Along the Atlantic Seaboard of the United States, approximately 20,000 known cultural sites are at risk of damage or complete loss due to coastal erosion, sea-level rise (SLR), and increasing numbers of catastrophic events such as hurricanes (Anderson et al. Reference Anderson, Bissett, Yerka, Wells, Kansa, Kansa, Myers, Carl DeMuth and White2017). Additionally, portions of the southeastern coastline are experiencing SLR at rates three times higher than the global average (Valle-Levinson et al. Reference Valle-Levinson, Dutton and Martin2017). Pockoy Island, the case study location presented in this article, is one of the most extreme examples of rapid erosion along the southeastern coastline, if not globally. The Pockoy Island Shell Ring Complex is located on the ancestral homelands of the Muscogee Nation, who still retain strong cultural ties to the region, and the Gullah Geechee people also live across our study area today. The island, over the past seven years, has experienced an erosion rate / shoreline retreat of 12.19 m/year inland, and as discussed below, more than 1 km over the past 171 years. Shoreline monitoring is important to landowners and land managers planning for conservation and research.
Beginning in February 2021, members of the Archaeology, Geology, and Wildlife Biology teams of the South Carolina Department of Natural Resources (SCDNR) began a collaborative shoreline monitoring project on Pockoy Island, a rapidly eroding barrier island that is part of SCDNR's Botany Bay Plantation Heritage Preserve / Wildlife Management Area (hereafter referred to as Botany Bay) on Edisto Island in Charleston County, South Carolina, USA (Figure 1). The long-term objectives of the project are as follows:
(1) To better understand the driving forces behind the landward movement of the shoreline (transgression)
(2) To collect data to inform and update the short-term rate of erosion of the shoreline in front of and adjacent to the Pockoy Island Shell Ring Complex (38CH2533), which has implications for archaeological mitigation
(3) To establish best practices for future long-term community science monitoring efforts, where “best” is defined as sustainable, efficient, and replicable practice
Three different shoreline monitoring methodologies—(1) high-resolution data acquisition (imagery and derivative elevation data) using an unoccupied aerial vehicle (UAV or drone), (2) high-resolution real-time kinematic (RTK) global navigation satellite system (GNSS) shoreline surveys using an Arrow Gold and smartphone or tablet with Eos Tools Pro and Esri ArcGIS Field Maps (formerly Esri ArcGIS Collector) applications installed, and (3) transect-based erosion measurements—were used to address these objectives and answer two specific research questions:
(1) How comparable are the data collected by the three monitoring methodologies?
(2) With the lessons learned from this project, which methods should be used to build a long-term community science volunteer program to assist with monitoring?
Although monitoring by agency staff is ongoing as of the writing of this article in August 2023, the results presented here represent the project's first two years, from February 2021 through December 2022. These first two years of shoreline monitoring serve as a how-to for other researchers, heritage stewards, and property managers, as the SCDNR team moves forward with establishing a long-term community science volunteer monitoring program based on our results.
Figure 1. Location map for Botany Bay Plantation Heritage Preserve / Wildlife Management Area. Top panel (Figure 1a) shows locations of Ring 1 and Ring 2 of the Pockoy Island Shell Ring Complex (38CH2533) on 2017 imagery, as well as the location of erosion transects used in analyses. Bottom panel (Figure 1b) shows locations of Ring 1 and Ring 2 over a digital elevation model derived from 2016 National Oceanic and Atmospheric Administration lidar data, as well as digitized lines indicating recent shoreline positions (2011 to 2022). Figure courtesy of Katie Luciano and Gary Sundin, SCDNR.
Management Context
SCDNR currently manages approximately 485,600 ha (1,200,000 acres) of land across South Carolina, and over 80,937 of those hectares (200,000 acres) are coastal. Some properties, like Botany Bay, are dedicated as Heritage Preserves, meaning that they are offered a high level of legal protection in perpetuity for both their natural and cultural resources. Heritage Preserves are considered, according to South Carolina's Heritage Trust Act, “irreplaceable as laboratories for scientific research; as reservoirs of natural materials . . . ; as habitats for rare and vanishing species; and as living museums” (South Carolina Code of Laws 2006). In essence, these places are a representation of the most significant parts of the state's natural and cultural heritage and are to be protected by the Heritage Trust Program for current and future generations.
