Iiwi (Drepanis coccinea)

Species Status Report

Pacific Islands Fish and Wildlife Office

U.S. Fish and Wildlife Service, Region 1

December 2016 ● Final v1.0

Introduction

The U.S. Fish and Wildlife Service was petitioned to list the Iiwi (Vestiaria coccinea) [now Drepanis coccinea] as an endangered or threatened species on August 24, 2010. On January 24, 2012, we published a finding that the petition presented substantial information indicating the listing of the species under the Endangered Species Act may be warranted; this finding initiated a status review of the species (77 FR 3423). We have prepared this summary document to help inform that status review. Here we present a review of the conservation status of the Iiwi, including: species abundance, distribution, and trends; stressors that may affect the Iiwi; conservation measures or regulations that may ameliorate those stressors; and projected future conditions for the species. We reviewed a large number of references to compile this report; however, for the discussion of population status and trend, we primarily draw upon a thorough and recent peer-reviewed report published by the U.S. Geological Survey (USGS), “Abundance, Distribution, and Population Trends of the Iconic Hawaiian Honeycreeper, the Iiwi (Vestiaria coccinea) throughout the Hawaiian Islands” (Paxton et al. 2013). For the climate change analysis portion of this status report, we relied particularly on two recently published papers, “Large-Scale Range Collapse of Hawaiian Forest Birds under Climate Change and the Need [for] 21st Century Conservation Options” (Fortini et al. 2015), and “Will a warmer and wetter future cause extinction of native Hawaiian forest birds?” (Liao et al. 2015).

This document is not intended to be an exhaustive review of all published scientific literature relevant to the Iiwi; rather, it is intended to capture and summarize the key points from the best scientific and commercial data available relevant to our evaluation of the current and future conservation status of the species.

This species status report was prepared by the staff of the U.S. Fish and Wildlife Service, Pacific Islands Fish and Wildlife Office, Honolulu, Hawaii; and the U.S. Fish and Wildlife Service, Pacific Region, Portland, Oregon.

Executive Summary

Executive Summary

The Iiwi is an Hawaiian forest bird in the endemic honeycreeper subfamily of the Fringillidae (finch family). It is found primarily in closed canopy, montane wet or montane mesic forests of tall stature, dominated by native ohia trees (Metrosideros polymorpha) or both ohia and koa trees (Acacia koa). Iiwi are nectarivorus and feed primarily on flowering ohia; ohia trees are also used for nesting.

Abundance, Distribution and Trend Although historical abundance estimates are not available, the Iiwi was considered one of the most common of the native forest birds in Hawaii by early naturalists, described as “ubiquitous” and found from sea level to the tree line across all the major islands. In the late 1800s, Iiwi began to disappear from low elevation forests, and by the mid-1900s, the species was largely absent from sea level to mid-elevation forests. Today Iiwi are no longer found on Lanai and only a few individuals may be found on Oahu, Molokai, and west Maui. Remaining populations of Iiwi are restricted to high-elevation forests on Hawaii Island, east Maui, and Kauai. The current population size of Iiwi rangewide is estimated as a mean of 605,418 individuals (range 550,972 to 659,864). Ninety percent of all Iiwi now occur on Hawaii Island, followed by east Maui (about 10 percent), and Kauai (less than 1 percent).

Iiwi population trends vary across the islands. The population on Kauai is in steep decline. Trends on Maui are mixed but generally appear to be in decline. On Hawaii Island, there is evidence for stable or declining populations on the windward side of the island. Apparent trends of increase have been documented on the leeward (Kona) side of Hawaii Islands, but these trends were inferred from a limited data set. Of the nine regions for which sufficient information is available for quantitative inference, five show strong or very strong evidence of declining populations; one, a stable to declining population; one, a stable to increasing population; and two, strong evidence for increasing populations. Four of the nine regions show evidence of range contraction. Overall, based on the most recent surveys (up to 2012), approximately 90 percent of remaining Iiwi are restricted to a narrow band of habitat between elevations of roughly 4,265 and 6,234 feet (ft) (1,300 and 1,900 meters (m)).

Stressors For the purposes of our evaluation, we look for evidence that a stressor, acting either singly or in concert with other factors, may be resulting in population-level impacts to the species. We evaluated many potential stressors to Iiwi and their habitat, including: stressors that may affect the extent or quality of their ohia forest habitat (ohia dieback, ohia rust, rapid ohia death, drought, fires, volcanic eruptions, nonnative plants, and feral ungulates), introduced diseases, predation by introduced mammals, competition with nonnative birds, ectoparasites, climate change, and the effects of small population size. Based on our assessment, disease is the primary driver in the ongoing declines in abundance and range of Iiwi, and climate change substantially exacerbates the impact of disease on the species. Some of the other stressors may

i

Executive Summary

have contributed to past declines in Iiwi, or may even have some negative effect on the species or its habitat today; however, of the additional stressors listed above, we found no information to suggest that any is currently a key factor in the ongoing declines in abundance and range of Iiwi.

We did, however, identify rapid ohia death (ROD), a type of Ceratosystis wilt fungal disease, as a potentially significant emerging habitat stressor. Based upon the most recent research, ROD-infected stands of ohia often show greater than 50 percent mortality initially and nearly 100 percent of trees in a stand succumb to the disease within 2 to 3 years. Rapid ohia death is presently reported only from the island of Hawaii; however, over roughly the last 5 years it has spread across the island, which is home to 90 percent of the Iiwi population, and in some areas, affected trees have been observed within the range of Iiwi. As of January 2016, ROD is estimated to have infected approximately 34,000 ac (13,759 ha), which is a 100 percent increase in affected area since 2015.

The best scientific data available consistently identifies introduced mosquito-borne avian diseases, including avian malaria and avian pox, —as the primary stressors driving the declines in abundance and distribution of Iiwi observed since the turn of the 20th century. Avian malaria is a disease caused by a blood parasite, and avian pox is caused by a virus; both are transmitted by the bite of the nonnative mosquito Culex quinquefasciatus, and both have serious impacts on native forest birds, including Iiwi. The two diseases often co-occur, and although avian malaria has received greater study, avian pox is likely also an important factor in Iiwi declines based on field observations and limited experimental studies in other native forest birds. Iiwi exhibit an extremely high mortality rate in response to avian malaria (95 percent) and are absent at low elevations where malaria is prevalent, despite the availability of otherwise suitable habitat. Both the life cycle of the mosquito vector and the development and transmission of the malaria parasite are temperature-limited; neither can be completed at cool temperatures, which prevail at high elevations. Iiwi are now found primarily in high elevation forests above 3,937 ft (1,200 m) that provide disease-free refugia, where malaria prevalence and transmission is only brief and episodic, or nonexistent, under current conditions. The Iiwi has not demonstrated any sign of developing resistance to avian malaria to date. Although there has been less research on the effects of avian pox virus, the limited results to date suggest it may also be a source of significant mortality. The fact that avian malaria and avian pox often infect the same individual simultaneously complicates the ability to discern the effects of each disease, as the two diseases may be acting both individually and synergistically.

Climate change is a stressor with potentially significant impacts on Iiwi when considered in conjunction with avian malaria. Air temperature in Hawaii has increased in the past century and particularly since the 1970s, with greater increases at high elevation. Increases are currently documented in (a) the elevation where the avian malaria parasite and its mosquito vector can survive and reproduce, and (b) the prevalence of avian malaria in forest birds at high elevation sites, where Iiwi are declining (e.g., on Kauai). The projections for future climate conditions in Hawaii describe a continued warming trend, especially at high elevations. The temperature barrier to the development and transmission of avian malaria will continue to move up in elevation in response to warmer conditions, resulting in the curtailment or loss of disease-free habitats for Iiwi. Three studies specifically address the future of native Hawaiian forest birds,

ii

Executive Summary

including Iiwi, in the face of the synergistic interactions between climate change and avian malaria:

• Benning et al. (2002) conclude that under optimistic assumptions (i.e., 3.6°F (2°C) increase by 2100), malaria-susceptible Hawaiian forest birds, including Iiwi, lose most of their low-risk habitat in the three areas considered in their projection of climate change impacts. For example, current disease-free habitat (below the 55°F (13°C) isotherm) at Hakalau Forest National Wildlife Refuge is reduced by 96 percent.

• Fortini et al. (2015) conducted a vulnerability assessment for 20 species of Hawaiian forest birds, based on an increase of 6.1°F (3.4°C) at higher elevations by 2100. All species were projected to suffer range loss as the result of increased transmission of avian malaria at higher elevations with increasing temperature. Iiwi was predicted to lose 60 percent of its current range by 2100, and climate conditions suitable for the species will shift up in elevation. Most of the remaining habitat for Iiwi would be restricted to Hawaii Island.

• Liao et al. (2015) generated temperature and precipitation projections under three alternative emissions scenarios, and projected future malaria risk for Hawaiian forest birds. Irrespective of the scenario modeled, by mid-century (beginning roughly 2040), malaria transmission rates and impacts to bird populations begin increasing at high elevations. By 2100, the increased annual malaria transmission rate for Iiwi was projected to result in population declines on the order of 70 to 90 percent for the species, depending on the emissions scenario.

Climate change may also exacerbate other stressors to Iiwi. Changes in the amount and distribution of rainfall in Hawaii likely will affect the quality and extent of mesic and wet forests on which Iiwi depend. However, changes in the trade wind inversion (which strongly influences rainfall) and other aspects of precipitation with climate change are difficult to model with confidence. In addition, potential increases in storm frequency and intensity in Hawaii as a result of climate change may lead to an increase in direct mortality of individual Iiwi and a decline in the species’ reproductive success. Currently, no well-developed projections exist of these possible synergistic effects.

Conservation Measures We evaluated regulatory or other measures in place today that are intended to ameliorate the stressors to Iiwi. Forest habitat protection, conservation, and restoration can benefit Iiwi by protecting and enhancing breeding and foraging areas for the species, and by reducing mosquito breeding sites. Efforts to restore and maintain large, continuous tracts of native forests have met with some success in this regard, especially when combined with fencing and ungulate removal. Although some local examples of forest restoration exist (e.g., at Hakalau Forest National Wildlife Refuge on Hawaii Island), we do not have any information to suggest that restoration efforts are currently planned on a large scale. Similarly, although some mosquito control efforts

iii

Executive Summary

are taking place on a localized basis (generally aimed at the mosquito responsible for transmission of dengue, Aedes aegypti, rather than the malaria vector, Culex), and new research is underway, methods for achieving large-scale eradication of mosquitoes in Hawaii do not yet exist.

Some high-elevation habitats currently occupied by Iiwi are owned and managed for the protection of native forests or native species (e.g., Hakalau Forest National Wildlife Refuge (NWR), Hanawi Natural Area Reserve (NAR) on Maui). However, much of the disease-free habitat currently within the boundaries of these managed areas will be lost as avian malaria moves upward in elevation in response to warming temperatures, or is already gone (e.g., Alakai Wilderness Preserve on Kauai is now within the malaria zone). No specific regulatory mechanisms exist that govern the protection of high altitude areas as refugia for the Iiwi and other Hawaiian forest birds to buffer them against the upslope advance of avian disease in association with rising temperature.

The United States has passed some regulations specifically intended to reduce the emission of greenhouse gases that contribute to climate change; the scope and effect of such regulations are limited, however. On a global scale, the United Nations Framework Convention on Climate Change (UNFCCC) met in December, 2015, to attempt to reach a binding agreement to reduce greenhouse gas emissions. All participating countries were asked to submit their Intended Nationally Determined Contribution (INDC) to cut net greenhouse gas emissions. Although the INDCs would slow future emissions growth, the UNFCCC has indicated that even if all of them were fully implemented and targets met, the goal of limiting the increase in global average temperature to 2°C (3.6°F) by the year 2100 would not be achieved.

Many of the efforts to ameliorate the primary stressors to Iiwi are still in the research and development stage, or are being implemented only on a small scale. As the primary stressor, avian malaria, continues to have impacts, and these impacts are increasing and exacerbated by climate change, we must conclude that no current conservation measures or regulations are sufficient to offset this stressor to the species.

Conclusion Introduced mosquito-borne disease, specifically avian pox and avian malaria, has been a primary driver in the declines and extinctions of many native Hawaiian forest birds since the late 1800s. Iiwi are known to be especially vulnerable, suffering particularly high mortality from avian malaria compared to other native bird species. The transmission of avian malaria is currently limited or absent at higher elevations, where temperatures are too cool for the development of the malaria parasite. With increasing temperature at high elevations resulting from global climate change, avian malaria is projected to continue moving upward into high- elevation forests that currently provide refuge from disease for Iiwi. Modeling of future conditions consistently predicts a significant loss of disease-free habitat for Iiwi, with consequent severe reductions in population size and distribution by the year 2100, although significant changes are likely to be observed as early as 2040. As the Iiwi’s numbers and distribution continue to decline, small, isolated populations may become increasingly vulnerable to the additional stressor of loss of their ohia forest habitat from rapid ohia death, as well as other

iv

Executive Summary environmental catastrophes and demographic stochasticity; this will particularly be the case should all remaining Iiwi become restricted to a single island (Hawaii Island), as some scenarios suggest.

v

Table of Contents Executive Summary ...... i Abundance, Distribution and Trend ...... i Stressors ...... i Conservation Measures ...... iii Conclusion ...... iv 1. Species Information ...... 1 Taxonomy ...... 1 Description ...... 1 Habitat Requirements, Foraging Behavior, and Seasonal Movement ...... 1 Breeding and Demography ...... 2 Natural Causes of Mortality ...... 2 Historical Distribution and Abundance - Summary ...... 2 2. Status - Abundance, Distribution, and Population Trends ...... 3 Iiwi Population Status—Methods ...... 3 Iiwi Population Status—Summary of Rangewide Results ...... 3 Iiwi Population Status by Island ...... 4 Iiwi Population Status – Island of Kauai ...... 4 Iiwi Population Status – Island of Oahu ...... 8 Iiwi Population Status – Island of Molokai ...... 8 Iiwi Population Status - Island of Maui ...... 10 Iiwi Population Status - West Maui ...... 10 Iiwi Population Status - East Maui ...... 10 Iiwi Population Status – Hawaii Island ...... 11 3. Potential Stressors Affecting the Status of the Iiwi ...... 23 Habitat Loss and Degradation—Overview ...... 23 Habitat Loss and Degradation—Impacts to Ohia Trees ...... 23 Habitat Loss and Degradation—Impacts to Ohia Trees—Ohia dieback ...... 24 Habitat Loss and Degradation—Impacts to Ohia Trees—Ohia Rust ...... 27 Habitat Loss and Degradation—Impacts to Ohia Trees—Rapid Ohia Death or Ohia Wilt ...... 30

Impacts to Ohia Trees—Rapid Ohia Death—Cause and Transmission ...... 30 Impacts to Ohia Trees—Rapid Ohia Death—Range and Effects ...... 31 Habitat Loss and Degradation – Impacts to Ohia Trees – Volcanic Eruptions ...... 32 Habitat Loss and Degradation—Impacts to Ohia Trees—Drought ...... 34 Habitat Loss and Degradation—Impacts to Ohia Trees—Fire and Invasive Grasses ...... 35 Habitat Loss and Degradation – Impacts to Ohia Forests – Feral Ungulates and Rats ...... 36 Habitat Loss and Degradation—Impacts to Ohia Forests—Competition with Nonnative Plants...... 40 Disease—Overview ...... 44 Disease—Avian Malaria and Avian Pox—History ...... 44 Disease—Avian Malaria—Development and Transmission...... 46 Disease—Avian Malaria—Temperature Dependence ...... 47 Disease—Avian Malaria—Mortality of Native Hawaiian Birds and High Susceptibility of Iiwi ...... 48 Disease—Avian Malaria—Lack of Resistance in Iiwi ...... 51 Disease—Avian Malaria—Seasonal Iiwi Movement to Lower Elevations Increase Disease Infection Risk ...... 52 Disease - Avian Pox ...... 53 Disease—Summary ...... 55 Predation ...... 56 Predation—Rats ...... 56 Predation—Feral Cats ...... 57 Competition ...... 58 Ectoparasites ...... 59 Small Population Size ...... 62 Climate Change ...... 63 Climate Change—Historical Record—Temperature ...... 64 Climate Change – Historical Record – Precipitation ...... 65 Climate Change –Temperature, Precipitation, and Elevation Tied to Avian Malaria Transmission ...... 68 Climate Change – Impacts to Iiwi with Documented Increase in Elevation of Malaria Transmission .... 68 Climate Change – Projected Future Temperature in Hawaii ...... 69 Climate Change – Projected Future Precipitation in Hawaii ...... 69 Climate Change – Projected Future Impacts to Iiwi Increased Elevation of Malaria Transmission ...... 70 Climate Change—Projected Increase in Severe Storms ...... 77

Climate Change—Effects on Ohia Forest Habitat ...... 78 Climate Change—Summary ...... 78 4. Conservation Measures to Address Stressors to Iiwi ...... 79 Conservation Measures to Address Habitat Loss and Degradation ...... 79 Impacts to Ohia Trees—Ohia Rust—Research and Conservation Measures ...... 79 Impacts to Ohia Trees—Rapid Ohia Death—Research and Conservation Measures ...... 79 Conservation Measures to Protect Ohia Forest Habitat ...... 80 Protection of High Elevation Habitats ...... 81 Iiwi Habitat Protection and Conservation – Restoration ...... 83 Conservation Measures to Address Avian Malaria ...... 83 Conservation Measures to Address Predation ...... 84 Conservation Measures to Address Competition ...... 85 Conservation Measures to Mitigate Climate Change ...... 85 Climate Initiatives – United States ...... 85 Climate Initiatives – International ...... 86 5. Summary ...... 90 6. References ...... 91 APPENDIX A ...... 110

Species Information

1. Species Information

Taxonomy

The Iiwi, Drepanis coccinea, is a nectarivorous honeycreeper in the family Fringillidae and the endemic subfamily Drepanidinae (which includes all Hawaiian honeycreepers) (Pratt et al. 2009, pp. 114, 122). The species was originally placed in the monotypic genus (a genus of only one species) Vestiaria. In 2015, the genus Vestiaria was merged with Drepanis based upon morphological similarities as revealed by recent analyses of specimens (Chesser et al. 2015, pp. 749; 757-758). The Iiwi is classified as a discrete species by the American Ornithologists’ Union (AOU 1998, p. 677).

