https://doi.org/10.18261/let.55.4.5 Copyright © 2022 Author(s). This is an open access article distributed under the terms of the Creative Commons CC-BY 4.0 License. Published by Scandinavian University Press on behalf of Lethaia Foundation. Interspecies and intraspecific variability in the trilobites Duyunaspis and Balangia from the Cambrian Series 2 (Stage 4) of Jianhe, South China ZHENGPENG CHEN, YUANLONG ZHAO, XINGLIAN YANG, JORGE ESTEVE, XIONG LIU, SHENGGUANG CHEN AND RONG FENG Abundant and diverse trilobite assemblages have been reported from the Cambrian of South China, especially from the Jiangnan Slope Belt (e.g. Zhang et al. 1980; Peng 2000, 2009; Yuan et al. 2002; Peng et al. 2012a, 2017). Given the great biological diver- sity, complex morphologies, relatively continuous change in evolution process, few transitional groups and relatively sequential fossil record in continuous geological successions, South China is arguably one of the best areas to explore the morphological diversity of trilobites. Cambrian oryctocephalid trilobites have played an important role in Cambrian chronostratigraphy and the morphological analysis of trilobites because of their diversity, richness, and broad distributional range (Zhao et al. 2019). Duyunaspis and Balangia are two familiar oryctocephalid taxa from the Cambrian Series 2, Stage 4 of South China which have earned much attention (e.g. McNamara et al. 2006; Lei 2016; Dai et al. 2017; Peng et al. 2017; Chen et al. 2018). Supplements and revisions have continually been made on the characteristics of Duyunaspis (Yin & Li 1978; Zhang et al. 1980; McNamara et al. 2006; Lei & Peng 2014; Dai et al. 2017; Chen et al. 2018), and stud- ies on the ontogeny of Duyunaspis duyunensis and D.  jianheensis have also been reported (McNamara et  al. 2006; Lei 2016; Dai et al. 2017; Chen et al. 2018). D. jianheensis was first described by Chen et al. (2018) based on specimens from the Tsinghsutung Formation. D. duyunensis and D. jianheensis have sim- ilar morphological structures and are easily confused (Chen et al. 2018). Therefore, a detailed discussion on interspecific variations of Duyunaspis is necessary. Balangia and Duyunaspis are morphologically simi- lar and are thought to be closely related (McNamara et al. 2006), thus the variations among them need to be quantified by more detailed studies. Geometric morphometrics not only provides great insight into biological and evolutionary processes, but also explores the morphology of organisms, their Though morphological variation among trilobites has been studied for some time, it is difficult to identify differences between interspecific and intraspecific variations. In this study, the intergeneric, interspecific, and intraspecific variations of the Cambrian oryctocephalid trilobites Duyunaspis and Balangia from Jianhe, South China are anal- ysed using geometric morphometrics. In addition to the number of thoracic segments and the relative size of the pygidium compared to the cephalon mentioned in previous studies, the results in this study suggest that the predominant variation of Duyunaspis and Balangia is the presence or absence of the glabellar furrow. The interspecific vari- ations of Duyunaspis (D. duyunensis, D. paiwuensis and D. jianheensis) mainly lie in the types of facial suture, occipital ring and glabella. Although the differences of D. duyunensis or D. jianheensis from adjacent sections have been observed, these vari- ations may be caused by ontogeny, and do not reach the magnitude of interspecific variations in Duyunaspis. This research provides a better quantitative understanding of the variability in oryctocephalid trilobites. □ Interspecies and intraspecific variation, oryctocephalid trilobite, Cambrian, South China. Zhengpeng Chen [chenzhengpeng2016@163.com], Yuanlong Zhao [zhaoyuanlong@126. com], Xinglian Yang ✉ [yangxinglian2002@163.com], Shengguang Chen [chenshenggu- ang2019@126.com], Rong Fong [fengrong0358@qq.com] and Xiong Liu [1098088784@ qq.com], College of Resources and Environmental Engineering, Guizhou University, Guiyang 550025, China; Xinglian Yang [yangxinglian2002@163.com], Key Laboratory of Karst Georesources and Environment, Ministry of Education, Guiyang 550025, China; Xinglian Yang [yangxinglian2002@163.com], Guizhou Research Center for Palaeontology, Guiyang 550025, China; ✉ Jorge Esteve [jorgeves@ucm.es], Departamento de Geodinámica, Estratigrafía y Paleontología, Facultad de Ciencias Geológicas. Geológicas, Universidad Complutense de Madrid, Madrid 28040, Spain; manuscript received on 05/10/2021; manuscript accepted on 08/07/2022; manuscript published on 23/11/2022 in Lethaia 55(4). https://doi.org/10.18261/let.55.4.5 https://creativecommons.org/licenses/by/4.0/ Chen et al. 2 ontogeny and phylogeny (e.g. Zelditch et al. 2004; Hammer & Harper 2006). Since cephalic features are important for oryctocephalidae classification (e.g. Whittington 1995; Sundberg 2014; Esteve et al. 2017; Peng et al. 2017, 2018), this morphological study (using landmark-based geometric morphometrics approach) will focus on the cephalon. We analyse three species belonging to Duyunaspis (i.e. D. duyunensis Zhang and Qian in Zhou et al. 1977, D. paiwuensis Lei & Peng 2014, and D. jianheensis Chen et al. 2018) and one species belonging to Balangia (i.e Balangia balan- gensis Qian, 1961) from the Balang and Tsinghsutung formations, Cambrian Series 2 Stage 4, South China. The intergeneric, interspecific and intraspecific varia- tions between Duyunaspis and Balangia are analysed, with the aim to provide a better quantitative under- standing of the variability in oryctocephalid trilobites. Geological setting Cambrian sediments are widespread in South China. One of the classic areas for Cambrian chronostrati- graphical studies is the continuous succession with abundant trilobites in the Jiangnan Slope Belt (Peng 2000, 2009; Peng et al. 2004, 2012b; Zhao et al. 2019). The Cambrian Stage 4 from eastern Guizhou and western Hunan (Fig. 1) comprises slope deposits, and contains the Balang Formation, the Tsinghsutung Formation, and the lower part of the Kaili Formation (eastern Guizhou Province) or the lower part of the Aoxi Formation (western Hunan Province). In east- ern Guizhou Province these strata reach a thickness of over 600 m. Greyish-yellow, greyish-green mudstone, shale and silty mudstone are dominant in the Balang Formation (eastern Guizhou Province), with more calcareous shale occurring in the Balang Formation in western Hunan. The Tsinghsutung Formation con- sists mainly of limestone and dolomite. The lower part of the Kaili Formation consists of calcareous marl and silty calcareous mudstone and the lower part of the Aoxi Formation, contemporaneous of the Kaili Formation, deposited in western Hunan, is composed of dolomite. Material and methods Material Duyunaspis jianheensis (Protoryctocephalus arcticus Zone; Figs 1B, 2A–C). A total of 198 specimens were selected for our analysis, including 154 specimens from the Songshan section and 44 specimens from the Malipo section from the Jianhe County, Guizhou Province. The specimens are housed at the Guizhou Research Centre for Palaeontology (GRCP). Duyunaspis duyunensis (Oryctocarella duyunensis Zone – Arthricocephalus chauveaui Zone; Figs 1B, 2G, H). A total of 27 specimens were selected for our Fig. 1. Study sample localities and stratigraphical horizons. 