Geographic Setting
Botany Bay is located within the Ashepoo-Combahee-Edisto (ACE) Basin National Estuarine Research Reserve (NERR), one of the largest remaining undeveloped wetland ecosystems on the US Atlantic coastline (Laurie and Harrigal Reference Laurie and Harrigal2009). In addition to the property's abundant and diverse natural resources—which includes federally endangered loggerhead sea turtles, neotropical songbirds (including painted buntings), and a variety of mammals, amphibians, fish, and shellfish—Botany Bay also contains multiple cultural sites dating back thousands of years. The most notable of these sites is the Pockoy Island Shell Ring Complex (38CH2533). The site was identified during a review of lidar imagery collected by the National Oceanic and Atmospheric Administration (NOAA) following Hurricane Matthew in 2016 (Figure 1b). At the time, the site consisted of two 60–70 m diameter shell rings and was constructed approximately 4,300 years ago (Smith et al. Reference Smith, Taylor, Ghaffar, Gaillard and Arrington2021). Pockoy Ring 1, the ring closest to the Atlantic Ocean, was lost to beachfront shoreline erosion over a period of approximately five years—from 2016 to 2021. This loss was verified by our team during on-the-ground shoreline monitoring of Pockoy Island (Figures 1b and 4).
Archaeological Context
Native Americans built shell rings along the Atlantic and Gulf Coasts of the southeastern United States between 3,000 and 5,000 years ago. In the area known as the Georgia Bight, shell rings are circular or C-shaped deposits of shell that range from approximately 40 m to 100 m in diameter. The rings were constructed from various shells such as oysters, whelks, arks, ribbed mussels, and periwinkles, as well as pottery sherds, stone, and animal bones. Although shell and other material make up shell rings, the centers of shell rings—also called the plazas—were purposefully maintained to be free of shell (Russo Reference Russo2002; Smith Reference Smith2023). In South Carolina, 24 shell rings have been recorded. SCDNR protects and manages one-third of them (Smith Reference Smith2023).
Coastal Geologic Setting and Erosion Trends
As recently as the mid-1990s, Pockoy Island was a hammock island surrounded by salt marsh and located landward of the active shoreline. Historic maps created by surveyors working with the United States Coast Survey (modern-day NOAA, National Ocean Service, and National Geodetic Survey) recorded the existence of an entire barrier island (Botany Bay Island) formerly located seaward of Pockoy Island in the mid-1800s (Figure 2). As stated above, since the late 1800s, the area has experienced one of the highest rates of erosion in coastal South Carolina, with the shoreline retreating over 1 km over the last 171 years (1851–2022). Higher rates of retreat have been measured to the north and south of Pockoy Island (Sexton and Hayes Reference Sexton and Hayes1983). The historical rate of erosion is −4.8 m per year (1851–2022). In the short term (2016–2022), the rate has averaged −12.19 m per year. During the short term (2016–2022), South Carolina's coastline has experienced increased exposure to the impacts of higher tides (king, or perigean, high tides) that have impacted coastal barrier island landforms including dunes (Harris and Ellis Reference Harris and Ellis2021).
Figure 2. US Coast Survey map (1851) of study area, overlying imagery from 2020. Figure courtesy of Katie Luciano, SCDNR.
The need for a rapid transdisciplinary research approach to help facilitate visitor interpretation and guide management decisions for both natural and cultural resources at Botany Bay is clear. Furthermore, because of the rapid erosion taking place at Pockoy Island, the property has served as a fast-track case study for our team so that we might develop a best-practice model moving forward. We feel that best practice for the monitoring of shoreline erosion at this and other sites is the ability for researchers and community science volunteers to perform safe and consistent repeat site visits that result in accurate quantifiable data, which, in turn, can enhance conversations about climate impacts and inform management decisions. Our project team was keenly aware of the need for researchers within our agency and state to develop techniques to measure risk that will benefit both natural and cultural resource managers. The results of our work have implications for property management plans, future research, community conversations and volunteer initiatives, and legislative policies that protect communities and resources and that provide a plan for resilience in the face of future climate-related impacts.
