Facies Reconstruction of a Late Pleistocene Cypress Forest Discovered on the Northern Gulf of Mexico Continental Shelf Suyapa Gonzalez1, Samuel J. Bentley, Sr.1, Kristine L. DeLong3, Kehui Xu2, Jeffrey Obelcz2, Jonathan Truong1, Grant L. Harley4, Carl A. Reese4, and Alicia Caporaso5 1 Coastal Studies Institute and Department of Geology and Geophysics, Louisiana State University, 331 & E235 Howe-Russell Geoscience Complex, Baton Rouge, Louisiana 70803 2 Coastal Studies Institute and Department of Oceanography and Coastal Sciences, Louisiana State University, 2165 Energy, Coast & Environment Bldg., Baton Rouge, Louisiana 70803 3 Department of Geography and Anthropology, Louisiana State University, 227 Howe-Russell Geoscience Complex, Baton Rouge, Louisiana 70803 4 5 Department of Geography and Geology, University of Southern Mississippi, 118 College Dr., Box 5051, Hattiesburg, Mississippi 39406 Bureau of Ocean Energy Management, 1201 Elmwood Park Blvd., New Orleans, Louisiana 70123 ABSTRACT A previously buried bald cypress forest (Taxodium distichum) was discovered on the continental shelf seafloor, offshore of Orange Beach, Alabama, USA, in ~20 m water depth. The forest was likely buried in the late Pleistocene, possibly exhumed by Hurricane Ivan in 2004, and is now exposed as stumps in life position. In August 2015 and July 2016, submersible vibracores and geophysical data were collected to investigate local stratigraphy and mode of forest preservation. This study focuses on analysis of the longest and most stratigraphically complete vibracore, DF–1 (4.78 m). This core revealed, from top to bottom, a surface of Holocene transgressive sands, underlain by interbedded sand and mud (potentially Holocene or Pleistocene), overlying a swamp or delta plain facies (likely Pleistocene) containing woody debris and mud that has been provisionally dated using radiocarbon to ca. 41–45 ka. One core collected in 2016 revealed a Pleistocene paleosol beneath Holocene sands in a nearby trough. We hypothesize that floodplain aggradation in the area was a key factor that might have allowed forest preservation. A sea-level rise pulse of 10–15 m occurred ca. 40 ka that could have produced widespread floodplain aggradation, likely burying the swamp and forest sediments. During the subsequent glacial lowstand, sediments that comprise the floodplain were eroded and paleosols were formed in other nearby areas. It is hypothesized that some swamp sediments located in paleo-topographic lows might have been preserved and buried due to the deep coverage of the eastern-trending channel infill sediments. Coastal wave erosion during transgression likely eroded high ground but enough sediment remained to keep the cypress forest blanketed and therefore allowed preservation of stumps and woody debris. INTRODUCTION The Gulf of Mexico is a complex and dynamic system that is governed by fresh water discharge (e.g., Morey et al., 2003), open ocean circulation (e.g., Hamilton, 1990), changes in sea level (e.g., Törnqvist et al., 2004), Gonzalez, S., S. J. Bentley, Sr., K. L. DeLong, K. Xu, J. Obelcz, J. Truong, G. L. Harley, C. A. Reese, and A. Caporaso, 2017, Facies reconstruction of a late Pleistocene cypress forest discovered on the northern Gulf of Mexico continental shelf: Gulf Coast Association of Geological Societies Transactions, v. 67, p. 133–146. 133 Gonzalez et al. tropical storms (e.g., Stone et al., 2004; Reese et al., 2008), and subsidence in some locations (e.g., Turner, 1991). Even though many studies have been conducted regarding the geologic setting and sea level variations in the area during the Holocene, relatively little is known about the Quaternary geology and stratigraphy of the northeastern Gulf of Mexico, especially along the continental shelf south of Alabama. Previous studies (e.g., McBride and Byrnes, 1995; McBride et al., 1996, 1999; Bartek et al., 2004) have determined that the regional stratigraphy of the inner continental shelf south of Alabama, our study area, consists of Holocene sands containing abundant shells, underlain by estuarine and other coastal deposits of Pleistocene to Holocene age (Figs. 1 and 2). During the