Over 4,600 species of plants have been introduced into the Hawaiian Islands over the last 200 years. Only 86, less than 2% of the total, have become serious pests of native ecosystems. Of these, the most significant are Andropogon virginicus, Clidemia hirta, Lantana camara, Leucaena leucocephala, Melinis minutiflora, Myrica faya, Passiflora mollissima, Pennisetum clandestinum, P. setaceum, Psidium cattleianum, Rubus argutus, and Schinus terebinthifolius. All 86 species are discussed with regard to their impact on the ecosystem, dispersal mechanism, fire tolerance, potential for biological control, and their distribution and principal infestation sites. Twenty-eight (32%) are invasive weeds; the remainder generally require some form of disturbance in order to become established.
The lowland ecosystems have suffered the most disruption from alien species because of agriculture, fire, and urbanization. However, all vegetation types have been affected to some degree. The ecosystems least impacted are alpine habitats, rain forests, and bogs, although they are coming under increasing pressure.
A number of strategies are discussed which may help to ameliorate weed problems. Greater effort by government is needed to educate the public on the need for importation control and to enforce regulations. Mechanical and herbicidal control is discounted except in small areas. Biological controls offer considerable hope, but there are many problems associated with this strategy.
The unique flora and fauna of the Hawaiian islands is seriously threatened by alien plants. Many native species have already been extirpated. Unless importation of aliens and the continuing disturbance of the native ecosystems is stopped, the prognosis for the remaining native biota is grim.
"The history of weeds is the history of man" (Anderson 1952); the development of the Hawaiian flora is a classical example. Prior to the colonization of the Hawaiian Islands by man, the rate of introduction of plants was very low because of difficulties of dispersal over 3,200 km of ocean and subsequent establishment. The approximately 272 plants that did become established adapted and diversified to produce a flora of 1,729 species and varieties, 95% of which are endemic (Fosberg 1948). The aboriginal Hawaiians accelerated the process of introduction by bringing with them plants necessary for their culture. These plants had a distinct advantage in the process of establishment because they were deliberately cultivated. Most were cultivars and less than 25 escaped, but St. John (1978) listed 7 additional weeds which he believed were introduced inadvertently. The major impact of the Hawaiians was restricted to the lowlands, resulting from the clearing and burning of the native vegetation (Kirch 1982).
On the arrival of Captain Cook in 1778, the rate of plant introductions increased. Each new culture arrived with plants necessary for their cuisine or pleasure. Agricultural interests accelerated the process even more by bringing in pasture grasses and forb seeds which generally were contaminated with weeds from other sources, a common problem until quite recently (Salisbury 1964). Finally, foresters brought in trees, initially to reforest watershed areas devastated by feral cattle (Bos taurus), but later to establish a forestry industry in the Islands. Introductions of horticultural interest continue even today with little control except for a specific ban on a number of drug-producing plants and officially declared "noxious" weeds. Statutes exist which regulate the importation of plants, but they are only cursorily enforced except for those plants that come in under permit. The State's overriding concern is to protect and promote agricultural and other economically attractive interests, whereas protection of native ecosystems has received little consideration.
There are over 4,600 alien species in Hawai‘i (St. John 1973), of which over 600 have become naturalized (W. L. Wagner, pers. comm.). In this paper 86 are considered pests in areas not cultivated or urbanized. Many other weeds are confined to agricultural areas (Haselwood and Motter 1983). Although the term "pest" normally connotes social, economic, and biological problems (Norton and Conway 1977), in this paper it is used solely in the context of a weed's disruptive impact on the natural processes of native ecosystems. Species that prevent the reestablishment of native communities are included, along with those which invade disturbed or undisturbed native communities. Several species that may be considered beneficial by agronomists or horticulturists as forage, ornamentals, or for timber are included in this listing because of their negative impact in native Hawaiian ecosystems.
Six terms (adventive, alien, exotic, introduced, naturalized, and weed) are commonly used to describe species not native to an area. In Hawai‘i most people use "exotic" and "introduced", or more rarely "adventive". It may seem overly critical to worry about which term is correct. However, we are discussing not only a problem of communication among scientists, but also with the general public. The semantic confusion, particularly for the non-scientist, could prevent understanding of some essential elements of the problem. The term "exotic", although literally correct, is inappropriate because it also implies something excitingly different. "Adventive" is a word with a specific botanical meaning, i.e., not native to the environment, but it is generally used to refer to accidental introductions which persist for a while and then disappear. It is also somewhat indiscriminate in that it can be used to refer to a species introduced from a neighboring ecosystem. "Introduced" is probably appropriate except that it implies a deliberate action and does not carry the negative implication of the term "alien." The term "naturalized" is unsuitable because it refers to alien species which have become established and self-sustaining in a new geographic area. Not all alien species are naturalized and not all naturalized species are pests.
In many parts of the world, "alien" is the preferred word. "Alien" is appropriate because its meaning is direct and it also has the connotation of not belonging, a strongly desirable implication. The term "weed" refers to the functional role of an organism and is inappropriate in this context because not all alien plants are weeds and native species can also be weeds. Two endemic species, Cuscuta sandwiciana Choisy (dodder) and Hesperocnide sandwicensis Wedd. (stinging nettle), have been declared noxious weeds by the State of Hawai‘i, and an additional 7 endemic and 8 indigenous species have been called weeds (Haselwood and Motter 1983; Hosaka and Thistle 1954). The word alien will be used in this paper and its use is encouraged elsewhere.
PLANT
PESTS OF HAWAIIAN NATIVE ECOSYSTEMS
PROBLEM ALIEN PLANTS IN HAWAI‘I BY ISLAND
The determination that any weed is a pest is somewhat subjective. However, there is usually a general consensus about the most important weeds on each island. Seven species (Florida prickly blackberry, Christmasberry, common guava, Guinea grass, koa haole, lantana, and molassesgrass) are serious pests on all islands (Table 1). They all occur in dry to mesic lowland environments, but lantana and molassesgrass are also found at higher elevations and in wetter habitats. Another 5 (common ironwood, swamp oak, Java plum, kiawe, and strawberry guava), although present on all islands, are a problem only on 5 islands. With the exception of swamp oak, they are not a problem on Lana‘i because it is too dry for their proper development. Another 7 species (bamboo, Chinese banyan, faya tree, Indian fleabane, roseapple, Glenwood grass, and sourbush) are a problem on 4 islands only, but they may be present on all islands. Again these are species which prefer wetter habitats and are therefore excluded from Lana‘i and some other islands.
All 19 species just noted are serious threats to native ecosystems or their reestablishment. They all smother vegetation, displacing the native species or preventing their reestablishment. Egler (1942) has suggested that koa haole will enhance recolonization of native species by ameliorating the habitat. There is little evidence to support this hypothesis. In fact, it appears that the native species are being excluded from these environments and their seed banks exhausted.
Fourteen other species (African tuliptree, ardisia, banana poka, bush beardgrass, broomsedge, cats-claw, fountaingrass, Indian rhododendron, kahili ginger, kikuyugrass, Kosters curse, palmgrass, velvetgrass, and white ginger) are a serious problem on 2 or 3 islands but have not yet filled their potential range. With the exception of bush beardgrass, catsclaw, fountaingrass, and velvetgrass, they all infest wet habitats. Broomsedge and kikuyugrass infest dry, mesic, and wet habitats.
Seven species (buffelgrass, gorse, hairy cats-ear, kahili flower, Rubus nivalis, sweet vernalgrass, and white ginger) are problems only on Maui and Hawai‘i. Hairy cats-ear, Rubus nivalis, and sweet vernalgrass are confined to high elevations. Gorse was introduced to the higher elevations of Haleakala and Mauna Kea only but would grow in mesic habitats above 200 m on all islands. Buffelgrass is a serious weed in very dry areas. It is not clear why white ginger infestations are confined to Maui and Hawai‘i, as suitable habitat exists on other islands.
Table 1
Twenty-eight species appear to be a problem on one island only. Fourteen (50%) of them occur on Hawai‘i, 5 (18%) on Kaua‘i, 4 (14%) on Maui, 3 (11%) on O‘ahu, and 1 each (4%) on Lana‘i and Moloka‘i. Most of these species occur on other islands but have not escaped, probably because they have not been introduced into a suitable environment.
