Pleistocene Plant Fossils in and near La Selva Biological Station, Costa Rica1

Radiocarbon dating and 40Ar/39Ar analysis of overlying tephra indicate that plant fossil assemblages exposed by stream erosion and well construction in and near La Selva Biological Station in eastern lowland Costa Rica are Pleistocene in age. We
of 8
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
  434 Horn, Sanford, Dilcher, Lott, Renne, Wiemann, Cozadd, and Vargas zyxwv Sow, V. L. 1987. Effects of predation and light on seedling establishment in zyxwv ustavia superba. Ecology 68: 1341- WAFCO, zyxwvuts . M 977. Armillariella mellea and Agrilus bilineatus and mortality of defoliated oak trees. For. Sci. 23: 1350. 485-492. zyxwvutsr Kaoru Kitajima Department of Botany University of Florida 220 Bartram Hall Gainesville, Florida 3261 1 USA. E-mail: kitajimaO BlOTROPlCA 35 3): 434-441 2003 Pleistocene Plant Fossils in and near La Selva Biological Station, Costa Rica ABSTRACT Radiocarbon dating and zyxwvutsrq 0Ar/39Ar analysis of overlying tephra indicate that plant fossil assemblages exposed by stream erosion and well construction in and near La Selva Biological Station in eastern lowland Costa Rica are Pleistocene in age. We identified plant taxa based on wood, leaves, fruits, seeds, pollen, and spores examined from three sites at ca 30 m elevation. Extrapolating from modern ranges and surface temperature lapse rates suggests paleotemperatures 2.5-3.1 C cooler than at present. RESUMEN Dataciones radiocarbbnicas y anilisis de argon (*OArI3 Ar) de la tefra sobrepuesta indican una edad PleistocCnica para las asociaciones de plantas f6siles expuestas por erosi6n fluvial y por la construcci6n de un pozo dentro y cerca de la Estaci6n Biol6gica La Selva en la bajura oriental de Costa Rica. Se identificaron 1 s tixones vegetales con base en rnadera, hojas, frutas, semillas, polen, y esporas de tres sirios ubicados a unos 30 m sobre el nivel de mar. Los resultados, basados en la extrapolacih de 1 s imbitos geogrificos y del gradiente vertical de la temperatura superficial modernos, sugiere paleotemperaturas 2.5-3.1 C mis frescas que en el presente. zyxwvut Ks word: tropical vegetation; tropical wet firest. Costa Rica; fissil plants; La Selva Biological Station; paleoclimate; paleoecology; Pleistocene; pollen; spores; WE REPORT HERE ON PLEISTOCENE LANT FOSSIL ASSEMBLAGES exposed by stream erosion and well construc- tion at and near La Selva Biological Station (10'26' N, 83 59' W) in the Caribbean lowlands of Costa Rica. Research within the 1536 ha tropical wet forest reserve at La Selva has had a strong botanical emphasis, and considerable data therefore exist on the modern flora and vegetation (Hartshorn Ham- me1 1994, Wilbur et zyxwvuts l 1994). Analyses of pollen grains and charcoal fragments in swamp sediments and soils have provided information on the last few millennia of vegetation history. This includes episodes of forest clearance, maize agriculture, and widespread fires associated with pre-Columbian human activity and possible late-Holocene droughts (Horn Sanford 1992, pers. obs.; Kennedy Horn 1997; Kennedy 1998; Horn Kennedy 2001). The new plant fossil assemblages we report here extend the time frame of ecological studies at La Selva into the Pleistocene, beyond the limit of radiocarbon dating. They provide a window on tropical forests in eastern Costa Rica before the arrival of humans in the Americas. The fossils we examined are the remains of ancient forests fortuitously preserved under clay-rich volcanic deposits exposed along the Rio Puerto Viejo and unearthed during well construction just outside the La Guaria annex (Fig. 1). Tree trunks over 30 cm in diameter are present at both the river and well I Received 26 December 2002; revision accepted 3 July 2003.  Short zyx ommunications 435 zy FIGURE 1 Station, Costa Rica. Map was adapted from the La Selva GIs. Location zyxwvutsrqp f the Puerto Viejo fossil exposures and Pozo zyxwv a Guaria well site in and near La Selva Biological sites (Fig. 2), along with other wood, leaves, fruits, seeds, pollen, and spores. Most of our samples and data come from site 1 along the Rio Puerto Viejo, but we have also observed and collected plant fossils at five other locations along the river (all sites at zyxwvu u 28 to 29 m elev.). One of us RLS) first observed and collected wood exposed along the Rio Puerto Viejo in 1986 (site 0); RLS and SPH examined and sampled the other sites in 1992. At the time, we were unaware that Alvarado (1990) had also examined the deposits and had obtained a radiocarbon date on a wood sample collected at