USDA Forest Service

Pacific Southwest Research Station

Pacific Southwest
Research Station

800 Buchanan Street
Albany, CA 94710-0011
(510) 883-8830
United States Department of Agriculture Forest Service. USDA logo which links to the department's national site. Forest Service logo which links to the agency's national site.

Research TopicsTropical Ecosystems

Invasive Species in Hawaiʻi

Examining insect-damaged albizia plants near Limbangan, Indonesia
A Forest Service postdoctoral research entomologist and collaborator from Bogor Agricultural University examine insect-damaged albizia plants near Limbangan, Indonesia.

The Institute of Pacific Islands Forestry (IPIF) conducts terrestrial, riparian and marine research on invasive plant, animal and microbial species in the diverse ecosystems of Hawaiʻi and the US Affiliated Pacific.

The work seeks to:

  1. Understand impacts on invasion on native biodiversity, ecosystem structure and function, and dynamics in response to changing or novel disturbance regimes.
  2. Develop effective biological control agents.
  3. Increase ecosystem resilience against invasions through restoration and threat mitigation.
  4. Disseminate information on non-native invasive species of across the region.

Global collaborations in weed biocontrol

Biological control, the introduction of natural enemies from a pest’s native range in order restore balance to an invaded ecosystem, can provide long-term, large-scale, highly selective suppression of invasive species. Biocontrol of weeds has a long history of noteworthy successes around the world - a history that began in Hawaiʻi in 1902 when insect enemies of lantana were imported from Mexico. Our IPIF biocontrol program builds upon past efforts, with a unique focus on plants that invade native forests of Pacific islands. We prioritize target weeds that have broad environmental impacts and for which other control methods are not effective.

International collaborations are essential to biocontrol. Forest Service scientists and partners in the Pacific work within the close-knit community of international biocontrol specialists to raise awareness and meet needs for Pacific island resource management. We lead multinational efforts developing and assessing biocontrols for the invasive trees, strawberry guava, miconia, and albizia, as well as clidemia, cane tibouchina, and related weeds. Additional targets for which we are supporting active biocontrol research include Himalayan ginger (Hedychium gardnerianum), Himalayan raspberry (Rubus ellipticus), Christmas berry (Schinus terebinthifolius), and faya tree (Morella faya).

Biocontrol of invasive melastomes
Melastome evaluation in the IPIF quarantine facility.

Melastome plants (family Melastomataceae) found in Hawaiʻi are all non-native and many are among the worst invasive species in Pacific island ecosystems. Following on surveys for natural enemies in Brazil and Costa Rica in the 1990s and early 2000s, our biocontrol program has focused on developing enemies which appear to hold the greatest promise for impacting miconia, clidemia and other weedy melastomes. Initial studies by teams of students and post-docs in the native range generated a wealth of information on biology, host range and impacts of prospective biocontrols. As resources and knowledge of each species’ biology have allowed, they have been brought to Hawaiʻi for evaluation in quarantine. Several species are in final stages of development as new tools for managing Miconia calvescens, Clidemia hirta, Tibouchina herbacea, Pterolepis glomerata, and Melastoma species.

Biocontrol of Strawberry Guava
Strawberry guava invasion and replacement of native Hawaiian forest.

Without natural enemies or competitors in Hawaiʻi, strawberry guava (Psidium cattleianum) grows quickly to form dense thickets that replace native forests. Its spread through hundreds of thousands of acres statewide has far exceeded capacity for control by conventional means. With colleagues in Brazil, where the plant is native, IPIF scientists imported and tested the highly specific leaf-galling insect Tectococcus ovatus. Following careful review T. ovatus was introduced to Hawaiʻi in 2012 to slow the spread of strawberry guava. We conduct further research to develop methods for incorporating the biocontrol into guava management and to determine its long-term impacts.

Albizia biocontrol
Potential biocontrol agent being evaluated for albizia.

Invasions by hazardous, ecosystem-altering albizia (Falcataria moluccana) across Hawaiʻi and other Pacific islands are so widespread that long term management will depend on biological control using host-specific natural enemies. Through exploratory surveys of albizia's natural enemies in its native range of Indonesia and Papua New Guinea, we have discovered a variety of arthropods that belong to taxonomic groups that tend to be highly specific, and therefore more likely to pass screening to determine their safety for biocontrol. Species with potential for limiting the spread and growth of albizia have been prioritized for further study. These include gall forming eriophyid mites and stem boring weevils.

Aquatic invasive species
Non-native fish found in a Hawaiian stream

Monitoring efforts have shown that > 80% of plant and fish species in Hawaiian coastal wetlands are exotic, with exotic fish densities up to 15x greater than native fish. High densities of exotic plants and fish outcompete natives for space and resources and contribute to increased nutrient loading. Tools, such as the coanda screen below, are being tested to determine effective ways to prevent exotic fish from invading native habitats. High water use invasive trees such as strawberry guava (Psidium cattleianum) or kiawe (Prosopis palida) in Hawaiian watersheds or coastlines can also significantly reduce water input to streams or nearshore waters.

Grass Invasion and wildfire in Pacific Islands
Wildland fire in a non-native grass dominated landscape of Guam

Wildfires in the region are exacerbated by invasion of fire-prone grasses and shrubs, which now make up 25% of Hawaiʻi's land area and represent a growing threat to native ecosystems across the region. Wildfires have greatly reduced the extent of tropical dryland forests (< 10% their original cover), increased the threat of extinction for hundreds of Hawaiʻi's T&E species – most of which are island endemics, and caused enormous erosion with impacts to streams and near shore areas. Our research program provides basic scientific information and practical tools for managing and restoring tropical dry forest landscapes in the Pacific. Results benefit our Pacific Island partners by increasing capacity and knowledge to restore native forests, thereby reducing wildfire and enhancing habitat for threatened and endangered species.

