science

Deer use in good times and in bad: A Fort Ancient case study from southwest Ohio Jacob Deppen1, Robert A. Cook2

1Department of Anthropology, University of Washington, Seattle, WA, USA, 2Department of Anthropology, The Ohio State University, Newark, OH, USA

White-tailed deer (Odocoileus virginianus) utilisation before and during increased moisture variability and pronounced drought conditions during the late prehistory of southwestern Ohio is examined to assess the fit with the expectations of foraging efficiency models. The focus is on three Fort Ancient sites in the upper portion of the Great Miami River, including SunWatch, a large village located in present-day Dayton, Ohio. SunWatch was occupied during the Late Prehistoric era in the region. The earlier uses (AD 1150–1300) occurred during optimal moisture conditions. The later uses (AD 1300–1450) occurred within the context of increased droughts and extreme moisture variability. To address questions related to changing deer utilisation in response to drought, deer remains are examined temporally for the SunWatch site and two smaller Fort Ancient sites in the region (Wegerzyn and Wildcat). Results from this preliminary and exploratory study indicate that through time, deer body size is stable or decreases, juvenile deer became more abundant in the hunted assemblages, and human butchering strategies became less selective. These support the conclusion that environmental stress on the deer population led to a change in the deer population and influenced the way humans used deer.

Keywords: Archaeozoology, Drought, Climate Change, White-tailed deer, Fort Ancient, Ohio, USA

This study is a preliminary and exploratory examin- ation of deer remains to assess whether environmental stress, specifically drought, affected deer (Odocoileus virginianus) hunting and utilisation strategies by Native American communities in southwestern Ohio. Our focus is on a case study derived from late prehis- toric contexts in the upper portion of the Great Miami River in southern Ohio, USA, but the impli- cations are general in that they can be applied any- where drought has been well documented and for which exist faunal assemblages from dated contexts. White-tailed deer utilisation before and during a

severe drought during the Late Prehistoric period in southwestern Ohio is examined to assess the fit with the expectations of foraging efficiency models. White-tailed deer were chosen for analysis because they are unquestionably the most abundant taxon in these assemblages. The focus is on three Fort Ancient sites in the upper portion of the Great Miami River, including SunWatch, a large village located in present-day Dayton, Ohio. SunWatch was occupied at various times throughout much of the Late Prehistoric era in the region (Cook 2007a, 2008). The earlier occupations (AD 1150–1300)

occurred during optimal moisture conditions and mostly during the warm months, with colder months presumably spent at smaller camps. The later occu- pations (AD 1300–1450) occurred on a year-round basis, during and after a prolonged drought. Dietary analyses have not been temporally examined for SunWatch or for smaller Fort Ancient sites in the region. This study addresses this issue by examining whether or not there are differences in the deer assem- blages between pre-drought and post-drought con- texts. Assemblages used to address this question are from dated features at SunWatch and the only other Fort Ancient sites in the region that have been exca- vated, Wegerzyn and Wildcat.

Theoretical Expectations Foraging theory forms the basis of our theoretical expectations. Optimal foraging theory assumes that human populations will try to maximise their strat- egies with respect to a certain currency, usually energy (Bettinger 1991; Smith and Winterhalder 1992; Kelly 1995; Winterhalder 2001). We draw specifically on the prey-choice model within foraging theory. The prey-choice model assumes that humans hunt the highest-ranked animal resources whenever they are encountered (Schoener 1971; Pulliam 1974; Charnov 1976; Pyke et al. 1977; Stephens and Krebs

Correspondence to: Jacob Deppen, Department of Anthropology, University of Washington, Denny Hall, Box 353100, Seattle, WA 98195, USA. Email: [email protected]

© Association for Environmental Archaeology 2014 DOI 10.1179/1749631413Y.0000000002 Environmental Archaeology 2014 VOL. 19 NO. 172

1986; Winterhalder and Smith 2000). The presence of lower-ranked animals in the hunted population depends on the encounter rate with the highest- ranked resources. Thus, if the encounter rate with high-ranking resources declines, the presence of lower-ranking resources would increase. White-tailed deer are the highest-ranking resource in

much of the eastern United States, including southwest Ohio. While other mammals like elk and bear provided the opportunity for more total meat per animal, their relative scarcity and/or the energy required to hunt them means that they likely would have had a lower net energy gain than white-tailed deer (Reidhead 1981). As the highest-ranked mammal resource, we expect the zooarchaeological assemblages from the study sites to reflect the abundance of high-ranked deer on the landscape. More specifically, Wolverton’s (2008: Fig. 2, 185) model gives us tools to examine the twin effects of human harvest pressure and changes in environmental carrying capacity on deer populations. During a drought, we expect that agricul- tural communities like those investigated here would face shortages (or at least instability and uncertainty) in the domesticated plant resources they had come to rely on. To mitigate this shortage, we expect that they would rely more heavily on wild resources, especially highly ranked wild resources like deer. In the terms of the model, this would be seen as an increase in human harvest pressure on the deer population and would be indicated by a steepening survivorship profile of the population. Steepening survivorship simply refers to a decrease in the number of adult deer relative to young deer because more and more adults are being removed from the population by preda- tors, in this case, hunters. The deer population itself would also have experi-

enced the effects of drought directly. A decrease in the amount of forage available to the deer population (i.e. a decrease in environmental carrying capacity) would generally be expected to result in a decrease in body size. However, increased harvest pressure, as pre- dicted above, normally has the opposite effect: an increase in body size. When these two phenomena are paired together, it is reasonable to expect either a modest reduction in body size or no change at all. In sum, according to expectations derived from the

