Fossil wood from the lower Miocene of Myanmar (Natma Formation): palaeoenvironmental and biogeographic implications

2022  44  28 geodiversitas Geodiversitas est une revue en flux continu publiée par les Publications scientifiques du Muséum, Paris Geodiversitas is a fast track journal published by the Museum Science Press, Paris Les Publications scientifiques du Muséum publient aussi / The Museum Science Press also publish: Adansonia, Zoosystema, Anthropozoologica, European Journal of Taxonomy, Naturae, Cryptogamie sous-sections Algologie, Bryologie, Mycologie, Comptes Rendus Palevol Diffusion – Publications scientifiques Muséum national d’Histoire naturelle CP 41 – 57 rue Cuvier F-75231 Paris cedex 05 (France) Tél. : 33 (0)1 40 79 48 05 / Fax : 33 (0)1 40 79 38 40 [email protected] / http://sciencepress.mnhn.fr © Publications scientifiques du Muséum national d’Histoire naturelle, Paris, 2022 ISSN (imprimé / print) : 1280-9659/ ISSN (électronique / electronic) : 1638-9395 Directeur De la publication / Publication director : Bruno David, Président du Muséum national d’Histoire naturelle réDacteur en chef / editor-in-chief : Didier Merle assistant De réDaction / assistant editor : Emmanuel Côtez ([email protected]) Mise en page / Page layout : Emmanuel Côtez coMité scientifique / scientific board : Christine Argot (Muséum national d’Histoire naturelle, Paris) Beatrix Azanza (Museo Nacional de Ciencias Naturales, Madrid) Raymond L. Bernor (Howard University, Washington DC) Henning Blom (Uppsala University) Jean Broutin (Sorbonne Université, Paris, retraité) Gaël Clément (Muséum national d’Histoire naturelle, Paris) Ted Daeschler (Academy of Natural Sciences, Philadelphie) Bruno David (Muséum national d’Histoire naturelle, Paris) Gregory D. Edgecombe (The Natural History Museum, Londres) Ursula Göhlich (Natural History Museum Vienna) Jin Meng (American Museum of Natural History, New York) Brigitte Meyer-Berthaud (CIRAD, Montpellier) Zhu Min (Chinese Academy of Sciences, Pékin) Isabelle Rouget (Muséum national d’Histoire naturelle, Paris) Sevket Sen (Muséum national d’Histoire naturelle, Paris, retraité) Stanislav Štamberg (Museum of Eastern Bohemia, Hradec Králové) Paul Taylor (The Natural History Museum, Londres, retraité) couverture / cover : Réalisée à partir des Figures de l’article/Made from the Figures of the article. Geodiversitas est indexé dans / Geodiversitas is indexed in: – Science Citation Index Expanded (SciSearch®) – ISI Alerting Services® – Current Contents® / Physical, Chemical, and Earth Sciences® – Scopus® Geodiversitas est distribué en version électronique par / Geodiversitas is distributed electronically by: – BioOne® (http://www.bioone.org) Les articles ainsi que les nouveautés nomenclaturales publiés dans Geodiversitas sont référencés par / Articles and nomenclatural novelties published in Geodiversitas are referenced by: – ZooBank® (http://zoobank.org) 853 GEODIVERSITAS • 2022 • 44 (28) © Publications scientifiques du Muséum national d’Histoire naturelle, Paris. www.geodiversitas.com Nicolas GENTIS Muséum national d’Histoire naturelle, Centre de Recherche en Paléontologie – Paris (CR2P), CNRS/MNHN/Sorbonne Université, CP 38, 57 rue Cuvier, F-75231 Paris cedex 05 (France) [email protected] (corresponding author) Alexis LICHT Aix Marseille Université, CNRS, IRD, INRAE, Collège de France, CEREGE, Europole Méditerranéen de l’Arbois, BP 80, F-13545 Aix-en-Provence cedex 4 (France) [email protected] Anaïs BOURA Dario DE FRANCESCHI Muséum national d’Histoire naturelle, Centre de Recherche en Paléontologie – Paris (CR2P), CNRS/MNHN/Sorbonne Université, CP 38, 57 rue Cuvier, F-75231 Paris cedex 05 (France) [email protected] [email protected] ZAW WIN Geology Department, Shwe Bo University, Sagaing Region (Myanmar) [email protected] DAY WA AUNG Geology Department, University of Yangon, Pyay Rd, Yangon (Myanmar) [email protected] Guillaume DUPONT-NIVET Géosciences Rennes, UMR CNRS 6118, Univ. Rennes, F-35042 Rennes cedex (France) and Potsdam University, Institute of Earth and Environmental Science, 14476 Potsdam (Germany) and Key Laboratory of Orogenic Belts and Crustal Evolution, Ministry of Education, Beijing (China) [email protected] Submitted on 28 January 2021 | accepted on 23 February 2022 | published on 29 September 2022 Fossil wood from the lower Miocene of Myanmar (Natma Formation): palaeoenvironmental and biogeographic implications urn:lsid:zoobank.org:pub:2611B0BC-F569-4135-A09C-6E527C2565A4 Gentis N., Licht A., Boura A., De Franceschi D., Zaw Win, Day Wa Aung & Dupont-Nivet G. 2022. — Fossil wood from the lower Miocene of Myanmar (Natma Formation): palaeoenvironmental and biogeographic implications. Geodiversitas 44 (28): 853-909. https://doi.org/10.5252/geodiversitas2022v44a28. http://geodiversitas.com/44/28 ABSTRACT The palaeobotanical record of Myanmar (Burma) is poorly documented despite its importance for understanding the evolution of Asian monsoonal ecosystems through time. Here, we describe 20 taxa of fossil wood from 30 silicified specimens collected in the upper lower to lowermost middle Miocene Natma Formation, central Myanmar. These fossils share affinities with modern Fabaceae, Dipterocarpaceae, Burseraceae, Moraceae and Cupressaceae. They include a great variety of fossil dipterocarps (eightspecies) as found in today’s Southeast Asian rainforests. The nearest living relatives of this asKEY WORDS Petrified wood, Fabaceae, Cenozoic, Southeast Asia, dipterocarp, monsoon. 854 GEODIVERSITAS • 2022 • 44 (28) Gentis N. et al. INTRODUCTION Myanmar is located at the eastern edge of the Bengal Bay and today lies in the Asian monsoonal domain (Fig. 1A); its Palaeobotanical record has been shown to be critical to reconstruct the evolution of monsoons in deep time and the biogeographic connections between South and Southeast Asia (Licht et al. 2015; Huang et al. 2020). Most of Myanmar is today characterized by intense seasonal rainfall (> 2m) over its coasts and highlands, whereas the central part of the country, lying in the rain-shadow of the Indo-Burman Ranges (IBR), receives less than 600mm of annual precipitation (Aung et al. 2017). Burmese ecosystems are dominantly considered as tropical to subtropical and are all strongly influenced by the seasonal monsoonal rainfall (Kress et al. 2003; Ashton 2014; Beck et al. 2018). While tidal and swampy forests are present with some mangroves along the coasts (Rhizophora-Sonneratia associations), and temperate forests occur at high altitude (Kress et al. 2003; ASFN 2016), the majority of forested areas of Myanmar are covered by mixed deciduous forests and temperate (semi-)evergreen forests of moderate altitude typical of Asian monsoonal climates (FAO 2007). Seasonally dry lowlands are composed of acacia woodlands, mixed deciduous forests (Tectona-Xylia-Terminalia associations), and dry subtropical forests dominated by few species of dipterocarps (Sal-type forests) or Tectona-Acacia associations, which are common in the driest part of monsoonal South Asia; subtropical evergreen and semievergreen forests dominated by a wider variety of dipterocarps are present in wetter areas. Fossil flora studies investigating the persistence of these seasonally dry ecosystems have consisted so far in few pollen studies (Morley 2000, 2018; Huang et al. 2020, 2021) and studies on fossil wood specimens (Prakash 1965a, b, d, 1973; Prakash& Bande 1980; Du 1988a; Gottwald 1994; Privé-Gill et al. 2004; Licht et al. 2014, 2015). The age of most previously studied sites is commonly poorly constrained, and it remains unclear when the modern seasonally dry ecosystems of central Myanmar, typical of the regional monsoonal climate, were set up. Studies of the Burmese palaeobotanical record with a clear stratigraphic context exclusively come from the late middle Eocene of central Myanmar, namely from the Pondaung Formation (Privé-Gill et al. 2004; Licht et al. 2014, 2015) and the overlying Yaw Formation (Huang et al. 2020, 2021). The Pondaung Formation has yielded numerous fossil wood specimens associated with three types of forest ecotones: dry dipterocarp forests (with fossil specimens related to modern Shorea Roxb. ex C.F. Gaertn., Pentacme A. DC., Amesiodendron Hu, Schima Reinw. ex Blume, Terminalia L., Bombax L.); riparian and open seasonal wetlands typical of the Terai ecosystem of South Asia (with fossil specimens related to modern Terminalia, Bombax, Bauhinia L., Acrocarpus Wight ex Arn., Ficus L., Pycnarrhena Miers ex Hook.f.& Thomson, Saraca L., Heritiera Aiton), and tidal / littoral forests (with fossil Pycnarrhena, Saraca, Heritiera, Sonneratia L.f., Gluta L., Cynometra L.). These assemblages follow as it is seen in the actual Bay of Bengal (Licht et al. 2015). Most of the species identified in the Pondaung Formation grow today in climates with a well-marked seasonality. In addition, pollen from the Yaw Formation highlights a high abundance of palms that remain absent from the fossil wood record (Huang et al. MOTS CLÉS Bois fossile, Fabaceae, Cénozoïque, Asie du Sud-Est, diptérocarpacée, mousson. semblage reflect different ecotones of seasonal forests with coastal, mixed to dry deciduous, and wet evergreen species. This reconstruction implies a wet, warm, and monsoonal climate in Myanmar during the late early Miocene. The presence of fossil dipterocarp species typical of wet evergreen forests contrasts with Burmese Eocene dry dipterocarp assemblages and indicates wetter conditions during the Miocene. Our reconstructions support a long-term change from seasonal to everwet ecosystems for dipterocarp trees. RÉSUMÉ Bois fossiles du Miocène inférieur du Myanmar (formation de Natma) : implications paléoenvironnementales et biogéographiques. Le registre botanique fossile du Myanmar (Birmanie) est faiblement documenté malgré son importance pour comprendre l’évolution des écosystèmes de mousson au cours des temps géologiques. Nous décrivons ici 20 taxa de bois fossile issus de 30 specimens silicifiés de la Formation de Natma (Myanmar central) datant du Miocène inférieur tardif au début du Miocène moyen. Ces fossiles sont affiliés aux familles modernes des Fabaceae, Dipterocarpaceae, Burseraceae, Moraceae et Cupressaceae. Nos spécimens incluent une grande diversité de diptérocarpacées fossiles (huit espèces) retrouvée aujourd’hui dans les forêts tropicales d’Asie du Sud-Est. Les analogues modernes de ces espèces fossiles révèlent différents écotones de forêts saisonnières avec des espèces côtières, de forêts mixtes ou décidues sèches, et de forêts humides sempervirentes. Cette reconstruction implique un climat chaud et humide de mousson au Myanmar durant le Miocène inférieur tardif. La présence d’espèces fossiles de diptérocarpacées typiques de forêts humides sempervirentes contraste avec les assemblages forestiers birmans de l’Eocène, dominés par des forêts sèches à diptérocarpacées, et indiquent des conditions plus humides. Ces reconstructions suggèrent un changement à long-terme des diptérocarpacées d’environnements saisonniers vers des environnements plus humides. 855 Fossil wood from Natma Formation, Myanmar (lower Miocene) GEODIVERSITAS • 2022 • 44 (28) 2020). The late Eocene palynoflora of the same formation reveals a sequence of three ecosystems similar to the one of the Pondaung Formation with seasonnaly dry forests; swamp, gallery and evergreen forests; and tidal / littoral forests (Huang et al. 2021). All the ecological requirements of the species were used as pieces of evidence for monsoonal rainfall and support palaeontological and isotopic findings highlighting intense Eocene seasonality (Jaeger et al. 2004; Licht et al. 2014). The Burmese floral diversity during the Neogene is less wellknown; published fossils with only cryptic age and location data but attributed to this period have been related to modern Dipterocarpus C.F. Gaertn., Shorea, Sterculia L., Gluta, Acacia Mill.., Afzelia Sm., Intsia Thouars, Cynometra, Cassia L., Lagerstroemia L. (Prakash 1973), Terminalia (Chowdhury& Tandon 1964; Mädel-Angeliewa& Müller-Stoll 1973), palms (Sahni 1964), Swintonia Griff., Albizia Durazz., Careya Roxb., Cynometra, Araucaria-Agathis (Prakash& Bande 1980), Diospyros L., Saraca, Grewia L., Acrocarpus (Gottwald 1994). It remains unclear if these specimens were found at the same sites or in the same geological units and it is thus impossible to reconstruct Neogene floral assemblages based on these incomplete data. Documenting the development and expansion of seasonally dry ecosystems in central Myanmar during the Miocene provides direct insights into the long-term evolution of monsoonal activity and of the rainshadow effect of the IndoBurman Ranges. It has been proposed that monsoonal intensity significantly increased during the early middle Miocene and decreased in the late Miocene (Clift et al. 2008); the uplift chronology of the Indo-Burman Ranges is less documented. Sedimentological and low-temperature thermochronology data indicate that uplift began between the late middle Eocene and the late Oligocene (Licht et al. 2019; Najman et al. 2020), but it remains unclear when high topography (>2000m) was acquired, and rain-shadow effects became significant. This paper is the first to examine fossil wood specimens from the upper lower to lowermost middle Miocene Natma Formation, central Myanmar. We describe twenty fossil wood taxa, we identify eighteen of them from a new collection and reconstruct Miocene forested ecotones based on their nearest living relatives. We then compare the Natma Formation ecosystems with late middle Eocene and modern ecosystems of central Myanmar to document the regional landscape evolution. 1000 km 30°N 100°80°E 28°N 24° 20° 16° 92°E 96° 10 20 30 0 40 50 60 (Ma) 70 Myanmar borders Main Fault 10° Convergence zone Sub-Basin Central Myanmar Basins Assam and Indian Foreland Basins Indo-Burman Ranges Himalayan Ranges Lhasa Terrane rocks Sibumasu Terrane rocks Songpan Ganze & Yangtze complexes Cretaceous-Paleogene Asian Volcanic Arc Triassic metamorphics and Cretaceous ophiolites 200 km Mb Shan Plateau Tibet Himalayas Andaman Sea Bengal Bay ITSZ Assam Basin Yunnan Cb WPA Cretaceous-Paleogene Asian Volcanic Arc Tibetan-Himalayan domain: Myanmar domain: Mogok-Mandalay Belt and other metamorphic belts: Paleozoic-Jurassic Granitoids, Metamorphics, Jurassic Ophiolites Pondaung Fm Paunggyi Fm Letkat Fm Yaw Fm Tabyin Fm Tilin Fm Laungshe Fm Kabaw Fm Natma Fm Pleisto. Plio. Miocene UpperMid.Lower OligoceneEocene Up. Middle UpperLower Lower Palaeocene Upper Cret. Tonhe Fm Irrawaddy Fm Mingin Gravels Chindwin Basin Shwethamin Fm AB fig. 1 . — A, Simplified map of Myanmar showing the location of the field sites. Location of the fossil site is indicated with a green star (). Abbreviations: Cb, Chindwin Basin; Mb, Minbu Basin; WPA, Wuntho-Popa Arc (volcanic arc of central Myanmar); B, stratigraphy of the Chindwin Basin, after Westerweel et al. (2020). The Natma Formation is indicated with a green star (). 856 GEODIVERSITAS • 2022 • 44 (28) Gentis N. et al. GEOLOGICAL AND PALAEOENVIRONMENTAL CONTEXT East of the Indo-Burman Ranges in central Myanmar, the Chindwin Basin (Fig. 1A) constitutes the northern basin of the Burmese forearc and is filled up with Cenozoic clastic sedimentary rocks. Neogene deposits consist in the fluvial Lektat, Natma, Shwethamin, and Irrawaddy Formations, in stratigraphic order (Fig. 1B; Bender 1983; Licht et al. 2019). The c.1km thick Natma Formation consists of afossiliferous sandstones and pedogenised finer-grained sands and mudstones. The sedimentology of the Natma Formation recalls the one of the earlier Pondaung Formation in the same basin (Licht et al. 2013). Sandstones occur in 10 to 50m wide channel bodies; finer-grained layers display well-marked palaeosol horizons rich in root traces and pedogenic carbonates, which indicate the occurrence of a well-marked dry season at the time of deposition (Retallack 2005; Zamanian et al. 2016). Growth temperatures from Natma pedogenic carbonates, derived from clumped isotope data, fall in the range of modern winter-to-spring temperatures and indicate a dry season during the coldest part of the year, similar to the modern monsoonal regime (Licht et al. 2022). The age of the Natma Formation is constrained by the age of geological units higher and lower in the Chindwin Basin stratigraphy. The Letkat Formation has yielded detrital zircons and apatite ages as young as 20-17 Ma, indicating a lower Miocene age or younger (Wang et al. 2014; Licht et al. 2019; Westerweel et al. 2020). The base of the Irrawaddy Formation, also named Mingin Gravels in the Chindwin Basin, has yielded fossil mammals indicating a late middle Miocene (14-11 Ma) age, coeval to the Chinji fauna of Pakistan (Bender 1983); a similar age for the base of the Irrawaddy Formation has also been found in the nearby Shwebo Basin (Chavasseau et al. 2006). There is c.1.5km of sediment thickness (Shwethamin Formation) between the top of the Natma Formation and the base of the Mingin Gravels, implying some significant time lag between both units. Based on these constraints, the Natma Formation is attributed to the upper lower Miocene to lowermost middle Miocene (Westerweel et al. 2020). At the time of deposition of the Natma Formation, the IndoBurman Ranges were already forming a topographic barrier, of yet unknown elevation (Licht et al. 2019; Najman et al. 2020). The Natma Formation reflects deposition from a palaeo-drainage flowing southward into the Minbu Basin and the Andaman Sea further south, in contrast to the Pondaung Formation, which was opened to the Bengal Bay (Licht et al. 2019). Though no specific study has focused on the provenance of the Natma Formation, Westerweel et al. (2020) propose the Indo-Burman Ranges, the Wuntho Popa Arc in central Myanmar, and the eastern Himalayan Syntaxis further north as potential sources for the underlying unit, the Letkat Formation. MATERIAL AND METHODS: Thirty specimens of fossil wood from the Natma Formation were collected in the Kalewa Township, Sagaing Region, Myanmar, exposed in a wide badland system along the Kalewa-Mawlaik road (23°16’23.1”N, 94°18’25.3”E; see Fig. 1A for location). They are composed exclusively of small fragments from 3to 10cm across, cut from bigger ones. They are completely silicified and were found in-situ in basal lags of sandy channel bodies or on the ground next to these bodies, in a badland where only the Natma Formation is exposed. Sections of transverse, radial and tangential surfaces of the fossil wood specimens were prepared following the standard techniques (Hass& Rowe 1999) at the Muséum national d’Histoire naturelle (MNHN), Paris, France. Fossils were glued with Araldite AY 103 and covered with Araldite 2020. They are described following the IAWA lists of microscopic features for hardwood and softwood identification (IAWA Committee 1989, 2004). A rough estimate of the minimum stem diameter was carried out using the average geometrical intersection of the rays to locate a virtual center, as already done by De Franceschi et al. (2008) and following Dufraisse et al. (2020), rather than using the radius of curvature of the growth limits. Botanical affinities were determined using the InsideWood database (InsideWood 2004-onwards; Wheeler 2011; Wheeler et al. 2020), literature on fossil (Gregory et al. 2009 and references therein), including fossil wood slides from the palaeobotanical collection of Sorbonne Université (specimen no. 5790) described by Vozenin-Serra& Privé-Gill (1989), extant asian and non-asian wood samples and slides from the xylarium of the MNHN Herbarium (P) (specimen number starting by MNHN-P) or from the Cirad Xylarium (specimen number starting by CTFT). Nearest Living Relatives (NLR) ecology was determined using tropical plant databases, floras and forestry guides (referencies are given in concerned paragraphs). All microscopic slides, including at least one transverse, one longitudinal radial and one longitudinal tangential section per specimen are deposited in the collection of the MNHN (Table 2). The remains of the original specimens are dedicated to be sent back to the collections of the Department of Geology at the University of Yangon (Myanmar). RESULTS/TAXONOMY Of the 30 fossil wood samples, we identified 18 species in 11 fossil genera representing five families. Two specimens are still undetermined at the family and genus level. CONIFERS Family CupressaCeae Rich. ex Bartl. Genus Cupressinoxylon (Göppert) Gothan Cupressinoxylon sp. (Fig. 2) M aterial . — MNHN.F.50171 (field number: 17FN15). Estimated minimal diameter: 15-25cm. loCality. — Kalewa Township, Sagaing Region, Myanmar age. — Upper lower to lowermost middle Miocene 857 Fossil wood from Natma Formation, Myanmar (lower Miocene) GEODIVERSITAS • 2022 • 44 (28) DesCription Growth ring boundaries distinct (Fig. 2A), marked by 1-5 lines of flattened tracheids (Fig. 2B). Transition from earlywood to latewood gradual or indistinguishable due to flattening. Tracheids 2100-3200µm long (average 2730µm, n=20), tangential diameter 20-50µm (average 30µm), thin walled with polygonal cross-section. Radial pits 1-seriate, rounded, spaced, 15-24µm in diameter (average: 21µm) (Fig. 2H). Tangential pits present, minute to small (Fig. 2F). Axial parenchyma diffuse and zonate in the vicinity of growth ring boundaries (Fig. 2A, C), both horizontal and vertical walls smooth (Fig. 2E), 17-33µm tangential diameter (average: 25µm). Rays 1to 2-seriate, biseriate parts are present here and there (⩽ 10 %), only