(cambrian) trilobite guangxiaspis guangxiensis zhou, 1977 and the ...

J. Paleont., 84(3), 2010, pp. 493–504 Copyright ’ 2010, The Paleontological Society 0022-3360/10/0084-0493$03.00

VENTRAL STRUCTURE AND ONTOGENY OF THE LATE FURONGIAN (CAMBRIAN) TRILOBITE GUANGXIASPIS GUANGXIENSIS ZHOU, 1977 AND THE DIPHYLETIC ORIGIN OF THE MEDIAN SUTURE XUE-JIAN ZHU,1 NIGEL C. HUGHES,2

AND

SHAN-CHI PENG1

1

State Key Laboratory on Palaeontology and Stratigraphy, Nanjing Institute of Geology and Palaeontology, Chinese Academy of Sciences, Nanjing 210008, China, ,[email protected]., ,[email protected].; 2 Department of Earth Sciences, University of California, Riverside, California 92521, USA, ,[email protected].

ABSTRACT—Articulated meraspid and holaspid exoskeletons of Guangxiaspis guangxiensis from the Guole Township, Jingxi County, Guangxi Province, China, are preserved in mudstone deposited during an obrution event. The species has a short dorsal pre-cranidial median suture that splits ventrally into a pair of posteriorly divergent connective sutures. The rostral plate of G. guangxiensis is thus triangular in outline, as in the co-occurrent Shergoldia laevigata, which also bore a conterminant hypostome. These two taxa appear to be closely related. The cephalic venter of Shergoldia laevigata has recently been interpreted to suggest a diphyletic origin of the median suture within the order Asaphida, but Guangxiaspis guangxiensis, Shergoldia laevigata and other tsinaniid trilobites display several characters reminiscent of members of the non-asaphide suborder Leiostegiina. These include swellings adjacent to the margins of the L1 glabellar lobe, the shape and furrows of the glabella, a semi-circular pygidium with a long and thin axis, and macrospinous first opisthopleurae of the holaspid pygidium. Based on these characters and on other new information on the early ontogeny of other tsinaniids, all these taxa likely belong within Leiostegiina. This suggests that the median suture arose independently in corynexochide and asaphide trilobites. The degree of convergence between S. laevigata and members of the derived asaphide family Asaphidae was remarkable. Guangxiaspis guangxiensis shows marked morphological change during both meraspid and holaspid ontogeny and might include more than a single morphotype.

INTRODUCTION

of late Cambrian age in the vicinity of Guole Township, Jingxi County, Guangxi Province, China, contain the articulated exoskeletons of the meraspid and holaspid stages of a number of trilobite species. These specimens are well preserved and provide important new data on the ventral structure of the cephalon. This information has relevance for higher-level trilobite systematics and particularly for assessing the monophyly of the order Asaphida Salter, 1864, emended Fortey and Chatterton, 1988 (e.g., Zhu et al., 2007). Here we describe Guangxiaspis guangxiensis Zhou in Zhou et al., 1977, highlighting details of the cephalic venter, of the preservation and taphonomy of this trilobite, and of its ontogeny. Guangxiaspis guangxiensis was erected based on an almost complete internal mold of the exoskeleton and the internal mold of a cranidium. Both were collected from the Sandu Formation in Hewen Village, Jingxi County, Guangxi Province, China (Fig. 1). In Guole Township of Jingxi County, some 4.3 km northeast of Hewen Village (Fig. 1), we collected abundant new material of G. guangxiensis. This locality, its geology, and associated specimens have been described by Han et al. (2000), Han and Chen (2004, 2008), and Zhu et al. (2007). It is from the equivalent of the Probinacunaspis nasalis-Peichiashania hunanensis Zone of northwestern Hunan and is about 491 Ma old (Peng et al., in press).

S

ILTSTONE BEDS

PRESERVATION AND TAPHONOMY

Although all the specimens were collected from a single locality, there is considerable variation in their form, some of which can be attributed to preservation. There are evident differences in characters expressed on the internal and external molds (e.g., compare Figs 3.1 and 3.2, 3.6 and 3.7, 4.2 and 4.3), differences in whether or not the genal and pygidial

spines collapsed during compaction (Figs 4.3 and 4.8), and differences in relief related to the amount of compressional flattening (compare Figs. 4.3 and 5.4). Specimens also show a variety of patterns of articulation. Some are complete exoskeletons that remained articulated with the ventral sclerites in place (Fig. 4.12). Such specimens ¨ pik, 1970). Many other speciplausibly represent carcasses (O mens show complete dorsal shields with all dorsal sclerites apparently in the life position (e.g., Figs. 4.1–4.3, 4.9–4.11, 5.1–5.3, 5. 7–5.9). Other specimens show disarticulation. These include apparently complete dorsal exoskeletons in which some portion of the trunk has been displaced with respect to the remainder of the exoskeleton (e.g., Figs. 4.8, 5.4, 5.10). These specimens show extension in the thoracic region, presumably associated with the rupture of arthrodial membranes, and commonly also with rotation of the posterior portion of the trunk. While such configurations have been considered to imply molting behavior (Henningsmoen, 1975), the presence of free cheeks in their original positions in some of these specimens hinders this interpretation. Some specimens are what Henningsmoen (1975) termed ‘‘axial shields:’’ the cranidium articulated with the trunk (Figs. 3.1, 3.2, 3.10, 3.12, 5.6). Displaced but associated free cheeks of appropriate size occur in one case (Fig. 3.10), suggesting that this specimen may indeed be an exuvium. Most, but not all, of the axial shields represent immature individuals. Most notable among other associations is that illustrated in Figure 4.6, 4.7, in which two free cheeks and a hypostome/rostral plate are associated. Both the proportions and disposition suggest association of these sclerites during life, but the two free cheeks are orientated in the opposite direction to one another, with the right hand cheek-in-life inverted with respect to the left free cheek, as is the hypostomal/rostral plate, suggesting that twisting of the doublure was associated with the rupture of the rostral suture.

493

494

JOURNAL OF PALEONTOLOGY, V. 84, NO. 3, 2010

FIGURE 1—Location map of collecting site in Jingxi County, southwestern Guangxi Province, China.