Description

Iiwi are medium-sized forest birds (total body length is approximately 14 centimeters (cm) (5.5 inches (in)) with bright scarlet feathers, black wings and tail, and a small white patch on the inner secondary flight feathers. The bill is long, deeply decurved, and salmon in color. Juveniles are differentiated by their buff color with black spots, and shorter bills which change in color from dusky yellow to salmon as they mature (Fancy and Ralph 1998, p. 2). Iiwi are widely recognized as one of the most spectacular and iconic of the extant Hawaiian forest birds. Iiwi songs are complex and include variable creaks often described as a “rusty hinge” sound, whistles, or gurgling sounds, and sometimes mimic other birds (Fancy and Ralph 1998, p. 5; Hawaii Audubon Society 2011, p. 97).

Habitat Requirements, Foraging Behavior, and Seasonal Movement

Iiwi are found primarily in closed canopy, montane wet or mesic forests of tall stature, dominated by native ohia (Metrosideros polymorpha) or ohia and koa (Acacia koa) trees. Iiwi nest in ohia trees, and also depend on ohia as a primary food source. Iiwi are nectarivorous; their diet consists predominantly of nectar from the flowers of ohia, but they may also feed on Sophora chrysophylla (mamane), and plants in the lobelia family (Campanulaceae) (Pratt et al. 2009, p. 193), as well as opportunistic feeding upon insects and spiders (Fancy and Ralph 1998, pp. 4–5; ; Pratt et al. 2009, p. 193). During one study at three different test sites, researchers noted that Iiwi spent about 80-90 percent of their time foraging on ohia for nectar and insects (Fancy and Ralph 1998, p. 5). The Iiwi’s long, curved bill is thought to be the result of coevolution with native Hawaiian plants in the lobelia family, renowned for their long, curved corollas (Fancy and Ralph 1998, p. 4, and references therein). Many of Hawaii’s native lobelias are now extinct or occur in very low numbers, leading to the hypothesis that the observed reduction in Iiwi bill size during the past 100 years may be the result of a dietary shift. Iiwi now feed primarily on the open, non-tubular flowers of ohia (Fancy and Ralph 1998, p. 4).

Iiwi are strong fliers that move long distances to locate nectar sources, and are well known for their seasonal movements in response to the availability of flowering ohia and mamane for feeding (Fancy and Ralph 1998, p. 3; Kuntz 2008, p. 1; Guillamet et al. 2016, p. 192). Such movements generally occur after the breeding season. This seasonal movement to

1

Species Information lower elevation areas in search of nectar sources is an important factor in the exposure of Iiwi to avian diseases, particularly malaria (discussed below).

Breeding and Demography

On the islands of Hawaii, Kauai, and Maui (the only islands that still support more than just a few individuals of the species), Iiwi currently breed and winter above approximately 3,937 ft (1,200 m) in mesic and wet forests dominated by ohia and koa trees. Between years, individual birds change breeding sites, presumably to take advantage of nectar availability (Fancy and Ralph 1998, p. 9). Fidelity to a local breeding area is considered low to moderate (Fancy and Ralph 1998, p. 9; Kuntz 2008). The Iiwi breeding season starts as early as October and continues through to the following August (Fancy and Ralph 1998, p. 7). However, the majority of breeding occurs from February through June, coinciding with peak flowering of ohia (Fancy and Ralph 1998, p. 2). Iiwi nest sites are typically found in the upper canopy of ohia trees and the cup-shaped nests are comprised of twigs and lined with lichens and moss (Fancy and Ralph 1998, p. 8). The average clutch size is 2 eggs, and only one brood is normally reared per season, although Iiwi may renest if the first nest is unsuccessful (Fancy and Ralph 1998, p. 9). Remaining together during the season, breeding pairs defend a small area around their nest, and then disperse after breeding (Fancy and Ralph 1998, p. 2).

Survival of Iiwi was the lowest of six species of native Hawaiian forest birds observed during a six-year study at four sites on Hawaii Island (Ralph and Fancy 1995, pp. 729-730). However, because of their seasonal movements, re-sight probability is low for Iiwi, and low re- sighting estimates can lead to imprecise estimates of survival (Fancy and Ralph 1998, p. 9; Hart et al. 2011, p. 118). Population modeling by Samuel et al. (2015, p. 11) produced a similar estimate of annual survivorship for adults, on the order of 55 to 60 percent. Longevity in the wild is unknown (Fancy and Ralph 1998, p. 9).

Natural Causes of Mortality

Prior to the introduction of nonnative predators and avian diseases to Hawaii, the primary natural cause of Iiwi mortality is believed to have been nestling death from storms. Nestlings may die from exposure during storms or be blown from the nest during high winds (Woodworth and Pratt 2009, p. 211). (The Hawaiian hawk or io (Buteo solitarius) may sometimes take iiwi, and now-extinct native predators such as the Hawaiian harrier (Circus dossenus) and long-legged owl (Grallistrix spp.) may also have preyed on the species.) The daily survival rate of nests for almost all forest birds is reduced during wetter years (observed at Hakalau Forest NWR; Cummins et al. 2014, p. 17), and severe storms are documented as the primary cause of nest failure for many species of Hawaiian forest birds (Woodworth and Pratt 2009, p. 211; Cummins et al. 2014, p. 22).

Historical Distribution and Abundance - Summary

Although now restricted mostly to remote, high-elevation forests above 3,937 ft (1,200 m), the Iiwi was once described by early European visitors as one of the most common bird species on all the main Hawaiian Islands (Banko 1981, pp. 1-2). Many accounts by early naturalists in the late 1800s described the Iiwi as being abundant in all wooded areas from near

2

Species Information sea level to the tree line (Banko 1981, pp. 1-2). However, beginning in the early 1900s, accounts of Iiwi documented a steady decrease in abundance and a steady contraction in distribution from low-elevation forests to high-elevation forests (Banko 1981, pp. 2-3). By 1905, Iiwi were reported as the rarest of the five species of native mountain birds found on Oahu (Fancy and Ralph 1998, p. 2). By the 1940s, Iiwi numbers were greatly reduced, with the species entirely extirpated from Lanai by 1929 and absent, or nearly so, on Molokai (Munro 1944, p. 94). By the middle of the 20th century, Iiwi had vanished from many mid-elevation areas. Iiwi, which were noticeably common in the 1940s in the main visitor area of Hawaii Volcanoes National Park (elevation 3,937 ft (1,200 m)), had retreated by the 1970s to higher elevation forests (elevation 5,577 ft (1,700 m)) (Banko 1981, pp. 27-29), and are now reported as rare or absent throughout much of the park (Fancy and Ralph 1998, p. 3; Judge et al. 2011, pp. 23–24).

2. Status - Abundance, Distribution, and Population Trends

The following population status summary is drawn from the USGS Open-File Report 2013-1150, “Abundance, Distribution, and Population Trends of the Iconic Hawaiian Honeycreeper, the Iiwi (Vestiaria coccinea) throughout the Hawaiian Islands” (Paxton et al. 2013). We refer the reader to that report for full details.

Iiwi Population Status—Methods

Paxton et al. (2013) analyzed data from population surveys over the past four decades to determine abundance, distribution, and population trends of Iiwi. Between 1976 and 1983, the Hawaii Forest Bird Survey (Scott et al. 1986) established a quantitative baseline from which changes in bird populations on Hawaii, Maui, Lanai, Molokai, and Kauai could be ascertained. Since the start of the Hawaii Forest Bird Survey, data from 649 separate surveys conducted across the main Hawaiian Islands using point-transect surveys have been entered in the Hawaii Forest bird Monitoring Database (Camp 2016, pers comm.).

Population size estimates for each island or region were obtained by extrapolating mean bird density for surveyed areas to the area of available habitat within the species’ range (densities were calculated from point-transect distance sampling data). For this reason, population estimates are reported as mean values. Importantly, to address the sensitivity of these estimates to shifts in sampling effort and location over time, the researchers delineated consistently sampled areas (CSAs). The CSAs allowed for the reporting of trends based on a standard metric, defined in the study as change in density over a modeled period of 25 years (to adjust for variability in survey effort, and the observed annual rate of change for each survey period was applied over a period of 25 years. Note, this was a generic period used for modeling trends between areas, not observed trends over a specific historical time period of 25 years). The researchers categorized observed population trends as increasing, decreasing, negligible (i.e., a stable population), or inconclusive (trend uncertain). The evidence for a particular trend was interpreted as weak (P < 0.7), strong (0.7 < P < 0.9), or very strong (P > 0.9).

Iiwi Population Status—Summary of Rangewide Results

3

Abundance, Distribution, and Population Trends

Most Iiwi are currently restricted to high-elevation montane forests of Hawaii Island (90 percent), followed by east Maui (about 10 percent), and Kauai (less than 1 percent), with only relict populations, if any, persisting on Oahu, west Maui, and Molokai. Rangewide abundance of Iiwi is estimated with a 95 percent confidence interval (CI) between 550,972 and 659,864 individuals, with a mean of 605,418 individuals. Ninety percent of Iiwi now live in a narrow band of native forest between 4,265 ft and 6,234 ft (1,300 and 1,900 m) elevation in closed- canopied, high-stature forest dominated by ohia trees (Paxton et al. 2013, pp. 5-6). Iiwi population trends vary across the islands (see Figure 1). The Iiwi population on Kauai has experienced sharp declines, with a projected trend of 92 percent decline over a 25-year period based on the 2000–2012 surveys. On East Maui, the northeastern region has experienced declines (34 percent over a 25-year period), while the southeastern region has been stable to moderately increasing. On Hawaii Island, population trends are mixed. On the windward side, populations are largely declining, although the northern section (Hakalau Forest) has stable populations. On the leeward side, results suggest a strongly increasing population, with estimates of up to a 147 percent increase over a 25-year period from the Puu Waawaa region. However, it is not clear whether these results from the leeward side of Hawaii show a likely population trend that is contrary to population trends in all other areas, or if they are an artifact of a sparsely sampled area (Paxton et al. 2013, p. 1).

Trends by elevation suggest a large decrease in numbers of Iiwi at elevations below 1,200 m (3,937 ft) on Kauai and northeast Maui. Low-elevation Iiwi populations also appear to have decreased in other regions, although low-elevation areas are not surveyed as often as other areas because of their lack of native forest birds. An exception to this pattern was the lower portion of the Kona Unit of Hakalau Forest NWR in the central leeward part of Hawaii Island, where populations appeared stable at the lower elevations. Based on the most recent analyses of survey data (up to 2012), approximately 90 percent of Iiwi today live in a narrow band at elevations of 4,265 – 6,234 ft (1,300–1,900 m). Iiwi persist mainly in montane forest, with 61 percent in montane wet forest, 35 percent in montane mesic forest, 3 percent in lowland wet-mesic forest, and remaining habitat constituting less than 1 percent of their distribution (Appendix E, Paxton et al. 2013).

Of the nine regions of the Iiwi’s current range for which sufficient information is available for quantitative inference, five showed strong or very strong evidence of declining populations, one stable to declining population, one stable to increasing population, and two, strong evidence for increasing population. Four of the nine regions showed evidence of range contraction. In general, however, the lower elevation Iiwi populations are declining, and even within most high elevation areas, populations are showing evidence of declines. Central Kona on Hawaii Island is one area where low elevation populations appear to be stable or increasing, and the entire Kona coast stands out as an area of stable to increasing populations (Paxton et al. 2013, p. 11). See Figure 1 below, which presents the distribution, mean population size, and trends by region for Iiwi across its range in Hawaii.

2

Abundance, Distribution, and Population Trends

Figure 1. Iiwi distribution, mean population size (with 95-percent confidence interval), and population trends by region across its range in Hawaii. See Table 1, below for mean abundances. (From Paxton et al. 2013).

3

Abundance, Distribution, and Population Trends

Table 1. Range size and mean abundance (with 95-percent confidence interval) by islands and region for Iiwi throughout the Hawaiian Islands. The abundance estimates are the mean of estimates derived separately from elevation and vegetation classifications (Appendix E), with 95-percent CIs. Abbreviations: ac, acre; ha, hectare; %, percent; CI, confidence interval; SE, standard error. (From Paxton et al. 2013, except where noted).

Area Area Mean Lower Upper Island/region (ac) (ha) abundance 95% CI 95% CI Kauai* 13,343 5,436 2,603 1,789 3,520 Oahu unknown unknown Few1 Maui, east 32,620 13,201 59,859 54,569 65,148 Maui, west 4,663 1,887 1762 Molokai 4,448 1,800 Few3 Hawaii (all regions) 432,039 174,840 543,009 516,312 569,706 Hawaii, north 61,593 24,926 277,055 258,075 296,035 windward Hawaii, central 100,752 40,773 71,524 62,662 80,386 windward Hawaii, Kau 83,225 33,680 28,325 23,138 33,512 Hawaii, south 30,861 12,489 3,489 2,059 4,918 Kona Hawaii, central 62,866 25,441 139,829 124,649 155,009 Kona Hawaii, north 52,463 21,231 22,787 18,444 27,130 Kona Kohala 13,838 5,600 8022 Mountains Mauna Kea 10,378 4,200 4822 Individuals, total4 605,418 550,972 659,864 * Estimate of iiwi numbers on Kauai from Paxton et al. 2016. 1 VanderWerf 2016, pers. comm. See discussion below under Iiwi population status – Island of Oahu 2 Estimates from Scott et al. (1986) surveys 3 VanderWerf 2016, pers. comm. See discussion below under Iiwi population status – Island of Molokai 4 Total estimates fom Paxton et al. 2013 and do not include Scott et al. 1986 or small changes to Kauai estimates between Paxton et al. 2013 and Paxton et al. 2016.

