The identification and analysis of non-fish vertebrate remains from three archaeological sites on Colington Island was undertaken at Winthrop College, Rock Hill, South Carolina. The sample consisted of 459 fragments from 31DR14; 236 fragments from 31DR15; and 122 fragments from 31DR33.

The initial step in the faunal analysis was the taxonomic identification of the remains. Specimens from the zooarchaeological collection at Winthrop College were available for comparative purposes. Specimens were identified whenever possible to species; however, many bones were very fragmentary and identification was often limited to class. Following identification to taxonomic level, fragments were identified as to skeletal element, e.g. right distal humerus. Any indicators of age and sex were also noted.

Three methods were used to quantify species remains: determination of the total number of identifiable fragments of each species; calculation of minimum number of individuals; and calculation of bone weights. Calculation of number of individuals of each species at each site was based upon the minimum-number-of-individuals (MNI) method first described by White (1953). This method involves counting the most common skeletal element of each species. This corresponds to at least the minimum number of individuals represented in the sample. Variations in age, sex, and size were also considered. In employing this method, the minimum distinction approach outlined by Grayson (1973) was used. All stratigraphic and horizontal excavation units were ignored and MNI was determined from the faunal assemblage of the site as a whole.

Live weights of the non-fish species were determined using bone weights of the archaeological remains. Linear regression formulae used in this calculation were taken from Fradkin (1979). In these formulae, x = body weight in kilograms and y = skeletal weight in kilograms.

The live weights calculated were then converted to maximum possible edible meat yield using the following percentages: turtles 50%; birds 70%; and mammals 60%. Edible meat yield weights for fish were taken from Swift (Appendix B).

Using the edible meat yield weights, the calories available from these species were determined. Estimated calories/100 grams used in the calculation were taken from Watt and Merrill (1963). They are:

Deer 126 calories/100 gm Other Mammals 184.25 calories/100 gm Aves 225 calories/100 gm Turtle 111 calories/100 gm Fish 110 calories/100 gm

Finally, a list of the various species identified (including fish) and the environment (aquatic versus terrestrial) in which they are naturally found was made. Three methods were used to determine the importance of the total fauna exploited within the two major ecological habitats. The first procedure consisted of a species check-list indicating those animals identified. The second method involved calculation of the maximum possible edible meat yields of these animals, while the third method compared the caloric yields of these species.

Knowledge of the natural history of the various species and their seasons of maximum abundance was used to make inferences concerning seasonality of the sites.

In Table 6 are recorded the non-fish species identif ied at each of the three sites. A total of 11 genera or species of three classes of vertebrates were identified including five species of mammals, two genera of birds, and six genera of reptiles. Four species, raccoon (Procyon lotor), white-tailed deer (Odocoileus virginianus), duck (Anas sp.), and diamond-back terrapin (Malaclemys terrapin) common to the three sites.

From 31DR14 were examined 459 non-fish fragments of 11 species (Table 7). Of these, 91 (19.8%) fragments were not identified beyond class. Mammalian remains were most abundant in terms of the number of bone fragments (235 or 51.2%), number of individuals (7 or 53.8%), and number of species (.5 or 45.5%). One species, pig (Sus scrofa), was intrusive. (Page 58)

The white-tailed deer was the largest mammal exploited at the site. A minimum of two individuals represented by 182 bone fragments were identified. All major parts of the skeleton were recovered. At least one animal was male as evidenced by a skull fragment with the antler base intact (Figure 7a), The articular surface of this specimen was badly eroded, and it was not possible to determine with any degree of certainty if the antlers had been cut or shed.

Age determination was based on degree of tooth wear (Sevringhaus 1949) and degree of epiphyseal ossification of selected long bones (Lewall and Cowan 1963). One right mandibular fragment with PM₃ - M₂ in situ was aged at 4-5 years. A right distal humerus epiphysis was present indicating an animal less than 14 months of age.

Other mammals present at the tite included: two raccoons represented by two right mandibles both with postmortem tooth loss; one dog (Canis familiaris) represented by two isolated teeth; and one rabbit (Sylvilagus sp.) represented by a right maxilla fragment and a left humerus. The long bone was fully ossified indicative of an animal greater than 9 months of age (Hale 1949). Two species of rabbit, the eastern cottontail (Sylvilagus floridanus) and the marsh rabbit (Sylvilagus palustris), are common residents of the area, and determination of species was not attempted.

