Nathan Henry, Archaeological Conservator
NC Underwater Archaeology Branch
January 2000 (Revised 2003)

Archaeologists are typically faced with the problem of determining whether apparent associations in an artifact assemblage are the result of cultural patterns or natural phenomena. The Queen Anne's Revenge shipwreck site is no different. By the fact that the wreck lies in an environment of extreme energy, that being the outer bar of a coastal inlet, all apparent artifact associations must be viewed with skepticism.

In 1978, Keith Muckleroy proposed a methodology for examining shipwreck sites that entailed looking at environmental parameters to better understand post-wreck artifact distributions. In his paradigm he named broad categories of events or processes that could potentially affect the "observed seabed distribution of any underwater site". These events and processes either disrupted the patterns of the assemblage so that the original artifact associations no longer existed (scrambling devices) or removed artifacts altogether (filters) (Muckleroy 1978).

All underwater sites are impacted to some degree by scrambling devices and filters. Some archaeological sites lay undisturbed under layers of sediment. Conversely, dredging activity may destroy all patterns in a site's artifact assemblage. Most sites fall somewhere in the middle. It is necessary for archaeologists to look hard at the variables that affect a site's integrity to interpret the degree to which cultural patterns of the artifact assemblage are preserved.

In the following paper, two broad categories of scrambling devices and filters will be discussed as they apply to the Queen Anne's Revenge shipwreck site. These are seabed movement and disintegration of perishables. Seabed movement is a scrambling device, but it is directly related to the disintegration of perishables (a filter). If the remains of a shipwreck are covered by sediment, they will be protected from destructive natural processes. Removal of the sediment subjects the shipwreck remains to physical, chemical, and biological destruction. It will be shown that these processes have affected the Queen Anne's Revenge site in the past, and threaten the site's future integrity.

Geographical Overview

The Queen Anne's Revenge (QAR) shipwreck site is situated on the ebb-tidal delta of Beaufort Inlet. Unlike most inlets occurring on the Outer Banks of North Carolina, the position and depth of Beaufort Inlet has been relatively stable throughout the historical period. Historical maps suggest its existence near its present position since 1672 (Cumming 1597, John Ogilby Map, 1672).
Shackleford Banks and Bogue Banks, two wave-dominated barrier islands, border the inlet to the east and west respectively and represent the southern terminus of the North Carolina Outer Banks. Because of the east to west orientation of Shackleford and Bogue Banks, Beaufort Inlet is well protected from northeast winds and associated storm waves that affect much of the North Carolina coast. Cape Lookout, one of three cuspate forelands found in North Carolina, is located ten miles east of Beaufort Inlet. This feature shelters the barrier islands to its west from easterly winds and influences longshore currents along their beaches (Sarl 1977:23).
Hydrologic energy (current) is the most influential factor affecting the near-shore geomorphology of the Outer Banks coastline. Two energy sources: tidal fluctuation and wind driven waves are responsible for these currents. The degree in which tidal currents affect the coast is related to the tidal range in a particular area. The North Carolina coast is considered a microtidal environment having a tidal range of less than 6 feet. Therefore the effects of currents generated by the tide are, generally speaking, secondary in importance to the effects of those produced by wind generated waves (Sarl l977:21; Davis 1994:3).
The area around an inlet is the exception to this generalization. The velocity of tidal currents is dramatically increased when the flow is concentrated through an inlet channel, resulting in significant scouring of bottom sediments. This scouring has an obvious effect on sediment transport around the inlet as well as the integrity of a shipwreck that may lie in the channel. Thus, tidal currents and wave-generated currents combine to influence the morphology of the ebb-tidal delta (Davis l994:24).

The Big Question

One of the questions facing researchers was how a ship that presumably ran aground on an inlet bar wound up in twenty-three feet of water. A ship the size of the Queen Anne's Revenge would have a depth of hold less than thirteen feet, making it impossible for the vessel to strike the bottom in twenty feet of water while afloat (Moore, 1999). If the shipwreck did not move dramatically from the position of its initial stranding, the depth of the water over the site has increased or the bottom has subsided over the last 282 years. Three processes have occurred to explain this.