Iiwi Population Status by Island

Iiwi Population Status – Island of Kauai

Paxton et al. (2013, p. 13) describe the Iiwi population on Kauai as in rapid decline, with range and abundance of populations, as well as total number of birds, decreasing at a fast rate. Bird density across combined regions, the Alakai Plateau core (interior CSA), and surrounding perimeter (exterior CSA), showed a very strong decreasing trend (P > 0.9) from 2000 to 2012; this equates to a 92 percent reduction in the size of the Kauai Iiwi population over a modeled 25- year period. The trend is particularly acute for the exterior region, with a 97 percent decline over

4

Abundance, Distribution, and Population Trends

25 years (see Figure 2, below). As of 2012, the total population size was estimated at 2,551 (range 1,934 to 3,167 individuals). An analysis of Iiwi survey detections as a function of elevation indicates that decreases in the species’ range on Kauai are concentrated at lower elevations. The Iiwi’s range on Kauai has contracted from 40,525 ac (16,400 ha) in the period between 1968 and 1973 to approximately 13,433 ac (5,436 ha) in 2012, or roughly a two-thirds loss of range (Figure 3). Decreases were most significant below 4,042 ft (1,232 m) in elevation; most Iiwi are now found in montane wet forest from 3,609 to 4,265 ft (1,100 to 1,300 m) in elevation (Figure 4). (Paxton et al. 2013, pp. 13–16).

Figure 2. Trends in Iiwi density (mean birds per hectare at 95-percent confidence interval) on Kauai, Hawaii, for the Alakai Plateau (interior CSA) and surrounding perimeter (exterior CSA) (per Fig. 3) (note different survey time periods). (From Paxton et al. 2013).

5

Abundance, Distribution, and Population Trends

Figure 3. Extent of Iiwi range on Kauai, demonstrating recent range contraction, based on surveys completed during 1968–1973, 40,278 ac (16,300 ha) blue outline; in 2000, 24,958 ac (10,100 ha), orange outline; and in 2012 13,590 ac (5,500 ha), red outline. Sites where Iiwi were detected (solid circles) and not detected (open circles) are shown for the 1968–1973 survey. Shading represents elevation from low elevation (browns) to high elevations (greens). (From Paxton et al. 2013).

6

Abundance, Distribution, and Population Trends

Figure 4. Iiwi distribution and abundance on Kauai showing density per station values (points) derived from survey counts conducted in 2012. The vegetation types and elevation contours shown correspond to strata for which density and abundance were estimated for the area within the species’ range (red outline). (Image from Paxton et al. 2013).

7

Abundance, Distribution, and Population Trends

Iiwi Population Status – Island of Oahu

Iiwi may occur as a small remnant population on Oahu, if at all (Paxton et al. 2013, p. 10). Iiwi were noted in decline on Oahu by the early 1900s, continuing through the 1970s, and by 1991 the species was not detected during an island-wide survey. Based on a range of values representing remnant population sizes and statistical confidence, estimates of the probability of persistence of a small relict Iiwi population on Oahu were produced from historical data, indicating with 95-percent probability that the population in 1991 was less than 922 individuals (Paxton et al. 2013, p. 19). A small number of Iiwi (at least six individuals) was observed on Oahu in the West Koolau Mountains in 1996 at 1,805 ft (550 m) (VanderWerf and Rohrer 1996, pp. 1–2), with the last detection of a single bird in the Koolau Mountains in 2001. In the Waianae Mountains of Oahu, Iiwi has been more recently and consistently observed, indicating either local breeding on Oahu or movement of birds between islands. At Palikea, located in the southern Waianae Mountains, a single juvenile bird was observed each year in between 2006 and 2009 (a different bird each time), and an older bird was observed in January 2013 (VanderWerf pers. comm. 2016). Other than these sightings on Palikea, no other Iiwi have been observed on Oahu during hundreds of hours of incidental searching and field time in both the Koolau and Waianae Mountains each year (VanderWerf pers. comm. 2016).

Iiwi Population Status – Island of Molokai

Iiwi may occur as a relict population on Molokai, if at all (Paxton et al. 2013, p. 10). Only a remnant population was believed to persist on Molokai through the last century. Estimates of an Iiwi population on Molokai were produced from historical data, indicating with 95-percent probability that the Molokai population during a 1979 survey was less than 141 individuals (Paxton et al. 2013, p. 19). Very few Iiwi have been detected on Molokai during official point-transect surveys between 1998 and 2004: the most recent detection during an official survey was a sighting of three birds in 2004, while the most recent official point-transect survey in 2010 failed to locate Iiwi (Paxton et al. 2013, p. 19). However, very recent Iiwi observations during incidental surveys indicate that a small population or a few individuals persist on Molokai. In December 2013 one bird was sighted at Kamakou, three birds were counted in December 2014 at Waikolu, and one bird was sighted again at Waikolu as recently as December 2015 (Dibben-Young in litt. 2016, p. 1). See Figure 5, below, depicting the distribution of Iiwi on Molokai.

8

Abundance, Distribution, and Population Trends

Figure 5. Iiwi distribution on Molokai (top panel), and West Maui (bottom panel), based on most recent surveys (2004 for Molokai, 2010 for west Maui). (Image from Paxton et al. 2013).

9

Abundance, Distribution, and Population Trends

Iiwi Population Status - Island of Maui

Iiwi distribution on the island of Maui occurs within two distinct and disjunct regions, one on the upper slopes of the West Maui Mountains and the other on the northeast and southeast windward slopes of Haleakala on East Maui. The West and East Maui populations are 19 mi (30 km) apart and separated by a wide expanse of highly altered lowland vegetation unsuitable for Iiwi.

Iiwi Population Status - West Maui

A very small population of Iiwi may persist on West Maui, if at all. A total of 5, 2, and 11 Iiwi were detected on West Maui in 1980, 1997 and 2010, respectively. Due to the small data set, no quantitative inference is determinable about the current population size or distribution.

Iiwi Population Status - East Maui

East Maui is estimated to support approximately 10 percent of the remaining Iiwi population rangewide, with an estimated mean of 59,859 individuals (range 54,569 to 65,148). Surveys have been done in two consistently surveyed areas (CSAs) in East Maui (Figure 7); however, one site located in the northeast was studied from 1980 to 2011, while a second CSA site located in the southeast was studied from 2000 to 2012. The Iiwi population showed mixed trends between these two CSAs. Bird density in the northeast CSA showed strong evidence (0.7 < P < 0.9) of a decline between 1980 and 2011, equivalent to a 34-percent reduction in population density over 25 years. Densities observed for the southeast CSA, on the other hand, showed a stable to increasing trend (P < 0.7) from 2000 to 2012, equivalent to a 22-percent increase in population density, see Figure 6 below.

Figure 6. Trends in Iiwi density (mean birds per hectare and 95-percent confidence interval) on Maui, within the Northeast and Southeast CSAs are depicted. (Image from Paxton et al. 2013).

10

Abundance, Distribution, and Population Trends

Analysis of Iiwi survey detections as a function of elevation indicated a contraction in range at lower elevations in the northeast CSA. In contrast, Iiwi occurrence at lower elevations of their distribution has not changed for the southeast CSA, even though Iiwi occurrence at higher elevations had decreased. Paxton et al. (2013, p. 17) hypothesize that declines in Iiwi occurrence at the highest elevations may indicate different trends in habitat quality between the drier margins of montane forests at higher elevations and wetter forests at mid elevations, as well as localized elevation-dependent changes in precipitation.

Iiwi Population Status – Hawaii Island

Hawaii Island supports an estimated 90 percent of the rangewide Iiwi population. The estimated current total Iiwi population size on Hawaii Island is a mean of 543,000 individuals (range 516,312 to 569,706 individuals). Iiwi are distributed on both the windward and leeward sides of the island. For the purposes of their study, Paxton et al. (2013) delineated the species’ range into an area encompassing 432,674 ac (175,097 ha) and six regions comprised of three windward regions, including north (Hakalau), central, and south (Kau) windward, and three leeward regions including south Kona, central Kona, and north Kona, each of which also includes one or more CSAs (see Figure 9 and additional discussion below). Trend results were highly variable among the six regions on the Hawaii Island (Figure 10). Excluded from the Paxton et al. (2013) analysis were regions of about 13,838 ac (5,600 ha) in the Kohala Mountains and 10,378 ac (4,200 ha) on Mauna Kea volcano (as estimated by Scott et al. 1986, p. 64). Iiwi are likely sparsely distributed across southwest Mauna Kea based on more recent surveys between 2008 and 2012. Iiwi persistence on Mauna Kea is related to flowering of the mamane tree (Hess et al. 2001, p. 154), and while a small resident population is possible (Paxton et al. 2013, p. 21), it is also likely that many Iiwi observed on Mauna Kea may be transitional and visiting temporarily to exploit this resource. The Kohala Mountain region was last officially surveyed in 1979 at which time only 10 Iiwi were detected there. At that time, the population was estimated at 802 birds (± 286 SE) (Scott et al. 1986, p. 64). According to Hawaii State Natural Area Reserve biologists, Iiwi have been consistently heard and observed each year between 1997 and 2016 within the small ohia forest (968 ac (392 ha)) located above 5,000 ft (1,524 m) in elevation surrounding the two summits, Puu O Umi (5,260 ft (1,603 m)) and Kaunu O Kaleiho Ohie (5,480 ft (1,670 m)). An official survey for Iiwi on the Kohala Mountain summit is planned for 2016 (Nicholas Agorastos, Hawaii NARS, pers comm. 2016).

11

Abundance, Distribution, and Population Trends

Figure 7. Boundaries of the two East Maui CSAs used for assessing trends in density and range decreases of Iiwi on east Maui are depicted. The red outline depicts the northeast CSA and the blue line depicts the southeast CSA. (Image from Paxton et al. 2013).

12

Abundance, Distribution, and Population Trends

Figure 8. Iiwi distribution and abundance on East Maui showing density per station values (points) derived from survey counts conducted in 2011 and 2012. The vegetation types and elevation contours shown correspond to strata for which density and abundance were estimated for the area within the species range (red outline). (Image from Paxton et al. 2013).

13

Abundance, Distribution, and Population Trends

Figure 9. Boundaries of Hawaii Island Iiwi CSAs: Species range is shown in red, with blue outlines designating some specific CSAs (not all CSAs are depicted). (Image from Paxton et al. 2013).

14

Abundance, Distribution, and Population Trends

Figure 10. Trends in Iiwi density (mean birds per hectare at 95-percent confidence interval) in the Hakalau, Keauhou, Mauna Loa, the lower and upper portions of the central Kona and the Puu Waawaa CSAs of the Big Island are depicted. The shaded band represents the 95-percentconfidence interval of the trend for the entire time series. Note that the 1977 and 1978 HFBS survey densities were not included in the trend assessment because of differences in the time of year during which it and all subsequent surveys were conducted. (Image from Paxton et al. 2013).

15

Abundance, Distribution, and Population Trends

Iiwi Population Status – Hawaii Island – North Windward (Hakalau) (Figures 9 and 11)

• Stable to declining

The north windward (Hakalau) region encompasses a 61,593 ac (24,926 ha) area with an estimated population ranging between 258,075 and 296,035 birds (mean of 277,055 birds). Density in the 5,577 to 6,234 ft (1,700 to 1,900 m) elevation range for the region was estimated at 13.9 to 18.4 birds per ha (mean = 16.1), constituting the highest values recorded for Iiwi across its range (Figure 11). Iiwi density within the Hakalau CSA appears to be stable-to- declining; that is, the slope for the 1999 to 2012 survey period showed strong evidence (0.7 < P < 0.9) of a stable trend and weak evidence (P < 0.7) for a declining trend. The trend would result in an average Iiwi population decrease of 20 percent over a 25 year period. The surveyed portion of Hakalau occurs above 4,600 ft (1,400 m), where disease transmission is rare or absent, therefore iiwi showed no changes in distribution during the 1977-2012 surveys within the CSA (Paxton et al. 2013, p. 28).

Iiwi Population Status – Hawaii Island –Central Windward (Figures 9 and 12)

• Declining

The central windward region encompasses a 100,752 ac (40,773 ha) area with an estimated population ranging between 62,662 and 80,386 birds (mean of 71,524 birds). There are distinct areas of Iiwi distribution within this region including two CSAs, Keauhou CSA and Mauna Loa CSA (Figures 9 and 12). Iiwi density within the Keauhou CSA showed strong evidence (0.7 < P < 0.9) of a negative trend, which would result in an average decline of 48 percent over a 25 year period. Iiwi in the Mauna Loa CSA showed highly variable densities with a stable-to-downward trend and projected decrease of 29 percent over 25 years. Despite the declines in density apparent in the central windward region, Iiwi occurrence did not show a corresponding elevation-range decrease within the Keauhou and Mauna Loa CSAs. However, both of these areas lie almost entirely above 4,921 ft (1,500 m) elevation, and consequently appear to be situated above the zone in which changes in distribution are evident (Figure 12).

Iiwi Population Status – Hawaii Island –South Windward (Kau) (Figures 9 and 13)

• Declining

Trends in bird populations in south windward (Kau) and the south Kona regions were determined partly by comparing results for the 1976 to 1978 Hawaii Forest Birds Survey (HFBS: Scott et al. 1986) and more recent surveys. The south windward (Kau) region encompasses an 83,225 ac (33,680 ha) area with an estimated Iiwi population ranging between 23,138 and 33,512 birds (mean of 28,325 birds), based on the densities observed in 2004, 2008, and 2010, with some bird detections occurring as low as 2,950 ft (900 m). Compared to data for the 1976-1978 HFBS, densities from the more recent 2004 to 2010 surveys were significantly lower for all three Kau CSAs (north, central, and south): HFBS surveys in 1976 within the Kapapala Forest Reserve (north Kau CSA) showed Iiwi to be about twice as abundant in 1976 as they were in

16

Abundance, Distribution, and Population Trends

2004; Iiwi densities were three-fold higher in 1976 than in 2008 for the central Kau CSA; and while Iiwi were uncommon in south Kau during the HFBS in 1978, no Iiwi were observed in the south Kau CSA in the South Windward region in 2005 and 2010 (Figure 13).

Iiwi Population Status – Hawaii Island – South Leeward Kona (Figures 9 and 14)

• Declining

The South Kona region encompasses 30,861 ac (12,489 ha) with an estimated Iiwi population ranging between 2,059 and 4,918 birds (mean of 3,489 individuals). The South Kona area, with an elevation gradient ranging between 3,150 ft and 5,380 ft (960 and 1,640 m), showed a greater than 20-fold decrease in Iiwi density when comparing the 1978 HFBS surveys to 2009 surveys (2.6 [+ 0.4 SE] versus 0.1 [+ 0.1 SE] birds per ha). Comparisons from the Honomalino area focused on the elevation gradient of 1,360-1,810 m, indicate that density may have diminished in this area from 2005 to 2010 (0.7 [+ 0.2 SE] versus 0.4 [+ 0.2 SE] birds per ha).

Iiwi Population Status – Hawaii Island – Central Leeward Kona (Figures 9 and 14)

• Increasing

The Central Kona population region encompasses 62,866 ac (25,441 ha) with an estimated Iiwi population ranging between 124,649 and 155,009 birds (mean of 139,829 individuals). The Central Kona CSA spans an elevation gradient between 3,280 and 4,920 ft (1,000 and 1,500 m) in the lower range and 4,920 to 6,233 ft (1,500 to 1,900 m) in the upper range. Iiwi density within both the lower and upper portions of the CSA appears to have increased during the 1995 to 2012 survey period. Slopes for both elevation areas showed strong evidence of positive trends (0.7 < P < 0.9). The average of these trends over a 25 year period would result in a 71 and 97 percent increase in population size for lower and upper elevations, respectively. The distribution of Iiwi (presence vs. absence) from 1978 to 2012 shows no decreases in Iiwi distribution at low elevations in the Central Kona CSA. Habitat below the lower portion of the CSA was not surveyed following the 1978 HFBS surveys, during which Iiwi were detected as low as 1,312 ft (400 m). The HFBS surveys were conducted during summer months, and these detections may have been post-breeding Iiwi dispersing to lower elevations. Paxton et al. (2013) suggest that as elsewhere in leeward Hawaii, it is unlikely that the current Iiwi range still includes habitat below elevations of 3,280 ft (1,000 m) (Figure 14).

17

Abundance, Distribution, and Population Trends

Iiwi Population Status – Hawaii Island – North Leeward Kona (Puu WaaWaa) (Figures 9 and 14)

• Increasing

The North Kona region encompasses 52,463 ac (21,231 ha) with an estimated Iiwi population ranging between 18,444 and 27,130 birds (mean of 22,788 individuals) based on survey results from 2003 to 2009. Iiwi density within the Puu Waawaa CSA in the North Kona region appears to have increased during the 1990-2009 survey period. There is strong evidence (0.7 < P < 0.9) of a positive trend in the North Kona population, which would result in an average increase of 147 percent over a 25-year period. Results from 1978 to 2009 surveys demonstrate no decreases in Iiwi distribution at lower elevations in the Puu Waawaa CSA, which is at approximately 3,937 ft (1,200 m) (Figure 14).

