Pacific Science (1978), vol. 32, no. 3© 1979 by The University Press of Hawaii. All rights reserved

The Introduced and Native Terrestrial Vertebrates of FijP

JOHN C. PERNETIA and DICK WATLING 2

ABSTRACT: A complete list of native and introduced Fijian terrestrial verte­brates has been compiled that includes a number of previously unrecordedreptiles. An analysis has been made of the habitat distribution of most species.The past and present status of the various native vertebrates is examined, andthe impact of post-European vertebrate introductions on the indigenous andendemic fauna is assessed in the light of current knowledge. It is concluded thatthe success ofcertain introduced bird species is due in part to their close associa­tion with human-modified habitats, while the native species are primarilyadapted to a forest environment. Although the introduced mongoose may beimplicated in the decline of some native vertebrate species, its effect may be lessimportant than previously stated and is certainly less than the effect of nativehabitat destruction and environmental modification by humans.

NUMEROUS STATEMENTS MAY BE FOUND in theliterature concerning the effect of introducedplant and animal species on the native com­munities found on Pacific islands (Berger1972, Guppy 1906, Thaman 1974, Williams1953). The bulk of such comments refer todeleterious effects in terms of causing reduc­tion in numbers of native species; or theelimination of such species from certainhabitats, areas, or islands through competi­tive exclusion (Mercer 1964, Parham 1956,Thaman 1974, Watson 1960, Wood andWetmore 1925); or direct mortality due tointroduced predators (Brewster 1935, Gor­man 1972, 1975a, b, Mercer 1964). Incontrast, Lack (1975) suggests that the intro­duced land birds of Jamaica have had nodirect or indirect competitive effect on thenative species and that the successful estab­lishment of certain post-European bird intro-

I This work was supported in part by a grant from theUniversity of the South Pacific to John C. Pernetta andin part by a grant from the British Ornithologists' Unionto Dick Watling. Manuscript accepted 28 June 1978.

2 University of the South Pacific, School of NaturalResources, P.O. Box 1168, Suva, Fiji. Present address ofPernetta: University of Papua New Guinea, Departmentof Biology, Box 4820 University P.O., Port Moresby,Papua New Guinea. Present address of Watling: Univer­sity of Cambridge, Department of Applied Biology,Pembroke Street, Cambridge, CB2 3DX, England.

ductions results from their close associationwith people.

The purpose of the present paper is toassess the extent of the interaction betweenintroduced and native species of terrestrialvertebrates in Fiji. It is based upon an exten­sive review of the often anecdotal literature,together with 9 years of observational dataon birds (D. W.) and the extensive collectionand observation of reptiles, amphibia, andmammals by both authors on various islandsof the Fiji group.

THE FIJI ISLAND GROUP

The Fiji group consists of approximately320 islands located between 16° and 20° S,177° W, and 175° E. The land area totals18,270 km2 , the bulk of which forms twomajor islands, Viti Levu and Vanua Levu(Figure 1). Southwest tradewinds blow stead­ily from September to April, resulting in ahot, wet summer and a drier, cooler winter.Mean monthly temperatures range from 23°Cin July and August, to 27°C in January;humidity ranges from 75% during winterto 88% in summer (Fiji Weather Bureaustatistics).

The geological history of the islands in­dicates that Viti Levu consists of an Eocene

223

224 PACIFIC SCIENCE, Volume 32, July 1978

HABITAT TYPES

volcanic base, Miocene clastic sediments,and volcanic intrusives (Houtz and Phillips1963). Later periods of uplift have resultedin extensive areas of reef-formed limestoneand coastal sedimentary deposits beinglocated far from the present coast. VanuaLevu and Taveuni appear to be formed fromcoalesced centers of volcanic activity of Mid­to Upper-Miocene age linked in part by reef­formed limestone (Rickard 1966). The islandsare, therefore, a mixture of high, large, vol­canic landmasses with a number of smallerlimestone islands and sandy cays. The highcentral plateaus and mountains of the mainislands (1200+ meters) create a significantrain shadow, such that the western sides ofthese islands are drier and have more distinctseasonal weather differences than the easternsides. The rugged and eroded landscape ofthe main islands is a consequence of theirrecent geological origin and high annualrainfall (mean: 300 cm on the east coast;165 cm on the west coast).

replaced in estuarine and soft sediment areasby mangrove forests of Rhizophora spp. andBruguiera spp. At elevations from 600 to1200 meters lie the montane forests, whichmay once have occurred down to 300 meterson the drier western side of Viti Levu. Suchforests are characterized by tree ferns,Cyathea spp., the tree Podocarpus vitiensis,and the valuable softwood, Agathis vitiensis.Smaller areas at higher elevations consist ofa high montane or cloud forest, with stuntedtrees and abundant moss and epiphyticgrowth. Another minor but important habitatis the inland swamp, often fringed withPandanus spp. and dominated by the reedPhragmites. A number of these swamps lackopen water and are little more than peat bogs(Twyford and Wright 1965).

On the western sides of the main islands,much of the original forest has been replacedby cultivated fields and grassland. Twyfordand Wright (1965) suggest that grasslandhabitats are maintained in an early seralstage as a result of periodic firing by humans.They are dominated by the introduced mis­sion grass, Pennisetum polystachyon, andwire grass, Sporobolus indicus, which have

The habitat types of Fiji have received a replaced the native reed grass, Miscanthusgreat deal of attention and have been clas- fioridulus, in most areas (Parham 1956).sified in various ways by different authors Between the grasslands and agricultural areas(Berry and Howard 1973, Gibbs 1909, Par- of the west and the forests of the east lies anham 1972, Smith 1951, Sykes 1933, Twyford intermediate rainfall zone with mixed cover.and Wright 1965). It would appear, however, The leeward hill slopes carry grasses as wellthat prior to the arrival of the Europeans and as reed or bamboo, Bambusa spp. or Schizo­the development of large-scale cash crops stachyon glaucifolium, while the windwardsuch as sugar cane and copra, much of these slopes are lightly forested. Several intro­islands was forested (Twyford and Wright duced, invasive, scrubby weeds, such as Piper1965). The major plant communities that aduncum, Psidium guajava, and Leucaenastill exist in quite large areas are the lowland leucocephala, are abundant in this zone whileand montane rain forests, coastal communi- the aboriginally introduced Makosoi, Canagaties and mangroves, inland swamps, mixed odorata, is a common tree of forest associa­grassland and light forest of the intermediate tions.zone, and grasslands of the- western Sioes:---------FigurETshows the present distribution of

The lowland forest is dominated by hard- habitat types on the main islands of Fiji. Itwoods such as Callophyllum vitiense and can be seen that the bulk of modern agricul­Endospermum macrophyllum and occupies tural activity is concentrated in the coastalthe wetter eastern regions from sea level to and dry zones. To suggest that over the last600 meters. On the coast are found communi- 150 years European colonizers have beenties of various species of Pandanus, Barring- entirely responsible for the destruction oftonia, and Hibiscus, together with Cocos original lowland forest cover and the presentnuc(lera and Ipomea braziliensis, which are extent of the grasslands is probably incorrect.

~GAU

DKORO

itO

0 ~~...

.()

t1,..

C-

O~

G'

l'

<) 0u

p

Q...

o

..'

~17

<J,'. 0

.? =~ 0 .

"

[) I ~ P

\J

\WAKAYA

'Q

'0

~

", .~

• t)lBEQA

KADAVU

~

~

J(/ ~

~. ~

(l0~.~

tJ' oSy

r y

'"<r.>..

"

FIGURE I, Map of the main islands of the Fiji group named in the text. Scale 1 cm = 45 km, The distribution ofmajor habitats is indicated as follows: • = Montane forest; !Wi = Lowland forest; ::: = Intermediate zonevegetation; 1111 = Agricultural land ; • = Mangrove swamps; 0 = Grassland.

226 PACIFIC SCIENCE, Volume 32, July 1978

TABLE I

HABITAT DISTRIBUTION OF FIJIAN TERRESTRIAL VERTEBRATES

FOREST

INTER- COASTAL/ FRESHWATER NUMBER OF TOTALLOW- MEDIATE AGRI- SUB- MANGROVE INLAND SPECIES NUMBER OF

MONTANE LAND ZONE CULTURAL URBAN SWAMPS SWAMPS OMITTED SPECIES

Amphibia 2 2 2* 2Reptilia 9 9 9 61 61 8 3 5 2JlAves 38 46 31 23 15 17 II 6 59Mammalia 3 4 3 I I 6

Total 52 61 43 30 21 26 16 II 88native

species

Introduced 5 5 10 16 13 5 10 18species, all unsuccessfulclasses introductions

NOTE: A number ofspedes that may be extinct or for which data are not available have been omitted (see Appendix).*In forest only.'Three species in association with buildings only.~ Data for the two nominal Ellgyrus spp. are combined, since the authors were unable to distinguish them (Burt and Burt 1932).

