Notes by Gardner 1982
This species has been confused with T. prodigiosa. It differs primarily in its erect inflorescence and lower branches spreading to only about 45º degrees (the lowest on T. prodigiosa spread to 90 degrees) . It also differs in having a dense border of large trichomes on the edges of the floral bracts. Unfortunately these fall off, or are easily knocked off, after drying. Inflorescence and branch posture are difficult to determine from dried material. The lowest primary bracts of T. cossonii also tend to be larger and more leaf-like, and is the character on which assignment of the following specimens is based.
Tillandsia cossonii by Sue Gardner in J Brom. Soc.34(1): 9-10. 1984
TILLANDSIA COSSONII was described in 1887 by J. G. Baker, curator of the herbarium at Kew Gardens in England. In 1935, when Carl Mez of Germany published his final monograph on the Bromeliaceae, he had found no distinctions between this species and T. prodigiosa (Lemaire) Baker that would support its maintenance as a separate species. Since T. prodigiosa had been described in 1869, although as a species in the genus Vriesea, Mez reduced T. cossonii to a synonym of that species. Smith and Downs (1977) also lacked evidence to contradict Mez's decision and, therefore, retained it as a synonym under T. prodigiosa.
During the winter of 1980, while collecting specimens of tillandsias in the Mexican state of Michoacan, we found a species that was new to us growing in a pine and oak forest at about 2,300 meters. The plants resembled T. prodigiosa but differed in having an erect inflorescence in contrast with the pendulous inflorescence of T. prodigiosa.
Later, I had the opportunity to examine the type specimen for T. cossonii and found that while this specimen consists only of a fragment of the inflorescence, certain characters seemed to be in more agreement with our new plant than with T. prodigiosa. The character differences that could be determined from the dry material were more subtle than those found among the living specimens. For instance, the lower primary bracts are more strongly laminate (bearing leaf-like extensions) than occurs in T. prodigiosa. Another characteristic, which at first seemed to be very useful for distinguishing dried specimens, was the presence of a dense border of very large, silvery trichomes on the margins of the floral bracts. These trichomes are, unfortunately, easily knocked off a specimen during handling; therefore, their absence cannot be considered to be a key characteristic. The foliar trichomes were also found to differ significantly from those of T. prodigiosa, which have longer wing cells and a prominent "collar" or thickening at the union of the wing and the outer ring of cells. The trichomes of T. cossonii are more nearly symmetrical. Our Michoacan specimens give support to the validity of T. cossonii as a species distinct from T. prodigiosa.
Large plants such as these cannot be preserved as whole specimens, but must be cut into pieces for drying and mounting. As a result, a typical good specimen will consist of a leaf or two and a portion of the infloresence. Under these conditions, the posture of the inflorescence is difficult to determine unless a photograph, or comments regarding the plant's habit, accompanies the specimen.
Tillandsia cossonii is an attractive plant of considerable ornamental value. It is not, however, adaptable to hot, low elevation climates.
Tillandsia cossonii Baker. J. Bot. Brit. & For. 25: 279.1887. by McVaugh in Flora Novo Galiciana 1989
Habitat and flowering season. as far as known. about as in T. prodigiosa. According to Gardner (p. 173) the plant flowers in Michoacan Feb-Mar .
Sin., Nay., Jal., Mich., Mar., D.F. ("Mexican Desert" [Desierto de los Leones], Bilimek 440, K. the holotype; US!, isotype).
SW of Talpa de Allende. below the peak called Piedra Rajada (McVaugh 14293,. cited by Smith and by Gardner as T. bourgaei, annotated by Gardner in 1981 as T. cossonii); Sierra de Manantlan. SE of El Chante, E of Manantlan (Wilbur 1976,. named T. prodigiosa by Smith. and T. cossonii by Gardner in 1981).
