In Support of Sustainability: The Historical Ecology of Vertebrate Biodiversity and Native American Harvest Practices in the Florida Keys, USA

Abstract

The Florida Keys are currently experiencing unprecedented loss of lifeways, biodiversity, and cultural heritage due to sea-level rise, catastrophic storm events, unsustainable traditions of resource exploitation, and land development. Yet, these islands have a long history of human occupation and socioecological systems underlying their current sustainability crisis that date back at least 2500 years. Here we report early results of ongoing zooarchaeological research from Upper Matecumbe Key designed to explore anthropogenic engagement with vertebrate fauna between AD 800 and 1250, providing an approximately 500-year window on marine fisheries and terrestrial faunal harvesting for this small island archipelago. Focusing on one of the few remaining, nearly intact Native American archaeological sites in the region, our research contributes to critically needed long-term anthropogenic perspectives on harvest patterns relevant to regional biodiversity conservation and sustainability initiatives.

1. Introduction

“America’s Climate Exodus is Starting in the Florida Keys”.

This recent header to a popular science news story focused on the Florida Keys (Keys) highlights its stigma as ground zero for observing the effects of climate change on sustainability in the United States. It exemplifies how, over the last 20 years, the Keys have emerged as a contemporary paragon of what the future holds for the approximately 15 million people living in vulnerable US coastal zones, and more broadly, the 2 billion people living in coastal settings worldwide (e.g., [1,2]). Located in the southeastern United States, the Keys are one of North America’s most taxonomically and ecologically rich island biodiversity environments, including several endemic subtropical flora and fauna species and the world’s third-largest coral barrier reef system. Regrettably, the region is also well known for its sustainability crisis characterized by accelerated losses of livelihoods, homes, jobs, and cultural heritage. This crisis is linked to concomitant losses of ecosystem functions and biodiversity due to sea-level rise, increased catastrophic storm events, unsustainable traditions of resource exploitation, and landscape development. Within the next 50 years, the Keys will endure approximately 36% human population displacement, the destruction of more than 500 cultural heritage sites, and continued degradation of the remaining 2% of its famed coral reefs. Although recent archaeological research in the region demonstrates a long history of human occupation across the islands beginning by at least 500 BC [3,4,5,6], an archaeological perspective of Native American human–environment relationships and associated biodiversity baselines in the past is lacking for the island chain (e.g., [7]). As a result, we are missing documentation for more than 2000 years’ worth of biocultural diversity and lifeways that preceded current Keys sustainability challenges.

To begin to fill this gap and build a long-term, historical basis for contemporary biodiversity conservation and sustainability efforts in the Keys, we present the first systematic vertebrate zooarchaeological analysis published from the region. Our study focuses on the Clupper site (8MO17) located on Upper Matecumbe Key. Clupper is a Native American archaeological site under investigation as part of the Matecumbe Chiefdom Project (MCP). Established in 2009, the MCP is an archaeological research agenda focused on elucidating the culture history and historical ecology of the Keys [3,8]. In our work, we consider sustainability to be the ability to mobilize, adapt, and maintain a thriving, integrated cultural, social, and ecological system (e.g., [9,10]). We believe that biodiversity conservation efforts are essential to supporting sustainability [11]. As argued for other world areas (e.g., [12,13,14,15,16,17,18,19,20]), we assert that archaeology and long-term historical perspectives have important roles to play in Keys and greater (sub) tropical island sustainability. They provide centennial- to millennial-scale histories of past human–environment relationships and biodiversity baselines that may reveal more sustainable lifeways in the past with potential applicability to conservation efforts in the present and future (e.g., [21,22,23,24,25]).

Because tropical and subtropical island regions have some of the world’s most ecologically sensitive marine habitats (e.g., coral reefs), support high rates of terrestrial species endemism, and are susceptible to accelerated outcomes of climate change and anthropogenic pressures, they can provide ideal model systems from which to investigate long-term archaeological perspectives of human sustainability and biodiversity (for similar arguments see [26,27,28,29,30]). Recent archaeological research across the Caribbean Archipelago (Greater Antilles, Lesser Antilles, Bahama Archipelago), Oceania (Melanesia, Polynesia, Micronesia, Australasia), and Madagascar, for example, indicate that currently vulnerable (sub) tropical island ecosystems and their associated fauna and flora were once able to support, or be managed to support, centennial-scale Indigenous human lifeways and socioecological systems without the extreme environmental degradation and biodiversity loss characteristic of more recent history (>1500 s) [31,32,33,34]. Zooarchaeology, isotopic ecology, and ancient phylogeny studies from these island regions are helping us to understand how relative sustainability in the past was linked to various human engagements with and impacts on animal species.

For example, in the adjacent Caribbean Archipelago, the study of ancient mammal DNA and tortoise biogeography from archaeological contexts shows variable patterns in genetic diversity loss or persistence as well as distribution across several endemic species that appear to correlate with Indigenous exploitation and/or translocation of species prior to European invasion [35,36]. These data provide new genetic and/or biogeographic baselines for understanding the scope of species extinctions before and after the sixteenth century. Conducting isotopic studies of commensal or introduced domestic animals across Indigenous archaeological sites in Oceania, Swift et al. [37] demonstrate significant human impacts on ecological nutrient flows following initial human settlement of the expansive island region. These works model how Indigenous groups were able to employ various adaptive strategies to sustainably manage ever-changing socioecological systems due to human migration events and landscape modifications. In southwestern Madagascar, zooarchaeological analysis of coral reef exploitation suggests relatively stable exploitation for hundreds of years during the late Holocene [38]. Compared to unsustainable Western coral reef exploitation practices in the present, sustainable exploitation of local coral reefs in the past was likely supported in part by intentional avoidance of apex predator capture. Works such as these provide baseline data for the construction of historic benchmarks applicable to goal setting in conservation efforts addressing species and ecosystem restoration, as well as the development of sustainable resource management/exploitation [12].

Within the context of the current Keys sustainability crisis, our vertebrate zooarchaeological assemblage provides an initial baseline for assessing the taxonomic diversity and habitats people encountered and interacted with on Upper Matecumbe Key between ca. AD 800 and 1250. Prior to this study, historical perspectives of human harvest practices and impacts on animal diversity in the Keys before European colonization early in the sixteenth century were largely relegated to generalized comparisons with Indigenous histories of the Caribbean Archipelago (e.g., [7]). Our research provides the first archaeologically derived biocultural baseline of vertebrate diversity and human engagement with marine, non-marine aquatic, and terrestrial fauna for the region. Although our work is ongoing, initial data contribute two significant and previously undocumented insights into the historical ecology of this region with relevance to local conservation science and sustainability initiatives:

(1)

The Keys supported a more generalist approach to marine taxa harvest in the past, relative to the exclusively targeted fisheries dominant in the archipelago since the nineteenth century. This finding supports contemporary observations from other coastal and island regions where generalist fisheries tend to be more resilient and adaptable to environmental and cultural pressures or changes (e.g., [39,40]), suggesting historical precedent for generalist fisheries in the Keys that may be applicable to contemporary fisheries management and conservation strategies.

(2)

The Keys likely supported more taxonomically diverse and geographically broad terrestrial, non-marine aquatic, and marine fauna in the past, including land-based turtle species that were possibly extirpated, more widely distributed endemic Keys raccoon and deer subspecies populations, and more abundant, regionally popular carnivorous fish populations. These findings indicate the additional loss of terrestrial and marine animal biodiversity than presently accounted for in contemporary conservation planning in the Keys (e.g., [41,42]).

We begin by providing background summaries of Keys’ environments and biota, Native American culture history, historical ecology, and sustainability efforts. We also review the Clupper site and the zooarchaeological study assemblage. Next, we present the zooarchaeological results to date and discuss their significance to the creation of long-term historical biocultural diversity baselines for the Keys. We conclude by tying our work to ongoing biodiversity conservation and sustainability research and outreach efforts.

2. Background

2.1. Florida Keys Environment and Vertebrate Biodiversity

Home to more than 73,000 residents, the Keys are part of Monroe County, Florida and comprise approximately 1700 low-lying limestone islands stretching 350 km from Miami on the southeastern Atlantic coast of peninsular Florida southwest to the Dry Tortugas (Figure 1 and Figure 2). Taken together, the total land area of the Keys is approximately 350 km². Despite its diminutive size, the region is one of 15 designated marine sanctuaries in North America and supports thousands of plant and animal species, including endemic and endangered taxa. The Keys sit at the boundary line between the Atlantic Ocean and the Gulf of Mexico, and while all the islands are limestone substrate, two distinct Pleistocene formations are present: Key Largo Limestone in the Upper Keys (e.g., Key Largo to Bahia Honda Key) and Miami Oolite in the Lower Keys (e.g., Big Pine Key to Key West). The vast majority of islands average 0.5–1 m in elevation, with only a few areas in the Upper Keys reaching 5 m or more [43]). Their location and low relief make them particularly susceptible to sea-level fluctuations, catastrophic storm events, tidal surges, flooding, and anthropogenic landscape development.

