Introduction
Fusichonetes Liao in Zhao et al., 1981 and Tethyochonetes Chen et al., Reference Chen, Shi, Shen and Archbold2000 both belong to the Rugosochonetidae, the most species-rich family among chonetidine brachiopods (Racheboeuf, Reference Racheboeuf2000). As a genus, Fusichonetes was first mentioned in Liao (Reference Liao1979), without description. Later, Liao in Zhao et al. (1981) formally proposed and defined this genus, with Plicochonetes nayongensis Liao, Reference Liao1980a from the Changhsingian strata at the Zhongling section, Nayong, Guizhou, South China as the type species (Fig. 1). Tethyochonetes was proposed by Chen et al. (Reference Chen, Shi, Shen and Archbold2000) with Waagenites soochowensis quadrata Zhan, Reference Zhan1979 from the Late Permian of Lianxian, Guangdong, South China as the type species (Fig. 1).
Figure 1 Type species of Tethyochonetes Chen et al., Reference Chen, Shi, Shen and Archbold2000 and Fusichonetes Liao, Reference Liao1981 . Tethyochonetes quadrata (Zhan, Reference Zhan1979), copied from Zhan (Reference Zhan1979, pl. 4, figs. 16–19), original designation as Waagenites soochowensis quadrata, and Fusichonetes nayongensis (Liao, Reference Liao1980a) copied from Liao (Reference Liao1980a, pl. 4, figs. 7–9) and from Xinmin section: (1) dorsal valve internal mold, CUG26123, bs: brachial scar, ms: median septum; (2) ventral valve internal mold, CUG25824; (3) dorsal valve external mold, CUG25224; (4) ventral valve external mold, CUG25524. All scale bars=2 mm.
Since Fusichonetes was proposed, only two species have been recorded, F. nayongensis (type species) and F. pygmaea (Liao, Reference Liao1980a), both designated to this genus originally by Liao (Reference Liao1981), and both restricted to Lopingian (late Permian) and Griesbachian (earliest Triassic) strata in South China (Fig. 2.3). In contrast, at least 14 species of Tethyochonetes have been recognized (online supporting data: Appendix 1), most of which had been assigned to either Chonetes Fischer von Waldheim, Reference Fischer von Waldheim1830 or Waagenites Paeckelmann, Reference Paeckelmann1930 in the old literature. The collective stratigraphical range of these 14 Tethyochonetes species is from the Capitanian (middle Permian) to Griesbachian (Early Triassic), while their paleogeographic ranges spread across the Paleo-Tethys and Neo-Tethys (Fig. 2.3). Notably, both genera apparently survived the Permian-Triassic boundary mass extinction (PTBME), accounting for 11.8% of the Permian brachiopod genera that persisted into the Griesbachian before their final extinction at the Griesbachian-Dienerian boundary (Liao, Reference Liao1980b; Yang et al., Reference Yang, Yin, Wu, Yang, Ding and Xu1987; Chen et al., Reference Chen, Kaiho and George2005; Clapham et al., Reference Clapham, Fraiser, Marenco and Shen2013). As such, these two taxa, especially Tethyochonetes, have attracted much attention in recent years, especially in connection to their extinction, paleobiogeographical distribution, and body-size change patterns (Chen et al., Reference Chen, Tong, Zhang, Yang, Liao, Song and Chen2009; He et al., Reference He, Twitchett, Zhang, Shi, Feng, Yu, Wu and Peng2010; Wu et al., Reference Wu, He, Zhang, Yang, Xiao, Chen and Weldon2016; Zhang et al., Reference Zhang, Shi, He, Wu, Lei, Zhang, Du, Yang, Yue and Xiao2016).
Figure 2 (1) Paleogeographical map of South China during the Permian−Triassic transition (from Feng et al., Reference Feng, Yang, Jin, Li and Bao1997; Yin et al., Reference Yin, Jiang, Xia, Feng, Zhang and Shen2014), also showing the location of the Xinmin section (marked by a star). (2) Stratigraphical column of the Xinmin section, showing the lithology of the Talung Formation and the lower part of Daye Formation, as well as the stratigraphic distribution of Fusichonetes, as revised here (including species that would have been assigned to Tethyochonets prior to this study). (3) Global paleogeographical distribution of Fusichonetes and Tethyochonetes according to previous literature (basal map after Blakey, Reference Blakey2008).