composed of parenchyma cells, 3-22 cells high (average: 8) (Fig. 2C, D); ray cells 13-30µm in tangential diameter (average: 30µm), both vertical and horizontal walls smooth (Fig. 2I), indentures not visible. Cross-field pits possibly cupressoid or podocarpoid, maybe both, seemingly 2-4 pits per cross-field (Fig. 2G); 6-15µm in diameter (average: 11µm). Crystals not found. Resin canals absent. AB H C D E G F I fig. 2. — Cupressinoxylon sp. MNHN.F.50171: A, transverse section (Ts), growth limits (white arrows), diffuse parenchyma (grey arrows) and zonate parenchyma (black arrows), large black areas are cells filled with deposit material; B, Ts, growth limit with gradual transition from earlywood to latewood; C, tangential longitudinal section (Tls), 1sometimes 2-seriate rays and parenchyma lines (arrow); D, Tls, detail of 1sometimes 2-seriate rays; E, Tls, smooth longitudinal parenchyma walls (arrows); F, Tls, 1-2-seriate rays and tangential pits (arrows); G, radial longitudinal section (Rls), possibly cupressoid and/or podocarpoid crossfield pits (arrow); H, Rls, uniseriate radial pits; I, Rls, smooth parenchyma cell walls. Scale bars: A, C, 1 mm; B, 200 µm; E-F, H, 100 µm; G, 50 µm; I, 20 µm. 858 GEODIVERSITAS • 2022 • 44 (28) Gentis N. et al. DisCussion This specimen of homoxylate wood is characterized by: 1)tracheids with a polygonal transverse section; 2)uniseriate and spaced tracheid pits; 3)1to 2-seriate and rather short rays; 4)presence of tangential pits; and 5) smooth axial parenchyma and ray cell walls. It is thus attributed to conifers. The presence of axial parenchyma is reported in Podocarpaceae and Cupressaceae (including Taxodiaceae). According to Greguss (1955, 1972), Trivedi& Srivastava (1989) and the IAWA Committee (2004), zonate axial parenchyma is more frequent in Cupressaceae but is not completely diagnostic. It is yet a diagnostic feature of Cupressaceae when it includes dark content in the transverse section and in the transverse end walls of parenchyma cells in longitudinal sections, which is the case for our specimen. Wood species of the cupressoid group are poorly distinguishable from each other based on wood anatomy alone (Phillips 1948; Wheeler& Dillhoff 2009). Kress et al. (2003) record 23 species of Cupressaceae (most of them are non-native) in Myanmar, including Cryptomeria D. Don which shares common traits with our specimen, especially unito biseriate rays and Libocedrus Endl. (Greguss 1972). Within the InsideWood (2004-onward) database, Chamaecyparis formosensis Matsumura and Taxodium distichum (L.) Rich. match partially with our sample (InsideWood input code: 44p 61a 72p 73p 74p 80p 98p 99p 103p 108p). With less restrictive searches (e.g., excluding criteria such as smooth cell wall, cross-field pits cupressoid or podocarpoid and / or ray width from 1to 2-seriate) most taxa still belong to Cupressaceae (Chamaecyparis Spach., Cupressus L., Juniperus L.). However, no clear modern relative among modern Cupressaceae can be found for this fossil. Nonetheless, our specimen shares all the diagnostic features of the fossil genus Cupressinoxylon (cross-field pits oblique and more or less narrow, ray walls mostly smooth, parenchyma often present, resin ducts absent), which are wide enough to include most of Cupressaceae species (Vaudois& Privé 1971). The poorly preserved cell walls and the relatively small size of our specimen makes it difficult to determine the affinity at the species level. Wheeler& Dillhoff (2009) also note that “there are more species of Cupressinoxylon than there are distinctive wood anatomical types within extant Cupressaceae” (Wheeler& Dillhoff 2009: 84). Cupressaceae are mainly mesic-hydric (from balanced hydric conditions to wet conditions) trees in temperate or seasonal climates (Wilf et al. 2009; Pittermann et al. 2012) and all kinds of climate adaptation can be found in this pangeographic family, from moist tolerant trees to very drought-tolerant (Pittermann et al. 2012). Today, few species of Asian and Western-Pacific Cupressaceae grow in tropical areas from lowlands (600m) to tree line: Papuacedrus H.L.Li, some Calocedrus Kurz and Libocedrus (De Laubenfels 1988; Earle 2004-onward). They are found in the tropics above 600m of elevation, but few species occur in lowland tropical areas (<600m): Libocedrus yateensis Guillaumin and Callitris pancheri (Carrière) Byng from New-Caledonia, Callitris macleayana (F.Muell.) F.Muell. from Australia or Glyptostrobus pensilis (Staunton ex D.Don) K.Koch in China and Southeast Asia (Farjon 2005). ANGIOSPERMS Family FabaCeae Lindl. Genus Albizinium Prakash Albizinium eolebbekianum Prakash (Fig. 3) Albizinium eolebbekianum Prakash, 1975: 197, pl.3, figs9, 11, 12. o riginal holotype . — Birbal Sahni Institute of Palaeosciences Museum no. 150/1014. M aterial . — MNHN.F.50172 (field number: 17FN04). Estimated minimal diameter: 30cm. loCality. — Kalewa Township, Sagaing Region, Myanmar age. — Upper lower to lowermost middle Miocene DesCription Wood diffuse-porous. Growth ring boundaries indistinct. Vessels mostly solitary (80%) or grouped by 2, rarely 3, oval, 0-5 per mm² (average: 2) (Fig. 2A, B); tangential diameter 140-260µm (average: 210µm). Tyloses absent. Vessel elements 180-460µm long (average: 290µm). Perforation plates simple, mostly horizontal to slightly oblique (Fig. 3C). Intervessel pits alternate, polygonal in shape and crowded, 6-10µm in size (Fig. 3C). Vessel-ray pits not preserved. Axial parenchyma lozenge-aliform, sometimes confluent, and seemingly marginal bands (Fig. 3A, B); also, maybe some diffuse parenchyma; parenchyma cells 100-150µm (average: 120µm) long in tangential plan, 20-50µm (average: 30µm) wide; crystals present in chambered cells (Fig. 3D). Rays 1to 3-seriate (mainly 2), non-storied, 5-8 rays per tangential mm (average: 6), 150-340µm (average: 220µm) or 5-20 cells high (Fig. 3D), homocellular made of procumbent cells (Fig. 3F). Fibres with poorly preserved walls, septa present in some fibres (Fig. 3D, E), 15-22µm wide (average: 18µm). DisCussion This specimen is characterized by: 1)diffuse-porous wood; 2)exclusively simple perforation plates; 3)aliform parenchyma; 4)the presence of septate fibres; and 5)homocellular and mostly 2to 3-seriate rays. These characters suggest affinities with modern Fabaceae and particularly with the traditional Mimosoideae subfamily (now in the recircumscribed Caesalpinioideae subfamily) (LPWG 2017). In InsideWood (2004-onward), 35% of described specimens of the traditional Mimosoideae have septate fibres. Ogata et al. (2008) also indicate that homocellular rays are mostly found in Mimosoideae. In modern wood of Mimosoideae, the lozengealiform parenchyma combined with the presence of septate fibres and homocellular rays is mainly found in genera Albizia, Inga Mill. and Pithecellobium Mart. (InsideWood 2004-onward). Pithecellobium, however, has higher and wider rays (Awasthi 1979). Albizia and Inga share a close anatomy. According to Evans et al. (2006), Inga has slightly more frequent radial multiples of vessels, as well as confluent parenchyma. A comparison with 859 Fossil wood from Natma Formation, Myanmar (lower Miocene) GEODIVERSITAS • 2022 • 44 (28) specimens and plates available on InsideWood (2004-onward) also reveal higher and thinner rays for Inga (frequently with uniseriate portion, and 2-seriate), with more often uniseriate rays, wheareas most of Albizia species have fusiform and shorter rays. Thus, our specimen shows closer similarities to the genus Albizia, including three specific species: Albizia lebbeck (L.) Benth., regarding its parenchyma arrangement, the presence of septate fibres and non-septate fibres. However, it often has wider rays (up to 5-seriate); A.ferruginea (Guill.& Perr.) Benth. for its lozenge-aliform parenchyma and vessel density, resulting in a few confluences, similar ray size; and A.procera (Roxb.) Benth. for its aliform parenchyma rarely confluent and similar ray size (1-3 seriate, fusiform). Fossil wood specimens resembling Albizia species are designated under the genus Albizinium (Prakash 1975) or described as Albizia (Gregory et al. 2009). The features of the present specimen are compatible with the diagnosis of Albizinium. Although the anatomically close genus Acrocarpoxylon Gottwald has aliform parenchyma and septate fibres, it has more often heterocellular rays (Gottwald 1994). Among Albizinium fossil species close to our fossil (Awasthi 1979; Prakash et al. 1994; Mehrotra et al. 1999), A.eolebbekianum has shorter rays (up to 280µm compared to 340µm in our fossil) and A.pondicherriensis Awasthi has a higher ray density (10-15 / mm as opposed to 5-8 / mm) and more frequent vessels in groups than our fossil. The diagnoses of A.eolebbekianum and A.pondicherriensis are remarkably similar. Our fossil resembles A.eolebbekianum from the Miocene of India for the proportion of solitary vessels and the parenchyma arrangement, whereas it resembles more A.pondicherriensis for ray and vessel size. Our specimen is in fine attributed to A.eolebbekianum given the propensity of A.pondicherrienses to form vessel groups. A DE F BC fig. 3. — Albizinium eolebbekianum Prakash, MNHN.F.50172: A, Ts, vessels and parenchyma arrangement, marginal parenchyma band (arrow); B, Ts, lozengealiform parenchyma, sometimes confluent, fibres radialy aligned; C, Tls, alternate intervessel pits, simple perforation plates; D, Tls, 1-3-seriate rays, septa (black arrows) and crystalliferous parenchyma (white arrow); E, Tls, detail of a septate fibre; F, Rls, homocellular ray. Scale bars: A, 1 mm; B, 500 µm; D, 200 µm; C, F, 100 µm; E, 25 µm. Abbreviations: see Fig. 2. 866 GEODIVERSITAS • 2022 • 44 (28) Gentis N. et al. grow up to 85m tall, with a diameter up to 290cm (Soerianegara& Lemmens 1993). Koompassia malaccensis lives in lowland forests and in freshwater peat and swamp forests, as well as in dry lands up to 150m altitude (rarely up to 800m; Soerianegara& Lemmens 1993; Hou et al. 1996). cf. Koompassioxylon (Fig. 7) M ateriel . — MNHN.F.50182; field number: NAT17-06. Estimated minimal diameter: 11cm. loCality. — Kalewa Township, Sagaing Region, Myanmar. age. — Upper lower to lowermost middle Miocene. DesCription Wood diffuse-porous. Growth rings indistinct, but possible marginal bands. Vessels mostly solitary (80 %) or grouped by 2 or 3 (Fig. 7A, C), oval, 4-9 per mm² (average: 7); tangential diameter 140-250µm (average: 180µm; mesured on less compressed vessels but still underestimated). The radial diameter (71-350µm, average: 225µm) assumes a large tangential diameter. Tylose absent or rarely present (Fig. 7C). Vessel ele - ments 220-460µm long (average: 300µm). Perforation plates simple. Intervessel pits alternate, bordered, 6-13µm (average: 10µm) in diameter, non-vestured (Fig. 7G). Vessel-ray pits rather similar in shape and size to intervessel pits (Fig. 7I), 4-11µm in diameter (average: 7.3µm). Parenchyma widely aliform and mostly confluent laterally or in diagonal, forming anastomosed shapes, no proper tangential bands (Fig. 7A-C); also, maybe in marginal or seemingly marginal thin bands merging with the paratracheal parenchyma; parenchyma cells 50-130µm long (average: 85µm), 10-30µm wide (average: 20µm); 3-8 (or more ?) cells per parenchyma strands (Fig. 7F); abundant crystals in chambered parenchyma cells (Fig. 7H, J), especially in the margin ones. Rays 1to 3-seriate (mostly 2-seriate, very few uniseriate) (Fig. 7E, F), irregularly storied (Fig. 7D) resulting in ripple marks visible with the naked eye, 8-16 rays per mm (average: 11), 150-870µm (average: 400µm) or 6-27 cells high (average: 16-17 cells), occasional interconnections of rays (Fig. 7E) with alternating uniseriate and 2-3-seriate portions probably resulting from end-to-end fusions, thereby some rays are up to 30 cells high, weakly heterocellular to heterocellular to with mostly one to sometimes 2-3 rows of upright or square cells at both ends (Fig. 7H) or at least larger procumbent cells, possibly very rarely with 1-2 crystals in marginal cells; a tendency to have biseriate portions as wide as multiseriate ones, but not to be considered as a distinctive feature. Fibres commonly thin-tothick walled (lumina 0.66 times the double wall thickness) to occasionally very thick-walled, non-septate, 12-24µm (average: 17µm) wide. DisCussion This wood is characterized by: 1)diffuse-porous; 2)aliform parenchyma; 3)crystals in parenchyma cells; 3)mostly 2-seriate rays; 4)non-septate fibres; 5)vessel-ray pits similar to intervessel pits; and 6)simple perforation plates; these features are typical of modern Fabaceae (Metcalfe& Chalk 1950; Baretta-Huipers 1981; InsideWood 2004-onward). A search on the InsideWood (2004-onward) database shows affinities with species from the traditional Caesalpinioideae subfamily because of the number of cells per parenchyma strand that goes up to four cells, compared to hardly exceeding two cells for the traditional Papilionoideae subfamily. Concerning the traditional Mimosoideae, most species have homocellular rays, often septate fibres and rarely storied structures. Nonvestured pits are uncommon in Fabaceae and are restricted to three recircumscribed basal subfamilies: Cercidoideae, Duparquetioideae and Dialioideae (Herendeen 2000; Gasson et al. 2003; LPWG 2017; Zimmerman et al. 2017), all previously in the traditional Caesalpinioideae subfamily. An investigation of wood anatomy of these subfamilies was made with available literature (Gasson et al. 2003; InsideWood 2004-onward; Ogata et al. 2008; Pérez-Lara et al. 2019). Duparquetia Baill., the only genus of Duparquetioideae, is ruled out as it is a vine with no mineral inclusion and very thin-walled fibres. In Cercidoideae, only Bauhinia (which has a wide range of morphology) can be related to this fossil, but it shows more frequent uniseriate rays, sometimes septate fibres, homocellular rays or regular bands of parenchyma. Moreover, the number of cells per parenchyma strand is restricted to 4 in average in Bauhinia (compared to more than 4 in our fossil). Among Dialioideae, only Kalappia Kosterm., Koompassia and Martiodendron Gleason share features with our fossil: storied or irregularly storied rays, aliform-confluent parenchyma, 1-3 seriate rays, crystals in parenchyma, at least medium size intervessel pits and heterocellular rays. Kalappia and Martiodendron yet differ in having a strongly storied parenchyma, vessel elements and rays, fewer or more wavy confluent parenchyma, as well as smaller rays for Kalappia (up to 350µm compared to 400µm in average in our fossil). Koompassia wood is thus closer to our fossil than the other identified modern analogues, but has more frequent crystals in marginal cells; however, no specific NLR really stands out in regards of the state of preservation of the present fossil. The fossil genus Tzotziloxylon Pérez-Lara& Estrada-Ruiz (Pérez-Lara et al. 2019) covers fossils sharing features of the Cercidoideae/Dialioideae which includes non-vestured intervessel pits, aliform to occasionally confluent parenchyma as well as diffuse and sometimes banded, 1-4-seriate rays, crystalliferous parenchyma and non-storied structure. Our fossil is thus incompatible with this genus. Among Cercidoideae, woods resembling Bauhinia are described under the name Bauhinia and Bauhinium Trivedi& Panjwani with regularly storied and mostly uniseriate rays as well as regular parenchyma bands fig. 7. — cf. Koompassioxylon, MNHN.F.50182: A, Ts, mostly solitary vessels with aliform and aliform-confluent parenchyma, sometimes forming anastomosed shapes; B, Ts, interpretation of (A) of the parenchyma pattern (dotted) showing mostly confluent parenchyma, aliform parenchyma, but no regular bands; C, Ts, detail of aliform parenchyma and one vessel with tyloses (although rare, arrow); D, Tls, storied tendency of rays (arrows); E, Tls, 1-3 seriate heterocellular rays, sometimes with some rays with alternating uniseriate and mutliseriate portions (arrows); F, Tls, detail of a 2-seriate ray bordered by a strand of parenchyma 867 Fossil wood from Natma Formation, Myanmar (lower Miocene) GEODIVERSITAS • 2022 • 44 (28) A B DC E GF I H J made of at least eight cells (each marked by arrow); G, Rls, non-vestured (arrows), alternate, polygonal intervessel pits; H, Rls, heterocellular rays, sometimes with upright marginal cells (black arrow) and crystalliferous parenchyma, with more than 11 crystals per strand (white arrow); I, Rls, vessel-ray pits with distinct border (arrows) of the same size and shape as intervessel-pits; J, Tls, detail of crystals in chambered (arrows) parenchyma cells. Scale bars: A, B, 1 mm; D, 500 µm; C, E, F, H, 200 µm; G, I, J, 20 µm. Abbreviations: see Fig. 2. 868 GEODIVERSITAS • 2022 • 44 (28) Gentis N. et al. (Ramanujam& Rao 1966; Prakash& Prasad 1984; Trivedi& Panjwani 1986; Awasthi& Prakash 1987; Awasthi& Mehrotra 1990). The diagnosis of these genera is yet incompatible with our specimen. Among Dialioideae, woods resembling Koompassia are described under the name Koompassioxylon (Kramer 1974; Srivastava& Awasthi 1996). Although most of the features of the genus diagnosis are compatible with our fossil, some differences have to be pointed out: the aliform to confluent parenchyma with only a few vessels, marginal ray cells can be subdivided and contain crystals (presence of which is ambiguous in the present fossil as it is hard to determine if crystals belong to ray cells or the underneath axial parenchyma cells). Moreover, intervessel pits are vestured in previously described specimens of Koompassioxylon. This specimen is different from the K.elegans described above (p.864) because of the absence of numerous crystals in marginal ray cells, mainly confluent parenchyma and mainly 2-seriate rays. Pending more comprehensive studies about fossil Cercidoideae/Dialioideae and other related Fabaceae, we name it cf. Koompassioxylon to emphasize its close affinity to this genus. Koompassia ecology has been previously described (p.864). Kalappia trees reach up to 40m tall, 90cm in diameter, and are restricted to wet lowland forests of Sulawesi, up to 300m (rarely 500m) altitude (Hou et al. 1996; Sosef etal. 1998). Martiodendron are South American canopy trees, often riparian, growing in rainforests, periodically inundated forests, but also in tropical savanna woodlands, deciduous or seasonally dry forests, below 600m altitude (Koeppen& Iltis 1962; Lewis et al. 2005). These genera have relatively hard and durable wood (Scheffer& Morrell 1998). Genus Pahudioxylon Chowdhury, Ghosh& Kazmi Pahudioxylon bankurensis Chowdhury, Ghosh& Kazmi (Fig. 8) Pahudioxylon bankurensis Chowdhury, Ghosh& Kazmi, 1960: 22, pl. 2, figs1-6, ill.1. Albizzioxylon sahnii Ramanujam, 1960: 118, pl. 21, ill. 25. Pahudioxylon sahnii Ghosh& Kazmi, 1961: 96, figs 1, 2. Pahudioxylon deomaliense Prakash, 1965c: 433, figs. 1, 2. Ingoxylon sahnii – Müller-Stoll& Mädel 1967: 112. Pahudioxylon welkitii Lemoigne& Beauchamp, 1972: 336, pl.12. Pahudioxylon assamicum Prakash& Tripathi, 1975: 63, pl. 2, figs7, 9, 11, 12. Pahudioxylon indicum Prakash, 1979: 66, pl. 2, figs 1-5. Pahudioxylon bengalensis Ghosh& Roy, 1982: 52, figs1, 2. o riginal holotype . — Geological Survey of India (G.S.I.) no.P2/126. M aterial . — MNHN.F.50183 (field number: 17FN03). Estimated minimal diameter: 14-34cm. loCality. — Kalewa Township, Sagaing Region, Myanmar. age. — Upper lower to lowermost middle Miocene. DesCription Wood diffuse-porous. Growth ring boundaries distinct, marked by marginal parenchyma bands (Fig. 8). Vessels solitary (55%) as well as in radial groups of 2-4 (45%) (Fig. 8A), round to oval, 1-5 per mm² (average: 3); tangential diameter 120-275µm (average: 185µm). Tyloses absent. Vessel elements 120-370µm (average: 225µm) long. Perforation plates simple. Intervessel pits alternate, polygonal shaped, 4-7µm (average: 6µm) wide (Fig. 8C). Parenchyma paratracheal lozenge-aliform up to 13 cells wide for each wing and 5 cells around vessels, also confluent joining 2-4 vessels or groups of vessels (Fig. 8A, B), marginal bands 4-5 cells wide, sometimes surrounding narrower vessels; parenchyma cells 35-160µm (average: 90µm) long, 15-40µm (average: 26µm) wide; around 3-5 cells per parenchyma strand (Fig. 8E), crystals in chambered parenchyma cells at the margin of the aliform parenchyma (Fig. 8E), crystals are visible as black dots on the transversal section (Fig. 8A, B); possibly some rare diffuse parenchyma cells. Rays 1to 3seriate (mostly 3) (Fig. 8E), a very faint storied tendency in some parts of the section (Fig. 8D), 5-7 rays per mm (average: 6), 140-370µm (average: 245µm) or 8-21 cells high, mostly homocellular made of procumbent cells (Fig. 8F), rarely 1 row of marginal square (or larger procumbent) cells with sometimes crystals in them (Fig. 8G). Fibres non-septate, 7-30µm (average: 19µm) wide, thin-to-thick walled (lumina 1.6 times the double wall thickness in average). DisCussion This specimen is characterized by: 1)diffuse-porous wood, 2)exclusively simple perforation plates; 3)homocellular and 2to 3-seriate rays; 4)aliform to aliform-confluent parenchyma; 5)marginal parenchyma; and 6) storied ray tendency in places. These combined features are shared with some Fabaceae (Baretta-Kuipers 1981; Gasson et al. 2003; Evans et al. 2006; InsideWood 2004-onward), and in particular with many traditional Caesalpinioideae because most of the genera have heterocellular rays as well as rays more than 500µm high and the number of cells per parenchyma strand often goes up to 4 and even