©Copyright 2018 GEOSCIENCE RESEARCH INSTITUTE
11060 Campus Street • Loma Linda, California 92350 • 909-558-4548
FOOTPRINTS IN THE GRAND CANYON
Leonard R. Brand
Chairman, Department of Biology
Loma Linda University
Related page | IN A FEW WORDS |
The Coconino Sandstone which is well exposed in the Grand Canyon has traditionally been interpreted as a wind-deposited sediment. In the layers of this formation there are tell-tale tracks that seem to say something else. Read on....
The Grand Canyon of the Colorado River is not only a spectacular slice of scenery, but it also poses many intriguing questions for the student of earth history. One of those questions concerns the way in which the Coconino Sandstone, a prominent rock formation in the Grand Canyon area, was deposited.
We will begin our discussion with a brief summary of the geologic history of the Grand Canyon, especially the Paleozoic rocks that form the majority of the canyon wall. At one time northern Arizona was a basin where layers of sediments were being deposited by wind and/or water. The floor of the basin was formed of the tilted and planed-off layers of Precambrian sediments. Many layers of sand, mud, and other sediment were deposited in the basin (Figure 1), with most of them being brought in by water. Some of these layers contained animals and plants. As other layers were being deposited, animals walked on them and left footprints which were preserved by the next layers. The sediments became cemented into rock, and the organisms and footprints became fossils. Then the Grand Canyon area was uplifted to form a plateau, and the canyon was carved into the layers of sediment by water. The cutting of the canyon exposed the sediments, making it possible to attempt to reconstruct the detailed depositional history of this area, by examining the characteristics of the sediments and their fossils.
Most of the Paleozoic sediments in the Grand Canyon are considered by geologists to have been deposited primarily by water (McKee 1966). An exception to this is the Coconino Sandstone. McKee (1933) studied the Coconino Sandstone and concluded that the sand forming the Coconino's slanted layers were deposited by wind, in the form of layer after layer of desert sand dunes. This theory would mean that a layer of water-deposited mud (the Hermit Shale) was followed by a layer of wind-deposited desert sand, and then the area was again covered by water and the Kaibab limestone was deposited.
The Coconino Sandstone is a homogeneous deposit of fine-grained, primarily quartz sand. It extends over much of northern Arizona, from the Mogollon Rim, northward to the Utah border. It is up to 1000 feet thick at its southern edge and thins to a few feet at its northern limits. It is a crossbedded deposit, with individual horizontal units (Figure 1) composed of many fine layers sloping at 20º-30º. The sloping layers of laminae are often 30 or 40 and occasionally up to 75 feet long (McKee 1933). Sloping, crossbedded layers such as these are generally thought to represent deposits formed on the down-current side of moving sand dunes either desert dunes or underwater dunes.
Scattered throughout most of the lower half of the Coconino Sandstone are numerous fossil footprints of vertebrate animals and less common trails of worms and arthropods (Gilmore 1927; Brady 1947). The vertebrate tracks have been referred to as amphibians and/or as reptiles, but from the structure of the tracks the majority of them are most easily interpreted as amphibians. No other fossils have been found in the Coconino Sandstone (McKee 1933). Out of the hundreds of trackways that have been observed, almost all of them are going up the slopes of the crossbedded layers (Gilmore 1927). The discovery of a downhill trackway in the DeChelly Sandstone (a similar Permian crossbedded sandstone in northern Arizona and southern Utah with the same type of vertebrate tracks) was considered significant enough to warrant a separate publication (Vaughn 1963).
The current explanation of the origin of the Coconino Sandstone was developed primarily by McKee. His initial study of the Coconino (McKee 1933, 1945) focused on the physical characteristics of the sandstone, and he concluded that it was an eolian, or wind-deposited, sand accumulation. Later he also studied the footprints of living vertebrates and compared them with the Coconino fossil footprints (McKee 1944, 1947). From this work he concluded that the fossil footprints were most likely formed in dry sand, thus supporting his earlier conclusion that the Coconino Sandstone was a desert deposit.
