A Long-term Perspective on Biodiversity and Marine Resource Exploitation in Fiji’s Lau Group1

Sharyn Jones2

Abstract: I present research investigating biodiversity and human interaction with the local environment through three perspectives on diverse islands in Fiji’s Lau Group. First, I generated long-term data on marine diversity and exploi- tation through zooarchaeological analyses of fauna from sites spanning the re- gion’s prehistoric human occupation. The study areas are representative of regional fauna and local geographic variation in island size and structure. Each island also varies in terms of human occupation and degree of impacts on marine and terrestrial environments. Second, my ethnographic work recorded modern marine exploitation patterns by Lauan communities. Third, marine bi- ological surveys documented living faunas. Together this information is used to explore the marine environment over the three millennia of human occupation. Using data derived from my multipronged study I discuss potential causes of ecological change in this tropical marine setting. My findings include the follow- ing: (1) data indicate that relative intensity of human occupation and exploi- tation determines modern composition and biological diversity of marine communities because human disturbance occurred more extensively on larger islands than on smaller islands in Lau; (2) Lauans appear to have targeted similar suites of marine fauna across their 3,000 years of history on these islands; (3) Lauans have had a selective effect on marine biodiversity because particular spe- cies are/were targeted according to local standards of ranking and preference; (4) marine resources existing today have withstood over 3,000 years of human impacts and therefore may have life history traits supporting resilience and making conservation efforts worthwhile.

Marine ecological studies are increasingly erating long-term perspectives on human- taking into account the biological complex- environmental interactions. The most pro- ity of the past to identify causes of environ- ductive scientific programs combine multiple mental change and demonstrate achievable approaches and methodologies to enhance goals for resource management and restora- understanding of environmental changes that tion ( Jackson et al. 2001, Pandolfi et al. are critical locally and globally ( Jennings and 2003). Adding an archaeological dimension Polunin 1996, Kronen and Bender 2007). expands the concept of biodiversity by gen- Resource exploitation viewed through the lens of historical ecology has been explored through recent research on marine and island ecosystems, including, for example, the work 1 Financial support for this research was provided of Thomas (2002, 2007a,b) in Kiribati, Fitz- by grants to S.J. from the National Geographic Society patrick and Donaldson’s (2007) study focused and UAB-NSF ADVANCE. Manuscript accepted 28 December 2008. on anthropogenic impacts in Palau, Rick and 2 Department of Anthropology, University of Ala- Erlandson’s (2008) broad collection of case bama at Birmingham, 1530 3rd Avenue S, Birmingham, studies on human-marine environment inter- AL 35294-3350 (e-mail: [email protected]). actions across the globe, and Steadman’s (2006) massive work reconstructing avian Pacific Science (2009), vol. 63, no. 4:617–648 historical biogeography in the tropical Pacific : 2009 by University of Hawai‘i Press islands. These works demonstrate that multi- All rights reserved disciplinary historical approaches hold much

617 618 PACIFIC SCIENCE . October 2009 promise for the future of research on island My multidisciplinary research in the Lau and marine ecosystems worldwide. Islands contributes a more robust under- Perhaps one of the biggest debates among standing of biodiversity and long-term archaeologists and biogeographers who study human interactions with the local marine en- the Pacific islands has centered on natural vironment through three perspectives on four versus human causes of environmental diverse islands. This multipronged approach change. Strong arguments have been made incorporates data derived from ethnographic, on both sides of the debate (Nunn 2000, archaeological, and marine biological re- Butler 2001, Allen 2006, Steadman 2006, search with the aims of characterizing the Morrison and Cochrane 2008). Determining reef environment, documenting species and causality from archaeological data is difficult biological diversity, and identifying potential due to equifinality, the principle that a single changes in the marine ecosystem. The Fiji Is- outcome—resource change and depletion— lands are rich in cultural and biological his- can result from different causes. Some re- tory, with human occupation extending back searchers have concluded that, ‘‘. . . scientists to approximately 3100 B.P. (Nunn et al. should stop arguing about the relative impor- 2004, Nunn 2007). Fiji’s Lau Group is an ar- tance of different causes of coral reef decline: chipelago of 80 islands, located about 200 km overfishing, pollution, disease, and climate east and 100 km south of the main Fijian is- change. Instead, we must simultaneously re- lands of Viti Levu and Vanua Levu (Figure duce all threats to have any hope of reversing 1). The Lau Group is composed of volcanic the decline’’ (Pandolfi et al. 2005:1725). Pan- and coralline limestone islands that are lo- dolfi and colleagues view the methods of cated relatively close together and support ex- historical ecology as especially valuable in tensive reef systems, rich in marine faunal providing comparisons between the past and resources. The region has an incredible diver- the present, while also providing a baseline sity of marine life and a vibrant traditional for determining if restoration efforts succeed culture that is actively engaged in marine- ‘‘in ameliorating or reversing change’’ (Pan- oriented subsistence activities, making it an dolfi et al. 2005:1726). (According to this ideal location to focus research examining framework, indicators of success would in- marine biodiversity and human exploitation clude the return or presence of key groups of marine environments over a broad tempo- such as parrotfishes, grazing sea urchins, ma- ral scale. Food is still largely locally derived in ture complex coral colonies, sharks, turtles, this contemporary Pacific community, and, large jacks, and groupers [Pandolfi et al. compared with other areas in Fiji, the coral 2005].) reefs of Lau have had relatively little impact Historical ecology fundamentally denies by tourism, extensive modern developments, an either/or dichotomy of human versus commercial fishing, and high population natural induced changes in the environment; levels. Moreover, central Lau has been the rather, environmental change should be ap- subject of archaeological and ethnographic proached through an understanding of the research by myself and others, which has laid ‘‘landscape,’’ the human-environment inter- the groundwork for this project (Hocart action sphere (Crumley 1994, Bale´e 1998, 1929, Thompson 1940, Best 1984, O’Day et Bale´e and Erickson 2006). This perspective al. 2003, O’Day 2004, Thomas et al. 2004, argues that the landscape is imbued with Jones et al. 2007a,b). meaning and is the product of the collision To explore biodiversity and marine exploi- between nature and culture, which may be ex- tation over the past three millennia, this amined as a form of architecture, or material study investigated four diverse islands in the culture (Bale´e and Erickson 2006). I adopt Lau Group, including Lakeba, Nayau, Aiwa this approach, extending the concepts of Levu, and Aiwa Lailai (Table 1). Each island historical ecology and landscape to the ‘‘sea- varies in terms of human occupation and the scape,’’ associated with islands in Fiji’s central degree of impacts on marine and terrestrial Lau Group. environments. Lakeba is the largest of the Long-term Perspective on Biodiversity in Fiji . Jones 619

Figure 1. Map of the Fiji Islands with the study area indicated in the square. study sites (55.84 km2) and is occupied by suitable for agriculture ( Jones et al. 2007b). about 2,100 people. Nayau (18.44 km2) has The archaeological sites on these four islands a population of approximately 400 people. In span central Lau’s prehistoric occupation, ex- contrast, the small islands of Aiwa Levu (1.21 tending from the Lapita period to European km2) and Aiwa Lailai (1.0 km2) are currently contact in 1774 (Best 1984, Jones 2007). uninhabited, lack a freshwater source, and are Like Lakeba, Nayau is almost entirely modi- used primarily as temporary camps for fisher- fied by human occupation, with continuous people from Lakeba (12 km away). Geologi- man-made features across the landscape (in- cally the Aiwas consist entirely of limestone cluding gardens, habitation sites, trails, and and have very little agricultural potential. resource-collection areas). The Aiwas also Lakeba is the most varied topographically exhibit signs of use islandwide, but this is and geologically, having limestone regions limited to light surface scatters of artifacts and large areas of volcanic soils and bedrock and ecofacts (stone tools, Tridacna shells, and (andesitic and dacitic lava) where freshwater midden, for example). The reefs surrounding is locally available and agriculture is practiced each of the four islands are similar in terms extensively. People living on Lakeba and of structure and potential habitat. However, Nayau maintain dryland and wetland agricul- modern development and occupation on tural crops (especially taro, cassava, sweet Lakeba are much more intensive than that potato, and yam). Conversely, Aiwa supports on Nayau (and the uninhabited Aiwas), there- diverse native forests with little if any area fore Lakeba’s inshore reefs are likely the most TABLE 1 Summary of Environmental, Physical, and Archaeological Characteristics of the Lau Study Islands