more (compared to hardly exceeding 2 for the traditional Papilionoideae subfamily). Traditional Mimosoideae are ruled out as many genera have septate fibres (about 30 %), exclusively homocellular rays with small ray cells, rare storied structures and no crystals in ray cells. A comparison with Caesalpinioideae genera reveals that our specimen looks similar to the genera Afzelia and Intsia (Gasson et al. 2003; InsideWood 2004-onward), especially for their well-defined aliform parenchyma, sometimes confluent, their marginal parenchyma bands, the distinct crystals present in the margin of the aliform parenchyma (even in transverse section), and the strands commonly from 2 to 5 cells long. Both genera are hardly distinguishable based on xylological characters (Prakash 1966; Müller-Stoll& Mädel 1967; Gasson et al. 2003; Ogata et al. 2008) although they have a different geographical distribution (Léonard 1950). Afzelia is a genus of Africa and South-East Asia (mostly in Africa) (Léonard 1950; POWO 2019). Intsia is a South and South-East Asian 869 Fossil wood from Natma Formation, Myanmar (lower Miocene) GEODIVERSITAS • 2022 • 44 (28) genus, with species also present in Australia, Madagascar or Tanzania (POWO 2019). Based on a comparison with some species of Afzelia and Intsia from the InsideWood (2004-onward) database and the Xylarium of the MNHN (specimen MNHN-P-P00396553), our specimen shows the closest affinity with Intsia bijuga (Colebr.) Kuntze because of the clear margin of crystalliferous aliform parenchyma cells, a tendency to irregular storied rays, clear marginal bands of parenchyma often in contact with small vessels and the fequency of vessels. Although, this species does not display any marginal square ray cells with crystals, this feature is only sporadically observed in our fossil and should not be used as a distinctive feature. Based on their similar wood anatomy, fossil wood resembling both Afzelia and Intsia are described under the genus Pahudioxylon (Chowdhury et al. 1960). When compared with fossil species of this genus (a synthetic table is provided by Feng et al. 2015), all features of our fossil are compatible with the wide diagnosis of P.bankurensis. Our specimen is thus attributed to P.bankurensis. Intsia trees are tropical and live in wet evergreen forests, with rainfall higher than 2000mm per year, and is associated with Anisoptera Korth. and Hopea Roxb. (Soerianegara& Lemmens 1993). Intsia are mostly found in coastal habitat, near beaches, mangroves (Loo& Tan 1997) and sometimes in peatlands (Tanjung et al. 2020). Intsia bijuga is found along coasts, river edges, tidal or temporarily inundated places (sometimes salty water), in mangroves and back-mangroves, primary and secondary forests, up to 600m altitude (Soerianegara& Lemmens 1993; Hou et al. 1996; Orwa et al. 2009). Its seeds are sea-dispersed (Lewis et al. 2005). A E F G B C D fig. 8. — Pahudioxylon bankurensis Chowdhury, Ghosh & Kazmi, MNHN.F.50183: A, Ts, solitary vessels or in radial multiples of 2-4, aliform parenchyma with marginal crystals (black arrow) and marginal parenchyma (white arrow); B, Ts, detail of aliform parenchyma with crystals in marginal cells; C, Tls, alternate intervessel pits, probably vestured (arrow); D, Tls, rays with storied tendency (arrows); E, Tls, 2-3 seriate, not storied rays with crystalliferous parenchyma (black arrow), strand of parenchyma made of 4 cells (white arrow); F, Rls, homocellular ray; G, Rls, sometimes but rarely crystals in marginal ray cells (arrow). Scale bars: A, 1 mm; D, 500 µm; B, E, 200 µm; F-G, 150 µm; C, 20 µm. Abbreviations: see Fig. 2. 870 GEODIVERSITAS • 2022 • 44 (28) Gentis N. et al. Pahudioxylon cf. bankurensis (Fig. 9) Same synonymy list as Pahudioxylon bankurensis. o riginal holotype . — Geological Survey of India (G.S.I.) no.P2/126. Material. — MNHN.F.50184 (field number: 17FN06), MNHN.F.50185 (field number: 17FN16), MNHN.F.50186 (field number: 17FN18). Estimated minimal diameter: 10-28cm, 38-60cm for specimen 17FN18. age. — Upper lower to lowermost middle Miocene. loCality. — Kalewa Township, Sagaing Region, Myanmar. DesCription Wood diffuse-porous. Growth rings present, marked by marginal parenchyma bands (Fig. 9A). Vessels mostly solitary (70-90%) or grouped by 2-4 (Fig. 9A), round to oval, 1-7 per mm² (average 3); tangential diameter 100-300µm (average 200µm). Tyloses absent. Vessel elements 160570µm (average 310µm) long. Perforation plates simple (Fig. 9E). Intervessel pits alternate, 4-12µm (average 8µm) in diameter, apparently vestured (Fig. 9G). Axial paren - chyma lozenge-aliform to weakly confluent when vessels are close to each-other, and in marginal bands 1-7 cells wide, sometimes merging with the aliform parenchyma of smaller vessels (Fig. 9A); diffuse sometimes present, larger than fibres cells, often in the vicinity of the rays but also scattered among fibres (Fig. 9A, B); parenchyma cells 40-145µm (average 80µm) long, 10-49µm (average 25µm) wide; 2-4 or more cells per parenchyma strands; crystals in chambered parenchyma cells (Fig. 9E) (up to 16 crystals per parenchyma strands), mostly in the margin of aliform parenchyma and diffuse parenchyma. Rays 1to 3-seriate (Fig. 9D), storied in some places (Fig. 9C), sometimes not storied, varying within the same specimen and among individuals, ripple marks visible with the naked eye to a greater or lesser extent for all specimens, 5-12 rays per mm (average 8), 130-400µm (average 240µm) or 6-20 cells high, homocellular made of procumbent cells (Fig. 9F). Fibres thin to thick-walled (lumina 1.2 times the double wall thickness in average, for best preserved fibres), nonseptate, 8-26µm (average 16µm) wide, a storied tendency when it is the case for the rays. DisCussion Two out of three specimens (MNHN.F.50184 and MNHN.F.50185) are very similar in color and mineralization, and display the same compression, suggesting that they might represent different parts of the same broken piece. The third specimen (MNHN.F.50186) has a different preservation and deformation, bigger and less dense vessels, but shares all the key characters of the first two. They are characterized by: 1) diffuse-porous wood; 2)exclusively simple perforation plates; 3)lozengealiform parenchyma; 4)marginal bands of parenchyma; and 5)1-3-seriate and storied rays. These specimens are very close to our Pahudioxylon bankurensis but display distinctive features that allow us to put them apart: the lozenge-aliform parenchyma is less extended, vessels (or vessel groups) are more separated from each other which leads to a lesser degree of confluence; clear diffuse parenchyma is present; the degree of storied structures is greater, as ripple marks are visible. No perfect match comes out on InsideWood (2004-onward) with the storiation of rays taken into consideration as well as diffuse-parenchyma. These characters apart, our fossils share most features of Afzelia and Intsia species as well as of Pahudioxylon bankurensis. The two living genera are yet rarely described with diffuse parenchyma or storied rays. Soerianegara& Lemmens (1993) mention diffuse parenchyma in only a few Afzelia and Adenanthera L. species. The latter having no storied structures nor more than four cells per parenchyma strands (InsideWood 2004-onward). Diffuse parenchyma is mentioned in Afzelia africana Sm. ex Pers. and A.javanica (Miq.) J. Léonard (Soerianegara& Lemmens 1993, InsideWood 2004-onward, Feng et al. 2015), but they have no or faint storied rays (as seen in figures of Gérard& Louppe 2011). Storied tendency of the rays is mentioned for Intsia (Ogata et al. 2008, Feng et al. 2015), A.rhomboideae (Blanco) Fern.-Vill. and can be seen in A.xylocarpa (Kurz) Craib. It is noteworthly that A.xylocarpa and A.martabanica (Prain) J.Léonard grow today in Myanmar (Léonard 1950; POWO 2019; Haw 2019). The latter is restricted to this area, but no information on its wood anatomy or ecology is available. We consider that no single species can be assigned as NLR, but the genera Afzelia-Intsia as a whole regarding diffuse parenchyma. Similarities are found between our specimens and Adenantheroxylon pavoninium Prakash& Tripathi (1968, 1969), but A.pavoninium has commonly 2-seriate rays, vasicentric parenchyma (rarely aliform) and no marginal parenchyma. Among species of Pahudioxylon, our specimens are very close to Pahudioxylon bankurensis as well as P.kiliani (Louvet) Prakash (Prakash et al. 1967). Storied rays (or a storied tendency) and diffuse parenchyma are only (and variably) found in P.bankurensis (Feng et al. 2015). The specimen no. 5790 of Vozenin-Serra& Privé-Gill (1989), attributed to the species P.sahnii (synonym of P.bankurensis) displays the same diffuse parenchyma and portions of the wood with storied rays; in addition, Feng et al. (2015) also described some specimens of P.bankurensis for which the parenchyma is not in contact with vessels and could be interpreted as diffuse parenchyma (Feng et al. 2015: 489, fig. 2K). These features seem to be variable among individuals of the same species and among the species of the same genus, and we do not consider them as diagnostic to establish a new species or to exclusively attribute the specimens to Pahudioxylon bankurensis. Consequently, we assign these specimens to Pahudioxylon cf. bankurensis. Afzelia are tropical trees present in Africa and Asia at low elevation up to 400m altitude (Ali 1973; Hou et al. 1996; Orwa et al. 2009). Afzelia trees grow in mixed deciduous 871 Fossil wood from Natma Formation, Myanmar (lower Miocene) GEODIVERSITAS • 2022 • 44 (28) and dry evergreen forests, on well-drained soils and in periodically inundated lowlands (Soerianegara& Lemmens 1993). The habitat of Intsia is given p.868. Among the available species with sometimes diffuse parenchyma, Afzelia africana is tolerant to a wide range of climate from humid to dry forests, but is essentially found in the savanna woodland/dense dry forests borders and in semi-deciduous forests up to 1400m altitude; it can also occur in lowland rainforests, dry forests, gallery forests, periodically inundated areas and savannas, but commonly requires annual rainfall above 900mm (Orwa et al. 2009; Gérard& Louppe 2011; Hills 2020); Afzelia javanica grows in primary and secondary forests, in dry places, mostly in lowlands up to 800m altitude (Hou et al. 1996). These two species, pending more comprehensive anatomical data on the genus, are considered as the most adequate NLR. Family MoraCeae Gaudich. Genus Artocarpoxylon Prakash& Lalitha Artocarpoxylon kartikcherraensis Prakash& Lalitha (Fig. 10) Artocarpoxylon kartikcherraensis Prakash& Lalitha, 1978: 132, fig.1. o riginal holotype . — Birbal Sahni Institute of Palaeosciences Museum no. 35317. Material. — MNHN.F.50187 (field number: NAT17-2). Estimated minimal diameter: 25-37cm. age. — Upper lower to lowermost middle Miocene. loCality. — Kalewa Township, Sagaing Region, Myanmar. A B C D E F G A fig. 9. — Pahudioxylon cf. bankurensis Chowdhury, Ghosh & Kazmi, MNHN.F.50184: A, Ts, mostly solitary vessels, aliform parenchyma with crystals in marginal cells (black arrow), marginal parenchyma (bottom) and diffuse parenchyma (white arrow); B, Ts, detail of diffuse parenchyma (white arrow), often in contact with rays; C, Tls, storied rays in some parts of the section; D, Tls, 1-3-seriate homocellular rays in a non-storied part of the section; E, Rls, crystals in parenchyma strands, the strand on the left being diffuse parenchyma; F, Rls, homocellular ray; G, Tls, alternate intervessel pits, apparently vestured. Scale bars: A, 1 mm; C, 500 µm; B, D-E, 200 µm; F, 100 µm; G, 50 µm. Abbreviations: see Fig. 2. 872 GEODIVERSITAS • 2022 • 44 (28) Gentis N. et al. DesCription Wood diffuse-porous. Growth rings absent. Vessels solitary (70%) as well as in groups of 2 to 6, evenly distributed (Fig. 10A), oval, 0-8 per mm² (average: 3); tangential diameter 85-300µm (average: 180µm). Tyloses present (Fig. 10B). Vessel elements 130-430µm (average: 260µm) long. Perforation plates simple. Intervessel pits alternate, 3-9µm diameter (mean 6µm) (Fig. 10C). Vessel-ray pits not preserved. Parenchyma paratracheal in vasicentric sheath, broad to aliform with no distinct limit between parenchyma and fibres (Fig. 10A, B), maybe diffuse and rarely confluent; parenchyma cells 45-160µm high (average: 90µm), 10-45µm wide (average: 25µm) in tangential section; 4-8 cells per parenchyma strands. Rays 1to 6-seriate, mostly 5-seriate (Fig. 10D, E), non-storied, 4-7 rays per mm (average: 5), 190-950µm (average: 480µm) or up to 50 cells high with end-to-end fusion possible (Fig. 10E), 1-seriate and some 2-seriate rays made of upright cells only or mixed with procumbent cells, other multiseriate rays heterocellular made of procumbent cells with 1-4 upright marginal cells (Fig. 10H). Some sheath cells present (Fig. 10E). Fibres non-septate, 6-30µm (average: 15µm) wide, thin-to-thick walled (lumina 1.46 times the double wall thickness) with visible lumen sometimes wide. Radial laticifer tubes present in rays, of the same size as ray cells, visible when filled with black content (Fig. 10F, G). In tangential section, much larger or longer cells are visible, with a non-circular shape in the flank of vessels, sometimes even modifying the shape of the rays or between two ray ends (Fig. 10E). DisCussion This specimen is characterized by: 1)diffuse-porous wood; 2)scattered vessels; 3)up to 6-seriate and heterocellular rays, occasionally with sheath cells; 4)diffuse and broad sheath vasicentric to aliform parenchyma; 5)no growth rings; 6)abundant tylose; and 7)radial latificer tubes of the same size as ray cells. These features are characteristic of the Moraceae family (InsideWood 2004-onward). Tetramelaceae are also anatomically close but can have growth rings, parenchyma strands or fibres storied, no laticifer tubes and no tyloses (InsideWood 2004-onward). Most of the Moraceae genera can be ruled out based upon some diagnostic characters: the presence of septate fibres, banded parenchyma (58% of the specimens described on InsideWood [2004-onward]), frequent homocellular rays, no diffuse-porous wood, often unilateral parenchyma or almost never sheath cells. One genus is clearly apart from other Moraceae genera because of the absence of any crystal in any type of cells: Artocarpus J.R.Forst.& G.Forst. This feature is also observed in our specimen. Among Artocarpus, 5 species have at least sometimes laticifers according to InsideWood (2004-onward), and three with sheath cells in addition: A.chama Buch.- Ham. (although it is unclear and varies among publications Pearson& Brown 1932; Purkayastha et al. 1976; Singh et al. 2017), A.dadah Miq., and A.integer (Thunb.) Merr. (which has no tylosis). According to Ter Welle et al. (1986), Artocarpus has common tyloses, non-septate fibres, heterocellular mostly 3-5-seriate rays with 1-2 (5) upright or square marginal cells, sometimes few sheath cells. Parenchyma is mostly aliform, sometimes confluent. Radial latex tube are common, and axial latex tubes can sometimes be present. Singh et al. (2017) described four species of Artocarpus and mention laticifers in A.chama, A.heterophyllus Lam., A.lamellosus Blanco and A.lacucha Roxb. Ex Buch.-Ham. Unfortunately, the figure does not provide clear illustrations (Singh et al. 2017: 76, fig. 1). Latex tubes usually appear in radial section as long lines inside rays (which is the case in our fossil) or among fibres, sometimes irregularly crossing both of them; their walls are irregular and the lumina bigger than that of parenchyma cells (Farías et al. 2009). All of these species are quite similar, and their characters (i.e. non-septate fibres, sheath cells, tyloses, confluent parenchyma and laticifers) seem to vary from authors to others. However, given the descriptions provided in references cited above and figures in InsideWood (2004-onward), A.chama appears the most anatomically close species to our fossil, as well as A.lacucha and A.lamellosus. Fossil Moraceae are mainly represented by the genus Ficoxylon Kaiser (Gregory et al. 2009) which shows mostly banded parenchyma. The fossil genus Artocarpoxylon, also in Moraceae, groups fossils resembling modern Artocarpus (Prakash& Lalitha 1978). It displays close similarities with our specimen as it shows no growth rings, mostly solitary vessels, tyloses, vasicentric to aliform parenchyma, 1-6 seriate heterocellular rays and some sheath cells. It also sometimes displays horizontal latex tubes, as in our specimen. Fossil remains related to modern Artocarpus are known as far as the Cretaceous from Asia to North America, Europe and even Greenland (Ball 1930; Mehrotra et al. 1984; Williams et al. 2017). Two species of this genus are described: A.kartikcherraensis and A.deccanensis Mehrotra, Prakash& Bande (Mehrotra et al. 1984). They both have laticifers in rays and strongly resemble our fossil. Artocarpus deccanensis has smaller and more numerous vessels (80-180µm and 9-20/mm²) and narrower sheath of vasicentric parenchyma than our fossil. Artocarpus kartikcherraensis has vessel diameter and density closer to our fossil (105-315µm and 2-3/mm²). The present fossil is thus attributed to Artocarpus kartikcherraensis. Artocarpus is a genus of Asian (also in Pacific and Australia) tropical trees growing in everwet climate or with a short dry season, in evergreen forests or in areas with mild monsoon climate, usually scattered in lowland dipterocarp forests below 1000m altitude (Lemmens et al. 1995; Berg et al. 2006). Artocarpus lamellosus lives in evergreen forests, rarely in semideciduous forests or savanna woodlands, commonly found in mixed dipterocarp and sub-montane forests up to 1500m altitude, on hillsides and ridges (Lemmens et al. 1995; Berg et al. 2006; Tropical Plants Database 2014-onward). Artocarpus chama lives in evergreen, semi-evergreen and moist deciduous forests up to 1500m attitude, in areas with a monsoon climate and rainfall of at least 2000mm a year (Gamble 1902; Jarrett 1959). Artocarpus lacucha lives up to 1200-1800m of attitude (Gamble 1902; Jarrett 1960) in evergreen, semievergreen and moist deciduous forests, with a distinct dry season (Jarrett 1960). 873 Fossil wood from Natma Formation, Myanmar (lower Miocene) GEODIVERSITAS • 2022 • 44 (28) Family DipteroCarpaCeae Blume Genus Anisopteroxylon (Ghosh& Kazmi) Poole Anisopteroxylon sp. (Fig. 11) M aterial . — MNHN.F.50188 (field number: 17FN07). Estimated minimal diameter: non-assessable due to compression. loCality. — Kalewa Township, Sagaing Region, Myanmar. age. — Upper lower to lowermost middle Miocene. DesCription Wood diffuse-porous. Growth rings indistinct or absent. Vessels mostly solitary (>90%) oval, 2-13 per mm² (average: 6) (Fig. 11A); tangential diameter 80-230µm (average: 150µm; measured on less compressed vessels, n=30). Tyloses present. Vessel elements 130-290µm long (average: 220µm). Intervessel and vessel-ray pits not found due to compression and rare occurrence of vessel multiples. Perforation plates simple (Fig. 11B). Vasicentric tracheids present (Fig. 11B). Axial parenchyma not well defined due to compression, probably vasicentric and/or diffuse; without mineral inclusions. Rays 1to 7-seriate, mostly 6to 7-seriate (Fig. 11C, D), rare unior biseriate made of upright or square cells, non-storied, 4-7 rays per mm (average: 5), 510-1500µm (average: 1050µm) or up to 60 cells high, heterocellular made of procumbent cells and continuous 1-seriate margin of sheath cells (Fig. 11C, D) which may appear as a mix of procumbent and upright cells in radial section, 2to 3-seriA B D C F GE H fig. 10. — Artocarpoxylon kartikcherraensis Prakash & Lalitha, MNHN.F.50187: A, Ts, sparse vessels, vasicentric parenchyma sometimes slightly aliform; B: Ts, detail of vasicentric parenchyma and vessels plugged with tyloses; C, Tls, alternate intervessel pits with lenticular appertures; D, Tls, 1-6 seriate rays with upright marginal cells; E, Tls, 1-6 seriate rays with sheath cells (black arrow) and end-to-end fusions (white arrow); F, Tls, latex tube in ray, mostly of the same size as ray cells (arrow); G, Rls, view of a latex tube in ray, recognizable as a continuous black line in radial section; H, Rls, heterocellular ray with 1-4 rows of marginal cells, sometimes appearing with upright and procumbent cells mixed due to sheath cells or end-to-end fusions. Scale bars: A, 1 mm; D, 500 µm; B, E, G-H, 200 µm; F, 100 µm; C, 50µm. Abbreviations: see Fig. 2. 