Several different types of relatively small dinosaur footprints have been found in slabs of limestone from the Purbeck Formation on the Isle of Portland, Dorset. They have already been figured to a limited extent in a report by Edmonds et al. (2002 ). More than 70 individual tracks, mostly three-toed, have been found so far. Some of these are in trackways of several prints. Both herbivorous Ornithopod and carnivorous Theropod dinosaur tracks have been found during the preliminary observations. Some anomalous tracks seem to have the central toe impression either missing or very feebly impressed. They are difficult to explain because they do not have the correct symmetry for a dromaeosaur print with a raised claw, although an origin from this type of dinosaur was first considered as a possible explanation. Small, apparently juvenile, prints from a "baby" ornithopod are present. Richard Edmonds of the Devon-Dorset World Heritage Coast (Jurassic Coast) section of Dorset County Council is supervising both study and conservation on the slabs. Grateful acknowledgement is made to Hanson for allowing access and for conserving the specimens, which are their property. This present webpage is merely an initial photographic record of the finds as they are made, not the full scientific study which will follow. Its purpose is to ensure that early observations are recorded and little is lost, but any interpretive comments should be regarded as provisional and subject to revision at a later stage. It is intended that the full scientific study will by dinosaur footprint specialist - Paul Ensom of the Natural History Museum, London. His investigation will commence in October 2003. A useful undergraduate project by Caroline Clasby (2003) has discussed the environment of deposition and the research potential of the tracks. Her work took place when the main set of tracks was visible, but before the moving of some of the slabs by crane in September 2003, which revealed some new types of footprints.

Other interesting fossil remains are present in associated strata. In addition to the footprints Richard Edmonds has a dinosaur bone from adjacent strata. An impressions that appear to be of reptile skin has been found by Stuart Tabner in marl associated with dinosaur slabs. To avoid damage or loss of tracks the exact location on the Isle of Portland is not given here.

Dinosaur footprints are well-known features of the Purbeck Formation, mostly of early Cretaceous age (Berriasian) of Dorset. The basal part of the Purbeck Formation is uppermost Jurassic. They have been found particular in quarries and cliffs which expose the Middle Purbeck lagoonal and shelly limestones. They deposits are sub-humid origin at a palaeolatitude of about 37degrees north. A few have been found in Upper Purbeck strata and they are well-known in the Wealden strata which follow but usually in other parts of southern England. For a valuable review and a comprehensive bibliography of Purbeck dinosaur tracks and trackways see Ensom (2002) .

Although tracks also occur in the Lower Purbeck Formation which originated in dryer, semi-arid conditions, they are much less common in this part of the sequence. The Isle of Portland, a peninsula in the middle of the Dorset coast, does not seem to contain any significant amount of Middle Purbeck strata because it has been removed by erosion. There are some blocks moved by solifluction into the debris of a Pleistocene raised beach and there is the possibility of a very small amount existing near Southwell, but this is unlikely. Thus the most promising horizons for footprints are missing and until recently no footprints had been found. However, in 1996 a single tridactyl dinosaur footprint was discovered by Dr Jane Francis in the Transition Bed at the very base of the Purbeck Formation near Freshwater on the east side of the peninsula. On the basis of recent dating (not an easy matter because of the lagoonal rather than marine facies) by (Feist, Lake and Wood, 1995) this would be of very late Jurassic age. No others were found on Portland until a few years later a dog-walker noticed footprints in large slabs of Lower Purbeck limestone which had been tipped on the grass at the top of an active quarry. The quarry manager, Ben Murray, was kind enough to initially protect and preserve these and draw attention to them by email. Thanks to his interest and that of Richard Edmonds of Dorset County Council, they became a subject of initial studies with the prospect of a full investigation later. These footprints occur in blocks of the so-called "Hard Slatt" which is clearly seen to be present in the Lower Purbeck sequence between the Caps and Dirt Beds of the basal Purbeck and the gypsum (Portland Alabaster) of the Soft Cockle Member of the Lower Purbeck Formation. The present author considers the Hard Slatt to correspond to the Hard Cockle Member of the Lower Purbeck Formation which is a prominant limestone with similar high-energy (relatively), cross-laminated, carbonates and similar pseudomorphs after halite. The Hard Cockle Member can be examined at Lulworth Cove and adjacent area. Using the correlation of (Feist, Lake and Wood, 1995) this would be basal Cretaceous. It is very close to the Jurassic-Cretaceous boundary, although the position of this is not known with great precision.

Very well-preserved, but relatively small, dinosaur footprints occur in the slabs. Several of them are associated with the salt crust of a salt lake. Photographs and preliminary descriptions are given here with emphasis on the unusual limestone in which they occur and the sedimentary structures which are present. These help in understanding the palaeoenvironment. The preservation of the footprints is now under the control of Richard Edmonds of Dorset County Council and he has already published photographs of them. This webpage is a branch of the general Geology of the Isle of Portland webpage, and that should be referred to for background information on the geology of the Purbeck Formation on Portland in which the dinosaur footprints occur. It also provides maps and sections and information on various geological localities on the peninsula and various fossil remains.

The reader wishing to understand dinosaur footprints in general should see the excellent introduction - "Tracking Dinosaurs" - by Lockley (1991) . For more advanced and detailed study see - "Dinosaur Tracks and Traces" - a symposium of many papers on the subject and edited by Gillette and Lockley (1989). The Lower Cretaceous Wealden dinosaurs of the nearby Isle of Wight are well-described by Martill and Naish (2001) . Lower Cretaceous dinosaur footprints are discussed here too.

Safety on Field Trips on Portland

Before discussing the footprints the matter of safety, therefore requires attention. General information regarding safety on Dorset geological field trips is provided in a separate webpage and you are requested to read this.

There are various hazards with regard to Portland and great care should be taken with cliffs, rocks and boulders, quarries and quarry machinery. Specific risk regarding the dinosaur footprints are mentioned here. There is some risk in scrambling over blocks of stone in the search for footprints. It is very easy to slip and fall from the sloping slabs when studying or photographing them and great care must be taken. Since at times relatively few people may pass the blocks incapacity resulting from a slip amongst the blocks could be very dangerous, especially in bad weather. While there is no safe and easy access to them they are not recommended for visits by school parties.

There is always a hazard of falling rocks, and in quarries the rock is often loose and even more likely to fall. Safety helmets should always be worn beneath cliffs and in quarries, and any places with loose rock or where it is clear that rock has recently fallen should be strictly avoided. Quarry regulations must be followed if these are visited and care taken to keep clear of heavy plant and machinery and loose rocks. High-visibility tabards may be needed. It is important not to hammer the chert in the Portland and Purbeck strata because dangerous splinters can easily penetrate the body and may cause blindness. When preoccupied in geological work take care not to walk backwards over a quarry cliff or fall down fissures in the Portland Stone. A hazard of injury exists if persons attempt to lift a large block of stone.

