Sometimes They Come Back: Recovery and Reinterpretation of a Trackway Slab from the Permian Coconino Sandstone of the Southwestern United States

This article was downloaded by: [Paolo Citton]On: 29 August 2012, At: 11:42Publisher: Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,37-41 Mortimer Street, London W1T 3JH, UK

Ichnos: An International Journal for Plant and AnimalTracesPublication details, including instructions for authors and subscription information:http://www.tandfonline.com/loi/gich20

Sometimes They Come Back: Recovery andReinterpretation of a Trackway Slab from the PermianCoconino Sandstone of the Southwestern United StatesPaolo Citton a , Eva Sacchi a & Umberto Nicosia aa Dipartimento di Scienze della Terra, “Sapienza” Università di Roma, Roma, Italy

Version of record first published: 29 Aug 2012

To cite this article: Paolo Citton, Eva Sacchi & Umberto Nicosia (2012): Sometimes They Come Back: Recovery andReinterpretation of a Trackway Slab from the Permian Coconino Sandstone of the Southwestern United States, Ichnos: AnInternational Journal for Plant and Animal Traces, 19:3, 165-174

To link to this article: http://dx.doi.org/10.1080/10420940.2012.702607

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions

This article may be used for research, teaching, and private study purposes. Any substantial or systematicreproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form toanyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representation that the contentswill be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses shouldbe independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims,proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly inconnection with or arising out of the use of this material.

Ichnos, 19:165–174, 2012Copyright c© Taylor & Francis Group, LLCISSN: 1042-0940 print / 1563-5236 onlineDOI: 10.1080/10420940.2012.702607

Sometimes They Come Back: Recovery and Reinterpretationof a Trackway Slab from the Permian Coconino Sandstone ofthe Southwestern United States

Paolo Citton, Eva Sacchi, and Umberto Nicosia

Dipartimento di Scienze della Terra, “Sapienza” Universita di Roma, Roma, Italy

When a fossil vanishes to a private collection, it must be consid-ered lost to science because, frequently, it is no longer available forstudy. Fortunately some fossils occasionally are regained. We hadthe opportunity to recoup an interesting footprint-bearing slab thatwas part of a private collection in Italy. The specimen, found in 1992near Seligman, Arizona (USA) was described, before disappearing,as one of the best fossil examples of vertebrate (Chelichnus[Laoporus])-on-invertebrate (Octopodichnus) predation. After acareful re-examination of the slab, the primary conclusions of theformer describers are demonstrably groundless. The reanalysis ofthe tracks, as well as peculiar sedimentary structures associatedwith the tracks, allowed obtaining new information about thedepositional environment and the complex interactions betweenthe type of substrate and trackmaker behavior. The re-examinationof the specimen also revealed interesting aspects about trackmakerbiomechanics.

Keywords Tetrapod tracks, Invertebrate tracks, Chelichnus, Permian,Coconino Sandstone, Biomechanics

INTRODUCTIONPaleontologists frequently worry about the disappearance of

fossil specimens of particular importance into private collectionsbecause such specimens are no longer available for empiricalresearch. Occasionally, some fossils come back into publiccollections. This was the case of a footprint-bearing slabfrom the Lower Permian Coconino Sandstone collected nearSeligman, Arizona (USA) (Fig. 1). The specimen has beennoted twice in the literature. The slab was initially describedand figured by Kramer et al. (1995) as one of two knownexamples of Permian vertebrate-on-invertebrate predation. Thesame paper reported that the specimen was part of a privatecollection in Italy, but no information was provided about theowner. Consequently, the specimen had to be considered lost

Address correspondence to Paolo Citton, Dipartimento di Scienzedella Terra, “Sapienza” Universita di Roma, P.le A. Moro 5 00185Roma, Italy. E-mail: [email protected]

to science because of the impossibility of further examination.Indeed, in the subsequent reference Hunt and Lucas (1998)stated that they were unable to examine the specimen firsthand,and their illustration of the specimen (Hunt and Lucas, 1998,fig. 3) was simply a copy of the original figure of Kramer et al.(1995, fig. 3).

Purely by chance, we were entrusted with the task of docu-menting a former private collection seized by law enforcementofficers and currently in the custody of the Land authorities onCultural Heritage (Soprintendenza Archeologica per le Marche,Ancona, Italy). The track specimen was part of this collection,and we recognized it. So, for the second time, the specimen isnow publicly reposited (Museo di Serrapetrona, Macerata, Italy,specimen MSP 309) and available for study. That serendipitousseries of events allowed us to re-examine this crucial specimen,with some unexpected results.

