Very high-resolution seismic facies, classified, mapped, and interpreted from 3.5 kHz echograms, reveal that turbidity-current, mass-transport, and bottom-current depositional processes have all contributed to the regional sediment distribution in the intraslope basin province of the northwest Gulf of Mexico. Piston cores from these deposits confirm the interpretations of the processes. Turbidity currents transport sands into the intraslope mini-basins via channels and canyons. A few turbidity-current pathways, such as Bryant Canyon, allow extensive volumes of terrigenous sediment to bypass through many mini-basins and be deposited beyond the Sigsbee Escarpment to form large submarine fans, such as Bryant Fan. Bryant Fan is a large mud-rich fan that extends hundreds of kilometers from the mouth of Bryant Canyon but has only one meandering channel on the modern seafloor that extends down the length of the fan. In contrast, the much smaller Rio Grande Submarine Fan is deposited on a plateau area of the continental slope. Prolonged 3.5 echo character and numerous small, unleveed channels suggest this is a sand-rich fan with a braided channel system. These two submarine fans appear to present unique architectural and growth patterns not previously described in the numerous fan descriptions of W.R. Normark or other workers. Thus, these two fans appear to represent two new types of fans, which may be related to the complex structures of the intraslope basin province. Mass-transport deposits (MTDs) are ubiquitous throughout the mini-basins. Extensive areas affected by MTDs also occur along the upper continental slope and at the base of the eastern portion of the Sigsbee Escarpment. Piston cores confirm that the majority of MTDs are debris flows, which are characterized by mud clasts of variable size, shape, and color. Most have a muddy matrix, but sandy debris flows also occur in a few mini-basins. Some cores show deformation, folds, and faults that indicate slump or slide deposits. The East Breaks Slide Complex is the largest MTD complex and extends downslope from the shelf edge for >100 km off central Texas. The western portion of the complex is slump and/or slide blocks and debris flows. In contrast, the proximal part of the eastern portion of the complex is characterized by a modern leveed turbidite channel system. However, extensive MTDs underlie the channel-levee deposits and occur at the seafloor on the distal part of the eastern portion of the complex. Three large regions of migrating sediment waves occur on the Sigsbee Abyssal Plain and eastern Bryant Fan and appear to have been formed by circulation of the Loop Current. Sediment waves also occur locally at the base of the Sigsbee Escarpment in conjunction with previously reported erosional furrows.
Dedication: This paper is dedicated to the memory of two exceptional scientists with whom the senior author (JED) had the good fortune to interact during their careers: William R. Normark and Charles D. Hollister. I met Bill Normark in the early 1970s when I was revising my first paper on Amazon Fan based on the excellent review he had just provided. He had already become one of the foremost authorities on modern fans. After that, I had the good fortune to interact with Bill numerous times during his career including at meetings, collaboration on publications, and describing Amazon Fan cores together during Ocean Drilling Program Leg 155. Early on I realized that Bill would always convey important ideas, suggestions, observations, or criticisms that had never occurred to me. I came to expect this whenever we interacted; I was never disappointed. Our field greatly misses his expertise and leadership. Charley Hollister was finalizing his landmark thesis on contourites at Lamont Geological Observatory (now Lamont-Doherty Earth Observatory) under Bruce C. Heezen, when I began my graduate studies under Bruce. As a senior graduate student, Charley provided invaluable guidance for operating in the Lamont environment both in the lab and at sea. Most importantly, he introduced me to the method of echo-character mapping with 12 kHz records, which he developed for his contourite studies. Charley and Bruce suggested that I use the same method, except with 3.5 kHz records, in my thesis research. This inspired me to develop a classification for 3.5 kHz echo character and to use echo-character mapping techniques in many different basins during my career; the most recent study is presented below. Charley was not only an exceptional scientist but also a great promoter and fundraiser within the marine geology community. His abilities and advice greatly helped me and others, who mainly relied on soft money. His untimely accidental death left a huge gap in the community.
The northern Gulf of Mexico Intraslope Basin Province off Texas and Louisiana is one of the most studied hydrocarbon provinces in the world. Previous studies have mainly addressed Cenozoic sedimentation and the tectonic history in conjunction with hydrocarbon exploration. This region has a very distinctive morphology, which is characterized by numerous diapiric features and intervening mini-basins caused by the presence of thick (up to 8 km), allochthonous salt masses (Louann Salt) overlain by Cenozoic sediments (e.g., Diegel et al., 1995; Hudec and Jackson, 2006; Hudec et al., 2013). The complex interactions between sediment loading, sediment deformation, and salt deformation form these structures through ongoing regional gravity-tectonic processes as the sediments of the continental margin move slowly downslope (e.g., Winker, 1982). These processes create growth faults in the zone of extension on the continental shelf, mini-basins in the zone of translation across most of the slope and thrust faults (toe thrusts) in the zone of compression at the base of slope (Sigsbee Escarpment). Mini-basins formed by these gravity-tectonic processes are the site of fill-and-spill turbidite systems. The turbidity currents flowed from mini-basin to mini-basin and deposited thick ponded sands in many basin floors (Galloway et al., 2000; Combellas, 2003; Combellas-Bigott and Galloway, 2006; Galloway, 2009). Several studies have documented the stratigraphic evolution of various mini-basins (e.g., Armentrout et al., 1991; Prather et al., 1998; Weimer et al., 1998a, 1998b; Twichell et al., 2000).
In contrast to these numerous studies, relatively few studies have focused on the latest Quaternary depositional processes of the intraslope basins. However, latest Quaternary or “modern” analogs derived from such studies can be extremely useful for developing accurate exploration models that predict sediment facies and sand-body geometries of more deeply buried productive reservoirs. In addition, a better understanding of the depositional processes and lithologies of the uppermost seafloor (upper 10–50 m) promotes improved risk assessment for selecting safe placement of seabed structures related to the hydrocarbon production, including platforms, templates, piles, and pipelines. The major objective of this study is to understand the regional distribution of the various deep-water depositional processes and their sedimentary facies. This study was limited to deep-water areas seaward from the present continental shelf edge to the Sigsbee Abyssal Plain (Fig. 1). A wide variety of high-resolution data sets (seismic, side-scan sonar, bathymetric swath maps, and shallow piston cores; Fig. 1) were utilized to achieve the following scientific objectives: (1) examine and describe all available piston cores in the Gulf of Mexico Intraslope Basin Province and define the depositional processes; (2) construct a regional 3.5 kHz echo-character (seismic-facies) map of the intraslope basin province to interpret and map the regional distribution of depositional processes; and (3) correlate (i.e., “ground truth”) the sediment facies of piston cores with the 3.5 kHz echo character (seismic facies) to present an integrated analysis of the core and seismic data to reveal the details of the depositional features present (e.g., submarine fans, mass-transport deposits, and contourite deposits), their architecture (e.g., channels, overbank deposits, lobes, slumps, debris flows, and sediment drifts), their depositional processes (e.g., turbidity currents, bottom currents, and mass-transport deposits), and, ultimately, their sediment-facies distributions (i.e., sand vs. shale). As deep-water hydrocarbon exploration has progressed during the past 40 years, structural and sedimentary details of numerous continental margins around the world have revealed that many of these margins have been subjected to gravity-tectonic processes similar to the Gulf of Mexico. For example, a less detailed study of the Niger Delta continental margin (Damuth, 1994) revealed similar deformation by gravity tectonics and mini-basins containing sedimentary deposits similar to those described in the present study. Thus, our study here should be quite relevant for predicting sedimentation processes in various other intraslope basin provinces worldwide.
This paper (and volume) is dedicated to William R. Normark as a tribute to his enormous contribution to understanding deep-sea sedimentation processes, and especially the large number of deep-sea fan studies that he and his various co-workers (including the senior author) have detailed in the past 45 years. His early studies of modern fans off California led to the first model for modern fans (Normark, 1970), and subsequent studies continued to provide increasingly more details on fan architecture, sediment facies, and growth patterns for California fans (e.g., Normark and Piper, 1972; Normark et al., 1979; Normark and Hess, 1980; Piper and Normark, 1983; Normark and Piper, 1991; Piper and Normark, 2001; Fildani and Normark, 2004). He also conducted or collaborated on important studies of many other fans around the world (e.g., Mississippi Fan, Normark et al., 1986; Amazon Fan, Normark et al., 1997) and attempted to synthesize results of various fan studies to produce better models for submarine fans and related turbidite systems (e.g., Normark, 1978; Bouma et al., 1985; Normark et al., 1993; Piper and Normark, 2001; Piper and Normark, 2009). Our study of the Gulf of Mexico Intraslope Basin Province builds on and adds to these previous studies. The fan syntheses in the previous studies (e.g., Piper and Normark, 2001) do not include any submarine fans from margins with intraslope basins, such as our Gulf of Mexico study area, which have been highly deformed by gravity-tectonic processes. We provide new initial descriptions of two deep-sea fans, the Bryant and Rio Grande fans, which have architectures and growth patterns that are apparently unique and are different than fans described in Normark’s previous studies. In addition, some intraslope basin deposits in our study area have been termed (rightly or wrongly) fans by some workers (e.g., the Trinity Brazos Turbidite System; Expedition 308 Scientists, 2005; Pirmez et al., 2012; Prather et al., 2012). Our core and seismic descriptions of many basin deposits presented here provide data to further evaluate the origins of these deposits (e.g., turbidite fans or mass-transport processes).
