The numerous dispersed and well preserved Annalepis specimens described in the present paper have been collected in the lower part of the Kayitou Formation at the Mide section, within Xuanwei City of the Yunnan Province and at the Tucheng section, Panxian County, Guizhou Province (Fig. 1). These sections include the Upper Permian Xuanwei Formation and the lower part of the Kayitou Formation. The Xuanwei Formation consists of terrestrial clastics sediments (sandstones and siltstones), interbedded with coal beds and/or seams. It is conformably overlain by the Kayitou Formation, mainly composed of siltstones and mudstones interbedded with sandstone.

The elements of the Gigantopteris-flora, such as Stigmaria, Lepidodendron, Paracalamites, Gigantopteris, Gigantonoclea and Pecopteris which are relics, come from the Kayitou Formation in the Mide and Tucheng sections. Annalepis, Stigmaria, Pecopteris and Lepidopteris (al. Callipteris) martinsii-like foliage associated with Peltaspermum radially symmetrical peltate ovuliferous discs have been found in the Beds 19 and 21 of the Mide section and in Beds 17 and 18 of the Tucheng section (Fig. 2) (Yu, 2008) where they are associated with Early Triassic marine bivalves ¹ 1

Some of those marine species are now considered as to appear in the Latest Permian and to extend into the Triassic since the new PTB based on conodonts has been accepted (i.e. “transitional forms”) (Rong and Fang, 2004).

Very recently (Kustatscher et al., 2010), studying an outstanding Triassic material from the Dolomites, considered “the two genera Annalepis and Lepacyclotes to be identical” following the emendation proposed by Retallack (1997). We do not follow, at the moment in the framework of this paper, this treatment. The emended diagnosis given by Retallack for the genus Lepacyclotes Emmons 1856, making Annalepis Fliche 1910 falling down as a junior synonym, is mainly based on the cormus organisation. He postulated that: “separation of Annalepis from Lepacyclotes can no longer be supported, considering that the latter probably: ( !) also has microspores of Aratrisporites”. Kustatscher et al. are, without any doubt, right to put their fossil remains into Lepacyclotes as this material is complete enough to do so. Furthermore, it appears that their new species Lepacyclotes bechstaedtii would had been a better new type species for this genus than the Rettalack's emendation of Lepacyclotes circularis Emmons, insofar as the ICBN would allow it. In our present work, we are dealing only with disarticulated sporophylls, undoubted related to quillworts but unable to provide enough comparative characters to be definitely reported to Lepacyclotes. In our mind the naming Annalepis is to be considered here only as a form genus to accommodate isolated sporophylls of still uncertain generic affinity (as initially stated besides). By the way, it is clearly out of the scope of this paper to perform an extensive revision of the numerous fossil remains named as Annalepis to attribute each to one or another genus within the complex Lepacyclotes–Annalepis–Tomiostrobus–Skilliostrobus Isoetites… being clear, besides, that we still need to discover which kind of cormus were bearing it, especially for the Chinese remains.

From the distribution pattern in time and space of the Annalepis-type fossils (Fig. 3) and the related genera (Grauvogel-Stamm and Lugardon, 2001) coupled with the extant Isoetes (Hoot and Taylor, 2001 and Liu et al., 2004), it has been inferred that the Annalepis-type sporophylls originated almost simultaneously from the Early Triassic in the Russian South-East Siberia, where they were named as Tomiostrobus by Neuburg (1936) and in Australia, named there as Cylostrobus by Helby and Martin (1965) and Skilliostrobus by Ash (1979). Moreover some authors suggested that they have migrated along the northern and the southern coasts of the Tethys, and spread to China (Annalepis), Germany (Annalepis) and France (Annalepis) on the northern coast of the Tethys during the Middle Triassic (Liu et al., 2004). Similarly, they suggested that the Middle Triassic Annalepis of China might have derived from the Annalepis-type Cylostrobus of the Early Triassic of Australia. However, since Annalepis is now known to have appeared in the Induan Kayitou Formation (such as the Mide and Tucheng sections) in South China, it seems that this taxon may also have derived from South China. Moreover, according to the distribution characteristics of the Annalepis horizon-bearing in the Yangtze River valley (Meng et al., 2000), it seems that the taxon Annalepis extended to this region during the Early Anisian (early Middle Triassic) and that its distribution gradually shrunk during the Middle Anisian-Ladinian (middle-late Middle Triassic), in response to the marine regression that occurred from east to west on the Yangtze Platform during that period (Meng, 1998 and Meng et al., 2000).

