PALEVO 1631-0683 Elsevier S1631-0683(02)00035-0 10.1016/S1631-0683(02)00035-0 Research article Systematic Palaeontology / Paléontologie systématique A palm-tree and Sciadopitys swamp-forest from the Neogene of Bresse (eastern France) Une forêt marécageuse à Sciadopitys et palmiers dans le Néogène bressan. Philippe Marc philippe@cismsun.univ-lyon1.fr a b Méon Henriette b Lambert Georges c Erdei Boglarka d Thevenard Frédéric a b Gomez Bernard a Université Lyon-1, 7, rue Dubois, 69622 Villeurbanne cedex, France UMR 5125 du CNRS, 7, rue Dubois, 69622 Villeurbanne cedex, France Laboratoire de chrono-écologie, université de Franche-Comté, route de Gray, 25000 Besançon, France Palaeobotany, Hungarian Natural History Museum, 1476 Budapest Pf. 222, Hungary 1 4 S1631-0683(00)X0004-8 221 225 © 2002 Académie des sciences / Éditions scientifiques et médicales Elsevier SAS 2002 Académie des sciences / Éditions scientifiques et médicales Elsevier SAS Full (PDF)

During the Tertiary, a thick sedimentary sequence was accumulated in the Bresse Basin, including some lignite that was locally excavated. Although the palynology of this lignite is relatively well known, the fossil plant macroremains were only poorly studied. Well-preserved material from a new Upper Pliocene (Reuverian) outcrop is reported herein. It allows for palaeoecological and biogeographical reinterpretation. The flora, which includes palm-trees and umbrella pine, is interpreted as a ‘Sequoia-Moor’, i.e. an acidophilic swamp-forest type that had never been evidenced from both micro- and macroremains in France.

Le bassin de la Bresse a piégé au cours du Tertiaire une grande quantité de sédiments, dont des lignites, qui ont été exploités. Si la palynologie de ceux-ci est relativement bien connue, les données sur les mégarestes sont rares. La découverte d’un gisement du Pliocène supérieur (Reuvérien) particulièrement favorable permet d’apporter de nouveaux éléments, significatifs d’un point de vue paléoécologique et biogéographique. La flore, incluant des palmiers et des Sciadopitys, caractérise un « Sequoia-Moor », un type de forêt marécageuse qui n’avait jamais été mis en évidence en France conjointement par les micro- et les mégarestes.

 palaeobotany, Late Tertiary, Sciadopitys , France paléobotanique, Tertiaire supérieur, Sciadopitys , France presented Philippe Taquet Introduction

On the western border of the Jura Mountains, the Bresse Basin was active during the Tertiary 〚12〛 and 〚19〛. The related sedimentary succession includes fluvial and lacustrine deposits. During the Late Tertiary, mainly coarse fluvial sediments were deposited. They are surmounted by finer-grained sands and marls that commonly include ‘lignite’ 〚3〛, 〚14〛 and 〚19〛. This term is confusing, however, being sometimes used for isolated woody axes and sometimes for organic-rich levels with diverse plant remains. Both types are encountered in the studied material. These lignitic beds of the Bresse were still exploited shortly after the Second World War. They were extensively studied from a palynological point of view 〚4〛, 〚10〛, 〚15〛 and 〚16〛. The plant macroremains are, however, poorly known, the only data reported being ‘wood of the Taxodioxylon type’ 〚14〛, 〚21〛 and 〚23〛. Another Neogene lignite from the same region Franche-Comté was studied from both palynological and palaeobotanical points of view 〚20〛, but the geological and palaeoecological setting was quite different.

New material was collected in 1999 and 2001 during the construction of the highway A 39, in road-cuttings that occurred near the town of Poligny (Fig. 1). This includes a matrix bearing woody axes, barks, isolated leaves and rootlets, and palynomorphs. The present results deal with both micro- and macroremains.

 Geological setting, material, and methods

The material comes from the vicinity of Colonne (Jura department, eastern France). It was found in road-cuttings along an access road to the highway A 39, at Bersaillin (Fig. 1), and at a rest area named ‘Aire de la Jument-Verte’ (in reference to the famous novelist Marcel Aymé), 2 km northward. At both localities, grey sandy marls cropped out, which include organic beds and a few coarser levels with gravel. One of the organic beds is particularly obvious, circa 50 cm-thick, with numerous trunks and branches, some woody debris being metric in size. The matrix is dark-grey, with a high content of organic matter and microcrystalline pyrite. Detritic sand fraction is low. The depositional environment was an organic swamp with low alluvial input. Plant remains include palynomorphs, leaf compressions without cuticle preserved, bark fragments, woody axes and mat-forming rootlets. Although the deposit is probably not autochthonous, as no rooting systems were observed in the underlying marls, transport was probably short, because the plant debris are unsorted and unrolled, whereas the alluvial input is low.