3.5 kHz Echograms (High-Resolution Seismic Profiles)
Precision depth recordings (echograms) recorded at 3.5 kHz can achieve acoustic penetration of up to 150 m below the seafloor, although penetration is generally <50 m. This frequency can resolve beds of ∼12 cm in thickness and, thus, provide very high-resolution seismic facies of the uppermost seafloor. Previous studies have shown that the 3.5 kHz data commonly provide extremely well-imaged examples of the seismic facies, depositional features, and stratal geometries of the modern seafloor (Damuth, 1975a, 1978; Damuth and Hayes, 1977; see Damuth, 1980, for a review and references). These echograms also complement the piston-core studies because the 3.5 kHz records can be directly correlated with discrete beds (>12 cm), physical properties, and deposits (e.g., slumps, debris flows, sand beds, etc.) in the cores. The present study primarily utilized short ping (<5 msec) 3.5kHz echograms acquired by The University of Texas Institute for Geophysics (UTIG) during a large number of research cruises conducted since the mid-1970s (Fig. 1). In addition, 3.5 kHz data collected by the U.S. Geological Survey (USGS) in 1982 and 1985 during their survey of the U.S. Exclusive Economic Zone with the GLORIA long-range side-scan sonar (EEZ-SCAN 85 Scientific Staff, 1987) were also utilized (Fig. 1). A total of >71,000 km of 3.5 kHz seismic data were interpreted (Fig. 1). Because of the complex structure and topography of the intraslope basin province, acoustic penetration is quite variable throughout the region along individual lines. In areas of gentle to flat relief (e.g., continental slope and rise, intraslope basin floors, and abyssal plains), penetration generally ranges from 20 to 100 m below the seafloor. However, penetration is absent or quite limited on steep walls of intraslope basins, diapirs, hilltops, and the Sigsbee Escarpment.
Air-Gun Seismic Profiles
Air-gun seismic-reflection profiles were available for all of the USGS lines (EEZ-SCAN 85 Scientific Staff, 1987), and some of the UTIG lines used in this study. All available air-gun seismic profiles were utilized in conjunction with interpretation of the 3.5 kHz seismic profiles. These air-gun seismic profiles were generally recorded at 50–200 Hz and commonly image the upper 1–2 km of the seafloor sediments.
Bed-by-bed descriptions were conducted on the 288 piston cores of this study (Figs. 1 and 2; see the Supplemental Table1 for core locations and data). Individual core lengths range between 85 and 1140 cm, and the average length is 430 m. The majority of the cores (201) were collected during 17 UTIG research cruises between 1974 and 1995. These cores were continuously archived in refrigerated conditions. An additional 38 cores were collected in 1997 during a USGS research cruise (R/V Gyre Cruise 9706) to study the Bryant Canyon turbidite pathway (Twichell et al., 2000). The senior author (JED) participated on this cruise and described the cores onboard ship. Subsequently, both authors re-described and sampled these cores at the USGS core archive in Woods Hole, Massachusetts. Data from an additional 49 piston cores in the Lamont-Doherty Earth Observatory archive were also used in the present study. These cores were taken between 1954 and 1969 and range from 67 to 1102 cm in length (average = 674 cm; total of 330 m of core; see the Supplemental Table). These cores were never refrigerated and have completely dried out and disintegrated. Thus, only the original Lamont-Doherty core descriptions and black and white photographs were available.
Bed-by-bed descriptions, graphic logs, and photographs were produced for all UTIG and USGS piston cores (e.g., Fig. 3). A number of these graphic logs are shown along with close-up color photos in figures throughout the text and Supplemental Figures 1 and 3. The colors of the clays and muds in the cores are mainly gradational or intra-gradational shades of gray and brown. All discrete beds composed of silt-, sand-, or gravel-sized particles are colored yellow on the core logs. The particle size of each bed is represented by the extension of the bed to the right of the log for a distance proportional to the particle size. The scale at the top right of each log shows the sizes for clay (C), silt (S), fine to coarse sand (F, M, and C), and gravel (G). Where possible, all individual discrete silt and/or sand laminae and beds are represented on the logs. In some cases, however, laminae are too numerous, thin, and closely spaced in a short interval (a few cm) to represent the individual laminae. Graded sand beds are also represented where thicknesses permit. A few other symbols are used in the graphic logs and are generally self-explanatory. Mud clasts, folds, and discordant, flowed, and truncated beds or layers are generally only represented schematically. However, close-up photographs of short (∼20 cm) intervals of the split core face are presented with the core logs to illustrate examples of these features. Red bars and the red letter “P” identify the location of each photographed interval along the core log. In addition, 97 cores were sampled and the ages of the sediments were determined by biostratigraphic zonation to establish the relationship of the deposits to glacioeustatic sea-level change. These results are presented in a companion paper by Olson et al. (2016).
REGIONAL SEDIMENTATION PROCESSES REVEALED BY 3.5 kHz ECHO-CHARACTER MAPPING (HIGH-RESOLUTION SEISMIC FACIES)
Construction of the Echo-Character Map
Numerous previous studies have demonstrated that by mapping the distribution of various types of 3.5 kHz bottom echoes recorded on a continental margin area, then combining this information with additional data (if available) from cores, seismic, side-scan sonar, swath-mapped bathymetry, bottom photographs, and hydrographic data, etc., it is generally possible to determine the types and regional influence of the various sedimentation processes that have shaped that region (e.g., Damuth, 1975a, 1978; Damuth and Hayes, 1977; see Damuth, 1980, for a review). Echo types identified on 3.5 kHz echograms are actually high-resolution seismic facies. Interpretation and mapping of 3.5 seismic facies began well before the introduction of seismic sequence and facies interpretation introduced by Vail, Mitchum, and co-workers in Payton (1977). Thus, marine geologists have routinely referred to the seismic facies recorded on 3.5 kHz echograms as “echo types” or “echo character,” rather than “seismic facies” (see Damuth, 1980, for a review).
In this study, a 3.5 kHz echo-character map was constructed for the Gulf of Mexico intraslope basin study area to delineate the regional sedimentation processes (e.g., mass-transport deposits, turbidites, and bottom-current deposits) in this area (Fig. 2). This map was constructed by classification of all available 3.5 echograms following the classification scheme and methods developed by Damuth (1975a, 1978, 1980; Damuth and Hayes, 1977). Echo types were interpreted and mapped along each individual ship track. Boundaries between various echo types were then connected from ship track to ship track to produce the map, which shows the regional distribution of each echo type (Fig. 2).
Echo character typically changes very rapidly along any given ship track, and echo types commonly grade from one to another across short distances because of the rugged topography. The final map cannot encompass all such minute changes because of the map scale. Thus, by necessity, the map (Fig. 2) represents a sometimes greatly simplified depiction of the detailed echo-character distribution or resolution. Small regions of various echo types, which are too small to be depicted at this scale, are eliminated or combined to produce patterns that can be displayed at the map scale. Compilation of the echo-character map was enhanced by use of the regional bathymetry to create a more realistic determination or prediction of the true distributions of the echo types in the rugged intraslope basin province. For example, an intraslope basin might have only one or two 3.5 kHz profiles across it (e.g., Fig. 1). By using the SeaBeam and other bathymetric maps, we could then confine this echo type on the map to the limits of the actual basin floor, rather than having to speculate on the shape of the basin floor.
Classification and Interpretation of Echo Types
The 3.5 kHz echo types were classified on the basis of: (1) internal reflection character (seismic facies) of the uppermost seafloor including the clarity and continuity of bottom and sub-bottom echoes and (2) morphology or microtopography of the seafloor using the classification and methods developed by Damuth (1975a, 1978,1980). The classification includes eight discrete echo types, which are described below and illustrated in Figure 2.
