- Social Sciences
- African Studies
- American Studies
- Asian Studies
- Communication Sciences
- Ethnic Studies
- European Studies
- Gender Studies
- Physical Sciences
- Life Sciences
- Animal Communications
- Cell Biology
- Evolutionary Biology
- Food Science and Technology
- Human Anatomy
White-Tailed Deer Meat and Marrow Return Rates and Their Application to Eastern Woodlands Archaeology
by Julie Zimmermann Holt, T. Cregg Madrigal
White-Tailed Deer Meat and Marrow Return Rates and Their Application to Eastern Woodlands Archaeology
Julie Zimmermann Holt, T. Cregg Madrigal
Latin American Antiquity
Updated: November 6th, 2012
WHITE-TAILED DEER MEAT AND MARROW RETURN RATES AND
THEIR APPLICATION TO EASTERN WOODLANDS ARCHAEOLOGY
T. Cregg Madrigal and Julie Zimmermann Holt
Zooarchaeological hypotheses concerning prehistoric transport, processing decisions, and social srrarijication are often tested by correlating archaeological element frequencies with indices of the economic utili~ of carcass parts. Such indices have not been developed for one of the largest and most important mammals in Eastern Woodlands prehistorq, the white-tailed deer (Odocoileus virginianus). We present kilocalorie (Kcal) yields and return rates of meat and marrow from a sample of several white-tailed deer We then compare the meat and marrow data with skeletal element abundance in two Late Archaic assem- blages from Nebv York and a Middle Woodland/early Late Woodland assemblage from Illinois. In both examples, archaeologi- cal element abundance is positively correlated with marrobv yield and return rate, but negatively correlated or uncorrelated with meat yield and return rate. These results do not provide evidence for differential transport of higher meat-yield carcass parts, but instead may indicate differential processing of high-yield marrobv bones after entire deer carcasses were transported to the sites.
A menztdo las hipdtesis :ooarqueoldgicas sobre transporte prehistdrico, decisiones de procesamiento y estratijicacidn social son contrastadas correlacionando las frecuencias de 10s elementos arqueoldgicos con 10s indices de utilidad econdmica de las partes de una carcasa. Estos indices no han sido desarrollados para el cien,o de cola blanca (Odocoileus virginianus), ztno de 10s mas grandes y el mas importante mamfero de la prehistoria de las Woodlands Orientales. Aqztipresentamos rendimientos kilocaldri- cos y tasas de retorno de carney mkdula para ztna muestra de varios cien~os de cola blanca. Luego, comparamos 10s daros sobre carne y mkdula con la abundancia de elementos esqzteletarios en dos conjuntos Arcaico Tardios de New York y ztno Woodland Medio /Woodland Tardio inicial de Illinois. En ambos ejemplos, la abundancia de elementos arqueoldgicos se correlaciona pos-
irivamente con el rendimiento y la tasa de retorno de mkdula, per0 negativamente, o no se correlaciona, con el rendimiento y tasa de retorno de carne. Estos resultados no suministran evidencia a favor del transporte diferencial de partes de mayor rendimiento de carne pero, en cambio, pueden indicar el procesamiento diferencial de huesos de alto rendimiento de mkdula despuis que las carcasas enteras de ciewos fueron transportadas a 10s sitios.
ooarchaeologists have long studied prehis- indicative of different exploitation and transport toric transport and processing decisions by strategies. examining skeletal-part profiles of prey Economic utility indices have been derived in species found at archaeological sites. The differ- recent years for a number of species, from Binford's ential representation of animal body parts has also original work with sheep and caribou (Binford 1978) been used to infer status differentiation, ritual feast- to more recent work involving, among others, bison ing, or the provisioning of elites (e.g., Bogan 1983; (Emerson 1990), odontocetes (Savelle and Friesen Jackson and Scott 1995; Kelly 1997; Welch 1991). 1996), and horse (Outram and Rowley-Conwy Hypotheses are often tested by correlating skeletal 1998). A serious omission in this body of literature element frequencies with indices of the economic has been utility indices for the most important game utility of carcass parts (e.g., Binford 1978; Metcalfe animal in the Eastern Woodlands, the white-tailed and Jones 1988; O'Connell and Marshall 1989; deer (Odocoileus virginianus). As a result, Speth 1983; Thomas and Mayer 1983). Plotting researchers attempting to explain skeletal frequen- element frequency by utility should result in dif- cies of white-tailed deer have been forced to rely on ferent utility curves (Binford 1978) that may be utility indices for caribou (e.g., Holt 1996; Kelly
T. Cregg Madrigal .New Jersey Dept. of Environmental Protection. Division of Water Quality, P.O. Box 425. Trenton, NJ 08625-0425 Julie Zimmermann Holt.Southern Illinois University Edwardsville. Department of Anthropology. Edwardsville. IL 62026
American Antiquity. 67(4), 2002, pp. 745-759
Copyright0 2002 by the Society for American Archaeology
1997; Purdue et al. 1989; Styles and Purdue 1991).
Caribou (Rangifer tarandus) and white-tailed deer are both cervids, so we might assime that their economic utility indices will be similar. However, Scott (1987) observes relatively longer limbs, espe- cially hind limbs, in smaller cervid species (includ- ing white-tailed deer) than in larger cervid species (including caribou); she attributes this to differences in habitat and locomotor habits. These differences, as well as differences in body size and behavior (e.g., Whitaker 1980), suggest that there might also be dif- ferences in the relative distribution of meat and mar- row on caribou and deer carcasses. Research on other ungulate species has clearly demonstrated that poten- tially significant nutritional variability exists both among and within species (Blumenschine and Caro 1986; Blumenschine and Madrigal 1993; Emerson 1990). There is therefore a serious need for quanti- tative data derived directly from white-tailed deer.
The extent to which a single deer carcass may be consumed is dependent on many factors, including the current nutritional needs of the hunters and the availability and desirability of alternate food sources. White-tailed deer are often small enough that it is possible for one person to transport a complete indi- vidual. The deer used in this study, for example, were transported by a single individual; the same hunter has transported adult deer unassisted for distances up to one mile. In addition, virtually all parts of a deer can be used by people (e.g., Swanton 1946:249): meat, marrow, and internal organs can be consumed; bones and antlers can be used for tools; skin can be used for clothing; and sinew and tendon can be used for binding. For these reasons, we expect that in many, perhaps most, situations hunters would trans- port all or most of a carcass away from the kill site.
There are, however, several situations where peo- ple might selectively transport only certain parts of a deer. Such differential treatment of deer parts may be related to both gross yield and return rate of deer car- cass parts. These situations may include mass kills or communal hunting of deer; hunting in contested "buffer" zones where the hunter is at risk of being attacked by members of an antagonistic group (Hick- erson 1965); when a hunter fears being attacked by, or losing the deer carcass to, large carnivores such as wolves or mountain lions; or when the distance between the kill site and campsite or habitation site is great enough that it is not efficient to transport an entire carcass (Metcalfe and Barlow 1992).