The Keys are considered subtropical with dry (November to April) and wet (May–October) seasons [43]. While contemporary annual rainfall averages 111 cm for the region [5], there is variation across islands. The Upper Keys receive greater annual precipitation (140 cm) compared to the Lower Keys (100 cm). In addition to freshwater discharges from surrounding drainage basins, rainfall is the primary natural source of freshwater across the Keys, contributing to an underlying freshwater lens, as well as freshwater ponds and seasonally filled limestone depressions across islands [43] (p. 95), [44]. Current models suggest sea-level changes were the most significant factor in the stabilization of coastlines around 1300–1000 BC, although climate fluctuations, topography, vegetative colonization, and sediment availability were also relevant [45] (p. 200). Between 13,000 and 7000 BC, sea level rose 2 m/100 years in south Florida, primarily along the Gulf of Mexico coast, but also in Florida Bay at the southern tip of the peninsula [45] (p. 207). Inundation slowed over the next 6000 years, and in the past two millennia, research in southwestern Florida documents shifts in sea levels associated with climate episodes such as the Roman Warm period (350 BC–AD 500), Vandal Minimum (VM) (AD 500–850), and Medieval Warm period (MWP) (AD 850–1200) [46,47,48].

2.1.1. Terrestrial and Non-Marine Aquatic Habitats

Beyond urban landscape development, the terrestrial habitats of the Keys include pine rocklands and tropical hardwood hammocks (hardwood tree stands surrounded by wetlands) in the Upper Keys, with Caribbean slash pine stands more common in the Lower Keys. The tree and shrub diversity of the Keys hammock stands includes 150+ species and is more similar to that found in the Caribbean islands than elsewhere in North America [49] (p. 240). Dominant circum-Caribbean species are gumbo limbo (Bursera simaruba), pigeon plum (Coccoloba diversifolia), and whitestopper (Eugenia axillaris), while more ubiquitous southeastern US taxa such as live oak (Quercus virginiana), sugarberry (Celtis laevigata), and red mulberry are absent (Morus rubra) [49] (pp. 240,244). However, due to geologic and rainfall variability differences, there are demonstrable differences in the plant biota of the Keys across the upper and lower islands. The Upper Keys (from Sands Key in Biscayne National Park to Lower Matecumbe Key) hammocks are dominated by 9–12 m tall canopies rife with species such as gumbo limbo, pigeon plum, poisonwood (Metopium toxiferum), black ironwood (Krugiodendron ferreum), and mahogany (Swietenia mahagoni). Two species now limited to the upper Keys include soldierwood (Colubrina elliptica) and lignum vitae (Guaiacum sanctum), but the latter is known to have been present in the lower Keys in the past [49] (p. 245). In the lower Keys, from Long Key through the Dry Tortugas, hammock species diversity is characterized as “low hammock” [49] (p. 246), where canopy height averages 6–7.5 m and includes species such as poisonwood, buttonwood (Conocarpus erectus), blolly (Guapira discolor), Key thatch palm (Thrinax morrisii), Spanish stopper, and white stopper (Eugenia axillaris) [49] (p. 246).

Although plant biodiversity is similar to the Caribbean Archipelago, the Keys’ terrestrial faunal diversity is more similar to Florida and the southeastern United States overall due to Florida never having been connected to the Caribbean, precluding the terrestrial dispersal of animals [49] (p. 266). Animals common to the Keys include white-tail deer (Odocoileus virginianus), raccoon (Procyon lotor), opossum (Didelphis virginiana), various snakes (e.g., black racer [Coluber constrictor]), green anole (Anolis carolinensis), pine warbler (Dendroica pinus), and northern cardinal (Cardinalis cardinalis) [49] (pp. 264–265). However, despite origin from and affinity with greater Florida terrestrial vertebrate faunal diversity, as expected, the fauna of the Keys is characterized by the “peninsula effect”, meaning the taxonomic diversity is lower than that of northern peninsular Florida and includes several endemic and sub-species-level taxa [49] (pp. 233,267). These include Key deer (O. v. clavium), Key raccoon (P. l. auspicatus), and Key Largo woodrat (Neotoma floridana smalli). Interestingly, as with other south Florida rockland ecosystems (e.g., Big Cypress region), the Keys have no endemic birds [49] (p. 269), and its avian diversity is heavily influenced by Caribbean diversity due to the natural dispersal of birds via flight. For example, zenaida doves (Zenaida aurita) and the Key West quail dove (Geotrygon chrysia) native to the Caribbean are reported to have also once been present in the Keys, possibly up until the 1800s, but are no longer found as nesting populations [49] (p. 266). Other Caribbean-derived bird taxa in the Keys include mangrove cuckoos (Coccyzus minor), black-whispered vireo (Vireo altiloquus), white-crowned pigeon (Patagioenas leucocephala), and gray kingbird (Tyrannus dominicensis). Similar to birds, the Florida tree snail (Lignus fasciatus) is native to the Keys hammock stands while also present in Cuba [49] (p. 267).

Across terrestrial habitats, freshwater ponds, swamps, and brackish water marshes are associated with several non-marine aquatic turtle species, including striped mud turtle (Kinosternon baurii), diamondback terrapin (Malaclemys terrapin), and Florida box turtle (Terrapene carolina). There are no known freshwater turtles endemic to the Keys. The Florida box turtle is found across the entire state of Florida. Overall, a general census of terrestrial and non-marine aquatic fauna demonstrates the unique position of the Keys as being influenced by multiple regions, but with a biogeographical context conducive to the development of numerous sub-species found nowhere else.

2.1.2. Marine Habitats

Marine environments dominate the Florida Keys in terms of geographic area and biodiversity, including coral reefs (e.g., patch, outer, or dead reef), inshore habitats (e.g., mangroves, sea grass meadows, oceanic channels, etc.), and offshore pelagic zones [43] (p. 175). The marine environments and biodiversity of the region are a mix of subtropical and tropical taxa, and similar to Keys plant diversity, are characteristic of greater Caribbean marine environments. Mangrove species, including white mangrove (Laguncularia racemosa), black mangrove (Avicennia germinans), and red mangrove (Rhizophora mangle) shorelines are interspersed with salt marshes and prairies, especially along the Gulf coast side, and are also characteristic of the island chain. Rocky shoreline environments due to natural erosion of the limestone bedrock or anthropogenic constructions such as canals and seawalls are common along sheltered shorelines. Naturally occurring sandy beaches are now few and far between, having been replaced by gravel or mixed sand gravel beaches due to modern sand dredging, development, and landscape modification.

Broadly speaking, the marine environments and habitats of the Keys can be organized according to proximity to shoreline and substrate type. Shoreline habitats are dominated by mangroves or beaches. Mangroves make up to 50% of Keys shorelines, are present on Atlantic and Gulf coasts, and are well adapted to fluctuating tidal levels and variable water salinity [50,51]. Because mangrove habitats sit at the interface of terrestrial, freshwater, and marine habitats, they are home to a variety of vertebrate taxa, including mammals, birds, reptiles, amphibians, and fishes. Sandy, rocky, or mixed-beach habitats are present on both Atlantic and Gulf coast sides of the Keys but are more common on the Atlantic coast. They support a variety of birds, sea turtles, and smaller-sized fish species. The beaches of the Keys are renowned sea turtle nesting habitats, boasting five of the world’s seven extant species, including loggerhead (Caretta caretta), green turtle (Chelonia mydas), leatherback (Dermochelys coriacea), hawksbill (Eretmochelys imbricata), and Kemp’s ridley (Lepidochelys kempii) [52]. Today, per the International Union for the Conservation of Nature Red List, loggerhead and leatherback turtles are classified as vulnerable, green turtles as endangered, and hawksbill and Kemp’s ridley as critically endangered.

Seagrass beds and coral reefs are the most ecologically important, taxonomically rich, and economically significant offshore marine habitats in the Keys. The seagrass species on the Gulf coast side of the islands are relatively more taxonomically diverse (e.g., five species) and geographically extensive compared to the Atlantic side, which is dominated by a single seagrass species, turtle sea grass (Thalassia testudinum) [53]. Seagrass beds are the spawning and early-life habitats for hundreds of fish species, as well as diurnal and nocturnal feeding grounds for adult fish [53]. Regardless of seagrass species, the taxonomic diversity of fishes in seagrass habitats is relatively consistent throughout the region, exhibiting very little seasonal variation [53]. Seagrass beds are also frequented by sea turtles and sharks for feeding. Turtle sea grass is the main food source for green sea turtles [54], and some requiem shark species (e.g., Galeocerdo cuvier) prefer seagrass beds for hunting [55].

The Florida Keys Reef Tract is encompassed within the Florida Key’s National Marine Sanctuary (FKNMS) and is the third-largest barrier coral reef system in the world. The tract includes both bank and patch reefs comprising soft and hard corals, with at least 63 stony coral taxa (1–45 m deep), 42 soft taxa (<30 m deep), and 70 species of marine sponge present [56] (pp. 593–594), [57]. The majority of the Keys reef tract is separated from the island chain by Hawk Channel, a 4–13 m deep channel characterized by largely bare benthic substrates [53]. However, throughout the channel and interspersed among sea grass and muddy substrates, patch reefs are present in waters up to 10 m deep and range between 30 and 70 m in diameter [56] (p. 608). Vertebrate fauna attracted to patch reefs include sea turtles, sharks, and both carnivorous and herbivorous fishes that move between reef and seagrass habitats. Beyond Hawk Channel boundaries, the Keys bank reefs are located between 1 and 8 km offshore from the islands and can be up to 6.5 km wide. The bank reefs follow the continental shelf margin along the Keys and range between 1.5 and approximately 15 m in depth. Both historically and today, star (Merulinidae), elkhorn (Acroporidae), and brain (Meandrinidae Gray, Mussidae) corals dominate [57].

