The purpose of this report is to present the results of observations on natural populations of two perennial, diploid species of Helianthus. This investigation has been supplemented by an analysis of chromosome behavior, comparisons of herbarium collections, experimental hybridizationis, and studies of transplants in the garden (Long 1960). The main objective has been to clarify the taxonomy of the species by bringing to bear facts obtained from diverse lines of work. A synopsis of the species in question has already been published (Long 1954), and the nomenclature of the present paper follows that of the earlier onie. In the introduction to his monograph of Ielianthits, Watson (1929) stated that related species have a most perplexing telndenicy to fade into one another. Extremes are easily recognized but there will always be large numbers of specimens that do not fully satisfy the definitioni of either of two species, and these can be as logically placed with one species as with the other. A plant taxonomist working in Indiana, Ohio, or Wisconsin is able to verify Watson's observation particularly as it applies to H. giganteus L. and H. grosseserratus Martens, the subjects of this report. Large numbers of specimenis of these species do riot fit conveniently into either category. The species are partially sympatric, occurring together in precisely those areas where taxonomic difficulties are encountered. Preliminary field studies showed that the two species can be found in the same habitat, and putative hybrids were identified in population samples from northern Indiana (Long 1955). Herbarium collections from recions of overlap suggested that hybridization was widespread. That natural hybridization could be the cause of the confusing variation was demonstrated by the production of fertile, vigorous artificial F1, F2, and backeross hybrid progeny. In 1956, 1957, and 1958 field investigations were undertaken to determine the extent of hybridization that took place between the two species, with special regard for Ohio populations. To assess properly the degree of interbreeding, population samples from Ohio were compared with those from areas where the species do not commonly occur together. Samples of giganteuts were taken in southern Michigan and of grosseserratus in central Illinois. These latter are assumed to be more or less "pure" collections of each species. Table 1 lists the materials used and their origins. Each population sample included approximately 25 individual specimens obtained by a method previously described (Long 1959). In addition, transplants were takeni for the experimental garden, soil samples made, and flower buds collected for cytoloo'ical observations. Nineteen characteristics were measured in samples, but after close study it was possible to measure species-differences adequately by means of 8 characteristics. A hybrid index based on 6 of these was constructed. The characters and their relative values are given in Table 2. Cytological preparations were obtained by using immature flower buds
Read full abstract