It is somewhat surprising that O‘ahu does not have the highest number of problem weeds. It is the principal port of entry for the Islands as well as the location of 3 botanical gardens. It is the most probable area where weeds would become established. Gerrish and Mueller-Dombois (1980) found some support for this hypothesis in that an area close to the metropolitan area (Tantalus) did have more weeds present than a more remote area (Pupukea). Also, juniper berry and several other plants which now infest the southern Ko‘olaus appear to have dispersed from the Lyon Arboretum. However, the largest number of weeds occur on the island of Hawai‘i. Although no obvious single explanation is available, it is known that several ranchers on Hawai‘i have introduced large numbers of plants and birds for various reasons, particularly at elevations between 800-1,700 m. The recent formation and volcanic activity of much of the Island and the large expanses of open forest may provide favorable conditions for alien plant establishment.
Like most Pacific islands, the northwestern Hawaiian Islands have been affected by man--either by guano mining or World War II. Nihoa and Necker have not been disturbed (Clapp, Kridler, and Fleet 1977). Some aliens are established on Pearl and Hermes Reef and French Frigate Shoals, but they are not doing well (Amerson, Clapp, and Wirts 1974; Amerson 1971). Laysan and Lisianski were devastated by rabbits. Some alien species, including Casuarina, were planted but they have not spread (Clapp and Wirtz 1975; Ely and Clapp 1973; Lamoureux 1963). Kure Atoll, on the other hand, has a predominantly alien flora (Woodward 1972).
PROBLEM ALIEN PLANTS IN HAWAI‘I BY VEGETATION ZONE
It is quite obvious that lowland ecosystems in Hawai‘i have suffered the most disruption from alien species because of agriculture, fire, and urbanization. These disturbances have created ideal conditions for the establishment of weeds. The lowlands are also the principal points of entry of most introductions to the Islands. Therefore, unless weeds survive in the tropical lowlands, they rarely become established. Occasionally, plants originally cultivated as ornamentals at lower elevations reach higher elevations, where they escape, e.g., senecio mikanioides. This generalization will not be tenable for much longer as larger communities become established in the cooler upper elevations at Koke‘e, Kaua‘i; Kula, Maui; and Volcano, Hawai‘i. Typical ornamentals will be introduced and then escape, e.g. the recent escape of nasturtium, Tropaeolum majus, at Hawai‘i Volcanoes National Park.
It is not possible to make any meaningful assessment of weed invasions of Hawaiian ecosystems on the basis of elevation because of the significant differences in climatic conditions around the major islands. Ripperton and Hosaka (1942) in an early description of Hawai‘i's vegetation subdivided it into 10 zones. Krajina (1963) later refined this system into 14 zones, principally by further subdividing the high rainfall categories. These vegetation zones provide a convenient framework for a discussion of the impact of alien plants on Hawai‘i's vegetation (table 2). Because disturbance in each zone has varied quite considerably, the following remarks are generalizations illustrating trends.
Land below 300 m elevation receiving less than 500 mm of rain each year (Ripperton and Hosaka's Zone A) is now almost totally dominated by alien forbs and shrubs (table 2). Formerly open native scrub grasslands, these areas were the principal habitation of the aboriginal Hawaiians. The lands were greatly disturbed by fire and in many instances have been severely eroded. When fires were suppressed as western civilization took hold, the native vegetation began to recover. However, the introduction of koa haole and its subsequent aerial broadcasting resulted in its rapid colonization and ultimate domination of these areas. Other species which are common in this zone include Formosan koa, Indian fleabane, klu, Java plum, and sourbush. In areas with subsurface water, kiawe is dominant. Along the west coast of Hawai‘i fountain-grass has become a serious pest and dominates many coastal and upland areas. Where cheap irrigation water is available and the soil is reasonably deep, sugar cane or other crops are grown. In the driest areas of most islands (e.g. Diamond Head and Ka‘ena Point, O‘ahu) a few native species still predominate.
Zone B, land below 1,000 m with an annual rainfall from 500-1,000 mm, has suffered a fate very similar to Zone A in areas where there is insufficient soil for agriculture. However, below 350 m and where irrigation water is available, sugar cane plantations, or more recently papaya and macadamia nut farms, have been developed. Above 200 m pineapple is cultivated where irrigation is not feasible. On land too steep for plantations, various timber trees were planted, principally for watershed management. Where cropping has been abandoned, the area is dominated by broomsedge, cats-claw, Christmasberry, Formosan koa, fountaingrass, guava, Indian fleabane, Java plum, klu, koa haole, Koster's curse, lantana, and sourbush.
Table 2
There are some notable exceptions to this general description of Zone B. Along the coastline in areas subjected to salt-water spray, there is a specialized ecosystem dominated by Fimbristylis. This community is still predominantly native because most weeds cannot tolerate the salt-water stress. Where cattle graze, the small shrubs have disappeared due to trampling and overgrazing. Many of these stress-resistant native species also occur elsewhere. Above 300 m on old unweathered lava flows or where there are many boulders, pockets of native dry forest have survived, e.g. Auwahi, Maui. Unfortunately, these woodlands frequently have a heavy ground cover of alien grasses, e.g., kikuyugrass, as well as heavy infestations of lantana. Survival of such woodlands is tenuous because kikuyugrass inhibits seed germination (Sanchez and Davis 1969) and cattle graze anything that grows.
In areas between sea level and 1,300 m where the annual rainfall ranges from 1,000-1,500 mm (Zone C) and where there is sufficient soil, sugar cane, pineapple, and more recently macadamia nut, plantations are found below approximately 650 m. Above this elevation most of the native forests have been converted to grassland. Some forest plantations have also been established in this zone. Below 1,300 m large areas have been infested and are now dominated by common guava from sea level to 400 m and strawberry guava from 300-1,300 m. Below 400 m the worst weeds are bamboo, broomsedge, Christmasberry, fiddlewood, glorybush, common guava, Indian rhododendron, Java plum, kahili ginger, koa haole, Koster's curse, roseapple, rosemyrtle, sourbush, white ginger, and yellow ginger. Above 400 m, Christmasberry, fiddlewood, Java plum, roseapple and sourbush drop out and fayatree moves in. Although much of the land in Zone C would be suitable for cattle grazing, it is overwhelmed by guava unless pastures are heavily grazed or the guava poisoned. Very few species can survive the competitive exclusion, principally by shading, although allelopathy is also suspected. Pockets of native vegetation still survive, but if feral pigs are not excluded from these areas they will soon be converted into guava woodlands (Diong 1983).
In areas between sea level and 2,300 m where the annual rainfall exceeds 1,500 mm (Zone D), the land below 1,350 m has been greatly altered by agriculture. Sugar cane is grown below 500 m although some plantations reach above 650 m. The gullies at these lower elevations are forested by a tropical weed flora (i.e., African tulip tree, ardisia, bush beardgrass, broomsedge, gingers, glorybush, Hilo grass, Indian rhododendron, Koster's curse, palmgrass, and roseapple) in which few native species survive. Between approximately 650 and 2,500 m, much of the land has been designated forest reserve for watershed protection. These forested areas are also maintained as hunting preserves, particularly for feral pigs. Areas not in reserve status have been converted to montane parklands by cattle ranching. Below 1,200 m strawberry guava is frequently dominant, forming monotypic stands. Between 1,000 and 1,650 m alien species are creating significant problems, e.g. banana poka, broomsedge, gingers, gorse, and velvetgrass. However, there are large tracts of ‘ohi‘a forest, generally in wetter areas, that have not been invaded by alien species other than a few relatively unimportant herbaceous plants.
Between 1,650 and 2,200 m there are numerous mires which have been undisturbed until quite recently. On Kaua‘i, few weeds have moved into the disturbed sites although Juncus planifolius R. Br. has invaded the summit area of Wai‘ale'ale (J. Canfield and R. L. Stemmermann, pers. comm.). On Haleakala, where feral pigs have destroyed the Oreobolus furcatus tussock community in some mires, hairy cats-ear and velvetgrass have become established, although they are partially displaced by the Oreobolus once disturbance ceases. The resistance of Hawaiian mires to alien plant invasion is unclear, but the environmentally suboptimal conditions of these areas for plant growth (Crawford 1983) may reduce the competitive advantage of the weeds.
