The Afneriun Asscciation of Petroleum Geologists Bulletin V. 56, No, 12 (December 1972), P.

2303-2322,13 Ras,

Late Paleozoic Sedimentation in West Texas Permian Basin'

JOHN M. HILLS' El Paso, Texas 79999 Abstract Rocks of Permian age have been known in West Texas for over a century, and in the last 40 years the drilling of numerous test holes has made subsurface information available concerning these rocks. These data have enabled geologists to correlate beds of late Paleozoic age across the basin, a distance of more than 200 mi. Results of these correlations are presented in a series of eight paleogeographic maps. In early Paleozoic time this region formed a broad, generally shallow basin. In the middle Paleozoic the basin slowly filled with carbonate and siliceous rocks. In Mississippian time a positive median belt developed in a north-south direction, separating the larger basin into the eastern or Midland basin and the western or Delaware basin. In Pennsylvanian time wide carbonate shelves developed around these basins, especially the eastern one. At the south end of the basins tectonic activity increased and foredeeps developed and were filled with flysch material. In the Early Permian, seas spread over much of the basin region, depositing shale in the low parts of the sea floor while limestones accumulated on the higher parts. However, by Middle Permian (Leonardian) time, marine circulation was restricted and evaporites began to form. This restriction became increasingly severe during the remainder of the period. Thus carbonate reefs and banks formed on the margins of the Delaware and Midland basins, as well as on the Central Basin carbonate platform, overlying the old median mountain range. In the latter part of this period, high limestone reefs were formed on the basin edges, separating basinal sediments from evaporltic and clastic lagoonal strata. The Capitan reef surrounding the Delaware basin Is a notable example. All these formations show traces of cyclical deposition which may be attributed to eustatic changes in sea level, perhaps glacial in origin. During the closing epoch of the period, thick sequences of anhydrite, halite, and potash salts accumulated In the basin areas. After a final marine transgression, continental redbeds covered the region and Permian sedimentation ended.

I^m^oDuc^oN The first recognition of the Permian rocks in the West Texas area was by Jules Marcou (1855) who identified fossils of this age collected on Captain John Pope's expedition of 1853. In 1855 Pope returned to the southern end of the Guadalupe Mountains with Dr. George G. Shumard as "surgeon and geologist." Shumard described the rocks of the Guadalupe Mountains and collected fossils, which his brother, B. F. Shumard (1858) identified as Permian. Little more was learned of these rocks until Girty's visit to the base of Guadalupe Point in 1901, where he made the extensive collections that formed the basis for his classic paper, "The Guadalupian Fauna" (Girty, 1908).

In the meantime, Permian rocks had been discovered in the Mid-Continent area, and during World War I and immediately afterward petroleum geologists traced them on the outcrop from Kansas and Oklahoma into north-central Texas. As the rocks of this region seemed to be entirely different from those cropping out in extreme western Texas and south-central New Mexico, both in lithology and fauna, little more could be said about them than that both were Permian (Beede, 1910). Furthermore, changes in facies, especially on the western side of the basin, made correlation even over short distances very difficult. Crandall's (1929) discovery of the remarkable Upper Permian section in Rocky Arroyo north of Carlsbad, New Mexico, although at first misinterpreted as containing an unconformity, was a step forward in understanding the relation of the evaporite sections to the carbonate rocks. Later Bates (1942) pointed out the gradation in this section from bedded dolomite of the back reef facies to the gypsum and red elastics of the saline lagoon. Lloyd (1929), then associated with the European geologists of the Shell group, realized that the massive limestones forming Guadalupe Point were parts of a great reef complex similar to that previously recognized in the Triassic rocks of the Tyrolean Dolomites. Also during the late 1920s Sidney Powers discovered a late Permian turritelUd in a core of thick-bedded dolomite from the Hendrick field (Hills, 1940). This indicated that perhaps the same upper Permian beds found on the surface also occurred in the dolomite reef facies of the subsurface. Further drilling confirmed this and also showed that carbonates in
'Manuscript received, November 7, 1971; revised and accepted, March 17, 1972. ^University of Texas at El Paso. My colleagues, William S. Strain and Karl W. Klement, have read this paper critically and made many suggestions. Many other geologists have, in innumerable discussions over the years, contributed to my understanding of the late Paleozoic in this region. To all these I am grateful. I especially acknowledge the suggestions made by the late John Emery Adams in the early stages of preparation of this paper. Sally W. Hills did the prehminary editing; Jerry King and Martha Thomas typed the manuscript.
1972. The American Association of Petroleum Geologists. All rights reserved.

2303

2304

John M. Hills
camp, Leonard, Guadalupe, Ochoa. This proposal was favorably received and the names now are generally used. Thus by 1942 the essential framework of our knowledge of the Permian basin was complete, as far as the upper part of the Permian system is concerned. However, many details of correlation and sedimentary facies remained obscure. Some of these have been clarified by later work and some remain uncertain today. Very little was known in 1942 about the geology of the Lower Permian and almost nothing of the Pennsylvanian and older Paleozoic rocks in the central part of the basin, although they had been studied to some extent on the outcrop along the eastern and western edges. Classifying the geology of the older rocks was to occupy the petroleum geologists of the basin for the next 20 years.
FRAMEWORK OF PERMIAN BASIN

the subsurface graded into redbeds and then into evaporites, much as they do in Rocky Arroyo. By 1930, Permian subsurface correlations were sufficiently advanced so that Cartwright was able to draw an east-west cross section through the southern part of the Permian basin. This work showed that the fine clastic materials of the Delaware basin were facies of the massive Capitan carbonate reef complex, as well as of the dolomites of the Central Basin platform. The similar, but older, elastics of the Midland basin also were facies of the Eastern platform carbonates which graded eastward into the red shale and gypsum of the long-known north-central Texas Permian. From 1934 to 1938 P. B. King studied the outcrop of the Permian rocks in the Guadalupe Mountains and adjacent Diablo platform, and published a paper (King, 1942) in which he combined surface information with subsurface data to construct a comprehensive overview of the Permian geology of the region. More details were given in a professional paper, the publication of which was delayed by World War II (King, 1948). In 1939 Adams et al, using both surface and subsurface information, proposed that the Permian of the West Texas region be divided into the following series in ascending order: Wolf-

Okla. N. Mexico
1 Permian \

N, \ Basin J

Chih,

^v_^^

Texas

MEXICO /^ \ 4 j Coah*\^
FIG.

Index map showing location of Permian basin region.

In early Paleozoic time, the region now known as the Permian basin (Figs. 1, 2) formed a broad, generally shallow basin which merged with the Ouachita-Marathon geosynchne in the south and shoaled on the north in south-central New Mexico and the Texas Panhandle. This area has been named Tobosa basin by Galley (1958). During most of the early and middle Paleozoic, deposition consisted largely of shallow-water shelf-type, carbonates, but was interrupted by shale sedimentation during the Middle Ordovician, Late Devonian, and Early Mississippian. The stratigraphic section from the Precambrian to the base of the Mississippian is thin (about 6,000 ft) over most of the region. Rocks representing many parts of this interval are missing from the stratigraphic column. As there is little angular discordance at the hiatuses, it is difficult to determine whether the missing beds are absent because of erosion or nondeposition. These rocks contain many zones of chert, novacuhte, and other siliceous rocks. These zones possibly represent long intervals of extremely slow deposition of silica from solution and suspension (Wilson and Majewske, 1960). These time intervals may fit into the hiatuses and reduce the length of the intervals considerably. During the early Paleozoic there seemed to have been no well-marked platforms within the basin. However, lines of weakness along strikeslip faults in the basement probably were present (Hills, 1970). Along these faults later vertical movement took place. In Early Mississippian time a different paleogeographic regime began. The ancient Tobosa basin began deepening on

Late Paleozoic Sedimentation in West Texas Permian Basin

2305

FIG. 2Map of Permian basin showing localities and features referred to in text.

