Lower Cretaceous sauropod trackways from Lishui City and an overview of dinosaur dominated track assemblages from Zhejiang Province, China

2.1 Cretaceous strata in Zhejiang Province

In Zhejiang Province, there are more than 30 basins with exposures of the Cretaceous terrestrial red beds. These basins are mainly distributed in the central and eastern areas and range from several to over 500 km² (Fig. 1). Their basements are formed by widely distributed volcanic rock series. Generally, the basins are dominated by fluvial−lacustrine detrital deposits, with a small quantity of pyroclastic components. However, the geologic classification of these strata has proven challenging. Different opinions exist, not only regarding the stratigraphic ages of different series, but also the categorization of the individual stratigraphic units (Jin et al., 2012; Fig. 2).

Zhejiang Bureau of Geology and Mineral Resources (1996) established the Upper Jurassic–Lower Cretaceous Jiande Group (Luocun, Huangjian, Shouchang and Hengshan Formations) and the Upper Cretaceous Qujiang Group (Zhongdai, Jinhua and Quxian Formations) and Tongxiang Formation, in the Yangtze stratigraphic area; and, the Upper Jurassic–Lower Cretaceous Moshishan Group (Dashuang, Gaowu, Xishantou, Chawan, Jiuliping and Zhucun Formations), the Lower Cretaceous Yongkang Group (Guantou, Chaochuan, Fangyan, Keshan and Xiaopingtian Formations) and the Upper Cretaceous Tiantai Group (Tangshang, Liangtoutang and Chichengshan Formations), in the Southeast China stratigraphic area (see Fig. 2). Chen (2000) divided the Cretaceous strata of Zhejiang Province into the Lower Cretaceous Luocun, Huangjian, Shouchang, Hengshan and Guantou Formations, and the Upper Cretaceous Chaochuan, Fangyan, Lanxi, Quxian and Tongxiang Formations. Based on a comprehensive analysis of lithostratigraphic sequence, paleontology, and isotope age, Cai and Yu (2001) suggested that the Yongkang, Tiantai and Hengjiang Groups are synchronous equivalents of different facies without overlap, accumulated between the late Early Cretaceous and the early Late Cretaceous. Cai and Yu (2001) re-established the Lower Cretaceous Jiande Group (Luocun, Huangjian, and Shouchang Formations) and the Upper Cretaceous Qujiang Group (Zhongdai, Jinhua, Hengshan [Zhongdai+Jinhua] and Quxian Formations) and the Tongxiang Formation, in the Yangtze stratigraphic area; and, the Lower Cretaceous Moshishan Group (Dashuang, Gaowu, Xishantou, Chawan and Jiuliping Formations), the Upper Cretaceous Yongkang Group (Guantou, Chaochuan and Fangyan Formations) and the Upper Cretaceous Tiantai Group (Tangshang, Liangtoutang and Chichengshan Formations), in the Southeast China stratigraphic area (Fig. 2).

2.2 Xiaqiao site

Isotope dating of volcanic rocks from the Jiande Group yields an age estimation of 134–115 Ma, indicating that the Jiande Group in this study area is similar to other Chinese Lower Cretaceous groups in age (Li et al., 2011). In addition, the Jiande biota is generally compared with the Jehol biota of the western Liaoning Province and the Jiande Group is also similar to its contemporaneous groups in the composition of biota, such as bivalves, gastropods, fish, insects, and plants (Zhejiang Bureau of Geology and Mineral Resources, 1996).

In terms of lithology and biological types, the Moshishan Group and the Jiande Group are particularly similar (Liu and Zhao, 1927; Zhejiang Bureau of Geology and Mineral Resources, 1996; Cai and Yu, 2001). Regardless of whether the Xiaqiao site belongs to the Shouchang Formation of the Jiande Group (125–121 Ma; Li et al., 2011) or to the Chawan−Jiuliping Formations of the Moshishan Group (120–109 Ma; Zhang et al., 2012), these dates indicate that the Xiaqiao site is the Aptian in age.

