Bioarchaeological analysis of diet during the Coles Creek period in the southern Lower Mississippi Valley.код для вставкиСкачать
AMERICAN JOURNAL OF PHYSICAL ANTHROPOLOGY 144:30–40 (2011) Bioarchaeological Analysis of Diet During the Coles Creek Period in the Southern Lower Mississippi Valley Ginesse A. Listi* Department of Geography and Anthropology, Louisiana State University, Baton Rouge, LA KEY WORDS dental pathologies; Coles Creek diet; dietary transition ABSTRACT The timing of the dietary shift from foraging to maize agriculture, and the speed at which such practices were adopted, are important considerations in the cultural evolution of the New World. In the southern Lower Mississippi Valley, maize agriculture traditionally was believed to have been practiced during the Coles Creek period (A.D. 700–1200); however, direct evidence for maize is lacking in the archaeological record prior to A.D. 1000. The present study examines Coles Creek diet from a bioarchaeological perspective. Oral-health indicators, including abscesses, antemortem tooth loss, calculus, carious lesions, periodontal disease, and tooth wear, were evaluated in a regional, temporal context. Data were collected from 288 dentitions from eight sites in the southern Lower Mississippi Valley that range in date from 800 B.C. to A.D. 1200. The sample then was sepa- rated into Pre-Coles Creek and Coles Creek categories and statistical analyses were used to assess temporal variation in pathology load. Results indicate that pathology load in the Coles Creek sample is slightly heavier than the Pre-Coles Creek sample; however, the differences are not substantial. Furthermore, data suggest that regional differences in resource exploitation existed between the Lower Mississippi Valley and populations elsewhere in the eastern United States. Speciﬁcally, the presence of starchy native plants other than maize in the diet likely contributed to a high pathology load for early hunter-gatherers. Ultimately, data from this study complement the archaeological, botanical, and zooarchaeological records and indicate that Coles Creek subsistence was not based on maize agriculture. Am J Phys Anthropol 144:30–40, 2011. V 2010 Wiley-Liss, Inc. Archaeologists have long recognized the relationship between diet, population size, and social complexity (e.g., Schoeninger and Shurr, 1994). In the Lower Mississippi Valley, complex societies, as indicated by increasing population density, regional mound centers, and the emergence of hierarchically ranked societies are ﬁrst recognized during the Late Woodland Coles Creek period (A.D. 700–1200; Kidder, 1992a, 2002). Based on the number and widespread distribution of sites and the assumed correlation between social complexity and agriculture, archaeologists originally believed that Coles Creek diet was based on maize agriculture (Byrd and Neuman, 1978; Haag, 1978; Williams and Brain, 1983). However, little direct evidence for maize has been found in the Lower Mississippi Valley prior to A.D. 1000 (Fritz and Kidder, 1993; Fritz, 1995; Kidder, 2002). The present study investigates Coles Creek diet from a bioarchaeological perspective within the context of the dietary transition in the southern Lower Mississippi Valley by examining the frequency and interaction of multiple oral health indicators (including dental caries, calculus, periodontal disease, antemortem tooth loss, abscesses, and tooth wear). Two hundred eighty-eight adult dentitions were examined from eight sites that range in date from 800 B.C. to A.D. 1200. The sample then was divided into Pre-Coles Creek and Coles Creek categories and statistical analyses were used to assess variation in pathology load between the two groups. As numerous bioarchaeological studies have demonstrated a relationship between high carbohydrate diets and a decline in oral health (Turner, 1979; Larsen, 1984, 1995, 1997; Norr, 1984; Perzigian et al., 1984; Smith, 1984; Berry, 1985; Powell, 1985; Schmucker, 1985; Lukacs, 1989; Larsen et al., 1991), the hypothesis underlying this research is that a change in diet during the Coles Creek period from hunting-gathering to maize agriculture would be evident in the pathology load. Speciﬁcally, when compared to the Pre-Coles Creek sample, the frequency of all pathological conditions would be higher in the Coles Creek sample, with the exception of tooth wear, which would be lower. C 2010 V WILEY-LISS, INC. C COLES CREEK CULTURE AND DIET The Coles Creek period in the Lower Mississippi River alluvial valley is delineated by radiocarbon dates from approximately A.D. 700 to 1200; thus, it straddles the divide between the Late Woodland and Early Mississippian periods as deﬁned for the southeastern United States as a whole (Jeter and Williams, 1989; Kidder, 1992a). The Coles Creek period often is described as a dynamic time during which signiﬁcant changes occurred throughout the Lower Mississippi Valley. These changes included variation in ceramic styles as well as increases in population size and site size and complexity (Kidder, 1992b, 2002). Along with these changes, archaeologists originally supposed that subsistence practices changed as well, namely, the dietary transition from huntinggathering to maize agriculture (Byrd and Neuman, 1978; *Correspondence to: Ginesse Listi, LSU Geography and Anthropology, 227 Howe-Russell Building, Baton Rouge, LA 70803. E-mail: email@example.com Received 13 January 2010; accepted 26 May 2010 DOI 10.1002/ajpa.21364 Published online 18 August 2010 in Wiley Online Library (wileyonlinelibrary.com). BIOARCHAEOLOGICAL ANALYSIS OF COLES CREEK DIET Haag, 1978; Williams and Brain, 1983; Neuman, 1984). The evidence for this supposition was indirect: the locations of Coles Creek sites along the fertile alluvial bottomlands beside streams and rivers would have been ideal for plant cultivation; the relative predictability and high product yield of an agricultural subsistence base not only would have supported the increase in population size that occurred during this time, but also would have been necessary to provide sustenance for the large labor force required to build the extensive earthworks that are characteristic of Coles Creek sites. While such indirect evidence is no longer accepted as a basis for inferring subsistence, this logic was used as recently as the 1980s (Williams and Brain, 1983). Meanwhile, archaeological and botanical data began to accumulate which suggested that