Effect of harvesting methods on the phytosociology of a boreal mixedwood forest community

Abstract

A naturally regenerating boreal mixedwood site was remeasured 37 years after four harvest treatments; clearcutting (CC), deferred 'softwoods only' cutting (DC), 'softwoods only' cutting (SC), and hardwoods poisoned, followed by deferred 'softwoods only' cutting (HP). The objectives were to determine whether ; 1) harvest treatments produced significantly different phytosociological communities: 2) harvest treatments altered the normal trend of hardwoods dominating early succession, and allowed Picea spp. to gain a competitive advantage; and 3) phytosociological trends and corresponding possible environmental influences could be identified. Median polish and/or ANOVA were used to compare; species density, dominance (basal area), frequency, cover, age, and height for the tree stratum; species density, frequency and cover for the shrub stratum; and species frequency and cover for the herb stratum. Median polish, species' diversity, resemblance measures, and canonical correspondence analysis (CCA) were employed to determine whether harvest treatments affected the overall phytosociological community. Ratios of softwood to hardwood (S;H), spruce to fir (Spr;F), and spruce to hardwood (Spr;H) were calculated for the tree stratum. Comparisons of pre-harvest data to current data were made. Correspondence analysis (CA), CCA and cluster analysis were used to search for the main phytosociological trends and to determine whether these were related to soil moisture and depth. Some treatment differences were evident for individual species. However, species did not form groups that occurred exclusively or abundantly in only certain harvest treatments. The four harvest treatments did not produce significantly different phytosociological communities. This may be due to several factors; 1) the Initial Floristics character of boreal forests: 2) chance factors in natural regeneration, and 3) the confounding of treatment effects by complex environmental gradients. In terms of improving spruce and softwood status, the results were not conclusive as to the advantage over clearcutting of cutting only softwoods, and of poisoning hardwoods in advance of cutting. Picea glauca (Moench) A. Voss, Populus tremuloides Michx., Populus balsamifera L. and the Spr;H ratios, supported the expected trend of softwoods being more prominent in HP than CC, and the reverse for hardwoods. However, Abies balsamea (L.) Mill, Betula papyrifera Marsh., Picea mariana (Mill.) B.S.P. and the S;H ratios did not support this trend. The Spr.F was generally higher in HP than CC. Conifers were mostly of advance-growth origin in DC and HP, but of post-harvest origin in CC, matching expectations that CC caused more damage to advance growth than did partial cutting (conifer age in SC did not support this). The strongest phytosociological trend identified was the change in canopy composition from hardwood to mixedwood to softwood types. Understorey species distribution was related to either environmental gradients created by canopy composition, or to some underlying gradients that determined both canopy and understorey compositions. Soil depth and/or moisture did not appear to be the determining factors. It is suspected that the boreal mixedwood community of the RC17 site is best described as a continuum of species' presences and abundances, determined by a complex set of interdependent environmental factors, which would need to be clearly defined and accurately measured to determine conclusively whether harvest treatments differentially affected the community.