FORTY YEARS OF GlENECOLOQY
205
range of pollen and propagules, and by influences governing reproductive periodicity.
At the one extreme are “colonial” species, occurring in panmictic
populations perhaps numerically large but isolated spatially from others.
This pattern is necessarily imposed on a species when suitable habitats
are widely scattered in comparison with the average dispersal range of
pollen and propagules. At the other extreme are species which tend to
form continuous wide-ranging communities. Within these, the probability of any two individuals mating depends largely upon their spatial
separation, and it may not be feasible to define breeding populations
lesser in extent than each major distributional area, as is commonly true
with widely distributed forest trees. However, it is possible to overestimate the extent of what might be termed “simultaneous” panmixis
in these species. The fact that individual pollen grains can be carried
great distances may be of limited significance, since the chance of fusion
between gametes of remote provenance is determined by the proportion
of foreign and local pollen in the atmosphere, and in a closed community
the latter is always bound to predominate overwhelmingly. The effective
pollination range may thus be quite small. Colwell (1951) found that
extremely little of the pollen of Pinus coulteri released at a height of
12 ft was carried more than 150 f t even downwind; and Bateman (1947),
following gene flow rather than pollen movement directly, found that
with maize a distance of 60 ft was enough to reduce crossing to 1%.
Moreover, the phenological gradients in continuous wide-ranging populations of potentially inter-fertile individuals become major barriers to
long-range crossing, since they limit the distance over which a grain
released at any one point is likely to encounter a receptive stigma.
Latitudinal adjustment to photoperiod will thus contribute to limiting
panmixis. The conclusion must be that in a continuously dispersed,
wind-pollinated species the average area within which there is an appreciable chance of two individuals mating will normally be quite small
in relation to the total range. What is important is that pathways for
slow, continuous gene migration do exist, uninterrupted by barriers,
through large parts of the species area.
Most distributional patterns lie between these extremes, with local
concentrations of individuals separated by thinly inhabited belts corresponding to topographical barriers or zones of unsuitable ecology. In
terms of gene flow, whether such a situation resembles one or other
extreme depends again upon the average dispersal ranges of pollen and
propagules. With insect pollination, ecological limitation of the vectors
may be so strong as effectively to exclude gene exchange between neighbowing concentrations of individuals ; “genetic mobility” (Darlington,
1939) may then IargeIy become a function of seed dispersion. On the
205
range of pollen and propagules, and by influences governing reproductive periodicity.
At the one extreme are “colonial” species, occurring in panmictic
populations perhaps numerically large but isolated spatially from others.
This pattern is necessarily imposed on a species when suitable habitats
are widely scattered in comparison with the average dispersal range of
pollen and propagules. At the other extreme are species which tend to
form continuous wide-ranging communities. Within these, the probability of any two individuals mating depends largely upon their spatial
separation, and it may not be feasible to define breeding populations
lesser in extent than each major distributional area, as is commonly true
with widely distributed forest trees. However, it is possible to overestimate the extent of what might be termed “simultaneous” panmixis
in these species. The fact that individual pollen grains can be carried
great distances may be of limited significance, since the chance of fusion
between gametes of remote provenance is determined by the proportion
of foreign and local pollen in the atmosphere, and in a closed community
the latter is always bound to predominate overwhelmingly. The effective
pollination range may thus be quite small. Colwell (1951) found that
extremely little of the pollen of Pinus coulteri released at a height of
12 ft was carried more than 150 f t even downwind; and Bateman (1947),
following gene flow rather than pollen movement directly, found that
with maize a distance of 60 ft was enough to reduce crossing to 1%.
Moreover, the phenological gradients in continuous wide-ranging populations of potentially inter-fertile individuals become major barriers to
long-range crossing, since they limit the distance over which a grain
released at any one point is likely to encounter a receptive stigma.
Latitudinal adjustment to photoperiod will thus contribute to limiting
panmixis. The conclusion must be that in a continuously dispersed,
wind-pollinated species the average area within which there is an appreciable chance of two individuals mating will normally be quite small
in relation to the total range. What is important is that pathways for
slow, continuous gene migration do exist, uninterrupted by barriers,
through large parts of the species area.
Most distributional patterns lie between these extremes, with local
concentrations of individuals separated by thinly inhabited belts corresponding to topographical barriers or zones of unsuitable ecology. In
terms of gene flow, whether such a situation resembles one or other
extreme depends again upon the average dispersal ranges of pollen and
propagules. With insect pollination, ecological limitation of the vectors
may be so strong as effectively to exclude gene exchange between neighbowing concentrations of individuals ; “genetic mobility” (Darlington,
1939) may then IargeIy become a function of seed dispersion. On the