Evan Peacock

Assessing Bias in Archaeological Shell Assemblages

Abstract It is frequently assumed that assemblages of shell from archaeological sites suffer from various types of bias that might affect interpretations of the data. Such biases potentially include both pre- and post-depositional alteration of the shell; prehistoric practices that led to nonrandom collecting; sampling and recovery methods; and differing analytic capabilities. Species representation and metric data show that the effects of bias on three prehistoric freshwater mussel shell assemblages from the SE United States are minimal.

Distinguishing between Artifacts and Geofacts: A Test Case from Eastern England

AbstractDistinguishing between naturally and culturally-produced flaked stone has been a recognized problem in archaeology since the 19th century. A group of possible Paleolithic flakes from Kirmington, England, is examined using general and local flake characteristics. The general characteristics are derived by comparing known archaeological material to known natural material; the local characteristics are derived by comparing the possible artifacts from Kirmington to a natural control sample from the same site. The presence of features deemed significant by these comparisons contributes to a composite score for each object examined; artifacts score higher than natural flakes. The results indicate that some of the flakes in question are artifacts.

Accelerator Mass Spectrometry Radiocarbon Dating of Temper in Shell-Tempered Ceramics: Test Cases from Mississippi, Southeastern United States

Direct dating of artifacts is preferable to dating via association in most archaeological situations. In this article we describe and discuss the first application of accelerator mass spectrometry (AMS) to directly date temper particles in shell-tempered pottery. This method can be applied to ceramics from around the world that are tempered with freshwater or marine shell. To assess the method, we examine 34 new or previously reported dates (AMS dates on shell temper, paired luminescence dates from the same ceramics, and associated radiocarbon dates on other biotic materials) from seven sites in a variety of geological settings in the state of Mississippi. The results show that a freshwater reservoir effect is present in most cases. Other potential complicating factors include variable carbon input and metabolic uptake rates over space and time and the effects of burning during temper preparation or pottery firing. Despite these problems, the method holds promise for dating surface materials, museum specimens, and burial accompaniments; for investigating settlement patterns; and for other areas of research.

The Distribution and Research Value of Archaeological Mussel Shell: An Overview from Mississippi

Abstract Archaeological mussel shell rarely has been employed beyond paleodietary analysis in eastern North America, and most published work is site-specific in nature. Broad-scale syntheses of data on freshwater mussel shells are needed to realize the research potential of this common archaeological material. Such syntheses should also reveal gaps in our knowledge and perhaps illuminate reasons for why such gaps exist. In this paper, data on freshwater mussel shells from sites in Mississippi are gathered, and broad patterns in temporal and spatial distributions as well as site structure are discussed. The spatial data clearly suffer from survey, recovery, analytical and reporting biases, while data on temporal trends and site structure are lacking or are available at only gross levels. Given that shell data can be used in a number of research areas (e.g., biogeography, paleoenvironment, chemical sourcing, measurement of environmental disturbance), such unsystematic treatment of shell remains is unwarranted and should be corrected.

A Prehistoric Freshwater Mussel Assemblage from the Big Sunflower River, Sunflower County, Mississippi

Abstract Zooarchaeological deposits present a unique insight into the trajectory of faunal communities through time. Freshwater mussels (Bivalvia: Unionidae) are one of the worlds most imperiled faunal groups, and, at the same time, constitute a significant component of the archaeological record in North America. Current conservation efforts can be informed by studies of prehistoric mussel assemblages that catalog communities as they existed prior to any extensive modern impacts, ultimately representing an ecological baseline against which current populations can be evaluated. Over 47,000 freshwater mussel valves were recovered from the late prehistoric (ca. A.D. 700–1200) Kinlock site (22SU526), in Sunflower County, MS. Analyses revealed that the Sunflower River once supported a mussel community with greater taxonomic richness and more evenness among all the major species than today. Additionally, the presence in the archaeological record of numerous species which are currently considered rare, endangered, or extinct in the Sunflower River is indicative of a habitat disturbed by increased human environmental impact in historic times.

