%/& Conservation for the Twenty-first Century Edited by DAVID WESTERN MARY C. PEARL Wildlife Conservation International New York Zoological Society New York Oxford Oxford University Press 1989 Extinction on Islands: Man as a Catastrophe STORRS L. OLSON From a paleontological viewpoint, extinction is a pervasive, nearly inescapable natural process, though one that is hardly better understood than such phenomena as speciation. The threat of ex- tinction has been among the primary motivators of the conservation movement, because most historic extinctions have been due to man and are thus regarded as unnatural and preventable. If we look at historically documented extinc- tion*, we And almost none of exclusively ma- rine organisms. Although there 1* now great concern for habitats on continents, com- paratively few extinction* have been docu- mented in continental areas during the historic period. It is only when we turn to islands that man's negative impact on biotic diversity can be truly appreciated so far. Recognition of man- caused extinctions on islands can be traced back at least as far as the disappearance of the dodo (Raphus cucullatus) from Mauritius in the 1680s. Since then, many other species and pop- ulations of organisms have been exterminated on islands, as perhaps best exemplified by birds (Greenway 1958). Because until recently there was no paleon- tological record for most oceanic islands, it was natural to assume that European man was chiefly responsible for the degradation of in- sular habitats that has resulted in historically documented extinctions. This, in combination with the "noble savage" fallacy, has led to a gross underestimation of the effects of man on insular biotas. Now, with the paleontological record being expanded to many more islands, we have sufficient data to hint at the true magni- tude of the losses. The most startling data have come from the islands of the Pacific. Bones of the gigantic flightless moas of New Zealand were dis- covered more than a century ago. Since then, much more of the previous avifauna of New Zealand has been documented from archae- ological, swamp, cave, and dune deposits. The chronology of extinction leaves little doubt that the Maoris, through hunting and burning, have had a devastating impact on the biota of New Zealand (Casaels 1984; Trotter and McCuHough 1984). I knfe calculated the number and percentage of extinctions of resident land birds for New Zealand and the Chatham Islands, excluding species known or likely to have colonized the islands since the arrival of man. My systematic judgments are probably at variance with those of others attempting the same calculations, but the overall results would probably not be much dif- AfenLlgetGgurcs of Qf^r-two extant remideot species of land birds, of which at least nine are endangered. Extinctions in the historic period number twelve, and thirty-two species were ex- terminated prehistorically. Thus, 46 percent of the original fauna is now extinct and 33 percent of the Anna became extinct prehktorically. A recently completed study of bird remains from late Holocene cave deposits in New Cal- edonia (Balouet and Olson 1988) documents the extinction of hawks, megapodes, pigeons, owls, a gallinule, and a snipe, in addition to the peculiar, gigantic, flightless galliform bird (Syl- wofTwf). These fossils show that at least 25% of the resident species of nonpasserine birds of New Caledonia were exterminated pre- 50 EXTINCTION ON ISLANDS: MAN AS A CATASTROPHE 51 hisloncally, almoat certainly as m remit of human disturbance. But the fbasD faunas arc de- monstrably incomplete, and again, thoe are a number of species that have probably colonized the island since the arrival ofman. Taking these factors into account, we have extrapolated ex- tinction at at least 40% (or nonpasscrinc bird*. To this may be added the extinctions of a monitorlizard, a very peculiar crocodilian, and me horned turtle (Mew&uxb). In the Hawaiian Islands, prehistoric man- caused extinctions were truly massive in scale (Olson and James 1982, 1984) and included large flightless geese, flightless ibises, many flightless rails, owls, an eagle, a hawk, a petrel, and many species of small passerines. Bones from archaeological sites, Holocene sand dunes, lava tubes, and sinkholes have mo far documenW me extinction of 50 species, equal- ing 51% of the total number of native land birds of the archipelago. This is in addition to the 17 species that became extinct in me historic peri- od, so for the total avifauna yet known, 69% of the species