874 GEODIVERSITAS • 2022 • 44 (28) Gentis N. et al. ate rays with over 4 rows of upright or square marginal cells (Fig. 11E), otherwise, only one row; silica bodies (?) present in some ray cells (Fig. 11F). Fibres non-septate, tangential diameter 9-25µm (average: 16µm). Secretory canals cannot be identified but we observed some scattered, very small pores (Fig. 11A); no signs of tangential bands of any kind that could indicate a presence of canal lines. DisCussion This specimen is characterized by: 1)diffuse-porous wood; 2)exclusively simple perforation plates; 3)solitary vessels; 4)heterocellular with mostly 5-6-seriate rays; 5)continuous sheath cells; and 6)vasicentric tracheids. These features suggest an affinity with modern Malvaceae and Dipterocarpaceae (Metcalfe& Chalk 1950). Most Malvaceae have vasicentric to confluent parenchyma, apotracheal parenchyma, rays of two distinct sizes (1and 4to 9-seriate; Metcalfe& Chalk 1950), and no vasicentric tracheid, which contrast with our specimens. Dipterocarpaceae share all these features, but are additionally characterized by vertical canals, except for non-Asian genera Monotes De Candolle, Marquesia Gilg and Pakaraimaea Maguire& P.S. Ashton (Chowdhury& Ghosh 1958; Ogata et al. 2008). The character of secretory canals apart, and searching for diffuse-porous woods with vasicentric tracheids, exclusively solitary vessels, rays more than 4-seriate, heterocellular and with sheath cells on InsideWood (2004-onward) (code: 5p 9p 60p 96a 98p 104a 105a 110p 118a) gives us back multiple results. But adding the features 159 and 160 (silica bodies present in ray cells), only the Dipterocarpaceae family matches with our specimen. Even without silica bodies in ray cells, the affinity with Dipterocarpaceae is the most likely. Among them, Cotylelobium Pierre and Dryobalanops C.F.Gaertn. genera are ruled out because of smaller vessels (< 150µm) and secretory canals in long lines respectively (Schweitzer 1958; Ogata et al. 2008). Although the preservation state of our fossil is not good enough to see individual secretory canals or parenchyma surrounding them, their long lines or seemingly bands should have been observable if they were present, as mentioned by Schweitzer (1958). Following the identification key of the main genera of Diptero carpaceae provided by Schweitzer (1958), we narrow down the identification to two genera: Anisoptera or Dipterocarpus. They are not always distinguishable from each other when the information about the secretory canals is not available. However, Anisoptera has solitary secretory canals or in short tangential lines (rarely more than 2-3 canals), and sheath cells are frequent in rays; Dipterocarpus has comparatively longer tangential lines (2-8 canals), more frequent uniseriate rays and uniseriate portions in rays and less sheath cells, although they can sometimes be present in the whole ray (Chowdhury& Ghosh 1958; Gottwald& Parameswaran 1966; Prakash& Tripathi 1970; Ogata et al. 2008). We also observed these trends on Anisoptera and Dipterocarpus specimens available at the MNHN (CTFT19039, CTFT18446, CTFT11366, MNHN-PP00402617, P00400540, P00415879, P00406060) and on InsideWood (2004-onward). Our specimen displays features closer to those of genus Anisoptera, and in particular with species Anisoptera costata Korth. and Anisoptera scaphula (Roxb.) Kurz. Both display the same wide rays (up to 7-8-seriate) as well as uniform and unfrequent uniseriate rays and solitary vessels. These two species are hardly distinguishable from each other based on xylotomy (Chowdhury& Ghosh 1958). Seven fossil genera of Dipterocarpaceae are described (Gregory et al. 2009). The genus Anisopteroxylon has been instituted by Ghosh& Kazmi (1958) and emended by Poole (1993) for fossil resembling Anisoptera, leaving the genus Diptero carpo xylon (Holden) Den Berger for fossil resembling Dipterocarpus only (see Prasad& Gautam 2016, p. 263). The differences between Anisopteroxylon and Dipterocarpoxylon are mostly based on secretory canals (solitary for Anisopteroxylon and in short tangential bands for Dipterocarpoxylon). The quantity of sheath cells is not mentioned in the diagnosis, although Ghosh& Kazmi (1958) describe them as “rather common”, and uniseriate rays as “not very siMpliFieD iDentiFiCation key For Fossil DipteroCarpaCeae bluMe This identification key is based on Schweitzer (1958); with additional informations of Anisopteroxylon Ghosh& Kazmi and Hopenium Awasthi from Poole (1993) and Awasthi (1980) respectively. 1. Secretory canals diffuse or in short tangential lines up to 8. Rarely long lines .............................................. 2 — Secretory canals almost always in long tangential lines. Rarely isolated ........................................................ 4 2. Narrow vessels (mean diameter < 150µm) ..................................................................................................... ....................................................................................... Vaterioxylon Trivedi& Misra/Vaticoxylon Schweitzer — Wide vessels (mean diameter > 150µm) ..................................................................................................... 3 3. Diffuse and small canals, sheath cells abundant ............................................ Anisopteroxylon Ghosh& Kazmi — Canals in short lines, solitary vessels ....................................................................... Dipterocarpoxylon Holden 4. Exclusively solitary vessels, fibre-tracheids ......................................................... Dryobalanoxylon Den Berger — None of these combined ............................................................................................................................. 5 5. Upright or square cells interspaced with procumbent cells ................................................ Hopenium Awasthi — Only procumbent cells in median portion (or sheath cells) ........................................ Shoreoxylon Den Berger 875 Fossil wood from Natma Formation, Myanmar (lower Miocene) GEODIVERSITAS • 2022 • 44 (28) common” (Ghosh& Kazmi 1958: 486). Even though our fossil has no visible canals, which is possibly due to poor preservation and compression, its solitary vessels, vasicentric tracheids, abundant sheath cells, broad rays with silica bodies in some cells, make it compatible with the genus Anisopteroxylon. Some specimens of this genus are close to our fossil (Appendix 1): A.garoense (Chowdhury) Prakash& Tripathi (Prakash& Tripathi 1970) from the upper Miocene of India, for its 7-9-seriate rays, 4-5 vessels/mm², and the presence of identical continuous sheath cells; A.oblongoides Yadav (Yadav 1989) and A.surmaensis Prasad, Agarwal& Mandaokar (Prasad etal. 2009) but the former has long tangential lines of canals of equal size as vessels and the latter has narrower rays (mostly 4-5-seriate); Anisopteroxylon jawalamukhi Ghosh& Ghosh (Ghosh& Ghosh 1958) has longer rays (up to 1950µm). Considering the lack of several diagnostic characters at the species level in our specimen, we only assign it to Anisopteroxylon sp. Anisoptera is a genus of south-east Asian trees, absent in India, living in evergreen or semi-evergreen forests; in mixed dipterocarp, mixed swamp or heath forests; on well-drained soils or in peat swamps, rarely above 1000m altitude (Ashton 1982; Soerianegara& Lemmens 1993). Anisoptera costata is a tropical tree that can be found in Myanmar, living in semievergreen dipterocarps forests and evergreen forests of seasonal areas. Mostly in moist or slightly dry areas. It rarely occurs in everwet forests but grows well along rivers and streams, at up to 700m altitude (Ashton 1982; Smitinand et al. 1990; Soerianegara& Lemmens 1993; Sam et al. 2004; Nguyen et al. 2017). Anisoptera scaphula grows in semi-evergreen and evergreen dipterocarp forests on foothills (Ashton 1982; Soerianegara& Lemmens 1993; Ly et al. 2020). A B C F D E fig. 11. — Anisopteroxylon sp., MNHN.F.50188: A, Ts, strongly compressed wood, poorly distinguishable vessels, seemingly exclusively solitary, small pores that could be solitary secretory canals (arrows); B, Tls, simple perforation plates, vasicentric tracheids (arrows); C, D, Tls, 1-7-seriate rays with continuous sheath cells all around multiseriate rays (arrow); E, Rls, heterocellular rays with procumbent cells in median portion (white arrow) and square or upright marginal cells (black arrow), sometimes appearing with both types mixed due to sheath cells; F, Rls, silica bodies (?) in ray cells. Scale bars: A, C, 500 µm; B, D-E, 200 µm; F, 100 µm. Abbreviations: see Fig. 2. 882 GEODIVERSITAS • 2022 • 44 (28) Gentis N. et al. with a storied tendency, and a similar size and frequency of vessels. The state of preservation of our fossil and its unclear canal distribution pattern does not allow us to attribute the specimen to an individual species with certainty nor to create a new species. Thus, we name this fossil Dryobalanoxylon sp. Dryobalanops general ecology and distribution is given p.878; Dryobalanops aromatica lives in lowland mixed dipterocarp forests, on dry sandy or gravelly soils near the coast up to 400m (Ashton 1982; Tropical Plant Database 2014-onward; Barstow& Randi 2018). Dryobalanops oblongifolia lives in lowland mixed dipterocarp forests, also in periodically inundated, freshwater swamps, near streams, in poorly drained forests or on hillsides below 600m (Ashton 1982; Soerianegara& Lemmens 1993; Barstow 2018a). Genus Shoreoxylon Den Berger Shoreoxylon cf. deomaliense Prakash& Awasthi (Fig. 15) Shoreoxylon deomaliense Prakash& Awasthi, 1971: 219, pl.1, figs3-4. holotype. — Birbal Sahni Institute of Palaeosciences Museum, India, specimen no. 34050. Material. — MNHN.F.50192 (field number: NAT17-4). Estimated minimal diameter: 25-37cm. loCality. — Kalewa Township, Sagaing Region, Myanmar. age. — Upper lower to lowermost middle Miocene. DesCription Wood diffuse-porous, showing lateral compression. Growth limits marked by tangential canal lines. Vessels about 77-90% solitary (Fig. 15A) and in radial groups of 2-4, oval due to lateral compression, 4-13 per mm² (average: 8; likely overesti - mated due to compression); tangential diameter 120-200µm (average: 150µm; likely underestimated due to compression). Tyloses present (Fig. 15F). Vessel elements 140-460µm (average: 330µm) long. Perforation plates simple. Intervessel pits alternate. Vessel-ray pits not preserved. Vasicentric tracheids present (Fig. 15F). Parenchyma mostly vasicentric and aliform with short wings, sometimes confluent (Fig. 15B) or rarely forming thin bands from several rays to rays; diffuse parenchyma with cells larger than fibres ones (Fig. 15K), sometimes gathered in small groups; thin bands of parenchyma are tangentially crossing the section at regular intervals (5-7mm). They can contain secretory canals but these ones are not always visible (Fig. 15H). When no canal is present, the bands are only 1-4 cells wide. Parenchyma cells 50-90µm long (average: 70µm), 15-35µm wide (average: 25µm) in tangential section; sometimes crystals in chambered cells (up to 8 crystals per strand seen) (Fig. 15E). Parenchyma cells can be more or less enlarged in the form of idioblasts. Rays 1to 5-(6-)seriate (mainly 4) (Fig. 15C), uniseriate about 15% of the rays, non-storied, 5-10 rays per mm (average: 8), 270-1600µm (average: 650µm) or up to 30-40 (even 70) cells high, heterocellular made of procumbent cells with 1-4 or more upright cells at the ends (Fig. 15G), end-to-end fusion possible resulting in very high rays (Fig. 15D). Fibres non-septate, 5-19µm (average: 12µm) wide, thin-to-thick walled (lumina 1 time the double wall thickness in average) (Fig. 15J). Secretory canals in long tangential lines surrounded by parenchyma (Fig. 15A, H, I), but also in seemingly short lines probably due to compression in concentric parenchyma bands, or very rarely scattered by 2 in the section, 30-100µm in tangential diameter (average 60µm). DisCussion This specimen is characterized by: 1)diffuse-porous wood; 2)mostly solitary vessels as well as in radial groups; 3)crystalliferous and mostly aliform parenchyma, as well as diffuse; 4)1-6-seriate heterocellular rays; 5)long tangential lines of secretory canals; and 6)vasicentric tracheids. As for our previous specimens (start p. 878), these features are diagnostic of the Dipterocarpaceae family. According to the identification key of Schweitzer (1958), long tangential lines of canals are found in the genera Shorea, Dryobalanops, Hopea, and Parashorea Kurz but the latter three can be dismissed: the genus Dryobalanops has exclusively solitary vessels and visible fibretracheids, Hopea has smaller and more frequent vessels (less than 200µm in average diameter for 10-20 or more vessels per mm²), while Parashorea has less vessels and larger rays (up to 7-seriate) (Metcalfe& Chalk 1950; Gottwald& Parameswaran 1966; Soerianegara& Lemmens 1993; Richter& Dallwitz 2000-onward; Ogata et al. 2008). The genus Shorea is divided into several sections that are more or less phylogenetically supported and roughly characterized by few features: section ‘Pentacme’ by big vessels, section ‘Richetioides’ (or ‘Richetia’) by the presence of radial canals, section ‘Anthoshorea’ by the presence of silica bodies in ray cells as well as short rays, thin-walled fibres and rare crystals in parenchyma, sections ‘Rubroshorea’ by solitary crystals or in short chain of non-chambered (or chambered) parenchyma cells and idioblasts, section ‘Shorea’ by short rays, few marginal ray cells and crystals in long chains of chambered parenchyma cells and idioblasts. The present wood would thus be close to the section ‘Shorea”. Among Shorea extant species, Shorea laevis Ridl. shares many features of our fossil including the aliform parenchyma (but without crystals), the vessel size and density, the similar rays (mostly 3-5-seriate) with few marginal cells and few uniseriate rays. The same arrangement of crystalliferous parenchyma and ray size is found in S.parvifolia Dyer, S.pauciflora King, S.atrinervosa Symington., and with a lesser extent in S.maxwelliana King and S.almon Foxw. fig. 15. — Shoreoxylon cf. deomaliense Prakash & Awasthi, MNHN.F.50192: A, Ts, mostly solitary vessels and long tangential line of secretory canals embedded in parenchyma bands (arrow); B, Ts, zoom in on solitary vessels, mostly aliform parenchyma as well as some diffuse parenchyma (arrow); C, D, Tls, long 1-6-seriate rays with end-to-end fusion (arrow); E, Rls, crystals in chambered parenchyma cells; F, Tls, vasicentric tracheids (black arrow) and tylose in vessel 883 Fossil wood from Natma Formation, Myanmar (lower Miocene) GEODIVERSITAS • 2022 • 44 (28) A D E F B C G J K IH (white arrow); G, Rls, heterocellular ray with upright marginal cells; H, Ts, synthetic drawing of the transversal section (with only some vessels and rays displayed); visible bands of parenchyma are shown as grey lines and secretory canals as red dots (orange when their identification is less clear), arranged mostly in long tangential lines; I, Ts, zoom in on secretory canals embedded in parenchyma; J, Ts, thin-to-thick walled fibres; K, Ts, diffuse parenchyma (arrows). Scale bars: H, 1 cm; A, 1 mm; B, D, I, 250 µm; C, G, 200 µm; F, K, 100 µm; E, J, 50 µm. Abbreviations: see Fig. 2. 884 GEODIVERSITAS • 2022 • 44 (28) Gentis N. et al. Shoreoxylon groups the fossil specimens close to all Shorea and Parashorea. The genus Hopenium was instituted (Awasthi 1980) for woods resembling Hopea, with upright ray cells in the middle of the rays. Species descriptions in Shoreoxylon are often overlapping and rarely consider interand intraspecific variations. In addition, they sometimes lack diagnostic characters or qualitative illustrations. Consequently, it is difficult to identify a unique species that could be attributed to our specimen. Some species display features that are close to our fossil (Appendix 1): Shoreoxylon burmense Prakash (Prakash 1965a, 1973; Licht et al. 2014) share the same type of rays, the parenchyma is also quite similar, but the secretory canals are grouped in very close lines, from 2 to 4, which is not the case in our fossil, and it has no crystal in parenchyma. Shoreoxylon indicum Awasthi (1974) has the same vessel and ray distribution, crystals in parenchyma cells, but its apotracheal and confluent parenchyma are more developed whith only 5 crystals per parenchyma strands (up to 8 in our fossil). Shoreoxylonposthumi Schweitzer (1958) has crystals in parenchyma as well as enlarged parenchyma cells, but it has much developed apotracheal parenchyma and its canal lines are irregularly distributed or superimposed. Shore oxylontipamense Prakash& Awasthi (1970) has similar vessel, parenchyma and ray arrangement, but it also has bigger vessels, larger canals, thinner fibres cell walls and sheath cells. Shoreoxylondeomaliense is the closest fossil species to our specimen (Prakash& Awasthi 1971; Licht et al. 2014), though the present fossil has a lesser frequency of vessels, more aliform parenchyma and slightly thinner rays (up to 6-seriate, compared to 7-seriate for S.deomaliense) with shorter rows of marginal ray cells. Considering its preservation, we attribute our fossil to Shoreoxylon cf. deomaliense. Shorea is a genus of tropical Asian trees growing in humid lowland areas, on podzols and peat swamps, mostly below 1000m altitude (Ashton 1982; Soerianegara& Lemmens 1993). Shorea laevis mostly grows on well-drained to dry soils, on ridges or hillsides up to 1000m. It is also found in lowland mixed dipterocarp forests and on alluvial sites (Ashton 1982; Soerianegara& Lemmens 1993; Pooma et al. 2017). All the other species cited above are found in mixed dipterocarp forests in lowlands or on rolling hills, on well-drained soils at up to 1000m altitude (Ashton 1982, 2004). Shoreoxylon cf. sumatraense Du (Fig. 16) Shoreoxylon sumatraense Du, 1988b: 342, pl.1, figs1-4, pl. 2, figs1-4, pl. 5, fig. 4. h olotype . — National Museum of Geology and Mineralogy, Leiden, specimen no. RGM B (RGM 383446). M aterial . — MNHN.F.50193 (field number: 17FN12). Estimated minimal diameter: 10-14cm. loCality. — Kalewa Township, Sagaing Region, Myanmar. age. — Upper lower to lowermost middle Miocene. DesCription Wood diffuse-porous. Growth rings marked by tangential canal lines. Vessels 45-70% solitary, otherwise in radial multiple of 2-4 and clusters of different size with sometimes many small vessels surrounding bigger ones (Fig. 16A), round to oval, 8-20 per mm² (average: 15); tangential diameter 80-240µm (average: 160µm); walls of the vessel cells are thick compared to other cells (about 6-11µm). Tyloses present, common and visible (Fig. 16A-C). Vessel elements 150-570µm (average: 350µm) long. Perforation plates simple. Intervessel pits alternate, 3-6µm of diameter (average: 4µm). Vessel-ray pits apparently irregular in size and shape, simple, 4-10µm (Fig. 16F). Vasicentric tracheids present (Fig. 16H). In cross section, parenchyma indistinguishable from fibres due to poorly preserved cell walls and no obvious difference of size. At least vasicentric and around canals appearing as little flattened cells; parenchyma easily recognizable in tangential section and abundant in places, 4-8 cells per strand (Fig. 16E), it seems as abundant as fibres; parenchyma cells 40-120µm long (average: 90µm) 10-30µm wide (average: 20µm) in tangential section; no crystal. Rays usually 1to 4-seriate (mainly 3-, very rarely 5-seriate) (Fig. 16C, D), uniseriate for about 20% of the rays, non-storied, 4-8 rays per mm (average: 6), 3401040µm (average 620µm) or 10-45 cells high; multiseriate rays are heterocellular, made of procumbent cells mostly with 1-4 but sometimes more (9) square or upright cells at the ends (Fig. 16E, I); no mineral inclusion seen. Fibres non-septate, 8-20µm (average: 13µm) wide, apparently thin-walled (lumina 1.9 times the double wall thickness in average where the walls are best preserved; or individual cell wall 3-4.4µm thick), fibres cell walls are poorly preserved; clearly aligned in radial rows. Secretory canals in long tangential bands surrounded by parenchyma, frequent and closely spaced (every 0.4-3mm) (Fig. 16B, G), 40-75µm of diameter (average: 60µm). Black spots and long lines crossing tangentially the section (well visible in low magnification) are interpreted as potential lines of canals, even though their origin remains to be confirmed (Fig. 16G). DisCussion This specimen is characterized by: 1)diffuse-porous wood; 2)exclusively simple perforation plates; 3)closely and frequently spaced long tangential lines of secretory canals; 4)vasicentric tracheids; 5)presence of sheath cells; 6)vessels mostly in groups and sometime forming clusters; and 7)1-4-seriate rays, mostly less than 1mm high. The poor preservation of the specimen makes our observations on ray and parenchyma arrangement uncertain. Nonetheless, its features indicate an affinity with modern and fossil Dipterocarpaceae. As discussed p.882, long tangential bands of canals, grouped vessels and no fibre-tracheids are features compatible with extant Shorea. Among this genus, the sections ‘Shorea’ and ‘Rubroshorea’ are the more compatible with our fossil: ‘Shorea’ for the frequency of vessels (4-10[16] per mm²), the frequency of solitary vessels (55-85%), common tylose, mostly 3-4-seriate rays; ‘Rubroshorea’ for the thickness of fibres, the composition of the rays (with 1-4 row of marginal cells and sometimes 885 Fossil wood from Natma Formation, Myanmar (lower Miocene) GEODIVERSITAS • 2022 • 44 (28) A GH I C D E F B fig. 16. — Shoreoxylon cf. sumatraense Du, MNHN.F.50193: A, Ts, vessel often in groups, sometimes in clusters (arrows); B, Ts, long tangential lines of secretory canals (arrows) close to each other; C, D, Tls, 1-4-seriate rays; E, Tls, 3-seriate rays with a uniseriate row of six marginal cells (white arrow), abundant parenchyma (black arrow), 4-8 cells per strand; F, Rls, vessel-ray pits irregular in shape, simple with reduced borders (arrows); G, Ts, synthetic drawing of the transversal section (with only some vessels and rays displayed), visible bands of parenchyma as grey lines and recognized secretory canals as red dots, arranged mostly in long and closely spaced tangential lines; H, Rls, vasicentric tracheids (arrows); I, Rls, heterocellular ray with square or upright marginal cells (arrows). Scale bars: G, 1 cm; A, 1 mm; C, 500 µm; B, 330 µm; D, I, 200 µm; E, H, 100 µm; F, 50 µm. Abbreviations: see Fig. 2. 