Abstract The Guadalupe and Delaware mountains of west Texas and southeastern New Mexico contain superb outcrops of the Capitan depositional system (Fig. 1). The Capitan depositional system as used here includes the reef/forereef of the Capitan Formation, shelfal equivalents (Seven Rivers, Yates and Tansill formations), and the basinal equivalent (Bell Canyon Formation) (King, 1948; Newell et al., 1953; Hayes, 1964) (Fig. 2). Exposures in the Guadalupe Mountains are immense (1000 m vertical; 5-10 km in dip direction; 70 km along strike), and have suffered minimal structural deformation. The Capitan depositional system of the Guadalupe Mountains has been visited and studied by many of the world’s premier geologists. In addition, it serves as one of the primary training grounds for earth scientists from the petroleum industry and academia. As a result, the Capitan depositional system has been the focus of numerous stratigraphic, facies, paleontologic, and diagenetic studies, many of which have become classic works in their respective fields. Many facets of the Capitan system have been and continue to be controversial, and much excellent geological research is still being done on the Capitan reef and related units. An informal meeting of stratigraphers, sedimentologists, and paleontologists was held in Carlsbad, New Mexico, and the Guadalupe Mountains in October of 1996. This international conference was cosponsored by the University of Texas at Austin and Cambridge University in England. The purpose of the conference was to let Capitan researchers meet each other and to share

Abstract The spectacular mixed siliciclastic/carbonate exposures of the Guadalupe Mountains include 30 high-frequency sequences (HFS) that stack together to form six composite sequences (CS), the CS9 through CS14. These sequences include carbonate ramps and reef-rimmed platforms as well as basin-restricted lowstand sequences. The Capitan Formation represents the shelf-margin and slope facies tracts of the upper 12 HFS. The Capitan is examined in the context of this late Leonardian-Guadalupian ramp-to-rimmed-shelf system by focusing on extrinsic controls on platform development. Eustatic changes initiate and punctuate larger scale changes in platform evolution. Rapid shifts of large magnitude, such as the latest Leonardian (L7-L8 HFS) eustatic rise, are a first-order control on platform architecture and reef formation. The model for the late Permian eustatic curve based on the present stratigraphic framework suggests that by the time the Capitan was established, eustatic amplitudes were in the range of 20 m or less. This amplitude variation does not cause major shifts in the shelf-margin location but is sufficient to affect critical accommodation factors that influence reef depth and faunal composition. Antecedent topography, whether of tectonic, depositional, erosional, or compactional origin, is the critical parameter in controlling the timing and development of the Capitan and other buildups in the Leonardian-Guadalupian sequences, as well as the primary control driving the ramp-to- rimmed-shelf transition. The shelf-slope break, whether a ramp or rimmed shelf, is only one of numerous geometric parameters that can be used to describe the dynamic evolution of carbonate platforms. Changing styles of carbonate-platform progradation and aggradation, which are responses to changes in platform and basin accommodation and sediment-supply, can be captured using P/A (ratio of progradation to aggradation for a given chronostratigraphic unit) and SMP/A (shelf-margin progradation/aggradation ratio) ratios. P/A values > 25 are characteristic of ramps and lowstand wedges, whereas P/A values < 25 are indicative of either transgressive-dominated ramps or reef-rimmed margins. SMP/A values within the Capitan- equivalent sequences can be used to document the complex but systematic and predictable progradational-aggradational-progradational response of the shelf margin to changing base level. Within the high-frequency sequence framework, other analytical tools, including facies tract substitution and facies proportions, can be used to better constrain interpretations of the dynamic water-depth setting of the Capitan margin and factors controlling its position on the profile. This holistic approach, which draws on relationships from outer-shelf and shelf-crest facies tracts in the interpretation of the Capitan margin, demonstrates the power of a stratigraphic framework for sedimentologic analysis.