Even if an entire carcass is returned to a camp- site, different parts of the deer may be shared among people and redistributed across the site. Different individuals may have rights to different parts of a car- cass, or elites may appropriate desirable carcass parts. If intrasite spatial analysis is able to distinguish between different households, it may be possible to determine if different households had access to dif- ferent parts of deer. Therefore, the creation of util- ity indices by ranking body parts by gross yield or return rate and comparing them to skeletal element abundances at archaeological sites may be used to study many aspects of subsistence-related behavior, including social stratification or status hierarchies.
Research has demonstrated, however, that den- sity-mediated attrition of skeletal elements will affect skeletal part representation, and, consequently, the shape of utility curves (Binford 1981; Brain 1981; Grayson 1989; Lyman 1985; Marean and Spencer 1991). Therefore, utility indices cannot be used uncritically: taphonomic analysis is essential to cor- rectly understanding the processes by which bones are introduced to a site, modified, and differentially destroyed or preserved (e.g., Binford 1981; Brain 1981; Lyman 1994). Before using differences in skeletal element abundances to argue for differential transport of deer, social differences, or other prehis- toric activities, we must consider taphonomic processes that affect skeletal part frequencies.
In this paper we present results of our data col- lection on white-tailed deer. With few exceptions (e.g., Jones and Metcalfe 1988:421; Lupo 1998), utility indices have been based only on the gross yield (weight or kcal) of carcass parts, but opti- mization models require information on the net yield, or post-procurement return rate, of different nutri- ents. Therefore, we present both gross yields and return rates for meat and marrow of white-tailed deer. To demonstrate how these data might be used to make inferences regarding transport and processing decisions, we then apply these results to analyses of deer remains from Late Archaic hunter-gatherer sites in central New York and a Middle Woodlandlearly Late Woodland mound complex in the Illinois River valley.
Flesh weights and processing times were obtained from three white-tailed deer. Gross yields of marrow are based on the average Kcal yields of eight car- casses, including OV11, described below, and seven other individuals that have been previously reported (Madrigal and Capaldo 1999). This latter group includes five females (two fawns, two yearlings, and one adult) and two males (one fawn and one year- ling), all lulled in New Jersey between December and March. Marrow net yields, or post-procurement return rates, have not been reported previously, and are based on six deer (NJlOl, RU10, RU11, RU12113, RU14115, OVll) for which processing times are available.
Data on deer flesh weight and processing time were obtained from three white-tailed deer (OV11, OV12, and OV13) killed by a licensed bow hunter in New Jersey in October of 1997 and January of 1998. All age estimates are based on tooth eruption and wear (Severinghaus 1949) and epiphyseal fusion (Purdue 1983). OV 11 is a male aged 18 months and killed on October 31, 1997, in Monmouth County, New Jersey. This deer had a total dressed body weight (i.e., weight after evisceration) of 54.3 kg (1 19.5 lbs) and a thick subcutaneous fat layer. OV12 is a small female fawn aged six to eight months and killed on January 3, 1998, in Hunterdon County, New Jersey. This deer had a total dressed weight of 18.6 kg (41 lbs). OV13 is a six-to-nine-month-old male fawn killed January 10, 1998, in Hunterdon County, New Jersey. OV13 had a total dressed weight of 37.2 kg (82 lbs). Meat weight and processing time were recorded for all three carcasses; wet bone weight was recorded for OV 1 1 and OV 13; and marrow weight and processing time were recorded for OV11.
The hunter eviscerated the three deer (OV11, OV 12, and OV13) at the kill sites and then transported the carcasses for further processing. Internal organs were not weighed before being discarded. The carcasses were suspended by their hind legs with the body cav- ity propped open and left to cool overnight. The fol- lowing day the hunter, an experienced butcher, skinned and then butchered the deer with a steel knife following a method learned during previous hunting seasons. As the hunter butchered, the authors recorded processing times (i.e., the time it took to remove each meat cut or bone), meat cut weights, and bone weights. Processing times were recorded to the nearest second using a digital stopwatch, and meat cuts and bones were weighed to the nearest 2 g using an electronic scale.
All usable meat was cut from the bone, except for small scraps of meat left on the vertebrae and innom- inate that were difficult to remove. Muscles were removed individually in order to eliminate unpalat- able tendons and fascia (prior to wrapping and freez- ing the meat), and elements were removed by taking advantage of the relative weaknesses of different joints.
The hunter first removed the loins, which lie along the dorsal side of the deer's thoracic and lumbar ver- tebrae, and then the hams, the meat from the femora. These cuts were removed first because they are con- sidered the best cuts of meat, in terms of personal preference, and are also the meatiest cuts (Table 1). He then removed the meat from the forelimbs, after which he disarticulated the forelimb (including scapula) from the rest of the carcass and set it out of the way. The hunter then removed the meat from the rest of the carcass, beginning with the neck and work- ing his way down to the hind legs.
We assume that this method is essentially simi- lar to prehistoric methods, given that it is dictated mainly by the anatomy of the animal (cf. Hill 1979). Some deviation from this order would probably have little effect on processing times. The processing times may be shorter in absolute terms since a steel knife was used; a stone knife might require more frequent resharpening. Future research should test these assumptions and examine potential differences in processing time when using steel versus stone knives. The major differences between the methods used in this study and those used by prehistoric hunters are most likely that the head and internal organs were discarded, instead of being processed for food.
The metapodials and feet contain virtually no meat, and so no meat weights were recorded for these parts. In addition, meat weights for the lumbar ver- tebrae are underestimated while those for the tho- racic vertebrae are overestimated. The back strap or loin (the longissirnus thoracis and the longissirnus lumborum) runs the length of both the thoracic and lumbar vertebrae on their dorsal sides; however, we arbitrarily assign this cut of meat to the thoracic ver- tebrae. The meat weights assigned to the lumbar ver- tebrae are from the tenderloin (the psoas major), which is attached to the ventral side of the lumbar vertebrae.
Kcal yields were derived from meat weights in order to permit comparisons with other types of food for which energy yields have been reported, and
Table 1. Meat Weight (g), and Processing Time (sec) of White-Tailed Deer
Scapula Humerus Radio-ulna Femur Tibia Cervical Thoracic Lumbar Rib PelvisiSacrum Total
because Kcal yield is a commonly used currency in optimal foraging studies. The energy content of deer flesh was determined by McCullough and Ullrey (1983) to be 6.09 KcaVg. Therefore, meat weights were multiplied by 6.09 to obtain the Kcal yield. Return rates were calculated for each individual by dividing the Kcal yield for each element by the aver- age processing time for each element. The return rates for the left and right sides of each individual were then averaged.
Marrow-processing procedures have been described previously (Blumenschine and Madrigal 1993; Madrigal and Capaldo 1999). A hammer stone on anvil technique (Blumenschine 1988) was used to fracture long bones and mandibles. Processing time (in seconds) was recorded for periosteum scraping (when applicable), hammer stone fracture, and mar- row extraction. Marrow from each bone was weighed to the nearest .O1 g using an electronic balance, except for OV11, which was weighed on an elec- tronic balance accurate to 2 g. We refer to this mea- surement as marrow wet weight.
