Downloaded from http://rspb.royalsocietypublishing.org/ on January 24, 2017

Proc. R. Soc. B (2012) 279, 1613–1620 doi:10.1098/rspb.2011.1986 Published online 9 November 2011

Skimming the surface with locomotion Nicholas J. Minter1,*, M. Gabriela Ma´ngano1 and Jean-Bernard Caron2,3 1Department of Geological Sciences, University of Saskatchewan, 114 Science Place, Saskatoon, Saskatchewan, Canada S7N 5E2 2Department of Natural History (Palaeobiology section), Royal Ontario Museum, 100 Queen’s Park, Toronto, Ontario, Canada M5S 2C6 3Department of Ecology and Evolutionary Biology, University of Toronto, 25 Willcocks Street, Toronto, Ontario, Canada M5S 3B2 The first arthropod trackways are described from the Middle Burgess Shale Formation of Canada. Trace fossils, including trackways, provide a rich source of biological and ecological information, including direct evidence of behaviour not commonly available from body fossils alone. The discovery of large arthropod trackways is unique for Burgess Shale-type deposits. Trackway dimensions and the requi- site number of limbs are matched with the body plan of a tegopeltid arthropod. Tegopelte, one of the rarest Burgess Shale , is over twice the size of all other benthic known from this locality, and only its sister taxon, Saperion, from the Lower Cambrian Chengjiang biota of China, approaches a similar size. Biomechanical trackway analysis demonstrates that tegopeltids were capable of rapidly skimming across the seafloor and, in conjunction with the identification of gut diverticulae in Tegopelte, supports pre- vious hypotheses on the locomotory capabilities and carnivorous mode of life of such arthropods. The trackways occur in the oldest part (Kicking Horse Shale Member) of the Burgess Shale Formation, which is also known for its scarce assemblage of soft-bodied organisms, and indicate at least intermittent oxygenated bottom waters and low sedimentation rates. Keywords: Burgess Shale; Cambrian; ichnology; locomotion; trackway; palaeoecology

1. INTRODUCTION Tegopelte, but it is also inferred to have been a benthic Discovered over 100 years ago, the world-famous Burgess predator or scavenger [6,13]. The aims of this paper are Shale continues to yield important new insights into the evol- to (i) describe the first known arthropod trackways from ution and ecology of animals during the Cambrian explosion the Burgess Shale Formation; (ii) identify their potential [1–5]. The exceptional preservation of soft tissues in the producers; (iii) integrate trackway and body fossil data Burgess Shale permits functional morphological inferences to derive the locomotory techniques employed in pro- on the modes of life and palaeoecology of early metazoans ducing the trackways; and (iv) test previous hypotheses [6,7]. Trace fossils, including trackways and burrows, on their locomotory capabilities and mode of life. provide direct evidence for behaviour in the fossil record, and can add further crucial insights for unravelling the palaeoecology of their producers and placing tempo- 2. GEOLOGICAL SETTING ral and palaeoenvironmental constraints on the origin of The Middle Cambrian (Series 3) Burgess Shale Formation functional novelties. Here, we report and discuss the signifi- in Yoho National Park (British Columbia, Canada) is a cance of the first arthropod trackways discovered from the basinal succession of mudstone and carbonate that has Burgess Shale Formation. yielded several fossil assemblages of soft-bodied animals Functional morphological studies of the benthic [14,15]. The trackways were collected from two separate Burgess Shale arthropods Olenoides [8,9], Naraoia [10], localities on Mount Stephen and Mount Field (electronic Sidneyia [11], Molaria [12] and Tegopelte [13] postulated supplementary material, figure S1) within the Kicking that they employed slow, low-geared, stable gaits with Horse Shale Member of the Burgess Shale Formation many propulsive limbs in contact with the substrate [14,15]. The Kicking Horse Shale Member comprises when walking, and could launch themselves off the sub- thinly bedded, calcareous muddy siltstone and thin slaty strate to swim or drift. Limb morphology, with spinose limestone [14,15]. With the exception of the Sanctacaris coxae and podomeres, also led to the suggestion that locality on Mount Stephen [16], the Kicking Horse Olenoides, Naraoia, Sidneyia and Molaria were predators Shale Member generally contains a low diversity assem- or scavengers [6,8–12]. The coxae are not known in blage of soft-bodied animals [14,15]. The Kicking Horse Shale Member is the oldest unit (Glossopleura Zone) of the Burgess Shale Formation [14,15]. *Author for correspondence ([email protected]). The exceptional preservation of animals within the Electronic supplementary material is available at http://dx.doi.org/ younger (Bathyuriscus Zone) Phyllopod Bed in the Walcott 10.1098/rspb.2011.1986 or via http://rspb.royalsocietypublishing.org. Quarry Shale Member [14,15] has been attributed to

Received 22 September 2011 Accepted 19 October 2011 1613 This journal is q 2011 The Royal Society Downloaded from http://rspb.royalsocietypublishing.org/ on January 24, 2017