In the inversion layer zone between 2,000 and 2,300 m, where the annual rainfall is less than 1,250 mm, the native koa forests have been converted to cattle ranches and few intact segments remain. Most of the grasses in these parklands are aliens, principally kikuyugrass, although a few native grasses remain. However, along with the introduction of favored pasture grasses, e.g. sweet vernal grass, a number of less desirable grasses (e.g., orchard grass and velvetgrass) became established and spread to other areas. The dense growth and allelopathic secretions of grasses prevent the successful germination of koa and other native seeds. In some instances, koa sucker growth is inhibited. However, even when suckers grow they are soon eaten back by cattle and goats.
Above 2,300 m (Zone E), depending on the elevation that the inversion level reaches, the annual rainfall is less than 1,250 mm. This is one of the least impacted areas and alien species are not common except on deep ash or where the forest has been opened by feral goats and sheep. Giant mullein, hairy cats-ear, sheep sorrel (Rumex acetosella L.), orchardgrass, and velvetgrass are the most common weeds in this environment. Above 2,650 m the climatic conditions are so severe that there is little vegetation (Hedberg 1951). Any plant that does emerge is promptly eaten by feral herbivores. These high elevation environments are generally protected from alien plant introductions because most of the ports of entry are in the tropical lowlands. Temperate species would have difficulty getting established at lower elevations. However, recently, 2 temperate weeds were introduced at Haleakala National Park (but fortunately immediately eradicated), indicating the danger of temperate weed introductions to Hawai‘i's high elevation ecosystems during construction of roads and facilities. It is hoped that similar monitoring and precautions will be taken during the construction of the new telescopes and support facilities atop Mauna Kea.
IMPACT OF ALIEN PLANTS ON HAWAIIAN ECOSYSTEMS
Weeds can have a number of very different effects on associated plants. Physical displacement of other species, either directly or indirectly, is the most frequent mode of action. They can also deprive associated plants of water or nutrients, particularly nitrogen, not so much the result of greater efficiency of uptake (Mahmoud and Grime 1976) but by absorbing their "fair share" of a generally limited resource. Allelopathy from root secretions (Martin and Rademacher 1960) or aerial portions (Sanchez and Davis 1969; Gogue, Hurst, and Bancroft 1974) is often a direct competitive action. Weeds also act as primary or alternate hosts for pests and diseases. All these impacts can undoubtedly be found in the activity of alien species on native ecological processes in Hawai'i. However, the most common impact is displacement when weeds colonize disturbed sites (e.g. pig diggings, burnt areas) and occupy them before the slower growing native species can reestablish. Other adverse effects not commonly associated with weed activity are discussed below.
Formation of Monotypic Stands
The replacement of a relatively diverse native ecosystem by monotypic
stands of alien species is a serious disruption of the ecosystem. In Hawai'i,
the loss of diversity in even small areas can have a devastating effect
on the survival of species, many of which are already endangered, almost
extinct, or confined to very small areas. For example, Cyanea superba,
a candidate endangered species, is now restricted to 2 areas of less than
0.1 ha each in the Mokulelia Forest Reserve on O'ahu. Most of its former
habitats are now occupied by strawberry guava. Where it survives it is
threatened by the overgrowth of weeds, fire, and feral animals. The highly
restricted distributions of many endemic species are not understood but
are probably related to loss of dispersability characteristic of many island
species (Carlquist 1965). Native primary consumers rarely adapt to aliens,
and in the majority of instances they are excluded from the alien ecosystems.
Strawberry guava was introduced in 1825 (St. John 1973) and soon established in the wild (Judd 1936). On most islands, nearly monotypic stands of this species infest hundreds of hectares of mesic and well-drained rain forest areas between 200 and 1,300 m. Vast areas of mature 'ohi'a and koa forests have a dense understory of strawberry guava. Native species regenerate rarely; some straggly specimens of native plants, e.g. Alyxia olivaeformis, Osmanthus sandwicensis, Psychotria mariniana, are able to survive under the guava. However, the prognosis for native ecosystem reestablishment in guava thickets is poor for 3 reasons. First, the shade is so deep that no seeds germinate, or if they do they die for lack of sufficient light. In old strawberry guava stands in mesic areas, there is no ground cover. In rain forest areas, Oplismenis hirtellus and Christella dentata dominate the ground cover. Second, the fruit is relished by feral pigs, which move into these thickets during the fruiting season (Diong 1983) and thoroughly disturb the ground. Seeds, unharmed by passage through the gut, are dispersed in pig feces, generally to another disturbed site where they have a competitive edge. Third, there is growing evidence of allelopathic activity in the fallen leaves.
The creation of monotypic stands of alien species may in fact be followed by a more disastrous event. As the weed exploits and exhausts the particular resource that it is able to use, it may outgrow itself. Or, as the natural processes of aging and diseases take their toll, the population may crash, resulting in accelerated erosion or further weed invasions, but rarely in the reestablishment of even a semblance of the original native ecosystem. These population crashes are not uncommon events elsewhere (Salisbury 1964) but as yet have not been recorded in Hawai'i. Such crashes rarely provide sufficient time for the orderly reestablishment of a diversified ecosystem.
Changing Fire Characteristics
Although Vogl (1969) proposed that fire is a frequent natural formative
agent in Hawaiian ecology, Mueller-Dombois (1981) found in one area of
Hawai'i Volcanoes National Park that carbon-dated charcoal deposits and
other evidence suggested a very low frequency. Human activity and the introduction
of weeds, particularly grasses, has changed that. Since 1910 most fires
in the Park have been started by man (Smith, Parman, and Wampler 1980).
Much of the area in the Park supports a scattered scrub forest separated
by patches of very sparsely vegetated lava. Fires, ignited by infrequent
lava flows, were not uncommon. However, these fires were small in area
because of the natural firebreaks in the vegetation mosaic.
In the early 1970's the situation changed. The National Park Service began a major resource management program to rid the park of feral goats (Baker and Reeser 1972). A few years before, both bush beardgrass and broomsedge had invaded the area. After most goats were removed, the grasses were no longer grazed and consequently colonized the many ash-filled cracks in lava flows, creating a continuous ground cover between islands of native scrub forest. The flowering stalks of these species are persistent after death, creating excellent fuel. These grasses are fire-stimulated, rapidly resprouting from basal sprouts after burning.
The total area burned by fires in lowland ecosystems is now 2 orders of magnitude larger than before the invasion of these grasses, even though an aggressive program of fire suppression by Park staff has significantly diminished the sizes of individual fires. Rapid regeneration of grass overshadows the reestablishment of most native species whose cover and abundance are drastically reduced for many years after the fire. As if to add insult to injury, the grasses also secrete a very persistent allelopathic agent (Rice 1972). The native Styphelia tameiameiae (Cham.) F. Muell., a common shrub in this ecosystem, is not firetolerant and will be excluded from burnt areas until it is reintroduced.
Changing Soil-Water Regimes
Many weeds introduced into oceanic islands come from temperate areas
of Europe and America. Their phenology is occasionally not in synchrony
with local climatic conditions, which results in significant changes in
the edaphic ecology of the area.
Broomsedge is a bunchgrass that became established in the Hawaiian Islands about 50 years ago. It quickly invaded open sites with a deep soil or ash deposit where the annual rainfall was above 100 mm. Originally from the southeastern United States, it has retained the phenological pattern of its native area (Sorensen 1980). However, as Mueller-Dombois (1973) pointed out, the dormant period for broomsedge coincides with the wettest months in the Islands, so that water is retained in the soil instead of being depleted through evapotranspiration. Surface runoff results in increased rates of erosion, with slumping on steeper slopes. Under native evergreen rain forest canopies water is transpired away rapidly and the remainder percolates through the soil. Erosion and slumping are rare.
Changing Nutrient Status
Volcanically active areas do not have mature soils. Instead they have
either lava or ash substrata characterized by low levels of nitrates. Plants
which grow in such areas are adapted to survive under these conditions.
