What is the capacity of both our institutions and our local ecosystems to mediate the pressures of demographic growth, climate change, and/or sustained exploitation of resources? What kinds of strategies are best positioned to facilitate the long-term sustainability of natural resources? How do the actions of local communities contribute to and structure sustainable strategies and outcomes? Nowhere are these questions and challenges more critical today than across our coastal zones, where populations continue to grow and climate change impacts—including sea-level rise and warming oceans—are being felt and experienced on a very human scale (Hauer et al. Reference Hauer, Evans and Alexander2015, Reference Hauer, Saunders and Shtob2022; Neumann et al. Reference Neumann, Vafeidis, Zimmermann and Nicholls2015). Understanding these issues at multiple temporal, spatial, and social scales is critical for successful analysis and implementation of sustainable practices (Gardner et al. Reference Gardner, Ferreira, Barlow, Lees, Parry, Vieira and Berenguer2013; Jonas et al. Reference Jonas, Ometto, Batistella, Franklin, Hall, Lapola and Moran2014).
For millennia, societies across the globe have navigated, shaped, and responded to the complexities and sensitivities of coastal marine and estuarine ecosystems (Erlandson Reference Erlandson2001; Holland-Lulewicz et al. Reference Holland-Lulewicz, Ritchison, Holland-Lulewicz, Howland, Thompson and Thompson2025). Although the archaeological record of Indigenous North America provides a key case to evaluate mechanisms of resource use and sustainability under varying ecological and social conditions, such contributions have been hindered by expectations set forth in the archaeological legacies of unilineal, evolutionary schemes. How people “get food” has served as a catchall set of categorical schemas, limiting our ability to investigate the vast variability in past Indigenous resource management (Kidder Reference Kidder, Sassaman and Holly2011:118–119; Lepofsky et al. Reference Lepofsky, Letham, Ritchie, Armstrong, Fitzpatrick and Erlandson2026; Lyons et al. Reference Lyons, Hoffmann, Miller, Martindale, Ames and Blake2021; Price and Brown Reference Price and James1985). Indeed, a reliance on categorical approaches inhibits our ability to explore relationships and mechanisms of phenomena (Holland-Lulewicz Reference Holland-Lulewicz2021). Terms, such as “hunter-gatherers” or “agriculturalists,” remain tied to these legacies, disproportionately affecting the kinds of questions we have asked about strategies of resource management and sustainability by nonfood producing societies. Thus, this study investigates resource management practices of eastern oyster (Crassostrea virginica) at a single, large Indigenous village of fisher-hunter-gatherers on the Georgia coast (USA) over the entirety of its approximately 425-year life-span (Figure 1). Using absolute chronologies and oyster morphometric analyses, this study represents the largest analysis of oyster resource management practices yet to be conducted using the archaeological record to evaluate sustainability at the scale of a single community, providing critical contributions to our understanding of the intersections of sedentary hunter-gatherers, resource management strategies, and sustainability that are often masked by the deployment of categorical schema.
Image of the Georgia coast (USA) showing the series of barrier islands, including those mentioned in the study (map by Isabelle Holland-Lulewicz). (Color online)
Sustainability, Resource Management, and Hunter-Gatherers
To understand how diverse resource management practices contribute to sustainable outcomes, it becomes especially important to move beyond essentialized approaches and to instead investigate variation in the institutions guiding governance—or rather, the regularized rules, norms, and traditions structuring human-environment interactions—given that resource management essentially relies on the effective management of people as they move across the landscape and interact with both each other and their broader ecosystems (Berkes and Folke Reference Berkes, Folke, Fikret and Carl1998; Holland-Lulewicz et al. Reference Holland-Lulewicz, Conger, Birch, Kowalewski and Jones2020; Kelly Reference Kelly1992, Reference Kelly2013:77–113; Ostrom Reference Ostrom2011; Wu Reference Wu2021). Broad generalizations of hunter-gatherers as highly mobile, egalitarian communities masks critical local (and global) variation in sociopolitical and socioeconomic institutional diversity that impacts our understanding of how people manage resources (see Kidder Reference Kidder, Sassaman and Holly2011; Price and Brown Reference Price and James1985; Sampson Reference Sampson2023; Thompson Reference Thompson2023). Our focus should not be on such categorizations (e.g., “hunter-gatherers do this, agriculturalists do this”) but rather on investigations concerned with the institutions that facilitated cooperation and collective action (Thompson Reference Thompson2023:507), and in turn, the key relational factors (sensu Blanton and Fargher Reference Blanton and Fargher2016) that facilitate sustainable outcomes. In fact, to understand how people “get stuff done,” we must understand these behaviors at varying scales and within different economies (Blanton and Fargher Reference Blanton and Fargher2016:6; Jennings Reference Jennings2017:1387; Kowalewski Reference Kowalewski and Birch2013; Thompson Reference Thompson2023:507). Throughout this article, I use these terms simply to facilitate discussion and to highlight how the institutions associated with resource management strategies within so-called sedentary fisher-hunter-gatherer villages represent a key source of variation currently underdeveloped and undertheorized.
How hunter-gatherers or fisher-hunter-gatherers develop and maintain relationships with resources has long been studied with primary reference to degrees of mobility and the access to resources across landscapes that such mobility affords. For example, Binford (Reference Binford1980) unpacked and defined a systematic relationship between mobility and a group’s interactions with particular resources. Kelly (Reference Kelly2013:77–113) built on Binford’s (Reference Binford1980) model to more functionally investigate how we might understand hunter-gatherer mobility as it relates to important social and economic variation, discussing direct evidence for hunter-gatherers as conservationists or active resource managers. We know sophisticated strategies of resource management existed in the past, much of which would have been sustainable through and dependent on generational knowledge of landscapes and resource ecologies (Arnold Reference Arnold1996; Grier Reference Grier, Finlayson and Warren2017; Holland-Lulewicz Reference Holland-Lulewicz2024; Jackley et al. Reference Jackley, Gardner, Djunaedi and Salomon2016; Lepofsky et al. Reference Lepofsky, Smith, Cardinal, Harper, Morris and Bouchard2015; Trosper Reference Trosper2002). However, how we understand the way mobility structures human-environment relationships, especially in the context of sustainability, has been explored most extensively in sedentary agricultural societies and, to a lesser extent, highly mobile hunter-gatherers. How people cooperate around the large-scale harvest of resources is not necessarily dependent on degree of mobility (Bettinger Reference Bettinger2015; Thompson Reference Thompson2023), but I argue that the village itself—not often associated conceptually with hunting and gathering societies—represents a key institution among such societies, and that sedentary fisher-hunter-gatherer villages represent an undertheorized context to explore how institutional organizations, like a village, structure (or are functionally related to) resource management strategies to promote sustainability. Villages, and their role in intra- and intercommunity cooperative relationships, certainly represented key institutional structures that either served to manage resources (Thompson Reference Thompson2023) or at the very least impacted how resources were managed.
