Extinctions of Polynesian Birds

IO Jack GoIson

Haberle et al. (1991:30, 38) and Brookfield (1991:207) contrast the situation in the I ri an Jaya section ofthe Highlands, where novolcanic ash is present and the land is more susceptible to degradation and erosion. Brookfield (1991 :207) describes the intensively managed and densely peopled wetlands of the Baliem Valley and Paniai lakes as being separated by a wide expanse of sparsely populated degraded slopeland from agrarian settlements in tributary valJeys and at the forest edge.

CONCLUSION The massive transfomlation that the practice of shifting cultivation wrought in the ecology of Highlands New Guinea-from which we have derived some basic features of Highlands societies-as well as in agricultural practice, pig 4 husbandry, and systems of exchange, took place over many millennia. The en Extinctions of Polynesian Birds: try of the sweet potato was a fairly recent event, and it brought about changes in settlement and pig husbandry whose ramifications were still being worked Reciprocal Impacts of Birds and People out at the time ofEuropean contact. Today, in the Papua New Guinea High lands the cash economy and rising populations are making new demands on David W. Steadman natural and social systems and are requiring adjustments from them of a scale and urgency unparalleled in the past.

The loss ofbird life is now well recognized as one of the major environmental consequences of the human colonization of Oceania. The decline of birds is related mainly to predation and to landscape changes wrought by prehistoric peoples to accommodate agriculture. For birds, perhaps the most influential landscape change has been the elimination and alteration of indigenous for ests through cutting, buming, and the introduetion of nonnative plants (Kirch 1983). These activities have eliminated natural habitats and have rendered tlle surviving birds even more vulnerable to predation from humans and non native mammals (rats, dogs, pigs). The erosion of topsoil associated with de forestation has removed large areas of nesting habitat for burrowing seabirds, such as shearwaters and petrels. Ethnographic information gathered over the past two centuries is crucial for interpreting Polynesian uses of birds in prehistoric times. I shall concen trate as much as possibIe on the prehistoric and early historie relationships be tween birds and Polynesians, knowing that events of the post-European period have only exacerbated what already was a dismal situation from the birds' standpoint.

II ,2 David W. Steadman Extlnctions of Polynesian Birds SJ

The geographic area eonsidered here is all of Polynesia except the Hawai HUMAN IMPACTS ON BIRDS ian Islands and New Zealand. Although Hawaii (Olson and James 1982a, Background Extinction Hl82b, 1991; James and Olson 1991; James et al. 1987) and New Zealand (An derson 1984, 1989a; Cassells Hl84; Trotter and McCulloch 1984; Holdaway The Holocene extinction of vertebrates ou Oceanic islands has been cali 1989) both experienced major prehistoric losses of birds, the biotas of these brated by the fossil record of the Galapagos Islands (Steadman et al. 1991). two isolated archipelagoes are not closely related to those elsewhere in Poly Dnlike any other group of islands in the tropical Pacific, the Galapagos never nesia. supported human populations before tlleir discovery by Europeans in 1535 The prehistoric record ofbirds is much better known in Polynesia than in (Steadman 1986). As a result, human impact in the Galapagos is confined to Micronesia or Melanesia. Bones ofbirds from archaeological or paleontologi the past five centuries, and it was relatively minor until about 1800. The Holo cai contexts are now known from at least 20 Polynesian islands besides Hawaii cene fossil record of the Galapagos comprises about 500,000 bones, more and New Zealand. In Micronesia, studies have been completed thus far on than 90 percent ofwhich predate the arrival ofhumans. These paleofauna~ re olll)' hvo islands. From Rota in the Mariana Islands, several small cave depos veal the loss of only zero to three vertebrate populations in the 4,000 to 8,000 its have yielded bones of 13 species of extinet or extirpated birds, including years before human arrival, compared to the loss of21 to 24 populations in the sheanvaters, tems, ducks, megapodes, rails, pigeons, parrots, swifts, and pas past 150 to 300 years. Thus the rate ofbackground (prehuman) extinction in serines (Steadman 1992a). Assemblages of archaeological bird bones exca the Galapagos was at least a hundred times less than the rate of human-re vated on Fais (Yap) represent species ofindigenous resident birds, 12 ofwhich lated extinction. When undisturbed by humans, the natural processes of dis probably or certainly no longer live on Fais (Steadman and Intoh 1994). In persal, colonization, and evolution may resuIt in a very low rate of extinction spite of the small amount of data currently available from Micronesia, I see no for reptiles, birds, and mammals on tropical Oceanic islands. cultural, geographical, or biological reason why the extent of human-caused The Polynesian Record a\lan extinctions in this region of Oceania, when more fully studied, will differ in ,Uly major way from that ofPolynesia. Mo\ing west from the Galapagos to Oceania, we find that \irtllally all islands Thc islands of Melanesia tend to be larger, and to support richer floras and were inhabited at one time or another in prehistory and that a significant faunas today, than the islands ofPolynesia or Micronesia. The prehistoric re amount of prehistoric extinction has taken place. The prehistorie record ol' cord of Melanesian birds has been studied only on Fiji (Steadman 1989b), Polynesian birds is based on data collected from arehaeological and paleon New Caledonia (Balouet and Olson 1989), New Ireland (Steadman, White, tological sites in the Marquesas Islands (Steadman 1988, 1989a, 1991a, and Allen n.d.), and Mussau (Steadman n.d.). As in Polynesia and Micronesia, 1992b; Steadman et al. 1988; Steadman and Zarriello 1987; Dye and Stead the Iimited record in Melanesia shows losses of a variety ofboth seabirds and man 1990), the Society Islands (Steadman 1989a, 1992b; Dye and Steadman landbirds, tllC latter dominated by hawks, megapodes, rails, pigeons, parrots, 1990; Steadman and PallIavan 1992), tlle Cook Islands (Steadman 1985, 1987, and owls but also induding herons, ibises, buttonquails, snipe, owlet-night 1989a, 1991b, 1992b; Allen and Schubel 1990; Allen and Steadman 1990, jars, and crows. Preliminary data from the very large, high island of New Ire Steadman and !Grch 1990; !Greh et al. 1991; !Grch et al. 1992), Henderson Is land indicate that evcn its relatively rich avifauna (currently about 108 species land (Steadman and Olson 1985; Schubel and Steadman 1990), Easter Island ol' resident landbirds) has lost about 25% of its species. A greater percentage (Steadman et al. 1994), Samoa (Steadman 1993b), Tonga (Steadman 1989a, ofbirds has survived on tlle larger islands of Melanesia than on typical Polyne 1989b, 1993a; Dye and Steadman 1990), and the Polynesian Outliers in Mela siml or Mierouesian ishUlds, or on small Melanesian islands. This seems to be nesia (Balouet and Olson 1987; Steadman et al. 1988; Steadman, Pahlavan, due to tlw buffering effects that large island size, steep terrain, and diseases and !Grch 1990). I shall review this record briefly, using specific examples (such as malaria) have had on hUmaIl impact. Common sen se dictates that in from several island groups. sular floras and faunas are easiest to deplete on low, small islands. In addition, The losses of birds in Polynesia fall into three categories: (l) extinetion many parts of island Melanesia have been occupied for tens of millennia (loss of all populations of a species); (2) extirpation (los s of a species Oll an indi longer than anywhere in Polynesia or Micronesia (Jones 1989; see Chapter 2, vidual island, although one or more populations of the species survive else this volume). where); and (3) reduced population (loss without replacement of indi\iduals 1-4 David W. Steadman Extlnctions of PoJynesian Blrds ss

