INTRODUCTION

Minor nontectonic anticlinal structures known as "tepees" because of their resemblance to Indian tents (Adams and Frenzel, 1950) are a characteristic and much discussed feature of many Upper Permian shelf carbonate rocks in the eastern Guadalupe Mountains of New Mexico and adjoining parts of Texas. They are present in each of the shelf formations of the Artesia Group, but those examined by the writer were mainly in beds assigned to the Seven Rivers, Yates, and Tansill Formations. There is every reason to believe, however, that conclusions concerning these tepees are equally applicable to those in older rocks of the Artesia Group. Field work was carried out mainly in 1970, when the writer was employed by the New Mexico State Bureau of Mines and Mineral Resources, but this paper incorpo ates the results of preliminary research carried out in 1969 in collaboration with Charles L. Jones of the U.S. Geological Survey, Menlo Park, California, whose help is gratefully acknowledged.

DISTRIBUTION AND CHARACTERISTICS OF GUADALUPIAN TEPEES

Tepees in Upper Permian shelf sedimentary rocks of the eastern part of the Guadalupe Mountains are present in a diachronous belt up to 3 mi wide, but they are most abundant and best developed between 150 and 2,000 yd shelfward of the northwestern edge of the Capitan Limestone. On the southeast, tepees decline rapidly in both number and complexity, but scattered minor tepees persist to within 30 yd of the Capitan Limestone in Dark Canyon and Junigan Draw. Similar but less abrupt decline in their number and complexity is a feature of the northwestern margin of the belt. There appears to be a broad coincidence between the distribution of tepees and that of caliche fabrics in shelf carbonate rocks.

Tepees are found in carbonate beds ranging from fine dolarenites to dolorudites, although they appear to be most common in, and in many areas are virtually restricted to, sedimentary rocks inferred from their lamination to have been algal-bound. They commonly include fenestral sedimentary rocks in which the major axes of the fenestrae now are inclined, and although the rock of many tepees shows evidence of intense pisolitic or nonpisolitic calichification, many and perhaps most tepees are in noncalichified sediment. Few of the tepees seen by the writer were in cross-laminated grainstones, and none are present within interbedded siltstones and sandstones. In large sections they are concentrated at certain horizons, at any one of which in a given area they tend to be of roughly the same size, pattern, and complexity. Careful tracing of marker horizons and comparison of measured sections within the tepee belt reveal, however, that even at these horizons tepees may be developed only patchily.

Examination of tepees in well-exposed sections in Dark, Walnut, Rattlesnake, and Slaughter Canyons and in intervening smaller

End_Page 63------------------------------

valleys revealed that they range from simple structures a few inches in height and width to extremely complex mounds up to 20 ft high and 20 yards across. A typical small tepee (Fig. 1) lies on a fairly flat base, and in transverse section has a characteristic roughly symmetrical chevronlike structure with bedding inclined upward toward a subvertical axial fracture which in many cases is filled partly or wholly with fine sandstone or lamellar or fibrous carbonate rock. There is generally some interdigitation of truncated beds from opposite sides of the axial fracture. Dips within such tepees generally are less than 50°, but in some, particularly high on the flanks, beds approach vertical or even are overturned slightly (Figs. 2, 3). Minor fracturing and distention of beds in tepe s are prevalent, and are greatest where dips are high. Except where they have been modified by later compaction, strata near the crests of tepees commonly are arched, and this tendency is most marked where there is no continuous axial fracture (Fig. 2). A few are markedly asymmetric, including some in which tilted-up beds rest against the broken edges of undisturbed strata. Tepees are linear features, lacking preferred orientation (Newell et al., 1953, p. 127). The writer was unable to confirm Kendall's observation (1969) that they are polygonal in plan, but from their number, spacing, and apparently random orientation in many large sections, such a configuration appears likely. The essential contemporaneity of tepees, indicated by wedging of associated sediments against their inclined f anks (Fig. 2), is not in dispute.

