This article was downloaded by: [Ingenta Content Distribution (Publishing Technology)] On: 07 October 2014, At: 10:00 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Grana Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/sgra20

Pollen, fruits, and leaves of Tetracentron (Trochodendraceae) from the Cainozoic of Iceland and western North America and their palaeobiogeographic implications Friðgeir Grímsson

a b

c

, Thomas Denk & Reinhard Zetter

d

a

Science Institute , University of Iceland , Reykjavík, Iceland

b

Icelandic Institute (Museum) of Natural History , Reykjavík, Iceland

c

Department of Palaeobotany , Swedish Museum of Natural History , Stockholm, Sweden

d

Department of Palaeontology , University of Vienna , Vienna, Austria Published online: 24 Apr 2008.

To cite this article: Friðgeir Grímsson , Thomas Denk & Reinhard Zetter (2008) Pollen, fruits, and leaves of Tetracentron (Trochodendraceae) from the Cainozoic of Iceland and western North America and their palaeobiogeographic implications, Grana, 47:1, 1-14, DOI: 10.1080/00173130701873081 To link to this article: http://dx.doi.org/10.1080/00173130701873081

PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http:// www.tandfonline.com/page/terms-and-conditions

Grana, 2008; 47: 1–14

Downloaded by [Ingenta Content Distribution (Publishing Technology)] at 10:00 07 October 2014

Pollen, fruits, and leaves of Tetracentron (Trochodendraceae) from the Cainozoic of Iceland and western North America and their palaeobiogeographic implications

FRIÐGEIR GRI´MSSON1,2, THOMAS DENK3 & REINHARD ZETTER4 1

Science Institute, University of Iceland, Reykjavı´k, Iceland, 2Icelandic Institute (Museum) of Natural History, Reykjavı´k, Iceland, 3Department of Palaeobotany, Swedish Museum of Natural History, Stockholm, Sweden, and 4Department of Palaeontology, University of Vienna, Vienna, Austria

Abstract Dispersed pollen, fruits, and leaves of Tetracentron (Trochodendraceae) are described from the Miocene of Iceland and assigned to a new species, Tetracentron atlanticum. The Icelandic fossils represent the first unambiguous record of Tetracentron for the Cainozoic of Europe. Well-preserved dispersed grains of Tetracentron are also identified from the Eocene and Miocene of western North America and compared to the Icelandic fossils. In general, Tetracentron pollen is rather uniform through the Eocene to Recent, only displaying minor variation in ornamentation. Leaves are diagnostic at the species level. The findings add substantial new information to the generally poor fossil record of Tetracentron and indicate a more circumpolar distribution of the genus than previously assumed. The north-eastern Asian – western North American Cainozoic distribution for plant taxa presently confined to East Asia is relatively common. In contrast, the extension to Iceland is remarkable, particularly in view of the absence of the genus in the fossil record from eastern North America, Greenland, and Central Asia and mainland Europe.

Keywords: Tetracentron, fossils, LM, SEM, palaeobiogeography, Cainozoic, Northern Hemisphere

Tetracentron, along with Trochodendron, belongs to the family Trochodendraceae, which forms one of the earliest branching eudicots (AGP II, 2003). Trochodendraceae comprise two extant species of woody angiosperms. At present, Tetracentron Oliv. is a monotypic genus with only one species, Tetracentron sinense Oliver. This species is today restricted to central and southern China, northern Vietnam, northern Burma, and south of the Himalayas to eastern Nepal, Bhutan, and northeastern India (Figure 1; Fu & Bartholomew, 2001; Hara & Kanai, 1964). The oldest fossils that clearly belong to Tetracentron are leaves from the Paleocene to Early Eocene of the Russian Far East (Kamchatka; Chelebaeva & Shancer, 1988) and from the Middle Eocene of western North America (Pigg et al., 2007). No reliable fossils are known from Europe, Central Asia, and eastern North America. From the

Miocene of western North America, Manchester & Chen (2006) described infructescences that are nearly identical to modern Tetracentron. These specimens also contained pollen adhering to the styles of the capsules. Although the pollen of Tetracentron is relatively characteristic the genus is completely absent from the palynological fossil record and dispersed pollen has not been reported from any fossil locality. For the present study we examined leaves, dispersed fruits, and pollen from the Middle to Late Miocene of Iceland (Table I), as well as dispersed pollen belonging to Tetracentron from the Middle Eocene and Middle Miocene of western North America. In addition to light microscopic observations, scanning electron microscopy (SEM) was used. SEM observations enabled a comparative study of pollen from different localities and of fossil and

Correspondence: Friðgeir Grı´msson, Science Institute, University of Iceland, Askja, Sturlugata 7, 101 Reykjavı´k, Iceland. E-mail: [email protected] (Received 19 October 2007; accepted 12 December 2007) ISSN 0017-3134 print/ISSN 1651-2049 online # 2008 Taylor & Francis DOI: 10.1080/00173130701873081

Downloaded by [Ingenta Content Distribution (Publishing Technology)] at 10:00 07 October 2014

2

F. Grı´msson et al.

Figure 1. Map showing part of the Northern Hemisphere. Localities where reliable Cainozoic Tetracentron fossils have been found and the present distribution of T. sinense are shown. Fossil occurrences indicated with symbols, present range with cross-hatching. Migration routes A-E discussed in text (map modified after The Times Atlas of the World, 1958, J. Bartholomew).

Table I. Tetracentron fossils from the Miocene of Iceland. Locality Þrimilsdalur Brekkua´ Hestabrekkusund Hru´tagil Miðdalur Gry´lufoss Hu´savı´kurkleif Surtarbrandsgil Tafla Þo´rishlı´ðarfjall Botn

Formation

Age (Ma)

Leaf

Fruit

Pollen

Hreðavatn-Stafholt Hreðavatn-Stafholt Hreðavatn-Stafholt Skarðsstro¨nd-Mo´kollsdalur Has not been studied Tro¨llatunga-Gautshamar Tro¨llatunga-Gautshamar Brja´nslækur-Selja´ Dufansdalur-Ketilseyri Sela´rdalur-Botn Sela´rdalur-Botn

7–6 7–6 7–6 9–8 10–9 10 10 12 13.5 15 15

+ + 2 2 + 2 2 2 2 2 2

2 2 + 2 2 2 2 2 2 2 2

Not studied + + Not studied Not studied 2 + + Not studied Not studied +

Tetracentron from the Cainozoic of Europe and North America extant pollen. The biogeographic history of the genus including its migration to Iceland and persistence on the island is discussed in light of the new findings.