However, despite their close connection to the PTBME and potential significance for a better understanding of this major bio-crisis, the identity of these two genera and their mutual relationship have been problematic and never properly been investigated. Significantly, the problem arose because when Liao in Zhao et al. (1981) proposed Fusichonetes, he appears to have based the definition of his new genus primarily on a combination of external characteristics, including a very transverse outline, acute cardinal extremities, and coarse and angular costellae with deep interspaces. Although no internal features were illustrated by Liao in Zhao et al. (1981), he did indicate that the dorsal interior of Fusichonetes possessed a cardinal process that is bilobate internally and quadrilobate externally. Interestingly, however, of the two Fusichonetes species recognized by Liao, only the type species, F. nayongensis, most consistently demonstrates the external morphological features that Liao used to define Fusichonetes. The second species, Fusichonetes pygmaea (see Liao, Reference Liao1979, fig. 14; Reference Liao1980a, pl. 4, figs. 4–6; Reference Liao1980b, pl. 1, figs. 5, 6; Reference Liao1981, pl. 8, figs. 7, 8), is in fact characterized by a subquadrate outline, cardinal extremity with an angle ~75–90°, and coarse costellae with a few bifurcations.
When Chen et al. (Reference Chen, Shi, Shen and Archbold2000) proposed Tethyochonetes, based on Waagenites soochowensis quadrata Zhan, they emphasized both external features as well as internal characteristics. Among the external features, Chen et al. (Reference Chen, Shi, Shen and Archbold2000) considered the complexity of costellae (i.e, whether costellae are simple and undivided, or increase by bifurcation), the shell width/length ratio and the cross-section shape of the interspace between costellae as the main characteristics to distinguish the two genera, supplemented by differences in internal structures. In comparing Tethyochonetes with Fusichonetes, Chen et al. (Reference Chen, Shi, Shen and Archbold2000) noted that the former had a ventral median septum, an internally bilobate and externally trilobate cardinal process, a dorsal median septum, and lateral septa and brachial scars. But, it should be pointed out that most of these features that supposedly are diagnostic of Tethyochonetes were unknown in Fusichonetes when it was first formally proposed by Liao in Zhao et al. (1981) as he only had limited amount of internal material to study (this is evident from the fact that he figured no internal shell characteristics, nor mentioned any internal features, except for a brief comment on the cardinal process). Furthermore, even though Chen et al. (Reference Chen, Shi, Shen and Archbold2000) mentioned that Tethyochonetes possessed a cardinal process that is bilobate internally and trilobate externally, they did not illustrate the trilobate aspect. But, when examining the illustrations of a well-preserved dorsal interior of Tethyochonetes cheni figured by Zhang et al. (Reference Zhang, He, Shi and Zhang2013, fig. 9J), we found that it clearly has an internally bilobate and externally quadrilobate cardinal process, which is the same as what Liao in Zhao et al. (1981) had described for the dorsal interior of Fusichonetes.
As for the supposed significance of a ventral median septum, brachial scars, and a dorsal median septum in distinguishing Tethyochonetes from Fusichonetes, they are equally dubious. This is because from the specimens we have recently collected from the Xinmin section in South China, the ventral interior of Fusichonetes possesses a median septum, and its dorsal interior has well-developed brachial scars and a median septum (Fig. 1). Although until now no lateral septa are known from Fusichonetes, this does not necessarily mean it was originally absent in this genus, because even in some species of Tethyochonetes (e.g., Zhang et al., Reference Zhang, He, Shi and Zhang2013, fig. 5O; He et al., Reference He, Shi, Zhang, Yang, Zhang, Wu, Niu and Zhang2014, fig. 4K, N) where lateral septa are supposed to exist, they are in fact absent, and presumably not preserved. With these observations, it suggests that the two genera not only resemble each other strongly in external morphology, they also share very similar internal structures.