While the biodiversity of the Keys reef tract includes more than 6000 species of plants and animals, the vertebrate biodiversity is dominated by ray-finned fishes. In a synthesis by Hepner [42] (p. 83) of a visual census of coral reef vertebrate taxa within the FKNMS (a no-take marine zone) between 1999 and 2016, 316 species were identified across the Keys. Significantly, out of 62 recorded families, species from five families comprised approximately 75% of the diversity: wrasses (Labridae), damselfishes (Pomacentridae), grunts (Haemulidae), gobies (Gobiidae), and parrotfishes (Scaridae). The next most abundant families comprised approximately 20% of the diversity, including silversides (Atherinidae), snappers (Lutjanidae), surgeonfishes (Acanthuridae), and jackfishes (Carangidae) [42] (p. 83). Multiple shark species inhabit Keys coral reefs including species of requiem sharks (Carcharhinidae), hammerheads (Sphyrnidae), and nurse sharks (Ginglymostomatidae) [55]. Coral reefs also attract sea turtles for feeding, such as hawksbill that feed on sponges and leatherbacks that consume jellyfishes [54]. Taken together, what contemporary marine data show is that although diminutive in terms of terrestrial land size, the marine ecosystems of the Keys support an immensely rich spectrum of habitats and are one of the most taxonomically diverse marine environments in North America.

2.2. Florida Keys Sustainability, Conservation, and Historical Ecology

Due to threats from both global climate change and regional anthropogenic impacts, sustainability in the Keys is a major focus of municipal, state, and US federal government agencies [41]. Since the mid-20th century, the Keys have had a long tradition of robust biodiversity conservation initiatives, and to date, 96% of the Keys’ landscapes and associated flora and fauna are protected for conservation purposes [58]. These areas include municipal, state, and federal lands (e.g., John Pennekamp Coral Reef State Park, Bahia Honda State Park, Key West National Wildlife Refuge, Great White Heron National Wildlife Refuge, National Key Deer Refuge, etc.), and in most cases, extend into marine habitats (e.g., FKNMS). Beginning in 2016, through the office of the Chief Resilience Officer, Monroe County established the Green Keys Sustainability Action Plan [58]. The goal of this initiative is to identify and begin formulating responses to the major threats facing the Keys, particularly climate change, sea-level rise, and major storm events. Among the eight priorities identified for sustainability efforts is a category called “natural systems” focused on the resilience and protection of the Keys’ collective biodiversity, including monitoring species vulnerability and loss. This emphasis on “natural systems” sustainability is contextualized within the need to protect the many marine and terrestrial habitats and ecosystems of the Keys, and their associated biota, because they play a central role in supporting the region’s economy, community lifeways, and the shared sense of maritime and island heritage. Despite such efforts, the low-lying geology, limited freshwater resources, and the high proportion of endemic plants and animals make the Keys home to some of the species most vulnerable to extinction due to sea-level rise and anthropogenic impacts in Florida [59,60].

The majority of biodiversity conservation and sustainability efforts in the Keys focus on marine environments. The FKNMS is the most well-known research and conservation-based administrative and legislative federal entity in the Keys (see [61]). The FKNMS is within the U.S. National Oceanic and Atmospheric Administration, and through federal and state coordination, oversees the greater Keys reef tract, creating and enacting regulations and conservation policies related to exploitation practices, habitat protection (e.g., coral reefs), educational outreach, and recreational activities. Sea turtles, Key deer, herons (and other migratory birds), and coral reef fishes are arguably the most popular biodiversity conservation foci in the region, including initiatives across the FKNMS, U.S. Fish and Wildlife Service, Florida State Parks, and numerous local organizations. Such efforts have successfully capitalized on the charismatic status of such taxa. For example, conservation efforts since the 1950s aimed at Key deer have resulted in the population growing from <50 to between 700–800 individuals in 2020 [62].

Within Keys biodiversity conservation and sustainability, historical ecology research has played a prominent role in providing decadal- to centennial-scale historical context and baselines for identifying human–environment interactions through time, anthropogenic impacts, and conservation management strategies for the present and future (e.g., [63,64,65,66,67,68,69,70,71]). Because of its attention to the role of human societies in shaping ecosystems through time, historical ecology in the region has helped highlight how humans have impacted and shaped biodiversity and ecosystem structure in the Keys for decades, and in some cases, centuries. Such impacts include the identification of historically native plants that were thought to be invasive but are critical to Miami blue butterfly conservation [63], the documentation of changes in taxonomic composition and decreasing fish sizes in sport fishing during the latter half of the 20th century [67], and the loss of up to 31% of tropical hardwood hammocks in the upper Keys within a 13-year time span (1991–2004) due to the development of highland areas [66].

Striking, however, is the lack of engagement or application of archaeology within historical ecological frameworks (e.g., [7]). As discussed next, this is in large part due to the dearth of more academically focused archaeological research in the region, but also the tremendous loss of sites suitable for professional investigation. While sixteenth-century Spanish historic accounts of the Native Americans living in the Keys describe people well versed in seafaring technology and travel, and adept at deep-water diving and fishing, they offer little else in terms of understanding the scope of culturally mediated practices and decision-making that guided socioecological systems, animal harvesting, and manifestations of biocultural diversity [72]. Unfortunately, our understanding of the region’s historical ecology earlier than the past 250 years has remained largely opaque, something the MCP aims to remedy.

2.3. Florida Keys Archaeology and Culture History

Archaeological evidence indicates that the Keys were settled by Native Americans by at least 500 BC, if not earlier [6,73]. Based on pottery typology, Native American sites in the Keys are chronologically associated with the Glades Period (ca. 500 BC–AD 1500), indicating at least 2000 years of Native American occupation and livelihoods prior to the arrival of Europeans. The Glades archaeological culture extends across greater South Florida, inclusive of the Keys, and is characterized by black-earth middens and the ubiquity of sand-tempered plain pottery along with less-abundant decorated types [73]. The vast majority of Glades sites in the Keys are located in close proximity to marine habitats. Due to coastal erosion, shoreline development, and rising sea levels, it is possible that archaeological sites indicating earlier Native American presence in the region are now lost. Paleoindian Period (12,000–6500 BC) and Archaic Period (6500–500 BC) sites are recorded in peninsular Florida, but to date, there is no verifiable evidence for occupation of the Keys during these earlier time periods [8].

We fully acknowledge that the analytical recognition and use of archaeological cultures and pottery typology is not a true reflection of how people in the past identified or organized themselves in relation to one another or the landscape. Rather, here we use the Glades chronology as a point of reference for orienting the temporality of site and landscape occupations and for recognizing data patterns through space and time. The Glades archaeological culture is currently divided into three periods: Glades I (500 BC–AD 800), Glades II (AD 800–1250), and Glades III (AD 1250–1563). Recently, Ardren et al. [3] (pp. 248–249) noted that Native Americans occupied the Keys during all three Glades periods with archaeological site trends associated with each (see also [4,5]). Glades I is typically associated with exclusively sand-tempered plain pottery, but this may be a pattern born out of the paucity of archaeological research or poor site preservation among exclusively Glades I sites. Glades II includes the earliest evidence of decorated pottery types in the region, and it is during this period that mound construction begins. Although few in number, one of the more curious cultural features of the Florida Keys are its “rock mound” sites. As their name implies, these are large artificial mounds constructed of limestone cobbles. Some of these features reached heights of three meters, such as Key Largo Site No. 3. Goggin [74] (pp. 17–19) described this site as having similar features to the Ten Thousand Islands sites in southwestern peninsular Florida; but instead of being made of shell, they were constructed of stone and included mound features, ramps, walls, and causeways. Glades III is characterized by greater abundances of decorated pottery types and the inclusion of extra-local pottery types and lithic tools indicative of extensive social networks with peninsular Florida. The majority of recorded sites for the Keys are associated with Glades III, although it is common for sites to include earlier components. By the time of Spanish arrival to the region in the early 16th century, Native Americans living in the Keys—referred to as the Matecumbe in historic documents—were part of a larger south Florida sociocultural landscape encompassing the Calusa of southwest Florida, occupants of the Ten Thousand Islands and Everglades, and the Tequesta of southeast Florida. The nature of the relationship between the people in the Keys and those on the peninsula remains unclear, with interpretations spanning relationships of dependency, tribute-paying, or autonomy. Keys people were also in contact with groups in Cuba by this time [72].

Professional archaeological research in the Keys began in the mid-20th century and was heavily focused on pre-Columbian pottery typology and chronological seriation (e.g., [74,75]); the Glades period chronology in use today was developed during this time. Early surveys of Keys pre-Columbian sites were focused primarily on the Upper Keys and demonstrated that Native American lifeways and human–environment relationships were largely oriented toward the sea and marine resources. Since this initial work, the majority of archaeological research in the Keys has been conducted through cultural resource management projects (see [4,5]). To date, there are at least 150 archaeological sites recorded from the Keys [8], the majority of which are scant representations of previously larger site locations now mostly lost to development (e.g., Stock Island site [6]). Two of the last remaining, nearly intact Glades period archaeological sites include the Clupper site on Upper Matecumbe, the focus of this study, and Key Largo 1 located within a wildlife refuge on its respective island (Figure 1). In summary, despite ample archaeological evidence indicating long-term, millennial-scale, pre-Columbian occupation of the Keys and interactions with peninsular groups, many questions remain as to how the island region fits into greater Florida archaeology, culture history, and socioecological history.