.

The Portland Dinosaur Footprints - The Footprint Slabs

The best of the dinosaur footprint slabs have recently been lifted out of the pile of blocks to be placed in a safe and secure site. In due course it is hoped that they will be on public display, but this will take time. The study of the slabs and arrangements for their future is in the hands of Richard Edmonds from the Dorset-Devon World Heritage Coast (Jurassic Coast) section of Dorset County Council. Richard is shown in the above photographs. The moving of the slabs is by courtesy of Hanson Quarries, who are the owners of the fossils.

A typical slab from the Hard Slatt is shown here, together with enthusiastic amateur geologists, including the Tabner family who have helped by searching the smaller blocks of laminated limestone associated with the dinosaur slabs. Their help is much appreciated. Stuart Tabner discovered the strange markings on a thin limestone, probably from above the dinosaur footprint bed, and discussed below.

Stratigraphy of the Hard Slatt, Dinosaur Footprint Bed, and Associated Strata

These images provide information on the Hard Slatt Bed on the west coast of the Isle of Portland, not far from the site at which the dinosaur footprints were discovered.

Left: The Hard Slatt Bed in situ in the cliffs south of Mutton Cove. Underneath the Hard Slatt is the Shingle, rather crumbly, ripple-bedded, argillaceous and sandy limestone. This is probably the equivalent of the Linsen Shale, beneath the hard limestone of the Hard Cockle Member and a conspicuous bed in the Lulworth Cove and Worbarrow area. This contain sand linsen or lenses of ripple origin and within shale. The sand is mostly carbonate sand.

Centre: The succession in the cliffs from Mutton Cove to Wallsend to show the stratigraphical position of the Hard Slatt in relation to associated strata. The succession ranges from mildly hypersaline strata to evaporites including gypsum.

Right: Cliff-top exposure of the Hard Slatt at Mutton Cove. There are well-developed ripple marks on the top surface of the bed. A dinosaur footprint is visible within the Hard Slatt.

The Hard Slatt can be seen in several quarries in the central part of the Isle of Portland. The images above show the bed in both the northern and southern parts of Coombefield Quarry. It is to some extent variable in thickness and often thinner than in the blocks with footprints. It is frequently fractured and partly obscured by associated marly debris. It is thus not as obvious in the quarry cliff face as might be expected.

The dinosaur footprints are present in some very large and heavy slabs that have been removed by machines with other overburden from above Portland Stone in a working quarry. The site is within the quarry property and is to some extent protected and under observation. It is not accessible to vehicles except through the working quarry. The slabs are not systematically arranged and footprints occur here and there in slabs at all angles. Some slabs are the right-way up and these normally show a rippled surface. The footprints are seen as natural casts on some of those slabs which are inverted. The footprints are generally at or near the base of the bed, but not necessarily confined to the exact base. The slabs are joint-bounded with some slickensides showing strike-slip movement and with some travertine with stalactitic structures on certain faces.

Bottom centre and right: The dinosaur footprint bed or Hard Slatt is clearly visible in several quarries. Shown here is the position of this bed in a sequence of Lower Purbeck marls and clays with limestones at Coombefield Quarry, near Southwell. Here, the bed is distinctive with a thickness of near a metre (at maximum) and with a rippled top and halite casts below.

Dinosaur footprints are shown in the photographs above. These have come from the Hard Slatt which seems to correspond to the Hard Cockle Member of the Lower Purbeck Formation, which is better known on the mainland of Dorset. There appear to be more than one type. The examples with the very rounded toes (type 1) could be from an ornithodpod dinosaur like the Iguanodonts but smaller. Several of the prints seem to have been impressed into a salt crust (discussed below), with a raised impact rim or bourrelet around them. This rim is now seen in the casts from beneath so that it appears to descend around the margins into the inverted block of limestone. Some footprints (type 2) have no impact rims and these seem to be from a higher level in the bed of calcarenite.

The footprints discussed here are present at or near the base of the Hard Slatt, the equivalent on the Isle of Portland of the main limestone, the Hard Cockle Oolite, of the Hard Cockle Member on the Dorset mainland. Some possible, but very uncertain, footprints from this bed at Worbarrow Tout are discussed below.

The Underlying Spore Beds

Beneath the Hard Slatt are thin-bedded marlstones or argillaceous limestones, each about a centimetre thick, alternating with very thin grey clays or marls (often less than 1 mm thick). These beds represent the relatively argillaceous top of the Shingle, a sequence of ripple-laminated limestones. If the sequence is regarded as cyclical this part is the regressive top of a cycle and analogous in terms of position within a cycle to the Great Dirt Bed of the Caps. In an undergraduate research project, Clasby (2003) studied the footprint slabs and in particular these argillaceous beds. Pseudomorphs after halite occur in these. Insect remains include wings and elytra.

Sedimentology - The Salt Flat and the White Sand Beach

On the Isle of Portland the Hard Slatt is a calcarenite limestone of rather less than a metre in thickness with cross bedding. It is partly oolitic, consisting of small ooids together with intraclasts and algal features. The ooids are not of marine Porland Stone type and not reworked from this lower formation. The carbonate allochems are particularly coarse-grained at the base, as shown in the photograph above. This basal layer, just above the salt-crust, consists of rather poorly-sorted, subangular to subrounded, carbonate grains. The deposit is unusually coarse for the Lower Purbeck lagoonal sediments, and may represent the nearest approach to "beach shingle" as the transgression washed carbonate material across the flooded sabkha. The main part of the Hard Slatt originated as a broad, low-profile (rather flat) storm beach of the lagoon (to the southeast). This beach was of white carbonate sand.

A feature of the lower part of the bed is an abundance of flat mud-clasts which decrease in abundance upwards. Halite pseudomorphs within the mud-clasts show that they originated as desiccated mud sheets on a salty sabkha, like that which preserves the dinosaur footprints. These mud-clasts are also numerous in the equivalent bed on the mainland. The limestone shows cross-bedding, in some cases of herring-bone type, which is also a notable feature of the bed on the mainland. Its origin is probably quite rapid storm deposition of high-energy lagoonal, carbonate sediment. This must have been previously accumulated over a period of time and probably includes debris from the erosion of nearly contemporaneous sediments, including those of supratidal origin.