Recently, we collected 1150 well-preserved rugosochonetid specimens from a stratigraphic section in South China that straddles the Permian-Triassic boundary (Fig. 2.2, also see below for more details about the section). Our initial inspection of these specimens suggested that this collection, from a single section, seemed to represent a continuum of morphological variations from Fusichonetes (as one end member) to Tethyochonetes (at the other end), and that it was not possible to distinguish them with great confidence. The inability to separate them was due to the fact that there are specimens in the collection that clearly represent transitional forms between typical Fusichonetes characterized by a transversely quadrate outline to typical Tethyochonetes characterized by a more quadrate outline (Fig. 3). On this basis, we hypothesized that the entire collection, in all probability, should only be assigned to one genus, not two. If this hypothesis is correct, we further propose that the two genera be synonymized with Fusichonetes taking priority because it was proposed before Tethyochonetes.
Figure 3 A range of Fusichonetes specimens from the Xinmin section showing a continuum of external morphological variation from a typical quadrate-shaped individual (1) to a more transform-shaped individual (12): (1, 2) Fusichonetes pygmaea, ventral valve exterior, CUG20309; ventral valve internal mold, CUG23702; (3, 4), Fusichonetes cheni, ventral valve exteriors, CUG22202, CUG23801; (5, 6) Fusichonetes rectangularis, dorsal valve external mold, CUG20011; dorsal valve interior, CUG21009; (7, 8) Fusichonetes chaoi, dorsal valve external mold, CUG19104; ventral valve internal mold, CUG21228; (9, 10) Fusichonetes soochowensis, ventral valve internal molds, CUG24110, CUG23910; (11, 12) Fusichonetes nayongensis, ventral valve internal mold, CUG25424; ventral valve exterior, CUG24924. All scale bars=2 mm.
Thus, the purpose of this paper is to test this hypothesis by applying several quantitative and analytical procedures (bivariate plots, Kolmogorov-Smirnov test, categorical principle component analysis, and cladistic analysis) to a dataset of quantified morphological variables comprising both external and internal characters from all well-known Fusichonetes and Tethyochonetes species plus newly collected specimens from the Xinmin section.
Materials and methods
In this study, 330 well-preserved specimens belonging to 13 Tethyochonetes species and two Fusichonetes species were selected from the literature, as well as from our own fieldwork at the Xinmin section. The 130 specimens measured from literature were selected from those only with adequate description and photographic illustrations. The remaining 200 specimens were systematically collected from the Talung Formation of the Xinmin section, which was located in the northern margin of a deep-water basin in South China during the Permian–Triassic transition (Fig. 2.1). All specimens figured in this paper from the Xinmin section are stored in the Faculty of Earth Sciences, China University of Geosciences (Wuhan), China, with prefixes CUG.
At this section, the Permian-Triassic boundary is defined by the first occurrence of the typical basal Triassic conodont species Hindeodus parvus (Kozur and Pjatakova, Reference Kozur and Pjatakova1976) in the middle of Bed 36 (Zhang et al., Reference Zhang, Jiang and Zhong2014), within the uppermost part of the Talung Formation. The Talung Formation is dominated by siliceous mudstone, siliceous limestone, and shale, intercalated with siltstone and calcareous mudstone, suggesting an outer-shelf to deep-basin paleoenvironment (Feng et al., Reference Feng, Yang, Jin, Li and Bao1997). Apart from brachiopods including the rugosochonetidines studied here, the Talung Formation also contains abundant ammonoids, bivalves, and radiolarians, as well as plant fragments.
For each studied specimen, six numerical variables and nine categorical variables were recorded (Fig. 4; Table 1). Some variables have been adopted from previous studies using similar morphometric and phylogenetic approaches for brachiopod taxonomy (e.g., Shi and Waterhouse, Reference Shi and Waterhouse1991; Bauer and Stigall, Reference Bauer and Stigall2016). Others, mainly those concerning the complexity of costellae and ears, represent morphological characters widely regarded as important for distinguishing chonetid genera and species (Brunton et al., Reference Brunton, Lazarev and Grant2000; Chen et al., Reference Chen, Shi, Shen and Archbold2000; Shen and Archbold, Reference Shen and Archbold2002). The quantitative values for all variables were collected from either complete ventral valves or complete dorsal valves. For body-size measurements (length, width) taken from the literature, they were measured from the published actual fossil illustrations with a digital caliper from the Foxit Reader to the nearest 0.1 mm. According to Krause et al. (Reference Krause, Stempien, Kowaleski and Miller2007), size estimates of brachiopod fossil shells from photographed images correlate well with their real sizes and can therefore be used for studies of body size variations. For specimens from the Xinmin section, the width and length were measured from the actual specimens with a vernier caliper to the nearest 0.1 mm.