The stratigraphic assignation of these marls is difficult, because the outcrop is restricted. As it overlays the ‘Sables de Foulenay’ 〚13〛, the level probably belongs to the ‘Marnes de Commenailles’, a lateral equivalent of the ‘Marnes de Foulenay’, according to C. Petit. The ‘Marnes de Commenailles’ are Upper Pliocene (Reuverian, top at ca 2.4 ± 0.2 Ma) in age 〚19〛. This is consistent with our palynological data from the level studied (see below).

Twelve samples of wood were studied. They all originate from axes exceeding 10 cm in diameter. One of them was taken from a trunk 35 cm in diameter. After 12 h in water, samples were boiled for 2 h, before being cut with a razor blade along the three xylological plans. Palynological preparation was performed, following the usual method. A gentle application of H2O2 on the matrix allowed for removing a few cuticle fragments. However, cuticle structures were almost impossible to observe due to the presence of a dark content that Schulze reagent (nitric acid plus potassium chlorate) was unsuitable to clear. This precluded from studying cuticle structure.

 Results Fossil woods

Almost all the samples have the same structure. It is a homoxylous wood, with narrow but clearly marked growth rings. In transversal view, the tracheid section is inconstant. Rays are mainly uniseriate, less than 2% being biseriate on one or two levels at most. Rays are numerous, narrow, and their height is rather low (ca 7.5 cells). The radial pitting of the tracheids is abietinean, uniseriate, rarely (less than 4%) biseriate and, then, with opposite pairs. Pits are rather large, covering the whole tracheid radial wall. Cross-fields have only one pore, rarely two. This pore is a large podocarpoid to taxodioid oculipore, occupying the whole field, in the early wood, whereas a single cupressoid oculipore is observed in the late wood cross-fields. The ray-cell walls are smooth and unpitted, except for cross-fields. Axial parenchyma has not been observed. The wood underwent fungal attacks, displaying the pattern of an early stage of brown-rot wood decay 〚25〛. This wood is clearly related to the extant genus Sciadopitys 〚18〛. From a nomenclatural point of view, the morphogenus Sciadopityoxylon Schmalhausen cannot be used, as there are some uncertainties about its type specimen. Thus, this type from the Jurassic of Mangyschlack Peninsula (former USSR) may be related to Xenoxylon Gothan (pers. obs.). To stay out of this taxonomical/nomenclatural imbroglio, the wood studied herein is assigned to the morphogenus Podocarpoxylon Gothan. A reliable specific name cannot be given without exhaustive reappraisal of the fossil wood complex of Podocarpoxylon/Taxodioxylon species, described up to now from the European Tertiary. The present data establish the first record of Sciadopitys type fossil wood in France. The fossil woods mentioned up to now from the Bressan Tertiary deposits all belonged to the type Taxodioxylon Hartig 〚14〛, 〚21〛 and 〚23〛.

The single specimen of fossil wood displaying a different anatomy belongs to this latter type, featuring common biseriate radial pitting and Sanio’s rims, taxodioid oculipores, as well as axial parenchyma with smooth transverse walls.

 Other plant macroremains

A fragment of a palm leaflet was recognised among the leaf remains. A few ‘needles’, resembling pedicellate conifer leaves, are also present. In some of them, the cuticles are preserved. However, any description or identification of these cuticles cannot be given herein because of methodological problems. Rootlets are outstandingly abundant all through the fossiliferous level, clearly indicating parautochtony.

 Palynomorphs

The palynological results are given in Table 1. Spores are quite abundant, with a significant contribution of the hygrophytic and acidophilic taxon Osmunda (king-fern), especially in the ‘Aire de la Jument-Verte’. At both sites, spores of Sphagnum, Anthoceroteae and Polypodiaceae were also recognised. Pollen grains of Sciadopytis were absent, whereas the conifers as a whole were diversified. The Abietaceae are dominant and diverse (Pinus, Picea, Abies, Cathaya, Cedrus). Taxodiaceae, Alnus, Arecaceae (palm-trees) and Ericaceae (heathers) are present. The Poaceae (grasses) are poorly represented. The palynological spectrum clearly indicates that the level studied is older than the Pretiglian, which always displays significant amount of Artemisia and Poaceae in the Bresse area 〚16〛.