At the risk of essentializing “village” itself as a reified concept, I use it here broadly, contextualized by regional sociopolitical histories. Indeed, across the ethnographic and archaeological records, villages do not represent singular social, political, or economic forms of any single type (though this variation is seldom explored outside of agriculturally based communities [Bandy and Fox Reference Bandy and Fox2010:6]). Here, the term “village” is used to describe a relatively permanent, multifamily residential aggregation that facilitated regular or semiregular interactions among the population (Bandy and Fox Reference Bandy and Fox2010; Birch and Thompson Reference Birch and Thompson2018; Copeland Reining Reference Copeland Reining, Reining and Lenkerd1980). Defining the scope of the term a bit further, “village” is used here not just as a spatial scale but as a social scale, representative of a meaningful institution within the coastal societies that are the focus of this study. “Villages as institutions” is certainly not universal (Holland-Lulewicz et al. Reference Holland-Lulewicz, Conger, Birch, Kowalewski and Jones2020). In the coastal American Southeast, however, villages as key social and economic institutions have long histories: they represent not just population aggregations but critical extra-household, extra-familial organizations. The US Southeast sees villages form as early as approximately 5000 BP along both the Atlantic and Gulf Coasts as circular settlements, with organized layouts and evidence for key institutional roles for the village itself (e.g., central cremations and structures; associated communities of practice in craft production; DePratter Reference DePratter1979; Gibson Reference Gibson2001; Kidder Reference Kidder, Sassaman and Holly2011; Pluckhahn Reference Pluckhahn2010; Russo Reference Russo1998; Sanger Reference Sanger2015; Sassaman Reference Sassaman2005; Thompson Reference Thompson2007; Thompson and Andrus Reference Thompson and Andrus2011; Thompson and Holland-Lulewicz Reference Holland-Lulewicz2026; Thompson and Worth Reference Thompson and Worth2011; Turck and Thompson Reference Turck and Thompson2016). Over the next five millennia, villages waxed and waned in their importance as institutions in Indigenous societies of the coast. By AD 800, large aggregate villages (re)emerged on the Georgia coast as key institutions, with planned settlement features such as plazas, organized burial and platform mound layouts relative to other settlement features, and a domestic “ring” of residential zone. This importance of villages to sedentary fisher-hunter-gatherers is not unique to the Georgia coast. Indeed, fisher-hunter-gatherers across North American hold long histories of villages as important social formations underwritten economically by nondomesticated resources (Grier et al. Reference Grier, Alessa and Kliskey2017; Holland-Lulewicz Reference Holland-Lulewicz2024; Martindale et al. Reference Martindale, Marsden, Patton, Ruggles, Letham, Supernant, Archer, McLaren and Ames2017; Prentiss et al. Reference Prentiss, Cross, Foor, Hogan, Markle and Clarke2008, Reference Prentiss, Foor, Hampton, Walsh, Denis and Edwards2023; Thompson and Worth Reference Thompson and Worth2011). Although the dynamics of agricultural villages, and the role of the village in the socioeconomics of resource management, have been extensively theorized for agriculturally based societies, these key local institutions and their articulations with sustainable exploitation by hunting and gathering societies remains underappreciated.
Such an appreciation, however, is key in that local institutions (such as the village) often represent a critical scale at which rules, regulations, and practices are situated that serve to structure control and access to resources, facilitate conflict management and mediation related to resource use, and are the very mechanisms through which management strategies are institutionalized (Aiuvalasit Reference Aiuvalasit, Lozny and McGovern2019; Jacobson Reference Jacobson2022). In this sense, village-scale processes, as both spatial and social scale, are key to understanding socially meaningful units of interactions (Lock and Molyneaux Reference Lock and Molyneaux2006; Peterson and Drennan Reference Peterson and Drennan2005; Yaeger and Canuto Reference Yaeger and Canuto2000). Institutions situated at these scales most directly determine use and access among a populace. Long-term trends and broad-scale, regional patterns are driven by these short-term, spatially small-scale decisions. Indeed, spatial heterogeneity in resource abundance and characteristics—as well as social, political, and economic heterogeneity across communities—necessitates nuanced understandings of the relationship between local (e.g., village) histories and regional trends (Prentiss et al. Reference Prentiss, Cross, Foor, Hogan, Markle and Clarke2008, Reference Prentiss, Foor, Hampton, Walsh, Denis and Edwards2023). Relationships between people and their resources, the concomitant institutions governing these relationships, and the sustainable outcomes of such actions must be considered across geographic scales and across the variation of social, political, and economic forms (Jacobson Reference Jacobson2022; LeFebvre et al. Reference LeFebvre, Erlandson and Fitzpatrick2022).
Sustainability, or the ability of the present to meet its needs without compromising the needs of future generations, remains a concept central to many scientific and academic disciplines, including archaeology, that aim to understand human-ecosystem dynamics (Berkes and Folke Reference Berkes, Folke, Fikret and Carl1998; Komiyama and Takeuchi Reference Komiyama and Takeuchi2006; LeFebvre et al. Reference LeFebvre, Erlandson and Fitzpatrick2022). Literature across the sustainability sciences similarly recognizes not only the importance and complexity of understanding these dynamics at different scales—usually defined as the global (planetary base for human survival), social (political and economic institutions of humans), and human (all the factors that affect human survival)—but also that these scales are inextricably linked (Clark and Harley Reference Clark and Harley2020; Komiyama and Takeuchi Reference Komiyama and Takeuchi2006; Wu Reference Wu2021). Processes situated at differing spatial and temporal scales may articulate in important, complex ways, in which the analytical isolation of processes at a single scale does not provide a holistic picture of key articulations (Wu Reference Wu2021). Alternatively, signatures of sustainability, or processes driving sustainability at one scale, may not translate across scales. For instance, patterns suggesting sustainability at the regional scale may not describe or reflect sustainable practices at local scales and within communities, or vice versa. Archaeology as a discipline has long addressed the long-term, complex relationships of people and ecosystems (Balée and Erickson Reference Balée and Erickson2006; Fitzhugh et al. Reference Fitzhugh, Butler, Bovy and Etnier2019; Redman Reference Redman1999) and continues to contribute substantially to sustainability sciences more generally (Costanza et al. Reference Costanza, Graumlich, Steffen, Crumley, Dearing, Hibbard, Leemans, Redman and Schimel2007; Crumley Reference Crumley2021; Frazier Reference Frazier2007; Lertzman Reference Lertzman2009). However, accomplishing this requires archaeological research that contextualizes resource management strategies beyond broad categorizations of past human societies (e.g., farmer vs. hunter-gatherer) and accounts for organizational variability (e.g., sedentary hunter-gatherers), the resource base itself, local geographic variation, population pressures, and more (Kelly Reference Kelly2013:112, 242).
Shellfisheries
Coastal fisheries, both fish and shellfish, serve as prime examples to archaeologically explore the relationships between sustainable resource management/extraction practices among sedentary hunter-gatherer groups. Much of this important institutional variation has been investigated along the US West and northwest coasts and southwestern Florida (Grier et al. Reference Grier, Kim and Uchiyama2006; Sampson Reference Sampson2023). Long-term histories of shellfishery management have been demonstrated across the United States, especially in the Chesapeake Bay region of the eastern Atlantic seaboard (Jenkins and Gallivan Reference Jenkins and Gallivan2019; Reeder-Myers et al. Reference Reeder-Myers, Braje, Hofman, Smith, Garland, Grone and Hadden2022; Rick et al. Reference Rick, Reeder-Myers, Hofman, Breitburg, Lockwood, Henkes and Kellogg2016), along the Pacific coast (Erlandson et al. Reference Erlandson, Rick, Braje, Steinberg and Vellanoweth2008; Groesbeck et al. Reference Groesbeck, Rowell, Lepofsky and Salomon2014; Grone Reference Grone2020; Lepofsky et al. Reference Lepofsky, Smith, Cardinal, Harper, Morris and Bouchard2015; Toniello et al. Reference Toniello, Lepofsky, Lertzman-Lepofsky, Salomon and Rowell2019; Whitaker Reference Whitaker2008), the Channel Islands of California (Braje et al. Reference Braje, Erlandson, Rick, Dayton and Hatch2009, Reference Braje, Rick and Erlandson2012; Erlandson et al. Reference Erlandson, Rick, Braje, Steinberg and Vellanoweth2008, Reference Erlandson, Braje, Rick, Jew, Kennett, Dwyer, Ainis, Vellanoweth and Watts2011), and globally (Campbell Reference Campbell2010; Gillies et al. Reference Gillies, McLeod, Alleway, Cook, Crawford, Creighton and Diggles2018; Giovas et al. Reference Giovas, Fitzpatrick, Clark and Abed2010; Kataoka Reference Kataoka1996; Milner et al. Reference Milner, Barrett and Welsh2007; Robson et al. Reference Robson, Hausmann, Laurie, Astrup, Povlsen, Sørensen, Andersen and Milner2024; Thomas Reference Thomas2019). Oyster reef estuarine ecosystems have characterized what is known today as the Georgia Bight, off the Atlantic Coast, for the last 5,000 years, and these were sustainably exploited for millennia despite research demonstrating the sensitivity of shellfish and other species in the intertidal zones to human predation (Garland et al. Reference Garland, Thompson, Sanger, Smith, Andrus, Lawres and Napora2022; Lulewicz et al. Reference Lulewicz, Thompson, Cramb and Tucker2017; Reeder-Myers et al. Reference Reeder-Myers, Braje, Hofman, Smith, Garland, Grone and Hadden2022; Ritchison et al. Reference Ritchison, Thompson, Lulewicz, Tucker and Turck2021; Thompson et al. Reference Thompson, Rick, Garland, Thomas, Smith, Bergh and Sanger2020). In fact, commercial oyster fisheries today face concerns regarding sustainable management (Tookes et al. Reference Tookes, Barlett and Yandle2018). Recent studies from the American Southeast suggest the presence of sustainable strategies of oyster harvest since the Late Archaic, around 5,000 years ago (Andrus and Thompson Reference Andrus, Fred and Victor2012; Garland and Thompson Reference Garland and Thompson2023). The majority of this extant work on fisheries and shellfisheries in each of these North American Indigenous contexts has generally been situated at large geographic scales of analysis, including the macroregion (e.g., the Atlantic coast or Pacific coast), mesoregional scales (e.g., the Georgia Bight or coastal British Columbia), or at the island-specific scale (e.g., Ossabaw Island on the Georgia coast or San Miguel Island off the coast of California), but rarely at the scale of a single community or village, where the institutionalization of certain practices and management strategies was likely situated.