2306

John M. Hills

INTER GLASS W.TEXAS NATIONAL SERIES MOUNTAINS

CAPITAN LS.

i<^^"^
-?-?-

s.s.a SH,

GUADALUPE DELAWARE NORTHWEST MIDLAND W.CENTRAL I OKLAHOMA I FIG. MOUNTAINS BASIN SHELFaC.B.P. BASIN TEXAS so. KANSAS DEWEY LAKE, DEWEY LAKE DEWEY LAKE DEWEY LAKE RUSTLER F M. RUSTLER F M, RUSTLER F M. RUSTLER FM SALADO F M. SALADO F M. SALADO FM. SALADO FM. CASTILE FM .11 TANSILr TANSILL FORMATION CAPITAN BELL YATES FORMATION LS. CANYON SEVEN RIVERS FM. WHITE HORSE CLOUD CHIEF CHERRY CN, Z QUEEN FORMATION RUSH SPRINGS MARLOW FM GRAYBURG FM. , DELAWARE^ BRUSHY CNi -i -- ^ MT 6fi DOG CREEK SAN ANDRES SAN ANDRES BLAINE BLAINE GYP L S a SH DOL. < ^ N ANGELO FLOWEFLB 6L0RIETAS.S. SS. DUNCAN SSjui HARPER SS. CHOZA VICTORIO PEAK LS^ FULLERTON DRINKARD VALE ARROYO SPRABERRY SS. LEUDERS CLYDE ' WELLINGTON BELLE, PL'NS PONTO^ ADMIRAL TOC GR, CHASE PUTNAM / GR. MORAN PUEBLO / /COUNCIL GROVE STONE CORRAL

HESS LS

BONE SPRING LS.

ABO LENNOX HIU.S WOLFCAMP NEAL RANCH LS. a SHALE PRESENT

SHALE PRESENT ~^RSUM

pow wg!ft

ADMIRE OR.

Flo. 3Correlation of rock units of Permian basin. Arrows indicate horizons used as datums for paleogeographic maps (Figs. 4-9, 11-13).

either side of a medial zone bounded by some of the old basement faults. In response to increasing tectonic activity, beginning in the later Mississippian, broad basins formed in the east and west parts of the Tobosa basin. Black shale was deposited in the deep central parts of the basins while broad carbonate shelves formed around the margins. The blackshale deposition probably was slow, and much of it may have taken place during times of slight sea-level sinking (Adams et ah, 1951). Peterson and Ohlen (1963) described a similar relation between the black shale and carbonate deposits in the Pennsylvanian of the Paradox basin. At the beginning of Pennsylvanian time strong reef and bank growth continued along the edges of the platforms, resulting in a great complex of reefs and banks at different levels and of different ages (Owen, 1962; Van Siclen, 1969). Each period of reef growth reflects a time of rising sea level because the reef-forming organisms must be covered with at least a minimum of water to survive, and lowering of the sea level will kill these organisms and stop reef growth. Theoretically a stillstand would cause horizontal basinward growth of the reef by overgrowing its own submarine talus. However, there probably are few stillstands that last for any appreciable span of geologic time, so such growth is rare. Marine recession and a subsequent partial rise of sea level would.

of course, result in new reef growth at a lower level on the forereef talus, unless afloodof terrestrial material smothered the reef organisms. Because most of the reef stages can be traced laterally for tens of miles, the changes in sea level probably were eustatic rather than tectonic. On the broad carbonate shelves of Pennsylvanian time there must have been wide channels between separate banks, with reefs growing at some points. This is inferred from the fact that the backreef sedimentary rocks do not show the changes to red and green shales and evaporites that are found in other reef complexes of other ages and in those of the same age in other regions. The western subbasin of the Tobosa basin is known as the Delaware. Along its northwestern edge in Eddy County, New Mexico, Pennsylvanian banks developed especially during Missourian and Canyon deposition. At other times many long stretches of strand were formed, upon which siliceous sand derived from highlands in central New Mexico accumulated. The Pennsylvanian also was a period of increased tectonic activity. South and east of the Permian basin a range of mountains grew out of the Ouachita-Marathon geosyncline (Flawn, 1961, p. 56-58), and great thicknesses of flysch materialsTesnus and Raymond Formationsaccumulated in the foredeeps in front of the

Late Paleozoic Sedimentation in West Texas Permian Basin

2307

FIG. 4Paleogeography of Texas and New Mexico in earliest Permian time.

range. At the same time, the median section of the Permian basin was folded into a mountain range separating the Delaware basin on the west from the Midland basin on the east (Hills, 1963). At approximately the same time the Matador uplift was rising to mark the northern boundary of the basin. This tectonic activity resulted in vertical movement along the ancient strike-slip faults, with some new faulting taking place in the recently deposited Pennsylvanian rocks. Most of the structural petroleum reservoirs in the Paleozoic rocks of the entire Permian basin region took final shape at that time, especially the deep gas reservoirs of the Delaware and Val Verde basins. Thus, in the latter part of the Pennsylvanian, the general framework of the Permian basin had been established, with two subbasins separated by a rapidly eroding mountain range. Broad carbonate shelves on the east side of the main Permian basin were backed by coal swamps in what is now north-central Texas.
LATEST PALEOZOIC SEDIMENTATION

Although many wells have penetrated into the pre-Pennsylvanian rocks, literally tens of thousands of holes have been drilled into the Pennsylvanian and Permian strata of West Texas and southeastern New Mexico. These tests have made possible detailed correlations of cutting samples, cores, and electric logs so that time lines have

been well established. Many of these correlations are shown on the cross sections published by the West Texas Geological Society (Jones, 1949; Davis, 1953; Scobey et a/., 1951; Feldman et al., 1962; Vertrees et al., 1964), the Roswell Geological Society Stratigraphic Research Committee (1956), and The American Association of Petroleum Geologists (Maher, 1960). The U. S. Geological Survey's "Paleotectonic Investigations of the Permian System" (McKee er al., 1967) has made available much information on areas outside the main Permian basin. Figure 3 is a correlation chart which shows the author's interpretation of the time relations of the various formations of seven regions in, and adjacent to, the Permian basin. In each of these basins the contemporaneous rocks are generally of diff"erent facies and have received different names. Many more names have been applied to these rocks, but Figure 3 shows only those in common use at the present time. From this correlation it should be possible to reach certain conclusions with respect to paleogeographic conditions at specific times in the late Paleozoic of the region. I made an earlier attempt at such a synthesis (Hills, 1942). The present work starts with paleogeographic conditions al the close of the Pennsylvanian (Fig. 4). This map is drawn on a time horizon in the

2308

John M. Hills
At the same time, in the Delaware and Midland basins between the platforms, thin beds of shale were laid down. In the Val Verde foredeep in front of the rising Marathon mountains, thick layers of sedimentary rocks accumulated, many of them terrigenous. Over these, thrust sheets advanced from the mountains. In extreme eastern Texas and southeastern Oklahoma a seaway may have existed some time during the Early Permian and possibly in the Late Pennsylvanian, in which the Eagle Mills and related beds were deposited (McKee et ai, 1967). Subsurface evidence as to the age of these beds is rather scarce, but is sufficient to justify showing this feature on Figure 4 and subsequent maps.
MIDDLE WOLFCAMPIAN TIME

middle Harpersville Formation, as found on the east side of the basin. On the west side of the basin in central New Mexico this time horizon probably is represented by the base of the Bursum hmestone, whereas farther south it is placed in the Hueco Limestone at the base of the Powwow Conglomerate Member. In the Marathon region it falls in the upper part of the Gaptank, and in Kansas it is in the upper part of the Waubaunsee Group which grades to the red clastic material of the Pontotoc facies of Oklahoma. The sea at this time had retreated from its Middle Pennsylvanian stand. Perhaps this retreat was mainly eustatic, but it was compKcated by increased tectonic uplift. Folding and uplift had left mountainous areas in central New Mexico, southern Oklahoma, and central Texas, and low mountainous islands projected above the sea in the central part of the basin in southeastern New Mexico and adjacent Texas. In the Texas Panhandle a granite ridge formed low-lying hilly islands around which debris accumulated, forming the arkosic "granite wash" gas and oil reservoir of the Panhandle field. The "granite wash" is analogous to, and partly coeval with, the Fountain sandstone of the southern Front Range of the Rockies. The Abo Redbeds of New Mexico were derived from the lowlands of the central upHft where the Precambrian crystaUine rocks were exposed in limited areas in the core of the central mountain range and were the source of small patches of arkose around the mountains. In spite of the constriction of the seas by eustatic sea-level change and tectonic activity, there is no evidence in the sedimentary rocks of this age that marine circulation was impaired to the extent of causing hypersalinity of the water and deposition of evaporites. However, salinity may have increased slightly on the platforms, resulting in increased amounts of calcium carbonate available to the reef- and bank-building organisms. Thus, on the platforms many reefs and banks caused limestone deposition over great areas. In the platform area of northern Lea, eastern Chaves, and northern Eddy Counties, sedimentation probably was continuous from Pennsylvanian to Permian time (Meyer, 1966). The great horseshoe atoll in the central part of the basin was in the latter stages of its growth at this time. This atoll and other reefs and banks developed large amoimts of porosity through the leaching action of circulating meteoric water. This porosity was added to remnants of primary porosity and to secondary porosity caused by recrystalUzation from limestone to dolomite. Porous zones originating in these three ways now form important petroleum traps.