However, species that can fix nitrogen will have faster growth rates than
natives. Eleven of the 86 alien plants listed in Table 1 fix nitrogen.
Nitrogen-fixers will also enrich the soil as their litter decomposes. The
outcome is an enriched soil in which other plants may grow, potentially
replacing the original occupants of the area.
Fayatree is capable of fixing nitrogen in root nodules containing the actinomycete Frankia (Mian, Bond, and Rodrigues-Barrueco 1976; Miguel and Rodriguez-Barrueco 1974). The invasion of Myrica is retively recent so that no floristic changes have been noted to date. Bradshaw et al. (1964), Rorison (1968), Mahmoud and Grime (1976), Higgs and James (1969), and Whelan and Edward (1975) have all shown that species native to poor soils are no more efficient at absorbing mineral nutrients than those native to good soils. Mahmoud and Grime (1976) further demonstrated the exclusion of plants native to poor soils, when grown on good soil with the plants adapted to good soils. The converse was not true. In the case of Myrica fava there is the distinct possibility that by enriching the soil it will enhance its own survival, perhaps to the exclusion of the native species. A monotypic infestation is already present in Hamakua, Hawai'i, which perhaps exemplifies this phenomenon; however, the history is insufficiently known.
Formation of Mutually Beneficial Interaction between Alien Plants
and Animals
The absence of terrestrial mammals in the Hawaiian fauna has resulted
in the lack of defense mechanisms against such animals. The introduction
of mammals, especially ungulates, produced a new selective pressure on
the native flora. Coupled with simultaneous introduction of alien plants,
many of which are components of early secondary succession in their native
habitat, Hawaiian ecosystems were faced with serious threats to their integrity.
Animals which dig up subterranean foods as part of their normal foraging activity are frequently important disseminators of plant propagules and play a role in nutrient cycling. In their native habitats, such disturbance is followed by a successional series of vegetation ultimately leading to some sort of climax community. In Hawai'i a natural successional series of this nature does not exist. Disturbance here has a significant deleterious impact on the native ecosystem, destroying ground cover, damaging roots, and opening up the understory. Most native species cannot tolerate this disturbance; however, many aliens are dependent on it for their establishment. When the alien species is an important food resource for a feral animal, a mutually beneficial interaction will develop between the 2 species. The relationship between feral pigs and such plants as strawberry guava, banana poka, and hairy cats-ear illustrate this point. Pigs are attracted to the abundant fruit of strawberry guava and banana poka. As pigs forage for other foods they disturb the ground, providing a seedbed for the guava and poka. Seeds are defecated up to 48 hours after consumption. They germinate and grow rapidly, occupying the site before the native plants establish. In the case of cats-ear, pigs destroy the weeds by digging up and consuming the roots. However, the digging provides an ideal seedbed for the establishment of the numerous, wind-dispersed seeds of this species. Why native species do not germinate or grow as rapidly as aliens is not yet known but can be related to their role in succession and the type of seedbed to which they are adapted. In both examples, original infestations are intensified or expanded, providing even more food resources for pigs.
IMPACTS
OF ALIEN PLANTS ON OTHER TROPICAL AND SUBTROPICAL ISLANDS
With a few exceptions--i.e., African Banks, Amirante (Feare 1979); Caroline Atoll (Clapp and silbey 1971); Gough Island (Wace 1966); Henderson Island (Melville 1979); Raine Island (Stoddart, Gibbs, and Hopley 1981); and Wilingili Atoll, Maldives (Spicer and Newbery 1979)--alien species have become weeds and serious pests on every island that man has visited. Island ecosystems have been invaded to a considerable extentl but generally only after disturbance by direct human influence, e.g., fire, plantations, and introduced herbivores. For brevity's sake, I will review the literature on tropical and subtropical areas only. There is an extensive literature on temperate and subantarctic islands which is beyond the scope of this paper.
Atlantic Ocean and Caribbean Sea
The effects of alien organisms have been studied on only a few islands,
but enough information is provided in floristic accounts and notes on many
other islands to allow a reasonable analysis of the status of weeds on
most islands. Although the prehistoric flora of St. Helena is poorly known,
one-third of the known endemic flora is extinct and no vestiges of former
ecosystems remain. New Zealand flax is the most serious pest. Ascension
Island was relatively barren on its discovery and apparently a dismal place
(Duffey 1964; Melville 1979). Over a thousand species have been introduced
to vegetate the island. Many of them are familiar weeds on tropical islands,
but only Melinis minutiflora (locally known as greasy grass)
forms extensive ground cover above 500 m. Barbados has been severely disturbed
by sugar cane cropping, but there is a sizeable remnant of the seasonal
forest (Watts 1970). Mahogany has invaded most communities and abandoned
fields. Anisomeris fasciculate K. Schum., Cordia glabra
L., and Haematoxylon campechianum L. are also common. In the outer
Leeward Islands of the West Indies, alien species have invaded all ecosystems
(Harris 1962, 1965; Loveless 1960). Haematoxylon campechianum and
koa haole are perhaps the most prevalent weeds along with kiawe and several
thorny Acacia spp. common guava, Guineagrass, and Bermudagrass (Cynodon
dactylon (L.) Pers.). Although "the position of aliens... is formidable,"
Harris concluded that once human interference stops, the native species
will replace the currently dominant aliens. The Cayman Islands have been
significantly disturbed by agriculture, and 2 species, false kamani and
Colubrina asiatica (L.) Brongn., are both invading inshore
areas of the islands (Sauer 1982). In the Tortuga Keys, common ironwood
has formed a dense woodland on Loggerhead Key, but the remaining 10 keys
have not been affected by the alien species (Stoddart and Fosberg 1981b).
Except for the 2 mid-Atlantic islands, most other islands appear to be
able to revert to near-native plant formations once human interference
ceases.
Indian Ocean
The native vegetation of the following islands has been almost totally
destroyed by human activity: Diego Garcia (Stoddart 1971), the coral islands
north of Madagascar, except Aldabra (Stoddart 1967), and Rodriguez Island
(Melville 1979). In addition, however, the situation on Madagascar is among
the worst examples of human mediated destruction of native ecosystems in
the world (Humbert 1927). Rauh (1979) stated that "in less than 200
years the green island of Madagascar has been transformed into a red sand
island simply by the activity of Man!" Aldabra (Stoddart and Wright
1967), Reunion (Melville 1979), and the Seychelles (Stoddart and Fosberg
1981a) have been occupied or exploited for sugar cane or copra for some
time, but substantial segments of native ecosystems are still present.
Mauritius has been severely disturbed by sugar plantations, but a number
of reserves were set up in 1944. However, their continued existence is
threatened by Ligustrum walkeri Decne, strawberry guava,
Rubus spp. and Ardisia spp.
Pacific ocean
Most of the central Pacific atolls have been severely disturbed by
man (Egler 1942; Fosberg 1953; Lee1974). The atolls are such specialized
habitats that alien species, although maintaining themselves, do not become
disruptive elements of the ecosystem. On the high islands the picture is
very different. Their diversity of habitats and isolation have resulted
in the evolution of new species and unique associations of species. Unfortunately,
due to "haphazard exploitation and development" the "sorry
state of the Hawaiian islands, many of the Galapagos, [and] Juan Fernandez
... will be repeated" (Wace 1966). The recent forest cropping in Fiji
is yet another example of destructive expoitation (Melville 1979). The
disturbance creates avenues for the invasion of alien species which, once
established, can dominate that area and invade other habitats. The weed
pests in Hawai'i have been discussed already. Clidemia hirta is
a serious weed in Fijian forests (Wester and Wood 1977) although it is
under partial biocontrol (Simmonds 1933). Klu, lantana, Mimosa invisa,
common guava, and Urena lobata have all been declared noxious (Mune
and Parham 1967). In the Galapagos, Cinchona succirubra Pav. ex
Klotz., common guava, and Digitaria decumbens Stent are all serious
pests (van der Werer 1979). Miconia magnifica is invading the forests
of Tahiti, overwhelming the native ecosystems (Whistler, in press). In
Samoa, Castilloa elastica, Koster's curse, Funtumia
elastica, mikania micrantha and Passiflora laurifolia
are serious forest weeds (Whistler, in press). Lantana, koa haole, and
common guava are also present but do not form extensive dense stands. Yet,
in Tonga they are the 3 worst weeds in the forests (G. Buelow, pers. comm.).
