Bull. Fla. Mus. Nat. Hist. (2003) 44( 1 ): 131 - 158 131 ZOOARCHAEOLOGY OF CINNAMON BAY, ST. JOHN, U.S. VIRGIN ISLANDS: PRE-COLUMBIAN OVEREXPLOITATION OF ANIMAL RESOURCES Irvy R. Quitmyeri The zooarchaeological remains from a stratigraphic sequence excavated from the ceremonial site of Cinnamon Bay, St. John, U.S. Virgin Islands, were studied. The samples were recovered using a fine-gauge (1/16 in) screen. During the course of this study, 443 minimum numbers of individuals and 99 species of vertebrates and invertebrates were identified. The fauna was analyzed by estimating the trophic level of reef, inshore, and pelagic zooarchaeological components from three strata representing the Monserrate (ca. A.D. 950), Santa Elena (ca. A.D. 570), and Chican (ca. A.D. 460) ceramic periods. The trophic level model shows an initial increase in the trophic level of taxa from the reef between the Monserrate and Santa Elena periods. This initial increase corresponds to the exponential growth of midden density. Relative to the earlier faunal assemblages, midden density and the mean trophic level of reef resources declines during the Chican period. Greater reliance on pelagic species from the deeper waters offshore and the increased use of mollusks from inshore habitats is also seen. The data show that at low levels of cultural complexity humans can alter their environments. This is particularly true of island biota where biological reservoirs are small. Key words: candy, Caribbean, island biogeography, trophic level, zooarchaeology This chapter presents a study of well-recovered Caribbean pre-Columbian people is not well zooarchaeological remains from the Cinnamon Bay site understood and should be considered in its formative (12Vam-2-3), St. John, U.S. Virgin Islands (Fig. 1). The stages of development. Recent baseline site contains a sequential record of human immigration zooarchaeological data are forming a body of and habitation that began nearly 1000 years ago (Wild information that may be used to ask and answer 1999). Pre-Columbian people first occupied the site ca. questions about the human ecology of pre-Columbian A.D. 1000 and abandoned it by ca. A.D. 1490. For nearly maritime people of the Caribbean (Wing 1995, 2001 a, 500 years Cinnamon Bay served as a ceremonial site 200lb; Wing and Wing 1995,2001). that evolved into a Classic Tafno chief's offering place, Island ecosystems are particularly fragile and or caney (Wild 1999). Early in the sixteenth century, the susceptible to human disturbances because their Spanish chronicler of the Indians, Bartolom6 de las biological reservoirs are small and not easily replenished Casas, described the function of the caney as house of (MacArthur and Wilson 1967). Too frequently, the impact the kings where the "first fruits of the crops" were of humans on these systems is associated with incursions offered (las Casas 1909; Rouse 1992). by Historic Period colonists, while little thought is given Over the past 50 years, Caribbean archaeology to the effects that pre-Columbian people could have had has focused on interpreting the material culture of on the biota (Redford 1990). The scientific community the pre-Columbian people of the region, while has global data that indicate few plant or animal zooarchaeological research has been relatively communities were unaffected by noncomplex societies scarce. The field of zooarchaeology has evolved from (Jackson et al. 2001, Lepofsky et al. 1996; Steadman providing simple presence or absence lists to a formal 1995; Wing 1995, 200la, 200lb; Wing and Wing 1995, discipline that examines the interrelation of humans 2001). They concur that human contact, even at very with their environment. Nonetheless, the ecology of low levels of cultural complexity, can degrade the environment (Quitmyer and Jones 2000). Such findings 'Environmental Archaeology Laboratory, Department of are in contrast to the widely held view that pre-industrial Natural History, Florida Museum of Natural History, people had little affect on their environments. It appears Gainesville, Florida 32611, USA. that environmental degradation seen in the Caribbean 132 ZOOARCHAEOLOGY: Papers to Honor Elizabeth S. Wing Anegada Passage St. John „ St. Thomas 1 . St. Martin - .. .:PE' Virgin Islands 4.- .WI Puerto Rico 0 -*- Saba Caribbean Sea ~ Ne vis Leeward o Islands 0Gulf of Mexico ~ North Atlantic Ocean a - Guadeloupe ~ e Greater Antilles Lesser , Antilles : 400 km 100 lan South America Figure 1. Map of the study area (redrawn from Wing 2001 a). started much earlier than otherwise believed (Jackson rates, high natural mortality, and low recruitment, such et al. 2001, Wing 200lb:481). The most dramatic territorial reef fish higher in the food web as the groupers evidence appears in faunal records of island midden (Serranidae) and snappers (Lutjanidae) are sensitive to deposits (Steadman 1995; Wing 1995, 200la; Wing and even moderate exploitation. The numbers and average Wing 1995, 2001). body size of these fish declined under human fishing The island groupsof the Caribbean Basin are pressure (Sale 1991; Wing 200la:125). Further excellent zooarchaeological laboratories where pre- exacerbating their decline, territorial predators are easily Columbian subsistence strategies can be examined caught with baited hooks and are readily attracted to against the constraints of the principles of island baited fish traps (Wing 200la, 200lb). As the size and biogeography (MacArthur and Wilson 1967; Wing and relative frequency of reef predators decline, there is an Wing 1995, 1997). In the zooarchaeological record of increase in the catch of reef herbivores and omnivores the Lesser Antilles and the Virgin Islands, Wing in the zooarchaeological record (Wing 200la:123). The (2001 a: 125) has repofted changes in species abundance resulting trend is a decline in the measured mean trophic and sizes of animals from pre-Colombian midden deposits level of reef resources between the early and late (St. Thomas, St. Martin, Saba, and Nevis) (Fig. 1). components of the sites. This is accompanied by an Archaeological evidence indicates that the faunal remains incfeased emphasis on taxa with large population represent common, everyday subsistence activities. The reservoirs, such as herrings (Clupeidae), jacks changes observed in the zooarchaeological record (Carangidae), and tuna (Scrombridae), from inshore and correlate with the length of time the environment was pelagic habitats. A shift in technology is indicated in those exposed to humans, and not with the archaeological instances Where there is an increase in the relative period (Wing 2001 a, 200lb). Because of low growth abundance of tuna, because humans would need QUITMYER: Zooarchaeology of Cinnamon Bay 133 watercraft to travel to inherently more dangerous offshore subseries were present (Rouse 1992; Wild 1999). Three habitats (Wing 200la: 125). These trends accompany a ceramic Styles occur within these two subseries (Table three- to four-fold increase in archaeological site size 1): Monserrate, found between 70 and 100 centimeters and number (Wilson 1989). below surface (cmbs); Santa Elena, within the 30 to 70 Relative to the data outlined above, the following cmbs levels; and Chican Ostionoid, identified between question arises: given that Cinnamon Bay functioned as the present ground surface and 30 cmbs (Wild 1999). a ceremonial site, do the site's faunal remains exhibit The Chican Ceramic Period is generally attributed to similar changes in size and quantity? In other words, are the Tafno people, who were the first to make contact the food remains of the elites affected in similar ways with Columbus (Rouse 1992), while the Monserrate and by the over-exploitation of marine resources, or are they Santa Elena ceramic periods represent the ancestors of rendered immune by cultural processes that favor the the Tafno. The three ceramic·styles are not mixed. Rather, elites? The focus of this study is to identify the animal they afe well constrained in the stratigraphic sequence remains that were used at Cinnamon Bay to next address (Wild 1999) (Tablel). three basic questions about Taino subsistence behavior: The archaeological data from Cinnamon,Bay show 1) what animals species were being used, 2) how that the site functioned as a ceremonial center for these animals were obtained, and 3) what evidence nearly 500 years (Wild 1999). From its beginning in is there for anthropogenic changes in the local the Monserrate Period, the site increased in environment? This last question serves as a test of importance. Just prior to European contact it Wing's (200la) model of sustainability of resources evolved into a caney, or temple (Wild 1999). The used by pre-Columbians on St. Thomas, St. Martin, sixteenth-century cleric, Bartolom6 de las Casas Saba, and Nevis. (1909), was a primary observer of the Tafno and reported that the caney was separate from other MATERIALS AND METHODS structures. The caney was where ceremony and ARCHAEOLOGICAL CONTEXT prayer were conducted by the elites. Food offerings Gudmund Hatt (1924) first identified the prehistoric were an important part of these rituals (las Casas component ofCinnamon Bay in 1922. In the intervening 1909), so the Cinnamon Bay zooarchaeological record years, researchers have continued to add information to presents an important opportunity to elucidate the the Cinnamon Bay archaeological record (Haviser 1978; signatures of foods deposited in a well-defined Rutsch 1970; Stoutamire et al. 1980). In 1992 Ken Wild ceremonial context. and Reigina Lebo, U.S. National Park Service, directed Ceramic dates from the top and bottom of the a systematic excavation of three 4x4 meter excavation 1992 Cinnamon Bay excavation place the units from the beachfront, then in danger of erosion. A accumulation of the 1 m formation between A.D. 1000 well-preserved archaeological section was exposed and 1490. A subsequent series of radiocarbon dates where Elenan Ostionoid and Chican