PREVIOUS INTERPRETATIONSAt first examination, we recognized some substantial prob-

lems with the description of the specimen by Kramer et al.(1995), so the need for a thorough reinvestigation was imme-diately evident. Because of the complexity of our analysis, ourresults, which differ from those of the initial study, necessitatedalmost complete redescription and illustration of the specimen(Fig. 2). On the slab, Kramer et al. (1995) described twovertebrate trackways (L1 and L2) that they assigned to theichnogenus Laoporus and a single invertebrate trail (trackwayOd) that they recognized as Octopodichnus didactylus (Fig. 2A).In the present redescription, we retain these trackway labels butadd new ones for tracks undescribed in the previous paper. Theseinclude a third tetrapod trackway (L0) and many invertebratetracks and partial trackways (IT; Fig. 2B).

Before proceeding to the description, it seems useful to clar-ify a minor nomenclatural matter. Kramer et al. (1995) ascribedthe tetrapod tracks on the slab to the ichnogenus Laoporus Lull,1918, an ichnogenus that, at that time, was considered endemicto North America. Laoporus was subsequently recognized asjunior synonym of Chelichnus Jardine, 1850 by McKeever and

165

Dow

nloa

ded

by [

Paol

o C

itton

] at

11:

42 2

9 A

ugus

t 201

2

166 P. CITTON ET AL.

FIG. 1. Museo di Serrapetrona (MSP) 309, the track-bearing slab from the Lower Permian Coconino Sandstone of Arizona recovered from a private collection.Scale bar in cm.

Haubold (1996), as was noted and rectified in the mentionof the specimen by Hunt and Lucas (1998). Thus, we useonly the name Chelichnus in the present treatment. A furthersmall problem is the attribution of this ichnogenus to a caseid(Haubold, 1971) or, more conservatively and inclusively, to abasal synapsid (Haubold, 2000) trackmaker; resolution of thisproblem is beyond the scope of this paper but is under study.

Previous interpretation of MSP 309Kramer et al. (1995) considered the specimen to be extremely

important as a rare case shedding light on predatory aspects ofextinct animals. Apart from this specimen, few other examplesof predatory interactions are known in the continental Permianichnological record among either vertebrate or invertebratetracks (Lockley and Madsen, 1993; Hunt et al., 1994; Krameret al., 1995). In MSP 309, the behavioral interaction was inferredfrom an Octopodichnus trail that seemingly terminated at itsintersection with a Chelichnus trackway. The scenario proposedby Kramer et al. (1995) (Fig. 2A) depicted a Chelichnustrackmaker L1 crossing a sand slope (identified on the basis ofcrescentic structures clearly evident on the slab) from roughlythe same point as Octopodichnus trail Od, which was headeduphill. The authors stated that there is no evidence for interactionbetween these two trackmakers. At a second point, anotherChelichnus trackmaker (L2), the presumptive predator, makesits entry, proceeding broadly in the same direction as L1. At astill later point, trackway L2 both truncates the Octopodichnustrackway and exhibits an interruption in its stride between the

second and third manus-pes print sets, particularly on the rightside. The interruptions of print sets 2 and 3 of L2 and thedisappearance of trackway Od were concluded to be evidenceof a predator catching its prey.

A few years later, the specimen was mentioned by Huntand Lucas (1998), who specified six criteria necessary for therecognition of predatory behavior from tracks. They seemedunconvinced by the interpretation of the specimen presentedby Kramer et al. (1995). After direct re-examination of thespecimen, we agree that the pessimism expressed by Hunt andLucas (1998) was entirely justified because the traces on theslab surface correspond only partially to the drawing of Krameret al. (1995), and many features of the specimen do not supporttheir interpretation.

Redescription of MSP 309The tracks are on a subtrapezoidal slab around 90 cm long

and 60 cm wide. It is regularly 4–5 cm thick and made of aliver-red in color, well-cemented and well-sorted, fine-grainedsandstone. The sand grains are so fine (around 200 µm) andso tightly packed that the undisturbed portions of the surfaceare almost completely smooth. The surface preserves manyvertebrate and invertebrate tracks running in different directions.There are, in addition, three parallel linear features that weascribe to small-scale sand avalanches, all of which appear tohave been gently smoothed by wind action after their formation.Such features, along with the fine-grained and well-sorted sand,are typical of sand dunes and have been already described

Dow

nloa

ded

by [

Paol

o C

itton

] at

11:

42 2

9 A

ugus

t 201

2

SOMETIMES THEY COME BACK 167

FIG. 2. Schematic map of track-bearing slab MSP 309 from the LowerPermian Coconino Sandstone of Arizona. A. Interpretation of the specimenmade by Kramer et al. (1995; redrawn from that source and slightly modified);a-b-c emphasize the sequence of events described by Kramer et al. (1995).a) Trackmaker L1 enters the scene crossing the sand slope; b) Trackmaker L2enters the scene traversing the slope; c) Presumed termination point of trackwayOd. B. The present interpretation. Trackway L0 denoted in gray; trackway L2denoted in black; dashed lines indicate directions of sand avalanches.

in the Coconino Sandstone (McKee, 1947). The avalanchegenesis can easily be attributed to sliding of sand-flows onthe lee-sides of dunes whose crests are perpendicular to thedominant wind direction, induced when grains are moved forwhatever reason or when they overtop the dune crest. These sandflows run in the direction of maximum slope. Aside from thesesedimentological features (avalanches, fine-grained sand, andsorting), other features affecting footprints (discussed below)strongly support eolian deposition.

