Search

Genetic structure (variance among and within populations) and geographic pattern of variations in growth and frost hardiness of mountain hemlock (Tsuga mertensiana (Bong.) Carr.) populations from coastal British Columbia (BC) were examined. Populations fell into two main groups — south coast (16 sources) and north coast (two sources). Height growth of greenhouse grown seedlings was measured during the first growing season on a weekly basis whereas frost hardiness and branch water content were evaluated during the fall at monthly intervals. Significant differences among the populations were found in maximum instantaneous growth rate, height at the end of the growing season and in October and November frost hardiness. There were no differences in date of maximum growth rate, branch water content and September frost hardiness. Most of the genetic variance was found within populations: 77% for height, maximum growth rate and November frost hardiness and 87% for October frost hardiness. Geographic trends were identified using multiple linear regression and canonical correlation analyses based on population means and latitude, longitude and elevation of the population origin. The uneven distribution of sources limited the extrapolation, but some general trends appeared. Maximum instantaneous growth rates increased with latitude, elevation and decreased with longitude (R2=0.64) and consequently were negatively correlated with the length of the growing season in the seed collection sites. Both populations from the north coast developed frost hardiness earlier than any population from the south coast. The relationship between climate coldness and growth rates has potential implications for seed transfer guidelines. Due to their higher growth rates, plants from higher altitudes may accomplish the same amount of growth within a shorter period compared to plants from lower altitudes when both are transferred north. Results of the frost hardiness tests indicate that seed transfer along the British Columbia coast of more than 3° northward will considerably increase chances of frost damage in plantations.

Patterns of genetic variation for adaptive and quantitative attributes in mountain hemlock ( Tsuga mertensiana (Bong.)) seedlings from British Columbia were examined at the population (provenance) and family levels. The population and family levels were represented by 12 provenances and 10 families from a single location, respectively. The adaptive attributes were related to gas exchange (net photosynthesis, transpiration rate, intercellular-to-ambient CO 2 concentration ratio, mesophyll conductance, stomatal conductance, and photosynthetic water use efficiency) and the quantitative traits were related to biomass allocation (height, root collar diameter, root dry weight, shoot dry weight and harvest index). Univariate and multivariate analyses of the studied attributes indicated the presence of significant differences at both, the population and family levels. Although the populations differed significantly in all studied traits, a relatively small portion (5–15%) of the total variation was attributed to among populations. On account of the observed large intra-population variation, no trends between the studied traits and geography were detected when data from individual seedlings were used; however, significant trends were detected on population means for the majority of gas-exchange attributes. Population-level analyses indicated lack of appreciable relationship between the quantitative and adaptive attributes. Intra-population analyses produced high narrow-sense heritabilities for the majority of the attributes studied. The observed high intra-population variation that is accompanied with high genetic control over the studied attributes present an opportunity for genetic improvement. On the other hand, the observed significant differences among populations represent a challenge for gene conservation efforts as well as to the methods deployed to exploit these variations.