from a surviving population on an island). A single locality (Mangaia, the Cook Table 4.1 Resident Birds from Ua Huka, Marquesas Islancl~ Islands) ean exemplif)' eaeh eatego!)'. The Conquered Lorikeet (Vini vidivici) is extinct because it has been extenninated on Mangaia as well as on eve!)' Bones from Exists today other island where it ever oceurred. The Society Islands Pigeon (Ducula HaneSite on UaHuka aurorae) is extirpated on Mangaia (and several other islands) but survives on SEABIRDS Makatea (Tuanlotus) and Tal1iti. Audubon's Shearwater (Puffmus lher Wedge-tailed Shearwater (Puffinus pacificus) x x minien), a pantropical seabird, has a reduced population on Mangaia today, Christmas Island Shearwater (Puffinus nativitatis) x sllIviving in numbers of less than 100, whereas arehaeological and ethno Audubon's Shearwater (Puffinus lhenninieri) x graphic evidence suggests that once it was com mon and widespread on the is Bulwer's Petre! (Bulweria ef. bulwerii) x x land. Tahiti Petre! (Pterodroma rostrata ) x Phoenix Petre! (Pterodroma cf. alba) The loss of seabirds in Polynesia has been particularly severe for petrels x Unknown petre! (Pterodroma smal! sp.) and shearwaters, although the ranges and numbers of many other kinds of x Polynesian Storm-Petre! (Nesofregetta fuliginosa) tropical seabird (various albatrosses, stonn-petrels, tropicbirds, frigatebirds, x x White-bellied Storm-Petre! (Fregetta gralloria) x boobies, tems, guIls) have been reduced as well. The losses oflandbirds in White-tailed Tropicbird (Phaethon lepturus) x x Polynesia have been greatest for rails, pigeons, doves, and parrots, although, Red-footed Booby (Sula sula) x x as with the seabirds, no fanlily oflandbirds has been spared. In Western Poly Brown Booby (Sula leucogaster) x x nesia, unlike Eastem Polynesia, the losses of landbirds also include herons, Masked Booby (Sula dactylatra) x

megapodes, hawks, shrikebills, whistlers, monarch flyeatehers, thrushes, '> l' Abbott's Booby (Papasula abbottii costelloi) x white-eyes, and honeyeaters, some of which represent taxa previously occur Great Frigatebird (Fregata minor) x x ring only in Melanesia, not in Polynesia. By eliminating the "Melanesian" taxa, Lesser Frigatebird (Fregata ariel) x x the anthropogenie extinetion of birds in Tonga, for example, has artificially Gray-baeked Tem (Sterna lunata) x sharpened the biogeographie distinetion between the avifaunas of Polynesia Sooty Tem (Sterna fuscata) x x and Melanesia (Steadman 1993a). Brown N addy (Anous stolidus) x x B!aek N addy (Anous minutus ) Quite deservedly, the "biodiversity erisis" has received much popular and x x BIue-gray Noddy (Procelsterna cerulea) seholarly attention in the past deeade. An understanding of the eurrent biodi x Little Fairy Tem (Gygis microrhyncha) x x versity erisis requires proper historic perspective, which in tum tells us that few plaees on earth have escaped environmental change at the hands of pre LANDBIRDS historie humans. From the standpoint af numbers of extinet species, the most Pacific Reef-Heron (Egretta sacra) x x dramatic sto!)' is that of flightless rails. Each island in the tropical Pacifie with Tuamotu Sandpiper (Prosobonia ef. cancellata) x "Undescribed erake (Porzana new sp.) a thorough prehistorie record of birds has yielded the bones of one to three x "Undeseribed rail (Galliralus new sp.) x unique (endemie) species of flightless or nearly llightless rails. At least five Marquesas Ground-Dove (Gallicolumba rubescens ) x genera are involved, POr".d1/la, Gallirallus, Nesoclopeus, Gallinula, and Por "Giant Ground-Dove (Gallicolumba nui) x phyno, with Gallirallus providing the most species discovered thus far and Red-moustached Fruit- Dove (Ptilinopus mercierii ) x ane of the world's most exeiting, albeit least understood, examples of adaptive White-capped Fruit- Dove (Ptiltnopus dupetithouarsii) x x evolutionary radiation. About 800 islands in Oceania are inhabitable by peo N uku Hiva Pigeon (Ducula galeata) x pie and therefore, presumably, by llightless rails. Thus rails alone might ac "Marquesas Cuekoo-Dove (Macropygia heana) x count for as many as 2,000 species ofbirds that would exist today had people Marquesas Lorilæet (Vini ultramarina ) x riot eolonized Oceania. "Conquered Lorilæet (Vini vidivici) x The first of several examples of depleted Polynesian avifaunas is from Ua "Sinoto's Lorikeet (Vini sinotoi) x Huka in ilie Marquesas Islands (Table 4.1), where the Hane site has yielded Marquesas Swiftlet (Collocalia ocista) x Marquesas Kingfisher (Halcyon godeffroyi) x 56 David W. Steadman Extinctions of Polynesian Birds 57 Table .J.1 (continued) Table 4.2 Resident Birds from Huahine, Society Islands Bones from Exists today Bones Hane Site on UaHuka Exists from today on • lllldesclibed monarch (ef. M yiagra new sp.) x Fa'ahiaSite Huahine Iphis Monareh (Pomarea iphis) x SEABIRDS Marquesas Reed-\Varbler (A.crocephalus mendnTwe) x Wedge-tailed Shearwater (Puffinus pacificus ) x TOTAL.'i Christmas Island Shearwater (Puffinus nativitatis) x Total sp<'cies of seabirds 20 14 Audubon's Shearwater (Puffinus lhenninieri) x Combined total species of seabirds 22 Tahiti Petrel (Pterodrorna rostmta) x Total species oflandbirds 15 5 Phoenix Petrel (Pterodroma alba) x Combined total species ofLandbirds 18 HeraId Petrel (Pterodroma anninjoniana) x White-tailed Tropicbird (Phaethonlepturus) SOURCE: Modified from Steadman 1991a x x Red-footed Booby (Sula sula) • = emnet species x Brown Booby (Sula leucogaster) x Great Frigatebird (Fregata minDt') x Lesser Frigatebird (Fregata ariel) x about 11,000 bird bones, the largest avian assemblage yet obtained from °Undeseribed gull (Lams new sp.) x Brown N oddy (Anous stolidus) tropical Polynesia. Most of these bones represent seven species of shear x x Black Noddy (Anous minutus) waters and petrels, easily obtained from their nesting burrows. While six of x Common Fairy Tem (Gygis candula) the 20 species of seabirds from the Hane site nolongeroceuron Da Huka, this x x number alone does not represent the loss of seabirds. Most of the 14 species UNDBIRDS of seabirds listed in Table 4.1 as "exists today on Da Huka" nest today not on Pacifie Reef-Heron (Egretta sacra) x x the main island but only on tiny offshore islets. Even the surviving species of Mangrove Heron (Ardeola striata) x seabirds exist on D a Huka or its islets in mueh redueed numbers. Gray Duck (Anas superciliosa) x Pigeons and doves (family Columbidae) are the most eommon landbirds at Sooty Crake (Por,.,aTIG tabuensis) x the Hane site, both in tenns of number of species and number ofbones. Only °Undeseribed rail (GaUirallus new sp.) x one of the six species of eolumbids in the Hane deposits survives today on D a Society Islands Ground-Dove (Gallicolumha erythroptera) x H uka. Rails and parrots are the next most common, eaeh of these families hav °Giant Ground-Dove (GaUicolumha nui) x Society Islands Fruit-Dove (Ptilinopus purpuratus) ing deelined on D a Huka from three to zero species. Although the exact ehro x x Nuku Hiva Pigeon (Ducula galeata) nology of the Hane site is in dispute (Sinoto 1979; Kireh 1986), the pattem of x Society Islands Pigeon (Ducula aurorae ) av1an e:..ploitation is dear (Dye and Steadman 1990:212): "The relative eontri x °Huahine Cuekoo-Dove (Macropygia arevarevauupa) x bution ofbirds to the diet [more than half of all animal protein in the earlype Society Islands Lorikeet (Vini peru viana ) e riod) is more than halved within the first 550 years and deelines to °Conquered Lorikeet (Vini vidivici) x insignifieanee in just over a thousand years." ·Sinoto's Lorikeet (Vini sinotoi) x The seeond example of a prehistorie avifauna is from the Fa' ahia site on Tahiti SwiftIet (Collocalia leucophaea) e Huahine in the Society Islands (Table 4.2). This assemblage is extremely rieh Chattering Kingfisher (Halcyon ef. tuta) x x in species, considering that it is based on only 336 identified bird bones. All Society Islands Reed-Warbler (Acrocephalus calfer) x e Huahine Starling (Aplonis diluvialis ) but three of the 14 species of seabirds reeorded from Fa' ahia oceur at the x e Hane site as well. The species-level similarity among landbirds is mueh less striking, with only three species shared. Again, rails, pigeons, doves, and par- 58 David W. Studman