This assemblage of G. guangxiensis apparently contains both carcasses and exuviae, but the majority of sclerite associations are hard to interpret definitively. Disarticulated sclerites are also present. The abundance of articulated material strongly suggests rapid burial, likely as the result of an obrution event (Speyer and Brett, 1985; Webster et al., 2007). THE RELATIONSHIP OF GUANGXIASPIS GUANGXIENSIS TO SHERGOLDIA LAEVIGATA AND THEIR PHYLOGENETIC AFFINITIES

A recent study of Shergoldia laevigata from the Guole locality has stressed the many similarities between that trilobite and the superfamily Asaphoidea, a derived group within the order Asaphida (Zhu et al., 2007). However, the cephalic venter of S. laevigata is unlike that of other asaphides because it possesses a triangular, posteriorly expanding rostellum-like plate, or an inverted rostral plate with Yshaped sutures (Fig. 2.1–2.2). As the form of the venter has great importance in assessing asaphide monophyly (Fortey and Chatterton, 1988), this observation is important. The cephalic venter of G. guangxiensis includes a triangular rostral plate, conterminant with the hypostome and narrowing toward the anterior, with the connective sutures converging at the anterior margin. This is similar to the condition in some specimens of S. laevigata Zhu et al., 2007 (Figs. 2.1–2.3). However, G. guangxiensis does not show the ventral median suture evident at the anterior of the triangular rostral plate in other specimens of Shergoldia laevigata. Nevertheless, as a triangular rostral plate of this form is known only in these two taxa, it is strong evidence of phylogenetic affinity between them. The newly found complete material of G. guangxiensis is also strikingly similar to S. laevigata in several other respects. The overall shape and proportions of the cranidium are closely similar, as are the wide librigenal doublure, the long frontal area, the form of the isolated glabellar furrows, the supramarginal dorsal suture, the prominent eye ridges, and the presence of both the bacculae and the eye socle. Furthermore, meraspids and early holaspids of S. laevigata have a pair of spines apparently associated with the opisthopleura of the first pygidial segment. Hence the particular similarities of the rostral structure, the supramarginal dorsal suture, the glabellar furrows, the bacculae, and the pygidial spines, in addition to the similarity in overall proportions, strongly suggest a close phylogenetic relationship between these taxa and also hint at an asaphide association.

FIGURE 2—Diagrams of ventral view of cephala of Shergoldia laevigata and Guangxiaspis guangxiensis. 1–2, early and late holaspid of Shergoldia laevigata respectively. Note variation of rostral plate shape. 3, Guangxiaspis guangxiensis.

On the other hand, G. guangxiensis is readily distinguished from S. laevigata in being less effaced and in retaining its pair of pygidial spines throughout holaspid ontogeny. The two species also differ in other details such as the shape of the frontal area, the position of the eyes, and the relative width of the pygidium. Guangxiaspis guangxiensis has ten rather than eight holaspid thoracic segments, and these show a degree of heteronomy not typical of derived asaphids. It is also subisopygous. Shergoldia Zhang and Jell, 1987 is assigned to Tsinaniidae Kobayashi, 1935, which has been transferred to Asaphoidea Burmeister, 1843 by Zhu et al. (2007). Guangxiaspis guangxiensis, on the other hand, is much less asaphidelike in overall appearance and more closely resembles the leiostegiinid family Kaolishaniidae Kobayashi, 1935. Details of the ventral structure of other members of these groups are unpublished. Zhu et al. (2007) suggested three possible explanations for the unusual rostral structure of S. laevigata. The first, which we rejected, was that it might be a taphonomic artifact, a possibility due to the variable nature of the median suture witnessed in the S. laevigata sample. Assuming that the similar structure of the rostral plate in G. guangxiensis reflects common ancestry of both taxa, the taphonomic explanation for the anteriorly narrowing rostral plate can be rejected with increased confidence. The second possibility was that the rostral plate in S. laevigata represents an evolutionary reversal from an asaphide ancestor with a complete median suture. We considered this improbable because the novel form of the rostral plate in S. laevigata seems unlikely to be an atavism. The condition of G. guangxiensis provides further evidence against this interpretation: if it were true, the several nonasaphide characters in G. guangxiensis listed above would require interpretation as apomorphies, which we consider to be unlikely when these characters are compared to those of leiostegiinid trilobites (see below). The last possibility, which we favored (Zhu et al., 2007, p. 246), was that the origin of the median suture in S. laevigata was independent from that in basal asaphides such as pterocephaliids. Based on the condition in G. guangxiensis, we concur with this view but revise our view of the taxonomic placement of S. laevigata. In 2007 we considered that the origins of S. laevigata lay among trilobites that are considered members of Asaphida (sensu Fortey, 1990) or within closely related libristomates.

ZHU ET AL.—VENTRAL STRUCTURES OF FURONGIAN TRILOBITES The several non-asaphide characters of G. guangxiensis broaden our consideration of relationships. Both G. guangxiensis and S. laevigata lack the pre-occipital tubercle, which is one of the most consistently preserved putative synapomorphies of the more derived asaphides (Fortey and Chatterton, 1988). As neither G. guangxiensis nor S. laevigata show an occipital tubercle, the absence of the pre-occipital tubercle could be due to effacement. However, other tsinaniid relatives of S. laevigata, such as Shergoldia necopina (Shergold, 1975), bear an occipital, as opposed to a pre-occipital, tubercle, indicating that this character was likely absent in Tsinaniidae. It has been suggested that tsinaniids are leiostegiinid trilobites (Shergold, 1975, 1991) thought to have been related only distantly to libristomates (Fortey, 1990, 1997). Leiostegiina is a sizable group common in the later Cambrian and early Ordovician that has yet to be defined in terms of apomorphic characters. It contains trilobites with a prominent rectangular glabella with weakly incised glabellar furrows, roughly 10 thoracic segments in the holaspid phase, and a relatively large holaspid pygidium. These features are common to G. guangxiensis and S. laevigata but also to many other non-leiostegiinid trilobites. Some leiostegiinid pygidia strongly resemble tsinaniid pygidia in their semi-circular form and long, thin axis that tapers markedly posteriorly and which bear multiple axial rings (e.g., Lloydia (Leiostegium) douglasi Harrington, 1937, see Harrington and Leanza, 1957, fig. 24.3a, c). Other leiostegiinids, such as Chosenia adamsensis Jell and Stait, 1985, shared these characters but also bore a pair of narrow pygidial spines likely associated with the second pygidial segment. This condition also resembles that seen in G. guangxiensis and in some tsinaniids. Recent work on the ontogeny of the tsinandiid trilobite Tsinania canens shows that its protaspis was not asaphoid (Park and Choi, 2009) but more similar to that of corynexochide trilobites, particularly in its upturned lateral margin. If correct, this supports phylogenetic placement of tsinaniids, including Shergoldia laevigata, within Leiostegiina. Tsinania canens is also inferred to have possessed a median suture that crossed most or all of the doublure (Park and Choi, 2009). Leiostegiinid overall body proportions are similar to those of some corynexochide trilobites from earlier in the Cambrian, and data from protaspid ontogeny (Lee and Chatterton, 2003) clearly indicates a phylogenetic link between Leiostegium and Illaenina, also considered to be a clade within Corynexochida (Fortey, 1997). The swellings adjacent to the L1 glabellar lobes, known as bacculae, in S. laevigata, G. guangxiensis, Mansuyites futiliformis, and Tingocephalus concavolimbatus (Endo, 1937) (see Lu et al., 1965, p. 423, pl. 81, fig. 9) merit comparison with the lunettes of illaenid trilobites because illaenids may have had a leiostegiinid-like common ancestor (Fortey, 1990, p. 564). Lunettes occur within the axial furrow (see Whittington, 1997, figs. 5.5, 7.7, 8.5), and their adaxial margins display thickening of the ventral surface of the exoskeleton, whereas the bacculae of S. laevigata and G. guangxiensis occurred on the fixed cheeks, and represent doming of the ventral exoskeleton. The extent to which these structural differences relate to position within or beyond the axial furrow is unclear, and the homology or convergence of these structures is unresolved. Both structures typically occur among trilobites with effaced glabellar furrows. The ornament of both genal and pygidial spines in G. guangxiensis is comparable to that in leiostegiinids such as Chosenia adamsensis Jell and Stait, 1985, and the glabella of G. guangxiensis is rectangular in form. Furthermore, as discussed below, G. guangxiensis is closely comparable in both cranidial