ROD, Ungulates, Fences
Aerial image of fenced, ungulate-free sections of 'Ōhi'a forest (fences denoted by white boundaries), surrounded by unfenced, ungulate impacted forest zones. Colored dots denote recently deceased 'Ōhi'a trees symptomatic of Ceratocystis-induced Rapid 'Ōhi'a Death (ROD). Image clearly depicts the preponderance of 'Ōhi'a mortality in unfenced, ungulate-impacted forest zones and the virtual absence of mortality inside fenced ungulate-free forest, providing strong evidence regarding the efficacy of fences for protecting 'Ōhi'a forests from ROD. (Greg Asner, Arizona State University)

IPIF led research has shown that fenced areas where ungulates have been removed are less vulnerable to Rapid ʻŌhiʻa Death, while analyses of ʻōhiʻa monitoring plot data show lower mortality rates where low levels of pig activity exist compared areas of high pig activity. Further research is ongoing to determine the role of feral ungulates in transporting and/or transmitting ROD. Such information is informing resource manager's actions to prioritize effective ungulate management (i.e., fencing and removal) for ecologically important forests to reduce unnecessary tree damage and enhance a given forests' ability to resist new fungal infections. Research is demonstrating that fencing and ungulate removal, or effective ungulate suppression, is likely to be among the most effective and practical means to reduce the spread of ROD at meaningful landscape scales.

Weed Risk Assessment
Conceptual model of how the Hawai’i Pacific Weed Risk Assessment Tool works (HPWRA).

We adapted a program developed in Australia and New Zealand for use on Pacific high islands known as the Hawaiʻi Pacific Weed Risk Assessment and evaluated its predictions against expert opinions of species in the horticultural trade. The system identifies pest and non-pest species with about 87% accuracy. It is now being used in Hawaiʻi to both avoid use of potentially invasive species in the nursery and horticultural trade and to guide restoration practitioners on the use of non-native but non-invasive species for use in restoring highly degraded ecosystems. The Hawaiʻi Pacific Weed Risk Assessment tool can be found online.

Ceratocystis-induced Rapid ʻŌhiʻa Death
A native Hawaiian forest ravaged by Rapid ʻŌhiʻa Death (ROD).

ʻŌhiʻa lehua (Metrosideros polymorpha) is the dominant tree species in Hawaiian watersheds that supports enumerable native plant and animal species and provides abundant, clean water. Rapid ʻŌhiʻa Death (ROD) caused by the fungal pathogens Ceratocystis lukuohia and C. huliohia threaten Hawaiʻi's ʻŌhiʻa forests statewide and have caused extensive stand level mortality in more than 10% of Hawaiʻi Island's ʻŌhiʻa-dominated forests. Our research on this topic is focused on both understanding the impacts of the disease on Hawaiʻi's forests and developing effective means of managing the disease and preventing its spread.

Detector dogs
Dogs used to detect the scent of Ceratocystis in Hawaiʻi's ʻŌhiʻa trees.

Canines have very sensitive and specific olfactory capacities that can be employed effectively for scent discrimination detection tasks including forest and agricultural pathogens such as Raffaelea lauricola that cause Laurel Wilt and potentially Ceratocystis lukuohia and C. huliohia that cause ROD and threatenHawaiʻi's ʻŌhiʻa forests. Application of canine ROD detection paired with qPCR methods may help to inform how fast the fungus is spreading, the prevalence of asymptomatic trees, dispersal agents and movement around infected trees, and to aid inspectors and managers tasked with preventing the spread of this pathogen. Stopping the spread of ROD using scent discriminating canines could save the state millions in mitigation costs. Our IPIF research team has demonstrated that certified scent-discriminating canines are reliably capable of detecting C. lukuohia and C. huliohia (CF) known to cause Rapid ‘Ōhi‘a Death (ROD) amongst Ōhi‘a trees (Metrosideros polymorpha) in both controlled settings and across a range of environmental and geographic conditions. This research has resulted in the development of safe training aide technology, standardized protocols and best practices for training and deployment of precision scent discriminating canines for ROD. Our work also has implications for responding to new and emerging ecosystem and agricultural threats (i.e. rhinoceros beetles, fire ants, snakes) to Hawaiʻi.

Ecosystem-scale response to invasion by and removal of albizia (Falcataria moluccana)
Field crew in American Samoa documenting albizia impacts to native forests

Albizia (Falcataria moluccana) is a highly invasive tree species that impacts native ecosystems across the western Pacific. This fast-growing tree can reach heights of up to 150 ft and are prone to being knocked over in storms. In Hawaiʻi, albizia caused more than $1 million dollars in damage after a recent tropical storm. The albizia invasion of native ecosystems also significantly increases N content in invaded forest soils and streams draining those forests, creating habitat prone to invasion by other non-native species. Removal of albizia can result in rapid recovery of some native forests. For example, biomass of native Samoan tree species following removal of F. moluccana accumulated rapidly following 8 years of post-removal recovery. Although F. moluccana is a daunting invasive species, it exhibits characteristics that make it vulnerable to successful control: it is easily killed by girdling or herbicides, and its seeds and seedlings do not tolerate shade.