A Nested Methodology Approach
Globally, coastal researchers are documenting increasingly severe rates of climate-driven processes that are impacting, or even erasing, cultural heritage. Whereas archaeologists have turned their attention to the issue of heritage at risk and monitoring impacts to coastal sites in recent decades (Barker and Corns Reference Barker and Corns2023; Dawson et al. Reference Dawson, Nimura, López-Romero and Daire2017; Florida Public Archaeology Network 2020), coastal geologists have documented shoreline transgression (Leatherman Reference Leatherman2003; Morton Reference Morton1991). Our project team came together in 2021 to explore how a transdisciplinary approach might enhance and strengthen our individual and collective research approach and methodologies moving forward.
Each research team participating in the project—Archaeology, Geology, and Wildlife Biology—used a different shoreline monitoring methodology. Two methods were designated as high tech: a high-resolution UAV survey used by the Wildlife Biology team and an Arrow Gold RTK-GNSS survey used by the Archaeology team. A third method was designated as low tech: the on-the-ground erosion transect method used by the Geology team. The three teams all targeted low-tide events for data collection. Low tide provided the most flexibility in terms of transiting the beach between measurements (Archaeology and Geology teams) and was necessary for setting out ground control points and capturing the most extensive view of the shoreline for UAV flights (Biology team).
The Pockoy Island shoreline, from the southern end of Pockoy to the access causeway, was mapped on nine occasions by the Wildlife Biology team using a DJI Phantom 4 Pro quadcopter UAV to collect overlapping photographs of the beach and shoreline. Pix4D Mapper photogrammetry software (v.4.8.1, Pix4D SA, Switzerland) was used to create georeferenced orthoimagery and digital surface models (DSMs) suitable for use in geographic information system (GIS) analyses of the shoreline (Figure 3). DSMs are elevation models that capture both the topography of a landscape and objects on that landscape, such as trees and human-made features. A Trimble RTK-GNSS was used to record the locations of visible ground control point (GCP) targets, which were placed in the flight area before each flight. Targets were used to georectify the GIS map products during the processing phase. Flights were completed at 61 m (200 ft.) above the ground. Table 1 provides information about the UAV flights. The position of the edge of the tree canopy on each occasion was manually digitized from these drone products using ESRI GIS software (ArcGIS v.10.8.0, ESRI, Redlands, California). A colorized and partially transparent DSM was overlaid on the orthoimagery to create a highly contrasting visual cue for manual digitizing (Figure 3a). Each shoreline was digitized as a single linear feature representing the position of the beach-vegetation interface at the time of the flight. Analyzing Moving Boundaries Using R (AMBUR) software (Jackson Reference Jackson2018) was used to calculate erosion rates for the shoreline over the project period (Figure 3c). AMBUR is an R software language package that allows users to calculate rates of change of linear features over time using inputted digitized linear features created in GIS. AMBUR uses observed changes in shoreline position at multiple user-defined transects to calculate end point rate (EPR), the total distance between the oldest and youngest shorelines divided by the number of years, and the linear rate of regression, which is the rate of change based on a line fitted through three or more points representing shoreline position at different times.
Figure 3. Examples of UAV-derived GIS products used for Analyzing Moving Boundaries Using R (AMBUR) analyses of shoreline change at Pockoy Island. Top panel (Figure 3a) shows the initial and final digitized shorelines analyzed for the project over December 2022 UAV imagery overlaid with a partially transparent digital surface model (DSM). This view was used to digitize shorelines. Middle panel (Figure 3b) shows digitized shorelines over initial and final digitized shorelines over February 2021 UAV imagery. Bottom panel (Figure 3c) shows an example of AMBUR transects (red lines) cast perpendicularly across digitized shorelines (black lines) at 5 m intervals along the southern shoreline of Pockoy Island, overlaid on UAV-derived imagery from December 2022. The length of each transect represented the total shoreline change (erosion) over the analysis period. Figure courtesy of Gary Sundin, SCDNR.
