Introduction
Substrate conditions of the shallow marine seafloor are critical to the survival of immobile benthic metazoans—organisms that rely upon life-long consistency in substrate for successful attachment and stabilization upon the seafloor (Bottjer et al. Reference Bottjer, Hagadorn and Dornbos2000). Substrate conditions remained relatively uniform in shallow subtidal marine settings through much of the Proterozoic, but a significant change took place coincident with the Proterozoic–Phanerozoic transition. This event, the agronomic revolution (AR), inexorably altered the physical nature of marine substrates and in part was responsible for the initial evolutionary changes that resulted in the morphologies of subsequent Phanerozoic immobile benthic metazoans (Seilacher and Pflüger Reference Seilacher and Pflüger1994; Bottjer et al. Reference Bottjer, Hagadorn and Dornbos2000). The AR was driven by increasing bioturbation that mixed sediment, incorporated water into the substrate, and created a “mixed layer”—the few centimeters of soft, soupy, low-density sediment at the water-surface interface that is a hallmark of the modern marine seafloor (Thayer Reference Thayer1983). The resulting soft Phanerozoic-style substrates stood in sharp contrast to the earlier firm, unlithified Proterozoic-style surfaces of the Precambrian (Seilacher and Pflüger Reference Seilacher and Pflüger1994).
Despite recognition of the AR and its implications for early Phanerozoic paleoenvironments, its exact timing and extent remain poorly defined. Many previous studies focused on changing adaptive strategies among benthic metazoans as a response to changing substrate conditions, and these suggest that the AR was already having an effect by the early Cambrian (Dornbos et al. Reference Dornbos, Bottjer and Chen2005) and was nearly complete by the late Cambrian (Bottjer et. al. Reference Bottjer, Hagadorn and Dornbos2000; Dornbos Reference Dornbos2006). These studies were limited either to individual taxonomic groups (e.g., molluscs, echinoderms) or to local faunal assemblages (e.g., Maotianshan and Burgess shales). Studies of discrete trace fossils and bioturbation levels also support the presence of firm Proterozoic-style sediment well into the Cambrian (Buatois et al. Reference Buatois, Narbonne, Mángano, Carmona and Myrow2014; Tarhan and Droser Reference Tarhan and Droser2014). This study expands on the premise of this previous work, analyzing the substrate-adaptive morphologies of benthic taxa from multiple Burgess Shale-type (BST) localities and associated bioturbation levels in order to reconstruct changing substrate conditions through the Cambrian on a global scale.
Paleoecological analysis was carried out on the adaptive strategies of benthic taxa from 14 Cambrian BST faunal assemblages in order to estimate, by proxy, the general substrate conditions encountered by each fauna at the time of deposition. Morphological adaptations were interpreted from data compiled via direct observation of fossils, the Paleobiology Database (PBDB; paleobiodb.org), and literature review. Adaptive strategies were compared with bioturbation levels in order to further refine substrate conditions for faunas preserved in the Spence, Wheeler, and Maotianshan shales. The goal of this study is to enhance our understanding of early metazoan evolution by testing the hypothesis that the AR varied temporally and spatially across multiple BST localities well into the Cambrian.
Previous Work
Bioturbation had a profound effect on the benthic metazoans living upon unlithified substrates during the Cambrian radiation. All organisms that interact with the seafloor are mitigated behaviorally and/or morphologically by substrate conditions. Most burrowing organisms, for example, are restricted to soft marine substrates or evolve specific adaptations for burrowing into hard substrates (e.g., modification of valve morphology in wood-boring “shipworm” mollusks and coral-boring bivalves [Kleemann Reference Kleemann1996; Evans Reference Evans1998]). For sessile filter-feeding benthic metazoans their relationship to the marine substrate is of greater significance, because most of these organisms are unable to relocate if conditions change unfavorably, and because initially locating suitable substrate conditions during the mobile larval stage prior to attachment is critical for survival.
