Pacific Science (1995), vol. 49, no. 1: 87-97 © 1995 by University of Hawai'i Press. All rights reserved

Why Do Introduced Species Appear to Devastate Islands More Than Mainland Areas?!

DANIEL SIMBERLOFF 2

ABSTRACT: Island biotas are viewed popularly as much more fragile than those ofmainland areas and much more prone to damage from invaders. There are far too few data to assess this view thorougWy; for example, failed invasions are often unrecorded, and claims that an introduced species has displaced a native one are often based on correlated population changes rather than experiment and/or detailed field observations. Ifthere is a tendency for invasions to affect island communities more than mainland ones, it is far from universal; virtually every kind of damage wrought by invaders on islands has also been wrought in mainland areas. It is unlikely that, by virtue of their reduced species richness alone, island communities pose less "biotic resistance" to invaders than main­ land communities do. Rather, certain entire groups of species, like terrestrial mammals, are often missing from islands, and these absences can predispose certain invaders to be especially likely to survive and to produce particular im­ pacts.

CONVENTIONAL WISDOM (e.g., Elton 1958, (1968) saw islands as evolutionary "back­ Carlquist 1965, Wilson 1965) is that islands waters and dead ends," populated by species are more easily invaded than mainland areas, whose relatives had long ago been ex­ and ecological effects of invasions are greater tinguished in mainland areas. Greuter (1979) on islands than in mainland situations, be- depicted islands as repositories for old relicts, cause island biotas are somehow fragile and relatively unimportant in plant evolution. the species weaker than those of the main- Williamson (1981) believed that slower ana­ land. Yoon (1992: 88) eloquently expressed genetic evolution on islands leaves island en­ this view with respect to the Hawaiian Is- dernics likely to be extinguished by competi­ lands: "But the isolation of these gentle is- tion from introduced mainland species. In lands from mainland predators and diseases addition to the susceptibility of island biotas created a flora and fauna ill equipped to engendered in the long term by evolutionary handle the rigors of competition with the processes, the small number of species and outside world. Hawaii is home to thistles supposedly simplified food web structures without prickles, blackberries without thorns were believed to make island communities and many flightless insects and birds." particularly invasible (e.g., Wilson and Bos- The idea is that island species have been sert 1971). This belief was part of the dogma subjected to much less intense selective pres- that diverse ecosystems are more stable by sures. In particular, the smaller number of virtue of their diversity (MacArthur 1955, species on islands suggests that competition is Elton 1958). The notion that introduced spe­ less intense, so island species evolve to be less cies would be much more likely to survive on "competitive." Mayr (1965) found island islands than in mainland areas even has avifaunas to offer "far less resistance" to im- management implications. For example, .-. migFants-than-mainland-faunas- do~-Preston .. Immst193-1-)-felt-that-intro-ducin-glfaturaI-en­ emies for biological control would be an ef­ fective method only in the simplified com­ 1 Manuscript accepted 27 April 1994. 2 Department of Biological Science, Florida State munities of islands. University, Tallahassee, Florida 32306-2043. Nowadays, key aspects of the arguments 87 88 PACIFIC SCIENCE, Volume 49, January 1995 underpinning this conventional wisdom are 1992). But these are only surviving introduced questionable on numerous empirical grounds. species; who knows how many propagules of First, is competitive pressure really stronger other species landed but did not lead to pop­ on the mainland? Though Crowell (1962) and ulations or produced small populations that MacArthur et al. (1972) have seen competi­ went extinct before being recorded? Surely tion among species in bird communities as the number must be greater for the con­ lower on islands, Schoener (1965), Grant tiguous United States, but how much greater? (1968), Keast (1970), and others have argued Probably a lot, so it seems as if the prob­ that competition on islands may be stronger ability of an introduced insect's surviving in than in mainland areas because the resource Hawai'i is greater. Because Hawai'i has base is reduced and that this more intense about 5000 described native species (Nishida interspecific competition is reflected in greater 1992) and the contiguous United States has morphological divergence of close relatives about 90,000 (Arnett 1983), the probability on islands than in mainland situations. Per­ of survival seems greater in the site with haps generic statements about the relative fewer species. Is this probability of survival intensity of competition on islands and the greater because there are fewer species there mainland are not very useful, and one must to resist the invader? always speak of a specific island and main­ An order-by-order examination of the in­ land and also determine the relevant resource troduced insects of Hawai'i is not consistent base empirically. with this hypothesis (Simberloff 1986a). For Second, the relationship between stability example, Hawai'i has proportionally more and diversity or complexity is now seen as surviving introduced Coleoptera and Lep­ complicated and not monotonic (Begon et al. idoptera species than does the contiguous 1990, and references therein); some diverse, United States, but it also has proportionally complex systems are very fragile and some more native Coleoptera and Lepidoptera. By simple ones seem stable. That islands tend to and large, even though the native insect or­ have fewer species than mainland areas is ders are very differently represented in Ha­ likely to be far less important, from the wai'i than in the contiguous United States standpoint of invasibility, than the nature of (many orders are even missing), the relative those species that are present (Goodman proportions of surviving introduced species 1975). in the different orders are very similar in the Aside from assessing the underlying rea­ two sites. Without knowing how many spe­ sons for believing that island communities cies of each order were introduced in each are more likely than mainland ones to be place, we cannot even say where the prob­ disrupted by introduced species, it is also ability of survival is greatest, so we cannot worth examining whether empirical data really assess the hypothesis as we would wish really demonstrate this heightened suscepti­ to, but the data on surviving species suggest bility. that the question is not trivial. Probably the best published data for sur­ vival and disappearance of introduced spe­ Probability That an Introduced Species cies come from the literature on biological Survives on an Island control, although even here there have been A problem in assessing the likelihood that many unrecorded, more or less ad hoc in­ an introduced species survives on both main­ troductions (F. G. Howarth, pers. comm., land and islands is general lack of informa­ 1991). From an encyclopedic list of bio­ tion on failed introductions (Simberloff 1981, control introductions (Clausen 1978), I found 198-6a}--Apprax-imately-3000--slll"Viving---in­ -six -genel'a-~five-parasiticwasp genera -and-a troduced insect species are known from the beetle genus) with the largest number of spe­ contiguous United States (Anonymous 1986), cies that had been introduced, in various di­ and about 2500 are known from the Ha­ rections, among islands and mainland, with waiian Islands (Funasaki et al. 1988, Nishida Australia considered a mainland (Simberloff