Table4.2 (continued) Bones Exists from today on Fa' ahia Site Huahine I I

TOTALS 4 Total species of seabirds 14 IC') Combined total species of seabirds 15 Total species oflandbirds 15 4 Combined total species oflandbirds 18 So\;RCE: Modified from Steadman and Pahlavan 1992 I I • ~ extinct species e ~ re<:orded in nineteenth century but nowextirpated 1.-<

rots predominate in the prehistoric avifauna. Except for bone assemblages re I I cently collected with screens of -is inch mesh on Aitutaki and Mangaia (Allen and Steadl1lan 1990;Steadman and Kirch 1990), all ofthe Eastern Polynesian C') t faunal assemblages are biased toward larger species ofbird. (Screens of ~ inch lI).-<~1fi I I I I t--..-! o; li) .-< mesh remove most, but not all, of this bias.) Tims tiny species, such as swifts ci and many passerines, typically are absent or underrepresented. To the west, in Tonga, we now know through sampling with screens of -is inch that passer .-<1 ines underwent losses comparable to those of larger landbirds (Steadman 1993a). The prehistoric exploitation ofbirds can be exarnined from a more detailed chronostratigraphic perspective at Tangatatau Rockshelter (site MAN-44) on u I I Mangaia in the Cook Islands (Table 4.3). This stratified site ranges from about

900 (zone J) to 200 or 300 years old (zone A). During this period, the human 00 II) co cl cO 1.-< li) co "'lI' consumption of vertebrates underwent dramatic change. Domesticates (the co .-< pig, Sus serafa, and chicken, Gallus gallus ) are rare in lower strata but in crease as native landbirds decrease. Both pig and chicken decline, however, in .-<1 2;11.-<1 the uppernlost strata, perhaps because of overconsumption by a relatively large human population. The number ofbones from native birds-that is, all species except chickens-is high in zones I and J, low in ilie middle zones, and high again in zones B and C. The species composition, however, is different between the early and late zones. Extinct or extirpated species oflandbirds dominate zones G to J, while seabirds, especially the Black-winged Petrel (PteradroTTUJ nigripen nis), account for the late increase in exploitation ofbirds. This differs from sites elsewhere in Polynesia, such as ilie Marquesas, the Society Islands,

59 Extinctløns øf Pølynesian Bird. 61 IQ David W. Steadman

Tonga. and Tikopia, where bones of seabirds are more abundant than those of hunt birds at a young age and often continue to hunt nearly throughout their landbirds in the oldest culturallevels of early sites (Steadman 1989a, 1989b; lives. I have leamed from personal experience tllat certain boys and young men Dyc and Steadman 1990; Steadman et al. 1990). A factor on Mangaia may be ean elimb cliffs with amazing skill, speed, and fearlessness (to grab birds, espe its precipitous, creviced limestone cliffs, which provided an extensive albeit eially tropiebirds and tems), ean tllfOW rocks at birds witll deadly speed and ac narrow band of relatively rat -free nesting habitat for seabirds. euraey, and mayvirtually run up a large tree to grab anesting bird. Based upon number of bones, predation on landbirds at MAN-44 was most Perhaps the most eomIllon way that prehistorie Polynesians eaught birds intense in zones I and Jand tape red off to practically nil by the end of rone G was simply by hand. The hand method ofgathering birds, known as tangotango (Table 4.4). Based on relative abundance ofbones (percentage of all verte on Tikopia and tango manu on Pukapuka (Beaglehole and Beaglehole brates), predation on landbirds was by far most intense in rone J, after which 1938:74), is rather straightforward: you simply "see a bird dozing in a tree, most species survived into rones I, H, or G, but only in reduced numbers. No elimb up behind it, grab it, and break its neek" (Feinberg 1981:34). Stieks are extinct or e:\.iirpated landbirds are recorded above rone G except the Cook Is used to strike birds in trees, along shorelines, and on cliffs. Seabirds are par lands Fmit-Dove (Ptilinopus rarotonge1l.sis) , which still survives on Raro tieularly vunerable. Some species of tropicbirds, boobies, and tems, for exam tonga and Atiu and is known historically from Ma' uke and Aitutaki. pIe, nest on the ground, while most species of shearwaters and petrels nest in Conforming to the general Eastem Polynesian pattem, the eight species of burrows or ereviees. Any ofthese seabirds could (and still ean) be plueked from extinet or extirpated landbirds from MAN-44 consist of rails, pigeons, doves, their resting piaces just by reaehing down or in and grabbing them. One sueh and parrots. Both of the extinct rails were /lightless. Their obligatory existence bird on Mangaia was the titi, probably a yet to be determined species of shear on the ground for foraging and nesting must have facilitated predation by hu water (Puffinus sp.) that no longer seems to exist on the island. In describing mans and rats. Pigeons, doves, and parrots are forest species favored as food the eating habits of two late prehistoric fugitives living in the Mangaian forest, by Polynesians. Parrots (and sometimes pigeons and doves) also were used for Gill (1894:26) stated: "The bird most easily eaught by [the fugitives] Uriitepi their brightly colored feathers, as I discuss below. tokura and Temoaakaui was the titi (so called from its ery). In the montllOfDe The five surviving species of landbirds recorded from MAN-44 are those cemberit leaves its burrowings in the red mountain soil, and eO/nes to the rocks that ean withstand substantial forest clearance. Because they prefer marshes near tllC sea to fatten its young on small fish. By day it hides in holes, and sleeps. or dense growtllS of fems, grasses, and sedges, the Gray Duck (Anas super The hunter has only to eall at the entrance to the dark eave, in a plaintive voiee, ciliosa) and the Sooty Crake (Porwna tabuensis) have probably benefited Etitie, when the foolish bird, imaginingit to be the voice ofits mate, cO/n es out from anthropogenie landscape changes that promote irrigated taro cultiva of its secure hiding-plaee, and, dazzled by the unwelcome light, allows itself to tion. The Lesser Golden-Plover (Pluvialis dominica) is a rnigratory shorebird be eaught by hand." In a Mangaian dirge, a defenseless titi is eompared to an that nests only on the tundra ofhigh northem latitudes. In Polynesia it occurs unarmed man faeing a man willi a war club (Gill1894:307). in open habitats like reefs and sand /lats, grassy fields, and airstrips. The Man Another example of predation on a defenseless seabird is Gill's description gaia Kingfisher (Halcy(m mangaia) seems to tolerate moderate forest clear (1885:108-9) ofthe Herald Petrel (Pterodronw [armiryonialw) heraldica): ance but requires cavities in trees for nesting. Its current scarcity on Mangaia A fishing-hawk, about the size of a pigeon, witll black eyes and dark plumage, ex Illay be related to nest-site competition with the Common Myna (Acri cepting undemeath, where it is white, was fonnerlyplentiful at Rarotonga (where it (Uilhcres tristis) , an abllndant and aggressive Asian species introduced to still survives; McConnack and Kunzle 1990b). Its home is in the crannies of a1most Mangaia earlier this century. The Mangaia Reed-Warbler (Acrocephalus inaccessibie rocks. The koputu (such is its name) lays but two eggs-exactly Iike kerearako) occurs in forests, thickets, and shmblands of varying disturbance those of a duck in size and color-in the season. It is considered by the natives to be and maturity. excellent eating. A favorite pastime of young men in the olden time was to catch these birds in the breeding season at t11e risk ol' their lives. The plan was to lower a Hllnting Methods lad over the edge ol' a cliff with a stout coir rope round his waist, the upper end On many Polynesian islands today, birds are hunted with shotguns. Prehistoric passed round the trunk of a tree and firm ly grasped by a near relative (usually a Polynesians were high ly skilled at catching birds bytraditional methods, which brother, as distant relatives were not trusted). A basket slung round the neck would are still used on some islands. Boys today (and presumably in the past) begin to SOQn be filled with dead birds and the lad hauled up again. -~