Ostioneid ceramic (Table 1) shows that the Monserrate, level 9 (80-90 Table 1. Carbon (14C) dates analyzed from Cinnamon Bay, St. John, U.S. Virgin Islands (Wild 1999). Analysis Level cmbs B.R A.D. Mean A.D. sigma Ceramic Period Beta 1 0-10 -160 1490 - - Chican no date-fauna analyzed 2 10-20 Chican Beta 69973 3 20-30 570+70 1290-1450 1370 23 Chican no date-fauna analyzed 4 3040. Santa Elena no date 5 40-50 Santa Elena Beta 73413 6 5060 520 +70 1300-1485 1393 23 Santa Elena no date 7 60-70 Santa Elena no date 8 70-80 Monserrate Beta 69974 9 80-90 860*80 1020-1290 1155 2a Monserrate Beta-fauna analyzed 10 90-100 -950 1000 - - Monserrate 134 ZOOARCHAEOLOGY: Papers to Honor Elizabeth S. Wing SAMPLING AND SIEVING-9.0 Ve rte br ae W id th ( m Ten 50 x 50 x 10 cm superimposed samples wereE 8.0 -7.0 - taken from the southeast corner of Unit 1. Each sample was water-sieved in the field through a nested 6.0 - pair of screens measuring 1/4 in (6.35 mm) and 1/16 in 5.0- (1.6 mm) gauge. Samples were allowed to air dry 4.0 - before packing and transport to the Florida Museum , of Natural History. Prior to study, the samples were Fine Coarse fumigated for 24 hours with the inorganic compound, Screen Gauge VikaneTM (sulfuryl fluoride). One sample from each of the ceramic components was randomly selected Descriptive Statistics Level 4, FS 111 Level 4, FS 111 for study: FS 117, level 10 (90-100 cmbs); FS 111, Fine Coarse level 4 (30-40 cmbs); and FS 109, level 2 (10-20 cmbs) Mean 3.32 7.35 (Table 1). Standard Deviation 1.60 2.36 The use of fine-gauge screens in faunal recoveryRange 1.04-16.32 1.88-16.32 Sample n 1528 80 represents an important advance in the study of 95% CI 0.08 0.52 zooarchaeological remains (Reitz and Wing 1999: 119- MIN of 95% CI 3.24 6.83 121). Historically, archaeologists have used 1/4 in gauge MAX of 95% CI 3.40 7.87 screens (coarse) in faunal recovery without realizing the possible ramifications of this choice. In the field, the skeletalremains of large animals are highly visible, while Figure 2. Lateral width (mm) of fish vertebrae (mm) from Unit 1, it is difficult to see the remains of small taxa lost throughlevel 4, Cinnamon Bay, St. John, U.S. Virgin Islands. The error coarse-gauge screens. Without the benefit of analysis, itbars represent the 95% confidence interval about the mean comparing screen gauge recovery using fine (1/4-1/16 in) and is easy to be misled into believing that large taxa, such coarse (1/4 in) gauge screens. as mammals, shellfish, and large fish, were the mainstay of the assemblage. When fine screen (1/16 in) is used in the recovery of fauna from sites associated with aquatic habitats, without exception the remains of small fishea cmbs), was deposited between A.D. 1020 and 1290 represent a major part of the sample (Reitz and (Beta 69974); Santa Elena, level 6 (50-60 cmbs), Quitmyer 1988; Reitz and Wing 1999; Wing and between A.D. 1300 and 1485 (Beta 73413); and Quitmyer 1985). Screening experiments from sites the Chican, level 3 (20-30 cmbs), between A.D. representing several cultural periods and across the 1290 and 1475 (Beta 69973). The radiometric ec) southeastern United States and Caribbean confirms this dates indicate that, relative to the Santa Elena and observation (Quitmyer and Massaro 2000; Reitz and Chican strata, the Monserrate levels probably formed Quitmyer 1988; Reitz and Wing 1999; Russo et al. 1991; quite slowly. In fact, the level 3 and 6 (30 cm) carbon Wing and Quitmyer 1985). dates overlap in time, perhaps representing only a To illustrate the importance of fine-gauge screen few years of midden accumulation. The faunal recovefy at Cinnamon Bay, the lateral width (mm) archaeological data do not seem to provide of all unbroken fish vertebral centra from level 4 (FS information about why Cinnamon Bay was abandoned 111) were measured with a Max Cal·IM caliper attached after A.D.1490. to a personal computer. Max CalTM caliper software facilitated the entry of the data into the Microsoft.spread- ZOOARCHAEOLOGICAL METHODS sheet Excel™ where descriptive statistics and graphs Three 4 x4 meter excavation units were opened in 1998 could be produced (figs . 2 and 3 ). The 95% confidence at Cinnamon Bay (Wild 1999). These excavations were calculated around the mean of the fish vertebrae from preparatory to taking a series of midden samples in the the 1/4 in gauge screen (coarse) was compared to those summer of 1999. KenWild and Irvy R. Quitmyer provided recovered in the nested screens measuring 1/4 in and oversight of the sampling. 1/16 in gauge (fine) (Fig. 2). QUITMYER: Zooarchaeology ofCinnamon Bay 135 Two significant observations can be drawn from the 70 - Unit 1, Level 2 90 --30), FS 109 - n=654 measured fish vertebra data. First, 96% of the Cinnamon 60 - Bay level 4 fish vertebrae are smaller than 1/4 in (6.35 60 - 6.35 mm - 1/4 in mm) and would have been lost during sieving if fine-4 40 - Pe rc en t ( 54 Pe rc en t ( 54 P er ce nt (9 4 30.927.8 gauge screen had not been used (Fig. 2). The same is 30 - true for all three samples of the measured fish vertebrae20 -,11 li11010 - 5.4 1 ~) U 12 0.3 0.5 0.0 0.0. 0.2 0.0 0.0 0.0 (Fig. 3). Second, the mean and 95% confidence interval 0- , I , I , , I , I , I , I , , of fish vertebrae recovered in the 1/4 in gauge screen 1 2 3 4 5 € 7 8 910111213141516 and in the 1/4 to 1/16 in gauge screens is significantly Ve tebrae Width (rnm) different (P 5 0.05) (Fig. 2). The mean lateral width of the vertebrae from the fine screen (1/4 and 1/16 in gauge combined) is 3.32 mm (n = 1,528) and those from the 70 - Unit 1, Level 4 (40 -- 50), FS 111 -- n=1520 coarse screens (1/4 in) average 7.35 mm (n = 80). At 60 - 50 - the 95% confidence interval of each of the two samples, 40 - they do not overlap (P 5 0.05), thus illustrating that they 29.2304 30 - are statistically significant and cannot be adequately 20 - compared (Fig. 2). The use of screens of different gauges 10 - 3.4 1 43 1.9 0,5 0.6 0.5 0.2 0.1 0.1 0.0 01 01 in faunal recovery yields very different results, which, in turn, affect the interpretation of subsistence behavior. 1 234 5 E 78 910111213141616 The choice of screen size affects not only the recovery Ve tebrae Width (mm) of representative size classes, but also the kinds of taxa, the number of taxa, and the count of minimum numbers 70 - Unit 1, Level 10 (100 -- 110), FS 117 -- n=121 of individuals that may be lost through coarse-gauge 60.3 60 - screen. 50 A comparison of the percentage of individuals 40 - recovered from level 2 in the coarse-gauge screen 30 - (1/4 in) versus those recovered in the fine screen 20 - 165 10 -41 ~ ~ 66 (1/4-1/16 in) supports the use of fine-gauge screen 1 12 0.8 00 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 sieving (Fig. 4). Fauna recovered in the coarse-gauge O . 11,11,1111,11, 1 2 3 4 5 6 7 8 910111213141616 screen is represented by 178 minimum number of individuals (MNI) and 40 taxa, while fine-gaugeVertebrae Width (mm) sieving yields 256 MNI and 66 taxa. Percentage of Unit 1, Level 2 (20 - 30 cmbs), FS 109 Vertebrae width (mm) 1 2 3 45678 9 10 11 12 13 14 15 16 n n 35 202 182 113 72 26 11 7 2 3 0 0 1 0 0 654 % 5.4 30.9 27.8 17.3 11.0 4.0 1.7 1.1 0.3 0.5 0.0 0.0 0.2 0.0 0.0 0.0 Unit 1, Level 4 (40 - 50 cmbs), FS 111 Vertebrae width (mm) 1 2 3 4 5 6 7 8 9 10 11'1213 1415 16 n n 52 446 465 264 176 66 29 7 9 7 3 1 1 0 1 1 1528 % 3.4 29.2 30.4 17.3 11.5 4.3 1.9 0.5 0.6 0.5 0.2 0.1 0.1 0.0 0.1 0.1 Unit 1, Level 10 (100 - 110 embs), FS 117 Vertebrae width (mm) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 n n 5.0 73.0 20.0 12.0 8.0 2.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0 0 0 121 % 4.1 60.3 16.5 9.9 6.6 1.7 0.8 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Figure 3. Lateral width (mm) of fish vertebrae (mm) from Unit 1, Levels 2,4, and 10, Cinnamon Bay, St. John, U.S. Virgin Islands. The vertical bar marks 6.35 mm or 1/4 in gauge. 136 ZOOARCHAEOLOGY: Papers to Honor Elizabeth S. Wing 100 1/4th in. Screen Unit 1, Level 4 (40 -- 50 cmbs), FS 111 n = 178 -- Taxa = 40 80 - (coarse gauge) P er ce nt (% ) 68.5 60 - 40 - 20 - 14.0 12.4 11 0.6 0.6 ~1711 0 -, - I Mammals Birds Reptiles Fishes Crabs Chions Gastropods Bivalves I 100 1/4th and 1/16th in. Screen Unit 1, Level 4 (40 -- 50 cmbs), FS 111 (fine gauge) n = 256 -- Taxa = 66 80 - P er ce nt ( % ) 60- 48.8 40 - 34.8 20- 13.5 20.8 0.4 0 -· Mammals Birds Reptiles Fishes Crabs Ch#ons Gastropods Bivalves Unit 1, Level 4 (40-50 cmbs), FS 111-1/4 in gauge screen Unit 1, Level 4 (40-50 cmbs), FS 111-1/4-1/16 in gauge screen Classes MNI % Classes MNI % Mammals 2 1.12 Mammals 2 0.8 Birds 1 0.56 Birds 1 0.4 Reptiles 1 0.56 Reptiles 2 0.8 Fishes 25 14.04 Fishes 89 34.8 Crabs 3 1.69 Crabs 9 3.5 Chitons 2 1.1 2 Chitons 5 2.0 Gastropods 122 68.54 Gastropods 125 48.8 Bivalves 22 12.36 Bivalves 23 9.0 Total 178 100.00 Total 256 100.00 Taxa 40 Taxa 66 Figure 4. A comparison of the minimum numbers of individuals of animals classes recovered in 1/4 in gauge screen (coarse) and 1/4 and 1/16 in gauge screen (fine) from Unit 1, level 4, FS 111, Cinnamon Bay, St. Thomas, U.S. Virgin Islands. QUITMYER: Zooarchaeology ofCinnamon Bay 137 fishes increases from 14% in the coarse gauge screen in varying stages of ontogenetic development. The to 35% in the fine-gauge screen. evidence indicates that the landsnails are commensal This sieving experiment shows that fine-gauge species that were probably not consumed by humans. A sieving ( 1/4- 1/16 in gauge) assures that large and small random sample of terrestrial gastropods was sorted from fauna have an equal chance of being recovered (Wing each sample to identify those species in the midden. and Quitmyer 1985), thus providing an optimal picture Barnacles, limpets, and corals were also identified in the of the zooarchaeological assemblage. Cinnamon Bay assemblages. These, too, are considered Minimum numbers of individuals. The three taxa that were probably not a common part of the diet . Cinnamon Bay faunal samples were hand-sorted