4.1 Description

Matsukawa et al. (2009) described these tracks but identified no trackway. However, the 3D photos analysed in this study are more discriminating in identifying track outlines and the patterns of trackways. On the mudstone level, we discovered at least 13 tracks. This is less than the 18 tracks identified by Matsukawa et al. (2009), and the reduced number is probably attributable to intervening weathering and damage caused by the roots of plants, which flourish in the wet environment near the river. However, among the 13 tracks, there were five isolated tracks (XQ-SI1–XQ-SI5; Table 2; Fig. 3) and three possible trackways (XQ-S1–XQ-S3; Table 2; Fig. 3). All tracks and trackways remain in situ.

Track	ML	MW	PL	SL	PA	ML/MW	WAP	WAP/P’ML
XQ-S1-RP1	49.4	32.8	101.8	174.7	124	1.5	45.3	0.9
XQ-S1-LP1	47.6	34.2	96.2	–	–	1.4	–	–
XQ-S1-RP2	44.4	32.9	–	–	–	1.4	–	–
Mean	47.1	33.3	99.0	174.7	124	1.4	45.3	0.9
XQ-S2-LP1	44.0	32.0	82.4	–	–	1.4	–	–
XQ-S1-RP1	–	–	–	–	–	–	–	–
XQ-S3-RP1	34.3	25.4	114.5	195.2	122	1.4	54.4	1.6
XQ-S3-LP1	38.3	24.1	108.8	157.5	94	1.6	71.8	1.9
XQ-S3-RP2	34.1	19.7	106.7	–	–	1.7	–	–
XQ-S3-LP2	40.8	21.7	–	–	–	1.9	–	–
Mean	36.9	22.7	110.0	176.3	108	1.6	63.1	1.7
XQ-SI2	53.5	40.0	–	–	–	1.3	–	–
XQ-SI3	34.7	38.4	–	–	–	0.9	–	–
XQ-SI4	30.4	21.3	–	–	–	1.4	–	–
XQ-SI5	34.5	30.9	–	–	–	1.1	–	–

Abbreviations: ML Maximum length, MW Maximum width, PL Pace length, SL Stride length, PA Pace angulation, WAP Width of the angulation pattern of pes (calculated value), P’ML Maximum length of pes. ML/MW and WAP/P’ML are dimensionless

Trackway XQ-S1 consists of three successive pes tracks and appears to be narrow-gauge, with a (WAP/P’ML)-ratio of 0.9 (P’ML stands for the maximum length of pes, Table 2; Marty, 2008). No manus impressions were recognized. The average length of the pes impressions is 47.1 cm, and the average (length/width)-ratio is 1.4. The pes impressions are oval and the metatarso−phalangeal region is smoothly curved. All tracks are poorly preserved without clear digit impression. The average pes pace angulation is 124°, which is relatively high for sauropods. Trackway XQ-S2 only preserves an incomplete single step and is morphologically similar to XQ-S1.

Trackway XQ-S3 consists of four poorly preserved pes tracks lacking distinct digit traces. The trackway pattern is between medium-gauge and wide-gauge (Marty, 2008), with a (WAP/P’ML)-ratio of 1.7. Impressions are oval in shape and have an average length of 36.9 cm and an average (length/width)-ratio of 1.6. The average pes pace angulation is 108°, which is characteristic of sauropods.

XQ-SI2 is 53.5 cm long and is the largest track at the Xiaqiao site. XQ-SI3 may be a manus impression with a (length/width)-ratio of 0.9 and is U-shaped with rounded marks inferred to represent traces of digits I and V. Only half of XQ-SI1 has been preserved. XQ-SI4 is only 30.4 cm long, possibly representing a smaller sauropod trackmaker. XQ-SI5 is 34.5 cm long with a (length/width)-ratio of 1.1, and may be a poorly preserved manus impression.

