202
J. HESLOP-HARRISON
interpreted as an evolutionary cul-de-sac (Darlington, 1939). It seems
probable, however, that autogamy is rarely complete, and many species
appear to benefit from an ability to interpolate outbreeding episodes
between long periods of selfing, the condition Fryxell (1959) has called
cyclical autogamy. The outbreeding episodes result in immediate increase in the level of heterozygosity, leading to the segregation of new
recombinant genotypes in future generations. In offering new pabulum
for selection, this facilitates further ecological adaption. The restoration
of the autogamous habit rapidly reduces heterozygosity again, but the
opportunity is available for the fixation of new adapted variants as pure
lines.
The sporadic release of variation in this manner is well-known both in
wild and cultivated autogams. The control must necessarily be an external one, and several environmental factors, alone or in consort, may
be concerned. Environmental control of cleistogamy has been known
since Darwin’s day, and Uphof (1938) has listed many examples. Cleistogamy is effective in promoting self-pollination, but it is not necessarily
an adaption to this end, since it can also serve as a device permitting
reproduction under circumstances adverse enough to preclude normal
pollination (Stebbins, 1957). In many described examples, the norm of a
species is chasmogamy and outbreeding, cleistogamy resulting from
unfavourable conditions of soil or moisture supply (e.g. stipa leucotricha, Brown, 1952). Of more interest from the evolutionary point of
view are cases where the norm is self-pollination, cross-pollination
being the rarer alternative. In the facultatively cleistogamous grass
Bromus carinatus, Harlan (1945a, b) found that chasmogamous panicles
were formed only under the most favourable growth conditions. Most
wild stands sampled in California and Arizona consisted of one or a few
practically homozygous lines, but highly heterozygous hybrids were
occasionally produced in consequence of chasmogamy and outbreeding.
These hybrids Harlan considered could act as the progenitors of “race
swarms’’ from which new, uniform populations might be derived.
A similar variation pattern to that described in Bromus carinatus
typifies many of the short-lived tropical and sub-tropical Andropogoneae, and environmentally governed facultative inbreeding mechanisms have been observed in these also. Bothriochloa decipiens, reported
by Blake (1944) as commonly cleistogamous in the wild, possesses an
adaptation ensuring the discharge of pollen within the flower. This takes
the form of a glume pit which prevents the emergence of the single
anther and ensures its dehiscence in contact with the stigmas if the
glumes do not open at anthesis (Heslop-Harrison, 1961). The extension
of the stem determines whether or not the inflorescence is exserted, and
so the possibility of chasmogamy; this is subject to control by photo-
J. HESLOP-HARRISON
interpreted as an evolutionary cul-de-sac (Darlington, 1939). It seems
probable, however, that autogamy is rarely complete, and many species
appear to benefit from an ability to interpolate outbreeding episodes
between long periods of selfing, the condition Fryxell (1959) has called
cyclical autogamy. The outbreeding episodes result in immediate increase in the level of heterozygosity, leading to the segregation of new
recombinant genotypes in future generations. In offering new pabulum
for selection, this facilitates further ecological adaption. The restoration
of the autogamous habit rapidly reduces heterozygosity again, but the
opportunity is available for the fixation of new adapted variants as pure
lines.
The sporadic release of variation in this manner is well-known both in
wild and cultivated autogams. The control must necessarily be an external one, and several environmental factors, alone or in consort, may
be concerned. Environmental control of cleistogamy has been known
since Darwin’s day, and Uphof (1938) has listed many examples. Cleistogamy is effective in promoting self-pollination, but it is not necessarily
an adaption to this end, since it can also serve as a device permitting
reproduction under circumstances adverse enough to preclude normal
pollination (Stebbins, 1957). In many described examples, the norm of a
species is chasmogamy and outbreeding, cleistogamy resulting from
unfavourable conditions of soil or moisture supply (e.g. stipa leucotricha, Brown, 1952). Of more interest from the evolutionary point of
view are cases where the norm is self-pollination, cross-pollination
being the rarer alternative. In the facultatively cleistogamous grass
Bromus carinatus, Harlan (1945a, b) found that chasmogamous panicles
were formed only under the most favourable growth conditions. Most
wild stands sampled in California and Arizona consisted of one or a few
practically homozygous lines, but highly heterozygous hybrids were
occasionally produced in consequence of chasmogamy and outbreeding.
These hybrids Harlan considered could act as the progenitors of “race
swarms’’ from which new, uniform populations might be derived.
A similar variation pattern to that described in Bromus carinatus
typifies many of the short-lived tropical and sub-tropical Andropogoneae, and environmentally governed facultative inbreeding mechanisms have been observed in these also. Bothriochloa decipiens, reported
by Blake (1944) as commonly cleistogamous in the wild, possesses an
adaptation ensuring the discharge of pollen within the flower. This takes
the form of a glume pit which prevents the emergence of the single
anther and ensures its dehiscence in contact with the stigmas if the
glumes do not open at anthesis (Heslop-Harrison, 1961). The extension
of the stem determines whether or not the inflorescence is exserted, and
so the possibility of chasmogamy; this is subject to control by photo-