495

and pygidial form to the kaolishaniid trilobite Mansuyites futiliformis Shergold, 1972, which also bears bacculae, a similarly shaped glabella with comparable isolated glabellar furrows, and a pygidium of the same proportions. Kaolishaniid trilobites have long been considered to be leiostegioideans (e.g., Shergold, 1972). The close relationship inferred between G. guangxiensis and S. laevigata is thus consistent with the idea that leiostegiinids including kaolishaniids and tsinaniid trilobites share close common ancestry (Fortey, 1990), a view recently championed by Park and Choi (2009) on the basis of the non-asaphoid protaspis and the inferred median suture in T. canens. If the ancestors of G. guangxiensis and tsinaniids lie outside Asaphida, then the extent of convergence shown by S. laevigata with the characters of some pre-occipital tubercle bearing asaphides was prodigious. SYSTEMATIC PALEONTOLOGY

Morphological terminology follows that of Whittington and Kelly (1997). Specimens listed here are housed in Nanjing Institute of Geology and Palaeontology, Chinese Academy of Sciences. Order CORYNEXOCHIDA Kobayashi, 1935 Suborder LEIOSTEGIINA Bradley, 1925 Superfamily LEIOSTEGIOIDEA Bradley, 1925 Family KAOLISHANIIDAE Kobayashi, 1935 Discussion.—Kaolishaniidae has been transferred to the superfamily Leiostegioidea Bradley, 1925 within the order Corynexochida Kobayashi, 1935 by Fortey (1997). Shergold (1972) divided the family into three subfamilies: Kaolishaniinae Kobayashi, 1935, Mansuyiinae Hupe´, 1955 sensu Shergold, 1972, and Tingocephalinae Hupe´, 1955, according to the structure of the preglabellar area, characteristics of the glabella and the nature of the pygidial shape, axis and spines. Both Mansuyiinae and Tingocephalinae are distinguished from Kaolishaniinae by having a long preglabellar field (sag.) and anterior branches of the facial sutures which meet sagittally. The preglabellar field (sag.) of Tingocephalinae is longer than that of Mansuyiinae, which differentiates one from the other. Shergold (1972) assigned Mansuyites to Kaolishaniidae Kobayashi, 1935, and his assignment was accepted by Jell (in Jell and Adrain, 2003). Both Zhou (in Zhou et al., 1977) and Qian (in Qiu et al., 1983) considered the familial assignment of Guangxiaspis to be uncertain, but Jell (in Jell and Adrain, 2003) ascribed it to Pterocephaliidae Kobayashi, 1935 based on Zhou’s original generic concept. The cranidial characteristics of Guangxiaspis and Pterocephalia are grossly similar but differ in many details such as the supramarginal suture, the presence of bacculae, and the isolated glabellar furrows. Their pygidia are also easily distinguished by the prominent spines present in Guangxiaspis. Hence we reject assignment of Guangxiaspis to Pterocephaliidae. Subfamily MANSUYIINAE Hupe´, 1955 sensu Shergold, 1972 Genera with relatively short preglabellar fields that have been ascribed to Mansuyiinae by Shergold (1972, 1975) include: Mansuyia Sun, 1924, Kaolishaniella Sun, 1935, Paramansuyella Endo, 1937, Mansuyites Shergold, 1972, and Hapsidocare Shergold, 1975. The anterior branches of the facial suture of Mansuyites futiliformis Shergold, 1972, the holotype of Mansuyites, meet at the sagittal line (Shergold, 1972, p. 57), and this species may have had a similar ventral structure to that of Guangxiaspis guangxiensis. Some other mansuyiine genera, such as Mansuyia and Hapsidocare, also

496


FIGURE 3—Guangxiaspis guangxiensis Zhou, 1977, Sandu Formation, Guole, Jingxi, Guangxi. 1–2, NIGPAS148317, latex cast and internal mold of incomplete meraspid degree 2 exoskeleton respectively. Note the division between thorax and transitory pygidium indicated by white arrow. X13.5; 3–4, NIGPAS148318: 3, latex cast of external mold of incomplete meraspid degree 6? exoskeleton, X7.2; 4, enlargement of the pygidium of 3, X10.8; 5, NIGPAS148319, complete internal mold of meraspid degree 5? exoskeleton. Note the division between thorax and transitory pygidium indicated by white arrow. X4.5; 6–7, NIGPAS148320, internal mold and latex cast of complete meraspid degree 3 exoskeleton. Note the division between thorax and transitory pygidium indicated by white arrow. X10.5; 8, NIGPAS148321, internal mold of incomplete meraspid degree 3 exoskeleton. Note the division between thorax and transitory pygidium indicated by white arrow. X16; 9, NIGPAS148322, internal mold of incomplete meraspid degree 5? exoskeleton. Note the division between thorax and transitory pygidium indicated by white arrow. X8; 10, NIGPAS148323, internal mold of incomplete meraspid degree 4 exoskeleton. Note the division between thorax and transitory pygidium indicated by white arrow. X10.5; 11, NIGPAS148324, internal mold of nearly complete meraspid degree 6 exoskeleton. Note the division between thorax and transitory pygidium and the position of pygidial pleural spine indicated by white arrows. X2; 12, NIGPAS148325, internal mold of incomplete meraspid degree 6 exoskeleton. Note the division between thorax and transitory pygidium indicated by white arrow. X14.