Marrow is composed of fat, water, and a small, constant amount of nonfat cell residue. The fat con- tent of marrow can be obtained by oven drying (Nei- land 1970) until all water is removed. All marrow from each bone was placed in a small metal pan in an oven heated to 200 degrees Fahrenheit. Marrow was periodically weighed during this procedure. When there was no weight change for three consec- utive weighings, a final reading was taken. This is the marrow dry weight (fat plus nonfat cell residue) and is used to determine Kcal yields.
OV12 OV13 Average
Because left and right values for each animal were similar, wet weights, dry weights, and Kcal yields were obtained by averaging the left and right sides for each animal. In cases where marrow was obtained from only a single bone, that value was used instead of an average. The amount of marrow fat was deter- mined by multiplying the dry weight by .93 to account for the presence of nonfat cell residue (Blu- menschine and Madrigal 1993; Brooks et al. 1977; Neiland 1970). The energy content of white-tailed deer marrow fat was determined by McCullough and Ullrey (1983) to be 9.37 Kcallg. Therefore, the mod- ified dry weight was multiplied by 9.37 in order to obtain the Kcal content. Return rates were calcu- lated using the methods used for meat.
Average meat weights and processing times for the three deer are presented in Table 1. The femur has the highest average gross weight, followed by the tho- racic vertebrae, ribs, cervical vertebrae, and scapula. The other long bones are among the lowest-yielding elements. After the femur, the second highest-ranked long bone is the tibia, which has an average weight that is only about one-fifth as great as the femur. The tibia is followed by the humerus and radio-ulna. The lumbar vertebrae are the lowest-ranked body part, but this is due to the fact that only the tenderloin was attributed to the lumbar vertebrae. The amount of meat associated with the heads and feet is relatively insignificant and therefore was not weighed.
Despite differences in the age, sex, and total dressed weights of the three deer, meat weights-
Table 2. Meat Kilocalorie (Kcal) Yield and Return Rate (Kcalmr) of White-Tailed Deer. OVll OV12 OV13 Average
and hence Kcal yields-are
strongly and signifi-
cantly correlated among the three individuals. Cal- culation of Pearson's correlation coefficient shows that the strongest correlation is between OV12 and OV 13 (5= 38;p =.0007), but correlations between OV 1 1 and OV 13 (r,, = 33; p = ,003) and OV 1 1 and OV12 (rp = .77;p = .009) are also very strong. The femur is the body part with greatest gross yield in all cases, and the five highest-ranked parts in all three deer are the same (i.e., the femur, thoracic vertebrae, ribs, cervical vertebrae, and scapula). There are, of course, major differences in gross weight among the deer. Meat from the femur of the largest deer (OV 1 I), for example, weighs more than twice that of the smallest deer (OV12), while OV13 is intermediate between the two.
Variation in the ranking of parts by return rate is greater than the variation in ranking of parts by gross yield (Table 2). In terms of return rate, the highest- ranked elements for OV 1 1 are the thoracic vertebrae, followed by the pelvis/sacrum, the radio-ulna, femur, and lumbar vertebrae. The highest-ranked elements for OV12 are the femur, cervical vertebrae, tibia, thoracic vertebrae, and scapula. The highest-ranked elements for OV13 are the pelvis/sacrum, tibia, femur, thoracic vertebrae, and scapula.
Some of this variation in rank of gross yields and return rates may be due to slight differences in the way the three deer were butchered and, for return rates, differences in the speed of butchery of partic- ular parts of different deer. Differences in sex and age of the deer presumably are also a factor, but given the small sample size, it is premature to try to attribute specific differences to age or sex at this time.
Ranking of body parts based on the average return rates (Table 2) differs from the ranking based on average gross yield (see Figure 1). To reiterate, the five highest-ranked parts based on average gross yield are the femur, thoracic vertebrae, ribs, cervical vertebrae, and scapula. The five highest-ranked parts based on average return rate are the thoracic verte- brae, femur, pelvis/sacrum, cervical vertebrae, and radio-ulna. Notably, the rib cage, which has the third- highest average gross yield, has the lowest average return rate. The radio-ulna has the fifth-highest aver- age return rate but has the second-lowest average gross yield. Differences between average gross yield and return-rate rankings are also somewhat disparate for the lumbar vertebrae and for the pelvis/sacrum, both of which have a higher rank in terms of return rate than in terms of gross yield. For all other ele- ments, average gross yield and return-rate rankings are identical (cervical vertebrae) or nearly identical (scapula, humerus, femur, tibia, and thoracic verte- brae).
Marrow weight, Kcal, and processing times for OVl 1 are presented in Table 3. The return rates for OVll and the average return rates for six deer (including OV11) are presented in Table 4. The tibia has the highest average gross marrow yield, followed distantly by the femur and radius. The humerus has only the fifth-highest gross yield, but the third-high- est return rate.
An earlier study (Madrigal and Capaldo 1999) of seven deer found that the ranking of long-bone gross marrow yields is consistent among individuals. The tibia has the highest Kcal yield, followed by the femur and radius. The relative ranking of the humerus and metatarsal does vary by individual. It appears that the humerus marrow yield may be higher than the metatarsal yield in individuals that are relatively
Element Figure 1. a) Average meat Kcal yield by element; b) average meat return rate (KcaVhr) by element.
unstressed. Marrow gross yield from OV11 is simi- lar to the other seven deer in that the tibia and femur have the two highest yields. However, based on dry marrow weight, the humerus, radius, and metatarsal have identical yields.
The tibia also has the highest average return rate, followed by the femur and humerus. Radius and metatarsal return rates are almost identical, while the mandible has ahigher return rate than the metacarpal. First and second phalanx return rates are very low compared to all other elements.
While the correlation between gross yields and return rates of meat is not significant (rp= .49;p > .l; see Figure I), gross yields and return rates of marrow are highly correlated (rp= .91;p = .0001;
see Figure 2). A likely explanation for this is that whereas there is relatively little variation in the amount of time needed to break open a particular bone for marrow, there is more variation in the meat- processing times of different body parts. Meat yields and return rates might also be uncorrelated due to small sample size. We postulate that as more deer are added to this sample, variation will be seen to be continuous and individual differences will appear less striking.
Data presented here give some indication of varia- tion in white-tailed deer meat and marrow yields. The information on marrow yields suggests that while
Table 3. Marrow Weight (g), Processing Time (sec). and Yield (Kcal) of OV11.
|I st Phalanx, ant.||1||1||252||8.70||8.70|
|2nd Phalanx, ant.||1||1||30 1||8.70||8.70|
Table 4. Marrow Return Rate (KcalIHour) of OV 11 and Average Marrow Return Rate of Eight Deer.