are extinct. For the two imlands with representative fossil Annas, prehistoric extinc- tion was 69% onOahu and 71% on Maui, where only 10 indigenous species of land birds were recorded in the historic period, all small ar- boreal passerines of me tribe Dmpanidini. Of a total of 163 known island populations of en- demic species ofland birds, 82, or 50%, became extinct prehistorically, and 31, or 19%, became extinct historically, for a total of 113 (69%). Archaeological deposits on remote and sup- posedly pristine Henderson Island in the Rt- caim group show that the island was once inhab- ited by Polynesians, who exterminated at least two species of pigeons, representing 33% of the total known land bird fauna, of which four spe- cies arc sdll extant. These two species of pigeons are the same as, or very closely related to, species known historically from the distant Society and Marquesas groups (Steadman and Olson 1985). Bones ofpigeons, parrots, and flightless rails from Holocene cave deposits on the Island of Mangaia in the Cook group have raised the number of species of land birds known on that island from two to ten, so that 80% of the fauna became extinct prehistorically (Steadman 1985, in press). Most impressive are Steadman's (in press) recent discoveries of extinct species of rails, pigeons, and parrots in purely archae- ological sites on four of the thirteen islands of the Marquesas. Only eleven species of native land birds were known previously from me Marquesas, but the archaeological deposit* %ww add at least seven species m the toW, forammi- mumavifaiinal loss of 39% for the whole archi- pelago. Many of the species from these archae- ological sites are not yet extinct but do not occur on the islands whom their booes were found. The islands of Hiva Oa, Tahuata, and UaHuka each have only four historically known specie* of birds, whereas the fossil record adds five, six, and nine species, respectively, to these islands, for local extinction rates of 55 to 69%. On the island of Huahine in the Society group, only a kingfisher (ffafcyoa) and a war- bler (Acrm%pAo6?) are known historically, whereas bones from archaeological sites have now added at least seven species of rails, pigeons, parrots, and passerines, for an extinc- tion rate of 78% (Steadman, in press). Other extinct species or populations of birds have been documented from archaeological sites in Tonga, Fiji, Wallis, Tikopia, and the Santa Cruz Islands. Extinctions of birds following in the wake of European explorations in the fifteenth and six- teenth centuries have been documented both historically and palcontologically on Islands in the Indian and Atlantic Oceans. The Ate of the fauna of the Mascarenes, home of the dodo and me solitaire, which were large HigMras pigeons, is all too familiar. On the small island of St. Hekna in the South Atlantic, only one native species of land bird has survived, where- as fbasD remains show that at least four others were probably present when humans first ar- rived in 1502 (Olson 1975). Many watbiat extinctions have been documented in the West Indies, although good chronologies arc lacking, so it is difficult to sort out environmental from human-induced Actors. Nevertheless, we have documented through the fossil record drastic changes in the composition of vertebrate faunas of the Lesser Antilles that have taken place with- in the past 2,000 to 3,000 years, and these arc almost certainly the result of man's interference (Steadman et al. 1984). Although I have stressed land birds, the ef- fects of man on scabirds have been just as dra- matic. We know of the extinction by man of certain species of scabirds, as on St. Helena (Olson 1975), but moat of the reductions have 52 THE BIOLOGY OF CONSERVATION been in numbers and sizes of populations. The effect: (? man-caused removal of miDkms of individual predators from oceanic ecosystems within the past 2,000 yean have never been cal- culated but should be of concern to fisheries biologists and others. Organisms other than birds must necessarily have been affected by the nearly complete con- version of lowland habitats to agriculture throughout Oceania and elsewhere in the world. The only extensive evidence from the pre- historic record for anything other than verte- brates comes from land snails, where the same pattern of massive extinctions is discerned (see, for example, Christensen and Kirch 1986). In- sects, other invertebrates, and plants have doubtless been just as severely affected. When extrapolation is attempted from thcpaleon- tological data, it