Characteristics Lakeba Nayau Aiwa Levu Aiwa Lailai

Island size and 55.8 km2 18.4 km2 1.2 km2 1km2 elevation 215þ m 160þ m 50þ m 30þ m Reef area 61 km2 20.15 km2 ca. 36.1 km2 (with multiple ca. 37 km2 (multiple fringing fringing reefs) reefs) Geology Large outcrop of Lau volcanics Geological composite of exposed Weathered raised reefal limestone Weathered raised reefal (andecites-dacites with some volcanics and weathered raised limestone basalts) and weathered reefal limestone limestone Human occupation Modern: 8 villages and the Modern: 3 villages Modern: No occupation; currently Modern: No occupation; very and archaeo- Lauan capital/chiefly seat of Prehistoric: Less-intensive used as a temporary fishing camp rarely used as a fishing logical sites power occupation than Lakeba; 34 Prehistoric: 7 sites, including two camp, although the reef is Prehistoric: Intensive archaeological sites identified, caves, a fortified hilltop occupa- frequently exploited, along occupations (gardens and consisting of inland tion, and four rockshelters of with that surrounding Aiwa habitation sites) spread rockshelters, hilltop fortified varied uses/occupation types; the Levu. continuously across the villages, and open villages on landscape exhibits some signs of Prehistoric: One rockshelter island; 209 sites identified the beach use (surface scatters of artifacts site with extensive occupa- ranging from coastal villages, and ecofacts) islandwide tion, but the landscape rockshelters, caves, and exhibits signs of use fortified hillforts islandwide Archaeological sites None; Best (1984) extensively Excavations at 13 sites; 3 m2 at 12 m2 excavated at 5 sites (6 m2 3m2 excavated at one long- excavated by the excavated 5 highly stratified the Lapita-period site of Na at Aiwa 1, the island’s largest term occupation site, Dau author and area sites and conducted test Masimasi, a deep stratified rockshelter and permanent Rockshelter; this site excavated excavations at 5 sites continuous Lapita occupation; occupation site; this site had exhibited very good an additional 13 m2 was very good preservation and preservation and a stratified excavated, and 17 shovel tests, highly stratified, deep subsurface deposit at mid-late prehistoric sites deposits) Long-term Perspective on Biodiversity in Fiji . Jones 621 heavily exploited of the study islands. It is research project (O’Day et al. 2003, Jones interesting that Lauans travel to exploit reefs 2007, Jones et al. 2007b). Archaeological near their islands. This was very likely the work has generated a basic chronology of hu- case in the past as well. For example, 17.7 man occupational history on each island. All km east of Lakeba is an area of shallow water excavations were conducted in 10 cm levels, where a reef encloses a lagoon; this reef is using trowels and following natural stratigra- called Bukatatanoa, a well-known Lakeban phy whenever possible. All sediments were fishing ground that is five times the size of dry screened through nested sieves of 1/2 the reefs surrounding Lakeba. Fisherpeople inch (12.8 mm), 1/4 inch (6.4 mm), 1/8 inch travel there in small boats almost daily. (3.2 mm), and 1/16 inch (1.6 mm) mesh. Although the most obvious and frequently Zooarchaeological analysis followed stan- exploited marine areas are those in close dard procedures described in O’Day (2001) proximity to the villages, Lauans do not look (publications by me from 2001 to 2004 are upon the sea as a barrier; rather they utilize a listed under my former name of Sharyn variety of environments and frequently move O’Day; fish bone identifications were based beyond their home shores. on all diagnostic elements [not limited to so- called ‘‘special elements’’], including atlases, materials and methods vertebrae, cranial elements, spines, scales, and ‘‘special elements’’) and utilized my com- My research contributes to the understanding parative collection housed at the University of biodiversity through three perspectives. of Alabama at Birmingham’s zooarchaeology First, zooarchaeological analyses of animal laboratory. (I have assembled a comparative bones and shells provide long-term data on collection of animal bones, especially fishes marine diversity and exploitation. Second, and invertebrates [>2,500 specimens], to ethnographic work recorded modern marine identify Pacific archaeological materials to exploitation patterns by Lauan communities. the most specific level possible. The collec- Third, I compare my archaeological and tion includes multiple individuals of the most ethnographic data to information derived common central Pacific families, genera, and from the examination of living marine faunas species and a variety of size and age classes. through detailed biological surveys of the The majority of these fishes are from Fiji reefs surrounding the study sites. Together and Lau in particular, but the collection also this information is used to understand and includes fishes and invertebrates from numer- characterize the causes and rates of ecological ous islands in Micronesia, the Philippines, change in this tropical marine context over Tahiti, Hawai‘i, Okinawa, and the Carib- the past 3,000 years. I view the data described bean.) The bones of all taxa were counted, herein as a starting point for the understand- weighed, and modifications recorded. The ing of these complicated processes. Although number of identified specimens per taxon research on these topics is preliminary, I have (NISP) is the basic specimen count used; this begun to generate a description of the marine includes bones and shells that were not environment and human exploitation patterns identified to a specific taxonomic level. The based on three datasets. At this point my pri- minimum number of individuals (MNI) was mary goal is to determine if there is evidence determined by paired elements and size and for changes in marine biodiversity and, if so, age classes whenever possible, following cal- what the relationship is between these shifts culations described by Reitz and Wing and human exploitation over time in each of (1999) and O’Day (2001). (Specifically, I esti- the island settings. mated MNI based on paired elements and evi- dence for symmetry, age and/or body sizes. MNI was estimated for the lowest taxonomic Archaeology level within a systematic hierarchy. Each pro- Since 2000, 23 sites on the four islands have venience [feature, natural stratigraphic layer, been excavated and analyzed as part of this and site]) was considered separately with at- 622 PACIFIC SCIENCE . October 2009 tention to stratigraphy when appropriate, in the sizes of collected modern fishes. In total, an effort to generate conservative MNIs. the ethnographic aspect of my research was Within a given provenience [layer, unstrati- conducted over 10 months in Fiji between fied unit, or feature] excavation levels were 2001 and 2007. (Over the course of my re- considered together to estimate MNI.) Mea- search I devoted about 3 months full-time to surements of faunal remains, particularly archaeology and the remainder of my time to the fish bones, were taken in an effort to illu- both archaeology and ethnography, with the minate long-term changes in the size and majority of my time spent doing ethnogra- makeup of exploited marine species assem- phy.) Fieldwork ranged in duration from 1 blages. The anterior width of complete fish month to 4 months. The fish catches also vertebral centra from both identified and un- serve as population cross sections of the exist- identified remains was measured for compar- ing fauna because Lauans frequently collect a ative purposes and to estimate the average wide range of inshore species with gill nets, weight and size of fishes in the assemblages. which are not size and species selective. For This procedure is based on the assumption this study, I recorded the numbers of modern that both the identified and unidentified fish taxa that are collected and eaten on each is- vertebrae represent a cross section of the spe- land. A variety of fisherpeople were inter- cies present in the assemblage (Pandolfi et al. viewed to determine what marine resources 2003, Newsom and Wing 2004). are targeted and why, and to discuss local Shannon’s H, diversity or niche breadth conservation issues. I recorded how the catch measure, was used to quantitatively determine is processed, divided up, consumed, and dis- the variety of animals consumed or exploited carded. at each of the analyzed sites over time. The Over the course of this research on formula used follows Krebs (1999:463). Lakeba, Nayau, and Aiwa, I accompanied 0 (H ¼S½ pi½loge pi is the formula for Shan- fisherpeople, including individuals and non’s H, where H 0 is the information content groups, on over 50 fishing expeditions to re- of the sample and pi is the relative abundance cord information on modern marine-oriented of individuals or resources for each taxon in subsistence activities. Different types of fish- the collection.) Because this diversity measure ing trips were documented (in notes, video re- ranges from 0 to y, it can be standardized on cordings, and still photos) at a variety of times a scale from 0 to 1 by using an evenness or and locations on all four of the study islands. equitability measure ðV 0Þ. Equitability ðV 0Þ, Following Thomas (2002:186), fishing expedi- or the evenness with which taxa were ex- tions were formally documented by collecting ploited, and the relative importance of each the following information: (1) members com- taxon was also calculated. (The equitability posing group; (2) collection strategy; (3) occa- formula used in the analysis is following Reitz sion for which the group is collecting; (4) and Wing [1999]. Because this is a compari- target prey; (5) actual prey, location, count, son of fauna representative of a wide range weight, and SL (standard length) of collected of organisms that have varied numbers of ele- items; (6) collection area (marine zone, type ments, diversity and equitability calculations description, location); (7) date, time, moon are based on MNI estimates; this places di- phase; (8) search time; (9) handling time, pro- verse organisms on a more uniform basis.) cessing time; (10) general weather and tidal V 0 is calculated on a scale of 0 to 1, thus conditions; (11) how the catch is divided up; values close to 1 indicate an even distribution (12) items consumed on shore or during fish- of taxa, and lower values suggest a dominance ing; (13) patterns of flesh and tissue disposal of one taxon. on the beach; (14) who will consume the fish.

Ethnography Biological Surveys Modern Lauan fishing expeditions provided Using both formal surveys with standardized opportunities to record species diversity and methods and informal surveys, marine biolo- Long-term Perspective on Biodiversity in Fiji . Jones 623 gists and I inventoried the marine vertebrates millennia of human activity and occupation, and invertebrates on each island (Maragos extending from 2,710 to 10 cal B.P., with and Cook 1995, DeVantier et al. 1998). A va- most of the dates suggesting site occupations riety of coastal ecosystems (inshore, forereef, before 500 cal B.P. ( Jones et al. 2007b). How- mangrove, sea grass flats, bays, fringing reefs, ever, none of these dates is from Lapita occu- and platform reefs) were surveyed and de- pations; therefore my recent determinations scribed. The majority of survey transects from the Lapita-period site of Na Masimasi were positioned on reef flats, areas that are are critical for providing a more complete frequently fished when in close proximity to chronology of the study area, with dates villages. Inshore areas located farther from ranging from the first human occupation of the villages and on the currently uninhabited the islands through the contact and historic islands of Aiwa Levu and Aiwa Lailai were periods. Radiocarbon determinations from also sampled. In total, 18 sites were surveyed Na Masimasi were provided by Beta Analytic, with line-transect methodologies on Lakeba, Inc., Miami, Florida, and are on a single bone Nayau, Aiwa Levu, and Aiwa Lailai ( Jones or piece of charcoal (Table 2). The conven- 2009a). Transects on Lakeba measured 200 tional 14C age is adjusted for 13C/12C ratios. m long and those on Aiwa Levu, Aiwa Lailai, (Calibration for atmospheric variation in 14C and Nayau measured 100 m in length from follows OxCal version 3.3 and INTCAL98 shore, and all transects extended 10 m deep. Radiocarbon Age Calibration [Bronk Ramsey Visual scans for fish extended 5 m on either 1995, Stuiver et al. 1998].) The two dates are side of the transect line and above to the sur- similar, according to a conventional 14C age face of the water. Invertebrate scans included determination, suggesting an early occupa- an area extending 1 m on either side of the tion of Nayau by about 2,580 G 40 years transect line. Five forereef scuba dives were B.P. (A recent article by Beavan Athfield et conducted around Lakeba, ranging from 18 al. [2008] discussed potential problems with to 26 m deep. In addition, 15 informal sur- the radiocarbon dating of human bones and veys were conducted on Lakeba, Nayau, suggested that these same problems apply to Aiwa Levu, and Aiwa Lailai at various tide the dating of chickens, dogs, and pigs, ani- levels and during both day and night. mals that like humans have diets of mixed terrestrial and sea biota. The authors argued that it is necessary to predetermine the type results of diet associated with both humans and do- mestic animals that are processed for radio- Archaeological Research carbon dates via isotope studies. This should radiocarbon dating and chronol- be done in advance of interpreting the radio- ogy. Earlier archaeological research on carbon results. I have submitted 61 archaeo- Nayau, Aiwa Levu, and Aiwa Lailai developed logical samples of bone from a broad range a chronology for the islands by obtaining of Lauan fauna for isotopic analysis. The re- accelerator-mass spectrometer (AMS) radio- sults are currently pending, but I plan to use carbon (14C) dates from seven sites. The the data to evaluate the Beavan Athfield et al. Nayau chronology was based on six AMS arguments in relation to data from central dates from excavations of 12 mid-late prehis- Lau.) toric archaeological sites (mid-late period ar- zooarchaeological data. To com- chaeological sites in my study area include pare local marine exploitation patterns and primarily fortified sites and rockshelter occu- prehistoric environments on each island over pations ranging in time from about 600–280 the past 3,000 years, I have summarized the cal B.P. Most of these radiocarbon dates clus- archaeological sites excavated, radiocarbon ter around 710–540 cal B.P.) and therefore dates, and the total number of fishes and extended only to ca. 700–600 cal B.P. (O’Day invertebrates identified in terms of count et al. 2003). On Aiwa Levu and Aiwa Lailai, (NISP), minimum number of individuals 11 AMS dates provided evidence of over two (MNI), measures of diversity and equitability TABLE 2 AMS Radiocarbon Dates from Nayau, Aiwa Levu, and Aiwa Lailai, Lau Group, Fiji

Unit, Layer/ Measured 14C 13C/12C Conventional 14C Beta no. Material Dated Island, Site Level Age (years B.P.) Ratio (o/oo) Age (years B.P.) OxCal Cal B.P. (2s)

235993 Metatarsal, Homo Nayau, Na Masimasi D8, II/8 2,400 G 40 14.5 2,570 G 40 2,760–2,700 and sapiens 2,640–2,610 235994 Charcoal Nayau, Na Masimasi G6, III/13 2,630 G 40 28.0 2,580 G 40 2,760–2,700 and 2,630–2,620 164249 Gallus gallus Nayau, Waituruturu E 1, I/1 470 G 40 19.5 560 G 40 650–580 and 570–510 164248 Ptilinopus porphyraceus Nayau, Waituruturu E 1, II/2 490 G 40 21.1 550 G 40 650–580 and 570–510 164247 Pteropus samoensis Nayau, Waituruturu E 1, II–III/3 610 G 40 19.9 690 G 40 690–620 and 610–550 164253 Homo sapiens Nayau, Qaranilulu 1, I/2 550 G 40 15.7 700 G 40 710–620 and 610–550 165468 Pteropus tonganus Nayau, NukutubuRS2 1, II–III/3 100.6 G 0.8 19.2 50 G 60 280–180 and <150 % modern C 173059 Homo sapiens Nayau, NukutubuRS2 1, IV/1 280 G 40 16.6 420 G 40 540–420 and 380–320 164251 Pteropus tonganus Aiwa Levu, 2, I/1 180 G 40 19.2 280 G 40 470–280 (.91) Rockshelter 1 170–150 (.04) 164258 Pteropus samoensis Aiwa Levu, 4, II/3 260 G 40 19 360 G 40 510–310 (.95) Rockshelter 1 164260 Wood charcoal Aiwa Levu, 2 (oven), II/5 210 G 40 25.7 200 G 70 430–360 (.06) Rockshelter 1 330–10 (.89) 164261 Wood charcoal Aiwa Levu, 2 (oven), III/10 580 G 40 25.6 570 G 40 650–510 (.95) Rockshelter 1 164252 Homo sapiens Aiwa Levu, 2, III/13 2,090 G 40 11.8 2,310 G 40 2,440–2,410 (.01) Rockshelter 1 2,370–2,290 (.68) 2,270–2,150 (.33) 165466 Pteropus samoensis Aiwa Levu, Cave 2 2, I/1 860 G 40 18.9 960 G 40 960–760 (.95) 165467 Pteropus samoensis Aiwa Levu, Cave 2 3, II/2 1,530 G 40 19.1 1,630 G 40 1,690–1,670 (.02) 1,610–1,410 (.94) 165465 Gallus gallus Aiwa Levu, Cave 2 1, II/2 2,180 G 40 12.5 2,380 G 40 2,710–2,630 (.16) 2,620–2,590 (.01) 2,500–2,330 (.78) 165469 Ducula pacifica Aiwa Levu, Goat 1, II/5 370 G 40 20.3 450 G 40 550–430 (.92) Rockshelter 360–330 (.03) 172192 Pteropus tonganus Aiwa Lailai, Dau 1, IIIb/4 1,410 G 40 10.5 1,510 G 40 1,520–1,310 (.95) Rockshelter 172191 Gallus gallus Aiwa Lailai, Dau 2, IV/5 2,060 G 50 19.1 2,300 G 50 2,460–2,150 (.95) Rockshelter Long-term Perspective on Biodiversity in Fiji . Jones 625