Because of the relationships among the various trackwaysand because they interfere with each other, in our newanalysis we describe the trackways in the order in which,in our interpretation, they were impressed. Moreover, in thedescriptions, we identify the tracks in sets with consecutive

FIG. 3. Chelichnus trackway L0 on slab MSP 309 from the Lower PermianCoconino Sandstone of Arizona. “Ringed slides” affecting sets 2-3-4-5. Scalebar in cm.

numbers. Terminology and measurements follow Leonardi(1987).

Trackway L0. The first trackway to have been imprinted,as determined by relative overprinting of trackways across theslab, is designated L0 (Fig. 2B; Fig. 3). The trackway, which is70 cm long and 10 cm wide, comprises nine sets of manus-pesprint pairs (two of which are incomplete) and, though readilyapparent, was not described by Kramer et al. (1995). In theirdrawing of MSP 309 (Kramer et al., 1995, fig. 3), a portion ofthis trackway is depicted as three unintelligible markings. Thesecorrespond to the right side of the trackway as recognized here;the left side is not drawn. The trackway comprises only poorlypreserved prints but is extremely important both because it isthe key to determining the relative timing of most of the trackson the slab and because it allows explanation of the type ofpreservation and the morphologies of other traces.

Trackway L0 is subparallel to the avalanche lineations. Eventhough the individual footprints are not as detailed as theothers tracks on the slab, their arrangement in alternating setsand their relative dimensions indicate they were made by atetrapod that was at least twice the linear dimensions of themakers of trackways L1 and L2. The primary importance ofthis trackway derives from its type of preservation, which ischaracterized by peculiar sedimentary features. Each manus-pesset is preserved as a single, roughly ovoid structure that iselongate in the direction of travel. Each ovoid structure is deepestin its center and is composed of a series of subconcentricallyarranged half-rings cranially and a V-shaped structure caudally.

Dow

nloa

ded

by [

Paol

o C

itton

] at

11:

42 2

9 A

ugus

t 201

2

168 P. CITTON ET AL.

FIG. 4. Photograph of a “ringed slide” on set 3 of Chelichnus trackway L0 onslab MSP 309 from the Lower Permian Coconino Sandstone of Arizona; A-B-Cemphasize successive stages of formation. A. First sand flow event filling thecavity left by the foot; B, C. Subsequent, rounded, half-ring structures formedby subsequent flow events. Scale bar in cm.

The surfaces of these structures, affecting both the sets andthe surroundings for an area four or five times the assumeddimensions of the set, are less smooth than any of the rest of theslab. These structures, which we define as “ringed slides,” weregenerated by the animal walking on the surface of a substratecomposed of dry, fine-grained, loose and barely stabilized sand,close to its angle of repose. The sand grains, in the small areawhere the animal trod, lost their stability and slid downslope.This movement triggered a multiphase process (Fig. 4), in someways similar to that shown by Loope (2006) in wind-blowncross-strata of the Lower Jurassic Navajo Sandstone. First,flowing sand that formed the walls of the footprint flowedinto and filled the cavity left by the foot, initiating a secondsliding event, external to and surrounding the first, had formedthe rounded, half-ring structure, a kind of “micro-barchan” inwhich the wings are better developed in the downslope direction.The process was then repeated, enlarging the destabilized areaand generating further rings. This multiphase process, variablein time and extent, was repeated until a new equilibrium wasreached. Obviously, this sort of process is possible only onan inclined surface on which fine-grained and loose sedimentcan freely slide downward. This reconstruction of events alsoenables us to infer that the trackways are true tracks (contra

Kramer et al., 1995) and that they were registered on a surfaceon which the loose sand grains were nearly stable, probably thelee side of a dune with a slope inclined up to 35◦ (Ciccacci,2010). This is, obviously, a maximum value, whereas manysurfaces of the Coconino Sandstone dip at lower angles (McKee,1947; Hunter, 1977). Taking into account the direction of themaximum dip, deduced from the aforementioned avalanches,the features of the “ringed slides,” and the trackway midline, itwas possible to establish that the trackmaker was walking moreor less directly uphill.