PREHISTORIC FRESHWATER MUSSEL FAUNAS FROM BAYOU BARTHOLOMEW, SOUTHEASTERN ARKANSAS

Abstract We present data on two prehistoric freshwater mussel (Unionidae) shell assemblages from Bayou Bartholomew, a free-flowing stream in Drew County, southeast Arkansas. The Taylor (3DR2) and Tillar Farms (3DR30) sites provided large shell samples from Baytown (seventh century A.D.) and Mississippian (fifteenth century A.D.) contexts, respectively. Comparison between prehistoric and modern assemblages shows some negative effects on fauna from increased turbidity, a phenomenon that may have begun in prehistoric times. In general, however, despite decades of intensive agriculture and other landscape modifications in the area, historical impacts on molluscan faunas in the Bayou have been unusually minor as compared to other waterways in eastern North America. These data emphasize the value of zooarchaeological remains as they pertain to modern landmanagement issues.

Assessing the Pre-modern Tree Cover of the Ackerman Unit, Tombigbee National Forest, North Central Hills, MS, Using GLO Survey Notes and Archaeological Data

Abstract General Land Office (GLO) survey data from the Ackerman Unit of the Tombigbee National Forest, MS are used to characterize early 19th-century tree cover in a part of the North Central Hills physiographic province. Archaeological settlement-pattern data indicate that the area was abandoned ca. A.D. 1000 and that early Historic-period settlement was minimal by the time the GLO surveys were done. The GLO data therefore represent forest conditions as they developed in the absence or near-absence of human influence. Tree cover consisted of oak-dominated hardwoods with a non-clustered pine component. The distributions of some hardwood species were related to geological controls. Based on previous archaeological work, the argument concerning minimal human impact can be extended to the entire North Central Hills province, with consequent implications for forest management on federal lands.

Confirmation of an Archaic Period Mound in Southwest Mississippi

Abstract Site 22LI504 is a predominantly Archaic period site in Lincoln County, Mississippi. One of its primary elements of interest is a single conical mound from which small-diameter cores revealed evidence of advanced pedogenesis. A radiocarbon sample from one soil core produced a date suggesting that the mound was Archaic in age (Fulmer 2001); however, it was unclear whether the sample came from within the mound or an underlying midden. In the spring of 2006, we excavated a 1-x-1-m unit in the mound to investigate this question. Diagnostic lithic artifacts, an advanced state of soil horizonation, and a lack of ceramics indicate that the mound is of Archaic period construction, with as many as five construction stages and artifact-rich features. We describe the soil profiles, features, and artifacts recovered from the mound, with comparisons to excavation units in an adjacent Archaic midden to show that there is no clear evidence for the mound being a functionally specific locus. We also present radiocarbon dates that indicate the earthwork is over 5,000 years old. These results are evaluated within the broader context of Archaic mound building, focusing in particular on environmental parameters underlying bet-hedging behavior.

Pre-Columbian Freshwater Mussel Assemblages from the Tallahatchie River in the Lower Mississippi River Alluvial Basin, U.S.A.

Abstract: The Tallahatchie River, Mississippi, southeastern U.S.A., is a medium-sized waterway originating in the North Central Hills and flowing into the Mississippi Alluvial Plain physiographic province. Although the rivers molluscan fauna remains relatively poorly studied, a survey by Haag and Warren (2007) of Lower Lake, an impounded, regulated segment of the Little Tallahatchie River high in the drainage, revealed a freshwater mussel assemblage that was surprisingly diverse and healthy given current stream management practices. Mussel assemblages from three prehistoric sites in Leflore County, Mississippi, further downstream on the main Tallahatchie River, yielded 32,303 valves representing 41 taxa, including 24 new river records, one of which, Quadrula fragosa (Conrad, 1835), represents a new state record and a notable range extension for this species. Comparison with modern data shows that about twice as many species existed in the waterway prior to Historic-era impacts, including five currently listed as threatened or endangered. These results highlight the importance of applied zooarchaeology for establishing conservation baselines.

DETERMINING THE SEASON OF MISSISSIPPIAN HOUSE CONSTRUCTION FROM PLANT IMPRESSIONS IN DAUB

Abstract It has been suggested that Mississippian farmsteads in some parts of the Southeast were abandoned in the late fall/early winter, with concomitant population concentration in large, palisaded villages. This assertion is contradicted by published data on plant impressions in daub which indicate farmstead house construction during the fall. Missing from this debate has been information on the season of house construction at mound centers. Plant impressions in daub from a structure at the Lyon’s Bluff site (22OK520) in Mississippi indicate that house construction took place during early to midspring. When coupled with other data, year-round construction at this mound and village complex is suggested. Although these results do not solve the question of the season of house construction at farmsteads, they do suggest that house construction at mound sites was not confined to any particular season of the year, as might be expected with seasonal influxes of farmstead inhabitants.