Given the marked physical difference between firm Proterozoic-style and soft Phanerozoic-style substrates, one would expect benthic metazoans to have evolved unique morphological adaptation to each. Thayer (Reference Thayer1975) described four distinct adaptive strategies of modern benthic metazoans from soft-substrate marine settings that are prototypical of Phanerozoic-style substrates. These adaptive strategies include the following: (1) Hard substrate attachment. Direct attachment to a hard substrate such as rock, coral reef, or exposed skeletal material through varying means; for example, brachiopods attach via a pedicle, and barnacles grow directly on hard surfaces or attach by means of a stalk (Doyle et al. Reference Doyle, Mather, Bennett and Bussell1996). (2) Root-like holdfast. Development of root-like cirri or other structures that embed some depth into the substrate and increase drag during attempted uprooting events. Many species of crinoid are well known for this style of adaptation, although some sponges and also algae (“seaweed”) also exhibit root-like holdfast structures (Milligan and Dewreede Reference Milligan and Dewreede2000; Seilacher and Macclintock Reference Seilacher and Macclintock2005). (3) Iceberg strategy. Attachment via a long skeletal extension inserted into the substrate. At greater depths in the sediment the density of the substrate increases sufficiently to stabilize the metazoan at the surface. (4) Snowshoe strategy. Attainment of very large size and a very high rate of growth, sufficient for the metazoan to remain buoyant atop the substrate and not become smothered by sediment deposition. Some bivalves are known for using this strategy, growing to immense size (Thayer Reference Thayer1975).
Early metazoans that were adapted to firm Proterozoic-style substrates required different adaptive strategies for stabilization. These adaptive strategies differ from those identified by Thayer (Reference Thayer1975, 1983) in modern marine substrates, and are described by Dornbos et al. (Reference Dornbos, Bottjer and Chen2005). (1) Shallow sediment sticker: Directly embedding the distal end into the sediment without any obvious means of attachment (e.g., root-like holdfasts, attachment disks) and not to a depth great enough to be considered an iceberg strategy (Thayer Reference Thayer1975). It appears that the substrate was sufficiently stable for the metazoan to be supported without additional morphological structures. The chancelloriids (Kloss et al. Reference Kloss, Dornbos and Chen2009), some groups of eocrinoids (Parsley and Prokop Reference Parsley and Prokop2004; Schlottke and Dornbos Reference Schlottke and Dornbos2007) and helicoplacoids (Dornbos and Bottjer Reference Dornbos and Bottjer2000) apparently used this strategy. (2) Sediment attacher: Direct attachment to the firm unlithified substrate using a suction-cup-like suctorial disk. The substrate would need to be cohesive enough to allow for adhesion to the seafloor through suction. The middle Cambrian edrioasteroid Totiglobus (Domke and Dornbos Reference Domke and Dornbos2010) utilized sediment attachment. (3) Sediment rester: Passively resting upon the sediment surface without obvious means of attachment. Several genera of Cambrian sponges (Choia, Crumillospongia) were sediment resters (Dornbos et al. Reference Dornbos, Bottjer and Chen2005).
Once the AR began, these Proterozoic-style adaptive morphologies would not have been successful on soft Phanerozoic-style seafloors with a mixed layer. Metazoans needed to adapt to changing substrate conditions in the wake of the AR, thus precipitating a second “revolution” in animal ecology and evolution, known as the Cambrian substrate revolution (CSR) (Bottjer et al. Reference Bottjer, Hagadorn and Dornbos2000). The CSR was not limited simply to sessile substrate attachers: some mobile benthic metazoans are notable for exhibiting behaviors or morphologies that are logical only in association with Proterozoic-style substrates. Mono- and polyplacophoran mollusks, for example, are mat grazers that rely upon layered microbial communities for sustenance; their presence in shallow subtidal substrates during the Cambrian is perplexing unless a microbial mat food source was also present (Bottjer et al. Reference Bottjer, Hagadorn and Dornbos2000). The unusual coeloscleritophoran Wiwaxia and the non-mineralized Odontogriphus were also likely mat grazers, given the presence of radula in both genera, and their inhabiting of fine-grained muddy siliciclastic sedimentary facies, and according to traditional paleoecological interpretations (Conway Morris Reference Conway Morris1985; Dornbos et al. Reference Dornbos, Bottjer and Chen2005; Caron and Jackson Reference Caron and Jackson2006). A second adaptive strategy observed in some mobile benthic metazoans is firmground walking, whereby the legs/ambulatory structures of the metazoan are adapted for locomotion across relatively firm substrates. Such structures make walking in softer, bioturbated sediment difficult. Lobopodians such as Hallucigenia, with its short unjointed legs, are a typical example of this firmground walking lifestyle.
