A comparison between dinosaur footprints from the Middle Jurassic of the
Isle of Skye, Scotland, UK, and Shell, Wyoming, USA
N. D. L. CLARK1 & M. K. BRETT-SURMAN2
1Hunterian Museum and Art Gallery, University of Glasgow, University Avenue, Glasgow, G12 8QQ,
UK (e-mail: nclark@museum.gla.ac.uk)
2Smithsonian Institution, Department of Paleobiology, PO Box 37012, MRC 121 Washington, DC
20013-7012, USA
Synopsis
Measurements of Middle Jurassic tridactyl dinosaur tracks from the Bathonian, Lealt
Shale, Valtos Sandstone, Duntulm and Kilmaluag formations of the Isle of Skye, UK, are
compared to the same measurements taken for dinosaur footprints from the Bajocian,
Gypsum Spring and the Bathonian, Sundance Formation of the Bighorn Basin, Wyoming,
USA. Principal component analysis of the data suggests that the smaller footprints from the
Valtos Sandstone and Kilmaluag formations are indistinguishable from the footprints of
the Sundance Formation. The single footprint from the Lealt Shale Formation is similar to the
larger footprints from the Valtos Sandstone Formation. The footprints from the Duntulm and
Gypsum Springs formations form distinct groupings from all other footprints. Four di?erent
groupings of dinosaur footprints can be recognized from the principal component analysis
that may represent at least four di?erent types of dinosaur.
Introduction
Dinosaur footprints are well known from the Middle
Jurassic rocks of the Isle of Skye, Scotland, UK (Clark
& Barco Rodriguez 1998; Andrews & Hudson 1984;
Clark et al. 2004, 2005; Marshall 2005) and the Big
Horn Basin, Wyoming, USA (Kvale et al. 2001, 2004;
Breithaupt et al. 2004) (Fig. 1).
The first recorded occurrence of dinosaur remains on
the Isle of Skye was the discovery of a large 49 cm long
footprint from the Lealt Shale Formation (Bathonian)
at Rubha nam Brathairean in 1982 (Andrews & Hudson
1984; Delair & Sarjeant 1985) (Fig. 2(1)). In 1996,
further footprints were found on a fallen block of the
overlying Valtos Sandstone Formation (Bathonian) near
to the original locality (Clark & Barco Rodriguez 1998;
Clark 2001a, 2004, 2005). Other footprints from the
Valtos Sandstone Formation have been found at Dun
Dearg and Kilt Rock, near Valtos (Clark et al. 2005)
(Fig. 2(2, 3)) and from a locality north of Elgol in the
southern part of the Isle of Skye (Marshall 2005) (Fig.
2(6)). The footprints from both these locations are
much smaller (<30 cm length) and have triangular claw
impressions rather than the broad spatulate digits of the
first recorded footprint from the Lealt Shale Formation.
Further footprints have been found since then in the
Duntulm Formation (Bathonian) at An Corran, Sta?n
Bay (Fig. 2(4)) and the Kilmaluag Formation (Bathonian)
at Score Bay, north of Uig (Clark 2003, 2005; Clark
et al. 2004, 2005) (Fig. 2(5)). The Duntulm Formation
footprints are all large footprints up to 53 cm in length
with narrow digits and triangular claw impressions
(Clark et al. 2004). These also di?er from the Lealt Shale
Formation footprint and are thought to have been
produced by a large theropod. In late 2002, dinosaur
footprints from the Kilmaluag Formation were discov-
ered on loose blocks of sandstone, as well as in situ, on
the foreshore at Lub Score, NW Trotternish Peninsula,
Isle of Skye. The majority of these footprints are less
than 14 cm long, and are closely associated with larger
footprints (about 22 cm long) of what seems likely to
be the same ichnospecies (Clark et al. 2005). These
footprints are stratigraphically younger than any other
dinosaur remains found in Scotland.
Dinosaur bones are also known from Scotland. A
theropod tibia was found in the Broadford Beds Forma-
tion (Hettangian) in the Strathaird Peninsula, southern
Isle of Skye (Benton et al. 1995), a thyreophoran ulna
and radius came from the Bearreraig Sandstone Forma-
tion (Bajocian) at Bearreraig Bay, northern Isle of Skye
(Clark 2001b), and cetiosaur bones and a coelophysoid-
grade tail bone were discovered in the Valtos Sandstone
Formation (Bathonian) at Dun Dearg near Sta?n
(Clark et al. 1995, 2004; Liston 2004). The latest discov-
ery has been of a sauropod tooth from the Kilmaluag
Formation (Bathonian) near Glen Scaladal, north of
Elgol, Isle of Skye (Barrett 2006).
In Wyoming, footprints from the Middle Jurassic are
known from the Gypsum Springs Formation (Bajocian)
and the Sundance Formation (Bathonian) in the Big-
horn Basin, northern Wyoming (Kvale et al. 2001, 2004;
Mickelson et al. 2006) (Fig. 3). The best known track-
site is the Sundance Formation Red Gulch Dinosaur
Tracksite (discovered by Cli? Manuel of Shell) (Fig.