Figure 4 Biometric measurements of external shell morphology used for this study. The full explanation of abbreviated morphological variables is given in Table 1.
Table 1 Numerical and categorical variables measured in Fusichonetes and Tethyochonetes (see Figure 4 for illustration of the chosen variables in a hypothetical rugosochonetid shell).
Four different approaches were applied to the dataset, in order to test and visually demonstrate whether species of Fusichonetes and Tethyochonetes could be distinguished with rigor. First, we used simple bivariate plots and linear regression models to analyze and visually depict the relationships between certain pairs of key morphological variables (e.g., shell length versus width, density of costellae versus shell outline, shell shape [outline] versus shell length) (see online supporting data: Appendix 2).
Second, the Kolmogorov-Smirnov test was used to analyze differences in shell size and width/length ratio between the two previously recognized genera; this was carried out using the software PAST (Hammer et al., Reference Hammer, Harper and Ryan2001). In this analysis, brachiopod shell size was approximated with the geometric mean of length and width, following Jablonski (Reference Jablonski1996).
Third, owing to the fact that our dataset is comprised of a mixture of both numerical and categorical variables (see online supporting data: Appendix 3), we performed a categorical principle component analysis (CATPCA) to conduct a multivariate analysis of the dataset. This procedure simultaneously analyzes numerical and categorical variables while reducing the dimensionality of the original data. CATPCA has been used for similar taxonomic studies (Domínguez-Rodrigo et al., Reference Domínguez-Rodrigo, De Juana, Galán and Rodríguez2009; Claerhout et al., Reference Claerhout, Dewaele, De Riek, Reheul and De Cauwer2016). The CATPCA was performed with the software SPSS Statistics 22 (SPSS Inc. Chicago, IL, USA). To maintain the category order in the quantifications on theoretical grounds, categorical variables were discretized and optimally scaled by ordinal transformation. Since we are interested only in the relations between variables and objects, or between objects, rather than the relationships between the variables, numerical variables were scaled by numeric transformation. The discretization method selected was symmetrical normalization, given that our aim was to examine the differences or similarities between the objects.
Finally, we employed cladistics to investigate the phylogenetic relationship of the 15 species that have hitherto been assigned to the two genera. For this analytical procedure, Neochonetes (Huangichonetes) substrophomenoides (Waagen, Reference Waagen1884) and N. (Sommeriella) strophomenoides (Huang, Reference Huang1932) were selected as outgroups. The cladistic analysis was conducted in TNT (Goloboff et al., Reference Goloboff, Farris and Nixon2008), treating continuous characters (numerical variables) ‘as such’ (Goloboff et al., Reference Goloboff, Mattoni and Quinteros2006). For each species, the total range of measured morphometric values was adopted in order to show data information on specimen level. All measured values were transformed by log(x+1) for standardization (Kitching et al., Reference Kitching, Forey, Humphries and Williams1998) (online supporting data: Appendix 4).
Results
The bivariate plots of width versus length and the number of costellae versus width/length ratio of the two genera are shown in Figure 5 and Figure 6, respectively. Figure 5 suggests that the two genera have very similar shell size variation trends through ontogeny. Figure 6, on the other hand, shows two interesting features. First, taking the dataset for both genera as a whole, the number of costellae appears quite stable and changes little with the variation of the shell outline, expressed by the width/length ratio. Second, based on this plot, the data points representing Fusichonetes cannot be well separated from those of Tethyochonetes, although F. nanyongensis, the type species of this genus, appears to stand out quite clearly from the rest of the plot. Also of note from Figure 6 is that although F. nayongensis is more transverse than Tethyochonetes, they have a similar number of costellae.
Figure 5 Graph of shell length to width of previously recognized Tethyochonetes and Fusichonetes species, plus some Fusichonetes nayongensis specimens from the Xinmin section.
Figure 6 Graph showing the relationship between the total number of costallae and the shape (here measured by the width/length ratio) of Tethyochonetes and Fusichonetes shells (data points are comprised of measurements from literature and some Fusichonetes nayongensis specimens from the Xinmin section).