and Consequently, these animals are not included as part of identified to the lowest possible taxon using the the analysis of human subsistence (MNI), but their comparative collections of the Horida Museumof Natural identification contributes information to the natural history History (Table 2). Standard zooarchaeological methods of the island (Table 1). were used to quantify the faunal remains (Reitz and Biomass estimates. In most cases , we are presented Quitmyer 1988; Reitz and Wing 1999; Ziegler 1973). with only a portion of the skeleton of each individual These include a count of the identified specimens and estimated in the zooarchaeological record and we are their weight in grams. A count of the MNI is used to unable to measure the whole organism. Fortunately, characterize the ranked frequency of animals that are these skeletal elements scale allometrically with body present in the faunal assemblages. MNI represents the size (Peters 1983). Allometry reflects the structural and fewest number of individuals that can be identified from functional consequences of a change in size or in scale the skeletal assemblage. MNI was detefmined by the among similarly shaped animals (Peters 1983; Reitz et use of the concept of paired elements and individual size. al. 1987; Schmidt-Nielsen 1984). Growth is a non-linear For example, four left frontals and five right frontals process through ontogeny and this allometric relationship of equal size from parrot fish (Scaridae) represent is describedby a mathematical power function, y = a(Xb) five MNI, while four large frontals and five small (Schmidt-Nielsen 1984). In order to produce a straight frontals represent nine MNI (Reitz and Wing 1999; line regression, this is transformed using the common Ziegler 1973). log. The resulting formula is: Commensal species. It is common within most zooarchaeological assemblages to identify a suite of taxa Log y=a+ b(logX) that may be regarded as commensal species. Some are Where: animals that are attracted to human habitation, where b = the slope of the line, they gain protection or food. Other taxa may find their a = the y intercept, way into the zooarchaeological record through the x = the independent variable, skeletal measurement, acquisition of targeted species. For example, clutches y = the dependantvariable, the estimated body mass. of oysters may yield a microcosm of taxa that are not directly intended as a human food resource. Such a Many vertebrate and invertebrate characteristics collection might include a large number of barnacles scale allometrically, but the most useful to this study is and limpets. It is difficult to determine which of these body mass (biomass) in relation to a measurable skeletal attendant species contribute to the human diet. The element. The constants used to estimate body weight soft tissue of commensals would unintentionally from the vertebrae of teleost fishes is presented in Table 3. contribute to the human diet if the clutches of oyster were boiled and the liquid and meat consumed. In ANALYSIS OF TROPHIC LEVEL contrast, the consumption of raw oysters would not Recent work by Pauley et al. (1998) has helped include the soft tissue of commensals adhering to the to identify the affects of modern fishing practices on oyster shells. worldwide stocks over the past 45 years. They In the three assemblages in this study, ten species conclude that modern fishing practices reduce the of terrestrial gastropods were identified (Table 1). With availability of taxa from high trophic levels, thus the exception of the terrestrial snail (Polydontes lima), requiring the exploitation of species from lower in none of these animals has a body dimension greater than the food web (Pauley et al. 1998). Pauley and his 7 mm and most are smaller than 2 mm. These taxa were associates (1998) have characterized these findings 138 ZOOARCHAEOLOGY: Papers to Honor Elizabeth S. Wing Table 2. Presence and absence of animal taxa identified from Cinnamon Bay, St. John, US Virgin Islands. Key: * = taxa, identified from the general levels and previous studies; X = present in the sample. Unit 1 - Faunal Sample Level 1 Level 4 Level 10 Taxa FS109 FS111 FS 117 Comments VERTEBRATES Mammals Mammalia mammals x x Mammalia (medium) medium mammals x x Nesophontes edithae extinct insectivore * introduced ; extinct Homo sapiens human * introduced ; Tafno extinct Rodentia rodents x Rodentia (large) lg, rodents (e.g., hutia) x Rattus %pp. European rat * Cavia porcellus guinea pig * introduced; extirpated Isolobodon portoricensis hutia x x x introduced ; extinct Cetacea porpoise * Monachus tropicalis monk seal * extinct Trichechus manatus manatee * threatened Birds Aves birds x x Aves (medium) medium birds - e.g., dove x Puffinus iherminieri Audubon ' s shearwater * Ardeidae herons * Rallidae possibly flightless rail * extinct Porphyrula martinica purple gallinule Otus nudipes Puerto Rican screech owl * . x rare Reptiles Squamata lizards x Iguanidae iguana x Iguana iguana iguana * rare on St . John Serpentes snake x x rare on St. John Testudines turtles x x Trachemys spp. pond turtle * introduced; rare or extirpated Cheloniidae sea turtle * Cartilaginous Fishes Chondrichthyes cartilaginous fishes x Rajiformes sates and rays x x Lamniformes shark * Carcharhinus spp. requiem sharks * Bony Fishes Osteichthyes bony fishes x x x Elops saurus ladyfish x Gymnothorax spp. nnoray x Clupeidae shads/herrings x x x Belonidae needlefishes x x Holocentrus spp. squirrelfishes x x Holocentrus adscensionis squirrelfish x Prionotus spp. searobins x Epinephelus spp. groupers x x x Epinephelus cuentatus graysby * Epinephelus striatus Nassau grouper * Carangidae jacks x x x Caranx spp. jacks x x Caranx latus horse-eyejack * Caranx ruber barjack x Lutjanus spp. snappers x x x cf. Lutjanus griseus gray snapper * (cont.) QUITMYER: Zooarchaeology of Cinnamon Bay 139 Table 2 (cont.). Taxa FS109 FS 111 FS 117 Comments Haemulon spp. grunt x x x Sparidae porgies x Calamus spp. porgy x x x Mugil spp. mullet * Sphyraena spp. barracuda x x x Labridae wrasses x Bodianus spp. hogfish x Halichoeres spp. wrasse x Scaridae parrotfishes x x Scarus spp. parrotfish x x Sparisoma spp. parrotfish x x x scans 4 coeruieus blue parrotfish * Sparisoma viride stoplight parrotfish x Acanthurus spp. surgeonfish x Scombridae tuna x x x Auxis rochei bullet mackerel Euthynnus alletteratus little tunny Balistidae leatherjackets x Balistes spp. triggerfish x x Ostraciidae boxfishes x x x Diodontidae porcupinefishes x x Diodon spp. porcupinefish x INVERTEBRATES Crustaceans Cirripedia barnacles x Balanomorpha acorn barnacles x Balanus spp. barnacle x Panulirus spp. spiny lobsters * Decapoda decapod crabs x x Brachyura crabs x x Callinectes spp. blue crab * Coenobita clypeatus land hermit crab x x x Gecarcinidae land crabs x x x Mithrax spp. spider crab x Mollusks Mollusca snails and clams x x x Chitons Chitonidae Chitons x x Acanthopleura granulata x x Gastropods Gastropoda snails x x Pleurotomariacea Archaeogastropoda x Fissurellidae limpets x Diodora spp. limpet x x Diodora listeri Lister 's keyhole limpet x Acmaea antillarum Antillean limpet x x x Turbinidae starsnails/turban snajls x x Turbo castanea chestnut turban x Cittarium pica West Indian topsnail x x x Nerita spp. nerite x x Nerita petoronta bleeding tooth x x x Nerita versicolor four-tooth nerite x x Neritina spp. nerites x Neritina virginea virgin nerite x x Cerithiidae ceriths x x Tectarius -muricatus beaded periwinkle x (cont.) 140 ZOOARCHAEOLOGY: Papers to Honor Elizabeth S. Wing Table 2 (cont.). Taxa FS 109 FS 111 FS 117 Comments Strombus gigas queencohch Crepidula spp. slipper shells x Crucibulum auricula West Indian cup-and-saucer x Naticidae moonsnails x Polinices hepaticus brown moonsnail x Muricidae murex x Cymatium muricinum knobbed triton x Chicoreus brevifrons West Indian murex x Plicopurpura patula widemouth rocksnail x Thais spp. rocksnail x Thais deltoidea deltoid rocksnail x Columbellidae dove-shell Columbellidae dovesnails x x Alcadia spp. X drop X X X Columbeila mercatoria West Indian dovesnail Olividae olive x Fasciolaria tulipa true tulip Conus spp. cone x x Pupoides modicus island dagger x x x Hinea lineatus dwarf planaxis x Subulinidae awlsnails x x x Lamellaxis micra tiny awlsnail x x x Opeas pyrgula sharp awlsnail x x x Subulina octona miniature awlsnail Bulimulus guadatupensis West Indian bulimulus x x Polydontes lima land snail - no common name x x X >< X X K X Succineidae ambersnails Xanthonycidae land snails - no common name x x Sagdidae mudcloak x x Tuskshells and Toothshells Dentalium spp. tuskshell x Bivalves Bivalvia clams/mussels/oysters xxx Mytilidae mussels x x Brachidontes spp. rrlussel x Brachidontes exustus scorehed mossel x x Anadara notabilis eared ark x Pinctada radiata Atlantic pearl oyster x x Arca spp. ark 5hells x Arca zebra turkey wing Glycymeris pectinata comb bittersweat x Nodipecten nodosus lions-paw seallop x Codakia orbicularis tiger lucine x x x Chamidae jewelbox x x Chama spp. jewelbox x Laevicardium spp. eggcockle x Donax denticulatus coquina x Asaphis deflorata gaudy sanguin x X * >< X X K Veneridae Venus clams x x Periglypta listeri princess verus Echinoderrns Echinoidea sea urchin x x Corals Anthrozoa corals x x QUITMYER: Zooarchaeology of Cinnamon Bay 141 as "fishing down the food web." Wing (2001 a, 200lb) was determined by multiplying the average biomass of has applied the methods of Pauley et al. (1998) to the the individualsin each species by the MNI. zooarchaeological record of five sites on islands in 3) The final step is to multiply the biomass of each the Caribbean: 1) Tutu, St Thomas, VI; 2) Hope for each species by the mean trophic level index (Pauley Estate, St. Martin; 3) Kelbey's, Saba; and 4) et al. 1998). Trophic level indices range from one to five. Indian Castle and Hichman's, Nevis (Fig. 1). The Plants are primary producers, trophic level (TL) = 1, while zooarchaeological remains record a decline