must always be remembered that fossil faunas are almost invariably in- complete. The data I have presented here are only minima, and the degree of extinction is always greater, sometimes much more so, than indicated by fossils. The fossil record has shown that most bio- geographical data based only on the historic rc- cord of islands are so misleadingly incomplete as to be all but useless for determining spe- ciea/arca curves or the natural distributions of individual speciea. Theoretical studies of island biogeography founded on such data, including studies that have been used in planning reserves in continental areas, are thus not likely to be particularly accurate or meaningful (Olson and James 1982; Stcadman 1986). It is still too early to make any realistically quantified estimate of the impact of man on in- sular ecosystem*. Too few Islands have been sampled paleontologically, and, as noted, most sample* are still quite incomplete. However, we can comider some examples that are illustrative of the scale of extinctions. Endemic species of flightless rails doubtless occurred on virtually every oceanic island in the world, with some islands having two or more species. Extrapolat- ing from the number and size of islands in Oceania, we may expect that hundreds of spe- cies of flightless rails have been exterminated in the Pacific in the past 2,000 yean or less. Ex- clusive of continental islands, New Zealand, and the Solomons, only fourteen species of fjighdea* rails, of all the hundred: predicted, were recorded in the historic period in the Pacif- ic, and all but three of these are already probably extincL About one-third of all the species of birds in the world are endemic to islands, and this figure is probably considerably underestimated by ap- plication of the so-called biological species con- cept, in which distinctive allopatric populations are considered as subspecies rather than full spe- cies. From the fossil record it is clear that spe- cics diversity of birds on virtually all oceanic islands was reduced by 30 to 50%, and some- times much more, within the period of man's occupancy. Thus, perhaps as much as one- quarter of all recent avian species were eradi- cated within an instant of geological time. Add to this the thousands of extinctions of other ver- tebrates, land snails, insects, and plants that must have taken place contemporaneously, cou- pled with historically documented extinctions, and we are faced with one of the swiftest and most profound biological catastrophes in the history of the earth. Unlike tropical rain forests, this catastrophic reduction in species diversity is not something that is projected into the future?it has already happened?and the remnants of insular biotas are continuing to be depleted at a voy rapid pace. Previously it was thought that mgh islands had greater species divenity because of their momlane rain forests. An important observation to emerge from recent studies, however, is that drier, more level, lowland habitats, the ones most susceptible to burning and clearing for ag- riculture, had greater species diversity than steep areas of high, wet forests. On islands, most species that persist in wet montane forests today do so not because this is their preferred habitat, but because it is the only habitat left that has not been too severely modified by man (see discussion in Olson and James 1982:42-49). Can this knowledge be applied to continental areas as well? Have we perhaps underestimated the diversity of mesic and arid environments in the tropics for lack of appreciation of the pre- historic influence of man? These environments, as on islands, are more susceptible to alteration by man than are rain forests, and their biotas may have experienced as yet undocumented prehistoric man-caused extinctions. Campbell (1979) has shown that the deseA west coast of South America harbored an extensive endemic avifauna that has largely disappeared since the end of the Pleistocene. These extinctions have EXTINCTION ON ISLANDS: MAN AS A CATASTROPHE 53 been tied to clirnadcdeterioradon and increas- ing desertification, processes that have ex- Bendedmto the Holoceoe and up to the present, but that wem also coincident with the arrival of man in that part of the continent. Can we really rule out the possibility that increased burning by man in habitats adapted to very little rainfall may have exacerbated and accelerated the pro- cess of extinction at the same time that the biota of the Pacific coast of South America was en- during climatic stress? It seems to me