Developmental trends of stand structure and tree mortality in coastal western hemlock forests

I examined forest structure and tree mortality in the Coastal Western Hemlock very wet maritime 1 variant (CWHvm1) of British Columbia, using data collected from 190 permanent sample plots representing stand ages from 13 to over 300 years old. I found that patterns of structural development of forest attributes in the CWHvm1 were similar to patterns observed elsewhere in the Pacific Northwest. For instance, percent cover of understory vegetation followed a so called "U" shaped curve with increasing stand age and, conversely, mean tree diameter, biomass oflive stems and production of dead stem biomass were consistent with the so called "s" shaped or sigmoid curve. Overall stem density declined with age and in diameter distributions of live stems shifted from a "reverse-J" type distribution in young stands, to a near normal distribution in intermediate-aged stands, and back to a reverse-J distribution in old stands. Tree mortality and many size classes of live stems, for both stem density and stem biomass, followed an inverted "U" shaped trend over stand age. My observations are consistent with other research demonstrating that old growth stands are structurally distinct from mature stands. Old growth stands were characterized by a wide range of tree diameters, high levels of understory cover, low density and high biomass of live stems, and a reverse-J diameter distribution. Further, tree mortality rates were low in old growth stands. In contrast, in intermediate-aged stands, the ranges of stem diameter, proportion of subcanopy stems, proportion of tree species other than hemlock, and understory cover were low. As well, stem diameter displayed a near-normal distribution and mortality rates were high. Developmental trajectories in the CWHvm1 were influenced by site productivity and other factors such as aspect, elevation and geographic location. The rate of change in the density of different size classes of stems, percent cover of understory vegetation and degree of canopy closure increased on more productive sites. Therefore, the rate that a stand moves through developmental stages depends to some extent on the productivity of the site. For tExample, a stand is likely to develop old growth characteristics more rapidly on more productive sites. Age, site productivity, aspect and geographic location do not explain all of the variation observed for the attributes examined in my study. When I examined the influence of these factors on density of live trees, I found that about 30% of the variation in stem density remained unexplained. It is possible that evaluating developmental trends at the site series level would eliminate much of this remaining variation, although I was unable to examine this with my data. The structural dynamics ofCWH forests require additional characterization. I was unable to examine trends in size and decay class distributions of standing and down dead wood, and differences in developmental trajectories resulting from different disturbance histories. Low sample sizes prevented me from exploring detailed developmental trends in stand structure during the transition from mature to old growth. Moreover I was unable to examine temporal trends among old growth CWH stands because ages of plots in these very old stands were not accurately determined. New silvicultural prescriptions need to be developed if very large stems and elevated levels of understory cover are to be retained in managed landscapes. Further, regulations, guidelines and prescriptions based solely on the assumption that certain attributes are associated with a given age class are inappropriate because of substantial variation in structural characteristics of forest stands.