in mean TL value for top predators is 5. These indices are derived trophic level of reef resources, while there is a from the feeding behaviors of the organisms (Pauley and subsequent increase or decrease in the mean trophic Christensen 1997). level of inshore and pelagic taxa between the early The habitat eategories of fishes associated with coral and late components of the sites. When there is a reefs follow Sale (1991). The mean trophic levels were measured increase in the mean trophic level of inshore calculated only for the aquatic vertebrate component of and pelagic resources, tuna and other large predators the Cinnamon Bay assemblages. This is comparable to predominate. In sites where a decrease in the mean Wing's (200la) analysis where her faunal samples did trophic level of inshore and pelagic taxa occurs, there not include the molluscan data. is a relative increase in the biomass of herrings Size class analysis of West Indian topsnail (Clupeidae) and other small fishes (Wing 200la ). (Cittarium pica). Through successive deposits of The trend toward the use of tuna or herring some pre-Columbian Caribbean sites the numbers of represents a shift to those species with larger land crabs declines with the length of exposure to biological reservoirs when compared to species human habitation while the numbers of West Indian from the reef habitats. topsnail (and other mollusks) increase (Wing 200la). In this study, I apply the methods presented by Measurements were taken from the shells of the Pauley et al. (1998) and adapted by Wing (200la) to topsnails to document two temporal Changes in their determine the mean trophic level of the catch of size classes: 1) the greatest distance from the notch vertebrate aquatic species identified in the Cinnamon in the umbilicus to the notch of the aperture; and 2) Bay faunal assemblage. The results of this experiment the greatest aperture height (mm). Where there are are compared to the studies of the St. Thomas, St. Martin, large numbers of fragmented shells, the notch of the Saba, and Nevis zooarchaeological records (Wing 200la, umbilicus to the aperture notch measurement assures 200la). a larger sample size. The formula for calculating the mean trophic level To facilitate the temporal comparison of the (TL) is as follows (Pauley et al. 1998): size classes of topsnails, the mean and the 95% confidence interval for the two shell measurementsTE = 6. n. y ./6 y.,' U U u U listed above were calculated for each of the 10 where: levels. It was then possible to rank the mean values TLj is the mean trophic level for year i, and ascertain which of the samples were (Y) is the landings by trophic levels of individual statistically different (P 5 0.05) from one another species groups j. by noting whether or not their confidence intervals overlapped. This procedure is straightforward,The application of the trophic level formula using easily interpretable, and conservative.the zooarchaeological specimens follows a three-part process (Wing 200la): 1) The appropriate allometric formula (Table 3) was used to calculate the average biomass for the various Table 3. Allometric constants used to estimate biomass (Y) in taxa in each sample from measurements of their grams. These are applied the formula log Y =log a + b(log X) where X is the measured width of the vertebrae in millimetersvertebrae. In those rare cases where there are no (Wing 1999).measurable vertebrae present, the mean vertebral width of unidentified fish was used with the assumption that Measurement n slope b Y intercept a ri the vertebrae came from a cross section of the identified X = width of teleostspecies. vertebrae (mm) 43 2.53 0.872 0.872) Estimated biomass of the catch for each species 142 ZOOARCHAEOLOGY: Papers to Honor Elizabeth S. Wing Table4. Fauna identified from Unit 1, Level 2 (10 - 20 cmbs), FS 109 Cinnamon Bay, U.S. Virgin Islands, NPS ACC# 191, UF Accession # 510. Key: * = present in the sample, but not quantified; + = weighed as Subulinidae. TaxoD Caunt % MNI % Weight &) % Vertebrata * - - - 4.22 0.37 Mammalia 6 0.41 - 0.29 0.03 Mammalia (medium) 3 0.20 - - 0.46 0.04 Rodentia (large) 3 0.20 - = 0.35 0.03 Isolobodon portoricensis 7 0,48 1 0.68 1.14 0.10 Aves (medium) 1 0.07 - - 0.05 0.00 Otus nudipes 1 0.07 1 0.68 0.05 0.00 Squamata 1 0.07 - - 0.01 0.00 Igunidae 2 0.14 1 0.68 0.01 0.00 Serpentes 7 0.48 1 0.68 0.18 0.02 Testudines 1 0.07 1 0.68 0.20 0.02 Osteichthyes 599 40.92 - 61.46 5.45 Clupeidae 15 1.02 1 0.68 0.04 0.00 Belonidae 5 0.34 1 0.68 0.15 0.01 Holocentrus spp. 7 0.48 2 1.35 0.16 0.01 Epinephelus spp. .6 0.41 3 2.03 1.21 0.11 Carangidae 10 0.68 - - 0.27 0.02 * Caranx spp. 13 0.89 3 2.03 1.93 0.17 Lutjanus spp. 17 1.16 6 4.05 1.70 0.15 Haemuton spp. 24 1.64 9 6.08 1.08 0.10 Sparidae 4 0.27 - - 0.·18 0.02 Calamus spp. 5 0.34 2 1.35 043 0.04 Sphyraena spp. 6 0.41 1 0.68 0.10 0.01 Bodianus spp, 1 0.07 1 0.68 0.01 0.00 Scaridae 28 1.91 - 3.19 0.28 Scams spp. 6 0.41 1 0.68 0.33 0.03 Sparisoma spp. 118 8.06 7 4.73 17.30 1.53 Sparisoma viride 15 1.02 8 5.41 2.99 0.27 Scombridae 20 1.37 2 1.35 3.96 0.35 ~a~~eP 12 0.82 1 0.68 0.48 0.04 5 0.34 1 0.68 0.14 0,01 Diodon spp. 1 0.07 1 0.68 0.13 0.01 Cirripedia 15 1.02 - - 0.87 0.08 Brachyura 35 2.39 - 2.13 0.19 Coenobim clypeams 32 2.19 2 1.35 2.83 0.25 Gecarcinidae 6 0.41 1 0.68 1.40 0.12 Mollusca * - - - 24.10 2.14 Acanthopleuragranulam 27 1.84 4 2.70 14.90 1.32 Gastropoda 8 0.55 - - 22.55 2.00 Fissurellidae 2 0.14 - - 0.02 0.00 Diodora spp. 4 0.27 * - 0.27 0.02 Acmaea antillarum 19 1.30 * - 1.39 0.12 Turbinidae 1 0.07 - - 2.96 0.26 Turbo casmnea 1 0.07 1 0.68 3.67 0.33 Cittarium pica 180 12 .30 29 19 .59 678 .05 60. 10 Nerita spp. 3 0.20 - - 0.89 0.08 Neritapeloronta 2 0.14 2 1.35 4.58 0.41 Nerim vers,color 3 0.20 2 1.35 1.77 0.16 Neritina virginea 30 2.05 14 9.46 10.94 0.97 Tecmrius muricams 4 0.27 4 2.70 3.18 0.28 Strombus gigas 1 0.07 1 0.68 7.84 0.69 Crepidula spp. 2 0.14 2 1.35 0.04 0.00 Naticidae 1 0.07 1 0.68 0.02 0.00 Chicoreus brevifrons 2 0.14 2 1.35 7.03 0.62 Thais spp. 1 0.07 1 0.68 11.03 0.98 Columbellidae 1 0.07 1 0.68 0.30 0.03 (cont.) QUITMYER: Zooarchaeology of Cinnamon Bay 143 Table 4 (cont.) Taxon Count % MNI % Weight(g) % Atcadia sp - 1.27 0.11 Conus spp. 1 0.07 1 0.68 0.10 0.01 Pupoides modicus * 0.04 0.00 Subulinidae * 0.75 0.07 Inmellaxis.micra * + + 8 + 1Opeas pyrgula Subutina octona * + + Bulimulus guadalupensis * 2.29 0. 20 Polydontes lima * 3.16 0. 29 Succineidae * 0.29 0.03 Xanthonycidae * 0.02 0.00 Sagdidae * - - - 0.06 0.01 Dentatium spp. 1 0.07 1 0.68 0.01 0.00 Bivalvia 8 0.55 - - 5.71 0.51 Mytilidae 10 0.68 1 0.68 1.53 0.14 Anadam notabilis 3 0.20 2 1.35 64.15 5.69 Pincmda radiam 15 1.02 3 2.03 5.38 0.48 Arca zebra 9 0.61 4 2.70 24.84 2.20 Lyropecten nodosus 1 0.07 1 0.68 0.49 0.04 Codakia orbicularis 57 3 .89 4 2 .70 97 . 01 8 . 60 Chamidae 2 0.14 - - 0.25 0.02 Chama spp. 4 0.27 3 2.03 3.75 0.33 Donax denticulatus 5 0.34 2 1.35 2.78 0.25 Asaphis defloram 2 0.14 2 1.35 0.89 0.08 Veneridae 1 0.07 1 0.68 0.07 0.01 Echinoidea 11 0.75 1 0.68 0.44 0.04 Anthrozoa 5 0.34 * 5.87 0.52 TOTAL TAXA 1464 100.00 148 100.00 1128.20 100.00 Taxa = 49 Summary by Class Class MNI % Class MNI % Mammals 1 0.68 Crabs 3 2.05 Birds 1 0.68 Chitons 4 2.74 Reptiles 3 2.05 Gastropoda 61 41.78 Fishes 50 34.25 Bivalves 23 15.75 TOTAL CLASSES 146 100.00 RESULTS identified in the sample. The skeletal remains are well LEvEL 2 (10-20 CMBS) FS 109 preserved, showing few signs of abrasion, pitting, or Archaeological context. The faunal sample from diagenesis. FeWer than 0.5% of the remains showed level 2 (FS 109) was excavated from 10 to 20 cmbs. signs of having been burned. The well preserved While no radiometric (140 date has been analyzed nature of the sample is consistent with materials from for level 2 (Table 1), level 1 dates radiometrically to the deeper levels of the site. -460 B.R (A.D. 1490), and level 3 to 570 *70 B.R Gastropods (41.4%), fishes (34.5%), and bivalves (A.D.1290 to 1450) (Beta 69973), thus indicating that (15.9%) were the most frequently identified animals level 2 formed sometime during this +110-year (MNI) from level 2 (Table 4, Fig. 5). West Indian interval. The artifacts indicate that the faunal topsnail (19.7%) was the most numerous animal remains are associated with the Chican Ceramic identified, while four-tooth nerite (Neritina virginea, Period . 9 . 5 %) and beaded periwinkle (Tectarius muricatus, Minimum numbers Of individuals. The level 2 2 . 7 %) were also frequent . Parrotfishes (Sparisoma faunal assemblage contains 49 taxa and 147 MNI spp., 10.2%) are the second most abundant taxon. (Table 4). This does not include commensal species The relative abundance of groupers (Epinephelus 144 ZOOARCHAEOLOGY: Papers to Honor Elizabeth S. Wing 100 Unit 1, Level 2 (20 -- 30 cmbs), FS 109 n= 145 80 - Pe rc en t ( % ) Pe rc en t ( % ) 60 - 41.4 40 - 34.5 20 - 15.9 07 0.7 2.1 21 28 0 -·-·-·-·--· Mammals Bircls Reptiles Fishes Crabs Chitons Gastropoda Bivalves 100 Unit 1, Level 4 (40 -- 50 cmbs), FS 111 n = 256 80 - Pe rc en t ( % ) 60 - 48.8 40 - 34.8 0.8 0.4 0,8 ~ 3,5 2.00 -· Mammals Birds Reptiles Fishes Crabs Chaons Gastropoda Bivalves 100 Unit 1, Level 10 (100 -- 110 cmbs), FS 117 n = 34 80 - 60 - 47.1 40 - 235 20 - 14.7 5.92.9 2.9 2.90.0 - i0-·I. · - Mammals Birds Reptiles Fishes Crabs Chitons Gastropoda Bivalves Figure 5. Summary of the minimum number of individuals (MNI) by class identified from levels 2,4, and 10, Cinnamon Bay, St. John, U.S. Virgin Islands. QUITMYER: Zooarchaeology ofCinnamon Bay 145 Table 5. Mean trophic level of aquatic vertebrate taxa identified from.Unit 1, Level 2 (10-20 embs), FS 109, Cinnamon Bay, US Virgin Islands. Measurennent Biomass Biomass X (g) MNI x MNI TL TLijYij REEF Hotocentrus spp. 3 . 14 . 134 . 66 2 269.32 3 . 5 942 . 61 Epinephelus spp. 3 . 31 153 . 87 3 461 .62 3 . 8 1754. 16 Lutianus spp. 3.14 134.66 6 807.95 4.6 3716.56 Haemulon spp. 3 . 53 181 . 08 9 1629.72 3 . 5 5704.01 Bodianus spp. 3 . 14 134 . 66 1 134.66 3 . 6 484. 