Downloaded by [Ingenta Content Distribution (Publishing Technology)] at 10:00 07 October 2014

Material and methods Macrofossils and sediment samples were collected from various Miocene formations in Iceland during 1998 – 2007. Additional undescribed material in existing collections was included as well. Leaves and fruits from Iceland described in the present study are housed in the collections of the Icelandic Institute of Natural History (Museum), Reykjavı´k (IMNH). Slide preparations and SEM stubs (nos. 3850/1–8) of fossil and modern pollen are stored in the collections of the Department of Palaeontology, University of Vienna (IPUW). The leaves were measured and described following Dilcher (1974). Herbarium material of Tetracentron sinense was examined in the herbarium of the Swedish Museum of Natural History (S) for comparison with the fossils. Pollen of T. sinense was obtained from the herbarium of the Institute of Botany of Academia Sinica, Beijing (PE). Sedimentary samples for pollen analyses were collected from most of the currently known Miocene formations in Iceland, 15–6 Ma. Additional samples were collected from the Middle Eocene Princeton Chert locality, southern British Columbia, western Canada (Cevallos-Ferriz et al., 1991), and the Middle Miocene of Clarkia, Idaho, western United States (Kvacˇek & Rember, 2000). Sedimentary rock was washed and dried and then ground by hand in a mortar. The resulting powder was boiled in concentrated HCl for 5 minutes to remove all carbonates and then cleaned with distilled water. The remaining material was gently boiled for 5 – 10 minutes in a copper pan in HF to remove all silicates. The solution was then decanted into a 4 l polyethylene beaker filled with water. After settling, the liquid was decanted and remaining material was boiled in concentrated HCl for 5 minutes in a glass beaker to prevent formation of fluorite crystals. The solution was centrifuged and the deposit washed 3 – 4 times with water. The sample was then treated with acetolysis (chlorination and acetylation), washed again, and centrifuged with water 3 – 4 times. Organic material was then mixed with glycerol and one drop was transferred onto a slide. Pollen grains were investigated by light microscope (LM). Single grains were picked using a preparation needle with a human hair mounted on it and placed on a separate slide for photography in LM. The same grain was then transferred to a SEM stub with a drop of absolute ethanol. Transferring

3

the same grain from a LM slide to a SEM stub ensures that the same taxon is photographed in LM and SEM. After LM observation, documentation, and mounting on a SEM stub, the pollen grains were sputter coated with gold and examined in a scanning electron microscope (SEM). In the present study Iceland is considered a European country. From a geological/geographical point of view Iceland could belong to either or both North America and Europe as it is on the boundary of the North American and Eurasian plates. On the other hand, politically, the island has always been part of Europe. Systematic palaeobotany Trochodendraceae Eichler Tetracentron Oliver Tetracentron atlanticum spec. nov. (Figures 2B, E; 3A, F, J, O, P; 4A–O; 5A–C) Diagnosis. Leaves petiolate, petiole long, widened at distal part, lamina wide-ovate, base deeply cordate to lobate, primary venation actinodromous, primary and secondary veins forming loops, margin serrate, teeth triangular, tooth acumen glandular, two veins running from gland to adjacent sinus; detached fruits are bilateral symmetric capsules, median axial lineation, apex emarginate, styles originating in lower third of capsule, one style per carpel (fruit originally 4-carpellate); pollen tricolpate, small, prolate to subprolate, tectate, striate to microstriatoreticulate, striate around apertures, open striatoreticulate on mesocolpium, colpus membrane covered with globular elements. Holotype. IMNH 6894, leaf (Figures 2E, 3A). Paratypes. IMNH 804, 6895a, 6895b (Figures 2B, 3F, 3J - leaves), 6896a, 6896b (Figure 3O, 3P - fruits). Type locality. Brekkua´ outcrop (7–6 Ma), close to the lake Hreðavatn, western Iceland. The Brekkua´ outcrop belongs to the Hreðavatn-Stafholt Formation (Table I). Additional localities. Botn, Sela´rdalur-Botn Formation, 15 Ma; Surtarbrandsgil, Brja´nslækur-Selja´ Formation, 12 Ma; Miðdalur (formation needs to be studied), 10– 9 Ma; Þrimilsdalur and Hestabrekkusund, HreðavatnStafholt Formation, 7–6 Ma. Etymology. The species is named after the Atlantic Ocean, emphasizing the novel geographic position indicated by its occurrence in Iceland.

F. Grı´msson et al.

Downloaded by [Ingenta Content Distribution (Publishing Technology)] at 10:00 07 October 2014

4

Figure 2. A – E. Extant and fossil Tetracentron leaves. A, C & D. Tetracentron sinense: (A) ovate medium-sized leaf, truncate base, large triangular teeth, A. N. Steward, Y. Chiao & H. C. Cheo, 663 (S); (C) elliptical medium-sized leaf, numerous small teeth, Z. Quing-Sheng, 0541 (S); (D) ovate, large leaf, cordate base, numerous small teeth, Z. Quing-Sheng, 0639 (S). B, E. Tetracentron atlanticum: (B) fossil leaf from the Þrimilsdalur locality, Hreðavatn – Stafholt Formation, 7–6 Ma, medium-sized leaf, large triangular teeth, IMNH 804; (E) holotype, fossil leaf from the Brekkua´ locality, Hreðavatn – Stafholt Formation, 7–6 Ma, ovate large leaf, part of lamina displaced, large triangular teeth, IMNH 6894. Scale bars – 2 cm (A–C); 3 cm (D, E).