ZHU ET AL.—VENTRAL STRUCTURES OF FURONGIAN TRILOBITES

497

FIGURE 4—Guangxiaspis guangxiensis Zhou, 1977, Sandu Formation, Guole, Jingxi, Guangxi. 1, NIGPAS148326, latex cast of complete early holaspid exoskeleton. Note the termination of exoskeleton indicated by white arrow. X6.5; 2–5, NIGPAS148327: 2, internal mold of complete early holaspid exoskeleton, X4.3; 3, latex cast of complete early holaspid exoskeleton, X4.3; 4, right anterior lateral view of 3, X4.3; 5, enlargement of anterior area of 2, X8.6; 6–7, NIGPAS148328; 6, latex of a pair of librigenae, rostral plate and hypostome, X5.5; 7, enlargement of rostral plate and hypostome of 6, X11; 8, NIGPAS 148329, latex cast of nearly complete early holaspid exoskeleton, X3.0; 9, NIGPAS148330, internal mold of incomplete later holaspid exoskeleton, X3.0; 10, NIGPAS148331, internal mold of incomplete later holaspid exoskeleton, X2.0; 11, NIGPAS148332, internal mold of incomplete later holaspid exoskeleton, X3.5; 12, NIGPAS148333, internal mold of complete later holaspid exoskeleton, X3.0.

498


FIGURE 5—Guangxiaspis guangxiensis Zhou, 1977, Sandu Formation, Guole, Jingxi, Guangxi. 1–2, NIGPAS148334, latex cast and internal mold of complete early holaspid exoskeleton, X3.2; 3, NIGPAS148335, internal mold of complete later holaspid exoskeleton, X2.1; 4, NIGPAS148336, latex cast

ZHU ET AL.—VENTRAL STRUCTURES OF FURONGIAN TRILOBITES have this type of facial suture (Shergold, 1975). Guangxiaspis Zhou in Zhou et al., 1977, with another two similar latest Cambrian genera, Anhuiaspis Qian and Qiu in Qiu et al., 1983, and Shidiania Luo, 1983, are added to Mansuyiinae herein. Genus GUANGXIASPIS Zhou, 1977 Guangxiaspis ZHOU in Zhou et al., 1977, p. 155; QIAN in Qiu et al., 1983, p. 101. Type species.—Guangxiaspis guangxiensis Zhou in Zhou et al., 1977 (p. 155, pl. 47, figs. 10–11), late Furongian, Hewen, Jingxi County, Guangxi Province, China. Other species.—Guangxiaspis midonodus Qian in Qiu et al., 1983, p. 101, pl. 34, fig. 9, late Furongian, Tangcun, Jingxian, Anhui, China. Emended diagnosis.—Glabella tapering forward slightly and truncated anteriorly; three pairs of isolated lateral glabellar furrows; palpebral lobe short, situated medially; bacculae small; librigenae separated dorsally by median suture; rostral plate triangular in outline, conterminant with the hypostome. Ten thoracic segments in maturity, axis slightly narrower than pleural region; outer part of pleura flexed down, facetted anterolaterally. Pygidium semicircular, with lateral marginal spine derived from opisthopleural region of the first segment. Discussion.—Qian (in Qiu et al., 1983) recognized the similarity of the cranidia of Guangxiaspis and Mansuyites. The former is distinguished from the latter by smaller palpebral lobes that lack a firmly incised palpebral furrow, the blade-like rather than band-like posterolateral limbs, and by bearing relatively wider fixigenae. Anhuiaspis is very similar to Guangxiaspis but differs from the former by having a more anteriorly situated palpebral lobe, and by less effaced glabellar furrows. Anhuiaspis may prove to be a junior synonym of Guangxiaspis, but until more material of Anhuiaspis is available, it remains a valid genus. Shidiania is differentiated from Guangxiaspis by having less divergent anterior branch of facial sutures and less effaced exoskeleton. In addition, Shidiania bears a median pygidial spine that distinguishes it from Guangxiaspis. The more anteriorly located palpebral lobe and less curved anterior cranidial margin differentiate Kaolishaniella from Guangxiaspis. Paramansuyella is readily distinguished from Guangxiaspis by the short (sag.) preglabellar. Hapsidocare differs by having a more anteriorly located palpebral lobe and broadly based pygidial spines. Guangxiaspis resembles Mansuyia, but differs from the latter by a smaller palpebral lobe and expanded bulb-like terminal segment of the pygidial axis. GUANGXIASPIS GUANGXIENSIS Zhou, 1977 Figures 3–5 Guangxiaspis guangxiensis ZHOU in Zhou et al., 1977, p. 155,pl. 47, figs. 10–11. Mansuyia laevigata ZHOU in Zhou et al., 1977, p. 203, pl. 60, fig.3. Mansuyia pulchra ZHOU in Zhou et al., 1977, p. 203, pl. 60, fig. 7. New Material.—Twenty articulated specimens, 20 isolated cranidia, and 20 isolated pygidia. Occurrence.—Guole Township, Jingxi County, southwestern Guangxi, South China, latest Furongian.

499

Diagnosis.—Palpebral lobes small, situated opposite midlength of the glabella; fixigena nearly as wide as glabella. Mature pygidial axis composed of at least six segments with bulbous terminal piece, with pleural furrows effaced on all except anteriormost segment and all interpleural furrows effaced. Description of meraspid period, degree 2 (Fig. 3.1–3.2).— about 2.0 mm in axial length. Glabella occupies most of cranidial length, rectangular in outline and sharply truncated anteriorly, occipital furrow entire and firmly incised. Axial furrow wide and deep, frontal area depressed. Glabellar furrows, bacculae and eye ridges apparently absent. Anterior branch of facial suture slightly convergent, posterior branch of facial suture short and running diagonally. Preglabellar field short (sag.) and concave, occupying about 0.17 of cranidial length; anterior margin narrow (sag.) and straight. Posterior border furrow wide (exsag.) and deeply incised, posterior border widening (exsag.) abxially. Thoracic segments transverse. Pygidium semicircular in outline, with slight emargination postaxially and pleural platform pronounced in posterior region; axis tapering backward gradually; pleural field divided by about six pairs of pleural and interpleural furrows; border narrow, separated from pleural field by narrow border furrow. Degree 3 (Fig. 3.6–3.8).—Exoskeletal axial length about 2.7 mm to 3.0 mm (excluding genal spine). Exoskeleton nearly rectangular in outline, except for long and robust genal spine that originated advanced from the glabellar posterior margin and extends about 1.5 times length of exoskeleton (exsag.). Form as in degree 2 except that preglabellar field is only weakly concave; anterior margin slightly arched forward; posterior border furrow wide and deep. Pygidial border discernible in external mold but obsolete in internal mold. Degree 4 (Fig. 3.10).—About 4.0 mm in axial length. Glabella slightly tapering forward and sharply truncated anteriorly; anterior branch of facial sutures slightly convergent; preglabellar field short (sag.), occupying about 0.23 of cranidial length, merging with fixigenae laterally; anterior border arched forward and widening adaxially to form slightly pointed sagittal projection; posterior border furrow wide and deep. Pygidium semicircular; axis tapering backward gradually; pleural field divided by at least five pairs of pleural furrows and interpleural furrows. Degree 5? (Fig. 3.5, 3.9).—Exoskeletal axial length ranges from 3.5 mm to 5.0 mm. Glabella slightly tapering forward and sharply truncated anteriorly; occipital furrow short (sag.) and deep; axial furrow deep, preglabellar furrow shallow; eye ridge very weak; anterior branch of facial sutures slightly convergent; anterior margin arched forward; posterior border furrow wide and deep, posterior border widening adaxially. Genal spine extending backward and outward about 1.6 times exoskeletal length. Pygidium semicircular; axis tapering backward gradually; six pairs of interpleural furrows deeply incised. Degree 6 (Fig. 3.3?–3.4, 3.11, 3.12).—Exoskeletal axial length ranges from 4.5 mm to 6.5 mm. Cranidium as in degree 5 except occipital ring long (sag.) and narrowing abxially; palpebral lobe prominent and crescentic, palpebral furrow deep; anterior branch of facial suture slightly divergent; anterior border medially long (sag.) and gradually