::::mwa:awsUih briM' thiS 1+5 iff'''i' jiei@ffl U2~a.'.4 MAt gw Impact of Introduced Species on Island Biotas-SIMBERLOFF 89

1986a). I tabulated successes and failures of 1983). There are probably some competitive introductions from mainland to island, interactions between introduced birds that do mainland to mainland, island to island, and invade native forest (especially Zosterops island to mainland. If there were less biotic japonicus Temminck & ScWegel) and native resistance posed by island biotas and island birds (Scott et al. 1986). But habitat restric­ species were generally less competitive, one tions surely greatly reduce potential com­ would expect mainland-to-island introduc- petition between introduced species and na­ tions to have survived far more frequently tive ones, so that it is unlikely that resistance than island-to-mainland ones, with the other or lack thereof by native birds is the reason two classes intermediate. These six genera for success or failure of species introduced to produced 281 introductions; although the Hawai'i. probabilities of survival are in the order pre- Similarly, 20 passeriforms have been in­ dicted, the probabilities are all very close. troduced to the Mascarene Archipelago, of For example, 49% of mainland-to-island in- which nine survived on all islands to which troductions survived, and 33% of island-to- they have been introduced, nine failed wher­ mainland ones. A multiway contingency test ever introduced, and two survive tenuously does not reject the hypothesis that survival or on La Reunion but disappeared from Mau­ disappearance of a propagule does not de- ritius after more than a century (Simberloff pend on whether source or target areas are 1992a). As in the Hawaiian Islands, there is island or mainland. general habitat separation between the in- The general relevance of this result is less- troduced species and the native ones, with the ened because most biological control efforts native species in upland native forest and the on both islands and mainland are in agri- introduced species in lowland open anthro­ cultural communities, which are both far pogenous habitat. Only Zosterops borbonicus from pristine and more similar to one an- borbonicus (Boddaert) among the natives in­ other, no matter whether on island or main- vades open lowland habitats, and only Fou­ land, than pristine communities would be. dia madagascariensis (L.) among introduced Thus the species introduced to islands for species is found in forests; for both species biological control purposes were not, by and the anomalous habitat is a very secondary large, faced at the outset with competition one (Barre 1983). The species whose inva­ and other interactions from the entire native sions failed were also primarily species of community. Rather, resistance was from a open habitat; it is difficult to believe that small subset of it (that lives in anthro- their failure was caused by competition from pogenous habitats), as well as from other in- native birds. troduced species. This same problem obtains The smaller number of species on islands when one considers introduced birds, for than in nearby mainland areas is the most which some data exist for failures as well as widely known characteristic of island biotas. successes. At least 50 passeriform species Invasion by a new species of a new site, is­ have been introduced to the Hawaiian Is- land or mainland, is often viewed in terms of lands, of which 30 survived on all islands to the species' having to find an "empty niche" which they were introduced, 17 failed on all (Udvardy 1969, and references therein). Al­ islands to which they were introduced, and 3 though the static "niche" that this view survived on some islands and failed on others entails has given way to a more dynamic (Simberloff and Boecklen 1991). However, conception, the general idea still surfaces fre­ for the most part the introduced birds are quently and has management implications. found in agricultural lands, exotic forests, For example, many game species are in­ and built··up-areas, and-thesewerealsotypi- --- treGueea-tefill-suppesedlyempty-nicheson­ cal habitats of those species that failed. The islands (Ebenhard 1988). The vacancy of native land birds are primarily found in the niches is notoriously hard to prove (Herbold native forests, almost wholly restricted to and Moyle 1986). Further, since resources uplands (Berger 1981, Moulton and Pimm are ultimately metabolized in some way or 90 PACIFIC SCIENCE, Volume 49, January 1995