Trackway Od. Octopodichnus tracks on the slab compriseboth the long trackway interpreted by Kramer et al. (1995)as having been terminated and several isolated prints, ofwhich one set is particularly interesting. We designate thepreviously described portion of the trackway Od1, and thenewly recognized, isolated set Od2 (Fig. 2B). Trackway Od1was imprinted after trackway L0 because it overprints with thelatter. This invertebrate trackway consists of 19 sets of tracksthat are generally well preserved, and at least 14 of which arecomplete (Fig. 5). Each set is made of four imprints, three ofwhich are arranged roughly in a line (imprints 1-2-3 accordingto the convention of Sadler, 1993) converging on the midline(imprint 1 is the nearest the midline) (Fig. 6). The fourth imprintis always medial, but its position (i.e., relative distance fromimprint 3) is highly variable among sets. The imprint lines lie ata roughly 45◦ angle to the midline, although this angle is highlyvariable on the trackway. All the sets alternate with each other.All the imprints appear as simple, singular pits or, in the bestpreserved traces, as bifurcate pits.

The trackway is clearly recognizable, though divided intofour portions composed respectively of four, six, six and threetrack sets; it goes uphill, more or less parallel to L0 in its lowerportion, but at the end of its second portion, it curves gentlyto the right, crossing trackways L0 and L2 and reaching theslab margin. After the first four sets, the trackway is effacedbut reappears a few cm farther along (sets 5–10). The thirdportion (sets 11–16) records the action of the trackmaker whilecrossing the destabilized area generated by trackmaker L0; inthat area, the prints of set 11 are elongate because the animal leftscratch marks or slid on the loose sand (Fig. 6B). Such repeateddisappearance and re-appearance events in themselves renderless credible the putative “termination” identified by Krameret al. (1995) and seem instead only to indicate that preservationof such tiny invertebrate traces strongly depended on the variablesubstrate conditions. Indeed, after having disappeared a secondtime, the trackway reappears: the positions of the last three setsare consistent with the direction of the curved midline of the trail.These three sets are clearly evident a few centimeters beyond thealleged “termination” point identified by Kramer et al. (1995)and the second set of trackway L2 (Fig. 2B; Fig. 6C).

Isolated Octopodichnus trackway Od2 (Fig. 2B), clearlyevident near trackway L1, is reversed with respect to the otherOd tracks, showing that the animal was moving downhill withoutchange in morphology and preservation. This complete left set

Dow

nloa

ded

by [

Paol

o C

itton

] at

11:

42 2

9 A

ugus

t 201

2

SOMETIMES THEY COME BACK 169

FIG. 5. Schematic drawing of Octopodichnus trackway Od1 on slab MSP 309from the Lower Permian Coconino Sandstone of Arizona. Light gray shadingindicates areas destabilized by the L0 track maker. Dashed line indicates theOd1 trackway midline. A. Set shown in Fig. 6A; B. Set shown in Fig. 6B; C.Set shown in Fig. 6C.

is quite interesting, since downhill tracks on a slope would havea rare chance of preservation (McKee, 1947).

Trackway L2. Chelichnus trackway L2 interferes with bothL0 and Od1, and therefore was imprinted after both of them.It is composed of 20 manus-pes sets and spans a total distanceof about 60 cm; its maximum width is 5 cm (Fig. 2B). Thetrackmaker crossed the dune side at an angle of nearly 20◦

with respect to the direction of maximum slope (as inferredfrom the sliding sand avalanches and disruption features of

FIG. 6. Individual track sets from Octopodichnus trackway Od1 on slab MSP309 from the Lower Permian Coconino Sandstone of Arizona. A. Close-up ofthe best preserved set. B. Close-up of the elongate tracks on the destabilizedarea. C. Close-up of a set beyond the putative “termination” point.

Dow

nloa

ded

by [

Paol

o C

itton

] at

11:

42 2

9 A

ugus

t 201

2

170 P. CITTON ET AL.

FIG. 7. The central portion (sets 9–16; see Fig. 2B) of Chelichnus trackway L2 on slab MSP 309 from the Lower Permian Coconino Sandstone of Arizona.Scale bar in cm.

trackway L0). The first two track sets are normal, albeit shallow,impressions. The third set, and the pes prints of the fifth and sixthsets, were imprinted on sand already destabilized by the L0trackmaker. They appear to have been filled by sliding sand justafter their formation and so can barely be recognized beneaththat cover. From the seventh set onward, the trackway seemsfairly regular, but careful examination shows many differencesbetween footprint and trackway parameters, indicating differentpreservations of the prints on the right and left sides (Fig. 7).

Footprints on the left side are well preserved; pedes aretridactyl to pentadactyl and semiplantigrade and have short,broad soles and roughly straight, short, pointed digits. Themanus prints are smaller than those of pedes, are tridactyl topentadactyl and semidigitigrade to digitigrade, and have shortand slender digits. Sometimes the manus prints are oversteppedby the pes prints. In sets 3, 9, 11, and 13, the left pes printsinclude sand crescents that have their concave sides orienteduphill. Similar although less-clearly preserved structures arealso associated with the pes prints in sets 1, 7, 15, 17, and 19.