Type 1: Sharp Bottom Echoes with Continuous Parallel (i.e., Conformable) Sub-Bottom Reflections (Fig. 2; Profiles A–F in the Legend)
This echo type is divided into two subtypes (1A and 1B) based on the relief or roughness of the seafloor. Echo Type 1A is recorded from flat or smooth to undulating, gently rolling, or slightly hilly, seafloor topography (Fig. 2; profiles A–C in legend). Acoustic penetration is 20 to >100 m. Individual sub-bottom reflections can commonly be traced for tens of kilometers and rarely >100 km. Echo Type 1A is mainly recorded seaward of the Sigsbee Escarpment from the continental rise, the Sigsbee Abyssal Plain, and Bryant and Alaminos submarine fans. In the intraslope basin area, this echo type is commonly recorded from the flat floors of many intraslope basins (Fig. 2). Depositional processes represented by this facies type include pelagic and hemipelagic sedimentation on the slope and rise; overbank or levee deposits of turbidity currents on submarine fans; ponded deposits of turbidity currents and related gravity-controlled flows in the floors of some intraslope basins and on much of the Sigsbee Abyssal Plain. On other continental margins, piston cores from this echo type generally contain few (0%–5%) or no interbedded silt and/or sand laminae and beds (Damuth, 1980, and references therein). In the study area, we do not have enough piston cores from Type 1A regions to determine if this statistical relationship holds for this region; however, the several cores recovered from Type 1A areas contain zero to a few silt and/or fine sand laminae. Thus, many regions of Type 1A echo type throughout the study area generally can be interpreted as sediments with little or no bedded silt and/or sand.
Echo Type 1B is recorded from gently hilly to steep irregular and very hilly seafloor topography (Fig. 2, profiles D–F in legend) throughout the intraslope basin province. Types 1A and 1B echoes are end members that grade into one another, and the division between the two is somewhat subjective and is based solely on the topographic relief on the seafloor. For example, Type 1A echoes on an intraslope basin floor commonly transition into Type 1B echoes on the hills of the basin walls. Acoustic penetration for Type 1B is 20 to >75 m. Individual sub-bottom reflections are conformable and commonly can be traced for tens of kilometers. Depositional processes represented by Type 1B echoes are mainly hemipelagic to pelagic sedimentation that conformably drapes topographic highs. Cores from Echo Type 1B are commonly composed of clays with abundant foraminifera, i.e., foraminiferal clays, marls, and oozes. Biostratigraphic analysis (foraminifera) and chemical stratigraphy (CaCO3) performed on many of these cores confirm that the sediments accumulated slowly and continuously as pelagic drape.
Type 2: Semi-Prolonged Bottom Echoes with Discontinuous, Parallel Sub-Bottom Reflections (Fig. 2; Profiles G, H, and I in the Legend)
Echo Type 2 is recorded from flat or smooth to undulating, gently rolling, or slightly hummocky seafloor topography (Fig. 2). Acoustic penetration is 10 to >50 m. This echo type is recorded from limited parts of the continental slope and rise and a very few intraslope basin floors (Fig. 2). Two larger areas are recorded from the Sigsbee Abyssal Plain, although data coverage is sparse in these areas (Fig. 1). This echo type is also recorded from the levees of the turbidite channel on the eastern side of the East Breaks Slide Complex (Fig. 2). In addition, Type 2 echoes are recorded from the upper Rio Grande Fan and adjacent continental slope (Fig. 2). Depositional processes represented by this echo type include hemipelagic sedimentation on the slope and rise; overbank or levee deposits of turbidity currents on submarine channels and fans; ponded deposits of turbidity currents and related gravity-controlled flows in the floors of intraslope basins and on the Sigsbee Abyssal Plain. On other continental margins, piston cores from this echo type generally contain moderate amounts (up to 30%) of interbedded silt and sand (Damuth, 1980, and references therein). In the present study area, we do not have enough piston cores from Type 2 echoes to determine if this statistical relationship holds for this region.
Type 3: Prolonged Bottom Echoes Generally with No Sub-Bottom Reflections; Or Rarely a Single Discontinuous Sub-Bottom May Be Present (Fig. 2; Profiles J, K, L, and M in the Legend)
Echo Type 3 is recorded from flat or smooth to undulating or gently rolling seafloor topography (Fig. 2). Acoustic penetration is generally >10 m. The largest area of this echo type is recorded from the Rio Grande Fan at the western edge of the study area (Fig. 2). This echo type is also recorded from the floors of a few intraslope basins (Fig. 2; profiles L and M in legend) and locally from floors of submarine channels; however, these areas are generally too small to be represented on Figure 2. Depositional processes represented by this echo type are interpreted as deposits of sandy to very sandy turbidity currents and related gravity-controlled mass flows. On other continental margins, piston cores from this echo type generally contain very high amounts (up to 100%) of bedded silt and sand (Damuth, 1980, and references therein). In the present study area, we do not have enough piston cores from Type 3 echoes to determine if this statistical relationship holds for this region.
Type 4: Sharp Bottom Echoes with Parallel to Subparallel or Disconformable to Convergent Sub-Bottom Reflections that Resemble Migrating Wave Forms (Fig. 2; Profiles N and O in the Legend)
Echo Type 4 is recorded from undulating, gently rolling, or slightly hummocky seafloor topography (Fig. 2). Acoustic penetration is 40–100 m. The subparallel to convergent reflections along with the regular hummocky seafloor and the focusing effects beneath some hummocks (e.g., Fig. 2; profiles N and O in the legend) indicate lateral migration of sediments. Thus, this echo type is interpreted to represent deposits of large, migrating sediment waves created by bottom currents. The wind-driven Loop Current in the northern Gulf of Mexico reaches to the seafloor (at least intermittently) and redistributes sediment into such wave forms (e.g., Behrens, 1994). These sediment waves can be considered deposits of contour currents (contourites) and related bottom currents. The few cores that penetrate this echo type are compatible with an interpretation of contourite processes because they contain numerous, thin silt laminae (see section on Bottom Current Deposits below).
Type 5: Sharp to Prolonged Bottom Echoes with Transparent or Non-Reflective Internal Character (No Sub-Bottoms) Commonly Contained in External Forms that Display Irregular Wedge, Mound, Fill, or Lens Shapes (Fig. 2; Profiles P and Q in the Legend)
Echo Type 5 (Fig. 2) is recorded from a wide variety of seafloor topography from smooth seafloor to steep hillsides. Acoustic penetration is <5 to >100 m. This echo type may occur overlying or interbedded with Echo Types 1, 2, and 7 and may truncate sub-bottoms of the latter facies. Figure 2 shows that areas of this echo type are ubiquitous throughout the study area but are commonly localized in a large number of intraslope basins. In contrast, much larger areas of this echo type occur along the upper continental slope, portions of the East Breaks Slide Complex, east of the base of the Sigsbee Escarpment, and on the southwestern portion of the Mississippi Fan (Fig. 2). Type 5 echoes are generally interpreted as the deposits of gravity-controlled mass-transport processes, mainly slumps and debris flows. Coring of these deposits at numerous locations around the world (e.g., Damuth, 1980; Mosher et al., 2010), as well as the present study area (Tripsanas et al., 2004a, 2004b, 2006, 2008; Olson and Damuth, 2010), confirm that these deposits are mainly debris flows. The cores in the present study confirm the MTD interpretation because they contain thick intervals of deformed, flowed, and folded mud and mud clasts of exotic colors and lithologies, which are clearly slump and/or debris-flow deposits (see Mass-Transport Deposits section below). Deposits of sandy debris flows also occur. Note that in some regions on the upper continental slope in the northern portion of the study area and east of the Sigsbee Escarpment, Echo Type 5 may not be as extensive as the map portrays because of the wide data spacing and poor quality of some records (lack of penetration). In addition, on some 3.5 kHz profiles of poor quality across intraslope basin floors, Type 3 echoes may have actually been interpreted as Type 5 echoes because the two types can be difficult to differentiate on poor-quality records. Regardless of the true limits of individual regions of mass-transport deposits, mapping of Type 5 echoes shows that deposition by mass-transport mechanisms is a dominant process in intraslope basins.