Leaves abaxially. and especially distally, conspicuously cinereous-lepidote; primary bracts minutely lepidote; floral bracts glabrous without any ciliate margin (in an extralimital specimen, Sinaloa. Breedlove 1698), or thinly scurfy with marginal laciniate scales; sepals glabrous; inflorescence erect as far as known; panicle and scape together 25-60 cm high; panicle dense, longer than the scape, its base more or less enveloped by the sheaths of the upper scape-bracts, these smaller than the rosette-leaves but otherwise very similar; lower internodes of the panicle 1 cm long or less, the spikes at these nodes divergent at an angle or 45° or less, 6-7 cm long, not evidently longer than the middle and upper spikes; sheaths of the primary bracts appressed to and almost wholly concealing the spikes, at least the lower ones; blades of the primary bracts markedly reflexed, leafy ,20-30 cm long, 1.5-2.5 cm wide at base.
Though this appears to be the plant described from central Michoacan by Gardner (1982, p. 171), she seems to have had a somewhat more inclusive species concept. She stressed the importance of the ciliate margins of the floral bracts in T. cossonii, and perhaps for this reason she identified Gregory & Eiten 237 (MICH), mentioned above under T. prodigiosa, with T. cossonii. Except for the bract character, Gregory & Eiten 237 seems quite out of place among the specimens cited above under T. cossonii.
It is not easy to evaluate this character of ciliate floral bracts in terms of its correlation with other features. It seems to be found in a majority of the known specimens of T. cossonii as described above, and in some plants having the inflorescences intermediate between those of T. cossonii and T. prodigiosa, i.e. , moderately elongated, erect and raised above the scape-bracts, but still dense with ascending spikes, the scape-bracts only moderately leafy , the primary bracts with colored sheaths subtending but scarcely concealing the spikes, and with greatly reduced blades. The floral bracts in such specimens may be not only ciliate but persistently scurfy-lepidote on the backs. When the bracts are lepidote, the sepals are often similarly and coarsely lepidote along the keels. otherwise glabrous:
Nay. , Mpio. Acaponeta, E of La Cienaga, ca "5 mi" NW of Mesa del Nayar (Norris 14336); Jal., 5-6 km S of Chiquilistlan (McVaugh 25984).
Yet another plant with obscurely ciliate and distally scurfy floral bracts also has the primary bracts scurfy , and the sepals coarsely lepidote along the keels; in this plant the panicle is massive and crowded, ovoid or ellipsoid, 15-23 cm long and when pressed 10-14 cm wide, with 10-20 appressed spikes; neither scape-bracts nor primary bracts are conspicuous. In the "broad, ellipsoid to subglobose" inflorescence 15-20 cm long, this suggests Tillandsia macrochlamys as described by Smith (in Smith & Downs, 1977). (T. macrochlamys Baker, J. Bot. Brit. & For. 26: 142. 1888, the holotype from central Mexico, Hahn 589, P). On the other hand, T. macrochlamys is said to have 1-4-ftowered spikes 4 cm long, and primary bracts much exceeding the axillary spikes, as well as sepals as long as or shorter than the floral bracts which are 20-25 mm long. In our plant the spikes are 10-15-flowered, 6- 7.5 cm long, about twice as long as the sheaths of the primary bracts, and the sepals are 26-29 mm long.
Habit and flowering season, as far as known, about as in T. prodigiosa and T. cossonii. Collected with mature fruit in mid-Aug and mid-Jan.
Nay.: Sierra de los Huicholes, 1800-2250 m, NE of San Juan Peyotan, and 510 km by road SW of the Zac.-Nay. boundary (McVaugh 25748); "10 mi" SE of Ahuacatlan, road to Barranca del Oro (Feddema 364).
Evidently this plant deserves more study. Feddema 364 (MICH) was named by Smith in 1964 as Tillandsia bourgaei, which see under Doubtful and Excluded Species. The specimen seems to be excluded from T. bourgaei by the shape of the inflorescence, which being ovoid and less than twice as long as wide can hardly be called "narrow, cylindric or fusiform," and by the sepals, which are glabrous except on the keels, not densely lepidote as shown by Smith (in Smith & Downs, 1977, fig. 302, F). —SeeGardner 1982p. 172