Although pre-European agriculture appearsto have been of a shifting, clearance type(still extensively practiced in many areas), itseffect on natural vegetation may have beengreat. Berry and Howard (1973) identifiedlarge areas of mature secondary forest withinthe lowland rainforest habitat of the interiorof the larger islands. This, combined with theobservation Of old, abandoned village sites;village names on early maps and in writtenand verbal records where no villages arefound today; and, in places, extensive hillterracing which is no longer practiced; sug­gests the occurrence of widespread forestclearance in pre-European times. This, inturn, indicates the presence of a far larger,more industrious population in the inlandparts of the major islands than has previouslybeen suspected. This suggests that secondaryhabitats may have been present for a longperiod in prehistory and thus have been animportant part of the environment of thenative animals. Secondary habitats may alsohave been significant on small islands sub­jected to not infrequent hurricane damage.

A possible sequence of events in habitatterms may be outlined as follows: initially,prior to human arrival, the islands were prob­ably forested from shore to mountain top,with some relatively small areas of native

grassland occurring in restricted areas of thedry zone. As the aboriginal populationincreased, patchy clearance of forests aroundthe coast and inland along river valleyscreated a mozaic of secondary plant growthfollowing agricultural clearance. Clearanceof forest in the intermediate and dry zonesby the use of fire created more extensiveareas of grassland. European forestry andagricultural activities further increased thearea of such secondary habitats; changed thefloral composition through the introductionof exotic weeds and the selective eliminationof some plant species such as sandalwood,Santa/urn yasi; and greatly increased the landarea under permanent cultivation.

HABITAT UTILIZATION BY TERRESTRIAL

VERTEBRATES

Table 1 and the Appendix present data onthe habitat distribution of native vertebrates.It is at once apparent that the bulk of thenative species of birds (87 percent, n = 53),amphibia (100 percent, n = 2), reptiles (59percent, n = 17), and mammals (66 percent,n = 6) are found predominantly in foresthabitats.

In the case of the reptiles, the skinks found

''''I )j !Qlfl t h t ItriP ::amm::&:¥ !Pi d; ;TimMS 1M i; Ii

Terrestrial Vertebrates of Fiji-PERNETTA AND WATLING 227

in agricultural and suburban habitats wereground-dwelling open habitat types such asEmoia cyanura, or species confined to coastalcommunities such as Lipinia noctua andEmoia nigra. Of the geckos, three of the fivespecies collected during the present studywere found only in association with humans,and their original, natural habitats are un­known. Of the 17 reptile species that are wellknown from the group, seven are confined toprimary forest or forest associations of theintermediate zone. The two native amphibiaare similarly confined to primary forest, andthe native mammals (all bats) feed mainly inforest associations. Although at least four ofthe six bats may occasionally be seen in agri­cultural or suburban situations, these are nottheir primary habitats but are utilized merelyas feeding areas on an intermittent basis.

The bulk of the native species of birds arefound in forest situations. Only 23 species(43 percent) may be observed in agriculturalareas and 15 (28 percent) in suburban habitats(Table 1). Eighteen of the 59 native land birdsare confined to forest habitats or the inter­mediate zone, and all those species observedin agricultural and suburban situations alsooccur in forest habitats (see Appendix).Exceptions are the swamp and coastal rails(four species), the reef heron, Demigrettasacra, the little mangrove heron, Butoridesstriatus, and the Pacific swallow, Hirundotahitica.

It is also apparent from Table 1 and theAppendix that in a majority of cases, nativeforest vertebrates may be found in both mon­tane and lowland wooded habitats, a pointnoted by Gorman (l975b) in the case of birds.This suggests that there is a lack of the habitatand ecological specialization among Fijianvertebrates that is characteristic of mainlandvertebrate faunas (Lack 1971).

INTRODUCED VERTEBRATES

Table 2 presents details of the successfulintroductions of vertebrates to Fiji. Theseanimals fall into two categories-the ab­original introductions and the post-Europeanintroductions, both deliberate and accidental.

It has long been accepted that the Melane­sian and Polynesian peoples transportedplant and animal species that were economic­ally or culturally important from island toisland. In the case of the terrestrial verte­brates, the jungle fowl, Gallus gallus, domesticpig, Sus serofa, and Polynesian rat, Rattusexulans, probably all represent introductionsof important protein sources (Bahr 1912,Gressitt 1956, Tate 1935). Since both thePacific boas, Engyrus (= Candoia) spp., andthe Fijian ground frog, Platymantis vitianus,were extensively consumed in Fiji prior tothe arrival of the Europeans (and are stilleaten in some areas), they may well representadditional, deliberate, prehistoric introduc­tions of potential food sources.

Except for such intentional introductions,the role ofaboriginal peoples in the accidentaltransport of reptiles and amphibians hasreceived little attention to date. In view ofthe extensive voyages undertaken in thePacific by such people, and the large foodand planting stocks carried by them, it isnot inconceivable that eggs or even adults ofmany small species may have been carriedfrom Papua New Guinea and the Solomonsto Fiji. A number of the geckos and skinkspresently found in Fiji lay their eggs in cracksin tree bark and might equally well lay eggsbetween the planks or in the cargo of voyag­ing canoes drawn up the beach for loading.In addition, eggs might be accidentally in­cluded in planting stocks (Pernetta andGoldman 1977) or leaf mold used to packplant roots for transport. While such acciden­tal transport of amphibians (Darlington1957) and especially small reptiles is probable,the accidental transport of birds is highlyunlikely.

In contrast to aboriginal introductions, thespecies of vertebrates introduced followingEuropean colonization are more numerousand varied. They include deliberate in­troductions as early attempts at biologicalcontrol. The cane or marine toad, Bufomarinus, was introduced from Hawaii in1936 to control various coleopterous pests ofsugar cane, banana, and other cash crops(Hinkley 1962). The small Indian mongoose,Herpestes auropunctatus, was introduced in

TABLE 2

INTRODUCED TERRESTRIAL VERTEBRATES CURRENTLY ESTABLISHED IN FIJI

DATE OFSPECIES INTRODUCTION DISTRIBUTION STATUS HABITAT DIET

Bufo marinus 1936 Main islands Common Agricultural, urban, suburban Millipedes, Orthoptera,marine or cane toad Hymenoptera

Gallus gallus Aboriginal Formerly all islands (?) Rare Forest, inland swamps Insects, seeds, fruitjungle fowl

Synoicus australis c. 1900 Viti Levu, Vanua Levu Rare Agricultural, grass Seedsbrown quail

Acridotheres tristis c. 1890 Main islands Common Agricultural, suburban, urban Insects, fruit, refuseIndian mynah

Acridotheres fuscus c. 1890 Main islands, except Taveuni Common Agricultural, suburban, urban, Insects, fruitjungle mynah scrub

Pycnonotus cafer c.1900 Viti Levu, Ovalau, Wakaya Common All habitats Fruit, insectsred-vented bulbul Beqa, Taveuni Present, but rare

Streptopelia chinensis c. 1900 Main islands Common Agricultural, urban, suburban Seeds, grainMalay turtle dove

Amandava amandava c.1900 Main islands Common Agricultural, urban, suburban, Seedsstrawberry finch grass

Padda oryzivora c. 1930 Viti Levu, Vanua Levu, Taveuni Locally abundant Agricultural, urban, suburban, Seeds, grainJava rice sparrow wet zone only

Gymnorhyna tibicen c.1900 Taveuni Locally abundant Coconut plantations Seeds, insects, fruitblack-headed magpie

Sturnus vulgaris c.1930 Ono-i-Lau, Votua Locally abundant Agricultural, villages Fruit, insects, seedsEuropean starling

Columba livia c.1900 Towns only Locally common Urban, suburban, agricultural Grain, food refusedomestic pigeon (rare)

Sus scrofa Aboriginal Main islands Common Forest, intermediate zone, Fruit, rootsdomestic pig grass

Rattus exulans Aboriginal All islands Abundant All habitats Vegetable seeds, roots,Polynesian rat insects

Rattusnorvegicus Nineteenth century Main islands Common Agricultural, urban, suburban, Vegetables, grain,brown or Norway rat coastal refuse

Rattus rattus Nineteenth century Main islands Locally abundant Agricultural, plantations, Fruit, vegetablesblack rat urban, suburban, coastal

Mus musculus Nineteenth century Viti Levu, Vanua Levu Locally abundant Urban, suburban, agricultural Vegetable seeds, fruit,house mouse refuse

