This thesis describes patterns of variation in several cone, seed and seedling characteristics of Pacific silver fir (Abies amabilis [Doug.] Forbes) sampled from Vancouver Island, B.C. Cone collections kept separate by tree were made at eight locations during the fall of 1983 to provide material for the study. The inheritance patterns of 13 enzyme loci were determined from seed tissues of 87 trees, and seven loci were found to possess at least two allozyme variants. These loci conformed to the assumptions of Mendelian-type inheritance, although AAT-2 displayed marked segregation distortion. No linkage groups could be established at the sampling intensity available (20 seeds per tree). Significant levels of inbreeding, based on a multilocus estimate of outcrossing, were detected in five of seven populations and indirect evidence suggests related matings other than selfing may be occurring. Variation among populations in outcrossing rate was evident (0.725 < tm ≤ 1.0) and appears positively correlated with seed size (measured by 1000-seed weight). High levels of allozyme variation were found to exist within populations (95-98%) and estimates of the extent of population differentiation were shown to differ depending upon the particular analytic method employed. Maternal (extant) trees appeared more heterozygous than did viable embryos and populations sampled on southern Vancouver Island appeared more genetically diverse than did populations sampled on northern Vancouver Island. A sub-sample consisting of two populations, each with seven trees, from northern, mid and southern Vancouver Island provided material for a germination test and open-pollinated progeny study. Seed dormancy was not pronounced among populations. Large family differences in germination responses were detected, irrespective of pregermination treatment, suggesting a high degree of genetic control of germination in Pacific silver fir. Anomalous germination behavior in one population was attributed to sub-optimal stratification conditions and proliferation of mold. Improvement in production of seedlings of Pacific silver fir may be achieved by collecting and germinating seeds on a family-by-family basis. Germinants from the first count of the germination test provided open-pollinated progeny for measurement of growth variables. Seedlings were grown in a greenhouse alongside production Abies stock for 29 weeks. Population differences accounted for a considerable part of variance in cone and seed size. The effect of population on height of seedlings at eight weeks was significant but declined to virtually zero by the end of the test. Populations had negligible influence on growth rate of seedlings as well. Variation in growth rate among open-pollinated families was statistically significant but accounted for only 20% of the total variation. Significant population differences were detected in root weight of harvested seedlings. Family differences in this and other biomass variables were at most 20%, with the majority of variation in seedling growth traits residing within families. / Forestry, Faculty of / Graduate

When making decisions on which areas to harvest in a sustained yield, even-flow manner in mountainous areas such those in coastal British Columbia, it is important to know how timber productivity changes with elevation. This information allows foresters to decide at what elevation to start increasing the rotation age and to decide at what elevation sustainable harvesting becomes infeasible due to low productivity. Since Pacific silver fir (Abies amabilis Dougl. ex Forbes) has an elevation range that extends from sea level nearly to the tree line (0 m to approximately 1,650 m; from the Coastal Western Hemlock zone, through the Mountain Hemlock zone; to the lower limits of the Alpine Tundra zone), productivity-elevation relationships are especially important. To acquire quantitative measures of productivity decrease with increasing elevation a regression equation relating site index (the height of the dominant trees at a base age of breast height age of 50 years) to elevation in southern coastal BC was developed. In turn, we used this regression as an input into the height driven yield model named the Variable Density Yield Prediction model (VDYP). The use of the VDYP model allows the site index values to be translated into actual productivity measures (e.g., volume per hectare, mean annual increment at culmination age).

Ecosystem-specific forest management requires comprehension of tree species productivity in managed settings, and how this productivity varies with the ecological determinants of site quality, i.e., the environmental factors that directly affect the growth of plants: light, heat, soil moisture, soil nutrients, and soil aeration. A good understanding of this variation is necessary for making species- and site-specific silvicultural decisions to maximize productivity. Productivity of a given species is usually measured by site index (tree height at 50 years at breast height age). Quantitative relationships between site index and these measures of site quality provide predictive models for estimating site index. Pacific silver fir (Abies amabilis (Dougl. ex Loud.) Forbes) is an important timber crop species in the coastal forests of British Columbia. In relation to climate, its range in southwestern British Columbia extends from sea level to almost timberline, and from the hypermaritime region on western Vancouver Island to the subcontinental region on the leeward side of the Coast Mountains. In relation to soils, its range extends from slightly dry to wet sites and from very poor to very rich sites. In view of this relatively wide climatic amplitude, a large variability in productivity can be expected. It is particularly important to consider the growth performance of Pacific silver fir when decisions are made regarding whether or not to cut stands on high-elevation sites. In the study summarized here, relationships between Pacific silver fir site index and selected ecological measures of site quality were examined, and site index models using these measures as predictors were developed.