Cenomanian Gulf of Mexico Paleodrainage from Detrital Zircons: Source-to-Sink Sediment Dispersal and Prediction of Basin-Floor Fans
Published:December 01, 2016
- PDF LinkChapter PDF
Mike Blum, Kristy Milliken, Bruce Frederick, John Snedden, 2016. "Cenomanian Gulf of Mexico Paleodrainage from Detrital Zircons: Source-to-Sink Sediment Dispersal and Prediction of Basin-Floor Fans", Mesozoic of the Gulf Rim and Beyond: New Progress in Science and Exploration of the Gulf of Mexico Basin, Christopher M. Lowery, John W. Snedden, Norman C. Rosen
Download citation file:
U-Pb dating of detrital zircons in fluvial sandstones provides a method for reconstruction of drainage basin and sediment routing systems for ancient sedimentary basins. This paper summarizes a detrital-zircon record of Cenomanian paleodrainage and sediment routing for the Gulf of Mexico and U.S. midcontinent. Detrital zircon data from Cenomanian fluvial deposits of the Gulf of Mexico coastal plain (Tuscaloosa and Woodbine formations), the Central Plains (Dakota Group), and the Colorado Front Range (Dakota Formation) show the Appalachian-Ouachita orogen represented a continental divide between south-draining rivers that delivered sediment to the Gulf of Mexico, and west- and north-draining rivers that delivered sediment to the eastern margins of the Western Interior seaway. Moreover, Cenomanian fluvial deposits of the present-day Colorado Front Range were derived from the Western Cordillera, flowed generally west to east, and discharged to the western margin of the seaway. Western Cordillera-derived fluvial systems are distinctive because of the presence of Mesozoic-age zircons from the Cordilleran magmatic arc: the lack of arc zircons in Cenomanian fluvial deposits that dis-charged to the Gulf of Mexico indicates no connection to the Western Cordillera.
Detrital zircon data facilitate reconstruction of contributing drainage area and sediment routing. From these data, the dominant system for the Cenomanian Gulf of Mexico was an ancestral Tennessee River (Tuscaloosa Formation), which flowed axially through the Appalachians, had an estimated channel length of 1200-1600 km, and discharged sediment to the east-central Gulf of Mexico. Smaller rivers drained the Ouachita Mountains of Arkansas and Oklahoma (Woodbine Formation), had length scales of <300 km, and entered the Gulf through the East Texas Basin. From empirical scaling relationships between drainage-basin length and the length of basin-floor fans, these results predict significant basin-floor fans related to the paleo-Tennessee River system and very small fans from the east Texas fluvial systems. This predictive model is consistent with mapped deep-water systems, as the largest fan system was derived from rivers that entered the Gulf of Mexico through the southern Mississippi embayment.
The modern northern Gulf of Mexico continental margin is dominated by the Mississippi River sedimentdispersal system (Fig. 1). The modern Mississippi source terrain stretches from the Rocky Mountains in the western U.S. to the Appalachian cordillera in the east, and sediment is routed through the Mississippi tributary system and main stem for ~5000 km to sinks in the well known alluvial-deltaic plain of south Louisiana and the basin-floor fan in the deep water Gulf of Mexico (see Bentley et al., 2015; Fildani et al., 2016). However, integration of continental-scale Mississippi drainage is a Neogene phenomenon, and sediment routing to the Gulf of Mexico has changed significantly over time (Galloway et al., 2011; Blum and Pecha, 2014).
This paper builds on a paper by Blum and Pecha (2014), which discussed mid-Cretaceous to Paleocene North American drainage reorganization, and summarizes a detrital zircon, record for the Cenomanian of the Gulf Coastal Plain and U.S. midcontinent so as to reconstruct Gulf of Mexico paleodrainage and sediment routing. Drainage-basin reconstruction is then used to predict the scale of linked basin-floor fans using scaling relationships established from modern systems by Somme et al. (2009), and predictions are then compared with empirical reconstructions. A complementary paper by Milliken et al. (2016) presents drainage-basin reconstructions from measurements of point bar thicknesses in well logs.