or near our site 1. His sample, and three additional wood samples that we collected at sites 0, 1, and 2, all yielded ages older than the range of radiocarbon dating, zyxwvu e., older than zyxwvu a zyxwvut 3 00046 000 years (Table 1). The wood at Puerto Viejo Site 1 occurs within a layer of gray, clay-rich volcanic material up to 1.4 m thick, which is in turn overlain by 11.8 m of other sedimentary deposits and the modern soil horizon at the top of the riverbank. Larger wood fragments including tree trunks are concentrated along with leaves, fruits, and seeds at the base of the gray sediment, with smaller wood fragments scattered through- out the deposit at higher levels. Excavation at site 1 showed the basal leaf layer to be up to 4 cm thick. Based on preliminary field inspection, Alvarado (1990) described the gray sediment with plant fossils as a fluvially reworked volcanic deposit; however, laboratory examination of a sample collected at site 1 for *OAr/39Ar analysis revealed it to be fine-grained tephra consisting of angular plagioclase crystals in a FIGURE 2 We estimate an srcinal diameter of ca 35 cm. A smaller piece of fossil wood sits atop the trunk. Compressed tree trunk at Puerto Viejo site 1 Hands mark edges of trunk, now about 50 cm across.  436 zyxwvutsrqp orn, Sanford, Dilcher, Lott, Renne, Wiemann, Cozadd, and Vargas zyxwv TABLE zyxwvutsrqp . zyxwvutsrqponm adiocarbon dates j-om zyxwvutsrqp lant fossil localities at and near La Selva Biological Station. Lab number zyxwvutsrqp r reference Sample and context Radiocarbon age ~~~~ ~~~ p-54282 >45680 years zy P p-54283 >46350 years BP p-54374 >44000 years HP Alvarado 1990 >43000 years BP p-115184 Wood from Pozo La Guaria site >45990 years BP Charred wood from Rio Puerto Viejo site Wood from Rio Puerto Viejo site 2 Wood from Rio Puerto Viejo site 0 Wood from or near Rio Puerro Viejo site 1 dark brown bentonitic (clay) matrix. Although the matrix was altered, the texture of the feldspar grains and absence of abraded detritus indicated that the sample represents a primary airfall deposit. Whether this interpretation is true for the entire unit that includes and overlies the plant fossils will require additional field and laboratory study. The *OAr/39Ar analysis was performed on the coarsest plagioclase crystals (250-425 pm size fraction), which were concentrated by washing and handpicking. Altered glass adhering to the plagioclase crystals was removed by ultrasonic cleaning in dilute HE Approximately 30 mg was irradiated for 1.5 hours along with Alder Creek (1.194 Ma) and Fish Canyon (28.02 Ma) sanidines as neutron fluence monitors, using the calibration of Renne et al. (1998). The sample was incrementally outgassed with a defocused Ar-ion laser using established methods (Renne 1995). The fine-grain size, young age, and low K content precluded single crystal analysis. The apparent age spectrum consists of 11 steps, all of which define a plateau in the sense that the ages are mutually indistinguishable. Most of the step ages have very large errors due to the extremely small amount of radiogenic zyxwvut OAr zyxwvutsrqpon ie., weight of the air-correction). The plateau age of 126 ? 789 ka, calculated as the inverse variance weighted mean of all step ages, is dominated by two steps that released ca 30 percent of the total 39Ar. The Ca/K spectrum, determined from 37Ar/39Ar, increases dramatically in the first 10 percent of 39Ar released to a plateau value of 10 to 13. This high Ca/K ratio poses additional difficulties in obtaining precise data due to influence of the Ca correction (Dalrymple et al. 1981). The initially low values of CdK suggest the possibility of minor grain marginal alteration to K- rich clays, but this effect is not reflected in the higher temperature data that contribute most precisely to the plateau age. Though relatively imprecise, the data show that the sample is less than 915 ka (915,000 years old) at the level of one standard deviation. Since multigrain pyroclastic samples are always potentially con- taminated with xenocrysts, 915 ka is truly a maxiumum age. This age, in conjunction with the radio- carbon dates, establishes that the material was deposited sometime between 915 and 46 ka, and is thus Pleistocene in age. The maximum age of 915 ka is consistent with a K-Ar date of 1.2 ? 