77 Scarus spp. 4.18 277.70 1 277.70 3.4 944.18 Sparisoma spp. 3.62 192.99 7 1350.92 3.5 4728.23 Reef Sparisoma viride 3 . 62 192 .99 8 1543 .91 3 .5 5403 .69 Biomass Balistes spp. 3 . 14 134. 66 1 134 .66 3 . 5 471 . 30 x MNI MeanTL TLijYij Ostraciidae 3.14 134.66 1 134.66 3.2 430.91 6879.77 3.64 25011.33 Diodon spp. 3.14 134.66 1 134.66 3.2 430.91 39.96% INSHORE, PELAGIC Clupeidae 1.67 27.26 1 27.26 2.6 70.87 Belonidae 3.86 227.02 1 227.02 3.2 726.47 Caranx spp. 9.85 2428.81 3 7286.43 4.0 29145.73 Inshore, Pelagic Calamus spp. 3 . 14 134 . 66 2 269.32 3 .4 915 . 67 Biomass Sphyraena spp. 2 .47 73 . 37 1 73 .37 4. 5 330. 17 x MNI MeanTL TLijYij Scombridae 7.52 1226.96 2 2453.92 3.8 9324.91 10337.32 3.92 40513.82 60.04% TOTAL SAMPLE 50 17217.09 65525.15 spp., 2.0%), snappers (Lutjanus spp., 4.1%), and turtle carapace was identified and its surface texture grunts (Haemulon spp., 6.1%) is somewhat less is not consistent with sea turtle. It is most likely that important to the calculation ofMNI in level 2. Bivalves the sample represents the pond turtle (Trachemys represent a minor component of the sample, with spp.), although land tortoise cannot be ruled out. Pond Atlantic pearl oyster (Pictada radiata, 2 .0%), turkey turtle , introduced to the island by pre-Columbian wing (Arca zebra , 2 . 7 %), tiger lucine (Codakia immigrants , has been previously identified in the orbicularis, 2.7%), and jewel box (Chama spp., Cinnamon Bay zooarchaeological remains (Table 2). 2.0%) being among the most significant. Two species The Puerto Rican screech owl (Otus nudipes), the of crabs are present : land hermit crab (Coenobita only volant species identified, currently is severely clypeatus, 1.4%) and land crab (Gecarcinidae, 0.7%). threatened, and possibly extirpated, from some of the Mammals (0.7%), birds (0.7%), reptiles (2.1%), Virgin Islands. and crabs (2 . 1 %) were among the rarest animals Trophic level analysis. Table 5 and Fig . 6 show encountered in the level 2 sample. The only mammal trophic level analysis of the level 2 aquatic fauna. that was identified is the hutia (Isolobodon The data show that 40 . 0% of the fish biomass comes portoricensis, 0. 68 %), introduced to the island by from reef fishes , while 60 . 0 % is obtained from pre-Columbian people and now probably extinct inshore/pelagic taxa. Among the reef species, (Wilson 1989; Wing and Wing 1997). Iguanid lizards parrotfish and grunts represent the largest biomass. (Iguanidae, 0.68), snakes (Serpentes, 0.68%), and The greatest portion of biomass from the inshore/ turtles (Testudines, 0.68%) are the three groups of pelagic species come from the tunas (Scombridae) reptiles present in the assemblage . A single piece of and jacks (Caranx spp .). The mean trophic level of 146 ZOOARCHAEOLOGY: Papers to Honor Elizabeth S. Wing Reef Fishes 4.00 3 3.go _ 0 M ea n Tr op hi c Le ve l M ea n Tr op hi c Le 3.80 - 3.70 - 3.60 - 3.50 - 3.40 , , , 0 500 1000 1500 2000 Years Before Present -li- St. Thomas -0-St. Martin -7,--Saba -0-Nevis a Cinnamon 1-2 o Cinnamon L4 O Cinnamon L10 Inshore, Pelagic Fishes 4.00 a. K 3.80 - 1.fl--0- 3.60 - 3.40 - -~3.20 - 3.00 , , , 0 500 1000 1500 2000 Years Before Present -*-St. Thomas -0- St. Martin -*-Saba -0- Nevis a Cinnamon L.2 o Cinnamon L4 o Cinnamon L10 Figure 6. Comparison of trophic levels of reef and inshore/pelagic species of early and late Caribbean Island archaeological sites (Wing 2001). Open symbols represent the Cinnamon Bay site, St. John, U.S. Virgin Islands. QUITMYER: Zooarchaeology of Cinnamon Bay 147 the reef faunal assemblage is 3.6; for the inshore/ snappers, and grunts-all reef species-represent the pelagic fauna it is 3.9. largest biomass. The greatest portion of biomass from the inshore/pelagic species came from the tunas and LEvEL 4 (30-40 CMBS) FS 111 jacks. Archaeological context. Fauna from level 4 (FS 111) was excavated from 30 to 40 cmbs. While no LEVEL 10 (90-100 CMBS) FS 117 radiometric ( 14C) date has been analyzed for level 4 Archaeological context. Level 10 (FS 117 ) fauna (Table 1), by inference its stratigraphic position, just was excavated from 90 to 100 cmbs. A radiometric date before level 3 (570 +70 B.P., A.D. 1290-1450) (Beta of ca. 950 B.P. documents the earliest arrival of 69973) and after level 6 (520 +70 B.P.,A.D. 1300-1485) Monserrate ceramic-bearing people. Relative to the much (Beta 73413), suggests it formed between those times. denser levels 2 and 4, level 10 represents a very loose Because the two radiometric dates overlap, level 4 fauna midden accumulation. was deposited in a relatively short interval of time. The Minimum numbers of individuals. The fauna in artifacts are consistent with the Santa Elena Ceramic Period. level 10 consists of 29 taxa and 35 MNI (Table 8). Minimum numbers of individuals. Fifty -eight Commensal species are not included in these counts . taxa and 256 MNI were identified from level 4 (Table Preservation of the skeletal remains and the presence 6). Commensal species are not part of the quantified of burned elements are consistent with materials from sample. The faunal remains are well preserved, the upper levels of the site. containing few signs of abrasion, pitting, or diagenesis. Table 8 and Fig. 5 show that gastropods (24.0%), Fewer than 0.5% of the remains showed signs of fishes (47.1%), and bivalves (14.7%) were the most being burned. frequently identified animals (MNI) from level 10. Gastropods (48.8%), fishes (34.8%), and bivalves West Indian topsnail (11.4%) is the most numerous (9.0%) were the most frequently identified animals animal identified, while all other individual gastropods (MNI) from leve14 (Table 6, Fig. 5). West Indian topsnail contribute 2.9% of the MNI (Table 8). Barracuda (31.9%) is most numerous. Four-tooth nerite (9.5%) also (Sphyraena spp., 5.7%), parrotfishes (5.7%), and represented a frequently identified gastropod. tunas (5.7%) are the most abundant fish species. All Parrotfishes (10.4%) are the second most abundant other fish species contribute no more than 2.9% each taxon, while groupers (2.7%), snappers (6.2%), and to the MNI. Bivalves represent a minor component grunts (4.2%) are important contributors to the level 4 of the sample and no single species dominates the classification of MNI. Bivalves are a minor component count of MNI. Relative to levels 2 and 4, in level 10 of the sample. Turkey wing (3.9%), tiger lucine, (1.9%), the land hermit cra (2.9%) and land crab (2.9%) and coquina (Donax denticulatus, 1.2%) are among reach their greatest percentages. Land crabs account the more significant bivalve species . The land hermit for a total MNI of 5 .9%. crab (1.5%), and land crab (1.5%) were present in the Mammals (2.9%) and birds (2.9%) were less sample. frequently identified in level 10 than in the upper Mammals (0.8%), birds (0.4%), reptiles (0.8%), and levels of the deposit. Hutia (2.9%) was the only crabs (3.5%) are among the rarest animals occurring in mammal represented in the sample. No reptiles were the leve14 assemblage. Hutia (0.8%) is the only mammal identified. present in the sample. A single snake and single turtle Trophic level analysis. The results of the trophic are the only two reptiles identified in level 4. Like the level analysis of the level 10 fauna is presented in specimen found in level 2, the fragment of turtle carapace Table 9 and Fig. 6. The mean trophic level for the probably represents the pond turtle, but land tortoise reef assemblage is 3.5; for the inshore/pelagic fauna, cannot be ruled out. it is 3.7. Just over one-half (51.0%) the fish biomass Trophic level analysis. Table 7 and Fig . 6 present comes from reef fishes and 49% is contributed by the trophic level analysis of the level 4 aquatic fauna. inshore/pelagic species. Among the reef fishes, The mean trophic level for the reef and inshore/pelagic snappers, grunts, and triggerfish (Balistidae) are the fauna is 3.9. Fish species common to the reef account most prevalent biomass contributors. The greatest for 79.0% of the biomass and the inshore/pelagic species portion of biomass from the inshore/pelagic taxa came represent 21.0% of the total fish biomass. Parrotfish, from the jacks, barracudas, and tunas. 148 ZOOARCHAEOLOGY: Papers to Honor Elizabeth S. Wing Table 6. Fauna identified from Unit 1, Level 4 (40 - 50 cmbs), FS 111, Cinnamon Bay, U.S. Virgin Islands, NPS ACC# 191, UF Accession # 510. Key: * = present in the sample but not quantified; + = Weighed as Subulinidae (a) = intrusive into the sample- not included in the quantification of the remains. Taxon Count % MNI % Weight (g) % ANIMALIA * - - - 1510.60 36.40 Vertebrata * - - - 0.08 Of© Mammalia 1 0.03 - - 0.02 O.00 Marnmalia (medium) 9 0.29 - - 0.37 0.01 1 Rodentia 1 0.03 - - 0.10 0.00 Isolobodon portoricensis 40 I.30 2 071 11 .84 0.29 AVES 6 0.19 1 0.38 0.85 om Serpentes 5 0.16 1 0.38 0.14 O.00 Testudines 2 0.06 1 0.38 OAO 0.01 Chondrichthyes 1 0.03 - - 0.06 O.00 Rajiformes 3 0.10 1 0.38 006 0.00 Osteichthyes 1823 59.25 - - 170.09 4.10 Elops saurus 1 0.03 1 0.38 ODD 0.00 Gymnothorax spp. 2 0.06 1 0.38 0.39 0.01 Clupeidae 31 1.01 1 0.38 0.08 O.00 Belonidae 8 0.26 3 1.15 0.51 0.01 Holocentrus spp. 6 0.19 - - 0.20 O.00 Holocentrus adscensionis 1 0.03 1 0.38 0.06 O.00 Prionotus spp. 7 0.23 1 0.38 0.30 0.01 Epinephelus spp. 35 1.14 7 2.69 6.13 0.15 Carangidae 51 1.66 - 4.17 0.10 Caranx spp. 12 0.39 4 1.54 1.04 0.03 Caranx ruber 13 0.42 1 0.38 1.62 0.04 Lutjanus spp. 48 1.56 16 6.15 3.94 0.09 Haemulon spp. 31 1.01 11 4.23 2.01 0.05 Calamus spp. 6 0.19 2 0.77 0.78 0.02 Sphyraena spp. 23 0.75 2 0.77 0.48 0.01 Labridae 1 0.03 - - 0.19 O.00 Halichoeres spp. 4 0.13 2 0.77 0.25 0.01 Scaridae 26 0.84 - - 1.76 0.04 Scarus spp. 7 0.23 1 0.38 0.44 0.01 Sparisoma spp. 282 9. 