Abstract The Capitan depositional system was studied in the subsurface using seismic and well data from the northeastern Delaware basin. Seismic data of the Capitan depositional system show characteristics that include (1) a massive prograding reef/slope, (2) back-reef/shelf reflectors that dip and diverge basinward before disappearing into the massive reef, and (3) layered bottomset beds that thicken basinward by the addition of younger reflectors. A wireline log cross-section of nearby wells illustrates the stratigraphy in more detail than the seismic line. Basinward-dipping shelf strata are interbedded sandstones and carbonates that diverge and pass basinward into massive carbonate of the reef. Correlative markers within the massive reef are difficult to find. Slope carbonate beds thin and basinal siliciclastics thicken toward the basin. Bottomset beds in the basin consist of interbedded sandstones/siltstones and low-porosity carbonates. This subsurface stratigraphy is very similar to outcrop stratigraphy described in the Guadalupe Mountains. Lithologic differences between outcrops and their subsurface equivalents are due largely to variations in dolomitization and evaporite dissolution on outcrops. Distribution of porosity in the Capitan depositional system is closely related to depositional facies. Back-reef sandstones and some shelf carbonates adjacent to the reef have good porosity and moderate permeability, but porosity and permeability in those strata generally decrease landward. The subsurface Capitan reef has moderate porosity and high permeability and is a regional aquifer. Carbonate beds in the basin are generally not porous, but some basinal sandstones filling elongate channels have good porosity and moderate permeability. Hydrocarbons are not present in the Capitan reef because it does not occur in a setting that allows structural or stratigraphic closure and/or isolation from active meteoric aquifers. Many oil fields (10-400 million barrels recoverable) occur in back-reef equivalents of the Capitan reef, primarily the Seven Rivers and Yates formations, on the Northwestern Shelf and western edge of the Central Basin Platform. Those reservoirs are generally in stratigraphic or combination stratigraphic-structural traps, where porous and permeable sandstones pass up-dip into impermeable sandstones/siltstones, carbonates, and/or evaporites. Oil also occurs in channelized basinal sandstones equivalent to the reef, but the basinal fields have <5-30 million barrels of oil recoverable, and hence are generally smaller than those of the back-reef.

Abstract The apparent prevalence of meter-scale lithologic cyclicity in shallow-marine carbonate and mixed carbonate-siliciclastic platform strata has led to the widely accepted paradigm that such meter-scale cycles are the fundamental building blocks of platforms. This view commonly is challenged, however, on the grounds that lithologic cyclicity or “ideal lithologic cycles” are difficult to demonstrate from statistical evaluations (for example, Markov chain analysis) of facies successions. This issue is of major importance in determining the accuracy of high-resolution chrono- stratigraphic correlations using present sequence stratigraphic techniques. In this study we have used traditional and modified Markov chain analysis to evaluate the presence of cyclicity in backreef strata, the Seven Rivers Formation, of the Capitan reef. Traditional Markov analysis, similar to that used in previous studies, fails to verify the presence of cyclicity. Through the addition of stratigraphic data on the distribution of significant surfaces (subaerial exposure and hiatal) and the two-dimensional distribution of facies bodies, however, Markov analysis verifies the presence of cyclicity and even allows the statistical identification of potential “ideal cycles” in the system. Thus, meter-scale cyclicity does exist in this system, with the implication that previous studies may have failed to incorporate stratigraphic information necessary for an adequate evaluation of the existence of cyclicity.

Abstract The Guadalupian Series, with its type section in the Guadalupe Mountains, west Texas and New Mexico, has generally become the accepted worldwide standard for the Middle Permian. The Cisuralian of Russia is the accepted standard for the Early Permian; and the Lopingian of China has been accepted as the Late Permian standard. Most recently, the task of securing this threefold worldwide standard has been pursued by a host of workers in many specialties in order that no major gaps or overlaps are left for chronostratigraphic correlation. Following established international procedures for defining a chronostratigraphic boundary, the base of the Guadalupian Series is recognized at the stratigraphicaUy lowest occurrence of the conodont Jinogondolella nankingensis in a section recording continuous deposition in the middle of the El Centro Member of the Cutoff Formation, Roadian Stage, Guadalupe Mountains. This designation is in basic agreement with the ranges of important species of three other major faunal groups: the ammonoids, brachiopods, and fusulinaceans. Designation of a formal top of the Guadalupian Series has remained more elusive inasmuch as the final establishment of the base of the overlying Lopingian in China has remained in a rather unsettled state until recently. Enough is now known to insure that the Lopingian base will fall chronostratigraphicaily somewhere between the upper Lamar Limestone, Bell Canyon Formation, and the base of the Castile Formation of the Delaware Basin. This interval includes the so-called “post-Lamar beds” of the upper Bell Canyon Formation. As a means of equating the lithostratigraphy to the chronostratigraphy, we describe herein the Reef Trail Member of the upper Bell Canyon Formation by establishing its type section at McKittrick Canyon, Guadalupe Mountains, as a formal replacement name for the informal “post-Lamar beds.” We also identify the fusulinacean and conodont faunas from composited sections of Lamar Limestone and Reef Trail members, and extend their correlations biostratigraphically from the Guadalupes to the Apache and Glass mountains. The McKittrick Canyon Limestone is a new name given to the “Middle” Limestone (Brown, 1996), following his recognition that this limestone had for too long been misidentified as the McCombs Limestone. Finally, correlation of faunal zones established in the Delaware Basin from the three mountainous areas are extended globally in an attempt to reconcile chronostratigraphic problems.