Most of the tracks are associated with casts of small, concave-faced, cubes of halite at the original surface, with clay beneath. It would seem that in most cases the dinosaurs walked over salt-encrusted, firm but slightly plastic mud. Some comparision with modern salt-crusts, like those common in sabkhas and playa lakes, is shown above (although it is important to note that the climate was not as extreme as that of modern Middle Eastern deserts - note the presence of trees). The seasonal Mediterranean climate would imply that the prints were formed in the dryer summer season, or perhaps spring or autumn, but probably not in the winter. A rapid spread of carbonate sediment over the sun-dried mud would have preserved not only the indentations but also the impressions of the halite crystals. Although the carbonate shows rather high energy conditions, there was no major erosion just here but relatively rapid deposition associated with a transgression. The large mud-flakes show that the surface of dried mud was ripped up nearby. The extent of carbonate sand accumulation at this particular level in the Lower Purbecks in the region is interesting. It is located at the northwestern margin of the basin on the Lulworth and Portland Swells, but is not developed in the Durlston Bay basin facies. The position of the relatively coarse-grained carbonate accumulations in relation to the broad palaeogeography suggests that southeasterly storm winds were driving waves onto the shallows of the structural highs. Those footprints which originated above the salt layer, and generally without impact rims, were formed on the beaches or shoal of carbonate sand that thus developed at the lagoon margin.

For more on Purbeck sediments associated with dinosaur footprints see the Durlston Bay, Middle Purbeck webpage.

Introduction to Dinosaur Footprints

Many examples of dinosaur tracks and trackways are known around the world. The prints are usually in sandstone, limestone or hardened clay. The original track may be seen or an undertrack in which parallel laminae beneath are pushed down. The footprint or track may be seen as a mould or a cast or as both.

Although there are many different types of dinosaur footprints, in very simple terms most, but not all, can be classified into one of three categories. These are: 1. Sauropod Footprints - large roundish prints like those of an elephant's foot and made by the large quadrupedal dinosaurs like Brachiosaurus or Diplodocus; 2. Ornithopod Footprints - tridactly (three-toed) prints of broad, rather flabby type made by the bipedal herbivorous dinosaurs like Iguanodon ; 3. Theropod Footprints - tridactlyl footprints of narrower type, with claws, made by the bipedal carnivorous dinosaurs. The first category - the Sauropod footprints, are quite distinctive. The diagram above shows the main differences between the Ornithopod and Theropod footprints. This is simplified and there are many variations of shape and complications of preservation which can make it difficult to place a track or trackway in one of these categories. This simple classification does not take into account some smaller quadrupeds. Apart from the description, in a full study the dimensions of the footprints must be measured, and, in addition, the pace between two prints and the stride between two prints of the same foot (i.e. two paces). Other variables include the angles between the toes, and the extent to which the prints turn out or turn in (pigeon-toed) in relation to the direction of the trackway. For more information on the study of dinosaur footprints particularly see Lockley (1991)

Notice in the images that follow that most of the prints (tracks) from this locality are relatively small for the Purbeck Formation and are not longer in the anterior-posterior direction than a man's foot. Some seem clearly to be of the broad Ornithopod type, although a small variety. Fortunately, one example, a small one reveals in cross-section that it is not an undertrack, but was made directly into a salt crust. Thus the broad shape is original and not due to some distortion at a depth beneath the true print. Since no print as large as that typical for an Iguanodon was seen an origin from smaller Ornithopod dinosaurs must be considered. Other tracks are narrower and may be of Theropod type. In one case there is some overlap of prints. Is this the result of one dinosaur following the trackway of another or has it been produced by some genus of quadrupedal dinosaurs. These apparently Theropod-type tracks are also small. It is possible that the semi-arid Lower Purbeck environment was less favourable for dinosaurs than the sub-humid Middle Purbeck environment. Examine the footprints, now, with these various ideas in mind.

Probable Ornithopod Tracks

The most impressive of the dinosaur footprints on Portland are those with well-defined impact rims. Examples are seen on a partially inverted slab here. These prints are present at or near the stratigraphical base of the Hard Slatt Bed and were impressed into a thin salt-crust over mud. The pseudomorphs or, more strictly, sediment casts, of halite are obvious. The impact rims were in some cases almost entirely of halite and in other cases involved mud. In one case, shown below, the impact rim was of a few millimetres of carbonate sand on salt-crust on mud. The features are clear in the photographs.

The footprints of this type have broad-toes and in this respect resemble Ornithopod prints. They are like small versions of "Iguanodon footprints". They are, however, about the same size as the Theropod-type prints shown further below. That is they are about 17 to 20 cm in length with a pace or step of about 0.8 cm. The impact rims are particularly well developed, and it is significant they may include a sand layer, as shown in a photograph below. This does mean that the toes appear wider than they really were. Thus although these prints seem to have originated from a broad-toed ornithopod, some degree of caution is needed in interpreting them.

Many of the footprints are impressed in a salt-crust, as shown by a surface of halite pseudomorphs or halite casts. These particular pseudomorphs are a surface type in which the downward protruding lower parts of the halite crystals have been filled with carbonate sediment. They are not where isolated halite crystals within a carbonate bed have been dissolved and the cavities filled with calcite. This latter type is common in the Soft Cockle dolomites and is of different origin. The pseudomorphs at the base of the Hard Slatt are in lateral contact, forming the preserved remains of the salt crust. Some halite has been squeezed up between the toes of the dinosaur footprints and preserved as pseudomorphs. There is clear evidence as shown by the photographs that the crust was deformed by dinosaurs walking over it.

Here are some more with impact rims, shown with different type of surfaces and in different light conditions. In both cases the footprints are relatively small for dinosaurs, with a short step or pace. As in the other examples there is no impact rim at the rear of the footprint. This is evidence that the dinosaurs were semidigitigrade (partly on the toes) with some impression of metatarsals but not the heel. They were not fully digitigrade as is the horse.

Shown in different light conditions, here are three prints of similar type and size, about 18cm in length, but with one them being offset and perhaps part of another trackway. Note the signs of slippage towards the large depression of contorted salt-crusts. The origin of the depression is not clear. It might have been a large sauropod print or it might have been produced by a dinosaur struggling in the mud and salt. If it already existed at the time when the small dinosaurs passed this way it might well have caused some minor sliding of one foot towards it. Notice incidently that there is some evidence of interference ripples and the environment had clearly had been very shallow water at one stage.

The first footprint seen in closer view. There are imperfect cubic casts of halite crystals, with concave faces, on the limestone surface (the under-surface). Notice how in this case mud has been pushed up into the impact rim or surrounding bulge. This is visible as some grey mudstone in the view from underneath. Notice the anterior-posterior ridges with parallel depressions developed in the central toe. The toe prints seem to end anteriorly in a rather indistinct rounded outlines, without any claw being obvious. The front of the footprint is not deep as is the print of a horse hoof when the horse is cantering. It is more the like the print of an animal that is just walking.