past decade, an example of tree stumps, mostly bald cypress trees (Taxodium distichum), in life position have been identified exposed on the continental shelf seabed, south of the Alabama shoreline. Their water depth and location initially suggested their ages to be late Pleistocene to early Holocene. The location of these well preserved tree stumps is approximately 13 km offshore of Gulf Shores, Alabama, at 20 m below present sea level (Figs. 1–3). This buried cypress forest can provide valuable insight into paleoenvironmental conditions, such as sea level variations, climate, and position of ancient shorelines. This is the first time a study has been conducted that focuses on the geologic framework and mode of preservation of this recently discovered buried cypress forest. This is an ongoing multidisciplinary research project in which paleoenvironmental observations and other analyses (pollen, foraminifera, absolute dating, and geophysical analyses) have been conducted by scientists and students from Louisiana State University and University of Southern Mississippi (Gonzalez et al., 2016; Ryu et al., 2016; Obelcz, 2017). This paper focuses on the geological setting and stratigraphy of this unique study area. STUDY AREA The study area for this project encompasses approximately 2 km2 of the northern Gulf of Mexico continental shelf (Fig. 1), which we have cored and mapped geophysically. It is approximately 15 km south of Gulf Shores, Alabama. To the north, it is bounded by the Alabama shoreline, to the west by the St. Bernard lobe of the Mississippi River Delta, and to the southeast by the DeSoto Canyon. Water depths at the study site for both surveys (August 2015 and July 2016) were approximately 18 to 20 m. The seabed encompassing the study area is known as the MAFLA (Mississippi-Alabama-Florida) sand sheet, which is also subdivided into two sub-provinces, the Fig. 2 Fig. 3 Figure 1. Map of study area showing approximate location of the ancient forest (blue box) within the context of the northern Gulf of Mexico coast. 134 Facies Reconstruction of a Late Pleistocene Cypress Forest Discovered on the Northern Gulf of Mexico Continental Shelf Figure 2. Time structure map of MIS (Marine Isotope Stage) 2 sequence boundary showing the two incised valleys located south of the modern Alabama shoreline (modified after Bartek et al., 2004). Contour interval is 10 ms. Modern shelf break is found at the 120 m contour. Red box shows approximate location of study site. 135 Gonzalez et al. Figure 3. Bathymetric map of study site (see blue box on Figure 1 for location) showing DF−1 core location, collected in August, 2015. 136 Facies Reconstruction of a Late Pleistocene Cypress Forest Discovered on the Northern Gulf of Mexico Continental Shelf Mobile and the Apalachicola sand sheets (Mazzulo and Peterson, 1989). From recent geophysical and seafloor surveys, the location of drowned cypress trees has been identified (Obelcz, 2017). GEOLOGIC SETTING The Gulf of Mexico during the Holocene Sea level rose rapidly from the Last Glacial Maximum, when global sea levels were as much as 125 m lower than present, to present late Holocene levels, decelerating around 6000 yr ago (Fig. 4). The Last Glacial Maximum was encompassed by Marine Isotope State (MIS) 2 (29–14 ka), followed by MIS 1 (14–0 ka) (Lisiecki and Raymo, 2005). The Holocene slowdown in rate of sea level rise during this time period allowed for the development of coastal land forms in the area. Since that time, shoreline position has been stable (Donoghue, 2011). The trees (Taxodium distichum), which produced stumps preserved in the study area, generally grow between 0–30 m above sea level in flood plain and riparian environments in the humid subtropics (Little, 1971). Based on this observation, approximate time frames can be identified when the forest may have grown, by studying the sea-level curves like the one in Figure 4. The depth range of 18–20 m intersects sea level near times of ~10,000 ka, 110,000 ka, and 125,000 ka. Therefore, the time between about 10,000 ka and about 110,000 ka thus represents a possible time interval during which this forest may have grown. Figure 4. Global sea level variations since 70 ka. Depth range of core DF–1 (20 m) and sample (DF1– 414; 4.14 m below seafloor; horizontal bar) and radiocarbon age of sample (vertical bars) plotted on eustatic sea level maximum and minimum range. Dashed lines represent the maximum and minimum sea level from Waelbroeck et al. (2002). 