OV 1 1 (Kcallhr) Combined Average
Element Dry Wet Kcal (dry) Kcallhr Humerus 1,702.8 2,226.7 46.1 3.163.2 Radius 1.268.5 1,463.6 63.0 1,824.4 Metacarpal 812.2 913.7 30.1 997.8 Femur 3,226.7 3,704.7 91.5 4,844.8 Tibia 2.886.4 3.342.2 171.5 5.498.4 Metatarsal 1.724.4 1,989.7 52.6 1,822.3
1st Phalanx. ant. 124.3 124.3 5.3 163.2 2nd Phalanx. ant. 104.1 104.1 4.6 104.2 Mandible 10.4 1,122.5 Note: Includes OV11 and seven deer from Madrigal and Capaldo 1999; based on dry weights of marrow. Combined average marrow return rate (Kcalhr) is based on specimens with complete time data (NJlO1, RU10, RUl1, RU12113. RU14115, OV11). Marrow Kcal yields for phalanges and mandible is from fewer than 8 carcasses. The gross marrow yield used here dif- fers from the average Kcal figures in Madrigal and Capaldo (1999:Table 5) because those averages are based on both wet and dry weights (i.e., 14 data points), while combined average is based only on dry marrow weights (i.e., 7 data points) plus the dry marrow weight of OV11 (8 total data points). See Madrigal and Capaldo (1999) for additional details on marrow weights.
larger deer will have more marrow than smaller ones, absolute yield and return rate between white-tailed the relative ranking of different elements does not deer and the larger caribou is great. These data are vary greatly. Meat yields are only available for one particularly important for studies that compare the yearling and two fawns, but there is some consistency relative importance of different types of animal and in the relative ranking of different elements. Adult plant foods in the diet. deer will obviously be larger, and we might expect Deer marrow return rate is positively and signif- some changes in the relative weights of different icantly correlated (rp = .90; p = ,0009) with caribou body parts as the muscles develop, strengthen, and marrow return rate as estimated by Jones and Met- age. calfe (1988:42 1). Caribou Kcallhr estimates are,
Deer meat Kcal yield is positively and signifi- however, much lower than deer return rates for the cantly correlated (rp = .78; p = ,008) with caribou same elements, despite the larger body size of cari- Food Utility Index (FUI; Metcalfe and Jones 1988), bou. For example, caribou tibia return rate is esti- and deer marrow Kcal is positively and significantly mated to be 1670 Kcallhr, compared to 5,498.4 correlated (rp = .91; p = ,002) with caribou marrow Kcallhr for the same element in deer (Table 4). Jones cavity volume (Binford 1978). These results appear and Metcalfe estimated return rates by "multiplying to indicate that the relative importance of elements the marrow cavity volume by a constant of 4 Kcal./rnl within a single animal is consistent between white- . . . and dividing that by the mean processing time tailed deer and caribou (both are cervids), although for each bone (Binford, 1978:26, table 1.7)" (Jones additional data from caribou would be desirable. It and Metcalfe 1988:421422). Our results suggest is important to note also that the difference in that this method greatly underestimates the actual
Element Figure 2. a) Average marrow (dry) Kcal yield by element; b) average marrow return rate (KcaVhr) by element.
caloric yield of caribou marrow, although the rela- tive ranking of caribou elements is similar to that of deer.
Because there can be significant nutritional vari- ability within a species depending on sex, age, and season of death (Blumenschine and Caro 1986; Blu- menschine and Madrigal 1993), in the future we hope to increase our sample to include adults and deer killed during different seasons. In the Northeast, sea- sonal weight loss in females and all fawns begins in January, while adult males, due to their increased activity and reduced food intake during the autumn rut, lose weight earlier. In New York, female deer weights are at a minimum in April; weight gain resumes in May with females reaching their maxi- mum weight in November and December (Madrigal 1999:294-295; Moen and Severinghaus 1981; Sauer 1984; Verme and Ullrey 1984). Nevertheless, our sample does provide an idea of the range seen in the modem deer population in New Jersey. Based on New Jersey Department of Environmental Protection condition parameters, the two males (OV11 and OV13) in our sample are substantially above stan- dard for their respective age and sex classes while the female deer (OV12) is below standard (New Jer- sey Department of Environmental Protection 2001).
Immature deer are more vulnerable to predation than most adults, and fawns and yearlings numeri- cally make up the majority of individuals in any sta- ble deer population (e.g., Emerson 1980).We suspect that in many cases prehistoric hunters may have killed fawns and yearlings more frequently than the less-numerous and more wary adult deer (see also Munson 1991). Furthermore, while deer can be hunted year-round, historic, ethnographic, and archaeological evidence indicates that deer procure- ment was often concentrated in the fall and winter months (e.g., Hudson 1976;McCabe and McCabe 1984; Swanton 1946).Thus our sample, despite its small size, is probably representative of the major- ity of deer killed by prehistoric hunters in terms of age and season of death.
The Late Archaic in New York
The data on white-tailed deer gross yields and return rates are applied to two Late Archaic sites in central New York. Both sites were occupied for multiple seasons by hunter-gatherers whose diet varied sea- sonally and included deer, passenger pigeon, tree squirrel, turkey, bullhead catfish, and other species (Madrigal 1999).The Lamoka Lake site (Hpt. 1) is located on the shore of a short stream connecting two small lakes in central New York. It dates to about 2500 B.C. and has been the subject of several exca- vations over the past seven decades (Gramly 1983; Madrigal 1999; Ritchie 1932, 1969). Deer bones used in this study are derived from William Ritchie's and Harrison Follett's original excavations at Lam- oka Lake in the 1920s (Ritchie 1932, 1969) and are curated by the Rochester Museum and Science Cen- ter. Detailed information on the original context of these bones is not available, but they are thought to be derived primarily from hearths, pits, and fire beds (see Madrigal 1999:74-90;Ritchie1932,1969).Sediments were not screened. The composition of the assemblage suggests that all deer bone fragments observed during excavation were collected, but very small elements, such as some carpals and tarsals, were not recovered (see discussion in Madrigal 1999).White-tailed deer is the most common species, represented by 1,042of the 3,758 faunal specimens and at least 37 individuals.
The Cole Gravel Pit (Hne 17-1) is located on a terrace above the Genesee River, south of the city of Rochester. Nearly 300 features were excavated by the Rochester Museum and Science Center in the 1960s. Most of the features are interpreted as earth ovens, although a small number of features may have been used for food storage (Madrigal 1999:95-100). Based on two radiocarbon dates, the site dates to approximately 1900 B.C. (Hayes and Bergs 1969). Of the 296 features excavated, 176 contained animal bone. Feature fill was screened through quarter-inch mesh. A total of 16,180 bones were examined, of which 8,174,or 50.5 percent, were identified to class (Madrigal1999; see also Brown et al. 1973). Deer is one of the most common species, consisting of 691 specimens and at least 20 individuals. An additional
109 specimens, mainly antler fragments, were iden- tified only as cervid, but most are probably also from deer, rather than wapiti (Cewus elaphus).
Long-bone Minimum Animal Units (MAU) for these sites (Table 5) was calculated using both epi- physes and midshafts, thereby minimizing the effects of density-dependent destruction (Blumenschine and Marean 1993; Bunn and Kroll 1986; Marean and Frey 1997; Marean and Spencer 1991). MAU was determined by first calculating the Minimum Num- ber of Elements (MNE), then dividing the MNE by the number of elements in a single individual. There- fore, long-bone MNE was divided by two, phalanx MNE was divided by eight, and cervical (excluding atlas and axis) MNE was divided by five.