Abstract The outer shelf sediments of the Capitan Reef Complex (Permian, Guadalupian) dip towards the basin at variable angles of up to 10°. High-resolution correlation of field logs from several canyons in the Guadalupe Mountains, New Mexico, has revealed a geometry that is most likely to have resulted from differential subsidence with the greatest rate of subsidence occurring at the shelf margin. During deposition of the upper Seven Rivers, Yates, and lower Tansill Formations, the reef accreted in several alternating phases of progradation and aggradation in response to fluctuating sea level. As the reef aggraded, loading of the underlying fore-reef and basin deposits gave rise to episodic compaction-induced differential subsidence, which tilted the outer shelf strata in a basinward direction. This process was controlled by the variable compactability of the forereef and basin lithologies as well as by the loading of the aggrading reef. Episodic differential compaction also influenced deposition of outer shelf sediments and induced small-scale relative sea-level changes near the shelf margin that were not experienced by the inner shelf. Mechanical failure of lithified outer shelf sediments occurred as the reef prograded over the compacting fore-reef and basin deposits. Water and liquefied sand from compacting basin sediments appear to have been forcefully injected upwards through mechanically initiated fractures, and the effects of dissolution are apparent along some fracture margins. The importance of differential compaction as a control on platform development must be recognized if the effects of sea level are to be properly understood.

Abstract Outer shelf carbonates in the Tansill Formation in Dark Canyon include the following inferred depositional environments landward of the Capitan reef: (1) subtidal: bioclastic wackestones, packstones, and graded and locally cross-stratified grainstones; and biostromes; (2) peritidal flats: fenestral and locally desiccated mudstone; and (3) shoreface: admixed subtidal and peritidal deposits. Two sequences are recognized in the section based on biotic diversity and parasequence thickness and facies-stacking patterns, and the boundary between them is at or near the base of the Ocotillo Member. Maximum flooding of the platform occurred during deposition of sequence 1 in the lower part of the middle Tansill. Environments were biostromes, mainly high energy shallow subtidal packstones and grainstones, and associated peritidal islands in the early HST in this sequence. Precipitation of abundant marine cements in these rocks may have been promoted by active seawater pumping through the sediments on this wide and shallow shelf. Microbial activity in the grainstones may have been promoted by restricted circulation around associated peritidal islands. In contrast, environments in the late HST of this sequence and in sequence 2 were more restricted in terms of shelf width and energy, and the rocks are dominantly micritic and contain little marine cement. The principal difference between depositional facies in the Carlsbad Embayment and along the Reef Escarpment appears to be the presence of patch reefs in the former area, which were deposited on an outer shelf that was wider and of more open circulation than to the south. Inferred marine cements are dominated by prismatic calcites, with microdolomite inclusions and some recognizable radiaxial-fibrous habit, interpreted as former high-Mg calcite. The 8 I8 0 and 8 13 C compositions of the least-altered of these cements (−1.6 %c and 5.8 %o PDB, respectively) suggest precipitation from marine pore fluids. Former aragonitic cement of similar isotopic composition is a volumetrically minor, first-generation marine cement in these rocks. Dominance of inferred former high-Mg calcite cement contrasts that in coeval platform-margin patch reef facies, the Capitan reef, and in pisolites on the shelf crest in which former aragonite marine cement dominates. Mean 8 18 0 composition of the dolomite that replaced peritidal deposits (0.1 %c PDB) suggests that it precipitated from ambient marine fluids of elevated salinity. Mineralogic stabilization of earlier diagenetic phases likely attended precipitation of equant calcites (spar 1) in the rocks, which is interpreted to have occurred in a meteoric phreatic system during the latest Guadalupian and/or early Ochoan. A second generation of equant cements (spar 0) is inferred to have precipitated later, during the Ochoan or later, in a deeper meteoric-dominated groundwater system. Replacement of syndepositionally precipitated evaporites by poikilotopic calcite is the last diagenetic phase present and, based on stable oxygen-carbon isotope data, appears to have attended meteoric dedolomi- tization during the Tertiary.