a dietary transition had not occurred. Botanical evidence from the interior Lower Valley indicates that maize was present as early as the eighth or ninth centuries A.D. (Scarry, 1993); however, its incidence in the archaeological record is rare. For example, at one Coles Creek site in which maize has been found, the overall low frequencies recovered and the contexts in which it was found (e.g., middens, specialized features with tobacco) led researchers to conclude that maize was a relatively minor food source and may have been tied to ritual behavior rather than subsistence (Fritz and Kidder, 1993; Kidder and Fritz, 1993). Conversely, substantial data exist which demonstrate that Coles Creek populations exploited a variety of other botanical resources, including nuts, fruits, and wild, or possibly, cultivated plants (Kidder and Fritz, 1993; Fritz, 1995; Roberts, 2000; Kidder, 2002). With regard to material culture, not only are Coles Creek pottery assemblages indistinct from other Late Woodland cultures, but also artifacts suggestive of plant cultivation (e.g., hoes) are not abundant at Coles Creek sites (Kidder, 1992a). On the other hand, artifacts related to the procurement (including the bow and arrow) or processing of food are ubiquitous throughout the Lower Mississippi Valley and provide ample support that Coles Creek populations exploited natural resources through hunting, gathering, and ﬁshing (Williams and Brain, 1983; Neuman, 1984; Kidder, 1992b; Fritz, 1995; Jackson and Scott, 2002). Correspondingly, zooarchaeological evidence from Coles Creek sites consistently attests to the importance of wild game and/or aquatic resources in the diet (Springer, 1980; Mariaca, 1988; Jeter and Williams, 1989; Colburn and Styles, 1990; Kelley, 1992; Smith, 1996; Jackson and Scott, 2002; Kidder, 2002). Bioarchaeologically, data from the Lower Mississippi Valley, including the Coles Creek period, generally are sparse: human remains typically are fragmented and poorly preserved, thus, sample sizes are small, individuals are often incomplete and commingled, and few pathological conditions have been assessed (Jeter et al., 1989; Rose and Harmon, 1999). Most of the previously available bioarchaeological data relating to the Coles Creek period, compiled and presented in two noteworthy publications (Jeter et al., 1989; Rose and Harmon, 1999), consisted of demographic proﬁles or the analysis of one or two pathological conditions appended to site reports, monographs, or unpublished manuscripts. The few osteological studies that have been undertaken report that Coles Creek skeletons exhibit a pathology load that is different from earlier groups, particularly for tooth wear 31 and carious lesions (Rose et al., 1984, 1991). These researchers conclude that the change in pathology load was due to an increase in carbohydrates and grit in the diet. However, since direct evidence of maize is virtually absent in the archaeological record, they suggest the high carbohydrate content came instead from native starchy plants or seeds (Rose et al., 1984, 1991). Lastly, unpublished stable isotope data, though rare, are available for several sites in the southern Lower Mississippi Valley and indicate that maize consumption was minimal or absent at these sites (Bruce Smith, personal communication to author, 2003). Prehistoric diet is best understood using a multidisciplinary approach that combines botanical, zooarchaeological, and bioarchaeological evidence (Sobolik, 1994). While archaeological and botanical research in the southern Lower Mississippi Valley challenges the traditional view of Coles Creek subsistence as maize agriculture, bioarchaeological research that incorporates a wide range of osteological data has not been undertaken. The current study addresses this deﬁciency by using multiple oral health indicators to assess Coles Creek diet in a comprehensive, systematic way. The distinct pathology proﬁles exhibited in hunter-gatherers and agriculturalists provide contrasting expectations by which data can be interpreted (Lukacs, 1989). The broad temporal span, yet relatively narrow geographic focus, of the skeletal populations included in this study allow pathology load to be analyzed through a regional diachronic framework. MATERIALS A total of 288 adult dentitions were examined from eight sites in the southern Lower Mississippi Valley. Skeletal remains from the individual sites were excavated at different times by different people over a span of nearly 40 years (from the mid 1930s to the early 1970s). Currently, remains from individual sites are catalogued and curated in museum collections in compliance with NAGPRA. The temporal span for the sites extends from 800 B.C. to A.D. 1200. Based on these dates, all remains were classiﬁed into two broadlydeﬁned temporal categories that separated the Coles Creek sample from the Pre-Coles Creek sample. Figure 1 depicts the location of each site in the Lower Mississippi Valley. Table 1 provides a summary of the skeletal samples and MNI used in this study. The Pre-Coles Creek sample is comprised of 144 individuals from ﬁve sites. These mound and/or midden sites include Little Woods (16OR1-16OR5) and Tchefuncte (16ST1), located in coastal regions of Louisiana, and Lafayette Mounds (16SM17), Greenhouse (16AV22), and Mount Nebo (16MA18), which are located in interior regions (see Fig. 1). The majority of the Pre-Coles Creek sample dates to the Tchula period, Tchefuncte culture (800 B.C. to 100 B.C.); however, the burials from Greenhouse and Mount Nebo date later to the Baytown period, Troyville culture (approximately A.D. 400 to 700), (Ford and Quimby, 1945; Ford, 1951; Giardino, 1977, 1982; Lewis, 1991; Kidder, 2002). Pre-Coles Creek burial patterns include both primary and secondary, single and multiple interments. Many of the remains were ﬂexed and/or bundled, but some individuals were extended. Grave goods generally were not found; therefore, burials were dated based on site stratigraphy (Ford and Quimby, 1945; Neuman, 1984). American Journal of Physical Anthropology 32 G.A. LISTI Fig. 1. Approximate locations of sites used in current study. TABLE 1. Sites and minimum number of individuals (MNI) for Pre-Coles Creek and Coles Creek samples Pre-Coles Creek (800 B.C. to A.D. 700) Site N Greenhouse Lafayette Little Woods Mount Nebo Tchefuncte Total 16 31 45 11 41 144 Coles Creek (A.D. 700 to 1200) Site Greenhouse Lake George Lake St. Agnes Morton