16 27 10.38 59.92 1 .44 Acanthurus spp. 1 0.03 1 0.38 0.10 O.00 Scombridae 43 1.40 3 1.15 13.25 0.32 Balistes spp. 0.37 0.01 8 0.26 1 0.38 0.51 0.01 Ostraciidae 16 0.52 1 0.38 Diodontidae 2 0.06 1 0.38 0.36 0.01 Balanomorpha 10 0.32 - - 1.04 0.03 Balanus spp. 1 0.03 - - 0.38 0.01 Decapoda 7 0.23 - - 0.00 O.00 Brachyura 11 0.36 - - 3.09 0.07 Coenobim ctypeatus 41 1.33 4 1.54 3.45 0.08 Gecarcinidae 44 1.43 4 1.54 11.67 0.28 Mithrax spp. 1 0.03 1 0.38 0.36 0.01 Mollusca - - - 10.63 0.26 Chitonidae 34 1.10 3 1.15 11.39 0.27 Acanthopleura granulata 10 0.32 2 0.77 8.87 0.21 GASTROPODA 3 0.10 - - 8.67 0.21 Diodora listeri 1 0.03 * - 029 0.01 Acmaea antillarum 43 1.40 * - 3.42 0.08 Turbinidae 2 0.06 1 0.38 0.15 0.00 Citmrium pica 99 3.22 83 31.92 1765.43 42.54 Nerita spp. 10 0.32 - - 1.92 0.05 (cont.) QUITMYER: Zooarchaeology ofCinnamon Bay 149 Table 6 (cont.) Taxon Count % MNI % Weight (g) % Nerita peloronta 2 0.06 2 0.11 0.68 0.02 Nerita versicolor 1 0.13 5 1.91 3.74 0.09 Neritina spp. 9 0.29 - - 0.44 0.01 Neritina virginea 30 0.97 23 8.85 23.73 0.57 Cerithiidae 1 0.03 1 0.38 0.20 0.00 Strombus gigas 3 0.10 1 0.38 20.39 0.49 Crucibulum auricula 1 0.03 1 0.38 1.00 0.02 Polinices hepaticus 1 0.03 1 0.38 3.30 0.08 Muricidae 2 0.06 - - 3.58 0.09 Cymatium muricinum 1 0.03 1 0.38 2.43 0.06 Plicopurpura patula 1 0.03 1 0.38 0.32 0.01 Thais deltoidea 1 0.03 1 0.38 6.68 0.16 Columbellidae 1 0.03 1 0.38 0.04 0.00 Alcadia spp. * - - - 2.13 0.05 Fasciolaria tulipa 1 0.03 1 0.38 95.41 2.30 Conus spp. 1 0.03 1 0.38 0.06 O.00 Pupoides modicus - - 0.61 0.00 Hinea lineatus 1 0.03 1 0.38 0.15 0.00 Subulinidae * 2.19 0.05 Lamellaxis micra * 0 - Opeas pyrgula * + - Bulimulus guadalupensis * 1.89 0.05 Polydontes Lima * 3.4 0.08 Xanthonycidae * ~ 0.25 0.01 Sagdidae * - 0.11 0.00 Bivalvia 5 0.16 4.54 0.11 Mytilidae 1 0.03 - 0.16 0,00 Brachidontes spp. 4 0.13 1 0.38 0.45 0.01 Brachidontes exustus 1 0.03 1 0.38 0.60 0.01 Pinctada radiata 29 0.94 2 071 4.61 0. 11 Arca spp. 10 3.85 121.17 2.92 1 0.03 - - 22.10 0.53 Arca zebra 14 0.45 Glycymeris pectinata 1 0.03 1 0.38 0.02 0.00 Codakia orbicularis 61 1.98 5 1.92 137.68 3.32 Chamidae 1 0.03 - - 0.18 O.00 Don~ denticulatus 1 003 1 0.38 31.45 0.76 5 0.16 3 1.15 1.16 0.03 Asaphis defloram 2 0.06 2 0.77 3.05 0.07 Veneridae 2 0.06 1 0.38 0.07 0.00 Periglypta listeri (a) 1 0.03 (a) 1 25.96 0.63 Echinoidea 12 0.39 0.26 0.01 TOTALTAXA 3077 100.00 260 100.00 4149.94 100.00 Taxa = 58 Summary by Class Class MNI % Mammals 2 0.78 Birds 1 0.39 Reptiles 2 0.78 Fishes 89 34.77 Crabs 9 3.52 Chitons 5 1.95 Gastropoda 125 48.83 Bivalves 23 8.98 TOTAL CLASSES 256 100.00 150 ZOOARCHAEOLOGY: Papers to Honor Elizabeth S. Wing Table 7. Mean trophic level of aquatic vertebrates identified from Unit 1, Level 4 (30-40 cmbs), FS 111, Cinnamon Bay, U.S. Virgin Islands. Measurement Biomass Biomass Taxon X (g) MNI X MNI TL TLijYij REEF Holocentrus adscensionis 3 . 3 153 .87 1 153 . 87 3 . 5 538 . 56 Epinephelus spp. 4 . 2 279 . 38 7 1955 .69 3 . 8 7431 . 61 Lutjanus spp. 5.8 630.49 16 10087.82 4.6 46403.98 Hae,nuton spp. 7.0 1019.83 11 11218.12 3.5 39263.41 Halichoeres spp . 3 . 3 153 . 87 2 307 . 75 3 . 6 1107 . 89 Scarus spp. 3 .3 153 . 87 1 153 . 87 3 . 4 523 . 17 Sparisoma spp. 3.1 131.43 27 3548.52 3.5 12419.83 Acanthurus spp. 3 . 3 153 . 87 1 153 .87 3 . 5 538 .56 Reef Balistes spp. 3 . 8 211 . 01 1 211 . 01 3 .0 633 .02 biomass Ostraciidae 3.3 153.87 1 153.87 3.2 492.40 x MNI TL TLii Yi i Diodontidae 3.3 153.87 1 153.87 3.2 492.40 28098.27 3.91 109844.83 78.93% INSHORE, PELAGIC Rajiformes 3.4 165.15 1 165.15 3.5 578.01 Elops saurus 2 . 2 52 . 87 1 52 . 87 3 . 0 158 . 62 Gymnothorax spp. 3 . 3 153 . 87 1 153 . 87 3 . 5 538 . 56 Clupeidae 1.7 26.85 1 26.85 2.6 69.80 Belonidae 3.1 130.36 3 391.08 3.2 1251.46 Prionoms spp. 4 .4 318 .00 1 318 .00 3 . 5 1113 . 01 Caranx spp. 5 . 6 579 . 36 4 2317 .46 4.0 9269. 83 Caranx ruber 8 . 5 1647 . 96 1 1647 . 96 44 6756 . 65 Inshore , pelagic Calamus spp. 3 . 3 153 . 87 2 307 .75 3 . 4 1046 .34 biomass Sphyraena spp. 2. 1 45 . 79 2 91 . 57 4 . 5 412 .08 x MNI TL TLitYii Scombridae 5.9 675.58 3 2026.75 3.8 7701.66 7499.32 3.85 28896.02 TOTAL 88 35597.59 138740.85 21.07% Mean trophic level 3.9 Size class analysis of West Indian topsnail trend correlates with the most intensive period of (Cittarium pica). The measurements from the notch in topsnail harvest (Fig. 8). Even so, their average size the umbilicus to the terminus of the aperture and the increases from level 6 through level 3. In level 4 there aperture height present virtually the same pattern of size is an inverse correlation between the percentage of class for the Cinnamon Bay topsnails. The mean values topsnails and their size. It appears that the collection and 95% confidence interval about the mean show that strategy was to obtain only the largest possible samples from level 1 through level 9 are statistically specimens. This seems to lead to another decrease similar and could have been collected from the same in shell size in level 2, followed by a subsequent population (Table 10 and Fig. 7). increase in level 1, this during a period that exhibits a Although not statistically significant (P S 0.5), relatively low percentage of topsnails. three trends seem to emerge in the nine samples. From level 9 to level 7 there is an increase in the DISCUSSION mean size of the topsnails. This event corresponds to The zooarchaeological record of Cinnamon Bay a small but persistent increase in the percentage of presents a pattern of subsistence activity similar to topsnails in levels 10 through 7 (figs. 7-8). In level 6 that observed in other Caribbean sites (Wing 1995, through level 4, specimens are smaller on average 2001 a, 200lb), even though for over 500 years it than the specimens from the previous levels. This functioned as a ceremonial site. In other words, there QUITMYER: Zooarchaeology of Cinnamon Bay 151 Table 8. Fauna identified from Unit 1, Level 10 (90 - 100 cmbs), FS 117, Cinnamon Boy, U.S. Virgin Islands, NPS ACC# 191, UF Accession # 510. Key: * = present in the sample but not quantified; + = weighed as Subulinida6. Taxon Count % MNI % Weight (g) % Vertebrata 2 0.32 - 0.91 0.89 Isolobodon portoricensis 2 0.32 1 2.86 1 .48 1 .45 Aves 1 0.16 1 2.86 0.05 0.04 Rajiformes 1 0.16 1 2.86 0.04 0.04 Osteichthyes 405 64.29 - 5.24 5.11 Clupeidae 22 3.49 1 2.86 0.05 0.05 Epinephelus spp. 1 0.16 1 2.86 0.13 0.13 Carangidae 7 1.11 1 2.86 0.33 0.32 Lutjanus app 2 0.32 1 2.86 0.10 0.10 Haemulon spp. 8 1.27 1 2.86 0.32 0.31 Ca/amus spp. 1 0.16 1 2.86 0.23 0.22 Sphyraena spp. 19 3.02 2 5.71 0.54 0.53 Sparisoma spp. 13 2.06 2 5.71 2.08 2.03 Scombridae 3 0.48 2 5.71 0.10 0.09 Balistidae 1 0.16 1 2.86 0.08 0.08 Ostraciidae 2 0.32 1 2.86 0.05 0.04 Diodontidae 2 0.32 1 2.86 0.02 002 Decapoda 56 8.89 - - 3.15 3.07 Coenobim clypeams 3 0.48 1 2.86 0.18 0.18 Gecarcinidae 22 3.49 1 2.86 3.54 3.46 Mollusca 7 1.11 - 4.65 4.54 Chitonidae 5 0.79 1 2.86 1.74 1.70 Pleurotomariacea 2 0.32 - 0.06 0.06 Diodora spp. 1 0.16 * 0.09 0.09 Acmaea antillarum 2 0.32 # - 0.50 0.49 Cittariumpica 17 2.70 4 11 .43. 47 . 94 46 . 77 Neritapeloronta 4 0.63 1 2.86 0. 87 0 . 85 Cerithiidae 1 0.16 1 2.86 0.12 0.12 Akadia spp. * - - 0.18 0.18 Columbellamercatoria 1 0.16 1 2.86 0.09 0.09 Olividae 1 0.16 1 2.86 0.02 0.02 Pupoides modicus 0.02 0.02 Subulinidae * 2.67 2.60 Lamellaxis micra * + Opeas pyrgula * + - Bulimulus guadalupensis 2 . 36 2 . 30 Polydontes lima 1 .64 1 . 60 Xanthonycidae * 0,58 0.57 Sagdidae * 0.23 0.22 Gastropoda (terrestrial) * - 0.19 0.19 Bivalvia 2 0.32 - 0.15 0.15 Brachidontes crustus 1 0.16 1 2.86 0.05 0.05 Arca zebra 1 0.16 1 2.86 0.06 0.06 Codakia orbicularis 5 0.79 1 2.86 16 . 87 16 .46 Laevicardium spp. 1 0,16 1 2.86 0.01 0.01 Veneridae 5 0.79 1 2.86 0.13 0.13 Echinoidea 1 0.16 1 2.86 0.03 0.02 Anthrozoa * 2.64 2.58 TOTAL 630 100.00 35 100.00 102.50 100.00 Taxa = 29 Summary by Class Class MNI % Class MNI % Mammals 1 2.9 Crabs 2 5.9 Birds 1 2.9 Chitons 1 2.9 Reptiles 0 0.0 Gastropoda 8 23.5 Fishes 16 47.1 Bivalves 5 14.7 TOTAL CLAsSES 34 100.00 152 ZOOARCHAEOLOGY: Papers to Honor Elizabeth S. Wing Table 9. Mean trophic level of aquatic vertebrates identified from Unit 1, Level 10 (10-20 cmbs), FS 117, Cinnamon Bay, U.S. Virgin Islands. Biomass Taxon Measurement Biomass MNI x MNI TL TI-ijYij REEF Epinephelus spp. 2 .79 99 . 86 1 99 . 86 3 . 8 379 .46 Lutjanus spp. 3.36 159.82 1 159.82 4.6 735.18 Haemulon spp. 2.82 102.60 1 102.60 3.5 359.09 Span'soma spp. 2.79 99.86 2 199.71 3.5 699.00 Reef Balistidae 3.75 211.01 1 211.01 3.0 633.02 biomass Ostraciidae 2.79 99.86 1 99.86 3.2 319.54 x MNI TL TLiiyli Diodontidae 2.79 99.86 1 99.86 3.2 319.54 972.71 3.54 3444.84 51.01% INSHORE, PELAGIC Rajiformes 3.8 218.20 1 218.20 3.5 763.70 Clupeidae 1.66 26.85 1 26.85 2.6 69.80 Carangidae 3.66 198.43 1 198:43 3.3 654.82 Inshore, pelagic Calamus spp. 2.79 99.86 1 99.86 3.4 339.51 biomass Sphyraena spp. 2.47 73.37 2 146.74 4.5 660.34 x MNI TL TLiiyli Scombridae 3.02 122.02 2 244.03 3.8 927.32 934.11 3.66 3415.49 48.99% TOTAL 16 1906.82 6860.32 Mean trophic level 3.60 are patterns of subsistence behavior that exist in the segments of the population and required little more zooarchaeological record of ritually deposited fauna than a collecting container. Watercraft would have and in a common midden. For example, terrestrial been necessary to effectively set nets and traps to animals were never an important part of subsistence catch fishes in the inshore, pelagic, and reef zones. because aquatic resources were the focal point of Changes in island fauna similar to those reported by the zoological part of the diet at all the sites (Fig. 5) Wing (200la) occur in the Cinnamon Bay (Wing 1995, 200la, 200lb). Collection of mollusks zooarchaeological record. The relative abondance of was probably an activity that was done by most hutia and land crabs declines over time (Table 11, Fig. 5) and there is a decrease in the relative abundance of fishes relative to the presence of Table 10 . Measured West Indian topsnail (Cittarium pica) from mollusks (Table 11 , Fig . 