1905

American Journal of Botany 98(11): 1905–1908. 2011.

American Journal of Botany 98(11): 1905–1908, 2011; http://www.amjbot.org/ © 2011 Botanical Society of America

Karatophyllum bromelioides L. D. G ó mez is the only mega-fossil considered to be convincingly attributable to Bromeli-aceae ( Smith and Till, 1998 ; Benzing et al., 2000 ). The type specimen was described in 1972 by the late Luis Diego G ó mez Pignataro (1944 – 2009) in a paper in Revista de Biolog í a Tropi-cal ( G ó mez, 1972 ). This publication indicates that the fossil was collected in middle Tertiary deposits near the town of San Ram ó n, in the Alajuela Province of Costa Rica.

Karatophyllum bromelioides is an impression of a leaf em-bedded in a matrix of travertine ( Fig. 1A -1), with external di-mensions of 195 mm in length, 85 mm in breadth, and ca. 20 mm in depth. A medial cavity running longitudinally through the block, corresponding to the position formerly occupied by the leaf, has a maximum height of 1.6 mm ( Fig. 1A -2) and a width of 84 mm. The imprint on the internal surface preserves traces of closely parallel but distally converging veins ( Fig. 1B -3), cuticle ( Fig. 1B -5), epidermal characters, and Bromeliaceae-like scales that are particularly prominent on the abaxial leaf surface ( G ó mez, 1972 ). The specimen was interpreted as possessing

pronounced, curved marginal spines ( Fig. 1B -4), 6 mm long, 3 to 3.5 mm wide at their base, and spaced 24 to 30 mm apart.

These characters, which we confi rmed by inspection of the specimen, led G ó mez (1972) to attribute the fossil to Bromeli-aceae. He noted its possible affi nity to extant genera such as Aechmea Ruiz & Pav. and Bromelia L. in subfamily Brome-lioideae, but assigned the fossil to the new organ-genus Karato-phyllum . Karatophyllum bromelioides has been regarded by various authors as the most plausible macrofossil attributable to Bromeliaceae (e.g., Smith and Downs, 1974 ; Smith and Till, 1998 ; Givnish et al., 2007 ). Six other possible bromeliad mac-rofossils were considered unconvincing by Smith and Downs (1974 , p. 57), who described the tally as “ one impossibility, one improbability, and four faint possibilities. ” Graham (1985) also described pollen grain assigned to “ cf. Tillandsia ” from the Gatuncillo Formation of the middle(?) to late Eocene in Panama.

Karatophyllum bromelioides is of considerable interest be-cause the inferred maximum leaf thickness of 1.6 mm suggests a degree of succulence in this organ. Leaf succulence in many fam-ilies of angiosperms is closely associated with the occurrence of crassulacean acid metabolism (CAM), a photosynthetic adapta-tion to water-limited environments ( Winter and Smith, 1996 ). This relationship to succulence also holds within Bromeliaceae ( Smith, 1989 ; Pierce et al., 2002 ): data from a survey of 49 spe-cies show that a leaf thickness > 1.0 mm is associated consistently with carbon-isotope ratios less negative than − 20 ‰ , indicative of full CAM activity ( Fig. 2 ). The fossil record of CAM plants is poor to nonexistent because the dry, oxidizing conditions of arid environments are not generally conducive to preservation of mac-rofossils. Karatophyllum bromelioides could thus potentially be valuable in reconstructing the evolutionary history of CAM plants ( Crayn et al., 2004 ; Givnish et al., 2007 , 2011 ).

1 Manuscript received 3 June 2011; revision accepted 18 August 2011. Financial support for this investigation was provided by the U.S. National

Science Foundation Partnerships in International Research and Education grant 0966884 (OISE, EAR, DRL), EAR 0824299, and EAR 0418042, the Mark Tupper Fellowship, Ricardo Perez SA, and the Smithsonian Tropical Research Institute. The authors thank Jorge Aranda and Aurelio Virgo for assistance and Camilla Crifo, Judy Jernstedt, John G. Conran, and an anonymous reviewer for useful comments and reviews of the manuscript.