(H and V ), and the most common fish fami- for bony fishes (Figure 2). In contrast, the lies identified (Table 3) as these data relate to NISP from the site (7,570) is very high and general periods of prehistory. (The NISP abundant enough to interpret the marine listed in Table 3 includes fish bones that faunal assemblage with some degree of confi- were not identifiable below the level of class; dence (Reitz and Wing 1999). The low num- the percentage of specific fish identifications, ber of MNI from Na Masimasi is due to an that is, to the level of family and below, varies extremely high frequency of vertebrae in the from 7% to 26%, depending on the site. For assemblage and the somewhat preliminary example 23% of the fish bone from the mid- nature of fish bone identifications. I antici- late period sites from Nayau was identifiable pate that the fish MNI will increase meaning- to the level of family, genus, or species.) The fully when my analysis is complete. same classes of data are provided for inverte- Nevertheless, the other sites yielded ample brates in Table 4. These tables provide com- MNI, and all of the sites produced sufficient parisons of the marine resources that were numbers of NISP to conduct interpretative exploited prehistorically over the entire pre- analyses. There are four major trends that historic human occupation of the study sites; are evident in the faunal data. the detailed individual site data are published First, the identified vertebrate taxa (and es- in O’Day et al. (2003), Jones et al. (2007a), pecially the families of fishes exploited) from and in a forthcoming manuscript currently the early to late period sites are dominated by under review ( Jones 2009b). Although the tangs, grouper, parrotfishes, triggerfishes, and argument could be made that the amalgam- emperorfishes (Table 3, Figures 2–5). These ation of data from multiple sites, especially families are common in Pacific island archae- the 12 mid-late period Nayau occupations, is ological assemblages and in material identi- too general to contribute useful information, fied from Fiji (Leach and Davidson 2000, I argue that the summary H and V 0 values Thomas et al. 2004, Jones et al. 2007b). Por- are both relevant and appropriate for the cupinefish occur in high frequencies in the following reasons. Intersite variation of early period sites on Lakeba and Nayau but identified fish remains is minimal within the were less frequently identified in sites from time periods. Moreover, within the mid-late mid-late time periods and in deposits from Nayau assemblages, 80% of the identified other islands. Second, for fishes, the V 0 values fish bones by count were recovered from two suggest that there is an even distribution of virtually contemporaneous fortified rockshel- taxa (note that values close to one indicate an ter occupations (Waituruturu East and Wai- even distribution). The marine vertebrate turuturu West produced the majority of the taxa exploited remained even over time, and mid-late period Nayau fish remains; these diversity ðHÞ was moderate to high, ranging sites also contributed over half of the MNI, from 0.88 to 2.03. The number of taxa re- 87 individuals [see O’Day et al. 2003].) The mains high across temporal periods (28–33). mid-late period sites include fortified occupa- Third, for invertebrates, the diversity and tions and rockshelters, all of which likely had equitability measures are variable (H ranges similar functions and durations of occupation. from 0.88 to 2.03 and V 0 ranges from 0.25 Finally, the mid-late Nayau occupations all to 0.6), much more so than in the fish data occur inland, away from the coast, and accord- (Table 4). Sites associated with the early ing to the zooarchaeological analysis the occu- time periods on Nayau and Aiwa produced pants of all these sites exploited the inshore assemblages dominated by particular species areas as their primary animal food source. (Figure 6); for example, Na Masimasi’s inver- The comparative abundance of fauna in tebrate assemblage is primarily composed of terms of primary and secondary data is evi- the small fighting conch (Strombus gibberulus). dent in the total amalgamated NISPs and On Aiwa Levu and Aiwa Lailai, the inverte- MNIs from sites representative of each time brates are dominated by turban snails (Turbo period. The Lapita site of Na Masimasi on spp.) throughout the sites (Figures 7 and Nayau has a relatively low overall MNI (59) Figure 8). Aiwa’s sites have high diversity TABLE 3 Summary of Fish Bone Data from Archaeological Sites on Aiwa and Nayau (Sites Listed from Early to Late Period)

Site(s) Island Agea Site Type NISP MNI No. of Taxa H, V 0b Dominant Fish Familiesc Reference

Aiwa 1 (units 1–4, Aiwa Levu Early (2,500– Rockshelters 3,138 110 28 3.29, 0.99 Acanthuridae (tangs), Jones et al. III–IV ), Dau and Lailai 960 cal Scaridae (parrotfishes), (2007b) Rockshelter, B.P.) Labridae (wrasses), Cave 2 Serranidae (groupers) Na Masimasi Nayau Early (2,610– Beach site/ 7,570 59 33 3.21, 0.92 Tangs, groupers, O’Day et al. 2,760 cal Lapita Diodontidae (2003) B.P.) occupation (porcupinefishes), parrotfishes 12 sites Nayau Mid-late Fortified 3,054 175 31 3.34, 0.97 Tangs, Balistidae O’Day et al. (600–280 sites and (triggerfishes), (2003), Jones cal B.P.) rockshelters groupers, parrotfishes, et al. (2007a) Lethrinidae (emperorfishes) Aiwa 1 (units 1–4, Aiwa Levu Late (550– Rockshelters 1,665 86 30 3.27, 0.96 Tangs, groupers, Jones et al. I–II and all of 280 cal parrotfishes, (2007b) units 5–6), Goat B.P.) triggerfishes, Rockshelter emperorfishes

a Age ranges based on AMS radiocarbon dates with OxCal cal B.P. correction. b H, Shannon-Wiener measure of diversity; V 0, equitability value. c Fishes listed in rank order of abundance. Long-term Perspective on Biodiversity in Fiji . Jones 627

TABLE 4 Summary of Invertebrate Data from Archaeological Sites on Aiwa and Nayau (Sites Listed from Early to Late Period)

No. of Dominant Site Island Age Type NISP MNI Taxa H, V 0 Invertebrate Taxaa

Dau Aiwa Early (2,460– Rockshelter 1,804 782 18 1.71, 0.6 Turbo sp., Modiolus Rockshelter 1,310 cal auriculatus, B.P.) Cellana spp. Na Masimasi Nayau Early (2,610– Beach site 1,732 1,026 32 0.88, 0.25 Strombidae (esp. 2,760 cal Strombus B.P.) gibberulus), Turbo spp. 12 sites Nayau Mid-late Fortified 2,625 1,703 35 2.03, 0.57 Turbinidae (esp. (600–280 sites and Turbo spp.), cal B.P.) rockshelters Cypraea spp., Strombus spp., Modiolus auriculatus Aiwa 1 Aiwa Mixed Rockshelter 2,726 1,001 21 1.63, 0.54 Turbo sp., Cellana (2,370–280 spp., Chitonidae cal B.P.)

Note: Biomass was based on the weight of the shell. a Invertebrates listed in rank order of abundance. levels comparable with those of the mid-late fishes. The number of taxa represented in period Nayau sites (Figure 9). Exploitation the invertebrate assemblages varies from 18 patterns in terms of diversity and equitability to 35; again, this is more variable than that for invertebrates are different than those of documented in the fish assemblages.

Figure 2. Early Nayau fish remains from the Lapita site of Na Masimasi, listed by family. 628 PACIFIC SCIENCE . October 2009

Figure 3. Mid-late Nayau archaeological fish taxa, listed by family.

Figure 4. Early Aiwa Levu and Aiwa Lailai fish remains, listed by family.

Fourth, it is surprising that measurements earliest occupations sites were actually smaller of more than 2,500 fish vertebrae, used as a than those consumed later in prehistory proxy for fish body size and weight, reveal (Table 5). This trend is particularly evident that the fishes collected and consumed at the on Nayau where the Lapita site of Na Masi- Figure 5. Mid-late Aiwa Levu and Aiwa Lailai fish remains, listed by family.

Figure 6. Early Nayau shellfish taxa from the Lapita site of Na Masimasi, listed by family. Figure 7. Early Aiwa Lailai shellfish remains, listed by family.

Figure 8. Summary of shellfish from the site of Aiwa 1, Aiwa Lailai, representing mixed time periods. Data listed by family. Long-term Perspective on Biodiversity in Fiji . Jones 631

Figure 9. Mid-late Nayau shellfish taxa, listed by family.

masi contained the majority of small bones in each context [sieves with 1/2 inch (12.8 (mean anterior vertebral centra width 2.98 mm), 1/4 inch (6.4 mm), 1/8 inch (3.2 mm), mm and mean estimated weight 118 g), rep- and 1/16 inch (1.6 mm) mesh].) Na Masimasi resenting small-bodied fishes. (I do not attri- also produced the broadest range in the size bute this variation in the size of fishes to of fishes with vertebral centra measuring differences in recovery of fauna because I 0.61–20.2 mm, although the vast majority of directed the excavation and screening of the fish vertebrae are small. Mid-late period sites study sites and used the same nested screens on Nayau produced a smaller range in the

TABLE 5 Fish Vertebral Centrum Widths (mm) from Early and Mid-Late Period Occupations on the Study Islands

Standard Mean Estimated Provenience n Mean Range Deviation Weighta (g)