Paxton et al. (2013, pp. 4, 29-30), did not explain the population increases in the Central and North Kona Iiwi population regions, but one hypothesis is that drier conditions in Central and North Kona during a decade-long drought may have limited mosquito breeding sites (Paxton pers. comm., 2015); or possibly that Iiwi within these populations are developing disease tolerance (see “Iiwi Stressors – Disease” for discussion of this latter concept).

18

Abundance, Distribution, and Population Trends

Figure 11. Iiwi distribution and abundance in the North Windward (Hakalau) region of Hawaii Island showing density per station values (points) derived from survey counts conducted in 2012. The vegetation types and elevation contours shown correspond to strata for which density and abundance were estimated for the area within the species range (red outline). (Image from Paxton et al. 2013).

19

Abundance, Distribution, and Population Trends

Figure 12. Iiwi distribution and abundance in the Central Windward region of Hawaii Island showing density per station values (points) derived from survey counts conducted in 2010 and 2012. The vegetation types and elevation contours shown correspond to strata for which density and abundance were estimated for the area within the species range (red outline). (Image from Paxton et al. 2013).

20

Abundance, Distribution, and Population Trends

Figure 13. Iiwi distribution and abundance in the South Windward (Kau) region of Hawaii Island showing density per station values (points) derived from survey counts conducted in 2004, 2008, and 2010. The vegetation types and elevation contours shown correspond to strata for which density and abundance were estimated for the area within the species range (red outline). (Image from Paxton et al. 2013).

21

Abundance, Distribution, and Population Trends

Figure 14. Iiwi distribution and abundance in the Kona region of Hawaii Island, depicted separately for the North Kona region (left panel) and the Central and South Kona regions (right panel) and showing density per station values (points) derived from survey counts conducted in 2003 and 2009 (North Kona), 2009 and 2010 (Central Kona), and 2003, 2009, and 2010 (South Kona). The vegetation types and elevation contours shown correspond to strata for which density and abundance were estimated for the area within the species range (red outline). (Image from Paxton et al. 2013).

22

Potential Stressors

3. Potential Stressors Affecting the Status of the Iiwi

Habitat Loss and Degradation—Overview

• Iiwi require native ohia forest habitat, which currently is degraded by nonnative plants, disease, and feral ungulates.

Iiwi depend on areas of intact native ohia forest for foraging and nesting. Iiwi generally are found only within forests comprised of predominantly native vegetation, and today their density is greatest at high elevations where the risk of avian malaria is lowest (Ralph and Fancy 1995, p. 740; Camp et al. 2010, pp. 198, 206; Paxton et al. 2013, p. 28). Historically, the Iiwi’s habitat was reduced and fragmented by the clearing of ohia forests by native Hawaiians, and later by American and European settlers, for food crops, sugar production, grazing, and other development, particularly below 4,000 ft (1,250 m) (Scott et al. 1986, pp. 371–373). The estimated loss of Iiwi’s forest habitat since human contact ranges from 52 percent on Hawaii Island to 85 percent on Oahu (Mountainspring 1986, p. 98). Although very little of the remaining upper elevation Iiwi habitat is at risk of clearing for development today, the species’ habitat continues to be negatively affected by nonnative species and compounding factors including diseases. Invasive plants outcompete and displace native ohia used by Iiwi for foraging and nesting. Feral ungulates, particularly pigs (Sus scrofa), goats (Capra hircus), and axis deer (Axis axis), degrade ohia forest habitat by spreading nonnative plant seeds and grazing on and trampling native vegetation, contributing to erosion, and creating mosquito breeding habitat (Mountainspring 1986, p. 95; Camp et al. 2010, p. 198). In addition to the effects of nonnative plants and animals on ohia and its habitat, ohia forest is impacted by several diseases and natural processes including ohia dieback, ohia rust, and ROD, the effects of which will likely be compounded by each other and by the effects of nonnative species and climate change.

Habitat Loss and Degradation—Impacts to Ohia Trees

• Ohia trees and forest habitat required by Iiwi are negatively affected by drought, volcanic eruptions, wildfire, diseases, wood boring beetles, and nonnative plants and animals.

Ohia remains Hawaii’s most common and widespread native tree, occurring from sea level to 8,200 ft (2,500 m) elevation in dry and wet forests and comprising about 62 percent of the total forest area on all of the major Hawaiian Islands (Hodges et al. 1986, p. 1) (see also Figure 15). The species is the most ecologically important native Hawaiian tree, defining native forest succession and ecosystem function over broad areas (Dawson and Stemmermann 1990, p. 967; Pratt and Jacobi 2009, p. 40).

23

Potential Stressors

Figure 15. Ohia Distribution by Moisture Zone (Image from USGS)

The health and distribution of ohia forest is affected by natural and nonnative processes and agents, such as volcanic eruptions, wildfire, extreme weather events, wood boring beetles (native and nonnative), nonnative invasive plants, and diseases. While ohia dieback and ohia rust do not currently appear to cause broad-scale rangewide effects to ohia trees (see detailed discussion below), there is growing concern that the impacts of dieback and rust could be exacerbated by the cumulative effects of stressors such as extreme drought, climate change, and invasive nonnative plants, such as strawberry guava (Psidium cattleianum) and albizia trees (Falcataria moluccana), which can prevent or retard regeneration of ohia forest. The combined effects of drought and nonnative, invasive grasses have resulted in increased fire frequency and the conversion of mesic ohia woodland to exotic grassland, particularly in sub-montane Hawaii Volcanoes National Park. More recently, researchers have recognized that rapid ohia death (ROD), caused by a species of Ceratosystis fungus, is an emerging disease that has the potential to rapidly kill large numbers of ohia trees over extensive areas of Hawaii Island.

Habitat Loss and Degradation—Impacts to Ohia Trees—Ohia dieback

• Ohia dieback is a complex, largely natural process that affects older and stressed ohia trees in a manner similar to cohort senescence. It poses a potential indirect risk to Iiwi should ohia dieback possibly be exacerbated or accelerated unnaturally, due, for

24

Potential Stressors

example, to climate change or any number of new or existing ohia diseases, including ohia rust or rapid ohia death, resulting in loss of habitat.

Iiwi rely on native ohia trees for nesting and foraging, thus any phenomenon that negatively affects ohia forest may negatively impact the amount of habitat available to Iiwi in a given area, depending upon the scale and severity of its impact. With ohia dominating 60 percent of the relatively intact remaining native forest in the Hawaiian Islands, the subject of dieback is an important issue. Attention to the phenomenon of ohia ‘dieback’ began as a topic of tremendous ecological concern after researchers working on Hawaii Island first began to notice it in the late 1960s. After 40 years of research, scientists concluded that ohia ‘dieback’ is an essentially a natural process in which small to relatively large stands of mature ohia trees age and die simultaneously (see Figure 16). We now know that ohia dieback is due in part to the species’ life history and is also a response triggered by a complex interplay of environmental factors and stressors. The most current and thorough explanation and summary of ohia dieback can be found in the book “Ohia Lehua Rainforest,” by Műeller-Dombois et al. (2013, 269 pp.), which served as a primary source in our review.

Researchers now mostly agree that the primary factor behind ohia dieback is the species’ trait of experiencing synchronized generational turnover following cohort senescence. In other words, in many areas ohia grow in large stands comprised of trees all approximately the same age (cohorts). These stands become large ohia trees, which create a mostly closed canopy that precludes its own shade-intolerant progeny from completing development and becoming part of the canopy, except for instances where canopy openings, natural or otherwise, are created. Eventually, all of these same-aged trees reach a point of natural senescence and lose their vigor, followed by either simultaneous slow or rapid death (Figure 17). With the demise of the canopy above, the underlying seedlings are finally able to obtain sufficient sunlight to reach sapling stage, thus continuing the ohia forest lifecycle. Ohia dies periodically over part of its range in response to a variety of environmental changes (Műeller-Dombois 1986, pp. 238–239). No native species is capable of replacing it as a dominant, however, and ohia seedlings usually reinvade vigorously (Jacobi et al. 1983, pp. 327, 336).

Researchers believe this process can be hastened or affected by numerous factors, which may act independently or in combination. These factors include many natural native elements such as volcanic activity, both localized and extreme, wide-ranging weather events such as drought or excessive rain, different soil condition issues including soil age and corresponding changes in nutrients or aluminum toxicity (a trait of volcanic soils), root fungus, and the effects of endemic wood boring beetles. Anthropogenic factors may influence the process as well, including the activities of feral pigs, the effects of nonnative plants, road building, impacts from nonnative wood-boring beetles, and diseases such as ohia rust, the latter of which was almost certainly introduced inadvertently by the nursery industry (Kilgore and Heu 2007, pp. 1–2; Anderson 2012, p. 1). Any of these factors acting independently or in conjunction, could trigger or exacerbate an ohia dieback event.

25

Potential Stressors

Figure 16. Showing ohia dieback within Hawaii Volcanoes National Park (photo by Jim Jacobi, USGS)

Pertaining to Iiwi, there is cause for concern despite the natural, evolved circumstances surrounding ohia dieback. First, it has been documented that loss of the ohia canopy in some areas has resulted in an observed decrease in population of native birds (Burgan and Nelson 1972, p. 1; Hodges et al. pp. 11, 21), along with an increased number of introduced birds (Hodges et al. pp. 11, 21). If we expect that ohia dieback and ‘growback’ (Mueller-Dombois et al. 2013, pp. 145–149) would continue as a natural element of native Hawaiian forest succession, then the temporary displacement of forest bird populations would not be cause for concern. However, because the Iiwi already appears to be restricted to an increasingly narrow band of suitable habitat, the real risk of ohia dieback, even over the short-term, is the unforeseen compounding effects with ever-increasing numbers of nonnative plants and animals altering the ohia dieback and regrowth cycle, which results in reduced suitability for Iiwi in its limited remaining habitat. Numerous studies (Mueller-Dombois et al. 2013, pp. 157–161; Mueller and Boehmer 2013, pp. 5,178, 5,181; Schulten et al. 2014, p. 267) have described and established these impacts, ranging from reduced recruitment ability of ohia in forests invaded by feral ungulates and nonnative weeds, to the complete inability for ohia to reseed in some weed- invaded areas due to lack of sunlight. In addition, the unknown compounded effects of anticipated climate change and nonnative species acting in concert upon the ability of ohia forests to naturally recover from these previously benign events may also result in loss or degradation of ohia habitat for the Iiwi (see below, discussion on Climate Change).

26

Potential Stressors

Figure 17. Photo showing cohort and landscape-level ohia dieback (photo by Jim Jacobi, USGS)

Habitat Loss and Degradation—Impacts to Ohia Trees—Ohia Rust

• The strain of ohia rust currently present in Hawaii likely causes very little impact to ohia trees. The risk to Iiwi include the indirect effect of a possibly more potent strain being introduced, and or the possibility of ohia rust acting in concert with any number of other ohia stressors such as drought, climate change, or rapid ohia death to compound cumulative effects resulting in overall ohia forest decline.

Ohia rust is a plant pathogen caused by the fungus species Puccinia psidii, which affects hundreds of plants in the Myrtaceae family including Eucalyptus spp., Melaleuca spp., and Hawaii’s native ohia (Figure 18). Rust caused by P. psidii was first documented in Brazil on guava (Psidium guajava) trees in the 1800s. It is still a serious pest of guava, and remains known internationally as “guava rust.” Since its discovery in 1884, research on P. psidii has revealed an ever expanding host range within the Myrtaceae, affecting hosts originally throughout much of South and Central America and the Caribbean, and now worldwide (Anderson 2012, p. 1). Puccinia psidii includes several strains that affect different species differently, and is now of serious economic and conservation concern due to mortality and other impacts to over 94 myrtaceous plants, including 16 endangered species in Australia alone. Symptoms of P. psidii rust on affected plants most often occur in tender, young growing points, and first begin as tiny bright yellow powdery eruptions in circular patterns on the leaf or stem surface. These infection loci or spots expand and develop into dying plant tissue. Leaves and

27

Potential Stressors stems may be deformed by the disease and growing tips can die back if the infection is severe (HDOA 2007, pp. 1–2; Anderson 2012, pp. 1–2; Silva et al. 2014, p. 1; see Figure 18).

Puccinia psidii was first diagnosed in Hawaii in 2005 on nursery grown ohia on the island of Oahu, and the rust spread quickly to become ubiquitous on all the islands (HDOA 2007, pp. 1–2). To date, Hawaii hosts only a single known strain of P. psidii, and although recently reported on multiple myrtaceous hosts on all of the islands, this strain has thus far caused only mild levels of damage to the State's ohia (Anderson 2012, pp. 1–2; Silva et al. 2014, p. 1) (Figure 19).

Figure 18. Left, early onset of ohia rust symptoms (left) and advanced disease conditions (right). (Image from Hawaii Department of Agriculture. (M. Killgore and R.A. Heu). 2007. Ohia Rust Puccinia psidii New Pest Advisory Factsheet No. 05-04, 2 pp. Used with permission)

Besides the possibility of reduced vigor to ohia from the present strain of Puccinia psidii compounded by other stressors, the larger risk to ohia and thus, the Iiwi’s habitat, is that a new, more virulent strain will be inadvertently introduced to Hawaii through the importation of plants or plant material by the nursery and horticulture industry (see Conservation Measures to Address Habitat Loss and Degradation section below). As noted above, Hawaii quarantine specialists suspect the existing strain was brought to Oahu by the nursery and horticulture industry (Kilgore and Heu 2007, pp. 1–2). Multiple strains of P. psidii have been identified from Brazil and characterized via extensive sampling and microsatellite analyses. Experiments have been conducted in Brazil to investigate a variety of potential effects of different P. psidii strains on Hawai'i's ohia (Silva et al. 2014, pp. 1–2). Assessments included pathological impact to seedlings, variation in seedling susceptibility, and the influence of the rust disease on growth and survival of ohia seedlings. Two of the five tested P. psidii strains investigated were not particularly virulent, but three strains were highly virulent on most of the inoculated ohia seedlings (93 percent–100 percent infection rates), and none of the ohia used in the test showed significant resistance to the rust. Infection by the highly virulent strains of P. psidii resulted, on average, in a 69 percent reduction in height growth and 27 percent increase in mortality of ohia seedlings at 6 months post-infection (Silva et al. 2014, pp. 1–2).

28

Potential Stressors

Figure 19. Ohia rust (Puccinia psidii) survey records and disease index for ohia on Hawaii for the years 2005– 2010. (Image from Anderson 2012; used with permission).

29

Potential Stressors

Habitat Loss and Degradation—Impacts to Ohia Trees—Rapid Ohia Death or Ohia Wilt

• Ohia wilt, commonly known as rapid ohia death or ROD in Hawaii, is a new, potentially significant, and poorly understood ohia disease occurring on Hawaii Island. With rapid spread and high stand mortality, all indications suggest that this particular ohia disease could alone, or in conjunction with other stressors such as climate change, have far- reaching negative consequences for ohia forests on the island harboring 90 percent of the remaining Iiwi.

In the past few years, a potentially serious new risk factor has emerged for the ohia habitat upon which the iiwi depends. In 2012, landowners in Puna District of Hawaii Island began reporting that previously healthy-appearing ohia trees were dying within a few weeks of all of their leaves inexplicably turning brown. In 2014, trees in more areas were found to be afflicted, and the range of the affected trees had spread to include the District of Hilo and areas near the Hawaii Volcanoes National Park (HAVO) (Figure 22). Pathogenicity tests conducted by the USDA Agriculture Research Service determined that the the disease, now commonly known in Hawaii as rapid ohia death (ROD), is a vascular wilt caused by the fungus Ceratocystis fimbriata (Keith et al. 2015, pp. 1–2). In 2015, a second species of Ceratocystis was found to be killing ohia; this new species is being described now (Keith 2016, pers. comm).