other, even if only by decomposers, the de­ specific interactions can be very difficult to piction of a niche as empty can always be elucidate without a controlled experiment or construed as simply a different way of de­ very detailed observations, yet such data are fining the niches of a particular community usually lacking. Suggestive observations can (Simberloff 1991). be very misleading. For example, the rapid Nevertheless, it is clear that the reduced spread ofthe North American mink (Mustela number of species on islands often leads to vison Schreber) in Great Britain in the 1950s the absence ofentire ways ofmaking a living. coincided approximately with the decline of The literature on disharmony ofisland biotas the otter. This synchrony led many to infer a (e.g., Carlquist 1974, Begon et al. 1990) de­ causal nexus; however, pesticide pollution is picts this same absence. So it is not surprising the chief reason otter populations have de­ that certain types of species are especially clined (Chanin and Jeffries 1978). likely to survive on islands ifintroduced. The The subtlety of effects of introduced spe­ African dung beetles introduced to Australia cies on natives, even some devastating effects, found a resource (300 million cowpats pro­ and the absence of detailed study in the vast duced each day) that was not used in the way majority of cases make any discussion of this they use it, because the native dung beetles phenomenon highly anecdotal. For example, are not adapted to this resource. And, of it is often said that introduction of natural course, the numerous successful introduc­ enemies, especially insect parasitoids and tions of predatory mammals such as rats, predators, for biological control is envir­ small mustelids, and the small Indian mon­ onmentally friendly. There is good reason to goose, Herpestes auropunctatus Hodgson, to be skeptical (Simberloff 1992b). Gagne and small islands lacking such species but con­ Howarth (1985) and Howarth (1985) argued taining abundant potential prey (often birds that such introduced insects might have and their eggs) can be viewed in this light (cf. caused the unintended extinction of several King 1984, Atkinson 1989). endemic Hawaiian moth species, a charge Thus, there may be a generic tendency for contested by Funasaki et al. (1988). The spe­ certain types of introductions to be more cifics of this case cannot be resolved with likely to survive on islands than in mainland available data (see Simberloff 1992b for de­ areas. About 1.6 times as many mammal tails), but it is difficult to believe that enough species and three times as many bird species attention was paid to species of pristine en­ have been successfully introduced to islands vironments, especially relatively rare insects than to mainland areas (Atkinson 1989). of no obvious economic importance, to sup­ Without a tabulation of how many attempts port the general conclusion of Funasaki et al. were made in each setting, one cannot say (1988) that biological control introductions with certainty that the rate of survival was are not harming native nontarget species greater on islands, but I think it is likely that (Simberloff 1992b). many of these island successes were facili­ Similarly, Groves (1989) claimed that no tated by absence of native species with sim­ insect introduced to control an invasive plant ilar "niches." has affected a native congeneric plant, but this claim is incorrect and, in any event, there Ecological Impacts ofSurviving has not been nearly enough field study of Introduced Species native plants, especially in pristine environ- ments, to justify lack ofconcern. A candidate With respect to degree of ecological im- endangered species, the semaphore cactus pact, the published record on introduced (Opuntia spinosissima Miller), was found in species-is-deficient-on-tw0-G0unts.-F-irst,it-is---the-United-States -only_on_LittleTorch-Ke¥, more likely that dramatic effects will be no- Florida. All individuals there have been de­ ticed, and the introduced species associated stroyed or are under attack by caterpillars of with them thus recorded, than if there is the moth Cactoblastis cactorum (Berg), which minimal ecological impact. Second, inter- was deliberately introduced to the island of