3. Material and Methods

3.1. The Clupper Site (8MO17), Upper Matecumbe Key

The Clupper site is situated on the southern end and leeward side of Upper Matecumbe Key along the Indian Key Channel. The site represents the remains of a Native American village and its associated midden or refuse mound. Today, it includes portions of the remaining midden deposits approximately 85 × 52 m in size [8] (p. 317) (see also [74]). Structural features (e.g., post holes, walls, hearths) have not been identified at the site. It was originally excavated by John Goggin and Frank Sommer [75] in the 1940s and was reported to be buffered by a mangrove swamp along the western boundary of the site, which is still common along Upper Matecumbe Key today. Goggin and Sommer’s [75] work included two large trenches, one of which extended 5 × 40 feet (1.5 × 12 m) from the mangrove swamp to the summit of the midden. Trenches revealed a black midden matrix primarily composed of vertebrate and invertebrate specimens, with smaller amounts of ceramic and shell artifacts. LiDAR topographic maps of the site suggest the presence of an identifiable apex on the summit of the mound (Figure 3); however, the function of the apex remains unclear, and the eastern edge of the site is cut off by a modern boat slip. Although now filled with sediment, two access points to freshwater (i.e., the Ghyben-Herzberg lens) were once located near the Clupper, one just north of the site boundaries and the other close by on the northern tip of Lower Matecumbe Key.

In 2014 and 2015, MCP participants revisited Clupper, conducting the only systematic archaeological investigation since Goggin and Sommer’s initial work. A total of five 50 × 50 cm test pits were excavated in 10 cm levels by trowel across the midden. Sediments were screened over 1/8-inch (0.318 cm) mesh. Historic or modern finds were present in the first levels of several test pits, though once excavation continued beyond the first 10 cm below the surface (cmbs) (e.g., Level 1), the matrix was clean of intrusions. The sediment across test pits was relatively homogenous in color and texture throughout with the primary stratigraphic difference occurring at approximately 20 cmbs (e.g., Level 2) where the soil changed from a loose humic layer to darker, greasier soil characteristic of intact black earth middens in the region. At approximately Level 6, the test pits were composed of a gray concretion layer, including pockets of concreted bone. This layer is not unique to the site or the Keys, but rather is ubiquitous across most south Florida archaeological sites and is believed to be the result of natural processes causing limestone to change to calcrete sediments (e.g., [76]). The concretion layer averaged 10 cm in thickness, and cultural material was recovered beneath the concretion for a total depth of approximately 1.10 m.

Although analysis from the 2014 and 2015 excavations is still ongoing, initial results provide an opportunity to begin to construct the first absolute chronology for the site and a biocultural baseline of animal biodiversity and harvest. Across the test pits, Glades II pottery (ca. AD 800–1250) was the dominant type, indicating the site was primarily occupied during this timeframe and corroborated by AMS radiocarbon dates with local marine reservoir correction from the site [77]. This chronology places site occupation during the transition of the VM to the MWP, a time of rising temperatures in the greater south Florida region culminating in regimes similar to or warmer than today. In addition to pottery, and one lithic artifact, the overwhelming majority of finds from Clupper were vertebrate and invertebrate faunal specimens, including bone and shell artifacts. While vertebrate specimens are the focus of this paper, the analysis of invertebrate specimens is currently underway. Similarly, bulk soil samples were collected from test pits and are the focus of paleoethnobotanical analyses. Any zooarchaeological specimens included within the bulk samples will be the subject of future study.

3.2. Zooarchaeological Sample and Methods of Analysis

The results presented below are based on the zooarchaeological analysis of vertebrate specimens from Test Pit 4 (TP 4). TP 4 was located at the highest elevation point of the midden and excavated to a depth of 80 cmbs where excavation was halted due to time constraints. The sediments present in the test pit across the eight excavated levels (Levels 1–8) were typical of black earth midden ubiquitous at Glades periods sites in the Keys. As was characteristic across all five test pits at the site, TP 4 had relatively little stratigraphic differentiation in terms of sediment color and texture (Figure 4). Nonetheless, there was a noticeable trend in the abundance of vertebrate specimens recovered across the levels. Levels 1–3 and 8 had noticeably fewer vertebrate specimens than Levels 4–7. Although a more complete understanding of the stratigraphic patterning in vertebrate specimen abundance requires additional zooarchaeological analyses from the Clupper site, for now, we suggest that the patterning does not reflect differential preservation across the levels, but rather more concentrated midden accumulation events and human occupation of the site.

Based on analysis of AMS radiocarbon dates with local marine reservoir correction from invertebrate shell (Codakia orbicularis) and white-tail deer bone specimens from TP 4 Levels 1, 2, 3, and 6, the chronological sequence of the test pit falls within the approximately AD 800–1250 time range attributed to the site [77]. It is important to note that two articulating pieces of a single white-tail deer metatarsus were recovered from levels 5 and 7. Both the radiocarbon chronology of TP 4 as well as the cross-mending metatarsus suggest that this portion of midden possibly formed due to rapid deposition events and/or post-depositional sediment mixing. As a result of these indicators, and the overall homogenous stratigraphy of all units at the Clupper site, vertebrate specimens recovered from TP 4 were initially analyzed by 10 cm excavation levels and then analytically combined for assemblage-level summary quantification and interpretation. The level-based analysis provides a comparative foundation for general chronological patterning and for future comparative analyses across additional vertebrate samples from other site contexts and test pits. However, in this study, we primarily focus on assemblage-level summary results as a first baseline for identifying general trends in vertebrate harvest at the site between AD 800 and 1250 and to compare with contemporary records of vertebrate animal biodiversity in the Keys.

All study specimens were identified at the Florida Museum of Natural History (FM), University of Florida (UF). Taxonomic identification was based on documented historical and contemporary vertebrate diversity in the Florida Keys, peninsular south Florida, and surrounding Caribbean islands in conjunction with the use of modern comparative skeletal collections at UF. Collections used include those in the Laboratory of Southeastern Archaeology, the FM Environmental Archaeology Laboratory, and the FM Herpetology, Mammalogy, and Ichthyology Collections. Together, the comparative collections comprise one of the most comprehensive southeastern United States and Caribbean-focused skeletal collections available (e.g., [78]). Specimens were identified to the lowest taxonomic level possible, including class, order, family, genus, and species identifications. Data collection included the recording of the number of individual specimens (NISP), weight (g), element present, element portion, element side, as well as the calculation of the minimum number of individuals (MNI) based on element identification, portion, side, size, age markers, and articulation across specimens. Because summary data were analytically combined across levels, total assemblage NISP and MNI values represent the potential maximum of each measure represented in the assemblage.

The species diversity (H′) and equitability (V′) of the total TP 4 assemblage was calculated using the Shannon–Weaver Index based on the total MNI (n = 377) and natural log_e [79] (p. 111).

$$\mathrm{H}\prime =-{\displaystyle \sum }_{i=1}^s\left(p_i\right) (\log p_i)$$

(1)

Principle component analysis (PCA) was conducted using MURAP software to evaluate and graphically show how taxa contributed to variance across the sample; in this case, relationships of abundance as measured by NISP and associations between taxa within the assemblage. For the PCA, identifications were integrated to represent the most taxonomically refined groupings in the assemblage.

4. Results

4.1. Assemblage Summary

A total of 5819 vertebrate NISP representing 377 MNI was identified from the TP 4 assemblage with a combined weight (Wt.) of 2826 g (

Table 1. Major vertebrate categories identified in Test Pit 4.

General Taxon	NISP	% NISP	Weight (g)	% Weight	MNI	% MNI
Unidentifiable vertebrate (Vertebrata)	424	7.29	89.28	3.16	0	0
Cartilaginous fish (Chondrichthyes)	273	4.6	136.59	4.83	18	4.77
Fish (Actinopterygii)	3027	52.02	1120.30	39.64	325	486.21
Reptile (Reptilia)	2075	35.66	1458.62	51.61	25	6.63
Bird (Aves)	8	0.14	2.50	0.09	5	1.33
Mammal (Mammalia)	12	0.21	18.73	0.66	4	1.06
Total	5819	100.00	2826.02	100.00	377	100

). Approximately 69% of the NISP (n = 4044) were identifiable beyond class, including taxonomic resolution at the level of 3 orders, 11 families, 17 genera, and 22 species. The most abundant taxonomic groups identified overall are ray-finned fishes followed by reptiles and cartilaginous fishes, and to a much lesser extent, mammals and birds (

Table 2. Vertebrate taxon identifications from Test Pit 4 at the Clupper site. * Indicates taxa associated with multiple habitats spanning general terrestrial, freshwater, brackish water, and marine habitats.