1, Jianhe County, Guizhou Province, China (Songshan and Malipo sections). 2, Huayuan County, Hunan Province, China (Bulin and Zila A sections). Bb, Balangia balangensis. Dd, Duyunaspis duyunensis. Dp, Duyunaspis paiwuensis. Dj, Duyunaspis jianheensis. Variability in the trilobites Duyuanaspis and Balangia 3 analysis including nine specimens from Bulin sec- tion, eight specimens from Zila A section and ten specimens from a fossil site near Zila A section in Huayuan County, Hunan Province previously studied by Lei & Peng (2014), Lei (2016) and Dai et al. (2017). This species has also been previously reported in the Balang Formation in eastern Guizhou Province (Yan et al. 2014). The specimens are housed at the Nanjing Institute of Geology and Palaeontology, Chinese Academy of Sciences (NIGP, CAS) and the Geology Department of Northwest University (NWU). Duyunaspis paiwuensis (Arthricocephalus chau- veaui Zone; Figs 1B, 2I). Three specimens were selected for our analysis. All the specimens were collected from Zila A section in Huayuan County, Hunan Province, previously studied by Lei & Peng (2014). The specimens are housed at the NIGP, CAS. Balangia balangensis (Arthricocephalus chauveaui Zone; Figs 1B, 2D– F). A total of 21 specimens were selected for our analysis. All the specimens were col- lected from the strata near the Jiaobang section in Jianhe County, Guizhou Province. The specimens are housed at the GRCP. Geometric morphometric methods Geometric morphometry is a quantitative method that combines morphology with multivariate sta- tistical analysis, as a tool for analysing morphology and morphological changes (e.g. Zelditch et al. 2004; Hammer & Harper 2006). Recently, it has been used on trilobites (e.g. Hopkins & Webster 2009; Webster 2011; Abe & Lieberman 2012; Gendry et al. 2013; Esteve et al. 2017; Álvaro et al. 2018; Bicknell et al. 2019; Oudot et al. 2019; Álvaro & Esteve 2020; Zhao et al. 2020). In this study, we used landmarks on the trilobite’s cephalon for geometric morphological anal- ysis. All the specimens were photographed with a Leica DVM6 microscope, and the photos were digi- tized by using tpsDig v.2.16 (Rohlf 2010). A total of 24 landmarks were selected for morphometric analy- ses in Duyunaspis (four along the central axis and ten pairs on each side of the axial lobe) and 16 in Balangia (four along the central axis and six pairs on each side of the axis) (Fig. 3). The landmarks related to the gla- bellar furrow of B. balangensis were removed from the analysis because the glabellar furrow of B. balangensis is not developed. These data were analysed using Past v.4.02 (Hammer et al. 2001). When digitizing landmarks, the specimens usually have different sizes and are also present in different positions and rotations within the measuring equip- ment (Zelditch et al. 2004; Hammer & Harper 2006). To remove differences attributable to size, position, and orientation from configurations, a Procrustes- fitting transformation of data was applied before anal- yses, thus leaving only differences in shape (Zelditch et al. 2004; Hammer & Harper 2006; Webster & Sheets 2010). Principal component analysis (PCA) was used to compare the deviation of changes among samples in the statistical analysis. A scree plot was used in PCA to visually determine which underlying com- ponents explained the largest variability in the data (Zelditch et al. 2004). To assess difference in mor- phology between the sample groups in this study, we used canonical variates analysis (CVA, describing differences between groups) to explore the data. To test for shape differences between groups, we carried out a multivariate analysis of variance (MANOVA, statistical significance of across group differences was assessed). Furthermore, analysis of similarities Fig. 2. Photographs of Duyunaspis Zhang and Qian in Zhou et al., 1977 and Balangia Qian, 1961. A-C, Duyunaspis jianheensis from the Songshan section (Q) and Malipo Section (QM), Guizhou Province, South China. A, exoskeletons, Q52-2583; B, exoskele- tons, Q51-2441; C, exoskeletons, QM33-856. D-F, Balangia bal- angensis Qian, 1961 from a fossil site near the Jiaobang section, Guizhou Province, South China. D, exoskeletons, JJB-B-20; E. exo- skeletons, JJB-B-1; F, exoskeletons, JJB-B-24; G-H, Duyunaspis duyunensis Zhang and Qian in Zhou et al. 1977. G, exoskeletons, NIGP 159524, Zila A section, Huayuan County, Hunan Province, South China (Lei & Peng 2014); H, exoskeletons, NIGP 159515, a fossil site near Zila A section, Huayuan County, Hunan Province, South China (Lei & Peng 2014); I, Duyunaspis paiwuensis Lei & Peng, 2014 from Zila A section, Huayuan County, Hunan Province, South China (Lei & Peng 2014), exoskeletons, NIGP 159531. The scale bars for all lines represent 1 mm except for H where the scale bar represent 0.5 mm. G–I courtesy of Qianping Lei. Chen et al. 4 (ANOSIM, comparing within-group and between- group distances) and thin-plate spline (TPS, visu- alizing shape change as a deformation) was used to compare the similarities and differences among the taxa. With PCA we can obtain information on mor- phological features that distinguish interspecific and intraspecific variations, that is, features that vary along the PC axis; CVA describes variations among groups, while ANOSIM is used for comparison of taxonomic composition in two or more groups of samples, both of which can be used to help determine the validity of taxon classification. With TPS we can visualize the average morphologic variations either in interspecific or different populations. Results Assessment of morphological variation in Duyunaspis and Balangia PC1 explains 27.44% of the variance and is primar- ily related to the posterior branch of facial suture and the occipital ring (Fig. 4A). Positive scores along PC1 are associated with posterior branches of the facial suture away from the occipital ring (i.e. proparian) as well as a longer (sag.) occipital ring. Negative scores along PC1 are associated with posterior branches of the facial suture closer to the occipital ring (i.e. gona- toparian or opisthoparian) and a shorter (sag.) occip- ital ring. Duyunaspis duyunensis and D. jianheensis are better distinguished on PC1, although with a lit- tle overlap. The samples of Balangia balangensis and D. duyunensis partially overlap on PC1. On PC1, most D. jianheensis samples occupy the –0.05 –0.1 interval; most