Type 6: Small, Regular, Overlapping to Singular Hyperbolic Bottom Echoes with Varying Vertex Elevations. Sub-Bottom Reflections Are Generally Rare to Absent (Fig. 2; Profiles R and S in the Legend)
Echo Type 6 is rarely recorded from the study area (Fig. 2). Acoustic penetration is generally <20 m. This facies type is returned from the floors of a few intraslope basins. The largest area is returned from the East Breaks Slide Complex (Fig. 2). In this area, Type 6 echoes are associated with Type 5 echoes, and the two types grade into one another. Type 6 echoes are generally interpreted as gravity-controlled mass-transport deposits, which are mainly slumps and slides. Coherent or deformed blocks of various sizes that make up these deposits produce a rough seafloor microtopography, which, in turn, is recorded as a series of small overlapping to single hyperbolae of slightly variable sizes on 3.5 kHz profiles. In rare locations, the slump or slide blocks form regular undulations of the seafloor to produce a “carpet-roll” effect, and in these cases, Type 6 is recorded as single hyperbolae with sub-bottoms (see Mass-Transport Deposits section below). Piston cores confirm the interpretation of Type 6 echoes as mass-transport deposits because they contain intervals of deformed, flowed, and folded mud and mud clasts of exotic colors and lithologies (see discussion of East Breaks Slide Complex below). In local studies of Bryant Canyon mini-basins, Tripsanas et al. (2004a, 2004b, 2006, 2008) confirmed with side-scan sonar data that Type 6 echoes are produced from large blocks in slumps and slides.
Type 7: Large, Single Irregular Hyperbolic Bottom Echoes with Widely Varying Vertex Elevations. Sub-Bottoms Are Generally Absent (Fig. 2; Profile T in the Legend)
Echo Type 7 is recorded from regions of very steep, irregular to hilly or rugged seafloor topography (Fig. 2). Acoustic penetration is generally 0 to <20 m. Individual sub-bottom reflections are generally not present, but in some cases, discontinuous, conformable sub-bottoms may be recorded across short distances. However, in most instances, the seafloor returning this seismic facies is steep with rugged outcrops where acoustic penetration is not achieved and because of the steepness of the seafloor topography, only single hyperbolic echoes are returned (Fig. 2; profile T in the legend). This echo type is recorded mainly from the steeper portions of the intraslope basin province (Fig. 2). Outcrops of older, more lithified rocks of diapirs, etc. are common. Echo Type 7 generally provides no information about depositional processes. Echo Type 7 intergrades with Echo Type 8, which is distinguished by overlapping hyperbolae.
Type 8: Large, Overlapping, Irregular Hyperbolic Bottom Echoes with Widely Varying Vertex Elevations and No Sub-bottom Reflections (Fig. 2; Profile U in the Legend)
In most instances, the seafloor returning Echo Type 8 is the steepest and most rugged topography in the intraslope basin province. Echo Type 8 is recorded from most of the steep Sigsbee Escarpment (Fig. 2). Outcrops of older more lithified rocks of diapirs, etc. on the most rugged hills in the intraslope basin province produce this echo type. This echo type generally provides no information about depositional processes. Thus, areas returning Echo Type 8, as well as Echo Type 7, are not discussed further in this paper.
DEPOSITS OF TURBIDITY CURRENTS AND RELATED GRAVITY-CONTROLLED FLOWS
“Hemipelagic” Sediments (Hemiturbidites)
The cores from the continental slope, intraslope basin floors, submarine fans, and the Sigsbee Abyssal Plain are predominantly composed of hemipelagic sediments, which are gray to brown-gray clays and muds. Previous biostratigraphic, chemical, and other stratigraphic zonation of similar sediments on other continental margins have demonstrated that they accumulated slowly and continuously throughout glacial cycles (e.g., Damuth, 1975b, 1977; Damuth et al., 1988). The terrigenous sediments in these deposits moved downslope as thin, near-bottom (turbidity) flows and then distributed laterally across the seafloor. Near-bottom flow is confirmed by the fact that these gray “hemipelagic” sediments do not occur on even low-relief topographic features that rise <200 m above the surrounding seafloor (e.g., Damuth, 1977). Although these sediments do not rain out of the upper water column like true hemipelagic sediments, Damuth (1977) retained the name hemipelagic for these sediments and proposed that the terrigenous sediment components were delivered by turbidity flows, entrained in the benthic boundary layer above the seafloor where they were distributed and deposited slowly and continuously along with planktonic components (e.g., foraminifera) across the continental margin (see Damuth, 1977, and Damuth et al., 1988, for detailed discussion and additional references). Eisma and Kalf (1984) and Van Weering and van Iperen (1984) observed similar sediments on Zaire Fan and attributed them to continuous low-velocity density (turbidity) flows. Subsequently, Stow and Wetzel (1990) described similar deposits on Bengal Fan, for which they proposed the term hemiturbidite and interpreted them as deposition from dilute clouds of suspended sediment produced from dying turbidites. Hemiturbidite is probably a good term for these sediments because it distinguishes them from true hemipelagic sediments, as well as episodic turbidites. Whatever the origin of these deposits, the hemiturbidites are present throughout the Gulf of Mexico intraslope basins, fans, and abyssal plains. We have previously documented the presence and age of these sediments using foraminiferal biostratigraphy and calcium-carbonate stratigraphy on 97 of our piston cores (Olson et al., 2000; Olson and Damuth, 2001; Olson et al., 2001; Olson and Thompson, 2005) and describe them in detail in a companion paper (Olson et al., 2016).
Color change is observed in many of the cores corresponding approximately to the Pleistocene/Holocene boundary (e.g., Olson et al., 2000; Olson and Damuth, 2001; Olson et al., 2001, 2016). Sediments in the Holocene are commonly a more brown or brownish-gray color and consist of pelagic to hemipelagic foraminiferal marl to ooze. In contrast, sediments deposited during the Wisconsin or Last Glacial are commonly gray to greenish gray and consist of terrigenous-rich hemiturbidite clays. This color change marks the transgression of sea level across the continental shelf at the end of the Last Glacial, which trapped terrigenous sediments on the inner shelf and shut off the supply to the deep sea during much of the Holocene (e.g., Ewing et al., 1958; Damuth, 1977; Kolla and Perlmutter, 1993). In contrast, during the Last Glacial (Wisconsin) and previous glacial cycles, sea level was lowered beyond the shelf edge and rivers flowed directly into the heads of submarine canyons where gravity-controlled processes delivered large quantities of sediments to the intraslope basins and the abyssal plains and fans beyond the Sigsbee Escarpment. These processes continuously deposited hemiturbidites as defined above, as well as discrete turbidites and sandy debris flows deposited by instantaneous events.
The intraslope basin province south of Texas and Louisiana is characterized by numerous rugged hills and intervening mini-basins (Figs. 1 and 2). These mini-basins were the site of fill-and-spill turbidite systems. The turbidity currents flowed from mini-basin to mini-basin and deposited thick ponds of sand in many of the basin floors (e.g., Armentrout et al., 1991; Prather et al., 1998; Weimer et al., 1998a, 1998b; Twichell et al., 2000). Some of these pathways eventually developed unconfined canyons that allowed sediment to bypass the mini-basins to the abyssal basin floors to form submarine fans beyond the Sigsbee Escarpment (e.g., Galloway et al., 2000; Combellas, 2003; Combellas-Bigott and Galloway, 2006; Galloway, 2009). The Quaternary-age Bryant Canyon Turbidite System is a modern example of a pathway that leads through a series of fill-and-spill mini-basins to feed the large abyssal Bryant Submarine Fan (see below) (Lee et al.; 1996; Twichell et al., 2000; Figs. 1 and 2).
Our piston cores show that sandy turbidity-current and related gravity-controlled deposits are common in the floors of intraslope basins throughout the study area. Eighty-eight cores contain silt and/or sand laminae and thin-to-medium beds (Supplemental Table). Most of these cores were recovered from the floors of mini-basins, and a few examples are shown in Figure 3. Graded beds, cross-stratification and mud-draped ripples are common in the silt and/or sand beds. Although thick sand beds were rarely recovered in the cores, this may be the result of their short lengths and the fact that piston cores have difficulty penetrating and recovering thick sands. The 3.5 kHz Echo Type 3 (Fig. 2), prolonged echoes, are recorded from many basin floors and suggest thick sand deposits. Biostratigraphic zonation of some mini-basin cores shows that the silt and/or sand beds were deposited during the late Last Glacial (Wisconsin) (Olson et al., 2016) (Fig. 3A).
Trinity-Brazos Turbidite System
The study of modern turbidite pathways in the intraslope basin province has focused in detail on the Trinity-Brazos Turbidite System, where a leveed channel (Figs. 2 and 4D) extends downslope from the Trinity-Brazos shelf margin delta and feeds four mini-basins (Satterfield and Behrens, 1990). Piston cores taken adjacent to this channel contain graded laminae and thin-to-medium silt and/or sand beds, which are characteristic of overbank deposits along a channel (Figs. 4A–4C). Some beds show cross stratification and mud-draped ripples. Subsequent to our core studies completed in 1999, several very detailed studies of the Trinity-Brazos system using seismic, side-scan sonar, bathymetric swath mapping, drilling, and coring were completed and revealed deposits interpreted as both submarine fans and mass-transport deposits (Badalini et al., 2000; Beaubouef and Friedmann, 2000; Pirmez et al., 2000; Beaubouef et al., 2003; Expedition 308 Scientists, 2005; Pirmez et al. 2012; Prather et al., 2012). Certainly these studies and ours show leveed channels (e.g., Figs. 2 and 4D) characteristic of submarine fans extending downslope into these basins.