that a historical perspective and much better knowledge of Pleistocene en- vironmcnts is absolutely essential to planning large-scale conservation efforts in South Amer- ica. It is increasingly evident that neotropical rainforests are much less stable and of much younger origin than has long been thought (Campbell andFrailey, 1984). During the last glacial period, rain forests were greatly reduced in area, whereas the dominant habitat type was probably mesic or arid savanna (Lewin 1984). In some cases, areas that had been postulated as forest refugia were shown to have been savanna in the Pleistocene. Because the great biological diversity of SouA American ram forests abot& which we marvel today was sustained through- out the period when such habitats were much lew extensive than at present, is it not essential to try to detenmnc the location and extent of the late Pleistocene rain forests as a model for pre- serving modern diversity? And should we not also be equally concerned about conservation of habitats other than rainforest, which may be even more "fragile" and in the past may have been much more diverse? Fossils have shown that the human-induced biological catastrophe predicted as the future of tropical rain forests has long been underway on islands and perhaps in less humid continental areas as well. From such studies it is in- creasingly evident that the element of time, the paleontological record, and human history are essential factors that must be woven into the fabric of any successful strategy for conser- vation. REFERENCES B a I o u e t , Jean Chrlstophe, and Storrs L. Olson. "1988" = 1989. Fossil birds from Quaternary deposits In New Caledonia. S m Ithsonlan Contributions to Zoo logy, 469: 38 pages. Campbell, Kenneth E., Jr. 19 7 9. The non-passerine Pleistocene avifauna of the Talara tar seeps, northwestern Peru. Royal Ontario Museum Life Sciences Con tr Ibutlon 118: 203 pages. Campbell, Kenneth E., Jr., and David Frailey. 1984. Holocene flooding and species diversity in southwestern Amazonia. Quaternary Research. 21:369-375. Caaaala, Richard. 1984. Th* role of prehistoric man In the faunel extinctions of New Zealand and other Pacific islands Pages 741- 767. J_n_ Paul S. Martin and Richard G. Klein. Qu aternary Extinctions: A_ Preh IstorIc Revolution. Tucson: University of Arizona Press. Chrlatemaea, Carl C, and Patrick Y. Kirch. 1986. Nonmarln* mollusks and ecological change at Barbers Point, O'ehu, Hewal'i . Occasional Papers the Bern Ice P. Bishop Museum, 26:92-8 0. """ Green way, James C. 1958. Extinct and Vanishing Birds of the World. New York: American Committee for International Wild Life Protection (Special Bulletin 131. 518 pages. Lewin, Roger po I ten. 1984. Fragile forests Implied by Science, 226:36-37. PI elstocene Olson, Storrs I. 1975. Pa IeornItholoqy of St . Helena Is I South Atlantic Ocean . Sm Ithsonlan Contributions to Pa I eob I o I oq y , 23: 49 pages. ~~ "- Olson, Storrs L., and Helen F. James. 1982. Prodromus of the fossil avifauna of the Hawaiian islands. Smithsonian Contributions to Zoology. 365: 59 pages. Olson, Storrs L., and Helen F. James. 1984. The role of Polynesians In the extinction of the avifauna of the Hawaiian I aland*. Page* 768-789. _Ll Pawl S. Martin and Richard 6. Klein. Quaternary Extinctions: .A Prehistoric Revo I ufIon . Tucson: University of Arizona Press. Steadman , David W. Cook Islands. 109:58-86. 1985. Fossil birds from Mangala, Southern BuI let In of t h # British Ornithologists' CI ub Steadman, Oavld W. 1986. Floreana, Galapagos. 413: 103 pages. Holocene fossil vertebrates from Isla Sm 11h son I an Contributions to Zoo Io$y, Steadman. David W. "In pre**" - 1989. Fossil bird* and blogeography In Polynesia. Acta XIX Con gross us International Is OrnItho I og I c1 . Volume 2: 1526-1534. Ottawa: University of Ottawa Press. Steadman, David W., and Storrs L. Olson. 1985. Bird remains from an archaeological site on Henderson Island, South Pacific: man-caused extinctions on an "uninhabited" Island. ProceedIngs of the National Academy of Sc lances U.S.A., #2:61*1-6195. """ Staadman. David W., Gragory K. Prmglll, and Storr* L. Olaon. 1984. Fossil vertebrates from Antigua, Lesser Antilles: evidence for late Holocene human-caused extinctions In the West Indies. Proceed Ings of the National Academy of Sc Iences U.S.A.. 82:6191-6195. Trotter, Michael M., and Beverley McCulloch. 1984. Moas, men, and middens. Pages 708-727. J_n Paul S. Martin and Richard G. Klein. Quaternary Extinctions: A^ Prehistoric Revolution Tucson: University of Arizona Press.