Abstract Although the rate of accumulation, topography, and stratigraphic setting of the Permian Capitan reef are similar to that of modem coral reefs, both the ecology of the reef communities and the relative sources of carbonate in their final geological expression are profoundly different. Like modem coral reefs, the Capitan reef was constructed by a patchy mosaic of many different communities, but most were dominated by relatively short-lived aggregations of metazoans (particularly calcified sponges and bryozoans), and some were strongly differentiated into distinct open surface and cryptic subcommunities. Although our understanding of the distribution of these communities is poor, much of the deeper water Middle Capitan reef framework appears to have been constructed initially by a scaffolding of large frondose bryozoans. This framework created cryptic niches that were colonized by a diverse community dominated by solitary (cylindrical) sphinctozoan sponges and bryozoans that grew downwards from the roofs of the cavities. Bathymetrical ly shallow areas of the Middle and Upper Capitan reef were characterized by large, platy calcified sponges. In the Upper Capitan, the platy inozoan sponge Gigantospongia discoforma reached up to 2 m in diameter, and individuals projected from the reef slope to form the ceilings of substantial open cavities. The undersurfaces of these large sponges were colonized by an extensive cryptic community, including downward-growing branching sphinctozoans up to 0.5 m long. In the absence of both the physical and biological destructive forces prevalent on modem reefs, much of the relatively fragile Capitan reef-building community remained in growth position. It is unlikely, however, that such a community could have withstood highly turbulent and agitated waters. Additional strength was imparted to the framebuilding community by encrustations of Archaeolithoporella and Tubiphytes, followed by the precipitation of abundant micrite of probable microbial origin. This cavernous framework was subsequently partially filled with internal sediment and Areheolithoporella, and botryoidal aragonite and other submarine cements. These cements and sediments account for up to 70% of the reef rock. So unlike modem coral reef carbonates, most of the Capitan Reef was derived from probably inorganic sources, and moreover none of the organic carbonate—with the exception of phylloid algae and possibly Archaeolithoporella —is interpreted to have formed as a result of light-enhanced (photosynthetic) calcification. Such observations suggest that models that simulate modem coral reef growth, which is dependent upon the distribution of light, may not be applicable generally to Paleozoic reefs.

Abstract Facies relations determined from outcrop studies and photomosaics have been used in conjunction with reef maps, polished slabs, and wicrofacies data to better understand the depositional facies, quantitative composition, and control mechanisms of the upper Capitan massive, which differs from the outer shelf in its biotic composition and unbedded nature. A three-stage model of the seaward shelf is established comprising (I) a sponge reef/algal cement/phylloid algal stage, (2) a Tubiphytes stage, and (3) prograding cyclic outer shelf beds with isolated reefbuilders Stage 1 is characterized by progressive shallowing as evidenced by a shift from aggradation to progradation, by changes in boundstone composition, and finally by disintegration of the biogenic framework. Stage 2. dominated by Tubiphytes obscurtis. bryozoans, microbes, and small reefbuilders. exhibits a lateral conation triggered by the seaward dipping outer shelf. Stage 3 is composed of cyclic outer shelf grainstones; scattered sponges, and Tuhtphytes microbial-level bottom communities are restricted to few horizons. The entire depositional sequence is controlled by a third-order sea-level fall as evidenced by exposure horizons. This sequence is in turn superimposed by three high-frequency cyclcs. Water was only deep enough for typical Permian reef types during stage I, which produced the thickest depo- silional unit Relative water depth decreased significantly during stage 2, leading to a small, tabular reef constructed by an impoverished fauna. During stage 3, water was too shallow for reef communities. In contrast to modem reefs, shallowing beyond a distinct level in subtidal depth limited reef growth. In addition, salinity fluctuations indicated by cyclic cementation caused the demise of the Tubiphytes microbial-level bottom communities. Our quantitative data from the reef maps suggest a reinterpretation of existing models. The most important constructional clement is the microframework (76.7% average coverage), a consortium of low-growing reelhuilders and manne-phreatic cements. In contrast to previous interpretations, neither macro-sponges nor syndepositional cement predominated, making the upper Capitan massive a poor analog for both modern well- skeletonized metazoan and Prccambrian cement reefs.