After the beginning of Permian time tectonic activity subsided, but the higher platform areas remained exposed. However, fine sand, silt, and mud were deposited in the Delaware and Midland basins. In the Marathon and Val Verde foredeep this material was interbedded with sandy calcareous deposits which Ross (1963) has called Neal Ranch Formation. At about the middle of Wolfcampian time the final episode of the Marathon orogeny occurred, during which the Dugout Creek thrust formed and carried lower Paleozoic folded rocks out over the Neal Ranch shale (Feldman et ai, 1962). The writer examined water-well samples from sites north of Marathon which show lower Paleozoic rocks lying on yellow-green shale of Wolfcamp lithology with little apparent deformation of the shale. This orogenic episode resulted in the deposition of the coarser clastic rocks of the lower Lennox Hills Formation (Ross, 1963) and a corresponding increase of sandstone deposition in the eastern part of the Val Verde basin.
LATE WOLFCAMPIAN TIME

After the erosion of the highlands raised by the Marathon orogeny in the Big Bend region and of hills formed by the same episode at the positive centers in central New Mexico and southwestern Oklahoma, a slow advance of the seas began. This flood covered all the basin and resulted in deposition of fossiliferous limestone and shale, except on the hilltops at the south end of the central mountains exposed in the Fort Stockton high. In the north and west, great aprons of arkose spread out from the exposed crystalline rocks of central New Mexico and the AmarilloWichita mountains. In north-central Texas, shale and some limestone were deposited. One of these beds is the

Late Paleozoic Sedimentation in West Texas Permian Basin

2309

FIG. 5Paleogeography of middle Wolfcampian time.

well-known Coleman Junction Limestone Member of the Putnam Formation, which is the bed on which Figure 5 is based. At the same time, in the central part of the basin, fringing reefs and carbonate banks were growing on the flanks of the old mountain range and converting it into the Central Basin platform. The late Wolfcampian was the last time during the Permian when marine circulation was unrestricted and normal marine sediments were deposited. Later Permian sedimentation was dominated by the growth of extensive carbonate reefs and banks which severely restricted marine circulation. This restricted circulation, combined with
Normal
Marine (Pontic) SEA LEVEL Free circulflfion1 _- Q o J Lime a Sand Block Gray , _ Lime Vitosaline Penesalint Saline

cyclic or repeated lowering of sea level, gave rise to shallow, barred basins (Woolnough, 1937), in which great thicknesses of evaporites were deposited.
LEONARDLAN TIME

Figure 6 shows how concentration of saline material takes place according to Lang (1937), whose concept fits the known occurrences in the Permian basin. Seawater enters a lagoon over a bank or reef, forming a shallow bar. The barrier hinders the return flow of the highly saline denser water that has been concentrated by evaporation in the shallow lagoon behind the bank. This proBrackish Supersaline Fresh Sand a Red clays Character of Seawater tRote of evaporation

Area of restricted circulation .< ' Q A I. J X Anhydrite Halite a Mg. Lime a Anhydrite g Potash solts Halite Red White Buff White a Red

Red beds

Composition of sediments Color of sediments

(AFTER LANG) FIG. 6Conditions of evaporile deposition in Late Permian time (modified from Lang. 1937).

2310

John M. Hills
fifth of its primary volume took place, and deposition of anhydrite began. The increased concentration of magnesium ions in the dense layers near the bottom probably caused the dolomitization of the previously deposited calcareous muds forming the floor of the lagoon (Adams and Rhodes, 1960). Still further concentration in the saline and supersahne zones, probably in water only a few feet deep, resulted in deposition of halite. In limited areas evaporation went essentially to completion and potassium salts were precipitated. Many of the supersahne lagoons were close to shore, so that red sand and clay interfinger with the various salts. Local thin limestones are found in association with these clastic beds, indicating that occasionally there was enough fresh water draining off" the land to reduce the salinity of the lagoonal water considerably. Though Lang's (1937, Fig. 24) diagram shows only one marine fades, I have shown an additional facies called "pontic." This term was introduced by Lloyd (1938) to denote black shales and very fine sandstones, a facies called "euxinic" by some authors. It is quite distinct from the grayer shales and coarser sandstones of normal marine basins, and seems to have been deposited in the depths of restricted basins where circulation was too poor to allow oxidation of organic material, but not restricted enough to make possible the deposition of evaporites. Some lighter colored shales and fine sandstones in this facies may be the result of turbidite flows which carried oxidized material from basin edges to deeper water environments. The effect of the restricted basins in the saUne and supersahne zones first becomes apparent in the lowermost Leonard rocks of the West Texas region. However, evaporite deposits are prominent in the older Wolfcamp beds of Kansas, indicating that recession of the Permian sea took place from north to south (Adams, 1963). The basal Leonard formation of west-central Texas is the Belle Plains. The Valera Member of this formation contains gypsum on the surface which, in the subsurface of Concho County, grades into nearly 200 ft of anhydrite and shale of the saline facies and then to dolomite of the bank-edge vitasaline zone. Beds above this member show increasing amounts of evaporite in the backreef facies and thick dolomite on the shelf edges. A few thin sandstone beds are present, which probably reflect occasional eustatic lowering of sea level with consequent spreading of terrigenous sediments

cess requires an evaporation rate far in excess of the freshwater inflow from any tributary streams, and if it operates on a large scale, the chmate must be warm and dry. This concept fits the known occurrences of saline material in the Permian basin, therefore, the Permian climate probably was similar to that prevailing there todaya horse latitudes desert. To move this area north more than 5 of latitude by any reconstruction under the theory of continental drift and polar wandering would place it in the belt of prevaihng westerhes and a more humid regime. To move the region 10 or more south would bring it into the tropical wet-dry zone. Neither of these chmatic zones is compatible with the type of sediments laid down during most of Permian time. The global circulation that determines the general location of climatic belts is dependent on differences in insolation from equator to poles, and on the earth's rotation. As these relations probably have not changed since the Permian, the climatic belts were generally the same then as today. Thus it seems that the Permian basin region must have been near its present latitudinal position during most of the Permian Period. Lang's (1937) reconstruction of evaporite-depositing conditions is not far from a later theory of Landes (1963) called the "long fetch" theory, in which the necessity of a bar is eliminated by gradual evaporation and increasing salinity over a wide shallow shelf However, in the case of the Permian lagoons, initial concentrations of the seawater by evaporation at the shelf edge made available calcium ions in greater than normal concentration for use by carbonate-secreting organisms. These organisms in turn built up the shelf-edge barrier, which further impeded circulation and promoted the formation of evaporites in the lagoonal area. Porosity which developed in these shelf-edge carbonates forms many important petroleum reservoirs in Middle Permian rocks. Life, largely in the form of calcium carbonatesecreting sponges and bryozoans, was confined to the shelf edge and part of the adjoining lagoon, where biohermal mounds were abundant. Away from the basin edge the salinity of the water increased abruptly and the only organic remains are those which were washed into this environment by occasional storms. Details of the reef and bank zone are shown by Silver and Todd (1969), although their models do not emphasize the importance of the broad areas of evaporitic deposition in the backreef zones. In this penesaUne zone, concentration of seawater to perhaps a

Late Paleozoic Sedimentation in West Texas Permian Basin

2311

FIG. 7Paleogeography of middle Leonardian lime. Shelf carbonate deposition is starting to separate the Delaware and Val Verde basins.

over the shelves (Silver and Todd, 1969). Figure 7 reflects conditions just prior to the spread of one of these sand sheets, the FuUerton-Drinkard (King, 1945). This sandstone is correlated approximately with the Tubb Sandstone of Silver and Todd (1969) and the middle Yeso Formation on the New Mexico side of the basin, as well as the Vale Formation on the east side of the basin. At this stage the seas completely covered the West Texas-New Mexico basin, and sedimentational fades were well marked. Brackish water sediments formed a fringe around the edge of the sea, and only in Oklahoma and northern New Mexico do they cover any extensive area. Saline seas covered most of the central part of the basin, extending far north into Kansas. This arrangement of facies indicates that the land surrounding this sea probably was both low-lying and deficient in rainfall. At the same time the vitasaline or limestonedepositing sea covered all of the southern Central Basin platform and most of the shelf area on the east and northwest side of the basin, probably extending south into Chihuahua. Black shales with subordinate amounts of fine sand were deposited in the depths of the Midland and Delaware basins. The southern extent of these sediments is unknown, but they appear to thicken

southward and may once have been continuous with open-sea beds in South Texas and northernmost Mexico. In late Leonardian time there was a considerable uplift of the lands in the three positive centers surrounding the Permian basin. This activity was reflected in the upper Yeso sandstone of New Mexico, the sandstones and shales of the Choza Formation of central Texas and the Harper sandstone of Oklahoma and Kansas. Clastic deposition culminated in the deposition of great sheets of coarse fan and deltaic material that we know as the Glorieta Sandstone in New Mexico, the San Angelo conglomerate in Texas, and the Duncan sandstone and Chickasha Formation in Oklahoma.
SAN ANDRES DEPOSITION

