Ontogenetic moulting behavior of the Cambrian oryctocephalid trilobite Arthricocephalites xinzhaiheensis

Identifying moulting configurations

Numerous criteria have been suggested for identifying fossil moults, based on observations from fossil material in previous studies (Henningsmoen, 1975; Whittington, 1990; Whittington, 1997; Daley & Drage, 2016; Drage et al., 2018; Wang et al., 2020). According to these authors, four important criteria need to be taken into consideration. Criterion 1. Evidence for the opening of exoskeletal sutures, especially cephalic sutures; Criterion 2. A peculiar arrangement of the portions of the exoskeleton, in particular in configurations where certain portions are inverted with respect to the remainder; Criterion 3. Recurring configurations of displaced exoskeletal units; and Criterion 4. Consideration of preservational environment (Wang et al., 2020). The abundance of disarticulated configurations of Ar. xinzhaiheensis allowed the distinction between carcasses and exuviae to be made with confidence (Wang et al., 2020). Specimens displaying Somersault and Henningsmoen’s configuration (Figs. 2F–2I) were considered to be slightly disturbed (or undisturbed) exuviae (Wang et al., 2020). Identifying moults of Ar. xinzhaiheensis is discussed briefly below.

Somersault configuration involving inversion of the LCU is highly unlikely to have been produced by chance. The specimens in Henningsmoen’s configuration consist of LCU and trunk, and the cranidia are inverted, rotated or missing. As Whittington (1997) noted, it is difficult to accept that burial could take place without some disturbance of the remains (Wang et al., 2020). The positioning of the LCU and missing cranidium imply that the specimens suffered slight disturbance. In this study, two configurations of Ar. xinzhaiheensis meet criteria 1–4 and these arrangements are interpreted as resulting from a slightly disturbed (or undisturbed) exuvia which are consistent with Wang et al. (2020).

Fossil description

Specimens of Ar. xinzhaiheensis range in our sample 0.52–9.34 mm in sagittal trunk length and 0.94–7.87 mm in transverse trunk width (Table S1). According to the morphological characteristics of specimens and previous work (Wang et al., 2020; Shen et al., 2016), the number of thoracic segments and corresponding developmental stage can be confirmed (Fig. 3). Specimens in Somersault configuration (n = 49) include individuals from meraspid degree one to the holaspid stage (Figs. 3A–3L), while specimens in Henningsmoen’s configuration (n = 82) include individuals from meraspid degree six to the holaspid stage (Figs. 3M–3P). Figures 3B–3L, examples of Somersault configuration, show the inverted LCU lying beneath the thorax, confirming that the LCU must have been inverted anteriorly during moulting. As these specimens are internal moulds, the underlapping LCU impressed upon the overlying trunk is a major characteristic for the recognition of Somersault configuration in Ar. xinzhaiheensis. Apart from this, inverted librigenae extending from two sides of the trunk and broken thoracic segements revealing part of the inverted LCU are also important signals for identification. The specimen in Fig. 3A is also considered as Somersault configuration based on the inverted rostral-hypostomal plate lying beneath the trunk though the inverted librigenae is not very clear. Specimens of Henningsmoen’s configuration display disarticulated cranidia (invertion, lateral rotation or missing) and the joined LCU, which was held in place via integument, remains largely in situ with respect to the trunk (Figs. 3M–3P). As such these specimens are generally easily recognizable as moults belonging to Henningsmoen’s configuration. This configuration showing inverted or rotated cranidia in close association preserves substantial information of moulting. Those with missing cranidia are more likely caused by slight disturbance (Whittington, 1997; Wang et al., 2020), including the possibility of being carried away by the freshly moulted trilobite.

Analysis of dataset

There are two groups of data, the number of thoracic segments and the size of trunk, which are key markers of the developmental phase of the trilobite. The measured specimens show a reasonably complete ontogenetic sequence. The composition of moult configuration in different growth stages is significantly influenced by development (Figs. 4A, 4B). There is gradual transition from Somersault to Henningsmoen’s configuration during development, both with reference to ontogenetic stage (Fig. 4A) and to body size (Fig. 4B). Somersault configuration accounts for 37% of all specimens which appear in all stages of Ar. xinzhaiheensis, while nearly two thirds of specimens display Henningsmoen’s configuration which only exists in meraspid degree six, meraspid degree seven and holaspid stage (Fig. 4A). Somersault configuration is exclusive till meraspid degree five and the proportion gradually decrease from meraspid degree six (Figs. 4A, 4C–4J). In later development, Henningsmoen’s configuration gradually becomes dominant with the development of individuals (Figs. 4A, 4C–4J).