Avian remains consisted of ten fragments of which four were identifiable. Two genera, duck (Anas, sp.) and clapper rail (Rallus longirostris) were represented. The clapper rail is a permanent resident of the area, but is more abundant in the winter because of the influx of northern birds.

Seven species of the genus Anas are found in the coastal area of the two Carolinas: the black duck, mallard, gadwall, pintail, green-winged teal, blue-winged teal, and the cinnamon teal. All are common winter residents, and can be found from November through January. Five of the species can be found from October through April, and one species may be found from September through May (Trinkley 1980). The presence of this genus at the site is indicative of winter occupation, although they can be procured from mid-fall through mid-spring.

Reptiles were well-represented in the non-fish faunal assemblage and accounted for 45.1% of the identified fragments, 30.8% of the MNI, and 36.4% of the identified species. Remains included carapace plastron, and limb elements of two fresh-water species, the musk turtle (Kinosternon subrubrum) and the snapping turtle (Chelydra serpentina); one brackish water species, the diamond-back terrapin (Malaclemys terrapin); and one terrestrial species, the eastern box turtle (Terrapene carolina). (Page 61)

Turtles undergo a period of inactivity from December through February and are most abundant from March through May. Their occurrence at the site suggests spring occupation, although summer and early fall occupation is also possible.

The faunal assemblage recovered from 31DR14 indicates that the aboriginal inhabitants of this site utilized many of the ecological habitats and vertebrate species available. Fish were the most intensively exploited vertebrate group (Appendix B) and remains of these animals comprised 47.8% of the identified species (Table 8), 96.4% of the available meat, and 95.8% of the available calories (Table 9).

Members of the class Mammalia were the most intensively exploited group of non-fish vertebrates. Mammalian remains accounted for 17.4% of the species identified, 3.3% of the meat yield, and 3.8% of the available calories. The white-tailed deer was the most important mammal exploited. All major parts of the animal were recovered suggesting that the entire animal was returned from the kill site to the living area.

The four species of turtle represented in the faunal assemblage provided less than 1% of the total meat yield and calories, although they did comprise 30.8% of the non-fish species identified, 7.1% of the non-fish meat yield, and 6.1% of the non-fish calories yield.

Although a variety of birds are residents of the area, only two species were captured by the Indians and these animals provided 2.6% of the non-fish meat yield and 4.4% of the non-fish calorie yield.

The species present in the faunal assemblage from 31DR14 indicate the extreme importance of the aquatic habitat as a source of food. Aquatic or semi-aquatic species accounted for 77.3% of the identified species, 96.8% of the available meat, and 96.2% of the calories.

From 31DR15 were examined 236 bone fragments of non-fish vertebrates of which 50 (21.2%) fragments were not identified beyond class (Table 10). Five species of three classes of vertebrates were identified, although one species, water snake (Natrix Sp.), probably was not utilized as food. One white-tailed deer, one raccoon, one duck, and one diamond-back terrapin comprised the identifiable food remains. Among the white-tailed deer fragments was a fully ossified proximal femur indicative of an animal greater than 15 months of age. The remains of racoon consisted of a left (Page 62)

mandible with post-mortem tooth loss and a limb fragment. No age estimate was possible. All remains of the diamond-back terrapin were carapace fragments.

The faunal remains recovered from 31DR15 suggests minor utilization of the animal species. Very surprising was the total lack of fish remains. Five species/species groups provided the edible animal food (Table 11) and the meat yield and calories available from these animals were very small (4,017.3 gm and 4,564.7 calories respectively).

The bulk of the animal protein and calories were derived from the aquatic habitat (Table 12). These animals (sharks and rays, diamond-back terrapin, and duck) provided 89.4% of the meat yield and 88.3% of the available calories.

