates of net N mineralization in the upper 0.15 m of bole is representative of the entire bole mass, I estimate that well-decayed boles contribute [approximately]0.16-0.25 g N[multiplied by][m.sup.-2][multiplied by][y.sup.- 1] of plant-available N. Total plant uptake in a nearby old-growth Douglas-fir forest has been estimated at about 4.0 g N[multiplied by][m.sup.-2][multiplied by][yr.sup.-1] (Sollins et al. 1980), suggesting that boles could contribute only [approximately]4-6% of plant N uptake. This N flux is lower than annual net rates of inorganic-N flux from the forest floor to the mineral soil in this stand (0.3 g N[multiplied by][m.sup.-2][multiplied by][yr.sup.-1]; S.C. Hart, unpublished data), total atmospheric N inputs (0.3-0.4 g N[multiplied by][m.sup.-2][multiplied by][yr.sup.-1]; Sollins et al. 1980), and net N mineralization from ground-layer mosses (0.4 g N[multiplied by][m.sup.-2][multiplied by][yr.sup.-1]; Binkley and Graham 1981) and from soil humus (0.8 g N[multiplied by][m.sup.-2][multiplied by][yr.sup.-1]; Sollins et al. 1980) in similar old-growth stands at the H. J. Andrews Experimental Forest. Although these comparisons suggest that net N release from well-decayed boles is a relatively small ecosystem N-flux, the construct that well-decayed boles are net sources, not net N sinks, of plant-available N should be explicitly included in forest N-cycling models (Comins and McMurtrie 1993).

Bole chronosequence studies conducted in forests of the PNW have generally found similar N contents in boles of decay classes 1 through 3 (Sollins et al. 1987, Means et al. 1992; but see Grief 1978), suggesting that little net N immobilization or mineralization occurs in CWD during the early stages of decay (mean age of [less than]40 yr). The apparent lack of net N dynamics in CWD contrasts with the very dynamic changes in N content that typically occur in fine litter within these ecosystems (Harmon and Chen 1991). However, the failure of these chronosequence studies to account for the process of fragmentation (i.e., decrease in bole volume without a change in density) results in an underestimate of bole decomposition rates, and hence may lead to an underestimate of the net rate of N release from boles (Harmon and Chen 1991). Recent time-series decomposition studies also have shown that boles can be net sources of N during each of the first 7 yr of decomposition, with fungal sporocarp production being the major pathway of N export (Harmon et al. 1994). All of these results suggest that my estimate of net N mineralization in CWD during the latter stages of decay is likely not offset by net N immobilization during the early stages of CWD decomposition in these forests.