Shell Mound Mount Nebo N 30 25 3 75 11 144 The exception is Mount Nebo: artifacts, which might have been grave offerings, were found with several of the individuals and were used to obtain a relative date for the burials at that site (Giardino, 1982). Detailed subsistence data are not available for any of the Pre-Coles Creek populations used in this study. However, during this period in the Lower Mississippi Valley, deer generally is the most common mammal species present in site middens (Neuman, 1984). Various smaller mammals, reptiles, ﬁsh, and birds are present as well, and Rangia remains are prevalent at coastal sites American Journal of Physical Anthropology (Neuman, 1984; p 118). Thus, subsistence for the PreColes Creek populations is believed to have been a mix of hunting, gathering, and ﬁshing (Ford and Quimby, 1945; Neuman, 1984; Jeter and Williams, 1989). The Coles Creek sample is comprised of 144 individuals from ﬁve sites, four of which are located in interior regions (Greenhouse (16AV22), Lake George (22YZ557), Lake St. Agnes (16AV26), and Mount Nebo (16MA18)), one of which is coastal [Morton Shell Mound (16IB3); see Fig. 1]. All sites are multicomponent, but the burials included in this sample date to the Coles Creek period (A.D. 700–1200). All are mound sites and may have served ceremonial or ritual purposes; none is strictly habitational or residential. Therefore, individuals buried at these sites could represent ‘‘elite’’ or ‘‘special’’ members of the population and may not be representative of the entire population that inhabited the area or utilized the sites (Ford, 1951; Belmont, 1967; Robbins, 1976; Giardino, 1977, 1982; Toth, 1979; Williams and Brain, 1983; Neuman, 1984; McGimsey, 2003); (Saunders and Byers, unpublished). A common characteristic of Coles Creek burials is the apparent lack of order in which human remains were deposited (Ford, 1951; Williams and Brain, 1983). Burials include single and multiple interments, bundled, BIOARCHAEOLOGICAL ANALYSIS OF COLES CREEK DIET ﬂexed, or extended individuals, and isolated elements. Grave goods generally were not present; therefore, either stratigraphy or radiocarbon dates obtained from samples taken from the same stratigraphic layers were used to date the burials (Ford, 1951; Belmont, 1967; Giardino, 1977, 1982; Toth, 1979; Williams and Brain, 1983; Neuman, 1984; McGimsey, 2003). METHODS Skeletal remains from all sites were in various states of preservation and fragmentation, ranging from friable fragments held together by dirt matrix and preservative to well-preserved specimens. Some elements were heavily fragmented, while others were whole. Some individuals were only partially represented; others were nearly complete. MNIs reported in Table 1 were determined by the number of dentitions present. For single interments, a whole or partial dentition represented one individual. In the case of multiple interments, if a maxilla and mandible could be reliably associated together based on similar size and tooth morphology, they were counted as a single individual; otherwise, each element was considered separately. All individuals were ‘‘Adults’’ (greater than 18 years of age), based on eruption of the third molars. However, speciﬁc age ranges and sex were not able to be assigned for the majority of the individuals due to fragmentation and commingling. In single burials, while some individuals had an os coxae complete enough to suggest an age range or sex, the ossa coxarum of others were too fragmented to assess, or were not present at all. Age ranges and sex were not assessed in commingled remains due to the impossibility of reconciling postcranial elements with dentitions. 33 Data collection procedures Every tooth or tooth position was assessed as either present, lost antemortem, lost postmortem, or not able to be analyzed. Each tooth or alveoli subsequently was examined macroscopically for pathological conditions. A low magniﬁcation binocular microscope was used if the determination of the pathological condition was questionable. For the assessment of carious lesions, the presence and location of each carious lesion were recorded following guidelines in Buikstra and Ubelaker (1994). To differentiate between true carious lesions and natural noncarious pits and ﬁssures (i.e., buccal pits), an attempt was made to determine whether or not and the extent to which the dentin underlying the enamel was affected (Aufderheide and Rodriguez-Martin, 1998; Ortner, 2003). If a buccal pit was present but the extent of dentin damage could not be discerned, the lesion was noted, but not included in the analyses of carious lesions prevalence. Tooth wear was evaluated for the ﬁrst molars only. This tooth was selected because it was preserved more often per individual than second or third molars and, therefore, was the most numerous. Wear was analyzed based on Scott’s methodology (1979; see also Buikstra and Ubelaker, 1994): the occlusal surface was divided into quadrants, a score of 1–10 was assigned to each based on the amount of enamel present, and the quadrant scores were then totaled. If the crown was broken or portions were missing, the tooth was not scored. For recording dental calculus, a score was assigned to each tooth based on the amount of calculus present (based on Brothwell, 1981; see also Buikstra and Ubelaker, 1994). The ordinal categories (examples of which are pictured in Fig. 2) ranged from no calculus present, through small, moderate, and heavy amounts present. Fig. 2. Dentitions showing different levels of calculus: none (A), slight (B), moderate (C), and heavy (D). American Journal of Physical Anthropology 34 G.A. LISTI Fig. 3. Dentitions showing different levels of periodontal disease: none (A), less than one-half root absorption (B), greater than one-half root absorption (C), alveolar remnants (D, arrow), and complete alveolar resorption (D, posterior corpus). The assessment of periodontal disease was based on a scale that standardizes the progressive deterioration of the alveolar region for each tooth (Lukacs, 1989). The ordinal categories (examples of which are pictured in Fig. 3) ranged from no resorption, through minimal (in which less than one-half the tooth root is exposed), moderate (in which more than one-half the tooth root is exposed), heavy (where only remnants of the alveolus remain), and complete resorption of the alveolus. Antemortem tooth loss (AMTL), which also corresponds