137 Gonzalez et al. METHODS Field Work and Data Collection Vibracoring Vibracores were collected offshore of Gulf Shores, Alabama, in the study area in August 2015 and July 2016 onboard the R/V Coastal Profiler of Louisiana State University’s Coastal Studies Institute (CSI), in water depths of about 20 m. A submersible vibracorer was used to collect cores up to 5 m in length. Geophysical Data Swath bathymetry, CHIRP sub-bottom, and sidescan sonar data (Obelcz et al., 2014) were collected during both surveys. Geophysical instruments used in both surveys (2015 and 2016) were: EdgeTech 2000 combined dual frequency side scan sonar and sub-bottom profiler on the starboard side, EdgeTech 512i sub-bottom profiler on the port side, as well as EdgeTech 4600 swath bathymetry and side scan sonar on the bow. Of these datasets, only swath bathymetry is presented here in Figure 3. Data collection and processing methods are described in Obelcz et al. (2014). Data Processing and Analysis Gamma Density and Core Imaging Soon after core collection, cores from the 2015 and 2016 surveys were logged using a Geotek Multi-Sensor Core Logger (MSCL) for whole-core gamma density of sediments. Digital images of freshly split core were collected using a Geotek Geoscan core-imaging camera system. Grain Size Sub-samples of DF–1 underwent grain size analysis on a Beckman Coulter LS 13–320 Laser-Diffraction Particle Analyzer. Each sub-sample was submerged in a 0.05% sodium phosphate solution to deflocculate particles. Samples were then sieved through 850 μm sieves to get rid of any particles too large for the instrument. An aliquot of 30% hydrogen peroxide was added to samples that were then placed in a hot bath overnight to remove organic matter. Samples were then further disaggregated prior to analysis with an ultrasonic probe. Loss on Ignition (LOI) and Organic Content Organic-content analysis for core DF–1 was conducted using the loss on ignition (LOI) technique (Heiri et al., 1999). Sub-samples were taken every 5 cm. Radiocarbon Dating Sediment geochronology using 14C dating methods was used to obtain dates for the Pleistocene sediments and bald cypress trees found in the study site. A sub-sample from core DF–1 was extracted at 4.14 m depth. This sub-sample was located within a peat layer (Fig. 5). Seven additional samples were also taken at the following depths: 3.22 m, 4.05 m, 4.19 m, 4.24 m, 4.56 m, 4.07 m, and a second sample at 4.14 m. All samples were sent to Beta Analytic Radiocarbon Dating Laboratory in Miami, Florida, for sample preparation and dating. 138 Facies Reconstruction of a Late Pleistocene Cypress Forest Discovered on the Northern Gulf of Mexico Continental Shelf Figure 5. (Left) Highresolution imagery of core DF–1 showing facies. (Right) Pollen (not part of this study) and radiocarbon sample locations in core DF–1. 139 Gonzalez et al. RESULTS Gamma Density and Core Imagery Radiocarbon Dating The initial dates for 14C obtained from the sub-sample found in coarsely fibrous peat in core DF–1 at 4.14 m depth revealed a radiocarbon age of 37.35 ± 0.33 ka (41.83 cal ka with a range of 42.235 to 41.350 cal ka at 95% probability, where cal ka is thousands of calendar years before 1950 Common Era (CE). Calibrated age was determined using INTCAL13 (Reimer et al., 2013) following the parameters of Talma and Vogel (1993). Of the other seven samples sent to Beta Analytic, six came back as radiocarbon ‘dead’ (Table 1). The sample at 4.05 m depth in DF–1, which is located at the boundary between interbedded peat and mud and overlying interbedded mud and sand (described below), had a radiocarbon age of 41.83 ± 0.88 ka, and median calibrated age of 45.210 cal ka, and 95% probability range of 46.690–43.625 cal ka). This sample is older than the deeper 4.14 m peat sample. Because these radiocarbon ages are near the