•rK RHODORA, Vol. 95, No. 881, 25-37, 1993 pp. A TAXONOMIC -Lb, STUDY OF AND ARISTIDA STRICTA X BEYRICHIANA A, i Robert K. Peet ABSTRACT Grasses traditionally assigned to Aristida stricta Michx. Include two morpho- logically and geographically separate species. The name Aristida Michx. stricta corresponds to a taxon confined to North Carolina and the northern ofcounties tier A & South in Carolina. second species, Aristida beyrichiana Trin. Rupr., occurs from southern South Carolina south throughout Florida and west to Mississippi. Key Words: Taxonomy, Aristida beyrichiana, Aristida stricta, Poaceae, coastal North America, plain. southeastern U.S. INTRODUCTION The name Michaux Aristida has long been stricta ap] known those southeastern plants colloquially as "wiregrass dominate the understories of most freauentlv-bumed Pi and Eneelm Mill, P. elliottii southern Florida No eastern Mississippi. intraspecific variation has previously been described However, for this taxon. the range of has stricta, sA. /I. a conspicuous gap South in central Carolina, Further, differences in sheath and blade indument allow plants from north of the gap to be readily distinguished from those south of the gap. In this paper examine morphological I the characteristics, distribution, and taxonomic status of these two types of Aristida stricta. s,l For reasons explained below in the section on taxonomy, I refer to the northern plants as A, stricta Michaux and to the southern & plants as A, beyrichiana Trinius Ruprecht. MORPHOLOGICAL DIFFERENCES The two variants oi Aristida stricta, s,i can be readily and unambiguously by examination of leaf indument- differentiated On plants of the southern A, beyrichiana, a densely woolly or villous bearding or of indument present at the base of the tuft is These prominent vary from being leaf blade (Figure hairs 1). comers of much more commonly) surrounding the the (and collar to also With blade, and uppermost portion of the sheath. age the collar, 25 Rhodora 26 95 [Vol. E I ssw I: d f Figure Characteristic blade and sheath pubescence of (a-c) Aristida beyri- 1. & chiana Trinius Ruprecht and (d-f Aristida stricta Michaux. In each case the ) first (a and d) of the three figures is the most typical and the other two illustrate some of the range of variation. on can be deciduous, but they are always evident hairs partially the younger foliage and, together with the involute leaves, allow from A. beyrichiana to be distinguished in vegetative condition North American other southeastern grasses. Individual plants all amount minimum, can of indument, but a differ substantially in at plants oi A. beyrichiana have numerous hairs protruding from about the throat and comers of the collar. In contrast, the northern may At A. stricta lacks this localized bearding or tuft. most, there be a few prominent hairs at the comers of the collars of the most and on most no hairy plants, plants there are hairs except for those on the back of the blade as described below. The northern Aristida readily distinguished in stricta, s,s. is vegetative condition from other native grasses by the presence all mm indument of a line of villous (hairs .6-1.5 long) adjacent to and on each side of the midrib along the length of the back of the The involute blade (Figure hairs can be deciduous with age, 1). but are always present on the younger The of the foliage. foliage southern A, beyrichiana usually lacks indument on the backs of the leaves except for the characteristic woolly to villous tufts or The bearding at the bases of the blades. one exception that is seven percent of the Florida specimens examined of 25 had (17 1) foliar indument along the midrib; the was not observed in trait collections of beyrichiana from any other In those few yl. state. Florida plants where the lines of villous indument were present on wooly the foliage, the diagnostic or bearding the base tuft at of was the blade particularly conspicuous. Peel— 1993] Aristida 27 t MILES 200 300 400 ' ' ' ' Figure County oi Michaux and 2. distribution Aristida stricta (O) Aristida & beyrichiana Trinius Ruprecht Unverified reports are indicated by (•). (•). Other differences between the taxa can be observed, but they are generally minor and often For example, the average statistical. culm on ligule length is less for Aristida stricta, as