N'f2LllJilbj s QN Impact of Introduced Species on Island Biotas-SIMBERLOFF 91

Nevis in the Lesser Antilles to control an­ tains National Park (references in Simberloff other cactus and island-hopped over about [1991]). Mack (1986, 1989) explained the de­ 30 yr all the way to Florida. The important struction of the native plant communities of lesson from this introduction is that the cac­ the North American intermountain West and tus would have gone quietly extinct but for a their replacement by introduced Old World very unlikely set of circumstances. It in­ plant species as a consequence of the in­ habited an area favored by plant collectors troduction of large congregating grazers, to and botanists, in an intensively studied ham­ which American plant species were not mock that is part of a Nature Conservancy adapted. A similar fate befell South Amer­ refuge, and a local botanist was tipped to the ican and Australian grasslands when faced possibility of the moth's presence by a col­ with Eurasian invaders. Is the ecological im­ league who had noticed it at Guantanatno pact of grazers more severe on Auckland Is­ Bay, Cuba (Simberloff 1992b). How many land than in the intermountain West? No one species have been reduced or even extin­ has addressed this issue. guished without anyone's having thought to It is difficult to find ironclad examples on look? either island or mainland in which intro­ Granted that our knowledge of the true duced species have replaced native ones by extent of ecological effects is very poor, what competition, rather than by occupying a do the anecdotes suggest? We can begin by changed habitat (Udvardy 1969, Ebenhard looking at one dramatic effect, extinction. 1988). Ebenhard (1988), acknowledging that The number of known extinctions appar­ many of the cases he tabulated are not ently caused by introduced species is much strongly supported by empirical evidence, greater on islands (Atkinson 1989). Probably found that, of 1559 introductions of 330 spe­ the best-known reason is predation on island cies of birds and mammals worldwide, 79 in­ species that had not evolved in the presence troductions may have manifested competi­ ofpredatory mammals like cats, rats, and the tion between the invader and one or more small Indian mongoose, the key culprits aside native species, in that at least resources were from humans (Atkinson 1989). Absence of jointly used. There were relatively more such direct antipredatory behavior combined with cases in mainland areas than on islands, but, habits such as nesting on the ground to pre­ for introduced mammals (not birds), docu­ dispose such species to extinction. mented decreases in the population of the In fact, destruction of habitat by browsing native species (whether or not caused by and grazing by introduced pigs, rabbits, competition) were more common on oceanic sheep, cattle, horses, and other species has islands. The inferences about real impact for probably led to more island extinctions than most of these cases are so weak that it seems predation has, either by direct elimination of impossible to say whether introductions are an entire plant species or by reducing habitat more likely to lead to competition on islands quality and amount so much for animal and than in mainland areas. This is a surprising plant species that they are more prone to ex­ result considering the conventional wisdom tinction from the various factors that place described at the beginning of this paper. All very small populations at risk (outlined by one can really say for mammals and birds is Simberloff [1986b]). Although such habitat that competition seems to result from an in­ destruction is often depicted as far greater on troduction very infrequently, yet both island islands than on the mainland (e.g., Atkinson and mainland examples are known. 1989), there is really no thorough study of Disease vectored by introduced species is this matter. For example, pigs devastated often viewed as a key element in the devas­ plant-- communities of-:A:uckland- Island; re~­ tation- wrought by-inva-stOlrs- (e.g.,erosby stricting some species to sites inaccessible to 1986). Some of the introduced Asian passeri­ pigs (Atkinson 1989). Pigs have also greatly form birds may have contributed to the de­ affected plant communities and several spe­ cline of native Hawaiian species by vectoring cies of animals in the Great Smoky Moun- avian malaria (Ralph and van Riper 1985, 92 PACIFIC SCIENCE, Volume 49, January 1995

Van Riper et al. 1986), and a similar sce- sons, relating to the geological history of nario, though with no supporting data, was Eurasia and the long association of this re­ suggested for native Mascarene birds by gion with human activities, that might pre­ Cheke (l987a,b). In neither instance, how- dispose its species to be particularly adept at ever, does the evidence suggest that these invading other regions, including islands. diseases facilitated the survival of surviving One must always consider the opportunities invaders. Rather, the more likely scenario is for dispersal to and from various sites before that habitat modification created favorable postulating a hypothesis of greater likelihood circumstances for survival of at least some of inherent superiority of species from one introduced species, simultaneously reducing biota over those of another (Simberloff native populations, and the introduction of 1989). It may well be that far more species of new diseases may have worsened the plight of both disease and free-living organisms are the natives. transported from mainlands to islands than Exactly why island species should be more vice versa and that the per-capita probability prone than mainland ones to contract patho- of surviving does not differ. The same argu­ genic diseases from mainland invaders is un- ment may hold for disease organisms from known. Similarly, mainland species are rela- Eurasia compared with those from elsewhere. tively impervious to island diseases. Though I Perhaps disease susceptibility on islands is know of no statistical study of these proposi- a probabilistic phenomenon related to the tions, they are surely true. In addition to the species-area relationship. One can ask ifthere Hawaiian bird story, other anecdotes about is a null per-species probability that a para­ island populations devastated by introduced site will be severely pathogenic, so that re­ diseases are numerous (e.g., Ebenhard 1988, gions with more species have, on average, Pimm 1991), but I cannot identify an exam- more pathogenic species and potential hosts pIe of a disease endemic to an island that has in such regions experience more natural se­ devastated a mainland species. Of course, in- lection for resistance. Another possibility is troduced diseases have devastated rich con- that the smaller size of island populations tinental communities-rinderpest from Asia means that total genetic diversity is, on aver­ invaded large parts of Mrica in the late age, lower than in mainland situations. Car­ nineteenth century with enormous, lasting son (1981) argued that total genetic diversity impact on the ruminants and, through them, of continental species is far greater than that entire ecosystems (Dobson and May 1986, of island species, by virtue of the larger Barbault 1992, and references therein). The number of local populations. Thus the prob­ phenomenon may not be only a mainland-is- ability that resistant genotypes exist is lower land effect, but may be wholly or partly a on islands, and they will evolve more slowly statistical tendency for Eurasian diseases to once an introduced species arrives. Carson be more devastating in other areas. Only (1981) saw the reduced genetic base as re­ syphilis among New World human diseases sponsible for a generic tendency for island has invaded the Old World, whereas no hu- species to be less "aggressive," not well­ man diseases have been exported from Aus- adapted "general purpose genotypes" that tralia to elsewhere (Crosby 1986). Almost all typify invaders from the mainland. He devastating plant disease epidemics have pointed out (H. L. Carson, pers. comm., been caused by Eurasian pathogens in- 1993) that this pattern is far from universal. troduced to other continents (von Broembsen For example, among the endemic Hawaiian 1989). The pathogenicity of Eurasian human picture-winged Drosophila, most species are diseases and resistance by Eurasian humans monophagous, but five are so polyphagous -- -rnight-bea-result-ef-greater--Eurasian-human . ·that-eaGh-feeds·-0n -at--ll::ast-tivl::- -h0st-plant population densities, but the analogous trend families, including some introduced plant for plant pathogens cannot be so explained. species (Montgomery 1975). There has been Di Castri (1989) suggested generic rea- no precise quantification of aggressiveness