Identical or very similar fossil footprints also occur in several other crossbedded sandstone formations, and a number of authors have cited McKee's footprint studies in support of the idea that these other formations, as well as the Coconino Sandstone, were wind-deposited desert sands (Faul and Roberts 1951; Sarjeant 1975; Vaughn 1963; Walker and Harms 1972).
The implications of this work must be considered as we develop geologic flood models, and on the other hand perhaps our flood models can suggest new ways of looking at the Coconino Sandstone. A model or an idea is useful to science if it can suggest new lines of research that can be done successfully, and that improve our understanding of the subject we are investigating. Perhaps our flood model can suggest useful, new types of research that need to be done and that might not have been thought of by someone who did not believe in a flood of worldwide geologic significance.
The Coconino Sandstone is classified as a Permian deposit in the upper Paleozoic (Figure 1). As was mentioned earlier, the Paleozoic strata above and below the Coconino Sandstone are believed to have been deposited by water. Some of the flood models that are being developed propose that much of the Paleozoic sequence was deposited in the early part of the flood activities, and in these models the Coconino Sandstone would be a deposit laid down during the main part of the flood. Could there be a large-scale deposit of wind-blown sand in the middle of predominantly flood water-deposited strata?
The geologic data tell us that even though the flood was a rapid geologic event, it was nevertheless a very complex geologic event. We cannot arbitrarily rule out the possibility of a deposit of wind-blown sand forming during an interval of lowered water level, but our flood models do suggest that it may be very profitable to reinvestigate the Coconino Sandstone to see if there might be another explanation of its origin.
But haven't the existing data demonstrated that the Coconino Sandstone was wind deposited? How can we justify questioning this conclusion? Actually some of the criteria that were used to identify wind or water-deposited sand are not clear cut, and also more recent research has produced some interesting findings relevant to our topic. The Jurassic Navajo Sandstone is also a crossbedded sandstone, similar in many respects to the Coconino Sandstone. The Navajo Sandstone was previously considered to be an eolian, desert deposit, but more recently several authors have restudied the Navajo and have interpreted it as largely formed by shallow marine sand waves, with part of the formation deposited as coastal dunes formed by onshore winds (Dott and Batten 1971; Marzolf 1969; Stanley et al. 1971). These authors made the following comments about this change in interpretation:
Since 1903, most of the Navajo sands were assumed to represent ancient wind dunes formed on a vast Sahara-like desert; this became a ruling hypothesis.... The Navajo problem originated years ago when geologists could conceive of large amplitude cross stratification as originating only in wind-formed dunes; no other modern processes that could form it had been studied. This highlights the major short-coming of reasoning by analogy, namely the limitation at a given time of known possible analogues. Today, knowledge of modern shallow marine sedimentation has broadened the spectrum of counterparts of analogues. Insight gained into remarkably large underwater dunes found on very shallow shelf areas provides as attractive a comparison for much of the Navajo sands as for lower Paleozoic quartz sandstones (Dott and Batten 1971, p. 359).
Inasmuch as geologists are forced to interpret ancient sediments chiefly by analogies with modern phenomena, interpretations are severely biased if all possible modern analogues are not known; such was the case when the Navajo was first studied (Stanley at al. 1971).
Sedimentary features that were formerly thought to be diagnostic of eolian deposits are now known to be non-diagnostic. Stanley et al. (1971) pointed out that "grain frosting is no longer considered a criterion of wind transport," grain size distribution statistics have been ambiguous (for the Navajo), and "it can no longer be assumed a priori that large festoon cross strata prove an eolian dune origin for the Navajo or any similar sandstone because of the essential identity of form and scale of modern submarine dunes or sand waves, as documented during the past decade" (e.g., see d'Anglejan 1971; Harvey 1966; Jordan 1962; and Terwindt 1971).