model, we expect a drought would have had the twin effects of an increase in human harvest pressure and a decrease in environmental carrying capacity. In the zooarchaeological record, we should see an increase in the number of younger deer relative to older deer and a slight decrease or no change in deer body size when comparing assemblages before and during the drought. Changes in either human harvest pressure or

environmental carrying capacity might also be

evident in the utility patterns. In most applications of foraging theory, more selective utility patterns would be interpreted as an increase in transportation distance as a result of a decrease in the encounter rate (i.e. a decrease in the prey population) (e.g. Broughton 2002; Nagaoka 2002, 2005). However, even in the worst of times, Fort Ancient hunters in southwest Ohio may not have had to travel so great a distance as to change their butchering strategies because deer populations would never have been so sparse as to demand it. It has even been suggested that agricultural settlements like those analyzed here would have been ideal environments for deer, possibly even drawing the deer to the hunters (Yerkes 2005; but see Ketchum et al. 2009). Thus, if we assume that trans- portation distance was not an important factor, changes in utility pattern could be caused by a change in the abundance of prime-aged deer. ‘Prime- aged’ is used here as a relative marker for fully mature deer that have reached their maximum body size and thus, would result in the most meat reward for a human hunter. The exact age at which deer reach their maximum body size is highly dependent on local conditions, but usually occurs around 4–5 years. As explained below, we proceed under the likely assumption that maximum (i.e. optimal) body size occurred sometime after the juvenile deer’s bones have fused. In a drought context, we expect the number of these prime-aged deer to decrease due to harvest pressure. With fewer of the optimal prey available, we expect to see a ‘take what you can get’ strategy where Fort Ancient hunters are less selective in their hunting and butchering practices. In sum, according to expectations derived from the

model, we expect a drought would have had the twin effects of an increase in human harvest pressure and a decrease in environmental carrying capacity. In the zooarchaeological record, we should see an increase in the number of younger deer relative to older deer and a slight decrease or no change in deer body size when comparing assemblages before and during the drought. In addition, as the number of mature deer decreased, we expect that Fort Ancient hunters adopted butchering strategies to get the most out of the less-than-optimal prey.

Study Sites The study sites used to address the cultural response in faunal utilisation after a severe drought belong to the last prehistoric culture to inhabit the middle Ohio River Valley. Archaeologists have named them ‘Fort Ancient,’ and they have also generally described them as ‘tribal’ societies (Henderson and Pollack 2001). The Fort Ancient culture area was initially identified in reference to villages in southwestern Ohio (Putnam 1886; Moorehead 1892), but was later

Deppen and Cook Deer use in good times and in bad

Environmental Archaeology 2014 VOL. 19 NO. 1 73

expanded to cover much of the Middle Ohio Valley (Griffin 1943) (Fig. 1). (Detailed overviews can be found in Cowan 1987, Drooker 1997, and Henderson 1992.) Fort Ancient sites date from approximately AD

1000 to AD 1650 (Drooker 1997). There was only

one major change in settlement patterning during this 650-year period. At around AD 1400, there was a marked abandonment of the neighboring region to the west, which became the ‘Vacant Quarter’ (Williams 1990; Cobb and Butler 2002). In tandem with this abandonment, Fort Ancient sites became

Figure 1 Locations of study sites within the Fort Ancient region and closest PDSI cell (#227).

Deppen and Cook Deer use in good times and in bad

Environmental Archaeology 2014 VOL. 19 NO. 174

more restricted in the distribution of their settlements. They became more focused on the main rivers, and particularly along the Ohio rather than many sites in locales in the upper reaches of its tributaries (Kennedy 2000). At the same time, there may have been a slight increase in social complexity that hap- pened alongside further reaching exchange networks (Pollack and Henderson 1992, 2000; Drooker 1997; Cook 2008). This shift seems to be partially related to the end of the Medieval Warm Period and the beginning of the ‘Little Ice Age’. This period of global cooling was one of cool and dry conditions that affected many world populations. In this study, we focus on three excavated Fort

Ancient sites in the upper portion of the Great Miami River (SunWatch, Wegerzyn, and Wildcat) (see Fig. 1). This is an area known for its proximity to a dense concentration of prairie openings (Wagner 1988), ideal habitat for white-tailed deer (Hesselton and Hesselton 1982). Before briefly summarising these sites, we first describe the locale used for recon- structing the drought conditions in the area.

Palmer Drought Severity Index Locale Environmental variability reconstructions created by Cook et al. (1999, 2004) reveal a significant change in the moisture regime in this region through time (Fig. 2). These studies compared dendrochronology samples from throughout North America with the Palmer Drought Severity Index (PDSI) to examine the occurrence and duration of drought over the last

2000 years. Relevant to this study, Cook et al. (1999, 2004) found that during much of the AD 1300s, south- west Ohio was notably drier than previous centuries, conditions that persist for much of the remainder of the Fort Ancient period (ending ca. AD 1650). (It is important to note that there were some good years during this generally poor time – relative moisture often continued to fluctuate greatly from year to year. The extreme variability coupled with the cumu- lative effect of bad years is the important pattern.) These data form the temporal basis for our drought consideration and are available for querying and download on NOAA Paleoclimatology’s World Data Center for Paleoclimatology and the Applied Research Center for Paleoclimatology (www.ncdc .noaa.gov/paleo/pdsi.html). For our study, we used PDSI grid point #227, as it was closest in proximity to the study sites (see Fig. 1).