!fiHf-lE iAdfi'iffilriit6MMW4 bMW' • Impact of Introduced Species on Island Biotas-SIMBERLOFF 93 and general purpose capability, much less a tially has enormous ecosystem consequences, comprehensive comparison of all island and as do the introductions of some plants that mainland species to see if these traits are enhance fire frequency in various Hawaiian more frequent among the latter. ecosystems (Vitousek 1986). One possible generic reason why island It is not clear, however, that introduced ecosystems may be more prone to damage species are more likely to play such "key­ from introductions is that their species' pop­ stone" habitat-modification roles on islands ulations are perforce smaller and more geo­ than in mainland areas. The introduction of graphically restricted, and so less likely to salt cedar (Tamarix spp.) and Russian olive have some refuge, some regions not exposed (Eleagnus angustifolia L.) along rivers in the to predators, diseases, grazers, and so forth, American Southwest has created new forests than would mainland species exposed to even more extensive than those of Hawaiian similar invaders. Thus, extinction of a native mangroves with major ecological impact species is more likely for a given introduction (references in Simberloff [1991]). And in­ on an island, and, after extinction has oc­ troduced nitrogen-fixers are a major threat to curred, reinvasion and succession as means dune plant communities in northern Cal­ of repairing the ecological damage are fore­ ifornia (references in Simberloff[1991]). Most closed. On the mainland, by contrast, even if of the northeastern coast of North America vast areas are devastated, some areas are has been changed from mud flats and salt more likely to be unaffected by an introduced marshes to rocks by an introduced European species, and, if the introduced species is re­ periwinkle snail, Littorina littorea L. (refer­ moved or controlled, reinvasion is possible. ences in Simberloff[1991]), and native forests Introduced species can have enormous of much of South Florida have been replaced impacts on target ecosystems without directly by Melaleuca quinquenervia (Cav.) from causing extinctions. Species that constitute Australia and Schinus terebinthifolius Raddi entirely new habitats can entrain major from South America (Ewel 1986). The latter changes in a biota. For example, native two species are particularly inimical to the mangroves cover soft intertidal substrates in native biota because of the enhanced fire re­ most tropical bays and estuaries, but there gimes they foster. were none in the Hawaiian Islands until an However, the same reduced number of introduction in 1902. Now mangroves, espe­ species and disharmony noted above as cially Rhizophora, have spread, largely on characteristic of islands, especially small is­ their own, to many suitable sites throughout lands, would probably make it more likely the archipelago, forming substantial forests that an introduced species would "fill an in some areas where none had previously ex­ empty niche" in such a way as to become a isted. There has been little study of the ter­ new habitat on an island than on the main­ restrial or marine ecology of these forests, land. but the effect of this introduction on some ecological functions must be enormous Conclusions (Simberloff 1991). For example, mangrove swamps typically drop about 10,000 kg of The data are inadequate to draw strong leaves per hectare per year and accumulate conclusions about the relative invasibility sediment that builds land (Holdridge 1940). and fragility of islands and mainland. In They are major fishery nurseries (Hutchings particular, one would have to know much and Saenger 1987, and references therein). more about frequency of different types of Thorough modification of an existing habitat introductions and number of failed introduc­ can also-cause-community-wide changes.--Fol" - ti0ns.--F0Fe~ample,- the-fact that no -island example, the invasion of young volcanic bird species has invaded continents to the areas of Hawai'i by the Atlantic plant Myr- extent that species like the house sparrow or ica faya Aiton, which fixes nitrogen, poten- the Java sparrow have invaded islands may 94 PACIFIC SCIENCE, Volume 49, January 1995