Early Nayau 1,432 2.98 0.61–20.2 1.76 118 Early Aiwa 281 4.65 1.5–13.0 2.05 363 Mixed Aiwa 174 4.2 1.7–14.3 2 281 Mid-late Nayau 464 4.3 3–18.7 3.9 298 Mid-late Aiwa 165 4.85 2.04–13.33 1.9 404 Total 2,516 4.2 0.61–20.2 — 281

a Estimates of total body weight (g) are made using the allometric formula log Y ¼ 2:53ðlog XÞþ0:872, where Y is the body weight and X is the vertebral width, following Newsom and Wing (2004:71). 632 PACIFIC SCIENCE . October 2009 size of fishes exploited, a mean estimated areas that range in depth and substrate. They weight of 298 g, and a mean vertebral centra utilize their nets in various ways, for example, size of 4.3. The fish remains from Aiwa ex- to corral the fish or to chase the fish into hibit a less-dramatic variation in size over the nets. Net fishing is conducted at various time (the mean estimated weight is 363 g times, day and night, and in every season. with a 4.65 mm centra width at the early When and where people fish depends on the period sites and 404 g and 4.85 mm in the occasion that the fisherpeople are collecting mid-late period sites). for, the target taxa (if any), the moon phase, and the tidal phase, among other social issues. The important point is that net catches pro- Ethnography vide an extensive array of fish taxa that may Over 50 fishing expeditions were documented be used to understand biodiversity of bony to record information on modern marine- fishes inhabiting the inshore area. oriented subsistence activities. Eleven fishing ethnographic interviews. A number expeditions were recorded on Lakeba, and of important issues were brought up in seven expeditions were recorded on Aiwa ethnographic interviews with fisherpeople on Levu and Aiwa Lailai. A total of 37 hr of fish- Lakeba and Nayau (Table 6). Findings in- ing was recorded (about 155 person-hours) clude information on a wide range of issues, on Lakeba. On Nayau, nine fishing trips, cov- such as the style of fishing particular to cer- ering a total of 20 hr (1,140 person-hours), tain environments and villages, variations in were formally documented. the way men and women fish, the types of Ethnographic investigations on the island fishes preferred by Lauans, and shifts in the of Lakeba in August 2007 confirmed that local availability of certain resources. Lauans collect and consume virtually every Traditional vono style fishing is the pri- fish that is caught by nets or hand lines in mary mode that people in the northern vil- the inshore area. Shellfish are sometimes col- lages on Lakeba (Nasaqalau and Vakano) use lected and consumed on the reef, but they to fish inshore; they state that this has been form a very minor component of the catch the case for many generations. The method and marine portion of the diet overall. These produces thousands of small inshore fishes findings support conclusions derived from over the course of 2 to 3 days (especially previous research on Nayau (O’Day et al. rabbitfishes, or Siganus spp.; emperorfishes/ 2003, O’Day 2004, Jones 2007). The total Lethrinidae; and convict tangs). Vono is docu- number of fish taxa recorded in modern fish- mented in detail in a short ethnographic ing expeditions is 112; 105 of these fishes are film that I and other individuals involved eaten, and only seven species are bycatch, in this research project helped to write and which are not consumed (see Appendix). I re- produce. (The fishing expedition docu- corded the numbers of modern taxa that are mented in the film can be viewed on Google collected and eaten on each island; the results Video at http://video.google.com/videoplay? are as follows: Aiwa Levu, 24; Aiwa Lailai, 15; docid=6257583410432053434&q=vono+fiji& Lakeba, 49; Nayau, 94. (Obviously, these data ei=q6eQSOPfAYSGqwPZtt2iCA.) In mod- describe only a portion of the taxa exploited ern Fiji, this traditional fishing method is and are limited by the fact that I recorded only practiced in the Lau Group on Lakeba. fishing and collection over the duration of Like all inshore fishing in the region, this my fieldwork rather than a lifetime of obser- complex net method is organized and run en- vation.) These data also suggest that the tirely by women. Many Lauan women spend majority of fish taxa are collected with nets part of each day on the inshore reef, collect- inshore (85.7% of the 112 fishes). ing marine resources. However, in each vil- Lauan fishing expeditions that utilize nets lage there is a clear division between the ‘‘sea inshore produce a broad cross section of the people’’ (the fisherpeople, yavusa wai or kai existing inshore species. On a single expedi- wai) and ‘‘land people’’ (primarily farmers, tion the fishing group will exploit numerous who rarely if ever fish, the yavusa vanua, or TABLE 6 Shifts in Marine Resources as Recorded in Ethnographic Interviews on Lakeba, and Resources Specific to Various Villages

Village

Shift Tubou Nasaqalau Vakano

Resources no longer Large-bodied parrotfishes, large Large-bodied mullets (Mugil cephalus and Valamugil Bumphead parrotfish (Bolbometopon available or in mullet, mangrove crabs, giant buchanani), coral cod (Cephalopholis spp. and muricatum), large (>50 cm TL) sweetlips marked decline clams (Tridacna spp.), strombids Epinephelus sp.), sea cucumbers (Holothuroidea)a, (Plectorhinchus spp., including P. picus), large (Strombus spp.), spider shell mangrove mussels (Modiolus agripetus), bent nose eels (Muraenidae), midnight snapper (Lambis lambis), top shells clam (Tellina palatum), mangrove clam (Gafrarium (Macolor macularis), mullet (all local species), (Trochus niloticus), Turbo (Turbo tumidum), clams (Periglypta sowerbyi and P. large barracuda (Sphyraena spp.), crabs, spp.), trumpet triton (Charonia puerperal ), cockles (Anadara cornea), spider shell giant clams (Tridacna spp.), strombids tritonis) (Lambis lambis), threespot reef crab (Carpilium (Strombus spp.), spider shell (Lambis lambis) maculates), redeye crab (Eriphia sebana), slipper lobster (Parribacus caledonicus); painted rock lobster (Panulirus versicolor) Other changes, People claim that there are fewer More men are collecting in the inshore area now In general there are fewer large fish to be including social large fish now (versus 10 years than they did in the past. In particular, men and caught in the inshore area; the change is ago), and they can see that the boys will spend many hours each week with spears, noted over the last 20 years. The chief and sea level is rising, coral is dying, small nets, and their hands collecting crabs, small men of the village claim that spear fishing is and the ocean is becoming fish, and occasionally shellfish in the mangroves still highly productive. People used locally warmer. These changes are noted between Nasaqalau and Vakano. People claim that handmade natural fiber nets and line until both inshore and offshore. there are fewer large fish inshore now (versus 10– about 10 years ago when people shifted to 20 years ago); they can see that the sea level is using microfiber nets and line. rising, coral is dying, and the ocean is warmer. Marine resources Mullet, herrings, crescentbanded Coral cod (Cephalopholis spp.), turtles (Cheloniidae). Mangrove crabs (Scylla serrata and Cardisoma and techniques grunters, and sea grapes (Caulerpa All forms of fishing techniques are used here. cardifex), swimming crab (Thalamita crenata), known to be racemosa). All forms of fishing However, women typically participate in vono sea urchins (Tripneustes gratilla), top shells productive in techniques are used here and expeditions (1–2 times each week) or fish alone or (Trochus niloticus), Turbo (Turbo spp.), bent each village many people have motorboats, in small groups using spears, hooks, and handlines. nose clam (Tellina palatum), mangrove clam more than in other villages in Men sometimes throw a spear to hit a fish; this is (Gafrarium tumidum), clams (Periglypta central Lau. People do more especially common in the mangrove areas. Men sowerbyi and P. puerperal ), cockles (Anadara offshore trolling from Tubou and also troll offshore and spearfish, both inshore (at cornea), pearlshell (Pinctada martensi and the they travel to Aiwa frequently. night) and offshore (night or day). Nasaqalau is large-bodied Pinctada maxima), coral cod They often catch large-bodied known for sharks and the practice of shark calling (Cephalopholis spp.), turtles (Cheloniidae). offshore fishes on the reefs far (done by female priests in one clan). This was last People fish vono style and with spears. They outside Tubou (fishes such as done in 1992 for a BBC film, but no sharks came. preserve the fish by drying and smoking it tuna, barracuda, trevally and It was last practiced successfully in 1967. The on a vesa, or fish dryer. The mangroves are jacks, wrasses, and sweetlips). mangroves are highly productive. highly productive.

a Most of the sea cucumbers are exported from Lau to the markets on Viti Levu. People eat them occasionally on Lakeba and Nayau. 634 PACIFIC SCIENCE . October 2009 kai vanua). Women of the sea people clan are larger villages on Lakeba, such as Tubou, are the keepers of traditional marine ecological in more obvious decline. Lauans, and espe- knowledge (TEK). They hold a wealth of in- cially the fisherpeople, are keenly aware of formation about marine ecology and biota, worldwide environmental shifts such as global and have intimate knowledge of the natural warming, the heating of the oceans, and sea- order as well as changes and fluctuations in level rise. The long-term sustainability of the system. According to informants, the sea Lauan fishing practices is something that all people have always been the fisherpeople of the fisherpeople I interviewed on Lakeba Lau, from the time of the first human occu- mentioned, and in each village on Lakeba pation. They are the descendents of the first there are active representatives from the Fiji Lauan fishers and the people who collected Ministry of Fisheries. This is not the case on the marine resources that occur in ancient ar- Nayau, where there is comparatively little chaeological sites. I hypothesize, although it discussion of conservation and the effects of would be difficult or impossible to prove, global warming; however, people are aware that women were the primary producers of of changes in the marine environment. The inshore marine resources in prehistory, just long-term impacts of intensive marine har- as they are in the present. vesting, fishing, and collecting are a frequent Small inshore fish species make up the subject of conversation among the village largest portion of all animal protein con- chiefs and fisherpeople. sumed by Lauans each day. For example, the convict tang (A. triostegus triostegus) is a favor- Biological Surveys ite fish of Lakebans. It is small, measuring an average of only 18–22 cm in total length Over 200 species of marine vertebrates (TL). This fish is locally considered to be as- (sharks, stingrays, sea snakes, bony fishes, sociated with the high chief of Lakeba and and turtles) from over 50 families, and almost the Lau Islands, or king, the late Tui Lau, be- 200 species of macroinvertebrates, were ob- cause he was very fond of this fish and com- served around the study sites. The reefs of missioned vono expeditions whenever he was Aiwa Levu and Aiwa Lailai have the widest on Lakeba. People on Nayau also claim to range and abundance of observed vertebrates love eating the convict tang ( Jones 2009a). and invertebrates (Table 7). The detailed Interviews indicate that the majority of sur- results of this biological survey will be pub- vey respondents on Nayau favor inshore reef lished in the future. Most of the identified species that are typically small-bodied in the fishes are relatively common in the waters of modern environment (a35 cm in TL). The the tropical Pacific (Myers 1991) and belong single most sought after fish among women I to the most abundant coral reef families in- interviewed is a group referred to collectively cluding the following: Acanthuridae, Chaeto- as kawakawa (small groupers, including Epi- dontidae ( butterflyfishes), Labridae (wrasses), nephelus merra and Cephalopholis sp., the Lethrinidae, Mullidae (goatfishes), damsel- honeycomb and hind groupers). Epinephelus fishes (Pomacentridae), Scaridae, Lutjanidae merra in Fiji measure 20 cm (TL) on average (snappers), and Serranidae. On average, the when caught, and Cephalopholis averages 24– inshore scuba and snorkel surveys identified 40 cm in TL (Froese and Pauly 2006). 41 fish species on Lakeba (per 1,000 m2 Most of those interviewed from Lakeba area), 55 species on Nayau (per 500 m2), 67 and Nayau stated that they have noticed a de- species on Aiwa Levu (per 500 m2), and 66 cline in the local availability of some fishes species on Aiwa Lailai (per 500 m2). and invertebrates (Table 6). For instance, the The presence of juvenile fishes in the in- villagers of Vakano and Nasaqalau on Lakeba shore areas varies according to differences in complained that people from other villages the marine environments, including the tide, frequently come to their inshore reefs to col- the moon phase, and the substrate. On the lect fish. This practice is becoming increas- north side of Lakeba, villagers have planted ingly common as the marine resources in extensive areas of mangrove trees. This ap- Long-term Perspective on Biodiversity in Fiji . Jones 635

TABLE 7 Summary of Marine Biological Survey Data from Lakeba, Nayau, Aiwa Levu, and Aiwa Lailai

Survey Type, Transect Total No. of Island Survey Site Measurement Fish Fish Species