Downloaded by [Ingenta Content Distribution (Publishing Technology)] at 10:00 07 October 2014

Tetracentron from the Cainozoic of Europe and North America Description. Leaves – petiolate, petiole up to 45 mm long, stout; lamina up to 124 mm long and 95 mm wide, asymmetrical or symmetrical, wide ovate to wide elliptical, length to width ratio approximately 1.2/1, length of petiole to length of lamina approximately 0.27/1 (measured in a single specimen); apex acute to acuminate, base cordate to lobate; extreme base entire-margined, margin serrate, evenly toothed, apical side of tooth concave to acuminate, basal side convex to acuminate, tooth with elongate apex, curving upwards, tooth acumen glandular, teeth relatively large, 4 – 8 teeth per 2 cm margin, sinuses usually angular, rarely rounded, principal veins entering teeth medially, two distinct veins connecting glandular region of tooth to adjacent sinuses; primary venation actinodromous, 5 – 7 veins, central primary vein straight in proximal part, 2 or 3 pairs of lateral primary veins, innermost primary veins forming angles of 22˚ to 27˚ with central vein, second pair of primary veins forming angles of 45˚ to 65˚ with central vein, outermost primary veins forming angles of 85˚ to 95˚ with central vein, lateral primary veins running in strongly developed re-curved arches, converging with other veins towards leaf apex; secondary veins brochidodromous and connecting to superadjacent secondary veins, secondary veins diverging from mid vein at 30˚ to 40˚, branches of secondary or higher order loops running into sinuses of teeth and tooth apex; tertiary veins originating at right to acute angles from primary or secondary veins and their branches, tertiary veins thin but distinct, irregularly percurrent, simple or forked; quaternary veins orthogonal, arising at right angles, marginal ultimate venation looped, areoles well developed, moderate to large, 0.76 – 1.52 mm across, irregular in size and shape, veinlets slender, branched two to three times. Fruits – detached capsules, eroded, compressed, two of four carpels visible, up to 3.9 mm long and 3 mm wide, length to width ratio approximately 1.28; apex emarginate, fruit with a median axial lineation; styles persistent, originating in lower third part of fruit, one style per carpel, styles broken in all specimens, style parts 0.7+ mm long, recurved; sepals preserved as imprint just below styles, alternating with styles, 4 sepals originally covering lowest part of fruit. Pollen – monad, often in groups of 2 – >14, prolate to subprolate, elliptical outline in equatorial view, circular to lobate in polar view, polar axis 12.0 – 16.7 mm under SEM and 14.8 – 17.8 mm under LM, equatorial diameter 9.6 – 14 mm under SEM and 10.5 – 16.3 mm under LM, tricolpate, microreticulate in LM, striate to microstriatoreticulate in SEM, striae 0.29 – 0.37 mm in diameter, striate around apertures,

5

showing a more open striatoreticulate pattern on mesocolpium, space between striae narrower to rarely slightly wider than diameter of striae, colpus membrane microverrucate to microrugulate. Remarks. We argue that the Tetracentron fossils from the Miocene of Iceland belong to a single species. This is based on the occurrence of only one type of leaves, fruits, and pollen from all the localities studied. All described organs are co-occurring in the Hreðavatn-Stafholt Formation, 7–6 Ma. Pollen and fruits were identified from the same bed in the Hestabrekkusund outcrop (Table I), and leaves and pollen were found in the same bed at the Brekkua´ locality. Leaves have also been collected from the Þrimilsdalur locality. The fossiliferous sedimentary beds of these outcrops have been correlated (Grı´msson, 2007) and have been shown to belong to the same facies association (same sedimentary type, same age, and formed during the same event). We use this strong association evidence to accommodate the different organs in a single species. In general, in Tetracentron leaves appear to be more diagnostic (at the species level) than pollen and fruits. Thus, it may be difficult to establish distinct fossil species based on pollen and fruits only. In combination with the leaves they confirm the generic identity of the fossil taxon. Tetracentron remberi Manchester & Chen (Figure 5D–I) Locality. Clarkia, Idaho, western United States. Age. Middle Miocene. Description. Pollen – Monad, prolate to subprolate, elliptical outline in equatorial view, circular to lobate in polar view, polar axis 13.8 – 16 mm under SEM and 15 – 18.3 mm under LM, equatorial diameter 11.2 – 12 mm under SEM and 13.3 – 15.2 mm under LM, tricolpate, microreticulate in LM, striate to microstriatoreticulate in SEM, striae 0.41 – 0.42 in diameter, striate around apertures, mesocolpium showing a more open striatoreticulate pattern, space between striae much wider than diameter of striae, colpus membrane microverrucate to microrugulate. Remarks. Identical pollen has been found in situ on styles of fruits of Tetracentron remberi (Manchester & Chen, 2006). Tetracentron sp. (Figure 5J–L) Locality. Princeton chert, British Columbia, southwestern Canada.

Downloaded by [Ingenta Content Distribution (Publishing Technology)] at 10:00 07 October 2014

6 F. Grı´msson et al.

Tetracentron from the Cainozoic of Europe and North America Age. Middle Eocene.

Downloaded by [Ingenta Content Distribution (Publishing Technology)] at 10:00 07 October 2014

Description. Pollen – Monad, spherical, circular to elliptical in equatorial view, circular in polar view, polar axis 11.5 – 13.2 mm under SEM, equatorial diameter 10 – 12.5 mm under SEM, tricolpate, microreticulate in LM, striate to microstriatoreticulate in SEM, striae 0.43 – 0.52 in diameter, striate around apertures, microstriatoreticulate in mesocolpium (SEM). Remarks. Pigg et al. (2007) described leaves of Tetracentron hopkinsii Pigg, Dillhoff, DeVore, & Wehr from the Okanagan highlands. The sediments yielding these leaves are of similar age as the Princeton chert. The pollen described here and leaves of T. hopkinsii may therefore have been produced by the same species. Discussion Comparison of the new fossil taxa to other Tetracentron species Leaves. – The general outline and type of venation of the fossil leaves described here are found in various other genera such as Populus, Trochodendron, and Cercidiphyllum. A number of characters, however, are diagnostic of Tetracentron. Among these characters are the marginal glands that are gradually thickening towards the tooth acumen and connected to a pair of distinct veins running from the tooth acumen down to the sinuses (Figure 3G). Also the way that the primary veins depart from the petiole at the leaf base is typical of Tetracentron. Leaves of the modern Tetracentron sinense were compared to the Icelandic fossils (Figure 2A, C, D; 3B–E, G–I, K; Table II). T. sinense leaves are variable in outline, shape of base, and marginal features, and overall more variable than previously noted (e.g. Pigg et al., 2007). Leaves are found on two kinds of shoots. Long annual shoots with alternately arranged leaves (leaf type not shown), and single leaves subterminal