0.1 Ma on an andesite flow that underlies much of the La Selva reserve (Alvarado 1990). We attempted to further constrain the age of the deposit using thermoluminescence dating, but the material was insufficiently silicic to be suitable for this method (G. W. Berger, pers. comm.). The Pozo La Guaria site was discovered during hand-excavation of a well (surface elev. ca 40 m) in April 1997. We were not present, but B. Paniagua reported that at a depth of 8.3 m, the well-digger encountered a very large tree (diam ca 1 m) with intact bark D. B. Clark, pers. comm; I? Sollins, pers. comm.). Smaller wood fragments and fruidseeds were also excavated, and a subsample of this material including blocks of adhering sediment was set aside for our later analysis. The deposit with macrofossils was described as dark in color, but overlying material appears to have included gray, clay-rich sediment similar to that exposed along the No Puerto Viejo. A sample of excavated wood submitted for dating proved to be older than the range of radiocarbon dating (Table 1). The Pozo La Guaria and Puerto Viejo sites 1 and 2 may preserve material from the same volcanic event. A well dug in the late 1980s in the vicinity of La Guaria seems to have reached a similar deposit of fossil-rich, gray clay I? Sollins, pers. comm.). Other deep excavations within the reserve have not encountered the material, including some that reached the lava flow dated by Alvarado (1990) to 1.2 ? 0.1 Ma P Sollins, pers. comm.). These observations indicate that the material, which could be of more than one age, has a discontinuous distribution.  Short zyx ommunications zy  7 zy TABLE 2. zyxwvutsrqpo lant material identified in the zyxwvut ossil deposit exposed along the Rio Puerto Vido. * indicates materialfiom site 2 exposed on south side of river; zyxwvutsrq ll other material s fiom site 1. Taxonomy ollows Wilbur et al. (1994). Leaveslleaflets: Chrysobalanaceae (Parinurz) Fabaceae-Mirnosoideae (tribe Ingeae) Lauraceae (Nectandru, Ocotea) Piperaceae Pollen grains Aquifoliaceae Zh) Asteraceae Begoniaceae (Begonia) Betulaceae (Ahus) Sapotaceae (Pouteria) Bornbacaceae Cecropiaceae (Cecropia) Fruitslseeds Chloranthaceae (Hedyosmum) Annonaceae (Xyopia) Cunoniaceae Weinmannia) Arecaceae (Astrocaryum) Cyperaceae Chrysobalanaceae (Purinurz) Ericaceae Fabaceae-Mirnosoideae (tribe Ingeae) Hurniriaceae (Sacoglottis) Malpighiaceae (Heteropteris) Malpighiaceae (Byrsonimu) zyxwvutsrq elastornataceael ornbretaceae Menisperrnaceae (Abuta) Myricaceae (Myrica) Sapotaceae (Pouteria) Poaceae Wood Polygalaceae (Polygula) Bignoniaceae (Jacaranda, Tubebuia) Cornbretaceae Teminalia) Fabaceae-Mirnosoideae (Znga or Stryphnodendron; Lauraceae (Nectandra or Ocotea) Sapotaceae (Pouteria) Ulmaceae Trema) Polypodiaceae Euphorbiaceae (Aculypha, Alchornea) Piperaceae (Piper) Sapotaceae Urticales Vochysiaceae zyxw chysiu *Ingu, Stryphnodendron, or Pithecellobiurn) Fern spores Cyatheaceae Cyatheu, Cnemidariu) Selaginellaceae (Selaginella) Other fern spores Tables 2 and 3 list plant taxa identified from leaves, fruits, seeds, wood, pollen, and spores in the Puerto Viejo and Pozo La Guaria deposits. A few macrofossils were identified in the field by authors RLS and OV, but most data reported here are derived from laboratory study of fossil leaves, fruits, and seeds by DD and TAL with assistance from OV; of fossil wood by MCW; and of fossil pollen and spores by SPH and DAC. Fossil leaf material collected at Puerto Viejo site 1 was washed with commercial detergent to separate the leaves and remove debris, then soaked in HF to remove silicates. Over 100 cuticles and whole-mount mesofossil slides of fossil leaves, and specimens of modern leaves from over 100 species, were prepared and examined to identify fossil leaves in the deposit. Wood samples were not silicified, and it was possible to cut smooth surfaces and make hand-sections with a razor blade. In a few cases, the samples were too degraded and crumbly to prepare adequate TABLE 3. Fruitslseeds: Pollen grains Arecaceae (Rtrocuryum) Arecaceae Hurniriaceae (Humiriastrum, Sacoglottis) Asteraceae Fern spores Cyperaceae Cyatheaceae (Cnemidaria) Ericaceae Other fern spores Euphorbiaceae (Aculypha) Plant material identified in the Porn La Guaria deposit. Taxonomy ollows Wilbur et a. (1994). Cecropiaceae (Cecropia) MelastornataceaelCornbretaceae Myricaceae (Myrica) Poaceae Rutaceae (Zanthoxylum) Sapindaceae Scrophulariaceae Ulrnaceae (Ceftis, Eernu) Verbenaceae Zh)  438 Horn Sanford Dilcher zyxwvutsrq ott, Renne Wiemann Cozadd and Vargas zyxwv sections. For the intact samples, cross, radial, and tangential sections were prepared and mounted on glass slides for microscopic examination. Macroscopic anatomy was observed using a zyx ox hand lens on smoothed surfaces, and microscopic anatomy (size and arrangement of vessels, rays, and axial parenchyma) was observed using a compound microscope. A