Dicksonia karsteniana (Klotzsch) T.Moore, Ind. Fil.

Lehnert, Marcus & Kessler, Michael, 2018, Prodromus of a fern flora for Bolivia. XXI. Dicksoniaceae, Phytotaxa 344 (1), pp. 69-74 : 71-72

Dicksonia karsteniana (Klotzsch) T.Moore, Ind. Fil. 190: 313. 1860.

Range: —Rare in Costa Rica; common in the Andes from Colombia to Venezuela and Bolivia (CO, LP, SC).

Ecology: —Common in cool mountain rainforests to sub-paramo and open grasslands at (1200–)1800–3300(– 3600) m.

Notes: —This species represents the taxon formerly known as D. sellowiana in Bolivia ( Lehnert 2006). The true D. sellowiana is largely restricted to southeastern Brazil (Noben et al., unpublished data) and has not yet been confirmed for Bolivia. However, it may be that true D. sellowiana also occurs in small overlooked relict populations in the eastern part of the country. It should be sought in the intermittent mountain ranges of Santiago de la Chiquitania and the Serranía de Huanchaca, which may act as stepping-stones for biota between the two forest areas. These ranges consist of sandstone, a substrate where Dicksonia is often present but overlooked, growing here patchily in inaccessible narrow gorges and sinkholes. The Uruguayan population of D. sellowiana grows in a small hidden valley ca. 350 km from the nearest occurrence in Brazil, and although the distance between suitable habitat in Bolivia and Brazil is at least three times larger, it exemplifies that very small populations of these plants can persist over centuries. Dicksonia sellowiana is distinguished from D. karsteniana by pinnae to 35 × 12 cm (vs. to 77 × 29 cm); frond axes abaxially glabrescent, surfaces smooth (vs. abaxially with matted or spreading hairs, the latter sometimes abraded and leaving a scabrous surface); most adnate fertile segments basally constricted, sinuses between them notably triangular and usually becoming wider towards the pinnule bases (vs. most adnate fertile segments basiscopically decurrent, sinuses between the segments acute or rounded, not becoming wider towards the pinnule bases); and sori 0.9–1.2 mm diam. (vs. 0.9–2.0 mm).

Dicksonia karsteniana is a wide-ranging species that shows two indument extremes that might be related to certain external factors. One has the outer layer of straight reddish brown hairs weakly developed on the petioles but with an underlying denser, longer-lasting woolly pale indument that extends all along the rachis. The other variant is just the opposite, with well-developed and exceptionally dark straight hairs but almost no underlying indument and glabrescent distal petioles and rachises. These extremes have names under D. sellowiana s.l.; the woolly form “arachneosa“ (= Dicksonia sellowiana var. arachneosa Sodiro ) occurs in Venezuela to northern Peru, and may be more widely distributed, in temperate upper montane rainforests to sub-paramo; the patent-haired form “spruceana “ (= Dicksonia spruceana Kuhn ) is found on the eastern Andean slopes of Ecuador and Peru and is expected further north ( Colombia) and south ( Bolivia) in lower montane rainforests at (1400–) 1800–2000 m, usually in small numbers under a dense canopy, in deep ravines or sinkholes. These two forms usually occur very close to more typical plants of D. karsteniana , and transitions can be observed especially in shrubby paramos, where typical plants grow in sheltered forest patches and the woolly “arachenosa” is found in the open. These forms may be recognized as varieties of D. karsteniana in the future (Noben et al., unpublished data). Although Bolivian specimens seen by us cannot be considered these extreme forms, such variation may be present, especially in the remote western parts of Dept. La Paz (LP).

Lophosoria C.Presl, Gefässbündel Farrn 36. 1847 (preprint from Abh. Königl. Böhm. Ges. Wiss., ser 5, 5. 344. 1848).

Large terrestrial plants with prostrate or erect rhizomes to 1.5 m tall, with solenostelic to dictyostelic vascular tissue; a report of a trunk 8 m tall from Belize (Hawkins 1540, MO, UC) could not be verified. Most likely the original field note read “. 8 m ”, meaning a more realistic 80 cm. Detailed studies on the rhizome anatomy of Lophosoria quadripinnata s.l. ( Lucansky 1974, 1982) may apply to all species in the genus, but branching and colony formation seem to depend more on environmental factors ( Cox & Tomlinson 1985) and less on genetic disposition. Currently, two species are accepted in Lophosoria : L. quesadae A.F.Rojas , from Costa Rica and Panama, and L. quadripinnata , which spans the geographic range of the genus. The latter has two varieties, with L. quadripinnata var. contracta (Hieron.) R.M.Tryon & A.F.Tryon present only in grassy paramos and similar vegetation in the Huancabamba region ( Weigend 2002) of southern Ecuador and northern Peru; it is characterized by erect stiff fronds with ascending pinnae. Taxonomic decisions on the status of the discernable morphotaxa need additional anatomical and molecular studies. In the meantime, we here recognize only one species for Bolivia.

Lophosoria shares a common chromosome base number with Dicksonia ( Smith et al. 2006) , i.e., diploids with n = 65 ( Walker 1966). Spores of Lophosoria are among the largest known in ferns, with a perine width up to 70(–100) μm ( Murillo & Bless 1974; Gastony & Tryon 1976). They are tetrahedral-globose and have a unique ornamentation, with a strong equatorial flange and raised triangular pads between the laesural arms ( Tryon & Lugardon 1991). Identical fossil spores can be dated back to at least 113 Mya, maybe even up to 145 Mya ( Dettmann 1986, 1989), and allow retracing the historical distribution of the genus during the Mesozoic and Cenozoic. Lophosoria persisted on Gondwanan fragments except in Africa, from where no certain fossils are known, and became extinct in Australia only 3 Mya, or more recently ( Macphail & Cantrill 2006). The longest, probably continuous presence was in southern South America and western Antarctica ( Cantrill 1998, Césari 2006, Macphail & Cantrill 2006), spreading northward into areas that it still occupies today, i.e., the Valdivian forests in Chile and Argentina, the Atlantic rainforests in southeastern Brazil, the montane rainforests of the Andes, Mesoamerica, and southern Mexico, and the Greater Antilles. Available phylogenies ( Noben et al. 2017), however, do not retrace this old biogeographic scenario of northward migration; rather the substitution rate in the chloroplast genome indicates that the whole genus began to spread and radiate in the Andes ca. 3 Mya, which is about the same time it became extinct in the Old World. The most plausible explanation would be that this old genus went through a population bottleneck that removed most genetic diversity and since then expanded to its modern range, forming disjunct populations from one original, relatively small ancestral area (yet to be identified).