Back to search

University of British Columbia

Ordination and classification of immature forest ecosystems in the Cowichan Lake area, Vancouver Island

Abstract

dc:description

The objectives of this study included: 1) classification of immature forest ecosystems surrounding Cowichan Lake on Vancouver Island, 2) investigation of relationships between floristic composition, selected site and soil properties, and ecosystem productivity as estimated by site index (m/100 years) of Douglas-fir (Pseudotsuga menziesii (Mirb.) Franco), and 3) evaluation of the usefulness of several multivariate analysis techniques for deriving the classification and investigating relationships. Methods employed included: 1) standard methods used in the biogeoclimatic ecosystem classification system as applied by the B.C. Ministry of Forests, and 2) multivariate analysis techniques, including ordinations (polar,reciprocal averaging, and detrended correspondence analysis) and cluster analysis of the vegetation data, and stepwise discriminant analysis of edatopic indicator species groups (EISG's), and site and soil properties. The classification was finalized after consideration of the environment data and results of multivariate analyses (particularly reciprocal averaging and detrended correspondence analysis) applied to the vegetation data. Three orders, five alliances, and six biogeocoenotic associations (BA's) were established. An objective, repeatable procedure for extracting characteristic combinations of species from summary vegetation tables was developed and applied. A comparison of the summary vegetation tables of three of the immature Cowichan Lake associations to the summary vegetation tables of three (climatically and edaphically) similar mature associations revealed strong similarities in understory species abundance and composition. This suggested that understory plant communities of the immature forest ecosystems had sufficiently stabilized to permit successful identification of probable climax associations. Estimated hygrotope and trophotope values of each plot suggested that axis 1 of the floristic data ordinations corresponded to a complex environmental gradient related to increasing availability of soil moisture and nutrients. This suggestion was supported by the results of indicator plant analysis, and by trends in a limited number of quantitatively assessed site morphological and soil physical and chemical propert ies. Discriminant analysis of EISG's and several site and soil properties selected linear combinations of variables which best characterized differences between the BA's. Soil properties proved more successful than site properties for this purpose. In addition, classification functions were produced which could be used to classify plots not used in the original analysis. Variation in site index values suggested an increase in productivity from BA's 2 to 4 but no differences between BA's 4 and 5. An investigation of relationships between axis 1 scores of detrended correspondence analysis of the floristic data, canonical variable 1 from the discriminant analysis of soil properties, and site index of Douglas-fir suggested that (for the forty-one intermediate plots in BA's 2 to 5): 1) 83% of the variation in understory vegetation was related to differences in the selected soil properties, 2) 78% of the variation in site index was related to changes in understory species abundance and composition, and 3) 71% of the variation in site index was related to changes in the selected soil properties. Mineral soil mineralizable N and exchangeable Ca values were highest, and C:N ratios lowest on the most productive sites. In addition, most of these sites had a mull humus form. This suggested that increases in site productivity were at least partially due to higher N availability. The higher levels of soil Ca on the most productive sites suggested more rapid nitrification rates and more nitrate, a form of N often found to be associated with the best growth of Douglas-fir. It was concluded that multivariate analysis techniques (particularly reciprocal averaging, detrended correspondence analysis, and discriminant analysis) were useful not only for classification purposes, but also for the investigation of trends in environmental properties and site productivity. These techniques provide a rapid, computer-assisted approach to data synthesis, and a more objective basis for interpretations. It was recommended that proponents of the biogeoclimatic system make greater use of these techniques in future studies.

Degree

thesis:*
Name thesis:degree_name
Master of Science - MSc
Level thesis:degree_level
master's
Discipline thesis:degree_discipline
Forestry
Grantor dc:publisher
University of British Columbia
Year dc:date
1984

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Roy, Roger Joseph James

Rights

dc:rights
Statement dc:rights
  • For non-commercial purposes only, such as research, private study and education. Additional conditions apply, see Terms of Use https://open.library.ubc.ca/terms_of_use.
Language dc:language
eng

Identifiers

dc:identifier.*
Handle dc:identifier
http://hdl.handle.net/2429/24761

Chain of custody

source
Harvested from
University of British Columbia
Base URL
circle.library.ubc.ca/oai/request
Last updated
2026-11-19
Source record
OAI-PMH GetRecord
related terms
citation

Roy, Roger Joseph James. Ordination and classification of immature forest ecosystems in the Cowichan Lake area, Vancouver Island. master's thesis, University of British Columbia, 1984. http://hdl.handle.net/2429/24761