SunWatch The SunWatch site (33MY57) is a relatively large (1·4 ha), circular Fort Ancient habitation site located in what is now Dayton, Ohio (Fig. 3). Excavations at the site began in the 1960s by interested avocational archaeologists (Allman 1968; Smith n.d.), and have been conducted since then under the direction of the Dayton Society of Natural History (Heilman et al. 1988; Cook 2008). Portions of the site have been recon- structed and an interpretive center welcomes visitors to the site. The site is often used as an example of the classic Fort Ancient ‘doughnut-shaped’ village, with rings of houses, burials, and storage pits surrounding an open plaza (Henderson and Pollack 2001). The site sits on the floodplain of the Great Miami

River on the Wea Silt Loam soil type (Davis 1976), ideal for maize agriculture. Based on a recent analysis of maize consumption (based on stable carbon isotopes), we know that the inhabitants of SunWatch consumed more maize than other sites in the region (Cook and Schurr 2009). The results of over 30 years of excavation and analyses have revealed details about village structure during two distinct phases of use. In the early phase (AD 1150–1300), the site was occupied during the warm seasons, with winters likely spent away at hunting camps (Shane 1988; Wagner 1996, 2008). Corporate leadership also characterised this period, with family groups holding influence over their small sections in the village circle (Cook 2008). During the later phase (AD 1300–1450), the site was occupied year-round and leadership developed on a village- wide scale, during a time of marked Mississippian pres- ence in the village (Cook 2008).

Wegerzyn The Wegerzyn site (33MY127) is a 0·3-ha habitation located along the Stillwater River 10·5 km north

Figure 2 Relative moisture between AD 1000 and 1600 for PDSI #227 (see Fig. 1 for location, and Cook et al. 1999, 2004). Note: Data presented above are based on the 20-year low pass interval for clarity of presentation. Shaded area denotes onset of our interpretation of increased moisture variability and drought conditions.

Deppen and Cook Deer use in good times and in bad

Environmental Archaeology 2014 VOL. 19 NO. 1 75

of SunWatch (see Fig. 3). Professional excavations under the direction of the Dayton Society of Natural History began with a few test excavations and have been ongoing most summers since the mid-1990s (Simonelli and Kennedy 2003; Kennedy 2007). The site is situated on the Ross Silt Loam soil type (Davis 1976). While this soil is also excellent for maize agriculture, stable carbon isotope analysis reveals a lower level of maize consumption by people there (Cook and Schurr 2009). Radiocarbon dates place the main occupation of the site in the AD 1300s (Bill Kennedy, personal communication 2009). The site appears to be similar in plan to other small circular Fort Ancient villages like Horseshoe Johnson (Hawkins 1998), although further fieldwork is needed before this is confirmed.

Wildcat The Wildcat site (33MY499) is located north of both SunWatch (18 km) and Wegerzyn (8·5 km). While SunWatch sits 0·25 km from the Great Miami River and Wegerzyn sits 0·16 km from the Stillwater River, Wildcat sits near a small intermittent stream, 1·7 km from the Miami River. Professional excavations were

conducted by The Ohio State University Department of Anthropology from 2007 to 2009. Extensive shovel testing, magnetic gradiometry and suscepti- bility survey, and excavations indicate that the site con- sists of a small habitation covering 0·3 ha (Cook 2007b) (see Fig. 3). The site as a whole was likely occu- pied between the mid-AD 1200s and late AD 1300s. The site is situated on the Corwin silt loam soil type (Davis 1976), which is not as productive as Ross or Wea silt loam. No stable carbon isotope tests have been conducted, but we do know based on a prelimi- nary analysis of macrobotanical remains that maize density is roughly four times lower than at SunWatch (Martin 2009).

Faunal Samples The soils in the study area range in pH from 7·0 to 8·0, excellent conditions for faunal preservation (Davis 1976; Shane 1988; see Results section for more discus- sion of taphonomy). All samples are derived from fea- tures of the same type (storage/trash pits), sealed contexts where animal scavenging does not appear to have altered the assemblage (i.e. no extensive evidence of gnaw marks).

Figure 3 Maps of the SunWatch, Wegerzyn and Wildcat sites, indicating sample locations. Note: Wegerzyn map based on excavations through 2007. Wildcat map based on excavations through 2008.

Deppen and Cook Deer use in good times and in bad

Environmental Archaeology 2014 VOL. 19 NO. 176

The sample included all of the white-tailed deer bones from a selection of trash pit features: Feature 2/05 and Pit Feature Group 2·1 at SunWatch, Feature 2/00 at Wegerzyn, and Feature 3/07 at Wildcat (see Fig. 3). Two of the contexts produced early dates: SunWatch Feature 2/05 produced a median calibrated AMS radiocarbon date of AD 1290 and Wildcat Feature 3/07 produced a median calibrated AMS radiocarbon date of AD 1273. Two others produced later dates: Wegerzyn Feature 2/00 produced a median calibrated AMS radiocarbon date of AD 1413 and SunWatch Pit Feature Group 2·1 produced a median calibrated AMS radiocarbon date of AD 1359. This latter date places it in the later stage (ca. AD 1300–1450) of SunWatch’s occu- pational sequence (Cook 2007a). These four samples allow temporal comparison of sites from which to assess whether utilisation practices changed from pre-drought (SunWatch [Feature 2/05] and Wildcat [Feature 3/07]) to drought conditions (and Wegerzyn [Feature 2/00] and SunWatch [Pit Feature Group 2·1]) (Fig. 4).