Large tepees invariably are more complex than small ones, and brecciation in them generally is more pronounced. From the interrelations of component beds, it is clear that the larger examples had a long and varied history, and that phases of growth of the structure and accompanying brecciation were separated by phases of further sediment deposition and cementation and by periods of contemporaneous erosion. Contrary to the views of Newell et al. (1953, p. 127), the writer has found that most tepees underwent considerable erosion after formation, and in many their crests almost were planed off (for illustrations see Boyd,

Fig. 1. Detail of simple tepee, showing empty or sand-filled axial fracture and wedging out of some beds against flanks. Tansill Formation, south side of Dark Canyon, about 1,200 yd from mouth; Sec. 24, T23S, R25E.

End_Page 64------------------------------

1958, pl. 4A; Jacka et al., 1969, pl. 19D; Kendall, 1969, pl. 4A).

The relief of individual beds within simple tepees rarely exceeds 6 ft and is generally 1-3 ft; most taller structures which at first glance appear to be simple tepees prove, on closer examination, to consist of several small tepees which have developed above each other but are separated by erosion surfaces (Fig. 4). Tepees rarely developed on the same site as earlier tepees whose crests had been eroded flat and which therefore did not form a minor topographic ridge. Many such crests survived erosion, however, and by their presence determined the site upon which succeeding tepees with the same orientation were able to develop. In this way were formed cedar-tree-like structures spanning more than 30 ft of strata, excellent examples of which are displayed on the north wall of Walnut Can on in Sec. 30, T25S, R24E. By their shape these are readily distinguishable from large mound-shaped tepee complexes, which are most common in central parts of the tepee belt and result from the superposition of many closely spaced tepees at short vertical intervals. The apparent complexity is, of course, partly a function of the orientation of the plane of section, oblique and strike sections invariably appearing more complex than transverse sections (see Fig. 5).

Although central parts of all tepees share the distinctive splayed-out upturned bedding, there is great variation in the proportion of internal fill. This is least abundant near the seaward margin of the tepee belt, where axial voids are common, although the possibility that at one time these contained water or soluble evaporite cannot be excluded. In most of the tepee belt, areas between disrupted beds in central parts of tepees are filled with fine-grained, pale buff silty sandstone, with laminar, massive, and fibrous carbonate, or with combinations of these; in extreme examples, fillings of these materials reach 30 percent of the area of section (Fig. 6). In transverse sections of tepees, void fillings extend outward from the central fracture in a characteristic pattern (Fig. 7), b t in oblique sections and in the large complex mounds they follow highly irregular anastomosing courses between beds and joint blocks and locally extend

Click to view image in GIF format. Fig. 2. [Grey Scale] Tall narrow tepee in partly calichified fenestral bioclastic dolarenite. Note arched crest and absence of continuous axial fracture. Tansill Formation; locality same as Figure 1.

End_Page 65------------------------------

for many yards (Fig. 5). In some places sandstone-filled feeder channels may be traced directly downward to one of the interbedded sandstones, although in many other places sand and carbonate grains appear to have been washed into eroded tepees from above. In places where the axial fractures of superimposed tepees coincide, sandstone or fibrous carbonate locally is continuous between them. The laminar carbonate within tepees commonly is present in forms that would be regarded as typical of L-L-H algal-stromatolites were it not that many of the botryoidal surfaces are convex downward. Much of the laminar carbonate is roughly bilaterally symmetrical; it clearly lined and ultimately filled cavities.

ORIGIN AND ENVIRONMENTAL SIGNIFICANCE OF TEPEES

Theories on the origin of tepees were summarized and discussed by Newell et al. (1953) and Kendall (1969). The former authors suggested that tepees were produced by the expansion of crystallizing salts, whereas Kendall concluded that they probably resulted from either desiccation or fibrous carbonate precipitation, or possibly from a combination of these two. Jacka et al. (1969, p. 167) and Dunham (1972, p. 11-24) implied that tepees might be fossil caliche anticlines, a view independently considered and rejected by Kendall (1969) on the grounds that such anticlines contain no voids. This criterion probably is not conclusive, but Kendall could have added that few caliche anticlines have continuous axial-plane fractures, and that even those which most resemble tepees in section exhibit many points of difference in detail. In any event, large-scale anticlinal structures are not typical of modern caliche profiles (Reeves, 1970, p. 359). The writer believes that none of the previous interpretations satisfactorily explains all the features of tepees, and that a further alternative explanation must be sought.