The large-scale, long-term resolutions of recent, regionally focused studies, although critical to our understandings of sustainability and long-term Indigenous ecosystem management, do not necessarily capture the ebbs and flows of resource management systems and the short-term, locally specific pressures relevant to how people sustainably manage resources and navigate environmental impacts. Extant synthetic studies essentially produce “global”/regional averages of changes to species sizes through time, which have been cited to demonstrate long-term stability in resources, usually shellfish. Much like global climate trends, however, local trends may deviate significantly from broader patterns. In other words, local resources (e.g., oyster beds) may have been severely depleted, and yet the averaging of oyster sizes from dozens of sites at a continental (or regional) scale may still yield results indicating stability and sustainability, which may not reflect the experiences of individuals and communities within their particular environmental (and sociopolitical) locales within a region. Simply put, “global” averages cannot necessarily be used to understand local manifestations of broader institutional processes, and at the same time, local trends are the foundational units of global averages. Consequently, what remains to be brought into focus are the specific practices and institutions situated at smaller, human scales that drove the patterns documented across macroregional and macrotemporal scales. For example, recent synthetic studies may take a shellfish assemblage from one context (a pit feature) from one site and aggregate it with other oyster measurements from individual contexts from other sites to compile a single data point for the entire region that represents an aggregate of time (e.g., Late Archaic; Mississippian). These sites themselves—individual communities and the sustainable (or not) outcomes that they drove via particular institutional arrangements—are not the units of analysis.
Although large-scale, regional studies certainly represent a recent response and reaction to long archaeological histories of a preoccupation with individual sites, this study is a return to these smaller scales equipped with new conceptual frameworks, information, and methodological approaches that have been developed over the last few decades at these regional scales. Even though this study borrows from and builds on the conceptual and methodological approaches that characterize the recent regional work along the Atlantic and Pacific coasts described above, it follows closely the lead of work along the Northwest Coast that explores themes of long-term Indigenous resource management from integrated socioecological perspectives at the community and microregional scales (Cove Reference Cove1982; Grier Reference Grier2014; Groesbeck et al. Reference Groesbeck, Rowell, Lepofsky and Salomon2014; Lepofsky and Caldwell Reference Lepofsky and Caldwell2013; Lepofsky et al. Reference Lepofsky, Smith, Cardinal, Harper, Morris and Bouchard2015; Letham et al. Reference Letham, Martindale and Ames2020; Smith et al. Reference Smith, Lepofsky, Toniello, Holmes, Wilson, Neudorf and Roberts2019; Toniello et al. Reference Toniello, Lepofsky, Lertzman-Lepofsky, Salomon and Rowell2019; Trosper Reference Trosper2002). Such work has convincingly drawn on interdisciplinary frameworks to explore Indigenous resource management and has demonstrated the ability of the archaeological record to delineate these histories at institutional scales such as the village.
Ancestral Muskogean Guale of the Georgia Coast
The Guale people of the Georgia coast were a Muskogean-speaking group who lived along the estuaries and coastal zone for over 10,000 years and who encountered and interacted with some of the earliest Spanish colonizers. Although “Guale” does not exist as an ethnic or language identifier today, the descendants of the Guale communities recorded in ethnohistoric accounts are the modern citizens of the Muscogee (Creek) Nation and Seminole Tribes (Jones Reference Jones, Thomas, Jones, Durham and Larsen1978:186–187; Swanton Reference Swanton1922). Circa 5,000 years ago, with stabilizing sea levels and the formation of barrier salt marshes, Ancestral Muskogean communities began to live in the first sedentary villages of North America, referred to as “shell ring” villages (DePratter Reference DePratter1979; Sanger Reference Sanger2015; Thompson Reference Thompson2007; Thompson and Andrus Reference Thompson and Andrus2011; Thompson et al. Reference Thompson, Sanger, Smith, Garland, Howland, Andrus and Holland-Lulewicz2024; Turck and Thompson Reference Turck and Thompson2016). Over the next millennia, settlement along the coast waxed and waned with shifting social, political, ecological, and climatic dynamics. For this entire time, however, Guale peoples built their economies around estuarine resources, exploiting marine and estuarine animal species and wild plant resources (Thompson and Worth Reference Thompson and Worth2011) almost exclusively. Guale communities represent a case of so-called complex fisher-hunter-gatherers, who built and deployed complex social, political, and economic institutions and organization that—by circa AD 1000—began to be reflected in the large towns that started to grow and fill in coastal landscapes (Thompson and Holland-Lulewicz Reference Holland-Lulewicz2026).
One such town, and the focus of this research, is the Guale town at Middle Place (henceforth referred to as Guale MP) located on Ossabaw Island, a barrier island off the Georgia coast that eventually became one of the first reservations specifically established for the Muscogee (Creek) Nation in 1739 (Figure 1). Occupied between approximately AD 1100 and 1525, Guale MP is the largest Indigenous town yet to be recorded on the Georgia coast (Pearson Reference Pearson2001). At over 110 ha in area, Guale MP is comparable to the largest “Mississippian” towns known from the interior American Southeast and Midwest (Figures 2 and 3). The town itself is made up of at least four mounds (burial and, likely, platform), situated around a likely plaza space. To the west of this mound/plaza complex is a pond sitting just inland from the salt marsh. To the north, east, and south of the mound/plaza complex lies roughly 110 ha of households as evidenced from intact, low-relief, household shell middens typical of this time period that are visible on the surface (Crook Reference Crook1978; Garland et al. Reference Garland, Ritchison, Tucker and Thompson2023; Keene Reference Keene2004; Pearson Reference Pearson1977; Ritchison Reference Ritchison2025). The eastern half of the town alone contains over 200 of these undisturbed domestic middens. Between the pond and the mound/plaza complex, extensive sheet midden of shell runs the length of the pond’s eastern shore, suggesting the pond’s use by community members of Guale MP. The town sits directly adjacent to the expansive salt marsh, extending between the barrier islands and the coast of mainland Georgia. Salinity in and around Ossabaw is highly dominated by seasonal fluctuations in the Ogeechee River and its tributaries, with modern oyster beds experiencing a high mixing of salinity at the upper limits of optimality for oyster growth. Towns such as Guale MP harvested oysters intensively for centuries, to such an extent as to shape the ecological and geological processes in these environments (Holland-Lulewicz et al. Reference Holland-Lulewicz, Ritchison, Holland-Lulewicz, Howland, Thompson and Thompson2025). For example, around 30 million oysters were deposited across only a roughly 33 ha portion of Guale MP over its approximately 425-year history (Holland-Lulewicz et al. Reference Holland-Lulewicz, Ritchison, Holland-Lulewicz, Howland, Thompson and Thompson2025). Similar to other sites along the Georgia coast—based on faunal, botanical, and optimal foraging models—Guale MP was likely occupied by a relatively permanent year-round population (Jones Reference Jones, Thomas, Jones, Durham and Larsen1978; Keene Reference Keene2004; Thomas Reference Thomas2008a:722, Reference Thomas2008b:1112–1113).