Table 4.4 Summary af Bil-ds by Anolytic Zones fA to j j . Main Trench (Squares C30 to G30J, Tangalatau Rochhdtcr (MA,,-44 J, Mangaia, Cook IsIo,.u1~, july to August 1989

TAXO" A B C D E F G H TOTAl.

SEABlRDS Audubon's Shearwater (Puffinus lhenninieri ) 2 3 °Blaek-winged Petre! (Pterodromo nigripennis) 3 17 14 2 37 Unidentified petre! or shearwater (Procellariidae sp. ) l Polynesian Stonn-Petrel (Nesofregetta fuliginosa) l 2 White-tailed Tropicbird (Phoetlwnlepturus) l 5 Lesser Frigatebird (Fregata ariel) 2 3 2 7 Brown Noddy (Anous stolidus ) l 2 l 4 Blue-gray Noddy (Procelstema cerulea) l Com mon FairyTem (Gygiscandido ) 3 6 °Little FairyTem (Gygis microrhyncha) l

LANDBlRDS Gray Duek (Anas superctliosa) l 5 3 9 Lesser Golden-Plove r (M) (Pluvialis dominica) Crueken (I) (Gallus gallus ) l 2 3 13 3 6 3 31 OO Ripley's Rail (Gallirallus ripleyi) 2 3 7 6 18 Sooty Crake (Porzana tabuensis) l l .oMangaian Crake (Porzana rua) 2 13 11 26 oSociety Islands Ground-Dove (Gallicolumba erythroptera ) 4 3 7

oOGiant Ground-Dove (Gallicolumba nui) l l °Cook Islands Fruit-Dove (Ptilinopus rarotongensis) l 3 °Nuku Hiva Pigeon (Duculo galeata ) l l 2 4 °Rimatara Lorikeet (Vini kuhlii) 2 3 5 10 20 o'Conquered Lorikeet (Vini vidivid) 2 3 5 9 19 • Rimatara or OOConquered Lorikeet (Vini kuhlii or V vidivid ) 4 2 7 Mangaia King fisher (H alcyon mongaia ) l Mangaia Reed-Warbler (Acrocephalus kerearako) 2 2

TOTALS All species 8 27 24 7 20 4 15 13 52 47 217 All native species 7 27 22 4 7 l 9 13 49 47 186 Seabirds 5 21 22 4 4 l 6 2 67 Native landbirds 2 6 3 8 12 43 45 119 Extinet or extirpated landbirds l 7 12 40 45 105 Extinet or extirpated landbirds as % of all vertebrates O 0.1 0.7 1.5 2.5 24 Extinet or extirpated landbirds as % of all nonfish 0.5 8 18 24 41 Extinet or extirpated landbirds as % of all birds 4 47 86 77 98 SOURCE: From Steadman and Kireh 1990 I: Introduced species M: Migrantspeeies • Extant species extirpated on Mangaia •• Extinet species ... David W. Steadman Extinction5 of Polyne5ian Bird5 4\

Gathering eggs of wild birds, particularly of seabirds, was another popular for freezers, even yard chickens are shunned if one ean purchase and store the aetivity throllghout Polynesia. By contrast, eggs of chickens seem to be ultratender frazen crnckens, already plucked and gutted, that are imported shunned in many pIaees, both today and in the past. Small, uninhabited is from New Zealand, Australia, or the United States, the sanle countries where lands were often visited to gather seabirds and their eggs. Among the many people have begun to pay extra for "free-range" chickens and their eggs. examples are Marquesans visiting Hatutu, Motu lti, and Fatuhuku, Tahitians Chickens need not crass the road to provide cross-cultural amusement. \isiting Mopelia, Eastcr Islanders visiting Motu Nui, Mangaians andAitu Bird lime was an effective way to eatch perching birds up to the size of pi ta1.ians visiting Manuae, Atiuans visiting Takutea, Ma'ukeans visiting Maria, geons (Gill1885:90-91). This sticky substance was prepared by mixing the 'Euans visiting Kalall , and the Tikapia and Anuta visiting Fatutaka (see Den sap of Lreadfruit trees (Artocarpus altilis) with cmshed, baked nuts af the ing 196:3: 121, 122). The relative abundance of seabirds an uninhabited islands candlenut tree (AletLrites rrwluccana). Bird lime was spread on limos in places provides an import

the prehistoric populations of R exulans may have been large, and they are and agencies that fund research, became seriously interested in Polynesian is likely to have been important predators on a variety ofbirds. lands other than Hawaii and New Zealand. Although my main concem here is prehistory. I should mention that the Another topie for research would be to detennine the potential impact on period ofEuropean inHuence has brought some other nonnative predators to native birds ofblood-bome parasites (hematozoa), especially those that cause parts of Oceania, such as the cat (Felis catus) and mongoose (Herpestes sp.). avian malaria. These diseases are transmitted by dipterans (mosquitoes in the Among the various European-introduced herbivores to tropical Oceania, the case of avian malaria). Steadman et al. (1990) found no hematozoa among goat (Capra hircus) has probably done the most damage to natural habitats blood smears from 79 native and introduced birds from the Cook Islands. through over-browsing native plants and dispersing the seeds of nonnative While this absence would seem to be good news from a conservation stand plants. point, it may also indicate that native species of birds would have little resis tance to blood-bome parasites should they be introduced to the islands. Atolls Understanding the Survivors ofthe northem Cook Islands lacked mosquitoes in pre-European times (Gill The human occupation of tropical islands has been detrimental to a wide vari 1885:199-2(0). Whether this was also tme in the southem Cook Islands and ety of hirds, leading to the extinction of many species and the elimination or elsewhere in Eastem Polynesia remains unknown. reduction of innumerable populations. So far, it would seem that the impact Yet another subject for additional research and monitoring concems tl1e of humans on Polynesian birds has been almost unfailingly negative. Most of forests themselves. Most of the endangered species of Polynesian landbirds the surviving species are those that prefer open or wetland habitats (herons, require forested or partly forested habitats, which continue to face threats ducks, volant rails, apd migrant shorebirds) or that tolerate some levelof for from logging and the encroachment of nonnative species. Detailed descrip est clearance (certain fmit-doves, kingfishers, and warblers). Even many of tions and analyses of Polynesian forest vegetation, such as those of Stoddart these species are declining, and some seem destined to extinction within dec (1975a), Merlin (1985, 1991), and Franklin and Merlin (1992), do not exist for ades. There are, however, opportunities to reverse some of the losses ofPoly most islands. The ecology of dispersal and pollination is virtually unknown for nesian landbirds, including the forest-Ioving species. many species ofPolynesian forest trees. I cannot imagine, however, that there One strategy would be to translocate species onto islands that they pre have not already been serious long-term impacts on the dispersal and repro viously occupied, after it has been determined that adequate habitat now is duction of forest trees caused by the loss of hundreds of populations of suitable and that current human activities are compatibie (Franklin and frugivorous, granivorous, and nectarivorous forest birds, such as megapodes, Steadman 1991). Uninhabited islands are especiallywell suited for such pro pigeons, doves, parrots, thmshes, starlings, and honeyeaters. Thus the loss of grams. As Polynesians move to major population centers, many relatively re birds has undoubtedly had a negative effect on forests, which in tum has had a mote islands are being abandoned by humans or are decreasing in population. negative irnpact on the human populations that have depended on the forests The chances of success for bird translocations would be enhanced by good for food, medicine, and building materials. background knowledge of the species involved. Unfortunately, we do not Flying foxes or fruit bats (Chiroptera: Pteropodidae) are also important know much about the ecology and behavior of the landbirds that survive in pollinators and seeddispersers in the South Pacific (Coxet al. 1991; Fujita and Polynesia. With but a few exceptions, data are lacking on the precise distribu Tuttle 1991). Like birds, 6ying foxes were hunted and eaten prehistorically, tions, current population sizes, habitat preferences, and vulnerabilities of reducing their numbers and range (Steadman and Kirch 1990; Steadman Polynesian birds. 1991b; Koopman and Steadman 1995). These large bats continue to be popu Surveys like those conducted in American Samoa by Amerson et al. lar food throughout much of the Pacific, resulting in dramatic further decIines (1982a, b) should be undertaken at regular intervals throughout Polynesia. in recent decades (Wiles and Payne 1986; Wiles et al. 1989). Detailed studies of nesting and food habits are lacking for most Polynesian landbirds. Where such studies have been carried out, as for the endangered THE I MPACT OF BIRDS ON HUMANS Rarotongan Monarch [Flycatcher ) or kakerori (Pomarea dimidiata), an effec tive conservation program was implemented (McCormack and Kunzle Prehistoric Polynesians had a major and negative impact on native hirds. In this 1990a). M any other similar programs could be undertaken if more scientists, the Polynesians were not unique; I am aware of no group of humans, past or .. David W. Steadman Extinclions of Polynesian Birds 69