The currently accepted interpretation for the deposition of the Coconino Sandstone was developed long before the above-mentioned work on submarine sand waves was available. My search of the literature has not yet revealed any significant recent studies on the Coconino Sandstone. Thus it does not seem unrealistic to propose other possible interpretations for the deposition of the Coconino Sandstone, and to carry out research to test those interpretations.
The research reported in this paper involved comparison of the Coconino fossil footprints with laboratory footprints of living animals made under a variety of conditions, to determine which conditions will produce footprints that are most similar to the fossil footprints (Brand 1977; in press).
Fossil footprints in the Coconino Sandstone were studied in Hermit Basin of the Grand Canyon. An investigation of the Coconino Sandstone located 82 vertebrate trackways along the Hermit trail. Each trackway was identified with a number; notes were taken on physical features of the tracks, such as the presence or absence of toe marks and impressions of the sole of the foot; and most trackways were photographed. At each in-situ trackway location the directional heading of the trackways and the slope angle of the bedding plane was noted. The study area was also surveyed and mapped.
Trackways of living amphibians and reptiles were studied in the laboratory. Sand slopes were formed in two experimental chambers, 8 feet long and 6 feet long respectively. The animals were allowed to walk up and down the slopes. Each of the 236 experimental trackways was photographed, and identified with a number, and notes were taken on condition of the sand, the slope of the sand surface, and physical features of the tracks.
Laboratory tracks were studied mostly on 25º slopes, with some observations on 15º and 20º slopes for comparison. Four experimental conditions were used: 1) dry sand (simulating a dry desert environment), 2) dry sand moistened with a fine spray of water (simulating desert sand moistened by dew or light rain), 3) wet sand, with standing water at the base of the slope (simulating sand near the water table), and 4) underwater sand.
Most of the laboratory trackways were made on sand collected near Mt. Carmel Junction, in southern Utah. This sand was apparently derived from the Navajo Sandstone that forms the surface topography in that area, and was used because of its similarity to the sand grains in the Coconino Sandstone.
Table 1 lists the animals used in the laboratory studies. The underwater locomotion behavior of 5 species of salamanders was observed in the laboratory, and one species was also observed in Tenaja Creek, in the Santa Rosa Mountains, Riverside County, California in March and April 1975. The amount of time spent swimming or walking on the bottom was recorded to the nearest second. In this study an animal walking on the bottom and also using swimming movements of the tail was defined as swimming. Each individual was observed for 70 seconds of locomotion time, or until it disappeared from view or stopped moving for a considerable length of time. Photographs the lengths and widths of the footprints were measured to the nearest 0.5 mm. The Mann-Whitney U Test (Siegel 1956) was used to analyze the data for significant differences.
FIGURES 2-7 and 10. Fossil footprints in uphill trackway in the Coconino Sandstone in Hermit Basin. All numbers and letters in all photographs are 4.5 mm high.
Figure 10 FIGURES 8 and 9. Fossil footprints in the Coconino Sandstone going across the slope. Figure 9 is an enlargement of part of Figure 8.
The fossil trackways (Figures 2-10) were distributed through the lower half of the Coconino Sandstone (Figure 1). Within the track-bearing section, trackways occurred on a large number of the exposed surfaces. As reported by Gilmore (1927) almost all of the trackways were going up the slopes of the crossbedded strata towards the northeast. One indistinct trackway appeared to be going downslope.
Previous work on footprints in the Coconino Sandstone has usually been taxonomic comparison, but the present study emphasizes analysis of the physical characteristics of the entire population of trackways. Some of the Hermit Basin tracks were well defined with good impressions and toe marks (Figure 2), but the majority of the tracks were not as well defined or as complete. In some trackways the individual footprints did not have toe marks or other details (Figures 3-4), and in other cases the sole impressions were missing or were incomplete (Figures 5-7). Even though most of the footprints were not as complete as those in Figure 2, the majority of them did have evident toe marks and sole impressions (Figure 11). Tracks with sole impressions sometimes had small ridges of sand pushed up behind them, but these never extended back into the tracks behind them.