Substrates Associated with Cambrian Faunas
The close relationship between substrate conditions and benthic adaptive strategies allows for interpretations of paleoenvironmental conditions. When fossils are demonstrably preserved in situ, direct comparisons can be made among sedimentary fabrics in order to interpret substrate conditions, although this approach may also be used even when fossils are not preserved in situ. Such comparisons are useful for tracking the progress of the AR during the Cambrian, even when sedimentary evidence of the original depositional environment is limited.
Dornbos et al. (Reference Dornbos, Bottjer and Chen2005) used this approach for a study of the paleocommunities of the Maotianshan and Burgess shales, shedding light onto the temporal progression of the AR during the early and middle Cambrian. The Maotianshan Shale, a lower Cambrian (520-516 Ma) unit that preserves an in situ BST fauna, exhibits extremely low levels of bioturbation, and cohesive sedimentary fabrics suggestive of Proterozoic-style substrates (Dornbos et al. Reference Dornbos, Bottjer and Chen2005; Kloss et al. Reference Kloss, Dornbos and Chen2009). The fauna itself contains a diverse sessile benthic community that overwhelmingly preferred morphological adaptations to Proterozoic-style substrates (89% of sessile benthic metazoans). Only two genera utilized Phanerozoic-style substrate adaptations: Micromitra and Nisusia, bivalves that were both hard substrate attachers (Dornbos et al. Reference Dornbos, Bottjer and Chen2005).
The Burgess Shale is a middle Cambrian (513-500 Ma) unit, the fauna of which, unlike the Maotianshan, was originally interpreted as not having been preserved in situ (Whittington Reference Whittington1985; Conway Morris Reference Conway Morris1986). Instead the fauna was thought to have been transported to the Burgess depositional setting from an adjacent shallow subtidal shelf via obrution events. More recent studies, however, interpret it as being a relatively in situ fauna (e.g., Powell et al. Reference Powell, Johnston and Collom2003; Caron and Jackson Reference Caron and Jackson2006). The Burgess Shale fauna includes numerous benthic metazoans that are dominated by Proterozoic-style adaptive strategists (64%), though not to the same level as in the Maotianshan Shale fauna. Similar to the Maotianshan Shale fauna, the Burgess Shale metazoans that utilized Phanerozoic-style substrates were bivalves, with the exception of the eocrinoid Gogia, and all were hard substrate attachers (Dornbos et al. Reference Dornbos, Bottjer and Chen2005).
Methods
Determining Adaptive Strategies of Benthic Metazoans
To assess the global-scale effects of the CSR on Cambrian faunas, we examined the morphologies of 58 individual benthic taxa from 14 Cambrian BST faunas to determine their substrate affinities (Fig. 1). The 14 Cambrian faunas were chosen because they were exceptionally preserved, and include non-mineralizing taxa that might not otherwise be represented or counted in an analysis. The analysis was carried out at the genus level in one of five ways: (1) direct observations and interpretations of observed morphologies of fossil specimens from Maotianshan, Spence, and Wheeler shale localities; (2) observations and interpretations of morphologies taken from published figures, illustrations, and plates of fossil specimens collected from the analyzed assemblages; (3) literature review of previously published ecological interpretations of fossil specimens from these assemblages; (4) literature review of 31 previously published ecological interpretations of fossil genera known to occur in these units (but interpretations not necessarily based upon observations of the fossils); and (5) ecological data of fossil genera culled from 257 collections in the Paleobiology Database (PBDB; paleobiodb.org) and the Center for Orsten Research and Exploration (CORE). In many cases multiple approaches were used to clarify questions regarding morphological adaptations.
Geographic distribution of Burgess Shale-type faunas utilized in paleoecological study. 1, Mount Cap Formation; 2, Burgess Shale; 3, Spence Shale; 4, Wheeler Shale; 5, Marjum Formation; 6, Pioche Shale; 7, Chisholm Shale; 8, Indian Springs Lagerstätte; 9, Latham Shale; 10, Kinzers Formation; 11, Parker Slate; 12, Sirius Passet; 13, Orsten; 14, Maotianshan Shale.