3(2)) between Greybull and Shell on Bureau of Land
Management property, Wyoming (Breithaupt 2001),
Scottish Journal of Geology 44, (2), 139?150, 2008
although further localities also include the ?Yellow Brick
Road? (which is on Wyoming State land and was
discovered by Rowena Manuel of Shell (Fig. 3(4))
(Adams & Breithaupt 2003) and Flitner Ranch (which is
on private land (Fig. 3(3)) tracksites. All the Sundance
Formation tracksites seem to occur in the Canyon Creek
Member if the basal Sundance Formation (Harris &
Lacovara 2004; Kvale et al. 2004). There are other
equivalent horizons to the Sundance Formation in Utah
from which dinosaur footprints are also known (Lockley
et al. 1998; Hamblin & Foster 2000; Kvale et al. 2004).
The Gypsum Spring Formation footprints are similar
sized tridactyl dinosaur footprints although the hallux
impression is sometimes visible (Kvale et al. 2001) are
Bajocian in age. The northernmost Gypsum Spring
Formation site was discovered by Erik Kvale in about
1997 (Fig. 3(1)).
Methods
The footprints used in this analysis are from the
Trotternish Peninsula, Isle of Skye and include examples
from the Lealt Shale, Valtos Sandstone, Duntulm, and
Kilmaluag formations. The footprints from the Valtos
Sandstone Formation included two sizes and varieties
(one less than 15 cm in length with narrow digits and
triangular terminations and the other over 25 cm in
length with broad digits with rounded terminations) that
were included separately in the analysis to see if they
would plot di?erently. All dinosaur footprints from the
Isle of Skye were measured from photographs taken in
the field, or from photographs of samples in the Sta?n
Museum and Hunterian Museum collections.
Photographs of footprints used in this study from the
Red Gulch, Yellow Brick Road and Flitner Ranch
F??. 1. (a) Map of the United States of America showing location of Shell, Wyoming. (b) Map of Great Britain showing location
of Sta?n, Isle of Skye.
N. D. L. CLARK & M. K. BRETT-SURMAN140
dinosaur tracksite were photographed during the 2006
summer season in the field, as well as a single footprint
from the Flitner tracksite at the Draper Museum of
Natural History in Cody, Wyoming (the Smithsonian
Institution has a mould of a six-track sequence of which
footprint no. 1 is in the Draper Museum and footprint
no. 4 was also collected (USNM 508544)). There are
track sequences of more than six footprints at both Red
Gulch and Flitner sites, but the majority of the rest are
individual footprints.
A landmark analysis was carried out on the footprints
using five points (Fig. 4a). The landmarks chosen were
the tips of the digits, not including claw impressions, the
back of the ?heel? (back end of the footprint produced in
the plantigrade posture (Thulborn 1990, fig. 4.6a), not
including any hallux impressions, and the posterior of
the proximal node of digit III. Landmark data were
produced from the photographs using tpsDig version 2
(Rohlf 2004). The resulting polygons were analysed by
flipping the left-handed footprints to allow a direct
shape comparison, and performing a 2D procrustes
transform to eliminate orientation and size anomalies
using PAST version 1.57 (Hammer et al. 2001, 2007).
The polygons were then subjected to principal compo-
nent analysis using PAST version 1.57 (Hammer et al.
2001, 2007) to compare the footprints from the di?erent
localities.
Principal component analysis was also carried out on
five di?erent measurements using PAST version 1.57
(Hammer et al. 2001, 2007) (Fig. 4b). A 2D procrustes
transform was also done to eliminate size anomalies.
The measurements included the width between the distal
points of digits II and IV (WII?IV); the length from the
line between the distal points of digits II and IV and
the distal point of digit III (hIII); the length between the
?heel? impression and the line between the distal points
of digit III (pL); the length between the posterior point
of the proximal phalange of digit III and the distal
point of digit III (LIII); and the angle between the distal
points of digits II, III and IV (
) (Fig. 4b, Tables 1, 2, 3).
None of these measurements included the claws as the
length of the claw impressions can vary greatly depend-
ing on the amount of drag as the animal moves, and is
less reliable in di?erentiating between di?erent track-
makers (Clark 2005). The lengths of the digit impres-
sions and 
 can also be a?ected by drag, but do not seem
to vary as much as is evidenced by the tight correlation
between width/length and 
 in footprints from the
Kilmaluag Formation (Clark 2005).
Palaeogeography and palaeoenvironments
During the Middle Jurassic, the dinosaur-bearing
localities in Wyoming have been estimated as being
within 15( to 20(N latitude (Kvale et al. 2001). In
F??. 2. Map showing the Middle Jurassic dinosaur localities of
the Isle of Skye (1?5 are in the Trotternish Peninsula and
6 is in the Strathaird Peninsula; both are outlined). (1)
Bearreraig Bay; (2) Rubha nam Brathairean; (3) Dun
Dearg; (4) An Corran; (5) Score Bay; (6) Elgol.
F??. 3. Middle Jurassic footprint localities near Shell in Big Horn County, Wyoming. (1) Gypsum Springs Formation; (2) Red Gulch
Tracksite; (3) Flitner Ranch Tracksite; (4) Yellow Brick Road.
COMPARISON OF DINOSAUR FOOTPRINTS 141
Scotland, the palaeolatitude was probably between lati-
tude 35( and 45(N (Callomon 2003; Cecca et al. 2005).