LABORATORY UPHILL TRACKS
Dry sand tracks (Figures 12-15) usually consisted of a series of depressions with no toe marks or other details (Figures 12-13). There was usually a ridge of sand pushed up behind each footprint, and often the sand in these ridges flowed back into the previous footprint, obscuring any details that may have been there. A few salamander trackways had toe marks at the back of each print, which were made as the animal's foot was lifted out of the print. Also a few tracks had marks that were made by the toes being dragged across the sand from one foot position to the next (Figures 14-15). All of these were counted as toe marks in Figure 11.
Damp sand trackways always had definite foot impressions distinct from each other, but toe marks were rarely present (Figures 16-17; Figure 11). The dampened surface formed a crust of sand that broke up into many pieces when the animals walked over it. The pieces sometimes pushed up into a pile at the back of the footprint, and in other trackways they were scattered on the surface of the sand. If the damp crust of sand was thick enough so that the weight of the animal did not break it up, the only tracks produced were a series of small dimples left by the toes.
These dry sand and damp sand laboratory tracks differed from the Coconino Sandstone fossil tracks in several important features. Dry sand and damp sand tracks rarely had toe marks or other details, while the fossil tracks usually had definite toe marks. Dry sand tracks also had large ridges of sand behind them, which often flowed back all the way into the previous footprint, whereas the fossil tracks did not have very prominent ridges behind them. Jumbled pieces of damp sand crust around the damp sand tracks were never observed in the fossil tracks. Also, the proportions of the fossil tracks were quite different from the dry sand tracks. The dry sand tracks were longer than their width, but most fossil tracks were short in relation to their width (Figure 28). The outlines of most of the damp sand tracks were too poorly defined to allow precise measurements to be made. Consequently the damp sand and dry sand tracks observed in this study do not seem to provide an adequate model for the origin of the fossil tracks.
Tracks in wet sand, above water, were quite variable (Figures 18-21). The water seeps down through the sand quite rapidly, producing a gradient from lowest water content at the top of the slope to highest water content at the base. Footprints at the base of the slope were poorly defined. A little above the water level the tracks were variable, and some had clear toe marks and sole impressions. Higher on the slope, on more firm sand, the tracks consisted of toe marks only (Figures 19-21).
FIGURES 19 and 20. Same as Figure 18, only using the lizard Dipsosaurus.
Many of the trackways on wet sand contained some footprints that were closely similar to the fossil tracks, but in several respects the wet sand trackways were consistently different from the fossil tracks. The wet sand trackways almost always made a marked transition from well-defined prints to toe marks only or almost no prints at all as they ascended the slope, as in Figure 18. This feature was not seen in the fossil tracks, even though some fossil trackways were several feet long. Laboratory wet sand trackways that were made some distance above the water table, where the wet sand was more firm, consisted of small scratches or other marks from the individual toes and were quite different from most of the fossil tracks.
McKee's (1947) photographs of dry sand, damp sand and wet sand trackways look very similar to my results. From his experimental results under those conditions and a personal communication from the paleontologist Peabody indicating that salamanders do not make tracks underwater, McKee (1947) concluded that the fossil tracks were most similar to the dry sand trackways, because only in dry sand were there definite prints of individual feet.
Peabody (1959) stated that salamanders usually swim from place to place rather than walk on the bottom; that when they do walk they are partially buoyed up by the water and do not leave footprints. There is no indication of how extensive his observations on this phenomenon were. My results were quite different from his. All five species used in my study walked on the bottom more of the time than they swam in the water (Table 2). In a laboratory tank, many of the salamanders swam vigorously along the surface, against the side of the tank, and tried to climb out. If the tank arrangement provided a resting place such as a sand bar or some other object, the salamanders' behavior was more like that observed in the field. They would commonly rest on this object, before swimming around under the water or walking on the bottom.