Lidar map showing the location and features of the Guale Town at Middle Place (9CH158), Ossabaw Island, Georgia, with the whole site boundary (thick, solid line), previously surveyed area (dashed line), mounds, and domestic middens denoted (map by Isabelle Holland-Lulewicz). (Color online)
Figure 2 Long description
The map shows the location and features of Guale Town at Middle Place on Ossabaw Island, Georgia. The site extent is marked with a thick solid line. Mounds are indicated by red circles and domestic middens are shown as black dots. A pond is labeled near the center. The area depicted in another figure is outlined with a dashed line. The map includes a scale bar indicating 450 meters and a north arrow for orientation. An inset map in the top left corner provides a broader view of the island′s location. Elevation is represented in shades of gray, ranging from 9.74 meters to 1.46 meters. The map highlights the spatial distribution of archaeological features within the site boundary.
Lidar map showing the location and features of the Guale Town at Middle Place (9CH158), Ossabaw Island, Georgia, zoomed into the previously surveyed area with denoted sampled middens (map by Isabelle Holland-Lulewicz). (Color online)
Figure 3 Long description
The LiDAR map displays the Guale Town at Middle Place on Ossabaw Island, Georgia. It highlights various features within the Pearson Survey Boundary. Tested middens are marked with solid black circles, while untested middens are shown as open circles. Mounds are indicated by red circles. The map includes a pond labeled on the left side. A gradient scale from -1.46 meters to 9.74 meters is present, representing elevation changes. The map is oriented with a north arrow and includes a scale bar indicating 200 meters. The spatial distribution of middens and mounds is concentrated within the survey boundary, with clusters near the pond and scattered throughout the area.
This study explores the rich history and record of eastern oysters as an important resource for fisher-hunter-gatherer villages for millennia along the Georgia coast. Recent work has demonstrated that, at the regional scale, eastern oyster resource management likely mapped onto institutions situated at the scale of the individual island and that at this scale, oyster size through time increased. This suggests increases in ecosystem productivity either through human resource management practices or through environmental amelioration (Thompson et al. Reference Thompson, Rick, Garland, Thomas, Smith, Bergh and Sanger2020). Similarly, recent work on the regional-scale resilience, sustainability, and management of other fisheries on the Georgia coast has also posited the presence of some form of common-pool resource management, but it calls for more explicit investigations into how people achieved these states (Reitz Reference Reitz2014, Reference Reitz2021; Reitz et al. Reference Reitz, Colaninno, Quitmyer and Cannarozzi2022). Based on extant studies, Indigenous communities likely instituted strict control over harvesting, with sustainable rules and rights-use access at the island scale, which structured regional stability over the long term. However, the scale of these datasets limit interpretations of the actual mechanisms behind such patterns (Reitz Reference Reitz2014, Reference Reitz2021; Reitz et al. Reference Reitz, Colaninno, Quitmyer and Cannarozzi2022; Thompson et al. Reference Thompson, Rick, Garland, Thomas, Smith, Bergh and Sanger2020). Although it may be unsurprising that Indigenous communities were able to achieve regional-scale sustainable resource management strategies, given the balance between regional population densities and oyster resources, we do not know much about how these processes unfolded at particular places—or rather, how the practices potentially instituted through village-scale institutions drove regional-scale outcomes under local social, political, and economic contexts.
Thompson and colleagues (Reference Thompson, Rick, Garland, Thomas, Smith, Bergh and Sanger2020:6) recognize that the patterns evident at the regional level point to such local-level practices, but they suggest that precisely what kinds of adaptive controls were in place to promote sustainability remain difficult to discern from regional-level datasets. This is a critical point to consider. Although oyster reefs at the regional scale are abundant, individual villages would have been constrained by the local availabilities of these resources, putting more direct and acute pressure on local ecosystems compared to a long-term, regional perspective. In other words, although it may have been difficult to deplete the totality of oyster beds at the regional scale, it would have been easy to decimate individual oyster beds in the course of feeding the population of a town. For this reason, I use data collected from across a single Indigenous village of fisher-hunter-gatherers on the Georgia coast to explicitly investigate the outcomes of village-level resource management strategies and their potential contribution to regional sustainability. To this end, this study examines trends of eastern oyster use through time at a single town, Guale MP, as it emerged, grew, declined, and eventually was depopulated through an examination of 23 distinct domestic contexts from across an undisturbed portion of the site (Figure 3).
Materials and Methods
In the summer of 2023, 23 discrete, domestic midden deposits of the previously mapped 200 domestic middens (Pearson Reference Pearson2001) at Guale MP were sampled through small test excavations (Figure 3). These domestic middens are typical of Indigenous household refuse patterns across the coast. They are roughly circular, ranging in diameter between 5 and 15 m, with depths between 20 and 60 cm. They are composed almost exclusively of discarded shell, intermingled with other material signatures of household/domestic practice. Whether or not a single midden would have been associated with a single house remains unclear, but our sampling strategy was determined in such a way as to sample extensively across space and, ideally, to capture patterns related to distinct households across this 33 ha preserved area of the site. Each test excavation, placed in the thickest portion of the midden as identified through systematic probing, measured 50 cm2 × 50 cm2 and was dug in arbitrary 10 cm levels until sterile subsoil was reached. All midden matrix was screened through ⅛-inch screen, and all artifacts and shell were kept for later analyses.
Radiocarbon Dating
To anchor temporal trends in the oyster morphometrics used to evaluate human pressure and predation of local ecosystems (described below), I employ an AMS radiocarbon dating program reliant on 71 new AMS age determinations from across the sampled middens (Table 1). The sampling strategy prioritized materials such as hickory-nut fragments, given that they are a short-lived tissue. In cases where hickory nut was not present, I was constrained to selecting singular charred wood fragments. For each midden, when possible, samples were chosen from several sequential arbitrary stratigraphic levels to facilitate basic chronological modeling to refine temporal placement for each midden without having to rely solely on ceramic data (Table 1; Supplementary Text 1). All middens lacked any discernable natural stratigraphic separation, characterized by a homogenous 30–50 cm layer of shell midden atop the natural sterile sand subsoil (see Holland-Lulewicz et al. [Reference Holland-Lulewicz, Ritchison, Holland-Lulewicz, Howland, Thompson and Thompson2025:Figure 5F] for a representative stratigraphic profile of such middens). All dates were run at the Pennsylvania State University AMS Radiocarbon facility and were then calibrated using IntCal20 (Reimer et al. Reference Reimer, Austin, Bard, Bayliss, Blackwell, Ramsey and Butzin2020) and modeled in OxCal v.4.4 (Bronk Ramsey Reference Bronk Ramsey2009). Because sample sizes for each midden were generally low (between roughly one and five dates each), a combination of formal Bayesian chronological modeling, contextual information (e.g., thickness and depth), and ceramic data were used to produce generalized temporal estimations. Model descriptions and code for each midden are included in Supplementary Text 1, and the results are presented in Table 2.
AMS Radiocarbon Dates from the Guale Town at Middle Place (9CH158), Ossabaw Island, Georgia (USA).
Table 1 Long description
The table lists AMS radiocarbon results from multiple midden deposits, organized by midden number, excavation level and stratum, and assigned occupation period (A through D). For each sample ID and material type (mostly unidentified nut charcoal and wood charcoal), it reports radiocarbon age in years before present with an uncertainty, estimated carbon yield, and isotopic values (fraction modern and delta 14C) with uncertainties. Period A (Pre-Village) samples cluster around roughly 1435 to 1550 years before present. Period B (Village Growth) is generally younger, commonly about 675 to 1140 years before present, but includes older results around 1135 to 1140. Period C (Peak Village) is mostly about 550 to 855 years before present, with notable older wood-charcoal outliers near 1210 and around 1555 to 1560. Period D (Village Decline) is largely the youngest, often about 330 to 635 years before present, but includes an older nut-charcoal result near 1515 and a wood-charcoal result near 1185. Across periods, fraction modern tends to be lower and delta 14C more negative for older ages, consistent with expected radiocarbon behavior. Because materials include wood charcoal, some older-than-expected ages may reflect inbuilt age or redeposited charcoal rather than the exact timing of deposition.
Summary Table of Bayesian Models of AMS Radiocarbon Dates from the Guale Town at Middle Place (9CH158), Ossabaw Island, Georgia (USA).
