A palynological approach to reconstruct El Niño-variability; preliminary results from a sub-tropical Florida wetland

Timme H. Donders, Frederike Wagner, and Henk Visscher
Department of Geobiology, Botanical Palaeoecology, Utrecht University, Budapestlaan 4, 3584 CD Utrecht, The Netherlands

A peat profile from a mixed cypress swamp in Florida (U.S.A.) has been examined for pollen and spores. The swamp ecosystem is known to be very sensitive to changes in water availability ¹. The presented record reveals several clear changes in vegetation composition throughout the last millennium. Three distinct phases are present in the profile, of which the most recent one can be attributed to human-induced lowering of the waterlevel by draining large parts of the Florida wetlands ². If this latest shift in vegetation patterns is a result of artificial changes in hydrology, the earlier phases may result from natural variability of the water resources. The different phases in the record therefore reflect past hydrological conditions of the area.

In addition, frequent sub-decadal fluctuations are observed in the palynological record. Pollen production has been shown to react to year-to-year climate variability. Therefore the highest-frequency shifts in the ratio between water-sensitive and other taxa are most likely connected to dry/wet spells of the climate.

The strength of the El Niño Southern Oscillation accounts for over 50% of the precipitation available to the Florida vegetation during the growing season ³. Thus the presented record has the potential to serve as a proxy for variability in frequency, persistence and magnitude of El Niño events during the last millennium. Dating of the profile (currently in process) to calibrate the pollen data with climate parameters obtained from direct measurements and historical archives will be a crucial step in establishing these connections. It will also provide insight in the general development of the swamp, which is one of the very few remaining virgin forest stands in Florida.

¹ Willard, D. A., Holmes, C. W. & Weimer, L. M. The Florida Everglades Ecosystem: Climatic and Anthropogenic Impacts over the Last Two Millennia. Bulletins of American Paleontology, 361, 41-55 (2001).
² Swayze, L. J. & McPherson, B. F. (U. S. Geological Survey, Water Resources Division, Tallahassee, 1977).
³ Sun, H. & Furbish, D. J. Annual precipitation and river discharges in Florida in response to El Niño- and La Niña-sea surface temperature anomalies. Journal of Hydrology, 199, 74-87 (1997).

Response of Late Carboniferous tropical vegetation to transgressive-regressive rhythms at Joggins, Nova Scotia

H. J. Falcon-Lang
Department of Earth Sciences, University of Bristol, Bristol, BS8 1RJ, UK
e-mail: howard.falcon-lang@bristol.ac.uk

Fossil plant assemblages are described in their sequence stratigraphic context from the Upper Carboniferous (Langsettian) Joggins Formation of Nova Scotia to elucidate ecosystem response to transgressive-regressive rhythms. Results show that rising base-level resulted in retrograding submerged mires co-dominated by Lepidodendron and Lepidophloios, which were replaced by short-lived Paralycopodites communities following mire drowning. Extensive brackish bays existed during early highstand, distally fringed by (pro)gymnospermous coastal/upland vegetation. Late highstand bay filling generated prograding distributary wetlands dominated by flood-disturbed lycopsid-pteridosperm-sphenopsid communities and cordaite mangroves. As base-level fell, well-drained alluvial plains were dominated by fire-prone cordaite and/or Sigillaria communities, which persisted until the next phase of base-level rise. This rhythmic ecosystem succession repeatedly occurred on a 100-500 ka timescale and was likely important in maintaining high tropical biodiversity.

Pollen and spores tell nearly everything...- and often nothing

Christa-Ch. Hofmann
Institute of Palaeontology, University of Vienna, Austria

There is nothing new in stating that palynomorph assemblages are generally taphonomically biased mixtures of taxa derived from a wide range of vegetation types that grew directly at, more-or-less adjacent to, and relatively remote from the depositional site. Consequently, it is quite difficult to get the right picture of the different vegetation types. Particularly challenging are palynomorph assemblages that are strongly dominated (percentages over 55%) by one taxa or group, e.g. Pinus peaks, Glyptostrobus peaks, fern peaks, etc., which either represent the result of taphonomical processes (transport and facies biased) or reflect a particular vegetation type. The following few examples can give only a very brief overview.