7

on short, many-year-old shoots (Figure 2A, C, D). Leaves of annual or elongation shoots are usually smaller, narrower, and often oblong with attenuate apex; their base varies from obtuse to truncate, and teeth are inconspicuous. The leaf dimorphism in the extant T. sinense would need to be studied in more detail, but based on the material available to us the fossil leaves represent short shoot leaves. Main distinguishing features of leaves of Tetracentron atlanticum to those of T. sinense are: 1) a lower length to width ratio in T. atlanticum, 2) The consistently large teeth with a long and slender tooth acumen in T. atlanticum (Figure 3A, F) as opposed to teeth in T. sinense that are usually smaller, with a blunt and short acumen (Figure 3B–E, G–I), and 3) The deeply cordate to lobate base in T. atlanticum as opposed to the cordate to truncate base in T. sinense. Few fossil leaves of Tetracentron can be compared to the leaves described here from Iceland. Tetracentron beringianum Chelebaeva (Chelebaeva & Shancer, 1988) from Kamchatka appears to be the oldest record of Tetracentron. The leaves differ from T. atlanticum by having a leaf base that is only slightly cordate in contrast to a lobate one, and the smaller teeth (Table II). Tetracentron hopkinsii (Pigg et al., 2007) is morphologically intermediary between the extant T. sinense and T. atlanticum (Table II). The fossil species share a lower length to width ratio, whereas T. hopkinsii and T. sinense share the blunt dentition. The Late Miocene to Late Pliocene Tetracentron masuzawaense (Murai) K. Ozaki (including Tetracentron ibei K. Suzuki; Suzuki, 1967; Ozaki, 1987) from the Kabutoiwa, Masuzawa, and Atagi Formations in Japan shows more resemblance to T. atlanticum (Table II). However, the apical region in T. masuzawaense is more attenuate than in T. atlanticum (Figure 2B, E). The teeth in T. masuzawaense are narrow and long, almost triangular in outline, while the teeth of the Icelandic fossils are generally wider at their base and have a much longer and narrower apical region (Figures 2E; 3A, F).

r Figure 3. A – P. Extant and fossil Tetracentron leaves, infructescenses, and fruits. A, F, J, O & P. Tetracentron atlanticum. (A) Detail of Figure 2E – looping primary and secondary veins and marginal venation, note large teeth with glandular acumen. (F) Detail of Figure 2B – looped venation and large triangular teeth with glandular tips. (J) Fossil leaf from the Brekkua´ locality, Hreðavatn – Stafholt Formation, 7– 6 Ma, basal part of a large leaf, base lobate, IMNH 6895a. O, P. Fruits detached from calyx, styles partly preserved (black arrows – area of style attachment; white arrows – imprint of calyx on capsule). (O) Fruit from the Hestabrekkusund locality, Hreðavatn – Stafholt Formation, 7–6 Ma, IMNH 6896a. (P) Fruit from the Hestabrekkusund locality, Hreðavatn – Stafholt Formation, 7–6 Ma, IMNH 6896b. B–E, G–I & K–N. Tetracentron sinense. (B) Detail of Figure 2A – large triangular teeth with glandular acumen. (C) Numerous medium-sized teeth, most teeth with a glandular acumen, H. T. Tsai, 63072 (S). (D) Detail of Figure 2C – numerous small teeth with glandular acumen, apical side much shorter than basal side, teeth relatively short. (E) Tooth extension consisting of glands, teeth are small and inconspicuous, Z. Quing-Sheng, 1142 (S). (G) Cleared leaf, relatively large triangular teeth, A.N. Steward, Y. Chiao & H. C. Cheo, 663 (S). (H) Teeth small, sinuses rounded, H. T. Tsai, 63072 (S). (I) Teeth are wide and rounded, margin crenulate, Z. Quing-Sheng, 0589 (S). (K) Deeply cordate base, curved outer primary veins, H.T. Tsai, 63072 (S). (L) Part of infructescence with sessile fruits, H.T. Tsai, 63072 (S). (M) Open mature fruit, part of the calyx fallen off, imprints of calyx seen on capsules, H.T. Tsai, 63072 (S). (N) Mature fruit with part of calyx missing and broken styles, H.T. Tsai, 63072 (S). Scale bars - 1 cm (A–K); 2 mm (L–P).

F. Grı´msson et al.

Downloaded by [Ingenta Content Distribution (Publishing Technology)] at 10:00 07 October 2014

8

Figure 4. LM (A, D, G, J & M) and SEM (B, C, E, F, H, I, K, L & N, O) micrographs of dispersed pollen of Tetracentron atlanticum from Miocene sediments of Iceland. A–C. T. atlanticum from the Botn locality, Sela´rdalur-Botn Formation, 15 Ma, IPUW 3850/1. (A) and (B) Equatorial view. (C) Detail of Figure 4B – mesocolpium and part of colpus. D–F. T. atlanticum from the Botn locality, Sela´rdalur-Botn

Downloaded by [Ingenta Content Distribution (Publishing Technology)] at 10:00 07 October 2014

Tetracentron from the Cainozoic of Europe and North America Fruits. – The only fossil Tetracentron fruits that have been described until now are T. remberi from the Middle Miocene Clarkia locality, Idaho, western United States (Manchester & Chen, 2006). Fruits are attached to an infructescence axis and are slightly better preserved than the detached fruits of T. atlanticum. Although quite similar in appearance, fruits of T. atlanticum are slightly narrower (Figure 3O, P), more elliptical, and longer than the fruits of T. remberi. Fruits of the extant T. sinense (Figure 3L–N), are tetralocular capsules with four styles (Doweld, 1998). The Icelandic fruits fall well within the size range of the extant T. sinense, which is up to 4 mm long (3 to 5 mm according to Doweld, 1998). The general organization of the fossil and extant fruits is identical (Figure 3M–P). Pollen. To our knowledge, the only genus having pollen similar to Tetracentron is Trochodendron. Pollen of these two genera differs from each other in size and ornamentation of the tectum. Pollen of Trochodendron is larger and the tectum is more reticulate than striate in comparison to Tetracentron (Praglowski, 1974; Endress, 1986; Manchester & Chen, 2006; Zetter, pers. obs.). Pollen assigned here to Tetracentron from the Eocene to Recent shows only few distinguishing characteristics (Table III). The pollen is tricolpate with a striate to microstriatoreticulate tectum. All grains studied display similar dimensions. The only obvious difference is encountered in the pattern of the mesocolpium. In T. atlanticum the striae are arranged as loose or tight networks in the oldest formation, while they form a tight network in specimens from the younger formations. In all other fossil and recent species examined the striation on the mesocolpium typically appears as a loose network (Table III). In addition, pollen of T. atlanticum shows a more parallel oriented striation than the other species. Critical review of the fossil record of Tetracentron The presence of the genus Tetracentron has been reported from late Early Cretaceous to Pliocene sediments in the Northern Hemisphere (Table IV, Figure 1). The oldest fossils referred to Tetracentron are leaves from the late Early Cretaceous (Albian) of