be wholly or partly a consequence of the fact It would be worth exploring whether this that such introductions of island species to pattern extends to other archipelagoes. In mainland areas have never been performed. general, it is far from clear that island species Virtually all attempts to introduce island do tend to be weak, unaggressive, un­ birds have been from one small island to an­ competitive evolutionary incompetents; if other, usually in the same region, and many there is such a tendency, it is far from uni­ have survived (Long 1981). versal, and the reasons are still unknown. Despite the paucity of quantitative data, it appears that introduced species are more likely to survive on islands than in mainland ACKNOWLEDGMENTS areas, and are more likely to produce major ecological impacts, but there are many coun­ I thank K. Kaneshiro, H. Carson, S. terexamples. If this tendency exists, it is Pimm, L. Hamilton, and two anonymous probably not generated by an inherent, com­ referees for suggestions on the manuscript munity-wide "biological resistance" posed and K. Kaneshiro for facilitating my partici­ by mainland communities to invaders but pation in this symposium. vitiated on islands. It is not simply a matter of greater species richness in mainland areas conferring greater ecological stability. LITERATURE CITED Rather, it seems that the absence of entire groups of species from islands may have pre­ ANONYMOUS. 1986. Western Hemisphere Im­ disposed certain invaders from the mainland migrant Arthropod Database (WHIAD). to be likely to survive on islands and certain USDA Agricultural Research Service, introductions to be likely to have greater Beltsville, Maryland. ecological effects. In other words, the prox­ ARNETT, R. H. 1983. Status of the taxonomy imate factor in determining the trajectory of ofthe insects Qf America north of Mexico: an introduction may be the specific type of A preliminary report prepared for the species and the specific community it is in­ subcommittee for the insect fauna of troduced to, rather than simply the size ofthe North America project. Privately printed. island community. However, over the long ATKINSON, I. 1989. Introduced animals and term, the size and composition of the island extinctions. Pages 54-69 in D. Western community may have led to the evolution of and M. Pearl, eds. Conservation for the a community likely to be susceptible to cer­ twenty-first century. Oxford University tain types of damage from invasions. It is Press, New York. also possible that, if there is some inherent BARBAULT, R. 1992. Ecologie des peuple­ invasibility and fragility of island commun­ ments. Structure, dynamique, et evolution. ities, at least part of it is really an aspect of Masson, Paris. an inherent invasibility and fragility of all BARRE, N. 1983. Distribution et abondance communities when faced with Eurasian in­ des oiseaux terrestres de l'lle de la Re­ troductions. union (Ocean Indien). Rev. Eco!. Terre This is not a very satisfactory conclusion Vie 37: 37-85. to a paper about the supposed vulnerability BEGON, M., J. L. HARPER, and C. R. TOWN­ of island species and communities to inva­ SEND. 1990. Ecology: Individuals, popula­ sion. Species of some native island commun­ tions and communities, 2nd ed. Blackwell, ities seem to have resisted invasion as well as Boston. continental ones do. For example, in Ha- BERGER, A. J. 1981. Hawaiian birdlife, 2nd wai'i~the-plants-ofupland-nativeforests-are-­---ea-o-Ynive-rsit-y 0t'-Hawaii-I>-ress,-Honolulu. not greatly affected by exotic plants (Mueller­ CARLQUIST, S. 1965. Island life. Natural His­ Dombois 1981), and plant communities of tory Press, Garden City, New York. extreme habitats seem not to be easily in­ --. 1974. Island biology. Columbia vaded (Loope and Mueller-Dombois 1989). University Press, New York.