The distance between the localities in Scotland and those
in Wyoming, during the Middle Jurassic, was approxi-
mately 4000 km (Fig. 5).
In Wyoming the palaeoenvironment was warm and
dry. Although many of the footprint-bearing horizons
are biomicrites with ripples suggesting the presence
of water, there are also large halite pseudomorphs,
especially at the Flitner Ranch site, indicating periods,
perhaps seasonal, of evaporation (Kvale et al. 2001).
Rhyzocorallium and Diplocraterion, from the overlying
sediments, disturb the footprint surface at the Red
Gulch tracksite locality (Kvale et al. 2001). These trace
fossils are also found associated with the Duntulm and
Kilmaluag Formation footprints on the Isle of Skye
(Clark et al. 2004). The dinosaurs in Wyoming lived in a
seasonally arid environment during the Middle Jurassic
of both the Gypsum Springs and Sundance formations
(Kvale et al. 2001).
In Scotland, the depositional environment during the
Lealt Shale Formation, as well as the Duntulm Forma-
tion, is interpreted as being dominated by brackish
marine lagoon conditions (Harris & Hudson 1980;
Andrews & Walton 1990). The Valtos Sandstone For-
mation is thought to have been more fluvio-deltaic with
the footprints associated with a period of emergent
desiccation indicated by mudcracks. The footprint-
bearing sediments of the Valtos Sandstone Formation
are calcareous sandstones containing abundant bivalves
(Clark & Barco Rodriguez 1998). The Kilmaluag For-
mation footprint-bearing sediments were deposited in a
more freshwater lagoonal setting with abundant marls
and mudstones; the footprints are found at two horizons
within a single sandstone unit at two localities (Clark
et al. 2005). The footprints at the base of the unit are
impressed into a mud-cracked mudstone which was
covered with a sandsheet. The second level is 14 cm
above the base where the dinosaur footprints occur in a
ripple-bedded sandstone (Clark et al. 2005).
The sediments and palaeontology at the Wyoming
localities appear to suggest that the dinosaurs lived
closer to a marine shoreline in a seasonally arid environ-
ment, whereas the dinosaurs at the Scottish localities
lived in a deltaic environment with brackish and fresh-
water lagoons that were prone to occasional reduction in
size due to desiccation (Kvale et al. 2001; Clark et al.
2005).
Results
It was hoped that, using landmark analysis, it would
be possible to distinguish between tridactyl dinosaur
footprints on the basis of five landmarks. All the land-
mark data from Wyoming (n=58) and Scotland (n=48)
were analysed using principal component analysis, but it
was not possible to distinguish between the di?erent
forms with confidence (Fig. 6). All the 95% confidence
circles overlap substantially and the 95% confidence
circles for the Kilmaluag and Sundance formations
contain over 96% of the data. It was hoped that the
larger footprints of the Duntulm (Fig. 7c), Lealt Shale
(Fig. 7a) and Valtos Sandstone formations would plot
di?erently to the smaller Sundance (Fig. 8a?c), Gypsum
Spring (Fig. 8d), Kilmaluag (Fig. 7d), and Valtos Sand-
stone (Fig. 7b) formations. Only the larger footprints
from the Lealt Shale and Valtos Sandstone formations
appeared to deviate slightly from the other footprints
(Fig. 9d). Further discoveries of these larger footprints
F??. 4. (a) Points used for landmark analysis. (b) Measurements taken of footprints for comparison: hIII, perpendicular length of
digit III from WII?IV; LIII, length of digit III; WII?IV, distance between the apices of digits II and IV; pL, footprint length; 
,
angle between digits II, III and IV.
N. D. L. CLARK & M. K. BRETT-SURMAN142
would need to be made and added to the data for this
deviation to be confirmed.
Principal component analysis of the measurements of
the footprints, however, seems to be more useful in
distinguishing between footprints from the various for-
mations (Tables 4, 5, 6). The footprints of the Sundance,
Kilmaluag Formation, and the smaller footprints of
the Valtos Sandstone Formation, all plot in a similar
position with nearly all the data from these three forma-
tions contained within the 95% confidence circle for
the Sundance Formation. Briethaupt et al. (2007a, b)
suggested that the Sundance Formation preserved a
monotaxonomic community of carnivorous dinosaurs as
the footprints exhibit a similar growth trend to modern
emu footprints. The tight correlation of the principal
component analysis of the footprints examined here
TABLE 1
Measurements taken from the footprints of the Kilmaluag
Formation. Each field identifier refers to an individual foot-
print (see Fig. 4 and text for definitions of measurements).