Bryant Canyon Turbidite System
The Trinity-Brazos Turbidite System does not extend across the entire intraslope basin province. However, Lee et al. (1996) showed that the Bryant Canyon system extends from a late Quaternary shelf-edge delta associated with the ancestral Mississippi River (Suter and Berryhill, 1985) down through numerous intraslope mini-basins across the entire province to the Bryant Fan at the base of the Sigsbee Escarpment (Fig. 2). Subsequently, Twichell et al. (2000) used GLORIA side-scan sonar imagery registered with multi-beam bathymetry to survey the mini-basins in detail and assess the distribution and continuity of the near-surface lithofacies of the Bryant Canyon system. Thirty-eight piston cores collected during this survey were described and interpreted during the present study (Gyre 9706 cruise, Supplemental Table). Thirteen of these cores contain silt and/or sand laminae and beds. Core 97G06-3 (Fig. 3D) shows some of the thickest sand beds. In addition, 21 cores contain intervals of mass-transport deposits, mainly debris flows with mud clasts, which usually underlie the upper intervals containing the turbidite silts and sands (Supplemental Table). Twichell et al. (2000) developed a conceptual fill-and-spill model for the evolution of the mini-basins along the Bryant Canyon pathway. More recently, Tripsanas and co-workers presented a series of detailed studies of portions of the Bryant Canyon system based on giant piston cores (∼20 m) and seismic data to understand the sedimentary history and the evolution of the canyon system in relation to the salt tectonics along the pathway (Tripsanas et al., 2004a, 2004b, 2006, 2007, 2008). In particular, these studies provide excellent detailed descriptions of sediment lithofacies and processes (turbidity currents and mass-transport deposits) during the Last Glacial and Holocene (135 ka to present).
Piston cores from a few mini-basins contain deposits of thick, structureless clay, which have been previously interpreted as unifites (Behrens, 1984; Tripsanas et al., 2004b). The basin floors where these deposits are cored return distinctive Type 5 echoes (Supplemental Figure 12). Type 5 echoes are generally interpreted as MTDs in most mini-basins based solely on echo type, and cores from these deposits generally confirm an MTD interpretation (see discussion below). However, in some instances, cores show Type 5 echoes are from fine-grained unifites (e.g., Supplemental Figure 1). Unifites are generally interpreted as fine-grained turbidity-current deposits because there are no structures (e.g., clasts, folds, or deformation) indicative of MTDs. Damuth (1977) first described these thick, structureless clay deposits from abyssal-plain cores of the western Equatorial Atlantic and interpreted them as fine-grained turbidity-current deposits. Subsequently, Stanley (1981) described similar deposits in the Mediterranean Sea, which he named unifites, and also interpreted them as deposits of turbidity currents. Behrens (1984) appeared to favor a turbidity-current origin for the Gulf of Mexico unifites and more recently, based on detailed core studies, Tripsanas et al. (2004b) interpreted unifites in Hedberg Basin as deposits of long-lasting, pulsating, fine-grained turbidity currents.
Bryant Submarine Fan
Turbidity currents that traversed the length of the Bryant Canyon mini-basin system during periods of lowered sea level of the last and previous glacial cycles built the large Bryant Submarine Fan, which extends hundreds of kilometers seaward from the Bryant Canyon mouth in the Sigsbee Escarpment across the Sigsbee Abyssal Plain (Fig. 2). The fan extends from ∼3000 m depth at the canyon mouth to 3550 m depth on the lower fan. The upper fan is ∼200 km wide, whereas the lower fan is up to 365 km wide (Figs. 2 and 5A–5C). The upper fan proximal gradients are ∼1:200, the middle fan gradients 1:300, and the lower fan gradients 1:400.
A GLORIA side-scan sonar mosaic of the fan (EEZ-SCAN 85 Scientific Staff, 1987) shows only a single distributary channel, which is highly sinuous and extends continuously southeast from the canyon mouth for at least 180 km, where it may end in a distal fan lobe (Fig. 5D). On the upper fan, this channel is ∼25 km wide and has relief of up to 200 m (Figs. 5A and 5E; profile W). The channel shows less relief down fan; widths are uncertain because the 3.5 kHz profiles cross the channel at oblique orientations because of multiple meanders (Fig. 5E). On mid-fan, the channel is ∼40 m in relief and appears to remain nearly constant down fan (Fig. 5E; profiles X–Z). On the uppermost fan, the channel is nearly straight but becomes very sinuous on the middle and lower fan with sinuosities high enough (>1.5) to classify some reaches as meandering (Fig. 5D).
Unlike most large deep-sea fans (e.g., Mississippi, Amazon, Bengal, Indus, and Zaire), the most recently active channel is the only distributary channel-levee system visible on the present Bryant Fan surface (Fig. 5). There have apparently been no recent avulsion events that caused channel switching across the fan. This is unusual since most other large fans show numerous channel avulsions (e.g., Amazon Fan, Damuth et al., 1988; Pirmez and Flood, 1995; Indus Fan, Kenyon et al., 1995; Zaire Fan, Savoye et al., 2009). The seismic profiles do show many apparent high-amplitude reflection (HAR) units within the Bryant Fan; these units indicate channel-fill deposits in older, buried channel-levee systems (Figs. 5A–5C). Faint reflections indicative of levee systems appear to be associated with some of these HAR units. Also, possible base-of-channel high-amplitude reflection packages (HARP units) apparently occur at some locations. Unfortunately, the sparse seismic control across the fan does not permit tracing trends of these older channels or determining relative ages. However, the occurrence of only a single extensive channel, which shows no evidence of avulsion events, on the modern fan surface suggests that this large fan is somewhat unique, compared to other large fans of similar size (e.g., Amazon, Zaire; e.g., Piper and Normark, 2001) and, therefore, may represent a new type of fan. Additional detailed seismic and side-scan sonar investigations will be required in the future to confirm this.
Only nine Lamont-Doherty Earth Observatory piston cores were available from the Bryant Submarine Fan (Fig. 2). Most of the cores are from the proximal to distal portions of the levees of the most recently active channel and are predominately composed of gray clay with rare to common silt laminae and thin beds. Some beds show cross stratification and mud-draped ripples. Organic material is rare, as is color banding. These sediments are consistent with levee and/or overbank deposits and distal turbidites. Core RC12-8 was taken in a large regional field of migrating sediment waves (Echo Type 4, Fig. 2) that extends across the western portion of the fan; thus the silt and/or sand laminae and beds may actually be the deposits of bottom currents at this location (see Bottom-Current Deposits section below).
Rio Grande Submarine Fan
The Rio Grande Fan is located in the western part of the study area seaward of the Rio Grande River (Figs.1 and 2). The fan is deposited on a plateau-like region on the middle of the continental slope and extends eastward for ∼100 km between 96° and 95°W in water depths of ∼1200 m to ∼1700 m. The fan width is generally ∼30–50 km and lies between 26°15′ and 26°45′N. During the late Oligocene through Miocene time, a salt canopy uplifted the lower continental slope and formed the mid-slope plateau on which the fan was deposited (Diegel et al., 1995; Peel et al., 1995). As a result, numerous salt diapirs occur in the Rio Grande Fan region. Rothwell et al. (1991) briefly described Rio Grande Fan based on GLORIA side-scan sonar imaging and 3.5 kHz echo-character mapping. Using Damuth’s classification (Damuth, 1975a; Damuth and Hayes, 1977), they reported that the Rio Grande Fan returned prolonged echoes (Echo Type 3 of the present study). They identified small valleys on the 3.5 kHz records, which they interpreted as small (∼25-m-relief) channels. They also observed channel-like features on the GLORIA images that they equated with the small channels on the 3.5 kHz records and observed that these coalesced on the lower fan where they entered a canyon that crosses the Perdido Escarpment (Fig. 2). Rothwell et al. (1991) interpreted the prolonged echoes as sandy overbank deposits.