present, rich or poor, Westemized or tradition al, that has not reduced popula protein to early human colonists through out Polynesia, the seemingly intense tions ofbirds and other animals, mainly through habitat alteration and hunting. search for undiscovered, pristine islands may have been stimulated, at least in We aIready know that, on island after island, bones from archaeological part, by the abundance of tame birds that greeted the discoverers of each new sites indicate that indigenous species ofboth seabirds and landbirds were ex island. Thus the pursuit of unexploited avifaunas may have been a factor in the tirpated within the nrst millennium ofhuman occupation. And, as the various remarkable rapidity with which early voyagers colonized much of the Pacinc. native birds became rare or extinct, the nonfish vertebrate diet ofPolynesians The tameness of Polynesian birds at first human contact must have been re depended more and more on domesticated and commensal species (chick markable, the various species having evolved in ecosystems. [ree of mammal ens, dogs, pigs, and rats). Here, however, I want to explore the reciprocal rela ian predators. Obtaining birds in the nrst centuries of human occupation may tionship: What sort of impact, aside from the ohvious one of providing food, have resembled our concept of gathering more than hunting. did native hirds have on the lives of prehistoric Polynesians? As we shall see, The extinction and extirpation of birds in Polynesia are reminiscent of Polynesians were interested in birds for many reasons other than as food. Be some aspects of the blitzkrieg model of vertebrate extinction (Mosimann and cause the POI)l1esian ethnobiology of birds is such an extensive suhject, my Martin 1975; Martin 1984, 1990): a rapidly dispersing people who hunted in coverage cannot be comprehensive. Rather, for each category, I shall provide tensively wherever they went, wiped out many species, and then moved on to a few examples and some discussion. richer huntinggrounds. Not everyone, of course, moved Oll. Polynesians had a Nothing is as effective as extinction to decrease the interest in, and uses for, rich and productive set of domesticated plants and animals; on many islands, a species ofbird. For each of the categories to be discussed, the process of avi some portion of the founding populations seems to have remained to establish faunal depletion thqt followed the human colonization of an island continually a more agriculturally based economy. narrowed the range of species that people could use, whether as navigational Navigation and Fishing or nshing aids, in legends and imagery, for their feathers or bones, or as pets. Tims the importance ofbirds in Polynesian societies, which was substantial at Seabirds were involved in traditional methods of navigation in Oceania. An European contact, must have been even much greater before many of the important aspect of detecting neamess to an island was to keep an eye out for species were lost. certain seabirds that seldom stray more than 20 to 100 km from the islands where theyroost atnight (Heyen 1963:71; Hilder 1963:90; Dening 1963:11~ Human Colonization 16; Lewis 1964:364; Sharp 1964:40; Gladwin 1970:180, 181, 188, 195-200; The mspersal and colonization of humans in what now is known as Polynesia Lewis 1972:162-73; Finney 1979:334). Useful in this regard were boobies began about 3,500 years ago, when a pottery-making people characterized as (Sula spp.), the Blue-gray Noddy (Procelstema cemlea), and especiaIly Brown "the Lapita cultural complex" arrived in Western Polynesia after a rapid mi and Black Noddies (Anous stolidus, A. minutus) and the Com mon Fairy Tem gration through Melanesia (see various chapters in Kirch and Hunt, eds., (Gygiscandida, = G. alba). Tropicbirds (Phaethon spp.), frigatebirds (Fregata 1988). Although the record of birds exploited by the Lapita people is incom spp.), and Sooty Tems (Stema fuscata) also were helpful. although their wan plete, evidence from Mussau, New Caledonia, Tikopia, Anuta, Fiji, Tonga, dering habits made them fallibie as indicators of neamess to land. In addition Futuna, and Sanloa indicate that a wide range of seabirds and landbirds were to true seabirds, the melodic two-note whistle of a migrant shorebird, the taken and that many species did not survive the first millennium ofhuman oc Lesser Golden-Plover (Pluvialis dominica), could be heard both day and cupation (Balouet and Olson 1987, 1989; Steadman 1989a, 1989b, 1993a, night, and indicated neamess ofland. 1993b and 1995; Steadman, Pahlavan, and Kirch 1990). Observing birds also assisted fishermen in finding nsh. In the Caroline Is In spite of uncertainties about the chronology and spatial pattem ofhuman lands, Gladwin (1970:30) described how nearshore surface feemng by nod colonizatioJl in Eastem Polynesia (Kirch 1986), there is a consistent pattem of dies and fairy tems would trap "the little fish in tlleir frenzy between enemies heavy exploitation of native birds early in the cultural sequence (which begins above and below. It is the birds which signal to the nshennen that the (larger l up to 2000 BP.), followed by an increased dependence on domesticated and fish are running in a school. All the canoes tum and pi unge toward the birds." cOllllllensal species (Dye and Steadman 1990). On Mangaia, by contrast, fishermen today say that tlle kakaia (Common Fairy Because birds provided such an important and easily available source of Tem) is a trustworthy bird whose feeding activities often reveal productive 70 David W. Steadman Extinctions of Polynesian Birds 71

