FORTY YEARS O F QENECOLOGY
203
period and temperature. In another grass of the Andropogoneae,
Rottboellia exaltata, both the incidence of cleistogamy and the total
production of pollen are affected by the photoperiodic regime in which
plants are grown (Heslop-Harrison, 1959b). The tendency towards selfpollination in these grasses is certainly to be associated'with the occupancy of habitats of limited permanency in areas of forest. It is of
some interest that they can be brought into a state permitting outbreeding by environmental factors, for their ecological habit may demand
not only the ability to build up large homogeneous populations when a
favourable habitat is available, but also to achieve, rather regularly,
enough recombination to permit the colonization of new habitats
suddenly made accessible by local catastrophic change in the forest
cover. Rottboellia exaltata shows great geographic variation, sometimes
on a very local scale (Hubbard, personal communication), and it would
be informative to know to what extent this is ecologically correlated.
The matter of ecological adaptation in apomictic complexes merits
comment. Obligate apomixis, by suspending entirely the capacity for
recombination, must, like obligate inbreeding, presage ultimate extinction once the circumstances for which a, lineage is adapted cease to be
available. Yet many major apomictic complexes have achieved considerable ecological success, and have done so by what is essentially
genecological differentiation (Turesson, 1943, 1956; Nygren, 1951 ;
Clausen, 1954). This must mean that some capacity for attaining recombination remains, and it is probable that in all successful apomictic
complexes some sexual potential is retained. This could be through the
survival of some sexual races acting as progenitors of apomicts, or
through the ability of mainly apomictic lineages to produce some progeny by sexual processes (Gustafsson, 1947; Stebbins, 1950). Haskell
(1953, 1959), has convincingly demonstrated quantitative variation in
ecologically significant characteristics like flowering time in the progeny
of the pseudogamous, largely diplosporous apomict Rubus nitidioides,
and has shown a response to selection. The parental clone is highly
heterozygous, and Haskell suggests that the limited and irregular
segregation observed might be explained by recombination during
oogenesis (autosegregation). With aposporous apomicts, complete sexual
competence may be preserved, so that the progeny are partly sexual and
partly clonal. Apospory is very widespread in grasses of the Paniceae,
Andropogoneae and related groups (Brown and Emery, 1958), and it
may be surmised that it is commonly facultative. In Dichanthium
aristatum of the Andropogoneae, the balance between sexuality and
apospory is subject t o environmental control (Knox and Heslop-Harrison, 1963), and there is evidence that it varies from population topopulation in the wild (Knox, personal communication).
203
period and temperature. In another grass of the Andropogoneae,
Rottboellia exaltata, both the incidence of cleistogamy and the total
production of pollen are affected by the photoperiodic regime in which
plants are grown (Heslop-Harrison, 1959b). The tendency towards selfpollination in these grasses is certainly to be associated'with the occupancy of habitats of limited permanency in areas of forest. It is of
some interest that they can be brought into a state permitting outbreeding by environmental factors, for their ecological habit may demand
not only the ability to build up large homogeneous populations when a
favourable habitat is available, but also to achieve, rather regularly,
enough recombination to permit the colonization of new habitats
suddenly made accessible by local catastrophic change in the forest
cover. Rottboellia exaltata shows great geographic variation, sometimes
on a very local scale (Hubbard, personal communication), and it would
be informative to know to what extent this is ecologically correlated.
The matter of ecological adaptation in apomictic complexes merits
comment. Obligate apomixis, by suspending entirely the capacity for
recombination, must, like obligate inbreeding, presage ultimate extinction once the circumstances for which a, lineage is adapted cease to be
available. Yet many major apomictic complexes have achieved considerable ecological success, and have done so by what is essentially
genecological differentiation (Turesson, 1943, 1956; Nygren, 1951 ;
Clausen, 1954). This must mean that some capacity for attaining recombination remains, and it is probable that in all successful apomictic
complexes some sexual potential is retained. This could be through the
survival of some sexual races acting as progenitors of apomicts, or
through the ability of mainly apomictic lineages to produce some progeny by sexual processes (Gustafsson, 1947; Stebbins, 1950). Haskell
(1953, 1959), has convincingly demonstrated quantitative variation in
ecologically significant characteristics like flowering time in the progeny
of the pseudogamous, largely diplosporous apomict Rubus nitidioides,
and has shown a response to selection. The parental clone is highly
heterozygous, and Haskell suggests that the limited and irregular
segregation observed might be explained by recombination during
oogenesis (autosegregation). With aposporous apomicts, complete sexual
competence may be preserved, so that the progeny are partly sexual and
partly clonal. Apospory is very widespread in grasses of the Paniceae,
Andropogoneae and related groups (Brown and Emery, 1958), and it
may be surmised that it is commonly facultative. In Dichanthium
aristatum of the Andropogoneae, the balance between sexuality and
apospory is subject t o environmental control (Knox and Heslop-Harrison, 1963), and there is evidence that it varies from population topopulation in the wild (Knox, personal communication).