Dalea bicolor

Title

Dalea bicolor

Authors

Rupert C. Barneby

Scientific Name

Dalea bicolor Humb. & Bonpl. ex Willd.

Description

108.  Dalea bicolor Humboldt & Bonpland ex Willdenow

(Plates XCIX-CII)

Fruticose when adult, but precociously flowering and highly variable in mature stature, usually bushy, rarely diffuse and trailing, occasionally becoming tall shrubs or treelets, (2) 4-20, rarely 35 dm tall, the old wood becoming fissured and striped lengthwise, the young branches prominently and usually densely tuberculate with mammiform or grainlike glands, pilosulous with ± curly, subappressed, or incurved, more rarely spreading hairs, the foliage green, greenish, or ashen, always at least thinly and inconspicuously pilosulous, often densely so, sometimes silky or satiny on one or both sides, or densely tomentulose, each main stem branching distally into panicle of spikes terminal to leafy branchlets usually surrounding and often surpassing the leaf-opposed, central spike; leaf-spurs 0.3-1.5 mm long; stipules narrowly triangular to subulate or linear-setiform, livid, either pubescent or glabrous, (0.5) 1-4.5 mm long; intra- and post-petiolular glands always present, often conspicuous; leaves shortly petioled, the primary cauline ones (0.6) 1-3.5 (4.5) cm long, with narrowly margined rachis and (1) 2-10 (11) pairs of obovate, oblong-elliptic, oblanceolate, obovate- or cuneate-obcordate, obtuse, retuse, or bluntly gland-mucronate, dorsally punctate leaflets (1.5) 2-9.5 mm long, the last pair arising at same point to form a palmate trefoil; peduncles 0-4 (5) cm long; spikes usually loose, with calyces sub- contiguous or well separated, but sometimes closely contiguous, in Baja California sometimes extremely dense and almost conelike, without petals or androecia (4.5) 5.5-15 mm diam, the villosulous axis becoming (0.5) 1-20 ( 23) cm long; bracts rather tardily deciduous or, especially when small and low on the spike (or both) persistent through and even after anthesis, in dense spikes sometimes deciduous but held fast between the flowers, rhombic-ovate to broadly ovate, ovate-acuminate, lanceolate-caudate, or sometimes narrowed downward from near or above middle and then oblanceolate, 1-5.5 (6) mm long, dorsally puberulent to densely pilose, livid and glandular beneath the vesture, the tips often glabrate, usually appressed- pilosulous or at least puberulent internally, in var. orcuttiana sometimes internally glabrous; calyx 2.5-5.6 (6.8) mm long, variably pubescent, the tube often yellow or castaneous, 1.6-3.2 mm long, the ribs thickened and usually prominent, the intervals charged with 1 row of ± 2-3 (4) glands, the deltate-ovate to triangular-apiculate or -acuminulate, often gland-spurred teeth unequal, the dorsal one nearly always a trifle, often much longer than the rest, 0.5-2.9 (3.8) mm long, with rare exceptions shorter than the tube; petals bicolored, the banner opening white or creamy yellow, early rubescent, the epistemonous ones rose-purple, phlox-purple, lilac, violet, or vivid cobalt blue, perched well below middle of androecium, usually all glandless but the banner or keel (or both) rarely tipped with a minute gland; banner 3.2-7.5 (8) mm long, the claw 1.4-3.4 (4) mm, the deltate-cordate blade (2.2) 2.8-4.8 (5.8) mm long, 2.4-5.2 mm wide; wings 3.4-7.5 mm long, the claw 0.8-2.7 mm, the broadly lance-oblong to ovate-elliptic blade 3.6-5.5 mm long, 1.8-3 mm wide; keel 5-10 mm long, the claw 1.2-3.6 (4) mm, the ovate-elliptic blades 3.6-7 mm long, 2.1-4.4 mm wide; androecium 10-merous, 5.6-10.5 (12) mm long, the filaments free for (1.7) 2-4 mm, the connective gland-tipped, the anthers 0.65-1 mm long; pod 22-2.1 mm long, variable in outline, in profile obovate to triangular, the ventral side as long or longer than the others, the slender prow convex or straight, the style-base excentrical- ly terminal or latero-terminal, the valves thinly papery above the hyaline base, sparsely gland-dotted, distally pilosulous or rarely glabrous.

As defined by the description, D. bicolor is an assemblage of many small races adapted to an immense altitudinal and climatic range. By accretion around a nucleus of what has been known historically as D. tuberculata Lag. it has come to embrace so much diversity in form and detail that a comprehensive portrait becomes an abstract formula, portraying the parameters of total variation (or attempting to do so), but no particular plant.

Characters common to the components of D. bicolor sens. lat. are: prominently warty stems; foliage at least puberulent, often densely gray or even silky pubescent; leaflets of primary leaves mostly 3-9, rarely 2 or up to 11 pairs, neither very small nor greatly thickened; a relatively short-toothed calyx subtended by bracts tardily deciduous or persistent; and epistemonous petals of some shade of pink, purple, or blue, never yellow, contrasting in early anthesis with a white or pale yellow, soon rubescent banner. The distinguishing characters of several close relatives in central Mexico and southern Baja California cannot be usefully emphasized in context of the whole specific complex but are mentioned under the appropriate heading.

Several forms of D. bicolor have been noted as attractive to bees, which are sometimes seen in great numbers working the flowers for pollen and nectar. Herbarium specimens often show the keel-petals of even young flowers trodden down so as to expose the anthers, something that does not normally occur unless they have been visited by a heavy insect. In many flowers the style at full anthesis grows out a little beyond the keel-tip, which must facilitate cross-pollination.

The adoption for the species of the old name D. bicolor requires comment. Willdenow ascribed the name to Humboldt & Bonpland and thought the seeds had been brought from South America, but Bonpland, who grew the same plant at Malmaison (PI. Rar. Malm. 142, tab. 56. 1816) from 1810 onward, states that he himself had brought the seed from Mexico. For reasons now impossible to trace the epithet bicolor was early transferred in horticultural literature to a cultivated form of D. obovatifolia passing at the time under the taxonomic synonym D. mutabilis. Bonpland grew the latter side by side with D. bicolor at Malmaison and Redoute’s admirable drawings of both can be compared in Pl. Rares Malm. tab. 57, 58. It appears that the genuine D. bicolor, which Bonpland recognized as more interesting than ornamental, did not endure in gardens, but the epithet did, at least for some years, attached to D. mutabilis. The last mention of genuine D. bicolor until modern times is perhaps that of George Don (1832). As early as 1823 we find D. obovatifolia (mutabilis) grown at Liverpool and figured by Hooker (Exotic Fl. 1: tab. 43) as D. bicolor; the next year at Chelsea and figured by Sims (Bot. Mag. tab. 2486) as D. mutabilis but with D. bicolor (at least sensu hortulano) given as synonymous. In 1830 Schlechtendahl (Linnaea 5: 612) identified the plate of Sims with the plant in Will- denow’s herbarium (No. 14117) and the plate in Hortus Berolinensis, stating further that the species was in general stove culture at the time under the horticultural names D. foetens or D. foetida. I have not seen the actual specimen in herb. Willdenow., and am unable to judge whether it is relevant or not. Certainly the contemporary figures of D. bicolor given by Willdenow and Bonpland do not correspond with any known form of D. obovatifolia (mutabilis) and are, in my judgment, most highly suggestive of some phases of the species here treated that occur on the Mexican plateau in Mexico and Hidalgo. It is true that the stems are not shown strongly warty, as is generally the rule in D. bicolor as I interpret the species, but this could well be an effect of cultivation under glass in a northern climate.