Trackways assigned to the ichnogenus Protichnites were first noted in the Potsdam (= Cairnside) Sandstone in the Montreal Gazette in 1847 [36, 37] and later described by Owen [24, 37] as tracks belonging to, “a species of Terrapene or Emydian Tortoise.” The following year, after further study of more specimens [27], he emended his interpretation to describe the producers as arthropods and named the trackways Protichnites [25]. Owen designated six ichnospecies: P. septem-notatus, P. octo-notatus, P. latus, P. multinotatus, P. lineatus, and P. alternans. Hyphens were subsequently dropped as improper per ICZN Article 32.5.2 [38].

Hitchcock [38] worked to standardize Owen’s [25] description by listing Protichnites according to the following criteria, possessing: (1) a medial impression, (2) tracksets of five or more, (3) variable track morphologies, (4) variable size, and (5) not limited to any single producer [41]. This designation was little-used to identify trackways, and was effectively nullified when Seilacher [41] suggested that Protichnites be applied to all track-ways with a medial impression and Diplichnites to all those without one. Häntzschel [34, 42] later designated P. septemnotatus the type ichnospecies for the ichnogenus. His illustrations, however, differed from Owen’s [25] written description and plates, notably including a specimen [42, fig. 61,1b] with two parallel medial impressions [43].

Keighley and Pickerill [40] presented the most recent attempt to tighten the diagnosis of Protichnites, but in doing so followed Seilacher [41]. The diagnosis presented herein is held to more closely represent the importance Owen placed upon the specific number of tracks per set and to make the identification of Protichnites specimens easier once a complete re-evaluation of the ichnogenus is completed. This is in progress but lies beyond the scope of the present paper.

Table 1 defines terminology used. Terminology used herein follows Minter et al.[47] in most cases, the exception being the difference between ‘tracks’ and ‘imprints.’ The terminology in Minter et al.[47] assumes that a distinction between ‘tracks’ (discrete marks produced by walking limbs) and ‘imprints’ (discrete marks produced by other anatomical structures) can be made, where previous terminologies [48] have included ‘tracks’ as a subset of ‘imprints’ (defined therein as any discrete mark produced by an organism). This latter definition of ‘imprints’ is preferred because it allows reference to ‘discrete marks produced by an organism which may or may not be tracks,’ something that is often the case. To be precise, one would always use ‘tracks’ when referring to marks interpreted as having been produced by a walking limb, ‘non-track imprints’ when referring to marks that are not being interpreted as tracks (but are being interpreted as being produced by a living organism), and ‘imprints’ when referring to marks that are interpreted as being produced by a living organism but are unable to be connected to a particular aspect of the anatomy of the producer.

Table 1

General Notation (Variable Throughout Trackway Literature) Herein Follows [48, 67, 68]. Notation Specific to theArrangement and Order of Individual Imprints Follows [40]

The outcrop was visited first in April 2002 and revisited in September and October 2004 to make direct measurements. Because snowfall and trackway access were early problems, the majority of data collection and interpretation relied on digital photographs (greater than 3.0 megapixels) taken perpendicular to the bedding plane in low-angle light, which proved to be very satisfactory when printed on a large format plotter. Two photographs were printed at near life size (92.9% and 87.2%); measurements made from these photographs were scaled accordingly. Individual imprints and medial impressions were traced onto clear acetate from whence a series of tracings and photocopies was made. Imprints with discernable edges were marked without regard for depth of the imprint.

Track series were designated by locating repetitions of the same tracks at fixed intervals along the length of the trackway in each case. This was accomplished by overlaying traced groups of imprints on other sections of the trackway to look for repetition of: (1) distance between imprints, (2) distance of each imprint from the mid-line, and (3) shape of imprint. Regular repetition of an imprint in the same position in each set led to its identification as a track produced by the same walking limb. These methods were used until the unassociated imprints were exhausted, or until no spatial relationship could be found between any of the remaining unassociated imprints. Each repeated group of tracks was designated a series. These methods were used to resolve trackways A, B and C and to correlate between the related tracks across different series and across different trackways. Correlated impressions are designated by the same label and color in Fig. (4A-C). Note that there are a variety of ways in which these tracks can be divided into series, but not all of them could have been produced by a living organism [25, 48,49]. Several methods of ‘idealized’ track arrangements were applied here before arriving at the interpretation described.