The Glorieta, San Angelo, and Duncan are basal clastic beds of a thick, widespread carbonate unit, the San Andres Limestone, that is present from central Texas to Arizona and Utah and is correlated with the Kaibab and Toroweap. The San Andres grades northward into anhydrite and salt in Oklahoma and Kansas where it is known as the El Reno Group. It reflects a major advance of the sea and a short-term reversal of the general secular contraction of the Permian seas. The San Andres also is important because it contains

2312

John M. Hills

FIG. 8Paleogeography at time of greatest spread of Late Permian seas (datum is base of Blaine).

some of the most prolific petroleum reservoirs of the province. Traps in these rocks are structural and stratigraphic, reflecting both the growth of carbonate banks and the gradation of porous dolomites into impermeable evaporites. The carbonate rocks of the San Andres represent a shelf facies, whereas the fine sandstones and black shales of the Delaware Mountain Group are basin facies (Fig. 8). The Delaware Mountain Group by definition (Adams et al, 1939) comprises the Guadalupian Series in the western part of the region. These two facies can be traced within a few miles of each other, but their age relations are still doubtful. Paleontologic evidence accumulated by Lewis (1941), Clifton (1945), Skinner (1946), and HoUingsworth (personal commun.) indicates that the San Andres is equivalent to the Word Formation of the Glass Mountains, as well as the Brushy Canyon and lower part of the Cherry Canyon Formations of the Delaware basin. A summary of the fusuline evidence is given by Silver and Todd (1969), with the stratigraphic position of the fossils shown on their cross sections. In the area south of Carlsbad, New Mexico, Hayes (1959) interpreted the lower San Andres as Leonardian or earhest Guadalupian, whereas the upper San Andres which disconformably overlies

these beds is equivalent to the Cherry Canyon (Guadalupian). My correlations indicate time equivalency of the upper San Andres dolomite of the southeastern part of the Central Basin platform with basinal beds bearing Guadalupian fusulinids in central Crockett County. Nevertheless, some Leonard-type fossils have been found in the San Andres of New Mexico (Beede, 1910), and recently Meissner (1967) has reemphasized physical evidence from New Mexico, which seems to show that the San Andres should be correlated with the Leonard. The San Andres probably is continuous and coeval with the Kaibab of Arizona, which is definitely dated as Leonardian (McKee and Breed, 1969). The solution to the problem of the exact age of the San Andres has yet to be found. On the other hand, the correlation of the San Andres dolomite with the evaporite and terrigenous rocks on the landward side of the shelf away from the basin is relatively good. There, as Figure 8 shows, the limestone and dolomite of the vitasaline seas grade through the anhydrite and salt of the saline zone into nearshore red shale and sandstone. Adjacent beds of the formation are lithologically distinct and can be traced for long distances so that their gradation into other facies can be seen plainly. The cross sections by the West Texas Geological Society (T. S.

Late Paleozoic Sedimentation in West Texas Permian Basin

Jones, 1949; Davis, 1953) and The American Association of Petroleum Geologists (Maher, 1960) show this in detail. Well-to-well correlations on the shelf around the north end of the Midland basin do not encounter the problem of basin facies, but yield strong evidence that the San Andres dolomite of the platform and shelf zones is continuous and probably contemporaneous with the Pease River Group of western and north-central Texas and the El Reno Group of Oklahoma and Kansas. The maximum San Andres flood took place early, at about the time of basal Blaine deposition (top of the "Slaughter" pay zone of Hockley County in the northern part of the basin; Acme of Oklahoma). Figure 8, a paleogeographic map of this time, shows an extensive area of limestone deposition. In this region carbonate banks were present in which scattered reef mounds grew on local sea-floor highs. This sea probably extended westward to join the Kaibab sea. Although beds of this age appear to have been removed by erosion in far western Texas and southwestern New Mexico, the stratigraphic position and hthology of the Kaibab are similar to those of the San Andres (McKee and Breed, 1969). This formation has not been found in Mexico, but seas of this time probably covered northern Chihuahua. Pontic black mud, fine sand, and silt accumulated in the Delaware and Midland basins, probably augmented by turbid flows from sediment banks on the basin margins. Although widespread sandstone zones are not present in the San Andres section, disconformities in the formation probably represent diastems during which terrigenous material was swept over the shelf and deposited on the basin slope. As previously noted, Hayes (1959) believed there was evidence for a major hiatus in the San Andres limestone of Andrews Countythe upper San Andres of the shelf edge. Supporting this view is the presence of small seams of coaly material that have been identified as fragmentary remains of land plants (A. C. Noe, personal commun., 1937). In north-central Oklahoma, Fay (1964) has called attention to the presence of carbonate rocks, as well as thick red clastic beds in the upper El Reno Group. These beds indicate a deepening Anadarko basin that may have connected with a sea on the north throughout most of San Andres time. This confirms earlier work by Chfton (1944). Along the eastern edge of the Permian and Anadarko basins, the San Andres evaporites grade into the red shales and sandstones of the El Reno and Pease River Groups. Because these sandstones become coarser toward the east, they

2313

probably lay near the San Andres shoreline. However, because the beds are truncated by erosion, the strand deposits are not present. No very coarse material has been found, so by this time the ancient Ouachita Mountains probably had been worn down to low relief On the northwest side of the main Permian basin the hmestone-depositing area adjoined a redbed zone, partly marine and partly terrestrial, perhaps including the lower Bernal Formation. The mountains of north-central New Mexico probably were very low at this time and shed largely fine sediments. Throughout San Andres time the sea retreated not continuously, but in cycles broken by a number of small advances. Each of these advances left a record in a dolomite bed overlain by anhydrite, halite, and red shale deposited in successive stages of retreat and desiccation of the shallow sea. The brines present in the later stages of the cycle probably were agents of the dolomitization of the underlying carbonates through the reflux action of magnesium-charged waters (Adams and Rhodes, 1960). A similar process on a smaller scale has been described by Kinsman (1969) from the sabkhas or supratidal salt flats of the Persian Gulf region. Naturally such cycles of advance and retreat are shown better in the San Andres rocks of the northern part of the Permian basin at the farther edge of the vitasaline zone. In Floyd County, Texas, nine such cycles can be detected and in Castro County, 13 cycles. On the south, the lower dolomite beds abruptly become thicker and merge into an unbroken brown granular dolomite 600 ft thick. This section may be correlative with Hayes' (1959) lower San Andres member. In the upper 800 ft of the formation, the evaporites grade less abruptly southward into lighter more finely crystalline dolomite, with some vuggy and intergranular porous zones. Locally, banks comprised of very light-colored, coarsely crystalline porous dolomite are developed, notably at Hobbs in east-central Lea County and at Penwell in southern Ector County. The upper member of the San Andres is missing, apparently by erosion, from the structurally high area of central Crane and northern Pecos Counties. This leaves only about 800 ft of the lower brown granular cherty member present, although the normal thickness is 1,400 ft. In the northern and northeastern part of the basin the San Andres thins by steps from the top, through abrupt disappearance of the entire upper part of each dolomite-evaponte cycle, where the resistant basal dolomite above it grades into red shale, anhydrite, and halite. Thus, both on the

2314

John M. Hills

FIG. 9~Paleogeography during Grayburg-Marlow deposition.