It is very significant that in eastern Yunnan and western Guizhou border the genus Annalepis occurred in marine sediments together with typical Induan bivalves, ostracods, brachiopods and ammonites (Fig. 4). These observations indicate that the first occurrence of Annalepis is Early Induan in southwestern China, thus much earlier than in the Yangtze River area (Fig. 3). During the late Early and Middle Triassic, they might have migrated to the north and widely spread over the middle and lower reaches of the Yangtze River, in relation to the marine regression and transgression.

Zhou and Li (1979) were the first to describe Olenekian plants in South China, and to report that the fossil plants of the Lingwen Formation from the Qionghai District, Hainan Island, though fragmentary, are rich in individuals, comprising more than 20 species belonging to 18 genera. Later, Li (1995) added 38 species belonging to 28 genera, among which Equisetites, Phyllotheca, Neocalamites, Neuropteridium, Albertia, Voltzia, Glyptolepis etc. The fossil plants of the Lingwen Formation are not only very similar to those of the “Buntsandstein” flora of western Europe, but also the plant-bearing beds of this formation are very similar to it in lithology. Thus, the Neuropteridium-Albertia-Voltzia assemblage present in the Lingwen Formation undoubtedly also belongs to the late Early Triassic, i.e. the Olenekian, in China.

Yao et al. (1980) indicated that the fossil fauna (including the bivalves Towapteria aff. scythica, Leptochondris bittneci, Neoschizodus ovata, Unionites fassaensis, Pteria ussurica variabilis; the brachiopod: Lingula borealis; the insect Tomia) and the fossil plants (Permian relics Paracalamites stenocostatus, Annularia shirakii, Lobatannularia multifolia, Pecopteris sp., Gigantopteris sp., and now Annalepis which was not yet reported) appeared simultaneously in the lowest part of the Kayitou Formation along the Yunnan and Guizhou border (Yao et al., 1980). This stratum corresponds to the Claraia wongi zone of the lowest part of the Feixianguan Formation, which is Induan in age (Fig. 4) (Zhou and Li, 1979). These elements of the Late Permian Gigantopteris-flora should have survived into the early Early Triassic (Wang, 1991 and Yu et al., 2007). Our study indicates that the fossils showed in Fig. 2 from the Mide and the Tucheng sections are consistent with the above-mentioned fossils, except that Annalepis spp. were not recorded in Yao et al. (1980) paper. Thus we demonstrate that the relics of the former Late Permian Gigantopteris-flora and the Early Triassic Annalepis representatives co-formed the floral assemblage of the Early Triassic in South China. This assemblage is undoubtedly older than the Neuropteridium-Albertia-Voltzia assemblage recorded from the Lingwen Formation. Therefore, the Induan age of this floral assemblage fills the biostratigraphic gap between the Changhsingian Gigantonoclea guizhouensis-Ullmannia cf. bronnii-Annularia pingloensis assemblage defined by Li et al. (1995) and the Olenekian Triassic Neuropteridium-Albertia-Voltzia assemblage reported by Zhou and Li (1979).

Until now the occurrence of the genus Annalepis was considered as a good biomarker for dating the late Early to Middle Triassic period. Therefore, it is highly relevant that in eastern Yunnan and western Guizhou, this genus occurred associated with typical Induan bivalves, ostracods, brachiopods and ammonites in the marine basal deposits of the Kayitou Fm (Fig. 5). This indicates that the First Appearance Datum (F.A.D.) of the “pioneering lycopsid genus” Annalepis is in fact Early Triassic (Induan) in southwestern China, where it newly occurred associated with surviving Permian plants, thus much earlier than in the Yangtze River domain and any other area in South-East Asia. It is now to be considered as marking the very beginning of Triassic deposition for the western Guizhou and eastern Yunnan domain. During the Early and Middle Triassic, it might have migrated to the north and spread widely over the other areas of South-East Asia.