Our 3.5 kHz echo-character mapping using the same USGS data as Rothwell et al. (1991) plus dense coverage of UTIG data (Fig.1) confirmed that prolonged echoes (Echo Type 3) are returned throughout most of the Rio Grande Fan (Figs. 2 and 6A–6C). Prolonged echoes normally indicate abundant amounts of bedded silt and/or sand in the upper 50 m below the seafloor. Semi-prolonged echoes (Type 2; Fig. 2) are returned from the upper fan and adjacent continental slope and are also indicative of moderately sandy sediments (Damuth, 1980). The 3.5 kHz profiles show that the small channel segments are 10–45 m deep, up to a few hundred meters wide (Fig. 6A–6C), and are much more numerous on the fan surface (Fig. 2) than observed by Rothwell et al. (1991). These channels do not appear to have well-developed levees. The air-gun seismic lines across the fan show apparent similar buried channels within the fan (Fig. 6D). A strong reflection (base of fan; Fig. 6D) is interpreted to be the base of the fan with the entire unit above composed of turbidites and related gravity-controlled flows. Multiple submarine canyons apparently feed the fan’s channels because swath bathymetry on the continental slope west of the fan shows several long canyons and/or channels extending from the shelf edge across the slope toward the fan (Fig. 2). The channels are most numerous on the middle fan, where it is ∼30–50 km wide, but they diminish in number on the lower fan where it becomes considerably narrower (∼15 km) (Fig. 2). Only two channels are observed near the lower end of the fan (Fig. 6A), and these channels appear to extend into a large structural canyon or valley (see black arrows on map, Fig. 2) that extends southeast into Perdido Canyon.
Eleven piston cores were described from the fan and adjacent continental slope (Fig. 2). Unfortunately, most of the cores are short and are composed of gray to brownish-gray hemipelagic sediments. Silt and sand laminae and thin beds are rare to common. One core also contains oriented wood fragments up to 1 cm long overlying medium sand. Two cores contain disturbed clays that display deformed and flowed beds, angular contacts, and a few thin, deformed, or flowed beds of fine sand indicating they contain mass-transport deposits. For example, Core IG38-9 at the north edge on the inner fan contains two thick silt and/or sand beds 83 cm and 18 cm thick, which are mainly very fine to medium sand with scattered mud clasts up to 3 cm in diameter and appear to be sandy debris flows (Fig. 6E). The presence of sand in the cores is consistent with the prolonged echo character (Type 3) that characterizes the fan and the numerous channels (Fig. 6A–6C). The relatively poor core recovery may indicate that thicker silt and/or sand beds are present and impeded deeper penetration of the coring device. Biostratigraphic zonation of Core IG38-9 shows that these sand beds were deposited during the Wisconsin or Last Glacial when sea level was lowered (Olson et al., 2016) (Fig. 6E).
In summary, the Rio Grande Fan has an unusual setting on a slope plateau. The average surface gradient of ∼1:250 throughout the Rio Grande Fan is relatively steep, which is typical of sand-rich fans (Nelson and Nilsen, 1984). This combination of factors appears to have resulted in deposition of a sandy fan characterized by relatively coarse-grained deposits and multiple, unleveed channels. The typical incised or erosional character of the channels, rather than aggradational leveed channels, and the occurrence of a maximum number of channels on the middle fan that diminish in number down fan suggests a sand-rich fan with a braided channel system. Deep-sea fans with apparent braided channel systems have rarely been observed (Belderson et al., 1984; Ercilla et al., 1998). If the distributary-channel pattern on Rio Grande Fan represents a sandy braided channel system, this would be in contrast to most similar sized submarine fans, such as those off California described by Normark and co-workers (e.g., Normark, 1970; Normark and Piper, 1972; Piper and Normark, 1983, 2001; Fildani and Normark, 2004), which are mud rich and have aggradational distributary channel-levee systems. Rio Grande Fan may, thus, represent a new type of fan, which should be subjected to detailed studies in the future.
Three large areas of the Sigsbee Abyssal Plain and western Bryant Fan return Type 4 echoes (Fig. 2) and represent sediment waves showing lateral migration (Figs. 2 with profile N in legend, and 7A–7H). These sediment waves have wavelengths of up to several kilometers and amplitudes of tens of meters. Based on previous studies of similar echo types (see Damuth, 1980, for review), we interpret Type 4 echoes as large, migrating sediment waves created by bottom-current activity. These waves appear to have been created by currents within the strong wind-driven Loop Current, which has been well documented to impinge on the seafloor, at least intermittently, at many locations throughout the deep Gulf of Mexico (e.g., Pequegnat, 1972; Shanmugam et al., 1993a, 1993b; Behrens, 1994; Kenyon et al., 2002). Behrens (1994) attributed the formation of a large field of sediment waves, whose northern tip is visible in our study area at 95°W, to the Loop Current. The Loop Current has also formed fields of large erosional furrows and related features along the Sigsbee Escarpment (Bryant et al., 2000; Bean et al., 2002; Bean, 2005). These furrows are large (∼30 m wide by 10 m deep), u-shaped, channel-like features eroded into the seafloor, indicate very strong (1–2 knots) bottom-current activity (Hamilton and Lugo-Fernandez, 2001), and are similar to erosional furrows reported from other areas (e.g., Embley et al., 1980; Flood, 1983). Bean et al. (2002) note that besides the present furrows, there are several older cycles of furrow formation that are related to stronger bottom currents at times of high sea level. Profile H (Fig. 7) shows a field of migrating sediment waves seaward of the Sigsbee Escarpment. Sediments have clearly been eroded by bottom currents, which impinged on the base of the escarpment and were accelerated. This apparently led to increased erosion or scour at this location and formed the small valley at the base of the escarpment (Fig. 7H). Just seaward of this valley, there is a zone of small regular overlapping hyperbolae that appear to be recorded from erosional furrows similar to those described by Bryant et al. (2000) and Bean (2005) from this region.
The westernmost area of sediment waves (Echo Type 4; Fig. 2) was previously identified by Behrens (1994) and extends continuously along the continental margin from ∼20° to 25.5°N. The largest area of sediment waves (Echo Type 4) extends eastward on the Sigsbee Abyssal Plain from ∼94.75°W to the distal portion of the western Bryant Fan near 92.25°W (Figs. 2 and 7A–7D). The third area is east of the Sigsbee Escarpment and west of the western edge of the Mississippi Fan (Figs. 2 and 7E–7G). This region overlaps the large areas of erosional furrows and related features studied by Bean (2005). These sediment waves can be considered deposits of bottom currents and/or contour currents. Two piston cores, V3-99 and RC12-8 (Fig. 8A and 8B), are from the sediment-wave fields shown in Figure 2. The cores are gray clays with numerous silt laminae and are sediments consistent with bottom-current deposition. An additional core, IG46-8, has similar lithology (Fig. 8C) and was taken in an area between the Sigsbee Escarpment and Green Knoll. Bean (2005) has reported erosional furrows and related strong current features in this area. The areal distributions of these large fields of sediment waves, in conjunction with the fields of erosional furrows documented in other publications, indicate that bottom-current activity is extensive throughout the northwest Gulf of Mexico seaward of the Sigsbee Escarpment. The location of the furrows along the base of the escarpment and the sediment waves farther seaward suggest that the strongest currents are focused along the escarpment and are erosional in nature, whereas relatively weaker currents prevail seaward in the abyssal areas and rework sediment into large sediment waves.
MASS-TRANSPORT DEPOSITS (MTDs)
Mass-transport deposits are ubiquitous throughout the Gulf of Mexico Intraslope Basin Province as evidenced by the 3.5 kHz echo character (Fig. 2) and confirmed by piston cores (Olson and Damuth, 2010). In this paper, we use the term mass-transport deposits to include slides (brittle deformation), slumps (plastic deformation), and debris flows (plastic flow), as well as a combination of these deposits (see Shanmugam, 2015, for a review of terminology). The MTDs return two distinctive 3.5 kHz echo types—Echo Type 5 (Fig. 2; profiles P and Q in legend) is widespread, whereas Echo Type 6 (Fig. 2; profiles R and S in legend) has a very restricted distribution (Fig. 2). As discussed above, numerous studies have documented that these two echo types, especially Type 5, are characteristic of MTDs (e.g., see Damuth, 1980, for review and references). Tripsanas et al. (2004a, 2004b, 2006, 2008) also confirmed this correlation for local areas within the Bryant Canyon mini-basin system. These echo types commonly grade laterally into each other. In addition to the regions returning these echo types shown on Figure 2, there are many additional smaller areas in intraslope basins that return Echo Types 5 and 6, but these areas are too small to be displayed on Figure 2 because of the map scale.
