Coastal Douglas-fir provenance variation: patterns and predictions for British Columbia seed transfer

We evaluated performance variability in two series of provenance trials of Pseudotsuga menziesii var. menziesii. EP 480 included 16 reciprocal provenances planted from British Columbia to Oregon, measured up to age 45. EP 599.03 featured five common provenances, plus the local source, planted at 23 British Columbia sites, measured up to age 33. Although residual variance was high, site accounted for 31 to 60% of the volume variance, while provenance accounted for 1-6%. Interactions were also significant across ages and trials. Genotype-environment interaction was evident in EP 599.03, but not EP 480, which may reflect differences in experimental design. The worst provenances always ranked low across sites and over time. No geographic or climatic (annual, seasonal, monthly) variables consistently explained patterns of volume across sites or ages, singly or in combination for either trial, similar to findings from other studies of coastal Douglas-fir. Provenances from Washington to central Oregon often performed as well as the local provenance or better at British Columbia trial sites. Populations from higher elevations, poor sites and submaritime provenances were less vigorous. Results support maintaining elevational and ecotypic transfer limits, emphasizing site-specific decision making, and permitting wider latitudinal transfer on similar quality sites.