1614 N. J. Minter et al. Burgess Shale trackways

(a) (b) (c)

(d)

os

im is

Figure 1. Burgess Shale arthropod trackways. (a) Detail of the beginning of ROM 61454 (see electronic supplementary material, figures S3–S5 for full trackway) from the Kicking Horse Shale Member on Mount Field. (b) Line drawing of the beginning of ROM 61454. (c) ROM 61455, from the Kicking Horse Shale Member on Mount Stephen, showing a gradual turn and intermittent paired medial imprints, im. (d) ROM 61456, from the Kicking Horse Shale Member on Mount Stephen, showing a sharp turn with outer splaying of tracks, os, and short inner spurs of tracks, is. All specimens preserved in negative epirelief. Scale bars, 100 mm. rapid burial by mud-rich slurry flows [17]. This indicates levels following soft-tissue stabilization under anoxic con- that the animals were transported, although recent tapho- ditions [21]. High rates of sedimentation have also been nomic analysis suggests transport and decay were invoked to explain the near absence of burrows in parts of negligible, and most benthic animals were probably buried theBurgessShaleFormation[17], although the emplace- within their habitats [18]. Geochemical analysis of Burgess ment of simple trails only requires short colonization Shale-type deposits indicates exaerobic conditions, with the windows [27]. The diminutive nature and shallow anoxic–oxic boundary occurring at the sediment–water penetration depth of the trace fossils in the Stephen For- interface, rather than anoxic bottom water [19]. Trace fos- mation are indicative of oxygen-deficient conditions, and sils are preserved in situ and therefore represent reliable suggest that oxygen rather then sedimentation rate is the indicators of local environmental conditions. Simple bur- primary control on their occurrence in Burgess Shale-type rows and trails within the Burgess Shale Formation deposits [23]. [14,18,20] and other Burgess Shale-type deposits of the USA [21] demonstrate at least intermittent oxygenation and in situ colonization of the sediment. Diminutive trace 3. TRACKWAY DESCRIPTION fossil assemblages are also found in association with cara- The five known trackway specimens are reposited at paces of soft-bodied animals in the Stephen Formation of the Royal Ontario Museum, Toronto, Canada (ROM). Canada [22,23] and Chengjiang [24] and Kaili biotas of The longest trackway (ROM 61454) is preserved for China [25,26]. Such occurrences in Burgess Shale-type over 3 m (figure 1a,b; electronic supplementary material, deposits have been attributed to increased benthic oxygen figures S2–S5). The trackway comprises overlapping

Proc. R. Soc. B (2012) Downloaded from http://rspb.royalsocietypublishing.org/ on January 24, 2017