5 Ed# ;aM,i; Impact of Introduced Species on Island Biotas-SIMBERLOFF 95

CARSON, H. L. 1981. Microevo1ution in in­ ELTON, C. S. 1958. The ecology of invasions sular ecosystems. Pages 471-482 in D. by animals and plants. Methuen, London. Mueller-Dombois, K. W. Bridges, and H. EWEL, J. J. 1986. Invasibility: Lessons from L. Carson, eds. Island ecosystems: Bio­ south Florida. Pages 214-230 in H. A. logical organization in selected Hawaiian Mooney and J. A. Drake, eds. Ecology of communities. Hutchinson-Ross, Strouds­ biological invasions of North America burg, Pennsylvania. and Hawaii. Springer-Verlag, New York. CHANIN, P. R. F., and D. J. JEFFRIES. 1978. FUNASAKl, G. Y., P.-Y. LAI, L. M. NAKA­ The decline of the otter Lutra lutra L. in HARA, J. W. BEARDSLEY, and A. K. OTA. Britain: An analysis of hunting records 1988. A review of biological control in­ and discussion of causes. Bio!. J. Linn. troductions to Hawaii: 1890 to 1985. Proc. Soc. 10: 305-328. Hawaii. Entomo!. Soc. 28: 105-160. CHEKE, A. S. 1987a. An ecological history of GAGNE, W. C., and F. G. HOWARTH. 1985. the Mascarene Islands, with particular Conservation status of endemic Hawaiian reference to extinctions and introductions Lepidoptera. Pages 74-84 in Proc. 3rd of land vertebrates. Pages 5-89 in A. W. Congress on European Lepidoptera, Diamond, ed. Studies of Mascarene Island Cambridge, 1982. Societas Europaea Lep­ birds. Cambridge University Press, Cam­ idopterologica, Karlsruhe. bridge. GOODMAN, D. 1975. The theory of diversity­ ---. 1987b. The ecology of the surviving stability relationships in ecology. Q. Rev. native land-birds of Reunion. Pages 301­ Bio!. 50: 237-266. 358 in A. W. Diamond, ed. Studies of GRANT, P. R. 1968. Bill size, body size and Mascarene Island birds. Cambridge Uni­ the ecological adaptations of bird species versity Press, Cambridge. to competitive situations on islands. Syst. CLAUSEN, C. P. 1978. Introduced parasites Zoo!' 17: 319-333. and predators of arthropod pests and GREUTER, W. 1979. The origins and evolu­ weeds: A world review. U.S. Dep. Agric. tion of island floras as exemplified by the Agric. Handb. 480. Aegean Archipelago. Pages 87-106 in D. CROSBY, A. W. 1986. Ecological imperialism: Bramwell, ed. Plants and islands. Aca­ The biological expansion of Europe, 900­ demic Press, London. 1900. Cambridge University Press, Cam­ GROVES, R. H. 1989. Ecological control of bridge. invasive terrestrial plants. Pages 437-461 CROWELL, K. L. 1962. Reduced interspecific in J. A. Drake, H. A. Mooney, F. di Cas­ competition among the birds of Bermuda. tri, R. H. Groves, F. J. Kruger, M. Rej­ Ecology 43: 75-88. manek, and M. Williamson, eds. Bio­ DI CASTRI, F. 1989. History of biological in­ logical invasions: A global perspective. vasions with special emphasis on the Old Wiley, Chichester. World. Pages 1-30 in J. A. Drake, H. A. HERBOLD, B., and P. B. MOYLE. 1986. In­ Mooney, F. di Castri, R. H. Groves, F. J. troduced species and vacant niches. Am. Kruger, M. Rejmanek, and M. William­ Nat. 128:751-760. son, eds. Biological invasions: A global HOLDRIDGE, L. R. 1940. Some notes on the perspective. Wiley, Chichester. mangrove swamps of Puerto Rico. Car­ DOBSON, A. P., and R. M. MAY. 1986. Pat­ ibb. For. 1: 19-29. terns of invasions by pathogens and para­ HOWARTH, F. G. 1985. Impacts of alien land sites. Pages 58-76 in H. A. Mooney and J. arthropods and mollusks on native plants A. Drake, eds. Ecology of biological in­ and animals in Hawaii. Pages 149-178 in vasions--·of~~North- Amensa aml-H-awaii; - G;-P: Stone and-;J~ M:- Scott;-eds. -Hawai'i's Springer-Verlag, New York. terrestrial ecosystems: Preservation and EBENHARD, T. 1988. Introduced birds and management. University of Hawaii Coop­ mammals and their ecological effects. Vil­ erative National Park Resources Studies trevy (Stockh.) 13(4): 1-106. Unit, Honolulu. 96 PACIFIC SCIENCE, Volume 49, January 1995