to the most advanced stage of periodontal disease, was assessed using the initial recording of the presence or absence of each tooth. Teeth lost prior to death were indicated by the complete resorption of the alveolus (depicted in Fig. 3D). For abscesses, Buikstra and Ubelaker (1994) recommend recording the presence or absence of periapical abscesses and their location (buccal or lingual). In this study, the same information was recorded for periodontal abscesses as well. Data on both types of abscesses were combined for analyses of temporal variation. Data analysis Several methods were used to analyze the pathology load in the two samples. Population frequencies were calculated using the tooth count and individual count methods for all pathological conditions except tooth wear (for which only the average per molar was calculated). With the tooth count method, the number of teeth affected with pathology is counted relative to the total number of teeth available for study (Lukacs, 1989). These frequencies are reported for each individual site as well as for the collective Coles Creek and Pre-Coles Creek samples. In the individual count method, the number of individuals with the particular pathology is counted and divided by the total number of individuals present in the American Journal of Physical Anthropology sample (Lukacs, 1989). This method is used to report summary statistics and the average number of carious lesions per individual for each individual site as well as for the collective Coles Creek and Pre-Coles Creek samples. However, for samples in which individual preservation is varied and many specimens are incomplete, the individual count method can result in skewed data. Thus, an alternative technique called the mean number of pathologies per specimen (MNPS) also was used to calculate pathology load (Moore and Corbett, 1971; Lukacs, 1989). In this method, the number of teeth or alveoli affected with the pathological condition is divided by the number of teeth or alveoli available for assessment per individual (Moore and Corbett, 1971). The MNPS values are reported for each individual site as well as for the collective Coles Creek and Pre-Coles Creek samples. Temporal variation in pathology load was tested statistically using an independent samples t test for the average tooth wear and MNPS for each pathological condition. A chi-square was used to assess variation in the severity of calculus and periodontal disease between Coles Creek and Pre-Coles Creek samples. For all analyses, the null hypothesis was that subsistence was the same for both Pre-Coles Creek and Coles Creek populations. That is, no signiﬁcant differences would exist between the two samples in the MNPS or average tooth wear. Signiﬁcant differences (if P \ 0.05), therefore, would be indicative of a change in diet. RESULTS Tables 2–5 list the individual and tooth counts for each site and for the collective Coles Creek and PreColes Creek samples. Based on individual counts, the frequency increased through time for carious lesions, abscesses, AMTL, and calculus, but decreased through time for periodontal disease. Based on tooth counts, the 35 BIOARCHAEOLOGICAL ANALYSIS OF COLES CREEK DIET TABLE 2. Frequencies of dental caries based on individual and tooth counts Individual count Tooth count Naff/Ntota % Avg. Naff/Ntota % 2/15 3/15 0/7 3/11 6/30 14/78 13.3 20.0 0 27.3 20.0 17.9 0.13 0.33 0 0.72 0.33 0.32 2/143 5/87 0/17 8/111 10/240 25/598 1.4 5.7 0 7.2 4.2 4.2 3/29 15/25 1/3 17/68 3/9 39/134 10.3 60.0 33.3 25.0 33.3 29.1 0.17 1.60 0.33 0.29 0.44 0.52 5/234 40/477 1/14 20/247 4/68 70/1,040 2.1 8.4 7.1 8.1 5.9 6.7 Pre-Coles Creek sites Greenhouse Lafayette Little Woods Mount Nebo Tchefuncte Total Coles Creek sites Greenhouse Lake George Lake St. Agnes Morton Shell Mound Mount Nebo Total a Naff 5 number of teeth or individuals affected with pathology; Ntot 5 total number of teeth or individuals available for assessment. TABLE 3. Frequencies of abscesses and antemortem tooth loss (AMTL) based on individual and tooth counts Abscesses Individual count Pre-Coles Creek sites Greenhouse Lafayette Little Woods Mount Nebo Tchefuncte Total Coles Creek sites Greenhouse Lake George Lake St. Agnes Morton Shell Mound Mount Nebo Total AMTL Tooth count Naff/Ntota % 5/14 4/24 0/18 4/11 12/29 25/96 35.7 16.7 0 36.4 41.4 26.0 7/164 7/178 0/71 12/95 25/193 51/701 11/29 11/25 0/3 14/65 2/10 38/132 37.9 44.0 0 21.5 20.0 28.8 31/374 19/422 0/20 29/382 4/67 83/1,265 Individual count Naff/Ntota Tooth count % Naff/Ntota % Naff/Ntota % 4.3 3.9 0 12.6 13.0 7.3 4/16 9/31 11/45 3/11 9/41 36/144 25.0 29.0 24.4 27.3 22.0 25.0 7/225 33/343 44/287 4/133 30/611 118/1,599 3.1 9.6 15.3 3.0 4.9 7.4 8.3 4.5 0 7.6 6.0 6.6 15/30 11/25 0/3 13/75 4/11 43/144 50.0 44.0 0 17.3 36.4 29.9 55/553 49/608 0/51 40/566 6/110 150/1,888 9.9 8.1 0 7.1 5.5 7.9 a Naff 5 number of teeth or individuals affected with pathology; Ntot 5 total number of teeth or individuals available for assessment. TABLE 4. Frequencies of dental calculus and periodontal disease based on individual and tooth counts Calculus Individual count Pre-Coles Creek sites Greenhouse Lafayette Little Woods Mount Nebo Tchefuncte Total Coles Creek sites Greenhouse Lake George Lake St. Agnes Morton Shell Mound Mount Nebo Total Periodontal disease Tooth count Naff/Ntota % Naff/Ntota 9/15 12/13 4/7 7/11 28/31 60/77 60.0 92.3 57.1 63.6 90.3 77.9 21/27 21/25 3/3 53/68 6/8 104/131 77.8 84.0 100.0 77.9 75.0 79.4 Individual count Tooth count % Naff/Ntota % Naff/Ntota 49/125 53/84 12/17 31/110 190/240 335/576 39.2 63.1 70.6 28.2 79.2 58.2 14/14 17/24 19/21 7/11 28/34 85/104 100.0 70.8 90.5 63.6 82.4 81.7 106/130 56/132 65/75 63/90 148/218 438/645 81.5 42.4 86.7 70.0 67.9 67.9 89/223 294/465 11/12 178/244 37/61 609/1,005 39.9 63.2 91.7 73.0 60.7 60.6 25/28 24/25 3/3 34/69 6/10 92/135 89.3 96.0 100.0 49.3 60.0 68.1 277/308 306/420 10/16 125/263 31/50 749/1,057 89.9 72.9 62.5 47.5 62.0 70.9 % a Naff 5 number of teeth or individuals affected with pathology; Ntot 5 total number of teeth or individuals available for assessment. American Journal of Physical Anthropology 36 G.A. LISTI frequency increased for carious lesions, AMTL, calculus, and periodontal disease, but decreased for abscesses and tooth wear. The decrease in tooth wear through time is signiﬁcant (Table 5). Table 6 displays the MNPS scores and the results of the independent samples t tests. Only periodontal disease (which