Quantified growth and possible heterochronic development of two corynexochid trilobites from the middle Cambrian (Miaolingian Series, Wuliuan Stage) Mount Cap Formation, eastern Mackenzie Mountains, northwestern Canada

Abstract

The ontogeny of two species of corynexochid trilobites from the middle Cambrian Mount Cap Formation of the eastern Mackenzie Mountains, northern Canada, is documented. Sahtuia carcajouensis (Dolichometopidae) and Mackenzieaspis parallelispinosa (Zacanthoididae) are both endemic to this formation and only known from one locality. They, along with several other corynexochid taxa, occur in a succession of mudstone with scattered carbonate interbeds, deposited in a weakly storm-agitated setting near the flank of a semi-enclosed basin. The ontogeny of both species is characterized by mainly normal cranidial development, but a unique distribution of segments in their thoraxes and pygidia. The number of trunk segments was typical for their respective families, whereas the final number of segments released into the thorax was reduced. This occurred in both species through timing modifications to segment release, indicating heterochrony. Sahtuia carcajouensis and Mackenzieaspis parallelispinosa are likely derived from two separate clades, and heterochrony probably arose separately but synchronously. The endemicity of both species probably reflects unique paleoecological conditions in this part of the basin. Preliminary results indicate that the fossil-bearing mudstone was deposited under well-oxygenated conditions that underwent high nutrient flux and possibly experienced varying salinity. These factors may have affected the organisms’ physiology, or perhaps provoked an adaptation to achieve early maturation.

Non-technical Summary

Adult and juvenile specimens of two species of trilobites, Sahtuia carcajouensis and Mackenzieaspis parallelispinosa, were collected from the Mount Cap Formation of the eastern Mackenzie Mountains. This formation was deposited within a semi-enclosed, epicontinental sea during the early and middle Cambrian (approximately 530–500 million years ago). These two species are endemic to one section at Carcajou Falls in strata of the Glossopleura walcotti Zone, where they also are prolific, suggesting they had some type of affinity to the locality. These species are characterized by their unusually low number of thoracic segments and high number of pygidial segments. We describe and quantify their growth and development here and compare them to those of other species from their families, Dolichometopidae and Zacanthoididae, respectively. The results indicate that S. carcajouensis and M. parallelispinosa underwent relatively normal growth and development of their cranidia and developed a typical number of segments in their trunks (the combined thorax and pygidium). However, both species underwent fewer occurrences of segment release, when segments are transferred from the pygidium to the thorax, than was typical for dolichometopids and zacanthoidids. We hypothesize here that these species are not closely related, and instead that their unique development was brought about by changes to the developmental timing of two probable ancestor taxa by local environmental conditions.

Introduction

Trilobites, due to their well-preserved mineralized exoskeletons and easily identifiable growth stages, have been studied extensively in order to reconstruct their ontogenetic development (e.g., Barrande, 1852; Whittington, 1957; Palmer, 1958, 1962; Robison, 1967; Hu, 1971, 1985a, b; Robison and Campbell, 1974; Chatterton and Speyer, 1997; Hughes et al., 1999, 2006, 2017; Sundberg, 2000; Hughes, 2003, 2007; Webster and Zelditch, 2005, 2011; Webster, 2007, 2011, 2015; Dai and Zhang, 2012a, b, 2013; Dai et al., 2014, 2017; Laibl et al., 2014, 2015, 2018, 2021, 2023; Hou et al., 2015, 2017; Hopkins, 2017, 2021; Holmes et al., 2019, 2021a, b, c; Webster and Sundberg, 2019). In this way, growth mechanisms during the early history of arthropods can be compared to those that operate on extant counterparts. In addition, morphological similarities between protaspides and meraspides of numerous taxa provide evidence for closer evolutionary relationships than may be apparent from disparate holaspid attributes (Fortey, 1990, 2001). Investigations into the growth and development of trilobites have also revealed how the evolution of a number of groups has been influenced by developmental changes during ontogeny (e.g., Hughes et al., 1999, 2006, 2017; Sundberg, 2000; Hughes, 2003, 2007; Webster and Zelditch, 2005, 2011; Webster, 2007, 2011, 2015; Dai et al., 2017, 2021; Hopkins, 2017; Webster and Sundberg, 2019).

Changes in the developmental schedule relative to an ancestor is an important mechanism contributing to the evolutionary history of a number of trilobite clades (McNamara, 1978, 1981, 1986b, 2009; Crônier et al., 1999; Clausen, 2004). Modifications that indicate heterochrony will result in the alteration of the rate of morphological change, or a change in the timing of a developmental event (Webster and Zelditch, 2005), either of which will produce a species that exhibits a conserved growth trajectory in that feature, absent any spatial alterations.

Heterochrony, in the classical understanding, facilitates paedomorphism or peramorphism in the descendent taxon (McNamara, 1986a). Paedomorphism is when a descendent adult taxon exhibits a morphology or size similar to that of juvenile stages of its ancestor, brought about by either hastened sexual maturation (progenesis), slowed development (neoteny), or the delayed onset of growth of one or more morphological characters (post-displacement). By contrast, peramorphism is when a descendent taxon exhibits a larger size or a novel stage of development that is not present in its ancestor, brought about by either delayed sexual maturation (hypermorphosis), accelerated development (acceleration), or the early onset of growth of one or more morphological characters (pre-displacement). Other types of developmental changes that can occur are related to features that reflect spatial changes during ontogeny. Allometric repatterning is recognized when a descendent taxon has a differing growth trajectory relative to its ancestor (Webster and Zelditch, 2005). Other changes that may occur to development include alterations to the number (heterometry), type (heterotypy), or position (heterotopy) of an anatomical feature (Webster and Zelditch, 2005).

When dealing with fossils alone, however, it is difficult to establish a relationship between developmental changes and the precise environmental or paleoecological conditions that may have triggered them. This is complicated further by the possibility that multiple factors may have contributed to the selection for the modifications. There has been some speculation as to the kinds of environmental stimuli affecting development, such as water temperature (McNamara, 1981, 1986b), or occupation of a specific niche in the ecosystem favoring certain morphologies (Hughes et al., 1999).

In this study, the growth and development of two species of Corynexochida, Sahtuia carcajouensis Handkamer and Pratt in Handkamer et al., 2022, and Mackenzieaspis parallelispinosa Handkamer and Pratt in Handkamer et al., 2022, from the upper-lower and middle Cambrian (Series 2, Stage 4 to Miaolingian Series, Wuliuan Stage) Mount Cap Formation of the Northwest Territories, Canada (Fig. 1.1, 1.2), are characterized and quantified. Following this, their phylogenetic position in separate families, Dolichometopidae and Zacanthoididae, respectively, is hypothesized in order to elucidate how alterations to development led to their unique morphologies. Finally, evolutionary selection for their morphologies is put in the context of the basin structure, sedimentary conditions, and paleoecological attributes. We consider future investigations into these aspects may reveal a possible link between intrinsic developmental changes in trilobites and extrinsic environmental factors.

Figure 1. Location of study area. (1) Map of Canada showing the study area in the Northwest Territories; (2) map of the central part of the Northwest Territories showing the study area southwest of Norman Wells; (3) map of the eastern Mackenzie Mountains (part of NTS 96D) showing the location of the section that contains the specimens in this study, Carcajou Falls, and another studied section, Dodo Canyon (Carcajou Falls = 64.670639°N, 127.161682°W; Dodo Canyon = 64.937525°N, 127.265209°W). The dashed line roughly delineates the eastern edge of the Mackenzie Arch, and the gray-shaded areas indicate the outcrop belts of the Mount Clark, Mount Cap, and Saline River formations (modified from Handkamer et al., 2022, section descriptions are in Handkamer, 2020).

Trilobite ontogeny

Traditionally, the ontogeny of trilobites has been subdivided into three major stages of development: the protaspis, meraspis, and holaspis (Beecher, 1895; Chatterton and Speyer, 1997; Hughes et al., 2006). A “phaselus” has been recognized in a few species as a calcified stage that proceeds the protaspis and lacks a facial suture, rostral plate, and hypostome, although its acceptance is contentious (Fortey and Morris, 1978; see also Chatterton and Speyer, 1997; Hughes et al., 2006, p. 612). The protaspid stage is defined when the animal consists of a single dorsal shield, a defined axial furrow, a hypostomal suture, and may have a shallow suture between the cranidium and trunk (Hughes et al., 2006, previously referred to as the protopygidium elsewhere), as well as a few other features (Chatterton and Speyer, 1997; Hughes et al., 2006). Beecher (1895) divided the protaspis stage into three substages: the anaprotaspis, metaprotaspis, and paraprotaspis. Chatterton and Speyer (1997) advocated instead that protaspid life stages should be categorized based on their morphology, separated into either the nonadult-like bulbous form or the adult-like, flat, disc-shaped form. These stages are not homologous among all trilobites and were probably related to ecological constraints (Chatterton and Speyer, 1997; Laibl et al., 2023).

Although many species exhibit both protaspid forms during ontogeny, some species do not (Park and Choi, 2011). For example, the nonadult-like form has yet to be documented in corynexochids. Robison (1967) recognized that the early stages of growth of Bathyuriscus fimbriatus Robison, 1964, formed four distinct groups based on size but could only discern two distinct morphologies: the earlier anaprotaspid stage and later metaprotaspid stage, separated by the appearance of the juvenile trunk. This was followed by Öpik (1982) with Fuchouia fecunda Öpik, 1982. For the ontogeny of Ptarmiganoides propinqua (Resser, 1939b), Glossopleura boccar (Walcott, 1916), Paralbertella limbata (Rasetti, 1951), and Fieldaspis quadrangularis Hu, 1985, Hu (1971, 1985a, b) placed the insertion of the trunk instead at the onset of the paraprotaspid stage. Hopkins and Webster (2009) did not adopt Beecher's (1895) terminology for Zacanthopsis palmeri Hopkins and Webster, 2009, opting instead to subdivide the protaspid stage into the early and late stages, distinguished by morphological differences, the most distinctive being the insertion of the axial furrow to define the glabella and the formation of a trunk. Because development of the trunk occurred consistently at a protaspis length of 0.45–0.55 mm, it probably represents an important stage of development in all corynexochids, and therefore their terminology is followed here. The original terminology will be used, however, when citing published examples.

The development of a fully formed articulation between the cranidium and the trunk demarcates the onset of the meraspid stage (Chatterton and Speyer, 1997; Hughes, 2003). Meraspid development is characterized by the stepwise growth and development of the cephalon and trunk, the latter meaning all post-cephalic, exoskeletal segments comprising the thorax and pygidium. Trunk segments were generated at the subterminal generative zone in the posterior region of the meraspid pygidium (sometimes called the transitory pygidium), the process referred to here as segment appearance (Hughes et al., 2021; “somitogenesis” in McNamara, 2009). The transition of trunk segments from the meraspid pygidium to the thorax occurred by the development of an articulation at the posterior margin of the anteriormost pygidial segment (Hughes, 2003), referred to here as segment release (Hughes et al., 2021; “tagmosis” in McNamara, 2009).

Growth in trilobites was incremental, reflecting episodic molting, and the stages of ontogeny are defined by the number of segments allocated to the thorax, termed instars (Chatterton and Speyer, 1997). Most trilobite species released one segment per molt during their meraspid phase (Chatterton and Speyer, 1997), although some species can have successive molts in which no segment was released (Zhang, 1989), or in which more than one segment was released (Chatterton, 1971). Thus, trunk development in trilobites can be described as the stepwise co-occurrence of two processes: the increase in the number of trunk segments, and the posterior migration of the thorax–pygidium boundary via the development of thoracic articulations (Hughes, 2007).

Segment release occurred until a stable number was attained in the thorax, defining the holaspid stage. Growth, in terms of size, usually continued in this stage and segment generation may have persisted, but segment release was terminated (Hughes et al., 2006). Most trilobites underwent hemianamorphic trunk development, meaning that they underwent an anamorphic phase in which segment generation occurred following molting, succeeded by an epimorphic phase in which molting occurred absent segment generation (Hughes, 2003, 2007). Three modes of growth in trilobites can be characterized by timing of the onset of the holaspid stage relative to that of epimorphic growth (Hughes, 2007): protarthrous mode is when the holaspid stage preceded epimorphic development, protomeric mode is when epimorphic development preceded the onset of the holaspid stage, and synarthromeric mode is when these two transitions occurred synchronously. All three modes of growth have been documented in trilobites (Hughes et al., 2006). Protarthrous mode characterizes the growth of B. fimbriatus (Robison, 1964, 1967) because complete holaspides can have a variable number of pygidial segments, although the modes of growth exhibited by other dolichometopids or by zacanthoidids are not known. A recent investigation into the ontogeny of the oryctocephalid Oryctocarella duyunensis Qian, 1961, identified determinate growth following anamorphic development in that species (Dai et al., 2021), revealing that trunk development in trilobites was more diverse than was previously understood.

Geological setting

Mudstone and variably interbedded lime mudstone, with minor dolostone and siltstone, and local sandstone and coarse-grained carbonates of the Mount Cap Formation crop out in the eastern Mackenzie Mountains, Franklin Mountains, Hornaday River Canyon, and the western shoreline of Great Bear Lake and other nearby, smaller lakes (Williams, 1923; Aitken et al., 1973; Aitken and Cook, 1974; MacNaughton et al., 2013; Bouchard and Turner, 2017a, b; Handkamer et al., 2022). The Mount Cap Formation is also present in the subsurface below the Mackenzie Plain and Northern Interior Plains (Balkwill, 1971; Aitken et al., 1973; Meijer-Drees, 1975; Sommers et al., 2020). This formation conformably but diachronously (MacNaughton et al., 2013; Handkamer et al., 2022) overlies sandstone, siltstone, and minor mudstone and dolomitic limestone of the lower to lower-middle Cambrian (Series 2, Stage 4 to Miaolingian Series, Wuliuan Stage) Mount Clark Formation, and is in turn unconformably overlain by mudstone, dolostone, and evaporite of the upper-middle Cambrian Saline River Formation (Miaolingian Series, Wuliuan to Drumian stages, possibly younger).