upper limit of reliable detection (ca. 40 ka), other cores from the site are currently in the process of being dated with alternate absolute dating methods, such as optical stimulated luminescence (OSL). This additional dating will be used to confirm these 14C results. With this caveat, we used these dates cautiously for a preliminary core chronostratigraphy. A grain-size profile versus depth of core DF–1 was generated from laser-diffraction analysis (Fig. 6). Grain size varies greatly throughout the core. The upper ~3.10 m recovered in this core are sand-sized sediments where sand makes up roughly 65% of the entire core. The mean grain size for this section is 255 μm. The scatter plot shows high variations in the interval from 3.10 to 4.05 m, where mean grain size through this interval varies from 37 to 221 μm (Fig. 6). This is consistent with the varying gamma density readings for this interval, representing interbedded sand and mud. Average grain size for this interval is 133 μm. The deepest section, 4.05–4.78 m, shows relatively uniform particle sizes of silty and muddy sediment (organics were removed as described above) with only slight variations around mean values of ~36 μm. Table 1. Radiocarbon dates for sediment samples taken from core DF–1. Results were obtained from Beta Analytic Radiocarbon Dating Laboratory in Miami, Florida. DF–1 Sediment Samples Radiocarbon Results Sample Type Sample # Sample Name δ13C 14 Sediment 322 cm Beta–433024 DF1–322 -30.1 >43500 Sediment 405 cm Beta–433025 DF1–405 -28.8 Sediment 414 cm Beta–429909 DF1–414 Sediment 414 cm Beta–433026 Sediment 419 cm ± Cal BP age Cal BP range 41830 880 45210 46690–43625 -23.0 37350 330 41830 42235–41350 DF1–414–2 -28.9 >43500 Beta–433027 DF1–419 -28.7 >43500 Sediment 424 cm Beta–433028 DF1–424 -31.0 >43500 Sediment 456 cm Beta–433029 DF1–456 -27.4 >43500 Sediment 466 cm Beta–433030 DF1–466 -28.5 >43500 140 C age Facies Reconstruction of a Late Pleistocene Cypress Forest Discovered on the Northern Gulf of Mexico Continental Shelf Figure 6. DF–1 comprehensive graph showing % organic content, mean grain size, and gamma density vs. depth profile. Organic content increases with depth downsection. Classification of facies is shown. Note: % organic content (horizontal axis) numbers were adjusted to show organic content variably throughout the core. 141 Gonzalez et al. Loss on Ignition (LOI) LOI results allow determination of organic content in core DF–1 (Fig. 6). Throughout Figure 6, variability in organic content can be observed downsection. The sandy, uppermost 3.10 m has an average organic content of 0.36%. The adjacent unit of interbedded sand and mud, from 3.10–4.05 m has an average organic content of 2.19%. Highest organic content occurs in beds in the 4.05–4.78 m depth range, with a maximum of 16.80% at 4.15 m depth, coincident with peat-rich beds visible in core imagery (Fig. 5), which also have low bulk density (Fig. 6). Overall, this unit (4.05–4.78 m depth) of the core showed the highest organic content percentage with an average of 8.91%. INTERPRETATION Lithofacies Based on gamma density, grain size, and sediment observations, three different lithologic units (facies) characterize the sedimentary deposits found in the study area. From the gamma-density profile of core DF–1, higher density readings are associated with larger grain sizes and lower porosity, like those in sand-sized sediments. In contrast, lower density readings correlate to smaller grain sizes and higher porosity found in sediments like clays and silts, and zones of high organic content. Facies descriptions are derived from the primary core of this study, DF–1. Two of the facies (facies 1 and facies 2 below) are consistent with the lithologic descriptions of McBride et al. (1999). Sediment ages from 14C geochronology (4.05 and 4.14 m) in core DF–1 help establish a provisional chronostratigraphy. Facies descriptions below are from the top to the bottom of the core. Facies 1: Holocene Sand Facies 1 (0–3.10 m) is characterized by light, beige to grey, fine- to medium-grained quartz sand containing abundant bioclasts and shell fragments. Color becomes progressively lighter