 Discussion

In the Bresse region, the occurrence of both Sciadopitys and Tsuga pollen grains is frequent in the Pliocene. Such an association agrees with the age usually proposed for the « Marnes de Commenailles » 〚19〛. Otherwise, the woods assert that Sciadopitys was well represented at this time-period. The palynological analysis, however, failed in recognising any Sciadopitys, although this pollen is typical 〚24〛. Thus, the taphonomy of wood and pollen is very different 〚7〛. Moreover, the fact that Sciadopytis pollens are usually in low percentages in palynological spectra of the Neogene of Bresse 〚4〛, 〚10〛 and 〚15〛 may be a result of entomophily and of the related low pollen production.

The single extant species of the genus Sciadopitys, S. verticillata, is an endemic of Southern Japanese mountain areas. It grows within old-growth forest with Chamaecyparis obtusa, although rather heliophilous, on drained sandy or loamy soils (ph = 3.7 to 6.0), and under temperate maritime climate with high rainfall and moisture 〚2〛, 〚9〛 and 〚27〛. In the Tertiary, Sciadopytis is, however, represented all over the Palaearctic 〚6〛, and mainly by its pollen grains. Such a type of wood was found, for example, from the Pannonian of Hungary 〚8〛 and the Oligo-Miocene of Düren in Germany 〚11〛.

Outstanding palaeoecological studies were performed in Westfalia and central Europe 〚17〛 and 〚22〛, which indicate that Sciadopitys trees grew in the ‘Sequoia-Moor’. This raised bog swamp-forest was encountered around lakes, slightly over the water-table level. The forest was mainly composed of palm-trees (Arecaceae), true pines (Pinus), bald cypress (Taxodium) and umbrella-pine (Sciadopytis). Our results assert that a forest very similar to the German model occurred in northeastern France during the Pliocene. It is the first time that such a palaeoenvironment is demonstrated for France, with clear evidence of parautochtony from both micro- and macroremains. The occurrence of king-fern, heather and alder agrees with the inference of an acidic and hydromorphic soil. From a biogeographical point of view, although Sciadopitys is already known as pollen grains farther south up to Italy 〚1〛, 〚5〛 and 〚6〛, our data extend significantly southwesternward the distribution of Pliocene Sciadopitys swamp-forest.

As extant umbrella-pines do not grow actually in mires, uniformitarism suggests a discrepancy between Sciadopitys and the rest of the palaeoflora. Such a discrepancy was already noted by Figueiral et al. 〚7〛, and it is obvious that, although this taxon always favoured mild climates 〚26〛, Sciadopitys experienced a shift of its ecological requirements since the Tertiary. Palm-trees are too often and erroneously used as indicators of tropical climate, and their presence in the assemblage is consistent with occasional frosts. The clearly allochtonous remains represented in the palynological spectra by Cedrus and Cathaya pollen grains, genera that nowadays grow under similar climate, although with different ecological requirements, corroborate these paleoclimatological inferences. The plant remains, as a whole, point out a wet mild-temperate climate, warmer than the present.

 Acknowledgements

The authors thank sincerely Geneviève Farjanel, Christophe Petit and Paul Roiron for their much appreciated help and suggestions.

 Version abrégée

Relativement bien connu d’un point de vue palynologique 〚4〛, 〚10〛, 〚15〛 and 〚16〛 le Tertiaire bressan, quoique lignitifère 〚3〛, 〚12〛, 〚14〛 and 〚19〛, n’a jamais été étudié du point de vue des mégarestes végétaux, et seuls quelques « Taxodioxylon » ont été signalés 〚14〛, 〚21〛 and 〚23〛. D’autres lignites néogènes franc-comtois ont fait l’objet d’études paléobotaniques et palynologiques conjointes 〚20〛, mais dans un contexte différent.

À la faveur des travaux de l’autoroute A 39, un niveau marno-sableux et ligniteux a été localisé à proximité de Colonne, Jura (Fig. 1). Il appartient probablement aux marnes de Commenailles, une formation d’âge Pliocène supérieur (Reuvérien, toit à environ 2,4 ± 0,2 Ma 〚13〛 and 〚19〛. Les prélèvements ont été effectués à deux endroits où affleure ce niveau, l’échangeur de Bersaillin et l’aire de la « Jument-Verte », dans une couche argileuse riche en matière organique et pyrite microcristalline, pétrie de débris végétaux divers (écorces, axes ligneux, bois, débris foliaires, feutrages racinaires) parautochtones, certains métriques. Les bois montrent un début d’attaque fongique de type pourriture brune 〚25〛.