The Mount Cap Formation was deposited in a semi-enclosed basin that was partially connected to open ocean (Fig. 2) (Aitken and Cook, 1974; Dixon and Stasiuk, 1998; MacLean, 2011; Sommers et al., 2020; Fallas et al., 2021). During deposition of the Mount Cap Formation, following a period of thermal subsidence, the basin underwent rifting and developed a graben system at its core (MacLean, 2011). Deposition of the overlying evaporites records basin restriction and an arid paleoclimate. Local thinning of formations indicates that the basin was bound by several regional paleotopographic highs, which are composed of Neoproterozoic sedimentary units. These paleotopographic highs subdivided the basin into various depocentres (Dixon and Stasiuk, 1998; MacLean, 2011; Sommers et al., 2020), and provided local sediment sources (Hadlari et al., 2012; Lane and Gehrels, 2014).

Figure 2. Paleogeographic elements of the Cambrian basin in northwestern Canada (modified from Dixon and Stasiuk, 1998; Sommers et al., 2020). AD = Aubry depocentre, BLA = Bulmer Lake Arch, CA = Coppermine Arch, GBD = Great Bear depocentre, GHD = Good Hope depocentre, HD = Horton depocentre, LMD = La Martre depocentre, LR = Leith Ridge, MA = Mahony Arch, MG = McConnell Graben, MPD = Mackenzie Plain depocentre, MR = Maunoir Ridge, MT = Mackenzie Trough. Stars denote sections and cores of the Mount Cap Formation mentioned in the text: 1 = well L-04 in the subsurface Mackenzie Plain (Pugh, 1993; Dixon and Stasiuk, 1998); 2 = outcrops in the Franklin Mountains (Aitken et al., 1973); 3 = wells D-76, P-02, and G-77 in the subsurface Horn Plateau (Meijer-Drees, 1975); 4 = outcrops near the western margin of Hottah Lake (Balkwill, 1971); 5 = multiple wells in the subsurface Colville Hills (Dixon and Stasiuk, 1998; Sommers et al., 2020); 6 = outcrop along the Hornaday River Canyon (Aitken et al., 1973; Bouchard and Turner, 2017a); 7 = outcrops in the eastern Mackenzie Mountains (Aitken et al., 1973; Bouchard and Turner, 2017b; Handkamer, 2020, Handkamer et al., 2022).

The thickest succession (595 m, MacLean, 2011) of the Mount Cap Formation is beneath the Mackenzie Plain in the Mackenzie Trough, one of the extensional grabens at the core of the basin that was active during the middle Cambrian, where it is composed of greenish-gray, argillaceous mudstone, minor dolostone and limestone (Pugh, 1993; Dixon, 1997; Dixon and Stasiuk, 1998), and local halite (Shell Canada Limited, 1965; MacLean and Cook, 1999). The strata deposited on the eastern flank of the Mackenzie Trough, exposed in the Franklin Mountains, are composed of similar lithologies (Aitken at el., 1973), but lack halite. East of the Mackenzie Trough in the Great Bear depocentre below the plains west of Great Bear Lake, and in the La Martre depocentre below the Horn Plateau, the Mount Cap Formation is composed of the same greenish-gray, argillaceous mudstone, but with a medial, thick dolostone unit (Meijer-Drees, 1975). These rocks pass eastward into mudstone and minor sandstone with abundant, horizontal burrows west of Hottah Lake (Balkwill, 1971; Meijer-Drees, 1975) near the eastern flank of the basin.

In the subsurface of the Colville Hills in the north-central part of the basin, where strata were deposited in the Good Hope depocentre, the Mount Cap Formation consists of similar mudstone, as well as carbonates, siltstone, and sandstone with intensely bioturbated intervals (Sommers et al., 2020). Mudstone–carbonate cyclicity is present in the Colville Hills (Pugh, 1993; Sommers et al., 2020), but this has not been recognized farther south in the Mackenzie Trough and Great Bear and La Martre depocentres (Meijer-Drees, 1975; Pugh, 1993). In contrast to those regions, in the Hornaday River Canyon area near the northeastern flank of the basin, the formation is composed of intensely bioturbated, coarsely crystalline dolostone and coarse-grained sandstone units (Bouchard and Turner, 2017a). The lithologically variable character of the Mount Cap Formation probably reflects a diversity of depositional conditions across the basin. Alginites have been documented in the subsurface of the Colville Hills (Wielens et al., 1990; Dixon and Stasiuk, 1998). These alginites are considered to be indicative of a restricted environment, suggesting that periodic phytoplankton blooms above normal levels of productivity occurred locally, and probably also throughout much of the basin (Dixon and Stasiuk, 1998).

The study area is in the Canyon Ranges of the eastern Mackenzie Mountains (Fig. 1.3), where Cambrian formations are present at the distal edge of the Mackenzie Plain depocentre (MPD) and onlap the Mackenzie Arch, which was a regionally extensive paleotopographic high that bounded the basin to the west (Aitken et al., 1973; Dixon and Stasiuk, 1998). There, the Mount Cap Formation is thinner, dominantly composed of gray or black, rarely green, silty mudstone, lime mudstone, and locally coarse-grained carbonate and siltstone (Aitken et al., 1973; Bouchard and Turner, 2017b; Handkamer, 2020; Handkamer et al., 2022). Biostratigraphy indicates deposition occurred during the upper Olenellus through upper Glossopleura walcotti zones (Fritz, 1969; Serié et al., 2013; Handkamer et al., 2022). Macrofossils appear to be more common in the eastern Mackenzie Mountains than in subsurface cores, the Franklin Mountains, and the eastern outcrop belt (Aitken et al., 1973; Serié et al., 2013; Bouchard and Turner, 2017b; Handkamer et al., 2022). Alginites like those from the Colville Hills are present in outcrops at Dodo Canyon (Fig. 1.3) along with acritarchs (Stasiuk, 2005).

All specimens of Sahtuia carcajouensis and Mackenzieaspis parallelispinosa are from the Glossopleura walcotti Zone (uppermost Delamaran Stage, Lincolnian Series of Laurentia; upper Wuliuan Stage, Miaolingian Series) of the Mount Cap Formation at Carcajou Falls (Fig. 1.3) (Handkamer et al., 2022). Samples were collected from the uppermost 8.7 m of dark gray, fossiliferous, silty mudstone (Fig. 3), with a few sclerites collected from interbedded lime mudstone and thin bioclastic wackestone beds, and bioclastic grainstone lenses. In lime mudstone, wackestone, and pyrite-bearing mudstone horizons, there are rare horizontal burrows (Planolites or Trichichnus; Bioturbation Index = 0–1). Trilobites occur with linguliformean brachiopods. Co-occurring trilobite species include Glossopleura youngi Handkamer and Pratt in Handkamer et al., 2022, Eobathyuriscus mackenziensis Handkamer and Pratt in Handkamer et al., 2022, E. macqueeni Handkamer and Pratt in Handkamer et al., 2022, Albertella levis Walcott, 1917a, Mackenzieaspis divergens Handkamer and Pratt in Handkamer et al., 2022, and an unidentified ptychoparioid. However, whereas G. youngi is moderately common, S. carcajouensis and M. parallelispinosa are much more numerous than the other taxa.

Figure 3. Outcrop of the Mount Cap Formation at Carcajou Falls. Stratal thickness visible is 28.2 m; the base of the formation is approximately five meters below the river level. Solid line indicates the contact between the locally recognized Albertelloides mischi Zone (see Handkamer et al., 2022) and the Glossopleura walcotti Zone. Strata above the dashed line consist of fossiliferous, silty mudstone with interbedded carbonates. These yielded the specimens studied herein, which were collected from both sides of the river about 100 m upriver (behind the observer).

Mudstone is interpreted as hemipelagic in origin (Handkamer et al., 2022), possibly along with some deposition of mud and silt by gravity-driven flows. The high percentage of fossils in the mudstone facies (Fig. 4) is suggestive of an overall low sedimentation rate (Speyer and Brett, 1986). Beds of lime mudstone and bioclastic wackestone are regarded to have been deposited by hemipelagic fallout and locally modified by occasional weak, storm-generated currents. Bioclastic grainstone lenses are interpreted as products of winnowing by these currents. The restricted trace fossil assemblage may record a stressful environment for the infauna, although the apparent absence of burrows in most beds could instead be taphonomic. Correlation and paleogeographic relationships suggest that Carcajou Falls, while still distal, was closer to the paleoshoreline than other outcrops studied (Handkamer et al., 2022).

Figure 4. Bedding plane of fossiliferous, silty mudstone (in Fig. 3) showing articulated and partially articulated exoskeletons as well as disarticulated cranidia, pygidia, free cheeks, and thoracic segments, mostly oriented dorsal side up. Scale bar = 10 mm.

Methods

Specimens examined were collected in two field seasons: the first in 2011 by Pratt and the second in 2019 by Handkamer when stratigraphic sections were measured in detail (Handkamer, 2020). Following mechanical preparation as needed, fossils were coated with ammonium chloride sublimate and photographed. Large specimens were photographed using a DSLR camera with a macro lens, whereas smaller specimens were photographed using an Olympus SZX-16 microscope. Photograph quality was optimized using Adobe Photoshop.

Traditional and geometric morphometric data were collected using ImageJ from individual sclerites, or if present, partially or fully articulated specimens, including 14 cranidia and 26 pygidia assigned to Sahtuia carcajouensis, 20 cranidia and 36 pygidia assigned to Mackenzieaspis parallelispinosa, and 11 juvenile sclerites unassignable to either species. Measurements were taken three times (standard error = 0.00153), and the mean value was utilized in descriptions and linear morphometrics (Fig. 5.1). Reconstructions of trilobite ontogeny ideally should incorporate models developed in previous studies (e.g., Hopkins, 2020; Hughes et al., 2021), but the paucity juvenile material recovered, as well as its disarticulated nature, precluded the recognition of specific instar stages. Instead, growth reconstructions are based on linear regression plots of the length and width of cranidia. Cranidial morphs were recognized by size clusters that show different morphologies, regarded as different stages of development of the cranidium. Multiple clusters of somewhat variable size that exhibit the qualitatively same morphology are regarded as the same morph. These morphs were compared to those of other dolichometopids and zacanthoidids (Robison, 1967; Hu, 1971, 1985a, b; Öpik, 1982; Hopkins and Webster, 2009; Handkamer et al., 2022).

Figure 5. Illustrative guide to morphology, measurements, and landmarks used herein. (1) Linear measurements of cranidia: W = width of cranidium passing through its mid-length; AB = anterior border length; ABF = anterior border furrow; AF = axial furrow; AFS = anterior branch of the facial suture; AL = anterior glabellar lobe; GA = width of anterior glabellar lobe; GL = glabellar length, measured from occipital furrow to anterior margin of glabella; GO = width of glabella just anterior to occipital furrow; IA = width of interocular area, measured at widest point, from interior edge of palpebral lobe to axial furrow; L = sagittal length of cranidium; L1–L4 = lateral glabellar lobes; O = ocular ridge; OF = occipital furrow; OR = occipital ring; PB = width of posterior border; PBF = posterior border furrow; PFS = posterior branch of the facial suture; PL = length of palpebral lobe; S1–S4 = lateral glabellar furrows. (2) Landmark distribution: 1 = intersection of sagittal line and anterior cranidial margin; 2 = intersection of sagittal line and posterior margin of occipital ring; 3 = intersection of sagittal line and anterior margin of glabella; 4 = intersection of anterior-facing side of ocular lobe and axial furrow; 5 = intersection of anterior tip of palpebral lobe and anterior branch of the facial suture; 6 = intersection of posterior tip of palpebral lobe and posterior branch of the facial suture; 7 = intersection of first lateral glabellar furrow (S1) and axial furrow; and 8 = intersection of occipital furrow and axial furrow. (3) linear measurements of pygidia: 1–8 = axial ring number; AA = width of axis at axial ring 1; AP = width of axis at anterior margin of terminal piece; L = sagittal length of pygidium; PA = width of pleural field adjacent to axial ring 1, not including adjacent pygidial border; PB = sagittal length of posterior border; PP = width of pleural field adjacent to anterior margin of terminal piece; TP = terminal piece; W = maximum width of pygidium.

Data on eight landmarks were collected (Fig. 5.2) to quantify shape changes between each cranidial morph during ontogeny, specifically of the anterior border, anterior and posterior portions of the glabella, interocular area, ocular ridge, palpebral lobe, and the occipital ring. Landmarks were fitted using Bookstein method, calculated in Past4.04, fitting landmark 1 to the origin of the graph (0, 0), and landmark 2 to a standardized position along the x-axis (1, 0). Landmark data were plotted in the form of vector-on-landmark graphs, with each graph illustrating the mean positional change of landmarks 3–8 between cranidial morphs. These graphs were rotated, changing the y-axis to the horizontal axis so that landmarks 1 and 2, which pass through the sagittal line, are parallel to the vertical axis. Measurements and landmark data were collected from the right half of the cranidium in dorsal view, although if the right half was not preserved, the left half was utilized, and values were transformed across the x-axis.

Pygidial ontogeny was characterized using the overall size and descriptive shape change, relative growth trajectories of different lobes of the pygidium, and length change relative to the number of axial rings (Fig. 5.3). Log-transformed linear regression for lobes of the pygidium was conducted to test for allometry. Log transformation was calculated in Past4.04.