Herpestes auropunctatus 1883 Viti Levu, Vanua Levu Common All habitats Invertebrates,small Indian mongoose amphibia, aves,

reptiles, rodents

--_._---

Terrestrial Vertebrates of Fiji-PERNETTA AND WATLING 229

1883 to control rats in sugar cane (Gorman1975a). The black-headed magpie, Gymnor­hyna tibicen, was introduced from Australiaaround 1900 to control phasmid pests ofcoconut (A. Tart, personal communication).The Indian and jungle mynahs, Acridotherestristis and Acridotheres fuscus, were intro­duced from India around 1890 to controlorthopterous pests of sugar cane (Veitch1923). In addition, deliberate introductionsof other species were made, some of whichwere successful-for example, the brownquail, Synoicus australis, and Sambar deer,Cervus unicolor (introduced to Wakaya Islandonly)-and some of which were not-suchas the turkey, Meleagris gallopavo, pheasant,Phasianus sp., an unidentified grouse species,laughing kookaburra, Dacelo gigas, tawnyfrogmouth, Podargus strigoides, and part­ridge, Perdix sp. (Wood and Wetmore 1925).The bulk of the remaining bird introductionsappear to have been made as cage birds orfor esthetic reasons. Unsuccessful introduc­tions of this kind include the blackbird,Turdus merula, spotted-sided finch, Stegono­pleura guttata, Australian crow-shrike, Stre­pera melanoptera, magpie lark, Grallina(cyanoleuca ?), and native companion, An­tigone rubicunda (Wood and Wetmore 1926).Of the eleven successful bird introductions,the jungle fowl, both mynahs, Malay turtledove, and strawberry finch are at presentfound on all main islands; the red-ventedbulbul is common only on Viti Levu but isalso found on Ovalau, Wakaya, Beqa, andTaveuni; the black-headed magpie is foundonly on Taveuni; the European starling isfound only on Ono-i-Lau and Votua; thedomestic pigeon is restricted to large towns;and the Java rice sparrow is found only on thewet sides of Viti Levu, Vanua Levu, andTaveuni. The brown quail is rare andrestricted to the grasslands of western VitiLevu and Vanua Levu. Mercer (1964) sug­gests that the European starling may repre­sent a ship-assisted natural colonization fromAustralia or New Zealand, via the KermadecIslands, because no record exists of itsdeliberate introduction and it appears to bea relatively recent addition to the avifauna[it was not mentioned by Wood and Wetmore

(1925, 1926)]. A similar explanation may beadvanced for the occurrence of the domesticpigeon.

Accidental introductions by Europeansinclude the three rodent species as commen­sals of humans (Rattus norvegicus, Rattusrattus, and Mus musculus), all of which havesucceeded in establishing feral populationsthat are closely associated with buildings andagricultural land. Domestic animals suchas horses and cattle are also widespreadthroughout the islands and may roam in arelatively unrestricted manner over the un­fenced grasslands of the dry and intermediatezones. In a number of localities in the inter­mediate zone of Viti Levu and Kadavu, feralgoats are found in small numbers. Theseanimals are also found on various islands,such as Goat Island in the Yasawas, andNamara Island in the Kadavu group. Al­though European introductions of dogs andcats have resulted in their widespread dis­tribution throughout the islands, no feralpopulations of these animals are known atpresent. European breeds of domestic pigsand sheep are maintained under agriculturalconditions, as are large numbers of fowl,ducks, geese, and turkeys.

STATUS OF THE NATIVE SPECIES

In assessing the present status of the nativevertebrates, one is faced with the problem ofa lack of good biological data on the ecologyand numbers ofsuch species. However, usefulhistorical information on the birds may befound in Bahr (1912), Finsch and Hartlaub(1867), Layard (1875, 1876), and Wood andWetmore (1925, 1926). Similar informationmay be obtained for the reptiles and amphibiafrbm Burt and Burt (1932), Mertens (1934),Schmidt (1923), and, more recently, Brown(1956), Gorham (1965, 1968), and Watson(1960). .

Although the literature on birds does notprovide sufficient data to form definitiveconclusions concerning changes in the statusof all Fijian native species, it is possible tostate conclusively that the banded rail, Rallusphillipensis, sooty rail, Porzana tabuensis,

230

white-browed rail, Poliolimnas cinereus, andpurple swamphen, Porphyrio porphyrio, wereall common on Viti Levu in the last century.These now survive in any numbers only onislands that are free of the mongoose.Similarly, the Australian gray duck, Anassuperciliosa, was once recorded as abundantby Brewster (1935) and Seeman (1862), butit is now relatively uncommon on Viti Levu.

The nominate subspecies of the rare,native, long-legged warbler, Trichocichla mfamfa is known from only four specimens, thelast of which was collected in 1894. However,a different subspecies, T. rufa clunei, hasrecently been discovered on Vanua Levu(Kinsky 1975). The pink-billed parrot finch,Erythrura kleinschmidti, and the red-throatedlorikeet, Charmosyna amabilis, have ap­parently always been rare (Mayr 1945,Mercer 1964), although E. kleinschmidti hasrecently been found attempting to nest insecondary forest on Viti Levu (Clunie 1973).The barred-wing rail, Nesoclopeus poecilop­terus, is now possibly extinct, and thewhistling tree duck, Dendrocygna arcuata, isalmost certainly extinct. These may alwayshave been rare (Mayr 1945, Mercer 1964),although Layard (1875) terms the formerspecies as "generally distributed." A. vonHugel notes on the label of a specimen hecollected on Ovalau in 1877 (now in theCambridge University Zoological Museum)that it was "difficult to procure," but whetherhe was alluding to the elusive habits of thisrail or to its rarity is not clear. The grass owl,Tyto longimembris, is known from only twospecimens collected on Viti Levu in thel860s and has not been seen since and maybe extinct (Clunie 1972b).

The smaller reptiles are, on the whole,common and widely distributed throughoutthe group regardless of the presence of themongoose, although the patchy distributionof many species makes comparison betweenislands with and without the mongoosedifficult. It is interesting to note that duringthe present work an undescribed species ofEmoia (Brown and Pernetta in prep.) wasencountered on Viti Levu, Taveuni, andKadavu. Also, two species of gecko, Hemi­dactylus garnotti and Hemiphyllodactylus

PACIFIC SCIENCE, Volume 32, July 1978

typus, previously unrecorded from the group,were collected in several localities. Thiswould seem to indicate either that thesespecies are recent introductions or thatearlier work on the reptiles of these islandshas been far from systematic. Both geckospecies were encountered only in close as­sociation with buildings. These species arewidely distributed throughout the Pacific(Brown 1956), and neither were recorded byWatson (1960); thus, the possibility of theirrecent introduction from other Pacific islandsvia interisland copra boats or other vesselsis not remote. In the case of the Emoia sp.novum, this possibility is less likely becausethe species was encountered only in inlandforest habitats on the three islands.

Apparently, a number of the larger reptilespecies have, like many of the birds, alwaysbeen rare. The elapid snake, Ogmodon viti­anus, has been only occasionally collected onViti Levu. During the present investigation,a single headless specimen (possibly a mon­goose kill) was collected in the intermediatezone. The agamid lizard, Goniocephalus god­freyi, is known from only two specimenscollected in the late nineteenth century(Boulenger 1887). W. C. Brown (personalcommunication) seriously doubts the exis­tence of a natural population of this speciesin Fiji and suggests that the two knownspecimens were not collected in Fiji. The twosemiarboreal boas are both difficult to find,rather than really rare, and may still beregularly encountered in the intermediatezone of Viti Levu. However, they are possiblyless common on this island than on Ovalauand Taveuni. Indirect confirmation of thecontinued existence of snakes on Viti Levucomes from Gorman's (l975a) analysis ofmongoose feces. Four percent of thosecollected from lowland forest contained snakeremains. The banded iguana, Brachylophusfasciatus (the only Fijian terrestrial vertebratewith neotropical affinities), has apparentlydeclined in numbers over the last century onboth Viti Levu and Taveuni (Cahill 1970,Cogger 1974, Gorman 1975a) but is stillcommon on Kadavu and various smallerislands (Cogger 1974).

According to Gorham (1965, 1968), the

Ubft6*dR 'GflffIih JIiI ' . Sf#fM9ii! " (.Ii Ii I"• "'M aM

Terrestrial Vertebrates of Fiji-PERNETTA AND WATLING

TABLE 3

231

INTERSPECIFIC AGGRESSION BETWEEN Pycnonotus cafeI' AND OTHER BIRDS IN THE SIGATOKA VALLEY, VITI LEVU, FIJI

BULBUL

AGGRESSOR

77853

2

Totals 34

SPECIES

Lalage maculosa, Polynesian trillerAcridotheres tristis, jungle mynahAcridotheresfuscus, Indian mynahFoulehaio carunculata, Wattled honeyeaterZosterops lateralis, gray-backed white-eyeArtamus leucorhynchus, white-breasted wood swallowAmandava amandava, strawberry finchHalcyon chloris, white-collared kingfisherStreptopelia chinensis, Malay turtle doveMyiagra vanikorensis Vanikoro broadbill

BULBUL

RECIPIENT

I4

12

19

NOTE: Data from 411 hr of observation. February 1974-January 1976.

two Fijian frogs are now both less commonin localities where early collectors recordedthem as abundant. Gorman (1975a) statesthat their decline on Viti Levu is attributableto the mongoose. However, the field inves­tigations of one of the present authors(J. C. P.) suggests that the tree frog is capableof maintaining high densities in areas wherethe mongoose is also abundant. Although itis probable that the arboreal behavior of thisspecies reduces its susceptibility to groundpredators in comparison with the moreterrestrial ground frog, both species may befound on Viti Levu (Gorham 1968, Pernettaand Goldman 1977) and there is no evidenceto suggest that they are significantly lesscommon in the presence of the mongoosethan in its absence. Gorham (1968) suggeststhat habitat destruction may be responsiblefor their decline in certain localities, parti­cularly on the island of Ovalau.