Lakeba Oru Scuba, 200 m/1,000 m2 988 54 Lakeba Tubou Snorkel, 200 m/1,000 m2 711 37 Lakeba Nasaqalau Mangrove Walked, 200 m/1,000 m2 37 2 Lakeba Waciwaci Snorkel, 200 m/1,000 m2 807 33 Lakeba Nukunuku Mangrove Walked, 200 m/1,000 m2 27 3 Lakeba Vakano Mangrove Walked, 200 m/1,000 m2 210 16 Lakeba Selesele Point Forereef scuba, 18 m maximum depth 366 54 Lakeba Vakano Forereef scuba, 20 m maximum depth 339 54 Lakeba Ucuiboagi Point Forereef scuba, 26 m maximum depth 430 64 Lakeba Nasaqalau Forereef scuba, 20 m maximum depth 480 64 Lakeba Oru Forereef scuba, 20 m maximum depth 357 62 Aiwa Levu Aiwa 1 Scuba, 100 m/500 m2 938 76 Aiwa Levu Aiwa 2 Scuba, 100 m/500 m2 651 58 Aiwa Lailai Aiwa LL1 Snorkel, 100 m/500 m2 703 60 Aiwa Lailai Aiwa LL2 Snorkel, 100 m/500 m2 784 72 Nayau Salia Snorkel, 100 m/500 m2 618 49 Nayau Narocivo Snorkel, 100 m/500 m2 831 66 Nayau Liku Snorkel, 100 m/500 m2 587 50 pears to have increased the local abundance the reefs of central Lau were affected by of fishes on the reef flats. Juveniles of several bleaching was in 2003 (Leon Zann, Univer- species were frequently recorded in close sity of the South Pacific, pers. comm., 2006). proximity to the planted mangrove trees, and (Apparently Lauan reefs were minimally af- they are especially abundant around Aiwa fected or unaffected by the 1997/1998 El Lailai and in the intertidal area of Aiwa Nin˜ o–Southern Oscillation mass-bleaching Levu. Aiwa Levu’s mangroves also act as event.) nurseries for juvenile fishes. Around Nayau, On both Lakeba and Nayau the reefs gen- the presence of young fishes appears to be erally exhibit a notable lack of gorgonians highly dependent on the tide, but the inshore and neptheids (soft tree corals, finger soft flats around all three of Nayau’s villages boast corals, etc.). The environments evidence se- high frequencies of juveniles in the families vere leather soft coral diebacks (Sinularia, listed here. Sarcophyton, and Lobophytum spp.). The scale The effect of harvesting invertebrates is and condition of the coral bleaching and clearly visible on most of Lakeba and Nayau’s death on Lakeba and Nayau is so severe that reef flats. Species of giant clam (Tridacna although local fishing pressures over the last spp.) and large snails (e.g., Trochus spp.) are 3,000 years have undoubtedly affected it, re- rarely encountered (note that when large- cent climatic shifts are likely to blame. Never- bodied invertebrates such as Tridacna and theless, small coral recruits were documented Trochus are encountered by Lauans they are and thus natural partial recovery of once immediately collected and often consumed larger colonies is beginning at all the study right on the reef ). In contrast, on Aiwa Levu reefs. Despite the poor state of much of the and Aiwa Lailai, where regular harvesting is coral on these islands, substantial populations less common, several species that are rare on of fishes were observed in pockets of healthy Lakeba flourish (Tridacna spp.). The majority reef and in association with the recovering of forereef dive sites had less than 5% coral reef colonies, including large-bodied species cover, which is likely due to a recent bleach- of groupers, juvenile groupers and parrot- ing event. Since records began, the first time fishes, wrasses, large-bodied parrotfishes, and 636 PACIFIC SCIENCE . October 2009 a small number of sea turtles. Large-bodied Masimasi may be indicative of two distinct fishes were relatively uncommon in the in- fishing technologies utilized contemporane- shore areas and common but not abundant ously, such as trolling (resulting in the cap- on the forereef of all the study sites. Reef ture of large-bodied fishes) and inshore net sharks and jacks are most frequently observed fishing (resulting in the small-bodied fishes). around the inshore reefs of Aiwa. These ani- The most obvious differences are in the mals were also documented off Nayau’s north taxa exploited and changes in the composition and south forereefs. of the assemblages through time. There is little temporal variation in the overall num- discussion bers of fish taxa exploited, regardless of the NISP and calculated MNI. By comparing This research has produced important results, fish assemblages with measures of diversity including descriptions of modern exploitation and equitability, it becomes apparent that the patterns, extant marine biodiversity, and the taxa exploited remained relatively even over outcomes of inshore fishing expeditions. In time, and diversity ðHÞ was moderate to addition, ethnoarchaeological studies have high. More variation is evident in the inverte- illuminated several long-term trends in ma- brate assemblages, but the shifts are minor rine resource exploitation and biodiversity. with the exception of the early exploitation There appears to be relatively little variation of fighting conch (Strombus gibberulus)on in the types of marine vertebrate species ex- Nayau at Na Masimasi. On Aiwa Levu and ploited over time. The vertebrates are rela- Aiwa Lailai, the invertebrate assemblages tively diverse and do not show a marked are consistently dominated by turban snails decline in size according to the zooarchaeo- (Turbo spp.). The invertebrate pattern may logical data. However, both the invertebrate be attributed to preference; that is, the first archaeological assemblages and modern in- inhabitants of the islands selected choice in- vertebrate communities provide evidence for vertebrates based on taste and/or ease of ac- shifts in exploitation and availability of ma- cess to these items. For example, the inshore rine resources over time. area by Na Masimasi is composed of coral reef on sand and limestone flats, exactly the habitat preferred by fighting conch. It should Archaeology be noted that today both fighting conch and Laboratory analysis on archaeological bone turban snails are relatively rare on Lakeba and shell samples from Nayau and Aiwa and slightly less rare on Nayau and the Aiwas has contributed to the understanding of the (this is based on biological surveys and ethno- makeup and sizes of exploited marine taxa. graphic data documented during fishing expe- Detailed analysis of over 24,000 fish bones ditions in the inshore area). Lauans claim that and shellfish remains indicates that the these taxa were more abundant 20–50 years indigenous inhabitants of the study areas in- ago. The change in prehistoric and modern tensively exploited relatively small-bodied availability could be due to a long history of inshore marine resources over the course of intensive exploitation, resulting in overexploi- their 3,000-year history. However, to date, tation, or a combination of environmental there is no solid zooarchaeological evidence and human-induced changes. However, in suggesting a major decrease in the overall the absence of data supporting a species-size sizes of exploited fishes and invertebrates reduction in middens over time it is difficult over time. Rather, the data from fish vertebral to distinguish human impacts from the effects measurements suggest an increase in the sizes of natural disturbance on invertebrate popu- and weights of fishes exploited on Nayau lations or decreased shell bed density. Ar- from the early to mid-late time periods, and chaeologists have identified sites exhibiting relatively little change on Aiwa Levu and reductions in shellfish abundance that are Aiwa Lailai. On Nayau this bimodal distribu- likely the result of terrestrial/agricultural de- tion pattern in the Lapita-period site of Na velopment, including increased siltation into Long-term Perspective on Biodiversity in Fiji . Jones 637 the marine environment (Kirch and Yen A number of complex ecological processes 1982, Spennemann 1987). Unfortunately, in may be related to the documented changes in most situations there is no simple explanation fish and shellfish exploitation. These include, for the presence or absence of a marine spe- but are not limited to, variations in fishing cies in a given habitat because this depends technologies, shifting climatic conditions, re- heavily on natural stochastic recruiting events cruitment, and changing predator-prey rela- (Sale 1980). tions. Undoubtedly the marine biota of Lau habitat. Between 80% and 95% of the experienced both natural and human-induced identified fishes from all the study sites in- shifts across space and time; these phenomena habit inshore coral reef environments (by are well documented throughout the Pacific measures of NISP and MNI). The vast ma- islands (Allen 2006). Allen (2006) discussed jority of the identified invertebrates inhabit new data from long-lived Pacific corals and the intertidal reef, the splash zone above general climate modeling for the central Pa- the high-tide line, tide pools, sand flats, and cific. She described climatic shifts in the re- fringing reefs. Turban snails (Turbinidae), gion and areas of particular instability. The the most common invertebrate at all the sites, data suggest increased El Nin˜ o–Southern Os- with the exception of Na Masimasi where cillation (ENSO) variability from A.D. 1100 fighting conchs are more frequent, inhabit to 1400, with a shift from cool conditions coral reef habitats and rocky coral areas. after A.D. 1200 and a short warm period at Turbo setosus, the most frequently identified about A.D. 1300 (Allen 2006:530). In Fiji the Turbo on Aiwa, is known to prefer exposed transition from the Little Climatic Optimum areas of coral reef and the sublittoral zone in (LCO) (1250–700 B.P.) to the Little Ice Age shallow waters (it accounts for 40–60% of the (LIA) (700–200 B.P.) at around 700 B.P. was invertebrate assemblages by measure of MNI a period of frequent and intense ENSO and NISP). The identified bivalves can be activity, resulting in increased storminess, found in shallow-water habitats, including low-pressure systems, sea-level decline, and silty or sandy inshore areas on fringing reefs heightened sea temperatures (Nunn 2000, (Kay 1979, Colin and Arneson 1995). Bivalves Hughes et al. 2003, Field 2004). Increased make up a small portion of all the zooarch- temperatures and ocean salinity are known to aeological assemblages (<20% at any site). heavily impact coral reefs through bleaching In his analysis of archaeological marine re- and, ultimately, tremendous coral die-offs as sources from Lakeba, Best (1984:498) argued seen in modern ENSO events (Hughes et al. that there is little evidence of any noteworthy 2003, Allen 2006). The high level of decline changes in the fish taxa exploited over a in marine productivity due to these natural 3,000-year sequence of data. Best’s conclu- climatic events is extreme and may outweigh sions are supported by the archaeological the potential negative impacts that relatively data in this study. One possible explanation small populations such as those occupying for this phenomenon is that given fairly small Aiwa and Nayau at any given time could and stable human population sizes over time, have exerted upon their rich marine environ- the Lauan data represent a relatively stable ments (Thomas makes a similar observation system of exploitation, where some of the in his paper in this volume, regarding the po- traditional fishing methods utilized by Lauans tential marine impacts of small populations and the animals exploited (e.g., convict tangs inhabiting atolls in Kiribati. However, this and rabbitfishes that are targeted in vono view and that of Allen stand in contrast to expeditions have short population-doubling work by Pandolfi and colleagues [2003:957], times) are suited to long-term sustainability who argue that their data trajectories of de- (also see Jennings and Polunin 1996). Al- cline in fisheries worldwide through time are though this may be true for some bony fishes, similar and that ‘‘All reefs were substantially the exploitation and availability of particular degraded long before outbreaks of coral dis- shellfish species appears to have decreased ease and bleaching,’’ and therefore, human through time. overfishing, along with associated land-based 638 PACIFIC SCIENCE . October 2009 pollution and runoff, is the only reasonable tions are coordinated and directed by elder explanation for the pre-1900 decline of women, who pass knowledge of how and worldwide coral reefs); additional faunal data where to fish on to the younger generations. from sites dating to the LCO-LIA transi- I worked closely with these fisherwomen in tion and fine-grained radiocarbon dates are an attempt to learn what marine resources needed to further illuminate this issue. are harvested and how the diversity of avail- In D’Arcy’s (2006) recent book he dis- able fishes and invertebrates has changed. Ar- cussed potential impacts of highly unpredict- chaeological fish bones indicate that Lau’s able annual and seasonal climatic variations first inhabitants intensively harvested rela- on the lives of Pacific islanders inhabiting Re- tively small inshore fishes, just as their de- mote Oceania. In terms of fishing patterns, scendants do today. There appears to be a biomass of fish taxa also changes annually direct correlation between the fish taxa that and over longer-term periods, with variability Lauans claim to prefer and those that are visible in terms of individuals. ‘‘The fact that actually collected and consumed. The evolu- instability in individual species’ numbers can tionary perspective afforded by combining occur alongside stability in the overall size of multiple perspectives has illuminated long- the ecosystem biomass suggests that there can term trends in Lauan exploitation and prefer- be continuity in marine harvests for fisher- ences. It is possible that some of the Lauan folk not rigidly tied to harvesting a few traditional systems for marine exploitation species only’’ (D’Arcy 2006:21). In a highly and management allow for long-term sustain- variable environment flexibility in terms of ability (especially in the case of small-bodied technologies, habitats exploited, and targeted inshore bony fishes), but others, such as the catch appears to be a great advantage. In par- harvesting of shellfish, do not. Pandolfi et al. ticular, net technologies are well suited to (2005) suggested that indicators of a healthy capture a broad spectrum of available inshore reef and successful management might in- species. I argue that Lauan fisherpeople’s clude the presence of taxa such as parrot- flexibility and natural shifts may both account fishes, grazing sea urchins, sharks, turtles, for the zooarchaeologically observed patterns large jacks, and groupers. These taxa are in the Lau marine fauna. In addition, Lauans found on all the study reefs but infrequently have long-standing elaborate systems of ma- on Lakeba and more frequently on Nayau rine tenure and rules regarding the use of and the Aiwas. specific areas of the reef. They consider even minor shifts in marine variability and shift fo- Biological Surveys cus accordingly in ways that manage use and temporarily conserve particular resources. The results of the marine biological surveys This may be achieved by declaring areas of in association with this project build on previ- the reef taboo or ‘‘off limits’’ for some time ous marine biological surveys from the Lau while local populations of marine resources region by Wells (1977), Jennings and Polunin are allowed to recover. Or a village elder may (1995, 1996), Vuki et al. (2000), Dulvy et al. decide that a particular type of fish should not (2004), Kuster et al. (2005, 2006), and Turner be captured over a given time period. et al. (2007). These workers and others have noted that traditionally managed fisheries in Fiji have expanded to accommodate escalat- Ethnography ing demands for fish in emerging market The inclusion of indigenous behaviors and economies ( Jennings and Polunin 1995, knowledge is a unique and critical component 1996), and that indigenous or informal of this project. Interactions between scien- knowledge is extremely useful for marine bi- tists and local communities are undoubtedly ologists and ecologists in Fiji and the Pacific crucial to understanding and the long-term islands in general (Dulvy and Polunin 2004, maintenance and stewardship of marine bio- Kuster et al. 2006, Middlebrook and Wil- logical communities. Inshore fishing expedi- liamson 2006). Long-term Perspective on Biodiversity in Fiji . Jones 639