Here are two more prints of broad-toed type. Shown more clearly in one photograph is the evidence of a thin carbonate sand impact rim. The presence of this shows that the salt crust was covered in this case by a thin sheet of carbonate sand and the animals were not walking directly on the salt. The evaporite was concealed under the sand layer. Notice the curved brittle-fractures. The dinosaur footprints were projecting downward as casts when compaction occurred. They would have impacted on the stratum beneath and consequently have been pushed upwards when the rock was in brittle-fracture condition.

These prints may give some indication of dinosaur behaviour. There seems to have been some testing of the the salt-crust by this Ornithopod, using the side toe and also the central toe. Presumably when satisfied that the creature would not sink deep it placed its foot fully on the crust. This was pushed down for a few centimetres and mud was squeezed up around the foot to form an impact rim. Seen from the underside of the Hard Slatt, this impact rim appears as a depression surrounding the footprint.

Ornthipod-type Tracks of Smaller Size

Some small prints of the same broad-toed type with impact rims are shown here. There are three small footprints in an approximate line and with regular spacing and two of these are shown here. On the far side of the slab is a larger footprint. There are impact rims and mud which was below the thin salt-crust has pushed up around the toes. This mud is now visible as grey mudstone. The directions of the larger and smaller tracks are not exactly the same, but both are in the same general direction and they are very close. Both have been impressed into the same salt crust. A cross-section of the first of the small footprints is shown above and this indicates that it has originated at the level shown and has not been pushed down from a higher layer. It is not an undertrack or underprint.

A cross-section of a larger footprint, already shown above, provides further confirmation that the tracks seen are true tracks and not undertracks formed several layers below the surface at which they originated. The cross-section is the natural break through the right-hand track of the left image. As noted above it is possible and, in fact, quite likely that a few mms of carbonate sand may separate the original surface from that which is seen, but the separation is small.

Ornithopod Iguanodon-type Track

A single large footprint of the broad, flabby Iguanodon type, like those well-known from the Middle Purbecks is present in the same slab as that containing the very small footprints. Compare this to the Middle Purbeck examples, shown below.

Theropod-type Footprints - without Impact Rims

Left: Here are four small footprints of the narrow-toed type. They seem to be in two pairs with offset, and slight angular change of pointing direction. This is presumably being because they are both right and left foot prints. For scale the two white bands on the pen are at 5 cm. These prints are deeper at the toes and the heel is not properly developed.

Right: Narrow-toed footprints, wettened to increase visibility.

Three photographs shown above, some taken in different light conditions, are all of the same slab. These prints are of narrow type without Impact Rims or marginal bulges. In particular the narrow-toed print at the front of the series seems almost theropod-like. The one at the rear has toes that are broader and flatter in appearance and although incomplete it is more like some of the others with broader toes. Clearly the prints from one animal can appear rather differently according to the behaviour of the mud and the details of preservation. Thus, interpretation has to be cautious.

This example is not from the pile of slabs at the quarry, but is in situ in the west cliff of Portland . The Hard Slatt is well-exposed in a cliff-top ledge directly about the centre of Mutton Cove . It is just above an old cliff quarry in the Portland Stone. The bed here is thicker than at some other places and is divided into two parts by a central parting of shale. At the base of the upper part is what appears to be a cast of a tridactyl dinosaur footprint. The surface at the top of the shale, or mud, as it was, has been indented by three toes and a slight impression of a heel. There is no obvious impact rim and halite pseudomorphs were not seen at this level (although they occur in associated strata beneath the Hard Slatt). The significance of this footprint is that dinosaur footprints are not confined to one locality and not confined to the base of the Hard Slatt. The conditions were not only favourable for dinosaurs in this region during deposition of the Hard Slatt, but that the special sedimentary characteristics of this bed, particularly the content of coarse carbonate sand, were appropriate for preservation of footprints. It is hoped that more footprints may be found in this bed elsewhere, although they are certainly not common enough to be obvious at other localities. For more information on the Mutton Cove locality, see the Mutton Cove webpage and the undergraduate project report by Will (2003), which contains some petrographic information.

More Theropod-type Dinosaur Tracks

These tridactly footprints are rather larger than most at this locality. Two of the prints are partially overlapped by others. In addition to these four there are two indistinct and incomplete tracks in a different direction. These footprints differ from most of those on Portland. They have the narrow-toed characteristics of theropod dinosaurs as shown by the diagram above. Notice the impression of the pads on the feet and of claws in some cases. They might have been produced by two theropods (bipedal) following the same route. (Note that in the comparison diagram the drawings of the tracks of Lockley have been specifically positioned for comparison, and one has been cut, but the shapes of the prints have not been altered. See the original, complete diagram in the recommended book of Lockley (1991) for more information.)

These footprints seem to be of theropod type. They are spaced at about 95cm and only two are visible. If this was the pace (between right and left footprints) then this was made by a larger dinosaur than indicated by the other footprints. The lack of a third print between them seems to show that this was the pace and not the stride, which would, of course, have been twice as great. However, this is not proven because there is an indistinct mark to the right of the trackway, and it is just possible that the right foot impression has not been clearly made in the salt crust. A dinosaur might have been turning slightly left with more pressure on the left foot.

Ridges on Theropod Footprints

Some footprints show ridges on the base of the central toe. An example is shown above. These are in a antero-posterior direction and presumably form a sort of "tyre-tread". The particular footprint has some resemblance to the theropod-type discussed elsewhere on this webpage.

When the slabs were moved under the direction of Richard Edmonds on the 16 September 2003 some new and interesting tracks were revealed. Brushing a dirty slab revealed prints that have only two well-defined toe impressions, not three as is normal. There central toe print is light and indistinct in all three tracks, i.e. for both right and left feet. Favourable lighting conditions, however, does show a narrow central toe with a claw in two of the three prints. The dinosaur seemed to have had rather narrow toes and to have been walking plantigrade or even heel-down rather than in digitigrade manner, as was the case for most of the tracks. It was difficult to explain this.

The possibility of a Deinonychus or a similar dromaeosaur type of theropod dinosaur was initially considered. The second toe with the large claw of Deinonychus was held clear of the ground, and this seemed to me to provide some explanation. Thus it was thought at first that perhaps the tracks were produced by a dinosaur like Deinonychus or similar dromaeosaur. However, the raised "killer claw" of Deinonychus is the second toe of four toes. The first toe on the outside was small and would probably have made no impression. Considering just the three, it would have been the outside one with the large raised claw that should have resulted in the weakest marking, if any. This does not seem to correspond with the images. As is clear from the images here, it is the middle one of these three toes which is very narrow and shallowly impressed.