wII?IV
(cm)
hIII
(cm)
pL
(cm)
LIII
(cm)


(degrees)
KF1a 7.6 5 11.7 8.1 70
KF1b 7.5 4.8 11.3 8 78
KF1c 8.1 5.4 11.4 8.1 74
K2a 14.5 8.5 20.8 14.3 83
K2b 6.6 3.4 9.2 8.6 90
K2c 6.3 2.8 8.4 6.1 72
K2d 4.7 3.1 7.4 5.4 75
K2e 5.8 3.7 8.4 5.8 77
K2f 5.4 3.5 8.3 6.3 78
K2g 6.1 4.5 9.9 7.8 75
K2h 5.9 3.3 8.1 5.9 82
K2i 6.2 3.6 9.3 6.5 80
K2j 6.5 5 10.7 7.2 80
K2k 10.7 6.5 13.5 10.4 80
K2l 5.6 3.6 8.2 5.4 83
K2m 6.2 4.4 9.6 6.5 75
K2n 6.1 3.6 8.8 6.1 78
K2o 6.3 3.9 9.2 6.7 78
K2p 6.5 4 9.8 6.2 77
K2q 6.8 3.5 9.5 5.9 84
K2r 8.9 5 13.8 9.3 85
K2t 5 3.1 8.5 6 77
K2u 5.2 3.9 8.7 7 72
K2v 6.2 4.1 8 6.7 78
K2w 5.4 3.4 9.4 6.3 76
K2x 6.6 4.3 9.6 7.3 82
KF3a 6.2 2.8 7.6 6.5 94
KF3b 1.2 0.7 1.8 1.6 78
KF3c 1 0.7 1.8 1.5 70
KF4 9.7 6 16.7 11.1 78
KF5 10.4 5.6 17.8 12.1 80
KF6 6 3.5 8.4 6.2 82
KF7 13.1 7.6 22.4 14.7 82
KF9a 19.2 12.6 27 20 92
KF9b 4.8 2.9 6.8 5.1 82
KF9c 5.6 3.6 10.7 7.4 80
KF9d 5.6 3.7 9.2 6.9 77
KF9e 7.7 4.2 11.4 8 84
TABLE 2
Measurements taken from the footprints of the Sundance
Formation. Each field identifier refers to an individual
footprint (SI refers to footprints in the collections of the
Smithsonian Institution, Washington) (see Fig. 4 and text for
definitions of measurements).
wII?IV
(cm)
hIII
(cm)
pL
(cm)
LIII
(cm)


(degrees)
YBR1 10.18 6.79 14.69 9.27 74
YBR2 10.25 9.45 17.85 13.09 83
YBR3 15.23 8.34 20.97 12.94 85
YBR4 8.76 5.81 13.29 9.24 76
YBR5 11.98 7.75 18.34 13.05 76
YBR6 8.74 6.83 13.99 10.73 68
YBR7 17.51 9.33 21.56 19.94 86
YBR8 14.44 8.75 18.77 17.67 77
YBR9 11.58 6.22 17.78 13.12 84
YBR10 17.07 8.81 20.27 15.84 86
YBR11 20.9 10.62 22.96 19.56 88
YBR12 12.71 9.29 17.97 14 82
YBR14 3.99 2.55 4.98 6.38 78
YBR15 10.47 8.31 19.45 14.57 66
YBR16 13.14 9.31 18.52 12.96 78
YBR17 10.74 7.81 19.68 15.28 74
YBR18 9.33 5.5 13.31 10.32 85
YBR19 14.36 7.68 20.82 13.87 85
YBR20 11.72 6.06 16.8 12.08 88
YBR21 9.64 6.06 13.91 11.47 82
YBR22 12.08 7.51 15.73 11.82 82
YBR24 5.72 3.85 8.75 6.55 73
YBR26 6.66 3.77 8.88 6.47 85
YBR27 14.8 7.03 16.23 12.87 91
YBR28 16.92 6.7 20.91 15.34 98
YBR29 15.38 6.6 15.9 11.89 98
YBR30 6.85 3.9 9.29 7.4 83
YBR31 14.85 8.7 17.5 15.36 82
YBR32 14.5 8.14 17.75 14.6 83
YBR33 10.94 5.04 13.72 10.69 92
YBR34 12.01 8.18 17.34 12.51 66
YBR35 7.06 3.05 9.4 7.98 98
YBR36 11.1 3.98 12.81 8.56 98
YBR37 13.58 8.16 17.76 12.94 80
YBR38 8.45 4.98 11.3 8.48 78
YBR39 10.56 5.51 13.98 10.42 88
YBR40 11.9 8.6 15.54 12.21 81
YBR41 10.5 4.17 13.8 10 98
CodyFR 16.82 9.53 19.87 15.58 92
FR1 15.48 9.05 21.11 16.34 86
FR2 20.26 9.18 19.24 14.98 93
FR3 21.87 10.31 23.93 17.98 88
FR4 26.16 8.72 28.1 18.27 91
SI508524a 12.72 7.78 18.49 13.87 84
SI508524b 10.59 6.5 14.45 10.99 77
SI508524c 11.44 7.97 17.92 12.9 76
SI508524d 10.33 6.6 13.23 11.56 84
RGTS1 16.31 12.09 26.59 19.35 77
RGTS2 9.54 7.42 17.7 12.74 74
RGTS3 14.89 10.11 25.08 16.25 78
RGTS4 13.67 7.92 23.15 14.61 88
RGTS5 18.38 8.76 29.08 22.04 98
RGTS6 20.46 10.66 26.6 21.65 84
COMPARISON OF DINOSAUR FOOTPRINTS 143
supports this view. The data from the older Gypsum
Spring Formation plot above the Sundance Formation
95% confidence circle, and the large footprints from the
Duntulm, Valtos Sandstone and Lealt Shale formations
TABLE 3
Measurements taken from the footprints of the Gypsum
Springs Formation (GS2?3), Duntulm Formation (DF1?8,
DFs1); Lealt Shale Formation (LSF) and the Valtos Sandstone
Formation (VSF). LSF1, VSF and VSF1 were similar large
footprints with rounded broad digits and were analysed to-
gether. Each field identifier refers to an individual footprint
(see Fig. 4 and text for definitions of measurements).