Today, Pinus produces huge amounts of wind-dispersed pollen, and the trees are planted in vast monocultures for pulpwood, etc. As a modern example, several Pinus spp. are planted on the graben shoulders surrounding the taxa rich wetland system of the Mobile Delta (Alabama, U.S.A.). Surface samples of the Mobile Delta sediments are mostly over-dominated by Pinus pollen (up to 80%) and one has to count at least over 2000 grains per slide to get a reasonable amount of taxa originating from the wetland. This situation might be comparable with the rather small Tertiary coal-bearing intramontane basins of Austria, where the zonal forests of the palaeo-"hinterland" surrounding these basins consisted, amongst other taxa, of different Pinus species. The input of wind transported "hinterland" taxa can be well observed in lacustrine sediments deposited at the basin margin (Pinus of 50 to 90%). The circumstance that wetland Pinus taxa existed as well complicates the matter further.

The over-representation of the wetland taxon Glyptostrobus is very common in palynomorph assemblages of the Austrian Tertiary e.g. (Koeflach-Voitsberg Linite area, Badersdorf) and is often interpreted to represent coal swamp facies. In this case, it is important to investigate the sedimentary and organic facies in order to distinguish between more-or-less autochthonous or allochthonous Glyptostrobus pollen, because lakes and ponds often function as a sink for windblown Glyptostrobus (and as well as Pinus) pollen.

Fern peaks also are quite enigmatic. Quantities of fern spores encountered in Recent (Orinoco Delta in Venezuela and Mobile Delta) and Quaternary (Kathmandu Basin in Nepal, Spiti Valley in India) fluvial-deltaic environments reach sometimes up to 80% (average ca. 30%) and are a result of the comparable transport und settling behaviour of both fern spores and sand grains. On the other hand, Tertiary fern peaks occur in fluvio-lacustrine environments and lignites/coals. The latter tend to be interpreted as autochthonous monotypic vegetation, in particular if accompanied by fusinite (post fire re-colonization). The interpretation of fern peaks from clastic sediments is dependent on the facies, as the transport and settling behaviour of fern spore compares more-or-less to that of sand. Assemblages from higher energy environments might be completely allochthonous whilst very distal floodplain assemblages might be authochthonous (big Osmunda spores in very fine clays). It is crucial to unravel the taphonomical bias of palynomorph assemblages when the palynological data base is used to demonstrate hazards in earth history such as meteorite impacts, volcanic activities, fires, and last-not-least climate changes.

Palynological investigations on Randecker Maar (Baden-Württemberg, Germany)

Sabine Kottik
Institut für Paläontologie, Universität Wien, Geozentrum, Althanstr. 14, 1090 Wien

Samples from laminated sediments of the Randecker Maar representing the euxinic basinal facies of the maar lake were analysed for pollen and spores. The age of Randecker Maar is Lower to Middle Miocene (uppermost Karpatian/ lowermost Badenian) dated by micromammals of the MN 5 zone. The palaeoclimate is considered to have been humid-subtropical with a comparatively short dry season.

Fifty different taxa of pollen and spores belonging to 24 families of ferns, gymnosperms and angiosperms have been discovered. The presence of Selaginella, Ephedra, Araliaceae, Empetraceae and Styracaceae is documented for the first time. The abundant occurrences of leaves, fruits and seeds attributed to the modern genus Gleditsia (Fabaceae) and the very abundant pollen grains of Tricolporopollenites wackersdorfensis (extinct genus Podocarpium, Fabaceae), particularly in situ in flower material (LIU et al. 2001) are discussed. Based on the floral and faunal remains an attempt is made to reconstruct the Miocene environment of the crater lake.