Shown above is another image of the slab photographed from a different angle. This view seems to indicate traces of the termination of the middle claw. Notice again the 'flat-footed' stance of the dinosaur and also the weak impression of claw-marks of the side toes. These claws seem to have been raised to a limited extent.

The left-hand photograph (slightly oblique and therefore apparently narrowing the print) shows one of the prints in some detail. It is effectively unmodified, just having had some adjustment to gamma and sharpness overall. The middle image (also slightly oblique) has been enhanced but in a subjective manner after study of the left image. This image has the apparent area of the footprint shown in higher contrast than the surrounding which have been intentionally faded. If the the selection is correct (and you can judge for yourself) then the footprint is very bird-like with a narrow and light impression of the central claw. The right-hand image (near normal direction but not precise with regard to x-y ratio) has been constructed on the computer by superimposing one image on another of the same footprint. This method also seems to produce a bird-like footprint.

The origin of these footprints is at present a mystery. Is the absence or faintness of an impression of a central toe the result of a dromaeosaur-type sickle claw, raised above the surface, or is there some other explanation?

Imperfect Footprints - with with flat bases

These examples are rather poorly preserved and are almost flat at the base, with very limited depression. This bottom layer is the salt crust. They are probably being viewed at a few centimetres below the level at which they were made. The result is that they are less distinct and well-shaped and one of those on the left photograph is very broad, probably because of some oblique shearing. The upward injected mud, the impact rim is at the margin of this broad area. These probably represent a later passage of dinosaurs across a few centimetres thickness of carbonate sand above the salt and mud. The difference in type of preservation supports the view that the footprints are not all at one horizon.

Very broad-toed Dinosaur Footprint

Here is a very broad-toed footprint with a circular print of some type in front of it. They might represent the tracks of pes (foot) and manus (hand) of a small quadrupedal dinosaur, but this is not confirmed and there could be some other explanation. There seems to be evidence of a varied assemblage of small dinosaurs of several different genera.

Large Depression of Unknown Origin

An unexplained feature of the base of the Hard Slatt at the dinosaur footprint locality on the Isle of Portland is discussed here. An unusual and large depression of sediment with salt-crust that had been disturbed when still plastic was found. It in association with some small dinosaur footprints, three of which have been described above (notice that there is also a poorly-defined fourth example on the left). The depression is at least a metre in size, but the containing slab is broken and its limit on one side is not known. It has zigzag distortions that seem to be folds in the salt-crust. Some collapse or mechanical force was necessary for this type of deformation. It was formed before the brittle-fracture stage, shown by some structures on the slabs, and is not, therefore, the result of deep burial processes or late tectonism. The feature might be some early sedimentary structure such as a spring-pit, or, alternatively, like the small tracks, it may have resulted from dinosaur activity. A large sauropod footprint origin is possible, but there is no certain evidence for this. At present it remains a mystery.

At Worbarrow Tout, on the Dorset mainland, depressions were found at about the same level, just under the Hard Cockle Oolite, the equivalent of the Hard Slatt. They are shown for comparison. These are of about the same size and are also impressed into evaporites, although in this case a bed of gypsum, since calcitised. These depressions also could be sedimentary structures or they might be sauropod footprints.

Tool Marks and Unexplained Grooves etc.

The base of the Hard Slatt is characterised by well-developed halite casts, not only at the main site of the dinosaur footprints but elsewhere on the Isle of Portland. The main salt crust at the base in which the footprints are present contains some other features of interest at another locality. Here there are tool marks made by some pointed or blade-like object cutting into the mud. The tool marks are preserved as casts, relatively narrow but often with a flat base. They were grooves cut in the mud and then infilled with carbonate sand like that of the main part of the Hard Slatt. Some streaks and the manner in which one part of the mark crosses another suggests that the object moved overall from left to right. It is particularly interesting to note, though, that there are two set of oscillations recorded by this ancient blade. This is shown in the images above, where the groove swings back again in the reverse direction. The two reversal grooves (right to left) are of about the same length. At the loops the grooves seem deeper and oblique. In other places the grooving is not well-developed and all that is visible is a disturbance of the mud by the moving object.

A probable explanation of the origin of the markings follows but it is not by any means certain. The grooves have most likely been made by a floating object such as a log with a projecting branch, or a dead saurian or bird body with a hanging spine or beak. As noted, the shallow lagoon was subject to desiccation and development of salt crystals and dinosaur footprints. At other times, as during deposition of the main part of the Hard Slatt it was flooded by shallow saline or hypersaline water. During the course of such a flood some floating debris would have been transported landwards at the margin of the lagoon. Small waves would be expected to have been present, particularly since oscillation ripple marks are common in the Hard Slatt. Transport of the floating and mud-penetrating object landward by the flood 'tide' at the lagoon margin was likely but it was disturbed by small waves with back-wash. Large waves would have been unlikely in the shallow Purbeck lagoon and would have destroyed delicate structures. The deeper loops in the grooves might indicate small breaking waves. Shallow indentation or mere disturbance of the mud suggest that at these phases the object was lifted by the brine, presumably near the crest of a wave.

Since such delicate and sharp grooves were not likely to be preserved in good condition for long, it is likely that the sedimentation of the carbonate sand of the Hard Slatt took place almost immediately after the grooving. The sand would have been drifted in with this transgression. It is probable that the grooving and filling took place at the same time as the dinosaur footprints were preserved as casts by the carbonate sand.

These markings warrant further study because they may reveal more information about the conditions during this transgression by the lagoon that brought in coarser carbonate sediment at the start of deposition of the Hard Cockle Member. The traces are in a loose block but it might be possible to establish some directional data, using structural features in the block for orientation.

This zigzag mark is present in the slab (at the main dinosaur footprint site) which has three broad-toed tracks and with a very large irregular depression. If this mark had been found with a direction corresponding to the footprints it would have been interpreted as a tail mark with some confidence. Unfortunately it trends at right angles to the small tracks. It is close to the large impression which might be of sauropod origin and could be due to some fracturing of the salt crust affected by the impact, or of some inorganic origin. Thus its origin is unknown but it is shown here in case some future finds shed light on it.