Of the six bat species, four are abundantthroughout the group, while Tadaridajobensisis known only from Taveuni, where it maybe seen hawking insects among the crowns ofcoconut palms at the southern end of theisland. The recently discovered Ptera/opexsp. is known from a single specimen collectedin montane forest on the same island (W.Beckon, personal communication).

In addition to the above, two furtherliterature records should be mentioned: themarine crocodile, Crocody/us porosus, and

the long-necked water tortoise (speciesunknown). In the latter case [listed byLoveridge (1945)], the record is unsub­stantiated by either specimens or detailedobservations, and thus may be taken as anerror. In the former case, however, Derrick(1965) gives a reliable account of the killing ofa single crocodile that may well represent avagrant individual of C. porosus from theSolomons. Despite the presence of largeareas of suitable habitat in the form ofmangrove swamp and river deltas, a breedingpopulation of this species does not appearto be present in the islands.

INTERACTION BETWEEN INTRODUCED AND

NATIVE VERTEBRATES

During the course of extensive work onthe ecology of the red-vented bulbul in Fiji(Watling 1977), records were kept of directinteraction between this and other species ofbirds. The numbers of encounters observedover 411 hr of transect counts are summarizedin Table 3. In 34 (64 percent) of the en­counters, the bulbul initiated attacks againsteight species, four of which were nativespecies. The wattled honeyeater, Fou/ehaiocaruncu/ata, and the Polynesian triller, La/agemacu/osa, were the most frequently attackednative species, but both were challenged lessfrequently than the two introduced mynah

232 PACIFIC SCIENCE, Volume 32, July 1978

TABLE 4

DIET, SIZE, FEEDING STATION, AND HABITAT OF POTENTIAL AVIAN COMPETITORS ON VITI LEVU, FIJI

SIZE NATIVE (n) OR(cm) INTRODUCED (i) FEEDING STATION HABITAT

GranivoresErythrura pealii 10 n Ground zone/understory Agricultural, grass, intermediate,

(partly insectivorous) forestAmandava amandava 10 Ground zone Urban, suburban, agricultural,

grass, intermediatePadda oryzivora 14 Ground zone Suburban, agricultural, wet zoneSynoicus australis 16 Ground zone Agricultural, grass, intermediateStreptopelia chinensis 30 Ground zone Urban, suburban, agricultural,

grassGallicolumba stairii 37 n Ground zone Forest

FrugivoresAplonis tabuensis 17 n Canopy/understory ForestPycnonotus cafer 18 i Scrub/ground zone Urban, suburban, agricultural,

grass, forest edgesPtilinopus perousii 20 n Canopy ForestPtilinopus luteovirens 20 n Canopy/understory Forest, intermediateColumba vitiensis 37 n Canopy/scrub Forest, intermediate,

agricultural, grassDucula latrans 40 n Canopy ForestProsopeia personata 48 n Canopy/understory Forest, intermediate

Flower/nectar feedersMyzomela jugularis 10 n Scrub/understory/canopy Forest, intermediate,

agricultural, suburban, urbanFoulehaio carunculata 16 n Scrub/canopy Forest, intermediate,

agricultural, suburban, urbanCharmosyna amabilis 16 n Canopy ForestPhigys solitarius 20 n Canopy Forest, intermediate,

agricultural, suburbanGymnomyza viridis 26 n Canopy Forest

OmnivoresAcridotheres fuscus 21 Ground Suburban, agricultural,

intermediateAcridotheres tristis 23 Ground Urban, suburban, agriculturalColumha Iivia 35 Ground Urban, suburbanGallus gal/us 55 Ground Forest, intermediate

InsectivoresPetroica multicolor 10 n Aerial, understory ForestZosterops explorator II n Scrub/understory/canopy Grass, intermediate, forestZosterops lateralis II n Scrub/understory/canopy Grass, intermediate, forestCol/ocalia spodiopygia 12 n Aerial All habitatsVitia ruficapilla 13 n Scrub/understory Forest, intermediateMyiagra vanikorensis 13 n Aerial, all All habitatsMyiagra azureocapilla 13 n Aerial, canopy/understory ForestMayrornis lessoni 13 n Aerial, canopy/understory Forest, intermediateHirundo tahitica 13 n Aerial UrbanRhipidura spilodera 13 n Aerial, understory Forest, intermediateArtamus leucorhynchus 16 n Aerial/ground All habitatsLalage maculosa 17 n Ground/understory All habitatsClytorhynchus vitiensis 18 n Understory Forest, intermediateCl)'torh)'nchus nigrogularis 18 n Canopy/understory ForestTurdus poliocephalus 20 n Ground zone/scrub Forest, intermediateCacomantis pyrrllOphanus 24 n Understory/scrub Forest, intermediate

NOTE: Data from personal observation (D. W.) and Mercer (1964). See Table 2 and Appendix also.

Terrestrial Vertebrates of Fiji-PERNETTA AND WATLING 233

species. The wattled honeyeater initiatedaggressive encounters more frequently thanit received them, and this species accountedfor 12 of 19 attacks directed toward thebulbul. A monthly breakdown of bulbulaggressive activity showed that most inter­specific aggression occurred during the breed­ing season. Of the total of 53 encounters, 8involved aggression by the bulbuls aroundtheir nests, 7 involved food, and 38 were ofunknown causes. This, combined with similarevidence for the two introduced mynahspecies (Watling 1975), suggests that directaggressive activity between introduced andnative species does not occur often enough toexplain entirely the habitat separation of thetwo groups.

Food is perhaps the most important itemof potential competition between introducedand native species. The bulk of the introducedbirds are medium-sized omnivores or grani­vores. The native bird fauna in the same sizerange is composed of small specializedinsectivores and a single granivore that alsofeeds a great deal on insects (Table 4). Therare finch Erythrura kleinschmidti appears tobe primarily an insectivore (Clunie 1973).In the case of the native frugivores, however,the red-vented bulbul constitutes a potentialcompetitor, except that the feeding strategiesand stations of the species are different. Thebulbul, which might be termed an "oppor­tunist frugivore" (McKey 1973), feeds pri­marily on the fruits of introduced, invasiveweeds found in secondary seral associations(Solanum torvum and Piper aduncum) andcultivated land (Physalis angulata) (Watling1977). Almost all native frugivores feedmainly in forest associations on native andaboriginally introduced fruits such as themakosoi, Canaga odorata, kauvula, Endo­spermum macrophyllum, and koka, Bischoffiajavanica. The exception, the white-throatedpigeon, Columba vitiensis, feeds extensivelyon the fruits of S. torvum and is frequentlyseen in agricultural areas.

Bulbuls may be readily observed in openforest situations in the intermediate zone,feeding in association with native species andwithout direct aggressive interactions. Suchsightings, however, are usually made in or

around areas of secondary vegetation suchas river banks, old land slips, or forestclearings.

In the case of the marine toad, no direct orindirect ecological interaction is apparentwith the native amphibia, as they are es­sentially separated on a habitat basis. Thetoad is a ground feeder and takes largenumbers of Diplopoda, Orthoptera, andHymenoptera (Hinkley 1962) in open situa­tions. The two native frogs are found eitheralong stream banks or in foliage in denseforest, and apparently feed on night-flyinginsects (Gorham 1968, Pernetta and Goldman1977). Although large adult Bufo marinus mayoccasionally be encountered in the lowlandforests and secondary forest areas, largeviable populations of this species are onlymaintained in the flatter coastal areas. Thismay be because the species has a free­swimming tadpole stage that would be sweptaway during the flash floods to which themontane streams are subjected. The nativefrogs overcome this difficulty by laying eggsin Pandanus leafaxils and rotting logs, andmetamorphosis occurs within the egg, result­ing in the emergence of a fully formed frog(Gorham 1968, Pernetta and Goldman 1977).

The more recently introduced rats andmice have no apparent effect on the dis­tribution of the prehistorically introducedRattus exulans, as all three species may befound in the same habitats. There is someindication that the Norway rat is moreabundant in suburban situations than theother two species (Table 5); that the blackrat may preferentially inhabit plantationsand coconut palm crowns (Williams 1971,1973); and that these species are restrictedin their distribution to a close associationwith humans (Table 5).