Ontogenetic moulting behavior of Arthricocephalites xinzhaiheensis

The reasonably complete ontogenetic sequence of moulting configurations provides an unparalleled opportunity to discuss the developmental moulting behavior of this trilobite. Somersault configuration and Henningsmoen’s configuration represent different moulting processes (Drage et al., 2018; Wang et al., 2020; Whittington, 1990; Chen, Han & Zhao, 2008). Based on previous studies, the sequence of events based on Somersault configuration is summarized here (Figs. 4K–4R, ①, ②). The trilobite began a relative severe dorsal flexure. Then opening of the cephalic sutures (facial and rostral sutures) would have allowed the emergence forward of the post-ecdysical animal. As the animal continued to withdraw from the old cuticle, the LCU tilted up vertically and inverted. Meanwhile, the process which may have led to formation of Henningsmoen’s configuration can be presumed as follows. The trilobite began an initial downward (towards the sediment) angling of the cephalon and upwards flexure of the dorsal thorax producing a partial enrollment (Figs. 4P–4R, ③, ④). This high angle of cephalon resting on the sediment led to the cephalothoracic joint to break. The facial and rostral sutures successively opened. Subsequently, the disarticulated cranidia which provided wide exuvial gape were inverted forward or rotated beside when the newly moulted trilobite anteriorly egressed the old exoskeleton. Finally, the old exoskeleton was shed and the LCU which connected to the articulated trunk by integument preserved in suit. The composition of moult configurations (Figs. 4A–4J) suggested gradual transition of moulting behavior in Ar. xinzhaiheensis during development. Here we summarise the ontogenetic moulting behavior of Ar. xinzhaiheensis (Figs. 4K–4R). Somersault configuration is exclusive till meraspid degree five (Figs. 4K–4O) and exists in subsequent growth stage. This suggests that opening of facial and rostral sutures allowing the emergence forward of the post-ecdysial trilobite is prevalent in early growth stage. With the subsequent development of individuals, the composition of Henningsmoen’s configuration is increased. This demonstrates that the operations of the cephalic sutures (facial sutures and rostral suture) and cephalothoracic joint for newly moulted trilobite gradually become dominant behavior in later growth stage (Figs. 4H–4J, 4P–4R).

The change from Somersault to Henningsmoen’s configuration in Ar. xinzhaiheensis does not imply a change in the importance of the cephalic sutures. The operation of the cephalic sutures in small specimens generally seems to have been adequate, whereas in larger specimens both the cephalic sutures and disarticulation of the cephalothoracic joint were required. Essentially what we are seeing is that in the Somersault configuration the ecdysial gape is opened by a downwards movement of the LCU relative to the cranidium, whereas with Henningsmoen’s configuration the LCU stays in place while the ecydisal gape is opened by the removal of the cranidium. This implies more dorsal flexure in small specimens (resulting in the tipping over of the LCU) and less flexure in larger specimens (at least when they are extracting themselves from the exuviae). Actually it is difficult to say why there was a change in moulting strategy during development in this study. Perhaps small specimens could enroll more efficiently than larger specimens (e.g., for protection). For example, in Estaingia bilobata from the Emu Bay Shale meraspides are often found enrolled, whereas this is less common in larger specimens (Holmes, Paterson & García-Bellido, 2021).

New insight of variability in trilobite mouting behavior

Compared with a signature ecdysial pattern in extant macrobenthic marine arthropods (crabs, shrimp, lobsters, horseshoe crabs), moulting behavior in trilobites reveals obvious variation (Drage, 2019; Phlippen et al., 2000; Ayali, 2009; Brandt, 2002; Drage et al., 2019; Zong, 2020). Brandt (2002) considered that trilobites seem to have been opportunistic moulters by discussing ecdysial efficiency and evolutionary efficacy among marine arthropods. The most common moulting behavior for trilobites is utilizing the cephalic sutures, as expected by the almost ubiquitous presence of librigenae with facial sutures in trilobite morphology (Drage, 2019). Devonian phacopids possessed fused facial sutures in holaspides, displaying the inverted complete cephalon (excluding the hypostome) in exuviae (Henningsmoen, 1975; Whittington, 1997; Zong & Gong, 2017; Rustán et al., 2011). Drage, Laibl & Budil (2018) considered that facial suture fusion in Dalmanitina (phacopid trilobites) was intraspecifically variable which necessitate unstable alterations to exoskeletal moulting behavior. These works lead a realization that ecdysial motifs for trilobites are not stable within species. Previous studies also indicated, however, that the same moulting behavior exists in different trilobite clades and intraspecific variation of this strategy also appears in some trilobites (Drage et al., 2018; Drage, 2019; Wang et al., 2020; Whittington, 1990; Chen, Han & Zhao, 2008; Zong & Gong, 2017; McNamara, 1986; McNamara & Rudkin, 1984; Speyer, 1985; Zong, 2020). Some disarticulated carcasses were considered as disturbed exuviae in some studies (Drage et al., 2018; Whittington, 1990; McNamara, 1986). Abundant exuviae of Ar. xinzhaiheensis show that variation in moult configurations is regular in this species, implying a gradual transition of moulting behaviour across ontogeny.