Each genus was classified according to adaptive strategy: (1) Proterozoic-style strategy (sediment sticker, sediment attacher, or sediment rester); or (2) Phanerozoic-style strategy (hard substrate attachment, root-like holdfast, iceberg strategy, or snowshoe strategy). Immobile benthic metazoans were the primary target of this study, as mobile organisms are less affected by changing substrate conditions. Select mobile metazoans with likely affinities toward specific substrate conditions (e.g., Hallucigenia and Wiwaxia) were also included in the analysis.
Once adaptive strategies were determined, we devised a means of quantifying substrate affinities for direct comparisons between Cambrian BST faunas and for measuring changes in morphological adaptations through time. The devised metric, the Substrate Adaptability Index (SAI), is a factor designed to give rough estimates of substrate dominance based upon the adaptive strategies of associated metazoans. SAI is simply
$${\rm SAI}=Ph/T,$$
where Ph is the total number of taxa known to exhibit Phanerozoic-style substrate adaptations, and where T is the total number of fossil sessile benthic taxa. An SAI of 0.0 (no Phanerozoic-style adaptive strategies present) suggests dominance of Proterozoic-style substrates, whereas an SAI of 1.0 (all observed metazoans exhibit Phanerozoic-style adaptive strategies) suggests dominance of Phanerozoic-style substrates. Most Cambrian benthic fossil communities will likely fall between the end-members of this spectrum: based upon Dornbos et al. (Reference Dornbos, Bottjer and Chen2005), the Maotianshan fauna has an SAI of 0.11 (2 Phanerozoic-style genera, 19 total) and the Burgess fauna has an SAI of 0.38 (8 Phanerozoic-style genera, 22 total). SAI is a useful metric because it allows for quick comparisons between disparate faunas, and it provides a numerical value for use in graphical or statistical analysis. SAI values were plotted over time (Fig. 2) to evaluate changing morphological adaptations resulting from the CSR.
SAI values for Cambrian Burgess Shale-type faunas and exceptionally preserved Mistaken Point and Winneshiek faunas. Geologic time axis is approximate and not to scale. n for each Cambrian fauna: Mount Cap, n=3; Burgess Shale, n=22; Spence Shale, n=12; Shallow Wheeler Shale, n=11; Deep Wheeler Shale, n=12; Marjum Formation, n=12; Pioche Shale, n=4; Chisholm Shale, n=2; Indian Springs Lagerstätte, n=4; Latham Shale, n=1; Kinzers Formation, n=14; Parker Slate, n=7; Sirius Passet, n=2; Maotianshan Shale, n=19. Orsten fauna not included in figure because of ambiguous data (see Supplementary Table 1).
Measuring Bioturbation Levels
One approach to analyzing bioturbation is the semi-quantitative ichnofabric index (ii) method first developed by Droser and Bottjer (Reference Droser and Bottjer1986). The index ranges from ii1 (no bioturbation) to ii5 (complete destruction of primary sedimentary fabric by bioturbation). The ii-method is a useful gauge of the relative dominance of Proterozoic- versus Phanerozoic-style substrates based on mixed-layer development, and as such has been extensively applied to numerous studies of Cambrian benthic faunas (Bottjer et al. Reference Bottjer, Hagadorn and Dornbos2000; Dornbos and Bottjer Reference Dornbos and Bottjer2001; Parsley and Prokop Reference Parsley and Prokop2004; Kloss et al. Reference Kloss, Dornbos and Chen2009; Domke and Dornbos Reference Domke and Dornbos2010; English and Babcock Reference English and Babcock2010).
Ichnofabric indices are used here for analyzing the relative burrowing intensity in three early and middle Cambrian shale units: the Maotianshan Shale (China), and the Wheeler and Spence shales (Idaho and Utah, U.S.A.). Rock samples for this portion of the study were collected in the field for the Spence and Wheeler shales and a core was analyzed for the Maotianshan Shale. All samples were from intervals of exceptional preservation. In order to better understand how the AR and CSR affected component faunas, we determined relative bioturbation levels to assess the relative substrate conditions present at the time of deposition.
Results
Individual Cambrian faunas are described in detail below. Full data for the paleoecological analysis are also provided (Supplementary Table 1).
