Am. Jour. Bot. 28:957 (1941)
Leslie Hubricht and Edgar Anderson
1 Received for publication August 11, 1941.

TO PROVIDE precise information on relationships between Tradescantia canaliculata and Tradescantia subaspera var. typica an experimental study was made of their cross-pollination relationships. Since it was known that both of these species are self-sterile, two adjacent plants of T. subaspera var. typica in the experimental garden, which were known to be cross-fertile, were selected for the experiment. All the seed which was set during the season of 1939 was saved and as many seedlings as possible raised to maturity. Of the 105 seedlings which survived, 65 were hybrids with T. canaliculata and 40 were T. subaspera var. typica. The predominance of hybrids is all the more remarkable considering the conditions of the experiment. The two plants of T. subaspera were large and vigorous; they were side by side and flowered at the same time. The plants of T. canaliculata had earlier blooming seasons and had already passed their prime when the other species came into flower. Furthermore, they were several feet away from the plants of T. subaspera which were being tested.

We have, therefore, good experimental evidence that these two species of Tradescantia are readily cross-pollinated by natural agencies when they are growing near each other, in spite of such barriers as differences in season of bloom. This result is the more interesting since our previous studies (Anderson and Hubricht, 1938) of Missouri localities, where the species grow near each other, had shown very few hybrids. The lack of hybrids under natural conditions must, therefore, be laid in this particular instance not so much to barriers preventing intercrossing as to a lack of intermediate habitats in which the hybrid plants might develop to maturity.

It is significant that the hybrids which we had previously encountered had been at points where such intermediate situations had arisen (usually through human intervention). Tradescantia subaspera var. typica grows in shade in rich deep soil at the base of cliffs, while T. canaliculata prefers the full sunlight and thin rocky soil to be found above the cliffs. The two species sometimes grow within a few feet of each other, but in the Ozarks it is only occasionally that an intermediate habitat presents itself.

Showalter has shown (1938) that T. subaspera and T. canaliculata hybridize much more actively along the Appalachians. In the light of the above results it would seem likely that the different habitat pattern of the Appalachians more readily provides intermediate zones for hybrids between these two species. From such collecting as we have done in the region we would hazard a guess that in the moister climate of the Appalachians damp, slightly-shaded spots are to be found on the faces of the cliffs in which spiderworts can grow readily. In Missouri such situations would nearly always be too dry and barren for most flowering plants.

In Riley's (1938) analysis of introgressive hybridization in Iris the circumstantial evidence pointed in the same direction as the experimental evidence reported above. He found that in the intermediate zone between Iris fulva and Iris heragona var. giganticaerulea hybrids were many times more frequent when the zone was being pastured and lumbered. Hybrids were undoubtedly being produced at other points, but only where lumbering and pasturing produced strange new habitats were they able to survive.


Tradescantia subaspera var. typica and Tradescantia canaliculata were shown to be hybridized readily by natural agencies in an experimental garden, though they seldom do so in nature. It is suggested that a lack of intermediate habitats may be the chief barrier between these species under natural conditions.

St. Louis, Missouri