Comparative linear morphometrics of the growth trajectories of the palpebral lobe and anterior and posterior glabella between both species was conducted to determine if they shared a common cranidial ontogeny. This utilized Reduced Major Axis (RMA) method of linear regression to quantify bivariate growth relationships (Sokal and Rohlf, 1995). RMA regression was chosen to account for possible measurement errors present in both dependent variables. Statistical significances for the equivalency of slopes in the comparative linear morphometrics were tested using standardized major axis estimation in R with the package smatr (α = 0.05) (Warton et al., 2012).

Repositories and institutional abbreviation

All figured specimens are deposited in the type collection of the Geological Survey of Canada, Ottawa. Specimens not illustrated are housed at the University of Saskatchewan. Institutional abbreviations: GSC = Geological Survey of Canada, Ottawa, Ontario, Canada.

Results

Distinguishing sclerites of Sahtuia carcajouensis from Mackenzieaspis parallelispinosa below a sagittal length of about 1.1 mm is not possible. For this reason, the earliest juvenile stages are unassigned and referred to as gen. and sp. indet. Other corynexochid taxa whose ontogenies have been reconstructed, such as Bathyuriscus fimbriatus, Fuchouia fecunda, Glossopleura boccar, Ptarmiganoides propinqua, Fieldaspis quadrangularis, Paralbertella limbata, and Zacanthopsis palmeri, exhibit similar morphologies during the metaprotaspid and earliest meraspid stages (Robison, 1967; Hu, 1971, 1985a, b; Öpik, 1982; Hopkins and Webster, 2009).

The increase in pygidial length accompanied by an addition to the number of axial rings is regarded to indicate an episode of segment appearance. Contrasting this, the increase in pygidial length accompanied by a subtraction in the number of axial rings may represent periods of segment release or may be size variation within the collection. Trunk growth models were then interpreted from the scheduling of episodes of segment release and segment appearance. These models assume that segment release episodes are characterized by the articulation of only one segment, which is the most common number in trilobites (Chatterton and Speyer, 1997). Because the instar stage of meraspides is defined by number of thoracic segments, there is a degree of uncertainty associated with interpreting growth stages by using other species as an analogy. For example, if a change to developmental timing only affects the thorax, this may not have any noticeable effect on the development of the cranidium. As such, the only stages that can be interpreted are the protaspid stage and the fully articulated holaspid stage.

The following subsection of the results is structured like a taxonomic section, but this is done for descriptive purposes only, and should not be considered taxonomic entries on either S. carcajouensis or M. parallelispinosa. The original authorships on the species, as well as the complete list of their type material, are in Handkamer et al. (2022, p. 24 and 35, respectively).

Genus and species indeterminate

Two cranidial morphs assigned to gen. and sp. indet. are recognized based on RMA regression plots of length versus width (Fig. 6).

Figure 6. RMA regression plot of length vs. width of gen. and sp. indet. that belong to either S. carcajouensis or M. parallelispinosa but cannot be distinguished, with interpreted cranidial morph groups: Cranidial morph 1 and Cranidial morph 2.

Cranidial ontogeny: Cranidial morph 1

Figures 7.1–7.3.

Figure 7. Genus and species indeterminant Cranidial morphs 1 and 2 from the Mount Cap Formation. (1–3) Genus and species indeterminant Cranidial morph 1: (1) protaspis (dorsal) GSC 142359; (2) protaspis (dorsal) GSC 143692; (3) protaspis (dorsal) GSC 142358. (4–6) Genus and species indeterminant Cranidial morph 2: (4) protaspis or early meraspid cranidium (dorsal) GSC 143693; (5) protaspis or early meraspid cranidium (dorsal) GSC 143694; (6) protaspis or early meraspid cranidium (dorsal) GSC 143695.

Description

Cranidium is 0.36–0.55 mm long, 0.42–0.59 mm wide, and circular or elliptical in outline. Axial furrow is moderately to well defined, the glabella comprising 64–84% of the cranidial length (sag.) and extending to the anterior border or terminating in an anterior depression. In smaller specimens (Fig. 7.1, 7.2) the axial furrow is nearly parallel to the sagittal line, whereas in larger specimens (Fig. 7.3) the glabella narrows (trans.) gently from the occipital furrow to S2 and widens (trans.) gently from S2 to the anterior glabellar lobe. Two pairs of lateral glabellar furrows are present as narrow indentations: S1 is moderately to poorly defined and oriented obliquely backwards 42–48° relative to the sagittal line; S2 is poorly defined and oriented obliquely backwards 54–63° relative to the sagittal line. Occipital furrow is weakly to moderately defined, the occipital ring widening (trans.) posteriorly. Anterior border furrow is moderately defined; length of the anterior border (sag.) comprises 9–10% of the total sagittal length. Facial suture is not visible, possibly marginal. Interocular area at the widest point (trans.) comprises 31–40% of the cranidial width. Palpebral lobe is gently curved, with the anterior tip merging with the ocular ridge and then intersecting the axial furrow at the anterior glabellar lobe, and the posterior tip opposite S2. Length (exsag.) of the palpebral lobe is 28–47% that of the cranidium. Posterior margin is oriented obliquely forward from the occipital ring to the posterior tip of the palpebral lobe. Posterior border furrow is moderately to well defined and in width (trans.) is 1.5 times to double that of the occipital furrow. Juvenile trunk absent or not visible.

Material

Six specimens.

Remarks

Genus and species indeterminant Cranidial morph 1 is regarded as a late protaspis, due to the presence of the axial furrow and distinct lateral glabellar furrows. A juvenile trunk is absent in all specimens, although this may be from compaction collapsing the trunk ventrally under the exoskeleton. However, the lack of a transversely oriented, posterior border furrow indicates that a full articulation between the protaspid cranidium and what would become the trunk has not developed.

Genus and species indeterminant Cranidial morph 1 is nearly identical to the metaprotaspis of B. fimbriatus (Robison, 1967, pl. 24, figs. 4, 5) and that of P. propinqua (Hu, 1971, pl. 10, figs. 6, 7), although it lacks the short marginal spines of the former species. The larger specimen (Fig. 7.3, GSC 142358) with a slightly wider glabella also resembles the smaller paraprotaspides of P. propinqua (Hu, 1971, pl. 10, fig. 8–10), although larger paraprotaspides of that species have a wider posterior border with a deeper furrow. The paraprotaspides of G. boccar differs noticeably by the significantly narrower interocular fixed cheek (Hu, 1985a, pl. 9, figs. 2–5). The metaprotaspides of F. quadrangularis (Hu, 1985b, pl. 20, figs. 7, 8) and P. limbata (Hu, 1985b, pl. 21, figs. 7–11) are similar in morphology to the smaller specimens of Cranidial morph 1 (Fig. 7.1, 7.2), whereas the larger specimens (e.g., Fig. 7.3) resemble the paraprotaspides of those species, albeit with a slightly wider glabella. Cranidial morph 1 also resembles late protaspides of Z. palmeri (Hopkins and Webster, 2009, fig. 10.6–10.8, 10.10) and to a lesser extent the early meraspid cranidium (Hopkins and Webster, 2009, fig. 10.21–10.23), although it lacks the transversely oriented posterior border of that species. The metaprotaspides of F. fecunda (Öpik, 1982, p. 12, figs. 1a–1d, 2) are also similar to this morph, but they have a more annulated glabella with more distinct lateral glabellar furrows.

Cranidial ontogeny: Cranidial morph 2

Figure 7.4–7.6.

Description

Cranidium is 0.64–0.90 mm long, 0.70–1.09 mm wide, and subquadrate in outline. Axial furrow is moderately to well defined, the glabella comprising 79–93% of the cranidial length (sag.) and extending to the anterior border furrow. Glabella narrows (trans.) gently from the occipital furrow to medial L2 and widens (trans.) gently from medial L2 to the anterior glabellar lobe. Three pairs of lateral glabellar furrows are present as narrow indentations: S1 is moderately to poorly defined and oriented obliquely backwards 55–59° relative to the sagittal line; S2 is poorly defined and oriented obliquely backwards 56–60° relative to the sagittal line; S3 is poorly defined and oriented transversely. Occipital furrow is moderately defined, the occipital ring widening posteriorly. Anterior border furrow is moderately defined; length of the anterior border (sag.) comprises 5–7% of the total sagittal length. Anterior branch of the facial suture is strongly convergent anteriorly or nearly marginal. Interocular area at the widest point (trans.) comprises 26–38% of the cranidial width. Palpebral lobe is gently curved, with the anterior tip merging with the ocular ridge and then intersecting the axial furrow at the anterior glabellar lobe, and the posterior tip opposite the medial part of L2. Length (exsag.) of the palpebral lobe is 58–65% that of the cranidium. Posterior branch of the facial suture is oriented obliquely outwards and backwards. Posterior border is moderately defined, oriented nearly transversely and in width (trans.) is 134–149% that of the occipital furrow. Juvenile trunk absent or not visible.

Material

Five specimens.

Remarks

Ontogenetic changes that occurred from gen. and sp. indet. Cranidial morph 1 to Cranidial morph 2 include the increase in the overall size, change in outline from circular or semielliptical to subquadrate, proportional widening of the glabella, especially the anterior lobe, development of S3, development of the anterior and posterior branches of the facial suture, slight relative narrowing of the interocular area, and proportional elongation of the palpebral lobe. The abaxial movement of landmarks 4, 7, and 8 indicate that narrowing of the interocular area is due to widening of the glabella (Fig. 8). The large vector magnitudes in landmarks 5 and 6 suggest that the posterior and slight abaxial rotation of the palpebral lobe also contribute to this. The larger vector on landmark 6 as opposed to landmark 5 suggests that the relative increased length of the palpebral lobe is primarily due to its elongation posteriorly. The absence of an articulated trunk would confirm the stage of this morph, but the substantial variation exhibited by the cranidial, glabellar, and palpebral lobe dimensions suggests that both the latest protaspid and earliest meraspid stages are represented.

Figure 8. Bookstein-fitted, vector-on-landmark plot of the mean shape change from gen. and sp. indet. Cranidial morph 1 to gen. and sp. indet. Cranidial morph 2. Dots represent the mean landmark position of gen. and sp. indet. Cranidial morph 1, whereas the vectors indicate the direction and magnitude of shape change to the mean landmark position of gen. and sp. indet. Cranidial morph 2.

Genus and species indeterminant Cranidial morph 2 resembles the late protaspides and early meraspid cranidia of B. fimbriatus (Robison, 1967, pl. 24., figs. 5–7) and P. propinqua (Hu, 1971, pl. 10, figs. 9–14), although it differs from those of G. boccar by the wider interocular fixed cheek (Hu, 1985a, pl. 9, figs. 6, 7). Genus and species indeterminant Cranidial morph 2 is similar to protaspides of F. quadrangularis (Hu, 1985b, pl. 20, figs, 9, 10, 12, 13) and P. limbata (Hu, 1985b, pl. 21, figs. 7–11), although it differs from the early meraspid cranidia of both species (Hu, 1985b, pl. 20, figs. 14–16, pl. 21, figs. 12–14, 16) by having a subquadrate as opposed to subelliptical outline, a shorter palpebral lobe, and no metafixigenal spine. Meraspid cranidia M0 and M1 of Z. palmeri (Hopkins and Webster, 2009, fig. 10.21–10.23) are also similar to Cranidial morph 2 but differ by the presence of a fixigenal spine. The early meraspid cranidium of F. fecunda (Öpik, 1982, pl. 12, figs. 3, 4) is similar in that it has an expanded anterior glabellar lobe and transverse posterior border but differs by its small metafixigenal spine. The ontogenetic changes that occurred from gen. and sp. indet. Cranidial morph 1 to gen. and sp. indet. Cranidial morph 2 also occurred in the transition from protaspides to meraspides of all of the above seven species, although the glabella widened more in gen. and sp. indet. cranidia than in B. fimbriatus, F. quadrangularis, Z. palmeri, and F. fecunda.

Sahtuia carcajouensis

Three cranidial morphs assigned to S. carcajouensis are recognized based on RMA regression plots of length versus width (Fig. 9).

Figure 9. RMA regression plot of length vs. width of cranidia of Sahtuia carcajouensis, with interpreted cranidial morph groups: S. carcajouensis Cranidial morph 1, S. carcajouensis Cranidial morph 2, and S. carcajouensis Cranidial morph 3.

Cranidial ontogeny: Cranidial morph 1

Figure 10.2.

Figure 10. Sahtuia carcajouensis Cranidial morphs 1 and 2, and juvenile or mature pygidia of S. carcajouensis from the Mount Cap Formation. (1, 3) S. carcajouensis Cranidial morph 2: (1) cranidium (dorsal) GSC 143696; (3) cranidium (dorsal) GSC 143697; (2) S. carcajouensis Cranidial morph 1, cranidium (dorsal) GSC 143698; (4–7) juvenile or mature pygidia; (4) pygidium (dorsal) GSC 143699; (5) pygidium (dorsal) GSC 143700; (6) pygidium (dorsal) GSC 143701; (7) pygidium (dorsal) GSC 143702.