The Spence Shale Fauna
The Spence Shale is typically interpreted as fine-grained siliciclastic deposits of a distal outer shelf setting adjacent to a proximal carbonate ramp (Liddell et al. Reference Liddell, Wright and Brett1997). A total of 12 genera of sessile benthic metazoan fossils with identifiable substrate adaptions were preserved in the Spence Shale (Supplementary Table 1). Six genera exhibit Phanerozoic-style adaptive strategies: the hard-substrate-attaching lingulids Acrothele and Dictyonina and the rhyconelliform Diraphora; the eocrinoid Gogia; and the infaunal burrowing lingulids Lingullela and Micromitra. Six genera exhibit Proterozoic-style adaptive strategies: the sediment-resting sponges Choia and Vauxia, the mat-grazing monoplacophoran Scenella, the mat grazer Wiwaxia, and the shallow-sediment-sticking chancelloriids Chancelloria and Allonnia.
The genus Allonnia has not been previously described in the Spence Shale, but one specimen of Allonnia was observed in the collection of amateur paleontologist Val Gunther of Brigham City, Utah (personal observation). Chancelloriids have been previously interpreted as shallow-sediment stickers (Kloss et al. Reference Kloss, Dornbos and Chen2009). Bengtson (Reference Bengtson2009) documented chancelloriid specimens attached to hard substrates in the Burgess Shale. However, an exhaustive study of more than 600 chancelloriid specimens from the Maotianshan Shale found no evidence of hard-substrate attachment (Kloss et al. Reference Kloss, Dornbos and Chen2009), and none of the specimens observed from the Spence Shale exhibit adaptation or attachment to hard substrates. The Spence’s SAI value (0.50) suggests a relatively heterogeneous substrate (Fig. 2).
The Wheeler Shale Fauna
The depositional environment of the Wheeler Shale is generally interpreted as an outer shelf setting of the House Range Embayment where fine-grained siliciclastic sediment accumulated (Rees Reference Rees1986; Robison Reference Robison1991; Halgedahl et al. Reference Halgedahl, Jarrard, Brett and Allison2009). The Wheeler contains 14 benthic genera with well-constrained ecologies (Supplemental Table 1). Eight genera exhibit Proterozoic-style adaptive strategies including sediment stickers and mat-grazing mobile benthic metazoans. Unlike in the Spence Shale fauna, Gogia is categorized here as a sediment sticker. The sole species of Gogia found in the Wheeler Shale is G. spiralis, which ontogenetically developed sediment sticking as a means of attachment during adulthood (Dornbos Reference Dornbos2006; Schlottke and Dornbos Reference Schlottke and Dornbos2007). Allonnia is here identified for the first time as occurring in the Wheeler Shale, by a single, poorly preserved, disarticulated specimen. It is classified among other chancelloriids as a sediment sticker. Latouchella and Pelagiella are small, millimeter-scale helcionellid mollusks that are generally reconstructed as epifaunal grazers (Berg-Madsen and Peel Reference Berg-Madsen and Peel1994; Parkhaev Reference Parkhaev2000; Atkins and Peel Reference Atkins and Peel2008). Although never explicitly identified as mat grazers in systematic treatments, their general ecology, association with muddy substrates, and rather small size suggest adaptation to firm Proterozoic-style substrates. The remaining six genera, including the infaunal burrowers Acrothele and Lingullela, utilized soft-substrate Phanerozoic-style adaptive strategies..
Other BST Faunas
The Marjum Formation is a carbonaceous shale overlying the Wheeler Formation that also contains its own BST fauna as diverse as both the Spence and Wheeler shales. Its 14 benthic genera are split between Proterozoic- and Phanerozoic-style adapters (Fig. 2). The Marjum includes sediment stickers, sediment attachers, sediment resters, mat grazers, and the firmground-walking lobopodian Aysheaia. Among the fauna with Phanerozoic-style adaptive strategies are two brachiopods that attached to hard substrates and five infaunal burrowing genera.