DISCUSSION

In discussing the effect of vertebrateintroductions on the native fauna of Fiji oneshould first make the point that all introducedspecies, with the exception of the smallIndian mongoose and to a lesser extent thered-vented bulbul, are closely associated with

234 PACIFIC SCIENCE, Volume 32, July 1978

TABLE 5

PERCENTAGE COMPOSITION OF RODENT SAMPLES FROM DIFFERENT HABITATS, FIJI

INTERMEDIATE BARN OWL

ZONE PELLET

COASTAL VEGETATION, NEAR REMAINS- NAMARA,

COCONUT NATIVE GARDENS INTERMEDIATE SUBURBAN UNINHABITED

PLANTATIONS, AND HOUSES, ZONE, VITI SUVA, ROCKY ISLET,

TAVEUNI VITI LEVU LEVU VITI LEVU KADA VU GROUP

Mus museu/us 14 10 36Rattus exu/ans 80 77 82 8 100R. rattus I 18R. norvegicus 20 8 8 38

Total numberof animals 20 113 61 39 21

human-modified habitats. In contrast, thebulk of the native species are confined to, ormore frequently encountered in, the foresthabitats. Of the 32 species of nonpredatoryland birds regularly seen on Viti Levu, onlyone-the Pacific swallow-was not observedin lowland or montane forest. However, only16 of these 32 native species are regularly seenin agricultural land and only 13 in suburbanand urban situations. While it may be reason­ably assumed that the Pacific swallow shouldhave increased in numbers since urbanization(following an increase in potential nest sitesand hunting areas), no concrete evidenceexists support this.

Several views may be advanced to explainthe habitat isolation of native and introducedbird species. First, the native species areadapted to forest habitats and are incapableof modifying their behavior and ecology toinvade the secondary human-modified habi­tats. Second, the native species are forced outof such habitats by competition for food andnesting sites with introduced species. Third,direct predation by the mongoose has reducedthe numbers of native species in habitatswhere the mongoose is abundant. Fourth,direct destruction by people or effects ofhuman activities (such as the destructionof nest sites or areas of former habitat, orthe use of insecticides) have caused a declinein the native species in human-modifiedhabitats.

The use of insecticides to control insect

pests of various cash crops has been notedas the cause for the decline of the bandediguana on Taveuni (Cahill 1970). In view ofthe fact that this species is a specializedflower feeder (most commonly encounteredin forest) and that the Fiji AgricultureDepartment strictly controls the types ofinsecticide imported and used in the country,such an explanation is unlikely. Similarly,since many native and introduced birdspecies appear to thrive in agricultural situa­tions, selective elimination or reduction innumbers of individuals of some native speciesdue to insecticide poisoning seems unlikely.

The mongoose has been blamed for thedecline of the banded iguana, native frogs,and various bird species on Viti Levu.Although some circumstantial evidence doesexist for the coincident decline of variousground-dwelling bird species on Viti Levuwith the spread of this predator, the work ofGorman (l975a) has shown that much of thepresent diet of this species is composed ofinvertebrates and introduced vertebrates suchas rats and the cane toad. Historical evidence,in combination with evidence from islandswhere these native species are presently com­mon but where the mongoose is absent, sug­gests that the mongoose is indeed implicatedin their decline. In contrast, a number of rarebird species that are now possibly extinct(see Appendix) were apparently rare priorto the introduction of the mongoose. Inconsidering the impact of this predator on

'@ilF/' BP nmett <MOt ill jik M $ ; 2 $;;_ A$ g; d· HiM

Terrestrial Vertebrates of Fiji-PERNETTA AND WATLING 235

the reptile fauna and native frogs, it seemsthat its effect may have been overemphasized.In the case of the smaller lizards and the treefrog, observations made during the presentstudy indicate that these animals may occurnormally in a rather patchy distribution andare able to maintain high population numbersin areas where the mongoose is also abundant.Gorman (1975a) records that in some areasof Viti Levu, the ground frog and snakes arestill eaten by the mongoose, indicating thatthey are still sufficiently abundant to con­tribute part of the diet of this species. Themongoose has been present on Viti Levu fornearly 100 years and its spread and numericalincrease were rapid, suggesting that its preyhave been subjected to the present level ofpredation for a considerable time. It wouldappear that the snakes and frogs are ableto maintain viable populations under suchpredation pressure.

In regard to the effect of competitionbetween introduced and native species, thepaucity of observations of direct interactionbetween these two groups, combined with thefact that some native vertebrates successfullymaintain populations in human-modifiedenvironments, suggests that the section ofthe native fauna capable of utilizing suchhabitats has done so. Whether competitionwith introduced species has resulted inlower population densities of these nativespecies in human-modified habitats thanmight be possible in their absence is unknown.One must conclude that interspecific com­petition is relatively unimportant in themajority of cases, because the present situa­tion has existed for about 80 years andbecause none of the rare native species com­pete in any way with the introduced species.

When one considers the native speciesthat are now confined to forest habitats, onesees that the possibility of competitive ex­clusion of such species from other habitatsexists. Examination of the Appendix showsthat species found only in forest environ­ments are the giant forest honeyeater andpink-billed parrot finch, while the generallyforest-inhabiting blue-crested broadbill andblack-faced shrikebill were only occasionallyencountered in the intermediate zone vegeta-

tion. An additional six species of birds thatare regularly encountered in the intermediatezone vegetation fail to range into agriculturalland, although potential competition withintroduced birds provides an explanation ofthis restricted distribution for only one nativespeCies.

In Fiji, the Polynesian starling, Aplonistabuensis, is restricted to forest areas. It isnot frequently seen in the intermediate zone,and never in agricultural or urban situations;this observation is supported by Gorman(1972, 1975b). However, on Tongatapu,Tonga, the same species is a common bird allover the island, which is denuded of naturalvegetation and intensively farmed. Thisstarling was also regularly seen in urbanNuku'alofa (Watling, personal observation).Although the frugivorous bulbul is a likelypotential competitor, the habitat range ofA. tabuensis on Vanua Levu, where thebulbul is absent, is no different from thaton Viti Levu. In addition, the bulbul is alsopresent on Tongatapu, Tonga, and on Vpoluand Savai'i in Samoa, where it frequentlyoccurs in the same habitat as A. tabuensis.The jungle mynah, Acridotheres juscus, is notpresent in Tonga and is only a recent arrivalin Samoa [Dhondt (1976) incorrectly iden­tifies it as being Acridotheres tristis (Watlingin press)], so the present restricted range ofA. tabuensis in Fiji may be due to thecombined competitive pressure from bothmynah species. Similarly, the present restrict­ed range of the European starling in Fiji maybe due to competitive exclusion from otherareas by these two species. On Tongatapu,in the absence of the mynahs, the Europeanstarling does well.

If, as is suggested above, the two gecko... species discovered during the present study

do represent recent additions to the Fijifauna, then evidence for competitive interac­tions between them and the species Lepido­dactylus lugubris and Gehyra oceanica isnotable by its absence, since all four speciesmay be encountered on the walls of the samebuildings. It is probable that the markeddifference in size between species reducespotential competition for the live insect foodupon which all four species prey.

236

The occurrence of six, or possibly eight,species of the skink genus Emoia is of interest.From the data available, it appears thathabitat isolation is important in the main­tenance of these species on the same islands.Both Emoia concolor and E. samoensis arearboreal; the former is confined to coastalareas and the open woodlands of the inter­mediate zone, while the latter is abundant inthe lowlands and montane forests. W. C.Brown (personal communication) considersthat these two nominal species are in factconspecific, as they are inseparable on thebasis of squamation. If this is the case and thetwo animals represent two color morphs ofE. samoensis, then it is of interest to notethe apparent habitat separation betweenthem. Emoia nigra, a large ground-dwellingspecies, is again confined to coastal areasand is abundant beneath coconut palms,whereas the smaller E. cyanura is found inopen situations from the coast to forestglades. The undescribed Emoia sp. novum issimilar in size to E. cyanura, but is apparentlyconfined to inland forests; occasionally, bothspecies may be found together. Emoia caeru­leocauda is smaller in size than the precedingtwo species and like them is found on theforest floor. Although both E. adspersa andE. cyanogaster are recorded from Fiji (Stern­field 1920, Werner 1899), they were notencountered during the present study andtheir ecological interactions with the otherspecies are unknown.