Les douze fragments ligneux étudiés ont tous la même anatomie, similaire à celle du Sciadopitys actuel 〚18〛, sauf un qui est rapporté à Taxodioxylon Hartig. Un fragment de folioles de palmier ainsi que quelques aiguilles pédicellées de Conifères, dépourvues de cuticules, ont été reconnus. Les spores sont abondantes (Tableau 1), avec une part importante d’Osmonde, notamment pour la « Jument-Verte ». Les pollens sont dominés par les Abiétacées (Pinus, Picea, Abies, Cedrus, Cathaya et Tsuga). Parmi les Angiospermes sont représentés, entre autres, Alnus, Quercus, les Ericacées et Palmiers, ainsi que des Graminées.

Sciadopitys n’a pas été trouvé, quoique son pollen soit facilement reconnaissable 〚24〛. Cependant, ce pollen est toujours faiblement représenté 〚15〛 et les taphonomies du bois et du pollen sont différentes 〚7〛. Aujourd’hui, Sciadopitys ne se rencontre plus qu’au Japon, en climat maritime tempéré humide sur sols sableux à limoneux 〚2〛, 〚9〛 and 〚27〛, mais, au Tertiaire, il était répandu dans le Paléarctique 〚6〛, 〚8〛 and 〚11〛.

Des études paléoécologiques en Allemagne 〚17〛 and 〚22〛 ont démontré que Sciadopitys occupait des tourbières bombées couvertes d’un boisement clair à Cyprès chauve (Taxodium), Palmiers (Arécacées) et pins (Pinus), une formation habituellement dénommée « Sequoia-Moor » 〚17〛. Quoique le pollen de Sciadopitys se rencontre isolément plus au sud, jusqu’en Italie 〚1〛, 〚5〛 and 〚6〛, le « Sequoia-Moor » est mis en évidence pour la première fois en France, ce qui étend largement vers le sud-ouest la distribution connue de cette phytocénose si particulière.

Dans la mesure où les exigences pédologiques du Sciadopitys actuel sont différentes de celle du Sciadopitys du Tertiaire européen, on peut affirmer que ce genre a changé d’écologie, même s’il reste un indicateur de climat doux 〚26〛. Quoique la présence de palmiers ne permette pas d’écarter la possibilité de gels occasionnels, le climat du Reuvérien, au moment de ce dépôt, était probablement plus maritime et un peu plus chaud qu’aujourd’hui. La présence d’éléments clairement allochtones, comme Cedrus et Cathaya, qui vivent sous des climats semblables, mais n’appartiennent pas du tout au même type de forêt, corroborent ces indications paléoclimatiques.

