This paper is part of the Proceedings of the Maritime Cultural Landscape Symposium, October 14-15, 2015, University of Wisconsin-Madison

Margo Schwadron
Southeast Archeological Center, National Park Service

This case study details a new, important example of prehistoric hunter-fisher-gatherers from the Ten Thousand Islands region of the Everglades, Florida. As the largest subtropical wilderness in the US, the Everglades are an unparalleled landscape which provides important habitats for numerous rare and endangered species. The Everglades are an international treasure recognized as a World Heritage Site environmentally, an International Biosphere Reserve, and a Wetland of International Importance. While the natural and environmental significance of the Everglades have long been recognized, the human history of the Everglades is much less understood. This study fills an important gap in understanding the role of humans within this rich ecosystem and stands as an excellent example of a prehistoric maritime cultural landscape.

Studies on midden sites typically focus on diet, subsistence and paleo-environmental studies with a normative and long standing view of shell middens as domestic refuse, simply the remains of daily meals discarded in garbage piles. However, recent work by some researchers has challenged this idea. At the Southeast Archeological Center (SEAC), we have expanded our interpretations to beyond these strictly garbage pile contexts. This case study examines a little known but significant type of shell midden site called “shell works”, which are among the world’s largest, most-complex, prehistoric shell built landscapes ever known. They deserve far more consideration than the simple garbage pile perspective.

Shell works are complex sites that were socially constructed landscapes that reflect a unique maritime-hunter-gatherer adaptation and tradition of shell construction. These shell works sites represent some of the world’s best examples of prehistoric maritime cultural landscapes, as their preservation is unparalleled. Preserved in almost their entirety, the Ten Thousand Islands region is a vast prehistoric domain of waterways, islands, seascapes and shellscapes that stretch for some hundred miles along the southwest Florida coast.

South Florida contains an immense wetland of marshes, swamps, rivers and estuaries dominated by the Everglades, the largest sub-tropical wetland in North America. The lower southwest coast contains the Ten Thousand Islands, a vast maze of lagoons, mangrove swamps and marine meadows comprising one of the most productive sub-tropical estuaries in North America. The region contains over 400 recorded shell middens sites. Shell middens take many forms, including small heaps, linear or mounded accumulations, and are traditionally viewed as either primary or secondary refuse, the results of daily refuse from domestic garbage accumulations.

Another type of site is shell works, a parallel term to earthwork. Shell works are more than just large, happenstance shell middens accumulations; they were purposefully constructed features, intentional borrowed, piled, arranged and formed into mounds, ridges, rings, platforms and depressions. Shell works suggest planned landscapes and terraforming to define public, domestic, sacred and ceremonial spaces, suggesting that organized labor, community planning and the ideological constructs of monumentality and ceremonialism shaped these complex maritime cultural landscapes.

This investigation offers the first large scale settlement pattern of the region and employs the only holistic maritime landscape approach. To date, 15 shell work complexes have been investigated, with over 200 radiocarbon dates generated for the region. Sites range from very small, less than half an acre, to architecturally non-complex ring shaped middens, to massive sites comprising entire islands constructed from elaborate shell work features measuring up to 100 acres in extent. Comparison of shell work forms throughout the region demonstrates significant similarities, including several recurring site forms such as ring shape features, mounds and linear ridges.

There are 13 major shell work sites, ranging in size from 30 to 100 acres in extent, which likely represent large, nucleated villages. These sites occur with a regular spatial frequency. Eight of the largest sites occur three to four miles within the northern part of the region and become less frequent toward the southern end of the region. Thirty-one small shell work sites and 12 shell rings are also present.

The most well-known and northernmost of shell work sites in the region was the Key Marco site. Unfortunately, it is now mostly destroyed by development. Frank Hamilton Cushing’s 1890s map, digitized and brought into ARC-GIS, shows the site’s occupants engineered the island landscape with shell, creating features such as radiating finger ridges, water courts, flat top mounds, plazas and canals. These shell work constructions suggest organization and a planned maritime community.