Description

Cranidium is 1.18 mm long, 1.51 mm wide, and subtrapezoidal in outline. Axial furrow is well defined, the glabella comprising 87% of the cranidial length (sag.) and extends to the anterior edge of the cranidium. Glabella narrows (trans.) gently from the occipital furrow to medial L2 and widens (trans.) from medial L2 to the anterior lobe. Four pairs of lateral glabellar furrows are present: S1 is moderately defined and oriented obliquely backwards 60° relative to the sagittal line; S2 is moderately defined and oriented obliquely backwards 67° relative to the sagittal line; S3 is poorly defined and oriented obliquely forewords 63° relative to the sagittal line; S4 is poorly defined and oriented obliquely forwards 61° relative to the sagittal line. Occipital furrow is moderately defined, the occipital ring widening posteriorly. Anterior border is absent or not preserved in the specimen. Anterior branch of the facial suture is strongly convergent anteriorly. Interocular area at the widest point (trans.) comprises 35% of the cranidial width. Palpebral lobe is curved, with the anterior tip merging with the ocular ridge and then intersecting the axial furrow slightly posterior of S4, and the posterior tip opposite of medial L2. Length (exsag.) of the palpebral lobe is 44% that of the cranidium. Posterior branch of the facial suture is oriented obliquely outwards and backwards, becoming more strongly posteriorly oriented near the posterior margin. Posterior border is oriented nearly transverse, in width (trans.) is 137% that of the occipital furrow and bears a metafixigenal spine. Posterior border furrow is moderately defined.

Material

One cranidium.

Remarks

Possible ontogenetic changes that occurred from gen. and sp. indet. Cranidial morph 2 to S. carcajouensis Cranidial morph 1 include the overall increase in size, change from a subquadrate to subtrapezoidal outline, inward curving of the axial furrow, development of S4, relative widening of the glabella and interocular area, proportional shortening of the palpebral lobe, and development of the metafixigenal spine. The dominantly abaxial migration of landmarks 4–8, but with the greatest magnitude on landmarks 5 and 6, reveals that the relative widening of the cranidium is mostly due to the widening of the interocular fixed cheek, but also the glabella (Fig. 11.1). Furthermore, the anterior and abaxial migration of landmark 5, and posterior and abaxial migration of landmark 6, indicate a relatively slight outward rotation of the palpebral lobe.

Figure 11. Bookstein-fitted, vector-on-landmark plot of the mean shape change of cranidial morphs of Sahtuia carcajouensis. Dots represent the mean landmark position of the smaller morph, whereas the vectors indicate the direction and magnitude of shape change to the mean landmark position of the larger morph. (1) Gen. and sp. indet. Cranidial morph 2 to S. carcajouensis Cranidial morph 1; (2) S. carcajouensis Cranidial morph 1 to S. carcajouensis Cranidial morph 2; (3) S. carcajouensis Cranidial morph 2 to S. carcajouensis Cranidial morph 3.

Sahtuia carcajouensis Cranidial morph 1 is similar to a juvenile specimen attributed to Eobathyuriscus mackenziensis (Handkamer et al., 2022, fig. 13.9). It is also similar to meraspid cranidia M-3, M-4, and M-5 of B. fimbriatus (Robison, 1967, pl. 24, figs. 9, 11, 13), although the length of the palpebral lobe in S. carcajouensis Cranidial morph 1 is closest to that of M-5. It is also similar to the larger early meraspid cranidium of P. propinqua (Hu, 1971, pl. 10, figs. 17, 18), but differs from the smaller specimens (Hu, 1971, pl. 10, fig. 16) by having a slightly wider interocular fixed cheek. The early meraspid cranidium of G. boccar has a distinctly narrower interocular fixed cheek and shorter metafixigenal spine (Hu, 1985a, pl. 9, figs. 3, 7). Sahtuia carcajouensis Cranidial morph 1 is similar to the early meraspid cranidium of F. quadrangularis (Hu, 1985b, pl. 20, figs. 14–16) in the shape of the glabella, but has a shorter palpebral lobe, shorter occipital ring, and has a more prominent metafixigenal spine. The early meraspid cranidium of P. limbata (Hu, 1985b, pl. 21, figs. 12–14, 16) differs from S. carcajouensis Cranidial morph 1 in having a narrower anterior glabellar lobe, longer and wider anterior border, more divergent anterior branch of the facial suture, longer palpebral lobe, and an occipital spine. The early meraspid cranidia M0 and M1 of Z. palmeri (Hopkins and Webster, 2009, fig. 10.21–10.23) are similar to S. carcajouensis Cranidial morph 1 but have longer anterior borders, more marginal anterior branches of the facial sutures, and shorter metafixigenal spines. Later stages M2 and M3 of Z. palmeri (Hopkins and Webster, 2009, fig. 11.1–11.20) differ from S. carcajouensis Cranidial morph 1 in having longer anterior borders, longer palpebral lobes, shorter metafixigenal spines, and occipital spines. Cranidial morph 1 resembles the early meraspid cranidium of F. fecunda (Öpik, 1982, pl. 12, fig. 5a, 5b) but differs by having a less annulated glabella with shallower lateral glabellar furrows, and a shorter metafixigenal spine.

Cranidial ontogeny: Cranidial morph 2

Figure 10.1, 10.3.

Description

Cranidium is 1.75–2.11 mm long, 1.97–2.58 mm wide, and subtrapezoidal in outline. Axial furrow is well defined, the glabella comprising 78–79% of the cranidial length (sag.) and extends to the anterior border furrow. Glabella narrows (trans.) gently from the occipital furrow to medial S1 and widens (trans.) from medial S1 to the anterior lobe. Four pairs of lateral glabellar furrows are present: S1 is well defined and oriented obliquely backwards 50–52° relative to the sagittal line; S2 is moderately to well defined and oriented obliquely backwards 72–75° relative to the sagittal line; S3 is poorly to moderately defined and oriented obliquely forwards 70–73° relative to the sagittal line; S4 is poorly defined and oriented obliquely forwards 68–70° relative to the sagittal line. Occipital furrow is well defined, the occipital ring widening posteriorly and bearing a medial tubercle. Anterior border furrow is moderately defined; length of the border comprises 6–8% of the total sagittal length. Anterior branch of the facial suture is parallel to the sagittal line, curving adaxially near the anterior margin. Interocular area at the widest point (trans.) comprises 29% of the cranidial width. Palpebral lobe is strongly arched, with the anterior tip merging into the ocular ridge and then intersecting the axial furrow at S4, and the posterior tip opposite of S1. Length (exsag.) of the palpebral lobe is 41–43% that of the cranidium. Posterior branch of the facial suture is oriented obliquely backwards, becoming more strongly posteriorly oriented near the distal tip of the postocular fixed cheek. Posterior border is oriented nearly transversely, in width (trans.) is 141–160% that of the occipital furrow and bears a short posterior protuberance near the distal tip. Posterior border furrow is well defined.

Material

Two cranidia.

Remarks

Ontogenetic changes that occurred from S. carcajouensis Cranidial morph 1 to S. carcajouensis Cranidial morph 2 include the overall increase in size, proportional widening of the glabella, slight elongation of the anterior border, appearance of a medial occipital tubercle, abaxial rotation of the anterior branch of the facial suture, proportional narrowing of the interocular area, relative widening of the posterior border, and relative shortening of the metafixigenal spine to a small posterior protuberance. Vectors indicating abaxial migration of landmarks 7 and 8, and those indicating adaxial migration of landmarks 5 and 6, suggest that the transverse growth of the cranidium was mostly brought about by widening of the glabella as opposed to the interocular fixed cheek (Fig. 11.2). The anterior components of the vectors on landmarks 7 and 8 probably indicate relative elongation of the occipital ring. The mostly posteriorly oriented vector on landmark 4 is interpreted as produced by the relative elongation of the anterior glabellar lobe, anterior border, and preocular fixed cheek.

The early holaspid cranidium of B. fimbriatus (Robison, 1967, pl. 24, fig. 15) differs from S. carcajouensis Cranidial morph 2 by the slightly more divergent anterior branch of the facial suture, slightly longer palpebral lobe, shorter posterior fixed cheek, wider posterior border, and long occipital spine instead of a tubercle. The M-5 meraspid cranidium of B. fimbriatus (Robison, 1967, pl. 24, fig. 13) has a wider anterior glabellar lobe, a more-curved palpebral lobe, narrower postocular fixed cheek, and a metafixigenal spine. The late meraspid cranidium of P. propinqua (Hu, 1971, pl. 10, figs. 19, 20) differs from S. carcajouensis Cranidial morph 2 by the wider glabella and presence of an occipital spine. The meraspid cranidium of G. boccar (Hu, 1985a, pl. 9, figs. 3, 7) differs from S. carcajouensis Cranidial morph 2 by being proportionally narrower, having a narrower interocular area, and a longer, less-curved palpebral lobe. Late meraspid cranidia of F. quadrangularis (Hu, 1985b, pl. 20, figs. 17, 18) and P. limbata (Hu, 1985b, pl. 21, figs. 17, 18, 21) differ from those of S. carcajouensis Cranidial morph 2 by their wider anterior borders, more divergent anterior branches of the facial sutures, longer palpebral lobes, and shorter posterior borders. The former species also has a wider interocular area than that of S. carcajouensis Cranidial morph 2, while the latter species has a narrower anterior glabellar lobe. Late meraspid cranidia M3 and post-M3 of Z. palmeri (Hopkins and Webster, 2009, fig. 11.1–11.20, 11.23, 11.26, 11.37–11.40, 11.44, 11.45) differ from those of S. carcajouensis Cranidial morph 2 by their longer and wider anterior borders, narrower anterior glabellar lobes, wider interocular fixed cheeks, longer palpebral lobes, more transversely oriented posterior facial sutures, wider posterior borders, and presence of occipital spines. Sahtuia carcajouensis Cranidial morph 2 is also similar to the late meraspid cranidium of F. fecunda (Öpik, 1982, pl. 12, figs. 7–11, pl. 13, figs. 1–3a), but differs by the less anteriorly rounded anterior glabellar lobe. The smaller examples of that stage of F. fecunda (Öpik, 1982, pl. 12, figs. 7–9) have narrower posterior glabellas, the medially sized examples of that stage (Öpik, 1982, pl. 12, figs. 10, 11) have narrower interocular areas and bear metafixigenal spines, and the larger examples of that stage (Öpik, 1982, pl. 13, figs. 1–3a) have longer anterior borders, less-curved axial furrows, more quadrate-shaped, effaced glabellas, narrower interocular areas, longer palpebral lobes, wider posterior borders, and bear metafixigenal spines and occipital spines.

Cranidial ontogeny: Cranidial morph 3

Figure 12.1–12.3.

Figure 12. Complete or nearly complete exoskeletons of Sahtuia carcajouensis with Cranidial morph 3 and mature pygidia from the Mount Cap Formation. (1) Holotype exoskeleton lacking one free cheek (dorsal) GSC 142342; (2) paratype complete exoskeleton (dorsal, latex mold) GSC 142343; (3) paratype complete exoskeleton (dorsal) GSC 142347.

Description

Cranidium is 3.36–10.39 mm long, 3.61–10.05 mm wide, and subtrapezoidal in outline. Axial furrow is well defined, the glabella comprising 77–86% of the cranidial length (sag.) and extends to the anterior border furrow. Glabella narrows (trans.) gently from the occipital furrow to medial S1 and widens (trans.) from medial S1 to the anterior glabellar lobe. Four pairs of lateral glabellar furrows are present: S1 is well defined and oriented obliquely backwards 53–58° relative to the sagittal line; S2 is moderately to well defined and oriented obliquely backwards 67–70° relative to the sagittal line; S3 is poorly to moderately defined and oriented obliquely forwards 84–88° relative to the sagittal line; S4 is poorly defined and oriented obliquely forwards 80–83° relative to the sagittal line. Occipital furrow is well defined, the occipital ring widening posteriorly and bearing a medial tubercle. Anterior border furrow is moderately to well defined; length of the border comprises 5–7% of the total sagittal length. Anterior branch of the facial suture is parallel or slightly divergent anteriorly, curving adaxially near the anterior margin. Interocular area at the widest (trans.) point comprises 24–27% of the cranidial width. Palpebral lobe is strongly arched, with the anterior tip merging with the ocular ridge and then intersecting the axial furrow at S4, and the posterior tip opposite medial L1. Length (exsag.) of the palpebral lobe is 33–42% that of the cranidium. Posterior branch of the facial suture is oriented obliquely outwards and backwards. Posterior border is oriented nearly transversely and in width (trans.) is 97–99% that of the occipital furrow. Posterior border furrow is well defined.

Material

Five complete exoskeletons, four exoskeletons lacking free cheeks, and three cranidia.

Remarks

Ontogenetic changes that occurred from Sahtuia carcajouensis Cranidial morph 2 to S. carcajouensis Cranidial morph 3 include the overall increase in size, slight proportional elongation of the glabella, slight widening of the anterior glabellar lobe, transverse rotation of S3 and S4, slight proportional narrowing of the interocular area, abaxial rotation of the anterior branch of the facial suture, proportional shortening of the palpebral lobe, relative narrowing of the posterior border, and disappearance of the posterior protuberance on the posterior border. The abaxial migration of landmarks 7 and 8, and adaxial migration of landmarks 5 and 6, indicate that the narrowing of the interocular area was achieved through the widening of the glabella and slight posterior rotation of the palpebral lobe (Fig. 11.3). The abaxial migration of landmark 4 is indicative of the widening of the anterior glabellar lobe, causing the posterior and inward rotation of the palpebral lobe. These specimens are regarded as holaspides because increasing exoskeleton sizes consistently retain four thoracic segments.