The non-fish vertebrate assemblage from 31DR33 consisted of 122 bone fragments of seven genera of animals (Table 13). Of these 122 fragments, 18 (14.7%) were not identified beyond class. Mammals were the most abundant non-fish vertebrate class represented and remains of these animals accounted for 62.3% of the bone fragments and 55.6% of the MNI. The white-tailed deer was the most abundant mammal represented. Among the skeletal remains of this animal were two left distal tibiae. One of the fragments exhibited complete epiphyseal ossification characteristic of an animal greater than (Page 64)

15 months of age. The epiphysis and diaphysis of the other tibia were not fused indicating the presence of an animal less than 15 months of age.

Rabbit was represented by two left mandibles and a long bone shaft fragment. Both mandibles had the dentition of animals greater than one month of age (Dice and Dice 1941).

Raccoon was represented in the faunal remains by two isolated teeth (both heavily worn), a distal humerus, and a proximal radius. Both post-cranial fragments were completely ossified indicating an animal greater than 6 months of age (Grau et al. 1970).

Three aquatic species of reptiles, the snapping turtle, pond turtle (Chrysemys sp.), and diamond-back terrapin were represented by carapace, plastron, and limb elements. These animals accounted for 31.1% of the non-fish skeletal fragments and 33.3% of the MNI.

The faunal sample from 31DR33 suggests that the inhabitants of this site like those from 31DR14 intensively exploited the aquatic habitat. Only three of the identified species are terrestrial (Table 14) and these animals provided 5.4% of the meat yield and 7.0% of the available calories. Fish were the most intensively exploited aquatic taxa and these animals provided 94.2% of the usable meat and 92.4% of the available calories (Table 15).

The faunal assemblages from the three Colington Island sites have provided insight into the subsistence activities of the Indians. A total of 25 probable food species/species groups were identified from the three sites (Table 16). The greatest species diversity was evident at 31DR14 with the occurrence of 23 species. A total of 13 species were identified from 31DR33. Eleven species including three mammals, one bird, three reptiles, three fish, and members of the Elasmobranchii were common to the two sites.

The available data suggest the existence of a similar pattern of subsistence at the two sites. There appears to have been more intensive exploitation of available animals as a source of protein and calories at 31DR14. The number of species recovered at that site was 1.8 times greater, the amount of edible meat 4.5 times greater, and the number of calories 4.4 times greater than that recovered from 31DR33. Despite these differences, the Indians at both sites depended primarily on the aquatic habitat. Aquatic species accounted for 77.3% of the utilized species, 96.8% of the available meat, and 96.2% of the available calories at 31DR14 compared to 76.9% of the utilized species, 94.6% of the available meat yield, and 93.0% of the available calories at 31DR33. (Page 67)

Fish and members of the Elasmobranchii provided the bulk of this yield. These animals accounted for 77.3% of the utilized species, 96.8% of the available meat and 96.2% of the calories at 31DR14 compared to 41.7% of the species, 94.2% of the usable meat, and 92,4% of the calories at 31DR33.

Although no direct evidence of seasonality was available from the faunal remains, some inferences may be drawn from an examination of the life history of the species. Ten species from 31DR14 exhibit periodic peaks of abundance (Zingmark 1978), and when these periods of abundance are plotted (Figure 16), a pattern is evident. Occupation of the site between April and September would have coincided with the maximum availability of these species. Six species from. 31DR33 also exhibit seasonal abundance. It is evident from a plotting of these periods (Figure 17) that spring and/or summer occupation is inferred from the presence of these species.

The faunal assemblage from 31DR15 is very dissimilar from those of the other sites. The species diversity present at the site was very limited with five species/species groups providing the animal protein. The data suggest minimal exploitation of the available animal species. Particularly striking is the lack of fish remains in the sample. Like the inhabitants of 31DR14 and 31DR33, the residents of 31DR15 did acquire the bulk of their animal protein from the aquatic habitat. These animals accounted for 60.0% of the identified species, 89.4% of the meat yield, and 88.3% of the available calories.

Evidence of seasonality was lacking. Only three species exhibit periodic abundance peaks. Ducks are available from October through Aoril; reptiles from March through May, and to a lesser extent into September; and rays are most abundant from May through September. Hence there is some slight suggestion of spring occupation of the site. (Page 69)