When MNE counts based on epiphyses and shafts are used, the Lamoka assemblage MNE is not cor- related (r,, =.07;p=,352)with bone density (Lyman 1984; 1994), while the Cole Gravel Pit assemblage has a positive correlation that is not significant at the .05 level (rp = .27;p= ,096).Marean and colleagues (Marean and Spencer 1991 ;Marean et al. 1992)have demonstrated experimentally that long-bone epi- physes and axial elements are preferentially destroyed by scavenging hyenas, but long-bone mid- shaft fragments will preserve nearly 100 percent of the original number of elements. Examination of long-bone shaft fragments, despite being more dif- ficult to identify to element than epiphyses, may therefore provide a more accurate record of the rel- ative frequency of elements discarded by humans. Correlations of long-bone epiphysis MNE only, and shaft MNE only, with bone density (Lyman 1984) indicate that there is evidence at Lamoka Lake and Cole Gravel Pit of density-dependent destruction of epiphyses, but not of shaft fragments (Madrigal
|Table 5. MAU of Deer Elements from Lamoka Lake, Cole Gravel Pit. and Baehr-Gust.|
|Lamoka RMSCa||Cole Gravel Pit||Baehr-Gust|
|Element Scapula Humerus||MAU 21.5 33.0||8MAU 15.5 23.9||MAU 6.0 15.0||L7c MAU 9.4 23.5||MAU 6.5 10.0||8MAU6.2 9.6|
Radius 12.0 8.7 Ulna 15.5 11.2 Metacarpal 2.5 1.8 Femur 7.5 5.4 Tibia 9.5 6.9 Metatarsal 2.5 1.8 First phalanx 1.9 1.4 Second phalanx .9 .7 Mandible 16.0 11.6 Cervical 1 .0 .7 Thoracic .5 .4 Lumbar 1 .0 .7 Rib .5 .4 Innominate 12.5 9.0
"Rochester Museum and Science Center.
1999: 159-1 67, 253-255). Therefore, whole-bone MNEs and MAUs that are derived from both epiph- ysis and shaft fragments should be relatively immune to the effect of density-dependent destruction (Madri- gal 1999).
While there is clear evidence of carnivore tooth marks on deer bones from both sites, the location and frequency of percussion marks and carnivore tooth marks indicate that marrow was removed by humans, after which dogs or other carnivores scavenged the bones (Madrigal 1999).
In performing the following correlations, ulna MAUs were not considered in the correlations with marrow yields. The combined radius and ulna meat value was correlated with both radius and ulna ele- ment frequency. Metapodials were not included in any of the meat yield correlations because they do not have any associated flesh.
There is a statistically insignificant negative cor- relation between MAU and both meat gross yield (Lamoka: r = -.36;p = .28; Cole: r = -.l 1 ;p = .74)
and return rate (Lamoka: r = -.24;p = .47; Cole: rP
= -.20;p = .56) in both assemblages. However, two factors complicate the issue. First, vertebrae (except for axis and atlas) and rib fragments are difficult to identify to specific element, which makes it difficult to accurately estimate the total number of elements present. This means that MAUs most likely under- estimate the abundance of these elements. Second, density-mediated taphonomic attrition is known to preferentially destroy the less-dense vertebrae and
- 7.1 16.5 15.8
- 7.8 11.5 11.0
2.5 3.9 3.0 2.9
8.0 12.5 7.0 6.7
8.0 12.5 13.0 12.4
1 .5 2.4 6.5 6.2
1.5 2.4 4.9 4.7 .8 1.3 3.5 3.3
5.5 8.6 5.5 5.3 .6 .9 .8 .8 .2 .3 1.6 1.5 .5 .8 2.0 1.9 .2 .3 1.2 1.1
4.0 6.3 11.0 10.5
ribs (Binford and Bertram 1977; Brain 198 1 ;Klippel et al. 1987; Lyman 1984, 1992; Marean et al. 1992; Morey and Klippel 1991). At both Lamoka Lake and Cole Gravel Pit, most axial fragments were likely destroyed by carnivore ravaging, therefore fur- ther biasing the estimates of element abundance. To compensate for this, correlations were also calculated using only long-bone (femur, tibia, humerus, radius) and scapula MAU and meat yields, because long- bone MAUs are based on both shafts and epiphyses, and therefore are not biased by carnivore ravaging (Marean and Spencer 1991; Marean et al. 1992). Meat gross yield is negatively but insignificantly cor- related with Lamoka Lake long-bone and scapula MAU (rl, = -.37; p = .48) and is uncorrelated with Cole Gravel Pit long-bone and scapula MAU (rl, = .07; p = .9). Meat return rate is also negatively but insignificantly correlated with both long-bone and scapula MAU from both sites (Lamoka: rP = -.59;p = .22; Cole: rP = -.36;p = .49). These results do not support the proposition that elements with high meat yield were preferentially transported to the site.
In contrast, Cole Gravel Pit MAU is insignifi- cantly correlated with marrow gross yield (rl, = .42; p = .27) and significantly correlated with marrow return rate (rp = .67; p =.05). Lamoka MAU is insignificantly correlated with both marrow return rate (r = .34;p = .37) and marrow gross yield (rp =
10;p = .79). Positive correlations between long-bone element abundances and marrow yields at both sites may
show more direct evidence of processing for mar- row after bones had been brought back to the site than they do about the original hunting and transport behaviors that preceded marrow processing. This echoes suggestions by Marshall and Pilgram (1991) that within-bone nutrients may be more important than meat yields in determining element abundances. In summary, based on skeletal element representa- tion, there is no evidence of differential transport of high-yield meat bones to the Cole Gravel Pit or Lam- oka Lake. In contrast, bones with higher marrow gross and return rates are more abundant at both sites and are more fragmented. This does not mean that high-yield marrow bones were preferentially trans- ported to the site. Instead, it is more likely that skele- tal-part profiles are indicative of the preferential processing of high-yield marrow bones after entire deer carcasses were transported to the site.
The Middle to Late Woodland Transition in Illinois
The data on white-tailed deer meat and marrow yields and return rates are next applied to a Middle and early Late Woodland site in Illinois. The inhabitants of the Illinois Valley at this time were horticulturalists, cul- tivating a variety of native plants, yet they remained heavily dependent on hunting and gathering. The Baehr-Gust site (1 lBR2; formerly known as the Baehr site) is located at the juncture of the lower and central Illinois Valleys, west of the river and south of the town of Beardstown. The site was first reported in the 1890s by John Francis Snyder, who recorded five mounds there (Snyder 1962). Snyder excavated several of these mounds, encountering burials and artifacts typical of Havana Hopewell in the Illinois Valley. Howard Winters of New York University con- ducted excavations at the site from 1987 through 1993. These excavations took place in habitation areas adjacent to the mounds, encountering over 30 pit features. Bulk charcoal samples from 13 of the pit features produced radiocarbon dates ranging between 50 B.C. and A.D. 530 (Holt 2000:Table 4.7). These dates indicate that the site was occupied from the Middle Woodland period through the early Late Woodland period, with the heaviest occupation occurring during the early Late Woodland White Hall phase.
In order to maximize sample size, white-tailed deer bones excavated from across the site by New York University personnel are considered here as a single sample. The deer remains analyzed here were recovered using quarter-inch mesh from pit features and from nonfeature deposits that were midden or possibly mound fill. For analysis of the Baehr-Gust deer bone with finer chronological and spatial con- trol, see Holt (2000, 2002).