On each of the study islands the biological tural development) will determine the current surveys revealed that there are patches of composition and biological diversity of ma- healthy, recovering reef associated with rela- rine communities. It appears that biodiversity tively large populations of diverse fish com- does not vary as a function of physical factors munities. However, the overall diversity and alone. Although Lakeba should be the most the numbers of fishes recorded varied de- diverse of the islands surveyed, due to the pending on the island (Table 7). The data large size and the natural physical variation indicate that Aiwa and Nayau have more in the types of reefs, substrates, and land diverse and possibly more abundant marine forms, Aiwa and Nayau appear to have more communities. For example, the Lakeba sur- diversity in modern marine communities (ac- veys found relatively fewer total fishes and cording to catch diversity in Appendix and fewer fish species in larger survey areas that observed in marine surveys [Table 7]). (1,000 m2) than the surveys on Nayau and Lakeba has the highest overall human popu- Aiwa (500 m2). The Aiwa reefs have some lation for the study sites, and it likely has healthy colonies of soft tree corals and finger since the island was settled. However, by soft corals, which are lacking on Lakeba and land area/m2 and reef area Lakeba is cur- in most areas on Nayau. Reef sharks and rently less densely settled than Nayau. A large-bodied fishes are also more abundant detailed comparison of the faunal material around Aiwa than the larger islands. from Lakeba, presented in Best (1984) (the comparison is hampered by the fact that conclusions Best’s collection and identification methods differ from my own in fundamental ways. Using the past as a baseline for comparison For example, Best did not use fine screens to with the present to explore change, the inter- collect faunal samples, shellfish was not col- view and fishing expedition data provide im- lected using the methods he used for verte- portant information about the status of the brates, and he did not identify fishes using coral reefs; methods and technologies of ex- the same elements that I use, rather he relied ploitation; changes in this environment; and only on so-called ‘‘special elements’’ for iden- presence, absence, or shifts in the local avail- tification), and that from Nayau and Aiwa will ability of particular resources. A wide range determine if this pattern of diversity is appar- of species is regularly exploited, that is, a total ent throughout the archaeological data. It is of 112 taxa for all four islands. The biological interesting that the diversity of the zooarch- surveys recorded over 200 species of fishes, aeological samples of both bony fishes and and the archaeological bone identifications shellfish is greater on Nayau than on Aiwa. include over 50 fish taxa. Although there are Assuming that inshore catches represent a many factors that influence the outcome of cross section of the available marine re- the summary data for each line of evidence, sources, this pattern may reflect greater di- these measures do indicate an overlap in the versity on Nayau in prehistory and a shift in available resources past and present, and con- contemporary times to a less-diverse marine tinuity in the way that the marine fauna have environment. Alternatively, the data could be been used throughout Lauan history. In the indicative of a broader exploitation pattern by following section I discuss four key findings the prehistoric occupants of Nayau versus based on the data derived from the fieldwork that on Aiwa. and laboratory analyses. More research is Second, human disturbance may have oc- needed to adequately address the complex curred more extensively and continuously on issues raised, but I begin with the following large islands, and therefore small islands may observations and interpretations. harbor species-rich communities. This find- First, the data from central Lau suggest ing is evidenced on the islands of Aiwa Levu that the relative intensity of human exploita- and Aiwa Lailai and relates to the point tion (intensity is related to population size, already made. Lakeba has a larger human land area and use, and the degree of agricul- population, more advanced technologies and 640 PACIFIC SCIENCE . October 2009 infrastructures (roads, cars, educational and examined in this study, according to the medical compounds, an airport, a large pine zooarchaeological data. (This statement tree farm, generators), and more develop- needs to be evaluated by isotope analysis, ment and runoff from agriculture than the which is forthcoming.) By determining local other islands studied. Nayau, Aiwa Levu, and standards of preference and rank and examin- Aiwa Lailai are much less disturbed in mod- ing social and environmental motivations ern times, but even in prehistory these for exploiting particular food items, we can environments appear to have had less im- better understand how humans make deci- pacts resulting from human occupation. It is sions about using their environment and its likely that these smaller islands had contin- resources. ually smaller population densities and less- Fourth, vulnerable species are currently intensive exploitation over time (note that overexploited or locally extinct; even rare the size of the reef on Nayau is approxi- resources have withstood 3,000 years of hu- mately the same size as the reef associated man impacts and thus may have life history with Aiwa Levu and Aiwa Lailai). The diver- traits supporting resilience, making conserva- sity of fishes and invertebrates species ex- tion efforts worthwhile ( Jackson et al. 2001, ploited on Nayau and Aiwa is greater across Steadman 2006). In all of the archaeological prehistory; the reefs of these islands currently sites, fish bones constitute the majority of have more diverse biological communities the recovered fauna, suggesting a heavy reli- than those of Lakeba. ance on the sea and relatively small inshore Third, humans appear to have a selective reef fish in particular. The bony fish families effect on marine biodiversity as particular that contribute the most to zooarchaeological species are/were targeted according to local assemblages throughout the archaeological standards of ranking and preference. For ex- sequence by count (NISP) and the numbers ample, people sometimes target large taxa, of individuals (MNI) are Acanthuridae but they also target fishes with specific physi- (tangs), Serranidae (groupers), Scaridae (par- cal characteristics such as big eyes and red rotfishes), and Lethrinidae (emperorfishes). lips; this includes, for example, emperorfishes Modern subsistence systems also rely heavily (O’Day 2004). It should be noted that optimi- on these families of fishes, and they form a zation criteria derived from foraging theory majority of the modern fauna, according to do not entirely dictate cultural preference in the biological surveys. Invertebrates make up this setting. In addition to exploiting large- a less-important portion of the diet. The bodied fishes with spears and while trolling most commonly exploited archaeological in- (trolling and fishing outside the reef is a vertebrates include species in the family Tur- male activity; women fish in the inshore binidae (turban snails), which are now rare on area), Lauan women target fishes in the in- Lakeba and Aiwa and becoming increasingly shore area (which are primarily small-bodied) rare on Nayau; Strombidae (strombs); Coni- and are motivated by characteristics that dae (cones); and Patellidae (limpets). My extend beyond the body size of particular study did not note any locally extinct re- animals (Tables 3–5 and Figures 2–9). This sources (species that occurred in the past but finding was apparent in the archaeological were absent in the modern community over- data (small-bodied fishes occur in the highest all), but according to ethnographic and bio- frequencies in all sites across time periods), logical survey data, many large-bodied fishes and the small-bodied fighting conch was in- and invertebrates appear to be in decline after tensively exploited at Nayau’s earliest known three millennia of exploitation (groupers, Lapita occupation site, Na Masimasi. The parrotfishes, sweetlips, snappers, mullet, co- majority of the animal protein consumed conut crab, giant clam, strombids, trumpet each day by Lauans comes from the inshore triton). These species and those that are area; therefore small fishes contribute the known to be locally productive on Lakeba, most animal protein to the diet and have Nayau, and Aiwa would be excellent targets since the earliest occupation of the islands for conservation (see Table 6). For example, Long-term Perspective on Biodiversity in Fiji . Jones 641 in the study area the local diversity of lethri- zooarchaeological assemblages; that is, these nids, despite 3,000 years of intensive exploita- climatic changes resulted in environmentally tion, is remarkable. There are 14 species of induced cultural change. On Viti Levu Field lethrinids documented in all of the Fiji (2004) documented shifts to fortified settle- Islands (World Wildlife Foundation 2003), ments around A.D. 1300–1500 that were and my research documented 10 species accompanied by a broadening diet breadth. from ethnographic observations, six from the The mid-late period archaeological data archaeological sites, and four in the modern from Nayau and Aiwa exhibit minor shifts in marine surveys. the diversity of exploited taxa. A slightly Convict tangs (Acanthurus triostegus) and greater diversity in bony fish exploitation oc- rabbitfishes (Siganus spinous) are well suited curred during the mid-late period occupa- to the types of exploitation practiced in the tions on Nayau, where a more diverse group northern villages of Lakeba and on Nayau of invertebrates was exploited over time. because they are less vulnerable than other However, there is no evidence for intensive species—their population doubling times are declines in marine productivity from the 1.4–4.4 years and less than 15 months, re- Lauan archaeological data. More obvious de- spectively (Froese and Pauly 2006). Appar- clines in invertebrates and large-bodied fishes ently, a component of the Lauan traditional appear in the modern data, rather than in the system for marine management, moving archaeological past. Perhaps the data illus- from various fishing grounds exploited by trate both Lauan stability and flexibility in group fishing practices, allows for long-term marine exploitation patterns, just as D’Arcy sustainability of particular types of resources. (2006:21) suggested for all the inhabitants of These ecological practices and traditional Remote Oceania, who have been subject to ecological knowledge (TEK) may be used to highly variable climatic shifts throughout guide the development of programs for sus- their history. tainable use of marine resources in Fiji and The responsible management and conser- elsewhere in the tropics. In fact, social atti- vation of marine ecosystems is critical for Fi- tudes toward marine resource management jians. In the remote Lauan archipelago, the are critical to the success of management entire population lives on the coast and relies and conservation programs (Middlebrook heavily on marine resources for food and and Williamson 2006, Turner et al. 2007). their livelihoods. Like coastal ecosystems Thomas Lovejoy, who coined the term worldwide, Fiji’s marine biodiversity faces biodiversity, claims that this is the single the growing threat of overfishing and impacts most important measure for evaluating the resulting from pollution, land development, impacts of humans on the environment climatic change, and coral bleaching. Com- (Lovejoy 1997:10). To understand the extent pared with other areas of Fiji, the Lau Group of the human impact on marine ecosystems is less well studied but exhibits great marine in Lau and the complicated relationships sur- diversity as the result of smaller human pop- rounding this issue, long-term research is ulations, less development, partial isolation, needed. It is clear that modern surveys and little or no commercial fisheries, and lack of interviews, as well as ethnographic documen- a tourist infrastructure. My multidisciplinary tation of fishing practices are all useful modes approach contributes to the understanding of determining the local reliance on and im- of Lauan biodiversity through three perspec- portance of marine resources. Each line of tives, archaeological, ethnographic, and bio- evidence contributes varied but overlapping logical. Together this information was used information, which should be compared with to better understand the marine environment the archaeological data. and human interactions with it over the three Allen (2006) predicted that marine pro- millennia of human occupation on the study ductivity declines due to heightened ENSO islands. Ultimately, I expect that an evolu- activity during the LCO-LIA shift around tionary perspective will facilitate the devel- 700 B.P. had an obvious signature on marine opment of programs for sustainable use of 642 PACIFIC SCIENCE . October 2009 marine resources in the study area and be- Literature Cited yond. Methodologically, this study indicates that Allen, M. S. 2006. New ideas about late Ho- each level of data (ethnographic, archaeologi- locene climate variability in the central cal, and marine biological) produces different Pacific. Curr. Anthropol. 47:521–535. but overlapping results. A better understand- Bale´e, W., ed. 1998. Advances in historical ing of environmental problems and solutions ecology. Columbia University Press, New for dealing with them will come from multi- York. disciplinary collaborations and the examina- Bale´e, W., and C. L. Erickson, eds. 2006. tion of biological complexity over the long Time and complexity in historical ecology: term (Lovejoy 1997, Jackson et al. 2001, Studies in the Neotropical lowlands. Co- Briggs et al. 2006). Combining multiple ap- lumbia University Press, New York. proaches and methodologies will enhance Beavan Athfield, N., R. C. Green, J. Craig, B. understanding of the issues related to marine McFadgen, and S. Bickler. 2008. Influence changes on local and global scales (Kronen of marine sources on 14C ages: Isotopic and Bender 2007, Turner et al. 2007, Rick data from Watom Island, Papua New and Earlandson 2008). A multidisciplinary Guinea inhumations and pig teeth in light historical ecological approach holds much of new dietary standards. J. R. Soc. N. Z. promise for the future of research in Fiji and 38:1–23. in marine ecosystems worldwide. Best, S. 1984. Lakeba: The prehistory of a Fijian island. Ph.D. diss., University of Auckland, Auckland, New Zealand. Uni- acknowledgments versity Microfilms, Ann Arbor, 1989. Briggs, J. M., K. A. Spielmann, H. Schaafsma, I gratefully acknowledge the late Na Gone K. W. Kintigh, M. Kruse, K. Morehouse, Turaga Na Tui Lau, Tui Nayau Ka Sau ni and K. Schollmeyer. 2006. Why ecology Vanua ko Lau, The Right Honourable Ratu needs archaeologists and archaeology Sir Kamisese Mara and his family for allow- needs ecologists. Front. Ecol. Environ. ing me to work on Lakeba, Nayau, and 4:180–188. Aiwa. I thank the Fiji Museum and especially Bronk Ramsey, C. 1995. Radiocarbon cali- Sepeti Matararaba for their important assis- bration and analysis of stratigraphy: The tance. I am grateful to the people of Nayau OxCal program. Radiocarbon 37:425–430. and Lakeba for welcoming and facilitating Butler, V. L. 2001. Changing fish use on my research, in addition to providing me Mangaia, southern Cook Islands: Resource with a wealth of local knowledge. Vinaka depression and the prey choice model. Int. vakalevu Tui Liku, Tui Naro, Jack, Sera, J. Osteoarchaeol. 11:88–100. Caka, Meli, Mote, Sina, Semeti Guvaki, Colin, P. L., and C. Arneson. 1995. Tropical Colati, and Rusila. The majority of the radio- Pacific invertebrates: A field guide to the carbon determinations (excluding those from marine invertebrates occurring on tropical Na Masimasi) were funded by National Pacific coral reefs, seagrass beds and man- Science Foundation awards to David Stead- groves. Coral Reef Press, Beverly Hills, man. I also thank Scott Fitzpatrick, Loretta California. Cormier, and three anonymous reviewers Crumley, C. L., ed. 1994. Historical ecology. for constructive comments on this paper; School of American Research, Santa Fe, Heather Walsh-Haney of Florida Gulf Coast New Mexico. University for identifying the human remains D’Arcy, P. 2006. The people of the sea: submitted for radiocarbon dating; Joseph Environment, identity, and history in Oce- Ortega for assistance with the illustrations; ania. University of Hawai‘i Press, Hono- and Lisa Kirkendale and Peter Middlefart lulu. for conducting marine biological surveys in DeVantier, L. M., G. De’ath, T. J. Done, and August 2008. E. Turak. 1998. Ecological assessment of a Long-term Perspective on Biodiversity in Fiji . Jones 643

