18. Glaciation and Avian Evolution
285
ing the Pleistocene, rather than accelerated, as previous theories predicted. It is
clear that we have much to learn about avian evolution during the recent past (see
Avise and Walker 1998).
4.2 Glaciers and Phylogeography
Simple comparisons of current breeding ranges of birds, and the extent of the
Laurentide Ice Sheet (Pielou 1991), reveal that bird distributions have changed
dramatically over the past 15,000 years. Although the precise distribution of habitats during the height of the last glacial period is unknown, it seems likely that
most habitats (perhaps excluding marsh) expanded from refugia as the glacier receded. Consequently we predict that many bird species experienced population increases as they recolonized deglaciated areas. Thus our null prediction is that the
genetic imprints of Late Pleistocene glaciations will be those associated with population increases.
Many studies have found unstructured haplotype trees within North American
species (Zink 1996, 1997), including the Song Sparrow (Fry and Zink 1998). Such
trees are consistent with recent population increases. Some species, such as the Fox
Sparrow (Passerella iliaca), show strong phylogeographic structure, but it is not
clear whether this is in fact one species owing to the relative depth of the mtDNA
divisions (Zink 1994). If widespread panmictic species were indeed fractured by
the Laurentide Ice Sheet, it is possible that we could not as yet detect genetic divisions (Avise and Walker 1998). It takes approximately 4Ne generations for a lineage, once split to yield daughter species that have mtDNA haplotypes that exhibit
reciprocal monophyly (Ne = effective population size). This process could require
greater than 18,000 yr. For example, Ball and Avise (1992) suggested an Ne value
of 52,000 for the Song Sparrow, and with a generation time of 1-2 yr, it might
require 200,000 yr for evidence of a Wisconsinian split to be detected with our
current mtDNA surveys. Thus the unstructured haplotype tree for the Song Sparrow could reflect an insufficient amount of time post-barrier. Alternatively, even if
Song Sparrows were isolated in refugia during the Wisconsinian, subsequent gene
exchange between refugial popUlations would negate the isolating effect of the glacier; Zink and Dittmann (1993) suggested that gene flow was high. Hence, although unstructured haplotype trees are consistent with post-glacial population
expansion, they might also reflect polymorphism in pre-glaciation populations. Other
analyses are required. Clearly, if isolation induced during the last glacial cycle has
not yet yielded detectable genetic imprints within species, it did not cause speciation itself.
Few mismatch distributions have been computed for North American birds (Zink
1997). The mismatch distribution for the Song Sparrow (Fig. 4) exhibits the general shape of an exponentially expanding population but differs from the theoretical expectation. Possibly, populations of Song Sparrows have been increasing but
not exponentially, which the predicted curve represents (Slatkin and Hudson 1991).
Furthermore, some geographic heterogeneity (Fry and Zink 1998) might cause de-
285
ing the Pleistocene, rather than accelerated, as previous theories predicted. It is
clear that we have much to learn about avian evolution during the recent past (see
Avise and Walker 1998).
4.2 Glaciers and Phylogeography
Simple comparisons of current breeding ranges of birds, and the extent of the
Laurentide Ice Sheet (Pielou 1991), reveal that bird distributions have changed
dramatically over the past 15,000 years. Although the precise distribution of habitats during the height of the last glacial period is unknown, it seems likely that
most habitats (perhaps excluding marsh) expanded from refugia as the glacier receded. Consequently we predict that many bird species experienced population increases as they recolonized deglaciated areas. Thus our null prediction is that the
genetic imprints of Late Pleistocene glaciations will be those associated with population increases.
Many studies have found unstructured haplotype trees within North American
species (Zink 1996, 1997), including the Song Sparrow (Fry and Zink 1998). Such
trees are consistent with recent population increases. Some species, such as the Fox
Sparrow (Passerella iliaca), show strong phylogeographic structure, but it is not
clear whether this is in fact one species owing to the relative depth of the mtDNA
divisions (Zink 1994). If widespread panmictic species were indeed fractured by
the Laurentide Ice Sheet, it is possible that we could not as yet detect genetic divisions (Avise and Walker 1998). It takes approximately 4Ne generations for a lineage, once split to yield daughter species that have mtDNA haplotypes that exhibit
reciprocal monophyly (Ne = effective population size). This process could require
greater than 18,000 yr. For example, Ball and Avise (1992) suggested an Ne value
of 52,000 for the Song Sparrow, and with a generation time of 1-2 yr, it might
require 200,000 yr for evidence of a Wisconsinian split to be detected with our
current mtDNA surveys. Thus the unstructured haplotype tree for the Song Sparrow could reflect an insufficient amount of time post-barrier. Alternatively, even if
Song Sparrows were isolated in refugia during the Wisconsinian, subsequent gene
exchange between refugial popUlations would negate the isolating effect of the glacier; Zink and Dittmann (1993) suggested that gene flow was high. Hence, although unstructured haplotype trees are consistent with post-glacial population
expansion, they might also reflect polymorphism in pre-glaciation populations. Other
analyses are required. Clearly, if isolation induced during the last glacial cycle has
not yet yielded detectable genetic imprints within species, it did not cause speciation itself.
Few mismatch distributions have been computed for North American birds (Zink
1997). The mismatch distribution for the Song Sparrow (Fig. 4) exhibits the general shape of an exponentially expanding population but differs from the theoretical expectation. Possibly, populations of Song Sparrows have been increasing but
not exponentially, which the predicted curve represents (Slatkin and Hudson 1991).
Furthermore, some geographic heterogeneity (Fry and Zink 1998) might cause de-