Table 1. Summary of Flight Data for Nine UAV Flights Collecting Data at Pockoy Island.
Note: Includes the flight date, area mapped, number of ground control point (GCP) targets used, and total number of photographs collected. Table courtesy of Gary Sundin, SCDNR.
An Arrow Gold RTK-GNSS survey was used by the Archaeology team on 16 occasions. Point data were collected along the scarped shoreline from the southern end of Pockoy Island to the causeway used for pedestrian access to the island (Figure 1b) and a few meters beyond the causeway when possible. The Arrow Gold offered the possibility of collecting points with subcentimeter accuracy when conditions were ideal. The Arrow Gold was connected to either an iPhone or iPad with the Esri Field Maps (formerly Esri Collector) application used to record shoreline measurements and attach notes and photos when needed. Approximately 30 points per visit were collected where the shoreline could be safely reached. Points were processed into lines (Figure 4) in ArcGIS to measure shoreline loss using AMBUR. Point data collected with the Arrow Gold during four of these 16 visits that showed some of the most extreme shoreline transgression was used to compare to the other two methods within this project (Figure 5, top panel).
Figure 4. This figure depicts the locations of Ring 1 and Ring 2 of the Pockoy Island Shell Ring Complex (38CH2533) with an overlay of the eroding shoreline. The shorelines of Pockoy Island were measured by Arrow Gold in May 2021, December 2021, and December 2022. Figure courtesy of Karen Y. Smith, SCDNR.
Figure 5. Top panel shows a comparison of three project methodologies. Bottom panel shows comparative results of the three methodologies used within this project from March 2021 to July 2022. Figure courtesy of Katie Luciano, SCDNR.
The methodology used by the Geology team was based on other transect-based methods used in previous studies to quantify dynamic boundary lines along shorelines and rivers (Buzard et al. Reference Buzard, Overbeck and Maio2019; Gatto Reference Gatto1988; Hudson Reference Hudson1982; Sandweiss and Kelley Reference Sandweiss and Kelley2012; Saynor and Erskine Reference Saynor and Erskine2006). Five shore-perpendicular transects were established along Pockoy Island in March 2021 (Figure 1a): two north (T1, T2) and two south (T4, T5) of a central transect (T3) established at the Causeway running through the center of the island. Transects were spaced 100–200 m apart. Given that a goal of the project was to better understand patterns of shoreline change adjacent to or in front of the Pockoy Island Shell Ring Complex (38CH2533), T4 and T5 were installed close to Ring 1 and Ring 2, respectively. Measuring change at the Causeway was considered important because it is a landmark for property managers and the public, who use it to traverse from an upland parking lot to the beach. T1 and T2 were established north of T3 to allow for measurements in a rapidly eroding section of the island, where washover is common. Each transect included a front and back stainless-steel benchmark rod. The distance between the benchmark rods at each transect varied from 14 m at T3, to 10 m at T1 and T2, to 5 m at T4 and T5. This distance between the rods was largely a function of line of sight along the transects. T1 and T2 are situated in settings clear of large trees and dense maritime forest. T4 and T5 are located in areas of Pockoy Island that have denser vegetation, where establishing and maintaining a longer line of sight is more difficult. At T3, transect benchmark rod spacing (14 m) was greater because these rods were visible to the public, and the spacing was necessary to ensure that the rods were out of direct sight. Transects were established along the same azimuth (138°, which is roughly shore perpendicular) with a Brunton pocket transit, using a similar methodology as described in Buzard and colleagues (Reference Buzard, Overbeck and Maio2019). A Trimble RTK-GNSS was used to record positional data (latitude, longitude, and elevation) at each transect.
Transects were measured monthly and after perigean, or king, tide events and tropical storm or hurricane events to capture episodes of rapid transgression. A metric measuring tape was attached to the back benchmark and run alongside the front benchmark to the seaward point of erosion (Figure 6)—approximately the same point at which the Arrow Gold would be placed along the shoreline. These measurements were recorded in a field book. This technique was used 26 times over the period of study.