is width;, and, more average, the glumes are unequal in A. beyrichiana. DISTRIBUTION examined specimens Michx. held labeled as Aristida stricta I all by the following herbaria: clems, duke, fla, flas, ga, gh, mo, l, VDB and Southcm Ncsc, Ncu, NY, UAL, US, uscH, Mississippi. p, County range maps were drawn using data on the 564 specimens examined that were assignable to county of origin (Figure 2). Nine which counties within the range o{Aristida beyrichiana for could not locate specimens to authenticate reliable reports are I map on by Sumter County, Georgia indicated the small dots: Rhodora 28 95 [Vol. Richmond and Camden, and (Jones Coile, 1988); Candler, Pierce Counties, Georgia (Bozeman, Harrison County, Missis- 1971); and and Barnwell and Charleston sippi (Eleuterius Jones, 1969); A from Counties, South Carolina (Radford aL, 1968), record et Oglethorpe County, Georgia, in Jones and Coile (1988) based is on an (ga No. 1331 Blomquist Incorrectly identified collection 02). Coun- Pasquotank (1948) indicated a record ofAristida stricta for my North Carolina. Despite examination of of the herbaria ty, all Blomquist reported consulting, have found no specimen to sup- I port that record. Because Pasquotank County located well north is known have of the range of the species, not indicated this record I on the map, although appropriate habitat does occur within the county. The and ranges oTAristida A. beyrichiana are not only stricta been but separated (Figure Neither species has distinct, clearly 2). The found South Carolina. one near intrusion of the in central range of ^. beyrichiana into the central part of South Carolina is known in Berkeley County, where the species only from a single is '^ from which appears have been Exami- locality to extirpated. it nations of likely habitats in central South Carolina, including the I sandhills around Columbia and the pine flatwoods of the Francis 4-- Marion National Forest nearer the coast failed to locate either I species. L' h Aristida stricta Michx. almost a North Carolina endemic, I is J V 7 occurring only in North Carolina and the northern most counties common South of Carolina. In this area the species particularly is on coarse sands dominated by Finns palustris, both in the fall- and on line sandhills the flatlands of the outer coastal plain (Frost and et 1986; Peet Allard, 1993). Examination of the distri- al., f bution of the southern wiregrass, Aristida beyrichiana, shows that in the northeastern portion of range largely concentrated its it is in those counties that contain the dune systems characteristic of the northeast sides of major rivers the Savannah and Al- (e.g., tamaha Rivers). In the western portion of range, the species its confined to the southernmost of counties along the Gulf of is tier Mexico. The make ranges of the two species that up the Aristida stricta, complex are not atypical of the ranges of other species of the s,l. many southeastern pinelands. Indeed, so species have similar ranges seems were two main of that likely that there centers it of one persistence the pineland during the flora last full glacial: V^^X-Aristida 29 1993] now near but perhaps of what North east southeastern is Carolina, and another along the Gulf coast. Species with ranges similar to that of ^. and endemic striata, essentially to the pinelands of s,s. North Carolina and adjacent South Carolina Dionaea include muscipula Kalmia Ell, cuneata Michx., Lysimachia asperulae- Lysimachia folia Poir., loomisii Torr., Pyxidanthera brevifolia Solidam M. Wells, pulchra Small, Solida20 verna A. Sno- Curtis. nov. teretifolius Harper; a dominant of (aff. glabra Vaccinium Tofieldia Nutt., crassifolium Dulverulenta (Bart, ex Willd.) Pollard. In la ad^ Calamovilfi Michx. have similar New Among disjunct populations in the pine barrens of Jersey. species that, like Aristida beyrichiana, have ranges that extend from Gulf the coast north to southernmost South Carolina are Asclepias cinerea Walt., Astragalus Michx., Baptisia lan- villosus ceolata (Walt.) Ell, Gaylussacia tomentosa (Gray) Small, Gay- Helianth Kalmia Pim hirsuta Walt., Liatris gracilis Pursh, gelm., Sabatia bartramii Wilbur, Sabatia brevifolia repens Sorghastrum secundum Small, (Bartr.) (Ell.) I cinium Lam. myrsinites TAXONOMY The two taxa included in Aristida are quite similar striata, sA. and