fishing grounds, whereas the ngaia (Brown Noddy) is a "cunning" bird that Feathers ofte n willlead fishermen to sterile waters. Items made of feathers occulTed throughout Polynesia. Although c1oaks, lnhabited islands tend to have far fewer seabirds than do uninhabited is headdresses, skirts, and fans are especially well known, feathers adomed lands. Derling (1963:114) noted that "birds in great numbers became ac many other items, such as wooden and tapa images and the tails ofkites (Buck cepted in the Pacific by the explorers as the sign of an uninhabited island. In 1944:80-102,258,311,318,320,327,332-36,345-47, etc. ). The color red this we might find an explanation of why almost every uninhabited island in was associated with major deities, such as Tangaroa in the Cook Islands (Buck the Pacific gives signs ofhaving been visited by the Polynesians. Lost voyagers 1934:16) and Oro in the Marquesas, Tahiti, and Tonga (Parsonson 1963:29). would be easily attracted by the sign of birds." Fortunately for returning se a . Thus red feathers, when available, were particularly important and valuable. farers, noddies and fairy tems have been more resistant to over-exploitation Red feathers were sometimes used to decorate canocs (GillI894:134). In the than have other seabirds and thus have retained their usefulness in locating Cook Islands, the word kura was used variously to mean "red," "feather head home islands in areas where sailing and deep-water fishing still occur. dress," or "parrot." I believe that the parrot signi fie d by kura is Vini kuhlii (the The lass, however, af most other species of seabirds an most islands must Rimatara Lorikeet), a predominantly red species tllat sUrvlves on Rimatara in have diminished the importance of seabirds as both navigational and fishing the Austral Islands but has been found in archaeological contexts on Aitutaki, aids for prehistoric sailors. With few exceptions, the populations of shear Atiu, and Mangaia and seems to have occulTed into early historic times on Pal waters and petrels have been eliminated or depleted throughout Oceania. Es merston, Manuae, and Rarotonga. Words ranging from priceless and precious pecially out of the nesting season, most species of shearwaters and petrels are to sacred and beautiful were used to describe red parakeet feathers in Man highly pelagic. The former role of these and other pelagic seabirds in blue gaian stories. After the Rimatara Lorikeet had been lost on Mangaia, its treas water navigation is uncertain (Hilder 1963:83,84; Sharp 1964:42,43,47), al ured feathers were imported in late prehistory (GillI894:235, 240--42, 255). though their feeding activities far offshore probably aided fishermen in Although theyare known only from bones, it is Iikely that the larger extinct locating pelagic species of fish and marine mammals. WhiJe I agree with Le parrots of Eastem Polynesia (Vini vidivici of the Marquesas, Society, and wis (1972: 172, 173) that the role of migratory landbirds in facilitating prehis Cook Islands, and V sinotoi of the Marquesas and Society Islands) also had toric long-distance voyaging is speculative and subject to criticism, those (like significant amounts of red in their plumage. There is ethnographic evidence Sharp 1964:59, 61) who doubt the potential utility of pelagic seabirds in long that red featllers were brought to Mangaia from as far away as Tahiti, where ocean voyages should bear in mind that a respectable understanding of this no red parrot exists today. (The Black-fronted Parakeet, Cyanoramphus matter is impossible today because: (1) seabirds are so drastically reduced in zealandicus, and Raiatea Parakeet, C. ulietanus, which sUrvlved into the early range and numbers (the total number of resident seabirds in the tropical or mid-nineteenth century in the Society Islands, had only minor amounts of Pacific today may be 100 to 1,000 times less than it was 3,000 years ago); and red in their predominantly brown-and-green plumages.) In describing a late (2) the persons who study Polynesian navigation nowadays almost certainly do prehistoric conflict, Buck (1934:131) stated: "The father [Rongo-arikiJ, ter not understand the habits and field identification of Pacific seabirds as well as af prehistoric sailors did. handing his son [Rori) some precious red feathers from Tahiti, stayed behind Related to this is the possibie role bf seabirds as food to help sustain early to delay the enemy and so make good his son's escape." lronically, Rori's own Polynesian voyagers. Before their population declines occurred, seabirds may three sons were later killed in battle whiJe wearing red feather headdresses have been a signi/kant food supplement, although methods ofluring them at (Buck 1934:206). sea near enough to be captured are unrecorded. Ethnographic accounts of With tlle loss of red parrots virtually through out Eastem Polynesia, the two long-distance voyaging virtually lack mention of taking seabirds at sea; the long central tail feathers of the Red-tailed Tropicbird (Phaethon rubricauda) only reports l have found are of a group of native mission aries from Aitutaki became the primary source of red feathers. Pursuit of these tail feathers took \Vho killed a "few sea-birds" during five months at sea in the 1820s (Dening Polynesians to uninhabited islands, sometimes hundreds of kilometers from 1963: 138) and of several Great Frigatebirds (Fregata minor) caught after home. In the Fiji-Tonga-Samoa-Niue region, four species of palTots with red landing in the rigging of a ship off Nassau in the Cook Islands in the mid- in their plumage sUrvlved into historic times in spite of human-caused range 1800s (GillI885:31). contractions. These species are the Red-throated Lorikeet (Charmosyna Il David W. Steadman Extinctions øf Pølyne5ian Birds 11

1II1labilis). Collared LOlY (Villi [Phigys) sulitarius), Blue-crowned Lorikeet Whistles were made most often from tlle thin-walled bones of large (V australis), and Red Shining-Parrot (Prosvpeia tabuensis). A large extinct seabirds, such as frigatebirds or boobies. From' Anatu on 'Eua, I have exca parrot, Ecledlls new sp., was exterminated prehistorically on ' Eua (Steadman vated a still functional whistle (length 48 mm) made from the radius of a Lesser 1993a). Based upon the bright red plumage in the females of its only sUlviving Frigatehird (Fregata ariel). From Hanatekua Shelter No. 2 site (MH-11) on dose relative, the Edectus Parrot (E. roratus) of Western Melanesia, it is Hiva Oa in the Marquesas, I have identified the cut proxlmal44 cm of another probable that the extinct species of Eclectus from' Eua was yet another source radius of F ariel. Presumably, the adjoining piece (not recovered) would have of red feathers for early Tongans. been the whistle. Certain skeletal elements of chickens and large native birds, Although red may have been preferred throughout PoIynesia, feathers of especially the humerus, ulna, and tibiotarsus, were cut in short sections to \ i rtual ly any other color were also used. Prehistoric N iueans wore girdles that serve as heads. From Tangatatau Rockshelter on Mangaia, P. V Kirch and l re illcImled white, yellow, and green as well as red feathers ofV australis (Loeb covered in 1991 the entire sternum of a chicken in which two holes had been 1\:126:93, 163). Gill (1894:26, 27) reported a eloak from Mangaia made of"the drilled. This probable pendant came from strata dated to about 600 BP beautiful white, green, blue, and yellow feathers of the birds they had eaten." Names, Imagery, and Legemls I!eathers of frigatebirds and pigeons (Ducula spp.) were used on many is hmds. Chicken feathers were used for a variety ofheaddresses and other orna These three subcategories are too intertwined to separate dearly. Polynesian ments on Easter Island (Metraux 1940:220-28). The choice of chickens was names for hirds are significant from both a linguistic (Clark 1982) and an orni an obvious one, as all native landbirds and most seabirds had been wiped out thological (Steadman 1985) standpoint. Many avian names reappear as cog prehistorieally on Easter Island itself. Certain seabirds could be obtained sea nates throughout much or all ofPolynesia. One of many such examples is the son ally on hvo offshore islets, and F. E. Eyraud described "a sea bird whose proto-Polynesian matuku (variations include mafu'u, mOluku, kotuku, and careass had been opened more or less carefully" and used as a headdress by an others; Clark 1982), which refers to the widespread Pacifie Reef-Heron Easter Islander in I86fH)7 (Metraux 1940:220). (Egretta sacra). Occasionally, some rather drastic name changes reflect !ocal onomatopoetic inteipretations, such as tanga' eo for the Mangaia Kingfisher Bones rather than ngotare or kotare, used for elosely related Eastern Polynesian spe Being light and hollow but very strong, bird bones were important raw mate cies. Other major name changes are difficult to explain, such as iwa for fri rials filr tooIs, toys, and ornaments. Because bones are readily preserved in ar gatehirds in Hawaii rather than the widespread name kata' a and its cognates. chaeoIogical contexts, OUT knowledge of them, unlike our knowIedge of In some areas ofPolynesia, one ean find names ofhirds in dictionaries that feathers, covers the entire period of prehistory. refer to species that no longer exlst on a particular island. Such names gener Sewing needles were made of the ulna or, more often, the radius, by cut ally have fallen into disuse. Dictionaries for Rarotonga (Savage 1980, actually ting one end peipendicular to the Ion g axis, drilling a small hole in the shaft compiled between about 1900 to 1940) and Tahiti (Andrews and Andrews near the cut end, and filing the other end to a hollow point. This type of needle 1944) are rich SOUTces of such names. To mention only two examples, Savage is commonly found in prehistoric contexts on Easter Island (Metraux (1980:117, 122) lists Rarotongan words that indicate the former presence 1940:213; Heyerdallll96l:412), where it is calledivitia nua. AtAhu Naunau, there of a kingfisher (kotare; presumabIy the Chattering Kingfisher, Halcyon Anakena, Easter Island, two sueh needles were recovered in July 1991 from tuta, or elosely related species) and a small red parrot (kura; presumably the sediments dated to ahout 700-800 BP (Steadman, Vargas, and Cristino 1994). Rimatara Lorikeet). These needles had been made from the radii of MUiphy's Petrel (Pterodroma Bones from archaeological sites ean eorroborate the former exlstence of ulti 1//(/ ), a tropical seabird that no longer occurs on Easter Island. birds otherwise known only from stories or linguistic evidence. Clark (1982) Tattooing needles were made of bird bone on Easter Island (skeletal ele dedueed thata hird similar to the PUiple Swamphen (Purphyrio pO/7Jhyrio, ment and species unrecorded; Metraux 1940:237'38, 241), the Cook Islands proto-Polynesian name kalae ) must have exlsted in the Marquesas or Society (details unrecorded; Buck 1944:128), and Tikopia (humerus, ulna, and radius Islands, the presumed source areas of Hawaiian people, because the Hawai of frigatcbirds; Steadman, Pal1Iavan, and Kirch 1990:147). Mammal bones ian cognate' alae refers to similar large rails (Common Gallinule, or Moorhen, also were used as tattooing needles in many places. Gallinula chlorvpus, and American Coot, Fulica am.ericaIla) . No rails in the 14 David W. Steadnlilll Extinctions of Polyneslan Birds 75