r of nearly complete later holaspid exoskeleton, X2.3; 5, NIGPAS148337, latex cast of librigena, X3; 6, NIGPAS148338, latex cast of incomplete later holaspid exoskeleton, X3.3; 7–8, NIGPAS148339, latex cast and internal mold of incomplete later holaspid exoskeleton, X2.7; 9, NIGPAS148340, latex cast of incomplete later holaspid exoskeleton, X3; 10, NIGPAS148341, internal mold of incomplete later holaspid exoskeleton, X2.

500


narrowing abaxially, separated from preglabellar field by shallow border furrow; posterior border furrow wide and deep; genal spine about 1.6 times exoskeletal length. Pygidium semicircular in outline; axis relatively narrow, tapering backward gradually; pleural field divided by five pairs of pleural furrows; interpleural furrows absent; border narrow; pleural spine slender, extending outward and backward. Description of holaspid period. Early holaspid (Fig. 4.1–4.5, 4.8; Fig. 5.1–5.2).—Exoskeletal axial length ranges from 5.2 mm to approximately 12 mm. Exoskeleton, excluding genal spines, an elongated oval in outline. Glabella gently tapering forward with three pairs of lateral glabellar furrows, all isolated from axial furrow and becoming shorter (exsag.) forward; S1 long, directed inward and backward; S2 and S3 pit-like and nearly transverse. Occipital furrow shallow; occipital ring of uniform length (sag.). Bacculae faint, situated opposite L1. Eye ridge stronger in internal mold than in external mold. Palpebral lobe moderately large, situated opposite the midlength of glabella (including occipital ring). Anterior branch of facial suture divergent forward. Anterior area triangular. Genal spine robust, curved posteriorly and abaxially to position opposite posterior of pygidium, then posteriorly and slightly inward for a distance about half the axial length of the exoskeleton, tip slightly anterior of glabellar posterior margin. Base of genal spine opposite to or posterior of midlength of occipital ring. Thorax with ten segments. Anterior segments transverse with sharp, transversely directed pleural spines, pleurae of posterior segments posteriorly curved, with tips posteriorly directed. Pygidial axis consist of seven rings with a terminal piece; rachial furrows deep in internal mold, and faint in external mold; pleural spine strong and extending outward and backward. Later holaspid (Fig. 4.9–4.12; Fig. 5.3–5.10).—Exoskeletal axial length ranges from about 15.5 mm to approximately 40 mm. Exoskeletal outline longitudinally elliptical. Glabella subrectangular, tapering slightly forwards and truncated anteriorly, with moderate convexity, occupying 55–65% of the total cranidial length (sag.). Three pairs of lateral glabellar furrows faint, pit-like and isolated from axial furrow. Occipital furrow shallow, gradually becoming obsolete distally, occipital ring of uniform length (sag). Bacculae small and slightly elongate (sag.), separated from glabella by axial furrow, with posterior end opposite the occipital furrow. Axial furrows deeply incised, preglabellar furrow shallower adaxially. Eye ridge, running backward and outward at angle of about 65u to sagittal line from the anterior lateral corner of glabella. Paradoublural line nearly parallel with eye ridge. Palpebral lobe crescentic, occupying one third of glabellar length (including occipital ring), opposite the midlength of glabella (including occipital ring), clearly defined by shallow palpebral furrow. Base of genal spine opposite to anterior of SO. Palpebral fixigenae gently convex, as wide as glabella. Anterior branch of the facial sutures divergent forward, retracted from anterior of exoskeleton, and connected sagittally on dorsal surface by median suture; posterior branch of the facial sutures running outward and backward, and enclosing a blade-like posterolateral limb. The preglabellar field concave and long (sag.), curving upward into a gently convex, poorly defined anterior border. Posterior border furrow long (exsag.) but shallow, posterior border narrower than occipital ring and gently narrowing to the point opposite the posterior corner of palpebral lobe, then gradually widening distally and extending backward. Librigena narrow, doublure relatively long, wide (exsag., tr.), lateral border wide, merging with posterior border to form a robust genal spine.