The substrate at my field study site was not suitable to produce footprints, but in the laboratory all five species produced tracks on the sand underwater (Figures 22-27). These trackways were composed of distinct footprints, which usually had toe marks, and sometimes had sole impressions also (Figure 11). In some cases the footprints had small ridges of sand pushed up behind them, but these ridges never extended back into the previous print.
FIGURES 22-24 and 26-27. Uphill underwater laboratory track on a 25º sand slope using the salamander Taricha.
Of all the laboratory trackways produced, the underwater tracks were most similar to the fossil tracks. Underwater trackways had toe marks as often as the fossil tracks, and they were uniform in appearance the full length of the sand slope, as the fossil tracks are. Also, the proportions of the fossil tracks were most similar to that of the underwater tracks.
Figure 28 compares the ratios of length to width of the fossil tracks and laboratory tracks. While dry sand tracks were longer than their width, fossil tracks, underwater tracks, and wet sand tracks were short in relation to their width. Wet sand tracks are often short because they are only toe marks. The underwater tracks tend to have sole impressions that are short in comparison to their width. This is because the animals are partially buoyed up by water, and they often push against the sand with their feet almost at right angles to the surface, rather than placing their feet flat on the surface. This produces tracks that usually have only toe marks or toe marks with a shortened sole impression features that are also found in many of the fossil tracks. Statistical evaluation of the data shows that the difference between the fossil tracks (complete tracks) and the dry sand tracks was highly significant (Z=5.89; p<0.00001), but the fossil tracks and the underwater tracks were not significantly different (Z=.07; p=0.47).
Conspicuous tail drags were found in 40% of the laboratory trackways, but very few of the Coconino fossil trackways have tail drags. The only laboratory tracks that rarely showed tail marks were underwater trackways of the hellbender, Cryptobranchus alleganiensis. The hellbender had a much shorter tail than the other animals that I used, and it usually did not drag its tail on the sand enough to leave noticeable tail marks. Some of the fossil amphibians from Permian deposits were heavy-bodied, short-tailed animals. If these animals made the Coconino Sandstone tracks, this may explain why they rarely have tail marks.
It has been suggested by McKee that the near-absence of downhill trackways resulted from the animals' tendency to slide downhill, causing their tracks to be obliterated by sliding sand. This does not seem to be an adequate explanation. Downhill as well as uphill trackways were produced under all four laboratory experimental conditions used (Figures 29-31). On underwater sand, wet sand, and damp sand almost all downhill trials produced easily recognized trackways. On dry sand, salamander downhill trackways were usually reasonably well defined, and lizards produced distinct downhill trackways when they moved at a walking pace or a slow run. If they were urged into running very fast then their tracks were almost unrecognizable.
Thus the downhill laboratory trackways were often not quite as well defined as the uphill trackways, but the majority of the downhill trackways, in all of the experimental conditions, were more distinct than many of the fossil tracks; so an adequate explanation for their near-absence from the sandstone needs to be found. If the fossil tracks were produced under water, the preponderance of uphill trackways might be the result of some behavioral characteristic of the animals. For instance, they may have been swimming when going with the water current but would drop down and walk on the bottom when moving against the current. Behavioral traits of extinct animals cannot be tested, but this example illustrates that behavior can affect the tracks under water in ways that are not possible above water.
Several trackways were headed directly across the slope or at an angle across the slope (Figures 2, 8, 9), but with the toe marks of both back and front feet pointed upslope. These trackways can perhaps be best explained by animals being pushed by a water current moving at an angle to the direction of movement of the animal.
The data presented by McKee (1947) have been used by him and others (Faul and Roberts 1951; Vaughn 1963; Walker and Harms 1972; Sarjeant 1975) as evidence that the Coconino Sandstone and some other crossbedded sandstones were deposited in a desert environment.
The data presented in this paper indicate that fossil footprints of the type found in the Coconino Sandstone should not be used as evidence for eolian wind-blown deposition of dry sand. If the Coconino Sandstone was indeed dry when deposited, then several important features of its fossil footprints remain unexplained.
