Clovis site distributions
By Mark Muniz
The goal of this paper is to further attempts at modeling Paleoindian subsistence/settlement strategies by first reconstructing paleohabitats based on fossil pollen evidence, then evaluating mammal diversity for those reconstructed areas, and finally plotting site distributions in relation to those habitats. Due to space limitations, only Clovis site concentrations will be addressed here, although data for Paleoindian and Early Archaic periods of Florida have been presented elsewhere in a 1998 paper by Louis Tesar and me.
Paleoenvironmental reconstruction is based on published pollen studies from Lake Annie, Lake Tulane, Sheelar Lake, Lake Louise, and Camel Lake. To estimate species diversity, 48 mammalian species were chosen from those occurring in the Late Pleistocene of Florida and matched to habitats that met the dietary requirements and proposed niches of each animal. The only mammals not included in the distribution are mice, voles, moles, and others of comparable size.
Because so much of Florida has been inundated by sea level rise over the past 12,000 years, consideration of Paleoindian subsistence strategies clearly needs to take into account resources then available in the now inundated region of the continental shelf, which would have then been coastal plain. Habitats that combined both terrestrial, freshwater, and saltwater resources (such as estuaries) offered a greater diversity of resources than upland habitats 11,000 years ago, as they do today, and were surely utilized by Paleoindian peoples. Unfortunately, we will have to wait until pollen data from inundated regions of the Gulf of Mexico and Atlantic Ocean have been collected before a total picture of the habitats available for Paleoindian utilization can be completed.
For the period 12,000 to 11,000 RCYBP the habitat surrounding Camel Lake (located about 60 Km west of Tallahassee) is characterized by mesic deciduous forest with many mast producing oak trees and abundant moisture. To the northeast, the Lake Louise habitat (located about 20 Km south of Valdosta) was much drier, with far fewer floral species, exhibiting prairie development and dominated primarily by oak. The region around Sheelar Lake (located just outside of Gainesville) was characterized by a mesic moisture regime with broad leaf trees declining and pine increasing. Oak and herbs were also present, and there was some prairie development as well. The south-central Peninsula, represented by Lake Annie and Lake Tulane, was a much drier scrub and prairie habitat with some sclerophyllous oak woodlands on the upland regions. This area also contained composites, grasses, and some pine.
Prior to the extinction event of the terminal Pleistocene, the distribution of the 48 mammals chosen for this study ranged from a high of 34 species in the Panhandle to a low of 22 species in the northern Peninsula and 27 species in the south-central Peninsula. In every habitat, species ranged in size from the Gray squirrel to the Columbian mammoth. During this period, sea level rose at least 15 meters.
The heaviest concentrations of Clovis sites occur along the borders between differing ecozones and drainage basins, and range from 75 Km to 250 Km apart (Figure 1). Many Clovis sites once located along natural levees and terraces of coastward river valleys are today inundated. As a result, this distribution may only illustrate a part of the subsistence/settlement strategy practiced by these people. Nevertheless, it seems that Clovis people in Florida did not focus their sites in a single kind of environment, but instead concentrated on areas where the widest diversity of floral and faunal resources could be obtained. For example, exploiting the region between Camel Lake and Lake Louise provides access to 92% of the mammalian species chosen for this study, versus 71% in the Camel Lake region and 46% in the Lake Louise region. There was undoubtedly increased diversity of floral resources along these border lands as well.
The southern Peninsula, which most closely resembled habitats found in the Southwestern U.S., with large grasslands, prairie, and scrub oak woodlands, appears completely uninhabited. This is unusual given the presence of faunal resources such as mammoth, mastodon, bison, deer, and sloth that were present (representing 56% of the 48 mammalian species), and suggests a subsistence/settlement strategy for Florida Clovis peoples different from that employed by their Northeastern counterparts.
The very weak presence along the Atlantic Coast emphasizes the Clovis focus on the Gulf of Mexico. Of the six drainage basins bordering the Gulf, Clovis sites are concentrated in the central four. Site concentrations along the borders of drainage basins once again indicate a focus on areas that allowed access to greater diversity of faunal and floral resources. Finally, distinct concentrations in both the Panhandle and near Tampa Bay may suggest the territories of at least two macrobands.
Following the arguments of David Meltzer and others, hunter-gatherer settlement in regions with high species diversity and accompanying low populations of individual species, should result in a generalist, rather than a specialist, subsistence strategy. In addition, quarry and/or tool production sites should be the only site types that are large and repeatedly visited by generalists. The location of Florida Clovis sites along transitions between different habitats supports a generalist subsistence strategy. Meltzerís generalist hypothesis is also supported in Florida by the many isolated finds characterizing the Clovis presence along karst river valleys, with very few base camp sites or sites exhibiting repeated occupations that often characterize a specialist subsistence strategy as employed, for example, by Paleoindians in the Northeastern U.S.
Hopefully this brief discussion sheds light on Clovis subsistence strategies in Florida. Due to the absence of faunal remains from Clovis sites, we are forced to rely on theoretical formulations and inference to derive our conclusions. However, with more pollen data (especially from the Gulf of Mexico) and the continued search for intact, stratified Clovis sites in Florida, we will eventually be able to flesh out the model presented here.
Editor's Note: This paper was presented at the 31st Chacmool Conference, on Novermber 14, 1998, in Calgary, Alberta, Canada. It is excerpted from a larger paper entitled "Peopling of the peninsula".