These small curved indentations resembling claw-marks were found in a relatively thin (a few cm thick) slab of white, pelletoidal, lagoonal limestone. This is amongst the debris of finer material dumped alongside the large slabs with the dinosaur footprints on the Isle of Portland. The probable reptile skin impression came from this pile of debris. The thinner-bedded limestone here is likely to be from above the Hard Slatt horizon, because some thin beds of limestone are present there, and the material seen does not resemble the Shingle which lies beneath. No thorough investigation of its source has been made though. The origin of these marks is unknown, but they might possibly be from the claws of turtles, mammals or lizards or the beaks of birds. Clearly recognisable footprints have not been found in the slab. There is some indication that the marks are paired with a separation of about 1.5 cm to 2 cm. There are three possible pairs present.

These markings were found in a replaced salt crust that is probably the base of the Hard Slatt, at a more northern locality. They are not typical footprints because they do have a distinctive shape with toes and are isolated and not parts of trackways. They have been calcium sulphate nodules like those of modern sabkhas but such nodules normally form beneath the surface and not at the surface. One of them has an impact rim which suggests that it results from some object pressing down into the salt-crust and underlying mud. Although found in blocks, not in situ, which are at present about 200 metres apart, and in slightly different sediment, they are similar in shape and size. There is no obvious explanation as to their origin. Could they have been coprolites?

These problematical markings are not from the Portland dinosaur site but are from the Middle Purbeck Formation of Durlston Bay and are shown for comparison. They might possibly be scratch marks made by swimming or floating reptiles because of some limited similarity to marks figured by McAllister (1991, fig. 37.1 - fig. 37.3, p. 345) in the Cretaceous Dakota Formation of Kansas. However, they are much more elongate and do not have clear features which prove them to be of reptilian origin. Another possibility is that they are impressions of some plant debris, perhaps of cycadophytes. They are shown here, merely as another puzzling trace fossil of the Purbeck Formation which, just might be of reptilian origin. The block of limestone (in the possession of Ian West) is ex situ from the debris of the upper part of the Middle Purbeck Formation (Upper Building Stones up to Chief Beef Member). It is of lagoonal shell limestone, a biosparrudite with Neomiodon shells, of a type common in the beach material of Durlston Bay. It is a rather distinctive, blue-hearted, somewhat irregular limestone. It was encountered in landslide debris at the Zigzag path in the centre of Durlston Bay, many years ago. Should the block prove to be of any significance it is likely that further study would enable the limestone to be matched to that of one of the biosparrudite beds and the exact horizon identified.

Small Upbulge Structures of the Hard Slatt

These circular and oval features occur at the base of the Hard Slatt at the dinosaur footprint slabs in association with the dinosaur footprints. The footprints, however, show no brittle-fracture cracks, whereas the circular or oval objects are marked out by such cracks and in some cases parallel pairs of cracks. The footprints are casts of downward impressions, but the brittle-fracture structures are upward (although they are seen now in overturned blocks and thus appear superficially as depressions). Because these strange features show brittle fracture presumably they are much later in origin than the footprints. The bed has presumably become partly or completely lithified before these were formed. The features represent bulges made by something that was resistant when the bed was compressed under burial and collapsed to a small extent around the circular or oval plates of a resistant substance. The most likely cause were large disc-shaped mud-clasts, but there are other possibilities.

Associated Sedimentary Structures - Ripple Marks etc

Oscillation ripple marks characterise the top of the Hard Slatt at most localities on Portland. They are also present on the mainland. The Hard Slatt, the bed from which the blocks containing dinosaur footprints have come, is well-exposed at the cliff top above Mutton Cove . There is also a dinosaur footprint here. At Mutton Cove the ripple marks on the top are particularly conspicuous.

Top-left image: Oscillation ripple marks are present on the slabs of Hard Slatt limestone at the dinosaur footprint site. These slabs are associated with those which contain dinosaur footprints and seem to show the top of the bed of limestone. They show some features of oscillation ripples. A very shallow-water environment, like that of a beach, is implied. Some truncation of the ripple marks shows that the water was only a couple of centimetres deep before burial and preservation.

Top-central and top-right images: Similar ripple marks are present on the top surface of the Hard Slatt in situ in the cliff at Mutton Cove . Ripple marks also occur at partings within the bed.

Bottom-left image: General view of the Hard Slatt Bed above Mutton Cove. Ripple cross-lamination is abundant in this bed and there are many rip-up clasts of clay.

Bottom-central image: Ripple cross-lamination close to the junction of the Shingle and the Hard Slatt, in the cliffs south of Mutton Cove

Bottom-right image: These vee-shaped markings were found on a slab of the Hard Slatt which has fallen to the undercliff in the climbing area, south of Mutton Cove, near the topple. They are probably two intersecting sets of ripple-marks at different levels.

It is clear that the Hard Slatt is an extremely shallow water deposit from bottom to top. The interesting petrography with abundant intraclasts and reworked ooids shows relatively high energy conditions for the lagoonal Purbeck Formation. Halite pseudomorphs and blue-green microbial ('algal') remains indicate high salinities. It seems to have formed as a type of lagoon-beach of white carbonate sand, but was very flat and extensive with little sign of development of any large bank or berm.

The Tabner Find - Reptile Skin Impression or Sedimentary Structure?

In loose blocks at the site of the dinosaur footprints some anomalous markings have been found by Stuart Tabner. These are present in thin-bedded Purbeck limestone where there is contact with a lamina of black shale. The facies is typical lagoonal Lower Purbeck (usually hypersaline to some extent) but fish remains in the form of a vertebra are associated. Another vertebra has been found in associated blocks.

The origin of these impressions is a matter of much interest. Early discoveries suggested that they seem to cover an area of at least 30 cm square, and perhaps more, as a thin continuous flat sheet. They have small whorls and roughly parallel rib-like features. There is no recognisable bone, other than an attached fish vertebra, and no significant quantity of carbonized plant debris is preserved. The present writer's first thought was that they represent an impression of reptile skin in the soft carbonate mud at the margin of the shallow lagoon. Comparison was made with the skin of an alligator, as shown above, and there seems to be some similarity but there has been no definate confirmation of a crocodilian origin. The markings do not include clearly-defined scales such as have been seen in dinosaur skin impressions, although Ankylosaur skin has been mentioned as a possibility. They have not been recognised as of plant, fish, burrowing ichnofossil or sedimentary-structure origin. The markings are not those of a calcified stromatolite or microbial mat; they are just an impression in pelletal carbonate silt. That they are possibly the remains of some thin organic mat has not been completely eliminated but nothing similar has yet been noticed.

It is perhaps worth mentioning that the rib-like features and the whorls seem to have fairly constant characteristics, although there is some variation. The rib-like features are mainly parallel, so there is a lineation to the structure. It gives an impression of having been made by some well-organised organic sheet or film. The whorls are rather cone-like with very thin darker laminae descending into them at low angles. No large tears, breaks or folds have been seen. The limestone is very ripple-laminated and mostly unfossiliferous like several of the Purbeck limestones of the Hard Cockle Members of Portland. The way-up has not yet been determined.