are obviously closely related. Because they are readily and and consistently distinguishable morphologically are geographi- them cally separated as well, have chosen to treat as separate I species. Amer. Michaux's of Bor. original description Aristida striata [Fl. pu- 1:41, 1803]. reads "A. culmis foliisque stride erectis: foliis non racemo spicam longam, angustam, con- bescentibus: in ita gluma Obs. Variat fertam coarctato: longioribus. foliis aristis Hab. Carolina convoluto-filiformibus. in inferi- planiusculis et The pubescentibus" certainly sugges- ore." reference to "foliis is tive of the northern species with its conspicuous pubescence along the back sides of the leaf blades, but the description alone is Michaux which taxon determine col- definitively insufficient to and lected described. Although Michaux (1803) reported that his type collection of Rhodora 30 [Vol. 95 was from "Carolina Michaux^s anno- Aristida inferiore," stricta M. on "Hab. Rich- tation the label reads only in Carolina." L. C. have anonymously Michaux's ard widely believed written to is some of Michaux's death (Hitchcock, 1908; Stafleu text, after it many and Cowan, 1981; Uttal, 984) and of the habitat locations 1 The from on specimen in the text differ those the labels. basis for designating South Carolina rather than simply Carolina as the Michaux remains collection locality in (1803) unclear. Michaux numerous described in his journal (Sargent, 1889) through pine barrens between Charleston, South Carolina, trips and North and Wil- Charlotte, Carolina, along the coast past mington, North both which today dominated Carolina, areas are known by Michaux have the northern Aristida also to stricta, is (' where collected widely in the regions the southern A, beyrichiana I. dominates. In short, does not appear possible to determine it from where Michaux published records in the Carolinas collected his Aristida stricta. The Michaux Herbarium type of Aristida stricta in the at is Paris. The Michaux Herbarium contains two specimens labeled Aristida However, only one has the annotation "Hab. in stricta. As Michaux from Carolina." (1803) reported the type as being "Carolina specimen inferiore," this appears to be the type, a conclusion reached by Hitchcock and Henrard also (1908) 928). (1 I The specimen mixed The culm have * a collection. larger bases is had some their inflorescences broken off and appear to represent species other than A. or A, beyrichiana. These culm bases stricta mm have ligules 1.5 to 2.0 long, significantly longer than the .02 mm to .2 range normally encountered in Aristida stricta and A. beyrichiana, and longer than any observed in the over 650 I herbarium specimens examined. from In addition, they differ I both species in that they are almost entirely devoid of indument. However, two there are separate young, sterile culms (with ligules mm long) and several separate but older leaf blades that have .1 indument Mi- the distinctive villous pubescentibus," ("foliis chaux, 1803) along the sides of the midrib that characterize the They northern Aristida also lack the of hairs the stricta. tuft at throat and comers of the collar typical of the southern species. Because and have Aristida stricta A. beyrichiana similar flowers and cannot be which inflorescences, absolutely certain species I the fertile portion of the type collection represents. The combined collection was interpreted as the type by Hitchcock (1908) and FcGt-Aristida 1993] 31 Henrard (1928), but the pubescent foliage the only portion is unambiguously assignable to a single species. Thus, the pubescent young culms and pubescent older foliage on this specimen should probably be taken as the type of ^. striata. The other specimen Michaux Herbarium in the labeled Aristida striata is a flat-bladed species, almost ccrtsiinly Aristida purpuras- cens Poir. The original Michaux description ofAristida has striata the confusing line ''Variat foliis planiusculis et convoluto-filifor- mibus" which suggests that the description mixes the leaf blade of two by two Michaux characteristics the species represented the Herbarium specimens The labeled Aristida "convoluto- striata. filiformibus" foliage trait (unlike "planiusculis" foliage) char- is acteristic of a single species in the region, and "foliis pubescen- tibus" consistent with that