886 GEODIVERSITAS • 2022 • 44 (28) Gentis N. et al. more), the propencity to have lines of canals spaced by less than 1mm and small canals (40-80µm). An examination of different species of Shorea shows that this fossil is similar to Shorea negrosensis Foxw. (‘Rubroshorea’ section). This species has closely spaced canals that are sometimes small (surrounded by only four parenchyma cells), thin-to-thick fibres walls, parenchyma cells mostly of the same diameter as fibres and flattened around canals, frequent groups of vessels, as well as similar vessel groups and clusters, and aliform to aliform-confluent parenchyma. The modern specimen in our possession (no. CTFT25647) has mostly 3-4-seriate rays with 1-4marginal rows of cells, sometimes more for thinner rays. As for the previous Shorea-like fossil, we compared this one with the genus Shoreoxylon. Our specimen is among the few to have a high frequency of vessels, closely spaced canal lines and thin rays with not so rare uniseriate ones. Four species of Shoreoxylon share many similarities with our fossil (Appendix 1; Awasthi 1974; Sukiman 1977; Trivedi& Ahuja 1979; Bande& Prakash 1980; Du 1988b): S.arcotense Awasthi has a high density of vessels and small canals, but exclusively solitary and smaller vessels; S.pachitanensis Sukiman has similar bands of canals, but higher rays and wider canals; S.ornatum (Trivedi& Ahuja) Bande& Prakash also has similar bands of canals, but less frequent and more solitary vessels, rays up to 5-seriate and higher (up to 1870µm) and sheath cells are present; S.sumatraense from the Quaternary of Sumatra shares most of the main features of our specimen. It has frequent vessels (9-14 per mm²), multiples up to 5 vessels, sometimes clusters, thick walls (15µm, 6-11µm in our fossil), parenchyma abundant, rays 1-4-seriate with similar composition and height, radially aligned and thin-walled fibres (3µm, compared to 3-4.4µm in our fossil), small canals (40-90µm compared to 40-75µm in our fossil). However, it displays wider vessels (200-360µm), a storied tendency in the parenchyma that cannot be observed in our specimen, and no visible vasicentric tracheids (but they seem present in the figures provided) and sometimes crystals in ray cells. The higher density and smaller vessels in our specimen could be explained by compression or environmental constraints. These differences are minor and can be related to intraspecific variations and the poor preservation of our specimen; we thus attribute our specimen to Shoreoxylon cf. sumatraense. Du (1988b) indicates that S.sumatraense shares most features with extant Shorea negrosensis, which is in adequacy with our own observations. Shorea negrosensis grows today in the Philippines in evergreen, semi-evergreen and seasonal dipterocarp forests at low elevation (Ashton 1982; Soerianegara& Lemmens 1993; EDC 2020). Shoreoxylon sp.1 (Fig. 17) Material. — MNHN.F.50194 (field number: 19NAT03-1). Estimated minimal diameter: 13-17cm. loCality. — Kalewa Township, Sagaing Region, Myanmar. age. — Upper lower to lowermost middle Miocene. DesCription Wood diffuse-porous. Growth rings indistinct. Vessels solitary (55-60%) or in groups of 2-4, sometimes in clusters made of a mix of small vessels and vasicentric tracheids (Fig. 17A, D), round to oval due to radial compression, 4-16 per mm² (average: 9); tangential diameter 70-270µm (average: 180µm). Tyloses present and common, especially in tangential section (Fig. 17C). Vessel elements 150-400µm (average: 300µm) long. Perforation plates simple (Fig. 17E). Intervessel pits alternate, 4-7µm. Vasicentric tracheids present, interspaced with parenchyma (Fig. 17F). Parenchyma frequent, mostly aliform or aliform-confluent, crossing rays and joining vessels together but without any particular arrangement (Fig. 17A), diffuse-in-aggregate or drawing a network between rays in the form of small bands or groups; 4 cells sheath around canals (Fig. 17D); parenchyma cells 40-110µm long (average: 80µm), 13-26µm wide (average: 20µm) in tangential section, 4-5 cells (possibly more) per parenchyma strand, possibly subdivided (Fig. 17F). Rays 1to 5-seriate (dominantly 4-seriate) (Fig. 17B, C), uniseriate few (< 10%) and short (mainly less than 10 cells high), 4-8 rays per mm (average: 6), 130-800µm (average: 460µm) or 6-50 cells high. Heterocellular made of procumbent cells with 1-2 upright cells at the ends (Fig. 17E), rarely more (up to 7). Sometimes rays appear weakly heterocellular as the marginal rows are composed of enlarged procumbent cells (Fig. 17E); rays homogeneous in shape (fusiform) and size (Fig. 17C), all ray cells filled with dark content. Fibres non-septate, 12-20µm in diameter (average: 17µm), (lumina 1 times the double wall thickness on average). Secretory canals few and small (Fig. 17A, D), in irregularly spaced short (to long) tangential bands up to 4-6 canals, embedded in parenchyma bands (Fig. 17G), canals possibly crushed, 30-50µm in tangential diameter (average: 40µm). DisCussion This specimen is characterized by: 1)diffuse-porous wood; 2)high density of vessels, often in groups and sometimes in clusters; 3)short fusiform rays; 4)short lines of very small secretory canals; 5)abundant parenchyma, mostly aliformconfluent and diffuse-in-aggregate; and 6) vasicentric tracheids. Like for previous specimens (from p.873), secretory canals and vasicentric tracheids are diagnostic of the Dipterocarpaceae family. Few canals are distinguishable, and it is hard to determine if they are arranged in short or long lines. Our specimen is only compatible with genera Hopea, Parashorea and Shorea as they have vessels smaller than 150µm of average diameter, not exclusively solitary vessels, no fibre-tracheids, and canals in short or long lines. A closer affinity to Shorea genus is indicated because of its homogeneous rays, common tyloses, vessel density sometimes more than 10/mm². Following the discussion of the page 882, the sections ‘Pentacme’ and ‘Shorea’ are the closest to this fossil although ‘Shorea’ is the best option with short rays sometimes weakly heterocellular, with frequent tyloses, a variable frequency of groups of vessels, sometimes short lines of canals and frequent apotracheal parenchyma. This specimen recalls: Shorea robusta C.F.Gaertn. 887 Fossil wood from Natma Formation, Myanmar (lower Miocene) GEODIVERSITAS • 2022 • 44 (28) for its abundant parenchyma, 4-5-seriate and short fusiform rays, and vessels often in groups and clusters; Shorea obtusa Wall. (ex Blume) for abundant parenchyma, rays with only one upright marginal ray cell, vessels often in group and in radially aligned clusters; Shorea siamensis Miq. for the mostly aliform parenchyma, yet without a defined arrangement, abundant diffuse to diffuse-in-aggregate parenchyma, short (yet shorter than 50 cells high) and fusiform rays. All of them have few and short uniseriate rays (often less than 10 cells high). In a lesser extend, we can find some similarities with Shorea parvifolia for aliform parenchyma forming very thin lateral lines, rays up to 5-seriate, quite short and fusiform with few marginal cells. The specimen shares all diagnostic features of fossil genus Shoreoxylon. Three fossil species with 1-5-seriate rays display similar features to the ones of our fossil (Appendix 1; Schweitzer 1958; Prakash 1965a; Ramanujam& Rao 1967; Sukiman 1977; Prakash& Bande 1980; Gurusamy& Kumarasamy 2007): S.indicum for parenchyma, rays and crystals but with bigger canals, less frequent but more commonly solitary vessels; S.posthumi has short and fusiform rays but broader (mainly 5-seriate), enlarged parenchyma cells are frequent (idioblasts) and clearly visible in figures and secretory canals are wider; S.burmensehas similar rays although slightly broader but its vessels are more solitary and canal lines are often grouped by 2-5. In a lesser extent: S.irrawaddiensis Pakash& Bande is described with larger vessels (up to 6-(7)-seriate) although mostly 3-5-seriate and solitary vessels, however the rays in the figures seem to be mainly 4-5-seriate. Even though it is probable that our fossil might belong to one of these speA B C E F G D fig. 17. — Shoreoxylon sp. 1, MNHN.F.50194: A, Ts, vessels often grouped, sometimes in clusters (white arrows), abundant parenchyma, aliform to aliformconfluent, short lines and diffuse secretory canals (black arrows); B, C, Tls, 1-5-seriate rays, tyloses (arrow); D, Ts, zoom in on secretory canals in short lines and vessels in clusters with vasicentric tracheids or small vessels; E, Rls, weakly heterocellular rays with one row of larger procumbent cells (white arrow) or square to upright cells (black arrow); F, Tls, vasicentric tracheid (black arrow) and vasicentric parenchyma (white arrow); G, Ts synthetic drawing of the transverse section (only a few vessels and rays are displayed); secretory canals are displayed as red dots, very few are preserved. Scale bars: A, 1 mm; B, F, 200 µm; C, E, 500 µm; D, 250 µm; G, 1 cm. Abbreviations: see Fig. 2. 888 GEODIVERSITAS • 2022 • 44 (28) Gentis N. et al. cies, their descriptions are overlapping, often incomplete or lack clear figures; they do not allow us to attribute a clear identification to our specimen with confidence, nor to create a new species. As a consequence, we attribute this fossil to Shoreoxylon sp. with a noted resemblance with S.indicum, S.posthumi or S.burmense. Shorea parvifolia is common in dipterocarp forests up to 1100m altitude (Soerianegara& Lemmens 1993; Tropical Plants Database 2014-onward; Barstow 2018b); Shorea obtusa lives in dry lowland deciduous dipterocarp forests, in savannas and in monsoonal forests with a marked dry season and waterlogged periods, up to 1000m altitude (Soerianegara& Lemmens 1993; Ghazoul 2016); Shorea robusta is a common semi-deciduous tree in South Asia in areas with a dry season lasting 4 to 8months (a monsoon climate). Thus, it is mainly found in dry deciduous forests and savannas, but also in evergreen moist forests on well-drained soil and riverbanks. It is usually found below 800m altitude (Wu et al. 2007; Timilsina et al. 2007; Orwa et al. 2009). Shoreoxylon sp. 2 (Fig. 18) Material. — MNHN.F.50195 (field number: 19NAT07-2). Estimated minimal diameter: 73-105cm. loCality. — Kalewa Township, Sagaing Region, Myanmar. age. — Upper lower to lowermost middle Miocene. DesCription Wood diffuse-porous. Growth rings indistinct or delimited by marginal parenchyma (Fig. 18A). Vessels 80% solitary or in groups of 2-3, occasionally small clusters (Fig. 18A); they display an oblique tendency, round to oval, 3-9 per mm² (average: 6); tangential diameter 100-270µm (average: 190µm). Tylose present (Fig. 18I). Vessel elements 180-400µm (average: 280µm) long. Perforation plates simple. Intervessel pits alternate, vestured, 5-8µm (Fig. 18J). Vessel-ray pits simple or minutely bordered, ovoid in shape, 8-12µm in diameter (only 5 were observed) (Fig. 18K). Vasicentric tracheids present (Fig. 18I). Parenchyma vasicentric (sheath of 2-7 cells) and frequently aliform, also diffuse, and sometimes diffusein-aggregate in short lines mostly starting from the edge of paratracheal parenchyma (Fig. 18A, B); seemingly marginal lines that could contain secretory canals (Fig. 18A, G); sometimes crystals in chambered cells (up to 12 per strands) (Fig. 18H), mostly observed in the diffuse parenchyma; parenchyma cells 50-130µm long (average: 100µm), 12-19µm wide (average: 16µm) in tangential section; 4-5 parenchyma cells per strand. Xylem rays 1to 5seriate (mainly 4) (Fig. 18C, E), uniseriate short (< 10 cells) and about 19% of the rays; 7-9 rays per mm (average: 8), 100-850µm (average: 430µm) or 6-40 cells high, heterocellular made of procumbent cells with mostly 1 upright cell at the ends, sometimes more (Fig. 18F). Fibres non-septate, thick to very thick-walled (lumina from 0.5 to almost 0 times the double wall thickness in average). Secretory canals in long tangential bands (more than 30 canals) surrounded by parenchyma (Fig. 18D, G), 40-85µm in tangential diameter (average 60µm). DisCussion This wood is characterized by: 1)diffuse-porous wood; 2)mostly solitary vessels with an oblique tendency; 3)heterocellular 1-5-seriate rays that are short (< 1000µm); 4)small canals (<100µm) in long tangential lines; 5)vasicentric tracheids; and 6) aliform and diffuse parenchyma. As for previous specimens (from p.873), secretory canals and vasicentric tracheids are diagnostic of the Dipterocarpaceae family. Following the discussion p. 882, these features recall the genus Shorea, together with the presence of groups of vessels and clusters, the absence of vascular tracheids, and crystals in ray cells. More specifically with the sections ‘Rubroshorea’ and ‘Shorea’ for having frequent tyloses, crystals in parenchyma, rays with few marginal cells and small canals. Regarding rays, the section ‘Shorea’ has shorter ones (under 1000µm) and less marginal cells, even though this observation is not a general fact. Our specimen recalls extant species S.laevis for its aliform parenchyma, similar ray width and length (mostly 3-5-seriate and up to 45 cells high), and small canals and S.balangeran Dyer for its wide vessels (sometimes >200µm) often solitary or in very small groups, its long and distinctive lines of canals, its rays of varying length and its aliform parenchyma sometimes forming a discrete network without particular arrangement between vessels. The specimen shares all diagnostic features of fossil genus Shoreoxylon and shares most features of Shoreoxylon sp. 1 (p.886) and Shoreoxylon cf. deomaliense (p.882); it is yet clearly distinguishable from our two previous specimens. It has more solitary and less numerous vessels than Shorexylon sp. 1 as well as less confluent parenchyma and clear long lines of canals; it has more grouped vessels and shorter, less seriated rays than Shoreoxylon cf. deomaliense. Among Shoreoxylon species that are close to our fossil (Appendix 1; Den Berger 1923; Schweitzer 1958; Prakash 1965a; Prakash& Awasthi 1970; Awasthi 1974; Sukiman 1977; Trivedi& Ahuja 1979; Prakash& Bande 1980), two species share similar features: S.burmense has slightly larger rays and multiple bands of canals, mostly vasicentric parenchyma, but it has very thick walls; S.tipamense shares similar vessel and crystalliferous parenchyma arrangement in addition to 1-5-seriate rays up to 66 cells high, but has longer rows of marginal ray cells (1 to 12), some sheath cells, larger intervessel pits (8-10µm compared to 5-8µm) and thinner walls. Considering the poor preservation of the specimen, it is attributed to Shoreoxylon sp. 2, with noted resemblance with S.tipamense or S.burmense. fig. 18. — Shoreoxylon sp. 2, MNHN.F.50195: A, Ts, vessel arrangement, often solitary, sometimes in groups and clusters, with one parenchyma band (arrow); B, Ts, mostly vasicentric to slightly aliform parenchyma as well as some diffuse parenchyma (arrow); C, E, Tls, 1-5-seriate rays; D, Ts, long tangential line of secretory canals embedded in parenchyma band (arrow); F, Rls, heterocellular ray with one line of upright marginal cells (arrow); G, Ts, synthetic drawing of 889 Fossil wood from Natma Formation, Myanmar (lower Miocene) GEODIVERSITAS • 2022 • 44 (28) A D E F J K IHG B C the transversal section (only some vessels and rays are displayed) with bands of parenchyma as grey lines and secretory canals as red dots, arranged in long tangential lines; doubtful canals are displayed as orange dots; H, Rls, crystals in possibly chambered parenchyma cells (arrow); I, tyloses in vessels (black arrow) and vasicentric tracheids (white arrow) J, Tls, alternate, vestured (arrow) intervessel pits; K, simple vessel-ray pits with reduced borders and ovoid shape (arrow). Scale bars: G, 1 cm; A, D-E, 500 µm; B-C, F, 200 µm; H-I, 100 µm; K, 50 µm; J, 20 µm. Abbreviations: see Fig. 2. 890 GEODIVERSITAS • 2022 • 44 (28) Gentis N. et al. Shorea laevis mostly grows on tropical ridges or hillsides up to 1000m altitude; it is also found in lowland mixed dipterocarp forests and on alluvial sites (Ashton 1982; Soerianegara& Lemmens 1993; Pooma et al. 2017). Shorea balangeran is common in peat-swamp forests up to 100m altitude (Soerianegara& Lemmens 1993; Robiansyah 2020). Family burseraCeae Kunth Genus Burseroxylon (Prakash& Tripathi) Lakhanpal, Prakash& Awasthi Burseroxylonsp. (Fig. 19A-J) Material. — MNHN.F.50196 (field number: 17FN05), MNHN.F.50197 (field number: 17FN11), MNHN.F.50198 (field number: NAT17-01). Estimated minimal diameter: > 30cm, as rays are almost parallel. loCality. — Kalewa Township, Sagaing Region, Myanmar. age. — Upper lower to lowermost middle Miocene. DesCription Wood diffuse-porous. Growth rings indistinct or absent. Vessels mostly solitary (75%) or in radial groups of 1-4 (25%) (Fig. 19A), oval to round, 2-13 per mm² (average: 7); tangential diameter 90-280µm (average: 190µm). Tyloses present in most vessels (Fig. 19A, D), budding from ray parenchyma cells. Vessel elements 130-540µm long (average: 310µm). Perforation plates simple. Intervessel pits alternate, polygonal shaped, 7-10µm wide (average 8.5µm). Vesselray pits apparently simple with different shapes and sizes, mostly horizontal (5-20µm, 13µm in average) (Fig. 19E). Parenchyma exclusively paratracheal, scanty and vasicentric forming a 1-cell sheath around vessels (Fig. 19A, D); parenchyma cells 45-90µm long (average: 60µm), 25-45µm wide (average: 35µm) in tangential section; some crystals are present in chambered parenchyma cells in the specimen MNHN.F.50197 Rays 1to 4-seriate, mostly 3or 2-seriate, always more than 50% of the rays are 3-seriate, rarely uniseriate, (Fig. 19B), non-storied, 3-12 rays per mm (average: 7), 150-640µm (average: 380µm) or 4-23 cells high, heterocellular made of procumbent cells with 1(-3) square or upright cells at both ends (Fig. 19C, F, G); crystals sometimes present in upright or enlarged marginal cells (Fig. 19F), maybe in procumbent cells as well (Fig. 19G). Fibres thin-to-thick to thin-walled (lumina 2.5 times the double wall thickness in average), almost all fibres are septate (Fig. 19B), tangential diameter 10-30µm (average: 18µm). Traumatic canals present (specimens MNHN.F.50197 and MNHN.F.50198) as holes of different size and shape surrounded by clusters of undifferentiated cells full of black content (Fig. 19H-J). The specimen MNHN.F.50198 displays an enlarged radial line of undifferentiated tissue embedding some rays that recalls a healed crack filled with tissue during the growth of the wood (Fig. 19J) that ends with a line of traumatic canals (Fig. 19I). Radial canals absent. DisCussion These specimens are characterized by: 1)diffuse-porous wood with isolated and tylosed vessels; 2)exclusively simple perforation plates; 3)septate fibres which although are not always very well preserved are clearly present throughout the whole sample; 4)exclusively vasicentric parenchyma; 5)growth rings indistinct or absent; and (6) 2to 4-seriate heterocellular rays. These features suggest affinities with Burseraceae as well as some Anacardiaceae (Metcalfe& Chalk 1950; InsideWood 2004-onward; Ogata et al. 2008). Both families share several similar anatomical features. However, the presence of septate fibres is more consistent in Burseraceae, as well as a greater homogeneity in rays and parenchyma patterns, and the presence of crystals in enlarged ray cells. Moreover, traumatic canals are not reported in Anacardiacae, whereas they are reported in Burseraceae (Metcalfe& Chalk 1950). One species of Anacardiaceae resembles our specimens: Lannea coromandelica (Houtt.) Merr., which has similar vessel, ray and parenchyma patterns, crystals in normal and enlarged marginal cells, in parenchyma and in procumbent cells; however, silica bodies are present in its ray cells and it is always described with radial canals (InsideWood 2004-onward; Gupta& Agarwal 2008), which are absent in our fossils. The three present specimens show slight differences in vessel and ray density, ray height and width, and crystal presence (the specimens MNHN.F.50197 has slightly shorter rays, more frequent 2-seriate rays, crystals in non-enlarged marginal cells, in parenchyma cells and possibly in procumbent ray cells) but the range of these variations is within the interand intraspecific diversity of Burseraceae (Metcalfe& Chalk 1950; InsideWood 2004-onward; Ogata et al. 2008). The wood anatomy of Burseraceae is very homogenous, both among extant and fossil species (Awasthi& Srivastava 1989; Prasad 1993; Ogata et al. 2008). Among Burseraceae, our fossils resemble the genera Canarium L. and Protium Burm.f. which display homogeneous 1-3-(up to 4-)seriate rays and sometimes traumatic canals (Metcalfe& Chalk 1950). The examination of Burseraceae from the Xylarium of the MNHN highlighted a resemblance between the fossil specimens MNHN.F.50196 and MNHN.F.50198 with Canarium bengalense Roxb. (specimen MNHN-P-P00395579), because of crystals only present in enlarged marginal ray cells, similar vessel size and frequency (sometimes < 5 / mm² and over 100µm of diameter) and ray pattern (1-3-seriate, 3-16 cells high). Ogata et al. (2008) mention that the genus Canarium can have rays up to 4-seriate. The fossil specimen MNHN.F.50197 is closer to Protium serratum (Wall. ex Colebr.) Engl. (specimens MNHN-PP00396706 and MNHN-P-P00396663) as it displays crystals in normal marginal cells as well as sometimes in procumbent cells and in parenchyma cells, smaller and more frequent rays than Canarium (sometimes > 10/mm2 and under 100µm of diameter). Protium serratum is described with and without radial canals (InsideWood 2004-onward). Burseraceae are not numerous in the fossil record (Gregory et al. 2009) and include two dominant fossil genera: 1)Burseroxylon which includes fossils related to Bursera Jacq. ex L. (and by extension Protium) and some Garuga Roxb. (Lakhanpal et 891 Fossil wood from Natma Formation, Myanmar (lower Miocene) GEODIVERSITAS • 2022 • 44 (28) A B C GFD E H I J fig. 19. — Burseroxylon sp., MNHN.F.50196 (C), MNHN.F.50197 (A, B, D-E, G, H), MNHN.F.50198 (F, I, J): A, D, Ts, vessels with tyloses (white arrow) and thin sheath of vasicentric parenchyma (black arrow); B, Tls, (1)-4-seriate rays, septate fibres (arrow); C, Rls, heterocellular rays with upright marginal cells; E, Rls, simple vessel-ray pits, oval in shape; F, Rls, heterocellular rays with upright marginal cells, some of them are enlarged and contain a single crystal (arrow); G, Rls, crystals in upright marginal cells (white arrow), maybe also in procumbent cells sometimes (black arrow); H, I, Ts, traumatic canals (arrows); J, Tls, tangential view of undifferentiated tissue surrounding a traumatic canal. Scale bars: A, 1 mm; B-D, F-H, J, 200 µm; E, 50 µm; I, 500 µm. Abbreviations: see Fig. 2. 