9

Kazakhstan, eastern Siberia, and eastern North America (Iljinskaja, 1972, 1974) assigned to two different species, Tetracentron vachrameevianum Iljinskaja and Tetracentron potomacense (Ward) Iljinskaja based on a few and small leaves. Tetracentron vachrameevianum shares characters with leaves of other taxa such as Populus, and particularly Cercidiphyllum. The figured leaves and drawings of venation details do not compare to modern Tetracentron. The length/width ratio in the fossils from Kazakhstan is too low and the leaves are suborbicular in shape. Leaves of extant Tetracentron are ovate to elliptical and always longer than wide. In addition, T. vachrameevianum does not show the typical glandular teeth that are innervated by three veins. The overall venation, shape, and tooth architecture matches better types assigned to Cercidiphyllum. Also T. potomacense is unlike extant Tetracentron. Marginal characters, tooth architecture, and venation correspond better to fossil leaves that have been assigned to Populus. Teeth in T. potomacense are rounded and wide apart, separated by wide and rounded sinuses; this is not found in T. sinense. Only on rare occasions are the teeth of T. sinense rounded (Figure 3I) but they are then more closely spaced and separated by narrow angular to rounded sinuses. As in T. vachrameevianum, Tetracentron-like teeth are not found in T. potomacense (Iljinskaja, 1972, 1974). Tetracentron fossils have also been reported from the Late Cretaceous of East Asia by Kryshtofovich (in Kryshtofovich & Baikovskaja, 1966) who reported leaves that he described as Tetracentron amurense Kryshtofovich. The morphology of these leaves is markedly different from the living species, particularly in venation and marginal features. At present, it is unclear whether this taxon can be assigned to Tetracentron. The fossil record of Tetracentron from the Paleocene and Eocene is richer than the Cretaceous one (Table IV, Figure 1). Leaves from the Paleocene to Early Eocene of Kamchatka have been described as T. beringianum (Chelebaeva & Shancer, 1988). These fossils show a strong resemblance to the extant T. sinense (see above) and younger fossil types (Table II). From the Middle to Late Eocene of Alaska, Wolfe (1977) assigned fossil leaves to Tetracentron piperoides (Lesquereux) Wolfe and emended the occurrence of this species in the

r Formation, 15 Ma, IPUW 3850/2. (D) and (E) Equatorial view. (F) Detail of Figure 4E – mesocolpium and part of colpus. G–I. T. atlanticum group of pollen from the Botn locality, Sela´rdalur-Botn Formation, 15 Ma, IPUW 3850/3. (G) and (H) Pollen cluster (I) Detail of Figure 4H showing a single grain. J–L. T. atlanticum from the Surtarbrandsgil locality, Brja´nslækur-Selja´ Formation, 12 Ma, IPUW 3850/ 4. (J) and (K) Equatorial view. (L) Detail of Figure 4K – mesocolpium and part of colpus. M–O. T. atlanticum from the Surtarbrandsgil locality, Brja´nslækur-Selja´ Formation, 12 Ma, IPUW 3850/5. (M) and (N) Equatorial view. (O) Detail of Figure 4N – mesocolpium and part of colpus. All LM micrographs ca61000. Scale bars – 10 mm (H); 1 mm (B, C, E, F, I, K, L, N, O).

F. Grı´msson et al.

Downloaded by [Ingenta Content Distribution (Publishing Technology)] at 10:00 07 October 2014

10

Figure 5. LM (A, D, G, J & M) and SEM (B, C, E, F, H, I, K, L & N, O) micrographs of dispersed fossil pollen of Tetracentron from Iceland and western North America and extant pollen of T. sinense. A–C. Tetracentron atlanticum from the Brekkua´ locality, HreðavatnStafholt Formation, 7–6 Ma, IPUW 3850/6. (A) and (B) Slightly oblique equatorial view. (C) Detail of Figure 5B – mesocolpium and part

Downloaded by [Ingenta Content Distribution (Publishing Technology)] at 10:00 07 October 2014

Tetracentron from the Cainozoic of Europe and North America Palaeogene of North America (see synonymy list in Wolfe, 1977). Pigg et al. (2007) questioned the determinations by Wolfe and stated that two of his specimens should be assigned to Zizyphoides, while the third was indeterminable. Tetracentron hopkinsii from the Middle Eocene of British Columbia (Pigg et al., 2007) clearly belongs in the genus Tetracentron (Table II). From the Eocene of Japan, Tanai (1981) described T. piperoides. The venation and marginal features of some of these fossils clearly place them in the genus Tetracentron (and these fossils are in need of taxonomic revision). The only European Palaeogene record is Tetracentron agriense (Andrea´nszky) Kvacˇek & Hably from the Upper Oligocene Eger flora in Hungary (Kvacˇek & Hably, 1991). These fossils are not well preserved and from both description and published figures it is unclear if veins are running from the tooth glands to the sinuses. In addition, the area where the primary veins depart from the petiole appears not to be indicative of Tetracentron. Complete infructescences with fruits and in situ pollen have been reported from the Middle Miocene of western North America (Manchester & Chen, 2006). This is the only record of Tetracentron pollen prior to the present study (Table IV, Figure 1). Leaves and wood have been described from the late Middle Miocene to Late Pliocene of Japan (Suzuki, 1967; Ozaki, 1987; Suzuki et al., 1991). These leaves unequivocally belong to Tetracentron. There is only a single record of Tetracentron from the Neogene of Europe, namely Tetracentron hungaricum Andrea´nszky from the middle Lower Miocene of Hungary (Andrea´nszky, 1959). The single leaf differs from those of Japan and Iceland and also from extant leaves in the shape of the apex, and especially in features of the leaf margin. The general outline as well as primary venation is similar to Tetracentron, but the fossil figured and the drawing and description of venation suggest an entire to undulate margin. Therefore, it is unlikely that this fossil belongs to Tetracentron. Iljinskaja (1968) later assigned this specimen to Cercidiphyllum. Palaeobiogeography of Tetracentron and migration to Iceland The fossil record of Tetracentron (Table IV, Figure 1) indicates that the genus was well established in the