Another lager shell work site, Dismal Key, is a massive crescent shaped shell work island measuring 75 acres, containing shell mounds, ridges, plazas, canals, water courts, finger ridges and sea walls. At the northern edge of the site is a small crescent-shaped shell ring, similar in size and shape to other southeastern shell rings. South of the ring is the main portion of the site, which contains elaborate shell work architecture, including extensive shell fields and a central district of shell mounds, ramps, and canals. Two 6-meter-tall flat top shell mounds are bisected by a long central canal leading into the center of the site, suggesting a high amount of coordinated labor to build and maintain a functioning canal.

Archaeological testing determined that Dismal Key’s inner shell ring, the earliest component of the site, was built rapidly and dates to the terminal archaic. Testing of 4 of the largest flat top shell mounds suggests that intensive mound building occurred between 580 and 900 AD, a series of shell midden finger ridges at the west margins of the site are the most recently built features, dating from AD 990 to 1290. Terminal radiocarbon dates and ceramic chronology suggest Dismal Key became abandoned just prior to AD 1300.

Fakahatchee Key is a massive 98-acre shell work site with several curvilinear or ring shaped shell midden ridges. Investigation determined it contains elaborate shell works including mounds, platforms, water courts, canals, and radiating finger ridges. The curvilinear site plan of the site appears to be oriented towards the interior of the site, facing a low central area of shell fields and a large, flat, plaza-like area. Much like the Dismal Key site, the nested inner ring shaped middens of the site were determined to be the earliest dated components of the site, from BC 350 to AD 260. Also, the radiating finger ridges are the most recent features of the site, dating from AD 710 to 1280. Yeoman’s Mound is an isolated shell mound complex that appears to be purposefully separated from the main portion of the site and is to be discussed later. In tandem with Dismal Key, Fakahatchee Key appears to be abandoned just prior to AD 1300.

Survey mapping of Sandfly Key show a series of large nested crescents and rings. The earliest components of the site are the northernmost ring arms and two isolated sand and shell mounds, one of which dates to the transitional period between the Late Archaic and Early Woodland period. At the southern end of the site, Sandfly Key contains some shell work features including a flat top mound, possible house platforms, fish traps, canals, water courts and extensive shell fields. The shell work features date most recently, suggesting that over time Sandfly Key residents shifted from constructing simple ring shape middens to construction of more elaborate shell work features, suggesting an expanding community population and perhaps an increasingly complex social organization.

Russell Key is a 60-acre site and like other shell work islands, is composed almost entirely of oyster shell. Like Dismal and Sandfly Key, the northern end of the site contains a large, low shell ring almost completely buried under mangrove swamp, suggesting a post occupational sea level rise. Testing of the shell ring suggests the ring is the earliest component of the site and likely has much deeper and earlier deposits, probably dating to late archaic. South of the shell ring is the main portion of the site. It displays bilateral symmetry with a central plaza-like area. The central plaza is flanked on the east, west and south sides of the site with a series of radiating shell finger ridges. The ridges occur in distinct groupings suggesting that they were constructed as part of planned, organized activity areas, residential zones or habitation areas. Archaeological testing of these features indicated that they were built rapidly and they are contemporaneous.

As is the pattern at all other shell work sites, the radiating finger ridges at the southern edge of the site were determined to date most recently, from about AD 900 to 1200. This suggest a regional, temporal significance to these feature in that over time, Russell Key inhabitants continually expanded the site in a southern seaward direction, constructing additional habitable landscapes by continuing to build a new site area out of shell.

One of the most perplexing of shell work features are basins or depressions found around the margins of many sites. Collectively called water courts, it is not yet known what these features functioned for. These features are almost always in association with finger ridges, suggesting, perhaps, some type of fish or shell fish storage or fresh water impoundment structure. Along the southern edge of the site is one single, large water court, the largest found on Russel Key, measuring 15 by 50 meters. Radiocarbon dating places construction of this feature around AD 1030 to 1290. The presence of one large water court may suggest a shift towards a centralization or control of resources, whether fish storage, water or another function. Like the other large shell work sites, Russell Key was abandoned by AD 1300.