genera Porphyrio, Gallinula, and Fulica , however, had ever been found in The Polynesian names ofbirds were sometimes used to name otller things. Eastem Polynesia (Holyoak and Thibault 1984), the nearest occurrence being Te Ana o Kakaia (the Cave of the Fairy Tem) is a large cave in the lvirua dis that af P porphyrio in Tonga, Samoa, and Niue (Pratt et al. 1987). Raynal trict af Mangaia that served as a place af refuge for Ruanae and his dan after (1980-1981) noted that the Marquesan name koau refers to a Bightless bird, they were defeated in a battle said to have occurred in 1718 (GiIl1885:74, with bluish purple plumage and yellow bill and feet, that existed earlier this 1894:167). True to the cave's name, each af the six bird bones I recovered century an Hiva Oa. Raynal proposed that the koau was related to the Bight from Te Ana Kakaia in 1984 were those af the kakaia (Gygis candida). The less swamphen af New Zealand, known as thetakahe (Porphyrio manteUi). In lack ofbones from extinct or extirpated species suggests that the cave was in 1986 and 1987, while examining bones from archaeological sites an Hiva Oa habited only late in prehistory. and Tahuata, I discovered 19 specimens that belonged to an undescribed spe Two other well-known Polynesian caves have names involving hirds. Ana cies af swamphen, subsequently named Porphyrio paepae (Steadman 1988). Taketake (var. Ana Takitaki) an Atiu is named after the sound made by echolo While this adds support to the proposals of Clark (1982) and Raynal (1980- cating Atiu Swiftlets (Collocalia sawtelli; Holyoak 1974, 1980; Holyoak and 1981), it does not solve the linguistic discrepancy between kalae and koau. Thibault 1978; Steadman 1991b). Ana Kena an Easter Island is named after The bones from Hiva Oa are all more than 1,000 years old, while those from the Masked Booby (Sula dactylatra). Ana Kena, originally the name af a cave Tahuata are at least 700 to 800 years old. Raynal and Dethier (1990) have sug only, has been merged into the single word Anakena to refer to the surround gested that a "cryptozoological" search be made on Hiva Oa to see if P. paepae ing region, which consists of a valley mouth and protected bay, a site of much still exists. archaeological and paleoecological importance (Kirch, Christensen, and A somewhat similar situation involves megapodes (Megapodius spp.). Steadman n.d.; Steadman, Vargas, and Cristina 1994). Clark (1982) pointed out that cognates of the Tongan and proto-Polynesian Two vanquished leaders in Mangaian prehistory were named aft er birds word for megapode, malau, are found in various places in Melanesia and In (Buck 1934:35): Mokara (Gray Duck, Anas superciliosa) and Kota' a (frigate donesia. Because megapodes are lacking in Fiji, through which proto-Polyne bird; Polynesians seldom distinguish between the two local species, F minor sian speakers must have trave led an their way to Tonga, use of the word malau and F ariel). The bird's nest fem (Asplenium nidus) of Cook Island forests in Tonga suggests that megapodes must have existed at one time in Fiji. (In is called rau kota'a (Gill 1894:224) . or, more aften today, simply kota'a Tonga today, a single species af megapode, M. pritchardii, is confined to Ni (Wilder 1931:9; pers. observation), in reference to its leaves, which resemble uafo' au.) Archaeological bones have now shown that megapodes did exist two in shape the lang, narrow wings of frigatebirds. The grace ful kites once made ar three millennia ago not only in Fiji but also on Ofu (American Samoa) and in the Cook Islands were called manu tukutuku ("letting go a bird"), lnanu on Lifuka and 'Eua in Tonga (Steadman 1989b, 1993a, 1993b). The most aka-rere ("flying bird"), or simply manu ("bird"; Buek 1944:257; Savage thorough record is in Tonga, where four species, three af them extinct (0), 1980:139). have been recorded (O M. molistructor, o M. alimentum, and M. pritchardii on The words rupe and lupe refer to llie PaciRe Imperial-Pigeon in the Cook Lifuka and Foa, OM. alimentum, M. pritchardii, and OM. new sp. on 'Eua). As Islands and Western Polynesia, respectively. On Mangaia, where no species of megapodes became increasingly localized in the Fiji-Tonga-Samoa region, pigeons survive, there is a valley named rupetau (GillI894:209). Tau means their eggs became one af the prestigious trade items of a well-orgaruzed long "to alight, to eome to rest" (Savage 1980:363), thus rupetau seems to be distance exchange network (Steadman 1991c; see Demng 1963 and Kirch named after a former pigeon roost. 1984:238-42, 1988:257-DO for detaiis af thi s network). All sorts af birds were involved in Polynesian stories and sayings, whieh Clark (1982) also used cognates af the proto-Polynesian lulu (Common often were based in part an the habits of a bird. On Niue, for example, many Barn Owl, Ty to alba) to propose that some sort of owl once existed in Eastem pekapeka (White-rumped Swiftlets, Collocalia spodiopygia) were said to Polynesia. This suggestion has not been substantiated by bones. Lastly, Clark leave their eaves and "dance" in the sky befare a rain (Loeb 1926:179). Thus (1982) interpreted cognates of the proto-Polynesian siwili and kula to suggest the swiftlets' period ofheavy aerial feeding on insects helped Niueans to pre that a larger parrot once existed in Fiji or Tonga, a suggestion now supported dict rain. Although Polynesian stories often reveal an appreciation and under by discovery on 'Eua of the large extinct parrot, Eclectus new sp. (Steadman standing ofbirds, they indicate a "dase harmony willi the natural world" (as 1993a). purported by Muse and Muse 1982:ix) only if sueh harm any ind udes preda- 16 David W. Steadman Extinctions of Polynesian Birds 11