wII?IV
(cm)
hIII
(cm)
pL
(cm)
LIII
(cm)


(degrees)
GS2a 11.97 13.06 25.73 17.57 49
GS2b 9.03 10.12 19.74 14.56 50
GS3 10.19 8.92 22.39 17.11 61
DF1 26.9 16.73 39.21 32.2 86
DF2 35.7 14.69 48.07 36.5 88
DF3 22.8 15.29 37.4 25 81
DF4 29.4 23.01 54.32 33.5 72
DF5 27.2 21.52 52.44 32.8 82
DF6 27 18.36 42.12 30.2 81
DF7 29.6 17.88 52.48 38.7 82
DF8 27.1 17.06 48.13 34.3 80
DFs1 20.2 12.76 26.87 19.7 75
LSF1 44.57 13.71 49 35.95 115
VSF 39.8 13.54 47.67 34.12 111
VSF1 27.06 10.41 36.11 24.31 102
VSF2a 9 6.16 12.2 11.9 69
VSF2b 9 6.16 12 11.9 67
VSF2c 11.6 8.26 16.5 16.59 66
VSF2d 11.1 5.67 13.3 13.3 87
VSF2e 15.1 10.85 17.6 18.13 69
VSF2f 9.2 6.3 12.6 12.88 65
VSF2g 9.1 6.3 13.9 13.51 61
VSF3 5.1 3.72 10.8 6.41 72
VSF4 11.9 7.02 20.4 13.25 88
VSF5a 14.93 5.58 22 13.76 100
VSF5b 14.5 5.77 17 12.38 109
F??. 5. Palaeogeographic sketch map of part of Laurentia showing the relative positions of Wyoming and Great Britain during the
Middle Jurassic about 170 million years ago (based on Kvale et al. 2001; Hesselbo & Coe 2000).
F??. 6. Centred principal component analysis of the landmark
data: (a) 95% confidence circles; (b) convex hulls.
N. D. L. CLARK & M. K. BRETT-SURMAN144
plot in di?erent space to the right of the Sundance
Formation (Fig. 10a). This is more easily seen when
using the convex hull plots of the data from the various
formations (Fig. 10b).
The measurements used in this analysis may provide a
more useful means of distinguishing between di?erent
types of dinosaurs on the basis of their footprints
(adapted from Clark et al. 2005). The sediments were
similar between the tracksite localities, and the tracks
were either surface tracks or shallow transmission
tracks, resulting in a good correlation between similar
tracks. Studies looking at more distinct sediment types
and variations in track shape and dimensions with
transmission depth would help determine whether these
measurements may be more widely useful. This method
is not used here to distinguish between dinosaur ichno-
species which may vary as a result of transmission,
sediment type, water content of the sediment, as well as
the size, weight and type of trackmaker. The width of
the digits, claw impressions, and digit divergence from
F??. 7. Typical dinosaur footprints from the Middle Jurassic of the Isle of Skye: (a) from the Lealt Shale Formation of Rubha nam
Brathairean (scale bar 10 cm); (b) from the Valtos Sandstone Formation at Dun Dearg (scale bar 2 cm); (c) from the Duntulm
Formation at An Corran (scale bar 10 cm); (d) from the Kilmaluag Formation at Score Bay (scale bar 5 cm).
COMPARISON OF DINOSAUR FOOTPRINTS 145
the rear of the footprint may vary as a result of these
factors (Clark et al. 2005).
Conclusions
It is possible to distinguish between di?erent dinosaur
footprints on the basis of morphometric analysis using
measurements of the width between the distal ends of
digits II and IV, various lengths and the angle between
the distal ends of digits II, III and IV. A landmark
analysis of the same footprints did not allow any distinc-
tion between footprints from di?erent formations. Per-
haps the use of more landmarks on the pad impressions
would produce better results, but better preservation
F??. 8. Typical dinosaur footprints from the Middle Jurassic of Wyoming (a?c are from the Sundance Formation): (a) from the Red
Gulch Tracksite; (b) from the Flitner Ranch Tracksite (USNM 508544); (c) from the Yellow Brick Road Tracksite; (d) from
the Gypsum Springs Formation. (scale bars 10 cm).
N. D. L. CLARK & M. K. BRETT-SURMAN146
would be required to be able to introduce further
landmarks.
The footprints from the Sundance, Valtos Sandstone
and Kilmaluag formations are indistinguishable and it is
thought that they may have been produced by a similar
type of dinosaur. The sharp claw impressions on prints
from both these localities and the discovery of a
coelophysoid-grade caudal vertebra from the Valtos
Sandstone Formation, indicates that the animal that
produced these footprints may have been a small thero-
pod morphologically similar to a coelophysoid (Clark
2001b, 2004, 2005; Clark et al. 2004). The high density of
footprints from the same level in the Sundance Forma-
tion (probably over 150 000 footprints per square kilo-
metre; Kvale et al. 2001) are represented by a range of
sizes from about 8 cm to nearly 30 cm from the Sun-
dance Formation near Shell. In Scotland, the equivalent
footprints from the Kilmaluag Formation range in size
from less than 2 cm in length to about 25 cm. This
F??. 9. Mean shapes of landmark data of footprints: (a) Kilmaluag Formation; (b) Valtos Sandstone Formation; (c) Duntulm
Formation; (d) large footprints from Valtos Sandstone Formation and Lealt Shale Formation; (e) Sundance Formation; (f)
Gypsum Springs Formation.
