步林在青海柴达木盆地的早期工作记录——经典脊椎动物化石地点与现代地层框架的解译(英文)

EARLY EXPLORATIONS OF QAIDAM BASIN (TIBETAN PLATEAU) BY BIRGER BOHLIN—RECONCILING CLASSIC VERTEBRATE FOSSIL LOCALITIES WITH MODERN BIOSTRATIGRAPHY

1931 ～ 1932 年,博格·步林两次率领中瑞考察团( 又名斯文·赫定考察团) 到青海省柴达木盆地东部进行考察,在克鲁克湖与托素湖地区首次发现了丰富的新近纪哺乳动物化石。这批化石成为我国早期青藏高原古生物研究的奠基石。直至今日,双湖地区仍是青藏高原脊椎动物化石最丰富、研究程度最高的地区。柴达木盆地新近纪的堆积巨厚且连续,地层构造相对简单,盆地北缘地层出露好,是研究陆生哺乳动物演化的理想地点。然而,双湖地区地层的时代跨度超过 13 My,厚度接近 4600 m,如果化石层位不明,很容易引起动物群成分的混乱。步林对此似乎缺乏足够的认识,对所有采集的化石未做分层处理,导致后人对"柴达木动物群"年代的解释引起不少异议。步林并非对地层不重视。恰恰相反,他如同在甘肃塔奔布鲁克盆地的工作中那样,在缺乏地形图和地质资料的条件下做了大量的基础地层工作。他利用明显的地物( 山顶)和地质构造( 背斜) 作为参照系统,仔细地记录了化石的出露地点,并编制了一些草图。步林去世后,家属把他的野外记录留给赫定档案馆( Svan Hedin Archives) ,在斯德哥尔摩分几处收藏。遗憾的是,如同塔奔布鲁克盆地的情况,步林未能发表很多十分关键的化石地点资料,而发表的地质记录却被多数人忽略,所出版的游记也鲜为人知,个人档案记录更是深埋于历史的文献中。直到 1999 年,本文第一作者才首次接触到一些步林的野外原始资料。经过近 10 年多次对赫定档案馆的访问和对步林原始记录反复的野外核实,最终才将步林的经典地点结合进现代地层顺序的框架中。本文是继对步林塔奔布鲁克盆地燕丹图动物群经典地点与地层的研究后,进一步对其柴达木盆地化石地点的厘定与复原,是本系列文章之二。目的是将过去混淆于各地点的动物群逐一解译,将不同时代的化石分子置于现代地层与年代框架之中。幸运的是,步林的诸多野外记录都保存完好,结合发表的个人游记及地质记录,他所发表的多数哺乳动物化石都可以恢复到一定地层范围之内,尚能尽量减小经典化石地点的不确定性。本文试图整理步林未发表的珍贵资料,通过综合考证其他已发表的信息,结合作者在 1998 ～2010 年多次野外考察的结果,以期理顺柴达木地区经典化石点与现代地层框架的关系。这里将着重探讨经典化石点的生物地层学问题,并将步林柴达木盆地化石地点的野外记录资料译成英文附入文后,至于我们在柴达木盆地发现的其他化石点暂不涉及。步林主要在三个地区采集了化石: 托素湖北岸、怀头他拉村南及泉水梁火车站附近。尽管步林清楚地意识到其化石来自几千米厚的地层,但他统称其为柴达木动物群,成为以后混乱的主要原因。根据对柴达木新近纪的地层古生物考察,我们至少可以辨认出 4 个动物群: 中中新世欧龙布鲁克动物群、晚中新世早期托素动物群、深沟动物群和早上新世怀头他拉动物群。( 1) 中中新世欧龙布鲁克动物群。化石主要产于剖面底部克鲁克背斜两翼接近轴部的红色泥岩及绿色砂岩中。该动物群的成员发现较少,步林和我们都找到不多,主要来自两个地点: 怀头他拉剖面步林 165 号营地附近及泉水梁火车站东北 800 m 左右处。包括以下几个种类( 括号内为步林的化石编号; 该系列号码在步林的柴达木专著中广泛应用) : Olonbulukia tsaidamensis ( Nr. 356) ,Lagomeryx tsaidamensis ( Nr. 361) ,Stephanocemas( Nr. 368,407) ,? Dicroceros ( Nr. 372,397) ,? Eostyloceros ( Nr. 396,399,401) ,Acero-rhinus tsaidamensis ( Nr. 330,346,374,381) 。( 2) 晚中新世早期托素动物群。该动物群最初发现于托素湖北岸的狭长露头。步林对该点的记录最详细,其中各化石层位都被我们一一找到。托素湖的含哺乳动物化石层位局限在诺令根河( 连接克鲁克湖与托素湖之间的小河) 北岸 100 m 的地层中,是步林所有化石点中能最精确地确定层位的一个点( 即步林托素湖化石点都可以恢复到该段 100m 地层中) 。根据目前古地磁测量结果,该点 100 m 地层可以精确到 0. 3 Ma 范围之内。相当于含托素动物群层位的化石地点在怀头他拉剖面中也少量存在,但多数化石不特征,估计步林在此也找到不多。托素动物群最丰富的化石点在泉水梁车站东南的狭长露头。步林称做"一般条带"( "General Strips") 。在厚度不到250 m,北西-南东方向延伸达8 km的地层中,化石均匀散布、保存完好,是柴达木目前最丰富的化石点。托素动物群包括以下种类( 括号内为步林的化石编号) : "Ictitherium sp. "( Nr. 542) ,Tsaidamotherium hedini( Nr. 447,457) ,Tossunnoria pseudibex ( Nr. 449,481,537) ,Qurliqnoria cheni ( Nr. 441) ,Qurliqnoria sp. ( Nr. 508) ,Olonbulukia tsaidamensis ( Nr. 429) ,Chalicotheriidae indet.( Nr. 317) ,Hipparion sp. ( Nr. 352,444,467,493,498,500,501,523,553) ,Acerorhi-nus tsaidamensis ( Nr. 418,430,438,456,459,488,502-4,511,516,525-7,536,538,555-6) ,? Tetralophodon ( Nr. 458,487,524,530,534) ,Giraffidae indet. ( Nr. 316,320) ,? Eostyloceros ( Nr. 334,349) 。另外我们在泉水梁地点也找到大量化石,研究后可望增加更多的成员。泉水梁是我国少数几个三趾马出现最早的地点之一。该剖面目前还没有古地磁资料,但通过与怀头他拉古地磁剖面的对比,泉水梁三趾马出现的最低层位应大致在 C5r. 1r反向带与 C5r. 1n 正向带之间( 约 11. 12 Ma) 。另外托素动物群所代表的时代,是青藏高原首次出现独特种类的时代。牛羊科中具有高原特色的种类包括 Tsaidamotherium hedini,Tossunnoria pseudibex,Qurliqnoria cheni 和 Olonbulukia tsaidamensis,其中 Qurliqnoria cheni可能是现代藏羚羊( Pantholops) 的祖先类型。( 3) 晚中新世早期深沟动物群。深沟动物群是根据德令哈南部深沟剖面的化石建立的。产该动物群的地层虽然在怀头他拉剖面中段出露,但目前还没找到特征的化石。步林的个别不确定的化石有可能属于深沟动物群。邱铸鼎和李强最近( 2008) 对深沟的小哺乳动物做了详细的描述,认为与陕西蓝田霸河组下部的小哺乳动物组合最接近。( 4) 早上新世怀头他拉动物群。怀头他拉动物群也处于怀头他拉剖面上部化石不丰富的地段。从步林的野外记录中,惟一能够比较确定的化石是一些驼鸟蛋片。我们也仅增加了几种小哺乳动物( 未研) 。