Taxon	Common Name	NISP	% NISP	Wt. (g)	% Wt. (g)	MNI	% MNI
Vertebrata	unidentified vertebrate	424	7.29	89.28	3.16	0	0.00
Carcharhiniformes	ground sharks	88	1.51	19.43	0.69	0	0.00
Ginglymostoma cirratum *	nurse shark	14	0.24	3.86	0.14	2	0.53
Carcharhinidae *	requiem sharks	40	0.69	49.98	1.77	4	1.06
Rhizoprionodon terraenovae	Atlantic sharpnose shark	3	0.05	3.75	0.13	1	0.27
Sphyrna sp. *	hammerhead shark	90	1.55	42.22	1.49	5	1.33
Pristis sp. *	Sawfish	9	0.15	4.61	0.16	3	0.80
Rajiformes *	flattened cartilaginous fishes	25	0.43	7.48	0.26	2	0.53
Aetobatus narinari *	spotted eagle ray	4	0.07	5.26	0.19	1	0.27
Actinopterygii	ray-finned fishes	1341	23.05	323.16	11.44	0	0.00
Elops saurus *	ladyfish	2	0.03	0.06	0.00	2	0.53
Megalops atlanticus *	Atlantic tarpon	3	0.05	0.74	0.03	2	0.53
Albula vulpes *	bonefish	65	1.12	10.73	0.38	7	1.86
Gymnothorax sp. *	moray eel	2	0.03	0.38	0.01	2	0.53
Ariidae	catfishes	132	2.27	50.5	1.79	0	0.00
Ariopsis felis *	hardhead sea catfish	470	8.08	230.44	8.15	162	42.97
Bagre marinus *	gafftopsail sea catfish	19	0.33	5.04	0.18	2	0.53
Centropomus sp. *	snook	4	0.07	4.5	0.16	2	0.53
Epinephelidae	groupers	27	0.46	12.38	0.44	0	0.00
Epinephelus sp. *	grouper	126	2.17	89.27	3.16	20	5.31
Mycteroperca sp. *	grouper	43	0.74	31.14	1.10	11	2.92
Strongylura marinus *	Atlantic needlefish	5	0.09	0.88	0.03	2	0.53
Tylosaurus crocodilus *	hound needlefish	12	0.21	1.57	0.06	3	0.80
Pomatomus saltatrix *	bluefish	1	0.02	0.1	0.00	1	0.27
Carangidae	jacks, pompanos, scads	72	1.24	77.48	2.74	0	0.00
Caranx sp. *	jack	101	1.74	58.14	2.06	16	4.24
Trachinotus sp.	pompano	20	0.34	19.26	0.68	6	1.59
Lutjanus sp. *	snapper	110	1.89	41.14	1.46	16	4.24
Haemulon sp. *	grunt	343	5.89	46.18	1.63	31	8.22
Sparidae	porgies	2	0.03	0.27	0.01	0	0.00
Calamus sp. *	porgy	23	0.40	17.32	0.61	7	1.86
Pogonias cromis *	black drum	6	0.10	4.87	0.17	1	0.27
Mugil sp. *	mullet	22	0.38	2.26	0.08	5	1.33
Pomacanthidae	marine angelfish	5	0.09	0.73	0.03	4	1.06
Lachnolaimus maximus *	hogfish	20	0.34	24.8	0.88	9	2.39
Sparisoma sp. *	parrotfish	2	0.03	0.34	0.01	1	0.27
Sphyraena sp. *	barracuda	30	0.52	57.84	2.05	6	1.59
Trichiurus lepturus *	largehead hairtail	1	0.02	0.13	0.00	1	0.27
Istiophorus albicans	Atlantic sailfish	4	0.07	5	0.18	2	0.53
Paralichthyes sp.	flounder	4	0.07	1.16	0.04	1	0.27
Ostraciidae *	boxfishes	7	0.12	0.79	0.03	1	0.27
Diodontidae *	porcupinefishes	3	0.05	1.7	0.06	2	0.53
Testudines	turtles and tortoises	210	3.61	63.43	2.24	0	0.00
Chelydra serpentina *	common snapping turtle	2	0.03	0.45	0.02	1	0.27
Kinosternidae	mud and musk turtles	2	0.03	0.57	0.02	0	0.00
Kinosternon sp.	mud turtle	5	0.09	1.7	0.06	2	0.53
Sternotherus sp.	musk turtle	1	0.02	0.14	0.00	1	0.27
Gopherus polyphemus *	gopher tortoise	1	0.02	1.98	0.07	1	0.27
Cheloniidae	sea turtles	1785	30.68	1270.75	44.97	5	1.33
Caretta caretta *	loggerhead sea turtle	4	0.07	49.71	1.76	3	0.80
Chelonia mydas *	green sea turtle	63	1.08	69.08	2.44	10	2.65
Apalone ferox *	Florida softshell turtle	2	0.03	0.81	0.03	2	0.53
Aves (Small)	birds (small; warblers)	1	0.02	0.08	0.00	1	0.27
Aves (Medium)	birds (medium; e.g., gulls)	4	0.07	1.11	0.04	1	0.27
Aves (Medium-Large)	birds (large; e.g., herons)	1	0.02	0.23	0.01	1	0.27
Ardeidae *	Herons	2	0.03	1.08	0.04	2	0.53
Mammalia (Medium-Large)	mammals (medium-large; e.g., raccoon, deer)	1	0.02	0.23	0.01	0	0.00
Mammalia (Large)	mammals (large; e.g., deer)	3	0.05	1.55	0.05	0	0.00
Procyon lotor	raccoon	1	0.02	0.17	0.01	1	0.27
Odocoileus virginianus cf. clavium	white-tailed deer (cf. Key deer)	7	0.12	16.78	0.59	3	0.80
Total		5819	100.00	2826.02	100.00	377	100.00

). Amphibian taxa were not identified in the assemblage. Together, specimens identified to the level of ray-finned fishes (Class: Actinopterygii, n = 1311) and sea turtles (Family: Cheloniidae, n = 1782) account for approximately 53% of all vertebrate taxonomic identifications. The majority of ray-finned specimens identified as Actinopterygii were undiagnostic vertebrae or fragmented spine elements and specimens identified as Cheloniidae were, by and large, fragmented carapace or plastron elements. The preponderance of fragmented ray-finned fish and sea turtle specimens identified to the level of class and family, respectively, is common within tropical marine zooarchaeological assemblages. Such assemblages often feature high taxonomic diversity represented by taxa with morphologically similar skeletal elements such as thoracic and caudal vertebrae in the case of fishes and costals for sea turtles (e.g., [80,81]).

The TP 4 assemblage is relatively high in taxonomic diversity (H′ = 2.580) due to the number of taxa represented overall, especially fishes. This diversity, however, is not equitable (V′ = 0.663) across the assemblage because of the predominance of select taxa representing the majority of specimens identified. These taxa include sea turtles, ground sharks (Carcharhiniformes), hammerhead sharks (Sphyrna sp.), catfishes (Ariidae, Ariopsis felis, Bagre marinus), groupers (Epinephelus sp., Myctoperca sp.), jackfishes (Carangidae, Caranx sp.), snappers (Lutjanus sp.), and grunts (Haemulon sp.). The pattern of relatively high taxonomic diversity and low equitability is also reflected in the PCA. The PCA reveals clear patterns in variance driven by select taxa within TP 4, as well as across levels that may indicate changes in faunal harvest through time at the Clupper site (Figure 5). Not surprisingly, ray-finned fish taxa and sea turtles are the major determinants of vertebrate variance in TP 4.

In addition, less-abundant taxa such as mammals and mud turtles (Kinosternon sp.) impact variance, showing inverse relationships between their relative lack of abundance compared to that of sea turtles. Notably, catfishes and sea turtle specimens contribute the most to taxonomic variance within the assemblage as well as across levels. The two taxa have an inverse relationship where sea turtle specimen abundance increases from Levels 1–8 and catfish specimen abundance decreases from Levels 1–8. From a temporal perspective, this suggests a possible decrease in sea turtle exploitation and an increase in catfish exploitation over time as represented in TP 4. Catfishes, mammals, and mud turtles together are more abundant in Levels 1–3 compared to the dominance of sea turtle representation in Levels 5–7. The remaining fish, reptile, and bird taxa cluster closely together, indicating a negligible impact on assemblage variance. Levels 4 and 8 are differentiated from each other and from Levels 1–3 and Levels 5–7 by a relatively coeval representation of sea turtles and less-dominant taxa.

In terms of habitats represented in TP 4, most taxa identified are associated with marine habitats (e.g., coral reef and non-coral reef habitats) followed by brackish, freshwater, and terrestrial environments (Figure 6). However, it is important to note that the habitat associations across taxa are not mutually exclusive, with 52% of the identified taxa from TP 4 known to be associated with multiple environments throughout their life cycle (see Table 2).

4.2. Marine-Associated Fauna

Per taxonomic identification, as measured by NISP and MNI, catfishes, groupers, jackfishes, snappers, and grunts are the most prevalent ray-finned fish families represented in our sample from Clupper. Together, taxa from these families comprise approximately 88% and 82% of the total fish NISP and MNI, respectively (Figure 7). All of these fishes are known to inhabit marine environments, including mangrove, sea grass, coral reef, and brackish water habitats during their life cycles. The less-abundantly represented ray-finned fish taxa in the TP 4 assemblage also typically inhabit more than one marine habitat (Table 2). The one exception to this generalized pattern is marine catfishes. They are not known to be common among coral reef habitats and were likely caught around nearshore waters with rocky outcrops. The hardhead catfish is known to frequent deeper offshore waters during cooler winter months and shallower waters beginning in the spring. Gafftopsail catfishes (Bagre marius) will enter brackish waters throughout the year.