Abstract The origin of the Massive Member of the Capitan Formation in the Guadalupe Mountains of West Texas and New Mexico is controversial. It has been inteipreted in many different ways, including as a barrier reef, a deep water skeletal wackestone mound, a cement boundstone mound, and a linear complex of buildups. The reality is that it is not simply one or the other, but a complex with characteristics of all of the above — a reefal complex that changes both in time and space. The Massive Member of the Capitan Formation contains the components of a framework reef: frame-building organisms — bryozoans, platy sponges, and Tubiphytes', binding organisms — Archaeolithoporella and microbialite; infilling internal sediment; and marine cement. Much of the Capitan was probably deposited below wave base as a diverse framework reef. The profile of the Capitan reef varied throughout its deposition. At the mid-point of Capitan development, the slope of the reef was very steep (80° or more) and the living reef extended to a depth of approximately 140 m. The organisms that lived on the reef varied with depth. The shallowest portions of the reef were inhabited by Collenella and a diverse assemblage of other organisms. In slightly deeper water, large platy sponges formed meter-scale over-hangs and cavities, which were inhabited by a cryptic community dominated by sponges. In the deepest parts of the reef, frondose bryozoa created decimeter-scale framework and internal cavities inhabited by cryptic pendant sponges. These internal cavities were filled by a succession of encrustations and cements, beginning with thick layers (1-3 cm) of microbialite followed by profuse botryoidal aragonite (up to tens of centimeters thick), thin (1—4 mm) layers of radiaxial fibrous cal- cite, and much later, large crystals of meteoric spar. The Capitan reef influenced deposition in surrounding environments. Throughout its growth the Capitan was a “sediment factory” that generated prolific amounts of sediment that accumulated in the fore-reef. Study of the upper Capitan at Walnut Canyon suggests that the living Capitan reef probably created a depositional high that modified sedimentation in the lee of the reef on the outer shelf. Study of the middle Capitan and equivalents at McKittrick Canyon suggest that at other times circulation on the outer shelf was unrestricted by the reef escarpment and its living surfaces. Many questions about the Capitan remain unresolved, for example, what was the response of the reef to sea-level change? How did the reef profile change through time? Is there consistent evidence of exposure in the reef? Does diagenesis vary throughout the reef? How diverse is the biota in the lower and middle Capitan, and does biotic diversity change through time?

Abstract The Capitan Formation represents a Permian (Guadalupian) reef complex deposited on the margin of the Delaware Basin. The reef was cemented by aragonite botryoids, radiaxial-fibrous (RFC) and fascicular-optic calcite (FOC), radiaxial-prismatic calcite (RPC), dolomite, and late calcite spar. Variability of fibrous cement textures correlates with the presence or absence of dolomite overgrowths. Zones that contain subequal amounts of RFC and FOC are characterized by radiaxial prismatic terminations lacking dolomite overgrowths, but contain <10 pm diameter dolomite microinclusions, In contrast, zones with <20% FOC and >80% RFC have dolomite cement overgrowths and contain dolomite microinclusions 10-100 pm in diameter. Trace element concentrations from ICP and stable isotope compositions correlate with stratigraphic depth. Mn concentrations in RFC and FOC increase from an average of 25±2.9 ppm in the upper Capitan Formation to 62±31 ppm in the lower Capitan, whereas average Mg concentration increases from 1.45±0.81 to 2.40±1.37 mole % MgC0 3 and average Sr concentration decreases from 210±72 ppm to 63±22 ppm. RFC and FOC also contain covariant carbon and oxygen isotope ratios, which trend towards isotopically lighter values from compositions of 5 18 0=-2.8 %c, 8 13 C=5.2%o in the Middle Capitan. Microprobe analyses on individual RFC and FOC crystals include the <10 pm dolomite microinclusions. Point analyses traversing crystals from substrate to termination define apparent trends of increase from 1 to 4 mole % MgCOj, and reflect relative abundance of dolomite microinclusions. Strontium and MgC0 3 do not covary, although both generally increase. Combinations of RFC, FOC, and RPC form complex cement stratigraphies. Anhedral dolomite microinclusions may be either primary (penecon- temporaneous with host calcite), or secondary (formed during later diagenesis). Euhedral dolomite inclusions within calcite resemble cement overgrowths. Changes in dolomite and kaolinite microinclusion abundance in these cements may reflect porewater changes during early diagenesis. Early diagenesis was probably driven by reef growth (aggradational and progradational), and sea-level change.