Method of labeling tracks follows that of Keighley and Pickerill[40], who used R1, R2, R3, etc. and L1, L2, L3, etc. to designate tracks belonging to the same series on right and left sides of the trackway, respectively. Paired tracks nearest the mid-line (i.e., having the smallest distance between internal edges) in all trackways were labeled R0/L0. The impressions second-nearest to the mid-line and closest (along the length of the trackway) to tracks R0/L0 were marked as R1/L1, etc. Order of track numbering does not imply order of track production.

Fig. (4) Schematic diagram of described trackways with correlative coloring of identified tracks. A) Trackway A. B) Trackway B. C) Trackway C. Dashed lines separate neighboring track series, labeled by Roman numerals. Dashed circles indicate possible multipart imprints. Coloring is the same across trackways. Direction of travel was toward top of figure. Trackways are not to same scale; scale bars are 5 cm.

Trackway A consists of repeated series of seven oval or longitudinally tapered tracks on each side of a medial impression 1.5 meters in length. Over the course of 0.92 m there are 152 imprints of which 110 have been identified as tracks (Fig. 4A).

Tracks from 10 series sets are apparent. The medial impression is continuous and remains in the middle of the paired series. Some non-track imprints are located in the medial impression (width 7 to 9 mm throughout). The average external width of the track farthest from the mid-line (R7/L7) is 116 mm (n = 6) (Table 2). The average stride on the right side of the trackway (outside of gentle curve) is 103 mm (n = 40) (Table 3). Strides on the left side of the track average 99 mm (n = 35). Track morphology on the left side of the medial line seems to be significantly different from that on the right, the tracks most proximal to the medial line being longer and more tapered. There is no change in stride length or opposite-track width through the area where this trackway intersects with trackway B. Maximum trackway width is ∼119% of mean stride length. Individual tracks are oval (long axis parallel to trackway axis) or tapered (shallow-and-narrow end pointing in the direction of travel). Some tracks seem bifid or trifid, although these multiple-imprint tracks are inconsistent. Tracks L0 and R4 each have no opposite counterpart.

Table 2

Morphometric External Width Data of Trackway A (Format after [65]). (*) Indicates Widths Estimated by Doubling the Distance between the Opposite Track and Axis of the Medial Impression. (–) Indicates a Missing Value

Table 3

Morphometric Stride Data of Trackway A (Format after [65]). (*) Indicates an Averaged Value Used for Estimated Data. (—) Indicates a Missing Value

Trackway B (Fig. 3B) consists of a discontinuous medial impression flanked by repeated series sets of seven tracks of varying morphology. The portion that is preserved in enough detail to study contains 112 imprints and 74 identified tracks in a 0.42 m distance (Fig. 4B). The trackway is oriented at a high angle relative to trackway A where it becomes visible on the outcrop (right of Fig.3), then abruptly changes irection and continues through a gentle curve. Tracks representing six series sets are apparent. The medial impression is discontinuous and does not remain on the mid-line throughout. The length of the medial impression, where it is present, ranges from 15 to 50 mm (mean 42, n = 6) with an intermediary distance ranging from 5 to 35 mm (mean 21, n = 6). In one instance (sets IV and V) the medial drag is erratic. The average external width of the trackway at its widest point (R6/L6) is 65 mm (n = 4) (Table 4). Stride lengths on the right side of the trackway average 79 mm (n = 24) (Table 5). The average stride length on the left side is 76 mm (n = 23). Maximum trackway width is ~90% of mean stride length. Most tracks are nearly subcircular or sub-oval with long axes parallel to the mid-line. Track identifications were also assigned to more linear than round imprints in some cases.

Table 4

Morphometric External width Data of Trackway B (Format after [65]). (*) Indicates Widths Estimated by Doubling the Distance between the Opposite Track and the Medial Impression. (–) Indicates a Missing Value

Table 5

Morphometric Stride Data of Trackway B (Format after [65]). (*) Indicates an Averaged Value Used for Estimated Data. (—) Indicates a Missing Value

Table 6

Morphometric External Width Data of Trackway C (format after [65]). (*) Indicates Widths Estimated by Doubling the Distance between the Opposite Track and the Medial Impression. (—) Indicates a Missing Value

Trackway C exhibits many characteristics similar to trackways A and B. The medial impression is discontinuous and 8 to 11 mm wide. Medial impression segments are 30 to 55 mm long (mean 38, n = 9) with an inter-segment distance of 10 to 40 mm (mean 28, n = 8). Tracks along 77 cm of the medial impression are clear enough to be examined, but series sets could only be resolved over 48 cm. This distance includes 127 imprints, 56 of which have been associated with series (Fig. 4C). Tracks are sub-circular or oval with varying angle of long axis to the trackway. The trackway curves slightly to the right (towards top of Fig. 3) viewed in the direction of travel. Maximum trackway width is ∼86% of mean stride length.

