植 物 分 类 学 报 43(2): 182–190(2005) Acta Phytotaxonomica Sinica Emendation of Sorosaccus gracilis Harris 1935, a gymnospermous pollen cone 1, 2 LIU Xiu-Qun 3 Francis M. HUEBER 1 LI Cheng-Sen* 1 WANG Yu-Fei 1(Laboratory of Systematic and Evolutionary Botany, Institute of Botany, the Chinese Academy of Sciences, Beijing 100093, China) 2(Graduate School, the Chinese Academy of Sciences, Beijing 100039, China) 3(Department of Paleobiology, Smithsonian Institution, Washington, DC 20560-0121, USA) Abstract Study of gymnospermous pollen cones, identified with Sorosaccus gracilis Harris 1935, from the Yangcaogou Formation, Late Triassic of China, has led to the identification of new and significant characteristics of the species. The new specimens show distinct variations in the morphology of the distal laminar portions of the microsporophylls. These well- preserved fossils are helpful in elucidating, reconstructing and revising the diagnostic features of S. gracilis. After comparing the characteristics of the specimens from China with those of S. sibiricus Prynada 1962, and of the pollen cones which were respectively assigned to Baiera longifolia (Pom.) Heer 1876 from Siberia, Russia, S. minor Harris 1935 from Greenland, and S. umaltensis Krassilov 1972 from Bureya River of Russia, we consider that these four names should be reduced to the synonymy of S. gracilis. Thus we revise the diagnosis of Sorosaccus. The significance of Sorosaccus in evolution of pollen cones is discussed. We consider that Sorosaccus is possibly basic to the evolution of the genus Ginkgo by the reduction of the number of microsporangia and of microsporophylls. Ginkgo liaoningensis Liu, Crane, Li & Wang 2005 from the Early Cretaceous of Liaoning Province, China represents likely one of morphologically intermediate steps between Sorosaccus and G. biloba in evolution. Key words Sorosaccus gracilis, pollen cone, evolution, Late Triassic, Liaoning. The genus Sorosaccus was established by Harris in 1935, based on pollen cones initially found from the Neil Cliff, Liverwort Bed belonging to the Thaumatopteris Zone, in the fossil flora of Scoresby Sound East Greenland (Harris, 1935). The pollen cones were always associated with ginkgoalean leaves in East Greenland, and “their microspores were of a type which occurred in the Ginkgoales and many Gymnosperms” (Harris, 1935). But the genus is dubious in taxonomic position or is unknown in evolutionary trends. So far, only four species have been contained in Sorosaccus: S. gracilis Harris 1935, S. minor Harris 1935, S. sibiricus Prynada 1962 and S. umaltensis Krassilov 1972. In this paper, the specimens of pollen cone Sorosaccus gracilis from Yangcaogou Formation (Late Triassic) in Beipiao County, Liaoning Province, China, were studied, and its likely taxonomic affinity and evolutionary trends were considered. The Late Triassic flora from Yangcaogou Formation contains Thallites, Neocalamites, Annulariopsis, Danaeopsis, Dictyophyllum, Todites, Cladophlebis, Nilssonia, Pterophyllum, Sphenobaiera, Baiera, Ginkgoites, Glossophyllum, Podozamites, Pityophyllum, Cycadocarpidium, Taeniopteris, Stenorachis (Zhou, 1981). ——————————— Received: 13 January 2005 Accepted: 26 January 2005 Supported by a Kowledge Innovation Project of the Chinese Academy of Sciences (KSCX2-SW-108). * Author for correspondence. Fax & Tel.: +86-10-62593385. E-mail: <[email protected]>. No. 2 LIU Xiu-Qun et al: Emendation of Sorosaccus gracilis Harris 1935, a gymnospermous pollen cone 183 1 Material and methods The specimens were collected from Yangcaogou Formation in Zhaomagou Village, Beipiao County, Liaoning Province, China (Fig. 1). Yangcaogou Formation is regarded as Late Triassic in age (Zhou, 1981; Zhang & Zheng, 1987). They are preserved as impressions in a matrix of gray to yellow mudstone. The specimens were observed under a stereomicroscope Stemi SV 11 and photographed using a Nikon FM2 camera. All specimens are housed in the National Museum of Plant History of China, Institute of Botany, the Chinese Academy of Sciences, Beijing, China. Fig. 1. Map showing fossil locality. 