898 GEODIVERSITAS • 2022 • 44 (28) Gentis N. et al. Formation (Shoreoxylon sp. 1) relates to these seasonal and dry forests. Burmese fossil dipterocarps seem thus to have undergone a change of diversity and distribution area through the Cenozoic. These results support the hypothesis that Dipterocarpaceae appeared first in seasonal and open ecosystems and later specialized to tropical wet conditions, based on flower and fruits studies as well as phylogenetic studies (Ashton et al. 1988; Maury-Lechon& Curtet 1998; Gunasekara 2004; Kurten et al. 2018; Bansal et al. 2022). It has been proposed that Dipterocarpaceae originated from Gondwana (tropical Africa) during the midor lateCretaceous and latter spread to eastern regions, reaching India during the Late Maastrichtian and Palaeocene (Madagascar, India, South-East Asia; Appanah& Turnbull 1998; Maury-Lechon& Curtet 1998; Morley 2000; Gunasekara 2004; Dutta et al. 2011; Bansal et al. 2022), potentially up to eastern Europe in the Eocene (Poole 1993; Morley& Ashton in Ashton 2014), although the attribution of European fossils has been disputed (Morley 2018). Regardless of the actual geographical origin of the dipterocarps, our results show that dipterocarp-dominated evergreen forests have been present in Southeast Asia since at least the late early Miocene, corroborating the long-standing feature of regional evergreen forests (Morley 2000). palaeoenvironMental iMpliCations The presence of wet evergreen forest taxa in the Miocene fossil wood assemblage indicates that the ecosystem in the central Myanmar drainage basin was likely wetter than in the Eocene. Further evidence for wetter environments is supported by the fossil wood morphological characteristics. According to Wheeler& Dillhoff (2009), trees occurring in modern lowland tropical rainforests often show few (< 10 vessels per sq mm) and wide vessels (>200µm). Only 4 out of 18 fossil species described here have wide vessels, but almost all of them have few vessels, unlike Pondaung fossil wood specimens, supporting higher humidity than in the Eocene. Three mechanisms could explain this wetter assemblage: 1)changes of drainage that would favor wetter areas; 2)increased summer monsoonal rainfall; and 3)taphonomic bias in our Miocene sampling that would favor higher-altitude or wetter taxa. Asexplained in the previous subsection, there is no evidence for a bias toward higher altitudes in the Natma Formation assemblage; some fossil specimens in the Pondaung Formation suggest higher elevations in the Eocene than in the Miocene. Several studies have shown that central Myanmar river drainages experienced significant restructuring during the Oligocene based on changes of sedimentary provenance in nearby basins (Licht et al. 2016; Zhang et al. 2019; Westerweel et al. 2020). None of these studies specifically addressed the provenance history of the Natma Formation. Still, they showed no evidence for long-distance sediment transport into Myanmar, far outside the modern drainage basin of central Myanmar. Central Myanmar river drainages must thus have been local and restricted to the modern Burmese low plains, the surrounding highlands of intermediate elevation, and the higher highlands of the Eastern Himalayan Syntaxis (Fig. 1; Zhang et al. 2019; Westerweel et al. 2020). The wetter ecosystems with no evidence for high elevation nor long-distance transport thus suggest a wetter climate in central Myanmar during the Miocene compared to the Eocene, with the majority of the taxa suggesting more than 2000mm of annual rainfall. This interpretation is supported by stable isotopic data from Natma pedogenic carbonates, which suggest a wetter-than-today monsoonal regime (Licht et al. 2022). More broadly, this interpretation agrees with data from other areas in South Asia indicating strong monsoons during the late early to early middle Miocene (Clift et al. 2008). CONCLUSION This study details the first fossil wood assemblage from the upper lower to lowermost middle Miocene of Myanmar. This assemblage is dominated by Fabaceae (7 out of 18 species) and Dipterocarpaceae (8 out of 18 species), similar to the diversity found nowadays in Southeast Asian forests. We show that fossil taxa can be divided into three different types of low altitude (mostly < 1000m) forests: tropical wet evergreen, tropical dry and deciduous, and tropical littoral ecosystems. The presence of taxa from these various ecosystems indicates some degree of mixing and transportation, though Miocene river drainages were likely local. The assemblage related to tropical wet evergreen forests is the most diverse and includes most of the fossil dipterocarps. The presence of these combined ecosystems indicates a monsoonal climate during the upper lower to lowermost middle Miocene with an alternance of a dry season and a wet season, with many taxa suggesting more than 2000mm of annual rainfall. These ecosystems appear wetter than Eocene Burmese ecosystems that yet yielded fossil dipterocarps. These results support an evolution of the distribution of Dipterocarpaceae from seasonal to everwet ecosystems and an increase of summer monsoon rainfall in an overall Miocene climate change toward warmer and wetter conditions. Acknowledgements This work was funded with the European Research Council consolidator grant MAGIC 649081. We thank V. Rommevaux (CNRS, UMR 7207) for preparing wood sections. We also acknowledge D. Pons (SU, UMR 7207), C. Privé-Gill (SU, UMR 7207), and M. Laudoueneix-Dupéron (SU, UMR 7207) for their comments and expertise on the softwood specimen. The precious work of M. Tengberg (MNHN, UMR 7209) at the Xylarium of the MNHN is gratefully thanked. We also thank N. Boonchai (FLMNH, UF) and one anonymous reviewer for their constructive comments that greatly enhanced this manuscript. 899 Fossil wood from Natma Formation, Myanmar (lower Miocene) GEODIVERSITAS • 2022 • 44 (28) REFERENCES ali s. i. 1973. — Caesalpiniaceae. Vol. 54. Karachi, Dept. of Botany, Univeristy of Karachi, 47 p. (Flora of Pakistan). http:// www.worldcat.org/oclc/2349346 asFn (ASEAN Social Forestry Network) 2016.— Country Brief: Myanmar. ASFN: 2. appanah s. & turnbull J. M. (eds) 1998. — A Review of Dipterocarps: Taxonomy, Ecology and Silviculture. Center for International Forestry Research (CIFOR), Bogor, 220p. https:// doi.org/10.17528/cifor/000463 ashton p. s. 2004. — Dipterocarpaceae, in soepaDMo l. g., saw l. g. & Chung r. C. K. (eds), Tree Flora of Sabah and Sarawak Vol.5. Forest Research Institute, Kuala Lumpur: 63-388. ashton p. s. 1982. — Dipterocarpaceae, in van steenis C. G. G.J. (ed.), Flora Malesiana, Series 1, Spermatophyta 9 (2): 237552. https://repository.naturalis.nl/pub/532631 ashton p. s. 2014. — On the Forests of Tropical Asia: Lest the Memory Fade. Kew Publishing, Kew, 672p. ashton p. s., givnish t. J. & appanah s. 1988. — Staggered flowering in the Dipterocarpaceae: new insights into floral induction and the evolution of mast fruiting in the aseasonal tropics. The American Naturalist 132 (1): 44-66. https://doi. org/10.1086/284837 aung l. l., Zin e. e., theingi p., elvera n., aung p. p., han t. t., oo y. & skalanD r. g. 2017. — Myanmar Climate Report no. 9/2017. Norwegian Meteorological Institute: 105. a wasthi n. 1967. — Fossil wood resembling that of Millettia from the Tertiary of South India. Current Science 36 (7): 180181. https://www.jstor.org/stable/24063269 a wasthi n. 1971. — Revision of some dipterocarpaceous woods previously described from the Tertiary of south India. The Palaeobotanist 18 (3): 226-233 (1969). https://doi.org/10.54991/ jop.1969.840 awasthi n. 1974. — Occurance of some dipterocarpaceous woods in the Cuddalore series of South India. The Palaeobotanist 21 (3): 339-351 (1972). https://doi.org/10.54991/jop.1972.1496 awasthi n. 1977. — Revision of Hopeoxylon indicum Navale and Shoreoxylon speciosum Navale from the Cuddalore series near Pondicherry. The Palaeobotanist 24 (2): 102-107 (1975). https://doi.org/10.54991/jop.1975.975 awasthi n. 1979. — Three new leguminous woods from the Cuddalore series near Pondicherry. The Palaeobotanist 26 (2): 157-166 (1977). https://doi.org/10.54991/jop.1977.1056 awasthi n. 1980. — Two new Dipterocarpaceous woods from the Cuddalore series near Pondicherry. The Palaeobotanist 26 (3): 248-258 (1977). https://doi.org/10.54991/jop.1977.1068 awasthi n. 1992. — Indian fossil legumes, in herenDeen p. s. & D ilCher D. l. (eds), Advances in Legume Systematics. Vol.4. Royal Botanic Gardens, Kew: 225-250. awasthi n. & ahuJa M. 1982. — Investigations of some carbonated woods from the Neogene of Varkala in Kerala coast. Geophytology 12 (2): 245-259. a wasthi n. & M ehrotra r. C. 1990. — Some fossil woods from Tipam Sandstone of Assam and Nagaland. The Palaeobotanist 38: 277-284 (1989). https://doi.org/10.54991/jop.1989.1662 awasthi n. & prakash u. 1987. — Fossil woods of Kingiodendron and Bauhinia from the Namsang beds of Deomali, Arunachal Pradesh. The Palaeobotanist 35 (2): 178-183 (1986). https:// doi.org/10.54991/jop.1986.1526 awasthi n. & srivastava r. 1989. — Canarium palaeoluzonicum, a new fossil wood from the neogene of Kerala with remarks on the nomenclature of fossil woods of Burseraceae. The Palaeobotanist 37 (2): 173-179 (1988). https://doi.org/10.54991/ jop.1988.1615 ayyappan n. & kokilavani v. 2020. — Canarium bengalense Roxb. Available from https://indiabiodiversity.org/species/ show/280052 [accessed 18 February 2022] ball o. M. 1930. — A partial revision of fossil forms of Artocarpus. Botanical Gazette 90 (3): 312-325. https://www.jstor. org/stable/2471090 banCroFt h. 1933. — A contribution to the geological history of the Dipterocarpaceae. Geologiska Föreningen i Stockholm Förhandlingar 55 (1): 59-100. https://doi.org/10.1080/11035893309448839 banCroFt h. 1935. — Some fossil dicotyledonous woods from Mount Elgon, East Africa. I. American Journal of Botany 22 (2): 164-183. https://doi.org/10.2307/2436007 banDe M. b. & prakash u. 1980. — Fossil woods from the Tertiary of West Bengal, India. Geophytology 10 (2): 146-157. banks h. & gasson p. 2000. — Pollen morphology and wood anatomy of the Crudia group (Leguminosae, Caesalpinioideae, Detarieae). Botanical Journal of the Linnean Society 134 (1-2): 19-59. https://doi.org/10.1111/j.1095-8339.2000.tb02344.x bansal M., Morley r. J., nagaraJu s. k., Dutta s., Mishra a. k., selveraJ J., kuMar s., niyolia D., harish s. M., abDelrahiM o. b., hasan s. elDin, raMesh b. r., DayananDan s., Morley h. p., ashton p. s. & prasaD v. 2022. — Southeast Asian Dipterocarp origin and diversification driven by Africa-India floristic interchange. Science 375 (6579): 455-460. https://doi.org/10.1126/science.abk2177 baretta-kuipers t. 1981. — Wood anatomy of Leguminosae: its relevance to taxonomy, in polhill r. & raven P. (eds), Advances in Legume Systematics. Vol.2. Royal Botanic Gardens, Kew: 677-705. b arstow M. 2018a. — Dryobalanops oblongifolia. IUCN Red List of Threatened Species 2018: e.T63023A68070307. https://doi. org/10.2305/IUCN.UK.2018-1.RLTS.T63023A68070307.en barstow M. 2018b. — Shorea parvifolia. IUCN Red List of Threatened Species 2018: e.T63045A68074947. https://doi. org/10.2305/IUCN.UK.2018-1.RLTS.T63045A68074947.en b arstow M. & r anDi a. 2018. — Dryobalanops aromatica (errata version published in 2020). IUCN Red List of Threatened Species 2018: e.T61998024A173026192. https://doi.org/10.2305/ IUCN.UK.2018-1.RLTS.T61998024A173026192.en beCk h. e., ZiMMerMann n. e., MCviCar t. r., vergopolan n., berg a. & wooD e. F. 2018. — Present and future Köppen-Geiger climate classification maps at 1-km resolution. Scientific Data 5 (1): 180214. https://doi.org/10.1038/ sdata.2018.214 benDer F. 1983. — Geology of Burma. Gebrüder Borntraeger, Berlin, Stuttgart, 93 p. b erg C. C., C orner e. J. h. & J arrett F. M. 2006. — Moraceae genera other than Ficus. Flora Malesiana, Series 1, Spermatophyta 17 (1): 1-146. https://repository.naturalis.nl/pub/579344 biswas a., khan M. a. & bera s. 2019. — Occurrence of Dryobalanops Gaertn. (Dipterocarpaceae) in the late Miocene of Bengal basin, India and biogeography of the genus during the Cenozoic of Southeast Asia. Botany Letters 166 (4): 434-443. https://doi.org/10.1080/23818107.2019.1672102 boDos v., haMiDi a., Juiling s., Maryani a. & tanggaraJu s. 2019. — Dryobalanops lanceolata. IUCN Red List of Threatened Species 2019: e.T33164A68070258. https://doi.org/10.2305/ IUCN.UK.2019-3.RLTS.T33164A68070258.en boonChai n. 2008. — The study of the biodiversity and comparative anatomy of petrified wood in the area of the Noartheastern Research Institude of Petrified Wood and Mineral Resources, Thailand. Master’s Thesis, Suranaree University of Technology, Nakhon Ratchasima, 235 p. boureau e. 1952. — Contribution à l’étude paléoxylologique de l’Indochine. V– Bois fossiles de l’Annam et du Cambodge. Mémoires du Muséum national d’Histoire naturelle, Série C, Sciences de la Terre 2 (1): 1-26. https://www.biodiversitylibrary. org/page/58166728 Carlquist s. 2001. — Comparative Wood Anatomy. Springer, Berlin, Heidelberg, 448 p. (Springer Series in Wood Science). https://doi.org/10.1007/978-3-662-04578-7 900 GEODIVERSITAS • 2022 • 44 (28) Gentis N. et al. Chavasseau o., ChaiManee y., tun s. t., soe a. n., barry J. C., M aranDat b., s uDre J., M arivaux l., D uCroCq s. & Jaeger J.-J. 2006. — Chaungtha, a new middle Miocene mammal locality from the Irrawaddy Formation, Myanmar. Journal of Asian Earth Sciences 28 (4): 354-362. https://doi.org/10.1016/j. jseaes.2005.10.012 C hoo l. M., F orest F., w ieringa J. J., b runeau a. & D e la e strella M. 2020. — Phylogeny and biogeography of the Daniellia clade (Leguminosae: Detarioideae), a tropical tree lineage largely threatened in Africa and Madagascar. Molecular Phylogenetics and Evolution 146: 106752. https://doi.org/10.1016/j.ympev.2020.106752 ChowDhury k. & ghosh s. 1946. — On the anatomy of Cynometroxylon indicum gen. et sp. nov., a fossil Dicotyledonous wood from Nailalung, Assam. Proceedings of the National Academy of Sciences of India 12 (8): 435-447. ChowDhury k. & ghosh s. 1958. — Indian Woods: Their Tdentification, Properties and Uses. Vol.1. Manager of Publications, Government of India, Dehli, 304p. http://www.worldcat.org/ oclc/898795910 ChowDhury k., ghosh s. & kaZMi M. h. 1960. — Pahudioxylon bankurensis gen. et sp. nov. – A fossil wood from the Miocene bed of Bankura district, West Bengal, India. Proceedings of the National Academy of Sciences of India 26B (1): 22-29. ChowDhury k. & tanDon k. n. 1964. — A fossil wood of Terminalia tomentosa W.& A. from the Tertiary of Burma. Annals of Botany 28 (3): 445-450. https://doi.org/10.1093/oxfordjournals. aob.a083904 CliFt p. D., hoDges k. v., heslop D., hannigan r., van long h. & Calves g. 2008. — Correlation of Himalayan exhumation rates and Asian monsoon intensity. Nature Geoscience 1 (12): 875880. https://doi.org/10.1038/ngeo351 D e F ranCesChi D., h oorn C., a ntoine p.-o., C heeMa i. u., F lynn l. J., linDsay e. h., Marivaux l., Métais g., raJpar a. r. & welCoMMe J.-l. 2008. — Floral data from the mid-Cenozoic of central Pakistan. Review of Palaeobotany and Palynology 150 (1-4): 115-129. https://doi.org/10.1016/j.revpalbo.2008.01.011 De la estrella M., Forest F., klitgårD b., lewis g. p., MaCkinDer b. a., De queiroZ l. p., wieringa J. J. & bruneau a. 2018. — A new phylogeny-based tribal classification of subfamily Detarioideae, an early branching clade of florally diverse tropical arborescent legumes. Scientific Reports 8 (1): 6884. https://doi. org/10.1038/s41598-018-24687-3 D e l aubenFels D. J. 1988. — Cupressaceae. Flora Malesiana, Series1, Spermatophyta 10 (3): 442-447. https://repository.naturalis.nl/ pub/532508 Delteil-Desneux F. 1980. — Sur l’association de trois bois fossiles dans un même gisement tertiaire du Djebel Nara (Tunisie centrale). Review of Palaeobotany and Palynology 31: 289-310. https://doi. org/10.1016/0034-6667(80)90032-9 Den berger l. g. 1923. — Fossiele houtsoorten uit het Tertiair van Zuid-Sumatra. Verhandelingen van het Koninklijk Nederlands Geologisch Mijnbouwkundig Genootschap. Geologische serie7: 143-148. Den berger l. g. 1927. — Unterscheidungmerkmale von rezenten und fossilen Dipterocarpaceen gattungen. Bulletin du Jardin botanique de Buintenzorg 3: 495-498. Du n. 1988a. — Fossil wood from the late Tertiary of Burma. Proceedings of the Koninklijke Nederlandse Akademie van Wetenschappen, B 91 (3): 213-236. Du N. 1988b. — On some silicified woods from the Quaternary of Indonesia. Proceedings of the Koninklijke Nederlandse Akademie van Wetenschappen, B 91 (4): 339-361. D uFraisse a., b arDin J., p iCornell -g elabert l., C oubray s., garCía-MartíneZ M. s., leMoine M. & vila Moreiras s. 2020. — Pith location tool and wood diameter estimation: Validity and limits tested on seven taxa to approach the length of the missing radius on archaeological wood and charcoal fragments. Journal of Archaeological Science: Reports 29: 102166. https://doi. org/10.1016/j.jasrep.2019.102166 D utta s., t ripathi s. M., M alliCk M., M athews r. p., g reenwooD p. F., rao M. r. & suMMons r. e. 2011. — Eocene out-of-India dispersal of Asian dipterocarps. Review of Palaeobotany and Palynology 166 (1-2): 63-68. https://doi.org/10.1016/j.revpalbo.2011.05.002 earle C. J. 2004-onward. — The Gymnosperm Database. Available from http://www.conifers.org. energy DevelopMent Corporation (eDC) 2020.— Shorea negrosensis. IUCN Red List of Threatened Species 2020: e.T33144A68074543. https://doi.org/10.2305/IUCN.UK.2020-1. RLTS.T33144A68074543.en evans J. a., gasson p. e. & lewis g. p. 2006. — Wood anatomy of the Mimosoideae (Leguminosae). IAWA Journal Supplement 5: 1-117. http://www.worldcat.org/oclc/636075638 FAO (Food and Agriculture Organization) 2001.— Global Forest Resources Assessment 2000, Main Report no. 140. FAO Forestry Paper, FAO, 479 p. Farías F. R., Williamson J. S., Rodríguez S. V., Angeles G.& Portugal V. O. 2009. — Bark anatomy in Croton draco var. draco (Euphorbiaceae). American Journal of Botany 96 (12): 2155-2167. https://doi.org/10.3732/ajb.0900035 Farjon A. 2005. — A Monograph of Cupressaceae and Sciadopitys. Royal Botanic Gardens, Kew, 643 p. http://www.worldcat.org/ oclc/741991738 F eng x., J in J., l iao w. & o skolski a. a. 2015. — Wood of Pahudioxylon (Fabaceae) from the Miocene of Jialai-Nanbao Basin, Hainan Island, South China. International Journal of Plant Sciences 176 (5): 486-498. https://doi.org/10.1086/680554 Fengel D. 1991. — Aging and fossilization of wood and its components. Wood Science and Technology 25 (3): 153-177. https://doi. org/10.1007/BF00223468 Forest Research and Development Bureau (FRDB) 2016. — General Characteristics of Thai Timbers. Royal Forest Department, Bangkok, 263 p. (in Thai). gaMble J. S. 1902. — A Manual of Indian Timbers. Sampson Low, Marston& Company, London, 856 p. gasson p. e. 1994. — Wood anatomy of the tribe Sophoreae and related Caesalpinioideae and Papilionoideae, in Ferguson i. & tuCker S. (eds), Advances in Legume Systematics. Vol.6. Royal Botanic Gardens, Kew: 165-203. g asson p., t raFForD C. & M atthews b. 2003. — Wood anatomy of Caesalpinioideae, in k litgaarD b. b. & b runeau a. (eds), Advances in Legume Systematics. Vol.10. Royal Botanic Gardens, Kew: 63-93. gérarD J. & louppe D. 2011. — Afzelia africana Sm. ex Pers. Prota4u. Available from https://www.prota4u.org/database/protav8. asp?h=M26&t=Afzelia_africana&p=Afzelia+africana#MajorRef erences ghaZoul J. 2016. — Dipterocarp Biology, Ecology, and Conservation. Oxford, Oxford University Press. 307 p. http://doi.org/10.1093/ acprof:oso/9780199639656.001.0001 Ghosh P. K.& Roy S. K. 1982. — Fossil woods of Caesalpinioideae from the Miocene of West Bengal, India. Acta Botanica Indica 10: 50-55. ghosh s. & ghosh A. K. 1958. — Anisopteroxylon jawalamukhi sp. nov. : a new fossil record from the Siwaliks. Science and Culture 24: 238-241. ghosh s. & ghosh a. k. 1959. — Dipterocarpoxylon malavii sp. nov. : a new fossil record from the pliocene of Kutch. Science and Culture 25: 328-332. ghosh s. & kaZMi M. h. 1958. — Anisopteroxylon bengalensis gen. et sp. nov. - new fossil wood from microlithic site of West Bengal. Science and Culture 23 (9): 485-487. ghosh s. & kaZMi M. h. 1961. — Pahudioxylon sahnii sp. nov., a new fossil record from the Miocene (?) of Tripura. Science and Culture 27 (2): 96-98. gottwalD h. p. J. 1994. — Tertiäre Kieselhölzer aus dem Chindwinn-Bassin im nordwestlichen Myanmar (Birma). Documenta Naturae 86: 1-90. 