upsection, with some intervals slightly darker than others. Sediments are moderately to well sorted. Minor planar bedding is present. The paleontological assemblage identified so far is comprised of various gastropods, bivalves, bryozoans, foraminifera, and ostracods; these are still in the process of being fully identified. Most of the bioclasts found are highly fragmented and vary in size. This facies is consistent with facies 5 and 6 from McBride et al. (1999) described in that study as primarily sand, and shelly sand. These sand sheet deposits are thickest on the ridges and thinnest or completely absent in the troughs of the study site (see Figure 3). Facies 2: Interbedded Sand and Mud Facies 2 (3.10–4.05 m) is characterized by light to medium dark grey mud interbedded with very finegrained sub-bioclastic sand, and to some extent shell fragments. Fine-grained sand occurs in beds that are well sorted and have no sedimentary structures present other than prominent sand-mud interbedding. Beds are mostly 1–3 cm thick. In some intervals, mud is more abundant than sand. Comparison of this bedding with the interbedded bioturbated sand and mud described by Bentley et al. (2002) and possibly consistent with facies 3 of McBride et al., (1999) suggests that this unit may be estuarine. Preliminary evaluation of microfossils suggests a brackish-water setting but the exact age and depositional setting are not yet known. If these sediments are estuarine, they must have been deposited when sea level was near to or shallower than 24 m below present. This in turn implies that the age of this unit is ~8,000–10,000 yr, or early Holocene at the oldest, based on age-depth distributions in Figure 4, from the global sea-level curve of Waelbroeck et al. (2002). This interpretation, combined with age estimates for the underlying unit, suggests that an unconformity most likely exists between facies 2 and facies 3, but the exact location and age spanning of the unconformity is not clear with this dataset, and considering possible differences between Gulf of Mexico sea levels and the Waelbroeck et al. (2002) global sea level record. 142 Facies Reconstruction of a Late Pleistocene Cypress Forest Discovered on the Northern Gulf of Mexico Continental Shelf Facies 3: Interbedded Mud and Peat This basal facies (4.05–4.78 m) is characterized by dark grey, tan brown to dark brown muds and peat. Some intervals are darker in color than others, especially those where wood debris is found. Organic material and wood debris are abundant. Wood fragments are light brown in color and roughly 1–2 cm thick. No sedimentary bed forms are present, but bedding is clear in alternating peat and mud layers. Wood debris visible in core may correlate stratigraphically to the bald cypress stumps exposed in the scoured basin evident in Figure 3. This is also the facies sampled for 14C dating. Macro and microfossils are absent in the samples studied so far, but pollen analysis is currently being conducted to further characterize this facies. Ages of 14C in this unit of core DF –1 range from ~41,000 to >50,000 yr (i.e., radiocarbon dead), suggesting an approximate time of MIS 3–4 (Fig. 4). This facies was not identified by McBride et al. (1999). Sediment properties of this unit are consistent with deposition in a bald cypress swamp (producing peat and woody debris) in a river floodplain (providing mud beds from episodic flooding) (Smith and Bentley, 2015; Heitmuller et al., 2017). Overall, there is a coarsening upwards sequence in core DF–1, where all three facies are bioturbated to some extent. DISCUSSION Regional Stratigraphy The depositional environments and regional stratigraphy of the study area during the Pleistocene and the Holocene have been greatly impacted by changes in sea level in the northern Gulf of Mexico. Consequently, events of marine transgression and regression have played a major role in shaping the stratigraphy that is observed at this site today. Core DF–1 (4.78 m long) contains the most complete and varied stratigraphic record of all cores collected to date for this study. The three facies identified in DF–1 are: a Holocene sand sheet; interbedded