Source-to-sink (source to sink) approaches to sediment-dispersal systems are grounded on understanding sediment production rates, transport, and storage through system segments (e.g., rivers, deltas, slope canyons, and basin-floor fans), as well as how the unsteadiness of sediment transport and accumulation is preserved in the ancient stratigraphic record (see Romans et al., 2016). Sediment-dispersal system segments develop scales and properties that correlate to water and sediment flux, and the scales and properties of one segment are therefore inherently related to, and can be predicted from, the scales and properties of another.
Somme et al. (2009) and Blum et al. (2013) quantify scaling relationships between drainage area and modern sediment-dispersal systems segments, including basin-floor fans (Fig. 2). Scaling relationships follow power laws, where dimensions of dispersal-system segments scale to drainage area and sediment flux, and parameters like grain size and slope scale inversely. Somme et al. (2009) show that the length of many modern fans are ~10-50% of the length of the fluvial feeder system. The modern Mississippi system is a good example of this relationship: the basinfloor fan is ~540 km in length, or ~10% of the maximum channel length of 5475 km.
The Gulf of Mexico is a data-rich, well-understood sedimentary basin, and can be used to test source to sink approaches to reconstructing ancient systems: first order paleogeography, patterns of sediment input, key stratigraphic units, and basin-fill architecture are known from several generations of industry activity (Galloway, 2008; Galloway et al., 2011; Fig. 3).
The late Cretaceous (Cenomanian, Tuscaloosa-Woodbine trend represents the first major episode of clastic shelf-margin progradation (the Eagle Ford-Tuscaloosa supersequence of Snedden et al., 2016): the sand-rich lower Tuscaloosa crops out from Alabama through Mississippi, whereas the sand-rich lower Woodbine crop outs through southern Oklahoma and north-central Texas. Slope to basin-floor equivalents of the Tuscaloosa-Woodbine have become increasingly well-known in the deep Gulf of Mexico due to recent hydrocarbon discoveries (e.g., Horn, 2011). Tuscaloosa-Woodbine systems were active when the Western Interior seaway, farther to the west, was fully connected from the Boreal sea to the Gulf of Mexico.
U-Pb dating of detrital zircons provides a fingerprint of source terrains (see Gehrels et al., 2011; Gehrels, 2014), and detrital zircons in fluvial sandstones can be used to constrain contributing drainage areas and sediment routing in a manner that complements and adds to traditional provenance studies (e.g., Dickinson, 1985). North America crustal and magmatic zircon sources are well known (Becker et al., 2005; Dickenson and Gehrels, 2008; Park et al., 2010; Fildani et al., 2016), and provide a necessary background template for this approach (Fig. 4; Table 1).
Eleven detrital zircon samples for the Tuscaloosa-Woodbine have been collected from outcrops across the northern Gulf of Mexico coastal plain (Fig. 5), which represent fluvial sandstones of old alluvial-deltaic plains, analogous to the Pleistocene alluvial-deltaic plains that comprise the modern Gulf of Mexico coastal plain (Blum and Price, 1998; Blum and Aslan, 2006). Tuscaloosa-Woodbine equivalents in the U.S. midcontinent are referred to as the Dakota Group or Formation (see Ludvigson et al., 2010), and crop out through Iowa, eastern Nebraska, and east-central Kansas, then again in the Colorado Front Range: eight detrital zircon samples have been collected from this outcrop belt, and complement published results from Iowa (Finzel, 2014). Cenomanian Tuscaloosa-Woodbine detrital zircon data have been initially presented in Blum and Pecha (2014): these data, as well as new data presented here for the Albian-Cenomanian of the midcontinent and Front Range, are available on www.geochron.org.
All detrital zircon samples were processed and analyzed using Laser-Ablation ICP-MS techniques at the Arizona Laserchron Center of the University of Arizona (Gehrels, 2014). Analyses were based on a target of n=100 grains per sample, and all samples produced between 90-110 concordant analyses (Table 2). Detrital zircon populations were illustrated using normalized probability curves, which calculate the probability of individual age spectra, so that all probabilities for the sample sum to 100%. Probability curves for samples that were in close geographic proximity and statistically indistinct were lumped, assuming they represented the same paleodrainage axes.