In the Northern Marianas, koa haole is almost ubiquitous because of aerial
seeding after World War II (C. S. Hodges, pers. comm.).
Three separate programs are needed to manage the alien pest problem in Hawai'i:
Prevent Further Introductions
The prevention of further importations of alien species is imperative
if we are to manage our native habitats effectively. The money spent on
controlling pests would be much better spent on productive enterprises.
Resources will still be needed to educate people not to import biological
material indiscriminately, as well as to intercept deliberate attempts
to smuggle material into the State. An immediate political step would be
to require that all future government-sponsored landscaping use native
species or plants that are known not to naturalize in Hawai'i.
A total prohibition on importations is not necessary. However, an outright ban on certain plant groups is imperative, e.g. all melastomes. Fifteen species in this group have already been introduced, 3 of which are serious weeds (Plucknett and Stone 1961). Further importation of Rubus, grasses, passion fruit, and members of the Myrtaceae should be banned also, as should species known to be problems in other tropical islands. These include the following species (from lists cited in this paper): Anisomeris fasciculate, Funtumia elastica, Ligustrum walkeri, Mikania micrantha, Operculina ventricosa, and Passiflora rubra. Species known to be part of primary or secondary succession in tropical or subtropical areas should be evaluated before they are permitted entry. The evaluations should be conducted by a group of botanists familiar with weeds in tropical and temperate areas. They should be constituted as a State Commission similar to the Animal Advisory Commission. They should have the authority to ban outright any importation. Their recommendations should be addressed to the Boards of Agriculture and Land and Natural Resources, who would have the right to veto any recommendation permitting entry but not overrule a negative decision of the Commission. Applications for permission to import should follow a format similar to an environmental impact statement.
Without the cooperation of the general public, there is little likelihood that any preventive program will work. Public education is a Federal and State responsibility. The best place to conduct the education of tourists and visitors is on the plane prior to arrival in the Islands. A 5-minute "commercial" explaining the problem and the importance of preventing plant and animal introductions would be much more effective than the printed form currently handed out in an almost cavalier fashion by the airlines. "Honesty" boxes should be provided in the baggage claim area where people can discard material prior to leaving the airport. Confiscation of material should be minimized except from people who are bringing potentially hazardous material to the Islands. Quarantine of imported material is important to minimize the introduction of associated pests and diseases. A more visible and concerned presence by the State's agriculture inspectors at ports of disembarkation is also necessary. Periodic inspection of baggage similar to the procedure on departure for mainlands would help keep people honest. The constant disclaimer by the State that it does not have sufficient funds to support such a program is a tacit acknowledgement that it does not consider alien plant introductions to be a major problem. Failure to enforce current regulations has resulted in the importation of Miconia magnifica and its consequent establishment in Hilo, which may turn out to be a very serious problem.
Two recent brochures, "Beware of the Noxious Weed" and "Are You a Carrier?", published by Foster Botanic Garden, Honolulu, are a valuable first step toward educating the general public on weed problems in Hawai'i. Much more needs to be done.
Stop Disturbance of Ecosystems
Many biologists have stressed the relationship between disturbance
of ecosystems and alien establishinent (e.g., Harper 1965; Stone, this
volume). Two separate actions are necessary to reduce disturbance. The
first is to change the State Constitution so that preservation of the State's
natural resources is really mandated of the land managers. The second is
to develop feral ungulate management programs to the point that the native
forests are no longer significantly disturbed by these animals.
The State's Constitution is somewhat ambiguous regarding its natural resources. On the one hand, it talks about preservation and on the other, about the benefit of the public. The latter is currently interpreted as permitting exploitation. Some steps have been taken in the right direction. The State's Natural Area Reserve System, although originally an enlightened program, has become emasculated by politics and lack of finances and manpower. The number of areas formally designated is low. But even after areas are designated, no management is conducted because there are no funds allocated for that purpose. The resources of these areas are poorly known and research is discouraged by a cumbersome, time-consuming bureaucracy. Yet disturbances continue, sometimes on a large scale, resulting in the further degradation of ecosystems and the continued spread of alien species. Cooperation among the various government agencies with responsibilities in this area could also result in more effective management of pests. The recent signing of a memorandum of agreement between various State and Federal agencies regarding forest pest control is encouraging. However, the contributions of each of the agencies will be largely influenced by the internal budgets of each.
We have to accept the fact that feral ungulates are here to stay. However, it is possible to keep them out of areas not now infested, e.g. Oloku'i, Moloka'i; much of the Alaka'i Swamp, Kaua'i; Pu'u Kukui, 'Eke Crater, and Lihau Peak, West Maui; and to exclude them from important areas, e.g. national parks, State Natural Area Reserves, and the critical habitats of endangered and threatened species. The practice of maintaining sustained yields of animals by regulating the number of animals taken in forest reserves and other conservation areas should cease. This approach aggravates the alien plant problem by ensuring some level of perpetual disturbance in the forests.
Develop Strategies to Encourage Native species Reestablishment
Two programs are needed. The first is the formulation and implementation
of research on the biology of the most troublesome weeds. The second is
the development of an integrated pest management system.
The most basic research questions concern location and effects of alien species. If sufficient historical information on infestations is available, and there generally is a wealth of anecdotal information, the dynamics of the invasion can be described. The most susceptible habitats can be identified and measures adopted to contain or prevent outbreaks in areas where management is possible. By evaluating the biology of the alien, it is sometimes possible to identify critical points in the life cycle when it is susceptible to control. Herbicides have significantly different impacts at different stages of a plant's growth and development. Likewise, not all biological control agents are effective in all habitats. For example, lantana has been contained in some areas by biological control agents but remains uncontrolled in others (Gardner and Davis 1982). The most important functions of research are to evaluate the role of weeds in island ecosystems, their impact on individual native species, their dependency upon disturbance for success, and management strategies.
Weed management needs much greater evaluation today than in the past. The indiscriminate use of herbicides is very dangerous, not only to human health but also to the well-being of the native ecosystems in which they are used. Agricultural weed control strategies are generally developed to eliminate all species other than the crop. In natural ecosystems we generally try to eliminate one species only and preserve the rest. In addition, we are ignorant of the longevity or secondary effects of herbicides in tropical areas because most chemicals are evaluated in temperate ecosystems. Also, weeds can develop resistance to herbicides (Hanson 1956, 1962). This problem becomes more acute the longer the herbicide program is conducted. The temptation to use higher dosages, a very common practice in agriculture, must be assiduously avoided. Physical damage to the ecosystem almost always occurs during the application of the herbicide. However, a particular advantage of herbicide use is that the soil is left undisturbed, and in many instances the dead plant tissues form a ground cover that will impede the growth of seedlings.
Mechanical control is very expensive because it requires manpower. Another negative factor is the accompanying damage to the ecosystem in the process of weeding. The disturbance of the soil under the plant generally stimulates weed seeds to germinate.
Natural areas are not amenable to some techniques of weed control, e.g. changing cultural practices. However, the extent of the infestation or the sensitivity of the protected area may preclude any disturbance. In these instances, biological control is a potentially powerful weapon, but it is not a panacea (Howarth 1983; Mellanby 1974) and sometimes operates only in very restricted ranges. It is very expensive initially and the agents are not confined within any political boundaries.
Biological control is not the final solution to alien species problems in Hawai'i, yet some successes make it extremely attractive because it seems to be the natural solution to a problem. Although most species have other organisms that parasitize or feed on them, the successful introduction of these organisms is a formidable problem. Many may not be suitable for importation for the following reasons.
7. Parasites of control agents may have been introduced earlier in relation to another problem. The early importation of general parasites of Lepidoptera may prevent the use of heliconiid butterflies as control agents of Passiflora in Hawai'i.