It is important to note the possible effect of preservation bias on the apparent abundance of the ray-finned fish taxa, hardhead catfish (Ariopsis felis) in particular. While all the most-abundant fish taxa are represented by cranial or post-cranial bone elements, catfishes are represented by 397 otoliths and 224 other bone elements (i.e., not otoliths), with otoliths accounting for the relatively large MNI (n = 164) calculated for catfishes compared to other fishes. Compared to the other ray-finned fishes identified, catfishes have large, dense otoliths that preserve exceptionally well in south Florida archaeological sediments (e.g., [82]). At this early stage of zooarchaeological investigation of the Clupper assemblage, it is not possible to conclude whether or not the abundance of well-preserved otoliths affects our perception of catfish abundance overall in comparison to other ray-finned fishes. Additional study of vertebrate samples from the site as well as from other Native American sites in the Keys will help shed light on whether otoliths are uniformly the most common catfish elements present in vertebrate assemblages, and thus potentially representing preservation bias. An alternate possibility is a correlation with elemental representation suggestive of differential processing and/or cooking techniques across fish taxa (e.g., catfish heads are more abundantly represented in assemblages compared to the cranial representation of other fishes).

Identified cartilaginous fishes (sharks and rays) are dominated by requiem and hammerhead sharks. Compared to ray-finned fishes, sharks and rays are far less abundant, with 273 NISP and 18 MNI. Here, again, the difference in abundance between the two classes of fishes may be due to preservation bias as cartilaginous fishes have fewer elements (e.g., vertebrae, teeth, and stingers) preserved in the archaeological record compared to the numerous bony skeleton elements of ray-finned fishes. While it is likely sharks and rays are underrepresented overall, this pattern at Clupper is common across south Florida coastal and island Native American archaeological sites, suggesting that sharks and rays were not as frequently exploited across the greater region. The shark taxa identified from the TP 4 assemblage would have been accessible among near-shore marine habitats such as mangroves, shallow sandy beach shorelines, and sea grass beds, as well as in deeper waters such as those in the nearby Indian Channel and coral reefs on the leeward side of Upper Matecumbe Key.

Sea turtles represented in the TP 4 assemblage (1848 NISP and 18 MNI) include at least two species present in the Keys today—loggerhead and green turtle. Although based on a small sample size to date, green turtle identifications were more common in the assemblage with 63 NISP and 10 MNI compared to loggerhead with 4 NISP and 3 MNI. While both species of sea turtles are present in the Keys throughout the year, they are encountered by people most commonly between April and September during nesting season along sandy beaches and inshore waters. At the end of nesting season, each species is also known to embark on foraging migrations among deeper, off-shore waters [83,84]. Both species’ nesting seasons overlap, and each frequents near-shore sea grass beds, coral reefs, and floating sargassum mats throughout their life cycles.

The only bird taxa identified beyond class (Aves; NISP = 8, MNI = 5) was heron (Family: Ardeidae; NISP = 2, MNI = 2). Herons are ubiquitous across south Florida coastal environments, including marine coastline habitats, mangroves, swamps, and freshwater sources. Although it is not possible to identify the TP 4 specimen beyond family at this time, the Great White Heron (Ardea herodias occidentalis), specifically, is native to the southernmost portions of peninsular Florida and the Keys [85].

4.3. Terrestrial and Non-Marine Aquatic Associated Fauna

Terrestrial-associated vertebrate fauna identified in the TP 4 assemblage include mammals, gopher tortoise, and non-marine aquatic turtles. Amphibian taxa were not identified, and while common among vertebrate assemblages in the Everglades, they are not common components of assemblages from coastal sites in southern Florida (e.g., The Pineland Site Complex, located on Pine Island, Lee County, FL, USA). Three species from three different turtle families and two genera from the mud turtle family (Kinosternidae) are present at Clupper. While not abundant, the diversity of turtles is intriguing. Of the threeturtle species considered native to the Keys (Kinosternon baurii, Malaclemys terrapin, Terrapene carolina), the mud turtle (genus Kinosternon sp.) is the only one identified in the TP 4 assemblage. The striped mud turtle (Kinosternon baurii) is endangered within the Lower Keys. The remaining taxa identified in TP 4 are not known to be present in the Keys, either historically or today [84], including the snapping turtle (Chelydra serpentina; NISP = 2, MNI = 1), musk turtle genus Sternotherus sp. (NISP = 1, MNI = 1), gopher tortoise (Gopherus polyphemus; NISP = 1, MNI = 1; U.S. federally classified threatened species), and the Florida softshell turtle (Apalone ferox; NISP = 2; MNI = 2). Although the abundance of these taxa is significantly lower than sea turtles, those identified are interesting from a biocultural diversity perspective because the lack of non-marine turtles in the Keys, compared to peninsular Florida, has long been attributed to the general scarcity of consistent freshwater sources across the islands [86].

Of the 12 mammal specimens present in the assemblage, only two species are identified: white-tail deer (NISP = 7, MNI = 3) and raccoon (NISP = 1, MNI = 1). Currently, there are five raccoon subspecies recognized within Florida, with three traditionally believed to be endemic to islands in the Keys: Matecumbe Key raccoon (P. l. inesperatus), Key Vaca raccoon (P. l. auspicatus), and Torch Key raccoon (P. l. incautus) [87], though recent genetic analyses suggest that these three subspecies are actually members of a single subspecies, P. l. auspicatus (now called Keys raccoon) [88]. There are no trends in skeletal morphological diversity available for Key raccoons; however, the P. lotor taxa in the Keys are regarded as the smallest sized of the species present in the United States [89,90]. It is likely the specimen from TP 4 is a Keys-endemic raccoon subspecies. The white-tail deer specimens likely represent individuals of the famous “Key deer,” an endangered subspecies endemic to the lower Florida Keys (e.g., Big Pine Key to Sugarloaf Key) and the smallest-sized white-tail deer in North America. The identifiable elements from the TP 4 assemblage (e.g., metatarsus) were fully fused, representing an adult deer, and were definitively smaller than fused white-tail deer metatarsi found in the FM modern deer comparative collection from peninsular Florida. Interestingly, four of the six white-tail deer specimens identified are modified, showing evidence of implement manufacture. While little is known about the historical distribution of Key deer prior to the 20th century, they are believed to have once been ubiquitous across the islands from Vaca Key to Key West [91]. Their contemporary distribution correlates with freshwater resources more readily available in the lower Keys than to the north.

5. Discussion: Toward a Millennial-Scale Biocultural Vertebrate Baseline

Our work represents the first systematic zooarchaeological investigation of the ecological habitats and vertebrate fauna that Native American people at Clupper engaged with and harvested between ca. AD 800 and 1250. Unique to the Keys region thus far, our work expands the temporal depth of current anthropogenic perspectives of Keys historical ecology over more than 1000 years. When considered within the context of marine, non-marine aquatic, and terrestrial biodiversity conservation, our results provide new data about the historical distribution of some of the region’s most economically important and ecologically sensitive animals. The data also shed light on Native American subsistence strategies and provide an opportunity to consider how they may have been linked to supporting the sustainable occupation of Clupper for at least 500 years.

Based on zooarchaeological identifications thus far, the dominant ray-finned fish, sea turtle, mammal, and bird taxa represented at Clupper are still extant in the Keys region. However, several of these species are now classified as vulnerable, endangered, or occupying threatened environments. Our biocultural baseline indicates that vertebrate fishing and hunting were mostly focused on a relatively high diversity of ray-finned fishes and sea turtles, with comparatively minor harvesting of birds, mammals, and non-marine aquatic turtles. The taxa identified would have been accessible across several of the region’s most vulnerable habitats today, including mangroves, sandy beaches, inshore sea grass beds, shallow sandy bottom habitats, and coral reefs. Terrestrial or non-marine aquatic taxa would have been available in habitats with proximity to freshwater sources, marshes, and hardwood or shrub vegetation.

5.1. Marine Vertebrate Diversity and Harvest

When compared to the dominant ray-finned fishes identified in Hepner’s [42] contemporary synthesis of coral reef fish diversity, there is little overlap with the dominant taxa identified in TP 4 at Clupper. Grunts are the only fish family common among the most abundant taxa on both lists (

Table 3. A comparison of the most abundant fish families represented in Hepner’s [42] (p. 83) coral reef survey and ray-finned fish families identified in Test Pit 4.

Hepner [42]		Test Pit 4
Family	% of Abundance of All Fish Families Observed (n = 64)	Family	% of Total NISP for Ray-Finned Fish Families (n = 24)
Labridae	20.46	Ariidae	36.83
Pomacentridae	17.35	Haemulidae	20.34
Haemulidae	16.86	Epinephelidae	11.63
Gobiidae	10.47	Carangidae	11.45
Scaridae	9.23	Lutjanidae	6.52

). Interpreting this apparent dissonance between the two different datasets is not straightforward in terms of building a long-term biocultural baseline of marine vertebrate diversity and exploitation. The datasets are not commensurate in temporal or taxonomic resolution, location of focus, or parameters of observation (e.g., live versus archaeological), but these analytical challenges are not uncommon when using long-term historical datasets within contemporary considerations of biodiversity and its conservation [92,93]. The point is to acknowledge the differences in scale between historical and more contemporary datasets and use historical data as a guide for discovering biodiversity loss and changes through time—and in the case of our work with ancient Keys fisheries—within the context of human lifeways and pressures across centuries and millennia not otherwise knowable outside of the archaeological record.