tor-prey relationships. Many Polynesian stories include or imply the death of therefore value) of seabirds and their eggs. The seabirds and landbirds on the bird, particularly if it is a highly edible species, such as a shearwater or pi Easter Island itselfhad aIready been exterminated. geon. Birds also appeared in Polynesian riddles, such as "Who witll a black skin is Birds were ofte n associated with Polynesian deities. Such an association ever clotlled in purest wrute? A species of tem [Common Fairy Tem], com may have been accompanied by a taboo on killing that particular species. The mon in the Pacific" (GillI885:268). Following tlle arrival of mission aries in the fact that bones of"deified" species are frequently recovered in archaeological nineteenth century, birds continued to have roles in stories that were Chris sites may reflect local or chronological differences in their protective status. tianized (Gill1885:92, 108, 110, 136-37; 1894:372--73). For example, a com mon migratory shorebird, the Lesser Golden-Plover, is re Pets garded as sacred by modem Tikopia, who also associate with deities the resi dent PaciRc Imperial-Pigeon and migratory Long-tailed Cuckoo (Eudynamis Polynesians enjoy keeping birds as pets, a habit that may be decreasing in taitellsis ). Bones of all three species were found on Tikopia in archaeological popularity. Generally the birds are obtained as nestlings and are hand-reared, contexis (Steadman, Pahlavan, and Kirch 1990). increasing their tanleness. On Mangaia I have seen a Brown Noddy (obtained The endemic Mangaian Reed-Warbler (Acrocephalus kerear'ako), onoma locally) and a Red-footed Booby (Sula mia, obtained on Penrhyn) kept as topoeically called kerearako on Mangaia, was regarded as the incamation of tame pets in sheds berund houses. The noddy is also sometimes kept on Tik the god Tane (Buck 1934:163, 171). The Bristle-tlughed Curlew (Numenius opia. Frigatebirds were kept as pets in Samoa (Armstrong 1932:17) and Tu tahitiensis) is a large migratory shorebird variously called kiu or kau' a, again valu (Gill1885: 17-18), as Were White-tailed Tropicbirds (Phaethon lepturns ) because of its melc;xlic voice. Prehistoric Mangaians regarded both the curlew in Tonga (Lewis 1972:169). The Pacific Imperial-Pigeon and Crimson and warbler as "mouthpieces" of the god Tane (GillI885:116-17, 1894:332; crowned Fruit-Dove were commonly kept in Samoa (Armstrong 1932:1, 58). Buck 1934:172). The kau'a was "considered to be good eating by the natives, The Red Shining-Parrot is indigenous to Fiji (Layard 1876; Rinke 1989) but the tribe ofTane excepted" (GillI885:117). Another vocal bird, the Mangaia was kept as a pet in Samoa (Armstrong 1932:91) and still is held captive (and KingRsher, or tanga' eo, was associated \vith the gods Utakea and Tekura' aki exported?) on 'Eua, Tonga (pers. observation). The Tahiti Lorikeet (Vini pe (Buck 1934:166, 171). ruviana) was noted in captivity on Aitutaki (Townsend and Wetmore 1919). According to legend, a PaciRc Imperial-Pigeon belonging to the god Tan garoa mated with "a fernale shadow of great beauty" and became the fatller of CONCLUSION the first man on Atiu (GillI894:262r-63). On Mangaia, the god Motoro was "proudly calledte io ara = 'living-god: as rus worsruppers were not eligibie for It is not surprising that birds were extremely important to early Polynesians, sacrifice" (GillI894:332). Motoro was represented in sennit work and carv who arrived on one pristine island after another, each inhabited by a variety of ings from tlle oronga (a small tree, Pipturns argenteus), which used the flW' o birds but few if any species of reptiles or mammals. The importance of birds, or mo'o nw'o (Sooty Crake, Pop..ana tabu.ensis) as rus incamation (Buck however, did not save them. The net effect of the human occupation of Poly 1934:166, 171 ). Association of the mo'o mo'o with life may have been be nesia has been the elimination of much of the original bird life. Also gone is cause tllis bird inhabits taro swamps, wruch for centuries have been of utmost most of the natural habitat that supports those species of birds that have sur important to Mangaian subsistence. vived. Easter Island is well J...-nown for wooden carvings and petroglyphs of styl We know from archaeological sites that birds were killed and consumed ized birds and birdmen, the latter consisting of the head of a bird (often the regularly by early Polynesians. Nevertheless, it is difficult from thi s evidence Great Frigatebird) attached to a human body (Metraux 1940:256-59, 270-72; to assess the relative contribution to avian extinction from direct human Lee 1986). Bird glyphs also appear in tlle famous T'Ongorongo wooden tablets predation versus other prerustoric anthropogenic factors, such as habitat loss, ol' Easter Island (Metraux 1940:389-411). The "bird cult" ofEaster Island in disease, or predation from rats, dogs, and pigs. The importance of each factor volves the small population of Sooty Tems that nests opposite Orongo on the probably varied from island to island. Regardless of the specifics of any offshore islet of Motu Nui (Metraux 1940:331-41). Trus cult may have in individual case, human presence on tropical islands has been called an envi creased in religious significance late in prerustory because of the rarity (and ronmental catastrophe (Olson 1989). Although tllis statement may bc con tro- 18 David W. Steadman Extinctions of Polynesian Blrds 79

versial from an anthropological standpoint, you would find few birds that 4001-89), the National Science Foundation (grant BSR-8607535), and the Smith would argue against it. I say this without judgment; I am not criticizing any sonian Institution. For research permits, logistics support, and other cooperation, I thank the govemments of Easter Island (Chile), French Polynesia, the Cook Islands, individual person, living or deceased, who has killed birds (I have killed som e and Tonga. T. L. Hunt, H. F. James, P. V. Kirch, P. S. Martin, N. G. Miller, and S. L. Ol myself, in the name of science) or who has destroyed forest (which we all do, son kindly commented on the draft manuscript. at least indirectly, for food, lumber, paper, and so forth). l have mentioned some of the conservation programs that have been or could be undertaken to improve tlle lot of Polynesian birds. Although conser vation agencies can perhaps save more species per dollar on the mainland tropics than on islands, there is sometlling undeniably special about islands and their unique biotas. Programs to conserve Polynesian hirds should be considered for the sake of the hirds themselves, not to mention preserving an important part of the hutnan heritage of Polynesia. These programs are im portant for the future of science as well. It is no coincidence that Charles Dar win and Alfred Russel Wallace cameup with many of their most hrilliant insights wmle studying island faunas, as did other great names of zoology, such as Ernst Mayr, David Lack, Robert MacArthur, Edward Wilson, and Jared Diamond. In anthFOpology, such prominent figures as Bronislaw Malinowski, Margaret Mead, Sir Rayn10nd Firth, and Marshall Sahlins were inspired to new heights of creativity tllrough studies ofisland peoples. Any attempts to preserve the remaining fragments of Polynesian bird life must consider the importance of involving local people (Ray 1986). Interest in nature seems to be waning as cash economies replace subsistence agricul ture. Ifenvironmental protection is to succeed in Polynesia, this interest in na ture must be kept alive in young Polynesians as they inherit their island environments. More than SO years ago, Aldo Leopold wrote: "Conservation ists have, I fear, adopted the pedagogical method of the prophets; we mutter darkly about impending doom if people don't mend their ways. The doom is impending, all right; no one can be an ecologist, even an amateur one, without seeing it. But do people mend their ways for fear of calamity? I douht it. They are more likely to do it out of pure curiosity and interest" (1991:99, 101). In the neJet few decades it will be fascinating to participate in the interplay of sci ence, culture, economy, and environment in the South Pacific.

Note

l thank T. L. Hunt and P. V. Kirch for the opportunity to make a contribution to their symposium at the seventeenth Pacific Science Congress, upon which this chapter is based. The research was sllpported in part by the National Geographic Society (grant