D. duyunensis samples occupy the –0.15 –0.05 interval; and most B. balangensis samples occupy the –0.10 –0.00 interval. PC2 explains 17.79% of the variance, primarily related to the palpebral lobe and glabella and the posterior branch of the facial suture (Fig. 4A). Positive scores along PC2 are associated with a longer (sag.) palpebral lobe as well as posterior branches of the facial suture closer to the occipital ring (i.e. gonatoparian or opisthoparian). Negative scores along PC2 are associated with shorter (sag.) palpebral lobe and posterior branches of the facial suture away from the occipital ring (i.e. proparian). D. duyunen- sis, D.  jianheensis and B. balangensis are overlapping Fig. 3. Landmarks selected for morphometric analyses. A, reconstruction of the Duyunaspis cephalon showing chosen landmarks. B, reconstruction of the Balangia cephalon showing chosen landmarks. C, locations of cephalic landmarks for analysed specimens. Variability in the trilobites Duyuanaspis and Balangia 5 Fig. 4. A, B, principal component analysis (PCA) of Duyunaspis duyunensis, Duyunaspis jianheensis, Duyunaspis paiwuensis and Balangia balangensis. Thin-plate spline indicates the extreme shape for each axis. C, D, scree plot from principal component analysis. BL, Bulin section, Hunan Province (see Dai et al. 2017); JB, a fossil site near the Jiaobang section, Guizhou Province; ML, Malipo section, Guizhou Province; SS, Songshan section, Guizhou Province; ZA, Zila A section, Hunan Province (see Lei 2016); ZB, a fossil site near the Zila A section (see Lei 2016). in PC2. PC3 explains 7.81% of the variance, which mainly correlates to the relative distance of the pal- pebral lobe from the glabella (Fig. 4B). Positive scores along PC3 are associated with the palpebral lobe far from the glabella. Negative scores along PC3 are asso- ciated with the palpebral lobe close to the glabella. D. duyunensis, D. jianheensis and B. balangensis are overlapping in the PC3. D. paiwuensis, D. jianheen- sis and B. balangensis are distinguished in the PCA, but D. paiwuensis and D. duyunensis are overlapping (Fig. 4A–B). Although the results from PC1, PC2 and PC3 only account for 53.04% of the variance in total, the scree plot of the PCA (Fig. 4C) shows that the eigenvalue curve begins to flatten after PC3 (i.e. the ‘inflection point’), so it is feasible to select and utilize the first three principal components. The CVA (Fig. 5) shows that variations among groups occur in the Duyunaspis duyunensis, D. jian- heensis, D. paiwuensis and Balangia balangensis. Although D. duyunensis and D. paiwuensis have a large overlap in the PCA analysis (Fig. 4A–B), they are Fig. 5. Canonical variate analysis (CVA) scatter plot of Duyunaspis duyunensis, Duyunaspis jianheensis, Duyunaspis paiwuensis and Balangia balangensis. BL, Bulin section, Hunan Province (see Dai et al. 2017); JB, a fossil site near the Jiaobang section, Guizhou Province; ML, Malipo section, Guizhou Province; SS, Songshan section, Guizhou Province; ZA, Zila A section, Hunan Province (see Lei 2016); ZB, a fossil site near the Zila A section (see Lei 2016). completely separated in the CVA scatter plot (Fig. 5). This results need to be verified becuase too few spec- imens of D. paiwuensis were used. MANOVA (Wilks’ lambda  =  0.0653, df1  =  96, df2  =  641.5, F  =  9.946, p (same) = 2.507 × 10–78; Pillai trace = 1.597, df1  = 96, df2  = 648, F = 7.683, p(same) = 1.887 × 10–60) reveal significant differences between the four species. ANOSIM (Permutation N: 9999, Mean rank within =  1.188  ×  104, Mean rank between  =  2.204 × 104, R = 0.6582, p (same) = 0.0001. Fig. 6A) indicate that Fig. 6. Analysis of similarities (ANOSIM) box plot. A, ANOSIM box plot of Duyunaspis duyunensis, Duyunaspis jianheensis, Duyunaspis paiwuensis and Balangia balangensis. B, ANOSIM box plot of Duyunaspis duyunensis, Duyunaspis jianheensis, and Duyunaspis paiwuensis. Chen et al. 6 the between-group variations are greater than the with- in-group variations, and within-group variations are all significantly smaller than the between-group variations in B. balangensis, D. duyunensis and D. jianheensis, indicating that the classification of the three groups is reasonable. D. paiwuensis shows large intra-group vari- ations, but the results need to be further verified due to a small sample size. Figure 7 shows the morphological differences of the mean shapes of cephalon among Duyunaspis duyun- ensis, D. jianheensis, D. paiwuensis and Balangia bal- angensis (landmarks related to the glabellar furrow were removed). The results show that the main vari- ations between B. balangensis and D. duyunensis and D. jianheensis lie in the posterior branches of the dor- sal facial suture (suture type) and the occipital ring. The major variations between D. duyunensis and D. jianheensis mainly occur in the posterior branches of the facial suture, the occipital ring and the palpebral lobe. B. balangensis is most similar to D. duyunensis (Fig. 7A) while it differs from D. paiwuensis (Fig. 7B; the results need to be verified due to limited speci- mens of D. paiwuensis). Although landmarks of the glabella furrow were removed, D. duyunensis share the highest similarity with D. jianheensis among the three species of Duyunaspis, and the differences are with the posterior section of the facial suture and the occipital ring (Fig. 7E). Interspecific and intraspecific variability of Duyunaspis PC1 explains 23.39% of the variance and relates pri- marily to the glabella, the posterior branch of the facial suture, and the occipital ring (Fig. 8A). Positive scores along PC1 are associated with a centrally expanded glabella, posterior branch of facial suture Fig. 7. Thin-plate spline (TPS) deformation from the mean shape of the cephalon of Duyunaspis and Balangia. A, TPS from the mean shape of specimens from Balangia balangensis (B) to Duyunaspis duyunensis (D); B, TPS from the mean shape of specimens from Balangia balangensis (B) to Duyunaspis paiwuensis (P); C, TPS from the mean shape of specimens from Balangia balangensis (B) to Duyunaspis jianheensis (J); D, TPS from the mean shape of specimens from Duyunaspis duyunensis (D) to Duyunaspis paiwuensis (P); E, TPS from the mean shape of specimens from Duyunaspis duyunensis (D) to Duyunaspis jianheensis (J); F, TPS from the mean shape of specimens from Duyunaspis paiwuensis (P) to Duyunaspis jianheensis (J). Fig. 8. A, B, principal component analysis (PCA) of Duyunaspis duyunensis, Duyunaspis jianheensis, and Duyunaspis paiwuen- sis. Thin-plate spline indicates the extreme shape for each axis. C, canonical variate analysis (CVA) scatter plot of Duyunaspis duyunensis, Duyunaspis jianheensis, and Duyunaspis paiwuen- sis. BL, Bulin section, Hunan Province (see Dai et al. 2017); ML, Malipo section, Guizhou Province; SS, Songshan section, Guizhou Province; ZA, Zila A section, Hunan Province (see