In lieu of an abstract, here is a brief excerpt of the content:

INTRODUCTION The archaeobotanical analyses of the Bottle Creek materials were designed to investigate how the people who lived at the site procured and produced plant foods. Speci¤cally, our goal was to collect data that could be used to determine what plant resources the inhabitants chose to use, to assess the relative importance of cultivated and wild plants in their diets, and to compare the subsistence practices of the Bottle Creek residents to those of people who were part of contemporaneous Mississippian societies of the interior Southeast, such as Moundville. The following pages describe the results of my analyses of the plant data from the 1993 and 1994 excavations at Bottle Creek. The plant data collected from test excavations in mounds A, B, C, D, and L provide a picture of late prehistoric plant use at Bottle Creek and yield information relevant to the issues noted above. FIELD AND LABORATORY METHODS Because we wanted to recover ¤ne-grained subsistence data, we collected®otation samples from both midden and feature contexts encountered during the excavations in the various mounds. Generally the samples were 10 liters in volume, although some were smaller or larger depending on the nature of the deposit. We took the soil samples to the laboratory at the University of South Alabama where we ®oated them to separate the plant remains from other materials. The system we used to process the samples was a ®otation machine manufactured by R. J. Dausman Technical Services . (We are indebted to Dr. Greg Waselkov for the loan of this machine, which made processing the ®otation samples a comparatively ef¤cient task.) The light fractions from the ®otation samples were caught on ¤nemesh nylon cloth. The heavy fractions were caught on a 1 mm mesh screen. 5 / Food Plant Remains from Excavations in Mounds A, B, C, D, and L at Bottle Creek C. Margaret Scarry Since our goal was to collect evidence about subsistence practices, we selected for analysis samples from midden and feature contexts that seemed most likely to contain food plant remains. In all, I analyzed 36 samples from ¤ve mounds, representing 31 different contexts. Table 5.1 provides basic information about the quantities of wood and plant material (wood plus food plant remains) in each sample. The samples that were ®oated but not analyzed for this report are held in reserve. The methods I used to sort and identify the plant remains were the same for all samples except Mound L Feature 172 (see below). I weighed the samples and then sieved them through geological screens (2 mm, 1.4 mm, and .7 mm) to make sorting easier. I inspected each size fraction, including the material that passed through the .7 mm screen, with the aid of a binocular stereoscopic microscope (10–20x magni¤cation). I sorted all plant fragments greater than 1.4 mm in size and scanned the plant materials in 104 / C. Margaret Scarry the fractions smaller than 1.4 mm. I removed and identi¤ed seeds or seed fragments found in these small fractions, but did not otherwise sort the materials. My sorting procedures differ from those generally used on plant remains from the eastern United States in one way. It is standard practice to use 2 mm as the dividing point between complete sorting and scanning. In my analyses of remains from the Moundville system in the Black Warrior Valley of Alabama (Scarry 1986:186–188), however, I used 1.4 mm as the dividing point between sorting and scanning. I chose to sort to a ¤ner level because experiments by Neal Lopinot (1984:111–112) had demonstrated that acorn was underrepresented when sorting stopped at the 2 mm level. For this study of the Bottle Creek remains, I chose to sort remains greater than 1.4 mm in size. This makes the data from Bottle Creek directly comparable to those from sites in the Moundville system and insures that acorns are not underrepresented in the analysis. Several of the light fractions were quite large. For these samples, I used a rif®e splitter to halve or quarter the materials. Once the subsample was sorted, I used the quantities of plants identi¤ed in the subsample to estimate the quantities in the sample as a whole. Seeds and other non-wood plant parts were identi¤ed to the lowest possible taxonomic level. Size, shape, and surface texture were the primary characteristics I used to classify the seeds. I made initial identi¤cations by reference to...

Share