southern Central Basin platform and in the northeastern basin, there seems to be evidence at the top of the San Andres for a widespread unconformity which results in the truncation of several hundred feet of beds. In the Anadarko basin the evidence does not seem to be as good for an unconformity at this horizon, although the top of the Dog Creek Shale may mark a considerable hiatus. Thus the San Andres dolomite appears to be bounded below by sandstone and deltaic sedimentary rocks and above by a well-marked unconformity. Above this unconformity is a succession of rocks, the lower part of the Guadalupe Series, which are quite different in aspect from the San Andres in that these strata reflect a sharp retreat of the sea from its San Andres expansion. Dolomite in the beds above the San Andres is limited to the southern half of the Central Basin platform and the Midland basin (Fig. 9). Evaporite beds become increasingly important above the top of the San Andres in the north-central part of the Permian basin and include halite, red sandstone, and shale. These nearshore saline lagoonal deposits contrast markedly with the thick dolomitic San Andres beds underlying them. This change in hthology indicates a major change in physical sedimentary conditions. Thus,

on physical evidence of unconformity and change in lithology, the San Andres beds might better be placed in the underlying Leonardian rather than in the Guadalupian, paleontologic evidence notwithstanding. If, however, one wishes to follow the paleontologic correlation and place the San Andres in the Guadalupian, one must postulate that the post-San Andres unconformity is not present in the Delaware basin. It may be that most of the deposition of the lower Cherry Canyon elastics took place during the hiatus at the end of San Andres deposition while the carbonate rocks of the Central Basin platform were being eroded by solution, leaving no coarse debris to accumulate in the adjacent basin. Such solution-type erosion is occurring in Bermuda today (Bretz, 1960). Probably porosity in most of the San Andres petroleum reservoirs was developed at this time and added to that formed by previous dolomitization. In any case, it is certain that the vitasaline sea at the end of San Andres time occupied a very restricted area in the south end of the basin. Shortly afterward the sea retreated completely from the region, except for the Delaware basin.
GuADALUPUN T I M E

Following the post-San Andres interval there

Late Paleozoic Sedimentation in West Texas Permian Basin

was another limited advance of the sea. This flood carried the carbonate-depositing waters over most of the Central Basin platform and the southern part of the Eastern shelf, laying down the Grayburg Formation. The basal part of the Grayburg consists of sandy dolomite filling the low areasremnants of the Midland basin and the San Simon channel. The upper part of the formation is a well-bedded dolomite with sandy interbeds and a few thin bentonites. The top of this zone is the datum on which Figure 9 is drawn. On the basin edges the dolomites thicken into porous bank deposits that form important oil reservoirs. The saline sea advanced northward and approached the south flank of the Amarillo uphft where it merged with red mud and sand deposited in brackish and terrestrial environments. The shoreline of this time is difficult to delineate, as most of the strand rocks have been eroded. However, in southwestern Oklahoma lagoonal gypsum and thin dolomite of the Marlow Formation probably were deposited behind the Verden Sandstone offshore bar (Fay, 1964). This indicates that shallow marine conditions prevailed at least that far north. Recession of the seas continued throughout the rest of Guadalupian time with minor advances, which probably reflected eustatic rises in sea level. Major Capitan reef growth continued around the edge of the Delaware basin above the basinward reef slopes of the Goat Seep reef, thus constricting the area of the basin. Growth was, of course, possible only during times of rising sea level. The great vertical dimension of the massive reef (King, 1948) indicates relatively rapid rise in sea level. The reef seems to have grown from centers between surge channels. Jacka e( al. (1967) presented evidence that these channels may have been true submarine canyons. At times of reduced sea level the top of the reef wall was brought above high tide and growth ceased. Then fine sand was swept through the channels or canyons into the Delaware basin where it formed most of the Bell Canyon Formation. Where local subsidence was active, such as in the northeastern corner of the Delaware basin in southwestern Lea County, New Mexico, the reef was less well developed and the channels larger. Probably in such localities greater amounts of sand were swept through the channels during recessions, and perhaps this sand transport may have been almost continuous throughout Guadalupian time. The general northeast trend of the petroleum-producing sandstones {e.g., Ramsey

2315

sand) of the upper Delaware Mountain Group (Bell Canyon) may indicate a source in this direction. In addition to this local subsidence there also was apparently some general subsidence in the basin area, which deepened the water in these structurally negative centers and adjoining shelf edges. However, the cyclic nature of the apparent sea-level changes would require very small-scale mobihty of the earth's crust to cause them tectonically. Therefore, short lived sea-level fluctuations probably were superimposed on long term subsidence to produce the stages of reef growth. It has been postulated (Jacka et al., 1967, p. 169-172; Wanless, 1967, p. 52-53) that such eustatic changes in sea level may well have been caused by the locking up of water in the ice caps of recurring glaciations in Gondwanaland during the closing periods of the Paleozoic. Jacka and his colleagues pointed out that increased rainfall connected with such glaciations could account for textural changes in the exposed reef limestones, as well as making increased amounts of fine terrestrial sediment available for deposition in the basins. The duration of the times of lowered sea level must have been relatively short, because no appreciable reef growth is found on the basin slope, although the great accumulation of reef talus on the slope should form a good substratum for such growth. With another rise in sea level, reef growth would again be resumed. Back of and parallel with the reef, sand accumulated in lenticular bodies waiting to be swept into the basin at the next sea-level recession. Such bodies can be seen clearly in the Rocky Arroyo section (Bates, 1942). They also form important petroleum reservoirs in Range 37, Lea County, New Mexico, and extend south through central Winkler and Ward Counties, Texas, to the vicinity of Fort Stockton in Pecos County. During the times of low sea level and reef-top emergence, sand advanced over the reef and backreef carbonates; after dolomitization these carbonates were leached by relatively fresh water. This resulted in well-crystallized, blue-gray dolomite with vuggy porosity, usually connected and permeable. These porous rocks commonly are overlain by sandy, finely crystalline and poorly permeable dolomite. Such porous carbonate zones form a row of prolific petroleum reservoirs between the sandstone reservoirs mentioned heretofore and the marine reef growth centers at the basin edge. These porous zones are illustrated

2316

John M. Hills

R38K

T-EXAS

Dolomite Dolomite, blue gray, coarsely crystalline, poroas m Sandstone, red to yellow Anhydrite

Flo. 10Generalized east-west cross section through Cooper-Jalmat field in northern part of T24S (location shown on Fig. 11), showing relation of porous zones in lower Capitan and Goat Seep reef dolomites to sandstone zones of backreef Artesia Group and lagoonal anhydrites. Sand zones and corresponding tops of reef porous zones may represent limes of arrested reef growth and development of porosity by groundwater action.

by the section of the Cooper field (Fig. 10) in which seven cycles are shown. From outcrop evidence, Bates (1942, p. 98) suggested pauses in carbonate deposition while gypsum and red clastic material were deposited in the backreef zone. Jacka et al. (1967) described closely related phenomena from outcropping rocks of the same age. With each episode of renewed reef growth the basin became more restricted, until the reef and its associated hmestones became merely a strip surrounding a shrunken Delaware basin as indicated in Figure 11, a paleogeographic map of late Capitan time. During this time a high and narrow reef with few channels, except in the northeast corner, cut off most circulation of marine water in the Delaware basin. This feature probably trapped most terrigenous sediment from the land areas. Ball et al. (1971) have postulated a depositional situation in which the quartz sand was laid down in lagoons between the dolomites of the backreef and the evaporites of the landward ponds. This interpretation seems in better accord with my observations than does that of Silver and Todd (1969) in which the sands are considered to be continuous blankets overlapping older carbonates. Lagoon sedimentation in the Delaware basin would be very slow, except near the base of the reef where thick talus slopes and

mud slides are preserved as wedges with steep original dips (Newell et al., 1953; King, 1948). Bottom flows spread the finer fractions of this sediment over the deeper parts of the basin. Limited water circulation might have been maintained through a narrow strait north of the Marathon uplift. Perhaps at the close of Capitan time this strait became constricted and saline waters accumulated in the depth of the basin, eventually becoming concentrated enough to deposit evaporites. Back of the reef zones the dolomites and sandstones grade abruptly into the anhydrites, red shales, and salts of the upper part of the Artesia Group. The stages of reef growth are reflected in the backreef zone by carbonate spreading followed by deposition of anhydrite and hahte (Fig. 10), whereas cessation of growth resulted in the spreading of clastic material. Back of the reef the eastern part of the basin was covered with a very shallow evaporite-depositing sea in which the concentration was great enough in the northern extension to precipitate polyhahte, the mixed potassium-bearing sulfate, as well as anhydrite and hahte (Davis, 1953). The local development of beds of red shale and sandstone interbedded with the evaporites shows that the shallow saline lagoons were at times filled with detritus washed or blown from lands surrounding the basin.

