New Combinations in Carex Section Acrocystis (Cyperaceae) J. H. Rettig Biology Department, College of the Ozarks, Pt. Lookout, Missouri 65726, U.S.A. William J. Crins Ontario Ministry of Natural Resources, P.O. Box 9(XX), Huntsville, Ontario PIH 2A3, Canada ABSTRACT. Carex communis and C. amplisquama can be distinguished based on achene micromor- phology. However, similarity in macromorphology and flavonoid chemistry suggest they are best treat- ed as two varieties of Carex communis. Carex ru- gosperma and C. tonsa are morphologically closely related and are best treated as two varieties of a single species. For reasons of nomenclatural pri- ority, the new combination C. tonsa var. rugosperma is required. Carex tonsa also is lectotypified. Carex communis L. Bailey is a wide-ranging tax- on of eastern North America, whereas C. amplis- quama F. J. Hermann is restricted to northern Georgia and extreme western South Carolina. The most striking feature of C. amplisquama is the large pistillate scales. The two taxa are very difficult to distinguish, and some specimens assignable to C. communis based on other macromorphological characters have pistillate scales that approach the length of those in C. amplisquama. Principal component (PC) and cluster analyses of macromorphological characters revealed overlap between the two taxa; however, study of a discrim- inant function analysis resulted in complete sepa- ration (Rettig, 1988). Pistillate scales of Carex am- plisquama are significantly longer than those of C. communis (p < 0.001), although there is some over- lap (Rettig, 1988). Carex amplisquama also is al- ways densely caespitose with culms erect and arch- ing over at the tip, whereas C. communis is usually loosely caespitose with prostrate culms. This field character ("lost" in pressing) is especially useful in separating taxa when combined with pistillate scale length. Ranges of 10 achene micromorphological char- acters examined in the multivariate study over- lapped; therefore no characters could be considered diagnostic, although five characters were signifi- cantly different (Rettig, 1988). PC and cluster anal- yses show complete separation of the two taxa with- out any overlap: all individuals of Carex amplisquama clustered together before clustering with individuals of C. communis. Some plants of Carex communis produce Luteo- lin 7-Methyl ether 4'-diglucoside, a Luteolin 5-sub- stituted glycoside and a 5-substituted flavone gly- coside not found in C. amplisquama (Rettig, 1988). Multivariate analyses did not separate the two va- rieties into groups that correspond to morphological delimitations; however, C. amplisquama specimens were grouped together. Achene micromorphology provides the strongest evidence for two distinct taxa. However, similarity in macromorphology and overall flavonoid chemis- try suggest that they are best treated as two varie- ties. Carex communis L. Bailey var. amplisquama (F J. Hermann) J. Rettig. comb. nov. Basionym: Carex amplisquama F. J. Hermann, Rhodora 57: 158. 1955. TYPE: U.S.A. Georgia: Gilmer Co., J. II. Pyron & Rogers McVaugh 2951 (ho- lotype, US; isotypes, GA, MICH not seen). KEY o TO TE VAII~tt oF (,I. o 'at' riOunMs'IS la. Pistillate scale shorter than perigynium or ex- tending beyond the terigyniurn no more than 0.8 mm; plants usually loosely caespitose with lax culms ............ C. communis var. communis Ih. Pistillate scale extending lwyound the pwerigyniumn more than 0.8 am; plants usually densely caes- pitose with culms erect to arching ......... .............. co. nmmunlu var. amnplistquoma North American caricologists acknowledge Carex rugosperma Mackenzie and C. tonsa (Fernald) Bicknell to be very close relatives. The taxa differ qualitatively in leaf texture, color, and indumrnt, and in the degree of pubescence of the perigynia. with C. tonsa having coriaeteous, light green, smooth leaves and perigynia that are virtually gla- brous. Fernald (190)2) was the first to combine these taxa at varietal rank, although at the time the typ- ification of C. umbellat Schkuhr ex W1illdent w had not yet been resolved (C. umbellata is now known to be a related but distinct species with short perig- NovoN 6: 117-118. 1996. Novon ynium beaks). Voss (1966) also recognized the close relationship between C. rugosperma and C. tonsa and treated the latter as a variety of C. rugosperma. We concur with the view that these taxa are clos- est relatives. However, the epithet "tonsa" predates "rugosperma" at the rank of species, and therefore a combination of C. rugosperma within C. tonsa is required. Carex tonsa (Fernald) Bicknell, Bull. Torrey Bot. Club 35: 492. 1908. Basionym: Carex umbel- lata Schkuhr ex Willdenow var. tonsa Fernald, Proc. Amer. Acad. 37: 507. 1902. TYPE: U.S.A. Connecticut: C. B. Graves s.n. (lecto- type, selected here, GH; isolectotype, GH). Carex tonsa (Femald) Bicknell var. rugosperma (Mackenzie) Crins, comb. nov. Basionym: Car- ex rugosperma Mackenzie, Bull. Torrey Bot. Club 42: 621. 1915. TYPE: U.S.A. New Jer- sey: Tuckerton, May 1911, K. K. Mackenzie 9871 (holotype, NY). The epithet Carex tonsa requires lectotypifica- tion. One of three sheets collected by C. B. Graves in Connecticut is here designated as the lectotype. It is one of the syntypes cited by Fernald (1902) and contains ample material with characteristic thick, relatively smooth foliage, and typical long- beaked, nearly glabrous, mature perigynia. The plant at the bottom left corner of the sheet is se- lected as the type. Acknowledgments. We thank A. A. Reznicek, Peter Ball, John McNeill, and Ed Voss for helpful comments on this manuscript, and the curators of several herbaria for facilitating loans of critical ma- terial. Literature Cited Fernald, M. L. 1902. The variations of some boreal Car- ices. Proc. Amer. Acad. Arts Sci. 37: 495-510. Rettig, J. H. 1988. A biosystematic study of the Carex pensylvanica group (section Acrocystis, Cyperaceae) in North America. Ph.D. Dissertation, University of Geor- gia, Athens. Voss, E. G. 1966. Nomenclatural notes on monocots. Rhodora 68: 435-463. 118