We acknowledge that Hepner’s data and the TP 4 zooarchaeological data are separated by approximately 800–1200 years, and that one dataset is focused on annual scales of observation and the other spans centuries. Hepner’s survey also focused on the Keys coral reef tract exclusively, which is home to many adult-stage fishes, while the Clupper zooarchaeological assemblage may represent juvenile or adult-stage fishes exploited from several marine habitats (e.g., mangroves, sea grass beds, coral reef) (e.g., [94]). Because we cannot yet conclusively interpret whether the most dominant fish taxa from Clupper were exploited from non-coral reef inshore or coral reef habitats, it is not possible to deduce the role of habitat when interpreting the differences between the two sets of taxonomic diversity data. Current zooarchaeological analysis is focused on the calculation of fish size and biomass across the most abundant taxa represented at Clupper to infer life stage and correlations with likely habitats and possible seasons of capture (e.g., [80]).

Despite the ambiguity regarding the influence of habitat on the TP 4 assemblage, the differences in taxonomic diversity among the most abundant taxa between the two datasets suggest there have been potentially significant changes in ray-finned fish diversity and equitability in the Keys over the past 1200 years. This is intriguing from the perspective of links between biodiversity and sustainability. Three of the most abundant and consistently represented fish taxa across the 500-year time span represented in TP 4 (groupers, snappers, and jacks) are among the most economically important and popular in recreational and commercial Keys fisheries today. Yet, species from these taxa contribute less than 25% to current coral reef fish diversity per Hepner’s study. Additionally, historical marine baselines of 20th century sport fishing in the region document precipitous decreases in the abundance and size of groupers and snappers since at least the 1950s [67]. Our study indicates that groupers, snappers, and jacks were available for relatively consistent exploitation at Clupper for at least 500 years and that sometime between AD 1250 and the 20th century, these groups of fishes began to decline in abundance.

Unlike groupers, snappers, and jacks, the two other abundant ray-finned fish taxa in the TP 4 assemblage, grunts and marine catfishes, are not traditionally popular recreational or commercial fishes in the Keys. Species of each fish type frequent both inshore and offshore habitats throughout their lifecycles and none are under direct stress from targeted fishing efforts. Grunts and marine catfishes are unregulated species per the Florida Fish and Wildlife Conservation Commission [95]. Although not as popular a source of food as snapper or groupers, grunts are considered an edible white fish in the Keys and are part of a regional dish called “grits and grunts.” Given that grunts are among the most abundant fishes documented on both the zooarchaeological and Hepner’s [42] lists, we assert that their apparent relative abundance through time and lack of current fishing regulations suggest overall population stability. This is not to say that grunts are immune to the negative effects of current anthropogenic impacts on Keys marine biodiversity and coral reefs overall (e.g., [96]), but rather that, from a long-term, century–millennium-scale perspective of abundance and human fishing pressure, their Native American harvesting appears to have been relatively sustainable through time. Marine catfishes, on the other hand, are not commonly sought after in commercial or recreational fisheries of the region today and are not a popular food source by local reputation. In contrast to the present, our data suggest marine catfishes were readily exploited by the residents of the Clupper site—a pattern common across Native American and Colonial coastal and mangrove sites in greater south Florida (e.g., [97,98]).

Following ray-finned fishes, the TP 4 assemblage from Clupper shows that sea turtle taxa were also a significance focus of Keys marine hunting by Native Americans. As with all sea turtle species globally, the two identified at Clupper—loggerhead and green turtle—are currently endangered. The apparent predominance of green turtles at Clupper suggests this species may have been more readily available or specifically targeted for hunting near the site. The capture of both loggerhead and green turtles could have occurred on sandy beaches, among inshore habitats such as sea grass beds, or along coral reefs. Our data are significant from a historical perspective of sea turtle fisheries in the Keys, where the region was ground zero for US sea turtle exploitation during the nineteenth and twentieth centuries in conjunction with those from nearby Cuba that decimated populations [99]. For approximately two centuries, green turtles were the preferred catch, followed by the less-desired loggerhead and hawksbill [100]. Since the 1970s and the implementation of the US Endangered Species Act, however, conservation efforts have vastly improved sea turtle population sizes across the Keys (see https://www.fisheries.noaa.gov/sea-turtles (accessed on 15 February 2022)). The relative abundance of loggerhead and green turtle identified at Clupper thus far suggests that these species have been subjected to anthropogenic hunting pressures since at least AD 800, providing new zooarchaeological data from which to begin to consider biocultural baselines of sea turtle diversity and exploitation several centuries prior to the devastating effects of more recent fisheries (e.g., [65]).

In summary, based on the taxonomic and habitat diversity represented in the TP 4 vertebrate assemblage, coupled with the dominance of a consistent suite of carnivorous fishes for 500 years, we assert that the Native American communities living at Clupper likely practiced a generalist marine fishing strategy characterized by the ability and/or need to exercise adaptability based upon variable environmental or cultural parameters (e.g., signs of resource depression, climate change, regulatory constraints, dietary taboos) to foster sustainable capture. Such a strategy often includes switching between opportunistic or targeted harvesting across different habitats and taxa throughout the year. Ethnographic and community-based participatory studies indicate that generalist fishing strategies among Indigenous and/or traditional small-scale subsistence fisheries are able to support the diversification of marine resource use and mitigate potentially negative impacts of persistent exploitation pressures [39]. Generalist strategies also correlate positively with flexibility in human responses to internal (i.e., cultural) and external (i.e., natural) forces of socioecological change that may impact ecosystem resilience as well as human access to marine resources [101]. For example, among Inuit (Canadian Arctic), Cree (Eastern Canada), and Coastal-Vedda (Sri Lanka) fisheries—despite their vastly different environments, geographies, histories, and cultures—diversification is key to the development of adaptive management strategies [101,102,103]. Among the Inuit, this includes diversification of fishing technology, livelihood diversification among the Coastal-Vedda, and methods of pulse-fishing among the Cree. In a complementary vein, archaeological investigations from several island regions worldwide indicate that generalist fishing strategies were also integral to long-term persistence and socioecological sustainability among Indigenous communities in the past. Exhaustive pressures on marine and coastal taxa, as well as environmental changes, were actively anticipated and managed through culturally mediated practices aimed at sustaining resource accessibility and use (e.g., Aleutian Islands [104]; California Channel Islands [105], Persian Gulf [106]).

Our suggestion of a generalist approach to marine vertebrate harvesting at Clupper contrasts with the taxonomically limited focus of commercial and recreational Keys fisheries today. Often using longline fishing techniques, contemporary vertebrate fisheries primarily target large-sized adult carnivorous fishes. The deleterious ecological effects of these targeted practices are well documented and are a major driver of the Keys sustainability crisis [71]. From the perspective of taxonomic diversity, at Clupper, we do not have evidence of species-specific fisheries that would have necessitated limited technological protocols or habitat-specific applications. Similar suggestions have been made for other south Florida coastal Native American sites with comparable taxonomic diversity (e.g., [47,97]). Nets and traps would have been effective fishing techniques among mangrove and inshore habitats and canoes would have facilitated hook and line or spear fishing along offshore coral reef patches and tracts as well as across pelagic waters. Both the most abundantly and less-represented marine species from TP 4 could have been exploited via a combination of opportunistic mass capture (e.g., nets) approaches or through more targeted (e.g., hook and line) efforts depending on taxa size, behavior (e.g., schooling, solitary), age, and habitat, as well as the aim of the people fishing. While our current interpretation remains a point of further archaeological analyses, including the documentation of possible paleoenvironmental or cultural changes that may have impacted Native American harvest practices at Clupper, as well as concerted use of ethnographic analogy across Indigenous fisheries, it provides a new perspective of Keys marine historical ecology and a biocultural baseline highlighting the antiquity of human engagement with marine vertebrate biodiversity in the region up to 1250 years ago.

5.2. Terrestrial and Non-Marine Aquatic Vertebrate Diversity and Harvest

Although not a major focus of vertebrate harvesting, the terrestrial and non-marine aquatic vertebrates identified at Clupper also present interesting considerations for biocultural baselines in the Keys because of what they may indicate about the historical biogeography and loss of endemic species. All the turtles present in the assemblage are commonly identified at archaeological sites in Florida with similarly low abundance compared to other taxa (e.g., [107,108,109]). Gopher tortoise, snapping turtle, musk turtle, and Florida softshell turtle distributions across peninsular Florida correlate closely with habitats near freshwater resources. The same correlation is identified among contemporary mud turtle distribution in the Keys. Two plausible hypotheses for the presence of the four turtles not currently present in the Keys but identified at Clupper are as follows: (1) Each type of turtle was native to at least the Upper Keys in the past and has since been extirpated from the subregion, or (2) Native American communities translocated each taxon from peninsular Florida to the Keys. If the first hypothesis is correct, then we assume the terrestrial landscapes of the Upper Keys included more freshwater habitats and access points to freshwater in the past, supporting more terrestrial or freshwater turtle diversity. Accordingly, the low proportion of these turtles in the TP 4 assemblage may indicate that the taxa were not a commonly sought resource in the Keys relative to other vertebrates, as has been archaeologically documented at peninsular Florida sites. It is also possible that if native to Upper Matecumbe Key, their population numbers were naturally restricted or experienced stress during the occupation of Clupper and therefore were unable to support robust harvest prior to their extirpation. The second hypothesis of possible Native American translocation of the four turtle taxa is not as parsimonious a suggestion as the first. Within this scenario, the low proportion of these turtles in the TP 4 assemblage may reflect that these individuals were brought in limited numbers to the Clupper site and did not naturally occur in the Keys.

