146b.  Dalea sericea Lagasca var. humistrata Barneby

(Plate CXXXII)

Stems usually numerous, humifuse, branching near base and distally, forming mats of silvery-gray foliage, each main axis usually several-headed; peduncles mostly shorter than the leafy axis, incurved under the subvertical spike; flower relatively small, and calyx-short-toothed (cf. key); androecium ±7-8 mm long.— Collections: 8 (iv).

Sunny openings and barren gravelly ridges in oak-pine forest, ± 1930-2130 m (6425-7100 ft), locally abundant, mountains of central and south-central Oaxaca, extending w. along Sierra Madre del Sur into centr. Guerrero. — Flowering September to December, probably later.—Representative: Guerrero. Chilpancingo: Cerro Alquitran near Mazatlan, Rzedowski23,641 (ENCB). Oaxaca. Etla: Las Sedas, Pringle 6020 (BR, ENCB, F, M, NY, UC, W); Sierra San Felipe, n. slope, Ripley & Barneby 14,610 (NY); n. of Telixtlahuaca, Ripley & Barneby 14,669 (NY). Asuncion Nochixtlan: cf. typus. Miahuatlan: s. of Miahuatlan, Ripley & Barneby 14,638 (NY).

Dalea sericea Lag. var. humistrata Barneby, var. nov., a var. sericea caulibus demum numerosis basi et supra medium ramosis omnibus prostratis flexuosim humi appressis, spicis angustioribus (petalis delapsis 1-1.3 nec 1.3-2 cm diam), calycisque dentibus brevi- oribus, dorsali tubo subaequilongo vel breviori absimilis. Habitat in querceto-pinetis supra graminum zonam cui var. sericea, si adsit, saepissime adhaeret. — Parosela reclinata sensu Rydb., N. Amer. Fl. 24: 101. 1920, quoad descr., exclus. basionym. Psoralea reclinata Cav.— Oaxaca: thin oak-pine forest, forming mats of foliage up to 1 m diam on granitic or andesitic bedrock, ± 2010 m (6700 ft), in the mountains of Mixteca Alta ± 27 km s.-e. of Nochixtlan, nov 10, 1966, Ripley & Barneby 14,588. — Holotypus, NY; isotypi, CAS, DAO, MEXU, MICH, UC, US.

The entity described herein as var. humistrata was already recognized in Rydberg’s revision of Parosela as P. reclinata, a name properly belonging to a related annual dalea (see ser. Reclinatae immediately preceding); it is new here only in name and taxonomic status. Until I had seen it growing I interpreted Rydberg’s scanty material as another phase, comparable with but more extreme than the so-called D. gracilis, of an inclusive D. sericea, for the supposed differences in flower-color and width of scarious margin to the bracts seemed, as they still do, inconsequential. Rydberg did point out that the stems of his P. reclinata branched toward the base, unlike those of P. sericea sens. str., but it was only in the field that I learned they are also pliantly prostrate, thus forming a mat of silvery foliage, surrounded by a ring of spikes incurved to vertical, unlike any known phase of genuine D. sericea whether robust or slender. In Oaxaca var. sericea and var. humistrata are close neighbors, but separated by life-zone. Ascending the Sierra Madre south of Miahuatlan I encountered on grassy hillsides in the lower fringes of the oak-belt characteristic, coarsely erect var. sericea with one or few massive spikes of long-toothed calyces, while only a few tens of meters higher, on cooler slopes in pine-forest, appeared var. humistrata, typical and abundant. Search in this region for intermediates was unsuccessful, and I formed the impression, reenforced by observation of both elsewhere in the valley of Oaxaca, of two truly independent populations.