The Pacific, largest of all oceans, contains some 25,000 islands, more than half the world's total. Because almost all of the food plants and pests are introduced, biological control of pests is highly appropriate for oceanic islands, particularly those with small human populations and extreme shortage of foreign exchange for purchasing pesticides. Indeed pest surveys in the south Pacific have shown that 85% or more of the major insect pests are introduced, particularly from Southeast Asia. Furthermore about 95% of the major weeds are introduced, most from Central or South America. In most instances these insect and weed pests are of much less importance in their area of origin because of attack by natural enemies. Classical biological control consists of introducing and establishing selected natural enemies when careful tests indicate that it is safe to do so.

Success rate in islands (Hawaii, in particular) or ecological islands (Australia, California) is the highest in the world, but has been very variable in most Pacific islands. For a decade after the late 1920's there was, however, a notable sequence of successes in Fiji, unmatched elsewhere in the South Pacific. It is very encouraging to report that, since the mid 1980's, a steadily growing and well supported momentum has been generated for biological control and this has already led to several significant successes, with more in prospect.

Nevertheless, the need for biological control programs in the Pacific is likely to increase greatly in the future as air travel and tourism expand. In Hawaii, for example, an average of 3.5 pests per year become established and in Guam about 2.5 pests per year, most nowadays coming from the Americas. In an effort to reduce the number of pest invasions the XIX International Congress of Entomology resolved to recommend

to all Pacific Nations that they take urgent action to ensure (i) effective inspection of all incoming packages, including first class mail and (ii) effective insecticide treatment of all aircraft flying to and between Pacific countries.

Over 100 species of parasitic and predacious insects have been deliberately introduced for biological control of pests on Guam. A pop­ulation of the butterfly Hypolimnas bolina was studied to determine mortality factors and to see if there were serious effects on this non­target butterfly. Only two species introduced for biological control, Tri­chogramma chilonis and Brachymeria lasus, were found attacking H. bolina. Both species are known generalists that attack a wide variety of hosts. H. bolina failed to reproduce in the study site in 17 of 22 months. This failure resulted from a combination of predation by ants on eggs, larvae and pupae, and parasitisation of eggs by two species of native parasitoid. The native parasitoids caused density dependent mortality. Ant predation was inversely density dependent. Populations of H. bolina were little influenced by T. chilonis, but B. lasus parasitized 24% of the pupae and was an important mortality factor in two of the months when reproductive efforts failed. Generalist parasitoids such as T. chilonis and B. lasus should not be imported for biological control purposes. 

Adults of the fruit piercing moth Othreis fulIonia are serious pests of ripening fruit in most islands of the western Pacific, eastern Australia and parts of Southeast Asia. In some countries egg parasitoids are abundant, contributing to biological control of the moth but few natural enemies of larvae are effective. In Papua New Guinea, O. ful­Ionia is not sufficiently abundant to be considered a pest due, it is believed, to the activity of natural enemies, particularly egg parasitoids.
The most active hymenopterous egg parasitoids from Papua New Guinea, a Telenomus sp. and an Ooencyrtus sp. have been established in Western Samoa in 1988 and in Fiji in 1990 respectively. The abun­dance of eggs of O. fulIonia and the number hatching have declined in Western Samoa since establishment of Telenomus sp. The contribution by egg parasitoids to biological control of O. fullonia in the Pacific is discussed.

Four primary fruit piercing moths (Noctuidae) are currently found in Micronesia. They are Othereis fullonia (Marianas, Palau, Pohn­pei and Kosrae), Ercheia dubia (Guam, Rota, Tinian and Saipan), Per­icyma cruegeri (Guam, Rota and Palau) and Platyja umminia (Guam). All inflict moderate to severe damage on a variety of citrus and soft fruits at  certain times of the year and, in  this regard, O. fullonia is considered to be the most important species throughout the region. This species is native to Asia where the larvae feed on various vines (Men­ispermaceae) in contrast to the Pacific region where they feed almost exclusively on certain Erythrina spp. (Fabaceae). Although both groups of plants are unrelated taxonomically, they contain similarly structured tetracyclic erythrina-type alkaloids which, we believe, provide the chem­ical stimulus for host plant selection and thus play a key role in the successful shift from primary to secondary host species. The larval host plants of P.  cruegeri are well known (Delonix regia and Peltophorum pterocarpum), whereas those of E.  dubia and P.  umminia remain to be determined.
Several natural enemies of O. fullonia are known and the micro­hymenopteran egg parasitoids, Telenomus sp., Ooencyrtus sp. (both den­sity  dependent) and  Trichogramma sp.  (density independent) rank among the most successful biological control agents. Trichogramma sp. also parasitize the eggs of P.  cruegeri and the pupae of this species are attacked by two other parasitoids, Exorista civiloides and Brachymeria sp. 

