Herbs, perennial, not viviparous, 7-10 dm, glabrous. Stems above ground (caudex) or underground (corm), usually erect, simple or 2-branched at apex, rarely with lateral vegetative branches, often fleshy; floral stems annual, from rosette leaf axil, simple or branched, overtopping rosette, with scattered smaller leaves. Leaves persistent or withering in early summer, drying persistent and often covering caudex for years, densely crowded in basal or terminal rosette, alternate, subclasping basally, petiolate or sessile; blade linear to orbiculate, nearly laminar to terete or subglobose, 0.2-30 cm, fleshy, base not spurred, margins entire, not ciliate; veins entering margins. Inflorescences lateral cymes, 2+-cincinnate. Pedicels present. Flowers (usually odorless), erect to pendent, 5-merous; sepals connate basally, all alike, shorter than petals; petals erect and forming cylindric to 5-gonal tube or spreading from near middle, connate basally or to middle, white, yellow, orange, or red; nectary disc truncate, wider than tall; stamens 10; filaments adnate to corolla base; pistils erect or ascending to spreading, nearly distinct; ovary base rounded; styles shorter than ovary. Fruits erect to spreading. Seeds narrowly ovoid, longitudinally ribbed, finely cross-ribbed. x = 17. Dudleya is distinguished by its erect sepals usually connivent to the corolla and its corolla convolute in bud, rarely imbricate. The epipetalous stamens are mostly shorter and adnate higher than the episepalous ones and are not reflexed. The center of distribution for all three subgenera of Dudleya is near the coast north and south of San Diego, California. Because this area is now largely urban, many of the endemics are more or less threatened: M. W. Skinner and B. M. Pavlik (1994) listed 27 species and subspecies of Dudleya as to some degree rare and endangered. However, most species, including narrow endemics, are abundant where they do grow.
The herbage of some species, such as Dudleya edulis, is reported to have been eaten by Native Americans. Fanciers grow some species, and although the plants have a reputation for being hard to grow, some are easy enough (P. H. Thomson 1993). Some find a place in Californian rock gardens; none is in general cultivation.
C. H. Uhl has studied nearly 400 collections of Dudleya, including all known species (C. H. Uhl and R. V. Moran 1953). All have a basic chromosome number of 17, and about 35% of populations are polyploid, some up to 16-ploid (Uhl 1994). D. Verity (pers. comm.) was able to cross Dudleya species in every combination he tried, regardless of morphology and level of ploidy, including such improbable hybrids as D. blochmaniae × pulverulenta, crossing two of the most unlike. Also, natural hybrids are known between most diploids that grow together; the hybrids are mostly rare (Moran 1951; Moran and Uhl 1952; P. H. Thomson 1993). All diploid hybrids that Uhl studied showed very good chromosome pairing, with no detectable abnormalities at meiosis.
In subg. Dudleya and Stylophyllum an occasional rosette may reroot after breaking off or after the common stem dies and decays; in contrast to the many Crassulaceae that readily multiply by leaves and plantlets, these plants come almost solely from seeds. In subg. Hasseanthus, the rosette leaves do sometimes root and form new plants (M. Dodero 1996).
The most primitive species would seem to be those large and much-branched plants of subg. Stylophyllum with broad, flat leaves, much-branched cymes, and open Sedum-like flowers, such as Dudleya virens subsp. insularis. At the other extreme is D. brevifolia, still with open flowers but with underground tuberous stems, tiny vernal rosettes of subglobose leaves on threadlike petioles, and relatively few flowers. The most-advanced inflorescence and flowers are in hummingbird-pollinated D. pulverulenta, of subg. Dudleya.
The first-named species of Dudleya were placed in Cotyledon, Echeveria, or Sedum; A. Berger (1930), P. A. Munz (1935), and W. L. Jepson (1909-1943, vol. 2) treated Dudleya species under Echeveria. The two are quite distinct and evidently have a long, separate history. Dudleya is centered on the Pacific tectonic plate, with only a few species extending onto the western edge of the North American plate; Echeveria and relatives are on the North American plate and southward (C. H. Uhl 1994).
Although all species of Dudleya readily hybridize, and also species of the Echeveria group, all attempts to cross Dudleya with Echeveria and relatives have failed (C. H. Uhl 1994). Furthermore, Uhl concluded that chromosome changes in Dudleya have been mainly through gene mutations at the molecular level, with few or no changes in gene sequences; in Echeveria and relatives, he found a wide range of diploid chromosome numbers, with evidence of rearrangements of gene sequences (Uhl 1992). These different mechanisms of chromosome change suggested that the two groups might not be so closely related as usually thought (Uhl 1993). From the different structures of their testas, U. Knapp (1994) agreed; H. ´t Hart (1995) thought that the two probably came from different mostly Old-World lineages: Dudleya from his paraphyletic Leucosedum-clade, with such genera as Pistorinia and Rosularia and with Sedum album, and allies; and Echeveria from his paraphyletic Acre-clade, with the other Mexican genera Graptopetalum, Pachyphytum, Sedum in part, and Villadia, and with Sedum acre, and allies, in the Old World.
My descriptions are based largely on live plants. The five species of subg. Hasseanthus and the six of subg. Stylophyllum are fairly well defined and distinct; subg. Dudleya, despite some help from chromosome numerology, is still notoriously difficult. Although such diploids as Dudleya stolonifera are distinct endemics, D. abramsii and D. cymosa are wide-ranging and polymorphic diploids, whose diverse populations do not sort out easily into geographic subspecies. Similarly, such polyploids as D. nesiotica and D. traskiae are distinct endemics; others form difficult complexes, thus far defying classification. Hence the treatment of some species, and the keys to them, remain quite unsatisfactory.
PLANT: Glabrous perennials. LEAVES: alternate, subclasping, entire, in dense basal rosettes. FLORAL STEMS: lateral, overtopping rosettes, annual, with scattered smaller leaves; cymes of 2-several one-sided branches. FLOWERS: 5-merous, in ours odorless; sepals equal, nearly separate, erect; petals convolute in bud, united below, in ours erect to form tube; stamens 10, on corolla base, included; pistils not narrowed at base, in ours erect. FOLLICLES: many-seeded. SEEDS: narrowly ovoid, brown, striate, to 1 mm long. X = 17. NOTES: Ca. 50 spp., sw OR and s NV to s Baja C. and nw Son., Mex. (for W. R. Dudley). REFERENCES: Moran, Reid. 1994. Bixaceae. J. Ariz. - Nev. Acad. Sci. Volume 27, 190-194.