901 Fossil wood from Natma Formation, Myanmar (lower Miocene) GEODIVERSITAS • 2022 • 44 (28) gottwalD h. p. J. & paraMeswaran N. 1966. — Das sekundäre Xylem der Familie Dipterocarpaceae, anatomische Untersuchungen zur Taxonomie und Phylogenie. Botanische Jarbücher fur Systematik 85 (3): 410-508. g regory M., p oole i. & w heeler e. a. 2009. — Fossil dicot wood names: An annotated list with full bibliography. IAWA Journal Supplement 6: 1-220. http://www.worldcat.org/oclc/717995010 greguss P. 1955. — Identification of Living Gymnosperms on the Basis of Xylotomy. Budapest, Akadémiai Kiadó. 263 p. http:// www.worldcat.org/oclc/996264 greguss p. 1972. — Xylotomy of the Living Conifers. Akadémiai Kiadó, Budapest, 329 p. http://www.worldcat.org/oclc/633079317 g uleria J. s. 1983. — Some fossil woods from the Tertiary of Kachchh, Western India. The Palaeobotanist 31 (2): 109-128 (1982). https://doi.org/10.54991/jop.1983.1713 guleria J. s., gupta s. s. & srivastava r. 2002. — Fossil woods from upper Tertiary sediments of Jammu region (Jammu& Kashmir) North-West India and their significance. The Palaeobotanist 50 (2-3): 225-246 (2001). https://doi.org/10.54991/ jop.2001.1825 gunasekara n. 2004. — Phylogenetic and Molecular Dating Analyses of the Tropical Tree Family Dipterocarpaceae Based on Chloroplast matK Nucleotide Sequence Data. Master’s Thesis, Concordia University, Montreal, 128 p. gupta k. M. 1935. — A review of the genus Dipterocarpoxylon of Holden, with description of a new species D.holdeni from the Irrawady system of Burma. Proceedings / Indian Academy of Sciences 1 (10): 633-639. https://doi.org/10.1007/BF03039778 gupta s. & agarwal M. 2008. — Wood anatomy of Anacardiaceae from India with special reference to systematic position of Rhus. IAWA Journal 29 (1): 79-106. https://doi.org/10.1163/2294193290000172 g urusaMy v. & k uMarasaMy D. 2007. — A new species of Shoreoxylon from the Cuddalore Series, Tamilnadu, India. Plant Archives 7 (1): 165-167. hass h. & rowe n. p. 1999. — Thin sections and wafering, in Jones t. p. & rowe n. p. (eds), Fossil Plants and Spores: Modern Techniques. The Geological Society, London: 76-81. http:// www.worldcat.org/oclc/441117426 haw s. g. 2019. — The genus Afzelia and the Belitung ship. Journal of the Royal Asiatic Society 29 (3): 505-518. https://doi. org/10.1017/S1356186319000105 herenDeen p. s. 2000. — Structural evolution in Caesalpinioideae, in herenDeen p. s. & bruneau A. (eds), Advances in Legumes Systematics. Vol.9. Royal Botanic Gardens, Kew: 45-64. hills r. 2020. — Afzelia africana. IUCN Red List of Threatened Species 2020: e.T33032A67742420. https://doi.org/10.2305/ IUCN.UK.2020-3.RLTS.T33032A67742420.en h ou D., l arsen k. & l arsen s. s. 1996. — Caesalpiniaceae (Leguminosae-Caesalpinioideae). Flora Malesiana, Series 1, Spermatophyta 12 (2): 409-730. https://repository.naturalis.nl/ pub/532557 huang h., Morley r., liCht a., Dupont-nivet g., gríMsson F., Z etter r., w esterweel J., w in Z., w a a ung D. & h oorn C. 2020. — Eocene palms from central Myanmar in a South-East Asian and global perspective: evidence from the palynological record. Botanical Journal of the Linnean Society 194 (2): 177-206. https://doi.org/10.1093/botlinnean/boaa038 huang h., péreZ-pineDo D., Morley r. J., Dupont-nivet g., philip a., win Z., aung D. w., liCht a., JarDine p. e. & hoorn C. 2021. — At a crossroads: The late Eocene flora of central Myanmar owes its composition to plate collision and tropical climate. Review of Palaeobotany and Palynology 291: 104441. https://doi.org/10.1016/j.revpalbo.2021.104441 iawa CoMMittee 1989. — IAWA list of microscopic features for Hardwood identification, in wheeler e. a., baas p. & gasson P. E. (eds). IAWA Bulletin n.s. 10 (3): 219-332. https://doi. org/10.1163/22941932-90000496 iawa CoMMittee 2004. — IAWA list of microscopic features for Softwood identification, in r iChter h. g., g rosser D., h einZ i. & gasson P. E. (eds). IAWA Journal 25 (1): 1-70. https://doi. org/10.1163/22941932-90000349 i nsiDe w ooD 2004-onward. — Published on the Internet. Available from http://insidewood.lib.ncsu.edu/search iuCn 2021. — The IUCN Red List of Threatened Species. Version 2021-3. ISSN 2307-8235. https://www.iucnredlist.org. Jaeger J.-J., ChaiManee y., taFForeau p., DuCroCq s., soe a. n., Marivaux l., suDre J., tun s. t., htoon w. & MaranDat b. 2004. — Systematics and paleobiology of the anthropoid primate Pondaungia from the late Middle Eocene of Myanmar. Comptes Rendus Palevol 3 (4): 243-255. https://doi.org/10.1016/j. crpv.2004.05.003 Jarrett F. M. 1959. — Studies in Artocarpus and allied genera, III. A revision of Artocarpus subgenus Artocarpus. Journal of the Arnold Arboretum 40 (2): 113-155. https://doi.org/10.5962/p.186026 Jarrett F. M. 1960. — Studies in Artocarpus and allied genera, IV. A revision of Artocarpus subgenus Pseudojaca. Journal of the Arnold Arboretum 41 (2): 111-140. https://doi.org/10.5962/p.185817 Jensen M. 1999. — Trees Commonly Cultivated in Southeast Asia: An Illustrated Field Guide. FAO Regional Office for Asia and the Pacific, Bangkok, 229 p. knaap-van Meeuwen M. s. 1970. — A revision of four genera of the tribe Leguminosae-Caesalpinioideae-Cynometreae in Indomalesia and the Pacific. Blumea: Biodiversity, Evolution and Biogeography of Plants 18 (1): 1-52. https://repository.naturalis. nl/pub/525616 koeppen r. & iltis h. h. 1962. — Revision of Martiodendron (Cassieae, Caesalpiniaceae). Brittonia 14 (2): 191-209. https:// doi.org/10.2307/2805226 kraMer k. 1974. — Die tertiären Hölzer Südost-Asiens (unter Ausschluß der Dipterocarpaceae) 1. Teil. Palaeontographica Abteilung B 144 (3-6): 45-181. kräusel r. 1922. — Ueber einen fossilen Baumstamm von Bolang (Java), ein Beitrag zur Kenntnis der fossilen Flora NiederlandischIndiens. Koninklijke Akademie van Wetenschappen te Amsterdam, Verslagen van de Gewone Vergaderingen der Wisen Natuurkundige Afdeeling 31; Proceedings of the Section of Sciences 25: 9-15. https:// www.biodiversitylibrary.org/page/53228365 kress w. J., DeFilipps r. a., Farr e. & kyi D. y. y. 2003.— Achecklist of the trees, shrubs, herbs, and climbers of Myanmar. Contributions from the United States National Herbarium 45: 1-590. http://www.jstor.org/stable/23493222 kurten e. l., bunyaveJChewin s. & Davies s. J. 2018. — Phenology of a dipterocarp forest with seasonal drought: Insights into the origin of general flowering. Journal of Ecology 106 (1): 126-136. https://doi.org/10.1111/1365-2745.12858 l akhanpal r. n., p rakash u. & a wasthi n. 1981. — Some more dicotyledonous woods from the Tertiary of Deomali, Arunachal Pradesh, India. The Palaeobotanist 27 (3): 232-252. (1978). https://doi.org/10.54991/jop.2001.1825 l eenhouts p. w. 1959. — Revision of the Burseraceae of the Malaysian area in a wider sense. Xa. Canarium Stickm. Blumea: Biodiversity, Evolution and Biogeography of Plants 9 (2): 275-475. https://repository.naturalis.nl/pub/525660 leguMe phylogeny working group (lpwg) 2017. — A new subfamily classification of the Leguminosae based on a taxonomically comprehensive phylogeny. Taxon 66 (1): 44-77. https://doi. org/10.12705/661.3 l eMMens r. h. M. J., s oerianegara i. & w ong w. C. (eds) 1995. — Timber Trees: Minor Commercial Timbers. Backhuys Publishers, Leiden, 655 p. (Plant resources of South-East Asia [PROSEA]; 5 [2]). https://www.sudoc.fr/046600140 leMoigne y. & beauChaMp J. 1972. — Paleoflores tertiaires de la region de Welkite (Ethiopie, province du Shoa). Bulletin de la Société géologique de France S7-XIV (1-5): 336-346. https:// doi.org/10.2113/gssgfbull.S7-XIV.1-5.336 902 GEODIVERSITAS • 2022 • 44 (28) Gentis N. et al. leo r. F. & barghoorn e. s. 1976. — Silicification of Wood. Botanical Museum Leaflets, Harvard University 25 (1): 1-47. https://doi.org/10.5962/p.295209 léonarD J. J. g. 1950. — Note sur les genres paleotropicaux Afzelia, Intsia et Pahudia (Legum.-Caesalp.). Reinwardtia 1 (1): 61-66. https://doi.org/10.14203/reinwardtia.v1i1.1039 lewis g. p., sChrire b., MaCkinDer b.a & loCk M. 2005. — Legumes of the World. Royal Botanic Gardens, Kew, 577 p. http:// www.worldcat.org/oclc/491996118 liCht a., FranCe-lanorD C., reisberg l., Fontaine C., soe a. n. & Jaeger J.-J. 2013. — A Paleo Tibet–Myanmar connection? Reconstructing the Late Eocene drainage system of central Myanmar using a multi-proxy approach. Journal of the Geological Society 170 (6): 929-939. https://doi.org/10.1144/jgs2012-126 l iCht a., b oura a., D e F ranCesChi D., D uCroCq s., a ung naing soe & Jaeger J.-J. 2014. — Fossil woods from the late middle Eocene Pondaung Formation, Myanmar. Review of Palaeobotany and Palynology 202: 29-46. https://doi.org/10.1016/j. revpalbo.2013.12.002 liCht a., boura a., De FranCesChi D., utesCher t., sein C. & Jaeger J.-J. 2015. — Late middle Eocene fossil wood of Myanmar: Implications for the landscape and the climate of the Eocene Bengal Bay. Review of Palaeobotany and Palynology 216: 44-54. https://doi.org/10.1016/j.revpalbo.2015.01.010 l iCht a., r eisberg l., F ranCe -l anorD C., n aing s oe a. & Jaeger J.-J. 2016. — Cenozoic evolution of the central Myanmar drainage system: insights from sediment provenance in the Minbu Sub-Basin. Basin Research 28 (2): 237-251. https://doi. org/10.1111/bre.12108 liCht a., Dupont-nivet g., win Z., swe h. h., kaythi M., roperCh p., ugrai t., littell v., park D., westerweel J., Jones D., poblete F., aung D. w., huang h., hoorn C. & s ein k. 2019. — Paleogene evolution of the Burmese forearc basin and implications for the history of India-Asia convergence. GSA Bulletin 131 (5-6): 730-748. https://doi.org/10.1130/B35002.1 liCht a., kelson J., bergel s., sChauer a., petersen s. v., C apirala a., h untington k. w., D upont ‐n ivet g., w in Z. & aung D. w. 2022. — Dynamics of Pedogenic Carbonate Growth in the Tropical Domain of Myanmar. Geochemistry, Geophysics, Geosystems 23 (7). https://doi.org/10.1029/2021GC009929 loo a. h. b. & tan h. t. w. 1997. — The angiosperm flora of Singapore Part 6. Caesalpiniaceae. Garden’s Bulletin, Singapore 49: 55-106. https://biostor.org/reference/140124 ly v., nanthavong k., pooMa r., luu h. t., nguyen h. n., vu v. D., hoang v. s., khou e. h. & newMan M. F. 2017.— Dipterocarpus retusus. IUCN Red List of Threatened Species 2017: e.T32400A2817693. https://doi.org/10.2305/IUCN. UK.2017-3.RLTS.T32400A2817693.en ly v., nanthavong k., pooMa r., luu h. t., nguyen h. n., barstow M., vu v. D., hoang v. s., khou e. h. & newMan M. F. 2020. — Anisoptera scaphula (amended version of 2017 assessment). IUCN Red List of Threatened Species 2020: e.T33006A177787286. https://doi.org/10.2305/IUCN. UK.2020-3.RLTS.T33006A177787286.en MäDel-angeliewa e. & Müller-stoll w. r. 1973. — Kritische Studien über fossile Combretaceen-Hölzer: Über Hölzer vom Typus Terminalioxylon G. Schönfeld mit einer Revision der bisher zu Evodioxylon Chiarugi gestellten Arten. Palaeontographica Abteilung B 142 (4-6): 117-136. ManDang y. i. & kageMori n. 2004. — A fossil wood of Dipterocarpaceae from Pliocene deposit in the West Region of Java Island, Indonesia. Biodiversitas Journal of Biological Diversity 5 (1). https://doi.org/10.13057/biodiv/d050106 M aury -l eChon g. & C urtet l. 1998. — Biogeography and evolutionary systematics of Dipterocarpaceae, in appanah s. & t urnbull J. M. (eds), A Review of Dipterocarps: Taxonomy, Ecology and Silviculture. Center for International Forestry Research (CIFOR), Bogor: 5-44. https://doi.org/10.17528/cifor/000463 Mehrotra r. C., prakash u. & banDe M. b. 1984. — Fossil woods of Lophopetalum and Artocarpus from the Deccan Intertrappean beds of Mandla District, Madhya Pradesh, India. The Palaeobotanist 32 (3): 310-320 (1983). https://doi.org/10.54991/ jop.1984.1385 M ehrotra r. C., a wasthi n. & D utta s. k. 1999. — Study of fossil wood from the upper Tertiary sediments (Siwalik) of Arunachal Pradesh, India and its implication in Paleoecological and phytogeographical interpretations. Review of Palaeobotany and Palynology 107 (3-4): 223-247. https://doi.org/10.1016/ S0034-6667(99)00029-9 MetCalFe C. r. & Chalk l. 1950. — Anatomy of the Dicotyledons Vol. 1& 2. Oxford, Clarendon Press. 1500 p. http://www. worldcat.org/oclc/489685836 Morley r. J. 2000. — Origin and Evolution of Tropical Rain Forests. Wiley, Chichester, New York, 384 p. http://www.worldcat. org/oclc/41211613 Morley r. J. 2018. — Assembly and division of the South and South-East Asian flora in relation to tectonics and climate change. Journal of Tropical Ecology 34 (4): 209-234. https://doi. org/10.1017/S0266467418000202 Müller-stoll w. r. & MäDel e. 1967. — Die fossilen Leguminosen-Hölzer. Eine Revision der mit Leguminosen verglichenen fossilen Hölzer und Beschreibungen älterer und neuer Arten. Palaeontographica Abteilung B 119 (4-6): 95-174. Mustoe g. 2017. — Wood Petrifaction: A new view of permineralization and replacement. Geosciences 7 (4): 119. https://doi. org/10.3390/geosciences7040119 naJMan y., sobel e. r., Millar i., stoCkli D. F., govin g., lisker F., garZanti e., liMonta M., veZZoli g., Copley a., Zhang p., sZyManski e. & kahn a. 2020. — The exhumation of the Indo-Burman Ranges, Myanmar. Earth and Planetary Science Letters 530: 115948. https://doi.org/10.1016/j.epsl.2019.115948 nguyen h. n., vu v. D., luu h. t., hoang v. s., pooMa r., khou e. h., nanthavong k., newMan M. F., ly v. & barstow M. 2017. — Anisoptera costata. IUCN Red List of Threatened Species 2017: e.T33166A2833752. https://doi.org/10.2305/IUCN. UK.2017-3.RLTS.T33166A2833752.en nielsen i. C. 1992. — Mimosaceae (Leguminosae-Mimosoideae). Flora Malesiana, Series 1, Spermatophyta 11 (1). https://repository.naturalis.nl/pub/532645 n ielsen i. C. 1981. — Légumineuses-Mimosoïdées. Muséum national d’Histoire naturelle, Paris, 159 p. (Flore du Cambodge, du Laos et du Viêtnam; 19). https://www.sudoc.fr/000421316 ogata k., FuJii t., abe h. & baas p. 2008. — Identification of the Timbers of Southeast Asia and the Western Pacific. Kaiseisha Press, Ōtsu-shi, 408 p. http://www.worldcat.org/oclc/1087889669 o rwa C., M utua a., k inDt r., J aMnaDass r. & s iMons a. 2009.— Agroforestree Database: a tree reference and selection guide version 4.0. World Agroforestry Centre, Kenya. Available from http://apps.worldagroforestry.org/treedb2/ pasCal J. p., raMesh b. r. & De FranCesChi D. 2004. — Wet evergreen forest types of the southern western ghats, India. Tropical Ecology 45 (2): 281-292. https://hal.archives-ouvertes. fr/hal-00021855 pearson r. & brown h. 1932. — Commercial Timbers of India: Their Distribution, Supplies, Anatomical Structure, Physical and Mechanical Properties and Uses. Vol. 1& 2. Government of India, Calcutta, 1150 p. http://www.worldcat.org/oclc/830775997 péreZ-lara D. k., estraDa-ruiZ e. & CastañeDa-posaDas C. 2019. — New fossil woods of Fabaceae from El Bosque Formation (Eocene), Chiapas, Mexico. Journal of South American Earth Sciences 94: 102202. https://doi.org/10.1016/j.jsames.2019.05.018 péreZ-lara D. k., estraDa-ruiZ e. & CastañeDa-posaDas C. 2021. — Kingiodendron and Enterolobium Eocene woods from the El Bosque formation, Chiapas, Mexico. Journal of South American Earth Sciences 111: 103477. https://doi.org/10.1016/j. jsames.2021.103477 903 Fossil wood from Natma Formation, Myanmar (lower Miocene) GEODIVERSITAS • 2022 • 44 (28) phillips e. w. J. 1948. — Identification of Softwoods by their Microscopic Structure. Her Majesty Stationery Office, London, 55p. (Forest Products Research Bulletin; 22). http://www.worldcat. org/oclc/474788365 pitterMann J., stuart s. a., Dawson t. e. & Moreau a. 2012. — Cenozoic climate change shaped the evolutionary ecophysiology of the Cupressaceae conifers. Proceedings of the National Academy of Sciences 109 (24): 9647-9652. https://doi. org/10.1073/pnas.1114378109 p ons D. 1980. — Les bois fossiles du Tertiaire supérieur de la région de Toluviejo-Corozal (Dépt. de Sucre, Colombie), in 105ème Congrès national des Sociétés savantes, Caen (Section des Sciences). Bibliothèque nationale: 163-182. poole i. 1993. — A Dipterocarpaceous twig from the Eocene London Clay Formation of Southeast England. Special Papers in Paleontology 49: 155-163 pooMa r., newMan M. F. & barstow M. 2017. — Shorea laevis. IUCN Red List of Threatened Species 2017: e.T33121A2833046. https://doi.org/10.2305/IUCN.UK.2017-3.RLTS. T33121A2833046.en powo 2019. — Plants of the World Online. Facilitated by the Royal Botanic Gardens, Kew. Available from http://www.plantsoftheworldonline.org [accessed 7 July 2022] prakash u. 1965a. — Fossil wood of Dipterocarpaceae from the Tertiary of Burma. Current Science 34 (6): 181-182. https:// www.jstor.org/stable/24062074 prakash u. 1965b. — Dipterocarpoxylon tertiarum sp. nov., a new fossil wood from the Tertiary of Burma. Current Science 34 (8): 254-255. https://www.jstor.org/stable/24063317 prakash u. 1965c. — Pahudioxylon deomaliense sp. nov., a new fossil wood from the Tertiary of Eastern India. Current Science 34 (14): 433-434. https://www.jstor.org/stable/24062930 p rakash u. 1965d. — Fossil wood of Lagerstroemia from the Tertiary of Burma. Current Science 34 (16): 484-485. https:// www.jstor.org/stable/24061761 p rakash u. 1966. — Some fossil Dicotyledonous woods from the Tertiary of eastern India. The Palaeobotanist 14 (1-3): 223-235 (1965). https://doi.org/10.54991/jop.1965.735 prakash u. 1967. — Fossil wood of Cassia and Cynometra from the Tertiary beds of Mikir hills, Assam. Publication of the Centre of Advanced Study in Geology, Punjab University 3: 93-100. p rakash u. 1973. — Fossil woods from the Tertiary of Burma. The Palaeobotanist 20 (1): 48-70 (1971). https://doi.org/10.54991/ jop.1971.887 prakash u. 1975. — Fossil woods from the Lower Siwalik beds of Himachal Pradesh, India. The Palaeobotanist 22 (3): 192-210 (1973). https://doi.org/10.54991/jop.1973.941 prakash u. 1978. — Fossil woods from the Lower Siwalik beds of Uttar Pradesh, India. The Palaeobotanist 25: 376-392 (1976). https://doi.org/10.54991/jop.1976.1025 p rakash u. 1979. — Fossil Dictyledonous woods from the Tertiary of Thailand. The Palaeobotanist 26 (1): 50-62 (1977). https:// doi.org/10.54991/jop.1977.1048 prakash u. 1981. — Further occurrence of fossil woods from the Lower Siwalik beds of Uttar Pradesh, India. The Palaeobotanist 28-29: 374-388 (1979-1980). https://doi.org/10.54991/ jop.1981.1424 prakash u. & awasthi n. 1970. — Fossil woods from the Tertiary of Eastern India. 1. The Palaeobotanist 18 (1): 32-44 (1969). https://doi.org/10.54991/jop.1969.816 prakash u. & awasthi n. 1971. — Fossil woods from the Tertiary of Eastern India, 2. The Palaeobotanist 18 (3): 219-225 (1969). https://doi.org/10.54991/jop.1969.839 prakash u. & banDe M. b. 1980. — Some more fossil woods from the Tertiary of Burma. The Palaeobotanist 26 (3): 261-278 (1977). https://doi.org/10.54991/jop.1977.1071 prakash u. & lalitha C. 1978. — Fossil wood of Artocarpus from the Tertiary of Assam. Geophytology 8 (1): 132-133. prakash u. & prasaD M. 1984. — Wood of Bauhinia from the Siwalik beds of Uttar Pradesh, India. The Palaeobotanist 32 (2): 140-145 (1983). https://doi.org/10.54991/jop.1984.1372 prakash u. & tripathi p. p. 1968. — Fossil wood of Adenanthera and Swintonia from the Tertiary of Assam. Current Science 37 (4): 115-116. https://www.jstor.org/stable/24072989 prakash u. & tripathi p. p. 1969. — Fossil woods of Leguminosae and Anacardiaceae from the Tertiary of Assam. The Palaeobotanist 17: 22-32 (1968). https://doi.org/10.54991/jop.1968.535 prakash u. & tripathi p. p. 1970. — Fossil woods from the Tipam sandstones near Hailakandi, Assam. The Palaeobotanist 18 (2): 183-191 (1969). https://doi.org/10.54991/jop.1969.833 prakash u. & tripathi p. p. 1975. — Fossil Dicotyledonous woods from the Tertiary of Eastern India. The Palaeobotanist 22 (1): 51-62 (1973). https://doi.org/10.54991/jop.1973.924 prakash u., boureau e. & louvet p. 1967. — Les plants ligneux convergents et la nomenclature de bois de Légumineuses Tertiaires du Sahara et d’Asie. Taxon 16 (6): 505-509. https://doi. org/10.2307/1216954 prakash u., b‐eZinová D. & awasthi n. 1974. — Fossil woods from the Tertiary of South Bohemia. Palaeontographica Abteilung B 147 (4-6): 107-123. prakash u., lalitha v. & tripathi p. p. 1994. — Plant remains from the Tipam sandstones of northeast India with remarks on the Paleoecology of the region during the Miocene. Palaeontographica Abteilung B 231 (1-6): 113-146. prasaD M. 1993. — Siwalik (Middle Miocene) woods from the Kalagarh area in the Himalayan foot hills and their bearing on Paleoclimate and phytogeography. Review of Palaeobotany and Palynology 76 (1): 49-82. https://doi.org/10.1016/0034-6667(93)90080-E prasaD M. & gautaM s. 2016. — Dipterocarpaceous macrofossils from Churia Group of Arjun Khola area, western Nepal and their phytogeographical and Paleoclimatical implications. The Palaeobotanist 65 (2): 247-270. https://doi.org/10.54991/jop.2016.313 prasaD M., agarwal a. & ManDaokar b. D. 2009. — New species of the genus Anisopteroxylon from the Lower Miocene sediments of Mizoram, India. Phytomorphology 59 (1-2): 1-6. privé-gill C., voZenin-serra C., DuCroCq s., naing soe a. & Jaeger J.