sand and mud (potentially estuarine, age unknown); and Pleistocene interbedded mud and peat, from an ancient river flood plain (this assessment is also supported by preliminary analysis of seismic data; Obelcz, 2017). Consistency was found between some of these facies and those described by McBride et al. (1999). However, our identification of Pleistocene river-swamp deposits appears to be unique. The base of core DF–1 may stratigraphically correlate the stump-bearing horizon, but this must be confirmed at a later date by integration of core data with geophysical data (e.g., Obelcz et al., 2014). Based on the 14C date obtained from the sediments collected at 4.14 m depth in DF–1, it is likely that the forest was buried at or before 42.2 to 41.4 cal ka. The fact that many of the radiocarbon dates for the bald cypress wood and sediments came back inconclusive, suggests that the trees and forest might be older than ca. 45,000 yr, but these results are inconclusive and other dating technique are required to further constrain the age of the forest. Effects of Sea Level Change & Flood Plain Aggradation on Forest Preservation This unique scenario in which bald cypress trees and enclosing river-swamp deposits have been preserved for likely over 45,000 yr is a geological puzzle. Age (see 14C results) and depth relationships (Fig. 4) cannot with certainty explain how the cypress forest could withstand periods of erosion during the late Pleistocene and Holocene. Sea level near that time was in the range of -50 to -90 m below present sea level (Fig. 4), and regional topographic relief is relatively low. One interpretation of these observations is that the swamp developed at an elevation above and possibly distant from sea level and the shore. This is also consistent with the structure map of the MIS 2 land surface in Figure 2, adapted from Bartek et al. (2004). These deposits were buried and protected from erosion during a period of perhaps 30,000 yr of regression, transgression, and likely extensive subaerial exposure. For this to happen, it is likely that these deposits were originally buried by a substantial (and unknown) thickness of sediments, which were mostly but not entirely eroded during Pleistocene exposure and Holocene transgression. One potential process that could have produced such sediment burial is flood plain aggradation. A temporary sea-level rise of 10–15 m occurred ca. 40 ka, which could have produced local flood plain aggradation that would have likely buried the swamp and forest sediments. Such rapid aggradation has been observed many kilometers inland in other coastal-plain alluvial valleys (Shen et al., 2012, 2015). During lowstands, 143 Gonzalez et al. sea level falls and the sediments that comprise the floodplain are eroded. Subsequently, paleosols form in other nearby areas as documented by McBride et al. (1999) as a basal facies in some of their cores. The results presented here are at the northern limit of the area surveyed by Bartek et al. (2004). Bifurcation of the Mobile-Tensaw River system during the MIS 2 resulted in two incised valleys being re-occupied south of Mobile Bay (Fig. 3). One eastern-trending incised valley and one western-trending incised valley were identified by Bartek et al. (2004). The present site is located near the area encompassed by the eastern incised valley (red box on Figure 2). Sea level rise patterns in the area during the Holocene are interpreted to have led to transgression of the depositional environments associated with the incised valleys, therefore changing the position of the coastline. Figure 2 shows the approximate location of our study site near the eastern trending incised valley from the Bartek et al. (2004) study. We propose that swamp sediments such as in DF–1, located in topographic lows, might have been preserved and buried by fluvial/flood plain sediment accumulation. Finally, as sea level rose and transgression occurred, sediments within the incised valleys were first eroded by transgression of the coastline and associated wave energy and later covered by Holocene shelf sands as sea level rose, thus causing the coastline position to retreat. Coastal wave erosion during transgression likely eroded high ground but enough sediment remained to keep the cypress forest blanketed, therefore fortuitously favoring preservation. CONCLUSIONS This unique scenario where cypress trees have been preserved for possibly over 45,000 yr represents a puzzling geologic setting. Relatively rapid and deep burial by flood plain aggradation prevented exposure by channel incision and coastal erosion during the latest Pleistocene regression and Pleistocene-Holocene transgression, thus preserving the tree stumps and wood debris. 