Abstract The diagenesis of the Capitan forereef can be divided into four overlapping episodes: 1) early marine diagenesis; 2) early burial diagenesis involving normal marine to hypersaline fluids; 3) late burial diagenesis; and 4) uplift related diagenesis. Early meteoric diagenesis was not identified in the Capitan forereef. Early marine diagenesis is limited to minor isopachous cement in upper forereef grainstones. The greatest amount of diagenesis in the Capitan forereef occurred during early burial diagenesis in marine to hypersaline pore fluids. The lower forereef facies was largely altered by marine pore fluids; fabrics include moldic porosity, fine blocky spar and overgrowth cementation, silicification, and rare aragonite neomorphism. The upper forereef facies, however, was nearly completely dolomitized by mesosaline fluids mainly derived by seepage reflux from the near-backreef carbonate lagoon facies. The middle forereef facies contains both styles of alteration. During and/or subsequent to dolomitization, hypersaline fluids completely cemented the reef and forereef with evaporites (gypsum and/or anhydrite). Deeper burial processes include stylolites and partial recrystal I ization of the early fabric-preserving dolomite to a fabric-destructive dolomite. During uplift, erosion of overlying Ochoan evaporites allowed the influx of meteoric water into the Capitan Formation. This led first to hydration of anhydrite to gypsum and kaolinitizarion of feldspars and later to complete dissolution of evaporites and precipitation of coarse blocky spars II and III in the resulting porosity. Most diagenetic models focus on timing of alteration and the fluid composition. In the case of the Capitan forereef facies, however, the sedimentology was a major factor controlling the distribution of diagenetic fabrics. For example, debris-flow deposits (unsorted packstones to rudstones) are partially to completely dolomitized while interbedded turbidity-current deposits (graded packstones to wackestones) are limestone with fine blocky spar and aragonite alteration. The changing patterns of diagenesis from the upper forereef to the lower forereef are at least partially controlled by the change from a predominance of debris-flow deposits in the upper forereef to turbidity-current deposits in the lower forereef.

Abstract Caves exist in the Capitan Formation in the Guadalupe, Apache, and Glass Mountains, southeastern New Mexico and West Texas. Four episodes of karsting have been identified in the Guadalupe Mountains: (1) Stage 1 fissure karst (Late Permian), (2) Stage 2 spongework karst (Mesozoic), (3) Stage 3 thermal karst (Miocene), and (4) Stage 4 sulfuric acid karst (late Miocene to present). The last sulfuric acid episode was responsible for the large cave passages and for the distinctive deposits within these caves (gypsum blocks; native sulfur; the pH-indicator minerals endeliite, alunite, and natroaiunite; and the uranium minerals tyuyamunite and metatyuyamunite). The hydrogen sulfide responsible for forming sulfuric acid in the Capitan reef originated from hydrocarbons in the Delaware Basin. Stage 4 caves in the Glass Mountains are also of sulfuric acid origin, showing that the basin probably degassed hydrogen sulfide around its entire margin. Caves also exist in the subsurface Capitan on the eastern and northern sides of the Delaware Basin. San Simon Swale/Sink on the east and breccia pipes/domes on the north may stope down to the cavernous Capitan. The same sulfuric acid mechanism that formed caves in the exposed Capitan may have also been partly responsible for caves in the subsurface Capitan.