In contrast, on the right side of the trackway, footprintsare much less detailed: both manus and pes prints consist ofsubcircular, shallow impressions without traces of individualdigits, except in the manus prints of sets 2 and 8 and the pesprints of sets 4 and 14. On this side, sand crescents are locallyapparent in the pes prints of sets 2, 10, and 14, but their concavesides are oriented in the direction of travel, not uphill as wouldbe expected.

The relative positions of the manus and pes prints in eachset are quite unusual: on the right side, each manus liescraniolateral to its corresponding pedes whereas on the left

side, they lie craniomedial to their corresponding pes prints.In other, well-known Chelichnus trackways, the manus printsare cranial to their corresponding pes prints (Gilmore, 1927;Haubold, 1971; McKeever and Haubold, 1996). Moreover, thedigits in trackway L2 are oriented uphill in all the footprints sothat the left pedal digits turn inward nearly 90◦ with respectto the direction of movement. This peculiar characteristicwas previously noted by Hunt and Lucas (1998) and will befurther discussed below. The upward orientation of digits in theCoconino tracks was also a key point in the Brand and Tang(1991) subaqueous interpretation of the Coconino Sandstonedepositional environment.

Trackway L1. Trackway L1, which is 67 cm long and about6 cm wide, consists of 15 preserved manus-pes sets (separatedby four, irregularly spaced gaps) and is roughly parallel totrackway L2. In the drawing of Kramer et al. (1995, fig. 3),this trackway is not completely drawn, lacking the first andsecond sets that lie near the edge of the slab and are quite clear(Fig. 2B). Immediately after these first two sets, the trackwaycrosses both trackways L0 and Od, but its third and fourth setsare completely missing, so it is impossible to establish eitherinterference among them or a relative chronology. Midwayalong its length, the trackway becomes more regular, while inthe final part footprints are less clear or lacking on both sides,represented only by small, shallow, incomplete footprints. Theonly exception is the 15th.

The most important feature of trackway L1 is, once again,the nature of its preservation: it shows both of the mainfeatures previously described for trackways L0 and L2. Itsbetter-preserved footprints are similar in morphology to those of

Dow

nloa

ded

by [

Paol

o C

itton

] at

11:

42 2

9 A

ugus

t 201

2

SOMETIMES THEY COME BACK 171

FIG. 8. Set 9 of Chelichnus trackway L1 on slab MSP 309 from the LowerPermian Coconino Sandstone of Arizona, showing both the ringed slide (top)and sand crescent (bottom).

trackway L2, possessing short, broad soles with short, straight,and slender digits. Another feature shared with trackway L2is the relative orientation of its footprints, which are orienteduphill roughly 90◦ to the direction of travel, indicating apeculiar, slope-dependent locomotory mode. Moreover, somesets are characterized by well-defined sand crescents dippingin directions that come close to that of slope, as seen intrackway L2. These features probably suggest that the sandbore a certain level of moisture, necessary to ensure sufficientcohesion between sand grains during deformation. On somefootprints, including the most poorly preserved, the samesedimentary structures that adorn the footprints in trackwayL0 are again seen. In this case, the “ringed slides” are lesswell preserved than in trackway L0 while the “V”-shapedstructures are always absent. The coexistence of a “ringed slide”at the front of a footprint and of a sand crescent at its rear isclearly appreciable in set 9 (Fig. 8). Such a wide variability infootprint preservation over a very short distance and in the sametrackway, combined with the co-occurrence of the describedsedimentary structures, suggests an extreme yet synchronousheterogeneity of the surface over a millimetrical or centimetricalarea. This complicated scenario, in which water content playedan important role, allows better interpretation of track makerbiomechanics (see below).

Tracks IT. In addition to the Od Octopodichnus trackways,invertebrate tracks of two morphologies are also present on the

FIG. 9. Unidentified invertebrate trackway IT2 on slab MSP 309 from theLower Permian Coconino Sandstone of Arizona. Scale bar in cm.

slab but were not previously reported; we refer to them as theIT1, IT2, and IT3 trackways (Fig. 2B). These tracks are scatteredacross the track-bearing surface apparently without any pattern,suddenly vanishing after two or three strides.