The footprints alone cannot provide the answer as to whether the Coconino Sandstone was water or wind deposited, but we can say that the tracks now point more in the direction of water deposition. The tracks suggest that it may be profitable for a sedimentologist to restudy the other characteristics of the Coconino also, in light of current knowledge about the deposition of crossbedded sandstones, to see if the data will indeed indicate that the Coconino was deposited by water.
- Brady, L.F. 1947. invertebrate tracks from the Coconino Sandstone of northern Arizona. Journal of Paleontology 21(5):466-472.
- Brand, L.R. 1977. Coconino Sandstone (Permian) fossil vertebrate footprints paleoecologic implications (abs.), American Association of Petroleum Geologists and Society of Economic Paleontologists and Mineralogists, Program and Abstracts, pp. 66-67, annual meeting, Washington, D.C.
- Brand, L.R. In press. Field and laboratory studies on the Coconino Sandstone (Permian) fossil vertebrate footprints and their paleoecological implications. Palaeogeography, Palaeoclimatology, Palaeoecology.
- d'Anglejan, B.F. 1971. Submarine sand dunes in the St. Lawrence estuary. Canadian Journal of Earth Sciences 8(11):1480-1486.
- Dott, R.H. and R.L. Batten. 1971. Evolution of the earth. McGraw-Hill Book Company, New York.
- Faul, H. and W.A. Roberts. 1951. New fossil footprints from the Navajo (?) Sandstone of Colorado. Journal of Paleontology 25:266-274.
- Gilmore, C.W. 1927. Fossil footprints from the Grand Canyon: second contribution. Smithsonian Miscellaneous Collection 80(3):1-78.
- Harvey, J.G. 1966. Large sand waves in the Irish Sea. Marine Geology 4:49-55.
- Jordan, G.F. 1962. Large submarine sand waves. Science 136:839-848.
- Marzolf, J.E. 1969. Regional stratigraphic variations in primary features of the Navajo Sandstone, Utah (abs.). Geological Society of America Abstracts with Programs, 1969, part 5:50-51.
- McKee, E.D. 1933. The Coconino Sandstone its history and origin. Carnegie Institution of Washington, Contributions to Paleontology, Publication No. 440:77-115.
- McKee, E.D. 1944. Tracks that go uphill. Plateau 16(4):61-73.
- McKee, E.D. 1945. Small-scale structures in the Coconino Sandstone of northern Arizona. Journal of Geology 53(5):313-325.
- McKee, E.D. 1947. Experiments on the development of tracks in fine cross-bedded sand. Journal of Sedimentary Petrology 17:23-28.
- McKee, E.D. 1966. Ancient landscapes of the Grand Canyon region. Northland Press, Flagstaff, Arizona.
- Peabody, F.E. 1959. Trackways of living and fossil salamanders. University of California Publications in Zoology 63:1-48.
- Sarjeant, W.A.S. 1975. Fossil tracks and impressions of vertebrates. In R.W. Frey, ed. The Study of Trace Fossils, pp. 283-324. Springer-Verlag, New York.
- Siegel, S. 1956. Nonparametric statistics. McGraw-Hill Book Company, New York.
- Stanley, K.O., W.M. Jordan, and R.H. Dott. 1971. New hypothesis of early Jurassic paleogeography and sediment dispersal for western United States. American Association of Petroleum Geologists Bulletin 55:10-19.
- Terwindt, J.H.J. 1971. Sand waves in the Southern Bight of the North Sea. Marine Geology 10:51-67.
- Vaughn, P.P. 1963. A downslope trackway in the DeChelly Sandstone, Permian of Monument Valley. Plateau 36:25-28.
- Walker, T.R. and J.C. Harms. 1972. Eolian origin of flagstone beds, Lyons Sandstone (Permian), type area, Boulder County, Colorado. The Mountain Geologist 9:279-288.