TABLE 4
Correlation eigenvalues as a percentage of their sum for the
measured variables (with a Jolli?e cut o? of 0.7, only the first
two principal components are considered to be significant).
Principal component Eigenvalues Variance (%)
1 3.87 77.57
2 1.00 20.03
3 0.06 1.29
4 0.03 0.71
5 0.02 0.40
TABLE 5
Correlation loadings of the measured variables for the first two
principal components.
Variable Loading (PC1) Loading (PC2)
wII?IV 0.97 
0.14
hIII 0.93 0.29
pL 0.99 0.09
LIII 0.99 0.10

 0.34 
0.94
COMPARISON OF DINOSAUR FOOTPRINTS 147
suggests that the dinosaur was gregarious and may
even have moved in family groups (Clark et al. 2005;
Breithaupt et al. 2007a, b), although this is disputed by
Roach & Brinkman (2007).
If the trackmaker genus in Wyoming is the same as
the trackmaker for the similar footprints in Scotland,
then its presence at these two distant locations needs to
be explained. One hypothesis is that they may have
migrated between these two locations following sauro-
pods which certainly existed in Scotland at this time
(Clark et al. 1995; Barrett 2006; Liston 2004). It has been
suggested that some Cretaceous hadrosaur dinosaurs
migrated, but this has been disputed (Fiorillo &
Ganglo? 2001; Lockley 1995). Caribou migrate about
700 km from their wintering grounds to their calving
grounds (Zalatan et al. 2006) and can accumulate up
to more than 5000 km in a year (Fancy et al. 1989) for
the round journey. It is unlikely that the individual
trackmakers migrated between the two sites, but it may
represent the full range of the dispersed trackmaker
genus. It is therefore suggested that this represents a
wider Laurasian distribution for this theropod track-
maker.
The other question to be considered is where all the
herbivores are, if these footprints are considered to be of
a theropod trackmaker. It is possible that they are living
further inland amongst the vegetation rather than close
to the inland sea or saline lagoons of Wyoming. In
Scotland there do appear to be herbivore remains, but
the footprints are rarely associated with those of thero-
pods. Only in the Valtos Sandstone Formation are large
spatulate digits on tridactyl footprints found in close
association with footprints with small narrow digits.
Similar patterns have been observed where there is a bias
towards the footprints of carnivorous dinosaurs by eight
to two (Leonardi 1989). It may also be that the carnivo-
rous dinosaurs feed on near-shore aquatic prey such as
fish, which would also explain why there is a predomi-
nance of carnivorous dinosaur footprints in arid near-
shore environments such as those found at both the
Wyoming and Scottish sites. The existence of herbi-
vorous dinosaurs in the Scottish localities can be due to
the variety and greater abundance of vegetation derived
from a nearby source into the fluvio-deltaic and near-
shore marine depositional environments (Dower et al.
2004). The other possibility is that there is just not
enough exposure of the track-bearing surfaces to have a
fully representative ichnofauna. If the trackways repre-
sent only a short period of emergence, it is likely that
only a few species will be represented on the shores of
receding lagoons or seas.
Acknowledgements
Rowena and Cli? Manuel of GeoScience Adventures, Shell,
are thanked for looking after us in Shell and for accompanying
us to localities. Kim Moeller is thanked for taking care of N.C.
whilst in Washington DC. M.K.B.-S. and Jim and Ruth
Bobon are also thanked for making the trip memorable. The
Smithsonian Institution, the Geological Society of Glasgow,
and the John Robertson Bequest (JR06/01) are thanked for
funding N.C. to visit the sites in Wyoming and to study the
collections in the Smithsonian Institution. M.K.B.-S. would
like to thank N.C. and Dugie Ross for hosting a tour of the
Isle of Skye footprint localities, and access to the collections in
Scotland. Part of this research was carried out with permission
from the Bureau of Land Management (Permit: PA99-WY-
053). Thanks are also due to the anonymous reviewer who
provided many useful comments.
References
ADAMS, T.L. & BREITHAUPT, B. 2003. Middle Jurassic
dinosaurs and undergraduate research: opportunities
from the Yellow Brick Road Dinosaur Tracksite. Geologi-
cal Society of America Abstracts with Programs, 35, 43.
ANDREWS, J.E. & HUDSON, J.D. 1984. First Jurassic
dinosaur footprint form Scotland. Scottish Journal of
Geology, 20, 129?134.
ANDREWS, J.E. & WALTON, W. 1990. Depositional envi-
ronments within Middle Jurassic oyster-dominated la-
goons: an integrated litho-, bio and paynofacies study of
TABLE 6
Eigenvalues as a percentage of their sum for the first five
principal components using landmark data (with a Jolli?e
cut-o? of 0.00018, only the first five principal components are
considered significant).