5 Author for correspondence (e-mail: ba.andres@gmail.com); current address: Department of the Geophysical Sciences, University of Chicago, 5734 South Ellis Avenue, Chicago, Illinois 60637 USA

doi:10.3732/ajb.1100261

BRIEF COMMUNICATION

KARATOPHYLLUM BROMELIOIDES L.D. G Ó MEZ REVISITED: A PROBABLE FOSSIL CAM BROMELIAD 1

Andres Baresch 2,5 , J. Andrew C. Smith 3 , Klaus Winter 2 , Ana Luc í a Valerio 4 , and Carlos Jaramillo 2

2 Smithsonian Tropical Research Institute, Balboa, Anc ó n, Republic of Panama; 3 Department of Plant Sciences, University of Oxford, South Parks Road, Oxford, OX1 3RB, UK; and 4 Departamento de Historia Natural, Museo Nacional de Costa Rica,

Apartado Postal 749-1000, San Jos é , Costa Rica

• Premise of the study: The fossil leaf Karatophyllum bromelioides L. D. G ó mez found in Costa Rica was proposed by G ó mez (1972) to belong to the Bromeliaceae and to date from the middle Tertiary. If the age and affi nity of this specimen were proven to be correct, it would constitute the oldest record of this large and ecologically diverse monocotyledonous family.

• Key results: Morphological features of the fossil (leaf dimensions, marginal spines, cuticular traces) indicate a close affi nity with the extant bromeliad Aechmea magdalenae (Andr é ) Andr é ex Baker. Leaf thickness (1.6 mm at maximum) suggests that K. bromelioides L. D. G ó mez performed CAM photosynthesis. The geological information does not corroborate the estimated age and location of the specimen; the fossil is suggested to be of more recent origin.

• Conclusions: The affi nity of this fossil to Bromeliaceae was confi rmed, but the uncertainties surrounding its age and collection locality mitigate against its use in inferences concerning the evolutionary history of the family.

Key words: Aechmea magdalenae ; Bromeliaceae; CAM photosynthesis; Karatophyllum bromelioides ; leaf morphol-ogy; travertine.

1906 American Journal of Botany [Vol. 98

Plant remains preserved in travertine have been found in localities 40 km NNE from San Ram ó n ( P é rez and Laurito, 2003 ) and have been dated to the late Pleistocene (50 to 13 ka).

Given the importance of this fossil, we decided to re-examine the specimen. The fossil is held in the collections of the Museo Nacio-nal de Costa Rica, Departamento de Historia Natural (collection number 46399). We discovered that Alvaro Castaing was the re-searcher who gave the fossil to G ó mez to be studied. We contacted Prof. Castaing, who indicated that he found the fossil in a small collection of Liceo de Costa Rica (a high school in San Jos é de Costa Rica, Costa Rica), and there was no associated information on their collectors, except the area where it was found. We con-tacted the principal of Liceo de Costa Rica, Milton Rojas, but the school did not possess any additional information on the fossil.

We searched the geological literature on the San Ram ó n area ( Fern á ndez and Sandoval, 1966 ; Denyer and Alvarado, 2007 ; Ž á č ek et al., 2010 ), where the fossil is supposed to have origi-nated. Most of the outcrops in the vicinity of San Ram ó n are of igneous origin ( Fig. 3 ). The few sedimentary deposits found in this area are tuffs and sandstones of Miocene age ( Fig. 3 ) that do not have the kind of lithology observed in the fossil (travertine). Other sediments in the area include Pliocene deposits near the town of Palmares (7 km SE of San Ram ó n), where vertebrate fossils have been found, but do not contain travertine deposits ( Laurito et al., 2005 ). There is a small outcrop of travertine de-posits described by Fern á ndez and Sandoval (1966) associated with hydrothermal springs on the eastern margin of the Barranca river, 10 km WSW from San Ram ó n ( Fig.3 ), from which the fos-sil could originate. We visited this outcrop, but unfortunately the entire outcrop has disappeared because the travertine has been quarried for ornamental stone. There is a small area with marine limestone that outcrops 24 km SE of San Ram ó n ( Fig. 3 ), but the fossil does not seem to have accreted in a marine environment.