Figure 2 shows that MTDs represented by Echo Type 5 are common within the intraslope basins (e.g., Figs 9A–9C, 10, 11B, and 12). Extensive areas of Type 5 echoes are also observed on portions of the upper continental slope and along the eastern base of the Sigsbee Escarpment (Fig. 2). The large area of Type 5 echoes in the southeast corner of the study area (Fig. 2) represents the toe of a much larger MTD on Mississippi Fan (Fig. 9D), which was first identified by Walker and Massingill (1970). Type 5 echoes have been confirmed to be debris-flow deposits by numerous studies around the world (e.g., Damuth, 1980, and references therein). Areas of Type 6 echoes are rare throughout the study area, and the most extensive area is returned from the East Breaks Slide Complex in the northwestern part of the study area (Fig. 2). Type 6 echoes are interpreted as mainly slump and slide deposits. Coherent or deformed blocks of various sizes make up these deposits and produce rough seafloor microtopography, which, in turn, is recorded as a series of small overlapping to single hyperbolae of slightly variable sizes on the 3.5 kHz profiles (see discussion on East Breaks Area below).
The piston cores provide confirmation or “ground truth” of the interpretation of Echo Types 5 and 6 as MTDs. Ninety-five cores have chaotic MTDs; 81 contain debris-flow deposits and 27 contain slumps and/or slides (Supplemental Table). Combinations of these deposits are observed in 11 cores (e.g., Figs. 10–12). We generally interpret the dipping, discordant, and folded intervals as slump or slide deposits and the deposits of mud or rock clasts as debris-flow deposits. Slump deposits contain thick intervals of disturbed, deformed bedding and layers that show variable dips, discordance, truncations, faults, soft-sediment flowage, and folds (Figs. 11C, 11D, and 12B). Exotic colored (e.g., brick reddish-brown) layers and variable lithologies are common. Debris flows contain angular to rounded mud clasts of various sizes, shapes, and colors (Figs. 10, 11A, and 11B).
The majority of slumps and debris flows are muddy deposits, but sandy debris flows are also present. A few cores taken in MTDs interpreted from the Type 5 echoes recovered sandy deposits, which consist of medium to coarse sands with larger coarse- to gravel-size grains and rock clasts, and/or mud clasts (Figs. 11A and 12A). Unfortunately, there currently are no criteria for predicting muddy versus sandy slump and/or debris-flow deposits based solely on the 3.5 kHz echo character alone. Previous interpretation of sandy deposits as sandy debris flows (plastic flow) rather than turbidity-current deposits (fluidal flow) has been controversial and much debated as to whether various deposits represent fluidal, laminar, or hybrid flows (e.g., Lowe and Guy, 2000; Shanmugam, 2000, 2015, and references therein; Haughton et al., 2009; Kane and Ponten, 2012). However, review and discussion of these processes are beyond the scope of the present study. The important observation is that these sandy deposits were recovered from core in Type 5 echoes (Fig. 12A). This confirms that these cores were taken in mass-transport deposits and provides evidence that the sandy deposits recovered were deposited by debris flows. The occurrence of sandy MTDs in modern intraslope basins has important implications for hydrocarbon exploration. Such MTD sand bodies would be expected to exhibit much different reservoir geometries than ponded turbidite reservoirs in intraslope basins because plastic debris flows tend to “freeze” in place and produce sand bodies of irregular shape. These MTD sand bodies will have much different reservoir geometries than laterally extensive turbidites.
East Breaks Slide Complex
The largest area of the continental slope affected by mass-transport processes is offshore Texas in the northwest corner of the study area and is called the East Breaks Slide Complex (Fig. 2). Studies by Lehner (1969), Woodbury (1977), Sidner et al. (1978), Woodbury et al. (1978), Tatum (1979), and Hardin (1986) described this area from seismic-reflection profiles and core data. However, all these studies focused primarily on the area of the uppermost slope just beyond the shelf break and did not describe the entire complex. These studies basically concluded that a slide originated from a Late Wisconsin shelf-edge delta of the ancestral Colorado and Brazos rivers. Sediment failure was assumed to be initiated by sediment overloading or over steepening and as the slide moved downslope, it encountered a bathymetric high that divided it into two individual lobes (Abdulah, 1995).
Rothwell et al. (1991) defined the downslope limits of the slide complex using GLORIA side-scan sonar images and 3.5 kHz echo-character mapping of USGS data based on Damuth’s (1975a, 1980) classification. However, this study only extended upslope (e.g., north) to ∼27.5°N (red ship tracks in Fig. 1), and thus only covered the downslope portion of the complex. They also interpreted these data as showing two separate (east and west) debris-flow lobes (their fig. 4). Subsequently, Piper (1997) used the UTIG 3.5 kHz records (blue tracks, Fig. 1) and piston cores to better define the entire slide complex. These data provide extensive coverage and extend north of ∼27.5°N. In contrast to the previous studies, Piper (1997) reported that the western lobe of the complex is mass-transport deposits, whereas the eastern lobe is actually a leveed channel formed by turbidity flows.
The present study utilized all the 3.5 kHz records from both the USGS and UTIG data sets (Fig. 1) and all the UTIG piston cores from the complex (Fig. 13). In addition, a National Oceanic and Atmospheric Administration (NOAA) SeaBeam swath-bathymetry mosaic is available for the upper portion of the East Breaks Slide Complex and clearly shows that the western portion of the complex is dominated by mass-transport activity (Fig. 13, inset). Massive erosion and apparent downslope movement of numerous slide and slump blocks are clearly visible. In addition, the SeaBeam bathymetric mosaic confirms Piper’s observations (1997) that the eastern portion of the complex is dominated by a channel, which appears to have two separate segments at the south end of the map (Fig. 13, inset). On the uppermost slope just beyond the shelf edge, there appear to be several other submarine canyons. In the center of the complex between these western and eastern features, there appear to be some older slump and/or slide scars, evacuation zones, and erosion.
Profiles AB (Fig. 14) and profiles OP, QR, ST, and UV (Supplemental Figure 23) are dip lines down the western portion of the East Breaks Complex. Profile AB displays largely Echo Type 6 from the head of the complex down to a small hill on the lower part of the complex. This echo type is returned from irregular, large slide and slump blocks that have moved downslope and are also visible in swath bathymetry (Fig. 13, inset). Profile OP (Supplemental Figure 2) shows variable MTDs reflected by Echo Types 5 and 6 including deformed, in-place sediments upslope and debris flows with zones with large blocks downslope. Profile QR (Supplemental Figure 2) on the far western side of the complex shows in-place sediments with continuous sub-bottoms that have been highly deformed by slumping or sliding (“carpet roll effect”).
Profile CD (Fig. 14) shows a strike-oriented 3.5 kHz profile across the entire upper portion of the complex. Undisturbed, well-stratified sediments are abruptly truncated and bound each side of the complex. The western portion of the complex is characterized largely by Echo Type 6, which indicates large slump and/or slide blocks and erosion. In contrast, the eastern portion of the complex appears to show Echo Type 5 beneath a thin layer of conformable stratified sediments. The channel visible on the swath bathymetry (Fig. 13, inset) is also visible on the 3.5 kHz profile. Profiles EF and GH (Fig. 14) show the echo character returned from the middle and lower areas of the western portion of the complex and return predominantly Echo Type 6, which indicates the presence of larger blocks within the debris flows. This echo type grades into Echo Type 5 at the edges of the complex where debris-flow deposits abruptly terminate at or slightly onlap the undisturbed sediments at the boundaries of the slide. Profiles IJ and KL (Fig. 14) appear to show that the eastern portion of the complex also contains debris flows that have subsequently been buried by a thin layer of well-stratified deposits. Note the sharp subsurface boundary of buried Echo Type 5 in profile IJ, which marks the eastern edge of the complex. Profiles IJ and KL (Fig. 14) also show the well-developed seafloor channel with small natural levees developed on top of the buried mass-transport deposits. These profiles indicate that although the eastern portion of the present-day complex is actually dominated by a channel-levee system as reported by Piper (1997), this portion of the complex has been subjected to mass-transport activity at an earlier time.
Three 3.5 kHz profiles (ST, UV, and WX; Supplemental Fig. 2) show the distal end of the East Breaks Slide Complex. Profile ST is from the western portion of the complex and shows Type 5 echoes indicative of debris flows onlapping Type 6 echoes that show deformation (“carpet-roll effect”) of well-stratified sediments by slumping or sliding. Profile UV is also from the western portion and shows the toe of the debris flow (Echo Type 5) overlapping and thinning downslope over undeformed, well-stratified sediments. Profile WX is from the distal end of the eastern portion of the complex. Thin debris-flow deposits (Echo Type 5) characterize the uppermost sediments, but well-stratified, undisturbed sediments can be observed below these thin deposits. The leveed channel present upslope does not extend this far down the eastern lobe. An apparent channel is observed near the downslope end of this profile; however, its regional trend is uncertain.