Sahtuia carcajouensis Cranidial morph 3 resembles the holaspid cranidium of Eobathyuriscus mackenziensis (Handkamer et al., 2022, figs. 12.1–12.9, 13.1, 13.2, 13.4, 13.11), but differs by having a slightly shorter palpebral lobe. Sahtuia carcajouensis Cranidial morph 3 differs from the early and late holaspid cranidia of B. fimbriatus (Robison, 1967, pl. 24, figs. 15, 25) by having a wider interocular area, narrower posterior border, particularly in regard to the late holaspis, and lacking an occipital spine. However, it also differs from the late holaspid cranidium of B. fimbriatus by having a shorter palpebral lobe, but that species loses its occipital spine. The holaspid cranidium of P. propinqua (Resser, 1939b, pl. 3, figs. 35, 36; Hu, 1971, pl. 10, figs. 21–23) differs by the slightly narrower interocular fixed cheek, longer palpebral lobe, narrower posterior border bearing a metafixigenal spine, and having an occipital spine. The holaspid cranidium of G. boccar (Walcott, 1916, pl. 52, figs. 1, 1a–f) differs from S. carcajouensis Cranidial morph 3 by having a narrower anterior glabellar lobe and interocular fixed cheek, longer, more curving palpebral lobe, and shorter postocular fixed cheek. Holaspid cranidia of both F. quadrangularis (Hu, 1985b, pl 20, figs. 20, 21, 27) and P. limbata (Rasetti, 1951, pl. 18, figs. 8–10, 14–16; Hu, 1985b, pl. 21, figs. 24, 27) both differ from S. carcajouensis Cranidial morph 3 by their wider anterior borders, more divergent anterior branches of the facial sutures, narrower interocular areas, longer, more curving palpebral lobes, more transversely oriented posterior branches of the facial sutures, and shorter postocular fixed cheeks. Paralbertella limbata also differs from S. carcajouensis Cranidial morph 3 by its narrower anterior glabella, longer anterior border, and wider posterior border. The holaspid cranidium of Z. palmeri (Hopkins and Webster, 2009, figs. 2.3, 2.4, 2.7–2.17, 2.20–2.26, 2.28–2.33, 3.1–3.20, 4.2–4.11, 4.13, 4.25–4.30, 4.32, 5.1–5.8, 5.15–5.21) differs from S. carcajouensis Cranidial morph 3 by its longer and wider anterior border, narrower anterior glabellar lobe, wider interocular area and ocular ridge, longer palpebral lobe, shorter postocular fixed cheek, narrower posterior border, and having an occipital spine. Sahtuia carcajouensis Cranidial morph 3 resembles the latest meraspid and holaspid cranidia of F. fecunda (Öpik, 1982, pl. 6, fig. 1, pl. 7, figs. 1–3, pl. 8, figs. 4–5, pl. 13, figs. 1–3a), but differs by its shorter anterior border, curved axial furrow, deeper lateral glabellar furrows, shorter palpebral lobe, narrower posterior border, and lacking an occipital spine and metafixigenal spine.

Trunk ontogeny

Pygidia (Figs. 10.4–10.7, 12.1–12.3) range from as small as 0.84 mm long and 1.57 mm wide to as large as 11.55 mm long and 15.35 mm wide (Fig. 13). There is a distinct change in shape from subtriangular to subelliptical and they can bear between four to nine axial rings and four to seven pleural furrows. Log-transformed data of RMA regression plots of the allometric trajectories of pygidia indicate that, relative to the axial lobe at axial ring one (Fig. 14.1) and the terminal piece (Fig. 14.2), the pleural lobes adjacent to them undergo near isometric growth transversely.

Figure 13. RMA regression plot of length vs. width of pygidia of Sahtuia carcajouensis.

Figure 14. Log-transformed RMA regression plot of the width of the axial lobe and pleural field of the pygidia of S. carcajouensis. (1) Axial lobe and pleural field at or adjacent to axial ring 1; (2) axial lobe and pleural field at or adjacent to terminal piece.

Three models of trunk growth of S. carcajouensis are interpreted here based on a cross plot of the overall length of the pygidia versus the number of axial rings (Fig. 15). In the first two of these models, larger pygidia that have fewer axial rings than smaller ones are inferred to indicate an episode of segment release following molting.

Figure 15. Length (sag.) of pygidia as a function of the number of axial rings in Sahtuia carcajouensis.

The first model supposes that each period of segment release is separated from the successive one by the occurrences of segment appearance (i.e., segment release at molt 1, followed by segment appearance at molt 2, and then by segment release at molt 3). Here, two periods of segment release are recognized: one during which the third thoracic segment is released from a pygidium containing six segments to one containing five, and the other in which the fourth segment is released from a pygidium containing eight segments to one containing seven. Therefore, ontogeny can only be reconstructed as far back as the meraspid stage bearing two thoracic segments (Fig. 16.1).

Figure 16. Models of growth of the trunk of Sahtuia carcajouensis. (1) Staggered segment release model; (2) continuous segment release model; (3) early segment release model.

The second model supposes that release of segments may have occurred in one unbroken sequence absent intervening periods of segment appearance. This may explain why the longest measured pygidia bearing five axial rings are longer than those bearing seven, although the dearth of specimens makes this uncertain. This model suggests there was an early period of continuous segment appearance from a pygidium that contains four segments to one that contains eight segments. This was then followed by the release of three segments, although whether this occurred in one, two, or three molting episodes is unknown. Finally, once S. carcajouensis attained four thoracic segments, segment release ceased, and segment appearance continued until the maximum number of trunk segments was attained (Fig. 16.2).

The third model supposes that the gaps in pygidial-length distribution relative to the axial ring number do not indicate the timing of segment release, but instead are regarded as products of size variation within the collection (Fig 16.3). If true, then the holaspid stage was attained at, or prior to, the appearance of the eighth trunk segment. All three models indicate that S. carcajouensis attained a stable number of thoracic segments prior to the onset of epimorphic growth, indicating the protarthrous mode (Hughes, 2007).

Mackenzieaspis parallelispinosa

Three cranidial morphs assigned to M. parallelispinosa are recognized based on RMA regression plots of length versus width (Fig. 17).

Figure 17. RMA regression plot of length vs. width of cranidia of Mackenzieaspis parallelispinosa, with interpreted cranidial morph groups: M. parallelispinosa Cranidial morph 1, M. parallelispinosa Cranidial morph 2, and M. parallelispinosa Cranidial morph 3.

Cranidial ontogeny: Cranidial morph 1

Figure 18.3.

Figure 18. Cranidial morphs 1 and 2, and mature pygidia of Mackenzieaspis parallelispinosa from the Mount Cap Formation. (1, 2) M. parallelispinosa Cranidial morph 2; (1) cranidium (dorsal) GSC 143703; (2) cranidium (dorsal) GSC 143704; (3) M. parallelispinosa Cranidial morph 1 (dorsal) GSC 143705; (4–7) M. parallelispinosa mature pygidia; (4) pygidium (dorsal, latex mold) GSC 143706; (5) pygidium (dorsal) GSC 143707; (6) pygidium (dorsal) GSC 143708; (7) pygidium (dorsal) GSC 143709. Note in specimens 4, 6, and 7, the border spines are broken.

Description

Cranidium is 1.31–1.57 mm long, 1.40–1.73 mm wide, and subquadrate in outline. Axial furrow is well defined, the glabella comprising 82–88% of the cranidial length (sag.) and extends to the anterior border furrow. Glabella narrows (trans.) gently from the occipital furrow to S1 and widens (trans.) gently from S1 to the anterior glabellar lobe. Three pairs of lateral glabellar furrows are present: S1 is well defined and oriented obliquely backwards 45–47° relative to the sagittal line; S2 is poorly to moderately defined and oriented obliquely backwards 57–58° relative to the sagittal line; S3 is poorly defined and oriented transversely. Occipital furrow is well defined, the occipital ring widening posteriorly. Anterior border furrow is moderately defined; the length of the border comprises 5% of the total sagittal length. Anterior branch of the facial suture is slightly convergent, curving adaxially near the anterior margin. Interocular area at the widest point (trans.) comprises 30–34% of the cranidial width. Palpebral lobe is crescent shaped, with the anterior tip merging into the ocular ridge and then intersecting the axial furrow at the anterior lobe, and the posterior tip opposite the occipital furrow. Length (exsag.) of the palpebral lobe is 56–66% that of the cranidium. Posterior branch of the facial suture is oriented transversely. Posterior border is oriented nearly transversely; width (trans.) is 75–78% that of the occipital furrow. Posterior border furrow is moderately defined.

Material

Five cranidia.

Remarks

Possible ontogenetic changes that occurred from gen. and sp. indet. Cranidial morph 2 to Mackenzieaspis parallelispinosa Cranidial morph 1 include the overall increase in size, proportional widening of the glabella, relative elongation of the palpebral lobe, transverse rotation of the posterior branch of the facial suture, and proportional narrowing of the posterior border. Vectors on landmarks 5 and 6 exhibit the greatest magnitude, indicating a narrowing of the interocular area by the proportional widening of the glabella (Fig. 19.1). Elongation of the palpebral lobe, indicated by the anterior and adaxial migration of landmark 5, and posterior and adaxial migration of landmark 6, was brought about by growth primarily in the anterior direction.

Figure 19. Bookstein-fitted, vector-on-landmark plot of the mean shape change of cranidial morphs of Mackenzieaspis parallelispinosa. Dots represent the mean landmark position of the smaller morph, whereas the vectors indicate the direction and magnitude of shape change to the mean landmark position of the larger morph. (1) Gen. and sp. indet. Cranidial morph 2 to M. parallelispinosa Cranidial morph 1; (2) M. parallelispinosa Cranidial morph 1 to M. parallelispinosa Cranidial morph 2; (3) M. parallelispinosa Cranidial morph 2 to M. parallelispinosa Cranidial morph 3.

Mackenzieaspis parallelispinosa Cranidial morph 1 is morphologically similar to the late meraspid cranidium of P. limbata (Hu, 1985b, pl. 21, figs. 17, 18, 21), and to a lesser extent to that of F. quadrangularis (Hu, 1985b, pl. 20, figs. 17, 18), differing from both species by its slightly narrower anterior border and preocular fixed cheek and less-divergent anterior branch of the facial suture, and from the latter by its slightly longer palpebral lobe. The later meraspid cranidia M2 and M3 of Z. palmeri (Hopkins and Webster, 2009, fig. 11.1–11.20) have shorter and narrower anterior glabellar lobes, slightly longer and wider anterior borders, more convergent anterior branches of the facial sutures, wider interocular areas and ocular ridges, slightly shorter and less-curving palpebral lobes, more obliquely oriented posterior branches of the facial sutures, and slightly wider posterior borders bearing longer metafixigenal spines. The early meraspid cranidia M-0 to M-3 of B. fimbriatus (Robison, 1967, pl. 24, figs. 6–9) and that of P. propinqua (Hu, 1971, pl. 10, figs. 14, 16–18) differ by their wider anterior glabellar lobes, wider interocular areas, shorter, less-curved palpebral lobes, more obliquely oriented posterior branches of the facial sutures, longer postocular fixed cheeks, and wider posterior borders with longer metafixigenal spines. Late meraspid cranidia M-4 and M-5 of B. fimbriatus (Robison, 1967, pl. 24, figs. 11, 13) and the late meraspid cranidium of P. propinqua (Hu, 1971, pl. 10, figs. 19, 20) differ from M. parallelispinosa Cranidial morph 1 by their wider anterior glabellar lobes, wider anterior borders, wider interocular areas, more obliquely oriented posterior branches of the facial sutures, longer posterior fixed cheeks, longer occipital rings, and wider posterior borders bearing longer metafixigenal spines. Furthermore, both species have shorter, less-curving palpebral lobes than those of M. parallelispinosa Cranidial morph 1, but this difference is greater in B. fimbriatus than in P. propinqua. The meraspid cranidium of G. boccar (Hu, 1985a, pl. 9, figs. 3, 7) has a slightly longer palpebral lobe, more obliquely oriented posterior branch of the facial suture, and a shorter posterior fixed cheek in comparison to M. parallelispinosa Cranidial morph 1. Mackenzieaspis parallelispinosa Cranidial morph 1 is broadly similar to early meraspid cranidia of F. fecunda (Öpik, 1982, pl. 12, figs. 5a, 5b, 6a–c), but differs by the narrower anterior glabellar lobe, longer palpebral lobe, shorter postocular fixed cheek, and shorter metafixigenal spine.

Cranidial ontogeny: Cranidial morph 2

Figure 18.1, 18.2.

Description

Cranidium is 2.22–2.24 mm long, 1.96–2.22 mm wide, and subquadrate in outline. Axial furrow is well defined, the glabella comprising 74–78% of the cranidial length (sag.) and extends to the anterior border furrow. Glabella narrows (trans.) from the occipital furrow to S1 and widens (trans.) from S1 to the anterior glabellar lobe. Four pairs of lateral glabellar furrows are present: S1 is moderately defined and oriented obliquely backwards 55–60° relative to the sagittal line; S2 is moderately defined and oriented obliquely backwards 63–66° relative to the sagittal line; S3 is poorly defined and oriented obliquely forwards 80–83° relative to the sagittal line; S4 is poorly defined and oriented obliquely forwards 54–68° relative to the sagittal line. Occipital furrow is well defined, the occipital ring widening posteriorly. Anterior border furrow is moderately to well defined; the length of the border comprises 3–6% of the total sagittal length. Anterior branch of the facial suture is slightly divergent anteriorly or parallel to the sagittal line. Interocular area at the widest point (trans.) comprises 25–26% of the cranidial width. Palpebral lobe is crescent shaped, with the anterior tip merging into the ocular ridge and then intersecting the axial furrow at S4, and the posterior tip opposite of the anterior occipital ring. Length (exsag.) of the palpebral lobe is 50–51% that of the cranidium. Posterior branch of the facial suture is oriented transversely, curving slightly posteriorly at the distal tip. Posterior border is oriented nearly transversely, in width (trans.) is 110% that of the occipital furrow and bears a short metafixigenal spine. Posterior border furrow is well defined.

Material

Two cranidia.