The Kinzers Formation of Pennsylvania is of equivalent age to the Wheeler Shale, and the fauna includes several unusual echinoderms that are unique to the formation. Lepidocystis and Kinzercystis are eocrinoids that lack “true” crinoid-like stems with ossicles, instead bearing stem-like structures that are extensions of the aboral surface with irregular arrangement of the skeletal plates (Sprinkle Reference Sprinkle1973). As the stems lack obvious means of attachment to hard substrates or stabilization in soft muddy substrates, it is likely that both utilized sediment sticking for attachment. Camptostroma is an unusual medusoid-shaped echinoderm with either a flattened or slightly concave upward basal end (Durham Reference Durham1966), suggestive of either a sediment-resting or sediment-attaching habit. Planutenia, Pelagiella, and Yochelcionella are helcionellid mollusks with morphologies and ecological affinities suggestive of a mat-grazing lifestyle. Atkins and Peel (Reference Atkins and Peel2008) suggest that some distinctions in helcionellid ecology should be made on the basis of shell morphology for Yochelcionella, where more strongly curved shells may be indicative of a semi-infaunal deposit-feeding habit, whereas conical or orthoconic shell designs are more typical of epifaunal grazers. The Kinzers Formation preserves three species of Yochelcionella that exhibit both shell morphologies. However, the body of literature on helcionellids, their diminutive size, and their association with muddy substrates, all suggest that most helcionellids utilized a mat-grazing lifestyle and so they are categorized accordingly. Fauna with Phanerozoic-style adaptive strategies include two genera of brachiopods that attached to hard substrates and three burrowing lingulids.
Other remaining faunas surveyed—the Sirius Passet Formation, the Pioche Shale, the Chisholm Shale, the Latham Shale, the Mount Cap Formation, the Parker Slate, the Indian Springs fauna, and the non-BST Orsten phosphatized assemblage—contain few known benthic genera, and thus SAI values are significantly skewed (Fig. 2). Taken as a whole, these units have a heterogeneous assemblage of Proterozoic- and Phanerozoic-style adaptive strategies, suggesting that intermediate heterogeneous substrate conditions were a global phenomenon during the early to middle Cambrian.
In addition to the Cambrian BST faunas surveyed, two additional faunal assemblages were included as controls. These assemblages were chosen for two reasons: (1) they represent geologic periods before and after the AR, and thus their assemblages should represent benthic faunas found in environments where substrate conditions were more homogeneous; and (2) like the BST faunas, these control assemblages are Lagerstätten and thus the degree of preservational fidelity most closely matches that of the Cambrian BST faunas. The pre-AR period is represented by the Mistaken Point Ediacaran assemblage located in Newfoundland. This assemblage includes 12 genera of Ediacaran soft-bodied metazoans that, through morphology or ecology, suggest affinities for Proterozoic-style substrates (Narbonne Reference Narbonne1998; Clapham and Narbonne Reference Clapham and Narbonne2003; Tarhan et al. Reference Tarhan, Droser and Gehling2010; Wilby et al. Reference Wilby, Carney and Howe2011), giving the Mistaken Point assemblage an SAI of 0.0 as predicted by the AR (Fig. 2). Conversely, the Winneshiek Formation, a Middle Ordovician unit found in Iowa, is known primarily for its exceptional preservation of trilobites and other mobile arthropods (Liu et al. Reference Liu, McKay, Young, Witzke and McVey2006). Although other benthic fauna are scarce, lingulid brachiopods, infaunal burrowers, are known to be abundant (Liu et al. Reference Liu, McKay, Young, Witzke and McVey2006), and the general ecology of trilobites and other early arthropods suggest that deposit-feeding behaviors were likely also common. The Winneshiek Formation fauna therefore has an SAI of 1.0 (Fig. 2)
Bioturbation Levels in the Maotianshan, Spence, and Wheeler Shales
Ichnofabric indices of the Spence and Wheeler shales are predominantly extremely low, with 92.3% of the Wheeler (1370 mm, n=1485 mm) and 91.9% of the Spence (2646 mm, n=2883 mm) scoring ii1 (Fig. 3). A fraction of the formations scored ii2 (4.8% and 5.5%, respectively) and ii3 (2.9% and 2.0%, respectively). A total of 0.7% of Spence Shale strata (located entirely within a single sample of the High Creek locality) scored ii4 (Fig. 3B), but no such ii value was recorded in the Wheeler Shale. Scores of ii5 were not observed or recorded in either shale over the entire length of both composite sections (Fig. 3). Because a score of ii4 or above is generally accepted as the threshold for mixed-layer development (Dornbos et al. Reference Dornbos, Bottjer and Chen2005; English and Babcock Reference English and Babcock2010), it is unlikely that Phanerozoic-style substrates were significantly present during deposition of either shale. These results are consistent with an earlier bioturbation analysis conducted on the Maotianshan Shale (Fig. 3C), which had extremely low bioturbation levels (Kloss et al. Reference Kloss, Dornbos and Chen2009), and other previous studies of Cambrian trace fossils and bioturbation levels (Buatois et al. Reference Buatois, Narbonne, Mángano, Carmona and Myrow2014; Tarhan and Droser Reference Tarhan and Droser2014).