Deer was the most important species consumed at the site throughout its occupation. There is, how- ever, a shift in the importance of deer through time, as fish (particularly bullheads and bowfins) became a much more important resource during the early Late Woodland period (Holt 2000). There were few absolute indicators of seasonality in either the Mid- dle Woodland or the Late Woodland faunal sample from Baehr-Gust. The presence of a deer aged less than five months indicates either summer or fall occu- pation during the Middle Woodland period. The pres- ence of a deer aged 23-29 months and a deer with intact antlers indicates either summer or fall occu- pation during the Late Woodland period. The abun- dance of fish remains suggests warm weather occupation during both the Middle and Late Wood- land periods (Holt 2000).
MAUs were calculated for the Baehr-Gust deer bone using the methods outlined above: both epi- physes and shafts were used to calculate MNE, then MNE was divided by the number of elements in an individual to determine MAU (Table 5). Correla- tions between MAU and meat and marrow yields were then performed using the same procedures described above.
Element abundance at Baehr-Gust is not signifi- cantly correlated with either meat gross yield (rp = -.48; p = .13) or return rate (rp = -.08; p = 31). A rel- atively large number of high-meat-yielding femurs is primarily responsible for the insignificance of the negative correlation between MAU and meat gross yield. When femurs are removed, the negative cor- relation becomes significant (rp = -.71; p = .02). When the vertebrae and ribs are removed (see above), results are essentially unchanged: MAU is not sig- nificantly correlated with either meat gross yield (rp = -.67; p = .14) or meat return rate (r = -.33; p =
Biases of identification and preservation alone do not explain the lack of preference for high-meat- yielding bone. Element abundance at Baehr-Gust may instead indicate a preference for high-marrow- yielding bone. MAU is positively correlated with both marrow gross yield (rD = .63;p = .07) and mar- row return rate (r = .53;p = .14). Neither correla-
tion is significant at the .05 level, primarily due to a relatively large number of radii. When radii are removed, the correlations become significant (for gross yield, rp = 23;p =.Ol; for return rate, rp = 34; p = .01). We again conclude from this, as we did in the New York example above, that within-bone nutri- ents are more important than meat yields in predict- ing archaeological element abundance.
Holt (2000,2002) analyzed the Baehr-Gust mate- rials for density-mediated attrition by testing for cor- relation between MAU and deer bone density (Lyman 1984). No correlation was found between MAU and bone density for two Middle Woodland samples (r = -.09; p = .64; r = -.02; p = .93) or for
one early Late Woodland sample (rp = .21;p = .25), but MAU and bone density were positively correlated (rp= .54; p = .01) for one undated sample. We con- clude that, for the most part, preservation conditions little affected element abundance at Baehr-Gust.
In an analysis comparing animal remains from Baehr-Gust with animal remains from other Middle Woodland sites in the Illinois Valley, it was hypoth- esized that nonelites who inhabited nonmounded sites provided venison to elites who inhabited mound centers such as Baehr-Gust (Holt 2000,2002). This hypothesis was tested by looking for negative cor- relations between element abundance and meat yieldslreturn rates at nonmounded (producer) sites, and by looking forpositive correlations between ele- ment abundance and meat yieldslreturn rates at mounded (consumer) sites. However, no correlation was found between element abundance and either meat gross yields or meat return rates at any of the sites analyzed. Thus, this analysis provided no evi- dence to support the hypothesis that elites at mound centers were provided with venison by nonelites at nonmounded sites.
Marrow yields and return rates are positively corre- lated with element abundance while meat yields and return rates are uncorrelated at both the New York and Illinois sites. This, however, does not necessar- ily suggest that marrow consumption was more important than meat consumption at any of these sites, and we do not claim that bone transport deci- sions were determined solely, or even primarily, by marrow yields. In fact, we think it likely that in all three cases entire deer carcasses, including both meat and marrow, were transported from the kill site to the habitation site.
Why, then, do correlations suggest a preference for high-marrow-yielding bone? The explanation may be found in the cultural and natural processes that affect deer bones from the time the deer is first obtained by hunters. If a whole deer were brought to a living site, the animal would first be skinned and butchered: most meat would be removed from the bones to be distributed, eaten, or preserved, thereby disappearing from the archaeological record. Some bones are saved and used to make tools or other arti- facts, but most are likely to be discarded once meat and other nutrients have been removed.
Vertebrae and ribs have high meat yields, but they contain no marrow, are not particularly useful for tool manufacture, and have low structural density. If not processed for their bone grease, vertebrae and ribs would be discarded immediately, perhaps by dump- ing them at the site's periphery or in secondary dis- posal areas where they might be destroyed or removed by scavenging dogs or other carnivores.
Bones with low marrow yields will be fractured by humans less often than high-yield marrow bones. Whether the low-yield marrow bones from any given deer carcass would be exploited for their marrow is presumably related to the need for marrow fat and the availability of other foods. Unfractured low-yield bones, since they would retain marrow grease, would remain attractive to scavenging carnivores. Unbro- ken low-yield bones, therefore, are more likely to be completely destroyed or removed from the site by scavengers than high-yield bones (cf. Blumenschine 1988).
Bones with high marrow yields are most likely to be fractured by humans to remove the marrow. The epiphy seal fragments may be rendered unidentifiable by additional fragmentation for grease processing. If not processed to remove bone grease, epiphyseal fragments retain some nutritional value and may be completely destroyed or removed from the site by dogs or other scavenging carnivores. In contrast, long-bone shaft fragments are not attractive to car- nivores, and hence are more likely to survive in the archaeological record. The end result, in archaeo- logical terms, will be a faunal assemblage with a greater proportion of high-yield marrow bone frag- ments than low-yield ones. If correct, this implies that bone-element frequencies provide more direct evi- dence of marrow processing decisions (which occur
after carcasses have been transported) than they do of bone transport decisions.
We stress that sites from different regions and dif- ferent time periods, and with different modes of sub- sistence, appear to have very similar patterns of element representation. In this paper we have dis- cussed sites separated by several thousand years and nearly a thousand miles. Late Archaic inhabitants of New York were strictly hunter-gatherers, while Mid- dle-to-Late Woodland inhabitants of the Illinois Val- ley were also dependent to some extent on plant cultivation. The similarity in element representation at these sites is most likely due both to the impor- tance of marrow as a source of fat and calories, and to taphonomic processes that structure faunal assem- blages. We predict that similar patterns will be found in other faunal assemblages oriented toward wild game.
We should also reiterate that although our sam- ple of three modem deer measured for meat yields is small, not only is it larger than the caribou-based index most frequently used, but also, because we have calculated both Kcal and Kcanour yields, these data are more suitable for comparison with other animal and plant foods in optimal foraging studies. Moreover, our sample might accurately reflect a typ- ical prehistoric assemblage, assuming that immature deer were the most abundant and the most vulnera- ble age class prehistorically, and that deer hunting primarily took place in the fall and winter. Never- theless, we hope in the future to add mature deer and deer taken during the spring and summer to this sam- ple, and we encourage other researchers to do the same.