decreases) shows signiﬁcant variation through time. Table 7 depicts the frequencies for each category of calculus and periodontal based on tooth and/or alveolar counts. For calculus, the distribution of data is similar for each sample. Also, though 58–61% of each sample has calculus, the majority of teeth have a ‘‘small’’ amount as opposed to ‘‘moderate’’ or ‘‘heavy’’ amounts. While the frequency of calculus increases slightly through time, the severity decreases. This temporal change is not signiﬁcant. For periodontal disease, as with calculus, the frequency increases through time; however, the severity decreases. This variation is signiﬁcant (P \ 0.01). DISCUSSION The present study explores Coles Creek subsistence from a bioarchaeological perspective within the context of the dietary transition in the southern Lower Mississippi Valley. As the transition occurred (and with it, an increase in the consumption of carbohydrates), the expectation was that all pathological conditions except tooth wear would increase from the Pre-Coles Creek to the Coles Creek period. Table 8 summarizes the temporal trends in data compared to expectations. Though results vary somewhat depending on methodology, data initially appear to support expectations. That is, 12 of 18 categories of data are consistent with expectations for an increase in carbohydrate consumption during the Coles Creek period. However, a straightforward interpretation TABLE 5. Mean wear scores for ﬁrst molars Pre-Coles Creek Sites (N 5 89, n 5 99) Greenhouse Lafayette Little Woods Mount Nebo Tchefuncte Average Coles Creek sites (N 5 132, n 5 156) Greenhouse Lake George Lake St. Agnes Morton Shell Mound Mount Nebo Average T-test: t 5 3.993 M1 28.4 26.4 20.6 33.3 30.0 29.1 M1 28.4 25.5 20.0 23.4 21.9 25.0 Sig. 5 0.000* ‘‘*’’ indicates that relationship is statistically signiﬁcant. of these data as indicative of dietary change is confounded by several factors which include sample biases and potential intersite and interregional differences in resource exploitation. The condition of the skeletal remains examined in this study is less-than-ideal; therefore, sample biases associated with preservation and demography (or lack thereof) could be impacting the results. The former biases are alleviated somewhat by using multiple methods to calculate population frequencies. Of the three methods used in this study, the individual and tooth counts, which are commonly reported in the literature, assume dentitions are complete. Only the MNPS statistic takes into account individual preservation. Nevertheless, obtaining consistent results from these different methodologies might indicate that the impact of differential preservation is minimal. As demonstrated in Table 8, temporal patterns are consistent only for two of the ﬁve pathological conditions assessed with multiple methodologies (e.g., carious lesions and calculus). Therefore, differential preservation likely is affecting the results for abscesses, AMTL, and periodontal disease. The lack of demographic data also is inhibiting data interpretation. Ideally, skeletal samples should be representative of the entire population, as opposed to a small, specialized subset (e.g., only adults, children, or elite members) though Wood et al. (1992, p 344) argue that no skeletal sample can completely represent the original living population, ‘‘even (one) that is a perfect random sample of all those who died.’’ Additionally, comparability in age and sex distribution among samples is important because of the inﬂuence of biological parameters (particularly age) on pathology load. Most pathological conditions are progressive; therefore, a population consisting primarily of older adults could exhibit higher frequencies of pathology than a population composed of younger adults, simply because they are older. In this study, because speciﬁc adult age categories could not be assigned for the majority of the individuals, the extent to which age may be affecting the results cannot be determined. Thus, the possibility that the temporal variation in pathology load seen in this study is related to differences in age rather than diet cannot be excluded. Despite the difﬁculties presented above, the collections included in this study, while not ideal, represent the larger and more well-preserved skeletal samples available for analysis from the Lower Mississippi Valley. As such, they currently are the best (and only) alternative for researching prehistoric diet in this region from a bioarchaeological perspective. However, in recognition of the potential biases affecting the samples, results presented here are conditional with the provision that, should additional materials become available in the future, hypotheses for temporal variation in diet can be reconsidered. TABLE 6. Results of the independent samples t test of the MNPS statistic for all dental pathologies AMTLa Abscesses Pre-Coles Creek Coles Creek t-value Signiﬁcance Calculus N Mean N Mean N 96 132 0.08 0.06 144 144 0.08 0.07 77 131 1.232 0.219 0.636 0.525 AMTL 5 antemortem tooth loss; POD 5 periodontal disease. ‘‘*’’ indicates that relationship is statistically signiﬁcant. a American Journal of Physical Anthropology Mean 0.58 0.60 20.410 0.682 PODa Caries N 78 134 Mean 0.05 0.09 21.596 0.112 N Mean 104 135 0.70 0.55 2.722 0.007* 37 BIOARCHAEOLOGICAL ANALYSIS OF COLES CREEK DIET TABLE 7. Summary statistics and results of chi-square for severity of calculus and periodontal disease Severity of calculus None Pre-Coles Creek Coles Creek Chi-square Small Moderate N % N % 241 396 41.8 39.4 270 520 46.9 51.7 Heavy Total % Naffa Ntota % 61 10.6 4 0.7 84 8.4 5 0.5 X2 5 4.514; df 5 3; Sig. 5 0.211 335 609 576 1005 58.2 60.6 N % N Severity of periodontal disease \½ Root None Pre-Coles Creek Coles Creek Chi-square N % N % 207 308 32.1 29.1 220 535 34.1 50.6 [½ Root Remnants N N % % Resorbed N 39 6.0 78 12.1 101 75 7.1 39 3.7 100 2 X 5 80.596; df 5 4; Sig. 5 0.000* Total % Naffa Ntota % 15.7 9.5 438 749 645 1057 67.9 70.9 a Naff 5 number of teeth or individuals affected with pathology; Ntot 5 total number of teeth or individuals available for assessment. ‘‘*’’ indicates that relationship is statistically signiﬁcant. With regard to intersite variation in pathology load, summary data are presented for individual sites. However, data were not assessed statistically not only because the focus of the research is regional in scope (rather than local), but also because