TRENDS IN SYSTEMATIC BOTANY
79
of these widely separated areas. The Asian and American populations
in these instances are absolutely indistinguishable to the taxonomist,
although all the evidence indicates that these nonruderal species have
existed in two completely isolated areas for at least ten million years!
This speaks for the great genetic stability of these species under relatively stable climatic conditions.
Then there are examples of species that have diverged from each other
morphologically under geographic isolation but have retained complete
interfertility. One extreme case is Platanus orientalis of Asia Minor,
which crosses with P. occidentalis of the eastern United States to produce
the London Plane, often called P. acerifolia, which is fully fertile and
exceptionally hardy, with normal chromosome pairing at meiosis (Sax,
1933) and segregation in F 2 . The fossil record shows that the genus
goes back to the Cretaceous, and that the parents have been separated
perhaps that long or at least since Miocene times. They both have 21
pairs of chromosomes (possibly amphiploid in origin) and differ by
many characters affecting all parts of the tree. Some experimentalists
would reduce them to subspecies of a single species because of the
interfertility, and one can agree that they have a close genetic, although
a distant phylogenetic, relationship.
At least as striking is the case of Layia discoidea Keck inedited. This
Composite is discoid and according to its technical characters one would
assume that it belonged in the tribe Helenieae. Lacking ray-florets and
their characteristic outer involucre, it does not have the technical characters to place it in the subtribe Madiinae and the tribe Heliantheae, yet
that is where it goes. It is clearly a member of the genus Layia by the
sum total of its characteristics, although there are no other discoid forms
in the entire subtribe, and no other Layia has a pappus that is exactly like
this one. Even though ray-florets are not found in it, L. disccidea has
genes governing the color and shape of these organs. Its closest relative
is L. glandulosa
(Hook.) Η. & Α., as shown by the production of a
completely fertile F x hybrid and a large and very vigorous segregating
F 2 generation, and this relationship is verified by morphological analysis.
Nevertheless, these two forms are very dissimilar morphologically and
they have strongly contrasting edaphic preferences—very sandy soil
for L. glandulosa, serpentine for L. discoidea. From the crossing experiments it was estimated that some 20 pairs of genes are responsible for
the differences between them. The two species occur within 15 or so
miles of each other, but their ecologic separation is apparently complete.
Judging from the extent of their differences, it seems highly probable
that they have had long separate phylogenetic histories without having
79
of these widely separated areas. The Asian and American populations
in these instances are absolutely indistinguishable to the taxonomist,
although all the evidence indicates that these nonruderal species have
existed in two completely isolated areas for at least ten million years!
This speaks for the great genetic stability of these species under relatively stable climatic conditions.
Then there are examples of species that have diverged from each other
morphologically under geographic isolation but have retained complete
interfertility. One extreme case is Platanus orientalis of Asia Minor,
which crosses with P. occidentalis of the eastern United States to produce
the London Plane, often called P. acerifolia, which is fully fertile and
exceptionally hardy, with normal chromosome pairing at meiosis (Sax,
1933) and segregation in F 2 . The fossil record shows that the genus
goes back to the Cretaceous, and that the parents have been separated
perhaps that long or at least since Miocene times. They both have 21
pairs of chromosomes (possibly amphiploid in origin) and differ by
many characters affecting all parts of the tree. Some experimentalists
would reduce them to subspecies of a single species because of the
interfertility, and one can agree that they have a close genetic, although
a distant phylogenetic, relationship.
At least as striking is the case of Layia discoidea Keck inedited. This
Composite is discoid and according to its technical characters one would
assume that it belonged in the tribe Helenieae. Lacking ray-florets and
their characteristic outer involucre, it does not have the technical characters to place it in the subtribe Madiinae and the tribe Heliantheae, yet
that is where it goes. It is clearly a member of the genus Layia by the
sum total of its characteristics, although there are no other discoid forms
in the entire subtribe, and no other Layia has a pappus that is exactly like
this one. Even though ray-florets are not found in it, L. disccidea has
genes governing the color and shape of these organs. Its closest relative
is L. glandulosa
(Hook.) Η. & Α., as shown by the production of a
completely fertile F x hybrid and a large and very vigorous segregating
F 2 generation, and this relationship is verified by morphological analysis.
Nevertheless, these two forms are very dissimilar morphologically and
they have strongly contrasting edaphic preferences—very sandy soil
for L. glandulosa, serpentine for L. discoidea. From the crossing experiments it was estimated that some 20 pairs of genes are responsible for
the differences between them. The two species occur within 15 or so
miles of each other, but their ecologic separation is apparently complete.
Judging from the extent of their differences, it seems highly probable
that they have had long separate phylogenetic histories without having