Trackway IT1, which exhibits the first of the two morpholo-gies, consists of only 12 circular imprints that are grouped inthrees and have alternate symmetry. They are located very closeto the end of the third set of trackway Od1. Trackways IT2 andIT3 exhibit the second morphology, which consists of groupingsof two tapered, tear-drop-shaped imprints 4–6 mm in length and2–3 mm in width. These paired impressions form V-shapedpatterns. They lie on the downslope side of the slab, close totrackways L1 and L0 (IT2; Fig. 9) and on the upslope side, nearthe left side of trackway L2 (IT3). The arrangement of imprintsand the relative stride lengths in each track row in both IT2 andIT3 suggest that these tracks probably constitute two different,incompletely preserved, trackways left by two trackmakersof the same type. Trackway IT2 is sumperimposed on thesand crescent of the fifth track set of trackway L1, indicatingthat the trackmaker passed by after the L1 trackmaker, thusadding further information about relative chronology. Partialpreservation of these trackways supports the assumption thatpreservation of the small invertebrate traces, like that of thetetrapod trackways, was greatly affected by substrate conditions.

In sum, re-examination of the specimen revealed a scenariovery different from that proposed by Kramer et al. (1995).First, there are three tetrapod trackways present on the slab,not two. Second, the “prey” trackway does not terminate butrather continues after crossing L2, so no predation is indicated.Third, because “ringed slides” can form only at the dune surface,all the tracks must be true tracks. Thus, all the main statementsof Kramer et al. (1995) are unsubstantiated.

BIOMECHANICSMorphological inconsistencies among otherwise similar

trackways, and between footprints in the same trackway, revealcomplex interactions between the type of substrate, trackmaker

Dow

nloa

ded

by [

Paol

o C

itton

] at

11:

42 2

9 A

ugus

t 201

2

172 P. CITTON ET AL.

behavior, and the resultant footprint morphologies. Trackway L2was the most informative in this regard because its variabilityallowed us repeatedly to discard different, simple explanationsand to hypothesize a more complex one. As previously noted, thetrackway exhibits many differences from a “normal” Chelichnustrackway (Haubold, 1971; McKeever and Haubold, 1996) as aresult of moving across a dune slope, as noted by Hunt and Lucas(1998). There are seven main points, all of them parsimoniouslyexplained by a single hypothetical scenario:

1) In Chelichnus, the manus usually is placed directly in frontof the pes, but in L2 the relative positions of the manus andpes are highly variable in position and mode of impressionalong the whole trackway and on both left and right sides.

2) The manus impressions on the right side were displacedlaterally with respect to the pes impressions, while the leftmanus is medially placed.

3) The soles and digits of both the manus and pes impressionsare oriented uphill, roughly perpendicular to the traveldirection.

4) Sand crescents are pronounced on the left pes prints but arebarely visible on the right side of the trackway.

5) The sand crescents in sets 3, 7, 9, 11, 13, 15, and 19 areoriented differently than those in the first two sets as well asin other Chelichnus trackways on the slab.

6) The pes frequently oversteps the manus on the left side ofthe trackway, but it does not on the right side.

7) The right and left track sets are not consistently arrangedalong the trackway but instead fall into one of two types.In one type left and right prints are uniformly separated byless than 2 cm; in the other type the left and right prints areseparated by more than 2.5 cm.

Based on these features, other parameters (e.g., foot di-mension, stride length), features of other trackways, and thesedimentary structures apparent on the slab (e.g., sand crescents,avalanches and sliding traces), we infer the following hypothesisfor trackway L2: a small tetrapod trackmaker (a few centimeterslong from shoulder to hip) slowly climbed in a constant directionnearly 20◦ to the slope of a dune that was inclined 30–35◦.Neither body nor tail traces is present, so the body was held welloff the ground. The trackmaker did not have a limb structuresimilar to that of the extant lizards but a more derived one,as shown by footprint morphology and trackway pattern, andprobably did not have the typical sprawling posture. The limbswere probably semi-erected, and the tail was either very shortor was held off the surface.

During the short portion of its odyssey that is preserved onthe slab, the trackmaker had to oppose the force of gravitywhile simultaneously displacing itself forward on a dynamic,constantly changing substrate. It further encountered a secondissue: differing vertical distances of the limbs on either side fromthe substrate on the longitudinal plane (Fig. 10). This verticaldistance, due to the inclined substrate was minimal on the uphillside and maximal on the downhill side.

FIG. 10. Cartoon showing the position of a Chelichnus track maker lefthindlimb during a cycle in dorsal (left) and caudal (right) perspectives.

During its march, the trackmaker continuously slid, if onlyslightly, down the inclined surface, and, at each step, all its feethad to be displaced upward to balance downslope slippage and tomaintain forward progression. This accounts for the anomalouspositions of the manus, which are displaced upslope relative totheir respective pedes (points 1–2).