Considering the presence of Key deer in TP 4 is similar to that of the non-marine turtles. Upper Matecumbe Key is well north of the purported historical range for the subspecies (Key West to Vaca Key) as well as the contemporary northern boundary at Big Pine Key. The survival of contemporary Key deer populations is directly tied to the accessibility of freshwater sources, hardwood hammocks, and mangrove habitats. Key deer are well known to swim between keys to maintain access to freshwater sources and food between dry and rainy seasons. If Upper Matecumbe Key had more readily accessible freshwater sources and variable hardwood hammock and mangrove forest plants (e.g., red and black mangroves, gumbo limbo, and pigeon plum) during the Glades period compared to the present, the island may have supported a local Key deer population. Within this scenario, the low proportion of deer in the TP 4 assemblage would suggest deer were not heavily targeted for exploitation, or the population was possibly in decline or extirpated during the timeframe represented in TP 4. It is worth noting that white-tail deer are also not prominent in coastal assemblages along the southwest coast of Florida despite ample availability. The fact that the majority of the few deer specimens at the site are modified suggests they were hunted in part for the manufacture of implements, but this does not shed light on the geographic origin of the represented individual or the possibility of a more expansive historical range for Key deer. It is also possible the deer specimens from Clupper represent an individual that was hunted in the Lower Keys and its bones were later transported to Clupper as a raw material for implement manufacture.

The heron and raccoon specimens from TP 4 are particularly interesting from the perspective of genetic diversity and understanding of the role of isolation in the evolution of island subspecies. As with the non-marine aquatic turtles and Key deer at the site, neither heron nor raccoon are abundant and do not appear to have been a focus of hunting. Relative to marine taxa, both birds and racoons are generally found in low numbers within faunal assemblages from south Florida archaeological sites and likely were not an important food resource for Native American people living at Clupper. While it is possible the specimens represent translocated animals, this is probably not the most parsimonious explanation at this point. Interestingly, the archaeological presence of these subspecies presents an opportunity to leverage ancient genetic analyses to extend the temporal, biogeographic, and potentially taxonomic breadth of current debates surrounding heron and raccoon species’ respective diversity and phylogeny in the Keys and greater North America (e.g., [85,88]).

Overall, it is possible that the low abundance of turtles, Key deer, Key raccoon, and heron in the TP 4 assemblage may represent depressed populations or translocated individuals. However, when considered within the context of broader trends of terrestrial and non-marine aquatic vertebrate hunting across southern peninsular Florida, none of these taxa are commonly abundant at coastal Native American sites during the Glades period nor is there evidence of translocation. Their scant abundance in TP 4 suggests that they were not important hunting targets at Clupper, and possibly surrounding Keys. Their presence may also indicate a greater overall taxonomic diversity and broader natural distribution of these animals than previously documented in the Keys. During Glades II, this would have likely been attributable to more sustainable terrestrial landscape use and lower anthropogenic impacts on freshwater, swamp, and brackish mangrove forest habitats compared to the present (e.g., [44]).

6. Conclusions: Supporting Biodiversity Conservation and Sustainability in the Florida Keys

Our zooarchaeological results contribute new insights to the understanding of Florida Keys’ biodiversity, historical ecology, and sustainability. Native Americans who inhabited the Clupper site ca. AD 800–1250 harvested a range of terrestrial and non-marine aquatic fauna, but consistently relied on a predominant suite of marine fishes and sea turtles. It is important to remember, however, that the TP 4 vertebrate assemblage reflects a sample that is an artifact of vertebrate animal availability coupled with Native American cultural practices of animal selection, harvesting, processing, and deposition, as well as post-depositional site formation processes. Nonetheless, these qualities enable us to use the assemblage to create an initial biocultural baseline of Glades II vertebrate diversity and Native American subsistence at the site.

Through the zooarchaeological record, we have established a far deeper temporal baseline for Keys vertebrate taxonomic diversity and history of human engagement. We have, in effect, “shifted” our conception of historical biocultural baselines for some of the Keys’ most culturally and ecologically important animals and associated habitats from approximately 200 years ago to 1000 years earlier. We now know that several vertebrate taxa from populations currently considered to be significantly depleted (e.g., snappers, groupers, jacks), not native to the region (e.g., Florida softshell turtle, snapping turtle), and/or locally extirpated (e.g., Key deer, Key raccoon) lived within ecosystems that were enmeshed with Native American lifeways and human–environment engagement for centuries prior to the European invasion of Florida. This shift is an important archaeological contribution to Keys biodiversity conservation and sustainability efforts because the zooarchaeological data provide unequivocal evidence of taxa that experienced anthropogenic pressures not otherwise documented in shallower historical sources (e.g., photographs, ethnohistoric accounts). The taxonomic diversity, equitability, and variance recorded in the TP 4 assemblage suggest that several of the region’s most vulnerable or endangered taxa, including endemic species, have been under anthropogenic pressures for at least a millennium, if not longer. Although early in our zooarchaeological research in the Keys, the data also suggest the outcomes of Native American engagement with vertebrate taxa during Glades II were likely more sustainable than recent historical and contemporary human–environment interactions that have contributed to various conservation issues.

As recently argued by Bjorndal [110], and long championed by others (e.g., [63,111]), shifting baselines can be a significant threat to biodiversity conservation where historical benchmarks for diversity in the past provide a comparative perspective for identifying successful conservation efforts and sustainable outcomes in the present. Temporal shifts in such baselines can reset or contradict previously established baselines and measures of success [112]. The threat of “shifting” baselines is particularly salient within the context of identifying historical ecological benchmarks from which to interpret and mitigate contemporary anthropogenic impacts on biodiversity and socioecological sustainability. For the Keys, a region with very few intact archaeological sites remaining, near-total economic reliance on marine and terrestrial resources for the regional economy, and cascading losses of land, biodiversity, and human lifeways, the continued study of Native American biocultural legacies is imperative for producing temporally deep and taxonomically robust biocultural baselines of vertebrate diversity and human subsistence.

To this end, in support of federal FKNMS, municipal Monroe County, and local conservation efforts (e.g., [113]), a central goal of the MCP is to integrate and make available our zooarchaeological results for applications in interdisciplinary conservation research as well as outreach efforts. For example, determining whether more robust populations of snappers, groupers, and jacks were available for harvest between 800 and 1250 years ago is important from a long-term perspective of fisheries diversity. Ongoing research is focused on refining our chronological understanding of how long these taxa have been under anthropogenic pressure, how they were culturally structured (e.g., subsistence regimes through time), the resulting ecological outcomes, and thresholds of ecological depression or recovery (e.g., [67,114,115]). Continued work focused on sea turtle historical ecology in the Keys and greater south Florida includes employing biochemical analysis of zooarchaeological sea turtle specimens (e.g., ZooMS) to render more taxonomically refined identifications (e.g., species) and baselines that are not typically possible based on morphological analysis alone (e.g., [81]). This will allow us to test whether the presumed abundance of green turtles at Clupper is accurate or a product of being able to morphologically identify more green turtle bone specimens compared to other species possibly present at the site. There is also great potential to contribute to conservation genomic studies of endemic terrestrial taxa, especially questions surrounding Key deer genetic diversity and connections to predicting population loss (e.g., [116]).

We are also engaged in discussions with municipal and community leaders, conservation practitioners, and educators involved in Keys sustainability to develop how to best leverage archaeology and our research as tools for public outreach and informal education opportunities. Together we are focused on communicating conservation and sustainability goals, including future participation in sustainability community forums, museum and not-for-profit open lectures, and visiting scientist K-12 classroom programs. We assert that sharing our research is essential to effectively mobilize archaeology as a sustainability science and we support the efficacy and use of locally focused approaches to biodiversity monitoring, education, and management within the context of sustainability practice (e.g., [117,118]).

In closing, it is clear that our world is on the precipice of an irreversible biodiversity crash that will directly impact the trajectory of sustainability for all humankind. To that end, there can be no doubt that the conservation of biodiversity and the sustainability of human lifeways are intimately linked. How we recognize this link and mobilize scientific practice in support and protection of it remains a challenge [119]. As the proverbial “poster child” for the effects of the global sustainability crisis in the U.S., and the impacts on island and coastal communities in particular, the tie between biodiversity loss and loss of human lifeways in the Keys has now taken center stage in regional initiatives linking biodiversity conservation and sustainability policy (e.g., [58,71]). Although the significance of historical ecology and long-term historical baselines are increasingly recognized as crucial to conservation and sustainability efforts in the Keys, the potential of archaeology to contribute to and bolster these perspectives has yet to be fully recognized. This is in large part due to a dearth of academic archaeological research over the past 50 years as well as a lack of published or easily accessible archaeological data from the region. Recent archaeological research in the Keys, however, including the MCP and zooarchaeological data presented here, is bringing renewed attention to the long-term Native American history underlying this ecologically unique and culturally rich archipelago. With climate change, rising sea levels, and marine erosion posing existential threats to coastal archaeological sites around the world [120,121], additional work on Keys archaeological sites is of utmost importance.