HUTCHINGS, P., and P. SAENGER. 1987. Ecol­ MOULTON, M. P., and S. L. PIMM. 1983. The ogy of mangroves. University of Queens­ introduced Hawaiian avifauna: Biogeo­ land Press, St. Lucia. graphic evidence for competition. Am. IMMs, A. D. 1931. Recent advances in ento­ Nat. 121: 669-690. mology. Churchill, London. MUELLER-DoMBOIS, D. 1981. Understanding KEAST, A. 1970. Adaptive evolution and Hawaiian forest ecosystems. Pages 502­ shifts in niche occupation in island birds. 520 in D. Mueller-Dombois, K. W. Biotropica 22: 762-784. Bridges, and H. L. Carson, eds. Island KING, C. M. 1984. Immigrant killers: In­ ecosystems: Biological organization in se­ troduced predators and the conservation lected Hawaiian communities. Hutch­ of birds in New Zealand. Oxford Uni­ inson-Ross, Stroudsburg, Pennsylvania. versity Press, Auckland. NISHIDA, G. M., ED. 1992. Hawaiian terres­ LONG, J. L. 1981. Introduced birds of the trial arthropod checklist. Bishop Museum, world. Universe Books, New York. Honolulu. LOOPE, L. L., and D. MUELLER-DoMBOIS. PIMM, S. L. 1991. The balance of nature: 1989. Characteristics of invaded islands, Ecological issues in the conservation of with special reference to Hawaii. Pages species and communities. University of 257-280 in J. A. Drake, H. A. Mooney, F. Chicago Press, Chicago. di Castri, R. H. Groves, F. J. Kruger, M. PRESTON, F. W. 1968. On modeling islands. Rejmanek, and M. Williamson, eds. Bio­ Ecology 49: 592-594. logical invasions: A global perspective. RALPH, C. J., and C. VAN RIPER III. 1985. Wiley, Chichester. Historical and current factors affecting MACARTHUR, R. H. 1955. Fluctuations of Hawaiian native birds. Biol. Conserv. animal populations, and a measure of 2:7-42. community stability. Ecology 36: 533­ SCHOENER, T. W. 1965. The evolution of bill 536. size differences among sympatric con­ MACARTHUR, R. H., J. M. DIAMOND, and J. generic species ofbirds. Evolution 19: 189­ R. KARR. 1972. Density compensation in 213. island faunas. Ecology 53: 332-342. SCOTT, J. M., S. MOUNTAINSPRING, F. L. MACK, R. N. 1986. Alien plant invasion into RAMSEY, and C. B. KEPLER. 1986. Forest the intermountain West: A case history. bird communities ofthe Hawaiian Islands: Pages 191-213 in H. A. Mooney and J. A. Their dynamics, ecology, and conserva­ Drake, eds. Ecology of biological inva­ tion. Cooper Ornithological Society, Las sions of North America and Hawaii. Cruces, New Mexico. Springer-Verlag, New York. SIMBERLOFF, D. 1981. Community effects of ---. 1989. Temperate grasslands vulner­ introduced species. Pages 53-81 in M. H. able to plant invasions: Characteristics Nitecki, ed. Biotic crises in ecological and and consequences. Pages 155-179 in J. A. evolutionary time. Academic, New York. Drake, H. A. Mooney, F. di Castri, R. H. ---. 1986a. Introduced insects: A bio­ Groves, F. J. Kruger, M. Rejmanek, and geographic and systematic perspective. M. Williamson, eds. Biological invasions: Pages 3-26 in H. A. Mooney and J. A. A global perspective. Wiley, Chichester. Drake, eds. Ecology of biological inva­ MAYR, E. 1965. The nature ofcolonization in sions of North America and Hawaii, birds. Pages 29-47 in H. G. Baker and G. Springer-Verlag, New York. L. Stebbins, eds. The genetics of coloniz­ ---. 1986b. The proximate causes of ex­ ing species. Academic Press, New York. tinction. Pages 259-276 in D. M. Raup MONTGOMERY;-- -S~~-h ---1-9'7-5~-Gomflar-at-ive- ­ ane-D.-JableBski, eds.-Fattems-and-pro­ breeding site ecology and the adaptive ra­ cesses in the history oflife. Springer-Verlag, diation of picture-winged Drosophila Berlin. (Diptera: Drosophilidae) in Hawaii. Proc. ---. 1989. Which insect introductions Hawaii. Entomol. Soc. 22: 65-103. succeed and which fail? Pages 61-75 in J.

6 Ii Impact of Introduced Species on Island Biotas-SIMBERLOFF 97

A. Drake, H. A. Mooney, F. di Castri, R. zootiology and ecological significance of H. Groves, F. J. Kruger, M. Rejmanek, malaria in Hawaiian land birds. Eco!. and M. Williamson, eds. Biological inva­ Monogr. 56: 327-344. sions: A global perspective. Wiley, Chi­ VITOUSEK, P. M. 1986. Biological invasions chester. and ecosystem properties: Can species ---. 1991. Keystone species and com­ make a difference? Pages 163-176 in H. A. munity effects of biological introductions. Mooney and J. A. Drake, eds. Ecology of Pages 1-19 in L. R. Ginzburg, ed. Assess­ biological invasions of North America ing ecological risks of biotechnology. But­ and Hawaii. Springer-Verlag, New York. terworth-Heinemann, Boston. VON BROEMBSEN, S. L. 1989. Invasions of ---. 1992a. Extinction, survival, and ef­ natural ecosystems by plant pathogens. fects of birds introduced to the Masca­ Pages 77-83 in J. A. Drake, H. A. renes. Acta Oecol. 13: 663-678. Mooney, F. di Castri, R. H. Groves, F. J. ---. 1992b. Conservation of pristine hab­ Kruger, M. Rejmanek, and M. William­ itats and unintended effects of biological son, eds. Biological invasions: A global control. Pages 103-117 in W. C. Kauffman perspective. Wiley, Chichester. and J. E. Nechols, eds. Selection criteria WILLIAMSON, M. 1981. Island populations. and ecological consequences of importing Oxford University Press, Oxford. natural enemies. Entomological Society of WILSON, E. O. 1965. The challenge from re­ America, Lanham, Maryland. lated species. Pages 7-24 in H. G. Baker SIMBERLOFF, D., and W. BOECKLEN. 1991. and G. L. Stebbins, eds. The genetics of Patterns of extinction in the introduced colonizing species. Academic Press, New Hawaiian avifauna: A reexamination of York. the role ofcompetition. Am. Nat. 138: 300­ WILSON, E. 0., and W. H. BOSSERT. 1971. A 327. primer of population biology. Sinauer As­ UDVARDY, M. D. F. 1969. Dynamic zoo­ sociates, Stamford, Connecticut. geography. Van Nostrand, New York. YOON, C. K. 1992. Alien species threaten VAN RIPER, c., III, S. G. VAN RIPER, M. L. Hawaii's environment. Page B8. New GOFF, and M. LAIRD. 1986. The epi- York Times, 29 December.