Dalea purpurea Vent.

Authors

Rupert C. Barneby

Authority

Barneby, Rupert C. 1977. Daleae Imagines, an illustrated revision of Errazurizia Philippi, Psorothamnus Rydberg, Marine Liebmann, and Dalea Lucanus emen. Barneby, including all species of Leguminosae tribe Amorpheae Borissova ever referred to Dalea. Mem. New York Bot. Gard. 27: 1-892.

Family

Fabaceae

Scientific Name

Dalea purpurea Vent.

Description

Species Description - Perennial herbs from tough brown or blackish, rarely yellow roots, sometimes developing a knotty subterranean caudex, 2-9 dm tall, either glabrous to the spikes or (more commonly) thinly pilosulous, especially on petiolule and lower face of leaflets, or densely softly pilosulous nearly throughout, the weak spiral, spreading or ascending hairs up to 0.3-0.8 (1) mm long, the striate-ribbed, glandless or remotely minutely microglandular, pale green or stramineous stems most often erect and virgately ascending, few-branched distally into a subcorymbose panicle of 2-several heads, but not seldom (especially n.- and n.-w.-ward) simple and monocephalous, when either erect or diffuse, or (especially s.-e.-ward) diffuse and branched, nearly always densely leafy, the primary cauline leaves subtending short-shoots or fascicles of smaller leaves, the foliage vivid green, greenish-glaucescent, or ashen-pilosulous, the leaflets smooth above, normally punctate beneath; leaf-spurs 0.4-1.6 mm long; stipules subulate to narrowly lance-acuminate or -caudate, (1.5) 2-6 (7) mm long, early brown or stramineous, in age dry and fragile, either glabrous or ciliolate; intra- petiolular glands 0; post-petiolular glands mostly minute and immersed, sometimes 0; leaves petioled, the primary cauline ones 1.7-4 (4.5) cm long, with narrowly thick- margined, remotely punctate rachis and nearly always 5, exceptionally 3 or 7 linear- oblanceolate, linear, or linear-elliptic, abruptly short-acuminate or acute, tightly involute or if explanate then marginally elevated, dorsally carinate leaflets up to (7) 10-24 (28) mm long, the terminal leaflet either subsessile or very short-stalked, usually a trifle longer than the rest, the leaves of spurs and some upper leaves mostly 3-foliolate, their leaflets of the same type but smaller; peduncles almost 0 up to 9, rarely 15 cm long, the first spike of each principal stem-axis usually pedunculate but the lateral ones subsessile, when stems monocephalous the spike either pedunculate or sessile; spikes very dense, conelike, ovoid becoming oblong-cylindroid, without petals 7-12 (13) mm diam, the densely villosulous axis becoming (1) 1.5-7 cm long; bracts disjointing only with the fruiting calyx, subdimorphic, the lowest firm, ovate- to lance-acuminate or -caudate, 2.5-7 mm long, the recurving tail 1.5-4 mm long, the interfloral ones (2.3) 2.7-5.8 mm long, 1-1.8 (2.2) mm wide, the papery, stramineous or castaneous-flecked, membranous-margined, glabrous or subglabrous body oblanceolate or spatulate in outline, sometimes charged dorsally with 1-2 small livid glands, at apex densely pilosulous on the back and ciliolate, abruptly contracted into a subulate, erect or recurved, usually livid, either glabrous, puberulent, or rarely pilosulous tail (0.5) 0.7-3 mm long; calyx 3.2-4.5 (5) mm long, densely pilosulous throughout with ascending (when dry golden) hairs up to 0.2-0.7 mm long, the tube (1.7) 2-2.8 (2.9) mm long, not recessed behind banner, bluntly pentagonal and not prominently ribbed, the ribs either castaneous or concolorous with the submembra- nous, heavily castaneous-flecked, glandless intervals, the teeth of nearly equal length but different widths, the 3 dorsal ones lanceolate to narrowly ovate, 1.1-1.9 (2.1) mm long, 0.1-1.4 mm shorter than tube, the ventral pair ovate, up to twice as broad, the blades of all flat, green or livid-castaneous, rarely glabrescent externally, connivent in age; petals concolorous, vivid rose- or magenta-purple, pale purple, rarely lilac, pinkish, or exceptionally white, glandless; banner (4.3) 4.7-6.7 (7.2) mm long, the filiform claw (2.6) 2.8-4.2 (4.5) mm, the scoop-shaped, apically hooded and obtuse to emarginate blade broadly cuneate to truncate or subcordate at base, 1.7-2.6 (2.8) mm long, 1.7-2.8 (3.2) mm wide; epistemonous petals all alike, (3) 3.2-4.5 (5) mm long, the claw 0.5-1.2 (1.6) mm, the oblong or oblong-oblanceolate, obtuse or truncate, apically concave blades 2.5-3.8 (4) mm long, 0.8-1.3 (1.5) mm wide, either cuneate or truncate at base; androecium 5.6-8.5 (9) mm long, the column (2.6) 3-4.3 mm, the purplish free filaments up to 3-5.2 (5.5) mm long, the orange- yellow anthers (0.7) 0.8-1.3 (1.4) mm long; pod obliquely obovoid, 2.1-2.6 mm long, the valves hyaline glabrous in lower 1/2 or 1/3 thence thinly papery, pilosulous, gland-dotted; seed brown or olivaceous, smooth, sublustrous, 1.6-2.1 mm long.

