GLACIATION, AN ISOLATING FACTOR IN SPECIATION

Abstract

Geograplical isolation seems to be a necessity for a to break up into daughter species. The evidence for this is so great that apparent exceptions are of particular interest, as providing opportunities to test the generalization. On continental areas when with widespread, continnous ranges break up into stubspecies the usual pattern is for a gradual change in characters, from one to the other, this area of change being wider or narrower, depending on the abruptness of ecological change in the habitat. In northern continental North America most forms agree with this, but there are a number of non-conforming examples. These are cases where two representative forms with a continuous range have evolved beyond the point where they behave as subspecies to each other, but have not yet attained all the attributes of in respect to each other. There are two alternative explanations. Either they have evolved in situ without the aid of geographical isolation, or some factor in their history has provided an isolation since removed. It is such cases, from the northern part of North America, which I discuss below, with the suggested explanation that former isolation provided by the ice age is sufficient to account for present conditions. Much of northern North America has been occupied, or re-occupied, onily recently by its present biota. Presumably before the ice age there was a northern transcontinental belt occupied by plants and animals. But when glaciers came to cover most of the northern part of the continent, the plants and animals living there were reduced to existing in refugia south of the ice, and in ice-free refugia within the glaciated area. Later with the melting and retreat of the ice the biota again spread to occupy the northern part of the conitinent. One effect of this glaciation was to provTicle isolation for the fragments of populations of the various species, in which they could evolve independently, with resulting comlplexities wlhen later the descendants of these isolated populations met, after the meltinog of the ice. This aspect of speciation has received some attention particularly by botanists (see Raup, 1946, and Hulten, 1937) and has been considered it] working out the taxonomy of certain birds, notably the relation of bronzed and purple grackles Q niscalus quiscula aeneus, quiscula and stonei (Chapman, 1940) and the yellow shafted and the red shafted flickers (Colaptes auiratuts and C. cafer) (Mayr, 1942, p. 265). In the European area a similar set of phenomenla has received considerable attention (see Mayr, 1942, pp. 263-5) but in North America this subject merits additional study. In a review of the history of the North American bird fauna Mayr (1946) does not discuss the effect of glaciation. The effects of this glaciation are particularly apparent where pairs of semispecies were formed. These are the cases in which two forms, or groups of forms', nmeet along a narrower or wider belt, and their relationships to each other are neither those of species, nor those of subspecies, but combine some characteristics of eaclh. They seem to be, and in all probability are, at a stage of evolution between that of and of subspecies. It has been suggested that these are species in the making, and present day barriers, such as the Rocky Mountains,. are sufficient to account for this process (Mayr, 1940, p. 260). However from personal examination of muclh of this