Adaptive variation in fall cold hardiness of aspen from northwestern Ontario

We investigated adaptive variation in fall cold hardiness development based on the electrical conductivity of tissue diffusates (EC) among 20 aspen provenances from northwestern Ontario. Provenance accounted for over 40% of the total variation in cold injury for seven dates from September through November in three provenance trials. Principal component analysis was performed to summarize the combinations of results for all sampling sites, dates and temperatures (traits). Principal component (PC)-1 represented fully developed cold hardiness differences among provenances; PC-2 represented differences in the timing of the onset of cold hardiness development; and PC-3 represented a site-related difference in cold hardiness development. Heat sum in early summer and late summer precipitation together were the best predictors of absolute degree of cold hardiness (PC-1), whereas temperatures for mid- to late summer were best for predicting onset of cold hardiness development (PC-2). In a second study, we assessed the efficacy of chlorophyll fluorescence (CF) as a simpler technique for determining the cold hardiness of aspen stem samples. Fall cold hardiness of stem samples of 12 of the original 20 provenances was estimated by CF, and the results were evaluated by a visual scoring (VS) method. Correlations between EC and CF measurements from the two studies were moderately strong based on the extent of cold hardiness in October of each year, but were negative for September dates because of a later onset of cold hardiness in the EC study year. Although the EC and CF methods gave similar cold hardiness values for stem samples from 12 provenances, the CF method may be preferred to the EC or VS method for species with chlorophyllous stems because of its greater ease of use.