the resulting samples sizes would have been too small to yield meaningful results. Nevertheless, some general observations can be made. For the Pre-Coles Creek populations, while none of the populations has a consistently light or heavy pathology load, Little Woods has no observed caries or abscesses and is in the lower range of data for AMTL and calculus (based on individual counts). Generally, intersite data for the Pre-Coles Creek sample are comparable, though the range of variation exhibited in the frequencies of some pathological conditions is greater than in others (i.e., calculus and periodontal disease are more variable than caries, AMTL, and tooth wear). For the Coles Creek sample, Lake George is conspicuously different from the other populations in its caries data: the individual count is considerably higher than the others and it is the only population that averages more than one carious lesion per individual (Table 2). For all other pathological conditions (as well as for caries tooth counts), none of the Coles Creek populations stands out as having a consistently light or heavy pathology load (including Lake George). On the other hand, the range of variation among Coles Creek populations exhibited in the frequencies of some pathological conditions is greater than that found in the Pre-Coles Creek sample. This latter fact, together with the caries data for Lake George, may suggest that intraregional variation in diet existed during the Coles Creek period with some populations consuming greater quantities of carbohydrates than others. In addition to the potential intraregional variation in diet during the Coles Creek period, apparent interregional variation in resource exploitation is hindering the assessment of temporal trends in pathology load based on ‘‘normal’’ expectations. All of the Pre-Coles Creek populations are believed to have been hunter-ﬁsher-gatherers (Ford and Quimby, 1945; Kidder, 2002). Elsewhere in North America, such populations typically exhibit low rates of dental pathological conditions (with the exception of tooth wear; Cohen and Armelagos, 1984). The pathology load in the Pre-Coles Creek sample, therefore, is higher than anticipated, particularly for calculus and TABLE 8. Expectations and data trends between Pre-Coles Creek and Coles Creek samples Dental pathologies (individual count) Abscess frequency AMTL frequency Calculus frequency Caries frequency Caries average/individual Periodontal disease frequency Dental pathologies (tooth count) Abscess frequency AMTL frequency Calculus frequency Calculus severity Caries frequency Periodontal disease frequency Periodontal disease severity Wear (M1) Dental pathologies (MNPS) Abscess average AMTL average Calculus average Caries average Periodontal disease average Expectationa Trend in data 1 1 1 1 1 1 1 1 1 1 1 2 1 1 1 1 1 1 1 2 2 1 1 2 1 1 2 2 1 1 1 1 1 2 2 1 1 2 a A ‘‘1’’ indicates an increase through time; a ‘‘2’’ indicates a decrease through time. periodontal disease (the frequencies of these two pathological conditions are 58 and 68% (based on tooth counts) and 78 and 82% (based on individual counts, respectively). These results suggest that resource exploitation for hunter-gatherers in the Lower Mississippi Valley differed from that of other regions in eastern North America. Archaeologists have long reported that populations in the Lower Mississippi Valley relied on locally-available resources that, in addition to the abundant animal and aquatic species, included a host of starchy plants other than maize (such as amaranths, chenopods, goosefoot, little barley, maygrass, and knotweed), (Smith, 1986; Fritz and Kidder, 1993; Kidder and Fritz, 1993; Fritz, 1995; Roberts, 2000; Gremillion, 2002; Kidder, 2002). The presence of these plants in the diet must be contributing to the higher-than-expected pathology load for the Pre-Coles Creek populations. American Journal of Physical Anthropology 38 G.A. LISTI Despite the unusually high Pre-Coles Creek pathology load, the Coles Creek sample nevertheless has higher frequencies of most pathological conditions. However, with the exception of periodontal disease and tooth wear, the differences are neither signiﬁcant (when tested statistically), nor substantial (when noted through general observation), regardless of which method is used to analyze data. For example, based on tooth counts, increases in pathology load between samples vary only by 0.5 to 3%. Increases in pathology load based on individual counts are slightly higher, ranging from 1.5 to 11.2%. These slight increases in pathology load could signify a minor increase in carbohydrate consumption during the Coles Creek period, but they would not indicate a complete change in diet. Thus, at a time when populations elsewhere in eastern North America were increasing their reliance on domesticated plants, those in the Lower Mississippi Valley continued to rely on abundant natural resources, some of which included starchy plants. The supposition that Coles Creek populations were not maize agriculturalists is further supported when comparing carious lesions data between the Coles Creek sample and agricultural populations from the southeastern United States. Larsen et al. (1991) report carious lesion frequencies of 58.9% (based on individual counts) and 19.4% (based on tooth counts) for Precontact Agricultural populations from the Atlantic Coast of Georgia and Florida (dating to approximately A.D. 1150–1550). Similarly, Powell (1991) reports carious lesions frequencies of 54% (based on individual counts) and 18.7% (based on tooth counts) for the agricultural population at Moundville, located in western, central Alabama (dating to approximately A.D. 1000–1450). Conversely, the frequencies of carious lesions for the Coles Creek sample are well below those values: 29% (based on individual counts) and 6.7% (based on the tooth counts). Though such comparisons to the literature can be problematic due to methodological differences among researchers, the similarity between the agricultural populations and dissimilarity between the agricultural and Coles Creek data are notable. CONCLUSION Coles Creek diet traditionally was believed to have been based on maize agriculture due to the number, size, and complexity of Coles Creek sites. However, direct evidence in the form of maize kernels and pollen generally has been lacking. Instead, zooarchaeological and botanical evidence suggests that Coles Creek