Sea-level rose dramatically at the end of the Pleistocene Glacial Epoch: around 1 meter per century until about 7000 years ago. Since that time, it has continued to rise, though at a slower rate (Davis, 1994). It has been estimated that sea level has risen about 1 meter since QAR's sinking in 1718 (Wells, 2001).

There has been a general lowering of the ebb-tidal delta as a result of several factors, chief of which is the dredging of the inlet channel. It has been estimated that the total volume of sand making up the ebb-tidal delta has decreased 34% since extensive dredging activity began in the 1930's (Cleary, 1996)

The third factor is the laterally shifting ebb channel. If the channel became oriented over the site, there would be scattering of lighter or more buoyant artifacts, sediment scoured from beneath larger artifacts and the remaining assemblage would subside as the new channel deepened. Historic charts were consulted to determine if and when the channel passed over the site.

Cartography

Historic maps indicate that the orientation of the channel through Beaufort Inlet's ebb-tidal delta has been subject to continuous change, prior to artificial deepening and alignment that began in 1936 (Sarle, 1977:87). The University of North Carolina's Institute of Marine Sciences (IMS) undertook a project to analyze historic maps of Beaufort Inlet to determine if the ebb channel passed over the QAR site. Maps from the eighteenth century to the present were collected and screened for accuracy. Twenty-five of these maps were scaled to modern horizontal coordinates, corrected for sea-level rise, and digitized using AutoCAD. Despite the difficulty in using eighteenth century maps that lack an accurate coordinate system, nine distinct changes in inlet channel orientation were revealed (Wells, 2001)

A closer look at some of the maps used in the IMS project reveals that the channel has probably migrated over the site. It has been estimated that the QAR site has been uncovered by inlet-channel scour for 57 of the past 282 years (Wells, 2001).

The James Wimble (1738) map indicated a channel passing through the southwest corner of the ebb-tidal delta: an orientation that would put the channel west of the QAR site. This map depicted a continuous bar extending from Shackleford Island to the eastern edge of the channel. It was by plotting this bar on a modern NOAA chart that researchers determined where to search for QAR.

The Blount Survey chart (1830) located the channel just to the east of the QAR site in a southerly orientation. Two decades later, the J.N. Maffit Survey chart (1851) of Beaufort Harbor indicated that the channel migrated southeasterly and was oriented over the far-eastern edge of the ebb-tidal delta, skirting Shackleford Banks.

By the early twentieth century the channel returned to the center of the ebb-tidal delta. The E.I. Brown USACOE Survey chart (1908) indicated a southwest orientation of the channel with the center of the channel passing about 1200 feet to the east of the QAR site. It is interesting to note that though the channel was very close to the site, the depth of water over the site was around 1 foot at low water. This was due to the build-up of sediment on the flank shoals adjacent and parallel to the ebb channel.

Dredging of the Beaufort Inlet channel began in 1911 but no effort was made to stabilize its position until 1936. In the years between 1909 and 1936 the channel migrated about 3000 feet to the east, where it is currently maintained less than a mile to the east of the QAR site. This suggests that dredging never occurred directly over the site. Supporting this evidence are the artifacts themselves. In a relatively shallow (1 meter) trench excavated across the site, intact wine bottles, scientific instruments and depressions containing gold dust were discovered (Lusardi 1999). These artifacts would be easily broken or dispersed as a result of dredging.

Wave-Generated Currents

The nature of the current regime around the QAR site was documented by an InterOceans Systems® S4 current meter mounted above the sea floor approximately 20 meters west of the QAR site. The S4 measured current speed, direction, and wave data that were stored in the instrument's integrated memory. The S4 was deployed between March 1998 and June 1999, documenting a range of conditions including several hurricane events ((McNinch 2001).

The data retrieved from the current meter indicated that wave-induced currents are the primary cause of sediment transport on the QAR site. It has been calculated that a current velocity just under one-knot is required for sediment transport to begin, given the median grain size of sediment samples retrieved from the site. The S4 data indicated that in all conditions but the most severe storm event, that being Hurricane Bonnie, current velocities remained below the calculated critical threshold for sediment transport (McNinch, 2001). The currents associated with the passing of Hurricane Bonnie were significant. Wave heights measuring 2.5 meters generated sustained currents of two knots and occasional surges of 8 knots (McNinch 2001).