In examining the native fauna of Fiji it isnoticeable that, with the exception of thereptiles, the number of sympatric congenersis low. One congeneric pair of bats and sixpairs of birds overlap in habitat and alti­tudinal range on Viti Levu: Pteropus, Ptili­nopus, Clytorhynchus, Myiagra, Zosterops,Erythrura, and, doubtfully, Tyto. The twoTyto species occur sympatrically from Africato Australasia, and thus competition is anunlikely explanation for the apparent demiseof the grass owl in Fiji. Although the nativecongeners overlap in habitat on Viti Levu,only the Clytorhynchus pair do not haveclearly separable optimum habitats. In theother cases, one species is restricted to foresthabitats and the other is a more generalized

PACIFIC SCIENCE, Volume 32, July 1978

species with wide habitat tolerance. In thecase of the two fruit bats, Pteropus tonganusis strongly colonial and feeds in forest areasand coconut plantations, whereas Pteropussamoensis is solitary, feeding in more openhabitats.

With the exception of the Mayornis pairthat occur together on Ogea Levu only, allother congeneric species are allopatricallydistributed within the Fiji Island group(Appendix). This overall low occurrence ofsympatric congeneric pairs is in accordancewith the findings of Lack (1969, 1975), whosuggests that due to the absence of suchpairs, birds on islands generally show relaxedecological isolation with consequent expan­sion of individual species' habitat utilization.This is in contrast to mainland situationswhere habitat and ecological isolation areimportant features of bird faunas. It ispossible that in the case of the birds, at least,these congeneric Fijian pairs represent specia­tion through double invasion, with the earlierarriving and speciating form being restrictedto the primary forest habitats for which theyhad become better adapted.

The importance of people in the declineof some of the native species should not beunderestimated. Both the barred-winged railand jungle fowl were once hunted with dogs(Bahr 1912, Mercer 1964) and, in inland areasof Viti Levu, the banded iguana was eatenby the hill tribes (Cahill 1970). In addition,as previously mentioned, boas and nativefrogs are still eaten in some areas and weremore extensively eaten in the past. Perhapsof greater impact on the native fauna is theextensive deforestation that has accompaniedthe expansion of modern agriculture. Suchhabitat destruction may have an importanteffect on the ecological communities towhich many of the native species are adapted.It should be noted, however, that large tractsof land are still covered with mature forest,and due to the nature of the terrain thecomplete deforestation of these islands isimprobable. This, combined with the protec­tion now afforded native species by the Fijigovernment, may be sufficient to maintainthe present surviving members of the nativefauna.

·",a ..11'9. PilHU .3 /iilfflOJ JS.J14 US§! IS mg hiiIilJMUMWii£4 PI§ $ $;;; Ittlll::: i 114$1li ; ;;1

Terrestrial Vertebrates of Fiji-PERNETTA AND WATLING 237

CONCLUSIONS

With the probable exception of the Poly­nesian starling, no evidence exists to supportthe claim that direct or indirect competitionbetween native and introduced species hasrestricted the distribution of the former. Itwould appear that the introduced species ofbirds, mammals, and amphibia have ex­panded in numbers and range as a result ofthe creation of suitable habitats throughprehistoric and historic agricultural activities.Further, most native species are adapted to aforest environment and a number may be

incapable of adapting-ecologically and be­haviorally-to human-modified habitats. Theintroduced small Indian mongoose has prob­ably caused the decline in abundance ofvarious ground-living birds on the islands ofViti Levu and Vanua Levu. However, thepossible extinction of the barred-winged rail,grass owl, and whistling tree duck, and thepresent rarity of the red-throated lorikeet,pink-billed parrot finch, long-legged warbler,and banded iguana cannot be directly at­tributed to this or any other introducedvertebrate.

APPENDIX

HABITAT, DISTRIBUTION, STATUS, AND ORIGINS OF FIJIAN VERTEBRATES

ORIGINS/STATUS SPECIES

Amphibia(A) E Platymantis vitiensis

Fiji tree frog(A) E Platymantis vitianus

Fiji ground frogP.E. Bufo marinus

marine or cane toadReptiles

Goniocephalus godfreyi

E Brachylophus fasciatusbanded iguana

3 (A) Engyrus bibronii and EngyrusaustralisPacific boa

(1) E* Ogmodon vitianusbolo

(A) Gehyra oceanicaoceanic gecko

(A) Lepidodactylus lugubrismourning or Pacific gecko

E Lepidodactylus manniFiji; Mann's gecko

(1) E Lepidodactylus gardineriRotuma gecko

PREFERRED HABITAT TYPES1 234 5 6 7

* *

* *

? ?

DISTRIBUTION AND COMMENTS

Viti Levu, Vanua Levu, Taveuni,Ovalau

Viti Levu, Vanua Levu, Taveuni,Ovalau

Main islands

Viti Levu; only two knownspecimens (Boulenger 1887)

Widespread

Widespread; two nominal speciesnot valid (Burt and Burt 1932)

Viti Levu; only 18 knownspecimens (Gorham 1970)

All islands

Buildings only; widespread

Previously known from fewspecimens. Abundant on rockfaces. Viti Levu only

Rotuma

NOTE: Data based on observations of birds over 9 years (D. W.); Collection of 124 specimens and approximately 230 additional recorded observationsof reptiles and amphibia; mist netting and observations of bats; and 254 trapped specimens of rodents.

KEY TO SPECIES ORIGINS: E, species endemic to Fiji islands only; E·. genus or subgenus endemic to Fiji; A. aboriginal introduction; (A), probable aborigi­nal introduction; P.E., post·European introduction; (P.E.), possible post-European introduction; N, indigenous species; M, migrant species; V, vagrantspecies; I, species colonized from Papua, New Guinea, region; 2, species colonized from Australia; 3, species colonized from Papua, New Guinea, orAustralia; 4, species endemic to central Polynesian subregion (Fiji, Samoa, Tonga, and outlying islands); 5, species of probable central Polynesian origin,now more widespread; 6, species widespread, entered the Pacific via the Philippines. Numbers in parentheses indicate possible origin on the basis of closerelationships between endemic species and those found elsewhere. For example, (I) 4 N Vin; australis, blue·crowned lory-native species to Fiji; endemicto the central Polynesian subregion; a closely related species indicating a Papua, New Guinean, origin.

KEY TO PREFERRED HABITAT TyPES: I, montane forest; 2, lowland forest; 3, intermediate zone vegetation (including grassland); 4, agricultural land;5, suburban and urban; 6, coastal and mangrove swamps; 7, freshwater and inland swamps; t, normal habitat; ., occasional habitat.

238 PACIFIC SCIENCE, Volume 32, July 1978

APPENDIX (Cont.)

HABITAT, DISTRIBUTION, STATUS, AND ORIGINS OF FIJIAN VERTEBRATES

ORIGINS/ PREFERRED HABITAT TYPESSTATUS SPECIES I 2 3 4 5 6 7 DISTRIBUTION AND COMMENTS

6 (P.E.) Hemidacty/us garnotli Buildings only; Viti Levu, Taveunifox gecko

6 (P.E.) Hemiphyllodacty/us typus Buildings only; Viti Levutree gecko

(A) N Cyrtodacty/us pe/agicus Forest floor; main islands onlypelagic gecko

(A) N Lipinia noctua Coastal/dry zonesmoth skink

(A) N Cryptob/epharus boutoni Supralittoral zonesnake-eyed skink

(A) N Emoia nigra Coastal/plantation ground zonesblack skink

(I) E Emoia conc%r * * Palm trees in agricultural andgreen skink suburbs

(I) 5 N Emoia samoensis WidespreadSamoan skink

(I) 5 N Emoia cyanura Widespreadstriped skink

(I) E Emoia sp. novum Viti Levu, Taveuni, KadavuI N Emoia caeru/eocauda Viti Levu, Taveuni

blue-tailed skinkEmoia adspersa and Emoia Not seen by present authors

cyanogaster (Sternfield 1920, Werner 1899)Long-necked water tortoise Record unsubstantiated by

(species unknown) specimens (Loveridge 1945)V Crocody/us porosus Single record of this species

estuarine crocodile (Derrick 1965)Birds

3 N Demigretta sacra Widespreadreef heron

3 N Butorides striatus Large islandslittle mangrove heron

3 N Anas superciliosa Large islands, decreasedAustralian gray duck

P.E. Anas p/atyrhynchos Under domestication onlydomestic duck

3 N Dendrocygna arcuata ? Extinct? Viti Levu, Vanua Levuwhistling tree duck (Mayr 1945)

P.E. Anser anser Under domestication onlydomestic goose

E Accipiter rujitorques WidespreadFiji goshawk

2 N Circus approximans Widespreadswamp harrier

3 N Fa/co peregrinus * (') Viti Levu, Wakaya (IClunieperegrine falcon 19720)

A Gallus gallus ? Common on large islands free ofjungle fowl the mongoose; otherwise

rare/absentP.E. Synoicus australis Viti Levu, Vanua Levu; grasslands

brown quail only, uncommonP.E. Meleagris gallopavo Under domestication only

domestic turkey3 N Porphyrio porphyrio Common on islands free of the

purple swamphen mongoose; otherwiserare/absent

ii' M E&

Terrestrial Vertebrates of Fiji- PERNETTA AND WATLING

APPENDIX (Cont.)