Pollen (and leaves): implications for quantitative climate reconstructions

Norbert Kühl
Institut für Paläontologie, Universität Bonn, Nussallee 8, D-53115 Bonn, Germany

Climate reconstructions using plant fossils depend on:

1. the quality of the fossil record
2. the determination of the modern climate requirements of the vegetation or individual taxa
3. a procedure to connect these two data sets to deduce the paleo climate (transfer functions).
Each of these three topics has a major impact on the results.

The fossil record is a result of the input and depositional conditions for plant remains. Its quality also depends on the taxonomic level to which the remains can be determined. Pollen is abundant and can be transported over long distances. In combination with suitable sediments, it reflects vegetation changes over time. The advantage of its availability is facing the fact that pollen can often be determined to genus or family level only. Since macrofossils can often be determined to species level, they have the potential to fill in this gap, however, they are less abundant and often reflect very local conditions of the sediment trap. Since the fossil record gives us only an incomplete picture of the paleo vegetation, a method for climate reconstruction preferably includes as much information as possible, i.e. pollen (including rare, but significant taxa) as well as macrofossils.

Determining the modern climate requirements of the proxy data is crucial for providing a vegetation-climate relationship that can be applied to the paleo situation. This modern relationship to climate is either expressed for an assemblage or for individual taxa, depending on the method. All methods assume that this relationship to climate did not change over time. Since that is not always certain, reconstruction methods should be robust to this uncertainty.

The assemblage approach uses modern surface pollen samples as modern analogues. The method interpretes the climate as analogous paleo climate, which exists at the locations from where the surface samples originate which are most similar to fossil samples. It therefore needs a good database of modern surface samples that covers as many vegetation types as possible. For the reconstruction, it depends on the existence of modern vegetation being analogous to the vegetation in the past.

The indication species method (ISM) connects distribution data to climate, thus being able to incorporate macrofossils and being quite robust to the lack of modern analogous assemblages. An improvement of this method which we have developed during the last few years in Bonn is based on probability density functions (pdfs). It yields a reconstructed most probable climate and weighs each taxon to its climatic significance. By using many taxa, it is robust to a potential difference in climate dependency of an individual taxon between today and the past. However, good modern chorological data are inevitable for both the ISM and the pdf-method.

Reconstructions of Eemian mean January temperatures based on a pollen diagram from Bispingen/Luhe are presented. The low winter temperatures in the later part of the Eemian as well as a major cold event, as reconstructed using the assemblage approach, is not reproduced by the reconstructions using the pdf-method. Reasons for the differences between the reconstructions using the assemblage approach and the pdf-method are proposed. It appears that a reconstruction method should on the one hand use a maximum of information and on the other hand should be robust to the uncertainties that exist when comparing fossil and modern data to reconstruct paleo conditions.

The Anisian macroflora from Kühwiesenkopf/Monte Prà della Vacca in the Northern Dolomites (Italy)

Evelyn Kustatscher ¹, Johanna van Konijnenburg-van Cittert ²,
Carmen Broglio Loriga ¹, and R. Posenato ³
1 Dipartimento delle Risorse Naturali e Culturali, Università di Ferrara,
Corso Ercole I d´Este 32, 44100 Ferrara, Italy
2 Laboratory of Palaeobotany and Palynology, Budapestlaan 4, 3584 CD Utrecht, Netherlands
3 Dipartimento di Scienze della Terra, Università di Ferrara, Corso Ercole I d´Este 32, 44100 Ferrara, Italy

The knowledge of Triassic macroflora from the Dolomites mostly concerns the Ladinian, while literature data on Anisian plants are scarce. This gap is filled by the discovery by M. Wachtler from San Candido (South Tyrol) of a rich plant deposit within the Dont Formation located at Kühwiesenkopf/Monte Prà della Vacca (Prags/Braies Dolomites).