Hurricane Scour

The remains of the QAR were discovered in 1996 following Hurricane Fran. The most evident feature on the site was a compact pile of concreted artifacts that included anchors, cannons, and ballast stones. Though the pile's maximum height did not exceed the surrounding seafloor, it was visible due to the scouring of adjacent sediment, creating the illusion of a large mound rising above a featureless plane.

The area to the north of the artifact pile has been the subject of special concern. Storm- generated waves produced currents passing over and around the pile causing considerable scour of bottom sediments. In 1999, U.S. Army Corp of Engineers researchers conducted a high-resolution side-scan bathymetric survey using a SubMetrics® interferometric seabed inspection system (ISIS). This survey revealed that the scour fanned out to the northwest of the artifact pile at least twenty meters. This scouring reached its greatest depth adjacent to the pile and gradually lessened toward the extent (McNinch bathymetric survey 1999). The result of this scouring was the exposure of hull structure underlying the most severely scoured area of the site.

This scouring phenomenon apparently occurred during Hurricanes Bertha and Fran in 1996 since hull timbers were noticed during the initial site inspection by state archaeologists. The first major field project was conducted during 1997. At that time the hull structure was not evident, probably due to the fact that no hurricanes occurred that year. The following year scour resulting from Hurricane Bonnie re-exposed the timbers allowing thorough documentation by archaeologists during the 1998 field season. At the end of the 1998 project, archaeologists covered the timbers with protective sand-filled bags and a layer of dredged sand. Over the next year, archaeologists monitored the status of the timbers and noted that the scour area appeared to be filling with the absence of strong storms. Following Hurricane Dennis (1999), some scour was evident, with the sandbags and the tops of several hull timbers exposed. Just prior to the 1999 field project (October 4 to 15), Hurricane Floyd struck the coast. Diver inspection revealed extensive scouring to the north of the mound, sandbags scattered and undercutting of sediment beneath the hull structure. Archaeologists re-covered the timbers at the end of the project but Hurricane Irene's storm-generated currents negated these efforts within the week.

Though no current meter was present during the 1999 hurricanes, the direction of the current was evident. The scour was produced by eddies that formed on the down-current side of the artifact pile. Hurricanes Dennis and Floyd produced scour primarily to the northwest corner of the pile, though the depth and extent of the resulting scour was considerably greater from Floyd. Hurricane Irene produced significant scour to the northeast of the artifact pile. The different intensities and orientations of the scour footprints probably reflect the differing tracks of the storms. In fact, it is nearly impossible to predict the impact of hurricanes on the site because the storms vary so much in intensity and direction of travel. Both of these factors are directly related to the size and direction of the storm-generated waves (Wells 1999: personal comm.).

Disintegration of Perishables

The cycle of covering and uncovering of the QAR site is directly related to the survival, not only of the artifact associations but to the artifacts themselves. As artifacts become exposed to the seawater they are subjected to an array of destructive forces: physical, chemical and biological.

All exposed artifacts are susceptible to physical damage from currents, abrasion from sediment suspended in the currents, and larger objects being propelled by the currents. The affect of physical damage is most noticeable on artifacts made of malleable metals such as pewter and brass. Fragmented organic artifacts are simply washed away in the current. Fortunately iron artifacts become entirely encrusted with concretion material protecting them, to some degree, from physical damage.

Electrochemical corrosion affects most metallic artifacts when in contact with seawater. Galvanic cells are created between metallic artifacts or trace metals in the seawater resulting in the corrosion of the less noble metals. The process is accelerated when an aerobic environment is created around the artifacts. Oxidation works in concert with electrochemical corrosion to remove positively charged metal ions, thus destroying the integrity of the metal artifacts (Hamilton 1976).

Exposure of artifacts to the aerobic seawater allows the biological decomposition of organic materials. Wood, rope, leather, and cloth may survive centuries of inundation if covered by seafloor sediment. Upon exposure to the sea, a host of organisms attack these artifacts, eventually removing them from the assemblage altogether.

It has been documented that various species of marine fungi invade wood almost immediately upon its exposure to seawater. These organisms feed on the lumen in the cell walls, thereby thinning and weakening the wood fibers. Fungal attack is a precursor to infestation by marine borers (Kohlmeyer 1999).