For the same reason, the left pes, while applying to the bodythe necessary force to maintain its direction, had to be eitherraised to be placed very near the body, or lowered, before beingpushed down (Fig. 10C). On an inclined surface, and with its

Dow

nloa

ded

by [

Paol

o C

itton

] at

11:

42 2

9 A

ugus

t 201

2

SOMETIMES THEY COME BACK 173

FIG. 11. Map of track-bearing slab MSP 309 from the Lower Permian Coconino Sandstone of Arizona showing the directions of sliding vectors. Solid arrowsindicate directions based on sand avalanches, dashed arrows indicate directions based on “ringed slides,” and dotted arrows indicate directions based on sandcrescent sliding.

right side in the normal position, the left side of the body of thetrackmaker would have been far enough off the substrate thatthe left hind limb was obligated to effect a major extension, andthe foot was pushed down to the mechanical limits determinedby the total length reached by the leg at the end of the extensionphase. At the end of that movement, pivoting of the weight-bearing part of the foot as sand slipped downslope, rotated itsdigits upslope (point 3). This complex dynamic is consistentwith the structure and function of the tarsus in Diadectidae(Berman and Henrici, 2003).

Thus, the inward position of the digits in the left footprintsdoes not record the position of foot at the touch-down phase,but only the last position held by the foot just before the takeoffphase. The larger sand crescents on the left side of the trackwaymust have been made during the takeoff phase because the lefthind leg, at the end of kick-off, was also strongly inclined withrespect to the surface (Fig. 10D). The sand crescents thereforedo not record simply the dip of the surface but also the thruststhe animal must have developed in order to oppose gravity, toadvance, and to compensate for the loss in position due to itsslide down (point 4). This seems to be confirmed by the differentorientations of the sand crescents in the first two sets of trackwayL2 compared to those in sets 3, 7, 9, 11, 13, 15, and 19, as wellas by the sand crescent directions measured in other footprintson the slab. Indeed, directions taken from sand crescents on theslab surface can diverge by more than 110◦, and depart fromthe maximum dip up to 60◦ (Fig. 11). This datum explains the

observed differences (point 5) and emphasizes that, in general,sand crescent direction cannot be used directly to determine thelocal dip (contra Lockley and Hunt, 1999, among others): thedirection is affected by numerous, different factors.

The other differences between the right and the left sides ofthe trackway (points 6–7) may be explained by the movementof the animal along the slope. Its uphill appendages partly boreits weight but could barely move due to constraints producedby the closeness of the substrate. In contrast, the downhilllimbs were forced to take a longer step. With an apparentdip of around 20◦ and with the feet placed 3.5 cm apart fromeach other (the maximum width of the trackway), the verticaldifference between the positions of the feet was around 10 mm,a huge distance for an animal of the L2 trackmaker size. Theperfect straightness of the trackway, even under such complexenvironmental conditions, seems to indicate strongly reactivebehavior of the trackmaker. The trackmaker was agile and verywell adapted to such conditions.

CONCLUSIONSA Permian, track-bearing slab once unavailable to scientific

scrutiny in a private collection has been recovered and returnedto a public repository. A meticulous review of the tracks allowedreconstruction of a complex track-making scenario, one that isalmost the polar opposite of an earlier hypothesis proposed byKramer et al. (1995) for the same specimen.

Dow

nloa

ded

by [

Paol

o C

itton

] at

11:

42 2

9 A

ugus

t 201

2

174 P. CITTON ET AL.

Our re-examination of the specimen also revealed someinteresting aspects of trackmaker biomechanics. Analysis ofdifferences in footprint morphology, extramorphological fea-tures, and trackway patterns and parameters revealed complexinteractions between the type of substrate and trackmakerbehavior. These interactions also concern the process offormation of sand crescents, casting doubt on the reliabilityof these features in determining dune slope direction.

Vertebrate and invertebrate footprints on the slab are char-acterized by a style of preservation and associated sedimentarystructures explicable only by the simultaneous occurrence ofdifferent substrate conditions (i.e., dry sand and damp sand indifferent parts of the tracked surface). These different conditionsaffected different footprints in the same trackway. We attributethis complex diversity of substrates to a desert environment inwhich such different moisture contents could co-occur.

Such an environment exists in the modern Namib Desert,where the water content of sediments changes at least twicea day. During the night, sea breezes from the Atlantic Oceanincrease the degree of moisture, up to many tens of kilometersinland (Kimura, 2005), allowing a relative and temporarydampness of surface and the survival of many small organisms(Henschel and Seely, 2008). Obviously this condition iscounteracted by daytime evaporation that creates the famousthick fog and returns sand bodies to a dry condition in whichthe grains are loose. A Permian desert, east of an inland sea,in the specimen source area (i.e., southwestern United States)is also consistent with various paleogeographic (e.g., Scotese,1997) and paleoclimatologic models (e.g., Parrish and Peterson,1988).