Figure 6. Erosion indicators function as the measuring points for on-the-ground transect methodology. Indicators can include highest points of debris line (a, b, c), highest points of dunes or berms (e), or distinct scarps (d, f). Figure courtesy of Katie Luciano, SCDNR.
All data collected for this project, including information that supplemented and informed this project, such as property manager correspondence (discussed below), were archived and shared among the research team through the SCDNR secure servers.
Discussion
Although these three methods appear neat and tidy on the surface, no project is without hiccups. In December 2022, the seaward rod of Transect 1 at the northernmost end of Pockoy Island, used by the Geology team, was lost. Overwash in the area, which includes piles of downed trees, following several transgressive events in late 2022—most notably, Hurricanes Ian and Nicole—made it challenging to find a measuring point marking the active beach face. Benchmark rods at Transect 1 and Transect 2 were both bent landward following the Hurricane Nicole transgression event in November 2022.
The seaward benchmark rods at Transect 3, Transect 4, and Transect 5 were within several meters of the active shoreline. These transects were leapfrogged landward in early 2023 so that those transect rods would not be easily lost and to allow for measurements to continue as the project progressed. To do this, the seaward rod at each transect was removed and a new rod was put in place landward of the current landward rod. A compass was used to keep transects on the established azimuth. Although Pockoy Island is an extreme case of rapid erosion, it is assumed that if this method were to be used on other properties with lower rates of erosion, benchmark rods would not be threatened by shoreline transgression so quickly.
Although the high-tech methods have the advantage of not requiring equipment to be left in the ground, where there are exposed to high tides and waves, these methods also have limitations. They are costly to purchase up front, require software licenses, and—in our experience—prove temperamental or unusable at times when weather conditions are poor. Whereas the Wildlife Biology team's UAV was grounded on occasion due to high winds and the Archaeology team waited for satellites to connect to the Arrow Gold knowing that time on the beach was limited due to the incoming tide, the transect method used by the Geology team allowed for accurate measurements to be collected even in adverse conditions.
In addition to these three methods, it should be noted that Botany Bay Volunteer Coordinator Bess Kellett served as our daily eyes and boots on the ground, providing the project team—which is based hours away from Botany Bay in Columbia and Charleston, South Carolina—with updates and photos as conditions on the island changed rapidly. Her emails and text messages containing photo and notes about major changes were and still are incredibly valuable to the project and showcase how critical transdisciplinary collaboration and local community involvement and near-daily monitoring is at heritage at-risk sites such as this. Local community science volunteer efforts have been essential at heritage at-risk sites globally (Dawson et al. Reference Dawson, Hambly, Lees and Miller2021; Graham et al. Reference Graham, Hambly and Dawson2017), given that the resources of professionals can be limited. Moving forward, a community science monitoring effort and community conversations—including surveys and interviews—about heritage at risk will be launched in late 2024.
The average rate of erosion at Pockoy Island from March 2021 to July 2022 ranged from 5.77 m to 6.46 m/year and was broadly similar among the three methods (Figure 5). Of course, impacts from hurricanes, storm surges, and king-tide events increase what might be considered “average,” as reflected in the 12.19 m/year average from 2016 to 2022 as stated above. Due to its rapid rate of erosion, Pockoy Island has allowed our project team to work on a fast-forward case study for shoreline monitoring and set wheels in motion to collaboratively address best practices for monitoring the impacts to cultural and natural resources on this and other properties in a rapidly changing environment. The objectives for this project were three-fold:
(1) We wanted to better understand the driving forces behind the landward movement of the Pockoy Island shoreline, or transgression.
(2) We wanted to collect data to inform and update the short-term rate of erosion of the shoreline in front of and adjacent to the Pockoy Island Shell Ring Complex (38CH2533), which has implications for archaeological mitigation.
(3) We wanted to establish best practices for future, long-term community science monitoring efforts so that we can expand monitoring statewide, incorporating community conversations about heritage at risk.
As stated above, we feel that best practice for the monitoring of shoreline erosion at this and other sites is the ability for researchers and community science volunteers to perform safe and consistent repeat site visits that result in accurate quantifiable data, which in turn can enhance conversations about climate impacts and inform management decisions. Our collaborative shoreline monitoring project on Pockoy Island has established practices that are replicable, efficient, and sustainable, and they are the foundation for an enduring community science program similar in many respects to national environmental monitoring programs already supported by our agency within the fields of climatology, marine science, and ornithology.
