70
DAVID D. KECK
Autoploidy.
This is essentially an intraspecific phenomenon; amphiploidy is interspecific. Autoploids rarely have taxonomic significance.
It can be seen that an autoploid which arises from an essentially
homozygous "diploid" has limited potentialities (Stebbins, 1940). There
are at least two ways, however, in which an autoploid can have a significant evolutionary future: (1) if it should arise by somatic doubling
from an intervarietal hybrid exhibiting hybrid vigor, in which case it
might augment this vigor (Stebbins, 1947); and (2) if, due to its increased number of chromosomes, it is able to hybridize with other, less
closely related polyploids. At the higher levels of polyploidy, types are
often found that recombine the characteristics of autoploidy and amphiploidy. For example, such plants may have one genome doubled plus a
third genome. The term autoallopolyploid has been given to such a
form. As pointed out by Stebbins (1947), Phleum pratense may be one
example of a wild species that arose in this fashion, but presumably
there are very many of them. Several variations on this pattern are
known, and Stebbins has treated representative cases.
Fortunately for the cytotaxonomist most species have only one
chromosome number. The difficulties arise when a species has chromosome races or other deviations in chromosome number. If such chromosome races actually belong within one species the expectation is that
they usually are autoploid, and morphologically the plants may be only
obscurely distinguishable from their diploid counterparts. Wilkinson
(1944) cites several examples among the British willows in which polyploidy occurs within a given species unaccompanied by taxonomic differences, and he refers to others in the literature. The well-known forage
grass, Holcus mollis, was thought for nearly 2 decades to be merely a
diploid, with 7 pairs of chromosomes. But studies at the Welsh Plant
Breeding Station (Jones, 1954) disclose that in Wales, at least, somatic
chromosome numbers run 28, 35, 42, and 49, with the pentaploid, 2n =
35, probably the most common and most widely distributed type of the
four. In this complex the tetraploid appears to be autoploid, but the
other polyploids, on cytological and morphological evidence, may be of
amphiploid origin, with the morphologically similar diploid H. lanatus
as the other parent.
Usually autoploids are detectably different from their diploid counterparts by their somewhat larger size, greater vigor, tendency to be longerlived, and also, when compared with their diploid ancestors, they can
have a different geographic distribution (Müntzing, 1936). This is well
illustrated by the case of Cuthbertia gramínea of the southeastern United
States, which contains natural autoploid races (Giles, 1942). The di-
DAVID D. KECK
Autoploidy.
This is essentially an intraspecific phenomenon; amphiploidy is interspecific. Autoploids rarely have taxonomic significance.
It can be seen that an autoploid which arises from an essentially
homozygous "diploid" has limited potentialities (Stebbins, 1940). There
are at least two ways, however, in which an autoploid can have a significant evolutionary future: (1) if it should arise by somatic doubling
from an intervarietal hybrid exhibiting hybrid vigor, in which case it
might augment this vigor (Stebbins, 1947); and (2) if, due to its increased number of chromosomes, it is able to hybridize with other, less
closely related polyploids. At the higher levels of polyploidy, types are
often found that recombine the characteristics of autoploidy and amphiploidy. For example, such plants may have one genome doubled plus a
third genome. The term autoallopolyploid has been given to such a
form. As pointed out by Stebbins (1947), Phleum pratense may be one
example of a wild species that arose in this fashion, but presumably
there are very many of them. Several variations on this pattern are
known, and Stebbins has treated representative cases.
Fortunately for the cytotaxonomist most species have only one
chromosome number. The difficulties arise when a species has chromosome races or other deviations in chromosome number. If such chromosome races actually belong within one species the expectation is that
they usually are autoploid, and morphologically the plants may be only
obscurely distinguishable from their diploid counterparts. Wilkinson
(1944) cites several examples among the British willows in which polyploidy occurs within a given species unaccompanied by taxonomic differences, and he refers to others in the literature. The well-known forage
grass, Holcus mollis, was thought for nearly 2 decades to be merely a
diploid, with 7 pairs of chromosomes. But studies at the Welsh Plant
Breeding Station (Jones, 1954) disclose that in Wales, at least, somatic
chromosome numbers run 28, 35, 42, and 49, with the pentaploid, 2n =
35, probably the most common and most widely distributed type of the
four. In this complex the tetraploid appears to be autoploid, but the
other polyploids, on cytological and morphological evidence, may be of
amphiploid origin, with the morphologically similar diploid H. lanatus
as the other parent.
Usually autoploids are detectably different from their diploid counterparts by their somewhat larger size, greater vigor, tendency to be longerlived, and also, when compared with their diploid ancestors, they can
have a different geographic distribution (Müntzing, 1936). This is well
illustrated by the case of Cuthbertia gramínea of the southeastern United
States, which contains natural autoploid races (Giles, 1942). The di-