Table 2 Long description
The table summarizes Bayesian modeled calendar-age boundaries for 24 shell middens, listing number of radiocarbon dates, modeled start and end ranges at 68 percent probability, modal start and end years, and estimated spans and ranges. Early activity is limited to Period A (Pre-Village), with midden 359 starting about AD 455 to 565 and ending about AD 550 to 635, and midden 347 starting about AD 550 to 640 and ending about AD 600 to 800. Most middens fall in Period B (Village Growth), generally beginning around AD 950 to 1220 and often ending between about AD 1220 and 1575; several have longer modeled spans, including midden 272 ending about AD 1275 to 1550 and midden 273 ending about AD 1285 to 1500. Period C (Peak Village) concentrates around late 1200s through 1400s, with many estimated ranges centered near AD 1290 to 1390; one entry, midden 311, has a very late modeled end extending far beyond the others, indicating high uncertainty or an open-ended boundary. Period D (Population Decline) entries cluster from the mid 1400s into the 1600s, with middens 19 and 20 showing starts around AD 1440 to 1630 and ends around AD 1500 to 1765. The number of dates per midden is usually two to four, and modeled spans vary widely from near zero up to several centuries, so comparisons should emphasize the probability ranges rather than single modal years.
Alongside the formally modeled ranges (where it was possible to do so) are a series of more general estimates for age ranges. Figure 4 includes a kernel density estimation model to evaluate the villages’ demographic trends alongside the generalized ordering of household middens. The exact sequential ordering of each of the 23 middens as presented in Figure 4 is an estimation. Though the exact ordering may vary slightly—and of course many of these are likely to be contemporary—generalized temporal associations in grouped blocks of time are more confidently determined than the orderings of individual middens. The four time periods across which middens were categorized do capture the general ordering of groups of middens through time, especially for evaluating the demographic growth, peak, and decline of the village for this research. The time blocks were generated independent of established ceramic chronologies, but they generally coincide well with these chronologies (Parbus et al. Reference Parbus, Thompson, Garland and Tucker2023; Ritchison Reference Ritchison2018), contributing to our confidence in our chosen blocks of time within which to group individual middens. Indeed, the grouping of middens into time periods is meant to account for contemporaneity and general uncertainty inherent in all absolute age estimations. In other words, these groupings account for a lack of even higher resolution that would allow us to confidently order each individual midden, one by one, through time with complete certainty.
Kernel density estimate (KDE) plot of radiocarbon data from Guale Town at Middle Place, Ossabaw Island, Georgia, USA. Bars represent modeled estimations for each shell midden, which are placed into time periods A (AD 550–650, pre-village contexts), B (AD 1100–1300, initial settlement and early growth), C (AD 1300–1400, maximal village extent), and D (AD 1400–1525, town depopulation). (Color online)
Figure 4 Long description
The image A showing a density plot titled Eastern Oyster Valve Height (LVH). The horizontal axis label is Height. The vertical axis label is Density. The horizontal axis shows tick labels 0, 50, 100, 150, 200, 250. The vertical axis shows tick labels 0.00, 0.01, 0.02, 0.03, 0.04. Multiple overlapping density curves are shown, distinguished by different fills and outlines. The curves have their highest density near the lower end of the Height scale and taper toward higher Height values, with long right tails extending toward 250. A legend with multiple entries is present. An inset drawing of an oyster valve is shown with the label LVH. The image B showing a density plot titled Eastern Oyster Valve Shape Ratio (LVH:LVL). The horizontal axis label is Ratio. The vertical axis label is Density. The horizontal axis shows tick labels 0.0, 0.5, 1.0, 1.5, 2.0. The vertical axis shows tick labels 0.0, 0.5, 1.0, 1.5, 2.0. Multiple overlapping density curves are shown, distinguished by different fills and outlines. The curves have their highest density near the lower end of the Ratio scale and taper toward higher Ratio values, with right tails extending toward 2.0. A legend with multiple entries is present. An inset drawing of an oyster valve is shown with the label LVH:LVL. The image C showing a distribution plot grouped by Period for Height. The horizontal axis label is Period. The vertical axis label is Height. The horizontal axis shows four category labels A, B, C, D. Each category contains a vertical violin shape with many horizontal jittered line marks and a central vertical line. The vertical axis shows tick labels 0, 50, 100, 150, 200, 250. The distributions in all four periods extend from near 0 up to values above 200, with most marks concentrated below 100 and fewer marks at higher values. The image D showing a distribution plot grouped by Period for Ratio. The horizontal axis label is Period. The vertical axis label is Ratio. The horizontal axis shows four category labels A, B, C, D. Each category contains a vertical violin shape with many horizontal jittered line marks and a central vertical line. The vertical axis shows tick labels 0.0, 0.5, 1.0, 1.5, 2.0. The distributions in all four periods extend from near 0.0 up to values near 2.0, with most marks concentrated below 1.0 and fewer marks at higher values. The top row and bottom row pair the same variables: Height is shown as a density plot in image A and as grouped distributions by Period in image C and Ratio is shown as a density plot in image B and as grouped distributions by Period in image D.
Oyster Morphometrics
Oyster paleobiology studies use size as a useful proxy to evaluate the health of oyster reefs as related to both human predation and environmental pressures (Campbell Reference Campbell2008; Claassen Reference Claassen1998; Rick et al. Reference Rick, Reeder-Myers, Hofman, Breitburg, Lockwood, Henkes and Kellogg2016). Oyster size is driven by numerous factors such as age and environmental variability, so healthier reefs and climatic stability generally produce larger oysters (Bartol et al. Reference Bartol, Mann and Luckenbach1999; Lulewicz et al. Reference Lulewicz, Thompson, Cramb and Tucker2017). Overall shape of oysters is driven by a complex mixing of factors, but in general, intertidal oysters (those variably inundated) tend to be rounder, whereas subtidal oysters (those permanently submerged) tend to be longer and skinnier (Bartol et al. Reference Bartol, Mann and Luckenbach1999; Hajovsky et al. Reference Hajovsky, Pollack and Anderson2021; Mizuta and Wikfors Reference Mizuta and Wikfors2019; Toro and Newkirk Reference Toro and Newkirk1991). To assess oyster morphology through time, a series of measurements was collected on a total of 18,317 left oyster valves (see Figure 5 for measurements collected). Measurements for oyster height (LVH; usually the largest dimension from the hinge to the growing edge) and length (LVL; usually the smaller dimension perpendicular to height) were recorded in millimeters following standard methods outlined in Lulewicz and colleagues (Lulewicz et al. Reference Lulewicz, Thompson, Cramb and Tucker2017). Oyster paleobiology studies commonly use LVH measurements, but here, LVL measurements are also collected to generate ratio values used to evaluate overall oyster shape (LVH:LVL), which is also sensitive to environmental and human predation pressures (Bartol et al. Reference Bartol, Mann and Luckenbach1999; Claassen Reference Claassen1998; Hajovsky et al. Reference Hajovsky, Pollack and Anderson2021; Kent Reference Kent1988; Mizuta and Wikfors Reference Mizuta and Wikfors2019; Toro and Newkirk Reference Toro and Newkirk1991). All oysters <35 mm usually represent juveniles/spat and were excluded from this dataset following standard modern and archaeological methods outlined in Rick and colleagues (Reference Rick, Reeder-Myers, Hofman, Breitburg, Lockwood, Henkes and Kellogg2016). Although this does increase overall averages in oyster size, this study focuses on those individuals targeted for consumption. Statistical analyses of measurements were conducted in R (R Core Team 2025). Given that the data are not normally distributed or homoscedastic, LVH and ratio shape values through time were compared through a nonparametric Kruskal-Wallis test, and a post-hoc pairwise Mann-Whitney U test was used to identify which time periods were distinguishable in regard to both LVH and ratio shape. Differences were considered statistically significant when p was ≤0.05.