The discovery by Birger Bohlin of a series of vertebrate fossil sites in the twin lakes region ( Tuosu Nor and Keluke Nor) of eastern Qaidam Basin during the Sino-Swedish Expedition in 1931 and 1932 was a major milestone in vertebrate paleontology for the Tibetan Plateau. Qaidam fossil mammals collected by Bohlin still represent the best collections from the plateau and serve as an important reference point for a period of time that saw dramatic climatic changes. The more than 4600 m of strata in eastern Qaidam, spanning over 13 million years in time, are ideal for establishing a detailed biostratigraphic record, but Bohlin's published specimens lack any reference to stratigraphy, which causes much confusion about the nature of his "Tsaidam Fauna."Bohlin did, however, make a fairly detailed documentation of locality information, much of which remains buried in archival records in Stockholm. This paper is an attempt to reconstruct Bohlin's fossil localities by synthesizing relevant archives and historical accounts, as well as field observations of our own during the past 13 years. Fieldnotes in Swedish are translated to English and several field sketches are published for the first time. As a result, we are able to relocate many fossil localities that are of stratigraphic and taxonomic importance.Bohlin mainly collected in three major areas, along the northwestern shore of Tuosu Nor, south of Huaitoutala, and south of Quanshuiliang train station, each with stratigraphic settings of their own. In his published descriptions, all fossils are included in his "Tsaidam Fauna." With the help of his notes and sketches and our field verification, we are now in a position to recognize at least three faunal horizons among Bohlin's collections: middle Miocene (Tunggurian) Olongbuluk Fauna, early late Miocene (early Bahean) Tuosu Fauna, and early Pliocene (Yushean) Huaitoutala Fauna.Of these fossil sites, Bohlin's Tuosu Nor locality is the best constrained within the modern stratigraphic framework due to the relatively short section present along the northern bank of Tuosu Nor. His fossil localities are confined to a narrow band of about 100 meters in total thickness. Bohlin's Huaitoutala (Ulan-utsur) localities are, however, more scattered both spatially and stratigraphically. Fortunately, only a few localities produced fossils of biochronologic significance, and most are low in the section,probably belonging to the Olongbuluk Fauna.The majority of Bohlin's Qaidam collections comes from what he informally called the "general strips," which is not far from the present day Quanshuiliang railway station. Approximately 250 m of strata extend laterally for several kilometers. Almost all endemic bovids that Bohlin described, such as Tsaidamotherium, Olonbulukia, Qurliqnoria, Tossunnoria, are produced from this area. The Quanshuiliang region also produces some of the earliest occurrences of Hipparion in China. Preliminary correlation of the Quanshuiliang section with an existing magnetic section in Huaitoutala suggests that the local first appearance of Hipparion is close to the boundary between chrons C5r. Lr and C5r. 1n, about 11.1 Ma.