The Dont Formation of a marine marginal basin environment is well-known through the study of Bechstädt & Brandner (1970) and Senowbari-Daryan et al. (1993). It is more than 200 m thick in this locality, and the plant-bearing beds are placed about 75 m from the base. In the plant horizon the fossil plants are concentrated in lens-shaped layers of siltstone, marly siltstone and carbonate siltstone. At the base of the horizon a number of trunks occur on the upper surface of a carbonate siltstone layer, roughly N-S oriented. Occurrence and preservation of terrestrial (plants and reptiles) and marine fossils (bivalves, brachiopods, ammonoids, gastropods) within the plant horizon are presumably related to very rapid burial events caused by gravity flows within a marginal marine basin, connected with storms in the terrestrial domain (Tintori et al. 2001, Broglio Loriga et al. 2002).

A preliminary systematic analysis allowed least 17 fossil plant genera (both leaves and fructifications) to be identified: ?Isoetites, Lycophyta new taxon (Lycophyta), Equisetites (Sphenophyta), Anomopteris, Neuropteridium, Cladophlebis, Crematopteris, ?Marattiopsis (Pteridophyta), ?Sagenopteris, Scytophyllum, Peltaspermum (Pteridospermae), Bjuvia, Taeniopteris, Dioonitocarpidium, Pterophyllum/Nilssonia (Cycadophyta), Voltzia, ?Voltzia, Albertia (Coniferophyta).

References
Bechstädt T. & Brandner R. (1970) - Das Anis zwischen St. Vigil und dem Höhlensteintal (Pragser-und Olanger Dolomiten, Südtirol). Festband des Geol. Inst., 300-Jahr-Feier Univ. Innsbruck, 9-103, Innsbruck.

Broglio Loriga C., Fugagnoli A., Gaetani M., van Konijnenburg - van Cittert J., Kustatscher E., Mantovani N., Posenato R., Renesto S., Tintori A. & Wachtler M. (2002) Il giacimento a piante di Kôhwiesenkopf / Monte Prà della Vacca (Anisico, Dolomiti di Braies): una proposta di salvaguardia. In: A. Tintori (ed.), Giornate di Paleontologia, Verona-Bolca-Priabona 6-8/6/2002, abstracts p. 13.

Tintori A., Posenato R., Kustatscher E. & Wachtler M. (2001) - New Triassic fish faunas from paralic environments in the Alps. 3rd International Meeting on Mesozoic Fishes, 26-31 August 2001, Serpiano (CH).

Senowbari-Daryan B., Zühlke R., Bechstädt T. & Flügel E. (1993) - Anisian (Middle Triassic) buildups of the Northern Dolomites (Italy): The recovery of reef communities after the Perm/Triassic crisis. Facies, 28: 181-256, Erlangen.

Structural and functional aspects of secondary growth in Palaeozoic plants: quantitative analysis and simulation

Tom Masselter ¹, Nick Rowe ², and Thomas Speck ³
1+3 Botanischer Garten der Universität Freiburg i. Br., Germany
2 Botany and Plant Architecture Bioinformatics, Montpellier II, France

The aim of this work is to understand aspects of different constructional strategies of extinct woody plants. Some Palaeozoic plants lack significant development of secondary outer tissue such as bark that allows an increase in stem diameter. As a consequence, secondary growth of internal tissues such as xylem and phloem may lead to mechanical destruction of parenchymatous inner cortical tissues without its replacement by secondary cortical tissues. Fossil plants studied include stems from a Palaeozoic seed fern (Lyginopteris oldhamia) that were measured in order to analyse changes of different tissues due to secondary growth during ontogeny. The results will be used as a basis for simulating the growth of these plants with both physical models and computer models (FEM & CAD). The investigations will help to understand developmental constraints influencing the appearance of tree-like growth forms in the Devonian radiations and help us to understand why the lignophyte/seed plant bauplan has been so successful and perhaps why other constructional strategies have become extinct.