These last points are often misunderstood by many casual exponents of biological control. The problem is further complicated by the fact that problematic alien species generally infest their "fundamental" niche (Hutchinson 1957), having escaped many of the constraints which confined them to the "realized" niche of their native environment. Thus, a suitable biological control agent will probably only be effective in a segment of the insular range of the alien.
Biological control is a science wrapped up in a restrictive bureaucracy, but with good reason. At times the rules appear overly stifling, but they are necessary to prevent abuse and to adequately demonstrate the safety of the proposal. It is tragic that such rules are not applied to the importation of alien plants in the first place. That biological control is not a panacea is further emphasized by its expense and the time necessary to verify that an agent is not only suitable but reasonably certain not to have undesirable side effects. It is, therefore, initiated only after all else fails.
Thus, there is no easy answer to controlling all alien plants in insular environments. Each species has to be managed on its own and generally by a number of different approaches. It is the successful integration of these different approaches that is the challenge to the research scientist and the manager. Solutions are not to be quickly found in most cases, but probably control is possible for most plants.
Introduced plants can be quite innocuous. For example, it is highly unlikely that such horticultural favorites as plumeria (Plumeria acuminata Ait.) or pua kenikeni (Fagraea berteriana Gray) will ever pose a threat to native ecosystems. On the other hand, a number of alien species, e.g. Koster's curse and strawberry guava, are very serious threats. My candidates for the 10 most serious weeds in Hawai'i, in order of priority, are: strawberry guava, Koster's curse, banana poka, fountaingrass, fayatree, kikuyugrass, Christmasberry, blackberry, molassesgrass, and bushbeardgrass.
It is generally believed that under natural conditions island ecosystems are stable, invasion-resistant assemblages of species whose combined resource exploitation is in balance with productivity (MacArthur 1972). Without other alien influences, island biogeographers and ecologists would predict that very few alien introductions would become established in native communities (Cockayne 1928; Allan 1936; Anderson 1952). However, island ecosystems are disturbed to varying degrees by man, fire, feral ungulates, introduced birds, and a vast array of alien invertebrates. In Hawai'i, a significant number of native species has become extinct due to alien influences. Thus the underlying ecological processes on which native communities are structured have changed, probably irreversibly, in most cases.
One problem facing managers of native ecosystems is the determination of the significance of negative impacts of alien plants in native ecological processes. If a weed does not affect, or only marginally affects, ecosystems, e.g., Euphorbia hirta L., then it can be tolerated. As the system recovers, such species will be contained. Species at the other end of the scale, e.g., strawberry guava, need immediate attention. The mechanism of entry and establishment of an alien in the ecosystem is very important. If a weed is dependent on disturbance, then control can focus on preventing such occurrences, where possible. However, if the weed is invasive, then management must attempt to reduce or eliminate the species. Probably no species is beyond control as long as time and money are available. However, political and practical considerations may preclude such optimism. Such considerations include whether: the target species is socially useful in some context; it is too closely related to a commercial crop species; its impact in a remote ecosystem is too unfamiliar to attract the necessary funding; or the control program interferes with the activities of a special interest group, such as hunters.
This paper identifies 86 alien plants which are serious weeds in Hawaiian
ecosystems. All but one, Hypochoeris radicata, displace native
species when growing in the same habitat. Twenty-six (30%) have, or are
suspected of having, allelopathic activity against native species. Twelve
(14%), principally grasses, are fire-enhanced species which invade the
fire-disturbed area much faster than the natives. In so doing, they increase
the fuel level in the ecosystem and carry fires over larger areas than
before and generally at higher intensities. On the other hand, another
10 (12%) species are known to inhibit fires.
Just over half of the significant aliens, 45 (52%), are phanaerophytes, 18 (21%) chamaephytes, 10 (12%) hemicryotophytes, 6 (7%) lianes, 5 (6%) are geophytes, and 2 (2%) therophytes. The preponderance of trees, particularly evergreen species, is of considerable concern because they become an integral part of the canopy. Here they have a much greater disruptive influence on ecological processes. The large number of vines, all of them photophilic canopy species, is particularly important in Hawai'i where this growth form is poorly represented in the flora and, until recently, not a significant feature in the ecosystem. Thus, instead of being a natural part of succession as in other tropical areas, they destroy the forest structure by shading or breaking native tree branches.
Seventy-five percent of the weed species are well-adapted for dispersal in the Islands. Thirty (35%) are dispersed by predominantly alien, frugivorous birds, 29 (34%) by wind, and 9 (10%) on clothing or animal hides. One species is dispersed by water. However, it is somewhat surprising that 23 (27%) are dispersed only by man in the Islands. Once established, their infestation intensifies with some local dispersal by physical means.
Forty-five (52%) species are confined to the highly altered lowland
(below 800 m) ecosystems, 10 (12%) range from sea level to mid-elevations
(between 800-1,700 m), 18 (21%) are principally confined to mid-elevations,
3 (3%) to both mid- and higher elevations (up to 2,700 m), and 10 (12%)
are confined to the higher elevations. The preponderance of pests in the
lowlands is expected due to the extent and variety of disturbance there,
as well as the increased opportunity for introduction.
Most of the weeds presented in this listing invade native communities only after some type of disturbance, generally the consequence of man but occasionally natural, e.g., landslides, hurricanes, and treefalls. However, approximately half of the species listed (particularly those dispersed by wind or birds) can invade native ecosystems but generally remain minor components until some disturbance occurs. Twenty-three species have the ability to invade and take over native ecosystems without any apparent disturbance. These are: African tuliptree, Ardisia, banana poka, blackberry, bush beardgrass, Christmasberry, fiddlewood, fountaingrass, glorybush, huehue-haole, Indian fleabane, Indian rhododendron, juniperberry, kahili ginger, Koster's curse, lantana, melochia, miconia, raspberry, roseapple, rosemyrtle, strawberry guava, and yellow Himalayan raspberry.
Egler (1942) suggested that weeds, such as koa haole, would ameliorate ravaged native ecosystems, allowing native species to reestablish themselves. Unfortunately, the time frame about which he was talking is so long that many natives would already be extirpated before conditions were favorable. Very few native species are able to maintain themselves in heavy infestations of arborescent weeds. Those that do are generally so weakened that seed production is severely diminished or absent. However, the critical weakness is that the native species cannot compete against the aliens in the germinant and early seedling stages. The outcome is that the seedbank of native species is exhausted, effectively excluding the species from that area; reinvasion would be the only means of reestablishment. Since weeds normally occupy all the available space in the habitat, the prognosis for the native species is dismal. Overall, it is almost hopeless because succession, if it occurs at all, is generally by alien species. One weed is replaced by another; in the case of koa haole, Christmasberry and Java plum are frequent invaders of the habitat. The native species are therefore still excluded.
With the exception of St. Helena and Madagascar, the Hawaiian Islands contain the most ravaged island ecosystems in the world. The introduction of more than 4,600 different plants can only have a devastating effect on the survival of the 1,700+ native species. It is fortunate that less than 100 aliens have become pests. However, that number will increase continuously if restrictions are not imposed. It is because of such considerations that it is imperative that the importation of alien plants be stopped and the continued disturbance of native ecosystems prevented.
I am greatly indebted to a number of people who have assisted me with various aspects of this paper. L. Cuddihy, T. Flynn, E. Funk, P. Higashino, R. Hobdy, G. Linney, A. Medeiros, S. Perlman, L. Stemmermann, and L. Whiteaker all helped considerably in determining which species were pests and the localities of their principal infestations. J. Canfield, D. E. Gardner, P. Higashino, C. S. Hodges, L. Loope, L. Stemmermann, C.P. Stone, and W. Wagner provided many helpful comments on drafts of the manuscript, for which I am grateful. I am very appreciative also of the work of R. Saito and L. Matsumori, who spent many hours in the library locating obscure literature references.
My thanks also to the National Park Service (CA8009 2 0001) for their partial support during the preparation of this paper.
Achhireddy, N. R., and M. Singh. In press. Allelopathic effects of lantana (Lantana camara) on milkweed vine (Morrenia odorata). Weed Sci. 32.
Allan, H.H. 1936. Indigene versus alien in the New Zealand plant world. Ecology 17:187-193.