Biological control has solved many major problems caused by floating weeds and insects have been the best control agents. The most important of these weeds originated in South America and caused problems when they were introduced elsewhere in the tropics and sub­tropics. The search for control agents started in the 1960's and now most infestations of alligator weed, salvinia and water lettuce, and many of water hyacinth, have been brought under biological control. There has not been a single case of an agent becoming a pest following control. Significant infestations of some floating weeds remain in Africa, Mad­agascar, the Philippines, Malaysia and Indonesia due to administrative rather than scientific difficulties. Several other floating plants, notably members of the fern genus Salvinia in South America and Asia, are likely to become important weeds if (when?) they become established outside their native continents. If quarantine measures fail to prevent this, prospects for their biological control are good. Reasons include past experience, extensive vegetative propagation/genetic uniformity in these plants, and, in free-floating plants, packing together of individuals by the wind which maximises the searching efficiency of control agents. 

Populations of several species of Lepidoptera consuming mango leaves and flowers increased after successful biological control of the noctuid caterpillar Penicillaria jocosatrix on Guam. These included the geometrids Anisodes illepidaria, Thalassodes sp. and Chloroclystis sp. and the noctuid Nanaguna breviuscula. A. illepidaria populations increased 10 fold, and became damaging at certain times of the year. A comparison of its larval biology with that of P. jocosatrix suggested that there was indirect competition, and the larger, earlier, P. jocosatrix was preempting resources. The other moths increased, apparently as a result of increased flowering of the mango trees as a result of suppression of P. jocosatrix. The introduced biological control agents acted as keystone natural enemies and increased the resources available for other Lepidoptera. Implications in relation to ecological effects of biocontrol introductions are discussed. 

Biological control has been one of the primary methods of managing pest problems in American Samoa since 1954. Early projects included the introduction of the toad, Bufo marinus, for mosquito and centipede control; Toxorhynchites sp. for mosquito control; two praying mantids for control of the coconut stick insect, Graejfea crouanii; bac­ulovirus of Oryctes for the rhinoceros beetle, Oryctes rhinoceros; Te­trasticus brontispae for the coconut leaf hispa, Brontispa longissima; and Liothrips urichi for control of the weed Clidemia hirta. A major project has been the control of the giant African snail, Achatina fulica, by the predatory snails, Gonaxis kibweziensis and Euglandina rosea. More recent introductions include the predatory beetles, Nephaspis ocu­latus, N.  bicolor and Delphastus pusillus, and the parasitic wasp, En­carsia haitiensis, for control of the spiraling whitefly, Aleurodicus dis­persus. Introductions to control the coconut scale, Aspidiotus destructor, include the predatory beetles, Pseudoscymnus anomalus, Rhyzobius sat­teles, Telsimia nitida and  Chilochorus nigritus, and a parasitic wasp, Aphytis sp. The fungus, Metarhizium anisopliae, is being used to control B.  longissima. 

There have been many introductions of parasites and pred­ators to the Federated States of Micronesia (FSM) but, in the majority of cases, their impact on the populations of the pests for which they were introduced has not been assessed.
In recent years, the release of biocontrol agents has reduced pest abundance of (i) the coconut scale on Kapingmarangi, Pohnpei, FSM, and Majuro, Republic of the Marshall Islands; (ii) spiraling whitefly on Nukuoro Island and the main island of Pohnpei; (iii) orange spiny whi­tefly on Pohnpei, Kosrae, Chuuk and Yap, FSM; (iv) green stinkbug, (v) bean leafminer, (vi) Siam weed, and (vii) fluted scale on Pohnpei, FSM.

Nezara viridula (Hemiptera: Pentatomidae) has, for many years, been considered a major insect pest of vegetables on Pohnpei, Chuuk and Yap States of the Federated States of Micronesia. The egg parasitoid, Trissolcus basalis, obtained from Hawaii, was released on Kolonia, Nett and Madolenimw, Pohnpei and was recovered readily from stinkbug eggs 4  and 7 months after release. As a  result of this introduction, the stinkbug population became so low that it is seldom observed on vegetable plantings on Pohnpei. 

