Factors determining species composition of post-disturbance vegetation following logging and burning of an old growth Douglas-fir forest

I investigated how the factors that affect the contribution ofpropagules from the seed bank, bud bank and seed rain influenced theinitial vegetation following logging and slash-burning of an oldgrowth Douglas-fir forest at 800 m elevation located in the westernCascades of Oregon.
Permanent field plot data showed species composition anddensity of establishing vegetation differed greatly between the firstand second year after disturbance. This difference was caused by ashift in the relative contribution of propagule sources, whichgenerally differed in species composition. Vegetative propagules ofold growth species dominated the first year (67%) and dispersed seedof early successional species dominated the second year (98%).
Seedling emergence from field soil samples placed in agreenhouse showed early successional species dominated the old growthseed bank, which included 88 seeds/m2 and 11 species; the disturbed
seed bank density was significantly reduced to 18 seed/m2 with 6species.
To further investigate the effect of heat on seed bank density,a laboratory study assessed the effect of experimental heattreatments on the germination of six old'grawth seed bank species.With one exception all species had the same response to heattreatment: at 50C germination was not reduced; at 100C germinationwas significantly reduced; and, at 75C, germination was significantlyreduced in the wet soil, but not in the dry soil.
Of the pre-disturbance species on the permanent field plots,89% have the ability to sprout after disturbance. These speciesconstitute the potential bud bank.
Seedling emergence from soil seed traps placed at the fieldsite showed the seed rain of the second year was dominated by earlysuccessional species.
To investigate why so few old growth species regenerated fromdispersed seed, seed reproduction, seed viability, dispersal ratesand emergence rates were examined for three old growth understoryspecies, Berberis nervosa, Linnaea borealis and Gaultheria shallon.Although B. nervosa had the highest cover, it produced no seeds inthe year of study. The seed production (227 seeds/m2) of G. shallonwas much greater than that of L. borealis (4 seeds/m2). The densityof dispersed seed of G. shallon, the only species to disperse seedinto the old growth (18 seeds/m2) and into the disturbed site (6seeds/m2), was much less than the density of seeds produced. None ofthe three species emerged from seed experimentally sown in thedisturbed site. Thus, if similar patterns of low seed production,
low seed viability, low dispersal and establishment rates hold forother old growth understory species, regeneration from dispersed seedwill be necessarily slow immediately after disturbance.
Both the low potential of the seed bank and the greaterpotential of the bud bank contributed to'the dominance of old growthspecies from vegetative propagules in the first year's establishment.The reduction of the bud bank propagules after the first year'sestablishment, the low regeneration of old growth understory speciesfrom dispersed seed and the on-site dispersal of early successionalspecies that colonized the first year contributed to both the greaterdensity of vegetation establishing the second year and the dominanceof early successional species.