Lei 2016); ZB, a fossil site near the Zila A section (see Lei 2016). Variability in the trilobites Duyuanaspis and Balangia 7 away from the occipital ring (i.e. proparian), and a longer (sag.) occipital ring. Negative scores along PC1 are associated with a subcylindrical glabella, posterior branches of the facial suture closer to the occipital ring (i.e. gonatoparian or opisthoparian), and shorter (sag.) palpebral lobes. Duyunaspis duyunensis and D. jianheensis are better distinguished on PC1, although with a little overlap. On PC1 (x-axis), most D. jian- heensis samples occupy the –0.05–0.1 interval; most D. duyunensis samples occupy the –0.15 –0.05 inter- val. PC2 explains 13.63% of the variance, primarily relating to the palpebral lobes, the glabella and the posterior branch of facial suture (Fig. 8A). Positive scores along PC2 are associated with longer (sag.) palpebral lobes, a narrower (tr.) glabella, and pos- terior branch of facial suture closer to the occipital ring (i.e. gonatoparian or opisthoparian). Negative scores along PC2 are associated with shorter (sag.) palpebral lobes, wider (tr.) glabella, and posterior branches of the facial suture away from the occipital ring (i.e. proparian). D. duyunensis and D. jianheensis overlap in PC2. PC3 explains 7.89% of the variance, which mainly relates to the glabella and palpebral area (Fig.  8B). Positive scores along PC3 are associated with a wide (tr.) glabella and a narrow (tr.) palpebral area. Negative scores along PC3 are associated with a narrow (tr.) glabella and a wide (tr.) palpebral area. D. duyunensis and D. jianheensis overlap in PC3. D. paiwuensis and D. jianheensis are distinguished in the PCA, but D. paiwuensis and D. duyunensis are over- lapping (Fig. 8A–B). Although the results from PC1, PC2 and PC3 only account for 44.91% of the variance in total, the scree plot of the PCA (Fig. 4D) shows that the eigenvalue curve begins to flatten after PC3 (i.e. ‘inflection point’), so it is feasible to select and utilize the first three principal components. The CVA results show that between-group varia- tions existed among the three species of Duyunaspis (Fig. 8C). Although D. duyunensis and D. paiwuen- sis have a large overlap in the PCA (Fig. 8A, B), they are completely separated in the CVA scatter plot (Fig. 8C), the results need to be verified as few speci- mens of D. paiwuensis were used. MANOVA (Wilks’ lambda  =  0.1085, df1  =  96, df2  =  356, F  =  7.552, p (same) = 8.603 × 10–46; Pillai trace = 1.24, df1 = 96, df2 = 358, F = 6.088, p (same) = 1.68 × 10–36) reveal significant differences between the three species. ANOSIM (Permutation N  =  9999, Mean rank within = 1.097 × 104, Mean rank between = 1.944 × 104, R  =  0.6544, p (same)  =  0.0001. Fig. 6B) means that the between-group variations are greater than the within-group variations, and the within-group vari- ations between D. duyunensis and D. jianheensis are both significantly smaller than the between-group variations, which indicate that the classification of the two groups is reasonable. While D. paiwuensis shows large within-group variations, the results need to be further verified owing to the small sample size. The analysis results of CVA (Fig. 9) show that Duyunaspis jianheensis from the Songshan section are different from those from the Malipo section, and the two do not fully overlap in the CVA scatter plot. D. duyunensis collected from the Bulin section, Zila A section and Zila B fossil site have only a slight differ- ence, which is shown by the large overlap in the CVA scatter plot (Fig. 9). Figure 10 (A, B) shows the PCA of Duyunaspis duyunensis, D. jianheensis and D. paiwuensis during different ontogenetic stages at different localities. PC1 explains 23.39% of the variance, and mainly relates to the glabella, the posterior branch of the facial suture, and the occipital ring (Fig. 8C). Positive scores along PC1 are associated with a centrally expanded glabella, posterior branch of facial suture away from the occip- ital ring (i.e. proparian), and a longer (sag.) occipital ring. Negative scores along PC1 are associated with a subcylindrical glabella, posterior branches of the facial suture closer to the occipital ring (i.e. gonato- parian or opisthoparian), and shorter (sag.) palpebral lobes. D. duyunensis at different ontogenetic stages from different localities overlap in PC1. D. jianheensis at different ontogenetic stages from different localities Fig. 9. Canonical variate analysis (CVA) scatter plot of Duyunaspis duyunensis, Duyunaspis jianheensis, and Duyunaspis paiwuen- sis from different localities. BL, Bulin section, Hunan Province (see Dai et al. 2017); ML, Malipo section, Guizhou Province; SS, Songshan section, Guizhou Province; ZA, Zila A section, Hunan Province (see Lei 2016); ZB, a fossil site near the Zila A section (see Lei 2016). Chen et al. 8 also overlap in PC1. However, D. duyunensis and D. jianheensis are better distinguished on PC1. On PC1, most D. jianheensis samples occupy the –0.05–0.1 interval; most D. duyunensis samples occupy the –0.15–0.05 interval. PC2 explains 13.63% of the vari- ance, primarily relating to the palpebral lobe, the glabella and the posterior branch of facial suture. Positive scores along PC2 are associated with longer (sag.) palpebral lobes, a narrower (tr.) glabella, and posterior branch of facial suture closer to the occipi- tal ring (i.e. gonatoparian or opisthoparian). Negative scores along PC2 are associated with shorter (sag.) palpebral lobes, a wider (tr.) glabella, and posterior branches of the facial suture away from the occip- ital ring (i.e. proparian). Both of D. duyunensis and D. jianheensis at different ontogenetic stages from dif- ferent localities overlap in PC2. PC3 explains 7.89% of the variance, which mainly relates to the glabella and palpebral area. Positive scores along PC3 are associated with wide (tr.) glabella and a narrow (tr.) palpebral area. Negative scores along PC3 are associ- ated with a narrow (tr.) glabella and a wide (tr.) palpe- bral area. D. duyunensis at different ontogenetic stages from different localities overlap in PC3. D. jianheensis at different ontogenetic stages from different localities also overlap in PC3. Meanwhile, D. duyunensis and D. jianheensis overlap in PC3. Although the results from PC1, PC2 and PC3 only account for 44.91% of the variance in total, the scree plot of the PCA (Fig. 10C) shows that the eigenvalue curve begins to flatten after PC3 (i.e. ‘inflection point’), so it is feasible to select and utilize the first three principal compo- nents. These results suggest that morphological traits of Duyunaspis at genus level were well established in the early meraspids stages whereas the cephalic shape shows a large morphological disparity. The CVA of different ontogenetic stages of Duyunaspis from different localities (Fig. 11) shows small variations between the