Comparison by time period of eastern oyster (Crassostrea virginica) valve height (LVH in mm; left) and valve shape ratio (LVH:LVL; right) data from the Guale Town at Middle Place (9CH158), Ossabaw Island, Georgia, with depiction of oyster measurements collected (adapted from Lulewicz et al. Reference Lulewicz, Thompson, Cramb and Tucker2017). Bottom box and whisker plots with histograms represent the following: the tall line represents the range of data; the horizontal central line represents the median of the data; the boxes represent the data between the first and third quartiles; and the single points represent data point outliers. (Color online)
Figure 5 Long description
The image contains four graphs comparing eastern oyster valve height (LVH in millimeters) and valve shape ratio (LVH:LVL) across time periods. The periods visible in the graphs are labeled Period A, Period B, Period C and Period D. Each period is visually distinguished by a separate overlapping density curve or distribution shape. Graph 1 is titled Eastern Oyster Valve Height (LVH). The x-axis is labeled height and the y-axis is labeled density. It displays overlapping density curves for each of the four periods. The distributions are right-skewed, with peaks occurring at lower height values and long tails extending toward higher values. An inset diagram of an oyster valve is present in the upper right area of the graph, with measurement lines indicating LVH and LVL. A legend identifies the four periods. Graph 2 is titled Eastern Oyster Valve Shape Ratio (LVH:LVL). The x-axis is labeled ratio and the y-axis is labeled density. It also displays overlapping density curves for the four periods. The distributions are similarly right-skewed, with peaks near lower ratio values. The same inset oyster diagram appears in the upper right area with a measurement indicator circled. A legend identifies the four periods. Graph 3 presents violin and half-violin distributions with embedded box plots for valve height (LVH in millimeters). The x-axis is labeled Period with categories Period A, Period B, Period C and Period D. The y-axis is labeled LVH. Each period shows a half-violin shape paired with a box plot. The central horizontal line within each box represents the median. The box spans the first to third quartiles. The tall vertical line represents the full range of data. Individual points outside the range represent outliers. Outliers are visible across all periods and are most numerous in Period B and Period C. Median LVH values appear highest in Period A and Period B relative to Period C and Period D, with visible downward shift in the median line across later periods. Graph 4 presents violin and half-violin distributions with embedded box plots for valve shape ratio (LVH:LVL). The x-axis is labeled Period with categories Period A, Period B, Period C and Period D. The y-axis is labeled Ratio. The structure mirrors Graph 3, with half-violin shapes, embedded box plots, range lines and individual outlier points. Outliers are present across all periods. The median ratio values show variation across periods, with spread appearing wider in earlier periods. Taken together, the top two graphs show the overall distributional shape of each measure by period, while the bottom two graphs summarize the same data using distribution shapes combined with median, quartile, range and outlier information for direct period-to-period comparison.
Results
Radiocarbon Dating
From the 23 domestic midden mounds sampled for this study, 71 AMS radiocarbon dates place the town’s most substantive occupation between approximately AD 1100 and 1525, with the average use life of each midden at roughly 30–40 years; two middens dated to a time prior to the establishment of the large village (Figure 4; Tables 1 and 2). To facilitate comparison and evaluations of trends in oyster management, domestic middens were separated into four time periods: Period A, approximately AD 550–650, representing the pre-town, earlier occupation of the site; Period B, approximately AD 1100–1300, representing the initial town formation; Period C, approximately AD 1300–1400, representing the demographic boom and peak of the town; and Period D, approximately AD 1400–1525, representing the last phase of the town’s occupation (Figure 4; Table 2). Spatially, results indicate that middens associated with Period B—the initial emergence of the large town (likely also associated with the initial construction of the mounds; Moore Reference Moore1897)—are constrained to a small, central core. Expansion is clearly documented during Period C, where middens populate the entirety of the known spatial extent of the site. Period D middens are also distributed widely.
Oyster Morphometrics
Of the 18,317 left oyster valves measured across all 23 middens, 9,783 left valves were whole enough to assess trends in overall oyster size (LVH) through time, and 8,995 possessed both measurable dimensions to assess trends in oyster shape through time (LVH and LVL; see Figure 5; Supplementary Tables 1 and 2). This analysis yielded a wide range of oyster heights, from 30 to 195 mm. A nonparametric Kruskal-Wallis test followed by a post-hoc pairwise Mann-Whitney U test demonstrated statistically significant differences in oyster height between all time periods except for between Period B (mean 62 mm, SD 20.5, n = 3,485) and Period C (mean 66 mm, SD 20.4, n = 2,880; X 2 = 38.468, p < 0.01). Period D has the smallest oysters (mean 61 mm, SD 19.1, n = 3,178; Supplementary Table 3). The same tests for oyster shape over time showed statistically significant differences in oyster shape between all time periods (X 2 = 133.9, p < 0.01; Supplementary Table 4). Period A stands out as the longest (mean 70 mm, SD 25.08, n = 244) and skinniest oysters of all time periods, although this the period has the smallest sample size but also the largest standard deviations for both length and shape (Supplementary Table 1). In general, though, over the 425 years of village occupation, oyster sizes decreased among the household deposits and were the shortest and roundest toward the end of the village occupation (Figure 5).
Although the results presented here demonstrate statistically significant differences in size and shape of oysters throughout the life-span of the village, with the largest oysters harvested during the peak of village occupation, it remains important to contextualize these results within the human experience. The purposeful location of the town along the marsh’s edge likely facilitated a certain ease of movement throughout the estuary, providing access to a wide range of oyster reefs. This ease of access, via water, to a wide range of resources likely made the implementation of sustainable strategies, for example, easier. For instance, the differences in the mean oyster height between Periods B and D only differ by 1 mm, which represents a scale not likely experienced or noticed by people. It is unlikely that these very minor changes in size represent substantive changes in the experiences of the community or ecological regime; instead, they demonstrate sustainable oyster resource management enabled or facilitated by the institutions associated with the emergence of the town, despite both climatic and demographic challenges. There is certainly no evidence for the collapse of these shellfisheries. Movement and exploitation throughout the estuary, the ease of which was facilitated by canoe, holds antiquity and is evidenced in the range of habitats from which resources were acquired millennia before—in the Late Archaic and through the contact period, with paired political centers (towns) on islands and the mainland (Andrus and Thompson Reference Andrus, Fred and Victor2012; Jones Reference Jones, Thomas, Jones, Durham and Larsen1978:194–196, 200).
Discussion
Local Institutions, Guale Villages, and Sustainable Oyster Exploitation
The first important takeaway is that long-term, regional-scale sustainability does not mean no anthropogenic impacts occurred; when impacts were felt at local scales (impacts that may be invisible at the regional scale or explained away as noise/variation), community- or household-scale institutions could adjust to ameliorate those impacts and continue to drive long-term sustainable use. In any case, these results demonstrate long-term sustainable exploitation of oyster resources for around 425 years at a single, large village, with no indication of local collapse or depletion of the core oyster fishery. This suggests the presence of intentional forms of shellfish management at the scale of the village that were potentially complementary to or embedded within—namely, the kinds of regional-scale institutions suggested in previous studies (Reitz Reference Reitz2014, Reference Reitz2021; Reitz et al. Reference Reitz, Colaninno, Quitmyer and Cannarozzi2022; Thompson et al. Reference Thompson, Rick, Garland, Thomas, Smith, Bergh and Sanger2020). In other words, oyster-size profiles and patterns as manifested at regional scales are likely the products of intentional decisions at local levels, made within villages, and in the context of specific ecological conditions meant to mediate the impacts of population pressure and to protect against overharvesting.
Although this dataset suggests that the impacts of humans on oyster resources were not drastic, it nevertheless does indicate that impacts to oyster fisheries were nonzero. Impacts to oyster sizes and shapes into Period D do correlate with the demographic history of the town, with smaller oyster sizes trailing peak population. Interestingly, Period C, associated with the demographic peak of the town, yields the largest and skinniest oysters harvested during the town’s occupation. Considering general trends in oyster shape with regard to the locations of oyster beds in relation to high water lines, these results could point to a couple of different driving mechanisms. These results could represent a diversion of harvesting practices into oyster beds more frequently inundated or permanently submerged, which would require increased efforts to access. For example, households may have had to begin exploiting difficult-to-access beds as the most convenient beds—those above water, yielding rounder oysters—experienced more pressure with population growth. Another complementary explanation, if only size is considered (as all extant studies have done), is that increased demographic pressure could have sparked heightened institutional management and oversight of resource use, creating healthier reefs and therefore larger oysters; however, we cannot currently point to the specific kinds of rules or strategies that may have been implemented without other kinds of data (e.g., isotopes to indicate specific habitats exploited or seasons of collection).
