The genus Gossypium has received considerable attention from botanists as a result of its well-known economic importance as a source of cotton fiber. However, continuing study of the genus has brought to light new taxa, and revealed new relationships among taxa that have yet to be incorporated into a taxonomic framework. Moreover, existing classifications, apart from this need for updating in a purely chronological sense, all have significant flaws that the present treatment hopes to correct. The first comprehensive taxonomic study of the genus was that of Todaro (1863, 1877). His important contributions were followed by a major treatise by Watt (1907), a significant conceptual advance by Zaitzev (1928) and Mauer (1930), and such relatively modern works as those by Hutchinson (1947), Prokhanov (1947), Wouters (1948), Roberty (1942, 1946, 1950), and Mauer (1950, 1954). Hutchinson's (1947) treatment is perhaps the most widely used and generally accepted of the modern treatments because it is useful i.e., it is in broad outline conceptually sound. Yet, many problems have resulted from its nomenclatural inadequacy. The treatments of Wouters (1948) and Mauer (1954) are generally quite similar to that of Hutchinson in these respects. On the other hand, the classifications presented by Prokhanov (1947) and Roberty (1942, 1946, 1950) have in general a technically satisfactory nomenclature, but are structurally so chaotic as to be neither conceptually adequate nor useful. It is my intention to provide in outline form a nomenclatural basis for a taxonomic view of Gossypium that differs only by minor modifications from the views presented by Hutchinson (1947) and Mauer (1954). In addition to simple up-dating and nomenclatural corrections in both cases, my treatment differs from that of Hutchinson principally in providing a more detailed infra-generic breakdown that more sensitively reflects known relationships, in changing certain species alignments (e.g., G. armourianum, G. harknessii, G. gossypioides, G. areysianum), nd in including the several Australian species of section Hibiscoidea that he omitted; my treatment differs from that of Mauer principally in changing certain species limits (e.g., the recognition of G. thurberi, the reduction of G. ellenbeckii and G. californicum), and in the exclusion of G. bakeri Watt [= Senra bakeri (Watt) Prokh.] and of section Thespesiastra Tod., which has been shown (Fryxell 1965) to belong in Thespesia Sol. ex Corr. These views lean heavily upon cytogenetic findings, as well as upon geographical and morphological criteria. Indeed, these three kinds of data are strongly correlated. The tetraploid species are separated at the subgeneric level (subgenus Karpas), and the remaining diploid species are segregated into three additional subgenera, coordinate with the major continental divisions of the genus, as follows: subgenus Sturtia Australia; subgenus Gossypium Africa; and subgenus Houzingenia the Americas. Full synonymies are given for all of the taxa, except for the four species that are in cultivation: 21. G. herbaceum L., 22. G. arboreum L., 33. G. hirsutum L., and 34. G. barbadense L. The nomenclature of these four species is too complex and involves too many unresolved problems to be included at this time. Continuing study is being devoted to these questions (e.g., Fryxell 1968, 1969), and a monographic treatment is projected. * Geneticist, Crops Research Division, Agricultural Research Service, U.S. Departmen of Agriculture, College Station, Texas.
Read full abstract