different ontogenetic series. Particularly, there are variations among sam- ples from the early meraspid series and the late mer- aspid series or holaspid. ANOSIM (Permutation N  =  9999, Mean rank within  =  9075, Mean rank between = 1.102 × 104, R = 0.1997, p (same) = 0.0001. Fig. 12A) means that the within-group variations of Fig. 10. A, B, principal component analysis (PCA) of Duyunaspis duyunensis, Duyunaspis jianheensis, and Duyunaspis paiwuensis during different ontogenetic stages at different localities. Thin-plate spline indicates the extreme shape for each axis. C, scree plot from principal component analysis. BL, Bulin section, Hunan Province (see Dai et al. 2017); ML, Malipo section, Guizhou Province; SS, Songshan section, Guizhou Province; ZA, Zila A section, Hunan Province (see Lei 2016); ZB, a fossil site near the Zila A section (see Lei 2016). Fig. 11. Canonical variate analysis (CVA) scatter plot of Duyunaspis duyunensis, Duyunaspis jianheensis, and Duyunaspis paiwuensis during different ontogenetic stages at different locali- ties. BL, Bulin section, Hunan Province (see Dai et al. 2017); ML, Malipo section, Guizhou Province; SS, Songshan section, Guizhou Province; ZA, Zila A section, Hunan Province (see Lei 2016); ZB, a fossil site near the Zila A section (see Lei 2016). Variability in the trilobites Duyuanaspis and Balangia 9 D. jianheensis samples from the Songshan and Malipo sections are similar to the between-group variations, and the within-group variations are slightly greater than the between-group variations in D. jianheen- sis samples from the Malipo section. This indicates the failure to distinguish D. jianheensis samples in both sections, which means D. jianheensis in the Songshan and Malipo sections should be treated as the same taxon. ANOSIM (Permutation N  =  9999, Mean rank within  =  159.8, Mean rank between: 183.3, R = 0.1335, p (same) = 0.0089; Fig. 12B) shows that the within-group variations are greater than the between-group variations in D. duyunensis samples from the Bulin and Zila A sections, and the with- in-group variations is less than the between-group variations in D. duyunensis samples from the Zila B fossil site, which indicates that the grouping of D. duyunensis samples from the Bulin and Zila A sec- tions is a failure, meaning that they should be the same taxon. Figure 13 shows the morphological variations of the mean shapes of the cephalon among Duyunaspis duyunensis, D. jianheensis, and D. paiwuensis, as well as the morphological differences of the mean shapes of the cephalon of D. duyunensis or D. jianheensis, from different localities. The variations between D. duyunensis and D. jianheensis are mainly seen in the glabella, posterior branches of the facial suture (facial suture type) and occipital ring (Fig. 13A). The vari- ations between D. duyunensis and D. paiwuensis are mainly seen in the glabella and posterior branch of the facial suture (Fig. 13B). The difference between D. paiwuensis and D. jianheensis mainly occurs in the central region of glabella, occipital ring, facial suture and palpebral lobes (Fig. 13C). However, the results of the comparison need to be further verified owing to the small specimen size of D. paiwuensis. There is almost no variation in the cephalon of D. jianheen- sis from the Songshan and Malipo sections in Jianhe County, Guizhou Province (Fig. 13D), and there is a slight variation among the cephalon of D. duyunensis from the Zila A section, Zila B fossil site and Bulin section. Fig. 12. A, analysis of similarities (ANOSIM) box plot of Duyunaspis jianheensis from two sections. B, analysis of similar- ities (ANOSIM) box plot of Duyunaspis duyunensis from three sections. BL, Bulin section, Hunan Province (see Dai et al. 2017); ML, Malipo section, Guizhou Province; SS, Songshan section, Guizhou Province; ZA, Zila A section, Hunan Province (see Lei 2016); ZB, a fossil site near the Zila A section (see Lei 2016). Fig. 13. Thin-plate spline (TPS) deformation from the mean shape of the specimens from four species of Duyunaspis. A, TPS from the mean shape of specimens from Duyunaspis duyunensis (D) to Duyunaspis jianheensis (J); B, TPS from the mean shape of speci- mens from Duyunaspis duyunensis (D) to Duyunaspis paiwuensis (P); C, TPS from the mean shape of specimens from Duyunaspis paiwuensis (P) to Duyunaspis jianheensis (J); D, TPS from the mean shape of specimens from Duyunaspis jianheensis (Songshan section, Jsh) to Duyunaspis jianheensis (Malipo section, Jml); E, TPS from the mean shape of specimens from Duyunaspis duyunensis (Bulin section, Dbl) to Duyunaspis duyunensis (Zila A section, Dza); F, TPS from the mean shape of specimens from Duyunaspis duyunensis (Bulin section, Dbl) to Duyunaspis duyunensis (Zila B fossil site, Dzb); G, TPS from the mean shape of specimens from Duyunaspis duyunensis (Zila A section, Dza) to Duyunaspis duyun- ensis (Zila B fossil site, Dzb). bl, Bulin section, Hunan Province (see Dai et al. 2017); ml, Malipo section, Guizhou Province; ss, Songshan section, Guizhou Province; za, Zila A section, Hunan Province (see Lei 2016); zb, a fossil site near the Zila A section (see Lei 2016). Chen et al. 10 Discussion Duyunaspis has three known species: D. duyunensis, D. paiwuensis and D. jianheensis; and the interspecific variations between these trilobites have hitherto been poorly understood. By means of quantitative mea- surement, similarities and variations in morphology (Figs 6B, 8) were found among them. PCA (Figs 4A, 8A) shows that samples of D. duyunensis and D. jian- heensis can be distinguished on PC1, whereas mor- phological features associated with PC2 and PC3 do not distinguish the two species (Figs 4A, B, 8A, B). Between-group variations between D. duyunen- sis and D. jianheensis were amplified by CVA, and neither of them shows overlap, indicating that they are different from each other, but CVA did not indi- cate the significance of the variations. The ANOSIM shows that both D. duyunensis and D. jianheensis had fewer within-group variations than between-group variations and the variations were significant; that is, both could be divided into two taxa based on CVA and ANOSIM. Based on CVA and ANOSIM, it suggests that there are significant variations between D. duyunen- sis and D. jianheensis, as can be seen in the type of facial suture, size of the occipital ring, and shape of the middle of the glabella associated with PC1. The variations were also reflected in the TPS of the mean morphology. The palpebral lobes of D. duyunensis and D. jianheensis also differ in the TPS (Figs 7D–F, 13A–C), but the palpebral lobes did not effectively distinguish the two species in the PCA (Figs 4A, B, 8A, B), so the variation in the palpebral lobes cannot show interspecific variation. The main features of the cephalon of D. duyunensis are: subcylindrical glabella, short (sag.) occipital ring, and