4.2 Comparisons and discussion

The morphology and trackway configuration of all the Xiaqiao tracks are typical of sauropods (Matsukawa et al., 2009). China’s sauropod trackways are mostly wide- or medium-gauge and are therefore referred to the ichnogenus Brontopodus (Lockley et al., 2002). The Xiaqiao site sauropod trackways share numerous features with Brontopodus type tracks from the Lower Cretaceous of the USA (Farlow, Pittman and Hawthorne, 1989; Lockley, Farlow and Meyer, 1994). These features include: (1) probably U-shaped manus prints; (2) large and outwardly-rotated, oval pes tracks, with length greater than width; (3) wide-gauge. Overall, there is no discernable difference between the Xiaqiao Brontopodus tracks and other Brontopodus tracks widely distributed at China’s Early and “Mid” Cretaceous tracksites, such as the Yanguoxia sites of the Hekou Group in northwestern China (Zhang et al., 2006; Xing et al., 2015a), more than 10 sites developed in Jiaguan, Feitianshan and Xiaoba Formations in southwestern China (Xing and Lockley, 2016), and more than a dozen sites of the Dasheng Group in eastern China (Xing et al., 2013a, 2015b). The wide-gauge of the Brontopodus-type trackways suggests titanosaurian sauropods as the most possible trackmakers (Wilson and Carrano, 1999; Lockley et al., 2002). Skeletal fossils of the titanosaurian Jiangshanosaurus lixianensis are known from the Jinhua Formation in eastern Zhejiang Province (Tang et al., 2001).

5.1 Dongyang sites of Jinhua Formation

In October 2008, researchers from the Institute of Geology, Chinese Academy of Geological Sciences, the Zhejiang Museum of Natural History, and the Dongyang Museum found numerous tracksites in Dongyang City. Some of these sites are very small, with only a few dinosaur tracks being preserved, as is the case with the Luoyanshan site, in Shanhuli Village. Larger sites are present on Fengchekou Mountain, in Wushan Village (No. 23 in Fig. 1 and Table 1), which produced a rich track assemblage, including theropod, pterosaur and bird tracks (Lü et al., 2010). The tracks were found in brownish-red siltstones of the Upper Cretaceous Jinhua Formation, Qujiang Group (the Jinhua Formation was once regarded as Fangyan Formation; see, Lü et al., 2008, 2010; Table 2).

Lü et al. (2010) briefly described the pterosaur, bird, small theropod, ornithopod, and sauropod tracks from Fengchekou site. The tracks were located at a construction site, and there were no large areas of exposed rocks. Except for the bird tracks, most tracks were observed in isolation and were not part of recognizable trackways.

Birds and sauropod tracks were rare and poorly preserved. Lü et al. (2010) provided two photos of theropod tracks, which were relatively well preserved. Although the lack of description makes further discussion difficult, these tracks are morphologically similar to Asianopodus (Matsukawa et al., 2005; Xing et al., 2014a) and Jialingpus (Zhen, Li and Zhen, 1983; Xing et al., 2014b; Fig. 4), and specifically share the traits of wide divarication angles and well-developed heels. In size, they are also similar to Jialingpus.

Azuma et al. (2013) described a bird footprint assemblage from Fengchekou site (GPS: 29°18′27.5″N, 120°9′54.7″E), and attributed them to Dongyangornipes sinensis and Koreanaornis cf. hamanensis. Buckley, McCrea and Lockley (2016), who looked into the main bird fauna based on multivariate statistical analyses, considered Dongyangornipes sinensis a subjective junior synonym of Uhangrichnus chuni. The latter is from lacustrine siliciclastic deposits of the Upper Cretaceous, purportedly Campanian, Uhangri Formation in South Korea (Yang et al., 1995).

Chen et al. (2013b) assigned the pterosaurs from the Fengchekou site to Pteraichnus dongyangensis. Pteraichnus is a classic pterosaur track and is relatively widely distributed in China (Lockley, Harris and Mitchell, 2008; Fig. 5). The first Pteraichnus specimen in China was found in the Yangouxia site, Gansu Province (Peng et al., 2004; Zhang et al., 2006; Xing et al., 2013b; Li et al., 2015), but has never been described in detail. Pteraichnus isp. Was also known from the Jimo site in the Laiyang Group of Shandong Province (Xing et al., 2012), from the Wuerhe site in the Tugulu Group of Xinjiang Uygur Autonomous Region (Xing et al., 2013c; He et al., 2013), from the Lotus site in the Jiaguan Formation of Chongqing Municipality (Xing et al., 2013b), from the Zhaojue site in the Feitianshan Formation of Sichuan Province (Xing et al., 2015c), from the Shimiaogou site in the Jiaguan Formation of Sichuan Province (Xing et al., 2016), and, the Yangmeikeng site in the Zhutian Formation of Guangdong Province (Xing et al., 2017). Except for the tracks from the Fengchekou and Yangmeikeng sites, all Pteraichnus specimens were the Early Cretaceous in age.