Cenomanian Gulf of Mexico Tuscaloosa-Woodbine trend
Mid-Cenomanian fluvial deposits of the Gulf of Mexico coastal plain represent the routing system for the first significant delivery of sediments to the deep water of the Gulf of Mexico (Galloway, 2008): the updip component of the Eagle Ford-Tuscaloosa supersequence (Snedden et al., 2016). Cenomanian strata in Alabama and Mississippi are referred to as the Tuscaloosa Group, and basal fluvial sandstones rest unconformably on Paleozoic strata (Mancini, 1988). In Arkansas, Oklahoma, and Texas, the same trend is referred to as the Woodbine Group, where basal fluvial sandstones rest unconformably on older Cretaceous rocks. Mancini and Puckett (2002; 2005), Mancini et al. (2008), and Woolfe (2012) discuss the subsurface Tuscaloosa in Mississippi and Louisiana, whereas Ambrose et al. (2009) describe the Woodbine in the east Texas basin. Olson et al. (2015) place the base of the Eagle Ford-Tuscaloosa supersequence at ca. 96 Ma.
Eleven detrital zircon samples were collected through the Tuscaloosa-Woodbine outcrop belt from Alabama to Texas (Fig. 6). Consistent with studies of pre-Cretaceous Appalachian-derived fluvial deposits (e.g., Errikson et al., 2003; 2004; Moecher and Samson, 2006; Becker et al., 2005; Park et al., 2010; Blum and Pecha, 2014; Weislogel et al., 2015), Tuscaloosa samples were dominated by Grenville ages, which comprise 50-80% of total, whereas Appalachian ages comprised ~30% of zircons in the southernmost sample, but <12% in samples farther north. Minor populations (<5% each) included ages of 700-500 Ma, derived from peri-Gondwanan terranes, and 1500-1300 Ma, which were ultimately derived from midcontinent granite-rhyolite province. These data indicated a primary Appalachian source and transport to the east-southeast.
Like the Tuscaloosa, Woodbine detrital zircon samples in southeastern Oklahoma and north-central Texas are dominated by Grenville ages (~50%); Appalachian ages comprise up to 15% of the population, but the peri-Gondwanan component is more prominent, comprising up to 10%. Additional age clusters include the midcontinent granite-rhyolite, Yavapai-Mazatzal (1800-1600 Ma), Trans-Hudson or Penokean (2000-1800 Ma, provinces, and Archean Superior or Wyoming craton sources (>2500 Ma). Woodbine samples in Oklahoma and Texas are statistically indistinguishable from each other, and mostly indistinguishable from Tuscaloosa samples in Alabama. However, the increased representation of peri-Gondwanan ages, and ages that represent the broader midcontinent and shield region, is most similar to populations recorded in Mississippian and Pennsylvanian strata of the Ouachita fold and thrust belt (Shaulis et al., 2012). Woodbine strata are therefore interpreted to reflect recycling of Pennsylvanian foreland-basin strata, plus minor local sources, rather than draining the Appalachians per se. Woodbine sediment transport was to the south; see Ambrose et al., 2009.
Albian-Cenomanian Dakota Group, US Midcontinent and Colorado front range
Detrital zircon populations from Albian-Cenomanian Dakota fluvial sandstones from Iowa, Nebraska, and Kansas closely resemble those from Tuscaloosa sandstones of Alabama and are dominated by Appalachian-Grenville signatures (>80% of total); ultimately, ~10% of grains are derived from the midcontinent granite-rhyolite province and the remainder reflect peri-Gondwanan, Archean shield, and other minor constituents (Fig. 7; see also Finzel, 2014). These data, coupled with paleocurrents and other indicators of paleoflow (e.g., Joeckel et al., 2005), indicate an Appalachian source and sediment transport to the west and northwest. According to Ludvigson et al. (2010), Dakota fluvial sandstones range in age from late Albian to middle Cenomanian, and therefore correlate generally to Gulf of Mexico Tuscaloosa-Woodbine fluvial deposition, but there are no grains younger than ca. 275 Ma, so no useful maximum depositional ages.
Dakota samples from the Colorado Front Range had a strong Appalachian-Grenville signal, but contained populations derived from the Mesozoic magmatic arc, as some are as young as ca. 100-96 Ma. Yavapai-Mazatzal basement (ca. 1800-1600 Ma) was also well represented and likely derived from the Mogollon Rim in present day Arizona. In this case, the Appalachian-Grenville population was transported by late Paleozoic fluvial systems from the Appalachian Cordillera to the western US passive margin, entrained in the Mesozoic fold and thrust belt, and then transported to the east during the Cretaceous (Laskowski et al., 2013). Dakota samples from the Front Range therefore had a Western Cordillera source, and sediment transport was to the east. U-Pb ages of ca. 100-96 Ma represented maximum depositional ages, and indicated deposition was slightly earlier than Gulf of Mexico Tuscaloosa-Woodbine fluvial deposition.