HABITAT, DISTRIBUTION, STATUS, AND ORIGINS OF FIJIAN VERTEBRATES

239

ORIGINS/ PREFERRED HABITAT TYPESSTATUS SPECIES I 2 3 4 5 6 7 DISTRIBUTION AND COMMENTS

E Nesoclopeus poeeilopterus Viti Levu, Ovalau; extinct?barred-wing rail

3 N Rallus phillipensis Common on islands free of thebanded rail mongoose; otherwise rare

3 N Poliolimnas einereus Rare; recorded from Viti Levu,white-browed rail Ovalau, Gau (Mayr 1945)

3 N Porzana tabuensis Rare; recorded from Viti Levu,sooty rail Ovalau, Gau, Kadavu

(Mayr 1945)4 N Ptilinopus perousii * Widespread

many-colored fruit dove4 N Ptilinopus porphyraeeus Outlying islands

crimson-crowned fruit doveE Ptilinopus victor Vanua Levu, Taveuni, and

orange dove adjacent islandsE Ptilinopus luteovirens * Viti Levu, Ovalau, Gau, Yasawas

golden doveE Ptilinopus layardi Kadavu group

velvet doveP.E. Streptopelia ehinensis Main islands

spotted; Malay turtle dove4 N Gallieolumba stairii * Widespread

friendly ground dove5 N Dueula pacifica Outlying islands

Pacific pigeonE Dueula latrans Widespread

Peale's pigeonP.E. Columba livia * Large islands

domestic pigeonN Columba vitiensis Widespread

white-throated pigeonE Charmosyna amabilis (I) Viti Levu, Ovalau, Taveuni

red-throated lorikeet (IGorman 1975b)(I) 4 N Vini australis Outlying islands

blue-crowned loryE* Phigys solitarius Southern Lau

collared loryE* Prosopeia tabuensis Koro, Gau, Kadavu, Taveuni,

red-breasted musk parrot Vanua Levu; Introduced VitiLevu (Wood and Wetmore1926)

E* Prosopeia personata Viti Levu, ?Ovalauyellow-breasted musk parrot

2 N Caeomantis pyrrhophanus (I) t Large islands (IGorman 1975b)fan-tailed cuckoo

M Eudynamis taitensis Migratory specieslong-tailed New Zealandcuckoo

3 N Tyto alba Widespreadbarn owl

3 N Tyto longimembris (I) Extinct? Viti Levu (lCiuniegrass owl 1972b)

N Co/loealia spodiopygia Widespreadwhite-rumped swiftlet

240 PACIFIC SCIENCE, Volume 32, July 1978

APPENDIX (Cont.)

HABITAT, DISTRIBUTION, STATUS, AND ORIGINS OF FIJIAN VERTEBRATES

ORIGINS/ PREFERRED HABITAT TYPESSTATUS SPECIES I 2 3 4 5 6 7 DISTRIBUTION AND COMMENTS

N Halcyon chloris Widespreadwhite-collared kingfisher

N Hirundo tahitica WidespreadPacific swallow

2 N Artamus leucorhynchus Not southern Lau, Kadavuwhite-breasted wood swallow

5 N Lalage maculosa WidespreadPolynesian triller

P.E. Pycnonotus cafer Common; Viti Levu, Ovalau,red-vented bulbul Wakaya; rare-Beqa, Taveuni

N Turdus poliocephalus Large islandsisland thrush

P.E. Sturnus vulgaris Ono-i-Lau onlyEuropean starling

5 N Aplonis tabuensis WidespreadPolynesian starling

P.E. Acridotheres fuscus Main islands, except TaveuniIndian mynah

P.E. Acridotheres tristis Main islandsjungle mynah

E* Lamprolia victoriae Vanua Levu, TaveuniSilktail

E Vitia ruficapilla * Large islandsFiji warbler

E Trichocichla m{a ? ? Nominate race; Viti Levu;long-legged warbler Extinct? rediscovered Vanua

LevuE Rhipidura personata Kadavu

Kadavu fantailN Rhipidura spilodera * Large islands

spotted fantailE Mayrornis lessoni * Widespread

slaty flycatcherE Mayrornis versicolor Ogea Levu only

versicolor flycatcher4 N Clytorhynchus vitiensis Widespread

Fiji shrikebill5 N Clytorhynchus nigrogularis * Large islands

black-faced shrikebill(2) E* Myiagra vanikorensis Widespread

Vanikoro broadbill(2) E* Myiagra azureocapilla * Viti Levu, Vanua Levu, Taveuni

blue-crested broadbill2 N Petroica multico10r * Large islands

scarlet robin2 N Pachycephala pectoralis * Widespread

golden whistlerE Myzomela jugularis Widespread

orange-breasted honeyeater2 N Myzomela cardinalis ? ? ? ? Rotuma only

cardinal honeyeater5 N Foulehaio carunculata Widespread

wattled honeyeaterE* Xanthotis provocator Kadavu

Kadavu honeyeaterE Gymnomyza viridis Viti Levu, Vanua Levu, Taveuni

giant forest honeyeater

l'li'U !# e i !ifiHt -"!!F' 4,***E$ '50 &It. -

Terrestrial Vertebrates of Fiji-PERNETTA AND WATLING

APPENDIX (Cont.)

HABITAT, DISTRIBUTION, STATUS, AND ORIGINS OF FIJIAN VERTEBRATES

241

ORIGINS/ PREFERRED HABITAT TYPESSTATUS SPECIES I 2 3 4 5 6 7 DISTRIBUTION AND COMMENTS

E Zosterops explorator * Large islandsLayard's white-eye

2 N Zosterops lateralis Widespreadgray-backed white-eye

E Erythrura pealii Widespread (see Ziswiler et al.Fijian parrot finch 1972 for nomenclature)

E* Erythrura kleinschmidti t (I) Viti Levu (IClunie 1973)pink-billed parrot finch

P.E. Amandava amandava Main islandsstrawberry finch

P.E. Padda oryzivora Viti Levu, Vanua Levu, TaveuniJava rice sparrow

P.E. Gymnorhyna tibicen Coconut plantations; Taveuniblack-headed magpie only

Mammals5 N Pteropus tonganus * * Widespread

Tongan fruit bat5 N Pteropus samoensis Viti Levu, Vanua Levu

Samoan fruit batN Notopteris macdonaldi * Large islands

long-tailed fruit batE Pteralopex sp. Taveuni only (W. Beckon,

personal communication)3 N Tadarida jobensis Taveuni only

long-tailed mastiff batN Emballonura semicaudata * Large islands

sheath-tailed batP.E. Herpestes auropunctatus Viti Levu, Vanua Levu only

small Indian mongooseP.E. Canis familiaris Domestic only; widespread

domestic dogP.E. Felis domesticus Domestic only; uncommon

domestic cat(A) Rattus exulans Widespread

Polynesian ratP.E. Rallus rallus Large islands in association with

black rat peopleP.E. Rallus norvegicus Large islands in association with

brown or Norway rat peopleP.E. Mus musculus Large islands in association with

house mouse peopleP.E. Equus caballus Domestic only; free-ranging in

domestic horse some grassland areasA Sus scrofa * * Aboriginally introduced pigs

P.E. domestic pig feral on main islands; post-European introduction ofvarious domestic strains

P.E. Bos taurus Domestic; free-ranging in somedomestic cattle grassland areas; widespread

P.E. Bos indicus Domestic; uncommonZebu cattle

P.E. Ovis aries Domestic only; uncommondomestic sheep

P.E. Capra hircus Domestic; feral in grassland anddomestic goat on some small islands

P.E. Cervus unicolor Wakaya Island onlySambar deer

242

ACKNOWLEDGMENTS

The senior author gratefully acknowledgesthe financial support of the University of theSouth Pacific, Fiji, and the enthusiasm andsupport of many members of its School ofNatural Resources during the conduct ofthis work. In addition, the senior authorwishes to thank R. A. Woods, Universityof Winnipeg, Department of Biology, formaking the resources of that departmentavailable during the writing of this paper.For financial support of the junior author,thanks are due to the British Ornithologists'Union for a generous grant.

In addition, the authors wish to thankHarold Cogger of the Australian Museum,Sydney, for his invaluable assistance in con­firming and identifying various reptile speci­mens; W. C Brown, Menlo College, MenloPark, California, for his comments and iden­tification of the Emoia material collectedduring the present study; and A. C. Ziegler,Bernice P. Bishop Museum, Hawaii, for hiscomments, critical and otherwise, of thedraft of this paper.

LITERATURE CITED

BAHR, P. 1912. Notes on the avifauna of theFiji Islands. Ibis 6: 282-314.

BERGER, A. J. 1972. Hawaiian bird life.University Press of Hawaii, Honolulu.270 pp.