Table 7

Morphometric Stride Data of Trackway C (Format after [65]). (*) Indicates an Averaged Value Used for Estimated Data.(–) Indicates a Missing Value

Tracks from seven series are present and have the same arrangement as in trackways A and B. The external width (R6/L6) of the trackway averages 72 mm (n = 5) (Table 6). The trackway is straight along section analyzed. Average stride length on the right is 82 mm (n = 22) and on the left is 84 mm (n = 26) (Table 7).

Trackways are size-dimorphic according to external width and stride length. Tables 2, 4 and 6 show two size classes, trackway A being larger (mean external width 116 mm, n = 6; mean stride 101 mm, n = 76) and trackways B and C smaller (trackway B mean external width 65 mm, n = 4; trackway B mean stride 78 mm, n = 47; trackway C mean external width 73 mm, n = 5; trackway C mean stride 83 mm, n = 48).

Fig. (5) Exposed bedding-plane textures at trackway locality. A) Microbial mat ‘elephant skin’ texture adjacent to trackway A. B) nearby ripple marks. C) laminar bedding overlaying preserved imprints at a nearby outcrop. D) A portion of trackway A, showing preservation of imprints.

Portions of the track-bearing surface exhibit textures such as “elephant skin” [45] (Fig. 5A, C) and erosional pockets characteristic of microbial mats [45, 46]. Ripple marks (Fig. 5B, D) have been cut through by tracks and associated imprints, and apparently underlie (and therefore were generated prior to) the microbial mat surface. Hagadorn and Bottjer [10] previously noted the presence of trace fossils (Diplichnites, Planolites, Taphrhelminthopsis and Paleophycus) made by tracemakers interacting with wrinkled surfaces and have noted the role of these mats in preservation of trackways. The bedding plane on which the trackways are preserved was part of a high-intertidal or low-supratidal environment, possibly inundated only at spring-tide-like intervals. High intertidal to supratidal habitats have been postulated for microbial mats in similar settings [45]. Tidal evidence in the Potsdam Sandstone in association with arthropod trackways was first noted by Logan[50], who described repeated sequences of water-ripple marks and windblown sand deposits directly below his trackways. Other examples of wind-deposited sand and microbial binding of sediment(biofilms and stromatolites) can be found in the Nepean Formation in and around Ottawa, Ontario, Canada [51]. Herringbone cross-strata of presumed tidal origin are also a common feature of this portion of the formation though not directly associated with these traces.

The organisms were present and producing trackways at (or very nearly) the same time. Ripple marks were produced by wind over a tidal pool. The track producers were probably partially supported by water, where the body of the organism could have been held afloat above the substrate [52]. The presence of the microbial mat texture suggests that these impressions are not undertracks. Although the mat may have formed and then have been subsequently inundated and buried, the presence and depth of a large number of impressions and the medial impression show that any penecontemporaneous sediment overlying the now-exposed bedding plane would not have been very thick. The size and orientation of ripples preserved under the algal mat seem unlikely to be related to the distances between adjacent sections of the intermittent medial track structure. It is postulated that the intermittency of the medial structure was caused by wind-wave action similar to that which caused the sediment ripples. Eventual covering of the bedding plane may have been accomplished by eolian processes [51].wind-oscillated water would be insuffcient to cause this preservation [53]. The presence at this site of trackways associated with microbial mats suggests that the preservation potential may have been enhanced by this association,similar to circumstances described by Hagadorn and Seilacher[54] No evidence of grazing was seen in association with these trackways.

The trackways described here are assigned to Protichnites septemnotatus based on their comparison with Owen’s [25] original description. They are very similar to the trackways illustrated [25, Plate 9] in number of tracks per series, arrangement, and angle of track series to mid-line (Figs. 2, 6). Opposing tracks in the same series set are in-phase, implying that the organism was adapted for swimming rather than terrestrial life [55]. While Owen found it most likely that the animal producing the tracks was hexapodous with bifurcate or trifurcate legs, he did recognize that the association between what he considered trifurcate tracks was not proven and that any animal with any number of legs between six and fourteen could have produced his original Protichnites trackways. Fig. (6) applies labels to Plate 9 of the original description of P. septemnotatus to show that the interpreted arrangement of tracks for the trackways described here can also be applied to the type specimen.