Tuscaloosa samples are interpreted to represent a drainage area similar to the modern Tennessee River, upstream from the Cenomanian outcrop belt (Fig. 8), which is upstream of the Neogene Tennessee River's north-northwest diversion to join the Ohio River. This paleo-Tennessee system flowed parallel to structural grain through, and drained, the Appalachians in the southeastern US, including the Cumberland drainage to the north, and headwaters of the present-day Tombigbee-Alabama system to the south; the paleo-Tennessee was the largest river system contributing sediment to the northern Gulf of Mexico margin and entered the Gulf of Mexico in the eastern part of the Mississippi embayment of south Louisiana (see Woolf, 2012). The Cenomanian paleo-Tennessee River was more integrated than earlier Cretaceous and Jurassic precursors of the eastern Gulf of Mexico (Weislogel et al., 2015; Snedden and Bovey, 2016).
The Woodbine, by contrast, represented a series of shorter river systems that drained the Ouachita Mountains, entered the Gulf of Mexico in the East Texas basin (see Ambrose et al., 2009), and was likely not significantly different than earlier Cretaceous Paluxy systems. The youngest U-Pb age within Woodbine detrital zircon populations was ca. 292 Ma, which indicated there was no connection with the Mesozoic Western Cordillera.
Albian-Cenomanian Dakota detrital zircon data presented here, and previous work from the U.S. midcontinent and Colorado Front Range, support this reconstruction. Detrital zircon data and mapped paleoflow directions (Witzke and Ludvigson, 1996; Brenner et al., 2000; Joeckel et al., 2005; Ludvigson et al., 2010) show that Albian-Cenomanian fluvial sandstones from Iowa, Nebraska, and Kansas represent an Appalachian source and westerly transport. By contrast, detrital zircon data and mapped paleoflow directions from the Colorado Front Range show Western Cordilleran source terrains and easterly transport (e.g., Weimer, 1984).
Collectively then, the midcontinent was traversed by fluvial systems that were flowing from the Appalachians to the eastern margin of the Western Interior seaway, whereas what is now the Front Range was traversed by fluvial systems flowing from the Western Cordillera to the western margins of the seaway. Neither of these systems was directly discharging sediment to the Gulf of Mexico per se, and during time periods when the seaway was at a minimum, these fluvial systems likely flowed north to the Boreal sea, rather than south to the Gulf of Mexico. The Appalachian-Ouachita Cordillera formed the eastern continental divide that separated paleodrainage to the seaway from that of the Gulf of Mexico. In this sense, mid-late Cretaceous drainage is a remnant of earlier patterns. During the Aptian, Appalachian-derived and Sevier-derived fluvial systems converged in the Sevier foreland-basin system backbulge prior to formation of the Western Interior seaway, then flowed north to the eastern margins of the Alberta foreland basin.
This continental-scale fluvial system is represented by the Aptian-Albian Mannville Group, which houses the Alberta Oil Sands (Blum and Pecha, 2014).
Implications for Cenomanian sediment-dispersal to deep water
Reconstruction of Cenomanian drainage areas from detrital zircon data provide for first-order estimates of the scales of basin-floor fans that are relevant to current exploration interests, and address the question of how far out into the basin sands can be expected. For example, most basin-floor fans in the modern world are 10-50% as long as the drainage basin that feeds them (Fig. 2). As applied to the Gulf of Mexico, this approach represents a test-of-method in a well-known data-rich basin, which can be then exported to basin margins that are data-poor.
Tuscaloosa detrital zircon data indicate the paleo-Tennessee system was the largest Cenomanian fluvial system contributing sediment to the northern Gulf of Mexico margin: these data are consistent with drainage basin lengths of 1200-1600 km, extending from the headwaters of the present Tennessee system to the Tuscaloosa-Woodbine shelf margin. Assuming scaling relationships of basin-floor fan to drainage-basin lengths of 0.1-0.5 from Somme et al. (2009), Tuscaloosa basin-floor fan length can be estimated at 120-800 km. Woodbine detrital zircon data indicate that smaller sub-regional systems characterized the western Gulf of Mexico, with maximum length scales of 200-300 km between the Ouachita source terrain, and the Woodbine shelf margin of the East Texas Basin. Assuming the same scaling relationships, basin-floor fan length can be estimated at 20-150 km, significantly smaller than predicted for the Tuscaloosa sediment-dispersal system.