Late Paleozoic Sedimentation in West Texas Permian Basin

2317

adNTTNESrjrt'

OR r '^ /'.. BRA^I^H .' \,'V/.,,'V/1^^,

FIG. UPaleogeography of late Guadalupian time (upper Capitan deposition). This was epoch of intensive, highly restricted reef growth. OcHOAN T I M E

The close of Guadalupian time marks a great change in physical conditions in the Permian basin. Organic reef building ceased and evaporite deposition spread over the entire region. These uppermost evaporite beds have been studied by Kroenlein (1939), Adams (1944), Anderson et al. (1972), and others. The paleogeographic conditions in the early and middle part of this epoch are shown on Figure 12. Reconstruction of conditions at the close of the limestone reef deposition of the Guadalupian and the beginning of Ochoan deposition presents some difficulties. At the close of the Capitan deposition there was nearly vertical buildup of the upper Capitan limestone. A,short hiatus is marked by the Ocotillo silt which hes 75-100 ft above the Yates Sandstone. It is possible that shallowing of the southwest entrance to the Delaware basin by tectonic action or sedimentary processes (Adams, 1944) may have interrupted the return flow of dense saline waters. This may have initiated a buildup of salts in the bottom of the basin overlying the black shales and limestones of the Lamar Formation (uppermost Delaware Mountain Group). In some deep parts of the basin 30-40 ft of dark anhydritic limestone beds overhe the Lamar. Possibly these beds mark the beginning of evaporitic deposition.

Such conditions would lead to the diminution of flow of nutriments to the reef around the Delaware basin. A rise in sea level of perhaps 100-200 ft may have enabled the Capitan reef to have a final, rather short, growth period, building the uppermost Tansill equivalent. Following this final growth episode eustatic lowering of sea level probably exposed the top of the reef which, together with the concomitant fiUing of the basin with supersaline waters,finallykilled the reef If this reconstruction is correct, the lowest Castile strata are contemporaneous with the uppermost Tansill beds. This correlation was suggested by Kroenlein (1939), but because the lowest part of the basin, where the first evaporite deposition took place, is widely separated geographically from the reef top, no substantial evidence has been found as to the time relations of these beds. The most striking feature of the Castile Formation is the laminated anhydrite, which is 1,5002,100 ft thick and widespread in the Delaware basin. These laminations were first described by Udden (1924) and consist of 176,800 laminations, 1/10 in. to 1/20 in. thick. Each contains a thin brown layer of calcite crystals stained by hydrocarbons grading downward into a slightly thicker lamina of anhydrite. The thickness of the calcite

2318

John M. Hills

PALEOGEOGRAPHY EARLY AND MIDDLE OCHOA TIME CASTILE AND SALADO

^ H ^ ^ ZONE UPPER Of UPPER SALADO CASTII.E AND L<WER POTASH SALAOO OVERLAP'^

COMMERCIAL

DEPOSITION

FIG. 12Paleogeography of early and middle Ochoan times showing conditions during deposition of two greatest salt formations of North American Permian. Geography beyond central part of Permian basin is unknown.

layers is relatively constant, but that of the anhydrite layers varies greatly. In parts of the formation there are especially thick anhydrite beds, which seem to have resulted from incomplete cycles. Adams (1944) studied these laminae and Anderson et al. (1972) made an extremely detailed study of a series of cores taken from this formation in the deepest part of the Delaware basin in Ward and Winkler Counties, Texas. The results of these studies show that there is uniformity of laminae over many hundreds of square miles. This argues for some sort of geochemical control, rather than a repeated opening and closing of a channel leading to the Delaware basin or a short-term fluctuation of sea level. Udden's (1924) suggestion that these laminae were caused by seasonal variation in the composition in the seawater seems to be confirmed by later workers (Adams, 1944; Anderson et al., 1S*72). If this hypothesis is correct, the whole section of nearly 2,000 ft of laminated Castile beds was deposited in the remarkably short time of less than 200,000 years. In the deep part of the basin, through some interruption of the flow of water into the basin, several zones of halite each several hundred feet thick, were deposited as part of the laminated sequence. Toward the reef, a few feet of the basal banded

Castile grades into laminated limestone and then into uppermost Capitan reef carbonate (C. L. Jones, 1954; Newell et al, 1953). Anderson et al. (1972) stated that the Bell Canyon Formation which underlies the Castile also is laminated. As the Bell Canyon is the basinal equivalent of at least part of the upper Capitan reef, it seems that the conditions causing the laminations originated before the close of Capitan deposition and continued for some time afterward. Correlation of beds of the upper Castile in the vicinity of the Capitan reef indicates that some of them may be equivalent to the lowermost Salado halite below the Cowden anhydrite (C. L. Jones, 1954; T. S. Jones, 1949). Anderson et al. (1972) showed that the lower 400 ft of the Salado in the basin is laminated hke the Castile. As can be seen in Figure 12, the area covered by the Castile evaporites differs considerably from that covered by the Salado. This can be explained most easily by northeastward tilting of the Castile in the Delaware basin through subsidence along the West Platform basement fault, which forms the east edge of the basin (Hills, 1970). This tilt would bring the western margin of the Delaware basin above sea level and spiU the brines over the top of the Capitan reef zone into

Late Paleozoic Sedimentation in West Texas Permian Basin

2319

' - ^ ^ v

FIG. 13Paleogeography of late Ochoan time, final marine flood of Permian time.

the central part of the Permian basin (Fig. 13). This idea was first pubhshed by Kroenlein (1939). Because the Salado is composed largely of halite beds with interbedded thin red shale and some polyhalite, these evaporites probably were laid down in small shallow lagoons. Evaporation was very great, but deposition often was interrupted by freshwater drainage from the land which sometimes flooded the lagoons with fresh water. Evaporation often was sufficient to permit the deposition of the potassium-bearing sulfates. Thin sandstone beds containing rather small, but well-rounded, frosted grains are found, indicating that these lagoons might have been filled occasionally with sands derived from the continental areas on the north and east and deposited by wind in coastal dunes. The water remained deeper in the northeastern Delaware basin and spread over the top of the reef in southwestern Lea and southeastern Eddy Counties, New Mexico, to form Ochoa Lake (Kroenlein, 1939). In this lake, evaporation went on at a high rate, uninterrupted by changes in water depth or invasion of clastic sediments, except for relatively small amounts of wind borne clay. Here the sylvite, langbeinite, and related salts that now form the commercial potash deposits of the Carlsbad area precipitated from un-

usually concentrated brines. In the potash areas the upper Salado beds continue over the reef zone, with only slight thinning. However, to the south in Texas and extreme southeastern New Mexico the reef seems to have been less affected by tectonic subsidence; therefore, there were no saline-depositing lagoons over the top of the reef during Salado deposition. In this part of the reef zone the Salado is composed almost entirely of anhydrite with thin salt beds. The whole Salado section is only 500-600 ft thick, compared with 1,200-1,300 ft in southeastern Eddy County. In parts of this reef trend in northern Pecos County the Salado contains several beds, 10-15 ft thick, of brown crystalhne dolomite. It is possible that on the bottom of the shallow sea the Capitan reef mounds formed buried protuberances, which raised the sea floor to the layer of fresher water near the surface, resulting in the deposition on the high of dolomite by bacterial or algal action. On the north and east the Salado gradually becomes more clastic, with less halite and anhydrite. This effect is marked along the eastern edge of the Permian basin in Irion, Mitchell, and Scurry Counties where almost the entire saline content of the formation is lost. Thus, it is very probable that the present eastern hmit of the Salado in these counties is very close to the origi-

