Astragalus Lindheimeri

Annuals, resembling forms of A. Nuttallianus but often coarser, with a slender taproot, thinly strigulose with straight, appressed or subappressed hairs up to 0.35-0.5 mm. long, the herbage green, the leaflets glabrous above; stems 1-20 from the root-crown but commonly less than 10, erect when solitary but mostly incurved-ascending or nearly prostrate, (0.2) 0.5-3 (3.5) dm. long, green or purplish, often striate, simple or spurred at 1-3 nodes preceding the first peduncle, floriferous upward from near or below the middle; stipules herbaceous or membranous-margined, becoming papery, triangular or triangular-acuminate, 1.5-5 mm. long, about semiamplexicaul; leaves (1) 2—6.5 cm. long, petioled but the uppermost quite shortly so, with (11) 13-21 broadly to narrowly cuneate, cuneate-oblong, or -oblanceolate, truncate-emarginate to deeply retuse, flat or loosely folded leaflets of rather thick texture and 2-14 mm. long; peduncles ascending or incurved, (1) 2-6.5 cm. long; racemes loosely but very shortly (1) 2-6 (8)- flowered, the flowers spreading, the axis scarcely elongating, 0.2-1.5 (2.4) cm. long in fruit; bracts submembranous becoming papery and pallid, ovate-acuminate, lanceolate, or linear-caudate, 1.3-5 mm. long; pedicels at anthesis slender, ascending, (1) 1.3-2.3 mm. long, in fruit thickened, arched outward, 1.8-3 mm. long, persistent; bracteoles usually present as minute scales at or below the base of the calyx; calyx (4) 5-8 mm. long, thinly strigulose with white hairs but the teeth glabrous or nearly so externally, the subsymmetric disc 0.5-0.9 (1) mm. deep, the campanulate, often purplish tube (2) 2.4-3.3 mm. long, 2.4-3.6 mm. in diameter, the erect or spreading, narrowly lance-subulate teeth (2) 2.3-5 mm. long, the whole becoming papery, marcescent unruptured; petals bicolored, the banner purple-margined around the prominent white but purple-veined eye, the wings white, the keel-tip purple-maculate, the purple turning violet on drying; banner recurved through about 35°, broadly ovate or ovate-oblong beyond the shortly cuneate claw, sometimes subflabellate, shallowly notched, (12) 13-18.5 mm. long, 7.2-15.5 mm. wide; wings 11.2-17 mm. long, the claws 2.6-4 mm., the broadly oblanceolate or obliquely obovate, nearly straight blades 9.8-14.2 mm. long, (3) 3.5-5.8 mm. wide; keel 9.5-12.9 mm. long, the claws 3.4-4.6 mm., the broadly triangular blades (7) 7.5-9.5 mm. long, 3.8-5 mm. wide, abruptly incurved through 90-95° to the obtusely deltoid apex; anthers 0.4-0.65 mm. long; style minutely barbellate all around just below the stigma, rarely puberulent its whole length; pod stipitate, the stout stipe (0.8) 1-2.6 (3) mm. long, ascending at a wide angle, horizontal, or a little declined in line with the pedicel, the linear-oblong or -oblanceolate body gently and lunately or more abruptly and subbasally incurved through about a right angle (the distal half then narrowly ascending and becoming perpendicular in a plane parallel to the raceme-axis), 1.7-2.7 cm. long, 3.5-6 (6.5) mm. in diameter, broadly cuneate or acuminately tapering at base, cuspidate at apex, compressed-triquetrous with flat or low-convex lateral faces and narrower, narrowly sulcate dorsal face, carinate ventrally by the prominent, thick suture, the thinly fleshy, green, glabrous valves becoming thinly leathery or stiffly papery, brownish or when ripe nearly black, inflexed as a complete septum 2.2-3.3 mm. wide; dehiscence apical and tardily through the ventral suture; ovules 8-12 (14); seeds pale greenish- or yellowish-brown, sometimes purple-speckled, smooth, 3.2-4 mm. long.—Collections: 63 (iii) representative: Shinners 9964 (OKLA, SMU); Reverchon distrib. Curtiss 601* (NY, SMU); McBryde 1006 (NY, TEX); D. McLean 167 (CAS, TEX); Ripley & Barneby 7489 (CAS, GH, RSA).

Low hills, barrens, prairies, and gullied bluffs, sometimes abundant in disturbed soil along highways, in dry calcareous gravels, sands, red clay, or gypsum, mostly below 2100 feet, locally plentiful in northcentral Texas and the valley of the Red River in southwestern Oklahoma, in Texas east to Dallas and Navarro Counties and west to Irion and Howard Counties, also apparently isolated on the Balcones Escarpment between San Antonio and Austin; isolated records from the north end of the Texas Panhandle (Ochiltree County) and northcentral Oklahoma (near Stillwater, Payne County) are probably adventitious; a record from Monterrey, Mexico (Jones, 1923, p. 268) has not been traced to the source.— Map No. 152.—Late March to early June.