Acknowledgments. First and foremost, we owe great thanks to Henry Holt for providing us with specimens OV11, OV12, and OV13; for butchering them; and for sharing with us his knowledge of deer anatomy, behavior, and hunting techniques. Madrigal's work was funded in part by the National Science Foundation (Dissertation Improvement Grant SAR 95-22828) and by grants from Rutgers University and the Rochester Museum and Science Center. Julie Holt's work was funded in part by dissertation grants from Wenner-Gren (Gr. 5881) and the National Science Foundation (SBR-9704018). Thanks also to Brian D. Darrow, D.V.M., for providing us with the proper muscular nomenclature, and to Guillermo Luis Mengoni GoAalons for translating the abstract into Spanish. Lee Lyman, Rick Purdue, James Savelle, and anony- mous reviewers provided helpful comments on earlier versions of this paper.
Binford, L. R. 1978 Nunami14t Ethnoarchaeology. Academic Press, New York. 1981 Bones: Ancient Men and Modern Myths. Academic Press, Orlando, Florida. Binford, L. R., and J. B. Bertram
1977 Bone Frequencies-and Attritional Processes. In For Theory Building in Archaeology, edited by L. R. Binford, pp. 77-153. Academic Press, New York.
Blumenschine, R. J. 1988 An Experimental Model of the Timing of Hominid and Carnivore Influence on Archaeological Bone Assemblages. Joltnlal ofArchaeologica1 Science 15:483-502.
Blumenschine, R. J., and T. M. Caro 1986 Unit Flesh Weights of Some East African Bovids. African Journal of Ecology 24:273-286. Blumenschine, R. J., and T. C. Madrigal
1993 Variability in Long Bone Marrow Yields of East African Ungulates and its Zooarchaeological Implications. Journal ofArchaeologica1 Science 20:555-587.
Blumenschine, R. J., and C. W. Marean 1993 A Carnivore's View of Archaeological Bone Assem- blages. In From Bones to Behavior: Ethnoarchaeological and Experimental Contributions to the Interpretation ofFa~tnal Remains, edited by J. Hudson, pp. 273-300. Center for Archaeological Investigations, Carbondale, Illinois.
Bogan, A. 1983 Evidence for Faunal Resource Partitioning in an East- em North American Chiefdom. In Animals and Archaeol- ogy Vol. I, Hunters and Their Prey, edited by C. Grigson and J. Clutton-Brock, pp. 305-324. British Archaeological Reports, International Series 163, Oxford.
Brain, C. 1981 The Hunters or the Hunted? An Introduction to African Cave Taphonomy. University of Chicago Press, Chicago. Brooks, P. M., J. Hanks, and J. Ludbrook 1977 Bone Marrow as an Index of Condition in African Ungu- lates. South African Jo~tn~al
of Wildlife Research 7:6146. Brown, C., J. Kelley, J. Penman, and J. Sparling 1973 Faunal Analysis of the Cole Quany Archaic Site. Bulletin of the New York Stare Archeological Association 58:2540.
Bunn, H. T., and E. M. Kroll 1986 Systematic Butchery by Plio/Pleistocene Hominids at Olduvai Gorge, Tanzania. Current Anthropology 27:431452.
Emerson, A. 1990 Archaeological Itnplications of Variability in the Eco- nomic Anatomy of Bison bison. Ph.D. dissertation, Depart- ment of Anthropology, Washington State University. University Microfilms, Ann Arbor.
Emerson, T. E. 1980 A Stable White-tailed Deer Population Model and its Implications for Interpreting Prehistoric Hunting Patterns. Mid-Continental Journal ofArchaeology 5(1): 117-1 32.
Gramly, R. M. 1983 Below-the-Watertable Archaeology at Lamoka Lake. North American Archaeologist 4(2): 127-1 39. Grayson, D. K. 1989 Bone Transport, Bone Destruction, and Reverse Utility Curves. Journal ofArchaeologica1 Science 16:643-652. Hayes, C. F., 111, and L. Bergs 1969 A Progress Report on an Archaic site on the Farrell Farm: The Cole Gravel Pit 1966-1968. Bulletin ofrhe New
York State Archeological Association 47: 1-1 1. Hickerson, H.
1965 The Virginia Deer and Intertribal Buffer Zones in the Upper Mississippi Valley. Publication 78. In Man, Culture, andAnimal.7, edited by A. Leeds, and A. Vayda, pp. 4345. American Association for the Advancement of Science, Washington, D.C.
Hill, A. 1979 Butchery and Natural Disarticulation: An Investigatory Technique. American Antiquity 44:739-744. Holt, J. Z.
1996 AG Church Site Subsistence Remains: The Procurement and Exchange of Plant and Animal Products during the Mis- sissippianEmergence.IllinoisArchaeologj 8(1-2): 146-180.
2000 Animal E.wploitation and the Middle to Late Woodland Transition: A Comparison ofAnimal Use at Mound Centers and Hamlets in the Lower and Central Illinois Valleys. Ph.D. dissertation, Department of Anthropology, New York Uni- versity. University Microfilms, Ann Arbor.
2002 The Havana Social Structure: Testing Winters' Model with Faunal Remains from the Baehr-Gust Site. In Papers in Memon of Howard Dalton Winters, edited by A.-M. Cantwell and L. A. Conrad. Center for American Archaeol- ogy, Kampsville, IL, in press.
Hudson, C. 1976 The Southeastern Indians. The University of Tennessee Press, Knoxville. Jackson, H. E., and S. L. Scott
1995 Mississippian Homestead andvillage Subsistence Orga- nization: Contrasts in Large Mammal Remains from Two Sites in the Tombigbee Valley. In Mississippian Conznz14ni- tiesand Ho14seholds, edited by J. D. Rogers and B. D. Smith, pp. 181-200. The University of Alabama Press, Tuscaloosa.
Jones, K.T., and D. Metcalfe 1988 Bare Bones Archaeology: Bone Marrow Indices and Efficiency.Journal ofArchaeologica1 Science 15:415423. Klippel, W. E., L. Snyder, and P. W. Parmalee
1987 Taphonomy and Archaeologically Recovered Mammal Bone from Southeast Missouri. Jounzal of Ethnobiology 7:155-169.
Kelly, L. S. 1997 Patterns of Faunal Exploitation at Cahokia. In Cahokia: Domination and Ideology in the Mississippian World, edited by T. R. Pauketat and T. E. Emerson, pp. 69-88. University of Nebraska Press, Lincoln.
Lupo, K. D. 1998 Experimentally Derived Extraction Rates for Marrow: Implications for Body Part Exploitation Strategies of Plio- Pleistocene Hominid Scavengers. Jounzal of Archaeologi- cal Science 25:657-675.
Lyman, R. L. 1984 Bone Density and Differential Survivorship of Fossil Classes. Jo14rnalofAnthropologicalArchaeology3:259-299. 1985 Bone Frequencies: Differential Transport, in Situ Destruction, and the MGUI. Jo~tnzal ofArchaeologica1 Sci- ence 12:221-236.