Remarks

Ontogenetic changes that occurred from M. parallelispinosa Cranidial morph 1 to M. parallelispinosa Cranidial morph 2 include the increase in overall size, proportional increase in the length of the glabella, slight elongation of the anterior border, addition of S4, abaxial rotation of the anterior branch of the facial suture, relatively slight shortening of the palpebral lobe, and relative widening of the posterior border. Landmarks 4–8 all exhibit vectors oriented dominantly posteriorly, brought about by the proportional elongation of the anterior portion of the glabella (Fig. 19.2). The abaxial portion of the vectors in landmarks 4–8 probably reflect slight widening of the glabella.

Mackenzieaspis parallelispinosa Cranidial morph 2 is similar to the early and late meraspid and early holaspid cranidia of P. limbata (Hu, 1985b, pl. 21, figs. 12–14, 16–18, 21, 24, 27), although differs by its more-expanded anterior glabellar lobe, shorter and narrower anterior border, and less-divergent anterior branch of the facial suture. The late meraspid cranidium of F. quadrangularis (Hu, 1985b, pl. 20, figs. 17, 18) is similar to M. parallelispinosa Cranidial morph 2, but differs by its slightly wider anterior glabellar lobe, wider anterior border, more-divergent anterior branch of the facial suture, wider interocular area, and slightly shorter palpebral lobe. The late meraspid cranidia M3 and post-M3 of Z. palmeri (Hopkins and Webster, 2009, fig. 11.1–11.20, 11.23, 11.26, 11.37–11.40, 11.44, 11.45) differ from M. parallelispinosa Cranidial morph 2 by their shorter and narrower anterior glabellar lobes, longer and wider anterior borders, wider interocular fixed cheeks and ocular ridges, wider posterior borders, and presence of occipital spines. The late meraspid cranidia M-4 and M-5 of B. fimbriatus (Robison, 1967, pl. 24, figs. 11, 13) differ from M. parallelispinosa Cranidial morph 2 by their wider anterior glabellar lobes, wider anterior borders, more-divergent anterior facial sutures, wider interocular areas, shorter, less-curved palpebral lobes, more obliquely oriented posterior facial sutures, longer postocular fixed cheeks, wider posterior borders, and longer metafixigenal spines. The early holaspid cranidium of B. fimbriatus (Robison, 1967, pl. 24, fig. 15) has a wider anterior glabellar lobe, slightly narrower interocular area, wider posterior border, and a longer occipital spine. The late meraspid cranidium of P. propinqua (Hu, 1971, pl. 10, figs. 19, 20) differs from M. parallelispinosa Cranidial morph 2 by its slightly wider anterior glabellar lobe and interocular area, slightly shorter palpebral lobe, more obliquely oriented posterior branch of the facial suture, slightly longer postocular fixed cheek, and a slightly longer metafixigenal spine. The early holaspid cranidium of that species (Hu, 1971, pl. 10, figs. 21–23) differs by its slightly wider anterior glabellar lobe and slightly narrower interocular area. The holaspid cranidium of G. boccar (Hu, 1985a, pl. 9, fig. 10) differs from M. parallelispinosa Cranidial morph 2 by its narrower interocular fixed cheek, slightly shorter palpebral lobe, and wider posterior border. Mackenzieaspis parallelispinosa Cranidial morph 2 differs from the smaller late meraspid cranidium of F. fecunda (Öpik, 1982, pl. 12, figs. 7–9) by its shorter anterior border, narrower anterior glabellar lobe, shorter palpebral lobe, slightly shorter postocular fixed cheek, and narrower posterior border bearing a shorter metafixigenal spine. The medium-sized late meraspid cranidium of F. fecunda (Öpik, 1982, pl. 12, figs. 10, 11) differs from M. parallelispinosa Cranidial morph 2 by having a longer anterior border, slightly wider anterior glabellar lobe, slightly longer palpebral lobe and postocular fixed cheek, and slightly wider posterior border with a small metafixigenal spine. Compared to M. parallelispinosa Cranidial morph 2, the largest late meraspid cranidium of F. fecunda (Öpik, 1982, pl. 13, figs. 1–3a) has a longer anterior border and preocular fixed cheek, wider glabella, shallower lateral glabellar furrows, straighter axial furrow, an occipital ring bearing a spine, and a wider posterior border bearing a longer metafixigenal spine.

Cranidial ontogeny: Cranidial morph 3

Figure 20.1–20.3.

Figure 20. Complete or nearly complete exoskeletons of Mackenzieaspis parallelispinosa with Cranidial morph 3 and mature pygidia from the Mount Cap Formation. (1) Holotype complete exoskeleton (dorsal) GSC 142404, note the left spine is broken off in this specimen; (2) paratype exoskeleton lacking one free cheek (dorsal, latex mold) GSC 142406; (3) paratype complete exoskeleton (dorsal) GSC 142408, note both pygidial spines are broken off in this specimen.

Description

Cranidium is 2.91–10.91 mm long, 2.76–9.29 mm wide, and subquadrate in outline. Axial furrow is well defined, the glabella comprising 76–85% of the cranidial length (sag.) and extends to the anterior border furrow. Glabella is subrectangular in outline. Four pairs of lateral glabellar furrows are present: S1 is moderately defined and oriented obliquely backwards 52–55° relative to the sagittal line; S2 is moderately defined and oriented obliquely backwards 64–66° relative to the sagittal line; S3 is poorly defined and oriented transversely; S4 is poorly defined and oriented obliquely forwards 72–77° relative to the sagittal line. Occipital furrow is well defined, the occipital ring widening posteriorly and bearing a medial tubercle. Anterior border furrow is moderately to well defined; the length of the border comprises 5–8% of the total sagittal length. Anterior branch of the facial suture is parallel to the sagittal line, curving adaxially near the anterior margin. Interocular area at the widest point (trans.) comprises 22–28% of the cranidial width. Palpebral lobe is crescent shaped, with the anterior tip merging into the ocular ridge and then intersecting the axial furrow at S4, and the posterior tip opposite of the occipital furrow. Length (exsag.) of the palpebral lobe is 48–56% that of the cranidium. Posterior branch of the facial suture is oriented transversely, curving slightly posteriorly near the distal tip. Posterior border is oriented nearly transversely, in width (trans.) is 115–125% that of the occipital furrow and bears a short metafixigenal spine. Posterior border furrow is well defined.

Material

Seven complete or nearly complete exoskeletons, three exoskeletons lacking free cheeks, and three cranidia.

Remarks

Ontogenetic changes that occurred between M. parallelispinosa Cranidial morph 2 and M. parallelispinosa Cranidial morph 3 include the increase in overall size, relative widening of the glabella, appearance of the medial occipital tubercle, relatively slight elongation of the anterior border, and proportional widening of the posterior border. Vectors on landmarks have low magnitudes, suggesting little shape change of the examined landmarks between M. parallelispinosa Cranidial morph 2 and M. parallelispinosa Cranidial morph 3 (Fig. 19.3). Minor abaxial migration of landmarks 4, 7, and 8 indicates slight proportional widening of the glabella. The slight anterior component of vectors 4, 5, and 6 indicates that the palpebral lobe underwent slight rotation in the anterior direction. These specimens are regarded as holaspides because increasing exoskeleton sizes consistently retain four thoracic segments.

Mackenzieaspis parallelispinosa Cranidial morph 3 is similar to holaspid cranidia of Albertelloides eliasi Handkamer and Pratt in Handkamer et al., 2022 (Handkamer et al., 2022, fig. 22.4, 22.5, 22.7) and A. mischi Fritz, 1968 (Fritz, 1968, pl. 38, figs. 1–3, 7; Palmer and Halley, 1979, pl. 10, figs. 8, 9, 13; Handkamer et al., 2022, figs. 21.7, 21.8, 21.10, 22.1, 22.2). However, M. parallelispinosa Cranidial morph 3 differs from A. eliasi by having a slightly shorter postocular fixed cheek and smaller occipital tubercle, and differs from A. mischi by having a narrower glabella, slightly longer anterior border, less-curved anterior branch of the facial suture, narrower posterior border, and having an occipital tubercle instead of a spine. Mackenzieaspis parallelispinosa Cranidial morph 3 differs from P. limbata by its slightly less expanded anterior glabellar lobe, parallel, as opposed to divergent, anterior branch of the facial suture, shorter and narrower anterior border, and shorter occipital ring (Rasetti, 1951, pl. 18, figs. 8–10, 14–16; Hu, 1985b, pl. 21, figs. 24, 27). The holaspid cranidium of F. quadrangularis (Hu, 1985b, pl. 20, figs. 20, 21, 27) is also similar, but differs from M. parallelispinosa Cranidial morph 3 by its wider anterior glabellar lobe, wider anterior border, more-divergent anterior branch of the facial suture, slightly shorter palpebral lobe, and narrower posterior border. The holaspid cranidium of Z. palmeri (Hopkins and Webster, 2009, figs. 2.3, 2.4, 2.7–2.17, 2.20–2.26, 2.28–2.33, 3.1–3.20, 4.2–4.11, 4.13, 4.25–4.30, 4.32, 5.1–5.8, 5.15–5.21) differs from M. parallelispinosa Cranidial morph 3 by its shorter and narrower glabella, longer and wider anterior border, wider interocular area and ocular ridge, and longer occipital spine. Mackenzieaspis parallelispinosa Cranidial morph 3 differs from the early holaspid cranidium of B. fimbriatus (Robison, 1967, pl. 24, fig. 25) by its slightly narrower anterior glabellar lobe, wider interocular area, slightly longer palpebral lobe, slightly shorter postocular fixed cheek, and narrower posterior border. It differs from the holaspid cranidium of P. propinqua (Resser, 1939b, pl. 3, figs. 35, 36; Hu, 1971, pl. 10, figs. 21–23) by the slightly narrower anterior glabellar lobe, slightly shorter palpebral lobe, and slightly narrower posterior border. Mackenzieaspis parallelispinosa Cranidial morph 3 differs from the holaspid cranidium of G. boccar (Walcott, 1916, pl. 52, figs. 1, 1a–f), by its wider interocular area, longer palpebral lobe, and transversely oriented posterior branch of the facial suture as opposed to oriented obliquely backwards. It differs from the latest meraspid and holaspid cranidia of F. fecunda (Öpik, 1982, pl. 6, fig. 1, pl. 7, figs. 1–3, pl. 8, figs. 4–5, pl. 13, figs. 1–3a) by its shorter and narrower anterior border, narrower glabella, shorter occipital ring with a tubercle instead of a spine, and wider posterior border bearing a longer metafixigenal spine.

Trunk ontogeny

Pygidia (Figs. 18.4–18.7, 20.1–20.3) range from as small as 0.78 mm long and 1.53 mm wide to as large as 12.85 mm long and 16.82 mm wide (Fig. 21). There is a distinct change in shape from nearly elliptical to semicircular and they can have four to eight axial rings and four to five pleural furrows. Log-transformed data of the RMA regression plots of the allometric trajectories of pygidia indicate that the anterior pleural lobes, relative to the adjacent axial lobe at anteriormost pleural lobe underwent isometric growth transversely (Fig. 22.1), whereas the posterior pleural lobe adjacent to the terminal piece underwent negative allometric growth (Fig. 22.2). The border spine orientation diverges slightly during ontogeny and then converges again later in development.

Figure 21. RMA regression plot of length vs. width of pygidia of Mackenzieaspis parallelispinosa.

Figure 22. Log-transformed RMA regression plot of the width of the axial lobe and pleural field of the pygidia of Mackenzieaspis parallelispinosa. (1) Axial lobe and pleural field at or adjacent to axial ring 1; (2) axial lobe and pleural field at or adjacent to terminal piece.

The gaps in size distribution of the pygidial length relative to the number of axial rings suggest that there are several possible models of trunk growth and segment release for M. parallelispinosa (Fig. 23), similar to those interpreted for S. carcajouensis. However, the positioning of the pygidial border spine, considered to demarcate a homologous segment in all pygidia, indicates that even the smallest pygidium collected (Fig. 18.5) belongs to a holaspis. This is because both the smallest and largest pygidia contain three segments that are anterior to the one that bears the spine, and therefore, thoracic segment release had already ceased for those specimens. This suggests that gaps in the size distribution reflect size variation within the collection instead of segment release events. Also, it indicates that all thoracic segments had been released from the pygidium prior to development of the ninth trunk segment, similar to the early segment release model for S. carcajouensis (Fig. 16.3). The growth of M. parallelispinosa can only be incompletely reconstructed (Fig. 24), and the scheduling of early segment appearance and release episodes are currently unknown. Following the appearance of the eighth trunk segment and the release of the four thoracic segments, segment appearance continued at the subterminal generative zone (Hughes et al., 2006). Mackenzieaspis parallelispinosa achieved a stable number of thoracic segments prior to the onset of epimorphic growth, indicating a protarthrous mode (Hughes, 2007).

Figure 23. Length (sag.) of pygidia as a function of the number of axial rings in Mackenzieaspis parallelispinosa.

Figure 24. Model of growth of the trunk of Mackenzieaspis parallelispinosa, similar to the early segment release model of Sahtuia carcajouensis in Figure 16.3. The last episode of segment appearance varies intraspecifically.