Methodology Shane (1988) identified deer bones from SunWatch Pit Feature Group 2·1. This methodology was used for the remainder of the samples (identified by Deppen). Deer bones were identified to skeletal element, side, comple- teness (proximal, medial, distal, complete), and epi- physeal closure using osteological reference materials (Olsen 1964; Gilbert 1980) and the comparative collec- tion in the Department of Anthropology at The Ohio State University.

Age and Body Size Ideally, to examine age in deer, an age profile of the assemblage would be constructed using standards for the wear on deer teeth. However, the sample sizes for teeth in these assemblages make constructing accurate profiles in this way impossible. Instead, we use an admittedly less precise index based on epiphyseal fusion. As with all mammals, white-tailed deer begin life

with the ends of many bones (epiphyses) not yet ‘fused’ with the shafts (diaphyses). As the animal grows, the bones will steadily fuse to form complete

Figure 4 Calibrated radiocarbon dates showing two-sigma ranges (OxCal4; Bronk Ramsey 2009).

Deppen and Cook Deer use in good times and in bad

Environmental Archaeology 2014 VOL. 19 NO. 1 77

bones. The time of fusion differs for each bone and even for the different ends (proximal, distal) of the same bone. Most white-tailed deer bones, however, are fused by the time the animal is around 24 months old, and nearly all before 36 months (Purdue 1983). After skeletal growth is complete though, deer continue to add body mass and do not reach their full body size (in other words, their full utility to hunters) until some- where around 4–5 years (Rue 1997). In the absence of better aging indices, unfused bones

are used here as a rough, relative measure of the number of juveniles in each sample. We compare the ratio of unfused to total bones for each assemblage. Results are also presented specifically for early-fusing bones. The values themselves should not be inter- preted as ratio scale data on the proportion of juveniles in the deer population. However, given relatively equal preservation conditions, the measure can be used as an ordinal scale measure with which to compare assem- blages. As we are only interested in change through time, not the exact amount of that change, an ordinal scale will suit our goals. Estimates of body size were made using measures of

the astragalus as this bone matures early, does not sig- nificantly change in size during development, and has been shown to be a strong indicator of size and sex (Purdue 1986, 1989). It should not be sensitive to any potential changes in the adult body size or the abundance of juvenile deer. Specific measurements used here are medial depth (ASMD) and lateral length (ASLLEN) (see Purdue 1989: Fig. 1).

Utilisation A food utility index was used to make intersite utility comparisons. Food utility indices have been used to

evaluate the butchering practices of prehistoric peoples (see Metcalfe and Jones 1988; Purdue et al. 1989; Yerkes 2005). An ideal butchering strategy, usually derived from experimental or ethnoarchaeolo- gical evidence, is compared to the observed archaeolo- gical assemblage to examine whether and how the prehistoric butchering practices diverged from the ideal. Following these studies, bones were grouped according their utility value (i.e. low, medium or high) and the relative abundances (based on the number of identified specimens (NISP)) were com- pared. From this analysis, we can examine whether there was any change in butchering practices through time.

Results At SunWatch, Feature 2/05 (early) contained 427 identifiable deer bones and Pit Feature Group 2·1 (late) contained 631 identifiable deer bones. Wildcat’s Feature 3/07 (early) produced 202 identifi- able specimens. Wegerzyn’s Feature 2/00 (late) pro- duced 1052 identifiable specimens.

Age and Body Size Epiphyseal fusion data support the expectation that later contexts contain more young deer. The patterns are best expressed by examining the number of unfused bones relative to the total number of bones in the sample (Fig. 5). Of the 427 bones in the SunWatch sample (early), only 22 (5·2 percent) are unfused. Wildcat (early) contains 202 identified deer bones, 25 (12·4 percent) of which are unfused, more than two times more than from SunWatch (early), which is a significant difference (χ2 = 10·35, P = 0·001). Wegerzyn (late) contains the largest amount

Figure 5 Percentage of juvenile deer for each sample context.

Deppen and Cook Deer use in good times and in bad

Environmental Archaeology 2014 VOL. 19 NO. 178

of unfused bones (n = 247 [21·5 percent]), more than four times the value for SunWatch (early) and over one and a half times greater than even the Wildcat sample, both of which are statistically significant (χ2 = 58·84, P < 0·0001, and χ2 = 8·92, P = 0·003, respectively). SunWatch (late) also contains a large amount of unfused bones (n = 112 [17·8 percent]), sig- nificantly more than the early SunWatch feature (χ2 = 36·54, P < 0·0001). This trend is also observed when limiting the analy-

sis to early-fusing bones (Table 1). The proximal radius, second phalanx, and first phalanx all begin to fuse before the deer is 1 year old, and thus should cer- tainly reflect deer that have not reached full maturity. Although sample sizes are not large enough for statisti- cal analysis, the ratio of unfused to fused occurrences

of these skeletal elements increases through time, as would be expected if an increasing number of young deer were being hunted. Consideration of the astragalus as an indicator of

body size reveals changes in body, if any, were very subtle (Fig. 6). The two early samples produced a mean ASLLEN of 41·81 mm (n = 19), while the two late samples produced a mean ASLLEN of 40·59 (n = 50). A two-tailed t test reveals these samples to be statistically different (P = 0·02). The two early samples produced a mean ASMD of 23·39 mm (n = 19), while the two late samples produced a mean ASMD of 22·82 (n = 50). A two-tailed t test reveals this to be of ambiguous statistical significance (P = 0·09). There was not a significant distinction for either measurement with respect to site size.