BERRY, M., and W. HOWARD. 1973. Fiji forestinventory. Vo!. 1. Land resources studyno. 12. Foreign and Commonwealth Office.O.D.A. London. 98 pp.

BOULENGER, G. A. 1887. Catalogue of thelizards in the British Museum. 2d Ed.British Museum. London. 382 pp.

BREWSTER, A. G. 1935. King of the CannibalIsles. Robert Hale. London.

BROWN, W. C 1956. The distribution of theterrestrial reptiles in the islands of thePacific Basin. Proc. Eighth Pac. Sci. Congr.3:1479-1491.

BROWN, W. C, and J. C. PERNETTA. in prep.

PACIFIC SCIENCE, Volume 32, July 1978

A new Emoia (Scincidae) from the FijiIslands.

BURT, C. E., and M. D. BURT. 1932. Herpe­tological results of the Whitney South SeasExpedition. VI. Bul!. Amer. Mus. Nat.Hist. 63 :461-597.

CAHILL, C. 1970. The banded iguana. FijiMus. Ed. Ser. 2. Govt. Printer, Suva. 14 pp.

CLUNIE, F. 1972a. A contribution to thenatural history of the Fiji peregrine. No­tornis 19: 302-322.

---. 1972b. Fijian birds of prey. Fiji Mus.Ed. Ser. 3. Govt. Printer, Suva. 14 pp.

---. 1973. Pink-billed parrot finches nearNailagosakelo Creek, S. Viti Levu. No­tornis 20: 202-204.

COGGER, H. 1974. Voyage of the bandediguana. Aust. Nat. Hist. 18: 144-149.

DARLINGTON, P. J., JR. 1957. Zoogeography.John Wiley & Sons, New York. 675 pp.

DERRICK, R. A. 1965. The Fiji Islands. Govt.Printer, Suva. 334 pp.

DHONDT, A. 1976. Bird observations in W.Samoa. N otorriis 23 :29-43.

FINSCH, 0., and G. HARTLAUB. 1867. Beitragezur Fauna centralpolynesiens: Ornitholo­gie der Viti-, Samoa- und Tonga-inseln.H. W. Schmidt, Halle.

GIBBS, L. S. 1909. A contribution to themontane flora of Fiji with ecological notes.J. Linn. Soc. Bot. 34: 130-212.

GORHAM, S. W. 1965. Fiji frogs (with syn­opses of the genera Cornufer and Platy­mantis). Zoo!' Beit. 11 :381-435.

---. 1968. Fiji frogs, life history data fromfield work. Zoo!' Beit. 14: 427-446.

---. 1970. The rare Fiji snake Ogmodonvitianus Peters. Fiji J. Agr. 32:49-51

GORMAN, M. L. 1972. The origin of theavifauna of urban and suburban Suva, Fiji.Fiji J. Agr. 34: 35-38.

---. 1975a. The diet of feral Herpestesauropunctatus (Carnivora, Viverridae) inthe Fijian Islands. J. Zoo!' Lond. 175:273-278.

---. 1975b. Habitats of the land birds ofViti Levu, Fiji Islands. Ibis 117: 152-161.

GRESSITT, J. L. 1956. Some distribution pat­terns of Pacific island faunae. SystematicZoo!. 5: 11-47.

i! PH klillS. , hi "",aft'#! PH ift9ff JN htJftrlfMlM ? eN

Terrestrial Vertebrates of Fiji-PERNETTA AND WATLING 243

Guppy, H. B. 1906. Observations of a natu­ralist in the Pacific between 1896 and 1899.Vol. 2. Plant dispersal. Macmillan, Lon­don. 627 pp.

HINKLEY, A. D. 1962. Diet of the giant toad,. Bufo marinus (L.), in Fiji. Herpetologica

18: 253-259.HOUTZ, F. C., and K. A. PHILLIPS. 1963.

Interim report on the economic geologyof Fiji. Fiji Geol. Surv. Dept. EconomicRept. Govt. Printer, Suva.

KINSKY, F. C. 1975. A new subspecies ofthe long-legged warbler, Trichocichla rufa(Reichenow), from Vanua Levu, Fiji. Bull.Brit. Orn. CI. 95 :98-101.

LACK, D. 1969. The numbers of bird specieson islands. Bird Study 16: 143-209.

---. 1971. Ecological isolation in birds.Blackwell Scientific Publications, Oxford.404 pp.

---. 1975. Island birds. University ofCalifornia Press, Berkeley. 445 pp.

LAYARD, E. L. 1875. Notes on Fijian birds.Proc. Zool. Soc. Lond. 1875 :423-442.

---. 1876. Notes on some little knownbirds of the new colony of the Fiji Islands.Ibis 6: 137-157.

LOVERIDGE, A. 1945. Reptiles of the Pacificwofld. Macmillan Co., New York.

McKEY, D. 1973. The ecology of coevolvedseed dispersal systems. Pages 158-191 inCoevolution of animals and plants. L. E.Gilbert and P. H. Raven, eds. Universityof Texas Press, Austin.

MAYR, E. 1945. Birds of the South WestPacific. Macmillan Co., New York.

MERCER, R. 1964. A field guide to Fiji birds.Fiji Mus. Spec. Publ. No. 1. Govt. Printer,Suva. 40 pp.

MERTENS, R. 1934. Die Insel-Reptileen, ihreausbreitung, Variation und Artbildung.Zoologica, Stuttgart 32: 1-209.

PARHAM, J. W. 1956. The grasses of Fiji. FijiDept. Agr. Bull. 30. Govt. Printer, Suva.166 pp.

---. 1972. Plants of the Fiji Islands. Govt.Printer, Suva. 462 pp.

PERNETTA, J. C., and B. GOLDMAN. 1977.Botaniviti: the elusive Fijian frogs. Aust.Nat. Hist. 18:434-437.

RICKARD, M. J. 1966. Reconnaissance ge­ology of Vanua Levu. Fiji Geol. Surv.Dept. Mem. 2. Govt. Printer, Suva.

SCHMIDT, K. P. 1923. A list of Fijian lizards.Copeia 116: 50-58.

SEEMAN, B. 1862. "Viti." An account of aGovernment mission to the Vitian or Fijianislands in the years 1860-1861. Macmillan& Co., London. 447 pp.

SMITH, A. C. 1951. The vegetation and floraof Fiji. Sci. Monthly 73: 3-15.

STERNFIELD, R. 1920. Zur tiergeographie Pa­puasiens und der pazifischen Inselwelt.Abh. Senekenb. Nat. Ges. 36: 373-436.

SYKES, R. A. 1933. The forests of the Colonyof Fiji. Fiji Legislative Council Pap. 9.Govt. Printer, Suva.

TATE, G. H. 1935. Rodents of the generaRattus and Mus from the Pacific islands.Bull. Amer. Mus. Nat. Hist. 63:145-178.

THAMAN, R. 1974. Lantana camara: its intro­duction, dispersal and impact on islandsof the tropical Pacific Ocean. Micronesica10:17-39.

TWYFORD, I. T., and A. C. S. WRIGHT. 1965.The soil resources of the Fiji islands. Govt.Printer, Suva. 2 vols.

VEITCH, R. 1923. The minor pests of sugarcane in Fiji. Colonial Sugar Refiners Co.Ltd. Agr. Rept. 7.

WATLING, D. 1975. Observations on the eco­logical separation of two introduced con­generic mynahs (A cridotheres) in Fiji.Notornis 22: 37-53.

---. 1977. The ecology of the red-ventedbulbul, Pycnonotus cafer in Fiji: D. Phil.Thesis. Cambridge University.

---. in press. A mynah matter. Notornis.WATSON, J. M. 1960. Notes on the wildlife

of Fiji and its conservation. Fiji J. Agr.30: 1-4.

WERNER, F. 1899. Beitrage zur herpetologieder pacifischen inselwelt und von klein­asien. Zool. Anz. 12:252-256.

WILLIAMS, G. 1953. The dispersal from NewZealand and Australia of some introducedEuropean passerines. Ibis 95: 676-692.

WILLIAMS, J. M. 1971. Assessing the effectof rat damage on coconut yields. Fiji J.Agr. 33: 55-56.

244

1973. Rat damage assessment andcontrol in cocoa. Fiji J. Agr. 35: 15-25.

WOOD, C. A., and A. WETMORE. 1925. Acollection of birds from the Fiji islands.Ibis 55:814-855.

---. 1926. A collection of birds from theFiji islands. Ibis 56: 91-136.

PACIFIC SCIENCE, Volume 32, July 1978

ZISWILER, V., M. R. GUTTINGER, andM. BREGULLA. 1972. Monographie derGattung Erythrura Swainson, 1837(Aves, Passeres, Estrildidae). Bonn. ZooI.Monogr. 2.

iiWt6ffi!itiiftGM ;;;;;P;I