complex natural system: A case study from tionally managed Fijian fishing grounds. the Great Barrier Reef. Ecol. Appl. 8:480– Fish. Manage. Ecol. 3:335–347. 496. Jones, S. 2007. Human impacts on ancient Dulvy, N. K., and N. V. C. Polunin. 2004. marine environments of Fiji’s Lau Group: Using informal knowledge to infer Current ethnoarchaeological and archaeo- human-induced rarity of a conspicuous logical research. J. Isl. Coastal Archaeol. 2 reef fish. Anim. Conserv. 7:365–374. (2): 239–244. Dulvy, N. K., N. V. C. Polunin, A. C. Mill, ———. 2009a. Food and gender in Fiji: Eth- and N. A. J. Graham. 2004. Size structural noarchaeological explorations. Lexington change in lightly exploited coral reef com- Books/Rowman & Littlefield (in press). munities: Evidence for weak indirect ———. 2009b. Contemporary subsistence effects. Can. J. Fish. Aquat. Sci. 61:466– and foodways in the Lau Islands, Fiji: An 475. ethnoarchaeological study of non-optimal Field, J. S. 2004. Environmental and climatic foraging and irrational economics. In U. considerations: A hypothesis for conflict Albarella, ed. Ethnozooarchaeology: The and the emergence of social complexity in present past of human-animal relation- Fijian prehistory. J. Anthropol. Archaeol. ships. Oxbow Books, Oxford, United 23:79–99. Kingdom (in press). Fitzpatrick, S. M., and T. J. Donaldson. 2007. Jones, S., P. O’Day, and S. Connaughton. Anthropogenic impacts to coral reefs in 2007a. New archaeological insights from Palau, western Micronesia during the the central Lau Group, Fiji. Domodomo Late Holocene. Coral Reefs 26:915–930. Vol. 20. Froese, R., and D. Pauly, eds. 2006. Fishbase. Jones, S., D. W. Steadman, and P. O’Day. www.fishbase.org, version 12/2006. 2007b. Archaeological investigations on Hocart, A. M. 1929. The Lau Islands, Fiji. the small islands of Aiwa Levu and Aiwa Bernice P. Bishop Mus. Bull. 62. Lailai, Lau Group, Fiji. J. Isl. Coastal Ar- Hughes, T. P., A. H. Braid, D. R. Bellwood, chaeol. 2 (1): 72–98. M. Card, S. R. Connolly, C. Folke, R. Kay, E. Allison. 1979. Hawaiian marine Grosberg, O. Hoegh-Guldberg, J. B. C. shells. Reef and shore fauna of Hawaii, Jackson, J. Kleypas, J. M. Lough, P. Mar- Section 4: Mollusca. Bernice P. Bishop shall, M. Nystrom, S. R. Palumbi, J. M. Mus. Spec. Publ. 64 (4). Pandolfi, B. Rosen, and J. Roughgarden. Kirch, P. V., and D. Yen. 1982. Tikopia: 2003. Climate change, human impacts, Prehistory and ecology of a Polynesian and the resilience of coral reefs. Science Outlier. Bernice P. Bishop Museum, Ho- (Washington, D.C.) 301:929–933. nolulu, Hawai‘i. Jackson, J. B. C., M. X. Kirby, W. H. Berger, Krebs, C. J. 1999. Ecological methodology. K. A. Bjorndal, L. W. Botsford, B. J. Bour- 2nd ed. Benjamin Cummings, New York. que, R. H. Bradbury, R. Cooke, J. Erland- Kronen, M., and A. Bender. 2007. Assessing son, J. A. Estes, T. P. Hughes, S. Kidwell, marine resource exploitation in Lofanga, C. B. Lange, H. S. Lenihan, J. M. Pan- Tonga: One case study—two approaches. dolfi, C. H. Peterson, R. S. Steneck, M. J. Hum. Ecol. 35:195–207. Tegner, and R. R. Warner. 2001. Histori- Kuster, C., V. C. Vuki, and L. P. Zann. 2005. cal overfishing and the recent collapse of Long-term trends in subsistence fishing coastal ecosystems. Science (Washington, patterns and coral reef fisheries yield from D.C.) 293:629–637. a remote Fijian island. Fish. Res. 76:221– Jennings, S., and N. V. C. Polunin. 1995. 228. Comparative size and composition of yield ———. 2006. Validation of the accuracy of from six Fijian reef fisheries. J. Fish Biol. household reporting of subsistence fishing 46:28–46. catch and effort: A Fijian case study. Fish. ———. 1996. Fishing strategies, fishery de- Manage. Ecol. 13:177–184. velopment and socioeconomics in tradi- Leach, F., and J. Davidson. 2000. Fishing: A 644 PACIFIC SCIENCE . October 2009

neglected aspect of oceanic economy. Southwest Viti Levu Island, Fiji. Archaeol. Pages 412–426 in A. Anderson and T. Oceania 39:139–143. Murray, eds. Australian archaeologist: O’Day, S. 2001. Excavations at the Kipapa Collected papers in honour of Jim Allen. Rockshelter, Kahikinui, Maui, Hawai‘i. Center for Archaeological Research and Asian Perspect. 40 (2): 279–304. Department of Archaeology and Natural ———. 2004. Past and present perspectives History, The Australian National Univer- on secular ritual: Food and the fisher- sity, Coombs Academic Publishing. women of the Lau Islands, Fiji. Pages Lovejoy, T. E. 1997. Biodiversity: What is it? 153–161 in S. J. O’Day, W. Van Neer, Pages 7–14 in M. L. Reaka-Kubla, E. O. and A. Ervynck, eds. Behaviour behind Wilson, and D. Wilson, eds. Biodiversity bones: The zooarchaeology of ritual, reli- II: Understanding and protecting our bio- gion, status, and identity. Oxbow Books, logical resources. National Academies Oxford, United Kingdom. Press, Washington, D.C. O’Day, S., P. O’Day, and D. W. Steadman. Maragos, J. E., and C. W. Cook Jr. 1995. The 2003. Defining the Lau context: Recent 1991–1992 rapid ecological assessment of findings on Nayau, Lau Islands, Fiji. Palau’s coral reefs. Coral Reefs 14:237– N. Z. J. Archaeol. 25:31–56. 252. Pandolfi, J. M., R. H. Bradbury, E. Sala, T. P. Middlebrook, R., and J. E. Williamson. 2006. Hughes, K. A. Bjorndal, R. G. Cooke, D. Social attitudes towards marine resource McArdle, L. McClenachan, M. J. New- management in two Fijian villages. Ecol. man, G. Paredes, R. R. Warner, and J. B. Manage. Restor. 7 (2): 144–147. Jackson. 2003. Global trajectories of the Morrison, A. E., and E. E. Cochrane. 2008. long-term decline of coral reef ecosystems. Investigating shellfish deposition and land- Science (Washington, D.C.) 301:955–958. scape history at the Natia Beach site, Fiji. Pandolfi, J. M., J. B. C. Jackson, N. Baron, J. Archaeol. Sci. 35:2387–2399. R. H. Bradbury, H. M. Guzman, T. P. Myers, R. F. 1991. Micronesian reef fishes: A Hughes, C. V. Kappel, F. Micheli, J. C. practical guide to the identification of the Ogden, H. P. Possingham, and E. Sala. coral reef fishes of the tropical central 2005. Are U.S. coral reefs on the slippery and western Pacific. 2nd ed. Coral Graph- slope to slime? Science (Washington, ics, Guam. D.C.) 307:1725–1726. Newsom, L. A., and E. S. Wing. 2004. On Reitz, E. J., and E. S. Wing. 1999. Zooarch- land and sea: Native American uses of bio- aeology. Cambridge University Press, logical resources in the West Indies. Uni- Cambridge. versity of Alabama Press, Tuscaloosa. Rick, T. C., and J. M. Erlandson, eds. 2008. Nunn, P. D. 2000. Environmental catastro- Human impacts on ancient marine eco- phe in the Pacific islands around A.D. systems. University of California Press, 1300. Geoarchaeology 15:715–740. Berkeley. ———. 2007. Echoes from the distance: Re- Sale, P. F. 1980. Assemblages of fish on patch search into the Lapita occupation of the reefs: Predictable or unpredictable? Envi- Rove Peninsula, Southwest Viti Levu, Fiji. ron. Biol. Fishes 5:243–249. Pages 163–174 in S. Bedford, C. Sand, Spennemann, D. H. R. 1987. Availability of and S. P. Connaughton, eds. Oceanic ex- shellfish resources on prehistoric Tonga- plorations: Lapita and western Pacific set- tapu, Tonga: Effects of human predation tlement. Terra Australis 26, Australian and changing environment. Archaeol. National University E. Press. Oceania 22:81–96. Nunn, P. D., R. Kumar, S. Matararaba, T. Steadman, D. W. 2006. Extinction and bioge- Ishimura, J. Seeto, S. Rayawa, S. Kuru- ography of tropical Pacific birds. Univer- yawa, A. Nasila, B. Oloni, A. R. Ram, P. sity of Chicago Press, Chicago. Saunivalu, P. Singh, and E. Tegu. 2004. Stuiver, M., P. J. Reimer, E. Bard, J. W. Early Lapita settlement site at Bourewa, Beck, G. S. Burr, K. A. Hughen, B. Long-term Perspective on Biodiversity in Fiji . Jones 645