Table 1 Ratios of unfused to fused for early-fusing skeletal elements

Wildcat SunWatch (early) SunWatch (late) Wegerzyn

Unfused Fused Unfused Fused Unfused Fused Unfused Fused

Radius, proximal 1 3 0 1 0 14 2 12 Humerus, distal 0 9 0 11 0 12 2 21 2nd phalanx 0 7 2 9 4 27 12 16 1st phalanx 2 8 1 4 7 33 5 25 Total 3 27 3 25 11 76 21 74 Ratio (unfused:fused) 0·11 0·12 0·14 0·28

Figure 6 Astragalus measurements for early and late sample contexts (ASMD = medial depth; ASLLEN = lateral length [see Purdue 1989: Fig. 1]).

Deppen and Cook Deer use in good times and in bad

Environmental Archaeology 2014 VOL. 19 NO. 1 79

In addition to results from the statistical tests, visual inspection of the data (Fig. 6) also suggests that there is some difference between the early and late samples. However, it is important to note that even if statisti- cally significant, differences of just 1–2 mm could also be misleading in this context (though see Purdue 1989). While the largest astragali tend to be among the early group and the smallest tend to be among the late group, the amount of overlap suggests that the populations were not substantially different in body size.

Utilisation The food utility index shows a pattern of decreasing selectivity by Fort Ancient hunters through time. Table 2 shows the proportion of each skeletal element in the assemblage organised in order of utility (after Yerkes 2005). The earlier assemblages, Wildcat and SunWatch Feature 2/05, were selectively avoiding elements with low utility while the later samples reflect butchering practices that were less selective. The two early samples are very similar (r = 0·985) to each other as are the two later samples (r = 0·969). There is not a strong correlation between any of the early and late samples. It is also worth noting that there is a strong correlation between each of the late samples and the pattern that we would expect in a complete deer skeleton (r = 0·987 for Wegerzyn; r = 0·996 for late SunWatch), providing solid evidence that the entire animal was utilised. The strongest differences in utility are between early

and late contexts rather than between the larger SunWatch site and the smaller Wildcat and Wegerzyn sites. This seems to indicate that, again, differences are not related to site size or function. To ensure the differences observed above are not

simply an artifact of density-mediated destruction of

the bones, element abundances from each assemblage were compared with three bone density values for white-tailed deer presented in Lyman (1984). This analysis helps us understand whether the pattern we

Table 2 Food utility values (modeled after Yerkes 2005)

Wildcat (%) SunWatch (early) (%) SunWatch (late) (%) Wegerzyn (%) Complete Deer (%)

Skull 2·20 0·00 1·75 1·34 9·80 Mandible 0·73 2·42 0·00 1·69 1·40 Atlas + axis 1·47 0·24 1·91 1·96 1·40 Metacarpal + carpals 6·41 7·73 11·48 11·73 11·20 Phalanges 13·55 14·25 24·88 18·73 16·80 Total low 24·36 24·64 40·03 35·46 40·60

Vertebrae 10·62 20·29 12·92 17·75 16·80 Pelvis + sacrum 2·20 2·90 1·91 1·96 2·10 Ribs 13·55 22·71 4·78 13·11 16·80 Scapula 3·66 3·62 3·99 1·34 1·40 Humerus 5·49 5·56 2·71 3·39 2·80 Radius/ulna 5·49 1·45 6·54 3·30 2·80 Metatarsal 12·27 6·52 8·61 6·65 2·80 Total medium 53·30 63·04 41·47 47·50 45·50

Femur 5·49 4·83 1·59 2·50 2·80 Tibia + tarsals 12·82 7·49 16·91 12·13 11·20 Total high 18·32 12·32 18·50 14·63 14·00

Table 3 Rank-order correlations between skeletal elements and three bone density measures (after Lyman 1984)

Bone mineral density

Linear density

Volume density

SunWatch (early) Number of

pairs 62 62 62

Spearman’s rho

0·0601 0·0995 0·0534

T 0·47 0·77 0·41 Df 60 60 60 One-tailed p 0·32003 0·222161 0·341633 Two-tailed p 0·640059 0·444321 0·683266

Wildcat Number of

pairs 63 63 63

Spearman’s rho

0·1485 0·2172 −0·0162

T 1·17 1·74 −0·13 Df 61 61 61 One-tailed p 0·123276 0·043451 0·0448497 Two-tailed p 0·246552 0·086902 0·896994

Wegerzyn Number of

pairs 66 66 66

Spearman’s rho

−0·0666 −0·0416 −0·0796

T −0·53 −0·33 −0·64 Df 64 64 64 One-tailed p 0·298973 0·371239 0·262229 Two-tailed p 0·597946 0·742478 0·524457

SunWatch (late) Number of

pairs 62 62 62

Spearman’s rho

−0·224 −0·082 0·1833

T −1·78 −0·64 1·44 Df 60 60 60 One-tailed p 0·04007 0·262305 0·077533 Two-tailed p 0·08014 0·524609 0·155065

Deppen and Cook Deer use in good times and in bad

Environmental Archaeology 2014 VOL. 19 NO. 180

see is a product of taphonomic forces or a seemingly real pattern in past human behavior. Three different density standards are used because the rank order of the bone densities is slightly different when measured in different ways and there is no way to predict which is the proper standard to use. A Spearman’s rank-order correlation coefficient was calculated and the results are presented in Table 3. Of the 12 calcu- lations, three showed a statistically significant relation- ship between density and element abundance: the linear density for the Wildcat assemblage (Spearman’s rho = 0·2172; one-tailed P = 0·043) and the bone mineral density for the late SunWatch assemblage (Spearman’s rho = −0·224; one-tailed P = 0·040). The other 10 do not show a significant relationship between bone density and there abun- dance in the archaeological record. These results, though not conclusive, suggest that density did not play an important role in the element abundances.