11

Paleocene and especially the Eocene of northwestern North America and north-eastern Asia. During this time interval, Tetracentron could migrate freely between the two continents (migration route A, Figure 1). This Early Cainozoic biogeographical pattern is typical for a number of angiosperm taxa that are confined to East Asia today. Examples are Dipteronia (Sapindaceae, McClain & Manchester, 2001), Davidia (Manchester et al., 1999), Ulmus furcinervis (Borsuk) Ablaev sensu Feng et al. (2003) and the closely related Ulmus okanaganense Denk & Dillhoff (Denk & Dilhoff, 2005). The Icelandic records of Tetracentron add to this pattern. Based on the currently known fossil distribution of Tetracentron the Icelandic range appears to be extremely isolated. In view of the poor macrofossil record of Tetracentron and the fact that it may easily have been overlooked in the palynological record, its distribution may have extended beyond this region. The only previous records for Tetracentron in Europe are the Late Oligocene, T. agriense (Kvacˇek & Hably, 1991) and the middle Lower Miocene T. hungaricum (Andrea´nszky, 1959) from Hungary. The latter has been excluded from the genus Tetracentron (Iljinskaja, 1968). Even though the generic identity of the Oligocene fossils is highly questionable we cannot exclude that they belong to Tetracentron. This suggests two alternative migration routes of Tetracentron to Iceland. The first, and, based on the current knowledge, more likely, is from the west (migration routes B-C, Figure 1). In north-eastern Asia and western North America Tetracentron has a more or less continuous fossil record from the Palaeogene to the Miocene. Given this it must have migrated to the Atlantic via northern Canada over the Ellesmere Island region and to Greenland, and from Greenland over to Iceland. A similar migration route has been suggested for Fagus friedrichii Grı´msson & Denk (Grı´msson & Denk, 2005). The alternative migration route to Iceland is from the east. This would require an initial dispersal of the genus via Central Asia to Europe (migration route D-E, Figure 1). If the record of Tetracentron from the Late Oligocene of Hungary is correct, a migration to Iceland from Europe would be equally plausible. A migration from the east to Europe in the course of the Early Oligocene would be achievable in light of the

r of colpus. D–I. Tetracentron remberi from the Middle Miocene Clarkia locality, Idaho, western USA, IPUW 3850/7. (D) and (E) Equatorial view. (F) Detail of Figure 5E – mesocolpium and part of colpus. (G) and (H) Equatorial view. (I) Detail of Figure 5H – mesocolpium and part of colpus. J–L. Tetracentron sp. from the Middle Eocene, Princeton chert locality, southern British Columbia, western Canada, IPUW 3850/8. (J) Polar view. (K) Equatorial view. (L) Detail of Figure 5K – mesocolpium and part of colpus. M–O. Tetracentron sinense, Nepal, Gupta & Chandhary, no. 145 (PE). (M) and (N) Equatorial view. (O) Detail of Figure 5N – mesocolpium and part of colpus. All LM micrographs ca61000. Scale bars – 1 mm (B, C, E, F, H, I, K, L, N, O).

Not preserved 58 – 135 mm645 – 90 mm 1.3 – 1.9 Ovate Attenuate Cordate to deeply cordate

Serrate Large with elongate apex

Concave to straight Convex, acuminate to straight Acute, angular to round

(45 mm, – (124 mm695 mm 1.2 Widely ovate to widely elliptic Acute to acuminate Cordate to lobate

Serrate Large with elongate apex

Concave to acuminate Convex to acuminate Narrow to wide, angular or round

closure of the Turgai Strait (Akhmetiev & Reshetov, 1996). To confirm or reject the European record of the genus, the Eger sediments ought to be checked for dispersed Tetracentron pollen. In addition, palynological studies of Oligocene to Miocene sediments containing rich macrofloras from Eastern Europe and Central Asia ought to be done. However, a previous screening of numerous Central European Cainozoic floras (Germany, Italy, Austria, Czech Republic, Hungary) by one of the authors have not yielded any Tetracentron pollen. Timing and mode of migration to (Proto)Iceland have been discussed in two previous papers (Grı´msson et al., 2007; Grı´msson & Denk, 2007) and may have occurred long before the accumulation of the oldest (15 Ma) sediments in today’s Iceland. Tetracentron is present in the 15 Ma to the 6 Ma sediments of Iceland. Future studies will show whether or not this element extended into the Pliocene of Iceland. At present, T. sinense occurs as accessory element of mixed evergreen-deciduous and broadleaved evergreen forests from Central China to the eastern Himalayas. The vertical distribution of T. sinense increases from its eastern to its western range (1100 – 3500 m a.s.l.; Fu & Bartholomew, 2001). This makes the single modern species a humid temperate element and suggests that T. atlanticum could have coped well with winter snow in a generally oceanic climate.

Straight and short Straight, concave to convex Wide, angular to round Tooth apical side Tooth basal side Sinus

Straight to convex Convex Narrow, angular

Serrate Small and sharp Margin Dentition

Serrate Large to medium size

(28 mm, – 70 – 100 mmX660 – 80 mm 1.3 Ovate to widely elliptic Acute to acuminate Round

43 mm, 2 – 5 mm (95 mm675 mm 1.3 Elliptic Acuminate Cordate

Conclusions

Petiole (length, width) Lamina (length6width) Length/Width ratio Shape Apex Base

Tetracentron masuzawaense Miocene to Pliocene Japan (Suzuki, 1967; Ozaki, 1987) Tetracentron atlanticum Tetracentron beringianum Tetracentron hopkinsii Miocene Paleocene to Early Eocene Middle Eocene South-western Canada Iceland Kamchatka (Chelebaeva & Shancer, 1988) (Pigg et al., 2007) (This study) Figure 2B, E; 3A, F, J

Table II. Comparative leaf morphology of Cainozoic Tetracentron.