Fig. 2. Relationship between carbon-isotope ratio ( δ 13 C value) and leaf thickness for 49 species of Bromeliaceae, replotted from the data of Pierce et al. (2002) .

Fig. 1. Images of (A – C) fossil Karatophyllum bromelioides L. D. G ó mez holotype and (D – F) leaf sections of living material of Aechmea magdalenae (Andr é ) Andr é ex Baker. (A) Cross section of the travertine matrix (1) showing the internal cavity remaining from the original leaf (2); maximum thickness of the leaf was 1.6 mm and 1.2 mm at the central furrow. (B) Internal view of the matrix showing imprint of the abaxial surface with close parallel venation (3), spines (note that the outline of the spine was drawn in ink on the specimen itself) (4), and cuticle traces (5). (C) External view of the matrix, adaxial surface. (D) Cross section of a fresh specimen near the widest point of the leaf blade. (E) Abaxial leaf surface . (F) Adaxial leaf surface. Scale bars = 10 mm.

1907November 2011] Baresch et al. — KARATOPHYLLUM BROMELIOIDES L.D. G ó mez revisited

to a seasonally dry climate dominated by oaks in Costa Rica in the mid- to late Pleistocene ( Cavender-Bares et al., 2011 ). It seems more probable that K. bromelioides was collected from similar Late Pleistocene to Holocene travertine deposits ( P é rez and Laurito, 2003 ) rather than Middle Cenozoic.

Despite uncertainty over its provenance and age, we consider that the fossil is convincingly assignable to Bromeliaceae on morphological grounds (leaf dimensions, cuticular imprint, and marginal spines). Although he chose to assign the specimen to a novel genus, G ó mez (1972) noted the possible affi nities of this fossil with Aechmea and Bromelia , which are represented by 17 species and 3 species, respectively, in the modern fl ora of Costa Rica ( Morales, 2003 ). The Bromelia spp. in question do

Travertine forms by rapid precipitation of calcium carbonate from solution in surface waters, near hydrothermal springs or limestone deposits ( Tucker and Wright, 1990 ). Fossil imprints and molds are often found in karstic topographies. Rain dis-solves calcium from nearby limestones during the rainy season, and then calcium precipitates again along small creeks during the dry season, preserving leaves and other remains of the sur-rounding fauna and fl ora. Pleistocene volcanic activity in this region ca. 600 ka led to uplift of the Guanacaste range to its present-day elevation and was associated with a change in vegetation from tropical rain forest to a more seasonally dry landscape, more prone to producing travertine. This is also con-sistent with genetic evidence that tropical rain forest gave way

Fig. 3. Geological map of Costa Rica showing the outcrops in the surroundings of San Ram ó n, Alajuela Province (modifi ed from Denyer and Alvarado (2007) ), travertine deposit according to Fern á ndez and Sandoval (1966) , and hydrothermal springs in Ž á č ek et al. (2010) .

1908 American Journal of Botany

Givnish , T. J. , M. H. J. Barfuss , B. Van Ee , R. Rina , K. Schulte , R. Horres , P. A. Gonsiska , et al. 2011 . Phylogeny and historical biogeography of Bromeliaceae: Insights from an eight-locus plastid phylogeny. American Journal of Botany 98 : 872 – 895 .

G ó mez P., L. D. 1972 . Karatophyllum bromelioides L.D. G ó mez (Bromeliaceae), nov. gen. et sp., del Terciario Medio de Costa Rica. Revista de Biolog í a Tropical 20 : 221 – 229 .

Graham , A. 1985 . Studies in Neotropical paleobotany. IV. The Eocene communities of Panama. Annals of the Missouri Botanical Garden 72 : 504 – 534 .