2 Description Cones are roughly catkin-like (Figs. 2-5), cylindrical (Figs. 3, 4, 10), 35-45 mm in length, 6-10 mm in diameter. They consist of a main axis with microsporophylls arranged helically (Fig. 6). Some specimens (Figs. 2, 6-9) clearly show the microsporophylls with their arrangement and microsporangia. The main axis, 1-2 mm in diameter, possesses faintly striated surface (Figs. 3, 4). Microsporophylls, 30-40 in number, arise and are directed upwardly at acute or obtuse angles to the main axis. The distance from the base of the axis to the lowest microsporophyll is 10-15 mm. The interval between microsporophylls is 1-1.5 mm. Microsporophylls consist of a lamina bearing microsporangia on the abaxial surface. The microsporophyll is divided into two parts: a petiole and an upwardly bending, expanded lamina. The petiole is 2-3 mm long and 0.1-0.5 mm wide. The lamina is 2-5 mm long and 0.2-2.5 mm wide, bending upward and exhibiting considerable variation in outline from linear-lanceolate to oval (Fig. 11). Some laminae are broken from their petioles away from the main axis (Fig. 2). Microsporangia, in clusters of 6 to 8, are attached to the lateral and the abaxial sides of the petiole. Each microsporangium is oval, 1-1.5 mm long and 0.5-1 mm wide, and dehisces longitudinally. 184 Acta Phytotaxonomica Sinica Vol. 43 Figs. 2-5. Hypotypes of Sorosaccus gracilis. The specimen in Fig. 4 is the counterpart of the specimen in Fig. 3. The numbers (2A, 2B, 2C, 2D, 2E, 2F, 2G; 3A, 3B, 3C, 3D; 5A, 5B, 5C) show the microsporophylls with microsporangia. Scale bar = 5 mm. 3 Systematic palaeobotany DIVISION GYMNOSPERMOPHYTA ORDER INCERTAE SEDIS FAMILY INCERTAE SEDIS Sorosaccus Harris 1935 emend. Liu, Hueber, Li & Wang Emended generic diagnosis: Pollen cone consisting of a main axis with microsporophylls arranged helically. Microsporophyll consisting of petiole and lamina bearing microsporangia. Microsporophyll arising to the axis; lamina bending upward. Microsporangia attached in a cluster on the lateral and abaxial sides of the petiole. Microsporangium oval, dehiscing longitudinally. Type: Sorosaccus gracilis Harris 1935. Sorosaccus gracilis Harris emend. Liu, Hueber, Li & Wang 1935. Sorosaccus gracilis Harris, pl. 24, figs. 3-7, 11, 12; pl. 28, fig. 11; text-fig. 50, E, F, H. Synonyms: 1876. Baiera longifolia Heer, Pars. 53, pl. IX, figs. 8, 9. 1935. Sorosaccus minor Harris, pl. 27, fig. 2. 1962. Sorosaccus sibiricus Prynada, pl. XIII, figs. 7, 8; pl. XVIII, figs. 4a, 5-7; pl. XXI, fig. 5. 1972. Sorosaccus umaltensis Krassilov, pl. XIX, figs. 1, 2, 4-7, 9. No. 2 LIU Xiu-Qun et al: Emendation of Sorosaccus gracilis Harris 1935, a gymnospermous pollen cone 185 Figs. 6-9. Magnification of the hypotypes of Sorosaccus gracilis Harris 1935 emend. Liu, Hueber, Li & Wang. 6. Magnification of the part of the specimen of Fig. 2 and the numbers are corresponding to those in Fig. 2. 7. Magnification of the left side of the specimen in Fig. 2 and the letter A is corresponding to that in Fig. 2. 8. Magnification of the middle part of the right side of the specimen in Fig. 2 and the letter F is corresponding to that in Fig. 2. 