2320

John M. Hilts
the north and east, where they grade first into anhydrite, then salt, and then redbeds. As these dolomites phase out, their protective function against later erosion is lost, and the Rustler thins abruptly from western Andrews County to eastem Garza County where it is absent (Fig. 13). Probably, as in the case of the Salado, the present limit of the Rustler is very close to the original depositional hmit. The Rustler incursion of the Late Permian sea has been indicated on Figure 13 as vitasahne. Walter (1953) described a good fauna from the Rustler of the Culberson County outcrop. Many of these fossil forms are preserved in anhydrite. The dominance of pelecypods and gastropods and their resemblance to backreef Guadalupian facies forms suggest a hypersaUne environment. Much of the Rustler dolomite is of the honeycomb texture, which may reflect leaching of pisolitic forms. The fauna is definitely Late Permian, but it cannot be correlated exactly with Late Permian faunas in other parts of the world. In Figure 13, I have shown the Rustler sea as advancing from the southwest and retaining an outlet in that direction to the open sea. The Tessey dolomite of the Glass Mountains may be correlative with the Rustler (Adams, 1944). If so, the Tessey forms a thickening of the Rustler dolomites in the basin, perhaps because the Tessey occupied a position closer to the entrance from the open sea where there was better access to nutriment and less saline water. In the Oklahoma and Kansas area no formation is present which can be correlated with the Rustler. However, possibly the Doxey of northcentral Oklahoma (Fay, 1965, PI. II) may correlate with some of the Ochoa formations and may have been deposited in a sea advancing from the north. Otherwise there is no evidence in the southwestern United States of the presence of Permian rocks as young as the Rustler. The seaway in East Texas probably was not receiving sediment at this time and was not reopened until middle Mesozoic time. Rustler sedimentation probably ended by complete withdrawal of the sea from the southwest Texas area. Because there is no evidence of tectonic movement at that time, the probabiUties are that this was another eustatic movement. In the few lagoons remaining in the southern part of the basin, anhydrite deposits formed, and the basin was quickly filled and covered by sand and silty shale carried in from the emergent shelf areas on the north and northeast. These orange-red sandy shales and siltstones, known as the Dewey Lake Formation, contain many large (0.5mm) wellrounded and finely frosted grains. Therefore,

nal depositional limit. On the north the same increase in clastic material is noted, but the entire Salado section seems to be lost by truncation along the south slope of the Matador uplift before the northernmost depositional limit is reached. On the south edge of the basin, truncation again occurs under Cretaceous beds against the north edge of the Marathon-Ouachita uplift. The Salado depositional limit probably was very much south of its present line, but distinct thinning is noticed in southern Pecos, northern Terrell, and central Crockett Counties. On the west edge of the Salado area along the Reeves-Culberson County boundary and west of Carlsbad, New Mexico, the saline rocks of the Salado have been truncated by later Permian erosion. This truncation has been accentuated by halite solution from the underflow of the Pecos River and related groundwater channels in late Cenozoic time. Thus the western depositional Umit of the Salado is uncertain. Nevertheless, the increase in clastic material in the formation and the thinning as far west as the Pecos River indicates that perhaps here, too, the present limit of Salado rocks is close to the original depositional edge. The time needed for deposition of the Salado strata is not so clearly indicated as the time required to deposit the varvelike laminations of the imderlying Castile. Probably, however, the salt of the Salado was deposited during a relatively short interval, and the entire Ochoan epoch covered a very small part of the total of the Permian period, perhaps less than a million years. Following the end of Salado sedimentation there was a period of erosion and some solution of the saline deposits, as the Salado beds on the east and west edges of the basin are truncated by the basal elastics of the overlying formation, the Rustler. Rustler deposition began with a final incursion of the Permian sea into the southwestem United States. This sea deposited a covering of red shale and sandstone from 10 to 100 ft thick over the truncated edges of the Salado evaporites. Probably this material was derived mainly from the clastic material once part of the Salado Formation. Adams (1944) has described the Rustler in detail, and Hills (1942) has pubHshed a cross section of the best developed subsurface occurrences correlated with the type section. These sections show that the basal Rustler elastics are overlain by evaporites which are generally anhydritic on the south but become more saline on the north. Overlying these evaporites are three dolomite beds which are thickest in the southern part of the Rustler area and become less prominent in

Late Paleozoic Sedimentation in West Texas Permian Basin wind probably had a prominent part in the deposition of this formation. However, the Dewey Lake is regularly bedded with traces of soft gypsum so that it probably was deposited in playa lakes. The Dewey Lake Formation usually is from 100 to 200 ft thick. With the deposition of these redbeds, Paleozoic sedimentation in the southwestern United States ceased.
SUMMARY

2321

In Late Paleozoic time slow sedimentation in broad basins was replaced by sedimentation in a system of smaller basins sepa'rated by tectonically positive belts. These basins restricted circulation, so that calcium-carbonate concentration in the seawater increased. Carbonate shelves formed along the basin margins, while fine clastic sediments were deposited in the deeper parts of the basins. This restriction was accentuated by tectonic action in Late Pennsylvanian and Early Permian times, contemporaneous with the Marathon-Ouachita orogeny. Thus, by Middle Permian time restriction of marine circulation, coupled with eustatic withdrawal of the sea in a southwest direction, resulted in growth of high reefs and carbonate banks. Behind these barriers were wide shallow lagoons containing highly saUne waters. Here evaporites were deposited, ranging from finegrained dolomite, through sulfates and chlorides, to potash minerals. Most of this deposition probably was rapid and cyclic, being dependent on eustatic changes in sea level, perhaps glacial in origin. The longterm trend was gradual withdrawal of the sea toward the southwest. However, this was interrupted by a strong and fairly long-lived advance of the San Andres seas, and the smaller and short-lived Rustler flood close to the end of the period. During the Ochoan, the closing epoch of the Permian, deep-water evaporitic deposition took place in the Delaware basin and the banded laminated Castile anhydrite was laid down. Later, the potash-bearing salts of the overlying Salado Formation were deposited in shallow lagoons. After Salado deposition ceased, the Rustler flood covered the evaporites. Then, a final marine recession initiated a short interval of continental deposition before the close of the period and era. Sedimentation was not resumed in the region until the Dockum continental redbeds were laid down in the Late Triassic.
REFERENCES CITED
Adams, J. E., 1944, Upper Permian Ochoa Series of Delaware basin, West Texas and southeastern New Mexico: Am. Assoc. Petroleum Geologists Bull, v. 28, no. 11, p. 1596-1625.

1963, Permian salt deposits of West Texas and eastern New Mexico, in Symposium on salt: Northern Ohio Geol. Soc, p. 124-130. and M. L. Rhodes, 1960, Dolomitization by seepage refluxion: Am. Assoc. Petroleum Geologists Bull., v. 44, no. 12, p. 1912-1920. et al, 1939, Standard Permian section of North America: Am. Assoc. Petroleum Geologists Bull., v. 23, no. 11, p. 1673-1681. H. N. Frenzel, M. L. Rhodes, and D. P. Johnson, 1951, Starved Pennsylvanian Midland basin: Am. Assoc. Petroleum Geologists Bull., v. 35, no. 12, p. 2600-2607. Anderson, R. Y., W. E. Dean. Jr.. D. W. Kirkland, and H. 1. Snider, 1972, Permian Castile varved evaporite sequence, West Texas and New Mexico: Geol. Soc. America Bull., v. 83, no. 1, p. 59-86. Ball, S. M., J. W. Roberts, J. A. Norton, and W. D. Pollard, 1971, Queen Formation (Guadalupian, Permian) outcrops of Eddy County, New Mexico, and their bearing on recently proposed depositional models: Am. Assoc. Petroleum Geologists Bull., V. 55, no. 8, p. 1348 1355. Bates, R. L., 1942. Lateral gradation in the Seven Rivers Formation, Rocky Arroyo, Eddy County, New Mexico: Am. Assoc. Petroleum Geologists Bull., v. 26. no. I, p. 80-99. Beede, J. W., 1910, The correlation of the Guadalupian and the Kansas sections: Am. Jour. Sci.. ser. 4. v. 30. p. 131-140. Bretz. J. H., 1960, Bermuda a partially drowned, late mature, Pleistocene karst: Geol. Soc. America Bull., v. 71, no. 12, p. 1729-1754. Cartwright, L. D., Jr., 1930, Transverse section of Permian basin. West Texas and southeast New Mexico: Am. Assoc. Petroleum Geologists Bull., v. 14 no. 8, p. 969-981. Clifton, R. L., 1944, Paleoecology and environments inferred for some marginal Middle Permian marine strata: Am. Assoc. Petroleum Geologists Bull., v. 28. no. 7, p. 1012-1031. 1945, Permian Word Formation, its faunal and stratigraphic correlatives, Texas: Am. Assoc. Petroleum Geologists Bull., v. 29, no. 12, p. 1766-1776. Crandall, K. H.. 1929. Permian stratigraphy of southeastern New Mexico and adjacent parts of western Texas: Am. Assoc. Petroleum Geologists Bull., v. 13, no. 8, p. 927-944. Davis, H. E., chm., 1953, North-south cross section through Permian basin of West Texas: West Texas Geol. Soc. Fay, R. O., 1964. The Blaine and related formations of northwestern Oklahoma and southern Kansas: Oklahoma Geol. Survey Bull. 98, 238 p. 1965, Geology and mineral resources of Woods County, Oklahoma: Oklahoma Geol. Survey Bull. 106, 189 p. Feldman, M. L., Jr.. et a/., 1962, Southwest-northeast cross section Marathon Region tt Midland basin: West Texas Geol. Soc. Pub. 62-47. Flawn, P. T., 1961a. The Marathon area, m The Ouachita system: Texas Univ. Pub. 6120, p. 49-58. 1961b, The subsurface Ouachita structural belt in Texas and southeast Oklahoma, in The Ouachita system: Texas Univ. Pub. 6120, p. 65-81. Galley, J. E., 1958, Oil and geology in the Permian basin of Texas and New Mexico, in Habitat of oil: Am. Assoc. Petroleum Geologists, p. 395 446. Girty, G. H., 1908, The Guadalupian fauna: U. S. Geol. Survey Prof. Paper 58, 651 p. Hayes, P. T.. 1959, San Andres Limestone and related Permian rocks in Last Chance Canyon and vicinity, southeastern New Mexico: Am. Assoc. Petroleum Geologists Bull., v. 43, no. 9, p.2197-2213. Hills, J. M., 1940, Megascopic fossils from the Permian reef trend of West Texas and New Mexico: Jour. Paleontology, v. 14, no. 2, p. 162-163. 1942, Rhythm of Permian seasa paleogeographic