Impacts to Ohia Trees—Rapid Ohia Death—Cause and Transmission

Ceratocystis fimbriata has been present in Hawaii as a pathogen of sweet potato for decades (Brown and Matsuura 1941 in Friday et al., p. 1), but recent genetic and other research indicates that the strains of the fungus that are ohia pathogens are new; ohia is not simply a new host for the strain known from sweet potato (Keith 2016, pers. comm.). The disease spreads from tree to tree or from forest stand to forest stand by many means, and some affected trees have been observed in areas 10 mi (16 km) from the nearest road (F. Hughes, USFS, pers. comm. 2016). In other plant hosts of Ceratocystis species, such as sweet potato, cacao, mango, and eucalyptus, the fungus is moved by the feeding of insects, soil, water, infected cuttings, pruning wounds, or tools; in other words, by extensive means of transmission. Researchers have found Ceratocystis in soils under infected ohia stands in Hawaii, as well as in frass (sawdust-like feces produced by wood-boring beetles). Thus, contaminated soil, windblown frass, and infected wood may all be contributing to transmission of the disease. Some researchers have expressed concern that feral ungulates including pigs and cattle may also be spreading the disease from one area to the next (F. Hughes, USFS, pers. comm. 2016), and recent observations in areas north of Hakalau Forest NWR indicate that feral cattle damage to the trunks and roots of ohia trees may be exacerbating the spread of the infection in that area (F. Hughes, USFS, pers. comm. 2016).

30

Potential Stressors

Figure 20. Ohia tree dying of rapid ohia death (Image from www.rapidohiadeath.org; used with permission)

Impacts to Ohia Trees—Rapid Ohia Death—Range and Effects

Rapid ohia death affects non-contiguous ohia forest stands ranging in size from <1 ac (1 ha) up to 247 ac (100 ha), with nearly100 percent of trees infected (Friday et al. 2015, p. 1). The largest affected area is located within the Puna District of Hawaii Island, where infected trees have been observed within approximately 4,000 discontinuous acres (1,619 ha). Rapid ohia death has not yet been reported on any of the other Hawaiian Islands. In 2014, approximately 6,000 ac (2,430 ha) of ohia forest from Kalapana to Hilo on Hawaii Island had experienced greater than 50 percent mortality (Friday et al. 2015, pp. 1–3). Ohia stands experience rapid and extensive mortality to ROD. Annual mortality rates measured in monitoring plots averaged from 24 percent (measured as stems) to 28 percent (measured as basal area) between 2014 and 2015 (Mortenson et al. 2016, p. 89). When these plots were established in the ROD-infected area in January and February of 2014, all had already experienced an average of approximately 39 percent ohia mortality (Mortenson et al. 2016, p. 89). Ohia does experience mortality as a result of other diseases, including the nonnative ohia rust (described above), which only kills ohia seedlings, and a natural stand-senescence phenomenon known as ohia dieback. However, the dramatic, landscape-scale increase in ohia mortality is ascribed to ROD (Mortenson et al. 2016, pp. 88, 90). Other trees in the forest, including both native and nonnative species such as Psychotria spp. (kopiko), Polyscias spp. (ohe mauka), Psidium cattleianum (strawberry

31

Potential Stressors

guava), Melastoma spp., and Clidemia hirta (Koster’s curse) are not known to be affected by the fungus.

In late October 2015, samples from dead trees from Holualoa and Kealakekua on the Kona (leeward) side of Hawaii were confirmed to be infected with ROD (CTAHR 2015, http://www2.ctahr.hawaii.edu/forestry/disease/ohia_wilt.html). Researchers had hoped, previously, that ROD might be restricted to the wetter, windward side of Hawaii Island. The elevation range of affected trees is reported to range from sea level up to approximately 5,000 ft (1,524 m) at Wailuku forest near Hakalau Forest NWR, which contains a healthy Iiwi population (stable to increasing; Paxton et al. 2013), and some areas occupied by the Iiwi are already affected by ROD (Hughes, 2016a, pers. comm.). In July of 2016, the area infected with ROD was estimated at 34,000 ac (13,760 ha) (Hughes 2016a, pers. comm.) As of November 2016, the amount of forest area affected on Hawaii Island was estimated to be approximately 50,000 ac (20,235 ha), and increase of more than 47 percent within one year (Hughes 2016b, pers. comm.; Keith 2016, pers. comm.). With its rapid spread across Hawaii Island (see below), observed high mortality within affected stands, and lack of control measures, ROD poses a potentially serious risk to remaining iiwi habitat.

Figure 21. Range of rapid ohia death on Hawaii Island as of December 23, 2015. (Image from University of Hawaii, http://www2.ctahr.hawaii.edu/forestry/disease/ohia_wilt.html)

Habitat Loss and Degradation – Impacts to Ohia Trees – Volcanic Eruptions

32

Potential Stressors

• As an entirely natural process, volcanic eruptions have been impacting ohia forest on Hawaii Island for millions of years. There is likely very little effect on Iiwi except for the possibility of some small amount of habitat loss due to lava flows.

Hualalai, Moana Loa, and Kilauea are the three most active volcanoes on Hawaii Island. A study in 1980 of nine species of native plants and three exotic plant species exposed to volcanic emissions from gas vents at Pauahi Crater, Hawaii Volcanoes National Park, found ohia was extremely resistant to the formation of visible leaf injury when exposed to volcanic sulfur dioxide (Winner and Mooney 1980, pp. 389–391). Lava flows do have the potential to destroy some forest area utilized by Iiwi. The North Kona Iiwi population is situated on the flanks of Hualalai, which last erupted in 1801. Several Iiwi populations occur on the various flanks of Mauna Loa which has erupted 33 times since 1843. Land area covered by lava during the large 1950 and 1984 Moana Loa eruptions was 112 km2 (43 mi2) and 48 km2 (18.5mi2), respectively. This indicates there is the potential for some loss of up to 1 percent of habitat area for Iiwi on Hawaii Island from lava inundation as result of a possible future large Mauna Loa eruption. Kilauea and associated vents have been erupting continuously since 1983; this area is only historical Iiwi habitat. There are no other islands with active volcanoes in the Iiwi’s range.

Figure 22. Volcanic eruption hazard map for Hawaii Island. (Image from John G. Van Hoesen, USGS)

33

Potential Stressors

Habitat Loss and Degradation—Impacts to Ohia Trees—Drought

• Localized drought impacts ohia trees, particularly those growing in less than optimal soil or at the margins of precipitation tolerance. The indirect effect on Iiwi involves the possibility of wider-ranging impact to ohia forests from prolonged and extreme drought events, such as that observed on Hawaii Island during the past 12+ years.

Ohia ranges in elevation from just above sea level to 8500 ft (2,590 m), is variable in form (for example, it often appears stunted and miniaturized on ridgetops), and is somewhat adapted to a wide variety of soil and precipitation conditions (Hodel and Weissich 2012, pp. 40– 41; Mueller-Dombois et al. 2013, pp. 36–43; Cavaleir et al. 2014, pp. 8–10). While ohia trees do not grow in coastal areas with annual rainfall less than 20 in (500 mm) (Adee and Conrad 2014, p. 1), well-established ohia trees with deep root systems can persist in higher elevation, drier areas receiving less than 16 in (400 mm) of rain annually (Friday and Herbert 2006, p. 2). Conversely, ohia trees growing within shallow lava soils in wetter areas can be killed by drought within a span of just a few weeks (Friday and Herbert 2006, p. 8). Since approximately 1970, precipitation in the Hawaiian Islands, including Hawaii Island, has been declining (Chu and Chen 2005, p. 4,801). According to the Hawaii Drought Monitor, from 2008 through 2014, Hawaii Island experienced an extreme and prolonged drought event brought on by the Pacific Decadal Oscillation (PDO), an episodic variation in sea-surface temperature that affects weather patterns across parts of North America and Hawaii. As fluctuating weather patterns cause winter storms to migrate northward, away from the Hawaiian Islands, the result is often less annual precipitation than is normal and prolonged drought (Giambelluca et al. 2013, p. 315). Recent climate models predict further climate drying for Hawaii that may affect ohia distribution in dry and mesic forest areas.

34

Potential Stressors

Figure 23. Mean November rainfall for Hawaii Island for 2011, showing the extreme 10-year drought conditions experienced. (Image from Giambelluca et al. 2013)

Habitat Loss and Degradation—Impacts to Ohia Trees—Fire and Invasive Grasses

• It is widely established that the spread and encroachment of nonnative grasses into and along the margins of native ohia forest leads to their decline by facilitating wildfires, which further open up areas for the spread of additional nonnative plants and their impacts.

Fire is a relatively new, human-related impact to the native Hawaiian forests, including dry to mesic ohia forests inhabited by the Iiwi. The pre-human, historical fire regime in Hawaii was originally characterized by infrequent, low severity fires (Cuddihy and Stone 1990, p. 91; Smith and Tunison 1992, pp. 395–397), with few natural ignition sources beyond those caused by eruptions or lightning strikes, the latter of which is comparatively rare in Hawaii (McCarthy 2014, pp. 1–3). Following the use of fire by the early Hawaiians to clear native vegetation in the drier plains and foothills (Kirch 1982, pp. 5–6, 8; Cuddihy and Stone 1990, pp. 30–31), European and American settlers in the 1800s introduced many plants and animals, which resulted in further degradation to native Hawaiian ecosystems and contributed to a new wildfire regime in Hawaii (D’Antonio and Vitousek 1992, p. 67; Smith and Tunison 1992, pp. 395–397; Vitousek et al. 1997, pp. 7–8; D’Antonio et al. 2011, p. 1,617).

35

Potential Stressors

Fires of all intensities, seasons, and sources are destructive to native Hawaiian ecosystems (Myers 2000, p. 173), and a single grass-fueled fire can kill most native trees and shrubs in the burned area (D’Antonio and Vitousek 1992, p. 74). Few native Hawaiian plants and animals are adapted to withstand fire, and none are known to depend on fire for their existence or regeneration. Mueller-Dumbois (1981) (in Cuddihy and Stone 1990, p. 91) stated that most natural vegetation types of Hawaii would not carry fire before the introduction of alien grasses, and Smith and Tunison (1992, p. 394) point out that native plant fuels typically have low flammability. The activities and impacts of feral ungulates contribute to the fire cycle by increasing erosion and removing native plant cover by grazing, which reduces soil moisture levels, all of which increase the likelihood of grass invasion into formerly intact native forest and shrubland (Banko et al. 2013, pp. 71, 76; Hess 2014, p. 22).

Grasses (particularly those that produce mats of dry material or retain a mass of standing dead leaves) often invade native forests and shrublands, providing fuels that allow fire to burn progressively further into the outer edges of native forests that might not otherwise easily burn (Fujioka and Fujii 1980, in Cuddihy and Stone 1990, p. 93). Native woody plants may recover from fire to some degree, but fire typically favors nonnative grasses, which aggressively re- sprout and crowd out less competitive native plants (National Park Service 1989 in Cuddihy and Stone 1990, p. 93; Smith and Tunison 1992, p. 398). Many nonnative invasive plants, especially fire tolerant grasses, outcompete native plants and inhibit their regeneration (D’Antonio and Vitousek 1992, pp. 70, 73–74; Tunison et. al. 2002, p. 122). Several species of nonnative grasses, where established, are widely documented to fuel a grass/fire cycle of intrusion into Hawaii’s native ohia forests, further degrading biodiversity. These invasive grasses include Melinus minutiflora (molasses grass) (O’Connor 1999, p. 1.562; Cuddihy and Stone 1990, p. 89); Axonopus fissifolius (O’Connor 1999, pp. 1,500–1,502; Cook et al. 2005, p. 4); Pennisetum setaceum (fountain grass); Andropogon virginicus (broomsedge), and Schizachyrium condensatum (beardgrass) (Tunison et al. 2011, p. 122).

Habitat Loss and Degradation – Impacts to Ohia Forests – Feral Ungulates and Rats

• Feral nonnative ungulates and invasive nonnative plants are the leading causes of the degradation of remaining native forests throughout Hawaii. Feral pigs, goats, deer, sheep, and cattle damage native forests by consuming native plants, grazing and trampling, causing erosion, spreading nonnative invasive plants, and creating habitat for mosquitoes. Among the species of feral ungulates that occur throughout the Hawaiian Islands, feral pigs probably have the most widespread and detrimental impacts to the mesic and wet ohia forest on which the Iiwi depends.

Pigs (Sus scrofa) Pigs have been described as the single most pervasive and disruptive nonnative influence on the unique native forests of the Hawaiian Islands, and are recognized as one of the key sources of degradation in the wet and mesic ohia forest ecosystems on which Iiwi depend (Mountainspring 1986, p. 95; Aplet et al. 1991, p. 56; Loope et al. 1991, p. 19; Anderson and Stone 1993, p. 195; Loope 1999, p. 56; Camp et al. 2010, p. 198). Feral pigs are established on all of the islands where Iiwi occur or occurred until recent decades including on Kauai, Oahu, Molokai, Maui, and Hawaii Island, where their trampling and rooting disturbs and destroys

36

Potential Stressors

native plant cover. Pigs also reduce or eliminate plant regeneration by damaging or eating seeds and seedlings.

Pigs are a major vector for the establishment and spread of invasive nonnative plant species, by dispersing plant seeds on their hooves and coats and in their manure, which also fertilizes the disturbed soil. Pigs feed preferentially on the fruits of particularly invasive nonnative plants, such as banana poka and strawberry guava, and disperse their seeds. In addition, rooting pigs contribute to erosion by clearing vegetation and creating large areas of disturbed soil, especially on slopes (Smith 1985, pp. 190, 192, 196, 200, 204, 230–231; Stone 1985, pp. 254–255, 262–264; Medeiros et al. 1986, pp. 27–28; Scott et al. 1986, pp. 360–361; Tomich 1986, pp. 120–126; Cuddihy and Stone 1990, pp. 64–65; Aplet et al. 1991, p. 56; Loope et al. 1991, p. 19; Wagner et al. 1999, p. 52).

In addition to contributing to habitat degradation, feral pig activity produces soil compaction, wallows, and downed, hollowed-out native tree ferns (Cibotium spp.) that hold water and create breeding sites for mosquitoes, which transmit avian disease (Scott et al. 1986, pp. 365–368; Atkinson et al. 1995, p. S68), creating mosquito larval breeding sites in areas where these would not occur naturally. The geological and hydrological nature of many Hawaiian landscapes naturally and normally precludes the production of larval stage habitat for mosquitoes. However, Culex quinquefasciatus mosquitoes, the sole vector for avian malaria in Hawaii, are now abundant in many wet forests where their larvae rely primarily on habitats created by pigs including compacted volcanic soils and wallows and downed, hollowed-out tree ferns knocked over and consumed by pigs for their starchy pith (Scott et al. 1986, pp. 365–368; Atkinson et al. 1995, p. S68). As a consequence, one of the most common and currently feasible conservation measures identified for the control of avian malaria is management of feral pigs (LaPointe 2006, pp. 1–3; LaPointe 2008, p. 600; Ahumada et al. 2009, p. 354; Atkinson and LaPointe 2009b, p. 60; Samuel et al. 2011, p. 2,971).

In studies of native forest plots where feral ungulates including pigs were removed by trapping and other methods, researchers have demonstrated a reduction in the abundance of Culex spp. mosquitoes when comparing pig-free, fenced areas to places where feral pig activity was high. Evidence from these studies indicated a relationship between pig activities and the abundance and distribution of C. quinquefasciatus. Aruch et al. 2007 (p. 574), LaPointe 2006 (pp. 1–3) and LaPointe et al. (2009, p. 409; 2012, pp. 215, 219) confirm and assert that management of feral pigs may be strategic to managing avian malaria and pox, particularly in remote Hawaiian rain forests where studies have documented that habitat created by pigs are the most abundant and productive habitat for larval mosquitoes. LaPointe (2006, p. 3–4; 2008, p. 600) suggests that reduction in mosquito habitat must involve pig management across large, contiguous landscapes due to the tremendous dispersal ability of C. quinquefasciatus. Because of their ability to disperse up to 1 mi (1.6 km) through closed canopy forest, a substantial amount of the managed area is easily invaded and occupied by mosquitoes originating from adjacent areas not managed for feral pigs, including for example, suburban and agricultural areas where abundance of C. quinquefasciatus mosquitoes is much greater (LaPointe 2006, p. 3; LaPointe 2008, p. 600; LaPointe et al. 2009, p. 409). See Figure 24 below, which highlights an example of this concept as observed in the relatively large northern and southern units of the Hakalau Forest NWR.