Dispersed cuticles from Lower Miocene coal seams in Lusatia and adjacent areas

Wilfrid Schneider
Alte Berliner Strasse 13d, D-02977 Hoyerswerda, Germany

Associations of dispersed cuticles found in different strata of coal seams describe different taphocoenoses including leaves and leaf fragments from peat forming plant communities. Cuticle associations from leaf coals as well as coarse or fine detritic sedimentary seam strata proved to be different from the composition of contemporaneous leaf bearing clays and sediments including carpological fossils. For that reason the evolution of swamp and bog floras throws light on a specific aspect of phytostratigraphy.

Cuticles from Lower Miocene seams belong to conifers, dicotyle woody plants, palms, graminifolious plants and herbaceous aquatic plants. The conifer cuticles are divided into two groups: Surviving taxa from Paleogene swamp communities (Sequoia, Glyptostrobus, Quasisequoia, ?Cryptomeria) and Late Oligocene immigrants (Cunninghamia, Sciadopitys, Taiwania, Cupressospermum). The stratigraphic distribution of some conifer cuticles is characterized by changes of striking advances and absence intervals. Particularly the first immigration of Cunninghamia and Sciadopitys with the beginning formation of the 4nd Lusatian (=Bitterfeld) seam horizon (Oligocene/Miocene boundary) proves a significant phytostratigraphic marker. After an absence interval including seam "Lübbenau" and 3rd Lusatian seam horizon Sciadopitys comes back during the "Lower rider" seam, which ist dated by dinocysts as top of Lower Miocene. But the second presence of Cunninghamia begins later: middle layer of 2nd seam horizon.

Kalmiophyllum marcodurense KRÄUSEL & WEYLAND (Ericaceae) immigrates into Central European swamps during the formation of seam ÒCalauÓ resp. ÒGršbersÓ on the final stages of 3th Rupelian oscillation. The cuticle of this evergreen shrub predominantes in leaf taphocoenoses of all Miocene seams. In contrast, the stratigraphic distribution of the cuticles of Magnolia liblarensis (KRÄUSEL & WEYLAND) KVACEK as well as Pilimparicutis thomsoni (KRÄUSEL & WEYLAND) SCHNEIDER (cf. Sapotaceae) is characterized by abundance in the thick seams of 4th and 2nd Lusatian seam horizons and real absence in thin seam layers of the 3rd horizon and seam "Lübbenau". The cuticles of sabaloid palms show an analogous pattern of distribution. On the other hand the cuticles from Glumiflorae are more frequently in the thin seams. Both the very frequent cuticle Glumophyllum spinosum WEYLAND and the carpological taxon Caricioidea iugata (NIKITIN) MAI belong to the same botanical taxon. Finally, Varipilicutis liblarensis (KRÄUSEL & WEYLAND) SCHNEIDER (not Dioscorea but probably belongs to Zingiberaceae) proves to be an important component of all Miocene seams.

Recent developments and applications of CLAMP

R. A. Spicer
Earth Sciences Department, The Open University, Walton Hall, Milton Keynes, MK7 6AA, UK

Climate Leaf Analysis Multivariate Program (CLAMP) is now a well established technique for determining a range of palaeoclimate parameters from the physiognomy of fossil leaves representing woody dicots. Because leaves are particularly sensitive to humidity and temperature such variables as mean annual temperature, warm month mean temperature, cold month mean temperature, precipitation during the growing season, three driest and three wettest months, relative and specific humidities and enthalpy can be postulated. Statistical uncertainties for all these parameters can be given for particular data sets and combinations of variables. Successful as it is the method does have some limitations. Principally these are the quality of the fossil assemblage in terms of the physiognomy in the source vegetation and the ability of leaf physiognomy to reflect extreme warmth and cold. With respect to the former the quality of the analysis rests on the recognition of the taphonomic processes involved in assemblage formation and selecting only those sites that meet appropriate criteria. With respect to the latter the reference data sets are being expanded to investigate these uncharted regions of physiognomic space. The results so far are intriguing and not yet fully understood.

Two contrasting applications of CLAMP for determining a Cretaceous continental interior climate and the uplift history of Tibet will be presented.