Numerous cores from the western lobe of the East Breaks Complex contain mass-transport deposits (Fig. 13, yellow squares). There are also a number of cores that contain no MTDs (circles). Most cores that do show MTDs have up to 300+ cm of gray to brown-gray hemipelagic clay overlying them (Fig. 15 and Supplemental Fig. 34). Many of the cores containing no MTDs are probably too short (<250 cm) to have penetrated the MTDs at those locations. Alternatively, some of these cores may have penetrated coherent slide blocks within the MTDs, which are undeformed. Most of the MTDs recovered in cores from the western portion of the complex are muddy deformed sediments with exotic mud clasts of variegated colors and sizes (e.g., Figs. 15A, 15B, and 15F). In most of these MTDs, the mud clasts are deformed and show flowage and folding. The sediments in the cores are predominantly debris flows, although the 3.5 kHz records show that slumps and slides are common. A few cores show graded silt and/or sand beds, some of which show deformation and contain mud clasts (e.g., Figs. 15C and 15E; Supplemental Figure 3A).
A number of cores were taken along the channel on the upper eastern portion of the East Breaks Slide Complex (Fig. 13). Most of these cores are very short (<250 cm) (Piper, 1997); thus, they contain only gray to brown-gray hemipelagic sediments and do not penetrate any MTDs. Core 95G06-14 does contain several silt laminae near the base, which may represent overbank deposits of turbidity currents moving down the channel. Four cores from the downslope part of the eastern portion of the East Breaks Complex contain MTDs. Three of these cores (Supplemental Figures 3C–3E) are from a region returning Type 5 echoes (Fig. 13) and confirm an MTD interpretation. These three cores contain sand beds with mud clasts. Core 91L575-2B (Supplemental Fig. 3F) contains a thick, deformed sand bed and was taken from the distal tip of the eastern complex in sediments that return Type 2 echoes. The other cores from this region are very short gray to brown-gray hemipelagic clay and probably did not penetrate deep enough to recover sands, if they are present. Biostratigraphic analysis of several cores on the East Breaks Complex (H/W beside the cores in Fig. 15E and Supplemental Figures 3A, 3B, 3D–3F; Olson et al., 2000; Olson and Damuth, 2001; Olson et al., 2001, 2016) show that the MTDs were emplaced during the Wisconsin or Last Glacial cycle during glacioeustatic sea-level lowstand.
(1) Regional 3.5 kHz echo-character mapping of high-resolution seismic facies (echo types) reveals that turbidity-current, mass-transport, and bottom-current processes have all contributed to the regional sediment deposition and distribution in the intraslope basin province and the adjacent Sigsbee Abyssal Plain of the northwest Gulf of Mexico. Piston cores from these various deposits confirm the interpretations of the processes indicated by the 3.5 kHz echo character.
(2) Turbidity currents transported sands into the intraslope mini-basins via channels and canyon pathways. Some turbidity flows traversed multiple mini-basin systems, such as the Bryant Canyon system, and spread their sediments across the Sigsbee Abyssal Plain to form deep-sea fans, such as Bryant Fan. Turbidites observed in the piston cores are generally thin to medium in thickness; although coring limitations may have prevented recovery of thicker sand beds. The turbidites in the piston cores commonly show normally graded beds. These beds are sometimes interbedded with sandy debris-flow deposits.
(3) Bryant Submarine Fan is a very large mud-rich fan formed by turbidity currents that have flowed from the mouth of Bryant Canyon mini-basin system and built the fan across the abyssal seafloor. The fan has only a single, meandering channel perched on a large levee system, which extends nearly the entire length of the fan. No other channel segments are observed and indicate that channel avulsion events have apparently not occurred during channel formation and evolution. This single distributary pattern is different from most large mud-rich fans such as the Mississippi, Amazon, Zaire, and Indus, which show numerous channel avulsions.
(4) In contrast to Bryant Submarine Fan, the much smaller Rio Grande Submarine Fan is deposited on a plateau area of the continental slope east of the Rio Grande River mouth. The prolonged 3.5 echo character (Type 3) suggests that it is a very sandy fan. Piston cores confirm that sand beds are present. The fan has numerous small, unleveed channels that suggest the Rio Grande Submarine Fan may have a braided distributary channel system, which is very rare for deep-sea fans.
(5) Our study of the Bryant and Rio Grande fans builds on and adds to the numerous landmark deep-sea fan studies by W.R. Normark during his long career. However, these previous fan studies (e.g., Piper and Normark, 2001) do not include any submarine fans from margins that have been highly deformed by gravity-tectonic processes such as our study area. The Bryant and Rio Grande fans described here appear to be unique types of deep-sea fans, which have architecture and growth patterns that are apparently different from those described in previous fan studies.
(6) Mass-transport deposits are ubiquitous throughout the mini-basins. Extensive areas affected by MTDs also occur along the upper continental slope and at the base of the eastern portion of the Sigsbee Escarpment (Fig. 2). Piston cores confirm that the majority of MTDs are debris flows with mud clasts. Most MTDs have a muddy matrix, but sandy debris flows also occur in the mini-basins and have important implications for production from such deposits. Many cores show deformation of beds or clasts, folds, and faults that indicate slump and/or slide deposits are also present.
(7) The East Breaks Slide Complex is the largest MTD complex. The western portion of the complex shows mainly slump and/or slide blocks and debris flows. Piston cores from this region confirm the MTDs. The proximal part of the eastern portion of the complex is characterized by a modern leveed, turbidite channel system. However, extensive MTDs underlie these channel-levee deposits.
(8) Three regions of large migrating sediment waves occur on the Sigsbee Abyssal Plain and eastern Bryant Fan, which were apparently formed by strong bottom currents of the Loop Current. Strong erosion has occurred locally at the base of the Sigsbee Escarpment where sediment waves and apparent erosional furrows are observed.
The data reported in this study are from our Gulf Intraslope Basin (GIB) Project Phase I Study, which was conducted jointly at The University of Texas at Austin Institute for Geophysics and the University of Texas at Arlington Department of Earth and Environmental Sciences from 1998 to 2001 as a proprietary study supported financially by an industry consortium. We thank these companies for their participation and support: ENI (AGIP), Anadarko, BP (Amoco, Vastar, and Arco), ConocoPhillips (Conoco), Marathon, Repsol (Maxus), Chevron (Unocal), Statoil, Texaco, and TotalFinaElf. Without their financial support, a project of this scale would not have been possible. Patty Ganey-Curry was extremely effective as our project manager and industry representative, and her efforts to assist us are greatly appreciated. E.W. Behrens of The University of Texas Institute for Geophysics (now retired) collected the major portion of the cores and 3.5 kHz seismic data used for the present study during several research and student cruises he initiated during the 1980s and 1990s. The Ocean Drilling Program kindly stored and archived these cores at their Gulf Coast Repository and allowed us to use their core lab facilities for sampling and core description of most of our cores. We thank Paul J. Fox, Past Director, John Firth, Core Curator, and especially Phil Rumford, Superintendent, for being very accommodating to our needs and providing critical support for using the Gulf Coast Repository. The Bureau of Economic Geology provided space for examination of some UTIG cores that are stored there. Cores from the Lamont-Doherty Earth Observatory Core Repository were collected and maintained through funding by the National Science Foundation and the Office of Naval Research. We thank David Twichell and Hans Nelson of the U.S. Geological Survey (now retired) for allowing J.E. Damuth to participate in their 97G06 cruise and then providing financial support to Damuth and Olson to study the 97G06 cores through U.S. Geological Survey grant 99HQ-RO177. David Twichell also arranged for us to examine the 3.5 kHz and air-gun records archived at the USGS Woods Hole Coastal and Marine Science Center and had selected sections photographed for use in our study. Peter Thompson, Graham Moss, Adam Skarke, George Sayre, Ilene Rex Corbo, and Bill Abbott helped with biostratigraphic analyses. Margaret Slawin, Jeffrey Olson, Laurie Powery, Don Campbell, Melissa Seay-Morales, Abiola Spring, and George Sayre assisted with sample preparation. Shana Wells ran the carbonate analyses under the direction of W.A. Balsam in his laboratory at The University of Texas at Arlington. David Dunbar, Tim Whiteaker, Gwen Watson, Lisa Gahagan, Randy Schmitz, and Peter Abel provided helpful support with graphics, drafting, mapping, and GIS compilation of the database. We thank Zane Jobe, Efthymios Tripsanas, David Piper, and Andrea Fildani for very constructive reviews that improved the manuscript. This is UTIG Publication No. 2714.