Natural enemies keep most pest populations in American Samoa below economically important levels. When new pests (such as Thrips palmi and Piutella xylostella) arrive, they initially cause severe damage. However, within one to two years, natural enemies or perhaps other environmental conditions reduce pest populations. Nevertheless, in some cases, such as that of Othreis fullonia, additional biological controls are needed in order to keep a pest below the economic thresh­hold. The success of natural enemies in American Samoa may be due to a number of factors: little insecticide usage, thus minimising adverse effects on natural enemies; mixed cropping systems (eg. interplanting with Coleus blumei); and the presence of the virgin rainforest, which may contain alternative hosts and suitable habitats for natural enemies. 

The Philippine lady beetle became a serious pest of solana­ceous crops on Saipan after its accidental introduction in 1948. In 1985, Pediobius foveolatus adults were reared from Epilachna varivestis pupae which had been parasitised as larvae in USA. These parasitoids were released into two heavily-infested mixed plantings of eggplant and to­mato. In a 1989 survey, P. foveolatus was reared from 80% of E. phi­lippinensis larvae collected. Prior to release of the parasitoid, E. philip­pinensis was typically controlled using biweekly insecticide applications, but chemical control is now required only occasionally. Rapid success of this program may be due to an abundance of solanaceous weeds (eg. Physalis spp., Solanum nigrum) providing an untreated refuge for par­asitoid production. 

The importance of agromyzid leafminers in the genus Lir­iomyza in the Pacific Basin has increased significantly since 1980. This change in pest status has been associated with the introduction of Lir­iomyza trifolii and L. sativae to many Pacific islands ranging from Ha­waii to Guam above the equator and French Polynesia to New Cale­donia below. These species are  predominantly pests of vegetable and ornamental crops. On most vegetable crops the  leafminers only affect the non-marketable, leafy portions of the plant, reducing photosynthetic capability. Relatively high populations can  be  tolerated without eco­nomic damage, so  the  potential to control Liriomyza through natural enemies is high.
Numerous parasitoid species have been introduced for Liriomyza control in Hawaii. These include the effective species Chrysocharis os­cinidis, Chrysonotomyia punctiventris, Ganaspidum utilis, and Halti­coptera circulus. An  indigenous species, Diglyphus begini, also occurs in Hawaii. Those providing high levels of control in Hawaii have been introduced into Guam, Pohnpei and Tonga. Research in Hawaii indi­cates that multiple natural enemies are  needed to  produce Liriomyza control in a  range of crops. Parasitoid presence and effectiveness can differ significantly between crops, with from 6  to  12  species occurring in  particular crops. One of the most significant species introduced to Hawaii is  Ganaspidium utilis, which quickly made an impact on  leaf­miner densities, successfully competing with previously established par­asitoids. Strains of Diglyphus begini, highly resistant to pesticides, have been found in Hawaii and laboratory selection has increased resistance in Ganaspidium utilis. Research is  proceeding on  the  use  of entomopathogenic nematodes against Liriomyza.

The Chinese rose beetle, Adoretus sinicus (Coleoptera: Scar­abaeidae), is a pest with a large host range in Hawaii. Damage is caused by the adults and is characterised by interveinal feeding on host plant leaves. Similarly, the rose beetle, Adoretus versutus, effects significant feeding damage on many plant types in American Samoa.
Studies are being conducted in Hawaii and American Samoa to develop control strategies for these rose beetles. These studies include: 1) feeding and mating observations, 2) screening of commercially avail­able lures for adults, 3) trials with fungi and nematodes for larval control, and 4) observations of adult feeding patterns. 

The banana skipper, Erionota thrax, a pest of Southeast Asian origin, became established in Papua New Guinea (PNG) in 1983. It subsequently defoliated banana plants throughout most of mainland PNG and, by 1991, had spread to New Britain. The larval parasitoid, Cotesia ( = Apanteles) erionotae, was introduced from Guam into quar­antine in PNG, where host specificity tests were conducted by exposing larvae of four species of indigenous Lepidoptera (3 Papilionidae, 1 Hes­periidae) to adult C. erionotae. The behavioural response of gravid fe­male parasitoids and the failure of parasitoids to develop in the these Lepidoptera indicated that, if C. erionotae was released, it would not attack any of the species other than E. thrax. C. erionotae was subse­quently released in five mainland Provinces between March and October 1990 and on New Britain between June and October 1991. The par­asitoid has since become established in the Central, Western, Morobe and West New Britain Provinces, attacking up to 67% of larvae of E. thrax. At all sites where C. erionotae has become established, popula­tions of E. thrax have declined. Five species of parasitoid have been recorded from eggs of E. thrax in PNG. Prospects for biological control of E. thrax by C. erionotae in PNG appear to be promising.