Discussion

(Plate LVIII)

The purple prairie-clover, next to D. Candida the most widespread of its kind and the prototype of the genus Petalostemon, is a familiar wildflower of the prairie states where, in a variety of open habitats its graceful wandlike stems, clad in fine, few-fingered foliage, carry aloft in summer a few hard knobs of silvery calyces ringed with a tassel of purple petals and gold or orange anthers. Its immense range coincides, east of Continental Divide, very nearly with that of D. Candida. In the rich lowlands of the Mississippi valley the purple and white clovers are often found together, in this region both tall virgate herbs, their stems branching distally into a narrow thyrse or wider, corymbosely flat-topped panicle of heads. Passing west into the colder and drier climate of the short-grass prairie, each is represented by forms of low stature and simpler branching. But where D. Candida has become modified further in number of leaflets and density of spike, D. purpurea remains essentially unchanged in leaf and flower, presenting however a great variety of minor variants differing among themselves in growth-habit, density of vesture, length of peduncle, and diameter of the spike. Already in eastern United States D. purpurea has been known from early times as variable in pubescence; plants with glabrous and thinly pilosulous foliage often grow together and both are represented in Michaux’s original material from Illinois. On the far western plains and Rocky Mountain piedmont between Alberta and Colorado the variation in pubescence becomes greatly exaggerated, culminating in a plant densely and softly villous-tomentose throughout, almost suggesting D. villosa except for the few leaflets and conelike spike. Plants of this type were described from Montana as Petalostemon mollis Rydb., others from Colorado, differing further in obsolescence of the usually conspicuous dots on back of the leaflets, as P. pubescens A. Nelson. As I have seen the latter in Colorado (e. of Trinidad, Barneby 14,997, CAS, GH, IA, NY, US) it was confined to a peculiar habitat on stiff gumbo clay flats and low bluffs, and with its diffuse stems and relatively broad leaflets and petals appeared sharply marked. But analysed, the morphological features of these pubescent types do not reveal any discontinuity separating them from sympatric glabrescent phases of D. purpurea, and I agree with Wemple (1970, p. 90) that they are best treated taxonomically as minor variants.

Over the greater part of its range D. purpurea has calyces uniformly pilosulous from base to points of the teeth, usually so densely so as to conceal the greenish-glaucous or castaneous-flecked epidermis. In the Ozarks, and sporadically north into central Kansas and (according to Wemple) even Wisconsin, the calyx-vesture becomes concentrated in lines running lengthwise along the angles of the tube leading to the sinuses and on margins of the teeth, while the intercostal intervals of the tube and the external face of the teeth may become only thinly puberulent, in consequence glaucous. A calyx of this sort is strongly reminiscent of that of the closely related D. compacta except that it is always smaller. I follow Steyermark (Fl. Missouri 901) and Wemple (1970, p. 91) in my interpretation. A form with small leaves and small heads, somewhat isolated from the main range of the species in middle Louisiana and Alabama, likewise noted by Wemple (1970), and ecotypes from dune habitats as far apart as the shore of Lake Michigan in Indiana (= fma. arenarium Gates), the upper Red River in Manitoba, and the Arkansas valley in central Kansas, seem to reflect the selective influence of habitat on a somewhat plastic genotype.

In the Texas Panhandle and adjoining New Mexico, w. Oklahoma, and thence north through far western Kansas and (well out from the Rocky Mountain piedmont) eastern Colorado into Nebraska, D. purpurea is represented by a series of forms, found mostly on dunes or sandy habitats connected with river bluffs and streambeds below the level of the sedimentary caprock, notable for their small narrow flower-heads, without petals about 7-9 mm diameter, raised above the foliage on a distinct peduncle. The flower is not different from that of D. purpurea elsewhere, and the foliage appears identical. Wemple has segregated this material as Petalostemon arenicola, emphasizing not only the narrow pedunculate spike which is, I agree, the principal differential character, but also the shorter, usually unbranched (therefore monocephalous) stems glabrous to the base of the spike, more conspicuously ciliolate bracts, and a more oblique, proportionately narrower ovary. These supporting characters attributed to P. arenicola are somewhat weakened by the phototype accompanying the protologue (Wemple, 1970, fig. 8) which shows a plant freely branching both at base and distally (all main stems several-headed), and peduncles little longer than found in some unquestioned D. purpurea from the northern prairies. Interfloral bracts in the narrow spike tend to be a trifle broader proportionately to their length than elsewhere in D. purpurea but I have been unable to confirm any substantial difference in pattern of their pubescence, a feature in any case so variable in D. purpurea and its relatives that it can command little respect in any particular case. In aspect, however, P. arenicola is indeed rather distinctive and recognizable, at first glance suggesting often some form of the allopatric, calciphile D. tenuis (or P. stanfieldii), for which it has been mistaken. It is obviously different from D. tenuis in the antrorse pubescence of the calyx-tube, and even more so from the partly sympatric but there ecologically separated D. tenuifolia, which has a loose spike, broader interfloral bracts, and a softly long- pilose calyx dissimilar in form. The range of P. arenicola occupies a not quite perfect enclave within that of several minor geographic variants of D. purpurea which surround it on all sides except the south. The epicenter for sharply focussed P. arenicola lies on the Staked Plains between the upper Red River in Texas and the Arkansas valley in eastern Colorado. Northward from the Arkansas the populations become less distinctive, in Nebraska fading out into the short-stemmed, high prairie form of D. purpurea mentioned earlier. The segregate is treated below as a fairly well defined geographic variety.