opisthopar- ian suture in the holaspid. The main features of the cephalon of D. jianheensis are: medially expanded glabella, longer (sag.) occipital ring, and proparian suture. D. paiwuensis shows large variations from D. duyunensis and D. jianheensis (Figs 4A, B, 5, 7D, F, 8, 13A, C), but the results have yet to be verified for the limited specimens of D. paiwuensis, and need to be further discussed. In addition to the cephalic mor- phology, D. paiwuensis (nine segments, Lei & Peng 2014) differed from D. duyunensis (ten segments, Dai et al. 2017) or D. jianheensis (ten segments, Chen et al. 2018) (Fig. 2G–I) in having nine–ten thoracic seg- ments in the holaspid stage. Thus, the holaspid tho- racic segments can also be used to show interspecific variations, but this needs to be assessed because there is little change in the number of segments between different species belonging to the same genus (Yuan et al. 2002; Peng et al. 2017). Intraspecific variation is the raw material of natural selection (West-Eberhard 2005; Webster 2007; Hopkins 2011). CVA (Fig. 9) is a magnification of the variations between groups and does not tell whether the varia- tions are significant. Duyunaspis jianheensis are from adjacent sections, which have the same lithology and burial environment. D. duyunensis are from three sec- tions, which have approximately the same lithology and burial environment (see Lei & Peng 2014; Lei 2016; Dai et al., 2017). Therefore, we argue that the variation in morphology of ontogenetic stages is the main rea- son for the observed intraspecific variation. The study of morphological changes in ontogeny may represent a good microevolutionary approach (McNamara 2009). We conducted PCA (Fig.  10) and CVA (Fig. 11) of different ontogenetic series in different selections to explore whether the reasons for intraspecific variation are related to geographical or ontogenetic variations. On PC1, samples from different ontogenetic stages of D. duyunensis and D. jianheensis are scattered and not all together, but this does not effectively distinguish between D. duyunensis and D. jianheensis (Fig. 10A, B). That is, the shape of the glabella, relative position of the posterior branches of the facial suture to the cephalon (facial suture type) and size of the occipital ring asso- ciated with PC1 can be regarded as intraspecific varia- tions, and although these features are also interspecific variations in Duyunaspis, they do not reach interspe- cific variations within the species. Both PC2 and PC3 did not distinguish between D. duyunensis and D. jian- heensis, and there was a span of different ontogenetic stages, so the palpebral lobes, glabella, and posterior branches of the facial suture associated with PC2 and PC3 could all be used as intraspecific variation. In the CVA (Fig. 11), there are variations in the different ontogenetic series, but these variations do not affect species differentiation, that is, they do not represent interspecific variations. Although the num- ber of ontogenetic stages of the samples used in the analysis was not the same, the failure to distinguish groupings for D. duyunensis or D. jianheensis in dif- ferent sections was shown in the ANOSIM (Fig. 12), indicating that they had some ontogenetic variation, but this variation was small and did not rise to the level of interspecific variations. In the TPS (Fig. 13D–G), the average morphology between D. duyunensis and D. jianheensis in different sections showed minimal or essentially no variation. Intraspecific variations in D. duyunensis and D. jian- heensis were expressed in the shape of the glabella, type of facial suture, size of the palpebral lobes and occip- ital ring, but although some characters are the same as for interspecific variations in Duyunaspis, intra- specific variation is small compared to interspecific Variability in the trilobites Duyuanaspis and Balangia 11 variations. For example, the glabella of D. duyunensis is subcylindrical in shape, and intraspecific variation revolves around this shape; the glabella of D. jian- heensis is expanded at the middle, and interspecific variation occurs within the degree of expansion. The posterior branches of the facial suture of D. duyunen- sis vary near the genal angle; the posterior branches of the facial suture of D. jianheensis vary above the genal angle. Because Balangia bears an effaced glabela (Balangia has been wrongly described as possessing glabellar furrow in past reports, see Qian 1961, pl. 1, figs 14–16, 18, McNamara et al. 2006, text–fig. 6) and the four species of trilobites have an approximately or nearly cylindrical glabella. Landmarks related to the glabellar furrow or the landmarks opposite to the axial furrow were excluded (landmarks 9–12, landmarks 19–22 in Fig. 3). CVA (Fig. 5, intergroup variations exist between B. balangensis and D. duyunensis and D. jianheensis and could be distinguished) and ANOSIM (Fig. 6A, within-group variations were less than between-group variations among B. balangensis, D. duyunensis and D. jianheensis, and significant) shows that B. balangensis differs from D. duyunensis or D. jianheensis, and the three can be treated as distinct species. However, in the PCA (Fig. 4A), on the PC1 axis (associated with the posterior branches of the facial suture and occipital ring) samples of B. balangensis could be distinguished to some extent from samples of either D. duyunensis or D. jianheensis, but partial overlap occurs between samples of B. balangensis and either D. duyunensis or D. jianheensis. In particular, a large overlap exists in the samples of B. balangensis and D. duyunensis (PC1; Fig. 4A) and B. balangensis was also closer to D. duyunensis in the CVA (Fig. 5). This suggests a close similarity in the cephalic morphol- ogy between B. balangensis and D. duyunensis. The type of facial suture and the size of the occipital ring could be used to distinguish between Balangia and Duyunaspis, but they are the same as the interspecific variations between Duyunaspis. Therefore, variations in the facial suture and occipital ring cannot be taken as variations between the Balangia and Duyunaspis. In addition, the landmarks associated with the glabel- lar furrow were not considered when conducting the analysis, as the glabellar furrow of B. balangensis is not developed. In contrast, all three species of Duyunaspis have glabellar furrows. The presence or absence of the glabellar furrow can be used as a feature to distinguish between Balangia and Duyunaspis. Holaspid of Balangia balangensis has only four seg- ments (Fig. 2J; see Qian 1961; McNamara et al. 2006), a relatively unusual species of oryctocephalid trilobites. In contrast, the holaspid of Duyunaspis have more thoracic segments (D. paiwuensis, nine segments, Lei & Peng 2014; D. duyunensis, ten segments, Dai et al. 2017; D. jianheensis, ten segments, Chen et al. 2018). In addition, all three