5 ). Unit 1, levels 1 to 10, Cinnamon Bay, St. John, U.S. Virgin Islands. A comparison of the mean via the.95% confidence interval. FAUNAL CHANGES BETWEEN THE MONSERRATE (CA. A.D. 1000) Mean Mean AND SANTA ELENA (CA. A.D. 1393).PERIODS. Level n MES 1 * 9596CI n MES 2" 95%CI In contrast to Caribbean sites examined by Wing Ll 11 13.90 3.17 6 33.20 12.86 (2001 a), the Cinnamon Bay data show a statistically L2 21 12.38 2.21 6 29.71 14.20 significant (P 5 0.05) initial increase in the body size L3 22 17.53 3.36 20 36.64 7.11 IA 70 15.09 1.06 18 29.93 3.87 of all fishes between the Monserrate (level 10) and L5 42 13.96 1.50 20 31.31 4.27 Santa Elena (level 4) periods (Fig. 9). During this I-6 85 13.84 0.96 30 28.96 3.32 time the calculated biomass of reef species increases L7 16 14.95 3.05 7 38.88 14.70 (Fig. 10). The mean trophic level of the reef L8 12 13.33 2.17 8 28.89 6.14 component of the sample (Fig. 6) also increasesL9 6 11.60 1.38 6 24.97 3.96 L10 3 18.21 - 1 9.30 - during the 300-year interval between the Monserrate (level 10) and the: Santa Elena (level 4) periods. (Fig.*MES 1 = Greatest distance from the notch in the unibilicus to the terminus of the aperture (mm), 6). In the time subsequent to 1000 years B.R, the **MES 2 = Greatest aperture height (mm). mean trophic level of the Monserrate (level 10) faunal QUITMYER.,Zooarchaeology of Cinnamon Bay 153 25 E E 5 20 1 +0 . ~15 * ifil.0 2 10 t ® 45 , i,"i,i, i L1 L2 L3 L4 L5 L6 L7 L8 L9 L10 Level -- Unit 1 60 55 A pe rtu re H ei gh t ( m m ) 50 45 40 - 35 * 30 - - 25 20 15 10 + 5 ",i,",i,, L1 L2 L3 L4 L5 L6 L7 L8 L9 L10 Level Unit 1 Figure 7. Measured West Indian topsnail (Cittariumpica) from Unit 1, levels 1 to 10, Cinnamon Bay, U.S. Virgin Islands. A comparison of the mean of the measured distance from the aperture to the umbilicus (mm) and the aperture height (mm). The error bars represent the 95% confidence interval calculated around the mean. 154 ZOOARCHAEOLOGY: Papers to Honor Elizabeth S. Wing groupers and snappers declines, while the number of 35 n = 288 grunts and porgies increases (Table 11). These 30 - 0 changes are reflected in the mean trophic levels of 25- the reef component of the assemblages, which declines 20 - Pe rc en t ( %0 between the Santa Elena (level 4) and Chican (level 15- • 2) periods (Fig. 6). There is little change in the mean10 - + 0 trophic levels of the inshore/pelagic components.V + 0 • between the Santa Elena and Chican fauna (Fig. 6).0 L1 L2 L3 L4 L5 L6 L7 L8 L9 L 10 Notably, the changes seen between the Santa Elena Level -- Unit 1 and Chican periods accompany a four-fold decrease in midden volume from 4,150 g (Table 6) to 1,128 g (Table 8). Figure 8 . Percentage of West Indian topsnail (Cittarium pica) At Cinnamon Bay, there was not the continuous from Unit 1, levels I to 10, Cinnamon Bay, St. John, U.S. Virgin decline in the trophic level of the reef fauna that Wing Islands. (2001 a) observed between the early and late period samples in her study group. Sampling protocol and assemblage is similar to other Late Period Caribbean archaeological context probably account for this sites (Wing 2001 a) (Fig. 6). A mean trophic level of important>difference. First, the Cinnamon Bay study 3.9 for the level 4 reef component of the fauna had the advantage of analyzing three samples exceeds all other values reported for the region (Wing representing a time series of the faunal record. This 2001 a). Taxa from inshore/pelagic habitats also show finer scale of analysis helps to illustrate an increasing an increase in the mean trophic level (Fig. 6). The demand for predatory reef fishes in the first 300 years pattern outlined above follows a forty-fold increase of the site's history and that demand's inherent in midden volume between the Monserrate Period consequences. Second, Cinnamon Bay represents a (103 g) (Table 4) and the Santa Elena Periods (4,150 ceremonial place where offerings of food were made g) (Table 6). The midden volume charts the to the elites, while Wing's analysis (200la) examined evolutionary direction of the site's function as a the fauna from the general middens, representing ceremonial center, implying that it required great and common subsistence behavior. The increase in the increasing quantities of food to fuel this enterprise. trophic level of the reef component highlights the However, there appear to have been consequences increasing importance of Cinnamon Bay as a chiefs' associated with intensive use of reef resources. caney between the Monserrate and Santa Elena periods. But the data also show that the elites were FAUNAL CHANGES BETWEEN THE SANTA ELENA (CA. A.D. not immune to the overexploitation of the environment, 1393) AND CHICAN (CA. A.D. 1490) PERIODS even though they probably had greater access to a In the approximately 97 years separating the diverse array of resources. Relative to their Santa Elena (leyel 4) and Chican periods (level 2) at predecessors, in the last years of site occupation, the Cinnamon Bay there are distinct changes observed Chican elites consumed animals lower in the food web in the faunal assemblages, similar to changes found because that was what was available in the elsewhere in the Caribbean by Wing (200la). Biomass environment. The alternative for the Chicans was to of reef taxa declines, while biomass of inshore/pelagic consume reef fishes lower in the food chain while species eventually exceeds that of the reef species collecting shellfish from the inshore habitats. Pelagic (Fig. 9). With the exception of the Tutu site on St. species, higher in the food chain, were also more Thomas, biomass from reef vertebrates always intensively caught. Fishing for pelagic species in the exceeds the inshore/pelagic zones in the sites reported offshore zones presents greater personal danger than by Wing (2001 a). The average size classes of all fishes inshore or reef subsistence activities. from the two levels (Santa Elena and Chican) are The Cinnamon Bay faunal assemblages show that not significantly different (P 5 0.05), but they trend changes also took place on land. With human toward slightly smaller individuals in the more recent colonizatioh came the introduction of such non- Chican period ( Fig . 9 ). The relative number of indigenous animals as an insectivore (Nesophontes QUITMYER: Zooarchaeology of Cinnamon Bay 155 Table 11. A summary of selected species and the percent of minimum numbers of individuals identified from Unit 1, Cinnamon Bay, St. John, U.S. Virgin Islands. Unit 1, faunal sample FS 109 FS 111 FS 117 Leve12 Leve14 Level 10 Taxon Common name % MNI % MNI % MNI Isolobodon portoricensis hutia 0 . 7 0. 8 2 .9 Aves birds 0.7 0.4 2.9 Reptilia reptiles 2.0 0.8 - Rajiformes sates and rays - 0.4 2.9 Clupeidae shads/herrings 0.7 0.4 2.9 Belonidae needlefishes 0.7 1.2 Holocentrus spp. squirrelfishes 1 .4 0.4 _ Epinephelus spp. groupers 2 .0 2 . 7 2 .9 Carangidae jacks - 2.9 Caranx spp. jacks 2.0 1.5 Caranx ruber barjack - 0 .4 - Lutjamis spp. snappers 4.1 6.2 2.9 Haemulon spp. grunt 6. 1 4 . 2 2 .9 Calamus spp. porgy 1 .4 0 . 8 2 . 9 Sphyraena spp. barracuda 0.7 0. 8 5 . 7 Scarits spp. parrotfish 0.7 0.4 - Sparisoma spp. parrotfish 10 . 2 10 . 4 5 . 7 Scombridae tuna 1.4 1.2 5.7 Balistidae triggerfishes 0.7 0.4 2.9 Ostraciidae cowfishes 0.7 0.4 2.9 Diodontidae porcupinefishes 0.7 0.4 2.9 Crustaceans Coenobita clypeatus land hermit crab 1 .4 1 . 5 2 .9 Gecarcinidae land crabs 0.7 1.5 2.9 Mollusks Chitons Chitonidae chiton 2.7 1.9 2.9 Gastropods Cittariumpica West Indian topsnail 19 . 7 31 . 9 11 .4 Nerita peloronta bleeding tooth 1 .4 0. 8 2 .9 Nerita versicolor four-tooth nerite 1 .4 1 .9 - Neritina virginea virgin nerite 9 . 5 8 .6 - Gastropods and Bivalves Gastropoda 41 .4 48 . 8 23 . 5 Biv(livia 15 . 9 9 .0 14 . 7 edith(le ), hutia , guinea pig ( Cavia porcellus), and understood . Within the Santa Elena (level 4 ) and pond turtle (Trachemys spp.) (Table 2). As with Chican (level 10) periods, the tiny awlsnail modern-day travelers , there were some unintentional (Lamellaxis micra ) and sharp awlsnail (Opeas introductions of animals , represented by some of the pyrgula)are abundant. Tiny awlsnail is originally from commensal species present in the deposit. During the South America, Mexico, and the West Indies, while course of history, terrestrial awlsnails (Subulinidae) the distribution of the sharp awlsnail is unknown. It have been introduced into Florida and through the appears that both species were introduced to Caribbean (Auffenberg and Stange 1988). The timing Cinnamon Bay some time before 950 B.R In level 2, or agents of these introductions are not well the miniature awlsnail (Subulina octona) suddenly 156 ZOOARCHAEOLOGY: Papers to Honor Elizabeth S. Wing 90 79.0 With exposure to humans, a suite of native and 80 - - introduced animals has become threatened, extirpated,70 - 60.0 Pe rc e ( ) or extinct from the island of St. John. This suite$ 60 - 51.049.0 includes the pre-Columbian human populations of the1 so- 40.0~ 40- .. E. E: E Caribbean, as well as some of the animals they hunted 30 - - E 21.0 E:E:E- and tended , such as Nesophontes, hutia , guinea pig , 20 - Iii:}I 2 monk seal (Monachus tropicalis), manatee.... - 1. . (Trichechus manatus), Puerto Rican screech owl,0 iguana lizard, and pond turtle. The long-termLevel 2 Level 4 Leve110 interactions of humans with their environment have 0 Reef • Inshore/Pelagic exacted changes in the aquatic and terrestrial realms of St. John, changes evident in the diverse faunal Figure 9. Lateral width of fish vertebrae (mm) from Unit 1, assemblages examined in this paper. Cinnamon Bay, St. John, U.S. Virgin Islands. The error bars represent the 95% confidence interval around the mean. SUMMARY The faunal record of Cinnamon Bay presents a rare and important opportunity to examine the appears around 570 +70 B.P., during the Chican zooarchaeological record of pre-Columbian people on Period. The origin of the miniature awlsnail is probably St. John. The purpose of this study was to address South America. Its introduction into most tropical and three basic questions pertinent to the faunal remains subtropical areas of the world has been the result of and to compare the results to other pre-Columbian horticultural and agricultural imports (Auffenberg and Caribbean Island faunas. Stange 1988). 