 P. Ambrosetti, G. Basilici, A. Ciangherotti, G. Codipietro, Corona, E. Esu D., Girotti O., Lo Monaco A., Meneghini M., Paganelli A., Romagnoli, M., The Dunarobba fossil forest (Terni, Umbria, central Italy): lithostratigraphic, sedimentologic, palynologic, dendrochronologic and palaeomalacologic characteristics, Palaeontol. Ichnol. 1 (1997) (http://tetide.geo.uniroma1.it/sciterra/pubdip/publicaz/05_paich.html) Dallimore W. Albert B.J. Harrison S.G. A handbook of Coniferae and Ginkgoaceae 4th 1967 St-Martin's Press New York Delafond P. Depéret C. Les terrains tertiaires de la Bresse et leurs gîtes de lignites et de minerais de fer 1894 Imprimerie nationale Paris Farjanel G. La flore et le climat du Néogène et du Pléistocène de Bresse (France) d'après l'analyse pollinique. Implications chronostratigraphiques 1985 thèse, université de Bourgogne Dijon Fauquette S. Quezel P. Guiot J. Suc J.-P. Signification bioclimatique de taxons-guides du Pliocène méditerranéen Geobios 31 1998 151–169 Ferguson D.K. On the phytogeography of Coniferales in the European Cenozoic Palaeogeogr. Palaeoclimatol. Palaeoecol. 3 1967 73–110 Figueiral I. Mosbruger V. Rowe N.P. Ashraf A.R. Utescher T. Jones T.P. The Miocene peat-forming vegetation of northwestern Germany: an analysis of wood remains and comparisons with previous palynological interpretations Rev. Paleobot. Palynol. 104 1999 239–266 Greguss P. Fossil gymnosperm woods in Hungary from the Permian to the Pliocene 1967 Akad. Kiado Budapest Ishikawa S. Watanabe N. An ecological study on the Sciadopitys verticillata forest and other natural forests of Mount Irazu, southern Shikoku, Japan Mem. Fac. Sci. Kochi Univ., Ser. D (Biol.) 7 1986 63–75 Jan du Chêne R. Étude palynologique du Néogène et du Pléistocène inférieur de Bresse Bull. BRGM 4 1974 209–235 Jurasky K. Paläobotanische Braunkohlen-Studien I.I.I. Ein neuer Fund von Sciadopitys (Koniferen) in der Braunkohle (Sciadopityoxylon wettsteini n.sp.) Senckenb. Lethaea 10 1928 255–264 Lefavrais-Raymond A. Contribution à l'étude géologique de la Bresse d'après les sondages profonds 1962 Mém. BRGM n° 16, BRGM Orléans Liénhardt G. Rat P. Groupe d'étude « Bresse », Le groupe d'étude « Bresse » et les tableaux stratigraphiques proposés pour le Miocène, le Pliocène et le Villafranchien du remplissage bressan C. R. somm. Soc. géol. France 4 1974 95–100 Mazenot G. Les lignites bressans 1945 Imprimerie nationale Paris Méon H. Palynologie des formations miocènes supérieures et pliocènes du bassin du Rhône (France) 1970 Doc. Lab. Géol. Fac. Sci. Lyon n° 38 université Lyon-1 Méon H. Bachiri N. Puisségur J.-J. Analyses sporopolliniques du sondage de Beaune (NW de la Bresse, France). Stratigraphie et restitution climatique Rev. Micropal. 32 1989 277–292 Minnigerode C. Klein-Reesink J. Das Dörentruper Braunkohleflöz als Zeuge eines fossiles Moores Geol. u. Paläont. in Westfalen 2 1984 Peirce A.S. Anatomy of the xylem of Sciadopitys Am. J. Bot. 22 1935 895–902 Petit C. Un bassin d'avant-pays de type pelliculaire, la Bresse au Plio-Pléistocène 1993 thèse, université de Bourgogne Dijon Philippe M. Contini D. Méon H. Les lignites « purbeckiennes » de Longemaison (Doubs) sont en fait néogènes ; étude paléobotanique C. R. Acad. Sci. Paris, Ser. II 314 1992 359–364 Philippe M. Bois néogènes de Taxodiacées à Saint-Étienne-du-Bois (Ain) 1994 rapport d'expertise inédit Schneider W. Floral succession in Miocene swamps and bogs of central Europe Z. Geol. Wiss. 20 1992 555–570 Touton J.B. La zone inondable des bords de Saône dans la Bresse de l'Ain Bull. Soc. Nat. Arch. Ain 72 1958 101–145 Ueno J. Morphology of pollen of Metasequoia, Sciadopitys and Taiwania J. Inst. Polytech. Osaka City Univ., Ser. D 2 1951 22–26 Wilcox W.W. Comparative morphology of early stages of brown-rot wood decay IAWA J. 14 1993 127–138 Willard D.A. Palynological record from the North Atlantic Region at 3 Ma: a vegetational response to a period of global warmth, US Geol. Surv. open-file report 94-23 (http://geochange.er.usgs.gov/pub/PRISM/OFR_94-23/07_Willard.html) 1994 Yamamoto S.I. Seedling recruitment of Chamaecyparis obtusa and Sciadopitys verticillata in different microenvironment in an old-growth Sciadopitys forest Bot. Mag. Tokyo 101 1988 67–71

Location of ‘Péage de Bersaillin’ and ‘Aire de la Jument-Verte’ along the highway A 39, Franche-Comté, northeastern France.

Situation des localités, péage de Bersaillin et aire de la Jument-Verte, le long de l’A 39, Franche-Comté, Nord-Est de la France.

Percentages of the different palynomorph types encountered.

Pourcentages des différents palynomorphes rencontrés.

Bersaillin Jument-Verte Spores 33 23 of which Osmunda 1.5 13 Pollens 65 69 Abietaceae 56.5 61 of which Pinus 12 13 Picea 1 4.5 Abies 6.5 1 Tsuga 0 0.5 Cathaya 2.5 0 Taxodiaceae 0.5 1 Total Gymnosperms 57 62 Arecaceae 0 1 Alnus 1.5 4 Ericaceae 0 1 Quercus 0.5 0.5 Herbaceous Angiosperms 1 0.5 Total Angiosperms 8 7 Undetermined 2 8 Number counted 232 327