-J. 2004. — Bois de la Formation Pondaung, Eocène moyen du Myanmar. Implications paléoenvironnementales. Palaeontographica B 267 (1-3): 57-65. purkayastha s. k., JuneJa k. b. s. & kaZMi h. s. M. 1976. — Anatomy of More Important Andaman Commercial Timbers: (with Notes on their Supply, Properties& Uses). Vol. 2. Controller of Publications, Delhi, 48p. (Indian forest records [new series], Wood anatomy; 1). http://www.worldcat.org/oclc/25375036 r aDosavlJeviC a. 2019. — The rise of Cynometra (Leguminosae) and the fall of Maniltoa: a generic re-circumscription and the addition of 4 new species. PhytoKeys 127: 1-37. https://doi.org/10.3897/ phytokeys.127.29817 raManuJaM C. g. k. 1960. — Silicified Woods from the Tertiary Rocks of South India. Palaeontographica Abteilung B 106 (4-6): 99-140. r aManuJaM C. g. k. & r ao M. r. r. 1966. — A fossil wood resembling Bauhinia from the Cuddalore Series of South India. Current Science 35 (22): 575-577. http://www.jstor.org/stable/24063970 r aManuJaM C. g. k. & r ao M. r. r. 1967. — A new species of Shoreoxylon S.krauseli sp. nov. from the Tertiary of South India. Current Science 36 (16): 439-441. https://www.jstor.org/ stable/24064157 raMos r. s., brea M. & kröhling D. M. 2017. — Fossil woods of Detarioideae subfamily (Fabaceae) from El Palmar Formation (Late Pleistocene) in South America. Journal of South American Earth Sciences 79: 202-214. https://doi.org/10.1016/j.jsames.2017.08.006 r anDi a., k usuMaDewi y., t anggaraJu s., J uiling s., b oDos v., haMiDi a. & Maryani a. 2019. — Dryobalanops keithii. IUCN Red List of Threatened Species 2019: e.T33379A143153282. https:// doi.org/10.2305/IUCN.UK.2019-3.RLTS.T33379A143153282.en 904 GEODIVERSITAS • 2022 • 44 (28) Gentis N. et al. r etallaCk g. J. 2005. — Pedogenic carbonate proxies for amount and seasonality of precipitation in paleosols. Geology 33 (4): 333. https://doi.org/10.1130/G21263.1 r iChter h. g. & D allwitZ M. J. 2000-onward. — Commercial timbers: descriptions, illustrations, identification, and information retrieval. In English, French, German, Portuguese, and Spanish. Version: 9 th April 2019. DELTA – DEscription Language for TAxonomy. Available from https://www.delta-intkey.com/ robiansyah i. 2020. — Shorea balangeran. IUCN Red List of Threatened Species 2020: e.T33103A68072336. https://doi. org/10.2305/IUCN.UK.2020-1.RLTS.T33103A68072336.en roDrígueZ-reyes o., gasson p., FalCon-lang h. J. & Collinson M. e. 2017. — Fossil legume woods of the Prioria-clade (subfamily Detarioideae) from the lower Miocene (early to mid-Burdigalian) part of the Cucaracha Formation of Panama (Central America) and their systematic and Paleoecological implications. Review of Palaeobotany and Palynology 246: 44-61. https://doi.org/10.1016/j.revpalbo.2017.06.005 s ahni b. 1964. — Revisions of Indian Fossil Plants, Part III: Monocotyledons. Birbal Sahni Institute of Palaeobotany, Lucknow 89p. (Monographs of the Birbal Sahni Institute of Palaeobotany; 1). saM h. v., nanthavong k. & kessler p. J. a. 2004. — Trees of Laos and Vietnam: A field guide to 100 economically or ecologically important species. Blumea - Biodiversity, Evolution and Biogeography of Plants 49 (2): 201-349. https://doi. org/10.3767/000651904X484298 s CheFFer t. C. & M orrell J. J. 1998. — Natural durability of wood : a worldwide checklist of species. Forest Research Laboratory, Oregon State University. Research Contribution 22: 1-58. http://hdl.handle.net/1957/7736 s ChweitZer h.-J. 1958. — Die fossilen Dipterocarpaceen-Hölzer. Palaeontographica Abteilung B 105 (1-4): 1-66. singh M. k., sharMa C. l. & sharMa M. 2017. — Comparative wood anatomy of four Artocarpus species of North East India with reference to theiridentification, in panDey k. k., r aMakantha v., C hauhan s. s. & a run k uMar A. N. (eds), Wood is Good. Springer, Singapore: 73-81. https://doi. org/10.1007/978-981-10-3115-1_8 sMitinanD t., viDal J. e. & phaM h. h. 1990. — Diptérocarpacées. Muséum national d’Histoire naturelle, Paris, 123p. (Flore du Cambodge, du Laos et du Viêtnam; 25). https:// www.sudoc.fr/001874985 soerianegara i. & leMMens r. h. M. J. (eds) 1993. — Timber trees: Major commercial timbers. Pudoc Scientific Publishers, Wageningen, 610p. (Plant resources of South-East Asia [PROSEA], 5 [1]). https://www.sudoc.fr/013093835 soseF M. s. M., hong l. t. & parwirohatMoDJo s. (eds) 1998.— Timber trees: Lesser-Known Timbers. Backhuys Publishers, Leiden, 859p. (Plant resources of South-East Asia [PROSEA]; 5 [3]). https://www.sudoc.fr/074624253 srivastava r. & awasthi n. 1996. — Fossil woods from Neogene of Warkalli beds of Kerala Coast and their Paleoecologica significance. Geophytology 26 (1): 89-98. srivastava r. & kageMori n. 2001. — Fossil wood of Dryobalanops from Pliocene deposits of Indonesia. The Palaeobotanist 50 (2-3): 395-401. https://doi.org/10.54991/jop.2001.1835 srivastava J. & prasaD v. 2019. — Evolution and paleobiogeography of mangroves. Marine Ecology 40 (6). https://doi. org/10.1111/maec.12571 sukiMan s. 1977. — Sur deux bois fossiles du gisement de la région Pachitan à Java, in 102ème Congrès national des sociétés savantes, Limoges (Section des Sciences). Bibliothèque nationale: 197-209. t anJung r. h. r., s uharno , r uMahorbo b. t., r eZa M. a. & akhMaD 2020. — Characteristics of peatland chemicals and their association with the diversity of dominant plants in Papua. IOP Conference Series: Earth and Environmental Science 575 (1): 012082. https://doi.org/10.1088/1755-1315/575/1/012082 ter welle b. J. h., koek-noorMan J. & topper s. M. C. 1986.— The systematic wood anatomy of the Moraceae (Urticales) V. Genera of the tribe Moreae without urticaceous stamens. IAWA Journal 7 (3): 175-193. https://doi.org/10.1163/22941932-90000983 tiMilsina n., ross M. s. & heinen J. t. 2007. — A community analysis of sal (Shorea robusta) forests in the western Terai of Nepal. Forest Ecology and Management 241 (1-3): 223-234. https://doi.org/10.1016/j.foreco.2007.01.012 triveDi b. s. & ahuJa M. 1978. — Cynometroxylon siwalicus n. sp. from the Siwalik Range. Current Science 47 (17): 638-639. https://www.jstor.org/stable/24082243 triveDi b. s. & ahuJa M. 1979. — Pentacmeoxylon ornatum gen. et sp. nov. from the Siwaliks of Kalagarh. Current Science 48 (14): 646-647. https://www.jstor.org/stable/24082731 triveDi b. s. & panJwani M. 1986. — Fossil wood of Bauhinia from the Siwalik Beds of Kalagarh U. P. Geophytology 16 (1): 66-69 triveDi b. s. & srivastava r. 1989. — Gymnospermous woods from early tertiary of Chhindwara district of Madhya Pradesh. Phytomorphology 39 (1): 61-68 tropiCal plants Database 2014-onward. — Ken Fern. Available from https://tropical.theferns.info. turlanD n., wierseMa J., barrie F., greuter w., hawksworth D., herenDeen p., knapp s., kusber w.-h., li D.-Z., MarholD k., May t., MCneill J., Monro a., praDo J., priCe M. & sMith g. (eds) 2018. — International Code of Nomenclature for Algae, Fungi, and Plants (Shenzhen Code). Regnum Vegetabile 159. Glashütten, Koeltz Botanical Books. https://doi. org/10.12705/Code.2018 v auDois n. & p rivé C. 1971. — Révision des bois fossiles de Cupressaceae. Palaeontographica Abteilung B 134 (1-3): 61-86. voZenin-serra C. 1981. — Les structures ligneuses Néogènes du Plateau de Di Linh (Sud-Vietnam). Palaeontographica Abteilung B 177 (5-6): 136-161. voZenin-serra C. & privé-gill C. 1989. — Bois plio-pléistocènes du gisement de Saropee plateau de Khorat, Est de la Thaïlande. Review of Palaeobotany and Palynology 60 (3): 225-254. https:// doi.org/10.1016/0034-6667(89)90045-6 voZenin-serra C. & privé-gill C. 1991. — Les terrasses alluviales pléistocènes du Mékong (Cambodge). II – Bois silicifiés hétéroxylés récoltés entre Stung-Treng et Snoul. Review of Palaeobotany and Palynology 68 (1): 87-117. https://doi. org/10.1016/0034-6667(91)90059-C voZenin-serra C. & prive-gill C. 2001. — Bois Plio-Pléistocènes du gisement de Ban Tachang (=Sarapee), Est-Thailande. Palaeontographica Abteilung B 260 (1-6): 201-212. w ang J.-g., w u F.-y., t an x.-C. & l iu C.-Z. 2014. — Magmatic evolution of the Western Myanmar Arc documented by U–Pb and Hf isotopes in detrital zircon. Tectonophysics 612-613: 97-105. https://doi.org/10.1016/j.tecto.2013.11.039 westerweel J., liCht a., Cogné n., roperCh p., Dupont-nivet g., kay thi M., swe h. h., huang h., win Z. & wa aung D. 2020. — Burma Terrane collision and northward indentation in the Eastern Himalayas recorded in the Eocene-Miocene Chindwin Basin (Myanmar). Tectonics 39 (10): e2020TC006413. https://doi.org/10.1029/2020TC006413 wheeler e. a. 2011. — Inside Wood – A Web resource for hardwood anatomy. IAWA Journal 32 (2): 199-211. https://doi. org/10.1163/22941932-90000051 wheeler e. a. & DillhoFF t. a. 2009. — The middle Miocene wood flora of Vantage, Washington, USA. IAWA Journal Supplement 7: 1-101. http://www.worldcat.org/oclc/837980866 w heeler e. a., w ieMann M. C. & F leagle J. g. 2007. — Woods from the Miocene Bakate Formation, Ethiopia. Review of Palaeobotany and Palynology 146 (1-4): 193-207. https://doi. org/10.1016/j.revpalbo.2007.04.002 w heeler e. a., g asson p. e. & b aas p. 2020. — Using the InsideWood web site: Potentials and pitfalls. IAWA Journal 41 (4): 412-462. https://doi.org/10.1163/22941932-bja10032 905 Fossil wood from Natma Formation, Myanmar (lower Miocene) GEODIVERSITAS • 2022 • 44 (28) wilF p., little s. a., iglesias a., CarMen ZaMaloa M., ganDolFo M. a., Cúneo n. r. & Johnson k. r. 2009. — Papuacedrus (Cupressaceae) in Eocene Patagonia: A new fossil link to Australasian rainforests. American Journal of Botany 96 (11): 2031-2047. https://doi.org/10.3732/ajb.0900085 williaMs e. w., garDner e. M., harris r. iii, ChaveeraCh a., pereira J. t. & Zerega n. J. C. 2017. — Out of Borneo: biogeography, phylogeny and divergence date estimates of Artocarpus (Moraceae). Annals of Botany 119 (4): 611-627. https://doi. org/10.1093/aob/mcw249 wooDCoCk D. w., Meyer h. w. & praDo y. 2017. — The Piedra Chamana fossil woods (Eocene, Peru). IAWA Journal 38 (3): 313-365. https://doi.org/10.1163/22941932-20170175 w orlD C onservation M onitoring C entre 1998. — Cynometra inaequifolia. IUCN Red List of Threatened Species 1998: e.T33201A9764872. https://doi.org/10.2305/IUCN.UK.1998. RLTS.T33201A9764872.en wu Z.-y., raven p. h. & hong D.-y. (eds) 2007. — Clusiaceae through Araliaceae. Beijing, Science Press, Beijing; Missouri Botanical Garden Press, St. Louis, 548 p. (Flora of China; 13). wu Z.-y., raven p. h. & hong D.-y. (eds) 2008. — Oxalidaceae through Aceraceae. Science Press, Beijing; Missouri Botanical Garden Press, St. Louis, 662 p. (Flora of China; 11). wu Z.-y., raven p. h. & hong D.-y. (eds) 2010. — Fabaceae. Science Press, Beijing; Missouri Botanical Garden Press, St. Louis, 577p. (Flora of China; 10). y aDav r. r. 1989. — Some more fossil woods from the Lower Siwalik sediments of Kalagarh, Uttar Pradesh and Nalagarh, Himachal Pradesh. The Palaeobotanist 37 (1): 52-62 (1988). https://doi.org/10.54991/jop.1988.1600 yang J., guo s. & wei g. 1995. — Fossil Dicotyledonous woods from the Dabie Mountains of China (II) – Burseraceae& Cornaceae. Journal of Hubei University (Natural Sciences) 17 (1): 1-5. ZaManian k., pustovoytov k. & kuZyakov y. 2016. — Pedogenic carbonates: Forms and formation processes. Earth-Science Reviews 157: 1-17. https://doi.org/10.1016/j.earscirev.2016.03.003 Zhang p., naJMan y., Mei l., Millar i., sobel e. r., Carter a., barFoD D., DhuiMe b., garZanti e., govin g., veZZoli g. & hu x. 2019. — Paleodrainage evolution of the large rivers of East Asia, and Himalayan-Tibet tectonics. Earth-Science Reviews 192: 601-630. https://doi.org/10.1016/j. earscirev.2019.02.003 ZiMMerMan e., herenDeen p. s., lewis g. p. & bruneau a. 2017. — Floral evolution and phylogeny of the Dialioideae, a diverse subfamily of tropical legumes. American Journal of Botany 104 (7): 1019-1041. https://doi.org/10.3732/ajb.1600436 Submitted on 28 January 2021; accepted on 23 February 2022; published on 29 September 2022. table 2. — Correspondences between the MNHN.F specimen numbers cited in the article and the corresponding pages on https://science.mnhn.fr. Specimen numbers Corresponding slides and URLs MNHN.F.50171 MNHN.F.50171.1, 50171.2, 50171.3, 50171.4, 50171.5, 50171.6, 50171.7 MNHN.F.50172 MNHN.F.50172.1, 50172.2, 50172.3 MNHN.F.50173 MNHN.F.50173.1, 50173.2, 50173.3 MNHN.F.50174 MNHN.F.50174.1, 50174.2, 50174.3 MNHN.F.50175 MNHN.F.50175.1, 50175.2, 50175.3, 50175.4, 50175.5 MNHN.F.50176 MNHN.F.50176.1, 50176.2, 50176.3 MNHN.F.50177 MNHN.F.50177.1, 50177.2, 50177.3 MNHN.F.50178 MNHN.F.50178.1, 50178.2, 50178.3 MNHN.F.50179 MNHN.F.50179.1, 50179.2, 50179.3 MNHN.F.50180 MNHN.F.50180.1, 50180.2, 50180.3, 50180.4 MNHN.F.50181 MNHN.F.50181.1, 50181.2, 50181.3 MNHN.F.50182 MNHN.F.50182.1, 50182.2, 50182.3 MNHN.F.50183 MNHN.F.50183.1, 50183.2, 50183.3 MNHN.F.50184 MNHN.F.50184.1, 50184.2, 50184.3, 50184.4, 50184.5, 50184.6 MNHN.F.50185 MNHN.F.50185.1, 50185.2, 50185.3 MNHN.F.50186 MNHN.F.50186.1, 50186.2, 50186.3, 50186.4 MNHN.F.50187 MNHN.F.50187.1, 50187.2, 50187.3 MNHN.F.50188 MNHN.F.50188.1, 50188.2, 50188.3, 50188.4, 50188.5, 50188.6 MNHN.F.50189 MNHN.F.50189.1, 50189.2, 50189.3, 50189.4, 50189.5, 50189.6 MNHN.F.50190 MNHN.F.50190.1, 50190.2, 50190.3 MNHN.F.50191 MNHN.F.50191.1, 50191.2, 50191.3, 50191.4, 50191.5, 50191.6 MNHN.F.50192 MNHN.F.50192.1, 50192.2, 50192.3 MNHN.F.50193 MNHN.F.50193.1, 50193.2, 50193.3, 50193.4, 50193.5, 50193.6 MNHN.F.50194 MNHN.F.50194.1, 50194.2, 50194.3 MNHN.F.50195 MNHN.F.50195.1, 50195.2, 50195.3 MNHN.F.50196 MNHN.F.50196.1, 50196.2, 50196.3 MNHN.F.50197 MNHN.F.50197.1, 50197.2, 50197.3 MNHN.F.50198 MNHN.F.50198.1, 50198.2, 50198.3 MNHN.F.50199 MNHN.F.50199.1, 50199.2, 50199.3, 50199.4 MNHN.F.50200 MNHN.F.50200.1, 50200.2, 50200.3, 50200.4, 50200.5 906 GEODIVERSITAS • 2022 • 44 (28) Gentis N. et al. appenDix 1. — Synthetic summary of main characters of the Dipterocarp fossil woods mentioned in the article, as described by the original authors (first name), the authors who redescribed a specimen of the same species or emmended the original diagnosis (second name), or as extracted from another publication (in parenthesis). Features in parenthesis are less commun. Species Growth rings Vessels t.diam. (µm); frequency (/mm2)Fibres Axial parenchyma Rays Multiseriate ray height (µm; cells) Canals Canals t.diam. (µm) Geologic age/ country Authors Anisopteroxylon oblongoides Ind. 120-230 4-5 solitary, tylose Nonseptate, thick-walled, vasicentric tracheids Vasicentric, diffuse, diffusein-aggregate in irregular lines, around canals (1-11)7-9 seriate, 4-8/mm, heterocellular 1-2+ marginal cells, continuous sheath cells 150-1080 5-56 Long tangential lines, (diffuse) 120-150 Miocene/ India Yadav 1989 Anisopteroxylon garoensis Ind. 110-255 8-12 solitary, (tylose) Nonseptate, thick-walled, vasicentric tracheids Scanty to vasicentric,diffuse, in short lines, around canals (1-8)5-6 seriate, 5-12/mm, heterocellular 1-8 marginal cells, sheath cells 180-1350 6-30 Diffuse, (short tangential lines of 2-3) 40-55 Miocene/ India Prakash & Tripathi 1970 Anisopteroxylon surmaensis Ind. 130-230 6-10 solitary, tylose Nonseptate, thinto-thick walled, vasicentric tracheids Diffuse, diffuse-in-aggregate in thin lines, scanty to vasicentric, around canals (1-6)4-5 seriate, 5-6/mm, heterocellular 1-3 marginal cells, continuous sheath cells 212-1190 5-55 Diffuse 102-130 Miocene/ India Prasad, Agarwal & Mandaokar 2009 Anisopteroxylon jawalamukhi Ind. 154-300 - solitary, tylose Nonseptate, vasicentric tracheids Vasicentric, diffuse, diffuse-in-aggregate, around canals 1-7 seriate, heterocellular, sheath cells 458-1946 15-67 Diffuse, (short tangential lines of 2-3) 98-168 Mio-Plio/ India Ghosh & Ghosh 1958 (Prakash & Tripathi 1970) Dipterocarpoxylon sivalicus Ind. 160-240 5-7, solitary, tylose Nonseptate, thick-walled, vasicentric tracheids Diffuse, diffuse-in-aggregate, (scanty paratracheal), around canals 1-5(6) seriate, often uniseriate, 7-14/mm, heterocellular, sheath cells 250-1125 – Short tangential lines of 2-6, diffuse 40-68 Miocene/ India Prakash 1975 Dipterocarpoxylon malavii Ind. 80-240, mean 190, 5-7, solitary, tylose Nonseptate, thick-walled, vasicentric tracheids Diffuse, diffuse-in-aggregate, scanty to vasicentric, (aliform), around canals (1-7)3-5 seriate, 6-8/mm, heterocellular 1-n marginal cells, (sheath cells) 120-1600, mean <1000 10-60 Diffuse/ paired, short tangential lines of 4-5 40-120 Pliocene/ India Ghosh & Ghosh 1959, Guleria 1983 Dipterocarpoxylon africanum Ind. – (solitary), tylose ? Thin-to thick walled?, vasicentric tracheids? Vasicentric, (marginal, diffuse?) around canals (1-5)1-3 seriate, heterocellular 1-6 marginal cells, (end-to-end fusion) – 8-30 Diffuse, short tangential lines of 2-6 - same size as vessels Tertiary/ Uganda Bancroft 1933 Dipterocarpoxylon sarapeense – 160-322 4-7, solitary, (tylose) Nonseptate, thin to thick walled, vasicentric tracheids Vasicentric, diffuse to diffuse-in-aggregate, around canals 1-5(4-5) seriate, 6-8/mm, heterocellular 2-7(23) marginal cells, sheath cells, (end-to-end fusion) <2400 up to 80 Short tangential lines of 2-7 80-120 Plio-Pleisto/ Thailand Vozenin-Serra & Privé-Gill 2001 Dipterocarpoxylon gracile Ind. 100-175 5-14, solitary Nonseptate, (very) thick-walled, vasicentric tracheids Around canals, scanty paratracheal, (diffuse) (1-5)1 seriate, “homogeneous”, (crystalliferous) - up to 40 Diffuse, short tangential lines of 2 60-75 Pliocene/ Java Schweitzer 1958 Dipterocarpoxylon jammuense Ind. 80-260 3-9, solitary, tylose Nonseptate, vasicentric tracheids Vasicentric, (diffuse), around canals (1-6)3-5 seriate, 5-9/mm, heterocellular 2-12 marginal cells, sheath cells 200-1260 5-45 Diffuse, short tangential lines of 2-5 65-80 Miocene/ India Guleria et al. 2002 Dryobalanoxylon holdeniae Ind. 60-280 9-15, solitary, tyloses Fibre/vasicentric tracheids, nonseptate, thick-walled Vasicentric to aliform (aliform-confluent), banded around canals 1-4 seriate, heterocellular 1-5 marginal cells, (sheath cells) <900 up to 30 Long tangential lines 40-60 Tertiary/ India Awasthi 1971 907 Fossil wood from Natma Formation, Myanmar (lower Miocene) GEODIVERSITAS • 2022 • 44 (28) Species Growth rings Vessels t.diam. (µm); frequency (/mm2)Fibres Axial parenchyma Rays Multiseriate ray height (µm; cells) Canals Canals t.diam. (µm) Geologic age/ country Authors Dryobalanoxylon cf. bangkoense Ind. 75-250 7-13, (solitary) Fibre/vasicentric tracheids, nonseptate, thick-walled Diffuse-in-aggregate, diffuse, vasicentric paratracheal (1-7)4-6 seriate, frequent uniseriate, heterocellular 2-5 marginal cells, crystalliferous - 4-80 Long tangential lines 25-125 Quaternary ?/Maluku Is. Schweitzer 1958 Dryobalanoxylon bogorense Ind. 120-286 mean 203 3-4, solitary, tyloses Fibre/vasicentric tracheids, nonseptate, thinto-thick walled Vasicentric to aliform, (aliform-confluent), banded around canals 1-6 seriate, uniseriate rare, heterocellular 1-4 marginal cells 500-1200 11-36 Long tangential lines 33-100 Pliocene/ Java Srivastava & Kagemori 2001 Dryobalanoxylon sumatrense Ind. 75-200 4-8, solitary Fibre/vasicentric tracheids, nonseptate, thin to thick-walled Vasicentric, (aliform), diffuse 1-5 seriate heterocellular 1-15 marginal cells, crystalliferous, silica ? - 4-90 Short tangential line of 5 (or more ?) 75-110 Pliocene/ Sumatra Schweitzer 1958 Dryobalanoxylon tambouense – 125-234 2-10, (solitary, tyloses) Fibre/vasicentric tracheids, nonseptate, thin to thick-walled Vasicentric to aliformconfluent, banded around canals (1)2-5 seriate, 6-8/mm, heterocellular 1-6 cells tailed, (sheath cells, storied tendency) 192-1194 5-38 Long tangential lines 30-80 Pliocene/ Vietnam Vozenin-Serra 1981 Dryobalanoxylon khmerinum Ind. 35-122 7-8, solitary, tyloses Fibre/vasicentric tracheids, nonseptate, thin-walled Vasicentric, banded around canals (1?)3-7 seriate, 7-8/mm, heterocellular, (big cells in the middle, sheath cells) 417-700 12-29 Long tangential lines 55-215 Tert-Quat/ Cambodia Boureau 1952 Schweitzer 1958 Dryobalanoxylon lunaris Ind. 195-257 5-8, solitary, tyloses Fibre/vasicentric tracheids, very thick-walled Vasicentric to aliform, banded around canals, (diffuse-inaggregate in short lines) (1-4)3-4 seriate, 5-7/mm, heterocellular 1-4 marginal cells, crystalliferous, (storied tendency, end-toend fusion) <2980 mean 1405 5-90 Long tangential lines every 0.5-2mm, (diffuse) 90-242 mean 129 Pliocene/ Java Mandang & Kagemori 2004 Dryobalanoxylon javanense Ind. 70-120/125225 (5)8-16, solitary, tyloses Fibre/vasicentric tracheids, very thick-walled Diffuse, diffuse-in-aggregate in short lines, scanty paratracheal (1-6)3-5 seriate, 2-19/mm, heterocellular 2-4 marginal cells, (crystalliferous ?), sheath cells, end-to-end fusions - 3-45 Long tangential lines, (short tangential lines of 2-4) 30-100 Pliocene/ Java Den Berger 1927 (Biswas et al. 2019; Schweitzer 1958) Dryobalanoxylon tobleri Ind. 80-230 (5)8-16, solitary, tyloses Fibre ?/vasicentric tracheids Diffuse ? (1-6)3-4 seriate, 6-8/mm, crystalliferous ?, sheath cells - (8)35-50(60) Long tangential lines 50-250 Pliocene/ Indonesia Den Berger 1927 (Biswas et al. 2019; Schweitzer 1958) Shoreoxylon deomaliense Ind. 64-256 8-16, tyloses Vasicentric tracheids, nonseptate, thin to thick walled Vasicentric, (aliform to aliform-confluent), banded around canals, (diffuse), crystalliferous 1-7 seriate, 6-9/mm, heterocellular 1-8 marginal cells - 4-65 Long tangential lines 35-240 Mio-Plio/ India Prakash & Awasthi 1971 Shoreoxylon burmense Ind. 135-240 5-6, tyloses Vasicentric tracheids, nonseptate, very thickwalled Vasicentric to aliform, (aliformconfluent), diffuse, in short lines, banded around canals (1-5)4-5 seriate, 7-10/mm, homocellular to weakly heterocellular <1275 up to 75 Long tangential lines, 1 to 4 rows 45-150 Tertiary/ Myanmar Prakash 1965a, Prakash 1973 appenDix 1. — Continuation.