To test whether the visual differences observed in Figures 5 and 6 are of any statistical significance, we applied the Kolmogorov-Smirnov test. The result suggests, while there is no significant difference in shell size among the 15 species that have previously been assigned to the two genera (Fig. 5; D=0.2849, p=0.0796), the difference in the width/length ratio between the two genera is statistically significant (Fig. 6; D=0.7765, p<0.001), suggesting a possibility of distinguishing the two genera based on their shell outline or shape.
To further investigate whether shell outline varies consistently with shell size, we used a subset of our dataset with specimens collected from discrete beds of the Xinmin section. This sub-dataset contains the measurements of shell length and width from 174 specimens representing what would be considered as T. pygmaea using traditional qualitative taxonomy from bed 9, and 26 specimens representing F. nayongensis from bed 24. As shown in Figure 7, there is no significant linear correlation between the width/length ratio and length of the shell for either of them, suggesting that the shell shape or outline, approximated by the width/length ratio, is a rather stable feature because it changes little with shell size. This would translate to mean that the shell shape does not fluctuate much with ontogeny and, therefore, could be used as a reliable parameter for the comparison of other morphological features.
Figure 7 Linear regression of shell width/length ratio to length of Tethyochonetes pygmaea from bed 9, and Fusichonetes nayongensis from bed 24 of the Talung Formation in the Xinmin section.
The result from CATPCA is summarized in Figure 8, with the 15 species previously assigned to the two genera arranged in a biplot defined by the first two principle components (PC). PC 1 explains 25.3% of the total variance while PC 2 accounts for 16.7%. In the biplot, the 15 variables (see Table 1 for abbreviations) are projected as vectors whose lengths vary according to their variance accounted for (i.e., the longer the variable line in the plot, the larger the variance it accounts for). It is evident from the figure that the positive side of PC 1 holds F. nayongensis, T. soochowensis, and part of T. quadrata, T. chaoi, T. guizhouensis, T. wongiana, and T. cheni. These specimens are represented by having a significantly larger lateral margin angle (LMA) coupled with a large umbonal region width (URW). Next to this, specimens of F. nayongensis and T. soochowensis, located on the negative side of PC 2 in the biplot (Fig. 8), are drawn together because they both have a high width/length ratio (WLR). Specimens of F. pygmaea are located on the negative side of both PC 1 and PC 2, and possess ears that are not well demarcated from the visceral region. Fusichonetes pygmaea and most T. pygmaea are both restricted to the left side of the plot and, in particular, cluster closely together in the lower left quadrate of the plot. This is because both species have ears that are not well demarcated from visceral area (TEV). Overall, CATPCA, as shown in Figure 8, suggests that data points for F. nayongensis and F. pygmaea are well merged with those of Tethyochonetes and, hence, the two genera cannot be well separated based on the multivariate analysis of both external and internal morphological characteristics.
Figure 8 Result of categorical principle component analysis (CATPCA) plotted on the first two principle axes (PC 1 and PC 2). The black lines represent vectors of the analyzed 15 variables comprised of six numerical variables (CA, DP, LMA, LWR, SL, SW) and nine categorical variables. See Table 1 for explanation of abbreviations. Dashed lines divide the biplot into four quadrants according to PC 1 and PC 2.
The cladistic analysis resulted in three most parsimonious trees with 29.2 TL (CI: 0.543, RI: 0.602). Their strict consensus tree (Fig. 9) demonstrates a monophyletic group including the F. nayongensis-T. chaoi-T. wongiana-T. cheni clade, and three species (T. soochowensis, T. quadrata, and T. guizhouensis) as sister taxa. This tight grouping, consisting of a clade and three sister taxa, is also supported by the biplot mentioned above in that all these species are located on the positive side of PC 1, grouped together by having larger values of LMA (3) and URW (14) (Fig. 8). Notably, in Figure 9, F. pygmaea is shown to group together with T. pygmaea and T. liaoi as sister taxa, rather than with its supposed co-generic species F. nayongensis. Furthermore, in accord with the cladistic analysis, F. pygmaea and F. nayongensis appear to represent a polyphyletic group (Fig. 9).
Figure 9 A strict consensus tree of three most parsimonious trees (TL: 29.2; CI: 0.543; RI: 0.602). Code number 3 represents variable LMA (lateral margin angle), 14 represents variable URW (umbonal region width).