Gribbles, isopods of the Limnoria genus occur in oceans throughout the world. Though small, typically 2mm in length, these animals are capable of inflicting considerable damage to exposed wood when large numbers of them are present. These creatures produce cellulase, much like terrestrial termites, that enables them to digest wood particles (Ray1959).

A far more insidious problem on the QAR site is wood boring mollusks of the Teredinidae family, commonly known as shipworms. These highly specialized molluscans make use of their shell to scrape a burrow through exposed wood, particles of which are utilized as food. Though the Terediniens siphon seawater, a requirement for respiration, they are capable of closing off their burrows with plumb-like pallets in times of stress. Laboratory tests have indicated that the animals can live in anaerobic conditions for a considerable time, continuing to burrow and ingest wood all the while (Ray, 1959).

All wood thus far examined on the QAR site shows evidence of damage from marine borers. The fact that hull structure is regularly exposed by storm induced scour, during prime Teredinien breeding season, has been a source of considerable frustration to archaeologists intent on preserving the wreck.

Conclusion

The inlet environment in which the Queen Anne's Revenge site exists has been and continues to be very dynamic. Sea-bed movement has been responsible for the loss of archaeological information, directly through scrambling of patterns in the artifact assemblage and indirectly by allowing the disintegration of perishable artifacts by exposing the artifact assemblage to the sea. On the other hand, sea-bed movement has protected the site for 225 of the past 282 years by covering the artifacts with sediment. Any meaningful interpretation of the present artifact assemblage must take the past effect of sea-bed movement into account. Likewise, management of the site requires consideration of future damaging impacts resulting from this naturally occurring process.


REFERENCES

CLEARY, W.J.
1996 Environmental Coastal Geology: Cape Lookout to Cape fear, North Carolina. Carolina Geological Society Fieldtrip Guidebook.

CUMMING, W.P.
1957 The Southeast in Early Maps. UNC Press, Chapel Hill, North Carolina.

DAVIS, R.A.
1994 Geology of Holocene Barrier Island Systems. Springer-Verlag Pub., New York.

HAMILTON, DONNY L.
1976 Conservation Of Metal Objects From Underwater Sites: A Study In Methods.
Texas Memorial Museum/Texas Antiquities Committee, Austin, Texas.

KOHLMEYER
1999 Notes from symposium held at Institute of Marine Sciences, 1999.

LUSARDI, WAYNE
1999 "Do Artifacts Identify the Beaufort Inlet Shipwreck as the Pirate Blackbeard's Flagship Queen Anne's Revenge? Underwater Archaeology. Society for Historical Archaeology Proceedings.

MCNINCH, JESSE E., J.T.WELLS, T.G. DRAKE
2000 "The Fate of Artifacts in an Energetic, Shallow-Water Environment: Scour and Burial at the Wreck Site of Queen Amme's Revenge. Southeastern Geology 40:1.

MOORE, DAVID D.
1999 "Historical and Archaeological Research Focused on the Hull Remains Associated with Site 0003BUI, Beaufort Inlet, North Carolina". Underwater Archaeology. Society for Historical Archaeology Proceedings.

MUCKLEROY, KEITH
1978 "The Archaeology of Shipwrecks" in Maritime Archaeology. Cambridge University Press, New York.

RAY, DIXIE LEE ed.
1958 Marine Boring and Fouling Organisms: Friday Harbor Symposia. University of Washington Press, Seattle, Washington.

SARLE, LAURA
1977 Processes and Resulting Morphology of Sand Deposits Within Beaufort Inlet, Carteret County, North Carolina. MA thesis, Duke University, Department of Geology.

WELLS, JOHN T. AND JESSE E. MCNINCH
2000 "Reconstructing Shoal and Channel Configuration in Beaufort Inlet: 300 years of Change at the Site of Queen Anne's Revenge. Southeastern Geology 40:1.

MAPS

Wimble, 1738 On file at N.C. Underwater Archaeology Branch.
Blount, 1830 On file at N.C. Underwater Archaeology Branch.
Maffit, 1851 On file at N.C. Underwater Archaeology Branch.
Brown, 1908 On file at N.C. Underwater Archaeology Branch.