ACKNOWLEDGMENTSWe thank the Land authorities on Cultural Heritage (So-

printendenza Archeologica per le Marche) that allowed thestudy of the specimen. A special acknowledgement is dueto our colleague S. Ciccacci (University of Rome), specialistof desert geomorphology, for the interesting discussion. TheCommunity and the Mayor of Serrapetrona and the carabineersof the Nucleo Tutela Patrimonio Culturale di Ancona selflesslyhelped us during the study. Mostly, we are indebted to JamesFarlow (Indiana–Purdue University Fort Wayne, USA) for hisenlightening and priceless comments and suggestions. Anyerrors in the text must be attributed solely to the authors.

REFERENCESBerman, D. S. and Henrici, A. C. 2003. Homology of the astragalus and structure

and function of the tarsus of Diadectidae. Journal of Paleontology, 77:172–188.

Brand, L. R. and Tang, T. 1991. Fossil vertebrate footprints in the CoconinoSandstone (Permian) of northern Arizona: evidence for underwater origin.Geology, 19: 1201–1204.

Ciccacci, S. 2010. Le Forme Del Rilievo: Atlante Illustrato di Geomorfologia.Milano: Mondadori Education; Roma: Sapienza Universita di Roma,514 p.

Gilmore, C. W. 1927. Fossil footprints from the Grand Canyon: Secondcontribution. Smithsonian Miscellaneous Collections, 80: 1–78.

Haubold, H. 1971. Ichnia amphibiorum et reptiliorum fossilium. In Kuhn, O.(ed.), Handbuch der Palaoherpetologie, 18: 1–122.

Haubold, H. 2000. Tetrapodenfahrten aus dem Perm – Kenntnisstand undProgress 2000. Hallesches Jahrbuch fur Geowissenschaften, B, 22:1–16.

Henschel, J. R. and Seely, M. K. 2008. Ecophysiology of atmospheric moisturein the Namib Desert. Atmospheric Research, 87: 362–368.

Hunt, A. P. and Lucas, S. G. 1998. Ichnological evidence for tetrapod predationin the Paleozoic: Is there any? In Lucas, S. G., Estep, J. W., and Hoffer, J. M.(eds.), Permian stratigraphy and paleontology of the Robledo Mountains,New Mexico. New Mexico Museum of Natural History and Science Bulletin,12: 59–62.

Hunt, A. P., Lockley, M. G., and Lucas, S. G. 1994. Trackway evidence fortetrapod predation from the Early Permian of southern New Mexico. NewMexico Geology, 16: 84–85.

Hunter, R. E. 1977. Basic types of stratification in small eolian dunes.Sedimentology, 24: 361–387.

Kimura, K. 2005. Origin of the fog in Namib Desert in dry season. AfricanStudy Monographs, Suppl. 30: 57–64.

Kramer, J. M., Erickson, B. R., Lockley, M. G., Hunt, A. P., and Braddy, S.J. 1995. Pelycosaur predation in the Permian: Evidence from Laoporustrackways from the Coconino Sandstone with description of a new speciesof Permichnium. In Lucas, S. G. and Heckert, A. B. (eds.), Early Permianfootprints and facies. New Mexico Museum of Natural History and ScienceBulletin, 6: 245–249.

Leonardi, G. 1987. Glossary and Manual of Tetrapod Footprint Palaeoichnology.Departamento Nacional da Producao Mineral, Brasilia, 75 p.

Lockley, M. G., with Hunt, A. P. 1999. Dinosaur Tracks and Other FossilFootprints of the Western United States. Columbia University Press, NewYork, 338 p.

Lockley, M. G. and Madsen, J. H. 1993. Early Permian vertebrate trackwaysfrom the Cedar Mesa Sandstone of Eastern Utah: Evidence of predator-preyinteraction. Ichnos, 2: 147–153.

Loope, D. B. 2006. Dry-season tracks in dinosaur triggered grainflows. Palaios,21: 132–142.

McKee, E. D. 1947. Experiments on the development of tracks in fine cross-bedded sand. Journal of Sedimentary Petrology, 17(1): 23–28.

McKeever, P. J. and Haubold, H. 1996. Reclassification of vertebrate trackwaysfrom the Permian of Scotland and related forms from Arizona and Germany.Journal of Paleontology, 70: 1011–1022.

Parrish, J. T. and Peterson, F. 1988. Wind directions predicted from globalcirculation models and wind directions determined from eolian sandstonesof the western United States—a comparison. In Kocurek, G. (ed.), LatePaleozoic and Mesozoic Eolian Deposits of the Western Interior of theUnited States. Sedimentary Geology, 56: 261–282.

Sadler, C. J. 1993. Arthropod trace fossils from the Permian DeChelly Sandstone, northeastern Arizona. Journal of Paleontology, 67:240–249.

Scotese, C. R. 1997. Paleogeographic atlas. PALEOMAP Project, University ofTexas. http://www.scotese.com/

Dow

nloa

ded

by [

Paol

o C

itton

] at

11:

42 2

9 A

ugus

t 201

2