Sánchez-Hernández, Przewislik and Benton (2009) recognized that the ichnogenus Pteraichnus was over-split into five ichnospecies. They recognized only two of the six ichnospecies described from Spain, P. longipodus (Fuentes, Meijide and Meijide, 2004) and P. parvus (Meijide and Fuentes, 2001) as valid, and distinct from P. saltwashensis (Stokes, 1957) and P. stokesi (Lockley et al., 1995) which were named previously from North America. P. nipponensis (Lee et al., 2010) is also a distinct ichnospecies. Chen et al. (2013b) diagnosed P. dongyangensis as “manus prints with 29° of divarication of digits II and III and that of the digits I and II is 52°. Ratio of width to length of the pes print is 0.17”. The divarication of digits I–II and II–III in manus prints of Pteraichnus vary largely depending on the quality of the sediment, as in the manus dominated Pteraichnus isp. From the Yangouxia site, Gansu Province (Li et al., 2015). Chen et al. (2013b) did not provide outlines of the manus, though Figs. 3 and 4 in Chen et al. (2013b) suggest variable values of divarication between manus digits. Thus, manus digit divarication cannot be used as a valid diagnostic character. There are two pes print specimens. They have four clear digits, just as in all Pteraichnus type tracks. Therefore, we here consider P. dongyangensis a nomen dubium.

5.2 Guanyintang site of Jinhua Formation

In 2010, researchers from the Institute of Geology, Chinese Academy of Geological Sciences, the Zhejiang Museum of Natural History, the Dongyang Museum and the Fukui Prefectural Dinosaur Museum discovered the Guanyintang site (GPS: 29°18′39.89″N, 120°9′42.11″E) on Houyantou Mountain, near Guanyintang Village of Jiangdong Subdistrict, Yiwu City (No. 24 in Fig. 1 and Table 1). In 2014, the Zhejiang Museum of Natural History found two pterosaur tracks, nearly 30 theropod dinosaur tracks, two ornithopod tracks, a group of possible ankylosaur tracks, a suspected bird track and more fragmented sauropod tracks, when excavating the Guanyintang site. These tracks are now covered by a shelter (GPS: 29°18′39.49″N, 120°9′37.85″E).

The Guanyintang site is located in the center of the Jinqu Basin, in the lower Upper Cretaceous Jinhua Formation. The sedimentation environment has been interpreted as coastal shallow−lake facies and fluvial delta, at the edge of a lake (Du et al., 2015). A total of 11 track-bearing layers are present in a 3 m-thick rock sequence.

Du et al. (2015) described the geological background of the Guanyintang site, and briefly described the specimens, but did not provide outline drawings, detailed comparisons, or ichnotaxonomic discussions. Judging from the illustrations of Du et al. (2015), the pterosaur tracks are morphologically consistent with the Pteraichnus type (Fig. 5). They are 17−19 cm long and similar in general morphology to those from the Dongyang sites. There is no detailed information about the other track types.

The reassignment of Dongyangornipes sinensis to Uhangrichnis chuni, and the evidence that Pteraichnus dongyangensis is indistinguishable from other Pteraichnus ichnospecies, indicates that the ichnospecies of Zhejiang Province are not distinct from others found from other localities in China and South Korea. It is obvious that previous work has created a misleading “provincial” ichnotaxonomy. The avian theropod and pterosaur tracks are in fact indistinguishable from previously named ichnotaxa in the region. Moreover, the erection of invalid ichnotaxa, not only obscures faunal similarities, but also can inflate diversity estimates, and confuse the regional palaeobiological−palaeoecological environment.