Based on cytological evidence, Taylor and others [1964) postulated the evolution 

 of A. tvidentata subsp. tridentata from A, tridentata subsp. vaseyana for the British 

 Columbia populations of these two subspecies. In British Columbia, A. tridentata subsp. 

 vaseyana populations are uniformly diploid, n = 9, whereas A. tridentata subsp. tri- 

 dentata are all tetraploid, n = 18. No introgression between subspecies was observed 

 by Taylor and others (1964) in contrast to the relatively common introgressed popula- 

 tions of the Great Basin. In the Great Basin and adjacent areas, diploid and tetraploid 

 populations of both subspecies occur (Ward 1953; Winward 1970; unpublished data^) . No 

 logical pattern of chromosomal evolution is apparent for the Tridentatae as a whole 

 either in the Great Basin or over its complete range of distribution. A cytological 

 study should help to clarify the evolutionary past of the Great Basin Tridentatae. 



The phylogenetic pathway of taxa similar to A. higlovii '^tridentata lie — 



vaseyana la including evolutionary adaptive radiations from subgroups tridentata lie 

 and vaseyana la is suggested for the Great Basin Tridentatae by chromatographic evidence 

 presented here (fig 40). Hybridi zation-polyploidization cycles such as exhibited in 

 section Tridentatae result in reticulate evolutionary patterns (Ward 1953) . 



Artemisia nova, A. arbusaula, A. oana, and A. tripartita apparently represent two 

 separate evolutionary lines from subgroups vaseyana la (fig. 40). The latter two 

 species not only appear chromatographical ly similar to vaseyana la, but they largely 

 occupy similar habitats. Artemisia oana and A. tripartita share a common chromatographic 

 spot, 51, which is not found in other Artemisia species (table 1). Since it is un- 

 likely that the same spot should have a separate origin in each, these species probably 

 had a common unknown ancestor, or one evolved from A. tridentata subsp. vaseyana and 

 became the parent of the second. Similar problems are encountered in determining the 

 origin of A. nova and A. arbuscuta. The common ancestry of these two species is 

 suggested by spots 52 and 53, not contained in chromatograms of A. tridentata subsp. 

 vaseyana, but present in chromatograms of both A. nova and A. arhusaula (table 1). 

 The connecting link to A. tridentata subsp. vaseyana may be through either of these 

 species or from an unknown ancestor. The older of these two lines is probably that of 

 A^ oana and A. tripartita since these species exhibit the greatest chromatographic 

 dissimilarity to A. tridentata subsp. vaseyana. Artemisia nova and A. arhusaula are 

 more similar to A. tridentata subsp. vaseyana and so may have more recent origin. 



The striking similarity of chromatograms of A. rothrookii to those of vaseyana la 

 and vaseyana Ic suggests a close relationship among them (table I; figs. 2, 4, 23, and 

 40; Ward 1953; Beetle 1960). Therefore, this species must have evolved from one of the 

 higher elevation subspecies vaseyana types, probably vaseyana Ic, since the elevational 

 range of this subgroup is nearest that of A. rothrookii. 



The relationship of A. pygmaea to other members of the section Tridentatae is not 

 clear. Rydberg (1916) placed this species in a separate section, Pygmaeae , although 

 more recent authors (Hall and Clements 1923;_Ward 1953; and Beetle 1960) have considered 

 it to be included in Tridentatae. Beetle (1960) suggests that it may have an early 

 link with A. nova; however, chromatographical ly, there is no evidence of such a relation- 

 ship. Chromatograms of this species bear greater resemblance to A. higlovii or sub- 

 group tridentata IIc than to other Tridentatae species (fig. 21) . The individual plants 

 of the species tend to be dwarf and usually occur on severe sites impregnated by a 

 fairly high degree of alkalinity. 



^E. Durant McArthur. Artemisia cytological data on file at Intermountain Forest 

 and Range Experiment Station, Ephraim, Utah. 



22 



ill 