Discussion

Cranidial development

The cranidial development of Sahtuia carcajouensis (Fig. 25) and Mackenzieaspis parallelispinosa (Fig. 26) shows that the morphologies of the earliest juvenile stages, belonging to either or both species, do not differ significantly from those of other zacanthoidids and dolichometopids. The similar morphology of genus and species indeterminant Cranidial morph 1 with the interpreted protaspid stages of Bathyuriscus fimbriatus, Ptarmiganoides propinqua, Glossopleura boccar, Paralbertella limbata, Fieldaspis quadrangularis, Zacanthopsis palmeri, and Fuchouia fecunda indicates an essentially identical early ontogeny shared by all these species. Correspondingly, the larger gen. and sp. indet. Cranidial morph 2 is broadly similar to subsequent stages of these same species, although there are some minor differences in the width of the anterior glabellar lobe, length of the palpebral lobe, width of the interocular area, and development of metafixigenal spines. Some species of Oryctocephalidae also have similar protaspides (Hou et al., 2015). This agrees with previous observations that the families in Corynexochida share a common protaspid morphology (Chatterton and Speyer, 1997), just as do many species of Cambrian ptychoparioids, for which protaspides are known, such as those of Crassifimbra walcotti (Resser, 1937) (Palmer, 1958), Amecephalus sp., Plagiura cercops (Walcott, 1917b), Kochina elongata Hu, 1985a, and Chancia conica Hu, 1985a (Hu, 1985a). By contrast, the earliest recognized juvenile specimens of both S. carcajouensis and M. parallelispinosa indicate a divergence in ontogenetic trajectory. Mackenzieaspis parallelispinosa underwent more rapid growth of the palpebral lobe relative to S. carcajouensis (Fig. 27.1), assuming size is a proxy for age of the specimens, whereas S. carcajouensis underwent more rapid widening of the glabella, particularly the anterior glabellar lobe (Fig. 27.2, 27.3).

Figure 25. Illustration of the development of the cranidium and pygidium of Sahtuia carcajouensis. Illustrations are not to scale.

Figure 26. Illustration of the development of the cranidium and pygidium of Mackenzieaspis parallelispinosa. Illustrations are not to scale.

Figure 27. RMA comparative linear morphometrics of the growth of the palpebral lobes and glabella of Sahtuia carcajouensis (S) (red dots and regression) and Mackenzieaspis parallelispinosa (M) (blue dots and regression). (1) Length of glabella vs. length of palpebral lobe; (2) width of interocular area vs. anterior glabellar lobe width; (3) width of interocular area vs. glabella width slightly anterior of occipital furrow.

On the other hand, both species exhibit proportional growth of other cranidial elements that are similar to other zacanthoidid and dolichometopid taxa. For example, the palpebral lobe of S. carcajouensis Cranidial morph 1 has a similar proportion to those of the early and late meraspides of B. fimbriatus and F. fecunda. However, the palpebral lobes of S. carcajouensis Cranidial morphs 2 and 3 have similar proportions to those of early holaspides of B. fimbriatus, but they are shorter than those of the latest meraspides and holaspides of F. fecunda. Glossopleura boccar, a dolichometopid, developed a long palpebral lobe relatively early in its ontogeny, like the zacanthoidids M. parallelispinosa, P. limbata, and Z. palmeri, but unlike the zacanthoidids F. quadrangularis and P. propinqua. Other morphological features appear to be shared by species of both families: Mackenzieaspis parallelispinosa developed a subquadrate glabella at Cranidial morph 1, like those of P. limbata, Z. palmeri, and G. boccar in their early meraspid stages. This glabellar shape did not develop until the late meraspid stage in F. fecunda. By contrast, the glabellas of S. carcajouensis, B. fimbriatus, P. propinqua, and F. quadrangularis are anteriorly expanded through ontogeny, although the degree of expansion may vary between species and ontogenetic stages. It has been suggested that the presence of metafixigenal spines distinguishes zacanthoidids from dolichometopids (Palmer and Halley, 1979). However, metafixigenal spines are present in S. carcajouensis Cranidial morph 1, as well as on the early meraspides of B. fimbriatus and F. fecunda, but they are absent in S. carcajouensis Cranidial morph 2 and early and late holaspides of B. fimbriatus. The metafixigenal spine is retained throughout all ontogenetic stages of F. fecunda.

The cranidial morphs of both species, and the interpreted ontogenetic changes that occurred between them, do not exhibit any novel types of development for zacanthoidids and dolichometopids. Sahtuia carcajouensis Cranidial morphs 1, 2, and 3 are broadly similar in size and in most morphological features to the early and late meraspid and early holaspid cranidia of B. fimbriatus, although the late holaspid cranidium of that species exhibits a narrower interocular area, slightly longer palpebral lobe, long occipital spine, and wider posterior border. Sahtuia carcajouensis Cranidial morph 3 also bears an even closer resemblance to the holaspid cranidium of E. mackenziensis, and S. carcajouensis Cranidial morph 1 has a similar size and morphology as the only juvenile specimen collected of E. mackenziensis (Handkamer et al., 2022, fig. 13.9). Mackenzieaspis parallelispinosa Cranidial morphs 1, 2, and 3 show an ontogenetic trajectory similar to early and late meraspid and holaspid stages of P. limbata, differing in the shape, size, and development of the anterior glabellar lobe and anterior cranidial border. Mackenzieaspis parallelispinosa Cranidial morph 3 also closely resembles the holaspid cranidium of Albertelloides eliasi, and slightly less so to A. mischi. However, the ontogenies of those species are unknown.

Trunk development

In contrast to the growth of their respective cranidia, S. carcajouensis (Fig. 25) and M. parallelispinosa (Fig. 26) both exhibit novel and nearly identical trunk development within Dolichometopidae and Zacanthoididae, respectively. Each species has four thoracic segments, with nine pygidial segments maximally in the former, and either seven or eight maximally in the latter. Reconstructions reveal that each underwent a protarthrous mode of growth. If the early thoracic segment release model for S. carcajouensis is correct (Fig. 16.3), they may even share a similar schedule of segment appearance and segment release: both accumulating four thoracic segments and four pygidial segments, followed by the termination of segment release and the onset of the holaspid stage, and then by the accumulation of more segments in their pygidia. The incorporation of mathematical growth models may help elucidate this (Hopkins, 2020).

Only one other known zacanthoidid species has four thoracic segments, Dodoella kobayashii Handkamer and Pratt in Handkamer et al., 2022, which occurs rarely in slightly older strata of the Mount Cap Formation at Dodo Canyon (Fig. 1). However, the cranidial morphology of D. kobayashii differs in that it resembles that of species of Albertella Walcott, 1908, such as A. helena Walcott, 1908, as well as species belonging to Caspimexis Özdikmen, 2005 (an objective senior synonym for Mexicaspidella Handkamer and Pratt in Handkamer et al., 2022, proposed as a replacement name for Mexicaspis, which was preoccupied).

The oryctocephalid Balangia balangensis Qian, 1961, possesses four thoracic segments and a relatively large pygidium. Species of Tonkinella Mansuy, 1916, and Microryctocara nevadensis Sundberg and McCollum, 1997, also oryctocephalids, have five thoracic segments and relatively large pygidia. However, they all possess cranidia that have been interpreted as paedomorphic (McNamara, 1986b; McNamara et al., 2006), whereas cranidia of the holaspides of S. carcajouensis and M. parallelispinosa do not exhibit paedomorphism. An extreme example of a low number of thoracic segments is Thoracocare minuta (Resser, 1939a), which probably belongs to Oryctocephalidae (Robison and Campbell, 1974; Sundberg, 2018), although this has been debated (McNamara, 1986b). Thoracocare minuta, like the other oryctocephalids, has a paedomorphic cranidium and pygidium, and is considered to have evolved for a pelagic lifestyle (Robison and Campbell, 1974). Thus, these examples of oryctocephalids with an abnormal number or distribution of trunk segments are not directly comparable to S. carcajouensis or M. parallelispinosa.

Phylogenetic positions of Sahtuia carcajouensis and Mackenzieaspis parallelispinosa

The discrimination of Zacanthoididae and Dolichometopidae has been regarded as arbitrary (Rasetti, 1951) or even invalid (Whittington, 2009; Robison and Babcock, 2011). Palmer (in Palmer and Halley, 1979) suggested that spinosity of the pygidium and the presence of a metafixigenal spine on the posterior cranidial border separate the two groups, zacanthoidids having both features and dolichometopids lacking them. Although this classification was adopted (Handkamer et al., 2022), it is contentious because: (1) some species of both families have a metafixigenal spine during one or more stages of their ontogeny; (2) some dolichometopid species with a non-spinous pygidium, such as F. fecunda, have a metafixigenal spine during their holaspid stage; (3) some zacanthoidid species with a spinous pygidium and thorax, such as Albertella helena Walcott, 1908, do not bear a metafixigenal spine; and (4) some dolichometopid species, such as B. fimbriatus, can have a spinous pygidium early in ontogeny, or exhibit interspecific variation in spinosity, like B. adaeus Walcott, 1916, or exhibit intraspecific variation in spinosity, such as the small spines that are present on some specimens of E. mackenziensis. Thus, the characters used to discriminate these families are unreliable, although accepted here for the sake of discussion.

The monophyly of Corynexochida has been called into question as well. A Bayesian- and parsimony-based phylogeny of Cambrian trilobite families (Paterson et al., 2019) indicates that the order may be polyphyletic, but this study did not test the monophyly of each family. Furthermore, it should be noted that a contrasting topology and position of corynexochid taxa in the parsimony- and Bayesian-based trees (Paterson et al., 2019, figs. S1, S2) do not endow confidence in their topological location.

Here, S. carcajouensis and M. parallelispinosa are considered to be not part of a monophyletic group, and instead the character of four thoracic segments is hypothesized as homoplasious. Two species similar to S. carcajouensis and M. parallelispinosa are proposed to have been their ancestors (i.e., sister taxa on a phylogenetic tree). Eobathyuriscus mackenziensis is hypothesized as the ancestor of S. carcajouensis, and either A. eliasi or A. mischi as the ancestor of M. parallelispinosa, because: (1) each hypothesized ancestor resembles its respective descendant in most morphological features of the cranidium, free cheek, certain features of the trunk, and number of total trunk segments; (2) the only trunk character that is shared by both descendent species is the presence of four thoracic segments; (3) the cranidial development of each descendent species is more akin to other taxa in their respective families than to each other; and (4) the presence of both hypothesized ancestor species in older strata of the Mount Cap Formation (Albertelloides mischi Zone [= Mexicella mexicana Zone]) immediately below strata that contain S. carcajouensis and M. parallelispinosa. McNamara et al. (2006) recognized points 1 and 4 as evidence of the ancestor–descendent relationship of B. balangensis and its interpreted ancestor, Duyunaspis duyunensis Chang and Chien in Zhou et al., 1977. However, in this case, the similarity in cranidial morphology is between the holaspid cranidium of B. balangensis and the meraspid cranidium of instar stage 4 of D. duyunensis. This suggests that, along with the trunk, development of the cranidium was slower for B. balangensis than for D. duyunensis.

The starkest distinction between each probable ancestor and descendant is the distribution of their trunk segments (Fig. 28), but there are a few others. Relative to E. mackenziensis, S. carcajouensis has a slightly shorter palpebral lobe and wider pleural fields of both the thorax and pygidium. Relative to A. eliasi, M. parallelispinosa has a narrower postocular fixed cheek, a slightly wider pygidial pleural field (the width of the thoracic pleurae are equivalent, except for the spines), and smaller axial tubercles. A complete thorax of A. eliasi is not known, but most other zacanthoidid genera of the same age do not exhibit interspecific variation in the number of thoracic segments (e.g., for Albertella and Paralbertella; Palmer and Halley, 1979). Mackenzieaspis parallelispinosa differs from A. mischi by having a narrower glabella and slightly longer anterior cranidial border, a less-curved anterior branch of the facial suture, narrower posterior border, no occipital spine, a slightly narrower thoracic and pygidial axis, and a longer pygidial border. The pygidial border spines of M. parallelispinosa and both species of Albertelloides Fritz, 1968, are present on the same pygidial segment, but the trunk segmentation is different: all three species have the spine on the fourth segment of the pygidium, but this corresponds to the eighth trunk segment in M. parallelispinosa and twelfth trunk segment in A. mischi and probably A. eliasi. To further support the hypothesis here, a phylogenetic analysis is required to interpret the sister-taxa and ancestor–descendent relationships (Adrain and Chatterton, 1994) that incorporates S. carcajouensis, M. parallelispinosa, E. mackenziensis, A. eliasi, A. mischi, and the type species of other zacanthoidid and dolichometopid genera.

Figure 28. Illustration of the morphology and allocation of the trunk segments of the hypothesized ancestor and descendant pairs. (1) Segment allocation in Eobathyuriscus mackenziensis and Sahtuia carcajouensis; (2) segment allocation in Albertelloides eliasi and Mackenzieaspis parallelispinosa. A complete thorax of A. eliasi has not been collected, therefore the total number of thoracic segments and morphologies of the first three trunk segments are unknown. Their morphology is inferred from A. mischi (Palmer and Halley, 1979, pl. 10, fig. 9), but the segments known from both A. mischi and A. eliasi are similar, albeit the pleural spines in A. eliasi are slightly longer.

Developmental changes in Sahtuia carcajouensis and Mackenzieaspis parallelispinosa

Reduction in the number of incidents of segment release, in each ancestor–descendant pair, was probably brought about by a timing modification (Webster and Zelditch, 2005), indicating heterochrony. Heterochrony is recognized as having a major influence on the evolutionary history of corynexochids. It was suggested that Corynexochida arose from Ptychopariida (Robison, 1967; Chatterton and Speyer, 1997) through neoteny, producing paedomorphism in the former clade. In the view that Corynexochida is probably polyphyletic (Paterson et al., 2019), this would have had to occur in multiple lineages. Meraspides of ptychoparioids are broadly similar in morphology to holaspides of corynexochids, typically with an elongate glabella terminating adjacent to a short anterior border, long palpebral lobe, conterminant hypostome, low number of thoracic segments, and a relatively large pygidium (e.g., Crassifimbra walcotti; Palmer, 1958, pl. 26, figs. 3, 4, 7, 8, 11, 12; Robison, 1967, table 1, fig. 5).