14C dates were near the upper limits of this dating method, therefore more absolute ages are needed to establish a geochronology that will enhance our understating of the area’s geology and means of forest preservation. If this hypothesis is ultimately proven correct by further study, this could help identify the locations of other similar drowned forests of similar age around the Gulf of Mexico, because the depth range of such forest preservation would be caused by the regional impact of sea-level rise and associated flood plain transgression, similar to our study area, forming a sort of ‘bathtub ring’ around the Gulf of Mexico at common age and depth. Also, these sites should be concentrated offshore of modern river systems that would have incised the shelf at lower sea levels and deposited sediments during the phase of flood plain aggradation around 40 ka ago that likely preserved our drowned forest. REFERENCES CITED Bartek, L. R., B. S. Carbote, T. Young, and W. Schroeder, 2004, Sequence stratigraphy of a continental margin subjected to low-energy and low-sediment-supply environmental boundary conditions; late Pleistocene−Holocene deposition offshore Alabama, U.S.A., in J. B. Anderson and R. H. Fillon, eds., Late Quaternary stratigraphic evolution of the northern Gulf of Mexico margin: Society of Economic Paleontologists and Mineralogists Special Publication 79, Tulsa, Oklahoma, p. 85–109. Bentley, S. J., T. R. Keen, C. A. Blain, and W. C. Vaughan, 2002, The origin and preservation of a major hurricane event bed in the northern Gulf of Mexico: Hurricane Camille, 1969: Marine Geology, v. 186, p. 426–446. Donoghue, J. F., 2011, Sea level history of the northern Gulf of Mexico coast and sea level rise scenarios for the near future: Climatic Change, v. 107, p. 17–33. Gonzalez, S., S. J. Bentley, K. DeLong, K. Xu, G. L. Harley, C. A. Reese, J. Obelcz, and J. Ryu, 2016, Geologic setting and preservation of a late Pleistocene bald cypress forest discovered on the northern Gulf of Mexico continental shelf: 2016 Ocean Sciences Meeting, New Orleans, Louisiana, February 21–26, Last accessed August 6, 2017. Hamilton, P., 1990, Deep currents in the Gulf of Mexico: Journal of Physical Oceanography, v. 20, p. 1087–1104. Heiri, O., A. F. Lotter, and G. Lemcke, 2001, Loss on ignition as a method for estimating organic and carbonate content in sediments: Reproducibility and comparability of results: Journal of Paleolimnology, v. 25, p. 101–110. 144 Facies Reconstruction of a Late Pleistocene Cypress Forest Discovered on the Northern Gulf of Mexico Continental Shelf Heitmuller, F. T., P. F. Hudson, and R. H. Kesel, 2017, Overbank sedimentation from the historic AD 2011 flood along the Lower Mississippi River, USA: Geology, v. 45, p. 107–110. Lisiecki, L. E., and M. E. Raymo, 2005, A Pliocene-Pleistocene stack of 57 globally distributed benthic 18O records: Paleoceanography, v. 20, Paper PA1003, 17 p. Little, E. L., Jr., 1971, Atlas of United States trees, volume 1, conifers and important hardwoods: U.S. Department of Agriculture Miscellaneous Publication 1146, 9 p., 200 maps. Mazzullo, J., and M. Peterson, 1989, Sources and dispersal of late Quaternary silt on the northern Gulf of Mexico continental shelf: Marine Geology, v. 86, p. 15–26. McBride, R. A., and M. R. Byrnes, 1995, Surficial sediments and morphology of the southwestern Alabama/western Florida Panhandle coast and shelf: Gulf Coast Association of Geological Societies Transactions, v. 45, p. 392– 404. McBride, R. A., M. R. Byrnes, L. C. Anderson, and B. K. Sen Gupta, 1996, Holocene and late Pleistocene sedimentary