Fractal dimension of leaves - a new proxy for palaeoclimate reconstructions

Jorinde Sprong and Wolfram M. Kürschner
Dept. of Geobiology, Botanical Palaeoecology, Utrecht University, Budapestlaan 4, 3584 CD Utrecht, The Netherlands

Clouds are not spheres, mountains are not cones, coastlines are not circles and bark is not smooth, nor does lightning travel in a straight line (Mandelbrot, 1983). All these patterns and structures in nature are either irregular or fragmented to such a degree that "normal" euclidic geometry cannot describe their forms. Fractals and the fractal dimension (FD) have proved to be useful for characterizing complex structures. As a corollary FD should be a quantitative mean to measure leaf shape complexity. In palaeoclimatic research, the ratio between toothed and entire-margined leaves have been proved to be very useful as a botanical proxy for past temperatures. Yet, the description of leaf shape complexity is rather arbitrary and only qualitative depending on the experienced eyes of the worker. Fractal analysis of fossil leaf assemblages may represent an unambiguous measure of terrestrial palaeotemperatures. Typical leaf shapes (Chrysobalanus icaco = lanceolate entire, Acer campestre = lobed entire, Acer pseudoplatanus = lobed toothed, Castanea sativa = simple toothed and Betula pendula = compound toothed) have been analyzed for their fractal dimension. (1) Leaf samples were digitized with a resolution of 600 dpi and (2) the fractal dimension was estimated using the fractal "outline reduction" routine (with standard 30 erosion steps) of the image analysis program AnalySIS 3.0. Fractal dimensions varied significantly between 1.054 and 1.13. Mean FD of the different leaf shapes was: Chrysobalanus icaco 1.062 +/- 0.00297, Acer campestre 1.059 +/- 0.002711, Acer pseudoplatanus 1.066 +/- 0.002043, Castanea sativa 1.107 +/- 0.004928, Betula pendula 1.112 +/- 0.008637. FD was significantly different between all leaf shapes (p< 0.0001), for example between toothed vs. entire, lobed entire vs. lanceolate entire, lobed toothed vs. lobed entire. Consequently, FD represents an excellent estimation of leaf shape complexity. Future work will show if FD is a good parameter to estimate palaeotemperatures alternatively to conventional methods.

Leaf physiognomy and climate: analysis of contemporary patterns and fossil leaf assemblages

Christopher Traiser, Stefan Klotz, and Volker Mosbrugger
Institut für Geowissenschaften, Universität Tübingen, Sigwartstr. 10, 72076 Tübingen, Germany e-mail: christopher.traiser@uni-tuebingen.de

The distribution pattern of leaf physiognomic characters in different geographic regions of the world is used in many environmental studies in order to analyse ecosystem interaction. n this approach, the leaf physiognomic composition of woody angiosperm floras of Europe is investigated. The leaf physiognomic grid data set is compiled from "synthetic chorologic floral lists" based on distribution maps of 108 extant plants.

The calibration data set considers only those grid cells with a minimum of 25 taxa and elevations lower than 400m. The leaf physiognomic composition of 25 different leaf characters is calculated for each grid cell. The spatial distribution of many leaf physiognomic characters clearly shows a more or less zonal organisation. Among other environmental parameters such as leaf area index, biomass and soil types, leaf physiognomic composition of floras is also correlated with climatic data.

Transfer functions using different multivariate statistical approaches (multiple linear regression and redundancy analysis in ordination) are calculated in order to predict climatic parameters on the basis of leaf physiognomy. The prediction of climatic values shows substantial similarity with present day climatic data. This is especially true for temperature related parameters e.g. mean annual temperature (MAT), mean temperature of coldest and warmest month, whereas precipitation related parameters are predicted insufficiently e.g. mean annual precipitation and precipitation of driest/wettest month. The residual plot of predicted and real MAT is used to verify the accuracy of estimations. Approximately 75% of the data is predicted within the standard error of estimate (SE= +/- 0.9°C).