Conclusion White-tailed deer (O. virginianus) bones from dated contexts at three Fort Ancient sites were examined to assess the fit with foraging models in relation to environmental stress. Cook et al. (1999, 2004) pro- vided evidence that we interpreted as supporting sig- nificant periods of drought in southwest Ohio, most notably in the AD 1300s and persisting through much of the remainder of the Fort Ancient period (until AD 1650). It is often difficult to differentiate between changes in the deer population that result from human activity and those that result from environmental factors. Wolverton (2008) provides a framework to determine whether a given assemblage is the result of changes in human harvest pressure or changes in environmental carrying capacity. This model suggests that a steepening of survivorship (a shift toward more young individuals) and stable or decreasing body size is indicative of a decrease in the environmental carrying capacity (Wolverton 2008: 185). This is exactly the pattern we have described for the assemblages presented here. More juvenile deer were hunted during the drought period while body size, as measured from the astragalus, was either stable or decreased slightly. Additionally, the utility patterns fit our expectations

outlined above. Pre-drought contexts indicate some selectivity for higher utility elements, but drought con- texts show a less selective pattern that closely matches the pattern of a complete deer skeleton. If one allows for our assumption that carcass transport distance was not an important factor to Fort Ancient hunters, these patterns may represent what we called the ‘take what you can get’ strategy that resulted from a decline in the abundance of prime-aged deer.

We conclude that the evidence presented here shows that drought episodes in southwest Ohio, beginning in the 14th century AD, caused a decline in the environ- mental carrying capacity for deer in the area at the same time that humans increased the harvest pressure on the deer population. Further, Fort Ancient hunters adapted their butchering strategies to make up for the reduction in the number of prime-aged, large-bodied deer. They shifted from a butchering strategy based on some selection for food utility to a strategy based on using the entire deer. We have offered drought as a possible cause for the

observed changes in the 14th and 15th centuries for three reasons: (1) a broadening of the hunted popu- lation to include less-optimal animals has been suggested for drought periods elsewhere (Bousman 2005); (2) the changes are independent of site size; and (3) the chronological agreement between the drought and the archaeological changes noted here. This interpretation, of course, should not be construed as definitive, but rather should be taken as a working hypothesis. Further research needs to address the scope of the drought as well as a variety of other factors (environmental and cultural) that may have contributed to the shift in strategies. Additionally, we recognize in this preliminary study

that the sample sizes presented are far from ideal. The feature contexts analyzed here were chosen because they represent the only dated Late Prehistoric features from sites in the area with adequately sized deer assem- blages. Faunal analysis of other dated contexts in the region will be vital to increase the sample size and to examine whether the pattern also occurs in other parts of the Fort Ancient world. Based on data we have presented, deer utilization

strategies in the upper portion of the Great Miami River drainage changed through time, possibly as a result of variable moisture and frequent deficits begin- ning in the 14th century AD These findings adhere well to the tenets of the foraging efficiency models. Only after examining more dated contexts, before, during, and after the drought period, will we be better able to examine its validity. We have presented a series of methods that seem promising for assessing changing hunting practices in Fort Ancient popu- lations, and advancing general knowledge of the impact of environmental change on faunal utilisation strategies.

Acknowledgements This study was made possible through several grants from The Ohio State University: an Undergraduate Research Scholarship (College of the Arts and Sciences), an Undergraduate Research Award (College of Social and Behavioral Sciences), a Summer Research Internship (University Honors

Deppen and Cook Deer use in good times and in bad

Environmental Archaeology 2014 VOL. 19 NO. 1 81

and Scholars Department), and a Research and Scholarly Activity Grant (Newark Campus). The landowners of Wegerzyn (Five Rivers Metroparks) and Wildcat (Cemex) deserve special thanks for allow- ing archaeological excavations to be conducted there, resulting in datasets examined here. Bill Kennedy, Curator of Anthropology at the Dayton Society of Natural History, provided essential access to the Wegerzyn collections and helped move the project along. Richard Yerkes provided invaluable comments on various aspects of the study and generously allowed open access to The Ohio State University Department of Anthropology’s comparative collection. Donald Grayson also provided many helpful comments to improve early drafts of the manuscript. We also thank the two anonymous reviews for their thoughtful and immensely useful comments. Any errors or omis- sions are the sole responsibility of the authors.

References Allman, J. C. 1968. The incinerator village site. Ohio Archaeologist

18, 50–6. Bettinger, R. L. 1991. Hunter–Gatherers, Archaeology, and

Evolutionary Theory. New York: Plenum Press. Bousman, C. B. 2005. Coping with risk: later Stone Age technologi-

cal strategies at Blydefontein Rock Shelter, South Africa. Journal of Anthropological Archaeology 24, 193–226.

Bronk Ramsey, C. 2009. Bayesian analysis of radiocarbon dates. Radiocarbon 51, 337–60.