9. Magnification of the top of the specimen in Fig. 3 and the letters A and C are corresponding to those in Fig. 3. Scale bar=2 mm. 186 Acta Phytotaxonomica Sinica Vol. 43 1972. Sorosaccus ex. gr. sibiricus Prynada 1962, in Krassilov, Mesozoic Flora of Bureya River (Ginkgoales and Czekanowskiales), pl. XIX, figs. 3, 8. Nauka, Moscow. Emended specific diagnosis: Pollen cone elongate-cylindrical, catkin-like, 20-45 mm long and 5-10 mm wide. Main axis with micro- sporophylls arranged helically, cylindrical, 1-2 mm in diameter, surface faintly striated. Microsporo- phyll divided into two parts: horizontal petiole and an expanded lamina. Lamina linear, lanceolate to oval in outline, bending upward. Microsporangia attached to lateral and abaxial sides of petiole, 6-8 in number. Microsporangium oval, 0.7-1.5 mm long, 0.5-1.0 mm wide and longitudinal in dehiscence. Holotype: Sorosaccus gracilis Harris, 1935. pl. 24, figs. 5, 6. Type locality: Scoresby Sound East Greenland, Denmark. Horizon: Neil Cliff Formation, Liverwort Bed of Rhaetic (Late Triassic) and Lower Liassic (Early Jurassic) (Harris, 1937). Epitypes: Figs. 2-5. LN-8486, LN-8487, ZMG-8488, ZMG-8489. Locality: Zhaomagou Village, Beipiao City, Liaoning Province, China. Horizon: Yangcaogou Formation, Late Triassic (Zhou, 1981; Zhang & Zheng, 1987). Repository: National Museum of Plant History of China, Beijing, China. Distribution: East Greenland, Denmark; Siberia, Russia; Liaoning, China. 4 Comparison We compare four species in Sorosaccus: S. gracilis Harris 1935, S. minor Harris 1935, S. sibiricus Prynada 1962 and S. umaltensis Krassilov 1972. Harris (1935) separated Sorosaccus minor from S. gracilis stating that the former was “differing in being more slender owing to its smaller sporophylls (basal region 2 mm long) and smaller sporangia (0.7×0.5 mm)”. The sporangium F ig. 10. Reconstruction of Sorosaccus gracilis H arris 1935 emend. Liu, Hueber, Li & Wang. of S. gracilis, 1.1×0.8 mm, is bigger than that of the type specimen of S. minor (Harris, 1935, pl. 27, No. 2 LIU Xiu-Qun et al: Emendation of Sorosaccus gracilis Harris 1935, a gymnospermous pollen cone 187 Fig. 11. Variation of microsporophylls without microsporangia of Sorosaccus gracilis in outline. The letters A-N show respectively the drawings of the microsporophylls without microsporangia in Fig. 3A, Fig. 2A, Fig. 2B, Fig. 2D, Fig. 3B, Fig. 2C, Fig. 5A, Fig. 2E, Fig. 3D, Fig. 5B, Fig. 2G, Fig. 5C, Fig. 3C, and Fig. 2F. fig. 2), 0.7×0.5 mm (Table 1). But the size of the sporangium in our specimens, 1-1.5× 0.5-1 mm, covers that of S. minor. We consider that the different sizes show the specimens in different phases of development. The size of sporophylls in S. minor, with the petiole 2 mm long, is similar to that in S. gracilis based on Harris’ description, which is 2-4 mm, and the size is in the range of that of S. gracilis based on our collections, which is 1-3 mm long. It is very difficult to differentiate S. minor from S. gracilis based only on the size of basal stalks and microsporangia. We transfer therefore the specimens assigned to S. minor to S. gracilis. 188 Acta Phytotaxonomica Sinica Vol. 43 Table 1 Comparison of specimens assigned to Sorosaccus gracilis Harris 1935 emend. Liu, Hueber, Li & Wang Specimen* Locality Age Size of Diameter Length of Angle of Numbe Size of Size of Reference male of axis petiole* petiole* r of MIC pollen cone (mm) (mm) with axis MIC (mm) grain (mm) (µm) S. gracilis Greenland, LT, 30×8 1 2-4 right angle 6 1.1×0.8 90×60 Harris, UK EJ 1935 S. minor Greenland, LT, 2 0.7×0.5 Harris, UK EJ 1935 S. sibiricus Siberia, MJ 20× right angle 8 Prynada, Russia 5-7 1962 S. umaltensis Bureya LJ 18×8 0.5 3.5-4.5 right angle 8 1.0×0.8 40×27 Krassilov, River, (Not - 1972 