Millimeter-scale bioturbation levels in the Wheeler, Spence, and Maotianshan shales. A, Ichnofabric indices (ii) for Wheeler Shale, Wheeler Amphitheatre, Utah. n=1485. B, Ichnofabric indices (ii) for Spence Shale. Composite of samples collected at Miner’s Hollow, Spence Gulch, Oneida Narrows, and High Creek. n=2485. C, Ichnofabric indices (ii) for Maotianshan Shale, Yunnan Province, China. Data compiled from Kloss et al. (Reference Kloss, Dornbos and Chen2009). n=2301.
Following the methodology of Schieber (Reference Schieber1999), we examined suspect microbially mediated sedimentary structures in both shale units during ii-analysis. Some individual laminae in the Wheeler Shale show wavy-crinkly geometries (Fig. 4B). Evidence of cohesive behavior also exists in association with possible burrowing structures in the Spence Shale (Fig. 4C). These structures have sharp contacts with the surrounding substrate, and laminae that are cross-cut by the structures display frayed, angular edges consistent with cohesive behavior. In addition, roll-up structures are also present, although rare, in the Spence Shale (Fig. 4A). The overall results of this analysis are similar to those for other fine-grained siliciclastic Cambrian units where Proterozoic-style substrates are suspected: the Maotianshan Shale, Chisholm Shale, and middle member of the Poleta Formation (Dornbos and Bottjer Reference Dornbos and Bottjer2000; Dornbos et al. Reference Dornbos, Bottjer and Chen2005; Domke and Dornbos Reference Domke and Dornbos2010).
Sedimentary features suggestive of microbial mediation in Wheeler and Spence shales. A, Roll-up structure in Spence Shale. B, Wavy-crinkly laminae exhibiting irregular width in Wheeler Shale. C, Possible burrow cross-cutting multiple laminations in Spence Shale. White arrows point to irregular boundary with apparent feathering of edges indicating firm surrounding sediment. Scale bars, 5 mm. Apparent reflectivity in some images is result of oil immersion technique used to highlight sedimentary structures.
Discussion
Bioturbation as an Indicator of Substrate Conditions
The Maotianshan, Spence, and Wheeler shale strata exhibit bioturbation levels consistent with the presence of Proterozoic-style substrates that lacked mixed-layer development. If Phanerozoic-style substrates existed in either depositional environment, they were neither abundant nor widespread enough to affect the sedimentary fabrics. Bioturbation was noted at several locations within both units, but in all cases the bioturbation was extremely shallow, penetrating no more than a few millimeters through the substrate, and in some cases failing to completely cut through individual laminae, similar to the simple shallow burrows found in Ediacaran strata (Narbonne Reference Narbonne1998; Jensen et al. Reference Jensen, Droser and Gehling2005; Seilacher et al. Reference Seilacher, Buatois and Mangano2005; Buatois and Mangano Reference Buatois and Mangano2011). There is some evidence for the presence of cohesive microbial communities, suggesting that microbial mats had at least a modest impact on the development of substrates during deposition of the Spence and Wheeler shales. The paucity of evidence for microbial mats does not invalidate the putative existence of Proterozoic-style substrates; numerous other Proterozoic-style substrates typically lack microbial mat structures (Dornbos and Bottjer Reference Dornbos and Bottjer2000; Dornbos et al. Reference Dornbos, Bottjer and Chen2005; Kloss et al. Reference Kloss, Dornbos and Chen2009; Domke and Dornbos Reference Domke and Dornbos2010), and microbial mat development is not a prerequisite for Proterozoic-style substrates. Furthermore, fine-grained Cambrian siliciclastic sediment was cohesive enough on its own to account for the presence of Proterozoic-style substrates (Droser et al. Reference Droser, Jensen, Gehling, Myrow and Narbonne2002). The ichnofabric index data thus support widespread distribution of Proterozoic-style substrates in all three shales.