Amerson, A. B., Jr. 1971. The natural history of French Frigate Shoals, Northwestern Hawaiian Islands. Atoll Res. Bull. 150.
Amerson, A. B., Jr., R. B. Clapp, and W. O. Wirts II. 1974. The natural history of Pearl and Hermes Reef, Northwestern Hawaiian Islands. Atoll Res. Bull. 174.
Anderson, E. 1952. Plants, man and life. Berkeley, Calif.: Univ. Calif. Pr.
Baker, J.K., and D.W. Reeser. 1972. Goat management problems in Hawai‘i Volcanoes National Park: a history, analysis, and management plan. Natl. Park Serv. Nat. Resour. Rep. 2.
Bradshaw, A.D., M.J. Chadwick, D. Jowett, and R.W. Snaydon. 1964. Experimental investigations into the mineral nutrition of several grass species: IV. Nitrogen level. J. Ecol. 52:665-676.
Carlquist, S. 1965. Island life. A natural history of the islands of the world. New York: Nat. Hist. Pr.
Chou, C.H., and C.C. Young. 1975. Phytotoxic substances in twelve subtropical grasses. J..Chem. Ecol. 1: 183-193.
Clapp, R.B., and F.C. Silbey. 1971. Notes on the vascular flora and terrestrial vertebrates of Caroline Atoll, Southern Line Islands. Atoll Res. Bull. 145.
Clapp, R.B., and W.O. Wirtz, II. 1975. The natural history of Lisianski Island, Northwestern Hawaiian Islands. Atoll Res. Bull. 186.
Clapp, R.B., E. Kridler, and R.R. Fleet. 1977. The natural history of Nihoa Island, Northwestern Hawaiian Islands. Atoll Res. Bull. 207.
Cockayne, L. 1928. The vegetation of New Zealand. Weinheim, New Zealand: H.R. Engelmann.
Crawford R.M.M. 1983. Root survival in flooded soils. In Ecosystems of the world. 4A. Mires, swamps, bog, fen and Tnoor:- general studies, ed. A.i.P. Gore, 257-283. Amsterdam: Elsevier Sci. Pub. Co.
Diong, C.H. 1983. Population ecology and management of the feral pig (Sus scrofa L.) in Kipahulu Valley, Haleakala National Park, Maui, Hawai‘i. Ph.D. Diss., Univ. Hawai‘i, Honolulu.
Duffey, E. 1964. The terrestrial ecology of Ascension Island. J. Appl. Ecol. 1:219-251.
Egler, F.E. 1942. Indigene versus alien in the development of arid Hawaiian vegetation. Ecology 23:14-23.
Ely, C.A., and R.B. Clapp. 1973. The natural history of Laysan Island, Northwestern Hawaiian Islands. Atoll Res. Bull. 171.
Feare, C.J. 1979. Ecological observations on African Banks, Amirantes. Atoll Res. Bull. 227.
Fosberg, F.R. 1948. Derivation of the flora of the Hawaiian Islands. In Insects of Hawai‘i, ed. E.C. Zimmerman, Vol. 1 , Introduction, 107-119. Honolulu: Univ. Hawai‘i Pr.
Fosberg, F.R. 1953. Vegetation of Central Pacific atolls, a brief summary, Atoll Res. Bull. 23.
Gardner, D.E. 1984. Kikuyu grass rust caused by Phakopsora apoda in Hawai‘i. Plant Dis. 68:826.
Gardner, D.E., and C.J. Davis. 1982. The prospects for biological control of nonnative plants in Hawaiian national parks. Univ. Hawai‘i Coop. Natl. Park Resour. Stud. Unit Tech. Rep. 48. Honolulu: Univ. Hawai‘i.
Gardner, D.E., and V.A.D. Kageler. 1983. Glyphosate in the control of kikuyugrass, and its effects on associated native and nonnative plants in Hawaiian national parks. Univ. Hawai‘i Coop. Natl. Park Resour. Stud. Unit Tech. Rep. 49. Honolulu: Univ. Hawai‘i.
Gerrish, G. and D. Mueller-Dombois. 1980. Behavior of native and non-native plants in two tropical rain forests on Oahu, Hawaiian Islands. Phytocoenologia 8:237-295.
Gogue, G.J., C.J. Hurst, and L. Bancroft. 1974. Growth inhibition by Schinus terebinthifolius. Am. Soc. Hort. Sci. 9:45.
Hanson, N.S. 1956. Dalapon for control of grasses on Hawaiian sugar cane lands. Down to Earth 12:2-5.
Hanson, N.S. 1962. Weed control practices and research for sugar cane in Hawai‘i. Weeds 10:192-200.
Harper, J.L. 1965. Establishment, aggression, and cohabitation in weedy species. In The genetics of colonizing species , ed. H.G. Baker and G.L. Stebbins, 243-265. New York: Academic Pr.
Harris, D.R. 1962. Invasion of oceanic islands by alien plants. Inst. Brit. Geogr. Trans. 31:67-82.
Harris, D.R. 1965. Plants, animals and man in the outer Leeward Islands. Univ. Calif. Pub. Geogr. 18.
Harris, P. 1973. The selection of effective agents for the biological control of weeds. Proc. 3rd Internatl. Biol. Contr. Weeds, 75-85. Montpelier, France.
Haselwood, E.L., and G.G. Motter. 1983. Handbook of Hawaiian weeds. 2nd ed. Honolulu: Univ. Hawai‘i Pr.
Hedburg, 0. 1951. Vegetation belts of the East-African Mountains. Svensk Bot. Tidsk. 45:140-202.
Higgs, D.E.B., and D.B. James. 1969. Comparative studies on the biology of upland grasses: I. Rate of dry matter production and its control in four grass species. J. Ecol. 57:553-563.
Hosaka, E.Y. 1958. Kikuyugrass in Hawai‘i. Univ. Hawai‘i Agric. Ext. Serv. Circ. 389.
Hosaka, E.Y., and A. Thistle. 1954. Noxious plants of the Hawaiian ranges. Univ. Hawai‘i Coop. Ext. Serv. Bull. 62.
Howarth, F.G. 1983. Classical biocontrol: panacea or Pandora's box. Proc. Hawai‘i. Entomol. Soc. 24 (2&3):239-244.
Humbert, G. 1927. La destruction dune flore insulaire Par le feu. Acad. Malagache Mem. V.
Hutchinson, Q.E. 1957. Concluding remarks. Cold Spring Harbor 22nd Symp. Quant. Biol., 415-427.
Jacobi, J.D. 1981. Vegetation changes in a subalpine grassland in Hawai‘i following disturbance by feral Pigs- Univ. Hawai‘i Coop. Natl- Park Resour. Stud. Unit Tech. Rep. 41. Honolulu: Univ. Hawai‘i.
Judd, C.S. 1936. Seed dispersal in Hawai‘i. Mid. Pac. Mag. 49:111-118.
Kirch, P.V. 1982. The impact of the prehistoric Polynesians on the Hawaiian ecosystem. Pac. Sci. 36:1-14.
Krajina, V.J. 1963. Biogeoclimatic zones of the Hawaiian Islands. Hawaii. Bot. Soc. Newsl. 2: 93-98.
Lamoureux, C.H. 1963. The flora and vegetation of Laysan Island. Atoll Res. Bull. 97.
La Rosa, A.M. 1983. The biology and ecology of Passiflora mollissima in Hawai‘i. M.S. Thesis, Univ. Hawai‘i, Honolulu.
Lee, M.A.B. 1974. Distribution of native and invader plant species on the island of Guam. Biotropics 6:158-164.
Loveless, A.R. 1960. The vegetation of Antigua, West Indies. J. Ecol. 48:495-527.
MacArthur, R.H. 1972. Geographical ecology: patterns in the distribution of species. New York: Harper and Row.
Mahmoud, A., and J.P. Grime. 1976. An analysis of competitive ability in three perennial grasses. New Phytol. 77:431-435.
Markin, G.P. 1984. Biological control of the noxious weed gorse Ulex europaeus L.: a status report. Proc. 5th Hawai‘i Volcanoes Natl. Park Nat. Sci. Conf., 77. Honolulu: Univ. Hawai‘i.
