Flora of North America species comparison

Endolepis covillei was treated within the synonymy of Atriplex phyllostegia (Torrey) S. Watson by H. M. Hall and F. E. Clements (1923).

H. C. Stutz et al. (1993) placed Atriplex covillei within Endolepis, based in large part on the presence of a perianth subtending the pistil within the fruiting bracteoles and on the lack of Kranz anatomy in the leaves. The pattern of venation is, nevertheless, very similar to that in species with Kranz anatomy. The presence of perianth scales in the pistillate flowers of A. covillei has been regarded as evidence of relationship with A. suckleyi. Despite placement of these taxa within Endolepis by Stutz et al. Atriplex covillei is possibly more closely allied to the morphologically similar and partially sympatric A. phyllostegia than it is to strongly dissimilar and the distantly disjunct A. suckleyi. Stutz and his associates placed great emphasis on the presence of reduced perianth segments subtending the pistil within the fruiting bracteoles of A. covillei. Calyces per se, otherwise known only in A. suckleyi and A. pleiantha, probably have arisen independently. Their presence does not necessarily indicate a close relationship. Stutz et al. pointed to other differences aside from the calyx of the pistillate flowers, and it is apparent that the two entities can stand as distinct species. To segregate A. covillei within a separate genus and to ally it with a species to which its relationships are obscure at best, stretches logic beyond reason.

(Discussion copyrighted by Flora of North America; reprinted with permission.)

Varieties 7 (7 in the flora).

This is a widely distributed complex of intergrading genotypes of great phenotypic plasticity. The members occur commonly in fine-textured saline substrates in much of the western Great Plains and in the Intermountain Region. Diploids, triploids, tetraploids, and hexaploids (and higher polyploids, all multiples of the base number 9) are known within the complex, and hybrids are known not only between the constituents but with the other woody species which they contact, i.e., Atriplex canescens, A. confertifolia, and A. corrugata. Indeed, a case can be made for treating both A. gardneri and A. canescens within an expanded A. canescens. They are regarded here as forming two intergrading complexes, with some of the constituent varieties placed equally well within either of the species aggregations. The treatment essentially follows the alignment of taxa suggested by C. A. Hanson (1962), with the exception that they are reduced to varietal status and var. bonnevillensis and var. aptera are placed within the A. gardneri phase and not with A. canescens.

(Discussion copyrighted by Flora of North America; reprinted with permission.)