Between the whorls there are double-Y shaped structures, almost like girder bracings on a bridge. A photograph of these is shown above, with some colour enhancement. Examination of the main specimen photographs higher up will reveal several examples of these markings. Closer study of the whorls suggests that they are layered in a shallow conical manner with two or three parallel 'dishes'.

Further searching by a group including Stuart Tabner, Debbie Tabner and Linda in October 2003 revealed more of the markings in several blocks found in different parts of the pile of Purbeck debris. This suggests that the markings cover an appreciable area, perhaps to be measured in metres. No folds, tears or special structures were found. Unfortunately, this seems to indicate a sedimentary structure origin, because it would be surprising for a skin impression to be so even and so extensive. The markings have been located at a very finely laminated, almost varved, argillaceous limestone layer at the margin of a 10 cm thick bed of Purbeck lagoonal limestone. This may be the "Coombefield Thin Limestone Bed" above the Hard Slatt. (more information will be added).

The markings resemble to some extent those in the shed skin (left image) of a gecko lizard, which my wife keeps. The comparison is shown in the image on the right, although, of course the gecko is small and the markings are small. The gecko skin is from the underbelly. The features are not identical, of course, but there is much resemblance. There is, indeed, sufficient similarity, in the view of the author, to support the view, which came initially from illustrations of crocodiles, that these fossil markings really are an impression of reptilian skin. If the markings are roughly in proportion, the fossil skin would have been of a creature at least an order of magnitude larger than a small pet gecko. Reptiles of this size were common in the Purbeck environment.

Confirmation of a reptilian and probably dinosaur origin comes from an illustration in Edmonds (1979) of an impression of Iguanodon skin found with 30 Iguanodon skeletons from the famous Cretaceous Bernissart locality in Belgium. The discoveries were made in 1878. The illustration is not of good quality and is not reproduced for copyright reasons (it is hoped to obtain a good copy in due course). The skin has ridges rather like those in Stuart Tabner's specimens, although not so straight. It has the Y-shaped junctions. The scale is not given. Edmonds wrote " According to experts the pattern of markings could mean only one thing: it was an impression of the iguanodon's skin. From this we can assum that the animal was not so much scaly as covered by a skin with regular little bumps." No typical Iguanodon footprint has yet been found at the Portland locality but the small broad prints are effectively smaller versions of Iguanodon - type prints and may well have been made by a related ornithopod. The origin of the Tabner impression is, of course, not firmly proven but it is very likely that it really is reptilian skin and may well be of dinosaur origin.

Sedimentary Structures: Small Channels or Gutter Casts

Some small channel-like features, resembling gutter casts, occur in some slabs of Purbeck lagoonal limestone at the dinosaur footprint site. These are not from the Hard Slatt with the footprints, but from a different and thinner bed of limestone, probably from a short distance above. At the base of the small curved channels are lag concentrates of ostracod valves and minute gastropods. Although these are small objects they are coarser than the general sediment of the bed which is probably pelletoidal limestone.

Right: Compare these to gutter marks or gutter casts at Staithes, Yorkshire, in the Middle Lias. There the small channels are in a marine deposits and may have been formed in subtidal sea conditions. The Purbeck example is from a very shallow and restricted lagoon. If further work shows that these are true gutter marks then the possibility should be considered of such sedimentary structures having been formed in very shallow environments such as the "intertidal" zone.

The Hard Slatt (Dinosaur Bed) in Some Quarry Sections

There are several quarry sections which include the Hard Slatt or dinosaur footprint bed. Footprints have not as yet been found in situ except in a cliff exposure at Mutton Cove, referred to above. They may well be found in place in one of the quarries sooner or later. The Hard Slatt is well seen and accessible at Windmill Quarry (North Perryfield); it is present at the main Perryfield Quarry, at Bowers Quarry at Coombefield Quarry at Godnor and at several other places. Some examples will be shown here and some more added later. It is well above the level of the Caps and Dirt Beds (which are always obvious) and stands out above the softer Shingle beneath it. It is best seen in the more central parts of the Isle of Portland, being eroded away in the northern part such as Kingbarrow, and in the south at the Bill.

Middle Purbeck Footprints for Comparison

Left: Dinosaur footprints have been found from time to time in the quarries inland to the west of Swanage. They are present mostly in the Intermarine Member of the Middle Purbeck Formation, particularly the Roach Bed and adjacent strata. These footprints, unlike those in the Hard Slatt, are in biosparrudite limestone composed of lagoonal bivalve shells such as Neomiodon. Some of the footprints are of Iguanodont type. These broad-toed footprints resemble the broad-toed examples from the Hard Slatt but they are much larger. An example is shown here.

Centre and Right: Here are some old photographs (1978) of dinosaur footprints in a dipping slab of biosparrudite of the Intermarine Member of the Middle Purbeck Formation at Worbarrow Tout. Mud has been placed in the footprints to give them emphasis for the photographs. These footprints are still visible but rather worn and degraded now.

BIBLIOGRAPHIES, References and Internet Links

Go to Bibliography and References on Geology of the Isle of Portland.. Most papers referred to in this webpage are in this bibliography.

Go to Bibliography of the Purbeck Formation. This in alphabetical order by author and not by topic.

Go to Bibliography of the Purbeck Formation, Topics. . This is split into topics in alphabetical order.

Go to Select Bibliography of Dinosaurs in the Purbeck Formation (in preparation) .

Go to Middle Purbecks of Durlston Bay. This also refers to Purbeck dinosaur footprints.