Adaptive Strategies of Cambrian Fauna and the Implications for Substrates
The plot of SAI values through geologic time (Fig. 2) reveals that Cambrian BST faunas exhibit some evidence for heterogeneous substrate conditions, though the degree of heterogeneity and the relative abundance of Proterozoic-style versus Phanerozoic-style substrates vary considerably. There is no discernable pattern regarding the progression of the AR from the early to middle Cambrian (SAI: 0.11-0.58 for faunas with genera N>7) (Fig. 2). Thus, the AR was a globally asynchronous event that is restricted to the Cambrian, whereas before and after the AR benthic faunas had adaptations for more homogeneous substrate conditions (very low and very high SAI).
It is possible that environmental factors played a role in controlling the progression of the AR, and that differences between depositional conditions associated with BST faunas contributed to varying rates of substrate turnover during the Cambrian. Paleoenvironmental conditions are known to have affected patterns of distribution of fossil organisms, e.g., onshore-offshore patterns in the evolution of ichnotaxa (Jablonski et al. Reference Jablonski, Sepkoski, Bottjer and Sheehan1983; Bottjer et al. Reference Bottjer, Droser and Jablonski1988), bathymetry (Ekdale and Mason Reference Ekdale and Mason1988), environmental gradients (Frey et al. Reference Frey, Pemberton and Saunder1990), and local transient changes in physical or chemical conditions, such as a fluctuating oxycline (Brett et al. Reference Brett, Allison, DeSantis, Liddell and Kramer2009) or changes in salinity (Babcock et al. Reference Babcock, Zhang and Leslie2001). BST faunas are remarkable in the sense that their exceptional fossil preservation occurs under disparate depositional conditions. The Wheeler and Burgess Shale faunas were located in deep water whereas the Chengjiang and Emu Bay shales were deposited in shallower settings (Whittington Reference Whittington1985; Conway Morris Reference Conway Morris1986; Babcock et al. Reference Babcock, Zhang and Leslie2001; McKirdy et al. Reference McKirdy, Hall, Nedin, Halverson, Michaelson, Jago, Gehling and Jenkins2011). This disparity in depositional conditions may explain, in part, the lack of linear progression of the AR when viewed from the SAI data.
The slight differences in SAI value between BST faunas, however, and the mere presence of both Proterozoic- and Phanerozoic-style adaptive genera in many of these localities indicate that substrate heterogeneity was widespread during the early–middle Cambrian. Nonactualistic substrates were still an important physical control on depositional environments and, considering the number of Proterozoic-style adapted genera observed at these localities, they were still influential for the morphological development of the benthic fauna until at least the Furongian.
Conclusions
Paleoecological analysis of exceptionally preserved early and middle Cambrian Burgess Shale-type (BST) faunas centered on morphological adaptations to substrate conditions, utilizing the SAI metric, demonstrates that Proterozoic-style adaptive strategies were still common among benthic metazoans through at least the middle Cambrian. Analysis of bioturbation levels in the Maotianshan, Spence, and Wheeler shales indicate extremely low levels of bioturbation (ii~1) and lack of mixed-layer development for all three units. Comparisons with the paleoecological data support the hypothesis that the depositional environments of the BST faunas examined in this study contained nonactualistic firm Proterozoic-style substrates through the middle Cambrian. The agronomic revolution and associated Cambrian substrate revolution were therefore globally asynchronous events that lasted well into the Cambrian and had a protracted effect on the evolutionary paleoecology of early benthic metazoans during the Cambrian radiation.
Acknowledgments
This research was supported by grants to S. Q. Dornbos from the University of Wisconsin-Milwaukee Research Growth Initiative and the University of Wisconsin-Milwaukee Graduate School Research Committee Award and grants to T. J. Kloss from the Geological Society of America, the Paleontological Society, Sigma Xi, and the Wisconsin Geological Society. A. Aase and the Gunther family provided invaluable help with collection of samples from the Wheeler and Spence shales. R. Gaines is thanked for Wheeler Shale locality information. Special thanks to P. Myrow and L. Buatois for thorough, constructive reviews of earlier versions of this manuscript. This is Paleobiology Database publication #212.
Supplementary material
Supplemental materials deposited at Dryad: doi:10.5061/dryad.760v9