populations were supported by native plant, animal, and aquatic natural resources rather than agriculture, though some cultivated plants may have supplemented the diet. The present study considered Coles Creek subsistence from a bioarchaeological perspective: the dentitions of 288 adults from eight sites in the Lower Mississippi Valley ranging in date from 800 B.C. to A.D. 1200 were examined for multiple oral health indicators and statistical analyses were used to assess temporal changes in pathology load. Results from this study, though provisional because of sample biases, demonstrate that hunter-gatherers in the southern Lower Mississippi Valley have a higher pathology load than typically found in other areas of Eastern North America due to the presence in the diet of starchy plants other than maize. Additionally, the pathology load during the Coles Creek period is slightly heavier than that of the Pre-Coles Creek period; however, the differAmerican Journal of Physical Anthropology ences are neither signiﬁcant, nor substantial. Thus, an increase in carbohydrate consumption, whether from starchy plants already present in the diet or, possibly, the introduction of maize in some areas (i.e., Lake George), was minimal and occurred on a local level, if at all. Ultimately, this study complements the botanical, zooarchaeological, and artifactual records and provides bioarchaeological support that Coles Creek diet was not based on maize agriculture. ACKNOWLEDGMENTS The author thanks the following individuals for their assistance with various aspects of this research: Dr. Rebecca Saunders and Mr. Steve Fullen, Louisiana State University Museum of Natural Science; Dr. Michele Morgan, Peabody Museum of Archaeology and Ethnology, Harvard University; Dr. John Verano, Tulane University; Ms. Mary Lee Eggart, Cartographer, Louisiana State University Department of Geography and Anthropology; Ms. Mary Manhein, Louisiana State University Forensic Anthropology and Computer Enhancement Services Laboratory; and the Editor, Associate Editor, and two anonymous reviewers for their constructive comments and suggestions. LITERATURE CITED Aufderheide AC, Rodriguez-Martin C. 1998. The Cambridge encyclopedia of paleopathology. Cambridge: Cambridge University Press. Belmont JS. 1967. The culture sequence at the Greenhouse Site, Louisiana. Proceedings of the Southeastern Archaeological Conference. p 27–35. Berry DR. 1985. Dental paleopathology of Grasshopper Pueblo, Arizona. In: Merbs CF, Miller RJ, editors. Health and disease in the prehistoric southwest. Tempe: Arizona State University. p 253–273. Brothwell DR. 1981. Digging up bones, 3rd ed. Ithaca: Cornell University Press. Buikstra JE, Ubelaker DH. 1994. Standards for data collection from human skeletal remains. Fayetteville: Arkansas Archaeological Survey. Byrd KM, Neuman RW. 1978. Archaeological data relative to the prehistoric subsistence in the Lower Mississippi River alluvial valley. Geosci Man 9:9–21. Cohen MN, Armelagos GJ. 1984. Paleopathology at the origins of agriculture. Orlando: Academic Press. Colburn ML, Styles BW. 1990. Faunal remains from the Bangs Slough Site. In: Schambach FF, editor. Coles Creek and Mississippi period foragers in the Felsenthal region of the Lower Mississippi Valley. Arkansas Archaeological Research Series No. 39. Fayetteville: Arkansas Archaeological Survey. p 95–108. Ford JA. 1951. Greenhouse. A Troyville-Coles Creek period site in Avoyelles Parish, Louisisana. Anthropological papers of the American Museum of Natural History, Vol. 44. Part I. New York. Ford JA, Quimby GI Jr. 1945. The Tchefuncte culture. An early occupation of the Lower Mississippi Valley. Memoirs of the society for American archaeology. Number 2. Menasha: Society for American Archaeology. Fritz GJ. 1995. New dates and data on early agriculture. The legacy of complex hunter-gatherers. Ann Miss Botan Gar 82:2–15. Fritz GJ, Kidder TR. 1993. Recent investigations into prehistoric agriculture in the Lower Mississippi Valley. Southeastern Arch 12:1–14. Giardino MJ. 1977. An osteological analysis of the human population from the Mount Nebo Site, Madison Parish, Louisiana. BIOARCHAEOLOGICAL ANALYSIS OF COLES CREEK DIET Unpublished MA thesis. Department of Anthropology. New Orleans: Tulane University. Giardino MJ. 1982. Temporal frameworks. Archaeological components and burial styles. The human osteology of the Mount Nebo site in north Louisiana. Louisiana Arch 9:99–126. Gremillion KJ. 2002. The development and dispersal of agricultural systems in the Woodland Period southeast. In: Anderson DG, Mainfort RC Jr, editors. The Woodland southeast. Tuscaloosa: University of Alabama Press. p 483–501. Haag WG. 1978. A prehistory of the Lower Mississippi River Valley. Geosci Man 9:1–8. Jackson HE, Scott SL. 2002. Woodland faunal exploitation in the midsouth. In: Anderson DG, Mainfort RC Jr, editors. The Woodland southeast. Tuscaloosa: University of Alabama Press. p 461–482. Jeter MD, Rose JC, Williams GI Jr, Harmon AM. 1989. Adaptation types. In: Jeter MD, Rose JC, Williams GI Jr, Harmon AM, editors. Archaeology and bioarchaeology of the Lower Mississippi Valley and trans-Mississippi south in Arkansas and Louisiana. Fayetteville: Arkansas Archaeological Survey. p 355–378. Jeter MD, Williams GI Jr. 1989. Ceramic-using cultures, 600 B.C. to A.D. 700. In: Jeter MD, Rose JC, Williams, GI Jr, Harmon AM, editors. Archaeology and bioarchaeology of the lower Mississippi Valley and trans-Mississippi south in Arkansas and Louisiana. Fayetteville: Arkansas Archaeological Survey. p 111–170. Kelley DB. 1992. Coles Creek period faunal exploitation in the Ouachita River Valley of southern Arkansas. Midcontinental J Arch 17:227–264. Kidder TR. 1992a. Timing and consequence of the introduction of maize agriculture in the Lower Mississippi Valley. North Am Arch 13:15–41. Kidder TR. 1992b. Coles Creek period social organization and evolution in northeast Louisiana. In: Barker AW, Pauketat TR, editors. Lords of the southeast. Social inequality and the native elites of southeastern North America. Archaeological Papers of the American Anthropological Association. Number 3. Washington, DC. p 145–162. Kidder TR. 2002. Woodland period archaeology of the Lower Mississippi Valley. In: Anderson DG, Mainfort RC Jr, editors. The Woodland southeast. Tuscaloosa: University of Alabama Press. p 66–90. Kidder TR, Fritz GJ. 1993. Subsistence and social change in the Lower Mississippi