1992 Anatomical Considerations of Utility Curves in Zooar- chaeology. Jo14ntal of Archaeological Science 19:7-22. 1994 Vertebrate Taphonomj. Cambridge University Press, Cambridge, U.K. McCabe, R. E., and T. R. McCabe
1984 Of Slings and Arrows: An Historical Retrospection. In White-tailed Deer: Ecology and Management, edited by Lowell K. Halls, pp. 19-72. Stackpole Books, Hanisburg, Pennsylvania.
McCullough, D. R., and D. E. Ullrey 1983 Proximate Mineral and Gross Energy Composition of White-Tailed Deer. Journal of Wildlife Management 47:430-441.
Madrigal, T. C. 1999 Taphonomy and Zooarchaeology of Late Archaic Hunter-Gatherer Complexity in New York. Ph.D. disserta- tion, Department ofAnthropology, Rutgers University. Uni- versity Microfilms, Ann Arbor.
Madrigal, T. C., and S. D. Capaldo 1999 White-Tailed Deer Marrow Yields and Late Archaic Hunter-Gatherers. Journal ofArchaeologica1 Science 26:241-249.
Marean, C. W., and C. J. Frey 1997 Animal Bones from Caves to Cities: Reverse Utility Curves as Methodological Artifacts. American Antiquity 62:698-7 12.
Marean, C. W., and L. M. Spencer 1991 Impact of Carnivore Ravaging on Zooarchaeological Measures of Element Abundance. American Antiquity 56:645458.
Marean, C. W., L. M. Spencer, R. J. Blumenschine, and S. D. Capaldo 1992 Captive Hyaena Bone Choice and Destruction, the Schlepp Effect and Olduvai Archaeofaunas.Journal of Archaeological Science 19: 101-121.
Marshall, F., and T. Pilgram 1991 Meat Versus Within-Bone Nutrients: Another Look at the Meaning of Body Part Representation in Archaeologi- cal Sites. Joz4rnal ofArchaeologica1 Science 18:149-163.
Metcalfe, D., and K. R. Barlow 1992 A Model for Exploring the Optimal Trade-Off between Field Processing and Transport. American Anthropologist 94:34&356.
Metcalfe, D., and K. T. Jones 1987 A Reconsideration ofAnimal Body Part Utility Indices. American Antiquity 53:486-504. Moen, A. M., and C. W. Severinghaus
198 1 The Annual Weight Cycle and Survival of White-Tailed Deer in New York. New York Fish and Game Journal 28:162-177.
Morey, D. F., and W. E. Klippel 1991 Canid Scavenging and Deer Bone Survivorship at an Archaic Period Site inTennessee.Archaeozoo1ogia4: 1 1-28. Munson, P. J.
199 1 Mortality Profiles of White-tailed Deer from Archae- ological Sites in Eastern North America: Selective Hunt- ing or Taphonomy? In Beanzers, Bobwhites, and Blue-Points: Tributes to the Career of Paul W Parmalee,
edited by J. R. Purdue, W. E. Klippel, and B. W. Styles, pp. 139-152. Illinois State Museum Scientific Papers, Vol. 23, Springfield.
Neiland, K. 1970 Weight of Dried Marrow as an Indicator of Fat in Cari- bou Femurs. Jounzal of Wildlife Management 34:904-907. New Jersey Department of Environmental Protection
2001 Guidelines for rating deer range and condition in N.J. Electronic document, www.state.nj.us/dep/fgw/dercndthb.htm
O'Connell, J., and B. Marshall
1989 Analysis of Kangaroo Body Part Transport among the Alyawara of Central Australia. Journal of Archaeological Science 16:393405.
Outram, A,, and P. Rowley-Conwy 1998 Meat and Marrow Utility Indices for Horse (Equus). Journal ofArchaeologica1 Science 25:839-849. Purdue, J. R. 1983 Methods of Determining Sex and Body Size in Prehis- toric Samples of White-Tailed Deer (Odocoileus virgini-
anus). Transactions ofthe Illinois State Academy of Science
Purdue, J. R., B. W. Styles, and M. C. Masulis 1989 Faunal Remains and White-tailed Deer Exploitation from a Late Woodland Upland Encampment: The Boschert Site (23SC609), St. Charles County, Missouri. Midcontinental Jo~tnzal ofilrchaeology 14: 146-163.
Ritchie, W. A. 1932 The Lamoka Lake Site: The ljpe Station of the Archaic Algonkin Period in New York. Researches and Transactions of the New York State Archeological Association, Lewis H. Morgan Chapter, Rochester.
1969 The Archaeology of New York State. Revised edition. Natural History Press, Garden City, New York. Sauer, P. R.
1984 Physical Characteristics. In White-Tailed Deer: Ecol- ogy and Management, edited by L.K. Halls, pp. 73-90. Stackpole Books, Harrisburg, Pennsylvania.
Savelle, J. M., and T. M. Friesen 1996 An Odontocete (Cetacea) Meat Utility Index. Jountal ofArchaeologica1 Science 23:7 13-721. Scott, K. M.
1987 Allometry and Habitat-related Adaptations in the Post- cranial Skeleton of Cervidae. In Biology and Management of the Cervidae, edited by C. Wemmer, pp. 65-80. Smith- sonian Institution Press, Washington, D.C.
Severinghaus, C. 1949 Tooth Development and Wear as Criteria of Age in White-Tailed Deer. Jozirnal of Wildlife Management 13:195-216.
Snyder, J. F. 1962 A Group of Illinois Mounds. In John Francis Snyder: Selected Writings, edited by C.C. Walton, pp. 193-215. The
Illinois State Historical Society, Springfield. Speth, J. D. 1983 Bison Kills and Bone Counts: Decision Making by Ancient Hunters. University of Chicago Press, Chicago. Styles, B. W., and J. R. Purdue
1991 Ritual and Secular Use of Fauna by Middle Woodland Peoples in Western Illinois. In Beanzers, Bobwhites, and Blue-Points: Tributes to the Career of Paul WParmalee,
edited by J. R. Purdue, W. E. Klippel, and B. W. Styles, pp. 421436. Illinois State Museum Scientific Papers, Vol. 23, Springfield.
Swanton, J. R. 1946 The Indians ofthe Southeastent United States. Bureau of American Ethnology Bulletin No. 137. Smithsonian Insti- tution Press, Washington, D.C.
Thomas, D. H., and D. Mayer 1983 Behavioral Faunal Analysis of Selected Horizons. In The Archaeology of Monitor Vallq 2: Gatecliff Shelter, edited by D. H. Thomas, pp. 353-391. American Museum of Nat- ural History Anthropological Papers 59(1), New York.
Verme, L. J., and D. E. Ullrey 1984 Physiology and Nutrition. In White-Tailed Deer: Ecol- ogy and Management, edited by L. K. Halls, pp. 91-1 18. Stackpole Books, Harrisburg, Pennsylvania.
Welch, P. D. 1991 Mo14ndville's Econonzy. University of Alabama Press, Tuscaloosa. Whitaker, J. O., Jr. 1980 The Aud~tbon Society Field Guide to North American Mammals. Alfred A. Knopf, New York.
Received July 18, 2000; Revised December 5, 2001; Accepted June 27, 2002.