Burgess Shale trackways N. J. Minter et al. 1615 series of up to 25 tracks. It is slightly dimorphic in form. The majority of Cambrian trackways are broadly The tracks on one side are elliptical to curvilinear in shape attributed to trilobites [30,31] on the basis of a corre- and orientated obliquely to the midline of the trackway, lation between trackway features and the general while those on the other are more elongate and orientated trilobite body plan (assignment level i), or more specific perpendicularly to the midline. The sets of tracks are also groups [32] on the basis of correlation with coeval body all offset and orientated obliquely relative to the long-axis fossils from the same palaeobiogeographic region and of the trackway and the direction of travel. The external environment (assignment level iii). Trackways from the width is 100–120 mm. Tracks are 2–27 mm long and shallow marine Bright Angel Shale of Arizona have been spaced 7–28 mm apart. The stride is 200–230 mm. correlated with a producer similar to the coeval (but ROM 61455 preserves a trackway making a gradual deeper-water) Burgess Shale arthropod Habelia [33] turn (figure 1c). The external width is 120–135 mm. (assignment level ii). A second trace fossil from the The trackway splays on the outside of the turn, making same formation was identified as a trackway, potentially the series more apparent. Highly overlapping series of related to a lobopodian animal resembling Aysheaia up to 25 tracks are observed and the stride is 90–95 mm [34], although this particular trace fossil may instead be on the outside of the turn, decreasing to 60–85 mm on an unusual expression of a serially and alternating straighter sections. Tracks are circular to comma-shaped, branched burrow system [34,35]. 2–15 mm long and spaced 5–21 mm apart. Intermit- The Burgess Shale trackways described herein are tent paired medial imprints are present. ROM 61456 unique in their large size and occurrence in Cambrian Bur- preserves a trackway making a tight U-turn (figure 1d). gess Shale-type deposits. The trackways are of different The series splay strongly on the outside of the turn, sizes (58–135 mm wide) and probably come from differ- although a high degree of overprinting obscures the precise ent stratigraphic levels within the Kicking Horse Shale number of tracks per series. Short spurs of tracks are pre- Member, and were thus produced by different individuals. sent on the inside of the turn. Tracks range from elliptical However, they all share the same characteristics of relative to curvilinear in shape and at least 20 are present within dimensions and numbers of tracks, and were therefore all each series. The external width is approximately 70 mm, made by the same type of animal. Comparison of key and tracks are 5–18 mm long and spaced 8–22 mm apart. trackway characters (including size and number of tracks) Several smaller trackways with differing orientations are with potential producers from the Burgess Shale is facili- preservedinROM61457(electronicsupplementary tated by the exceptional preservation of soft tissues in material, figure S6) with external widths of 58–70 mm census and time-averaged communities [18]. This provides and overlapping series of up to 22 elliptical to curvilinear the potential opportunity to correlate the trackways with tracks. Tracks are 3–10 mm long and spaced 7–22 mm body fossils from the same formation (assignment level iv). apart. The stride is 145–150 mm. ROM 61458 is a small The producers of the Burgess Shale trackways were large slab that preserves a partial trackway similar to those of (up to 300 mm long and 135 mm wide) and possessed at ROM 61457. All trackways can be ascribed to the ichno- least 25 pairs of walking limbs. Among known Burgess genus Diplichnites (electronic supplementary material). Shale and other Lower–Middle Cambrian arthropods Helminthoidichnites-like trails are also associated with the with walking limbs (electronic supplementary material, trackways (electronic supplementary material, figure S7). table S1), only Te g o p e l t e [13](figure 2) fits these criteria, reaching 280 mm in length, 140 mm in width and posses- sing 33 pairs of walking limbs. The posteriormost limbs are 4. PRODUCER IDENTIFICATION reduced and are unlikely to have been used in locomotion Trace fossil morphology is influenced by the anatomy of [13]. Fragmentary remains of Sidneyia from the Middle the producer, its behaviour and the substrate conditions. Cambrian of Utah are estimated to approach the appropri- Compared with other types of trace fossils, trackways are ate size [36], although Sidneyia only has nine pairs of more strongly influenced by, and therefore more reflective walking limbs [11]. All other known Burgess Shale arthro- of, the anatomy of the producer [28]. Fortey & Seilacher pods with walking limbs are too small to have made all but [29] established a set of criteria for assigning a particular the smallest trackways, but still possess insufficient num- trace fossil taxon to a specific producer: (i) close associ- bers of appendages (electronic supplementary material, ation in the field; (ii) concurrent stratigraphic range; table S1). Saperion,fromtheLowerCambrianChengjiang (iii) minimal choice of available candidates; (iv) consistent fauna of China, reaches 151 mm in length and possesses size range; and (v) consistent biogeographic range. These approximately 25 pairs of walking limbs [37]. ideas can be built upon to create criteria for assigning indi- Te g o p e l t e is the largest benthic arthropod from the vidual trackways or assemblages of trackways to producers Burgess Shale (figure 2; electronic supplementary material, with varying levels of confidence and corresponding figure S8) and is at least twice the size of any other known approximate biotaxonomic rank. In increasing order, benthic Burgess Shale bilaterian. It is also one of the rarest, these assignment levels are: (i) general correlation of track- with the only two known specimens occurring from the way features with the body plan of an animal group (class Walcott Quarry Shale Member [13]. Tegopelte was orig- or higher equivalent taxonomic rank at the stem or crown- inally described as a ‘soft-bodied trilobite’ with a thin group levels); (ii) correlation of features with coeval body unmineralized divided exoskeleton [13]. Subsequent fossils (class–order); (iii) correlation of features with analysis has determined that the divisions are folds, arising coeval body fossils from the same palaeobiogeographic from compaction during preservation, in an undivided region and environment (order–family); (iv) correlation dorsal shield [37]. Tegopelte is often regarded as belonging of features with body fossils from the same formation to the lamellipedians [38], a group that includes trilobites; (order–genus); and (v) producer at the end or in close however, the position of this group is largely unconstrained association with the trackway (species). [39]. Phylogenetic analysis has resolved a sister relationship

Proc. R. Soc. B (2012) Downloaded from http://rspb.royalsocietypublishing.org/ on January 24, 2017

1616 N. J. Minter et al. Burgess Shale trackways

La n Ra (a) (b) (d) Ra La? Le H Re H

al

gd

(c)