Broughton, J. M. 2002. Prey spatial structure and behaviour affect archaeological tests of optimal foraging models: Examples from the Emeryville Shellmound Vertebrate Fauna. World Archaeology 34, 60–83.

Charnov, E. L. 1976. Optimal foraging: attack strategy of a Mantid. American Naturalist 110, 141–51.

Cobb, C. R. and Butler, B. M. 2002. The vacant quarter revisited: late Mississippian Abandonment of the lower Ohio Valley. American Antiquity 67, 625–41.

Cook, E. R., Meko, D. M., Stahle, D. W. and Cleaveland, M. K. 1999. Drought reconstructions for the Continental United States. Journal of Climate 12, 1145–62.

Cook, E. R., Woodhouse, C. A., Eakin, C. M., Meko, D. M. and Stahle, D. W. 2004. Long-term aridity changes in the Western United States. Science 306, 1015–8.

Cook, R. A. 2007a. Single component sites with long sequences of radiocarbon dates: the SunWatch Site and Middle Fort Ancient Village Growth. American Antiquity 72, 439–60.

Cook, R. A. 2007b. Ohio State University Field School at the Wildcat Site. Paper presented at the Ohio Archaeological Council Fall Membership Meeting, Newark.

Cook, R. A. 2008. SunWatch: Fort Ancient Development in the Mississippian World. Tuscaloosa, Alabama: University of Alabama Press.

Cook, R. A. and Schurr, M. R. 2009. Eating between the lines: Mississippian Migration and Stable Carbon Isotope Variation in Fort Ancient Populations. American Anthropologist 111, 344–59.

Cowan, C. W. 1987. First Farmers of the Middle Ohio Valley: Fort Ancient Societies, A.D. 1000–1670. Cincinnati, Ohio: Cincinnati Museum of Natural History.

Davis, P. E. 1976. Soil Survey of Montgomery County, Ohio. Washington, DC: U.S. Department of Agriculture.

Drooker, P. B. 1997. The View from Madisonville: Protohistoric Western Fort Ancient Interaction Patterns. Museum of Anthropology, Memoirs No. 31, Ann Arbor, Michigan: University of Michigan.

Gilbert, B. M. 1980. Mammalian Osteology. Columbia, Missouri: Missouri Archaeological Society.

Griffin, J. B. 1943. The Fort Ancient Aspect. Ann Arbor, Michigan: University of Michigan.

Hawkins, R. A. 1998. Coming full circle: Plowzone assemblages and the interpretation of Fort Ancient settlement structure, pp. 91–109 in Sullivan, A. P. (ed.), Surface Archaeology. Albuquerque, New Mexico: University of New Mexico Press.

Heilman, J. M., Lileas, M. and Turnbow, C. (eds.) 1988. A History of 17 Years of Excavation and Reconstruction: A Chronicle of 12th Century Human Values and the Built Environment: Volume I. Dayton, Ohio: Dayton Society of Natural History.

Henderson, A. G. (ed.) 1992. Fort Ancient Cultural Dynamics in the Middle Ohio Valley. Madison, Wisconsin: Prehistory Press.

Henderson, A. G. and Pollack, D. 2001. Fort ancient, pp. 174–94 in Peregrine, P. and Ember, M. (eds.), Encyclopedia of Prehistory, vol. 6. New York: Kluwer Academic/Plenum Press.

Hesselton, W. T. and Hesselton, R. A. M. 1982. White-tailed Deer, inChapman, J. A. and Feldhammer, G. A. (eds.). Wild Mammals of North America. Baltimore, Maryland: The Johns Hopkins University Press.

Kelly, R. L. 1995. The Foraging Spectrum. Washington, DC: Smithsonian Institution Press.

Kennedy, W. E. 2000. Interpreting Fort Ancient Settlement Variability. M.A. thesis, Department of Anthropology, Kent State University, Kent, Ohio. University Microfilms, Ann Arbor, Michigan.

Kennedy, W. E. 2007. Boonshoft Museum of Discovery Field School at the Wegerzyn Site. Paper presented at the Fall Meeting of the Ohio Archaeological Council, Newark, Ohio.

Ketchum, S. A., Schurr, M. R. and Garniewicz, R. C. 2009. A test for maize consumption by Fauna in Late Prehistoric Eastern North America. North American Archaeologist 30, 87–101.

Lyman, R. L. 1984. Bone density and differential Survivorship of Fossil Classes. Journal of Anthropological Archaeology 3, 259–99.

Martin, K. R. 2009. Eastern Agricultural Complex Traditions in Small Fort Ancient Communities – The Wildcat Example. M.A. thesis, Department of Anthropology, The Ohio State University, Columbus.

Metcalfe, D. and Jones, K. T. 1988. A reconsideration of animal body-part utility Indices. American Antiquity 53, 486–504.

Moorehead, W. K. 1892. Primitive Man in Ohio. New York: G. P. Putnam’s Sons.

Nagaoka, L. 2002. Explaining subsistence change in Southern New Zealand using foraging theory models. World Archaeology 34, 84–102.

Nagaoka, L. 2005. Declining foraging efficiency and Moa Carcass Exploitation in Southern New Zealand. Journal of Archaeological Science 32: 1328–38.

Olsen, S. J. 1964. Mammal remains from archaeological sites. Papers of the Peabody Museum of Archaeology and Ethnology, Harvard University vol. 56, No. 1. Cambridge, Massachusetts: Harvard University.

Pollack, D. and Henderson, A. G. 1992. Toward a model of Fort Ancient Society, pp. 281–94 in Henderson, A. G. (ed.), Fort Ancient Cultural Dynamics in the Middle Ohio Valley. Madison, Wisconsin: Prehistory Press.