Russia intact) 50×36 Specimen Liaoning, LT 35-45 1-2 3 right or 6-8 1-1.5× In this from China China × obtuse 0.5-1 paper 6-10 angle Notes: Specimen*=Specimens assigned to Sorosaccus gracilis Harris 1935 emend. Liu, Hueber, Li & Wang in this paper; Petiole*=Petiole of microsporophyll; MIC=Microsporangium; LT=Late Triassic; EJ=Early Jurassic; MJ=Middle Jurassic; LJ=Late Jurassic. Specimens of Sorosaccus sibiricus was collected from the Middle Jurassic of eastern Siberia (Prynada, 1962). Prynada discussed that “the sporophylls of Ust’-Baley specimens are similar with those from the Lower Jurassic deposits of Greenland but differing from them in the size of the whole cones and close arrangement of sporophylls”. In fact, the sporophylls of Ust’-Baley specimens are similar to those of Greenland in the size of the entire cones and arrangement of sporophylls. Firstly, the size of the specimens assigned to S. sibiricus, 20× 5-7 mm, is covered in that of S. gracilis based on new materials from China, 35-45×6-10 mm (Table 1). The number of microsporangia in S. sibiricus (8 in number) is similar to that of S. gracilis (6-8 in number). Krassilov (1972) thought that dimensions of pollen cones ascribed to S. sibiricus by Prynada were insignificantly distinguished from those by Harris. Thus we move the specimens assigned to S. sibiricus to S. gracilis. The specimens of Sorosaccus umaltensis were collected from the sediments of the Upper Jurassic at the right bank of Bureya River, Russia (Krassilov, 1972). Krassilov (1972) considered that the reproductive organs from Bureya River coincide almost completely with those of S. gracilis in dimensions of the pollen cones, microsporophylls and microsporangia. But the pollen grain size of S. umaltensis, 40×27-50×36 µm, is smaller than that of S. gracilis (90×60 µm) (Harris, 1935). It is uncertain to describe a new species based on the size of pollen grains since the size of pollen grains varies distinctly in different phases of development. We also transfer the specimens assigned to S. umaltensis to S. gracilis. Four specimens (Heer, 1876, pl. IX, figs. 8-11) were described as the pollen cones of Baiera longifolia based on their general agreement with the pollen cones of Ginkgo and on their association with the leaves. Harris (1935) considered that they were exactly is the same nature of Sorosaccus gracilis. After our observation, two specimens (Heer, 1876, pl. IX, figs. 8, 9) from the four show that the cone axis bears the bracts with extended laminae that bend upward and bear microsporangia. Therefore, we considered that the two belong to Sorosaccus gracilis. No. 2 LIU Xiu-Qun et al: Emendation of Sorosaccus gracilis Harris 1935, a gymnospermous pollen cone 189 5 Discussion We are trying to relate the genus Sorosaccus with ginkgoaleans and deduce its evolution based on the oval microsporangia dehiscing longitudinally and the pollen like that in Ginkgoales. Comparing the whole pollen cone of Sorosaccus gracilis with the pollen cone of Ginkgo biloba (the only extant species of the order Ginkgoales) and Ginkgo liaoningensis (a pollen cone collected from the Yixian Formation of Early Cretaceous, Liaoning, China) (in our another paper which is accepted by Botanical Journal of the Linnean Society, “The pollen cones of Ginkgo from the Early Cretaceous of China, and its bearing on the evolutionary signific- ance”), it is found that 6-8 micro- sporangia per cluster are attached to lateral and abaxial sides of micro- sporophyll in Sorosaccus gracilis, and 3-4 pendulous microsporangia are attached to the abaxial side of microsporophylls in Ginkgo liaoning- ensis and 2 in G. biloba. As these pollen cones