Today the site is thickly surrounded by mangroves. ARC-GIS spatial analysis is used to model a two-foot rise in sea level. With this scenario, the site appears more approachable by canoe and one can visualize how some of the sites finger ridges and water courts may have looked and functioned. With a two-meter high sea level rise, the long finger ridges are no longer encased in mangroves and are surrounded by water. The finger ridges likely functioned as canoe docks or jetties or functioned as platforms for people to engage in group fish netting with the nearby water courts functioning as temporary storage ponds.

Shell works demonstrate similar spatial and temporal patterns. Regionally there are strong temporal similarities and site structures, forms and layouts that imply nearby settlements must have been socially connected communities, sharing similar social, political and ideological characteristics that became manifested within their socially constructed landscapes. These constructed landscapes reflect a dynamic and recursive relationship with the environment, the sea, communities and their shellscapes. Shell works demonstrate not only a maritime cultural landscape that reflects changes in social organization over time but that the landscape itself is a repository for social memory and history and may be imbued with meaning and significance connected to a larger system of monuments and ceremonial landscapes, seascapes, and shellscapes.

For example, the Fakahatchee Key 3 site shows evidence of a possible ritual landscape suggested by the re-appropriation of the landscape features with the placement of a conical mound and two ramp projections superimposed on top of a much earlier, previously abandoned shell ring. This association or re-appropriation of the earlier features suggests that the builders of the conical mound may have viewed their earlier shell ring feature with some kind of significance, perhaps reflecting a material persistence of memory that now marks the landscape. The mound may represent a communal mortuary moment, perhaps to memorialize ancestors or it may mark a boundary, territory or forbidden place for the settlement.

A similar association is also found at Russell Key, with a flat top mound and ramp superimposed on a much earlier shell ring. These mounds may be suggestive of monuments which may have served a functional role, such as a special structure for elites or for religious use, or may have served a more ontological, cosmological or symbolic purpose.

Sandfly Key is also suggestive of a ritualized or ceremonial landscape, with a pair of conical burial mounds, out of view and deeply hidden within the mangrove swamp, surrounded by an extensive, protective ring of shell midden and separated from the rest of the site by water. The hidden nature of the mounds suggests a sacred context and their placement within a watery swamp may have further symbolic significance as water is often viewed by Native Americans as a sacred or protective supernatural barrier or portal to another world.

Lastly, the Yeoman’s Mound complex is another example of a ritual or sacred maritime landscape. The site contains a pair of two six-meter tall conical shell mounds, set along the edge of a ring or bowl-shaped midden within an arena-like complex. The interior is open and flat and is encircled by a raised ring of shell along its outer perimeter. At the southwestern edge of the site is a ramp of shell that gradually leads up into the complex, suggesting a directed entrance or perhaps a processional route into the complex. Its isolated position and separation by water also suggest secrecy or symbolic importance—purposefully separated from the secular, domestic areas of Fakahatchee Key. Human remains reported from the mounds and found within the plaza of the site suggest that it served special mortuary functions for the community.

In conclusion, the shell works of the Ten Thousand Islands represent some of the largest and most complex prehistoric shell constructions in the world and are unique, preserved prehistoric landscapes that reflect important hunter-gatherer-fisher histories. These represent an exceptional example of a prehistoric maritime cultural landscape. Nomination of these sites as a maritime cultural landscape and as a National Historic Landmark would fill an important gap in documenting and understanding the important histories of prehistoric maritime people of the world.


National Center for Preservation Technology and Training
645 University Parkway
Natchitoches, LA 71457

Email: ncptt[at]
Phone: (318) 356-7444
Fax: (318) 356-9119