Valley. The Reno Brake and Osceola Sites, Louisiana. J Field Arch 20:281–297. Larsen CS. 1984. Health and disease in prehistoric Georgia. The transition to agriculture. In: Cohen MN, Armelagos GJ, editors. Paleopathology at the origins of agriculture. Orlando: Academic Press. p 367–393. Larsen CS. 1995. Biological changes in human populations with agriculture. Ann Rev Anthropol 24:185–213. Larsen CS. 1997. Bioarchaeology. Interpreting behavior from the human skeleton. Cambridge: Cambridge University Press. Larsen CS, Shavit R, Grifﬁn MC. 1991. Dental caries evidence for dietary change. An archaeological context. In: Kelley MA, Larsen CS, editors. Advances in dental anthropology. New York: Wiley-Liss. p 179–202. Lewis BA. 1991. Analysis of pathologies present in the 16ST1 Tchefuncte Indian skeletal collection. Unpublished MA thesis, Department of Geography and Anthropology. Baton Rouge: Louisiana State University. Lukacs JR. 1989. Dental paleopathology. Methods for recon_can MY, Kennedy KAR, edistructing dietary patterns. In: Is tors. Reconstruction of life from the skeleton. New York: Wiley-Liss. p 261–286. Mariaca MT. 1988. Late Marksville/ Early Baytown period subsistence economy. Analysis of three faunal assemblages from northeastern Louisiana. Unpublished MA thesis, Department of Anthropology. Boston: Boston University. McGimsey C. 2003. The Morton Shell Mound project and other stories of southwest Louisiana history. Regional Archaeology Program. Management Unit III. 2002/2003 Annual Report. Moore WJ, Corbett ME. 1971. The distribution of dental caries in ancient British populations. I. Anglo-Saxon period. Caries Res 5:151–168. 39 Neuman RW. 1984. An introduction to Louisiana archaeology. Baton Rouge: Louisiana State University Press. Norr L. 1984. Prehistoric subsistence and health status of coastal peoples from the Panamanian isthmus of lower Central America. In: Cohen MN, Armelagos GH, editors. Paleopathology at the origins of agriculture. Orlando: Academic Press. p 463–490. Ortner D. 2003. Identiﬁcation of pathological conditions in human skeletal remains, 2nd ed. San Diego: Academic Press. Perzigian AJ, Tench PA, Braun DJ. 1984. Prehistoric health in the Ohio River Valley. In: Cohen MN, Armelagos GH, editors. Paleopathology at the origins of agriculture. Orlando: Academic Press. p 347–366. Powell ML. 1985. The analysis of dental wear and caries for dietary reconstruction. In: Gilbert RI Jr, Mielke JH, editors. The analysis of prehistoric diets. Orlando: Academic Press. p 307–338. Powell ML. 1991. Ranked status and health in the Mississippian chiefdom at Moundville. In: Powell ML, Bridges PS, Wagner-Mires AM, editors. What mean these bones? Studies in southeastern bioarchaeology. Tuscaloosa: University of Alabama Press. p 22–51. Robbins LM. 1976. Analysis of human skeletal material from Morton Shell Mound (16IB3), Iberia Parish, Louisiana. Unpublished manuscript on ﬁle, Department of Geography and Anthropology. Baton Rouge: Louisiana State University. Roberts KM. 2000. Plant remains. In: Ryan J, editor. Data recovery excavations at the Hedgeland Site (16CT19), Catahoula Parish, Louisiana. Report submitted to the US Army Corps of Engineers. Vicksburg District. Baton Rouge: Coastal Environments. p 10.1–10.33. Rose JC, Burnett BA, Nassaney MS, Blaeuer MW. 1984. Paleopathology and the origins of maize agriculture in the Lower Mississippi Valley and Caddoan culture area. In: Cohen MN, Armelagos GH, editors. Paleopathology at the origins of agriculture. Orlando: Academic Press. p 393–424. Rose JC, Harmon AM. 1999. Louisiana and south and eastern Arkansas. In: Rose JC, editor. Bioarchaeology of the south central United States. Arkansas Archaeological Survey Research Series No. 55. Fayetteville: Arkansas Archaeological Survey. p 35–82. Rose JC, Marks MK, Tieszen LL. 1991. Bioarchaeology and subsistence in the central and lower portions of the Mississippi Valley. In: Powell ML, Bridges PS, Wagner-Mires AM, editors. What mean these bones? Studies in southeastern bioarchaeology. Tuscaloosa: University of Alabama Press. p 7–21. Scarry CM. 1993. Variability in Mississippian crop production strategies. In: Scarry CM, editor. Foraging and farming in the eastern Woodlands. Gainesville: University Press of Florida. p 78–90. Schmucker BJ. 1985. Dental attrition. A correlative study of dietary and subsistence patterns in California and New Mexico Indians. In: Merbs CF, Miller RJ, editors. Health and disease in the prehistoric southwest. Tempe: Arizona State University. p 253–273. Schoeninger MJ, Schurr MR. 1994. Interpreting carbon stable isotope ratios. In: Sobolik K, editor. Paleonutrition. The diet and health of prehistoric Americans. Center for Archaeological Investigations. Occasional paper No. 22. Carbondale: Southern Illinois University. p 55–66. Scott EC. 1979. Dental wear scoring technique. Am J Phys Anthropol 51:213–217. Smith BD. 1986. The archaeology of the southeastern United States: From Dalton to de Soto, 10,500-500 B.P. In: Wendorf F, Close A, editors. Advances in world prehistory, Vol.5. Orlando: Academic Press. p 1–92. Smith BH. 1984. Patterns of molar wear in hunter-gatherers and agriculturalists. Am J Phys Anthropol 63:39–56. Smith RL. 1996.Vertebrate subsistence in southeastern Louisiana between A.D. 700 and 1500. Unpublished MA thesis, Department of Anthropology. Athens: University of Georgia. Sobolik KD. 1994. Introduction. In: Sobolik KD, editor. Paleonutrition. The diet and health of prehistoric Americans. Carbondale: Southern Illinois University. p 1–20. American Journal of Physical Anthropology 40 G.A. LISTI Springer JW. 1980. An analysis of prehistoric food remains from the Bruly St. Martin Site, Louisiana, with a comparative discussion of Mississippi Valley faunal studies. Mid-Continental J Arch 5:193–223. Toth EA. 1979. The Lake St. Agnes site. A multi-component occupation of Avoyelles Parish, Louisiana. Louisiana State University Museum of Geoscience. Mèlanges No. 13. Baton Rouge. American Journal of Physical Anthropology Turner CG. 1979. Dental anthropological indications of agriculture among the Jomon people of central Japan. Am J Phys Anthropol 51:619–636. Williams S, Brain JP. 1983. Excavations at the Lake George Site Yazoo County, Mississippi, 1958–1960. Papers of the Peabody Museum, No. 74. Cambridge: Harvard University Press. Wood JW, Milner GR, Harpending HC, Weiss KM. 1992. The osteological paradox. Cur Anthropol 33:343–370.