facies of a sand rich continental shelf: a standard section for the northeastern Gulf of Mexico: Gulf Coast Association of Geological Societies Transactions, v. 46, p. 287–299. McBride, R. A., L. C. Anderson, A. Tudoran, and H. H. Roberts, 1999, Holocene stratigraphic architecture of a sandrich shelf and the origin of linear shoals: Northeastern Gulf of Mexico, in K. M. Bergman and J. W. Snedden, eds., Isolated shallow marine sand bodies: Society of Economic Paleontologists and Mineralogists Special Publication 64, Tulsa, Oklahoma, p. 95–126. Morey, S. L., P. J. Martin, J. J. O'Brien, A. A. Wallcraft, and J. Zavala‐Hidalgo, 2003, Export pathways for river discharged fresh water in the northern Gulf of Mexico: Journal of Geophysical Research: Oceans, v. 108, no. C10, Paper 3303, 15 p. Obelcz, J., B. Brothers, J. Chaytor, U. T. Brink, S. W. Ross, and S. Brooke, 2014, Geomorphic characterization of four shelf-sourced submarine canyons along the U.S. Mid-Atlantic continental margin: Deep-Sea Research II, v. 104, p. 106–119. Obelcz, J., 2017, Sediment transport and slope stability in the northern Gulf of Mexico: Ph.D. Dissertation, Louisiana State University, Baton Rouge, 89 p. Reese, C. A., T. P. Strange, W. D. Lynch, and K. Liu, 2008, Geologic evidence of Hurricane Katrina recovered from the Pearl River marsh, MS/LA: Journal of Coastal Research, v. 24, p. 1601–1607. Reimer, P. J., E. Bard, A. Bayliss, J. W. Beck, P. G. Blackwell, C. Bronk Ramsey, C. E. Buck, H. Cheng, R. L. Edwards, M. Friedrich, P. M. Grootes, T. P. Guilderson, H. Haflidason, I. Hajdas, C. Hatté, T. J. Heaton, D. L. Hoffmann, A. G. Hogg, K. A. Hughen, K. F. Kaiser, B. Kromer, S. W. Manning, M. Niu, R. W. Reimer, D. A. Richards, E. M. Scott, J. R. Southon, R. A. Staff, C. S. M. Turney, and J. van der Plicht, 2013, IntCal13 and Marine13 radiocarbon age calibration curves 0–50,000 years cal BP: Radiocarbon, v. 55, p. 1869–1887. Ryu, J., K. DeLong, S. J. Bentley, K. Xu, G. L. Harley, C. A. Reese, J. Obelcz, and T. Guilderson, 2016, New evidence on an ancient bald cypress forest on the inner shelf of northern Gulf of Mexico: 2016 Ocean Sciences Meeting, New Orleans, Louisiana, February 21–26, Last accessed August 6, 2017. Shen, Z., T. E. Tornqvist, W. J. Autin, Z. R. P. Mateo, K. M. Straub, and B. Mauz, B, 2012, Rapid and widespread response of the Lower Mississippi River to eustatic forcing during the last glacial-interglacial cycle: Geological Society of America Bulletin, v. 124, p. 690–704. Shen, Z., T. E. Törnqvist, B. Mauz, E. L. Chamberlain, A. G. Nijhuis, and L. Sandoval, 2015, Episodic overbank deposition as a dominant mechanism of floodplain and delta-plain aggradation: Geology, v. 43, p. 875–878. Smith, M., and S. J. Bentley, Sr., 2014, Sediment capture in flood plains of the Mississippi River: A case study in Cat Island National Wildlife Refuge, Louisiana, in Y. Jun Xu, M. A. Allison, S. J. Bentley, A. L. Collins, W. D. Erskine, V. Golosov, A. J. Horowitz, and M. Stone, eds., Sediment dynamics from the summit to the sea: International Association of Hydrological Sciences Publication 367, p. 442–446. 145 Gonzalez et al. Stone, G. W., B. Liu, D. A. Pepper, and P. Wang, 2004, The importance of extratropical and tropical cyclones on the short-term evolution of barrier islands along the northern Gulf of Mexico, USA: Marine Geology, v. 210, p. 63– 78. Talma, A. S., and J. C. Vogel, 1993, A simplified approach to calibrating 14C dates: Radiocarbon, v. 35, p. 317–322. Törnqvist, T. E., J. L. González, L. A. Newsom, K. van der Borg, A. F. de Jong, and C. W. Kurnik, 2004, Deciphering Holocene sea-level history on the U.S. Gulf Coast: A high-resolution record from the Mississippi Delta: Geological Society of America Bulletin, v.116, p. 1026–1039. Turner, R. E., 1991, Tide gauge records, water level rise, and subsidence in the northern Gulf of Mexico: Estuaries and Coasts, v. 14, p. 139–147. Waelbroeck, C., L. Labeyrie, E. Michel, J. C. Duplessy, J. F. McManus, K. Lambeck, E. Balbon, and M. Labracherie, 2002, Sea-level and deep water temperature changes derived from benthic foraminifera isotopic records: Quaternary Science Reviews, v. 21, p. 295–305. 146