Kromer, G. McCormac, J. van der Plicht, Turner, R. A., A. Cakacaka, N. A. J. Graham, and M. Spurk. 1998. INTCAL 98 Radio- N. V. C. Polunin, M. S. Pratchett, S. M. carbon age calibration, 24,000–0 cal BP. Stead, and S. K. Wilson. 2007. Declining Radiocarbon 40:1041–1083. reliance on marine resources in remote Thomas, F. R. 2002. An evaluation of South Pacific societies: Ecological versus central-place foraging among mollusk socio-economic drivers. Coral Reefs gatherers in western Kiribati, Micronesia: 26:997–1008. Linking behavioral ecology with ethno- Vuki, V., L. P. Zann, M. Naqasima, and M. archaeology. World Archaeol. 34 (1): Vuki. 2000. The Fiji Islands. Pages 751– 182–208. 764 in C. R. C. Sheppard, ed. Seas at the ———. 2007a. The behavioral ecology of millennium: An environmental evaluation. shellfish gathering in western Kiribati, Vol. II. Pergamon, Oxford, United King- Micronesia. 1: Prey choice. Hum. Ecol. dom. 35:179–194. Wells, J. 1977. A preliminary report on the ———. 2007b. The behavioral ecology of littoral copepods of Moce Island, Lau shellfish gathering in western Kiribati, Mi- Group, Fiji. R. Soc. N. Z. Bull. 17:91–92. cronesia. 2: Patch choice, patch sampling, World Wildlife Foundation, Fiji Pro- and risk. Hum. Ecol. 35:515–526. gramme, November 2003. Fiji Islands Thomas, F. R., P. D. Nunn, T. Osborne, R. marine ecoregion: An overview of out- Kumar, F. Areki, S. Matararaba, D. Stead- standing biodiversity, threats, opportunities man, and G. Hope. 2004. Recent archae- and key stakeholders for conservation. ological findings at Qaranilaca Cave, Prepared by V. Nair. Online at http:// Vanuabalavu Island, Fiji. Archaeol. Oce- conserveonline.org/workspaces/era.peer. ania 39:42–49. review/workshops/9thbrisbane/workplans/ Thompson, L. M. 1940. Southern Lau, Fiji: FIME%20Reconnaissance.pdf An ethnography. Bernice P. Bishop Mus. Bull. 162.

Appendix Fish Species Collected in Modern Fishing Expeditions on the Islands of Aiwa Levu, Aiwa Lailai, Nayau, and Lakeba

Aiwa Aiwa Taxa: Scientific Name Common Name Levu Lailai Lakeba Nayau

Carcharhinidae Requiem sharks Carcharhinus spp. Reef sharks La LL Galeocerdo cuvier Tiger shark L L Dasyatidae Stingrays Dasyatis kuhlii Bluespot stingray G Muraenidae Moray eels Gymnomuraena zebra Zebra moray G* Gymnothorax favagineus Blackblotched moray S, G G. meleagris Guineafowl moray G* G. zonipectus Bartail moray S, G Enchelynassa canina Longfang moray S, G Enchelycore schismatorhynchus Whitemargined moray G* Echidna nebulosa Snowflake moray G S, G Ophichthidae Snake eels Myrichthys colubrinus Harlequin snake eel G* G* Elopidae Ladyfishes Elops sp. Ladyfish G Plotosidae Eel catfishes Euristhmus lepturus Longtailed catfish G 646 PACIFIC SCIENCE . October 2009

Appendix (continued)

Aiwa Aiwa Taxa: Scientific Name Common Name Levu Lailai Lakeba Nayau

Clupeidae Herrings Herklotsichthys quadrimaculatus Bluestripe herring G Batrachoididae Toadfishes Chelonodon patoca Mikspotted toadfish S, G Exocoetidae Flyingfishes Cypselurus sp. Flyingfish N, T Belonidae Needlefishes Platybelone argalus platyura Keeled needlefish G G Strongylura incisa Reef needlefish G Tylosurus crocodilus crocodilus Hound needlefish G Hemiramphidae Halfbeaks Hyporhamphus acutus acutus Halfbeak N Holocentridae Squirrelfishes, soldierfishes Myripristis sp. Soldierfish S S, G S, G S, G Sargocentron spiniferum Longjawed squirrelfish S G S S, G S. violaceum Violet squirrelfish S, G Platycephalidae Flatheads Onigocia spinosa Spiny flathead S, G Scorpaenidae Scorpionfishes G* Serranidae Groupers Cephalopholis argus Peacock grouper (rockcod) S S G, T C. sonnerati Tomato hind G, T Epinephelus merra Honeycomb grouper S, G, H S, G S, G G E. lanceolatus Giant grouper S, T E. howlandi Blacksaddle grouper G E. polyphekadion Marbled grouper S, G G, T E. tauvina Reef cod S Therapontidae Grunters Terapon jarbua Crescentbanded grunter G G Carangidae Jacks, trevallys Alectis ciliaris Pennantfish S, G, T Caranx ignobilis Giant trevally S, G, T C. melampygus Bluefin trevally H, T S S, G S, G, T Gerreidae Mojarras Gerres oyena Common silver-biddy G Lutjanidae Snappers Lutjanus spp. Snappers S, H, T Lutjanus gibbus Paddletail S, H, T Haemulidae Sweetlips, grunts Plectorhinchus spp. Sweetlips S, H, T S, H, T Nemipteridae Threadfin bream Scolopsis bilineata Twolined monocle bream S G Coryphaenidae Dolphinfishes Coryphaena hippurus Common dolphinfish T T Lethrinidae Emperorfishes Gnathodentex aureolineatus Yellowspot emperor G G Gymnocranius grandoculis Bluelined largeeye bream S S, G Lethrinus conchyliatus Redaxil emperor G L. erythropterus Orangefin emperor S, G G L. harak Blackspot emperor G G L. miniatus Trumpet emperor G L. obsoletus Yellowstripe emperor G L. xanthochilus Yellowlip emperor G G L. atkinsoni Yellowbrown emperor G Monotaxis grandoculis Bigeye bream S, G G G Long-term Perspective on Biodiversity in Fiji . Jones 647

Appendix (continued)

Aiwa Aiwa Taxa: Scientific Name Common Name Levu Lailai Lakeba Nayau

Mullidae Goatfishes Mulloidichthys flavolineatus Yellowstripe goatfish G G Mulloides vanicolensis Yellowfin goatfish G G Parupeneus barberinoides Half and half goatfish G G G P. cyclostomus Yellowsaddle or Yellow goatfish G P. indicus Indian goatfish G P. macronemus Longbarbel goatfish G P. multifasciatus Multibarred goatfish S G P. trifasciatus Goatfish S P. pleurostigma Sidespot goatfish G G Kyphosidae Sea chubs Kyphosus bigibbus Gray sea chub S, G Ephippidae Batfishes, spadefishes Platax teira Longfin spadefish G Chaetodontidae Butterflyfishes S G G G Chaetodon vagabundus Vagabond butterflyfish G G Pomacanthidae Angelfishes Pygoplites diacanthus Royal angelfish S G Pomacentridae Damselfishes Abudefduf septemfasciatus Banded sergeant S S Labridae Wrasses Anampses spp. Wrasses G Bodianus vulpinus Blackspot pigfish S, G Cheilinus undulatus Doubleheaded Maori wrasse S, G Hologymnosus doliatus Pastel ringwrasse G Labrichthys unilineatus Tubelip wrasse G Novaculichthys taeniurus Carpet wrasse S, G Thalassoma spp. Wrasses G G Iniistius pavo Peacock wrasse S, G Scaridae Parrotfishes Calotomus carolinus Carolines parrotfish G Chlorurus microrhinos Steephead parrotfish S, G C. sordidus Bullethead parrotfish S S, G Scarus globiceps Globehead parrotfish S S. rubroviolaceus Redlip parrotfish S G S. prasiognathos Dusky parrotfish S, G S. schlegeli Yellowband parrotfish G S. oviceps Blue parrotfish S, G G S, G Mugilidae Mullets Mugil cephalus Flathead mullet G G Valamugil buchanani Bluetail mullet G G Liza vaigiensis Squaretail mullet G Sphyraenidae Barracudas Sphyraena barracuda Great barracuda G S, H, T Sphyraena obtusata Striped seapike S, H, T Blenniidae Blennies Salaris sinuosus Fringelip blenny G Acanthuridae Surgeonfishes Acanthurus lineatus Blueband surgeonfish S G A. guttatus Whitespotted surgeonfish G G A. nigricauda Epaulette surgeonfish S A. triostegus triostegus Convict tang G, S G S, G G Ctenochaetus striatus Striped bristletooth S G S S, G Naso lituratus Orangespine unicornfish G, S S, G N. unicornis Bluespine unicornfish S G, S S, G 648 PACIFIC SCIENCE . October 2009

Appendix (continued)

Aiwa Aiwa Taxa: Scientific Name Common Name Levu Lailai Lakeba Nayau

Zanclidae Moorish idol Zanclus cornutus Moorish idol G G Siganidae Rabbitfishes Siganus argenteus Forktail rabbitfish G G S. spinus Little spinefoot rabbitfish S G G Scombridae Tunas, mackerels, bonito Scomberomorus spp. Mackerel T Bothidae and Soleidae Flounder and sole S, G Balistidae Triggerfishes S, G, H Balistoides viridescens Bluefinned triggerfish S Pseudobalistes fuscus Yellowspot triggerfish G Melichthys niger Ebony triggerfish S, G Ostraciidae Trunkfishes Ostracion cubicus Yellow boxfish G* G* Lactoria diaphana Roundbelly cowfish G* G* Tetraodontidae Puffers Arothron reticularis Reticulated pufferfish G Arothron stellatus Starry pufferfish G S, G Triodon macropterus Threetoothed puffer S, G Diodontidae Porcupinefishes Chilomycterus reticulatus Porcupinefish S, G Diodon hystrix Porcupinefish G S, G Istiophoridae Billfish, marlin T

Note: Only seven species are bycatch, which are not consumed; these are marked with an asterisk (*). a Modes of collection: L, longline; G, gill net; S, spear; N, hand net; T, trolling; H, hand line.