species, which provides further reason is name leaves. There Michaux is a third collection labeled Aristida striata in the General Herbarium at Paris which has been referred to by Henrard an from Richard Herbarium. This (1928) as isotype the more Michaux collection complete than the collection in the is Herbarium, has conspicuous indument villous along the length My of the midvein, and unambiguously the northern form. is suspicion specimen was examined when Richard that this also is Michaux or wrote the original description of Aristida for striata Michaux's 1803 simply because only specimen with the text, it is the leaves conspicuously all hairy. new In 849, Trinius and Ruprecht described a species, Aristida 1 & (Mem. beyriahiana Trin. Rupr. Acad. Petersb. VI. Sci. Nat. St. which from 7(2): 104. 1849), they recognized as distinct Aristida striata. However, their description of A, beyriahiana has not been make adequate for others to a distinction; as a consequence, all synonym subsequent authors have viewed A. beyriahiana as a of A. Further, because the description only addresses the striata. inflorescence, not possible to assign the description to either it is monograph of of the taxa recognized in the present paper. In his North hmtncdi Hitchcock (1924) stated that he thought Aristida, immature the type of Aristida beyriahiana to be an A. striata, He had examined fragment although he admitted uncertainty. a had of the type in us (No. 8101 but apparently not seen the 1), monograph holotype Henrard (1926) reported in his of in le. awns "This with not developed, Aristida that a plant yet fully is I 1. Rhodora 32 95 [Vol. but agrees for the rest with A. striata Michx., having the striking villous tuft at the throat of the sheaths," a sure indication that he was examining the southern species. have examined the fragment us (No. 81011), and a Beyrich at I collection l (No. 908.83-1016) that Henrard (1926) referred to at an found them both with Henrard's de- as isotype, consistent I scription. The villous tufts are present on the us fragment, in- dicating that Aristida beyrichiana, although the tufts are small- it is er than average, and the blades are both a wider and not little However, quite as closely involute as normally encountered. is the L specimen unambiguously the southern species with villous is bearding at the leaf base extending onto the collar, and with typical leaf dimensions. The spikelets on both specimens are, as Hitch- cock and Henrard not mature. (1924) (1926) reported, fully The type locality for Aristida beyrichiana remains uncertain, I but probably Georgia, Trinius and Ruprecht reported that the is If-: Beyrich specimen they described originated "In pinetis Georgiae -t may Arkansas." However, simply et in territorio this description 4 r represent the general area which Beyrich he also in collected; is - { s known to have collected in both North and South Carolina (Sayre, 1975). Henrard (1926) reported that the label on the type reads simply "Georgia almost in pinetis (Beyrich)," but this observation from on which I* certainly derives the label the putative isotype at l^ k matches exactly. it modem match Virtually descriptions oi Aristida the all striata i'^ species call A, beyrichiana. For example, Hitchcock (1924, 935), I 1 t Henrard Godfrey and Wooten and (1932), (1979) Allred (1986) make -«*: all reference to the distinctive villous tuft at the base of the Henrard -ii leaf blades as characteristic of the species. Interestingly, (1932) reported the range oi Aristida striata as South Carolina to and There Florida Mississippi, exactly the range of^. beyriahiana. no specimens oi Leiden Herbarium are Aristida in the striata, s.s. where Henrard worked that are old enough to have been in the collection when he was active. Except for the type in Paris, there no evidence that Henrard ever examined specimens of the true is A. striata, Trinius and Ruprecht (1849) Aristida lanuginosa Clarion listed Mem, and (Trinius Ruprecht, Acad. Petersb. VI. Nat. St. Sci. synonym The 46. 