Overall, the initial project results have shown that the three methods provided very similar results in terms of the final erosion rate (Figure 5, bottom panel). Our work also shows that regular, repeat visits to properties and repeat measurements are necessary to establish a longer-term erosion rate, regardless of method and property. Finally, and to drive home the point of repeat measurements even more, collecting measurements and real-time on-the-ground observations of the loss of cultural materials monthly using the erosion transect method allowed for the impacts of specific events—such as the November 2021 king-tide event—to be captured. Monthly monitoring may not be practical or even possible for many heritage stewards, property managers, or community scientists. However, for sites that are eroding as quickly as Pockoy, repeat visits following events such as storms and king tides capture important data that can help managers understand physical factors (i.e., tide heights and water levels, wind patterns) that may be causing erosion. Analyzing these data points, when and if they are available, can help in understanding the impacts that climate-related events have on sites at risk. At the end of the day, although any data consistently and accurately captured and shared is a positive step toward better understanding the loss of sites threated by erosion, we hope this article serves as a call to action for those who face the growing threat of climate-related impacts to find a steadfast path forward to document and address the loss of cultural heritage.
Moving forward, the low-tech, on-the-ground erosion transect approach tested within this project will serve as the foundation for community science monitoring for at-risk sites on SCDNR properties. We found, after comparing data collected by this method to data collected by the Arrow Gold, that the two methods produce comparable results (Figure 5, top panel). In the end, the measurements from the two methods were within 10 cm of one another, but the difference in price between the two methods is substantial. Whereas the on-the-ground transect method costs a few hundred dollars to purchase benchmarks, tape measures, field books, and a good pair of tall boots, the Arrow Gold alone is about $10,000, plus the cost of a smartphone or tablet and software. Even with labor costs added in, the differences are substantial. For those who are not able to leave benchmarks in the ground and elect to use a high-tech on-the-ground method, we would like to note there are lower-cost RTK GNSS receivers (e.g., Emlid) and free GIS software options (e.g., QGIS) that can be substituted for ArcGIS and Arrow Gold (used in our study).
Predictable erosion is compounded annually by increasing numbers of catastrophic events that result in impacts to living communities and the loss of cultural sites. As both researchers and community members of these impacted landscapes, how can we rapidly mobilize and strengthen practical, collaborative, and data-driven methods to monitor and predict climate and erosional impacts to cultural heritage? In the interest of developing accessible, cost-effective community science monitoring methods to document erosion at heritage at-risk sites, we hope that this study has shown how we need to put all monitoring options on the table, examine them with a collaborative lens, and always keep our options open as new opportunities emerge. Sometimes simple, low-tech methods that have stood the test of time for decades truly are our best way forward as we face climate change together.
Acknowledgments
SCDNR would like to acknowledge and pay respect to the Indigenous inhabitants of the land on which we work. Since time immemorial, Tribes inhabited, protected, and preserved these lands until they were forcibly removed from them. We respect their histories and the connections that they maintain to this area. We thank the following individuals for their assistance with field work, GIS mapping, and monitoring: Jessica Boynton (SCDHEC-OCRM), Evan Cook (Moore Farms Botanical Garden, formerly SCDNR-MRD), Will Doar (SCDNR-LWC, Geological Survey), Lauren Faulk (SCDNR-MRD, Shellfish Research Section), Bess Kellett (SCDNR-MRD, Botany Bay Volunteer Coordinator, Retired), and Aaron Ward (SCDNR-WFF, Botany Bay Property Manager). We would like to thank Emma Paz for Spanish language translations. We would also like to thank the numerous staff of the ACE Basin NERR. This publication represents Marine Resources Research Institute Contribution Number 878. Permits were not required for this research.
Funding Statement
This work was supported by the South Carolina Department of Natural Resources (SCDNR).
Data Availability Statement
Data are housed at the South Carolina Department of Natural Resources (SCDNR).
Competing Interests
The authors declare none.