n Figure 2. Tegopelte gigas from the Walcott Quarry Shale Member at the Walcott Quarry. (a) Holotype (USNM 189201) showing the hypostome, H, left antennae, La, right antennae, Ra, notches, n, at the anterior and posterior end and the alimentary canal, al. (b) Counterpart of the holotype showing the gut diverticulae, gd. (c) Detail of gut diverticulae. (d) Paratype (USNM 189200) showing the additional presence of a left eye, Le, and right eye, Re. Eyes are tear-shaped and probably ventral in position. Scale bars: (a,b,d ) 30 mm; (c) 10 mm. of Tegopelte with Saperion and position in the Order would have moved in advance of the limb in front by Helmetiida/Conciliterga [40,41]. Body fossils of Te g o p e l t e 88 per cent of the step cycle. The trackways have opposite are not known from the Kicking Horse Shale Member, symmetry and so paired limbs on either side of the body but the trackways can be attributed to Te g o p e l t e (or, at the moved synchronously. These parameters indicate that very least, a tegopeltid) on the basis of correlation with only one to two pairs of limbs were in contact with the sub- body fossils from the same formation (assignment level iv). strate at any one point in time and that metachronal waves of eight limbs appeared to travel backwards along the body with short gaps between the placement of proceed- 5. LOCOMOTION ing limbs (figure 3; electronic supplementary material, Animal locomotion can be defined by three parameters: figures S9a and S10, and animation S1). the gait ratio (p : r), which is the proportion of the step ROM 61455 demonstrates a slower but still high-geared cycle that a limb spends in the forward, recovery, protrac- gait. Analysis of the straighter section yields a gait ratio of tion period (p) to that spent in the backward, propulsive, 7.6 : 2.4, where limbs spent 24 per cent of the step cycle retraction period (r); the successive phase difference (suc), in the propulsive backstroke phase and successive limbs which is the proportion of the step cycle that a limb moves moved in advance of the limb in front by 69 per cent of before the limb in front; and the opposite phase difference the step cycle. The trackway has opposite symmetry and (opp), which is the proportion of the step cycle that a limb so paired limbs on either side of the body moved synchro- on one side of the body moves after the paired limb on the nously. These parameters indicate that six pairs of limbs opposing side of the body [42,43]. These parameters can were in contact with the substrate at any one point in be derived from trackways and a reconstruction of the time, and metachronal waves of three limbs appear to producer (electronic supplementary material), and have have travelled backwards along the body with short gaps been used to reconstruct the locomotion of numerous between the placement of proceeding limbs (electronic extinct animals [44–46]. supplementary material, figure S9b). The Burgess Shale trackways reveal that their producers ROM 61456 indicates the producer was capable of were capable of performing fast, high-geared walking gaits performing tight turns (figure 1d); however, unlike trilo- and also able to ‘skim’ across the substrate with few limbs bites, Tegopelte has an unarticulated dorsal shield rather in contact with the substrate at any one point in time. Track- than an articulated dorsal exoskeleton. Detailed examin- way analysis (electronic supplementary material) yields gait ation of Tegopelte and the trackway demonstrates how ratios of 9.3 : 0.7 and 9.4 : 0.6, indicating that limbs only the execution of a tight turn can be ameliorated with an spent 7 and 6 per cent of the step cycle in the propulsive arthropod with an unarticulated dorsal exoskeleton. In backstroke phase, for ROM 61454 (figure 1a,b and elec- Tegopelte, the compression of gill filaments on the surface tronic supplementary material, figures S3–S5) and ROM of the cuticle suggests that its dorsal shield was thin and 61457 (electronic supplementary material, figure S6), would have had some flexibility. In the trackway, the respectively. Successive limbs on one side of the body track series on the outside of the turn are splayed, but

Proc. R. Soc. B (2012) Downloaded from http://rspb.royalsocietypublishing.org/ on January 24, 2017