The exact kinds of social and political institutions or organizations that facilitated decision-making within the village of Guale MP remain difficult to explore, though the ethnohistoric record may be informative in producing some hypotheses for future investigation. Current evidence suggests that Ossabaw was depopulated prior to European colonization, by circa AD 1525 (Parbus Reference Parbus2026), so understanding exactly how Guale MP might have fit into the ethnohistorically documented political system is unclear, but there likely existed similar institutions as those witnessed by European colonizers across slightly earlier Guale communities (Jones Reference Jones, Thomas, Jones, Durham and Larsen1978:242). Accounts of political organization at the time of colonization support the idea of polities defined by two capitals—one located on the mainland and one located on a barrier island across the marsh—operating as a twin political unit, potentially unified under larger, regional administrative control (for discussion, see Jones Reference Jones, Thomas, Jones, Durham and Larsen1978:194–196; Worth Reference Worth2004). Governance of resource exploitation may have been situated at the scale of this political unit or within the individual villages themselves. Further ethnohistoric references to ownership might be informative. For instance, at the time of European colonization, Guale communities participated in agricultural production, and productive lands and resources were owned by chiefly matrilineages (Worth Reference Worth2004). However, the timing of maize agriculture is likely quite late to the coast, and no evidence of such agricultural practices or maize cultivation is currently known for Ossabaw Island. Whether the role of matrilineal or kin-based control over key resources extended into earlier periods, and to other kinds of key resources other than land, remains unclear. Even so, and despite the potential presence of chiefly matrilines later on, overall systematic organization was likely more corporate and collective (Anderson Reference Anderson1994; Ritchison Reference Ritchison2018; Saunders Reference Saunders2000; Thomas Reference Thomas2008c:23; Thompson Reference Thompson2009). In any case, the results presented here suggest that whatever the articulation was between corporate (chiefly matrilineages) and collective (household agreement, buy-in, and flexibility), institutions within the Guale MP village seemed to have successfully promoted sustainability.
These results track well with other recent work that has focused on other kinds of fishes exploited along the coast by the Guale and that have argued it is unlikely people instituted no control over access or efforts to protect shellfish resources (Reitz et al. Reference Reitz, Colaninno, Quitmyer and Cannarozzi2022). In other words, a completely “open access” (sensu Moritz et al. Reference Moritz, Behnke, Beitl, Bird, Chiaravalloti, Clark and Crabtree2018) regime was likely not the case. For instance, top-down rules may have potentially included restrictions on habitats, locations, or seasons of collection—possibly instituted by matrilines—or in the case of other Mississippian villages across the southeast and along the coast, village councils. Whatever the case, cooperative institutions related to resource use and access likely were those that held the most flexibility in harvesting practices to deal with increasing demographic pressures and any kinds of shifting environmental conditions. Although speculative, it is intriguing to suggest that the Guale political form, anchored by twin towns situated across diverse components of the estuaries, may have been an explicit adaptive strategy of governance that was built within the context of the estuarine system of the Georgia coast. This large geographic extent translates to a large expanse of access to fisheries that would have fostered flexibility within the larger system. In this way, the adoption of the village as a key institution in fact facilitated sustainable fisher-hunter-gatherer economies, similar to the way hypermobility serves in nonsedentary hunter-gatherer societies.
Even with the statistically significant diversions of oyster sizes and shapes over time at Guale MP, the overall ability of these villages to contribute to large-scale sustainability efforts remained, suggesting that the village represented a key institution in the management of fisheries (Reitz Reference Reitz2014). It is here that village-scale governance served as a mechanism for regional sustainability that transcended the scale of the individual village itself.
Environmental Drivers and Socioecological Dynamics
Long-term, large-scale perspectives have likely masked variability, ups and downs, good times and bad, defying a complete understanding of the effectiveness and outcomes of particular resource management strategies. Resolving time and space at higher resolutions is critical when dealing with resources highly sensitive to environmental and climatic changes and exploitation pressures, especially shellfisheries. For resources such as oysters, which respond rapidly to even minor environmental changes such as sea levels and rainfall inputs, understanding the flexibility of resource management strategies associated with this resource to adapt to shifting targets and baselines of sustainability is critical. Sustainable harvesting practices require intimate ecological knowledge of not only the local landscape but also the resource itself.
For oysters, modern studies demonstrate that size and shape are a result of a complex mixing of variables and are undoubtedly linked in part to environmental conditions such as salinity variations related to sea-level and rainfall patterns, especially due to the sessile nature of the organism (Hajovsky et al. Reference Hajovsky, Pollack and Anderson2021; Kent Reference Kent1988; Lenihan Reference Lenihan1999; Mizuta and Wikfors Reference Mizuta and Wikfors2019). Geohistorical data suggest similar relationships in the past, prior to anthropogenic engagement. For instance, oyster life-span decreases with warmer temperature, oyster colonization within estuaries tracks with past warm–cool cycles, and oysters in the Pleistocene lived much longer (Carbotte et al. Reference Carbotte, Bell, Ryan, McHugh, Slagle, Nitsche and Rubenstone2004; Durham Reference Durham2017; Kirby Reference Kirby2001; Lockwood and Mann Reference Lockwood and Mann2019). During the earlier occupation of the village, sometime in the thirteenth century (Period B, initial occupation), tree-ring data from the region suggests significant variance in spring rainfall values, potentially indicating a decreased predictability and stability in how people experienced and navigated environments (Stahle and Cleaveland Reference Stahle and Cleaveland1992). When combined with the Palmer Hydrological Drought Index (PHDI), there seems to be clear periodicity with which the Georgia coast experienced periods of droughts, with a pronounced drought and relatively cooler temperatures circa AD 1176–1220, coinciding with the initial formation of Guale MP in Period B (Blanton and Thomas Reference Blanton and Thomas2008:800) and potentially hinting at the adaptive advantages of coalescing into large social institutions (e.g., villages). However, these large temporal- and spatial-scale climatic datasets represent environmental conditions at expansive spatial and temporal scales and not at the scales comparative to this archaeological dataset. Moreover, generally speaking, archaeologists primarily theorize about the impacts of such climatic events only as they relate to the sustainability of agricultural production and food supplies (e.g., Blanton and Thomas Reference Blanton and Thomas2008:800), not how these fluctuations would have affected sustainable practices for peoples who structured their societies around fishing, hunting, and gathering. The further development of appropriately scaled climate datasets is key to further unraveling the complexity of institutions of sustainable resource management at the scale of the human experience.
Extant regional, long-term studies have suggested that the spatial patterning of impacts to oyster sizes up and down the Georgia coast is related to local environmental variables. For instance, differences exist between deltaic areas, where rivers empty into the estuaries, and nondeltaic areas, which lack riverine input (Turck and Thompson Reference Turck and Thompson2016). Deltaic areas and river discharge carry higher sediment and nutrient loads, and they experience greater hydrological dynamics, forcing salinity variability throughout the estuary and supporting marsh progradation during times of sea-level regression (Di Iorio and Castelao Reference Di Iorio and Castelao2013; Reed Reference Reed2002). Because nondeltaic areas lack the necessary sediment influx to combat long-shore drift and other processes shaping the coastlines, these areas lack a resiliency required for stable populations of the resources on which humans rely (such as oysters; Turck Reference Turck2011). Ossabaw Island and Guale MP are located at the southern end of the Ogeechee River deltaic region and the northern end of the nondeltaic area around St. Catherines Island directly to the south (Turck and Thompson Reference Turck, Thompson, Reeder-Myers, Turck and Rick2019). The back barrier salt marshes between Ossabaw Island and the mainland are designated as nondeltaic environments—meaning that these environments are vulnerable and that the marshes adjacent to Guale MP would have had a hard time maintaining themselves through changes to sea-level rise and fall (Turck Reference Turck2011). Overall, these nondeltaic environments, such as those available to the Guale MP populace, would have been more sensitive to climatic fluctuations, variation, and unpredictability. It is therefore more likely that the long-term sustainability of the local oyster beds around Guale MP were not the result of naturally favorable conditions and abundance but of resource management practices of Guale communities, who adjusted their use and conservation strategies to both anthropogenic and environmental pressures.