Laurito , C. , A. L. Valerio , and E. A. P é rez . 2005 . Los xenarthras f ó siles de la localidad de Buenos Aires de Palmares (Blancano Tard í o-Irvingtoniano Temprano), Provincia de Alajuela, Costa Rica. Revista Geol ó gica de Am é rica Central 33 : 83 – 90 .

Lincoln , K. , and B. Orr . 2011 . The use and cultural signifi cance of the pita plant ( Aechmea magdalenae ) among Ng ö be women of Chalite, Panama. Economic Botany 65 : 13 – 26 .

Monteiro , R. F. , R. C. Forzza , and A. Mantovani . 2011 . Leaf structure of Bromelia and its signifi cance for the evolution of Bromelioideae (Bromeliaceae). Plant Systematics and Evolution 293 : 53 – 64 .

Morales , J. F. 2003 . Bromeliaceae. In B. E. Hammel, M. H. Grayum, C. Herrera, and N. Zamora [eds.], Manual de plantas de Costa Rica, vol. II, Gimnospermas y monocotiled ó neas (Agavaceae – Musaceae), 297 – 375. Missouri Botanical Garden Press, St. Louis, Missouri, USA.

P é rez , E. A. , and C. A. Laurito . 2003 . Quercus corrugata Hooker (Fagaceae) como indicador paleoclim á tico del Pleistoceno de Costa Rica. Revista Geol ó gica de Am é rica Central 28 : 83 – 90 .

Pfitsch , W. A. , and A. P. Smith . 1988 . Growth and photosynthesis of Aechmea magdalenae , a terrestrial CAM plant in a tropical moist for-est, Panama. Journal of Tropical Ecology 4 : 199 – 207 .

Pierce , S. , K. Winter , and H. Griffiths . 2002 . Carbon isotope ratio and the extent of daily CAM use by Bromeliaceae. New Phytologist 156 : 75 – 83 .

Rossi , M. R. , V. H. Mendez , and J. Monge-N á jera . 1997 . Distribution of Costa Rican epiphytic bromeliads and the Holdridge Life Zone System. Revista de Biolog í a Tropical 45 : 1021 – 1031 .

Skillman , J. B. , M. Garcia , A. Virgo , and K. Winter . 2005 . Growth irradiance effects on photosynthesis and growth in two co-occurring shade-tolerant neotropical perennials of contrasting photosynthetic pathways. American Journal of Botany 92 : 1811 – 1819 .

Smith , J. A. C. 1989 . Epiphytic bromeliads. In U. L ü ttge [ed.], Vascular plants as epiphytes: Evolution and ecophysiology, 109 – 138. Springer-Verlag, Berlin, Germany.

Smith , L. B. , and R. J. Downs . 1974 . Flora Neotropica, monograph no. 14, Part 1, Pitcairnioideae (Bromeliaceae). Hafner Press, New York, New York, USA.

Smith , L. B. , and W. Till . 1998 . Bromeliaceae. In K. Kubitzki [ed.], The families and genera of vascular plants, vol. 4, Flowering plants, mono-cotyledons: Alismatanae and Commelinanae (except Gramineae), 74 – 99. Springer-Verlag, Berlin, Germany.

Ticktin , T. , and P. Nantel . 2004 . Dynamics of harvested populations of the tropical understory herb Aechmea magdalenae in old-growth versus secondary forests. Biological Conservation 120 : 461 – 470 .

Tomlinson , P. B. 1969 . Bromeliaceae. In C. R. Metcalfe [ed.], Anatomy of the monocotyledons, III Commelinales – Zingiberales, 193 – 294. Oxford University Press, Oxford, UK.

Tucker , M. E. , and V. P. Wright . 1990 . Carbonate sedimentology. Blackwell, Oxford, UK.

Winter , K. , and J. A. C. Smith , eds. 1996 . Crassulacean acid metabo-lism: Biochemistry, ecology and evolution. Springer-Verlag, Berlin, Germany.