Burgess Shale trackways N. J. Minter et al. 1617

(a)(time b) 33 25 18 17 91 t0 suc r p 9 1 33 25 19 18 10 2 t1

18 10 t t t t t t t t t t 1 2 0 1 2 3 4 5 6 7 8 9 33 25 20 19 11 3 t2

19 11 1 1 3 33 25 21 20 12 4 t3

2 20 12 1 4 t 33 25 22 21 13 5 3 4 21 13 1 5 t 33 25 23 22 14 6 4 5

22 14 1 6 5 33 25 24 23 15 7 t6 limb

23 15 7 6 1 33 25 24 16 8 t7

24 16 8 7 1

33 25 17 9 1 t8

8 25 17 9 1 t 33 25 18 10 2 1 9 9

25 18 1 10 1

Figure 3. Locomotion and trackway production of a tegopeltid animal, based on ROM 61454 using a gait ratio of 9.3 : 0.7 and successive phase difference (suc) of 0.88. (a) Manton gait diagram. Horizontal axis represents progression through time. The phase relationships of the anterior 10 limbs on one side of the body are depicted. Each complete step cycle of a limb is rep- resented by a couplet of a thick downward sloping line, corresponding to the relative duration of the propulsive, backward, retraction phase (r) when the limb is in contact with the substrate; and a thin upward sloping line corresponding to the relative duration of the recovery, forward, protraction phase (p) when the limb is lifted above the substrate. The successive phase differ- ence is the proportion of the step cycle that a limb n þ 1 moves in advance of the limb n in front. Vertical time lines (t0 2 t9) capture snapshots of the positions of the limbs that are depicted in (b). (b) Time-lapse trackway production. ‘Gait still’ snap- shots represent the positions of the limbs and development of the trackway between the placement of limbs n and n þ 1 in the proceeding step cycle. Light-coloured limbs indicate those in contact with the substrate. The en-echelon arrangement of the series of tracks has been exaggerated to show the individual tracks. only with a low degree of curvature, and short spurs of subject to a unidirectional current. The alternative is tracks are present on the inside of the turn (figure 1d). that the producers were less able to control the yaw of Such a tight turn could therefore be achieved by a their body at these higher-geared gaits, owing to fewer tegopeltid through a combination of limited lateral flexure limbs being in contact with the substrate at any one and side-stepping. point in time, compared with more normal trackways pro- The reconstructed method of trackway production duced by lower-geared walking gaits (figure 1c). shows that the anterior tracks of each series are produced by the posterior limbs (figures 3 and 4). In ROM 61454, the anterior tracks of each series are situated more 6. DISCUSSION medially to the midline of the trackway (figure 1a,b; elec- The discovery of trackways in the Burgess Shale permits tronic supplementary material, figures S3–S5), which is direct insights into animal behaviour and environmental consistent with the method of production and the poster- conditions during the Cambrian. The trackways can be iormost limbs being shorter in Tegopelte. The en-echelon attributed to the Burgess Shale arthropod Tegopelte, arrangement of tracks in this trackway and ROM 61457 or at least a tegopeltid (figure 4). No other types of results from the body of the producer being offset obli- trackways have been found so far in the Burgess Shale quely relative to the long axis of the trackway and Formation. On the basis of functional morphology, direction of travel. This is caused by the animal being Tegopelte was hypothesized as employing a walking gait partly buoyed up while producing the trackway and that was slow and low-geared (p : r ¼ 3.75 : 6.25, could be related to a current or the producer being less suc ¼ 0.125), with many propulsive limbs in contact able to control the yaw of its body while employing a with the substrate at any one point in time, but could high-geared gait ratio. There are no tool marks or also perform higher-geared gaits to launch from the sub- scours preserved with the trackways to indicate the strate and drift or swim [13]. The trackways provide action of a current. These may not be observed if direct evidence to largely support this hypothesis, the trackways are undertracks [47]; however, undertracks showing that such animals could indeed employ fast, are unlikely to be produced by an animal that was buoyed high-geared gaits akin to ‘skimming’ across the substrate. up. Also, in ROM 61457, several trackways have dif- The trackways also demonstrate that they were capable of fering orientations (electronic supplementary material, a slower and lower-geared walking gait that is similar to figure S6), but were all produced by high-geared gaits, those calculated for subaqueously walking myriapods and this is inconsistent with the producers all being [45]. This walking gait is higher-geared than that

Proc. R. Soc. B (2012) Downloaded from http://rspb.royalsocietypublishing.org/ on January 24, 2017

1618 N. J. Minter et al. Burgess Shale trackways

Figure 4. Reconstruction of Tegopelte gigas producing a trackway. Copyright q 2011 Marianne Collins. previously proposed for Tegopelte, although it falls within Burgess Shale localities [14], the fossil-bearing layers of the hypothesized locomotory end-members [13]. the Kicking Horse Shale Member do not lie directly at On the basis of functional morphology, Tegopelte is the base of the Cathedral Escarpment, and are of low considered to have been much like the large Sidneyia diversity and abundance. This might suggest that the and smaller Olenoides, Naraoia and Molaria in being a Kicking Horse Shale Member represented an environ- benthic predator or scavenger [6]. The trackway evidence ment generally unfavourable for a rich community to supports the hypothesis of a predatory or scavenging thrive. The presence of large trackways also suggests mode of life, demonstrating that they could employ fast, that the general scarcity of soft-tissue preservation high-geared gaits and were also capable of performing may be taphonomic, with a greater prevalence of oxygen tight turns. Fast, high-geared gaits could potentially be and/or low sedimentation rates throughout deposition associated with predator avoidance, although the produ- of the Kicking Horse Shale Member. cers of these trackways would have been twice the size The Burgess Shale biota provides key insights into the of any other benthic Burgess Shale arthropod and over evolution and ecology of body plans during the Cambrian half the estimated size of the largest Anomalocaris [48]. explosion. The discovery of arthropod trackways from the In addition, our re-examination of Te g o p e l t e identified the Kicking Horse Shale Member highlights their application presence of gut diverticulae (figure 2b,c). This provides as indicators of environmental conditions. The results strong evidence of a predatory or scavenging mode of life also provide insights into the locomotory capabilities of [49,50]. The presence of eyes in Te g o p e l t e [13] was con- extinct animals that can be used to test functional mor- sidered to be uncertain [40], but they are confirmed phological hypotheses on modes of life. Integrated trace herein and are probably ventral in position (figure 2d and and body fossil evidence provides an innovative and electronic supplementary material, figure S8a). dynamic picture of arthropods such as Tegopelte as active The presence of trackways, along with interface trails, carnivores. Early marine communities were therefore in the Kicking Horse Shale Member of the Burgess occupied not just by large nektonic predators such as Shale Formation also permits inferences on environ- Anomalocaris, but also by large benthic predators, empha- mental conditions. Burrows have been observed in other sizing the importance of such lifestyles in shaping early members of the Burgess Shale Formation and indicate complex marine communities and evolution during the at least intermittent oxygenation of the sediment [20], Cambrian explosion. while the diminutive and shallow nature of burrows and trails in the adjacent Stephen Formation suggests that We thank Parks Canada for collecting permits (to ROM— oxygen was a major control on their occurrence [23]. Desmond Collins and J.-B.C.), and Doug Erwin, Mark The trackways described herein are the first to be discov- Florence and John Pojeta (Smithsonian Institution) for access to the Tegopelte specimens and technical assistance. ered from the Burgess Shale Formation and they are N.J.M.’s research is supported through a Government unique for Burgess Shale-type deposits in their large of Canada Post-doctoral Research Fellowship under the size. They provide evidence of arthropods living within Canadian Commonwealth Scholarship Programme. the environment, extending the stratigraphic range of M.G.M.’s research is supported by Natural Sciences and tegopeltids, and indicate that the bottom waters must Engineering Research Council (NSERC) Discovery Grant have been sufficiently oxygenated, at least intermittently, 311727 and University of Saskatchewan StartUp fund to sustain large benthic animals during deposition of the 411435. J.-B.C.’s research is supported by NSERC Discovery Grant 341944 and ROM research and fieldwork oldest member of the Burgess Shale Formation. The pres- grants. Marianne Collins produced figures 3b and 4, ervation of such trackways would also have required that and electronic supplementary material, figure S10 and the substrate was dewatered and stiff owing to breaks in, animation S1. This is Royal Ontario Museum Burgess or slow rates of, sedimentation. Contrary to younger Shale project number 34. The manuscript benefitted from