Downloaded by [Ingenta Content Distribution (Publishing Technology)] at 10:00 07 October 2014

(14–40 mm, robust 40 – 120 mm618 – 90 mm 1.38 – 2.22 Ovate to elliptic Acuminate to acute Acute, truncate, round, or cordate to deeply cordate Serrate to bluntly serrate Small and numerous teeth, rarely large and few Convex, straight or concave Straight, concave to convex Wide to narrow, angular to round

F. Grı´msson et al. Tetracentron sinense Recent Central China to E. Nepal Figures 2A, C & D; 3B–E, G–I & K

12

This study adds an important component to the palaeobiogeography of Tetracentron. At the moment, it is not completely clear whether the genus migrated to Iceland from the west or from the east although current fossil evidence favours migration from the west. The study also highlights the importance of combining investigations of macrofossils with palynological data. Stratigraphic gaps noticed in the macrofossil record may be artificial and, in the case of Iceland, the timing of migration deduced from the macrofossil record might be incorrect (Grı´msson & Denk, 2007). For example, Tetracentron macrofossils have only been recovered in the 10–6 Ma sediments, but pollen of Tetracentron constitutes one of the most common pollen types in sediments from 15 Ma and 12 Ma. In general, one would expect anemophilous plants to be better represented (more diverse) in the pollen record than in the macrofossil record. In Icelandic sediments taxa such as Ilex and Sciadopytis are known only from the pollen record (work in progress). Tetracentron is an entomophilous genus (Endress, 1986) and the mass occurrence of its pollen in most of the Miocene sediments is

Tetracentron from the Cainozoic of Europe and North America

13

Table III. Comparative pollen morphology of Cainozoic Tetracentron.

Downloaded by [Ingenta Content Distribution (Publishing Technology)] at 10:00 07 October 2014

Polar axis SEM / LM (mm) Equatorial diameter SEM / LM (mm) Tectum, pattern of striation Striae diameter (mm) a

Tetracentron sp. Middle Eocene southwestern Canada Figure 5J–L

Tetracentron remberi Miocene western United States (Manchester & Chen, 2006) Figure 5D–I

Tetracentron atlanticum Miocene Iceland Figures 4 & 5A–C

Tetracentron sinense Recent Central China to eastern Nepal Figure 5M–O

11.5–13.2 / 17.2

13.8–16 / 15–18.3

12–16.7 / 14.8–17.8

9.5–15.0a / 14.7

10.0–12.5 / 16.8

11.2–12.0 / 13.3–15.2

9.6–14.0 / 10.5–16.3

7.9 / 12.7

Striate close to aperture; striato-reticulate mesocolpium 0.43–0.52

Striate close to aperture; striato-reticulate mesocolpium 0.41–0.42

Striate close to aperture; Striate close to aperture; striate to striato-reticulate striato-reticulate mesocolpium mesocolpium 0.29–0.37 0.33–0.42

Second value from Endress (1986).

interesting. The presence of Tetracentron pollen clusters (Figure 4G–I) from the 15 Ma sediments suggests that this pollen had not been transported over long distances prior to sedimentation. The finding of Tetracentron fruits and leaves together with dispersed pollen in younger Icelandic sediments further testify that this pollen is autochthonous. The generally very rare occurrence of leaves and fruits in Cainozoic sediments may have various reasons. The deciduous thin leaves of Tetracentron could have a low potential for preservation as is often the case for riparian elements (e.g. Fraxinus). Furthermore, the small detached fruits may be difficult to recognize in the macrofossil record. Finally, this study shows that small and inconspicuous (in LM) pollen, such as Tetracentron, is likely to have been overlooked or misidentified in previous LM palynological surveys of Cainozoic

sediments from the Northern Hemisphere. Studies focussing on systematic and biogeographical aspects of Cainozoic plant taxa should therefore include SEM investigations of microfossils as scanning electron microscopy allows for a much higher taxonomic resolution than light microscopy. Acknowledgements Nadja Kavcik is thanked for invaluable assistance with preparation of palynological samples and image processing. Ruth Stockey kindly provided sediment samples from the Princeton chert. Maria Tekleva provided access to Russian literature and translations into English. TD’s and FG’s research have been financially supported by Riksmusei Va¨nner and Swedish Research Council (Grant no. 2006–5571) grants to TD. FG currently holds a postdoctoral

Table IV. Tetracentron fossils from the Cainozoic of the Northern Hemisphere recognized in the present study. Species

Organ a

Locality, formation

T. beringianum Chelebaeva

Leaves

T. hopkinsii Pigg, Dilhoff, DeVore & Wehrb Tetracentron sp. Tetracentron sp.b

Leaves

Kamchatka, Sosophanskaya Formation Allenby Formation

Pollen Leaves

Princeton Chert Republic, Washington

T. piperoides (Lesquereux) Wolfec

Leaves

Ishikari and Kyji coal fields

T. atlanticum

Leaves, fruits, dispersed pollen

T. remberi Manchester & Chend

Infructenses, fruits, insitu pollen, dispered pollen Wood

Sela´rdalur - Botn Formation, Brja´nslækur - Selja´ Formation, Hreðavatn - Stafholt Formation Race Track, Clarkia, Idaho

T. japonoxylum M. Suzuki, L. Joshi & S. Noshiroe T. masuzawaense (Murai) K. Ozakif Tetracentron sp.g a

Leaves Leaves

Country Russia Southwestern Canada Southwestern Canada Northwestern United States Japan Iceland

Period Paleocene to Early Eocene Middle Eocene Middle Eocene Middle Eocene Middle to Late Eocene Middle to Late Miocene

Northwestern United States

Middle Miocene

Yanagida Formation

Japan

Kabutoiwa Formation, Masuzawa Formation Atagi Formation

Japan

Middle Miocene Late Miocene

Japan

Late Pliocene

Chelebaeva & Shancer, 1988; bPigg et al., 2007; c Tanai, 1981; dManchester & Chen, 2006; eSuzuki et al., 1991; fSuzuki, 1967, Ozaki, 1987; gOzaki, 1987.

14

F. Grı´msson et al.

position (grant no. 070236041) financed by the Icelandic Research Fund (Rannis). Most helpful comments on the manuscript were provided by Peter Endress, Steven Manchester and Else Marie Friis. We also thank Wieslaw Smolenski for assistance with typesetting and layout.