are quite similar in outline, we consider that Sorosaccus could have evolved to Ginkgo through the reduction of the number of sporangia and shortening the distal portion of the lamina at the distal portion of the microsporophyll (Fig. 12). Ginkgo liao- ningensis from the Early Cretaceous Fig. 12. Schematic drawin g showing the evolutionary lineage of several pollen cones from Sorosaccus gracilis to could represent one of morphologically Ginkgo biloba. intermediate steps between Sorosaccus and Ginkgo biloba in the evolutionary course of the ginkgoalean pollen cones. Another transitional type between Sorosaccus and Ginkgo (G. liaoningensis), probably occurring in the period of the Middle-Late Jurassic (Fig. 12), is interpreted to contain a shortened and narrowed lamina at the distal portion of the microsporophyll. It suggests that the number of sporangia may have decreased in the evolutionary process, and the length of the lamina at the distal portion of the microsporophyll may have shortened. The hypothesis of the evolutionary trend of the pollen cones needs further confirmation after obtaining specimens of pollen cones in organic connection with foliar axes. 190 Acta Phytotaxonomica Sinica Vol. 43 Acknowledgements We thank Professor Albert Ablaev (Pacific Oceanological Institute, Far East Branch, Russian Academy of Sciences, Russia), Professor Svetlana V. Syabryaj (Institute of Geological Sciences, National Academy of Sciences of Ukraine) and Professor Jan J. Wocicki (W. Szafer Institute of Botany, Polish Academy of Sciences, Poland) for providing and translating Russian literature. We also thank Mr. Ying-Bao Sun (Institute of Botany, the Chinese Academy of Sciences, Beijing, China) for his drawings. The work was supported by a Kowledge Innovation Project of the Chinese Academy of Sciences (KSCX2-SW-108). References Harris T M. 1935. The Fossil Flora of Scoresby Sound East Greenland, Part 4: Ginkgoales, Coniferales, Lycopodiales and Isolated Fructifications. Meddelelser om GrØnland 112 (1): 145-146. Harris T M. 1937. The Fossil Flora of Scoresby Sound East Greenland, Part 5: Stratigraphic Relations of the Plant Beds. Meddelelser om GrØnland 112 (2): 67. Heer O. 1876. Beiträge zur Jura-Flora Ostsibiriens und des Amurlandes. Mémoires de L’académie Impériale des Sciences De St.-Pétersbourg, VIIE Série. 22: 53. Krassilov V A. 1972. Mesozoic Flora of Bureya River (Ginkgoales and Czekanowskiales). Moscow: Nauka. 96-97. Prynada V D. 1962. Flore Mesozoique de Siberie Orientale et de Transbaikalie. Moscow: Nauka. 368. Zhang W (张武), Zheng S-L (郑少林). 1987. Early Mesozoic fossil plants in Western Liaoning, China. In: Yu X-H (于希汉), Wang W-L (王五力), Liu X-T (刘宪亭), Zhang W (张武) eds. Mesozoic Stratigraphy and Palaeontology of Western Liaoning, China (辽宁西部中生代地层古生物). Beijing: Geological Publishing House. 239-338. Zhou H-Q (周惠琴). 1981. Discovery of the Upper Triassic flora from Yangcaogou of Beipiao, Liaoning. In: The Palaeontological Society of China ed. 12th Annual Conference of the Palaeontological Society of China Selected Papers (中国古生物学会第十二届学术年会论文集). Beijing: Science Press. 188. 裸子植物雄球花——纤细堆囊穗的修订 1, 2刘秀群 3 Francis M. Hueber 1李承森* 1王宇飞 1(中国科学院植物研究所系统与进化植物学重点实验室 北京 100093) 2(中国科学院研究生院 北京 100039) 3(Department of Paleobiology, Smithsonian Institution, Washington, DC 20560-0121, USA) 摘要 研究了采自中国东北晚三叠世羊草沟组的一种裸子植物雄球花——纤细堆囊穗。通过对新材料 的研究,发现这种雄球花小孢子叶末端的裂片在形态上差异很大,这是原来没有发现的十分重要的特征。 这些保存精美的化石对于纤细堆囊穗特征的阐明和修订以及物种复原很有帮助。将中国的标本与西伯 利亚堆囊穗、小堆囊穗、乌尔马堆囊穗和被定为长叶拜拉的雄球花进行了比较,发现它们与纤细堆囊穗在 特征上一致,故将它们处理为纤细堆囊穗的异名。修订后的纤细堆囊穗包括上面所提到的所有种。同时, 也讨论了堆囊穗属可能的演化意义。它可能是银杏属的远祖,经过小孢子囊数目的减少和小孢子叶长度 的缩短而演化到现在的银杏,而产自辽西早白垩世的辽宁银杏可能代表了堆囊穗和现代银杏在形态演 化上的一个中间步骤。 关键词 纤细堆囊穗; 雄球花; 演化; 晚三叠世; 辽宁