The climatic transfer functions are applied to three fossil localities of Late Oligocene and Middle Miocene age. The palaeoclimatic results are compared to those derived from the coexistence approach and other existing climatic analyses of this fossil sites. These results reveal that palaeotemperature estimates are basically in agreement with other reconstruction methods. Summarising, this approach permits the analysis of contemporary leaf physiognomic distribution patterns and their correlation with different environmental parameters and provides viable palaeoclimatic estimates of European fossil leaf assemblages.

Stomatal-density and -index in Upper Permian conifers

Dieter Uhl ^1,2 & Hans Kerp ¹
1 Forschungsstelle für Paläobotanik, Westfälische Wilhelms-Universität, Hindenburgplatz 57, D-48143 Münster, Germany
2 Present address: Institut für Geowissenschaften, Eberhard-Karls-Universität, Sigwartstrasse 10, D-72076 Tübingen, Germany e-mail: dieter.uhl@uni-tuebingen.de

Changes in atmospheric CO₂ concentrations over the geological past have attracted a considerable amount of scientific interest in the last decade, due to the fact that this substance is an important, anthropogenically influenced, greenhouse-gas. It has been demonstrated by WOODWARD (1987) and many other authors, that at least in some extant plants stomatal-densities and -indices are negatively correlated with atmospheric CO₂. Similar changes of stomatal-densities and -indices in fossil plants are considered by many authors to be excellent proxies for changes of CO₂ concentrations in the geological past, as far back as the Palaeozoic (e.g. McELWAIN & CHALONER, 1996; RETALLACK, 2001). However, up to now systematic investigations on the variability of these parameters in most Palaeozoic plants are still lacking. Here we present the first results of an initial study on these parameters in some conifers from the Upper Permian of Central Europe.

As it has been demonstrated by McELWAIN et al. (2002) stomatal-density and -index from conifers with stomata arranged in distinct rows are not useful as indicators of atmospheric CO₂, due to the disturbing effects of such a pattern. This type of stomatal arrangement is present in most conifers from the Upper Permian of Central Europe, i.e. in the genera Ullmannia GÖPPERT, Pseudovoltzia FLORIN and Ortiseia FLORIN/Culmitzschia ULLRICH. Fortunately there are other widely distributed conifer taxa from this period, like the different species belonging to Quadrocladus MÄDLER, which show a more uniform distribution of stomata on both surfaces, without such disturbing patterns as in the other taxa mentioned.

Within individual leaves of this taxon variations of stomatal-density and -index were up to 50%, with a more or less constant increase from the base to the apex of a leaf (both surfaces). In cuticles of Quadrocladus from the locality Culmitzsch in Thuringia (E-Germany), which consists of fluvial sediments contemporary to the Upper Permian marine Zechstein, the overall range of both parameters was smaller than in cuticles from the marginal marine locality Frankenberg-Geismar (NW-Hesse, W-Germany). This difference is probably due to the wider variety of (micro-)climatic and edaphic conditions from which the cuticles may have originated in the latter locality. Stomatal-densities from the Upper Permian of S-Tyrol, pooled from all conifer taxa investigated, showed a significant increase of stomatal density during the Upper Permian, parallelizing an aridization evidenced by geological parameters. However, within individual taxa no such increase could be detected.

This discrepancy can be explained by a different mixing of cuticles, with more countable cuticles of taxa with generally higher stomatal densities (e.g. Ortiseia) in the upper part of the profile investigated. All in all our data suggest that taphonomic distortions (fragmentation of individual leaves, mixing of material from different micro-habitats, mixing of different taxa) of the data have to be taken into account, when stomatal densities and -indices of fossil plants, with no really close nearest living relatives (i.e. same genus or even species), are used as palaeoatmospheric proxies, especially when only a small number of cuticles is available for investigation.

REFERENCES

MCELWAIN, J., CHALONER, W.G., 1996. The fossil cuticle as a skeletal record of environmental change. Palaios, 11, 376-388.

