Vegetation - environment relationships of forest communities on central eastern Vancouver Island

• Main Author: Beese, William John
• Language: English
• Published: 2010
• Online Access: http://hdl.handle.net/2429/22318

The Habitat Type classification system, as introduced in coastal British Columbia by MacMillan Bloedel Limited, was used for classification of old-growth plant communities in four vegetation zones on central eastern Vancouver Island. A total of 14 forest habitat types were identified. A combination of ordination and traditional association table synthesis was used in defining types. Polar ordination and an improved reciprocal averaging technique named DEtrended CORrespondence ANAlysis (DECORANA) were found to be useful for both the classification and evaluation of environmental relationships among the types. Findings substantiate claims that DECORANA is the best general purpose, indirect ordination technique currently available. In most cases, DECORANA produced superior results to polar ordination with the same data input. The most successful ordinations of individual vegetation zones were obtained for understory percent coverage data transformed to an 8-point scale. Ordinations of tree species alone, or tree and understory species combined were less useful. The distribution of old-growth vegetation in response to environmental gradients was interpreted. A moisture gradient accounted for most of the variation between habitat types within any single vegetation zone. Interpretation of the moisture gradient was tested using a Water Stress Index (WSI), which predicts potential soil moisture stress from site and climatic data. Climatic data used for input in the WSI model were extrapolated from the nearest available Resource Analysis Branch (RAB) or Atmospheric Environment Service (AES) climatological stations. Soil texture, rooting depth, coarse fragment content and depth of the surface organic horizons were used to calculate Available Water Storage Capacity (AWSC) for use in the model. The WSI proved to be a useful integrator for general site comparison of potential moisture stress. In most cases, the WSI prediction showed a closer relationship to ordination axes interpreted as moisture gradients than any other single factor, except for sites with primary moisture inputs from seepage, high water table, or those with complex orographic rainfall influences. Lack of accurate climatic data, difficulty in quantifying soil moisture relationships and the influence of complex microclimatic effects reduced the utility of the index. The temperature data available were suitable for quantifying some of the broad macroclimatic differences between zones, and for evaluating differences between samples at the elevational extremes within the Tsuga heterophylla and Abies amabi1is - Tsuga heterophylla vegetation zones. They were not adequate for a comparison of many of the plots on a site-specific basis because of problems in extrapolation from a limited data base and because of the exclusion of soil temperature and cold air drainage relationships. Potential solar radiation was used as an integrator of slope, aspect and latitude. Some habitat types were found to occur more frequently in certain radiation environments. The elevational location of vegetation zones varied by as much as 80 metres as a result of solar radiation effects. Detailed transect sampling of continuous old-growth forests in leave-strips in the Cameron and Nanaimo River valleys revealed substantial differences in species composition and distribution between the two valleys that were accounted for by climate and overall soil texture. The wetter climate and finer-textured soils of the Cameron valley supported habitat types dominated by Polyst ichum muniturn (swordfern) and Achlys triphylla (vanilla leaf) with an absence of Gaultheria shallon (salal) dominated habitat types. The drier climate and coarser-textured soils in the Nanaimo valley supported salal-dominated habitat types along the entire south-facing slope and lower north-facing slope. Significant differences in tree species distribution were also observed. Habitat types were related to units in the existing plant community classifications in B.C. and the Pacific Northwestern U.S. Most of the habitat types could be related to types previously described in British Columbia, and many were similar to those found in coastal Oregon and Washington. Some differences in the flora of Vancouver Island and the mainland were noted. There is a need for a compilation of the descriptions of all of the plant communities on Vancouver Island in a single volume. === Forestry, Faculty of === Graduate