1838) as a of ^. original 1838 4(2): striata. name reference contains the only, and a formal description first appeared Trinius and Ruprecht where was in (1849) reference ?QQi-Anstida 1993] 33 made specimen to a in the Mertens Herbarium with name (le) the A, lanuginosa Bosc. (thus the appropriate citation is^. lanuginosa & Bosc ex Trin. Rupr.). Hitchcock (1924) reported that there a Bosc specimen from is South Carolina Padua Herbarium in the labeled Aristida lanu- which ginosa, really^, lanosa Muhl. is ex EIL In 1924, Hitchcock declared A, lanuginosa to be a synonym of ^. lanosa, but there no is evidence that he ever examined the type specimen. Like Hitchcock, Henrard (1927) stated that Bosc's plant unambig- is uously the previously described ^m//V/^ lanosa, a conclusion that probably from derives observation of a Bosc collection in p labeled A, lanuginosa (probably an isotype), but which certainly A. is name Thus, lanosa. extremely unlikely that the A. lanuginosa it is refers to the southern species oi Aristida si, but did stricta, if it name would be the oldest other than A, a be applied to to it strict the southern taxon. However, there has been described an A. lanuginosa Burch. which synonym of (Burchell, 1824), a later is A. vestita Thunb. As the 1824 name validly published (see is & Henrard, 1927; 287), A. lanuginosa Bosc ex Trin. Rupr. a is homonym and later could not be applied to the A, beyrichiana material, even if the type proved to be this taxon. KEY AND CHARACTERIZATION SPECIES Base of blade, collar, and upper sheath lacking a conspicuous tuft or bearding of woolly indument; to villous current-year leaves with villous indument along the sides of the midrib on the lower surface for most of the length of the blade Aristida striata 1. Base of blade and collar (and often upper sheath) with conspicuous tuft or bearding of woolly to villous indument, always present on current-year but sometimes deciduous on older fo- foliage cm leaves usually glabrous above the basal 2 of the blade liage; beyrichiana Aristida 2, Amer Michaux, Bor. Aristida stricta Fl. 1 . (Michx, striata Graminum. Muhlenberg, Descriptio uberior Michaux Herbar- (Holotype: 174. 1817. Type: Carolina p, ium!; fragment us No. 81246, but insufficient material for General Herbarium!). determination: Isotype: p. 1 34 Rhodora 95 [Vol. culms 60-120 cm; Plants perennial; cespitose, erect, sheaths glabrous, occasionally with a few prominent hairs at the comers of the collars but typically absent with the collars glabrous; blades mm cm when closely involute, .3-1.0 thick rolled, to 50 long, mostly emerging near base, firm and somewhat flexuous, upper mm surfaces scabrous with some hairs up to 1.0 long, hairs on the upper surface hidden within the rolled leaf or with a few emerging along the lower 5 cm, villous on the lower surface along mm sides of midrib with hairs ,6-1.5 often over twice the (i.e., width of the along the length of the blade or along leaf) all at least the lower 20 cm, hairs sometimes deciduous with maturity but always on young minute present foliage; ligules to nearly absent, mm, mm; membrane composed to lacerate or of hairs to flat 1 1 . . cm and 20-35 panicles long slender, long, branches appressed, from glumes somewhat awns floriferous the base; unequal, 1.5- mm, awns emerging from glume 7-10 2.5 usually bifid tips; first mm, -nerved or with an additional nerve on one scaber- 1 side, mm, ulous on the keel but otherwise glabrous; second glume 9-12 on 1 -nerved, glabrous, if scabrous the keel then only near the mm, lemma summit; 6-9 shorter than the second glume, glabrous mm; except for the densely short-pilose callus, callus .4-,6 awns lemma somewhat of unequal, sharply divergent with matura- 3, mm mm, awn 10-15 awns tion, central (22) lateral usually 1-2 shorter. Range. Coastal plain of the Carolinas from the Pamlico River on the north to the northern tier of counties in South Carolina, & Mem. Aristida beyrichiana Trinius Ruprecht, Acad. Pe- 2. St, tersb. VI. Nat. 104. 1849. Type: Georgia? ("In Sci. 7(2): pinetas Georgiae et in territorio Arkansas.'') (Holotype: le, Trinius Herbarium, not seen, Henrard, 1926; fragment fide No. 8101 No. us! Isotype: 908.83-1016, see Henrard, l!. 1; 1926). culms cm; sheaths glabrous except the summit, copiously somewhat on at to villous the upper cm, especially about throat and always with .5 collar, prominent comers protruding hairs the of the blades at collar; mm cm when 20-40 closely involute, .3-1.0 thick rolled, to long, mostly emerging near and somewhat upper base, firm flexuous, many surfaces scabrous with few to hairs protruding near base