Conclusions: Archaeology and Interdisciplinary Sustainability Science
Modern work to evaluate the success of sustainable practices often prioritizes understanding the scale at which communities feel the outcome of sustainable resource management practices. Today, sustainability scientists easily recognize the challenges of scale, especially in geographies that cross-cut national, state, and local governmental entities (Barkin and Moseley Reference Barkin, Moseley, Bressers and Rosenbaum2003; Bressers and Rosenbaum Reference Bressers, Rosenbaum, Bressers and Rosenbaum2003). If archaeology and investigations into the historical trajectories of human-environment dynamics are to contribute to sustainability science, we must work to not only identify sustainability in the past at large scales but also investigate these principles at multiple temporal and spatial scales so that we might gain a greater understanding of the actions and consequences that are enacted by and felt by communities (Thakar Reference Thakar, Wolverton and Lyman2012; Wu Reference Wu2021). This becomes even more pressing in the context of common pool resources, such as fisheries, which are expected to inevitably experience the “tragedy of the commons” (Hardin Reference Hardin1968). However, the scale at which these processes unfold matters and may or may not translate spatially and/or temporally. For example, although fisheries may be sustainable—or not—at the large regional scale over the long term, this does not necessarily translate to the experience of individual places or communities at a certain time. This work requires data appropriately scaled and centered on the foundational structures that regulate resource use. The institutions created by sedentary fisher-hunter-gatherers, who relied primarily on fisheries for millennia to support large populations, serve as critical contexts in which to investigate the variability and flexibility of management strategies within the venues and scales that matter.
The emergence of large villages, such as the one on Ossabaw Island, can be understood to potentially represent the emergence of a key institution—the village itself—as an important human organization critical for the sustainability of livelihoods in the face of potential climatic impacts on estuarine resources. The (re)emergence of this new institutional form (large aggregate villages), built on millennia’s worth of experience and knowledge, likely bolstered successful and sustainable outcomes as the village grew. This pattern, now demonstrated at a single community, supports ideas put forth by both Reitz (Reference Reitz2021) and Thompson et alia (Reference Thompson, Rick, Garland, Thomas, Smith, Bergh and Sanger2020) that suggest that the emergence of adaptive controls such as common-pool management and resource proprietorship coincided with the development of villages (Thompson and Birch Reference Thompson, Birch, Birch and Thompson2018). It also contributes to and supports investigations surrounding questions related to common-pool resources and the institutions that are built around—or that facilitate—cooperation as people settle down and population density increases (Feinman and Neitzel Reference Feinman and Neitzel2023; Feinman and Thompson Reference Feinman and Thompson2025; Thompson Reference Thompson2023:544).
The data presented from Guale MP also provides fodder for investigating the linkages between long-term trends in mobility, scales and degrees of movement across the landscape, and the ways people solve collective action problems in the face of what Thompson (Reference Thompson2023:545) also refers to as large-scale external systemic changes, such as population increase and climate change. Finally, the case study presented here for the Guale MP village serves as a way to track these strategies through the foundation, growth, maintenance, and depopulation of a village that implemented overall successful strategies of resource management despite the pressures of these various social and (potentially) environmental forces. In fact, this research continues to support extant studies demonstrating that these fisheries only collapsed once Euro-American peoples began to commercialize and industrialize the harvesting of such resources (e.g., Kirby Reference Kirby2004; Reeder-Myers et al. Reference Reeder-Myers, Braje, Hofman, Smith, Garland, Grone and Hadden2022).
Analyses of historic oyster fisheries (associated with Euro-American industrialization) of the eastern North American coast demonstrate unsustainable use and rapid depletion, especially in relation to the development and growth of urban centers (Kirby Reference Kirby2004). The impacts of unsustainable harvesting of oyster resources include creating cascading trophic-level disequilibrium, such as phytoplankton blooms, which can lead to eutrophication in an ecosystem (Jackson et al. Reference Jackson, Kirby, Berger, Bjorndal, Botsford, Bourque and Bradbury2001). Understanding long-term historical impacts on oyster populations does provide critical context for understanding current ecosystem function, but the specific contexts for these impacts cannot be overlooked in the creation of conservation strategies. Both historical (post-European colonization) and Indigenous patterns of oyster fisheries, at large and small temporal and spatial scales, offer insight into, for example, how demographic pressures and varying economic strategies influence the sustainability of a resource system. For Guale MP, institutional management of oyster-bed resources facilitated endurance in oyster resources that likely did experience pressures, both environmental and demographic. On the one hand, the subtle (yet statistically significant) differences (ca. 1–3 mm) in oyster sizes were likely unnoticeable to people, and this likely demonstrates the successful outcomes of purposeful and directed resource management. Indeed, these differences suggest both a significant pressure placed on resources and the concomitant practices implemented to ameliorate these pressures. Zooarchaeological research, such as presented here, provides modern sustainability scientists the ability to identify and anticipate both estuarine organism vulnerability to anthropogenic pressures at various scales and the broader evolutionary trajectory of marine ecosystems (Lyman Reference Lyman1996; Thakar Reference Thakar, Wolverton and Lyman2012; Wolverton and Lyman Reference Wolverton, Lyman, Wolverton and Lyman2012). More specifically, such work has the potential to provide a critical multiscalar perspective on how Indigenous management decisions regarding the methods of harvesting and their intensity may have affected populations of critical resources, and such knowledge and practice may be implemented to affect future outcomes of policy aimed to facilitate sustainable marine exploitation.
Acknowledgments
I extend my sincerest thanks to the numerous students who assisted in collecting oyster measurements, including Matthew Picarelli-Kombert, Andrew Yuscinsky, Lakelyn Smith, and the numerous field research assistants from the 2023 and 2024 field seasons. I also wish to thank Jake Holland-Lulewicz and Victor Thompson for their guidance and for their comments on a draft of this manuscript. Finally, I wish to thank the Pennsylvania State University College of Liberal Arts and Institute for Energy and the Environment (and the Radiocarbon Lab) for providing the funding for this research, along with the Department of Anthropology for unwavering institutional support. Thanks also to Rachel Black and the Georgia Department of Natural Resources, as well as the Ossabaw Island Foundation for their support in conducting the fieldwork necessary to undertake this research. Finally, thanks to the Muscogee (Creek) Nation Historical and Cultural Preservation Department for their continued support of and involvement in our research on Ossabaw Island and elsewhere across their ancestral homelands. I did not use any AI tools to generate text or images or to analyze or extract insights from data or other materials. I also wish to thank the three anonymous reviewers for their insightful and helpful comments in improving this manuscript.
Funding Statement
Funding for this research was provided by the College of Liberal Arts and the Institute for Energy and the Environment at Pennsylvania State University.
Data Availability Statement
Shell measurement data and R code used to produced the data, figures, and analyses is archived at Zenodo (10.5281/zenodo.21728385).
Competing Interests
The author has no competing interests to disclose.
Supplementary Material
The supplementary material for this article can be found at https://doi.org/10.1017/aaq.2026.10222.
Supplementary Table 1. Summary table of eastern oyster (Crassostrea virginica) height (LVH) measurements (mm) for overall size evaluation for the Guale Town at Middle Place, Ossabaw Island, Georgia, USA.
Supplementary Table 2. Summary table of eastern oyster (Crassostrea virginica) height:length ratio (LVH:LVL) measurements (mm) for overall shape evaluation for the Guale Town at Middle Place, Ossabaw Island, Georgia, USA.
Supplementary Table 3. Results of Kruskal-Wallis test and pairwise Mann-Whitney U test comparing time periods for eastern oyster (Crassostrea virginica) length comparisons (LVH) for the Guale Town at Middle Place, Ossabaw Island, Georgia, USA.
Supplementary Table 4. Results of Kruskal-Wallis test and pairwise Mann-Whitney U test comparing time periods for eastern oyster (Crassostrea virginica) shape comparisons (LVH:LVL) for the Guale Town at Middle Place, Ossabaw Island, Georgia, USA.
Supplementary Text 1. Bayesian Radiocarbon Modeling Text and Code.
Open access