Fig. (6) Plate IX from Owen´s [25] original description of Protichnites septemnotatus. Track series are labeled according to the interpretation applied to the trackways described herein. Direction of travel was toward top of figure. Scale bar is 5 cm.

Ichnotaxa should be based on morphology, not upon the identity of the interpreted producer. Therefore it should be noted that an organism defined as the producer of these particular trackways has no effect on their designation as P. septemnotatus. Keighley and Pickerill [40] summarized very well the criteria upon which a named ichnotaxon must be based. Trewin [48] added that interpreted direction of movement should also be disregarded when defining traces. If direction of movement is an applicable diagnostic tool, new ichnotaxa could be generated from the same trackways simply by interpreting the direction of movement differently, which would be counter to the idea of keeping ichnotaxonomy as unambiguous as possible. Many of the criteria and philosophies behind ichnotaxonomy were standardized by [56] in hopes of improving communication.

Trackways A, B, and C are made up of series consisting of tracks that are similar in order and relative position. These trackways are therefore assigned to the same ichnospecies and are likely to have been produced by the same kind of organism. Variation in substrate or preservation of the exposed bedding plane is minimal, making it improbable that those factors have contributed to making these trackways more similar than they actually are. Trackways B and C were produced under the same environmental conditions (tide-pool wavelets causing discontinuous medial impressions), while trackway A was probably produced under different conditions (calm water allowing a continuous medial impression) and hence at a different time.

We suggest an invertebrate animal with seven pairs of walking limbs, with the possibility of additional limbs held out of contact with the substrate. The regularity and large number of tracks, in addition to their geologic age, rules out most organisms, leaving arthropods as the most likely producers [57, 58]. Although Walcott[36] attributed Protichnites trackways to species of Dikelocephalus, a trilobite, it seems unlikely that a trilobite of this size would have produced such well-defined trackways with no extraneous impressions (e.g., gill marks) in an intertidal pool. Absence of any trace produced by the genal margins of the cephalon is also reason to exclude Dikelocephalus and most other Cambrian trilobites, and Dikelocephalus did not possess a telson. The absence of trilobite body fossils from the Potsdam Formation in the St. Lawrence Valley is well known. Cruziana, a potential trilobite trace, appears only rarely in the section in the Theresa-equivalent March Formation of Ontario, Canada, well above the Potsdam [30].

MacNaughton et al. [19] and Braddy [30] attributed similar trackways to members of the Euthycarcinoidea, which possessed 11 pairs of uniramous pre-abdominal limbs (as reconstructed by Trewin and McNamara [43] and Vaccari et al.[17]) and ranged from the Late Cambrian to Middle Triassic [17]. The euthycarcinoid body plan seems discordant with the trackways presented herein because the limbs of the euthycarcinoids decrease in length both anteriorly and posteriorly from the middle of the preabdomen [17] whereas the limbs in question here appear to have increase in length from anterior to posterior.

Arrangement of the tracks described here is reminiscent of those made by Braddy and Almond’s [55] reconstruction of the walking pattern of the eurypterid Onychopterella augusti, but the Eurypterida were limited to 5 pairs of limbs used for locomotion [59]. The Burgess Shale fauna seems to yield no good candidates for producers that would either be able to support themselves out of water, or be large enough to produce these trackways. The trackways described here are consistent with an organism with generally shorter anterior limbs. A number of tracks could be indicative of bifurcate or trifurcate walking limb termination (Fig. 4A,4B). While these examples have been marked as such, the only evidence for their association comes from their relative location to tracks one stride away in either direction. Depending on the ability of the producer to manipulate the terminal morphology of specific limbs, any apparent reduction from proven bifid or trifid tracks (c.f. Kouphichnium) may be a result of poor preservation. We are confident that all of the trackways described here were produced by the same type of organism. In determining a producer this is presently unimportant; no known fossil exactly conforms to a morphological description able to produce these trackways.

During the Late Cambrian a wide array of arthropod clades were available to supply the track makers including stylonurans, synxiphosurans, phyllocarids [60], aglaspidids [60], trilobitomorphs, euthycarcinoids [17,60], xiphosurans [18] and possible as yet undiscovered merostomes [61], their morphologic requirement being presence of seven or more ambulatory appendages and a telson. When found at the end of a trackway, the producer will finally be certainly known.