Genal spine approximately as long as the exoskeleton, and with external surface ornamented with terrace lines forming anteriorly directed chevrons, ornament particularly marked on outer rim. Rostral plate triangular in outline, widening posteriorly. Quadrate hypostome conterminant, anterior lobe of middle body circular, with median swelling, medial furrow wide and deep, posterior lobe of median body wide and crescentic; anterior and lateral border wide, lateral border with terrace ridge ornament, posterior border short (sag.). Thorax of 10 segments. Thoracic segments having axial rings similar in form to occipital ring. Pleural furrow diagonally bisects pleurae. Pleurae wide (tr.), inner portion transverse, anterior margin of part abaxial to fulcrum broadly and steeply faceted, curving backward with a marked inflection at base of pleural spine. Pleural spines short (exsag.), long (tr.) with posterior margin curved anteriorly at spine base. Spines transverse anteriorly, gradually more posteriorly directed toward rear. Pygidium approximately triangular in outline. The axis tapering backward to the posterior border, composed of seven segments and bulbous terminal piece. Pleural furrows and interpleural furrows effaced except for the first pleural furrow. Robust spine pair derived from pleural field of the opisthopleurae of the first pygidial segment, about as long as pygidial axis, running posteriorly and inward, straight or slightly curved, surface covered with terrace lines. Discussion.—Our complete material resembles the holotype of G. guangxiensis except in bearing a pair of lateral pygidial spines that were not preserved in the holotype, which is an internal mold. Zhou (in Zhou et al., 1977) also described two pygidia which bear lateral spines, one of which bears seven thoracic segments. Both were collected from the same locality as the holotype of G. guangxiensis, and each was ascribed to new species of Mansuyia Sun, 1924, M. laevigata and M. pulchra respectively. These two pygidia are closely similar to our material, and their differences are accommodated within the range of intraspecific variation observed in our new material. Because G. guangxiensis has page priority to both M. laevigata and M. pulchra, the latter two species are here considered to be junior synonyms of G. guangxiensis. In addition to the type species, Qian (in Qiu et al., 1983) described a new species, G. midonodus from Tangcun in Anhui Province. The position of the palpebral lobes and triangular posterolateral limbs distinguish G. guangxiensis from G. midonodus. In G. guangxiensis these are situated opposite the midlength of the glabella, a little further forward than those in G. midonodus. The ontogeny of G. guangxiensis is notable for the transition in body proportions that took place during the holaspid phase. The transition between meraspid degrees 3 and 4 marks the appearance of several characteristics that become distinctive of mature cranidial morphology of G. guangxiensis: in degree 4 the length of the frontal area expanded and the distinctive pointed appearance of the anterior border became evident. Meraspid degree 6 (Fig. 3.11–3.12) may mark the beginning of epimorphic phase, because at this degree the number of trunk segments reaches the maximum observed: the development of G. guangxiensis is thus protomeric (Hughes et al., 2006). Although each successive meraspid degree was not recovered, a number of ontogenetic trends are apparent in the meraspidholaspid ontogeny of G. guangxiensis: the glabella changes from rectangular to trapezoidal shape; the ratio of the anterior area to glabellar length increases; anterior branch of facial suture transits from convergent to divergent; the dorsal furrow


501

TABLE 1—Linear measurements for the nine specimens included in the principal component analysis. Occipital-glabellar length (mm) 5.0 3.4 1.7 4.5 4.4 5.5 3.5 8.5 3.2

Frontal area length (mm) 3.0 2.0 1.2 2.4 2.5 3.5 2.5 5.5 2.3

Posterior occipital Palpebral lobe length Pygidial length maximum width (mm) (mm) (mm) 4.2 2.0 1.2 3.2 3.6 3.5 3.4 6.5 2.9

becomes faint; bacculae emerge; and the pleural spines deflect from pointing outward to more inward. In late meraspid stages the pygidium bears a general resemblance to that of many raphiophorids (except for the pygidial spines), but in later holaspid ontogeny is it more like that of a ceratopygid. This transition reflects the protomeric pattern of development, for the late meraspid pygidium is more segment rich than the holaspid pygidium. The similarity to raphiophorid form in the meraspid and even early holaspid stages extends to the overall body shape, and is particularly notable in the extreme relative length of the gracile genal spines in the younger forms, in which the genal spines extend about one axial body length beyond the posterior of the pygidium. This condition is different in some of the larger holaspids, in which the genal spine terminates at about one half of the axial body length beyond the posterior of the pygidium (Fig. 5.7,5.8). On the other hand, other mature specimens maintain long genal (and pygidial) spines (Figs. 4.12, 5.1–5.2). The overall morphology of the G. guangxiensis suggests a benthic form living in an illuminated environment, and the conterminant hypostomal condition may permit processing of relatively bulky food items (Fortey and Owens, 1999). The marked changes during meraspid and holaspid ontogeny likely indicate ontogenetic changes in habits, but what these might be is unclear.

FIGURE 6—Bivariate relationships between occipital-glabellar length of Guangxiaspis guangxiensis and three other linear dimensions of the cranidium, showing positive allometry of the frontal area and glabellar width, and slight negative allometry of the palpebral lobe as the occipitalglabellar length increased, n 5 27.

1.6 1.2 0.9 1.4 1.2 2.0 1.1 2.5 1.4

5.6 3.4 1.5 6.5 6.2 7.0 4.5 6.0 4.0

Maximum pygidial width (mm)

Pygidial spine length (mm)

11.0 5.8 4.0 9.0 11.0 12.0 7.5 19.0 7.5

4.0 5.5 1.0 3.0 5.2 6.0 3.5 15.0 3.0

Although the transition in morphology is profound, there is little doubt that all these specimens are associated taxonomically: particularly telling are the presence of the bacculae and macropleural pygidial spines in both meraspid and holaspid stages. However, it is possible that the collection contains more than a single morphotype, and we have thus employed morphometrics to investigate the growth relationships and morphological variation among specimens. Given the limited sample size and preservation in shale, which is accompanied by compaction-related deformation, morphometric investigation of the growth and variation of G. guangxiensis was restricted to a series of linear measurements (Table 1). Bivariate plots of relationships among some of these variables (Figs. 6, 7) reveal no clear indication of more than a single trend line for each variable pair. After logarithmic transformation of the original measured distances variable pairs were investigated by calculating

FIGURE 7—Bivariate relationships between axial pygidial length of Guangxiaspis guangxiensis and two other linear dimensions of the pygidium, the width and the length of the pygidial spine, n 5 27 for length/width comparison, n 5 14 for length/pygidial spine length comparison. There is no marked allometry in the pygidial length/ width relationship.

502


TABLE 2—Reduced Major Axes for relationships among linear variables indicating significant allometry in the shape of the glabella, and in the lengths of the palpebral lobe, frontal area and genal spines during meraspid and holaspid growth of Guangxiaspis guangxiensis. Each analysis based on log transformation of original linear measurements.

TABLE 2—Continued. 95% confidence intervals: linear 0.7607, 0.9176 bootstrap 0.7883, 0.9030

0.5399, 0.8523 0.5731, 0.8180

0.841, 0.984

1. Glabellar length vs Frontal Area length, 27 cases

99% confidence intervals: linear 0.7230, 0.9553 bootstrap 0.7471, 0.9251

0.4649, 0.9273 0.5352, 0.8555

0.593, 0.998

RMA RMA

Intercept

Slope

R‘2 5. Pygidial length vs Pygidial width, 22 cases

Linear model: estimate st.error

0.3949 0.0175

0.5853 0.0352

0.910

Jackknife: estimate st.error

0.3982 0.02165

0.5785 0.04618

0.906 0.0287


0.5128, 0.6577 0.5186, 0.6991


0.4872, 0.6833 0.4985, 0.7692

Slope

0.939

0.858, 0.959


20.3962 0.05161

1.123 0.07269

0.936 0.0292

0.803, 0.966

95% confidence intervals: linear 20.5064,20.2914 0.998, 1.258 bootstrap 20.5100,20.3103 0.998, 1.277

0.889, 0.981

99% confidence intervals: linear 20.5456,20.2523 0.951, 1.305 bootstrap 20.5580,20.2679 0.955, 1.371