Examples of heterochrony have been documented in other groups of corynexochids. Paedomorphism has been recognized as important developmentally in the evolutionary history of the zacanthoidids from the Carrara Formation of the Great Basin (Palmer and Halley, 1979). There, the stratigraphic succession of species seems to coincide with gradually more paedomorphic forms, which was termed a paedomorphocline by McNamara (1986b). Features recognized as paedomorphic in that group include a reduction in exoskeletal size and number of thoracic segments, a shorter anterior cranidial border, a more convergent anterior branch of the facial suture, a wider interocular area, and a shorter palpebral lobe. This gradual evolution of paedomorphism was regarded to have arisen through progenesis (McNamara, 1986b).

Heterochrony also has been interpreted as a mechanism in the evolutionary history of oryctocephalids (McNamara, 1986b, 2009; MacNamara et al., 2003, 2006). McNamara (1986b) recognized a paedomorphocline in Laurentian oryctocephalids, from the primitive Lancastria Kobayashi, 1935, to the seemingly derived Tonkinella, and interpreted this as arising from progenesis. McNamara et al. (2006) suggested that B. balangensis arose from D. duyunensis, both from the Balang Formation of South China, by the complex interaction between heterochrony of segment appearance and thoracic segment release. A similar mechanism was suggested to have influenced a putative paedomorphocline of several species of Oryctocephalidae, also from the same formation (McNamara et al., 2003). McNamara et al. (2003) identified these as different species of Arthricocephalus Bergeron, 1899: Arthricocephalus chauveaui Bergeron, 1899, A. xinzhaiheensis Qian and Lin in Lu et al., 1974, A. balangensis Lu and Qian in Yin and Li, 1978, and A. pulchellus Zhang and Qian in Zhang et al., 1980. McNamara et al. (2003, 2006) were able to interpret these developmental schedules because of the availability of articulated juvenile specimens. However, these taxa were later re-identified by Dai et al. (2021) as Oryctocarella duyunensis, O. balangensis Lu and Qian in Yin and Li, 1978, A. xinzhaiheensis, and A. chauveaui, respectively, indicating that there may have been more complexity to the ancestor–descendant relationships than was previously considered.

With the corynexochids in the Mount Cap Formation, however, the classical mechanisms of paedomorphosis (progenesis, neoteny, post-displacement) cannot be applied with certainty. Progenesis, or the earlier onset of maturation, usually produces a size reduction in the descendent taxon, yet S. carcajouensis and M. parallelispinosa both exhibit similar maximum exoskeletal sizes relative to their hypothesized ancestors. Also, it is not conclusive that trilobites achieved sexual maturation at a perceived point of morphological maturation (Hughes et al., 2006). However, the abrupt changes in the growth trajectories in some species of ellipsocephaloids (Laibl et al., 2015; Holmes et al., 2021a, b) have been suggested as evidence of the onset of sexual maturation, either during the transition from anamorphic to epimorphic growth, or at the onset of the holaspid stage (Holmes et al., 2021a). One specimen of Olenoides serratus Rominger, 1887, has been shown to have two pairs of unique appendages below the posteriormost thoracic segment and the anteriormost pygidial segment (Losso and Ortega-Hernández, 2022, fig. 1). These were interpreted as serving a reproductive function, similar to claspers in horseshoe crabs (Losso and Ortega-Hernández, 2022), and it is possible that following the transition to the holaspid stage, these limbs adjacent to the thorax–pygidium boundary were modified, and thus, sexual maturity was reached.

Neoteny, or delayed development, cannot be applied due to the lack of complete juvenile exoskeletons, which could prove or disprove whether the mature, descendent taxon resembles the juvenile, ancestor taxon (e.g., evaluating if the holaspis of S. carcajouensis resembles meraspid instar 4 of E. mackenziensis, albeit larger). Post-displacement cannot not be applied (contra Handkamer et al., 2022) because it depends on the delayed onset of development of one or more structures. Here, this cannot be tested due to the lack of articulated specimens, which would ascertain if the first segment-release event occurred after more segments had accumulated in the pygidium of the descendent taxon than had accumulated in that of the ancestor taxon.

Thus, the unique trunk morphologies of S. carcajouensis and M. parallelispinosa are only interpreted as arising from timing modifications brought about by heterochrony. If articulated juveniles of both ancestor and descendent taxa can be collected, a rate modification can be tested for by documenting the number of segments released into the thorax relative to each molt (Webster and Zelditch, 2005). In this way, it can be determined if, for example, S. carcajouensis had, in the same number of molts as E. mackenziensis, released half as many thoracic segments. Other possible alterations to development may have occurred alongside these modifications to the timing of segment release. Almost all of the other morphological differences between the hypothesized ancestor–descendant pairs appear to show spatial variation in the development of certain structures, probably brought about by allometric repatterning (Webster and Zelditch, 2005). The exception to this is the substitution of a long occipital spine in A. mischi by an occipital tubercle in M. parallelispinosa, which may indicate heterotypy (Webster and Zelditch, 2005).

Ecological causation for the suppression of segment release

Although their ancestry is yet to be proven, it is considered likely that S. carcajouensis and M. parallelispinosa arose from populations of E. mackenziensis and A. eliasi or A. mischi, respectively, expressed by the decrease in the number of episodes of segment release. This leads to two questions: (1) why did these timing modifications to segment release occur; and (2) why are the two heterochronic species prolific at Carcajou Falls and essentially absent elsewhere in basin? The general, although not universal, reduction in segment release of the trunk region, or caudalization, gradually through the Paleozoic has been suggested to be an adaptive response to increase efficiency of enrollment, or protection from predators (Raymond, 1920; Stubblefield, 1959; Bruton and Haas, 1997; Hughes, 2007; Esteve et al., 2011, 2013; Ortega-Hernández et al., 2013; Dai et al., 2019; Hopkins and To, 2022). Oryctocephalids of the Balang Formation of South China seem to be an example of this (McNamara et al., 2003, 2006; McNamara, 2009), because increased convexity, and subsequently tighter enrollment, was achieved in the descendent taxon by the narrowing of the width (trans.) between the axis and the thoracic fulcra. The fulcrum is the point on the posterior margin of each thoracic segment, where it changes from horizontal to downward flexing. By narrowing the distance to the fulcra and widening the distal portion of the thoracic pleura, it has been proposed that the axial portion of the exoskeleton would have been raised relative to the sediment surface. This would have exposed the posterior part of the trilobite, requiring elongation and inclination of the pygidium to compensate for the lost protection (McNamara et al., 2006; McNamara, 2009).

However, this does not adequately explain the adaptive rationale that must have laid behind the evolution of S. carcajouensis or M. parallelispinosa, because no specimens were observed enrolled or partially enrolled, and their morphology does not indicate a propensity for enrollment (Ortega-Hernández et al., 2013). Relative to their presumed ancestors, neither M. parallelispinosa nor S. carcajouensis have proportionally narrower proximal portions of their pleurae between their axes and fulcra, which in the latter species is wider. Also, the descendent species are not more convex relative to the presumed ancestors. No exoskeletal architecture was identified in both descendent taxa that may have reduced shear strain during enrollment (Ortega-Hernández et al., 2013). Another possible selective pressure may have been to reduce the number of articulated segments that were targeted by predators (Hughes, 2003). However, the endemicity of these species in the Mount Cap Formation does not support a simple response to predation, because enhanced defense was presumably a universal pressure on Cambrian trilobites and global in scale (e.g., Babcock, 1993; Pratt, 1998; Nedin, 2004; Bicknell and Paterson, 2018; Bicknell et al., 2022). If the development of this morphology in both S. carcajouensis and M. parallelispinosa had optimized and improved their inherent defense, it would be expected that parallel changes would be present in other elements in this fauna, as well as elsewhere in Laurentia. Thus, predator–prey interaction does not adequately explain these developmental changes.

As an exemplar, the Silurian trilobite Aulacopleura koninckii (Barrande, 1846) shows a degree of intraspecific variation in its number and distribution of trunk segments: 18–22 segments comprise the thorax and 3–6 segments comprise the pygidium (Hughes and Chapman, 1995). Also, there is no correlation between the size of the specimen and the number of segments distributed to each region of the trunk (Hughes and Chapman, 1995). Variance in the total number of trunk segments and their distribution indicates intraspecific modifications to the timing of segment release and segment appearance (Webster and Zelditch, 2005). Specimens with more thoracic segments had proportionally shorter segments, suggesting that there was an alteration to their growth trajectory, and therefore, allometric repatterning (Webster and Zelditch, 2005). Intraspecific variation in the number of episodes of segment appearance and segment release also has been documented in other ‘olenimorph’ trilobites (Hughes et al., 1999), indicating that this may have been characteristic of that morphotype and determined by their respective ecological niche. Olenimorph trilobites typically have many thoracic segments with wide pleurae, small pygidia, and are usually present in strata interpreted as deep-water, oxygen-poor settings. Although the olenimorphs exhibit different morphology, trajectory of heterochronic development, and apparent paleoenvironmental preference than those of S. carcajouensis and M. parallelispinosa, they nonetheless demonstrate that changes to development can be controlled by paleoecological conditions.

The hypothesized, coincident modifications to the timing of segment release may be due to the specific paleoecological conditions that were present only at Carcajou Falls, and hence accounting for the absence of these species elsewhere. Sedimentological and biostratigraphic evidence (Handkamer et al., 2022) indicates that this locality was closer to the western paleoshoreline than other sections in the eastern Mackenzie Mountains, and the fossiliferous strata there suggest it was more favorable to a skeletonized benthos than other areas of the Cambrian basin. Preliminary analysis of trace elements in the silty mudstone facies indicates that this locality may have been deposited under well-oxygenated (cf., Tribovillard et al., 2006; Kimmig and Pratt, 2016), occasionally brackish-water (cf., Wei and Algeo, 2020) conditions, with a high flux of organic matter (cf., Tribovillard et al., 2006). The Mount Cap Formation across the Mackenzie Plain depocentre (MPD), and elsewhere, lacks echinoderm bioclasts and fossils, which are common in rocks deposited under normal-marine conditions. Their absence here, relative to their presence in synchronous formations of Laurentia (Palmer and Halley, 1979; Aitken, 1997; Pratt and Bordonaro, 2007), could indicate variable salinity conditions in the basin. Submarine halite deposits in the Mackenzie Trough (MacLean and Cook, 1999) also may suggest that salinity in the deepest part of the basin was above normal-marine levels, contrasting the brackish conditions at Carcajou Falls. The occurrence of alginites and acritarchs in the Mount Cap Formation at Dodo Canyon (Stasiuk, 2005) and the Colville Hills (Dixon and Stasiuk, 1998), part of the more northern Good Hope depocentre (Sommers et al., 2020), support the geochemical evidence of high nutrient flux.

Preliminary geochemical data suggest that the MPD, and possibly the whole Cambrian basin, during the Albertelloides mischi and Glossopleura walcotti zones, was characterized by variable paleoenvironmental conditions. Carcajou Falls was likely less hostile than other settings, as indicated by the prolific epibenthic fauna there, but possible variation in salinity and nutrient flux are suggestive of seasonal fluctuations like those in modern epicontinental basins (Szaniawska, 2018), and the fauna would need to adapt to tolerate fluctuating conditions. The timing modification to segment release may have caused early onset of the holaspid stage, and possibly sexual maturity, as an adaptation to compensate for a high mortality rate. Another possibility is that organic matter, and concomitant phosphorus concentrations, may have had an effect on the water chemistry and subsequently the metabolic rates of the fauna there.

Conclusions

The corynexochid trilobites Sahtuia carcajouensis and Mackenzieaspis parallelispinosa from the middle Cambrian Mount Cap Formation at Carcajou Falls, eastern Mackenzie Mountains of northwestern Canada, both exhibit a trunk morphology that is unique among dolichometopids and zacanthoidids: holaspides have a thorax composed of four segments and a pygidium composed of nine segments in S. carcajouensis and seven or eight segments in M. parallelispinosa. This is in contrast to their cranidia, which are typical of other species in their respective families, and their cranidial ontogenies do not exhibit any novel development. Linear morphometrics, however, indicate that the two species underwent different growth trajectories of the glabella and palpebral lobes, and their morphologies at each ontogenetic stage are more similar to other taxa in their respective families than to each other. By contrast, their trunk development through segment appearance and segment release, is nearly the same, and it is hypothesized that the few thoracic segments but many pygidial segments is likely homoplasious.

A future phylogenetic analysis may show that this morphology could have arisen by timing modifications to segment release, indicating heterochrony. Heterochrony has been recognized as important in the evolutionary history of trilobites, including the early Cambrian oryctocephalids of South China, early and middle Cambrian oryctocephalids of Laurentia, middle Cambrian zacanthoidids of Laurentia, and the paedomorphic Thoracocare minuta. Furthermore, if this character trait is indeed homoplasious, this evolutionary convergence, and the endemicity of the species displaying it, may be related to the unique environmental conditions that were present in that part of the basin. Preliminary work has identified probable fluctuating and low-salinity conditions and elevated levels of organic matter at Carcajou Falls. Future work should investigate other paleosalinity proxies there (Wei and Algeo, 2020; Jewuła et al., 2022) and in other parts of the basin. Further work also should explore possible temperature proxies from oxygen isotopes in linguliformean brachiopods (Hearing et al., 2018), and thin sections to study any ichnofabric, which could be an independent line of evidence of salinity conditions or oxygen levels in the sediments (Buatois and Mángano, 2011; Pratt and Kimmig, 2019). Continued investigation into the phylogeny of these species, the evolution of their development, and the local and basinal environmental conditions may identify a link between heterochrony in trilobites and the paleoenvironment.