In rejecting Kendall's interpretation, the

Fig. 3. Detail of simple tepee, showing empty or sand-filled axial fracture, oversteepened margins and wedging out of some beds against flanks. Tansill Formation, north side of Dark Canyon, about 1,800 yd from mouth; Sec. 23, T23S, R25E.

Fig. 4. Superposed simple tepees separated by erosion surfaces. Tansill Formation; locality same as Figure 3.

End_Page 66------------------------------

writer is guided by the fact that in desiccation polygons the surface layers of sediment contract, breaking away from surrounding sediment along vertical cracks that are widest at the top and taper downward. Curling of the edges of polygons further widens the crack, so that if tepees were the edges of fossil-desiccation polygons seen in section, then the axial fissure should be vertical or nearly so and should widen upward. Hundreds of tepees were examined by the writer, who found few with strictly vertical planar axial fissures, and none with fissures appreciably wider at the top than near the bottom.

Furthermore, contraction of the beds by desiccation would mean that, if the structure subsequently were pressed down again by overburden pressure, opposite sides of the polygonal cracks would fail to meet. Even cursory examination of many tepees (see Figs. 1, 3 for example) shows that if the upturned beds were forced back into position, they would in fact overlap (see also Boyd, 1958, p. 51), and this relation, whether resulting from flattening or from other causes such as direct impaction, is evident in many tepees (see Fig. 8).

Finally, in most fossil-desiccation cracks, beds affected are steepest near the edge of the crack, whereas in many tepees beds near the crest are arched over near the axial fracture; exceptionally, even an axial channel is absent, a notably rare situation in desiccation cracks of any age. The writer therefore concludes that, far from being desiccation cracks, tepees display clear evidence of having been caused by expansion, as Adams and Frenzel (1950) first suggested and Dunham (1972) reiterated. They are, if this interpretation is correct, polygonal pressure ridges. Similar but smaller ridges are present on parts of the sabkhas margining the Persian Gulf (Evans et al., 1969, Fig. 2) and are indeed said (P. Bush and R. Park, personal commun.) to result from expansion of the sediment i volved. Furthermore, very large polygonal expansion ridges with overthrust relations at fractured margins are known on parts of the floor of the Persian Gulf (Shinn, 1969). Smaller tepeelike structures ("megapolygons") also have been described from Triassic limestones of northern Italy by Assereto and Kendall (1971) who ascribed their formation to either desiccation-contraction or cementation-expansion

Fig. 5. Large complex tepee in oblique section. Void fill of very fine sandstone is directly traceable to Yates Sandstone about 3 ft below section illustrated. Tansill Formation, south side of Dark Canyon, about 1,100 yd from mouth; Sec. 24, T23S, R25E.

End_Page 67------------------------------

or to a combination of both these processes.

In the case of the Guadalupian tepees, there seems no good reason why the formation of cracks by desiccation should have played a substantial part in their formation, for a polygonal pattern is produced as readily by uniform expansion as by uniform contraction. Interpretation of the tepees as fossil-pressure ridges also is compatible with their common superposition, for whereas desiccation cracks developing from the surface downward are unlikely to have coincided with completely buried eminences, the presence of subjacent minor ridges obviously would have influenced the position of ridges formed by expansion of overlying sediments.

The configuration of beds within tepees shows that only about 6 in. of carbonate sediment was necessary to initiate a tepee, and marginal thickness changes show that the tepee continued to be raised as new sediment was added. Lithification of some of the beds appears to have been well advanced during final stages of tepee growth, for although beds were able to bend through angles of 20° to 30°, they were broken if subjected to further bending stresses. For this reason, beds tended to fracture near the apex of the tepee where the sharpest bending occurred, giving rise to the characteristically circuitous axial fracture. Once formed, this fracture would have provided a route by which previously trapped formation fluids could escape and also would have allowed free downward pas age of water at times when the standing water level declined.