Principal component Eigenvalues Variance (%)
1 0.00094 30.00
2 0.00075 24.05
3 0.00054 17.37
4 0.00044 14.03
5 0.00032 10.21
6 0.00011 3.63
F??. 10. Centred principal component analysis of the measured
data: (a) 95% confidence circles; (b) convex hulls.
N. D. L. CLARK & M. K. BRETT-SURMAN148
the Duntulm Formation (Great Estuarine Group, Inner
Hebrides). Transactions of the Royal Society of Edinburgh:
Earth Sciences, 81, 1?22.
BARRETT, P.M. 2006. A sauropod dinosaur tooth from
the Middle Jurassic of Skye, Scotland. Transactions
of the Royal Society of Edinburgh: Earth Sciences, 97,
25?29.
BENTON, M.J., MARTILL, D.M. & TAYLOR, M.A. 1995.
The first dinosaur from the Lower Jurassic of Scotland: a
limb bone of a ceratosaur theropod. Scottish Journal of
Geology, 31, 171?182.
BREITHAUPT, B.H. 2001. In the study of the most extensive
dinosaur tracksite in Wyoming. In Santucci, V. L. &
McClelland, L. (eds) Proceedings of the 6th Fossil Re-
sources Conference, United States Department of Interior,
National Park Services, 107?112.
BREITHAUPT, B.H., MATTHEWS, N.A. & NOBLE, T.A.
2004. An integrated approach to three-dimensional data
collection at dinosaur tracksites in the Rocky Mountain
West. Ichnos, 11, 11?26.
BREITHAUPT, B.H., GREEN, T., SOUTHWELL, E. &
MATTHEWS, N.A. 2007a. Footprints and growth
rates of emus and theropods: Ichnological evidence for
family groups of Middle Jurassic dinosaurs in Wyoming.
(abstract) Journal of Vertebrate Paleontology, 27, 52a.
BREITHAUPT, B.H., MATTHEWS, N.A. & GREEN, T.L.
2007b. Growing up in the Middle Jurassic: ichnological
evidence for family groups of theropods in Wyoming;
comparison of footprints and growth rates of emus and
dinosaurs. In Liston, J.J. (ed.) 55th Symposium of Verte-
brate Palaeontology and Comparative Anatomy and the
16th Symposium of Palaeontological Preparation and Con-
servation held at the University of Glasgow, 28th August?
1st September 2007: Abstracts of Presentations, 2007,
University of Glasgow Press, Glasgow, 9.
CALLOMON, J.H. 2003. The Middle Jurassic of western and
northern Europe: its divisions, geochronology and
correlations. Geological Survey of Denmark and Greenland
Bulletin, 1, 61?73.
CECCA, F., GARIN, B.M., MARCHAND, D.,
LATHUILIERE, B. & BARTOLINI, A. 2005. Paleo-
climatic control of biogeographic and sedimentary events
in Tethyan and peri-Tethyan areas during the Oxfordian
(Late Jurassic). Palaeogeography, Palaeoclimatology,
Palaeoecology, 222, 10?32.
CLARK, N.D.L. 2001a. Dinosaur tracks, helicopters, and
broken bones. The Geological Curator, 7, 163?166.
CLARK, N.D.L. 2001b. A thyreophoran dinosaur from the
Bearreraig Sandstone Formation (Bajocian, Middle
Jurassic) of the Isle of Skye. Scottish Journal of Geology,
37, 19?26.
CLARK, N.D.L. 2003. Dinosaurs and shifting sands. Earth
Heritage, 19, 19?20.
CLARK, N.D.L. 2004. Tracking dinosaurs in Scotland.
Teaching Earth Sciences, 29, 22?25.
CLARK, N.D.L. 2005. Tracking dinosaurs in Scotland. Open
University Geological Society Journal, 26, 30?35.
CLARK, N.D.L. & BARCO RODRIGUEZ, J.L. 1998. The
first dinosaur trackway from the Valtos Sandstone For-
mation (Bathonian, Jurassic) of the Isle of Skye, Scotland,
UK. Geogaceta, 24, 79?82.
CLARK, N.D.L., BOYD, J.D., DIXON, R.J. & ROSS, D.A.
1995. The first Middle Jurassic dinosaur from Scotland: a
cetiosaurid? (Sauropoda) from the Bathonian of the Isle
of Skye. Scottish Journal of Geology, 31, 171?176.
CLARK, N.D.L., BOOTH, C., BOOTH, P. & ROSS, D.A.
2004. Dinosaur Footprints from the Duntulm Formation
(Bathonian, Jurassic) of the Isle of Skye, Scotland, UK.
Scottish Journal of Geology, 40, 13?21.
CLARK, N.D.L., ROSS, D.A. & BOOTH, P. 2005. Dinosaur
Tracks from the Kilmaluag Formation (Bathonian,
Middle Jurassic) of Score Bay, Isle of Skye, Scotland, UK.
Ichnos, 12, 93?104.
DELAIR, J.B. & SARJEANT, W.A.S. 1985. History and
bibliography of the studies of fossil vertebrate footprints
in the British Isles: Supplement 1973?1983. Palaeogeogra-
phy, Palaeoclimatology and Palaeoecology, 49, 123?160.