0.863, 0.991

R‘2

0.2058 0.0175

0.8586 0.0387

0.949


0.2060 0.02068

0.8579 0.03807

0.949 0.0197


0.7788, 0.9383 0.7908, 0.9383

0.911, 0.979


0.7507, 0.9665 0.7725, 0.9684

0.874, 0.986

3. Glabellar length vs Palpebral lobe length, 27 cases RMA Intercept

Slope

R‘2


0.4246 0.0208

1.218 0.089

0.867


0.4249 0.02369

1.213 0.09206

0.865 0.0524


1.035, 1.400 1.065, 1.415

0.775, 0.937


0.971, 1.465 1.027, 1.463

0.743, 0.942

4. Genal spine length vs Glabellar length, 9 cases RMA RMA

Intercept

Slope

R‘2

1.128 0.062


RMA

Slope

20.3989 0.0515

RMA Intercept

Intercept


2. Glabellar length vs Glabellar width, 27 cases

RMA

RMA RMA

R‘2


0.8391 0.0332

0.6961 0.0661

0.937


0.8350 0.03578

0.7016 0.07084

0.940 0.0379

reduced major axis using the RMA1.17 program of Andrew Bohonak of San Diego State University. Bootstrap resampling, completed 400 times in each case, permitted the calculation of confidence intervals for both the slope and intercepts of each reduced major axis. The results (Table 2) show that both the length of the frontal area and the width of the glabella show significant positive allometry with respect to the length of the glabella, and the length of the palpebral lobe shows slight negative allometry relative to the glabellar length that is significant only at the 95% confidence level. The positive allometry of the frontal area is particularly strong (Figs. 4, 5, 6), but there is no indication that any of these growth allometries differed markedly from linearity during growth. Accordingly, size related shape change appears to have been progressive during ontogeny, at least at the level of resolution afforded by this analysis. Analysis of length and width of the pygidium (Fig. 7) indicates that its growth does not differ significantly from isometry (Table 2), but the length of the pygidial spine shows notable variance. The degree of scatter about the spine growth trend is large, some specimens bearing anomalously long spines (e.g., Fig. 4.10; Fig. 5.1–5.2) while the spines of others are unusually short (e.g., Fig. 4.12; Fig. 5.7–5.8). This distribution hints that sample may contain more than a single morph, one of which is characterized be relatively short pygidial spines, while another has long spines. Both long- and short-spined forms occur at a range of sizes, so this does not appear to be a characteristic that distinguishes different ontogenetic stages. To consider this further a principal component analysis (PCA) was performed using the Systat 8.0 package and was based on the correlation matrix that synoptically considered logged data for the 7 linear measures for the 9 specimens that preserved all these attributes (Table 1). Although the sample size was too small to permit firm conclusions, the eigenvectors of the first principal component, which accounts for some 89% of total variance, are all positive and almost equal in value (Table 3). This


503

TABLE 3—Principal component analysis based on logged data for 7 linear measures (see Table 1) of Guangxiaspis guangxiensis, n 5 9. Latent Roots (Eigenvalues) PC1 6.261 PC6 0.14 Percent of Total Variance Explained (for those axes .1% of total variance) PC1 89.436

PC2 0.363 PC7 0.006

PC3 0.197 . .

PC4 0.126 . .

PC2 5.180

PC3 2.820

PC4 1.805

PC5 0.033 . . . .

Latent Vectors (Eigenvectors) Occ.-glabellar length Frontal area length Occ.-glabellar width Palpebral lobe length Pygidial length Pygidial width Pygidial spine length

1 0.397 0.394 0.383 0.365 0.349 0.392 0.363

2 0.000 0.183 20.233 0.489 20.745 20.060 0.338

3 20.056 0.073 0.086 0.530 20.045 0.199 20.813

4 0.123 20.198 20.657 0.436 0.508 20.236 0.102

5 20.234 0.388 0.352 0.166 0.156 20.787 20.005

Occ.-glabellar length Frontal area length Occ.-glabellar width Palpebral lobe length Pygidial length Pygidial width Pygidial spine length

6 20.866 0.168 20.096 0.093 0.195 0.354 0.201

7 0.141 0.767 20.476 20.347 0.012 0.053 20.205

. . . . . . . .

. . . . . . . .

. . . . . . . .

suggests that the first component accommodates much of the size-related variation in the sample, as is typical for PCA analyses of single trilobite species (see Hughes, 1994). Eigen vectors for the subsequent principal components indicate that, additional variation related to pygidial spine length is an important contributor to total variance, but that this character does not stand out markedly compared to other attributes (Table 3), nor did scores on these components suggest distinct long and short spine groups. For these reasons, and pending a larger sample, we consider it prudent to group all specimens together until more is known. ACKNOWLEDGMENTS

We are grateful to Wei Renyan from Guangxi Institute of Geology and Chen Tingen from Nanjing Institute of Geology and Palaeontology, Chinese Academy of Sciences, for their kind help in the field. We are also indebted to Zhou Zhiyi and Yuan Jinliang for their much helpful discussion. Duck Choi and Steve Westrop provided astute reviews, and Brian Pratt and Tae-Yoon Park provided additional helpful discussion. The present study is supported by the Chinese Academy of Sciences (KZCX2-YW-122), National Natural Science Foundation of China (40602002, J0630967). N.C.H.’s contribution is made possible through the support of NSF grants EAR053868 and EAR-0616574. REFERENCES

BRADLEY, J. H. 1925. Trilobites of Beekmantown in the Phillipsburg region of Quebec. Canadian Field Naturalist, 39:5–9. BURMEISTER, H. 1843. Die Organisation der Trilobiten aus ihren lebenden Verwandten entwickelt; nebst einer systematischen Uebersicht aller zeither beschriebenen Arten. Reimer, Berlin, 147p. ENDO, R. 1937. Addenda to Part 1 and 2. Manchurian Science Museum Bulletin, 1:302–369, 435–461. FORTEY, R. A. 1990. Ontogeny, hypostome attachment and trilobite classification. Palaeontology, 33:529–576. FORTEY, R. A. 1997. Classification, p. 289–302. In R. L. Kaesler (ed.), Treatise on Invertebrate Paleontology, Part O, Trilobita (Revised). Geological Society of America, Boulder and University of Kansas Press, Lawrence. FORTEY, R. A AND B. D. E. CHATTERTON. 1988. Classification of the trilobite suborder Asaphina. Palaeontology, 31:165–222. FORTEY, R. A AND R. M. OWENS. 1999. Feeding habits in trilobites. Palaeontology, 42:429–465.