Sand and silt infilling of voids in tepees probably was accomplished by sediment carried in suspension in fluid moving up and down along these and other fractures, and it probably is significant that void-filling sediment is least common in areas nearest the Capitan Limestone where interbedded sands and silts are thin and fairly uncommon. Fibrous and laminar carbonate doubtless was formed by multiphase precipitation from aqueous solution on the walls of voids not filled with sand, although the presence of sand and silt lenses within such carbonate (Kendall, 1969) shows that loose sediment

Click to view image in GIF format. Fig. 6. [Grey Scale] Detail of fracture pattern and sandstone void fill of simple tepee in calichified, algal-laminated, fenestral fine dolarenite. Dark Canyon, south side, about 2,300 yd from mouth; junction of Sections 23 and 26, T23S, R25E.

End_Page 68------------------------------

was still available before carbonate infilling was complete. In many tepees abundant voids remain.

The filling of voids by laminar, fibrous, and micritic (massive) carbonate and by terrigenous sand and silt may have taken place during subaerial calichification, the most intense zone of which corresponds roughly to the tepee belt, and may in some areas have led to further minor development of the structure of the tepees. Thus, although the writer differs from Newell et al. (1953), Jacka et al. (1969), and Kendall (1969) in interpretation of the principal cause of tepee formation, there is a measure of agreement on possible contributory causes.

A further possible contributory factor may have been water under pressure, restricted within loaded but still unconsolidated interbedded sands by continuous algal mats in semilithified overlying carbonates. Such a mechanism was hinted at by Boyd (1958). Escape of such water, once concordant and discordant planes of weakness had been formed in overlying carbonates, at times might have been forceful or even explosive, but had this been a major factor in the formation of tepees, beds disrupted at the time surely would have been capable of returning to, or nearly to, their original position after pressure release was complete. The sandstone and siltstone beds probably also would show some evidence of movement and of thickness variation beneath tepees if really significant sediment flowage and depletion had occurred. Nevertheless, it is clearly possible that the sands were at one time in a quick condition, and that if this were so, some beds of sand might have been expelled totally, and others might have become homogenized and thus lost primary sedimentary structures, thickness variations, and evidence of flowage.

The writer believes that the most likely cause of expansion in sediment featuring tepees was the force exerted by the contemporaneous

Fig. 7. Detail of sandstone void fill in simple tepee. Tansill Formation; locality same as Figure 5.

Click to view image in GIF format. Fig. 8. [Grey Scale] Impacted core of simple tepee in fenestral bioclastic dolarenite. Tansill Formation; locality same as Figure 3.

End_Page 69------------------------------

growth of crystals in sediment partly trapped within algal mats, and that expansion was facilitated by the lack of adhesion between the mat-bound sediments and interbedded wet terrigenous sediments. In the modern polygons of the Abu Dhabi sabkha, the raised edges of expansion polygons are said (P. Bush, personal commun.) to collapse following rainfall, as a result of the solution of their halite cement; in this case, the interstitial growth of halite presumably provides the expansive force. There is no evidence that halite could have fulfilled this role in the sediments of the Carlsbad subgroup, however, it is probable that the growth of aragonite or, less likely, gypsum provided the force required. The close association of tepees with fenestral mat-bound sediments showing abundant ev dence of repeated subaerial exposure strongly argues against wholly subaqueous formation, and the writer therefore favors intertidal or supratidal formation of the tepees rather than the subaqueous environment of the polygonal pressure ridges described by Shinn (1969).

If the writer's view of the origin of the tepees is correct, the environmental implications of their presence are relatively clear. They began to form in the high subtidal to low supratidal zone, after fenestrae had formed in the mainly algal-bound, sand-grade carbonate sediment and when the sediment was at least partly lithified, as intermittent deposition continued, and mostly before calichification and erosion took place. Their distribution suggests that at any one time they were formed patchily in parts of a belt a few hundred yards wide on a flat or gently sloping platform within a few feet of mean sea level; that from time to time sea level fell relative to the platform, and tepee-bearing sediments were exposed subaerially for lengthy periods. Few other conclusions may be drawn, except that the depositional surface probably sloped gently toward the basin rim in outer parts of the belt of tepees. The occasional presence of tepees associated with caliche within 30 yd of the Capitan Limestone shows that the environment favoring their formation constantly shifted backward and forward across the platform with the varying interplay of sediment accretion, subsidence, and relative sea-level changes. Their association with caliche suggests a contemporary semiarid climate with seasonal rainfall.