Astragalus Lindheimeri (Ferdinand Lindheimer, 1801-1879) Engelm. ex Gray, Pl. Wright. 1: 52. 1852.—"Sand bar in the Colorado near Austin, Texas; April, (on rich Muskit soil near water, April, 1850, Lindheimer)."—Lectotypus, Lindheimer 746, collected at "Santa Clara 10 miles s. of New Braunfels, April, 1850," MO! isotypi, mostly labeled "Comanche Spring near New Braunfels" GH (2 sheets, not annotated by Gray + 1 sheet, renumbered 258 and named in Gray’s hand), K, MO, NY, PH, SMU, TEX, US! The paratypus at GH with penciled label "found on sand-bar of the Colorado near Austin, Wright"=A. Nuttallianus var. macilentus.—Tragacantha Lindheimeri (Engelm.) O. Kze., Rev. Gen. 946. 1891. Hamosa Lindheimeri (Engelm.) Rydb. ap. Small, Fl. S. E. U. S. 1332. 1903.

Astragalus Lindheimeri var. bellus (pretty) Shinners in Field & Lab. 25: 33. 1957.—"3.2 miles west of Archer City, Archer Co., Shinners 18,568, 24 April, 1954 ... "—Holotypus, SMU!

The Lindheimer milk-vetch is the showiest by far of the annual astragali of Texas, easily distinguished (almost always) from the other Leptocarpi by flower-size alone. There is a small overlap in length of banner between A. Lindheimeri and the largest-flowered populations of A. Nuttallianus var. macilentus and A. leptocarpus, although the keel is always shorter in both species. At anthesis A. leptocarpus differs further in its triangular keel-tip and many (at least 17) ovules, and var. macilentus by its more numerous flowers; moreover the style in both is glabrous and the fruit (as in all forms of A. Nuttallianus) is of narrower outline and sessile or nearly so. The only other American Astragalus beside A. Lindheimeri in which a puberulent style has been observed is A. brazoensis, by coincidence also a Texan annual, but readily recognized by its many and much smaller flowers and its almost circular, dorsiventrally compressed pod deciduous from a stipelike gynophore.

The flower of A. Lindheimeri is uncommonly pretty, because of the contrasts between the large, white eye and the band of lively purple (variable in width) around the margin of the banner, and the white-spotted or white-tipped wings against the broad, purple keel. The range of the species is apparently bicentric, with a large area in northcentral Texas and southwestern Oklahoma and a smaller disjunct one along the Balcones Escarpment between the Colorado River and San Antonio. According to Shinners the flower in the southern area is smaller and paler (only 11-15 mm. as opposed to 13-18 mm. long) and is combined with shorter peduncles and leafier stems; he has accordingly segregated the northern populations as var. bellus, the typus of A. Lindheimeri having originated in the southern area. I have not had opportunities of re-examining a large series of specimens since the publication of var. bellus, and have not seen fresh material from near Austin, so I feel at a disadvantage in trying to evaluate the supposed differences. Certainly these are far from absolute (and are not even claimed to be so), witness Reverchon 601 (NY) from Brown County, Texas, with banner 12-13 mm. long, as compared with Thurber’s collection from San Antonio (in 1853, NY) in which the banner is 12-14 mm. long and even after a century still brightly colored distally. Possibly Shinners had partly in mind as var. Lindheimeri the large-flowered phase of var. macilentus, in which the keel is prominent and the general proportions of the pale flower are very suggestive of A. Lindheimeri. Plants of this sort differ, however, in their looser, many-flowered racemes, narrow, subsessile pods, many ovules, and glabrous styles. Considering the relatively large size of the pod, the paucity of ovules is a remarkable feature of the Lindheimer milk-vetch, but the large seeds are of a size to fit the fruit. An albino form has been found rarely in the Red River Valley.

Typification of A. Lindheimeri has proved unexpectedly difficult. The material (GH) annotated in Gray’s hand is a mixture of the genuine A. Lindheimeri as known to Engelmann, acknowledged author of the name, and the large-flowered phase of A. Nuttallianus var. macilentus, represented by Wright’s plant from the Colorado River at Austin. There is a specimen of true A. Lindheimeri collected by Wright at GH and NY, but both lack Gray’s autograph and are apparently irrelevant to our problem. The sheets of Lindheimer 746 at GH were probably distributed after publication of Plantae Wrightianae, for they also lack annotations. Gray’s material from Lindheimer consists of two sheets numbered 258 (1850) and 542 (April, 1851). I take them to be duplicates of what Engelmann later sent out under the serial Nos. 746 and 747, respectively. The best choice of holotypus seems to be the sheet in the Engelmann herbarium that bears Lindheimer’s own note about the locality, slightly different, it should be noted, from that given on the printed label distributed afterwards. Blankinship (in Rep. Mo. Bot. Gard. 18: 165. 1907) reached a similar conclusion.