Downloaded by [Ingenta Content Distribution (Publishing Technology)] at 10:00 07 October 2014

References Akhmetiev, M. A. & Reshetov, Y. V. (1996). The Eocene to early Miocene evolution of plant assemblages and large mammals in the extratropical Asia. Stratigr.Geol. Correl., 4, 371–377. Angiosperm Phylogeny Group (2003). An update of the Angiosperm Phylogeny Group classification for the orders and families of the flowering plants. APG II. Bot. J. Linn. Soc., 141, 399–436. Andrea´nszky, G. (1959). Die Flora der sarmatischen Stufe in Ungarn. Budapest: Kiado´ Akad. Cevallos-Ferriz, S. R. S., Stockey, R. A. & Pigg, K. P. (1991). The Princeton chert: Evidence for in situ aquatic plants. Rev. Palaeobot. Palynol., 70, 173–185. Chelebaeva, A. I. & Shancer, A. E. (1988). Litologija i stratigrafija mezozoja I kajnozoja vostochnyh rajonov SSSR [Litology and stratigraphy of the Mesozoic and Cenozoic of the USSR eastern regions]. In P. P. Timofeev & J. B. Gladenkov (Eds.), Sbornik nauchnyh trudov [Collection of scientific papers] (pp. 135–148). Moscow: Geol. Inst. Denk, T. & Dilhoff, R. M. (2005). Ulmus leaves and fruits from the Early-Middle Eocene of northwestern North America: Systematics and implications for character evolution within Ulmaceae. Can. J. Bot., 83, 1663–1681. Dilcher, D. L. (1974). Approaches to the identification of angiosperm leaf remains. Bot. Rev., 40, 1–157. Doweld, A. B. (1998). Carpology, seed anatomy and taxonomic relationship of Tetracentron (Tetracentraceae) and Trochodendron (Trochodendraceae). Ann. Bot., 82, 413–443. Endress, P. K. (1986). Floral structure, sytematics, and phylogeny in Trochodendrales. Ann. Mo. Bot. Gard., 73, 297–324. Feng, G.-P., Ablaev, A. G., Wang, Y.-F. & Li, C.-S. (2003). Palaeocene Wuyun flora in Northeast China: Ulmus furcinervis of Ulmaceae. Acta Bot. Sin., 53, 146–151. Fu, D. & Bartholomew, B. (2001). Tetracentraceae. In Edit. Comm (Ed.), Flora of China. Vol. 6. Caryophyllaceae through Lardizabalaceae (p. 125). Beijing: Sci. Press/St. Louis: Mo. Bot. Gard. Grı´msson, F. (2007). Sı´ðmı´o´sen setlo¨g við Hreðavatn. Na´ttu´rufræðingurinn, 75, 21–33. Grı´msson, F. & Denk, T. (2005). Fagus from the Miocene of Iceland: Systematics and biogeographical considerations. Rev. Palaeobot. Palynol., 134, 27–54. Grı´msson, F. & Denk, T. (2007). Floristic turnover in Iceland from 15 to 6 Ma extracting biogeographical signals from fossil floral assemblages. J. Biogeogr., 34, 1490–1504.

Grı´msson, F., Denk, T. & Sı´monarson, L. A. (2007). Middle Miocene floras of Iceland – the early colonization of an island. Rev. Palaeobot. Palynol., 144, 181–219. Hara, H. & Kanai, H. (1964). The discovery of Tetracentron in East Nepal. J. Jpn Bot., 39, 193–195. Iljinskaja, I. A. (1968). Neogene floras of the Transcarpathian region of the U.S.S.R. Moskva: Nauka. Iljinskaja, I. A. (1972). Correction of the volume of the genus Trochodendron and new fossil species of Cocculus. (In Russian) Bot. J., 57, 17–30. Iljinskaja, I. A. (1974). Tetracentraceae. In A. Takhtajan (Ed.), Fossil flowering plants of the USSR. Vol 1, MagnoliaceaeEucommiaceae (pp. 113–115). Leningrad (St. Petersburg): Nauka. Kryshtofovich, A. N. & Baikovskaja, N. T. (1966). Selection from the works of Kryshtofovich, 3, 281–283. Kvacˇek, Z. & Hably, L. (1991). Notes on the Egerian stratotype flora at Eger (Wind brickyard), Hungary, Upper Oligocene. Ann. Hist.Nat. Mus.Natl Hungar., 83, 49–82. Kvacˇek, Z. & Rember, W. C. (2000). Shared Miocene conifers of the Clarkia flora and Europe. Acta. Univ. Carol. Geol., 44, 74–86. Manchester, S. R. & Chen, I. (2006). Tetracentron fruits from the Miocene of western North America. Int. J. Pl. Sci., 167, 601–605. Manchester, S. R., Crane, P. R. & Golovneva, L. B. (1999). An extinct genus with affinities to extant Davidia and Camptotheca (Cornales) from the Palaeocene of North America and Eastern Asia. Int. J. Pl. Sci., 160, 188–207. McClain, A. M. & Manchester, S. R. (2001). Dipteronia (Sapindaceae) from the Tertiary of North America and implications for the phytogeographic history of the Aceroideae. Am. J. Bot., 88, 1316–1325. Ozaki, K. (1987). Tetracentron leaves from the Neogene of Japan. Trans. Proc. Palaeontol. Soc. Japan, 146, 77–87. Pigg, K. B., Dillhoff, R. M., DeVore, M. L. & Wehr, W. C. (2007). New diversity among the Trochodendraceae from the Early/Middle Eocene Okanogan highlands of British Columbia, Canada, and northeastern Washington State, United States Int. J. Pl. Sci., 168, 521–532. Praglowski, J. (1974). The pollen morphology of the Trochodendraceae, Tetracentraceae, Cercidiphyllaceae and Eupteleaceae with reference to taxonomy. Pollen Spores, 16, 449–467. Suzuki, K. (1967). Discovery of Tetracentron leaves from the Neogene in Japan. Proc. Japan Acad., 43, 526–530. Suzuki, M., Joshi, L. & Noshiro, S. (1991). Tetracentron wood from the Miocene of Noto Peninsula, central Japan, with a short revision of homoxylic fossil woods. Bot. Mag. (Tokyo), 104, 37–48. Tanai, T. (1981). The revision of the so-called ‘‘Cercidiphyllum’’ leaves from the Paleogene of North Japan. J. Fac. Sci. Hokkaido Univ., 4 Ser., 19, 451–484. The Times Atlas of the World. (1958). The World, Australasia & East Asia, Vol. 1 Bartholomew, J. (Ed.), London: Times Publ. Co. Wolfe, J. A. (1977). Paleogene floras from the Gulf of Alaska region. USGS Prof. Pap., 997, 1–107.