Experimental Data for a Revision of the North American Wild Roses
- 1 December 1934
- journal article
- research article
- Published by University of Chicago Press in Botanical Gazette
- Vol. 96 (2) , 197-259
- https://doi.org/10.1086/334469
Abstract
Transplant experiments, the cultivation of several offspring of individual wild plants, pollen analysis and cytological examinations were employed to test the spp. of Rosa in America. Many of the characteristics commonly used to distinguish spp. proved to be fluctuating and minor variations which sometimes vary on a single plant. Spp. of the Linnean type exist in Rosa; some are distinctive and easily delimited, others are highly variable polymorphic spp. The latter have wide ranges and show parallel series of varieties. Chromosome number is an important diagnostic characteristic, but cytology has not radically affected classification. The native American roses belong to 3 sections: (1) Synstylae[long dash]1 sp.; (2) Minutifoliae[long dash]3 spp.; (3) Cin-namomeae[long dash]all other spp. Crepin''s sections Carolinae and Gymnocarpae are now included in the Cinnamomeae, which section contains forms with 14, 28, 42 and 56 somatic chromosomes. Morphological and physiological characteristics are discussed with reference to their value in species differentiation. Tabular analyses of the offspring of single wild plants of R. blanda, R. woodsii, R. calijornica and R. pisocarpa show that several of the spp. described previously were usually represented in a single culture. Variability in minor characteristics and in pollen fertility was as great among the offspring of isolated plants of R. woodsii in Utah and Nevada as among offspring of R. blanda from Michigan where chances for hybridization in the field were numerous. Hybridization is probably not the most important source of variation. Giant pollen grains (diploid) as well as triploid individuals in diploid cultures have been observed in R. blanda, R. woodsii and R. pisocarpa. Sterility arose from (a) reciprocal translocation of chromosome segments, (b) internal polyploidy, (c) hybridization with polyploid spp. Interspecific F1 hybrids among 13 spp. were raised and their pollen analysed. Six spp. were identified with F1 hybrids of widespread spp. Minor parallel variations in habit, armature, indument and fruit characters belong to each Linnaean sp. and can be conveniently treated only by using numerical symbols. Sixteen spp. only are considered worthy of specific rank in the American Cinnamomeae; 6 of these are relatively stable (R. joliolosa, R. nitida, R. palustris, R. rugosa, R. spithamaea and R. virginiana), 10 are polymorphic (R. acicularis, R. arkansana, R. blanda, R. calijornica, R. Carolina, R. durandii, R. gymnocarpa, R. nutkana, R. pisocarpa, and R. woodsii). Four of these spp. are compared in a table of numerical variations. Seven other spp. are retained tentatively as ecotype spp. R. acicularis consists of a hexaploid and an octoploid race indistinguishable in herbarium material. The 16 Linnaean spp. are given alphabetically with synonyms and an analytical key. R. calijornica v. bidenticulata (R. b. Rydb.) (p.222) is proposed.This publication has 6 references indexed in Scilit:
- A Group of Tetraploid Roses in Central OregonBotanical Gazette, 1931
- Cyto-Genetics and the Species ConceptThe American Naturalist, 1931
- STERILITY IN WILD ROSES AND IN SOME SPECIES HYBRIDSGenetics, 1931
- Studies inPrunus, I and IIJournal of Genetics, 1928
- Zytologische Studien über die Gattung RosaPublished by Biodiversity Heritage Library ,1922
- HYBRIDISM AND CLASSIFICATION IN THE GENUS ROSA.New Phytologist, 1920