Discussion
In view of the preceding descriptions and interpretations derived from the four different analytical procedures applied to the same dataset, it is clear that: (1) F. pygmaea and F. nayongensis, originally recognized as two distinct species within the same genus by Liao (Reference Liao1979, Reference Liao1980b, Reference Liao1981), in fact have considerably disparate morphological features and thus cannot be treated as forming a closely related coherent group; and (2) when the currently known species of both Fusichonetes and Tethyochonetes are analyzed together based on a common set of both external and internal morphological features, regardless of the analytical method used, they cannot be separated as two distinct genera with consistency and rigor. Instead, as particularly demonstrated by the cladistic analysis (Fig. 9), they appear to constitute a highly coherent monophyletic clade evolved from Neochonetes. Consequently, we conclude that Tethyochonetes should be considered as a junior synonym of Fusichonetes and, as such, be suppressed as an invalid genus name.
Following from this, it is necessary to provide an updated and refined diagnosis for Fusichonetes. In part, this is essential because the original genus diagnosis given by Liao contained little information about the internal features. On the other hand, merging Tethyochonetes with Fusichonetes by suppressing the former means that all previously recognized Tethyochonetes species need to be transferred to Fusichonetes, thus substantially increasing the species composition of this genus. Another strong reason for updating and expanding the genus diagnosis of Fusichonetes is due to the availability of a large number of exceptionally well-preserved materials from the Xinmin section, including both dorsal and ventral interiors that show important internal features unknown to previous studies.
Emended genus diagnosis for Fusichonetes
Small to medium in size, 1.47–11.77 mm long and 2.22–20.14 mm wide (online supporting data: Appendix 5); transversely rectangular to reverse trapezoidal in outline, width/length ratio 1.13–3.26, and lateral margin angle 14°–101° (online supporting data: Appendix 5); concavoconvex to almost planoconvex in lateral profile; ears normally smooth, flattened or slightly swollen; external surface ornamented by simple costellae, occasionally with micro-ornament of tubes; sulcus and fold variable. Ventral interior with median septum; dorsal interior with median septum, lateral septa (possibly) and brachial scars, cardinal process quadrilobate; internal surface of both valves with radially distributed papillae.
Distinguishing Fusichonetes from similar chonetid genera
As defined, Fusichonetes bears some similarities with several other rugosochonetid genera (Fig. 10). The genus is similar to Quinquenella Waterhouse, Reference Waterhouse1975, because both have a subquadrate outline, a moderately concavoconvex profile, and similar ventral interior, but differ in the latter having concentric stria, a long accessory septa in the dorsal interior, and lacking costellae. Rugaria Cooper and Grant, Reference Cooper and Grant1969 can be easily distinguished from Fusichonetes by having a strongly concavoconvex profile and much coarser papillae and costae, although Rugaria also has a subquadrate outline and short dorsal median septum. Prorugaria Waterhouse, Reference Waterhouse1982, a Mississippian genus, is similar to Fusichonetes in having a subquadrate outline, but can be distinguished from the latter in having coarser papillae and costae and long accessory septa in the dorsal interior. Neochonetes Muir-Wood, Reference Muir-Wood1962 differs from Fusichonetes in having a larger size, denser and bifurcate costellae, and papillae increasing in number and decreasing in size towards margin (Wu et al., Reference Wu, He, Zhang, Yang, Xiao, Chen and Weldon2016). Pygmochonetes Jin and Hu, Reference Jin and Hu1978 could be easily distinguished from Fusichonetes by having a semicircular outline, being strongly concavoconvex, and lacking a sulcus and long accessory septa in the dorsal interior. Linshuichonetes Campi and Shi, Reference Campi and Shi2002 differs from present genus in lacking any median, accessory or lateral septa. Waagenites Paeckelmann, Reference Paeckelmann1930 differs from Fusichonetes by its much coarser costae, a more highly convex ventral valve and very incurved umbo. Waterhousiella Archbold, Reference Archbold1983 is similar to the present genus in its simple costellae, but differs in having a more convex ventral valve and vascular trunks developed in the ventral interior.