Additional References (not in the other bibliographies)
.
Lockley , M. 1991. Tracking Dinosaurs: A New Look at an Ancient World. Cambridge University Press, Cambridge. 238pp. ISBN 0-521-39463-5 hardback, ISBN 0-521-42598-0. By Martin Lockley, University of Colorado at Denver. Extract from Preface: "The main objective of this book is to convey to the lay reader the substantial contribution that tracks make to our understanding of the Age of Dinosaurs. The professional reader should also benefit from the book because it deals with many new discoveries and synthesises much of what is known." [This is a very good book for understanding the basis of the study of dinosaur footprints and clearly explains terms like stride, step, pace, pace-angulation, plantigrade, digitigrade, true-tracks, undertracks, impact rims etc. and has many illustrations of tracks made by different genera of dinosaurs. It is a recommended introduction to the subject.]
.
Gillette , D.D. and Lockley, M.G. 1991 (first paper edition - 1991, hardback first edition - 1989). Dinosaur Tracks and Traces. Cambridge University Press, Cambridge. 454pp. ISBN 0-521-36354-3 hardback, ISBN 0-521-40788-5 paperback. [This is a volume of 51 papers by various authors which originated at a conference: First International Symposium on Dinosaur Tracks and Traces. It was held at Albuquerque, New Mexico, at the New Mexico Museum of Natural History on May 23-24, 1986. Topics in the volume include: Introduction; Historical Perspectives; Locomotion and Behavior; Eggs and Nests; Paleoecological, Palaeoenvironmental and Regional Synthesis; Biostratigraphy; Experimentation and Functional Morphology; Site Reports; Systematic Ichnology; Conservation and Preservation. It contains a large quantity of information with details and references to previous work. It is recommended for further study.]
.
Martill , D.M. and Naish, D. (Eds.) 2001. Dinosaurs of the Isle of Wight. Palaeontological Association, Field Guides to Fossils: Number 10. The Palaeontological Association, London. 433pp. ISBN 0 901702722. Edited by David M. Martill and Darren Naish, technical editing by David J. Batten, photography by Robert Loveridge and computer generated artwork by Stig Walsh. Price of paperback in 2003 - £16. Contents: Acknowledgements; Preface; 1. Introduction by Martill, D.M., Naish, D. and Hutt, S.; 2. The Geology of the Isle of Wight by Martill, D.M. and Naish, D.; 3. The Global Significance of the Isle of Wight Dinosaurs by Martill, D.M. and Naish, D.; 4. Taphonomy and Preservation by Martill, D. ; 5. Ornithopod Dinosaurs by Naish, D. and Martill, D.M.; 6. Boneheaded and Horned Dinosaurs by Naish, D. and Martill, D.M.; 7. Armoured Dinosaurs: Thyreophorans by Naish, D. and Martill, D.M.; 8. Saurischian Dinosaurs 1: Sauropods by Naish, D. and Martill, D.M.; 9. Saurischian Dinosaurs: Theropods by Naish, D., Hutt, S. and Martill, D.M.; 10. Dinosaur Trace Fossils: Footprints, Coprolites and Gastroliths by Martill, D.M. and Naish, D.; 11. Pterosaurs by Howse, C.B., Milner, A.R. and Martill, D.M.; 12. An Aid to the Easy Identification of the Commoner Wealden Dinosaur Bones by Martill, D.M. and Naish, D.; 13. A Dinosaur Trail by Martill, D.; Glossary of Terms; References, Appendix; Further Reading; Useful Information; Systematic Index. [The book is mostly in monochrome with numerous photographs and diagrams, but there are 16 colour plates in the central part. It contains a mass of information and is exceptionally good value for money. In addition to its main use regarding Wealden strata it is relevant to the Purbeck Formation of Dorset, because it discusses the dinosaur contents of a higher part of the Lower Cretaceous in the same general region, and some similar dinosaurs might be expected to have been present in the Purbeck (Berriasian). There is specific reference to Purbeck dinosaurs and Purbeck dinosaur footprints in places, as for example on pages 134 and 135 where Echinodon and Nuthetes are discussed. The chapter on dinosaur footprints, pages 310 et seq. is clearly of significance regarding Purbeck footprints and these are mentioned briefly (particularly of Iguanodon).]
.
McAllister , J.A. 1991. Dakota Formation tracks from Kansas: implications for the recognition of tetrapod subaqueous traces. Paper 37, pp. 343-351 in: Gillette, D.D. and Lockley, M.G. 1991 (first paper edition - 1991, hardback first edition - 1989). Dinosaur Tracks and Traces. Cambridge University Press, Cambridge. 454pp. ISBN 0-521-36354-3 hardback, ISBN 0-521-40788-5 paperback. Abstract: Criteria for the recognition of tetrapod trails are compiled from the literature and are illustrated by one example of footmarks from the Dakota Formation [Cretaceous] of Kansas. The Kansas footmarks are most easily recognised by criteria which indicate tracemaker bouyancy. Such foomarks are characterised primarily by posterior overhangs and reflectures of the individual digit impressions; and secondarily by striations and claw marks along their length and the often incomplete nature of the trails. These swim footmarks would be expected to grade into subaqueous traces formed by more typical terrestrial propulsion and demonstrate less bouyancy as the water becomes shallow, and disappear as the tracemaker becomes fully bouyant in deeper water and digits no longer reach the substrate.
.
Zhang, J. Lib, D., Li, M., Lockley, M.G. and Baib, Z. 2006. Diverse dinosaur-, pterosaur-, and bird track assemblages from the Hakou Formation, Lower Cretaceous of Gansu Province, northwest China. Cretaceous Research, Vol. 27, Issue 1, February 2006, pp. 44-45. By Jianping Zhang, Daqing Li, Minglu Lia, Martin G. Lockley, and Z. Baib. Available online 10 February 2006.
Abstract: Diverse and well-preserved assemblages of dinosaur (theropod, sauropod and ornithopod), pterosaur, and bird tracks from the Hekou Formation (Lower Cretaceous) in the Yellow River (Huang He) valley represent the first significant fossil footprint discoveries in Gansu Province, China. However, the sites are large, visually spectacular, and well-exposed thanks to labor-intensive hand excavation. The sites have the potential for development as educational and tourist destinations. These sites have become one of the National Geoparks in China.
Dinosaur tracks include at least two theropod morphotypes that range in size from about 5 to >30 cm in length. Wide-gauge sauropod tracks (Brontopodus) range in size from 25 to 90 cm (pes length) and are the best-preserved examples known from China, with clear claw impressions. One trackway suggests an accelerating/running individual. Parallel ornithopod trackways indicate gregarious behavior. An enigmatic trackway may be a manus-only ornithopod trackway.
A pterosaur trackway (cf. Pteraichnus), the first reported from China, consists of 24 consecutive footprints, and is the longest, well-preserved trackway on record. Bird tracks (cf. Aquatilavipes) are also very well preserved.
The tracks occur at multiple stratigraphic levels in fluvio-lacustrine sequences of paleosol mudstones and sandstones with mud cracks and wave ripple marks. A minimum ichnodiversity of eight, the highest reported from the Cretaceous of China, is estimated. The saurischian component (theropods and sauropods) compares well with Inner Mongolia ichnofaunas from the Jing Chuan Formation. However, the co-occurrence of ornithopod and sauropod tracks is rare in Asia and globally, and compares with assemblages from South Korea, a similar Cretaceous paleolatitude (ca. 30°).

Addendum - Ensom and Delair Paper in 2008

See also Ensom and Delair (2008), a printed paper on the topic of the Portland Dinosaur Prints.