Proc. R. Soc. B (2012) Downloaded from http://rspb.royalsocietypublishing.org/ on January 24, 2017

Burgess Shale trackways N. J. Minter et al. 1619 the comments of Greg Edgecombe as the Associate Editor, 18 Caron, J.-B. & Jackson, D. A. 2006 Taphonomy of the and two anonymous reviewers. greater phyllopod bed community, Burgess Shale. Palaios 21, 451–465. (doi:10.2110/palo.2003.P05-070R) 19 Powell, W. G., Johnston, P. A. & Collom, C. J. 2003 Geochemical evidence for oxygenated bottom waters REFERENCES during deposition of fossiliferous strata of the Burgess 1 Caron, J.-B., Scheltema, A., Schander, C. & Rudkin, D. Shale Formation. Palaeogeogr. Palaeoclimatol. Palaeoecol. 2006 A soft-bodied mollusk with radula from the Middle 201, 249–268. (doi:10.1016/S0031-0182(03)00612-6) Cambrian Burgess Shale. Nature 442, 159–163. (doi:10. 20 Allison, P. A. & Brett, C. E. 1995 In situ benthos and 1038/nature04894) paleo-oxygenation in the Middle Cambrian Burgess 2 Conway Morris, S. & Caron, J.-B. 2007 Halwaxiids and Shale, British Columbia, Canada. Geology 23, 1079– the early evolution of the lophotrochozoans. Science 315, 1082. (doi:10.1130/0091-7613(1995)023,1079:ISBA 1255–1258. (doi:10.1126/science.1137187) PO.2.3.CO;2) 3 Daley, A. C., Budd, G. E., Caron, J.-B., Edgecombe, 21 Gaines, R. R. & Droser, M. L. 2010 The paleoredox G. D. & Collins, D. 2009 The Burgess Shale anomalocar- setting of Burgess Shale-type deposits. Palaeogeogr. idid Hurdia and its significance for early euarthropod Palaeoclimatol. Palaeoecol. 297, 649–661. (doi:10.1016/j. evolution. Science 323, 1597–1600. (doi:10.1126/ palaeo.2010.09.014) science.1169514) 22Caron,J.-B.,Gaines,R.R.,Ma´ngano, M. G., Streng, M. & 4 Caron, J.-B., Conway Morris, S. & Shu, D. 2010 Tenta- Daley, A. C. 2010 A new Burgess Shale-type assemblage culate fossils from the Cambrian of Canada (British from the ‘thin’ Stephen Formation of the southern Canadian Columbia) and China (Yunnan) interpreted as primitive Rockies. Geology 38, 811–814. (doi:10.1130/G31080.1) deuterostomes. PLoS ONE 5, 1–13. 23 Ma´ngano, M. G. 2011 Trace-fossil assemblages in a Burgess 5 Smith, M. R. & Caron, J.-B. 2010 Primitive soft-bodied Shale-type deposit from the Stephen Formation at Stanley cephalopods from the Cambrian. Nature 465, 469–472. Glacier, Canadian Rocky Mountains: unraveling ecologic (doi:10.1038/nature09068) and evolutionary controls. Palaeontogr. Can. 31, 89–107. 6 Briggs, D. E. G. & Whittington, H. B. 1985 Modes of life 24 Zhang, X. G., Bergstro¨m, J., Bromley, R. G. & Hou, of arthropods from the Burgess Shale, British Columbia. X. G. 2007 Diminutive trace fossils in the Chengjiang Trans. R. Soc. Edin. Earth Sci. 76, 149–160. Lagersta¨tte. Te r r a N o v a 19, 407–412. (doi:10.1111/j. 7 Caron, J.-B. & Jackson, D. A. 2008 Paleoecology of the 1365-3121.2007.00765.x) greater phyllopod bed community, Burgess Shale. 25 Wang, Y., Lin, J.-P., Zhao, Y.-L. & Orr, P. J. 2009 Palaeogeogr. Palaeoclimatol. Palaeoecol. 258, 222–256. Palaeoecology of the trace fossil Gordia and its interaction (doi:10.1016/j.palaeo.2007.05.023) with nonmineralizing taxa from the early Middle 8 Whittington, H. B. 1975 Trilobites with appendages from Cambrian Kaili Biota, Guizhou province, South China. the Middle Cambrian, Burgess Shale, British Columbia. Palaeogeogr. Palaeoclimatol. Palaeoecol. 277, 141–148. Fossils Strata 4, 97–136. (doi:10.1016/j.palaeo.2009.02.017) 9 Whittington,H.B.1980Exoskeleton,moultstage,appen- 26 Lin, J.