Morphological variation in subalpine fir and its relationship to balsam fir in western Canada and the United States

Abstract

To help clarify the taxonomy of the two closely related North American firs subalpine fir {Abies lasiocarpa (Hook.) Nutt.) and balsam fir {Abies balsamea (L.) Mill.) in western Canada and the United States, 10 populations of subalpine fir from Yukon Territory, northern British Columbia, and Alaska were compared on the basis of morphological traits to fir populations from Washington to northern Ontario. Principal components and discriminant analyses were carried out separately on cone and needle data for various subsets of these populations. Multiple regression analysis and analyses of variance were used to investigate which evolutionary processes may have been important in the recent development of subalpine fir at its northern extreme. The regression analyses and analyses of variance suggest that variation is expressed primarily within populations, throughout a single large genepool at the northern extreme of subalpine fir's range, and that selection due to local environmental pressures has produced the existing morphological variation among populations. The multivariate analyses of cone and needle data produced different results for several of the population subsets, indicating that vegetative and sexual features of these firs respond to different selection pressures at the various sites. Although the results are not entirely consistent, the combined cone and needle data indicate that 1) north coastal British Columbia and interior Rocky Mountain populations of subalpine fir are generally distinct with the exception of a Vancouver Island population; 2) northern subalpine fir populations exhibit an affinity to interior populations and are very distinct from coastal populations. Additional geographic trends are 1) two Washington Cascades subalpine fir populations are distinct from all other populations in needle morphology, but one of these populations (Mount Baker) is similar to the interior populations in cone morphology: 2) all balsam fir populations are indistinguishable from interior subalpine fir populations on the basis of needle morphology, but well-differentiated from these and coastal populations in cone morphology. These results support 1) continued recognition of subalpine and balsam fir as separate species and 2) recognition of coastal and interior subalpine fir as separate taxa, perhaps at varietal rank. The present results refute the theory that subalpine fir survived the Wisconsin glaciation in one or more northern refugia. Rather, the close affinity between northern and interior subalpine fir populations suggests that they originated from a common Rocky Mountain refugium.