37

Potential Stressors

Figure 24. This figure depicts both the main, southern unit and the northern (Maulua) unit of the Hakalau Forest NWR and the extent of each hypothetically invaded by Culex quinquefasciatus from surrounding non-pig-managed lands (see red shading). Even though mosquito breeding sites are greatly reduced in both units as a result of pig control, 60 percent of the main (southern) unit and 100 percent of the northern unit of the Refuge is theoretically occupied by mosquitoes originating from outside of the unit boundaries given that Culex spp. mosquitoes can disperse an average of 1 mile (1.6 km). (Image from LaPointe, D.A. 2006. Feral pigs, introduced mosquitoes, and the decline of Hawaiian birds. U.S. Geological Survey, Fact Sheet 2006–3029).

Goats (Capra hircus) Native to the Middle East and India, goats were introduced to the Hawaiian Islands in the late 1700s. Feral goats now occupy a wide variety of habitats on Kauai, Oahu, Molokai, Maui, and Hawaii Island where they consume native vegetation, trample roots and seedlings, accelerate erosion, and disperse the seeds of invasive plants in their fur, on their hooves, and through their feces (van Riper and van Riper 1982, pp. 34–35; Stone 1985, p. 48). Feral goats consume most plants, and are instrumental in the decline of native vegetation in many areas (Cuddihy and Stone 1990, p. 64). Goats are able to forage in extremely rugged terrain and have a high reproductive capacity, formidable factors in the alteration of Hawaii’s forests (Clarke and Cuddihy 1980, p. C- 20; van Riper and van Riper 1982, pp. 34–35; Tomich 1986, pp. 153–156; Cuddihy and Stone 1990, p. 64). Exclusion of goats has been documented to benefit native plants, including ohia (Spatz and Mueller-Dombois 1973, p. 873; Loope et al. 1988, p. 277; HDOFAW 2002, p. 52).

38

Potential Stressors

Feral Cattle (Bos taurus) Introduced to Hawaii Island in 1793 (Fischer 2007, p. 350), feral cattle were considered a factor in the destruction of native Hawaiian forests as early as the 1940s (Baldwin and Fagerlund 1943, pp. 118–122). Historically feral cattle were established on the islands of Kauai, Oahu, Molokai, Maui, Kahoolawe, and Hawaii Island, but are found today only on Maui and Hawaii Island, from coastal lowland to montane mesic and wet forests (Tomich 1986, pp. 140–144). The effects of cattle, both feral and domestic, on native plants are well documented. Cattle eat native vegetation, trample roots and seedlings, cause erosion, create disturbed areas conducive to invasion by nonnative plants, and spread seeds of nonnative plants in their feces and on their bodies. Cattle have been observed breaking down ungulate exclosure fences at Hakalau Forest NWR, which provides important Iiwi habitat (Tummons 2011, p. 4). Similar to management involving other feral ungulates, the removal of cattle yields significant benefits to native plants and plant communities (Skolmen and Fujii 1980, pp. 301–310; Cuddihy 1984, pp. 16, 34).

Sheep and Mouflon (Ovis spp.) Sheep were brought to Hawaii Island in the early 1950s, and beginning in 1962, mouflon were crossbred with sheep and released onto the slopes of Mauna Kea. Today scattered mouflon hybrids occur around Mauna Loa, and a large population inhabits the region between the northern part of Mauna Loa and the lands surrounding Mauna Kea, the site of either a small, or possibly transitional, Iiwi population. Due to their great agility and high reproductive rates, mouflon sheep have the potential to occupy most forest ecosystems on Hawaii Island, from sea- level to very high elevations (Hess 2010, pers. comm.; Ikagawa 2011, in litt.), and are known throughout the world for their capacity to destroy vegetation (Ikagawa 2011, in litt.). Feral sheep browse and trample native vegetation and have destroyed large areas of native forest and shrubland (Tomich 1986, pp. 156–163; Cuddihy and Stone 1990, p. 65–66). Large areas of Hawaii Island have been devastated by sheep, and sheep impacts to native and endangered native plants are well-documented (Degener et al. 1976, pp. 173–174; Scowcroft and Sakai 1983, p. 495; Mehrhoff 1993, p. 11; Benitez et al. 2008, pp. 59, 61; Hess 2008, p. 1). Within the area of Mauna Kea inhabited by Iiwi, sheep browsing reduced seedling establishment of Sophora chrysophylla (mamane) so severely that it resulted in a reduction in tree line elevation (Warner 1960 in Juvik and Juvik 1984, pp. 191–202).

Black-tailed deer (Odocoileus hemionus) Black-tailed deer (also known as mule deer) were first introduced to Kauai in 1961 for the purpose of sport hunting. These deer are currently limited to the western side of Kauai, where they feed on a variety of native and alien plants (van Riper and van Riper 1982, p. 42–46). In addition to directly affecting native plants through browsing, deer likely serve as a vector for the spread of introduced plants. Deer feed on many alien plant species, and likely distribute seeds through their feces as they travel. Black-tailed deer have been noted as a cause of habitat alteration in the Kauai ecosystems (HBMP 2007).

Axis deer (Axis axis) Axis deer were first introduced to Molokai in 1868, Lanai in 1920, and Maui in 1959 (Hobdy 1993, p. 207; Erdman 1996, pers. comm. cited in Waring 1996, in litt., p. 2; Hess 2008, p. 2). Recently (2010–2011), axis deer were illegally introduced to Hawaii Island as a game animal (Kessler 2011, in litt.; Aila 2012a, in litt.). They have been observed in the regions of

39

Potential Stressors

Hawaii Island occupied by the Iiwi including Kohala, Kau, Kona, and Mauna Kea (HDLNR 2011, in litt.). Documented habitat destruction by axis deer on the islands of Kahoolawe, Lanai, and Maui raise concern about their becoming established on Hawaii Island (Swedberg and Walker 1978, cited in Anderson 2003, pp. 124–125; Mehrhoff 1993, p. 11; Anderson 2002, poster; Hess 2008, p. 3; Perlman 2009, in litt., pp. 4–5; Hess 2010, pers. comm.; Kessler 2010, pers. comm.; Medeiros 2010, pers. comm.).

Axis deer are primarily grazers, but also browse numerous plant species, including commercial crops (Waring 1996, in litt., p. 3; Simpson 2001, in litt.). Generally, they prefer lower elevation areas with low, open vegetation for browsing and grazing; however, during episodes of drought (e.g., from 1998–2001 on Maui (Medeiros 2010, pers. comm.)), axis deer can shift into urban and forested areas in search of food (Waring 1996, in litt., p. 5; Nishibayashi 2001, in litt.). Like goats, axis deer can be highly destructive to native vegetation and contribute to erosion by eating young trees and young shoots of plants before they can become established, creating trails that can damage native vegetative cover, promoting erosion by destabilizing substrate and creating gullies that convey water, and by dislodging stones from ledges that can cause rockfalls and landslides and damage vegetation below (Cuddihy and Stone 1990, pp. 63– 64).

Rats (Rattus spp.) Although introduced rats are best known for their predation on island birds, rats also eat fleshy fruits, seeds, flowers, stems, leaves, roots, and other parts of native plants (Atkinson and Atkinson 2000, p. 23), and thus can impede regeneration of native plant communities. Research on rats in forests in New Zealand has demonstrated that, over time, differential regeneration as a consequence of rat predation may alter the species composition of forested areas (Cuddihy and Stone 1990, pp. 68–69). Rats have caused declines or even the total elimination of many island plant species (Campbell and Atkinson 1999, p. 265). In the Hawaiian Islands, rats may consume as much as 90 percent of the seeds produced by some trees, or in some cases prevent the regeneration of forest species completely (Cuddihy and Stone 1990, pp. 68–69). All three species of rats introduced to the Hawaiian Islands (black, Norway, and Polynesian; see Predation section, below) have been reported to be a serious threat to many endangered or threatened Hawaiian plants (Stone 1985, p. 264; Cuddihy and Stone 1990, pp. 67–69; Gon III and Tierney 1996; Lorence and Perlman 2007, pp. 357–361; Pratt 2008, in litt.; HBMP 2010a; HBMP 2010b; HBMP 2010c; HBMP 2010d; PEPP 2010, pp. 101, 113; Bio 2011, pers. comm).

Habitat Loss and Degradation—Impacts to Ohia Forests—Competition with Nonnative Plants

• It is well established that numerous invasive, nonnative plant species widely impact and disrupt ohia forest ecology. Several of these plants, such as strawberry guava and albizia trees, possess a variety of traits allowing them to outcompete Hawaii’s native, shade-intolerant, slow-growing ohia. Because Iiwi rely on native ohia forest, the continued conversion of their habitat by nonnative plants remains a risk to the species.

The original native (pre-human) flora of Hawaii, including ohia, consisted of about 1,000 taxa, 89 percent of which were endemic. Since human arrival, over 800 plant taxa have become

40

Potential Stressors

established from elsewhere, and nearly 100 species have become pests due to their particularly invasive nature (Smith 1985, p. 180; Cuddihy and Stone 1990, p. 73; Gagne and Cuddihy 1999, p. 45). Besides accidental introductions, many of the 800 nonnative plant species (including some of the 100 pests) were purposely introduced to Hawaii, by the original Hawaiians, well- intentioned reforesters, ranchers, agriculturalists, or those involved in the horticultural industry (Scott et al. 1986, pp. 361–363; Cuddihy and Stone 1990, p. 73).

Nonnative plants adversely impact native Hawaiian ecosystems, including wet and mesic ohia forests, by modifying the availability of light, altering soil-water regimes, modifying nutrient cycling, altering fire characteristics of native plant communities (e.g., successive fires that burn increasingly further into native habitat, destroy native plants, and remove habitat for native species by altering microclimatic conditions to favor alien species), and ultimately converting native dominated plant communities to nonnative plant communities (Smith 1985, pp. 180–181; Cuddihy and Stone, 1990, p. 74; D’Antonio and Vitousek 1992, p. 73; Vitousek et al. 1997, p. 6). This directly and indirectly impacts the ecology of the native forests by modifying or altering or destroying habitat and reducing food sources. Feral ungulates (see discussion above), greatly exacerbate the spread of invasive plant species by transporting them externally on their bodies or through their feces.

Montane Mesic Ohia Forest Impacts The most common nonnative plant threats to the montane mesic ohia forests include the understory and subcanopy species, Axonopus fissifolius, Blechnum appendiculatum, Christella parasitica, Cyperus meyenianus, Ehrharta stipioides (meadow ricegrass), Erigeron karvinskianus, Hedychium gardnerianum, Holcus lanatus (common velvet grass), Kalanchoe pinnata, Lantana camara, Lonicera japonica (Japanese honeysuckle), Melastoma septemnervium, Paspalum urvillei, Passiflora tarminiana (banana poka), Rubus argutus, R. ellipticus (yellow Himalayan raspberry), and R. rosifolius; and the canopy species, Corynocarpus laevigatus (karakanut), Eucalyptus robusta (swamp mahogany), Falcataria moluccana (albizia tree), Psidium cattleianum, Rhodomyrtus tomentosa, and Ricinus communis (castor bean) (HBMP 2007).

Montane Wet Ohia Forest Impacts The most common nonnative plant threats to the montane wet ohia forests include the understory and subcanopy species, Andropogon glomeratus (bushy bluestem), Andropogon virginicus (broomsedge), Axonopus fissifolius, Clidemia hirta (Koster’s curse), Cyperus meyenianus, Erechtites valerianifolia (fireweed), Erigeron karvinskianus, Hedychium gardnerianum, Juncus planifolius, Kalanchoe pinnata, Lantana camara, Paspalum urvillei, Passiflora tarminiana, Rubus argutus, R. ellipticus, R. rosifolius, Sacciolepis indica, Setaria parviflora, and Xyris complanata (yellow-eyed grass); and the canopy species, Falcataria moluccana (albizia tree), Morella faya (firetree), and Psidium cattleianum (HBMP 2007).

41

Potential Stressors

Strawberry Guava and Albizia Trees

Invasion of low and middle elevation Hawaiian forests in the last century by introduced strawberry guava (Psidium cattleianum) and albizia (Falcataria moluccana) trees has resulted in the broad-scale conversion of large areas of native ohia forest to forests dominated by nonnative species throughout the main Hawaiian islands.

Strawberry Guava: Since its introduction in the early 19th century to the Hawaiian Islands from its native range in Brazil, strawberry guava has become invasive and has expanded into most of the native lowland forests of Hawaii, becoming the dominant species in these areas (DLNR 2010). Strawberry guava forms impenetrable stands of close-standing trees, to the exclusion of all native tree species at elevations up to 2,100 ft (640 m) in the Hamakua region on Hawaii Island, and has begun to invade native forests in the region to elevations as high as 3,200 ft (975 m) (USFS 2015). At a study site on Hawaii Island at an elevation of 3,000 ft (915 m), average annual increases of over 12 percent stem density and 9 percent total basal area were measured in areas of recent strawberry guava infestation (DLNR 2010, p. 14). With changing climate, strawberry guava is predicted to invade native wet and mesic forests up to 5,997 ft (1,828 m) in elevation, thereby potentially negatively affecting virtually all remaining native forest habitat for Iiwi by the end of the 21st century (Price et al. 2009, pp. 394–395).

Albizia: Albizia is a fast growing tree native to Indonesia, Papua New Guinea, New Britain, and the Solomon Islands that was introduced to Hawaii in 1917 (Big Island Invasive Species Committee (BIISC) 2014). Ohia tolerates only light shade and can grow under a koa canopy, but does not regenerate in its own shadow or under other overstory trees, relying on tree fall gaps in mature native forest for recruitment (Friday and Herbert 2006, pp. 8–9). Albizia trees can reach heights of more than 20 ft (6 m) in their first year, 45 ft (13.7 m) in their third year, and 60 ft (18 m) by the end of their tenth year, and can quickly overtop slower growing ohia trees. Albizia trees thus shade and kill ohia, enhancing the environment for other invasive species that exclude ohia, including strawberry guava (Sumida et al. 2013, p. 2). Extensive areas of the Puna District on Hawaii have been invaded by albizia from coastal areas to elevations of 492 ft (150 m), and the estimated potential distribution of albizia in the Puna and Hamakua Districts on Hawaii is up to 1,475 ft 450 m) in elevation (see Figure 25).

42

Potential Stressors

Figure 25. Map showing existing and potential albizia tree habitat on Hawaii Island (white areas near the coast (not on Mauna Loa and Mauna Kea summits) (Image from Sumida et al. 2013; used with permission)

43

Potential Stressors

Disease—Overview

• Several studies indicate that Iiwi, like most Hawaiian honeycreepers, have been greatly impacted by introduced diseases, including avian malaria and poxvirus. Iiwi show acute mortality when infected by malaria, and its populations are now mainly limited to high elevations where ambient temperature is low enough to inhibit the reproduction of mosquitoes and the replication and transmission of the parasite that causes the disease. Infection with avian pox, another mosquito-borne disease, can also result in high levels of mortality in native Hawaiian forest birds. Seasonal downslope movement of Iiwi in pursuit of flowering ohia continually exposes the species to mosquito-borne diseases. Based upon several climate change models projecting future increased temperatures, all existing research indicates that Iiwi’s risk from avian diseases will increase as a consequence of climate change.

Disease—Avian Malaria and Avian Pox—History

The introduction of avian diseases transmitted by the introduced southern house mosquito (Culex quinquefasciatus), including avian pox (Avipoxvirus sp.) and avian malaria (caused by the protozoan Plasmodium relictum), has been a key driving force in both extinctions and extensive declines over the last century in the abundance, diversity, and distribution of many Hawaiian forest bird species, including declines of Iiwi (e.g., Warner 1968, entire; Van Riper et al. 1986, entire; Benning et al. 2002, p. 14,246; Atkinson and LaPointe 2009a, p. 243; Atkinson and LaPointe 2009b, pp. 55–56; Samuel et al. 2011, p. 2,970; LaPointe et al. 2012, p. 214; Samuel et al. 2015, pp. 13–15). With no natural immunity to these introduced pathogens, native Hawaiian honeycreepers are highly susceptible to avian malaria, although there is growing evidence that a few of the remaining species may be showing signs of increased resistance, particularly the Amakihi and Apapane (Warner 1968, p. 117; Woodworth et al. 2005, pp. 1,531–1,536; Atkinson et al. 2009, pp. 57–58; Krend 2011, entire; LaPointe et al. 2012 pp. 220–221; Atkinson et al. 2014a, pp. 366, 369; Samuel et al. 2015, pp. 12–14). Iiwi, however, appear to be especially vulnerable to avian malaria, demonstrating an extremely high level of mortality in response to this pathogen (Atkinson et al. 1995, p. S65; Samuel et al. 2015, p. 12).