-P., Zhao, Y.-L., Rahman, I. A., Xiao, S. & dage morphology, and habits of the Middle Cambrian Wang, Y. 2010 Bioturbation in Burgess Shale-type trilobite Olenoides serratus. Palaeontology 23, 171–204. Lagersta¨tten: case study of trace fossil–body fossil associ- 10 Whittington, H. B. 1977 The Middle Cambrian trilobite ation from the Kaili Biota (Cambrian Series 3), Guizhou, Naraoia, Burgess Shale, British Columbia. Phil. China. Palaeogeogr. Palaeoclimatol. Palaeoecol. 291, Trans. R. Soc. Lond. B 280, 409–443. (doi:10.1098/ 245–256. (doi:10.1016/j.palaeo.2010.03.048) rstb.1977.0117) 27 Buatois, L. A., Ma´ngano, M. G., Maples, C. G. & Lanier, 11 Bruton, D. L. 1981 The arthropod Sidneyia inexpectans, W.P. 1997 The paradox of nonmarine ichnofaunas in tidal Middle Cambrian, Burgess Shale, British Columbia. rhythmites: integrating sedimentologic and ichnologic data Phil. Trans. R. Soc. Lond. B 295, 619–653. (doi:10. from the Late of eastern Kansas, USA. 1098/rstb.1981.0164) Palaios 12,467–481.(doi:10.2307/3515384) 12 Whittington, H. B. 1981 Rare arthropods from the 28 Minter, N. J., Braddy, S. J. & Davis, R. B. 2007 Between Burgess Shale, Middle Cambrian, British Columbia. a rock and a hard place: arthropod trackways and Phil. Trans. R. Soc. Lond. B 292, 329–357. (doi:10. ichnotaxonomy. Lethaia 40, 365–375. (doi:10.1111/j. 1098/rstb.1981.0033) 1502-3931.2007.00035.x) 13 Whittington, H. B. 1985 Tegopelte gigas, a second 29 Fortey, R. A. & Seilacher, A. 1997 The trace fossil Cruzi- soft-bodied trilobite from the Burgess Shale, Middle ana semiplicata and the trilobite that made it. Lethaia 30, Cambrian, British Columbia. J. Paleontol. 59, 1251–1274. 105–112. (doi:10.1111/j.1502-3931.1997.tb00450.x) 14 Fletcher, T. P. & Collins, D. H. 1998 The Middle 30 Seilacher, A. 1955 Spuren und Lebensweise der Cambrian Burgess Shale and its relationship to the Trilobiten. Abh. Math. Naturwiss. Kl. 10, 342–372. Stephen Formation in the southern Canadian Rocky 31 Alpert, S. P. 1976 Trilobite and star-like trace fossils from Mountains. Can. J. Earth Sci. 35, 413–436. (doi:10. the White Inyo Mountains, California. J. Paleontol. 50, 1139/e97-120) 226–239. 15 Fletcher, T. P. & Collins, D. H. 2003 The Burgess Shale 32 Crimes, T. P. 1970 Trilobite tracks and other trace fossils and associated Cambrian formations west of the Fossil from the Upper Cambrian of North Wales. Geol. J. 7, Gully Fault Zone on Mount Stephen, British Columbia. 47–68. (doi:10.1002/gj.3350070104) Can. J. Earth Sci. 40, 1823–1838. (doi:10.1139/e03-057) 33 Elliott, D. K. & Martin, D. L. 1987 A new trace fossil 16 Briggs, D. E. G. & Collins, D. 1988 A middle Cambrian from the Cambrian Bright Angel Shale, Grand Canyon, chelicerate from Mount Stephen, British Columbia. Arizona. J. Paleontol. 61, 641–648. Palaeontology 31, 779–798. 34 Lane, A. A., Braddy, S. J., Briggs, D. E. G. & Elliott, 17 Gabbott, S. E., Zalasiewicz, J. & Collins, D. 2008 Sedi- D. K. 2003 A new trace fossil from the Middle Cambrian mentation of the phyllopod bed within the Cambrian of the Grand Canyon, Arizona, USA. Palaeontology 46, Burgess Shale Formation of British Columbia. J. Geol. 987–997. (doi:10.1111/1475-4983.00329) Soc. Lond. 165, 307–318. (doi:10.1144/0016-7649 35 Seilacher, A., Buatois, L. A. & Ma´ngano, M. G. 2005 2007-023) Trace fossils in the Ediacaran-Cambrian transition:

