Discussion

North american representatives of this section (except Rumex venosus, a distinctive species) form a polymorphic complex consisting of often interfertile races (N. M. Sarkar 1958). In most cases those races are separated morphologically and restricted geographically. They may be arranged into at least three aggregates (groups) that approximate the subsections described by K. H. Rechinger (1937): R. verticillatus aggr. (subsect. Verticillati Rechinger f.), R. altissimus aggr. (subsect. Salicifolii Rechinger f., in part), and R. salicifolius aggr. (subsect. Salicifolii, in part). In case of difficulty with identification, inevitable with immature or intermediate specimens, determination to the aggregate level is recommended. Those preferring more broadly circumscribed species may use existing infrageneric combinations (see J. C. Hickman 1984). Attempts have been made to submerge various “microspecies” (especially in the R. salicifolius group) into more broadly circumscribed taxa, usually R. salicifolius and R. mexicanus, and various combinations at subspecific and varietal ranks have been proposed (see Hickman). In my opinion, results of this approach are inconsistent (e.g., some taxa are treated as varieties while others, not less distinct, are accepted as species or subspecies) and in most cases less convincing than the original treatment by Rechinger and the thorough study by Sarkar. I agree with Sarkar (p. 993) that “any drastic taxonomic revision of the species of the Axillares section of Rumex should be postponed until more complete cytogenetic data have accumulated concerning the interrelationships of all the taxa in this section.”

Discussion

North american representatives of this section (except Rumex venosus, a distinctive species) form a polymorphic complex consisting of often interfertile races (N. M. Sarkar 1958). In most cases those races are separated morphologically and restricted geographically. They may be arranged into at least three aggregates (groups) that approximate the subsections described by K. H. Rechinger (1937): R. verticillatus aggr. (subsect. Verticillati Rechinger f.), R. altissimus aggr. (subsect. Salicifolii Rechinger f., in part), and R. salicifolius aggr. (subsect. Salicifolii, in part). In case of difficulty with identification, inevitable with immature or intermediate specimens, determination to the aggregate level is recommended. Those preferring more broadly circumscribed species may use existing infrageneric combinations (see J. C. Hickman 1984). Attempts have been made to submerge various “microspecies” (especially in the R. salicifolius group) into more broadly circumscribed taxa, usually R. salicifolius and R. mexicanus, and various combinations at subspecific and varietal ranks have been proposed (see Hickman). In my opinion, results of this approach are inconsistent (e.g., some taxa are treated as varieties while others, not less distinct, are accepted as species or subspecies) and in most cases less convincing than the original treatment by Rechinger and the thorough study by Sarkar. I agree with Sarkar (p. 993) that “any drastic taxonomic revision of the species of the Axillares section of Rumex should be postponed until more complete cytogenetic data have accumulated concerning the interrelationships of all the taxa in this section.”

Discussion

North american representatives of this section (except Rumex venosus, a distinctive species) form a polymorphic complex consisting of often interfertile races (N. M. Sarkar 1958). In most cases those races are separated morphologically and restricted geographically. They may be arranged into at least three aggregates (groups) that approximate the subsections described by K. H. Rechinger (1937): R. verticillatus aggr. (subsect. Verticillati Rechinger f.), R. altissimus aggr. (subsect. Salicifolii Rechinger f., in part), and R. salicifolius aggr. (subsect. Salicifolii, in part). In case of difficulty with identification, inevitable with immature or intermediate specimens, determination to the aggregate level is recommended. Those preferring more broadly circumscribed species may use existing infrageneric combinations (see J. C. Hickman 1984). Attempts have been made to submerge various “microspecies” (especially in the R. salicifolius group) into more broadly circumscribed taxa, usually R. salicifolius and R. mexicanus, and various combinations at subspecific and varietal ranks have been proposed (see Hickman). In my opinion, results of this approach are inconsistent (e.g., some taxa are treated as varieties while others, not less distinct, are accepted as species or subspecies) and in most cases less convincing than the original treatment by Rechinger and the thorough study by Sarkar. I agree with Sarkar (p. 993) that “any drastic taxonomic revision of the species of the Axillares section of Rumex should be postponed until more complete cytogenetic data have accumulated concerning the interrelationships of all the taxa in this section.”