Martin, P., and B. Rademacher. 1960. Studies on the mutual influences of weeds and crops. In The biology of weeds, ed. J.L. Harper, 143-152. Oxford, England: Blackwell Sci. Pub.
Mellanby, K. 1974. The future of biological control in Britain: a conservationist's view. In Biology in pest and disease control, ed. D. Price-Jones and M.E. Solomon, 349-353. Oxford, England: Blackwell Sci. Pub.
Melville, R. 1979. Endangered island floras. In Plants and Islands, ed. D. Bramwell, 361-378. London: Academic Pr.
Mian, S., G. Bond, and C. Rodrigues-Barrueco. 1976. Effective and ineffective root nodules in Myrica faya. Proc. Royal Soc. Lond., Ser. B, 194:285-293.
Miguel, C., and C. Rodrigues-Barrueco. 1974. Acetylene reducing activity of detached root nodules of Myrica faya Ait. Plant and Soil 41:521-526.
Mueller-Dombois, D. 1973. A non-adapted vegetation interferes with water removal in a tropical rainforest area in Hawai‘i. Trop. Ecol. 14:1-18.
Mueller-Dombois, D. 1981. Understanding Hawaiian forest ecosystems: the key to biological conservation. In Island ecosystems: biological organization in selected Hawaiian communities, ed. D. Mueller-Dombois, K.W. Bridges, and H.L. Carson, 502-520. Stroudsburg, Penn.: Hutchinson Ross Pub. Co.
Mune, T.L., and J.W. Parham. 1967. The declared noxious Weeds of Fiji and their control. Fiji Dep. Agric. Bull. 48.
Nakahara, L.M., and P.Y. Lai. 1984. New state record... Psyllids. Hawai‘i Pest ReT). 4:2-9.
Neal, M.C. 1965. In Gardens of Hawai‘i. B.P. Bishop Mus. Spec. Pub. 50.
Norton, G.A., and G.R. Conway. 1977. The economic and social context of pest disease and weed problems. In Origins of Pest, Parasite, disease and weed problems, ed. J.M. Cherrett and G.R. Sagar, 205-226. Oxford, England: Blackwell Sci. Pub.
Pimentel, D. 1963. Introducing parasites and predators to control native plants. Can. Entomol. 95:785-792.
Plucknett, D.L., and B.C. Stone. 1961. The principal weedy Melastomaceae in Hawai‘i. Pac. Sci. 15:301-303.
Rauh, W. 1979. Problems of biological conservation in Madagascar. In Plants and islands, ed. D. Bramwell, 405-422. London: Academic Pr.
Rice, E.L. 1972. Allelopathic effects of Andropogon virginicus and its persistence in old fields. Am. J. Bot. 59:752-755.
Ripperton, J.C., and E.Y. Hosaka. 1942. Vegetation zones of Hawaii. Hawai‘i Agric. Exp. Stn. Bull. 89.
Rorison, I.H. 1968. The response to phosphorus of some ecologically distinct plant species: I. Growth rates and phosphorus absorption. New Phytol. 67: 913-923.
St. John, H. 1973. List and summary of the flowering plants in the Hawaiian Islands. Pac. Trop. Bot. Gard. Mem. 1.
St. John, H. 1978. The first collection of Hawaiian plants by David Nelson in 1779. Hawaiian plant studies 55. Pac. Sci. 32:315-324.
Salisbury, E. 1964. Weeds and aliens. 2nd ed. London: Collins.
Sanchez, J., and F. Davis. 1969. Growth inhibitors in kikuyu (Pennisetum clandestinum) as factors of its competitive ability. [Abstract] I Semm. Soc. Colomb. de Contr. des Malezas y Fisiol. Vegetal (COMALES), 23-24, 58-59.
Sauer, J.D. 1982. Cayman Island seashore vegetation: a study in comparative biogeography. Univ. Calif. Pub. Geogr., Vol. 25.
Simmonds, H.W. 1933. The biological control of the weed Clidemia hirta D. Don. in Fiji. Entomol. Res. Bull. 24:345-348.
Simpson, B.B., ed. 1977. Mesquite: its biology in two desert scrub ecosystems. U.S. Internatl. Biol. Prog. Synth. Ser. 4. Stroudsburg, Penn.: Dowden, Hutchinson and Ross.
Smith, C.W., T. Parman and K. Wampler. 1980. Impact of fire in a tropical submontane seasonal forest. In Proc. 2nd Conf. Sci.- Res. Natl. Parks. Vol. 10, Fire Ecology, 313-324. San Francisco, Calif.: Natl. Park Serv.
Sorensen, J.C. 1980. Phenology of Andropogon virginicus in Hawai‘i. M.S. Thesis, Univ. Hawai‘i, Honolulu.
Spicer, R.A., and D. Mc. Newbery. 1979. The terrestrial vegetation of an Indian Ocean coral island: Wilingili, Addu Atoll, Maldive Islands. 1. Transect analysis of the vegetation. Atoll Res. Bull. 231.
Stoddart, D.R. 1967. Summary of the ecology of coral islands north of Madagascar (excluding Aldabra). In Ecology of Aldabra Atoll, Indian Ocean, ed. D.R. Stoddart, 53-62. Atoll Res. Bull. 118.
Stoddart, D.R. 1971. Land Vegetation of Diego Garcia. In Geography and ecology of Diego Garcia Atoll, Chagos Archipelago., ed. D.R. Stoddart and J.D. Taylor, 127-142. Atoll Res. Bull. 149.
Stoddart, D.R., and F.R. Fosberg. 1981a. Bird and Dennis Islands, Seychelles. Atoll Res. Bull. 252.
Stoddart, D.R., and F.R. Fosberg. 1981b. Topographic and floristic change, Dry Tortugas, Florida 19041977. Atoll Res. Bull. 253.
Stoddart, D.R., and C.A. Wright. 1967. Geography and ecology of Aldabra Atoll. In Ecology of Aldabra Atoll, Indian Ocean., ed. D.R. Stoddart, 11-52. Atoll Res. Bull. 118.
Stoddart, D.R., P.E. Gibbs, and D. Hopley. 1981. Natural history of Raine Island, Great Barrier Reef. Atoll Res. Bull. 254.
Stone, C.P. Alien animals in Hawai‘i's native ecosystems: toward controlling the adverse effects of introduced vertebrates. [This volume]
van der Werff, H. 1979. Conservation and vegetation of the Galapagos Islands. In Plants and Islands, ed. D. Bramwell, 391-404. London: Academic Pr.
Vogl, R.J. 1969. The role of fire in the evolution of the Hawaiian flora and vegetation. Proc. Ann. Tall Timbers Fire Ecol. Conf., April 1969, 5-60. Tallahassee.
Wace, N.M. 1966. The last of the virgin islands. Discovery 27:36-42.
Warshauer, F.R., J.D. Jacobi, A.M. La Rosa, J.M. Scott, and C.W. Smith. 1983. The distribution, impact and potential management of the introduced vine Passiflora mollissima (Passifloraceae) in Hawai‘i. Univ. Hawai‘i Coop. Natl. Park Resour. Stud. Unit Tech. Rep. 48. Honolulu: Univ. Hawai‘i.
Watt, T.A. 1978. The biology of Holcus lanatus L. (Yorkshire fog) and its significance in grassland. Herbage Abstracts 48:195-204.
Watts, D. 1970. Persistence and Change in the vegetation of oceanic islands. Can. Geogr. 14: 91-109.
Wester, L.L., and H.B. Wood. 1977. Koster's curse (Clidemia hirta), a weed pest in Hawaiian forests. Environ. Conserv. 4:35-41.
Whelan, B.R., and D.G. Edwards. 1975. Uptake of potassium by Setaria anceps and Macroptilium atropurpureum from the same standard solution culture. Austral. J. Agric. Res. 26:819-829.
Whistler, W.A. In press. Weed handbook for Western Polynesia. Eschborn, W. Germany: Geselischafr fur Techniche Zussamznarbeit.
Woodward, P.W. 1972. The natural historv of Kure Atoll, Northwestern Hawaiian Islands. Atoll Res. Bull. 164.