species of Duyunaspis are micropygous in the holaspid (Lei 2016; Dai et al. 2017; Chen et al. 2018), while B. balangensis is isopy- gous (Qian 1961; Yuan et al. 2001; McNamara et al. 2006). Therefore, the number of thoracic segments and the relative size of the pygidium can also distin- guish between Balangia and Duyunaspis. Balangia is thought to be a descendant of Duyunaspis (McNamara et al. 2006), but this evolutionary relationship is based on heterochrony, which is difficult to determine in trilobites (Hughes et al. 2006). Thus, we suggest that Balangia is likely related to Duyunaspis, but further discussion on their ancestral relationship is needed. Conclusions Based on geometric, morphometric methods (PCA, CVA, MANOVA, ANOSIM, TPS), the studies carried out here indicate that interspecific variations among Duyunaspis jianheensis, D. duyunensis and D. paiwuen- sis mainly lie in type of facial suture, occipital ring and glabella. The variations of D. duyunensis or D. jianheen- sis from different localities are smaller than interspecific variations, which are principally due to their ontogeny variation. Balangia balangensis shared similarities with D. duyunensis and D. jianheensis in the cephalon, how- ever, the variations between Balangia and Duyunaspis in holaspid phase mainly lay in the glabellar furrow, the number of thoracic segments and the relative size between pygidium and cephalon. Acknowledgements. – We would like to express our thanks to Mingkun Wang and Dezhi Wang for assistance with the morphom- etry. Harriet B Drage and an anonymous reviewer are thanked for providing critical comments. We thank to Jillian Pearse for her comments and improvement of written English; to Qianping Lei for trilobite images. Financial supports by the National Natural Science Foundation of China (grant numbers 41962002, 41772021, 42162005), the Guizhou Bureau of Science and Technology (grant numbers Gui. Sci. Sup. [2020]4Y241; Gui. Sci. Tal. [2017] 5788), and the strategic Priority Research Program of Chinese Academy of Sciences (XDB26000000). Supporting Information Additional supporting information may be found online in the Supporting Information section at the end of the article. Table S1. Intergeneric variation data. Table S2. Interspecific variation data. Table S3. Intraspecific variation data. http://\\172.16.0.22\d\From_Customer\0133\Input\2022\Journals\09_Lethaia\Lethaia-4-2022\First files\05_Chen\S1._Intergeneric_variation_data http://\\172.16.0.22\d\From_Customer\0133\Input\2022\Journals\09_Lethaia\Lethaia-4-2022\First files\05_Chen\S2._Interspecific_variation_data http://\\172.16.0.22\d\From_Customer\0133\Input\2022\Journals\09_Lethaia\Lethaia-4-2022\First files\05_Chen\S3._Intraspecific_variation_data Chen et al. 12 References Abe, F.R. & Lieberman, B.S. 2012. Quantifying morphological change during an evolutionary radiation of Devonian trilobites. Paleobiology 38, 292–307. https://doi.org/10.1666/10047.1 Álvaro, J.J. & Esteve J. 2020. Reply to Comment on: Álvaro J.J., Esteve, J. & Zamora, S. 2019. Morphological assessment of the earliest paradoxidid trilobites (Cambrian Series 3) from Morocco and Spain by Geyer G, Nowicki J, Żylińska A & Landing E. Geological Magazine 157, 1971–1982. https://doi. org/10.1017/S0016756820000217 Álvaro, J.J., Esteve, J. & Zamora, S. 2018. Morphological assess- ment of the earliest paradoxidid trilobites (Cambrian Series 3) from Morocco and Spain. Geological Magazine 155, 1566–1595. https://doi.org/10.1017/S0016756817000449 Bicknell, R.D.C., Paterson, J.R. & Hopkins, M.J. 2019. A trilobite cluster from the Silurian Rochester Shale of New York: pre- dation patterns and possible defensive behavior. American Museum Novitates 2019, 1–16. https://doi.org/10.1206/3937.1 Chen, Z.P., Zhao, Y.L., Yanng, X.L. & Wang, M.K. 2018. Duyunaspis Zhang and Qian in Zhou et al., 1977 (trilobite) from the ‘Tsinghsutung Formation’ (Series 2, Stage 4) of eastern Guizhou, South China. Acta Palaeontologica Sinica 57, 181– 191. [in Chinese]. Dai, T., Zhang X.L., Peng, S.C. & Yao, X.Y. 2017. Intraspecific vari- ation of trunk segmentation in the oryctocephalid trilobite Duyunaspis duyunensis from the Cambrian (Stage 4, Series 2) of South China. Lethaia 50, 527–539. https://doi.org/10.1111/ let.12208 Esteve, J., Zhao, Y.L. & Peng, J. 2017. Morphological assessment of the Cambrian trilobites Oryctocephalus indicus (Reed 1910) from China and Oryctocephalus ‘reticulatus’ (Lermontova 1940) from Siberia. Lethaia 50, 175–193. https://doi.org/10.1111/ let.12185 Gendry, D., Courville, P., Saucède, T., Laffont, R. & Paris, F. 2013. Contribution of morphometrics to the systematics of the Ordovician genus Neseuretus (Calymenidae, Trilobita) from the Armorican Massif, France. Journal of Paleontology 87, 456–471. https://doi.org/10.1666/12-046.1 Hammer, Ø. & Harper, A.T. 2006. Paleontological Data Analysis. Blackwell Publishing, Oxford, 351 pp. Hammer, Ø., Harper, D.A.T. & Ryan, P.D. 2001. PAST: paleonto- logical statistics software package for education and data anal- ysis. Palaeontologia Electronica 4, 1–9. Hopkins, M.J. & Webster, M. 2009. Ontogeny and geographic variation of a new species of the corynexochine trilobite Zacanthopsis (Dyeran, Cambrian). Journal of Paleontology 83, 524–547. https://doi.org/10.1666/08-102R.1 Hopkins, M.J. 2011. How species longevity, intraspecific mor- phological variation, and geographic range size are related: a comparison using late Cambrian trilobites. Evolution 65, 3253– 3273. http://dx.doi.org/10.1111/j.1558-5646.2011.01379.x Hughes, N.C., Minelli, A. & Fusco, G. 2006. The ontogeny of tri- lobite segmentation: a comparative approach. Paleobiology 32, 602–627. https://doi.org/10.1666/06017.1 Lei, Q.P. 2016. New ontogenetic information on Duyunaspis Zhang and Qian in Zhou et al., 1977 (Trilobita Corynexochida) from Cambrian and its possible sexual dimorphism. Alcheringa 40, 12–23. https://doi.org/10.1080/03115518.2015.1069485 Lei, Q.P. & Peng. S.C. 2014. Duyunaspis Zhang and Qian in Zhou et al., 1977 (Trilobita) from the Balang Formation (Cambrian, Qiandongia) in northwestern Hunan Province and its intra- specific variations. Acta Palaeontologica Sinica 53, 352–362. [in Chinese]. McNamara, K.J. 2009. The importance of developmental repattern- ing in the evolution of trilobites. Journal of the Royal Society of Western Australia 92, 389–398. McNamara, K.J., Yu, F. & Zhou, Z.Y. 2006. Ontogeny and heteroch- rony in the Oryctocephalid trilobites Changaspis, Duyunaspis and Balangia from China. Palaeontology 49, 1–19. https://doi. org/10.1111/j.1475-4983.2005.00525.x Oudot, M., Neige, P., Laffont, R., Navarro, N., Khaldi, A.Y. & Crônier, C. 2019. Functional integration for enrolment con- strains evolutionary variation of phacopid trilobites despite developmental modularity. Palaeontology 62, 805–821. https:// doi.org/10.1111/pala.12428 Peng, S.C. 2000. Cambrian of slope facies (of China). In: Nanjing Institute of Geology and Palaeontology. (Ed.) Stratigraphical Studies in China (1979–1999). University of Science and Technology of China Press, Hefei, pp. 23–38. [in Chinese] Peng, S.C. 2009. Review on the studies of Cambrian trilobite faunas from Jiannan Slope Belt, South China, with notes on Cambrian correlation between south and North China. Acta Palaeontologica Sinica 48, 437–452. [in Chinese] Peng, S.C., Babcock, L.E. & Cooper, R.A. 2012a. The Cambrian Period. In: Gradstein, F.M., Ogg, J.G., Schmitz, M.D. & Ogg, G.M. (Eds), The Geologic Time Scale 2012. Elsevier, Amsterdam, pp. 437–488. Peng, S.C., Babcock, L.E., Robison, R.A., Lin, H.L., Ress, M.N. & Saltzman, M.R. 2004. Global Standard Stratotype- section and Point (GSSP) of the Furongian Series and Paibian Stage (Cambrian). Lethaia 37, 365–379. https://doi. org/10.1080/00241160410002081 Peng, S.C., Babcock, L.E., Zhu, X.J. & Dai, T. 2018. A new orycto- cephalid trilobite from the Balang Formation (Cambrian Stage 4) of northwestern Hunan, South China, with remarks on the classification of oryctocephalids. Palaeoworld 27, 322–333. https://doi.org/10.1016/j.palwor.2018.04.005 Peng, S.C., Babcock, L.E., Zhu, X.J., Lei, Q.P. & Dai, T. 2017. Revision of the oryctocephalid trilobite genera Arthricocephalus Bergeron and Oryctocarella Tomashpolskaya and Karpinski (Cambrian) from South China and Siberia. Journal of Paleontology 91, 933–959. https://doi.org/10.1017/jpa.2017.44 Peng, S.C., Babcock, L.E., Zuo, J.X., Lin, H.L., Zhu, X.J., Yang, X.F., Qi, Y.P., Bagnoli, G. & Wang, L.W. 2012b. Global Standard Stratotype-Section and Point (GSSP) for the base of the Jiangshanian Stage (Cambrian: Furongian) at Duibian, Jiangshan, Zhejiang, Southeast China. Episodes 35, 462–477. https://doi.org/10.18814/epiiugs/2012/v35i4/002 Qian, Y.Y. 1961. Cambrian trilobites from Sandu and Duyun, southern Guizhou. Acta Palaeontologica Sinica 9, 91–139. [in Chinese] Rohlf, F.J. 2010. tpsDig. Version 2.16. Department of Ecology and Evolution, Sate University of New York at Stony Brook. http:// life.bio.sunysb.edu/morph Sundberg, F.A. 2014. Phylogenetic analysis of the spiny orycto- cephalids (Trilobita, Corynexochida?, Oryctocephalidae), Cambrian. Journal of Paleontology 88, 556–587. https://doi. org/10.1666/12-130 Webster, M. 2007. A Cambrian peak in morphological variation within trilobite species. Science 317, 499–502. http://dx.doi. org/10.1126/science.1142964 Webster, M. 2011. The structure of cranidial shape variation in three early ptychoparioid trilobite species from the Dyeran – Delamaran (traditional ‘lower – middle’ Cambrian) boundary interval of Nevada, USA. Journal of Paleontology 85, 179–225. https://doi.org/10.1666/10-075.1 Webster, M. & Sheets, H. 2010. A practical introduction to land- mark-based geometric morphometrics. The Paleontological Society Papers 16, 163–188. https://doi.org/10.1017/S10893 32600001868 West-Eberhard, M.J. 2005. Developmental plasticity and the origin of species variations. Proceedings of the National Academy of Sciences of the United States of America 102 (suppl. 1), 6543– 6549. http://dx.doi.org/10.1073pnas.0501844102 Whittington, H.B. 1995. Oryctocephalid trilobites from the Cambrian of North America. Palaeontology 38, 543–562. Yan, Q.J., Peng, J., Zhao, Y.L., Wen, R.Q. & Sun, H.J. 2014. Restudy of sedimentary and biostratigraphy of the Qiandongian (Cambrian) Balang Formation at Jianhe, Guizhou, China— Example for Jiaobang Section of the Balang Formation. Geological Review 60, 893–902. [in Chinese] https://doi.org/10.1666/10047.1 https://doi.org/10.1017/S0016756820000217 https://doi.org/10.1017/S0016756820000217 https://doi.org/10.1017/S0016756817000449 https://doi.org/10.1206/3937.1 https://doi.org/10.1111/let.12208 https://doi.org/10.1111/let.12208 https://doi.org/10.1111/let.12185 https://doi.org/10.1111/let.12185 https://doi.org/10.1666/12-046.1 https://doi.org/10.1666/08-102R.1 http://dx.doi.org/10.1111/j.1558-5646.2011.01379.x https://doi.org/10.1666/06017.1 https://doi.org/10.1080/03115518.2015.1069485 https://doi.org/10.1111/j.1475-4983.2005.00525.x https://doi.org/10.1111/j.1475-4983.2005.00525.x https://doi.org/10.1111/pala.12428 https://doi.org/10.1111/pala.12428 https://doi.org/10.1080/00241160410002081 https://doi.org/10.1080/00241160410002081 https://doi.org/10.1016/j.palwor.2018.04.005 https://doi.org/10.1017/jpa.2017.44 https://doi.org/10.18814/epiiugs/2012/v35i4/002 https://doi.org/10.1666/12-130 https://doi.org/10.1666/12-130 http://dx.doi.org/10.1126/science.1142964 http://dx.doi.org/10.1126/science.1142964 https://doi.org/10.1666/10-075.1 https://doi.org/10.1017/S1089332600001868 https://doi.org/10.1017/S1089332600001868 http://dx.doi.org/10.1073pnas.0501844102 Variability in the trilobites Duyuanaspis and Balangia 13 Yin, G.Z. & Li, S.J. 1978. Trilobita. In: Stratigraphical and Palaeontological Working Group of Guizhou Province (Ed.), Palaeontological Atlas of Southwest China, Guizhou Province (1). Geological Publishing House, Beijing, pp. 385–595. [in Chinese] Yuan, J.I., Zhao, Y.L. & Li, Y. 2001. Notes on the classification and phylogeny of oryctocephalids (Trilobita: Arthropoda). Acta Palaeontologica Sinica 40, 214–226. Yuan, J.L., Zhao, Y.L., Yue, Li. & Huang, Y.Z. 2002. Trilobite fauna of the Kaili Formation (uppermost Lower Cambrian lower- Middle Cambrian) from Southeastern Guizhou, South China. Shanghai Science and Technology Publishing House, 423 pp. [in Chinese] Zelditch, M.L., Swiderski, D.L., Sheets, H.D. & Fink, W.L. 2004. Geometric Morphometrics for Biologists: a Primer. Elsevier Academic Press, San Diego, 443 pp. Zhang, W.T., Lu, Y.H., Zhu, Z.L., Qian, Y.Y., Lin, H.L., Zhou, Z.Y., Zhang, S.G. & Yuan, J.L. 1980. Cambrian Trilobite Fauna of Southwestern China. New Series B. Science Press, Beijing, 479pp. [in Chinese] Zhao, W.Y., Liu, J.N. & Bicknell D.C. 2020. Geometric morphometric assessment of Guanshan trilobites (Yuannan Province, China) reveals a limited diversity of palaeolenid taxa. Palaeontologia Electronica 23, a22. https://doi.org/10.26879/1062 Zhao, Y.L., Yuan, J.L., Babcock, L.E., Guo, Q.J., Peng, J., Yin, L.M., Yang, X.L., Peng, S.C., Wang, C.J., Gaines, R.R., Esteve, J., Tai, T.S., Yang, R.D., Wang, Y., Sun, H.J. & Yang, Y.N. 2019. Global Standard Stratotype-Section and Point (GSSP) for the con- terminous base of the Miaolingian Series and Wuliuan Stage (Cambrian) at Balang, Jianhe, Guizhou, China. Episodes 42, 165–184. https://doi.org/10.18814/epiiugs/2019/019013 Zhou, T.M., Liu, Y.R., Meng, X.S. & Sun, Z.H. 1977. Trilobita. In: Hubei Institite of Geological Sciences, Geological Bureau of Henan Province, Geological Bureau of Hubei Province, et al., (Eds), Palaeontological Atlas of Central and Southern China (1), Early Palaeozoic. Geological Publishing House, pp. 104–266. [in Chinese]. https://doi.org/10.26879/1062 https://doi.org/10.18814/epiiugs/2019/019013