2322

John M. Hilts
Mear, C. E., 1963, Stratigraphy of Permian outcrops. Coke County, Texas: Am. Assoc. Petroleum Geologists Bull., v. 47, no. 11, p. 1952-1962. Meissner, F. F., 1972, Cyclic sedimentation in Middle Permian strata of the Permian basin. West Texas and New Mexico, in Cyclic sedimentation in the Permian basin: West Texas Geol. Soc. Pub. 72-60 (2d ed.), p. 203- 232. Meyer, R. F., 1966, Geology of Pennsylvanian and Wolfcampian rocks in southeast New Mexico: New Mexico Bur. Mines and Mineral Resources Mem. 17, 123 p. Newell, N. D., et al., 1953. The Permian reef complex of the Guadalupe Mountains region, Texas and New Mexico: San Francisco, W. H. Freeman. 236 p. Owen, E. W., 1962, Regional stratigraphic concepts, with some speculative implications, in Contributions to the geology of South Texas: South Texas Geol. Soc, p. 7-17. Peterson, J. A., and H. R. Ohlen, 1963, Pennsylvanian shelf carbonates, Paradox basin, in Shelf carbonates of the Paradox basin: Four Comers Geol. Soc, p. 65-79. Ross. C. A., 1963, Standard Wolfcampian Series (Permian), Glass Mountains, Texas: Geol. Soc. America Mem. 88, 205 p. Roswell Geological Society Stratigraphic Research Committee, 1956, West-east correlation section San Andres Mountains to New Mexico-Texas line, southeastern New Mexico: Roswell Geol. Society. Scobey, W. B., et al., 1951. North-south cross section through Permian basin of West Texas (Platform): West Texas Geol. Soc. Shumard, B. F., 1858 (1860), Notice of new fossils from the Permian strata of New Mexico and Texas: St. Louis Acad. Sci. Trans. V. I, p. 290-297. Shumard, G. G.. 1858 (1860), Observations on the geological formations of the country between the Rio Pecos and the Rio Grande: St. Louis Acad. Sci. Trans., v. 1, p. 273-289. Silver, B. A., and R. G. Todd. 1969, Permian cyclic strata, northern Midland and Delaware basins. West Texas and southeastern New Mexico: .^m. Assoc. Petroleum Geologists Bull., V. 53, no. 11, p. 2223-2251. Skinner, J. W., 1946, Correlation of Permian of West Texas and southeast New Mexico: Avn. Assoc. Petroleum Geologists Bull., V. 30, no. II, p. 1857-1874. Udden, J. A., 1924, Laminated anhydrite in Texas: Geol. Soc. America Bull., v. 35, no. 2. p. 347-354. Van Siclen, D. C , 1969, A depositional model of late Paleozoic cycles on the Eastern shelf, in Cychc sedimentation in the Permian basin: West Texas Geol. Soc. Pub. 69-56, p. 17-27. Vertrees, C. D., et al., 1964, Cross section through Delaware and Val Verde basins from Lea County, New Mexico to Edwards County, Texas: West Texas Geol. Soc. Pub. 64-49. Walter, J. C , Jr., 1953, Paleontology of the Rustler Formation, Culberson County, Texas: Jour. Paleontology, v. 27, no. 5, p. 679-702. Wanless, H. R., 1967, Eustatic shifts in sea level during the deposition of late Paleozoic sediments in the United States, in Cyclic sedimentation in the Permian basin: West Texas Geol. Soc. Pub. 69-56, p. 41-54. Wilson, J. L., and O. P. Majewske, I960, Conjectured middle Paleozoic history of central <ind West Texas: Texas Univ. Pub. 6017, p. 65-86. Woolnough, W. G., 1937. Sedimentation in barred basins, and source rocks of oil: Am. Assoc. Petroleum Geologists Bull., v. 2 l , n o . 9 , p. 1101-1157.

study: Am. Assoc. Petroleum Geologists Bull., v. 26, no. 2. p. 217-255. 1963, Late Paleozoic tectonics and mountain ranges, western Texas to southern Colorado: Am. Assoc. Petroleum Geologists Bull., v. 47, no. 9, p. 1709-1725. 1970, Late Paleozoic structural directions in southern Permian basin. West Texas and southeastern New Mexico: Am. Assoc. Petroleum Geologists Bull., v. 54, no. 10, p. 18091827. Jacka, A. D., et al., 1967, Guadalupian depositional cycles of the Delaware basin and Northwest shelf, in Cyclic sedimentation in the Permian basin: West Texas Geol. Soc. Pub. 6956, p. 152-196. Jones, C. L., 1954, The occurrence and distribution of potassium minerals in southeastern New Mexico, in Guidebook of southeastern New Mexico: New Mexico Geol. Soc. 5th Field Conf Guidebook, p. 107-112. Jones, T. S., chm., 1949, East-west cross section through Permian basin of West Texas: West Texas Geol. Soc. King, P. B., 1930, The geology of the Glass Mountains, Texas: pt. 1, Descriptive geology: Texas Univ. Bull. 3038, 167 p. 1942, Permian of West Texas and southeastern New Mexico: Am. Assoc. Petroleum Geologists Bull., v. 26, no. 4. p.535-763. 1948, Geology of the southern Guadalupe Mountains, Texas: U.S. Geol. Survey Prof. Paper 215, 183 p. King, R. E., 1945, Stratigraphy and oil-producing zones of the pre-San Andres formations of southeastern New Mexico: New Mexico School Mines Bull. 23, 31 p. Kinsman, D. J. J., 1969, Modes of formation, sedimentary associations, and diagnostic features of shallow-water and supratidal evaporites: Am. Assoc. Petroleum Geologists Bull., V. 53, no. 4, p. 830-840. Kroenlein, G. A., 1939, Salt, potash, and anhydrite in Castile Formation of southeast New Mexico: Am. Assoc. Petroleum Geologists Bull., v. 23, no. 11, p. 1682-1693. Landes, K. K., 1963, Origin of salt deposits, in Symposium on salt, Northern Ohio Geol. Soc, p. 3-9. Lang, W. B., 1937, The Permian formations of the Pecos Valley of New Mexico and Texas: Am. Assoc. Petroleum Geologists Bull.,v.21,no. 7, p.833-898. Lewis, F. E., 1941, Position of San Andres Group, West Texas and New Mexico: Am. Assoc. Petroleum Geologists Bull., v. 25, no. 1, p. 73-103. Lloyd, E. R., 1929, Capitan limestone and associated formations of New Mexico and Texas: Am. Assoc. Petroleum Geologists Bull., v. 13, no. 6, p. 645-658. , 1938, Theory of reef barriers (abs.): Am. Assoc. Petroleum Geologists Bull., v. 22, no. 12, p. 1709. Marcou, J., 1855, Geological notes of a survey of the country comprised between Preston, Red River, and El Paso Rio Grande del Norte, in John Pope, Report of exploration near the 32nd parallel: U. S. 33rd Congress, 1st sess.. House Exec. Doc. 129, V. 18, pt. 2. Maher, J. C , ed., 1960, Stratigraphic cross section of Paleozoic rocks West Texas to northern Montana: Am. Assoc. Petroleum Geologists. McKee, E. D., and W. J. Breed, 1969, The Toroweap Formation and Kaibab Limestone, in The San Andres Limestone, a reservoir for oil and water in New Mexico: New Mexico Geol. Soc. Spec. Pub. 3, p. 12-26. et a/., 1967, Paleotectonic investigations of the Permian System in the United States: U. S. Geol. Survey Prof. Paper 515, 271 p.