Recent measurements of Cenomanian basin-floor fans from Snedden et al. (in review) fall within these domains. Basin-floor fan lengths were >600 km in length for the Tuscaloosa system, and <<100 km in length for the Woodbine system (Fig. 8).
This study uses a large set of U-Pb ages from detrital zircons to illustrate how detrital zircon data from the modern Mississippi system faithfully record source terrains, and defines Cenomanian drainage and sediment dispersal for the northern Gulf of Mexico and U.S. midcontinent. Results illustrate that Cenomanian drainage to the northern Gulf of Mexico was restricted to the Appalachian-Ouachita Cordillera and farther south. At this time, the Appalachian-Ouachita Cordillera formed the continental divide for eastern North America, the boundary that separated sediment routing through Tuscaloosa-Woodbine fluvial systems to the northern Gulf of Mexico, from sediment routing through Dakota Group fluvial systems to the eastern margins of the Western Interior seaway. Dakota fluvial systems in what is now the Colorado Front Range reflect source terrains in the Western Cordillera and flowed east to deliver sediment to the western margins of the seaway.
From detrital zircon data, Cenomanian Tuscaloosa strata in Alabama are interpreted to represent the paleo-Tennessee system, the largest sediment-dispersal system for the northern Gulf of Mexico during this time, having length scales of 1200-1600 km from source terrains in the Appalachians of Virginia and Tennessee to the shelf margin. This interpretation is consistent with point-bar thicknesses in Milliken et al. (2016), which indicate a drainage area of ~750,000 km2 : for comparison, this drainage area is roughly equivalent to that of the modern-day Ohio River but significantly smaller than the present-day Mississippi system. Cenomanian Woodbine strata from Oklahoma and Texas are interpreted to indicate a series of smaller river systems that headed in the Ouachita fold and thrust belt to the north, with maximum length scales of 200-300 km between source areas and the contemporaneous shelf margin.
These interpretations of fluvial-system scale are consistent with previous work from more traditional means, as well as point-bar measurements in Milliken et al. (2016). Moreover, predictions of basin-floor fan dimensions, using input from detrital zircon data, measured point-bar thicknesses, and published scaling relationships, agree well with existing data. Predicted drainage basin scales and basin-floor fan extents were modest during the Cenomanian, compared with those of the Paleocene Wilcox, or the Neogene Mississippi depocenter (Snedden et al., in review; Blum et al., in review).
The source to sink approach outlined and tested here first quantifies scales of the contributing drainage area and sediment routing system, then predicts scales of basin-floor fans using established scaling relationships. The Gulf of Mexico provides a unique opportunity for a test-of-method of this kind: this approach is robust at the first order, and can be applied in frontier basins, or in otherwise data-limited settings to predict potential reservoir presence. This approach can also be used to examine inconsistencies in preexisting paleogeographic interpretations.
Detrital zircon data were collected in 2011 when the senior author worked at ExxonMobil Upstream Research. We thank ExxonMobil for supporting that research, and for releasing data for publication. We also thank Mark Pecha and the Arizona Laserchron Center (University of Arizona) for processing the samples, and for numerous discussions. Last, Delores Robinson and an anonymous reviewer provided comments that improved the manuscript and figures.
Figures & Tables
Mesozoic of the Gulf Rim and Beyond: New Progress in Science and Exploration of the Gulf of Mexico Basin
- absolute age
- alluvial fans
- Atlantic Ocean
- clastic rocks
- Dakota Formation
- drainage basins
- fluvial features
- Front Range
- Gulf Coastal Plain
- Gulf of Mexico
- ICP mass spectra
- mass spectra
- North America
- North Atlantic
- Ouachita Mountains
- sediment transport
- sedimentary rocks
- Tuscaloosa Formation
- United States
- Upper Cretaceous
- Western Interior
- Western Interior Seaway
- Woodbine Formation
- zircon group
- detrital zircon
- Dakota Group