McELWAIN, J.C., MAYLE, F.E., BEERLING, D.J., 2002. Stomatal evidence for a decline in atmospheric CO₂ concentration during the Younger Dryas stadial: a comparison with Antarctic ice core records. Journal of Quaternary Science, 17, 21-29.

RETALLACK, G.J., 2001. A 300-million-year record of atmospheric carbon dioxide from fossil plant cuticles. Nature, 411, 287-290.

WOODWARD, F.I., 1987. Stomatal numbers are sensitive to increases in CO₂ from pre-industrial levels. Nature, 327, 617-618.

Palaeogeographic affinity of the Permian Jambi flora from Sumatra

I. M. van Waveren ¹, J. H. A. van Konijnenburg-van Cittert ¹, and R. H. Wagner ²
1 Naturalis, National Museum of Natural History, P.O. Box 9517, 2300 RA Leiden, The Netherlands
2 Jardín Botanico de Córdoba, Avenida de Linneo, s/n 14004 Córdoba, Spain

At Naturalis we are revising the main genera from the Jambi flora (Jongmans and Gothan, 1936) from the Lower Permian of Sumatra (Indonesia). This revision will enable us to compare the Jambi flora with more recent work on Permian South East Asian floras from Shansi (China), Thailand, Malaysia and from Australia. The comparison will give an indication on the palaeogeographic affinity of the region of Sumatra on which the Jambi flora is located.

Sumatra is composed of two micro-continents (the Sibumasu Terrane and the Indochina Terrane) and the boundary between both terranes runs more or less through the region where the Jambi flora was collected. The Sibumasu Terrane was separated from Gondwana during the Permian. This implies that if the Jambi flora is located on this terrane, it will resemble other Gondwana floras from the same age (e.g. in Australia). On the other hand, if the Jambi flora is located on the Indochina Terrane, which was already separated from Gondwana in the Devonian and joined Cathaysia (including parts of e.g. China and Thailand), then the flora should contain Cathaysian elements.

Comparison of the different palaeofloras in the different regions will be achieved through clustering of the plants according to morphological features. Localities will be clustered according to their plant clusters proportions.

Reconstruction of 400 years land use in St. Odiliënberg based on historical and palynological archives

J. G. M. Waucomont ¹, F. P. M. Bunnik ², T. B. van Hoof ¹, and W. M. Kürschner ¹
1 Dept. of Geobiology, Botanical Palaeoecology, Utrecht University, Budapestlaan 4, 3584 CD Utrecht, The Netherlands
2 Netherlands Inst. for Applied Geoscience, P.O.Box 80015, 3508 TA Utrecht, The Netherlands

Comparative studies of pollen records and historical information on vegetation and land use are an important tool for the validation of palynology as a quantitative proxy for vegetation patterns in the landscape. However, whereas documentation of vegetation cover and land use in historical archives improve for the (sub-)recent, natural archives become less abundant due to intensive anthropogenic impact. An exception has been found in the abandoned oxbow lake of the Roer meander at St. Odiliënberg near Roermond consisting of up to 4 m fluvio-lacustrine siliclastic sediments (mainly gyttja with some sandy intercalation) covering the last 1000 years. A palynological study has been carried out on two cores located in the abandoned channel of the Roer. Samples were taken from the uppermost 150 cm at 5 cm intervals. Forestry is reflected in the pollen record by changes in Pinus sylvestris type (pine plantations for coal mining), and the preceding cultivation of Lupinus luteus, used as a natural fertilizer. Furthermore, changes in the abundance of cereals such as Secale cereale (rye), Avena/Triticum type (oat, wheat), and Fagopyrum esculentum (buckwheat) reflect agricultural activities and crop failure due to climatic events or warfare. Additionally, topographical maps and cereal prices in the provincial archive of Limburg and historical municipal archives at Roermond and St. Odiliënberg have been studied in order to gather information on vegetation and land use of the research area. Examples of events in both the palynological record, as well as in the historical record are the pine plantations (distinct changes of pine cover on topographical maps), and the intensity of anthropogenic land use is reflected in Cerealia pollen, and cereal price fluctuations.