list of potential identifications was generated using the computerized GUESS wood identification program (Wheeler zyxwv t al. 1986). These were compared with authentic specimens to verify genus. Pollen and spores were prepared for microscopic analysis using standard palynological procedures (HCI, HF, KOH, acetolysis; Berglund 1986), and identified based on an extensive reference collection of tropical taxa. If our interpretation of a primary ashfall is correct, the plant taxa identified from macrofossils exposed along the Puerto Viejo represent plants that were growing at or near the collection sites at the time of burial; however, if the deposit includes fluvially transported volcanic sediment (Alvarado 1990), it may also include plant fossils transported from upslope forests. The distribution of macrofossils within the volcanic deposit at site 1, and the presence of some charred wood at its base, is suggestive of a forest buried in place (Fritz 1986), but more detailed taphonomic and geological studies would be required to establish this. For the Pozo La Guaria samples, we have less information on sample context and cannot rule out some downslope movement of macrofossils in a stream or volcanic flow. Given the possibility of downslope transport, it is interesting to note that all but two taxa identified from macrofossils are found within modern forests of the La Selva Biological Station (Wilbur et al. 1994). The fossils of Parinari (Chrysobalanaceae) found at Puerto Viejo site 1 and Humiriastrum (Humiriaceae) found at Pozo La Guaria are outside the modern ranges of the genera. The online taxonomic database for Costa Rica’s Instituto National de Biodiversidad zyxwv nd the Vas- cular Tropicos database of the Missouri Botanical Garden html) indicate a modern lower elevational limit for Parinari Z? excelsa or Z parvifolia) on the Caribbean slope of 500 m. Hurniriastrum (represented by one species, H. diguense) has been collected only above zy 00 m on the Caribbean slope. The La Selva checklist (Wilbur et al. 1994) lists one taxon identified from macrofossils (Byrsonima) as a cultivated species. Although more often associated with drier habitats of the Pacific slope, Bryronima can be a natural component of wetter forests (Anderson 1983). Several large specimens of B. crass olia occur today along the Rio Puerto Viejo, quite likely having established naturally, and B. cripa has been collected in primary premontane wet forest at 1000 m elevation on the Caribbean slope (INBIO data- base). While microfossils (pollen and spores) can be carried long distances by wind, our previous studies of pollen in surface soils at La Selva (Horn et al. 1998) and elsewhere in Costa Rica (Rodgers Horn 1996) indicate that soil pollen spectra within forests are to a large degree dominated by the pollen of local plants. If the Puerto Viejo and Pozo La Guaria macrofossil assemblages represent ancient forests buried in place, the associated pollen and spore assemblages should primarily reflect the plants that once grew at the sites. If macrofossils were transported downslope, however, microfossils may also have been transported. As is true for the plant macrofossils, most of the pollen and spores at the Puerto Viejo and Pozo La Guaria sites represent plant taxa in the modern La Selva flora (based on Wilbur et al. 1994, with additional information on tree ferns from G6mez 1983). Three genera represented by fossil pollen (Alntls, Myrica, and Weinmannia) do not grow at La Selva today. These wind-pollinated trees and shrubs are presently restricted to premontane and montane forests growing ca. 500-1000 m or more upslope from La Selva (Horn Rodgers 1996-1997, Burger 1977, and online databases cited above). Together, these three genera account for 9 percent of the total pollen enumerated in five samples prepared from sediment associated with site 1 macrofossils. This percentage of “extralocal” pollen is higher than we have found under modern closed forest canopies at La Selva (Horn et al. 1998; S. Horn, pers. obs.), but is similar to pollen percentages in surface and recent samples from the open-canopy Cantarrana swamp (Kennedy 1998), which receives pollen from upslope plant communities carried by downslope winds. The impor- tance of premontane/montane pollen types in the Puerto Viejo and Pozo La Guaria sediments is consistent with the presence of macrofossils of two genera presently found only upslope from La Selva. Because of the potential for long-distance wind dispersal of pollen, the plant macrofossils, which are generally dis-
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!