Classification of plant communities in Gwaii Haanas National Park Reserve and Haida Heritage Site

Plant communities were classified and described for Gwaii Haanas National Park Reserve and Haida Heritage Site located on the Queen Charlotte Islands in British Columbia, Canada. Existing vegetation and environment data were compiled from various sources and assessed for their compatibility as an integrated database. The combined data set was subjected to floristic and physiognomic - floristic methods for vegetation classification using TWINSPAN. The results of each were assessed using multiple response permutation procedures (MRPP) and vegetation tabular analysis (VTAB) to determine which method was superior, based on the homogeneity of individual vegetation units and the presence of meaningful diagnostic species in each. Gradient analysis (Cannonical Correspondence Analysis) was also utilised in conjunction with classification results to explore vegetation - environment relationships. Each association was described individually and in a landscape context. Data considered for analysis in this study were deemed to be acceptable for amalgamation. A bias was detected in sampling distribution overall, favouring upland forests. Species identification effort varied between data sets, particularly with respect to the detection of bryophytes and graminoids. The physiognomic - floristic approach was determined to be moderately superior. This classification resulted in a hierarchical structure with 9 alliances, 20 associations and 12 subassociations. These associations range in an elevational gradient from estuarine meadows and rocky shoreline communities through upland forests and wetlands to subalpine parkland. Subalpine and alpine communities are limited in the study area and were poorly represented in the data. Most communities appear simplified, homogenous and poorly defined, in floristic composition. This has been ascribed to browsing pressure by introduced deer. This effect is particularly evident in the upland forest communities. Recommendations are made for future expansion of classification work in the protected area. Assessment of plant association distribution and abundance for conservation purposes is desirable but not possible due to a lack of spatial accuracy and poor correlation between plant associations and environmental attributes. Additional sampling which takes into consideration geographic stratification and positional accuracy will allow for more rigorous spatial analysis in the future.Forestry, Faculty ofGraduat