Ž á Č ek , V. , T. Vorel , P. Kycl , and S. Huapaya . 2010 . Mapa geol ó gico 1: 50 000, hoja 3246-II Miramar, Rep ú blica de Costa Rica. Servicio Geol ó gico Checo, Prague, Czech Republic.

not exceed 50 mm in leaf width, but several species of Aechmea possess leaves that are 80 mm or wider. One modern species provides a particularly convincing match to the fossil: Aechmea magdalenae (Andr é ) Andr é ex Baker, an understory plant pos-sessing CAM photosynthesis that is often found in riparian habitats ( Pfi tsch and Smith, 1988 ; Skillman et al., 2005 ). The species is common from Mexico to Ecuador and Venezuela, and its distribution in Costa Rica is consistent with likely loca-tions for the origin of the fossil ( Rossi et al., 1997 ; Morales, 2003 ). Aechmea magdalenae is also known as the pita plant and has long been used for its fi ber ( Ticktin and Nantel, 2004 ; Lincoln and Orr, 2011 ).

The physical match of the fossil to both herbarium specimens and fresh material of A. magdalenae collected in the fi eld ( Fig. 1D – F ) was striking. A leaf ca. 2.6 m long had a pronounced central furrow, a maximum thickness of 1.64 mm at the middle of the leaf, and of 0.64 mm between the middle and the edge ( Fig. 1D ), and a maximum width (without following the contour of the central furrow) of 80 mm, tapering toward the tip and more gradually toward the base. The marginal spines had a maximum spacing of 45 mm at the position of greatest leaf width, decreas-ing as the leaf tapered both toward tip and base (e.g., spacing of spines 4 mm at 100 mm from leaf tip and 14 – 16 mm toward the leaf base). The largest spines, at the point of maximum leaf width, were 4 – 5 mm long and 3 – 4 mm across at their base. Leaf ribbing on the abaxial surface can also be a useful taxonomic character in Bromeliaceae subfamily Bromelioideae ( Tomlinson, 1969 ; Monteiro et al., 2011 ). Here again there was a close match between fresh material of A. magdalenae (16 to 17 parallel veins per 10 mm at the widest parts of the leaf, tapering to 21 per 10 mm at 100 mm from the leaf tip) and the fossil (18 ribs per 10 mm: Fig. 1D, E ). All of these features indicate excellent concordance between the characteristics of Karatophyllum bro-melioides and extant Aechmea magdalenae .

In conclusion, this fossil, although convincingly assigned to Bromeliaceae, is likely to be of relatively recent origin and should not be used for dating phylogenetic reconstructions of the family.

LITERATURE CITED

Benzing , D. H. , G. Brown , and R. Terry . 2000 . History and evolution. In D. H. Benzing [ed.], Bromeliaceae: Profi le of an adaptive radiation, 463 – 541. Cambridge University Press, Cambridge, UK.

Cavender-Bares , J. , A. Gonzalez-Rodriguez , A. Pahlich , K. Koehler , and N. Deacon . 2011 . Phylogeography and climate niche evolution in live oaks ( Quercus series Virentes ) from the tropics to the temperate zone. Journal of Biogeography 38 : 962 – 981 .

Crayn , D. M. , K. Winter , and J. A. C. Smith . 2004 . Multiple ori-gins of crassulacean acid metabolism and the epiphytic habit in the Neotropical family Bromeliaceae. Proceedings of the National Academy of Sciences, USA 101 : 3703 – 3708 .

Denyer , P. , and G. E. Alvarado . 2007 . Mapa geol ó gico de Costa Rica. Librer í a Francesa Press, San Jos é , Costa Rica.

Fern á ndez , M. , and F. Sandoval . 1966 . Evaluaci ó n de un dep ó sito de travertino localizado en la zona de Nagatac, regi ó n oeste de San Ram ó n. Ministerio de Industria y Comercio, Direcci ó n de Geolog í a, Minas y Petr ó leo [internal report], San Jos é , Costa Rica.

Givnish , T. J. , K. C. Millam , P. E. Berry , and K. J. Sytsma . 2007 . Phylogeny, adaptive radiation, and historical biogeography of Bromeliaceae inferred from ndh F sequence data. Aliso 23 : 3 – 26 .