Figure 10 Comparison of Fusichonetes and Tethyochonetes with selected other morphologically similar chonetid genera. Gray blocks show the distinctive characteristics dividing genera from each other. Question marks indicate information that was not provided by the original authors when the genera were proposed. Holotype specimens of all genera are from the original references (Cooper and Cooper, 1969; Waterhouse, Reference Waterhouse1975; Jin and Hu, Reference Jin and Hu1978; Zhan, Reference Zhan1979; Liao, Reference Liao1980a; Waterhouse, Reference Waterhouse1982; Campi and Shi, Reference Campi and Shi2002), except for Neochonetes from Racheboeuf (Reference Racheboeuf2000) and holotype specimens of Waagenites and Waterhouseiella separately from Waagen (Reference Waagen1884) and Waterhouse and Piyasin (Reference Waterhouse and Piyasin1970). Scale bars=5 mm.
Implications for survival of brachiopods in the aftermath of the PTBME
A final point that is worth mentioning is the implication of this study in relation to the survival of brachiopods in the aftermath of the PTBME. According to some previous studies on the Permian–Triassic brachiopods (Yang et al., Reference Yang, Yin, Wu, Yang, Ding and Xu1987; Shen and Shi, Reference Shen and Shi1996; Chen et al., Reference Chen, Kaiho and George2005; Shen et al., Reference Shen, Zhang, Li, Mu and Xie2006; Clapham et al., Reference Clapham, Fraiser, Marenco and Shen2013; Ke et al., Reference Ke, Shen, Shi, Fan, Zhang, Qiao and Zeng2016), there are 17 Changhsingian brachiopod genera that survived into the aftermath of the PTBME, but did not play a role in the post-extinction recovery. After merging the two genera, the revised number of surviving Permian brachiopod genera is now 16, including Fusichonetes, as revised and its species expanded here. Following this revision, the expanded Fusichonetes includes a total of 15 species at present time, comprised of 13 species transferred from Tethyochonetes and two original Fusichonetes species. Among the 15 species, two originated from the Capitanian (late Guadalupian), eight from the Wuchiapingian (early Lopingian) and five originated from the Changhsingian (late Lopingian), seven of which survived the PTBME until their final disappearance at the early Griesbachian (Chen et al., Reference Chen, Kaiho and George2005) (Fig. 11). Paleogeographically, these species were all restricted to the shallow- to moderately deep-marine environments of the Paleo-Tethys and Neo-Tethys region (Wu et al., Reference Wu, He, Zhang, Yang, Xiao, Chen and Weldon2016).
Figure 11 Stratigraphic ranges of all Fusichonetes species as recognized in this paper, including species that had previously been assigned to Tethyochonetes.
Conclusions
Four different approaches (bivariate plots, Kolmogorov-Smirnov test, categorical principle component analysis, and cladistic analysis) were employed to analyze a dataset composed of 15 species belonging to Fusichonetes and Tethyochonetes by 15 variables. The variables chosen are a mixture of both numerical and categorical characters and include important external and internal morphological features. Except for the Kolmogorov-Smirnov test demonstrating the possibility of separating the two genera in terms of shell outline approximated by the shell width/length ratio, all other analytical procedures suggest that species of the two genera cannot be separated as two distinct taxa with consistency and statistical rigor. Consequently, we conclude that: (1) Tethyochonetes and Fusichonetes be merged, and (2) Fusichonetes be maintained as a valid genus in recognition of its chronological priority over Tethyochonetes, while the latter be suppressed as an invalid genus. With this revision and the merger of the two genera, the diagnosis of Fusichonetes is updated and refined, in part based on observations of new well-preserved material from South China. Additionally, the number of brachiopod genera that survived the Permian−Triassic boundary mass extinction is revised from 17 to 16.
Acknowledgments
The authors thank S. Lee for helpful instructions on usage of TNT and revisions on the paper, J.B. Waterhouse for discussion on the paper, Y. Zhang for insightful suggestions on data collecting, and T. Yang, M. Yue, Y. Xiao, B. Chen, L. Zhu, B. Su for their help in fieldwork. Furthermore, we thank the journal editor and reviewers Y. Sun and C. Rasmussen for their constructive comments, and reviewer S. Shen for his valuable comments and help taking photos for the holotype of Fusichonetes. Support for this research was provided by NSFC (Grant Nos. 41372030, 41772016, 41602017), the Ministry of Education of China (B08030 of 111 Project), and an Australian Research Council grant to GRS (ARC DP150100690), as well as by Deakin University.
Accessibility of supplemental data
Data available from the Dryad Digital Repository: https://doi.org/10.5061/dryad.vb051