1. What animal species were being consumed? The faunal data show that fish and shellfish were the most commonly consumed species during the 500- year occupation. Groupers, jacks, snappers, grunts, and parrotfishes were the most frequently identified 4.00 3.80 - vertebrate species. Among invertebrates, West Indian Ve rte br ae W id th (m rn ) 3.60 - topsnails were the most common constituents (MNI) 340 - f of assemblages, although a diverse number of bivalves3.20- ~ 3.00 - and gastropods were also consumed. Land hermit 2.80 -260 - crabs and land crabs were a minor component of the 240 diet. Terrestrial species were uncommon, but 2.20 - zoo , , Nesophontes, hutia, guinea pig, and pond turtle were Level 2 Level 4 Level 10 transported from other island localities. Hutia and Samples guinea pig were probably kept by the Tafno, perhaps representing incipient domestication. Descriptive Statistics 2. How were these animals obtained? Level 2 Level 4 Level 10 The Tafno were maritime people whose adaptation FS 109 FS 111 FS 117 to the sea represents a very long tradition with roots on Mean 3.22 3.32 2.57 the South American coast. Based on the most common Std Dev 1.54 1.60 1.17 species and their habitats, boats, fish traps, nets, hook Range 1.04-12.94 1.04-16.32 1.1-7.36 and line, spears, and gathering were the common Sample n 654 1528 121 methods of subsistence. The evidence suggests that the 95% CI 0.12 0.08 0.21 technology changed little over the 500 years of pre- Columbian human settlement at Cinnamon Bay and was Figure 10. A comparison of the percentage of biomass con- similar to many other Caribbean Island sites. tributed by reef fishes vs. inshore/pelagic fishes, Cinnamon 3. Is there evidence of human impact on the local Bay, St. John, U.S. Virgin Islands. environment? QUITMYER: Zooarchaeology of Cinnamon Bay 157 Early in the human history of Cinnamon Bay there LITERATURE CITED was an exponential increase in the acquisition of Auffenberg, K., and L.A. Stange. 1988. The Subulinidae of aquatic resources. Vertebrates from inshore and Florida. Tallahassee: Florida Department of pelagic habitats were secondary to the reef Agriculture and Consumer Servide, Division, community. The size-class of fishes under exploitation Entomology, Circular no. 305. Hatt, G. 1924. Archaeology of the Virgin Islands. Theincreased, accompanied by an increase in the relative Hague: Proceedings of the Twenty-first International abundance of snappers, grunts, and parrotfishes. Congress of Americanists 1:29-42. Subsequently the mean trophic level of animals taken Haviser, J.B., Jr. 1978. The prehistory of the Virgin Islands: from the reef increases, along with an increase in the From excavations at Cinnamon Bay, St. John. M.A. mean trophic level of the inshore/pelagic zone. This Thesjs, Department of Anthropology, Florida State represents a growth period during which increasing University, Tallahassee. amounts of food were needed to sustain Cinnamon Jackson, J.B.C., M.X. Kirby, W.H. Berger, K.A. Bjorndal, Bay as a ceremonial center. L.W. Botsford, B.J. Bourque, R.H. Bradbury, R. Cooke, The pressure ef this initial growth period is J. Erlandson, J. A. Estes, T.R Hughes, S. Kidwell, C. B. Lange, H.S. Lenihan, J.M. Pandolfi, C. H. Perteson,recorded in the latest deposits of fauna analyzed in R.S. Steneck, M. J. Tegner, and R. R. Warner. 2001.this study. Over time the abundance of hutia and land Historical overfishing and the recent collapse of coastal crab remains decreases. A decline in the relative ecosystems. Science 293:629-638. abundance of fishes is accompanied by an increase las Casas, Bartolome de. 1909. Apologetica historia de in the use of mollusks. Relative abundance of groupers las Indias. Historiades de Indias, Vol. 1. Madrid: and snappers diminishes, while grunts and porgies M. Serrano Y Sanz, Bailly/Bailliere e Hijos. Pp. 170- increase. Animals from the reef are no longer the 180 in S.C. Grisold, translator. Tafno: Precolumbian dominant component of the assemblage, rather those Art and Culture from the Caribbean. New York: species from inshore/pelagic habitats predominate. Monacelli Press. The trophic level declines during the later part of the Lepofsky, D., P. V. Kirch, and K.P. Lertzman. 1996. site occupation. Stratigraphic and paleobotanical evidence for prehistoric human-induced environmental disturbanceThe faunal data from this later period in the history on Mo'orea, French Polynesia. Pacific Science of Cinnamon Bay most resembles the changes that took 50(3):253-273. place with the passage of time at other Caribbean sites. MacArthur, R. H., and E. 0. Wilson. 1967. The Theory of While all the Caribbean sites represent different time Island Biogeography. Princeton, New Jersey: periods and places, the changes in their faunal Princeton University Press Monographs in Population assemblages are similar. Biology. Pauley, D., and V. Christensen. 1997. Trophic levels of fishes. ACKNOWLEDGMENTS R 127 in R. Froese and D. Pauly, eds. FishBase 97: The Friends of the Virgin Island National Park, John Concepts, Design and Data Sources. Manila: ICLARM. Garrison, Director, funded this study. I gratefully Pauley, D., V. Christensen, J. Dalsgaard. R. Froese, and F. Torres, Jr. 1998. Fishing down marine food webs.acknowledge additional financial help from William Science 279:860-863. Keegan, Florida Museum of Natural History. Ken Wild Peters, R.H. 1983. The Ecological Implications of Body Size. of the U.S. National Park Service supported this research New York: Cambridge University Press. along with his volunteers. Our early inclusion in the Quitmyer, I.R., and D.J. Jones. 2000. The over-exploitation excavation process had the advantage of allowing for of hard clams (Mercenaria spp.) from five the extensiveexchange of ideas and planning of research archaeological sites in the southeastern United States. on these materials. Robin Brown and Yancy Hudson Florida Anthropologist 53(2-3): 160-167. assisted the author throughout the study. I thank Elizabeth Quitmyer, I.R., and M. Massaro. 2060. Zooarchaeology of S. Wing for her support and for making unpublished data Useppa Island. /n W. H. Marquardt, ed. Archaeology available. of Useppa Island. Gainesville: University of Florida Institute of Archaeology and PaleoenvironmentalCuration note. The National Park Service, St. Studies Monograph 3. John, Virgin Islands, will curate the Cinnamon Bay Redford, K. H. 1990. The ecologically noble savage. Orion faunal remains along with other items of material Nature Quarterly 9(3):24-29. culture. Reitz, E.J., and I.R. Quitmyer. 1988. Faunal remains from 158 ZOOARCHAEOLOGY: Papers to Honor Elizabeth S. Wing two coastal Georgia Swift Creek sites. Southeastern at the 18th International Congress for Caribbean Archaeology 7(2):95-108. Archaeology, Guadeloupe. Reitz, E.J., I.R. Quitmyer, H.S. Hale, S.J. Scudder. and E.S. Wilson, S.M. 1989. The prehistoric settlement pattern Wing. 1987. Application of allometry to Nevis, West Indies. Journal of Field Archaeology zooarchaeology. American Antiquity 52(2):304-317. 16(4):427-450. Reitz, E.J., and E.S. Wing. 1999. Zooarchaeology. Wing, E.S. 1995. Land Crab Remains in Caribbean Sites. Cambridge: Cambridge University Press. Paper presented at the XVIth International Congress Rouse, I. 1992. The Tafno: Rise and Decline of the People for Caribbean Archaeology, Guadeloupe. Who Greeted Columbus. New Haven, Connecticut: Wing, E.S. 2001 a. The sustainability of resources Used by Yale University Press. Native Americans on four Caribbean Islands. Russo, M., A.S. Cordell, L. Newsom, and S. Scudder. 1991. International Journal of Osteoarchaeology 11:112-126. Faunal Report on Horr's Island: the Archaeology of Wing, E.S. 200lb. Native American use of animals in the the Archaic and Glades Settlement and Subsistence Caribbean. Pp. 481-527 in C.A. Woods and F.E. Sergile, Patterns. Marco Island, Florida: Key Marco eds. Biogeography of the West Indies: Patterns and Developments. Perspectives, second edition. New York: CRC Press. Rutsch, E.S. 1970. Report on salvage operations during Wing, E.S., and I.R. Quitmyer. 1985. Screen size for optimal 1969 and January 1970 at Cinnamon Bay, St. John, data recovery: A case study. Pp. 49-58 in W.H. Adams, U.S.Virgin Islands. Tallahassee: National Park Service ed. Aboriginal Subsistence and Settlement Department of the Interior, Southeastern Archaeology of the Kings Bay Locality, vol. 2: Archaeological Center. Zooarchaeology: Reports of Investigations, No. 2. Sale, RR (ed.). 1991. The Ecology of Fishes on Coral Reefs. Gainesville: University of Florida, Department of San Diego, California: Academic Press. Anthropology. Schmidt-Nielsen, K. 1984. Scaling: Why Is Animal Size So Wing, E.S., and S.R. Wing. 1995. Prehistoric ceramic age Important. Cambridge: Cambridge University Press. adaptation to varying diversity of animal resources Steadman, D.W. 1995. Prehistoric extinctions of Pacific along the West Indian Archipelago. Journal of Island Birds: Biodiversity Meets Zooarchaeology. Ethnobiology 15(1):119-148. Science 267:1123-1131. Wing, E.S., and S.R. Wing. 1997. The introduction of Stoutamire, J.W., ED. Cring, K.J. Dinnel, and J.B. Haviser, animals as an adaptation to colonization of islands: Jr. 1980. Archaeological investigations at Cinnamon An example from the West Indies. Anthropozoologica Bay, St. John, U.S. Virgin Islands. Tallahassee: National 25,26:269-283. Park Service, Department of the Interior, Southeastern Wing, E.S., and S.R. Wing. 2001. Prehistoric fisheries in Archaeological Center. the Caribbean. Coral Reefs 20:1-8. Wild, K.S. 1999. Investigation of a "caney" at Cinnamon Ziegler A.C. 1973. Inference from Prehistoric Faunal Bay, St. John, and social ideology in the Virgin Islands Remains. Reading. Massachusetts: Addison-Wesley as reflected in Precolumbian ceramics. Paper presented Module in Anthropology no. 43.