DOWER, B.L., BATEMAN, R.M. & STEVENSON, D.W.
2004. Systematics, ontogeny, and phylogenetic implica-
tions of exceptional anatomically preserved cycadophyte
leaves from the Middle Jurassic of Bearreraig Bay, Skye,
Northwest Scotland. The Botanical Review, 70, 105?120.
FANCY, S.G., PANK, L.F., WHITEN, K.R. & REGELIN,
W.L. 1989. Seasonal movement of caribou in arctic
Alaska as determined by satellite. Canadian Journal of
Zoology, 67, 644?650.
FIORILLO, A.R. & GANGLOFF, R.A. 2001. The caribou
migration model for Arctic Hadrosaurs (Dinosauria:
Ornithischia): a reassessment. Historical Biology, 15,
323?334.
HAMBLIN, A.H. & FOSTER, J.R. 2000. Ancient animal
footprints and traces in the Grand Staircase-Escalante
National Monument, South-Central Utah. Utah Geologi-
cal Association Publications, 28, 557?568.
HAMMER, ?., HARPER, D.A.T. & RYAN, P.D. 2001.
PAST: Palaeontological Statistics software package for
education and data analysis. Palaeontologica Electronica,
4, 1?9.
HAMMER, ?., HARPER, D.A.T. & RYAN, P.D. 2007.
PAST ? palaeontological statistics version 1.57. Available
at: http://folk.uio.no/ohammer/past.
HARRIS, J.P. & HUDSON, J.D. 1980. Lithostratigraphy of
the Great Estuarine Group (Middle Jurassic), Inner
Hebrides. Scottish Journal of Geology, 16, 231?250.
HARRIS, J.D. & LACOVARA, K.J. 2004. Enigmatic fossil
footprints from the Sundance Formation (Upper Jurassic)
of Bighorn Canyon National Recreation Area, Wyoming.
Ichnos, 11, 151?166.
HESSELBO, S.P. & COE, A.L. 2000. Jurassic sequences of the
Hebrides Basin, Isle of Skye, Scotland. In Graham, J.R. &
Ryan, A. (eds) Field Trip Guidebook, International Sedi-
mentologists Association Meeting, Dublin, 2000, Univer-
sity of Dublin, Dublin, 41?58.
KVALE, E.P., JOHNSON, G.D., MICKELSON, D.L.,
KELLER, K., FURER, L.C. & ARCHER, A.W. 2001.
Middle Jurassic (Bajocian and Bathonian) Dinosaur
Megatracksites, Bighorn Basin, Wyoming, USA. Palaios,
16, 233?254.
KVALE, E.P., MICKELSON, D.L., HASIOTIS, S.T. &
JOHNSON, G.D. 2004. The history of dinosaur footprint
discoveries in Wyoming with emphasis on the Bighorn
Basin. Ichnos, 11, 3?9.
LEONARDI, G. 1989. Inventory and systematics of the South
American dinosaurian ichnofauna and its paleobiological
interpretation. In Gillette, D. D. & Lockley, M. G. (eds)
Dinosaur Tracks and Traces, Cambridge University Press,
Cambridge, 165?178.
LISTON, J.J. 2004. A re-examination of a Middle Jurassic
sauropod limb bone from the Bathonian of the Isle of
Skye. Scottish Journal of Geology, 40, 119?122.
LOCKLEY, M. 1995. Track records. Natural History, 104,
46?51.
LOCKLEY, M., HUNT, A., PAQUETTE, M., BILBEY,
S-A. & HAMBLIN, A. 1998. Dinosaur tracks from the
Carmel Formation, northeastern Utah: Implication for
Middle Jurassic paleoecology. Ichnos, 5, 255?267.
MARSHALL, P. 2005. Theropod dinosaur and other foot-
prints form the Valtos Sandstone Formation (Bathonian,
Middle Jurassic) of the Isle of Skye. Scottish Journal of
Geology, 41, 97?104.
MICKELSON, D.L., MICKELSON, K.A., KING, M.R. &
GETTY, P. 2006. Jurassic dinosaur tracksites from the
COMPARISON OF DINOSAUR FOOTPRINTS 149
American west. In Lucas, S. G., Speilman, J. A., Hester,
P. M., Kenworthy, J. P. & Santucci, V. L. (eds) Fossils
from Federal Lands. New Mexico Museum of Natural
History and Science Bulletin, 34, 138.
ROACH, B.T. & BRINKMAN, D.L. 2007. A reevaluation of
cooperative pack hunting and gregariousness in Deinony-
chus antirrhopus and other nonavian theropod dinosaurs.
Bulletin of the Peabody Museum of Natural History, 48,
103?138.
ROHLF, F.J. 2004. tpsDig, digitize landmarks and outlines,
version 2.0, Department of Ecology and Evolution, State
University of New York at Stony Brook.
THULBORN, R.A. 1990. Dinosaur Tracks, Chapman and
Hall, London.
ZALATAN, R., GUNN, A & HENRY, G. 2006. Long-term
abundance patterns of barren-ground caribou using
trampling scars on roots of Picea Mariana in the North-
west Territories, Canada. Arctic, Antarctic, and Alpine
Research, 38, 624?630.
MS. accepted for publication 24 June 2008
N. D. L. CLARK & M. K. BRETT-SURMAN150