Mosquitoes, avian pox, and avian malaria all became established in Hawaii between the early 1800s and early 1900s. Mosquitoes were accidentally introduced to the Hawaiian Islands in 1826, and spread quickly to the lowlands of all the major islands (Warner 1968, p. 104; Van Riper et al. 1986, p. 340). Early observations of birds with characteristic lesions suggest that avian poxvirus was established in Hawaii by the late 1800s (Warner 1968, p. 106; Atkinson and LaPointe 2009b, p. 55). Poxvirus was demonstrated to be in Hawaii as early as 1900, based on recent PCR (polymerase chain reaction) amplification of a partial poxvirus 4b core protein gene from a museum specimen Elepaio (Chasiempis sandwichensis) originating from Hawaii Island (Jarvi et al., 2008, p. 339). Avian malaria most likely arrived in Hawaii in the 1920s or 1930s (Warner 1968, p. 107; Van Riper et al. 1986, pp. 340–341; Atkinson and LaPointe 2009b, p. 55; Banko and Banko 2009, p. 52; Paxton 2016, pers. comm.), likely in association with imported cage birds (Yorinks and Atkinson 2000, p. 731 and references therein), or through the deliberate introduction of nonnative birds to replace the native birds that had by then disappeared from the lowlands (Atkinson and LaPointe 2009b, p. 55).

44

Potential Stressors

Figure 26. A generalized model of native bird abundances, malarial parasite incidence, and mosquito vector levels along an elevation gradient on Mauna Loa, Hawaii. (Image from Van Riper et al. 1986; used with permission).

The most convincing indirect evidence of the major impact of avian disease (pox and malaria) on forest bird populations is the strong negative correlation between local prevalence of vector and pathogens and the diversity and density of extant native birds (Warner 1968, entire; Van Riper et al. 1986, entire; Atkinson and LaPointe 2009b, entire; Samuel et al. 2011, p. 2,960). Warner (1968, p. 102) was the first to describe a striking pattern in the disappearance of native forest birds from the lower elevations of the Hawaiian Islands, and to tie the phenomenon to the high density of disease-carrying mosquitoes in the warmer low-elevation areas. Rapid declines and the disappearance of many forest birds from low and middle elevation forests on all Hawaiian islands in the early 1900s (Banko and Banko 2009, pp. 52–53) and the high susceptibility of native honeycreepers to avian malaria and pox provide strong circumstantial evidence that these diseases have substantially affected wild populations (Atkinson et al. 1995, pp. S65–S66; Samuel et al. 2011, pp. 2,965–2,967; Samuel et al. 2015, p. 1). The work of Van Riper et al. (1986) strongly supported the general pattern observed by Warner, demonstrating a clear altitudinal relationship between the occurrence of the mosquito vector, prevalence of the malaria parasite Plasmodium, and the abundance of native birds (see Figure 26); this relationship has since been confirmed in several recent studies (e.g., Eggert et al. 2008; Atkinson and LaPointe 2009a; Samuel et al. 2011; LaPointe et al. 2012; Atkinson et al. 2014; Samuel et al. 2015; Paxton et al. 2016).

We now know that both the mosquito life cycle and the development of Plasmodium respond positively to increased temperature, such that malaria transmission is greatest in warm, low-elevation forests with an average temperature of 72 degrees Fahrenheit (oF) (22 degrees Celsius (oC)), and is largely absent in high-elevation forests above 4,921 ft (1,500 m) with mean annual temperatures around 57oF (14oC) (Ahumada et al. 2004, p. 1,167; LaPointe et al. 2010, p. 318; Liao et al. 2015, p. 4,343). These high-elevation forests thus currently serve as disease-free refugia for Hawaiian forest birds, including Iiwi. Once described as “ubiquitous,” Iiwi are now

45

Potential Stressors rarely found at lower elevations, and are increasingly restricted to high-elevation mesic and wet forests, where cooler temperatures limit both the development of the malarial parasite and densities of its mosquito vector (Scott et al. 1986, pp. 367–368; Ahumada et al. 2004, p. 1,167; LaPointe et al. 2010, p. 318; Samuel et al. 2011, p. 2,960; Liao et al. 2015, p. 4,346; Samuel et al. 2015, p. 14;).

Disease—Avian Malaria—Development and Transmission

Here, we briefly summarize the life cycle and mechanism for transmission of Plasmodium relictum and the pathology of avian malaria. The Plasmodium sporozoites (infective mosquito stages of the protozoan) are transferred to a bird when it is bitten by an infective mosquito. Inside the bird, the sporozoites invade tissues, differentiate through asexual reproduction into meronts and produce merozoites that penetrate the bird’s red blood cells (erythrocytes). After one or more cycles of asexual reproduction in the bird’s red blood cells, merozoites develop into infectious gametocytes (the sexually reproductive form of the parasite).

These gametocytes are then transferred to a new mosquito when it bites and feeds upon the infective bird. Inside the new mosquito, gametocytes produce gametes that undergo sexual reproduction to produce a motile zygote called an ookinete. The ookinete penetrates the mosquito’s midgut wall and develops into an oocyst. Sporozoites develop within oocysts through asexual reproduction, and are released when oocysts mature and rupture. The sporozoites move through the haemocoel (main body cavity) of the mosquito and subsequently penetrate the salivary glands. At this point the mosquito is considered infective, and once it bites a new bird, the sporozoites are transferred through the salivary gland and the disease cycle continues (see Figure 27).

Figure 27. Life cycle and transmission of the Plasmodium parasite in birds.

46

Potential Stressors

Birds undergo an acute phase of infection during which parasitemia, a quantitative measure of the number of parasites in the circulating red blood cells, increases steadily. Because the parasite destroys red blood cells during asexual reproduction, anemia and decline of physical condition can result. In native Hawaiian forest birds, death may result either directly from the disease or indirectly when birds, weakened by anemia, become vulnerable to predation, starvation, or a combination of other stressors (LaPointe et al. 2012, p. 213). Studies have demonstrated that native Hawaiian birds that survive avian malaria remain chronically infected, thus becoming lifetime reservoirs of the disease (Samuel et al. 2011, p. 2960; LaPointe et al. 2012, p. 216). In contrast, the survivorship of nonnative birds is little affected by avian malaria, and they can clear their infections after a short acute phase, subsequently becoming incapable of disease transmission (LaPointe et al. 2012, p. 216).

Disease—Avian Malaria—Temperature Dependence

Of note, the production of sporozoites from the oocysts inside the mosquito is strongly temperature dependent during the Plasmodium life cycle; the incubation period of the Plasmodium is inversely related to temperature and may last from 6 to 28 days. This characteristic ultimately restricts the altitudinal distribution of infectious mosquitoes and disease transmission in the Hawaiian Islands (LaPointe et al. 2012, p. 217). Plasmodium relictum requires a minimum of 13oC (55oF) for development, which results in limited development during cooler seasons or at high elevation. The incubation period is described as rapid at 82oF (28oC), slowing considerably at 70oF (21oC), and extends beyond 30 days when temperatures are less than 63oF (17oC) (LaPointe et al. 2010, pp. 318, 322; LaPointe et al. 2012, p. 217).

Laboratory studies of thermal constraints on the production of Plasmodium sporozoites in experimentally infected Culex quinquefasciatus mosquitoes showed dramatically slowing development of the parasite at 59oF (15oC) and a cessation of development at 55oF (13oC) (LaPointe et al. (2010, p. 321). On the landscape, the 59oF (15oC) isotherm closely follows the 4,921 ft (1,500 m) contour line that generally corresponds to the current Iiwi lower elevation range. This finding indicates that current high-elevation refugia from disease are the result of thermal limits on malarial development in the mosquito vector, while at elevations below 4,921 ft (1,500 m), the key factor driving epizootics of avian pox and malaria is the seasonal and altitudinal distribution and density of C. quinquefasciatus (Atkinson and LaPointe 2009b, p. 57).

Temperature also affects the life cycle of Culex quinquefasciatus. Lower temperatures slow the development of larval stages and can affect the survival of adults as well (Ahumada et al. 2004, p. 1,165–1,168; LaPointe et al. 2012, p. 217). Although closely tied to altitude and a corresponding decrease in temperature, the actual range of mosquitoes varies with season. Generally, as temperature decreases with increasing elevation, mosquito abundance drops significantly at higher altitudes. In the Hawaiian Islands, in 1986 VanRiper et al. (1986, p. 338) had estimated its elevational boundary occurring at 4,921 ft (1,500 m), while more recently in 2012, LaPointe et al. (2012, p. 218), placed this boundary at 5,577 ft (1,700 m) (likely due to increasing temperatures; see additional discussion on this upward shift in the climate change discussion). Rainfall also impacts malaria transmission by affecting mosquito larval habitat

47

Potential Stressors

availability and larval and adult survivorship of mosquitoes. In general, increased precipitation and humidity increases mosquito populations, however excessive rainfall can reduce mosquito abundance by flooding larval habitats and killing adult mosquitoes (Aruch et al. 2007, p. 573; LaPointe et al. 2012, p. 217). In addition to suitable temperature and precipitation, mosquitoes also require suitable habitats for the development of their aquatic larvae, which are generally provided by areas of standing water. Normally Hawaii’s porous volcanic soils are not conducive to forming such habitats. However, actions that result in the compaction of soils or otherwise provide relatively impervious surfaces capable of holding water can create suitable larval habitat for mosquitoes. For example, feral pigs create such habitats by wallowing and by knocking down and eating hapuu fern (tree fern) logs, both of which form cavities that collect rain water. The complex relationship between temperature, rainfall, the mosquito vector, availability of mosquito breeding sites, and the avian host, among other factors, together influence the dynamics of the avian malaria disease cycle (Ahumada et al. 2009, pp. 333–336).

Prevalence of mosquitoes and disease at low, middle, high elevation sites is known to vary by season, although generally, mosquito populations are highest within lowland wet forests. At elevations between 4,000 and 5,000 ft (1,200 and 1,500 m), peak mosquito populations occur during the warmer fall months of September through December concurrent with the more common seasonal epidemics of malaria and pox (Atkinson and LaPointe 2009b, p. 57). LaPointe et al. (2012, p. 217) characterized the key seasonal patterns in avian malaria transmission as follows: high malaria transmission in low elevation forests with minor seasonal or annual variation in infection; episodic transmission in mid-elevation forests with site-to-site, seasonal, and annual variation; and disease refugia in high-elevation forests with slight risk of infection only during summer. In addition to elevation, the time of year and forest type (mesic vs. xeric) strongly influence Plasmodium prevalence and intensity. Researchers have observed that increases in prevalence in native birds were highest from July to December, and that wet forests consistently supported higher prevalence levels (Van Riper et al. 1986, p. 330).

Disease—Avian Malaria—Mortality of Native Hawaiian Birds and High Susceptibility of Iiwi

• Iiwi are highly susceptible to avian malaria and experience a high rate of mortality in response to the disease.

Wild Iiwi infected with malaria (between 2 and 6 percent prevalence) are rarely captured, apparently because the onset of infection leads to rapid mortality, precluding their capture (Samuel et al. 2011, p. 2,967; LaPointe et al. 2016, p. 11). However, controlled experiments with captive birds have demonstrated the susceptibility of native Hawaiian honeycreepers to avian malaria, and the extremely high mortality in some species experimentally infected with the disease. For example, Warner’s (1968) early controlled experiments with both Laysan finches (Telespiza cantans) and several species of honeycreepers demonstrated 100 percent mortality from malaria in a very short period of time (Warner 1968, pp. 109–112, 118; see Figure 28). Nonnative Japanese white-eyes (Zosterops japonicas), by comparison, were unaffected (Warner 1968, p. 112). Subsequent studies by Van Riper et al. (1986) were the first to specifically test the response of Iiwi to malaria. Similar to the results of Warner (1968), they found nonnative

48

Potential Stressors birds largely resistant to contracting malaria, while native birds showed varying levels of response. All native birds contracted malaria, but Iiwi exhibited among the lowest resistance and survivorship (Van Riper et al. 1986, p. 339).

Figure 28. Cumulative mortality curves in Laysan Finches exposed to lowland mosquitoes or kept in a mosquito- proof enclosure (control group) at Lihue, Kauai. (Image from Warner 1968; used with permission).

In a study specific to Iiwi, Atkinson et al. (1995, entire) demonstrated that the species is highly vulnerable to avian malaria, exhibiting an “extraordinarily high mortality” (Atkinson et al. 1995, p. S65). During a controlled experiment that duplicated natural routes of exposure through infective mosquito bites, 33 juvenile (hatch year) Iiwi were split into three groups, including low dose (single bite from an infected mosquito); high dose (multiple bites from infected mosquitoes); and a control group (mosquito bites from uninfected mosquitoes). For comparison, seven spice finch (or nutmeg mannikins) (Lonchura punctulata), a species introduced to Hawaii, were assigned to a separate and fourth group which received multiple infected bites (high dose). Following exposure to biting mosquitoes, food consumption, weight, and parasitemia were monitored for all groups. None of the seven nutmeg mannikins developed malarial infections, while all Iiwi exposed to infective bites developed infections within 4 days (Atkinson et al. 1994, p. S63). Mortality of the high dose Iiwi reached 100 percent by day 29 and 90 percent of the low dose birds by 37 days, an average of 95 percent mortality between the two groups (see Figure 29). A single male Iiwi survived the initial infection and was re-exposed with the same Plasmodium isolate. No subsequent increase in parasitemia was detected, suggesting that the bird had developed some immunity (Atkinson et al. 1995, p. S66).

49

Potential Stressors

Figure 29. Mortality and parasitemia for Iiwi experimentally exposed to malaria through low-dose and high-dose bites of infected mosquitoes. (Image from Atkinson et al. 1995; used with permission).

The authors noted that the young age of the birds in the study may have contributed to the high mortality observed, since juvenile birds are known to be more susceptible to malarial infections (Atkinson et al. 1995, p. S65). They also suggested that Iiwi may lack immuno- genetic traits and variation capable of recognizing and responding to malarial antigens, an important factor in Iiwi’s susceptibility to introduced disease (Atkinson et al. 1995, pp. S65-S66, and references therein).

A more recent study by Samuel et al. (2015) modeled avian malaria in Hawaiian forest birds and confirmed the findings of Atkinson et al. (1995) with regard to the particularly high mortality of Iiwi from the disease. The authors combined multi-state capture-recapture models with age-prevalence models to create estimates of disease transmission, mortality, and survival. The study examined three species of Hawaiian honeycreepers chosen for their differences in malaria susceptibility, with Iiwi being highly vulnerable, and both Apapane and Amakihi moderately vulnerable. The three species also exhibit different foraging and movement patterns, with the Amakihi being relatively stationary compared to the highly mobile Iiwi and Apapane, both which make altitudinal movements in search of seasonal or ephemeral food sources. Blood samples to test for malarial infection were obtained from 5,353 captured Amakihi, 2,116 Apapane, and 1,046 Iiwi. Because no Iiwi were captured in low-elevation forests and only a few were captured in mid-elevation forest, the evaluation for Iiwi was restricted to high-elevation captures. Results of the study indicated an average of 93 percent malaria mortality for both adult and hatch-year Iiwi, with a range of 87 to 98 percent. Mortality of Iiwi at 93 percent was significantly greater than estimates for either Amakihi at 66 percent or Apapane at 47 percent

50

Potential Stressors

(Samuel et al. 2015, pp. 12-13). The results of the study suggest that the estimated 93 percent Iiwi mortality rate may reduce the adult population by 16-20 percent annually and the hatch year population by 55-73 percent annually. The annual survival rate for Iiwi not infected by malaria was estimated at 55 to 60 percent (Samuel et al. 2015, p. 12). The study also estimated vector (mosquito) feeding preference based on relative annual infection rates in adult birds adjusted for population density, concluding that of the three native bird species, mosquitoes prefer feeding on Iiwi (Samuel et al. 2015, p. 15).