Pollack, D. and Henderson, A. G. 2000. Insights into fort ancient culture change: a view from South of the Ohio River, pp. 194–215 in Genheimer, R. A., Cultures Before Contact: The Late Prehistory of Ohio and Surrounding Regions. Columbus, Ohio: Ohio Archaeological Council.

Pulliam, H. R. 1974. On the theory of optimal diets. American Naturalist 108, 59–75.

Purdue, J. R. 1983. Epiphyseal closure in white-tailed Deer. The Journal of Wildlife Management 47, 1207–13.

Purdue, J. R. 1986. The size of white-tailed Deer (Odocoileus virgi- nianus) during the Archaic Period in Central Illinois, pp. 65–95 in Neusius, S. W. (ed.), Foraging, Collecting, and Harvesting: Archaic Period Subsistence in the Eastern Woodlands. Occasional Paper 6. Center for Archaeological Investigations, Southern Illinois University-Carbondale.

Purdue, J. R. 1989. Changes during the Holocene in the Size of White-tailed Deer (Odocoileus virginianus) from Central Illinois. Quaternary Research 32, 307–16.

Purdue, J. R., Styles, B. W. and Masulis, M. C. 1989. Faunal remains and white-tail deer exploitation from a Late Woodland Upland Encampment: The Boschert Site (23SC609), St. Charles County, Missouri. Midcontinental Journal of Archaeology 14, 146–63.

Putnam, F. W. 1886. Report of the curator. Peabody Museum Annual Report 18, 401–19, 477–502.

Deppen and Cook Deer use in good times and in bad

Environmental Archaeology 2014 VOL. 19 NO. 182

Pyke, G. H., Pulliam, H. R. and Charnov, E. L. 1977. Optimal fora- ging: a selective review of theory and tests. Quarterly Review of Biology 52, 137–54.

Reidhead, V. A. 1981. A Linear Programming Model of Prehistoric Subsistence Optimization: A Southeastern Indiana Example. Prehistory Research Series Vol. 6, No. 1. Indianapolis, Indiana: Indiana Historical Society.

Rue, L. L., III 1997. The Deer of North America. Guilford, Connecticut: The Lyons Press.

Schoener, T. W. 1971. Theory of feeding strategies. Annual Review of Ecology and Systematics 2, 369–404.

Shane, O. C. 1988. An interim summary report of the vertebrate faunal remains from the Incinerator Site, pp. 157–213 in Heilman, J. M., Lileas, M. and Turnbow, C. (eds.), A History of 17 Years of Excavation and Reconstruction: A Chronicle of 12th Century Human Values and the Built Environment. Dayton, Ohio: Dayton Museum of Natural History.

Simonelli, L. M. and Kennedy, W. E. 2003. “My other site is a National Historic Landmark”: current research at the Late Prehistoric Wegerzyn Garden Center Site (33My127). Paper presented at the annual meeting of the Society for American Archaeology.

Smith, C. n.d. Field notes on file, Anthropology Department, Dayton Society of Natural History, Dayton, Ohio.

Smith, E. A. and Winterhalder, B. 1992. Evolutionary Ecology. New York: Aldine de Gruyter.

Stephens, D. W. and Krebs, J. R. 1986. Foraging Theory. Princeton, New Jersey: Princeton University Press.

Wagner, G. E. 1988. Local environment, inHeilman, J. M., Lileas, M. and Turnbow, C. (eds.). A History of 17 Years of

Excavation and Reconstruction: A Chronicle of 12th Century Human Values and the Built Environment. Dayton, Ohio: Dayton Museum of Natural History.

Wagner, G. E. 1996. Feast or famine? Seasonal diet at a fort ancient community, pp. 319–37 in Reitz, E. J., Newsom, L. A. and Scudder, S. J. (eds.) Case Studies in Environmental Archaeology. New York: Plenum Press.

Wagner, G. E. 2008. What seasonal diet at a fort ancient community reveals about coping mechanisms, pp. 277–96 in Reitz, E. J., Scudder, S. J. and Scarry, C. M. (eds.). Case Studies in Environmental Archaeology (second edition). New York: Springer.

Williams, S. 1990. The vacant quarter and other late events in the lower Valley, pp. 170–80 in Dye, D. H. and Cox, C. A. (eds.). Towns and Temples along the Mississippi. Tuscaloosa, Alabama: University of Alabama Press.

Winterhalder, B. 2001. The behavioral ecology of hunter–gatherers, pp. 12–38 in Panter-Brick, C., Layton, R. H. and Rowley- Conwy, P. (eds.). Hunter–Gatherers: An Interdisciplinary Perspective. Cambridge: Cambridge University Press.

Winterhalder, B. and Smith, E. A. 2000. Analyzing adaptive strat- egies: human behavioral ecology at twenty-five. Evolutionary Anthropology 9, 51–72.

Wolverton, S. 2008. Harvest pressure and environmental carrying capacity: an ordinal- scale model of effects on Ungulate Prey. American Antiquity 73, 179–99.

Yerkes, R. W. 2005. Bone chemistry, body parts, and growth marks: Evaluating Ohio Hopewell and Cahokia Mississippian Seasonality, Subsistence, Ritual, and Feasting. American Antiquity 70, 241–65.

Deppen and Cook Deer use in good times and in bad

Environmental Archaeology 2014 VOL. 19 NO. 1 83