Proc. R. Soc. B (2012) Downloaded from http://rspb.royalsocietypublishing.org/ on January 24, 2017

1620 N. J. Minter et al. Burgess Shale trackways

behavioral diversification, ecological turnover and environ- Linn. Soc. Lond. 42, 93–117. (doi:10.1111/j.1096- mental shift. Palaeogeogr. Palaeoclimatol. Palaeoecol. 227, 3642.1952.tb01854.x) 323–356. (doi:10.1016/j.palaeo.2005.06.003) 43 Manton, S. M. 1977 The Arthropoda: habits, functional 36 Briggs, D. E. G. & Robison, R. A. 1984 Exceptionally morphology, and evolution. Oxford, UK: Clarendon Press. preserved nontrilobite arthropods and Anomalocaris 44 Briggs, D. E. G., Rolfe, W. D. I. & Brannan, J. 1979 from the Middle Cambrian of Utah. Univ. Kansas A giant myriapod trail from the Namurian of Arran, Paleontol. Contrib. 111, 1–23. Scotland. Palaeontology 22, 273–291. 37 Ramsko¨ld, L., Chen, J.-Y., Edgecombe, G. D. & Zhou, Q. 45 Smith, A., Braddy, S. J., Marriott, S. B. & Briggs, 1996 Preservational folds simulating tergite junctions in D. E. G. 2003 Arthropod trackways from the Early Devo- tegopeltid and naraoiid arthropods. Lethaia 29, 15–20. nian of South Wales: a functional analysis of producers (doi:10.1111/j.1502-3931.1996.tb01832.x) and their behaviour. Geol. Mag. 140, 63–72. (doi:10. 38 Hou, X. & Bergstro¨m, J. 1997 Arthropods of the Lower 1017/S0016756802006982) Cambrian Chengjiang fauna, southwest China. Fossils 46 Minter, N. J. & Braddy, S. J. 2006 Walking and jumping Strata 45, 1–116. with Palaeozoic apterygote insects. Palaeontology 49, 39 Budd, G. E. & Telford, M. J. 2009 The origin and evol- 827–835. (doi:10.1111/j.1475-4983.2006.00571.x) ution of arthropods. Nature 457, 812–817. (doi:10.1038/ 47 Goldring, R. & Seilacher, A. 1971 Limulid tracks and nature07890) their sedimentological implications. N. Jb. Geol. Pala¨ont. 40 Edgecombe, G. D. & Ramsko¨ld, L. 1999 Relationships of Abh. 137, 422–442. Cambrian Arachnata and the systematic position of 48 Whittington, H. B. & Briggs, D. E. G. 1985 The largest Trilobita. J. Paleontol. 73, 263–287. Cambrian animal, Anomalocaris, Burgess Shale, British 41 Paterson, J. R., Edgecombe, G. D., Garcı´a-Bellido, Columbia. Phil. Trans. R. Soc. Lond. B 309, 569–609. D. C., Jagos, J. B. & Gehling, J. G. 2010 Nektaspid (doi:10.1098/rstb.1985.0096) arthropods from the Lower Cambrian Emu Bay Shale 49 Butterfield, N. J. 2002 Leanchoilia guts and the Lagersta¨tte, South Australia, with a reassessment of interpretation of three-dimensional structures in Burgess lamellipedian relationships. Palaeontology 53, 377–402. Shale-type fossils. Paleobiology 28, 155–171. (doi:10. (doi:10.1111/j.1475-4983.2010.00932.x) 1666/0094-8373(2002)028,0155:LGATIO.2.0.CO;2) 42 Manton, S. M. 1952 The evolution of arthropodan 50 Vannier, J. & Chen, Y. 2002 Digestive system and feed- locomotory mechanisms. II. General introduction to ing mode in Cambrian naraoiid arthropods. Lethaia 35, the locomotory mechanisms of the Arthropoda. Zool. J. 107–120. (doi:10.1080/002411602320183971)

Proc. R. Soc. B (2012)