FORTY YEARS O F GENECOLOGY
211
trast, if the various sites have been colonized by a single race the populations may be expected to be genetically homogeneous except in so far as
secondary differentiation has progressed. Clausen (1951) reported a
study of prostrate maritime populations of Layia platyglossa of the
Californian coast in which an attempt was made to discover whether the
complex had had a monotopic or a polytopic origin. Representatives of
two phenotypically similar populations 140 miles apart were intercrossed, and a large F, family was raised. The adaptive characteristics
- prostrate habit, succulence and late flowering - appeared uniformly
in the F,, and showed no F, segregation. The F, was, however, variable
in respect to other characters such as density of internodes and number
of disk and ray florets. This result certainly suggests that the genetical
basis of the adaptive characteristics is identical in the remote populations, yet the occurrence of some F, segregation does indicate heterogeneity. Clausen concluded that the maritime race does in fact, represent
a single and rather ancient evolutionary entity. If so, the dissimilarities
between the populations in non-adaptive characteristics must have
arisen through secondary differentiation.
Where two or more races, originally allopatric, have attained coincident or overlapping ranges by secondary migration it must be supposed
that there are factors operative preventing miscegenation. The sequence
is, of course, that widely accepted as being usually involved in speciation
(Mayr, 1942) : divergence in geographic isolation, followed by the emergence of some form of reproductive isolation, followed in turn by further
migration to produce overlapping distributions. It is not within the
scope of the present paper to discuss the broader problems of isolating
mechanisms and speciation, but some consideration of the factors acting
to preserve the identity of sympatric ecological races is merited.
Ecological specialization may itself form the most potent such factor.
Where a predilection for highly distinctive habitats has been evolved,
this must impose spatial isolation to the extent that the habitats themselves are dispersed geographically. Beyond this, the fact that each race
has attained an adaptive peak for its particular habitat will mean that
recombinants will be at a disadvantage in competition with their
parents and uncontaminated progeny : they will become the victims of
stabilizing selection in each habitat. While the pattern of selective
forces remains unchanged, hybridization w i l l be disfavoured.
Many of the described cases of isolation imposed by strong habitat
specialization concern pairs or groups of entities sufficiently distinct in
morphology and distribution to have merited recognition as taxonomic
species (p. 187). Well-documented examples are Quercus ilicifolia and
Q. marilandica (Stebbins et al., 1947) and Silene vulgaris and S. maritima
(Marsden-Jones and T u r d , 1957). With both of these pairs hybridiza-
211
trast, if the various sites have been colonized by a single race the populations may be expected to be genetically homogeneous except in so far as
secondary differentiation has progressed. Clausen (1951) reported a
study of prostrate maritime populations of Layia platyglossa of the
Californian coast in which an attempt was made to discover whether the
complex had had a monotopic or a polytopic origin. Representatives of
two phenotypically similar populations 140 miles apart were intercrossed, and a large F, family was raised. The adaptive characteristics
- prostrate habit, succulence and late flowering - appeared uniformly
in the F,, and showed no F, segregation. The F, was, however, variable
in respect to other characters such as density of internodes and number
of disk and ray florets. This result certainly suggests that the genetical
basis of the adaptive characteristics is identical in the remote populations, yet the occurrence of some F, segregation does indicate heterogeneity. Clausen concluded that the maritime race does in fact, represent
a single and rather ancient evolutionary entity. If so, the dissimilarities
between the populations in non-adaptive characteristics must have
arisen through secondary differentiation.
Where two or more races, originally allopatric, have attained coincident or overlapping ranges by secondary migration it must be supposed
that there are factors operative preventing miscegenation. The sequence
is, of course, that widely accepted as being usually involved in speciation
(Mayr, 1942) : divergence in geographic isolation, followed by the emergence of some form of reproductive isolation, followed in turn by further
migration to produce overlapping distributions. It is not within the
scope of the present paper to discuss the broader problems of isolating
mechanisms and speciation, but some consideration of the factors acting
to preserve the identity of sympatric ecological races is merited.
Ecological specialization may itself form the most potent such factor.
Where a predilection for highly distinctive habitats has been evolved,
this must impose spatial isolation to the extent that the habitats themselves are dispersed geographically. Beyond this, the fact that each race
has attained an adaptive peak for its particular habitat will mean that
recombinants will be at a disadvantage in competition with their
parents and uncontaminated progeny : they will become the victims of
stabilizing selection in each habitat. While the pattern of selective
forces remains unchanged, hybridization w i l l be disfavoured.
Many of the described cases of isolation imposed by strong habitat
specialization concern pairs or groups of entities sufficiently distinct in
morphology and distribution to have merited recognition as taxonomic
species (p. 187). Well-documented examples are Quercus ilicifolia and
Q. marilandica (Stebbins et al., 1947) and Silene vulgaris and S. maritima
(Marsden-Jones and T u r d , 1957). With both of these pairs hybridiza-