Use of Phenology for Examining Provenance Transfers in Reforestation of Douglas-Fir

Abstract

'Seed transfer' in reforestation is the process of moving seed, or seedlings, from place of origin to alternative planting sites. Strong support for the theory of optimality of local populations has been presented by Langlet (1936) and Clausen, Keck & Hiesey (1948). Consequently, use of non-local seed is avoided if possible to minimize losses of vigour or fitness; but in cases where local seed is not available, transfer guides are necessary. For Douglas-fir (Pseudotsuga menziesii (Mirb.) Franco) in western Washington and Oregon, present guides are rules of thumb taken from Swedish models. Therefore, a method for early examination of effects of moving seed would be helpful. This report presents a simple predictive model which uses timing of vegetative bud burst for examining responses of seedlings to transfer. The main assumptions included in the model were: (1) genetic differentiation in timing of bud burst is in response to adaptive selection; (2) the timing of bud burst in an adapted phenotype is a function of the physiological response of that provenance to temperature; (3) chilling requirements of all provenances will be satisfied by normal winter temperatures, regardless of the part of the region to which they are moved. Only the first assumption is examined in this report, by the method of provenancehabitat correlations. The other two assumptions are supported in the literature; for example, timing of bud burst appears to be almost exclusively temperature mediated if chilling requirements are met (Wommack 1964; Lavender & Hermann 1970). Also the chilling requirement of Douglas-fir is less than 80 days of temperatures lower than 10? C (Wommack 1964), whilst in any winter most of the region under consideration will have more than 100 days in which maximum temperatures will not rise above 120 C (Wakefield 1969). The exceptions are in south-coastal Oregon and California, a negligible part of the Douglas-fir region. The model required two sets of equations. The first, based on an experiment using forty-four provenances, describes timing of bud burst as a function of average daily temperature. The second describes pattern of average daily temperature at stations where provenances were collected. Together, the equations provided functions for predicting bud-burst dates. Predicted dates were then used to examine hypothetical effects of transferring seed along elevational and latitudinal gradients.