HAN NAIREN AND C. GUIYING. 2004. Discovery of a complete Dictyella (Trilobite) carapace fossil in Guangxi and its significance. Acta Palaeontologica Sinica, 43:416–419. HAN NAIREN AND C. GUIYING. 2008. New stylophorans (Echinodermata) from the upper Cambrian of Guangxi, South China. Science in China Series D: Earth Sciences, 51:181–186. HAN, N., LAN TANG, RENSHAN WEI, AND GUIBIN WANG. 2000. Stratigraphy of Upper Cambrian from Guole, Jingxi, Guangxi. Journal of Guilin Institute of technology, 20:350–354. HARRINGTON, H. J. 1957. On some Ordovician fossils from northern Argentina. Geological Magazine, 74:97–124. HARRINGTON, H. J. AND A. F. LEANZA. 1957. Ordovician trilobites of Argentina. Special Publications of the Department of Geology, University of Kansas, 1:1–276. HENNINGSMOEN, G. 1975. Moulting in trilobites. Fossils and Strata, 4:179–200. HUGHES, N. C. 1994. Ontogeny, intraspecific variation, and systematics of the Late Cambrian trilobite Dikelocephalus. Smithsonian Contributions to Paleobiology, 79:1–89. HUGHES, N. C., A. MINELLI, AND G. FUSCO. 2006. The ontogeny of trilobite segmentation: a comparative approach. Paleobiology, 32:602– 627. HUPE´, P. 1955. Classification des trilobites. Annales de Pale´ontologie, 41:91–325. JELL, P. A. AND J. M. ADRAIN. 2003. Available generic names for trilobites. Memoirs of the Queensland Museum, 48:1–552. JELL, P. A. AND B. STAIT. 1985. Tremadoc trilobites from the Florentine Valley Formation, Tim Shea area, Tasmania. Memoirs of the Museum of Victoria, 46:1–34. KOBAYASHI, T. 1935. The Cambro-Ordovician Formations and Faunas of South Chosen. Journal of the Faculty of Science University of Tokyo, 4:49–344. LEE, D.-C. AND B. D. E. CHATTERTON. 2003. Protaspides of Leiostegium and their implications for membership of the order Corynexochida. Palaeontology, 46:431–445. LU YANHAO, WENTANG ZHANG, ZHAOLING ZHU, YIYUAN QIAN, AND LIWEN XIANG. 1965. Trilobites of China. Science Press, Beijing, 766 p. (in Chinese) LUO HUILIN. 1983. New finds of trilobites from Late Cambrian in western Yunnan, p. 1–18. In Contribution to the geology of the Qinghai-Xizang (Tibet) Plateau. Science Press, Beijing. ¨ PIK, A. A. 1970. Redlichia of the Ordian (Cambrian) of Northern O Australia and New South Wales. Bureau of Mineral Resources Geology & Geophysics Australia Bulletin, 11:41–66. PARK, T.-Y. AND D. K. CHOI. 2009. Post-embryonic development of the Furongian (late Cambrian) trilobite Tsinania canensis: implications for life mode and phylogeny. Evolution and Development, 11:441–455. PENG S., L. E. BABCOCK, AND A. COOPER. In press. Chapter 18, Cambrian Period. In F. Gradstein, J. Ogg, and A. Smith (ed.), A Geologic Time Scale 2010. Cambridge University Press, Cambridge.

504


QIU HONGAN, YANHAO LU, ZHAOLING ZHU, DECHANG BI, TIANRUI LIN, ZHIYI ZHOU, QUANZHONG ZHANG, YIYUAN QIAN, TIANYIN JU, NAIREN HAN, AND XIUZHE WEI. 1983. Trilobita, p. 28–254. In Nanjing Institute of Geology and Mineral Resources (ed.), Palaeontological Atlas of East China, (1). Geological Publishing House, Beijing. (In Chinese) SALTER, J. W. 1864. A monograph of British trilobites. Part 1. Palaeontographical Society Monograph, 1–80. SHERGOLD, J. H. 1972. Late Upper Cambrian trilobites from the Gola Beds, western Queensland. Bureau of Mineral Resources of Australia, Bulletin, 112:1–127. SHERGOLD, J. H. 1975. Late Cambrian and Early Ordovician trilobites from the Burke River structural belt, western Queensland, Australia. Commonwealth of Australia, Bureau of Mineral Resources, Geology and Geophysics, Bulletin, 153:1–251. SHERGOLD, J. H. 1991. Late Cambrian and Early Ordovician trilobite faunas and biostratigraphy of the Pacoota Sandstone, Amadeus Basin, Central Australia. Bureau of Mineral Resources, Geology and Geophysics Bulletin, 237:15–76. SPEYER, S. E. AND C. E. BRETT. 1985. Clustered trilobite assemblages in the Middle Devonian Hamilton Group. Lethaia, 18:85–103. SUN, YUNZHU. 1924. Contributions to the Cambrian faunas of North China. Palaeontologia Sinica (series B), 1:1–109. SUN, YUNZHU. 1935. The Upper Cambrian trilobite faunas of north China. Palaeontologia Sinica, series B, 7(2):1–93. WEBSTER, M. 2007. Ontogeny and evolution of the Early Cambrian trilobite genus Nephrolenellus (Olenelloidea). Journal of Paleontology, 81:1168–1193.

WEBSTER, M., N. C. HUGHES, AND R. R. GAINES. 2007. Microstratigraphy, trilobite biostratinomy, and depositional environment of the ‘‘lower Cambrian’’ Ruin Wash Lagersta¨tte, Pioche Formation, Nevada. Palaeogeography, Palaeoclimatology, Palaeoecology, 264:100–122. WHITTINGTON, H. B. 1997. Illaenidae (Trilobita): Morphology of the thorax, classification, and mode of life. Journal of Paleontology, 71:878–896. WHITTINGTON, H. B. AND S. R. A. KELLY, 1997. Morphological terms applied to Trilobita, p. O313–329. In R. L. Kaesler (ed.), Treatise on Invertebrate Paleontology, Part O, Arthropoda 1, Trilobita, Revised, Volume 1. Geological Society of America, Boulder and University of Kansas Press, Lawrence. ZHANG WENTANG AND P. A. JELL. 1987. Cambrian trilobites of North China: Chinese Cambrian trilobites housed in the Smithsonian Institution. Science Press, Beijing. ZHOU TIANMEI, YIREN LIU, XIANSONG MENG, AND ZHENHUA SUN. 1977. Class Trilobita, In Palaeontological atlas of Central and South China. 1, Early Palaeozoic. Geological Publishing House, Beijing. 104– 266. (in Chinese) ZHU XUEJIAN, N. C. HUGHES AND SHANCHI PENG. 2007. On a new species of Shergoldia Zhang & Jell, 1987 (Trilobita), the family Tsinaniidae and the order Asaphida. Memoirs of the Association of Australasian Palaeontologists, 34:243–253.

ACCEPTED 29 NOVEMBER 2009