Disease—Avian Malaria—Lack of Resistance in Iiwi

• There is to date no clear indication that Iiwi have evolved substantial tolerance or resistance to avian malaria, unlike some related honeycreepers.

In recent years there have been signs that some species of Hawaiian honeycreepers may be developing some level of resistance or tolerance to malaria including Amakihi and Apapane, which have been recently observed in lowland habitats from which they have long been absent (Woodworth et al. 2005, p. 1,531; Spiegel et al. p. 2006, p. 175; Eggert et al. 2008, p. 7-8). For Amakihi in particular, molecular studies have documented population differences in gene frequencies between high elevation susceptible and low elevation tolerant populations (Woodworth et al. 2005, p. 1,531; Atkinson et al. 2014a, p. 366), suggesting the low elevation population evolved tolerance in isolation from the high elevation population (Foster et al. 2004, entire; Eggert et al. 2008, pp. 7-8). As a species, Amakihi has maintained relatively high levels of genetic diversity and populations at different elevations exhibit genetic differentiation (Jarvi et al. 2001, p. 259; Jarvi et al. 2004, p. 2,163; Foster et al. 2007, p. 4,741; Eggert et al. 2008, pp. 2- 4). The stationary nature of Amakihi may have played a role in the relatively rapid response to selection pressure for disease resistance, as selection acting on a closed population would not face the counteracting effects of gene flow (Foster et al. 2007, p. 4,744). By comparison, populations of mobile species such as Apapane and Iiwi would be slower to evolve as directional selection would be offset by gene flow.

Iiwi consistently exhibit lower levels of genetic variability than Amakihi in most measures (Jarvi et al. 2001, p. 255; Jarvi et al. 2004, Table 4, p. 2,164; Foster et al. 2007, p. 4,744), and suffer high avian malaria mortality compared to both Amakihi and Apapane (Samuel et al. 2015, pp. 12-13). Some measures of Iiwi mitochondrial DNA demonstrate an apparent lack of variation (Tarr and Fleischer 1993, 1995; Fleischer et al. 1998; Foster et al. 2007, p. 4,743). In contrast, Jarvi et al. (2004, p. 2,166) reported diversity in the antigen-binding sites of Iiwi’s major histocompatibility complex (MHC) genes, which are extremely important in generating an immune response, comparable to that observed in both Apapane and Amakihi. This dichotomy of lack of variation in mitochondrial DNA and maintenance of variation in nuclear DNA suggests that Iiwi may have experienced a genetic bottleneck in the past (Jarvi et al. 2004, p. 2,166; Foster et al. 2007, p. 4,744). Despite the reduction in genetic variability that might result from such a bottleneck, the finding that Iiwi have maintained genetic variability at the MHC suggests that the MHC is under intense selection from malaria, but other immune systems may be important for tolerance or resistance to evolve, particularly innate immunity rather than adaptive immunity.

51

Potential Stressors

Despite extremely high mortality of Iiwi from avian malaria in general, past studies have demonstrated that a few individuals are capable of surviving the infection (Van Riper et al. 1986, p. 334; Atkinson et al. 1995, p. S63; Freed et al. 2005, p. 759). If a genetic correlation can be identified, it is possible that surviving individuals could serve as a potential source for spreading malaria pathogen resistance across the range of Iiwi. Eggert et al. (2008, p. 8) reported a slight but detectable level of genetic differentiation between Iiwi populations located at mid and high elevation, potentially the first sign of selection acting on these populations in response to disease. The infrequent but occasional sighting of Iiwi on Oahu indicates a possible developed resistance or tolerance to avian malaria, and study of these individuals should be a priority if possible. Despite these encouraging observations, there is, to date, no indication that Iiwi have developed significant resistance to malaria such that individuals can survive in areas where the disease is strongly prevalent, including all potential low elevation forest habitat and most mid-elevation forest habitat (Foster et al. 2007, p. 4,743; Eggert et al. 2008, p. 2). In one study, for example, four years of mist-netting effort on Hawaii Island resulted in the capture of 5,353 Amakihi, 2,116 Apapane, and 1,046 Iiwi (Samuel et al. 2015, p. 7). Despite the very high numbers of captured individuals and extensive area mist-netted over multiple years, no Iiwi were captured in low- elevation forests and only a few were captured in mid-elevation forests (Samuel et al. 2015, p. 11). Furthermore, Eggert et al. (2008, p. 9) noted that gene variations (alleles) that may confer disease resistance in Iiwi appear to be rare in the species. This homogeneity of this portion of Iiwi’s genome in conjunction with the high mortality rate of Iiwi in response to avian malaria and the high levels of gene flow as a consequence of the wide-ranging nature of the species, suggest that Iiwi would likely require a significant amount of time for development of genetic resistance to avian malaria – assuming the species retains a sufficiently large reservoir of genetic diversity for natural selection upon which to act.

Disease—Avian Malaria—Seasonal Iiwi Movement to Lower Elevations Increase Disease Infection Risk

• Iiwi exposure to malaria infection during seasonal migration to lower elevations in search of ohia nectar limits the protection provided by high-elevation, disease-free habitat.

Early on, both Ralph and Fancy (1995, p. 741) and Atkinson et al. (1995, p. S66) suggested that the seasonal movements of Iiwi to lower elevation ohia blooms may result in increased contact with mosquitoes carrying avian malaria, and that their susceptibility to the disease may explain their observed low annual survivorship relative to other native Hawaiian birds. Hart et al. (2011, pp. 121–122) also document these seasonal movements, and described Iiwi’s descent to lower elevation and increased exposure to avian malaria as an “ecological trap.” Compounding the issue, other bird species, including Apapane, which are relatively resistant and carry both Plasmodium and avian pox virus, have become reservoirs of these diseases, which they carry upslope where mosquitoes are less abundant but still occur in sufficient numbers to facilitate and continue transmission to Iiwi (Ralph and Fancy 1995, p. 741).

Subsequent studies have confirmed that altitudinal migrations of Iiwi in search of ephemeral nectar resources may result in birds from high elevation refugia becoming exposed to avian disease. Kuntz (2008, p. 3) found Iiwi populations at an upper elevation study site (6,300

52

Potential Stressors

ft (1,920 m)) at Hakalau Forest NWR diminished in abundance during the non-breeding season when birds departed the refuge for lower elevations, traveling up to 12 mi (19.4 km) over contiguous wet forest. These results suggest that upper elevation forest reserves in Hawaii may not adequately protect mobile nectarivores such as Iiwi, since individuals traveling to lower elevations face a higher probability of exposure to introduced mosquito-borne diseases. Guillamet et al. (2016, p. 192) used empirical measures of seasonal movement patterns in Iiwi to model how movement across elevations increases the risk of disease exposure, even affecting breeding populations in disease-free areas. In another study, La Pointe et al. (unpublished data 2015) examined mosquito abundance data and the rate of infection in captured mosquitoes to track and compare malaria prevalence and risk of forest bird migration between high-elevation forests in Hakalau Forest NWR to lower elevation forest in the Laupahoehoe Unit of the Hilo Forest Reserve. The researchers report that based on the prevalence of malaria in all forest birds, those migrating from upper elevations in Hakalau Forest NWR to lower elevations in Laupahoehoe Forest face a dramatically greater risk of contracting avian malaria. The increased risk was correlated with a much higher abundance of mosquitoes at lower elevations, which in turn was significantly attributable to the higher abundance of pigs and their activities in Laupahoehoe Forest.

Disease - Avian Pox

Avian pox (or bird pox) is an infection caused by the virus Avipoxvirus, which produces large, granular and eventually necrotic lesions or tumors on exposed skin or diphtheritic lesions on the mouth, trachea, and esophagus of infected birds. Avian pox can be transmitted through cuts or wound upon physical contact or through the mouth parts of blood-sucking insects. In Hawaii, the introduced mosquito Culex quinquefasciatus is believed to be a common vector for both the virus that causes pox and avian malaria (LaPointe et al. 2012, p. 221). Tumors or lesions caused by avian pox can be crippling for birds, and may result in death. Depending on the physical site and severity of the lesions, birds suffering from pox may encounter difficulty seeing, feeding, breathing, or perching (Atkinson et al. 2012, p. 808). Although not extensively studied, existing data suggest that mortality from avian pox may range from 4 to 10 percent observed in Oahu Elepaio (for birds with active lesions (VanderWerf 2009, p. 743) to 100 percent in Laysan Finches and Hawaii Amakihi (Warner 1968, and Atkinson et al. 2012, both detailed below). Similar to avian malaria, Van Riper et al. (2002, pp. 933, 936-937) found significantly greater prevalence of avian pox in native birds as compared to introduced species. Out of a total of 3,122 wild birds captured on the island of Hawaii, 34.9% of Apapane, 24.3% of Omao, and approximately 20% of Hawaii Amakihi, Elepaio, and Iiwi captured were infected with pox, whereas infection in four species of nonnative birds ranged from 0 to 2% (although the House Finch exhibited a prevalence of about 20%, similar to rates observed in the continential U.S.). Both the Iiwi and Omao exhibited especially low resistance to avian pox compared to the other native bird species examined, based on the ratio of active to healed lesions observed (61.1% active vs. 38.9% healed for Iiwi) (Van Riper et al. 2002, p. 937). VanderWerf (2009, p. 743) suggested that mortality levels from pox may correlate with higher rainfall years, and at least in the case of the Elepaio, observed mortality may decrease over time with a reduction in susceptible birds. Molecular work has revealed two genetically distinct variants of the pox virus affecting forest birds in Hawaii that differ in virulence (Jarvi et al. 2008, p. 347); one tends to produce fatal lesions, and the other appears to be less severe, based on the observation of

53

Potential Stressors recurring pox infections in birds with healed lesions (Atkinson et al. 2009, p. 56; see also Atkinson et al. 2012, p. 817).

Avian pox was introduced to the Hawaiian Islands earlier than avian malaria with observations of native birds suffering from avian pox documented by 1902 (Warner 1968, p. 106). Warner (1968, p. 106) reports that epizootics of avian pox “were so numerous and extreme that large numbers of diseased and badly debilitated birds could be observed in the field.” Jarvi et al. (2008, p. 339) were able to amplify the more pathogenic of the two forest bird pox variants from museum skins of Elepaio that were collected near the turn of the 20th century. As the initial wave of post-European extinctions of native Hawaiian birds was largely observed in the late 1800s, prior to the introduction of avian malaria (Van Riper et al. 1986, p. 342), it is possible that avian pox played a significant role, although there is no direct evidence (Warner 1968, p. 106; Van Riper et al. 2002, pp. 938-939).

In his investigation of introduced disease as a possible mechanism for the extinction of native Hawaiian birds, Warner (1968) conducted experiments on the avian pox susceptibility of Laysan finches, a species of honeycreeper. In 1958, he transported 24 Laysan finches from the mosquito-free island of Laysan, located in the Northwest Hawaiian Islands, to the island of Oahu, where they were exposed to a lowland environment inhabited by mosquitoes. Within two weeks, six birds had contracted avian pox and within a month, all of the birds had developed lesions with bleeding observed as the tumors progressed in severity. Eventually, all of the Laysan finches weakened and died (Warner 1968, p. 108). During the same studies, Warner (1968, p. 109) also noted other honeycreepers’ susceptibility to pox. Eight individuals of upper elevation Kauai honeycreepers (Amakihi, lesser Amakihi (now Anianiau), and Apapane) were captured and exposed to lowland mosquitoes and rapidly developed avian pox. Unfortunately, that portion of the study was terminated before the potential response to the virus could be determined (Warner 1968, p. 109). More recently, however, Atkinson et al. (2012) examined the response of another native honeycreeper, the Hawaii Amakihi, to avian pox in a series of experiments designed to test the potential efficacy of a live-attenuated vaccine against the disease. The researchers used 31 Hawaii Amakihi captured at high elevation, above 2,000 m, where prevalence of avian pox is very low at 0.4%, and ensured the birds had no physical sign of current or past infection (serological tests to determine current or prior infection are not yet available) (Atkinson et al. 2012, pp. 809-810). Test birds were vaccinated and exposed to one of two pox virus isolates, PV1 or PV2 (the former suspected of being the less virulent form); control birds received a “sham” vaccination of sterile distilled water. Unvaccinated birds exposed to PV1 developed the tumor-like swellings that are typically associated with avian pox, and only one bird developed fatal lesions (Atkinson et al. 2012, p. 817). Following challenge with PV2, both vaccinated and control birds developed pox lesions within 9 days; 60% of the vaccinated birds died with 40 to 50 days, and 100% of the control birds died within 24 to 30 days (Atkinson et al. 2012, pp. 811-812, 817). The significant difference in mortality between unvaccinated birds exposed to PV1 and PV2 further confirmed the suspected disparity in pathogenicity of the two known variants of pox virus (Atkinson et al. 2012, p. 817). The vaccine success rate following initial inoculation was low (63%), and many of the vaccinated birds developed life-threatening lesions, indicating the vaccines tested here are not a viable means of protecting Hawaii’s native forest birds from avian pox (Atkinson et al. 2012, p. 816).

54

Potential Stressors

Compared to avian malaria, notably fewer studies have examined the effects of avian poxvirus on native Hawaiian forest birds, but both limited experimental studies and field observations are consistent in documenting that the virus can, by itself, be a significant cause of mortality (Warner 1968; Atkinson et al. 2012). The largest study of avian pox in scope and scale took place between 1977 and 1980, during which approximately 15,000 native and nonnative forest birds were collected and examined for pox virus lesions from 16 different locations on transects along Mauna Loa on Hawaii Island (Van Riper et al. 2002, pp. 929-942). The study made several important determinations, including that native forest birds are indeed more susceptible than introduced species, that all species were more likely to be infected during the wet season, and that pox prevalence was greatest at mid-elevation sites approximately 3,937 ft (1,200 m) in elevation, coinciding with the greatest overlap between birds and vectors. Of the 107 Iiwi captured and examined during the study, 17 percent showed signs of either active or inactive pox lesions (Van Riper et al. 2002, p. 932). Many studies of avian pox have documented that native birds are frequently infected with both avian pox and avian malaria (Van Riper et al. 1986, p. 331; Atkinson et al. 2005, p. 537; Jarvi et al. 2008, p. 347). This may be due to mosquito transmission of both pathogens simultaneously, because documented immune system suppression by the pox virus renders chronically infected birds more vulnerable to infection by, or a relapse of, malaria (Jarvi et al. 2008, p. 347), or due to other unknown factors. The relative frequency with which the two diseases co-occur makes it challenging to disentangle the independent impact of either stressor acting alone (LaPointe et al. 2012, p. 221). However, pox virus infections often have an immunosuppressive effect in animals, and under experimental conditions avian pox and malaria act synergistically to increase malarial parisitemia and mortality in Hawaiian honeycreepers (Atkinson et al. 2012, p. 808).

Disease—Summary

The relatively recent introduction of avian pox and avian malaria, in concert with the introduction of the mosquito disease vector, is widely viewed as one of the key factors underlying the loss and decline of native forest birds throughout the Hawaiian Islands. Evolving in the absence of mosquitoes and their vectored pathogens, native Hawaiian forest birds, particularly honeycreepers, lack natural immunity or genetic resistance, and thus are more susceptible to these diseases than are nonnative bird species (van Riper et al. 1986, pp. 327-328; Yorinks and Atkinson 2000, p. 737). Researchers consider Iiwi one of the most vulnerable species, with studies showing an average of 95 percent mortality in response to infection with avian malaria (Atkinson et al. 1995, p. S63; Samuel et al. 2015, p. 2). Many native forest birds, including Iiwi, are now absent from warm, low elevation areas that support large populations of disease-carrying mosquitoes, and these birds persist only in relatively disease-free refuges in high elevation forests, above roughly 4,921 to 5,577 ft (1,500 to 1,700 m), where both the development of the malarial parasite and the density of mosquito populations are held in check by cooler temperatures (Scott et al. 1986, pp. 85, 100, 365-368; Woodworth et al. 2009, pp. 1,531; Liao et al. 2015, pp. 4,342-4,343; Samuel et al. 2015, pp. 11-12). Even at these elevations, however, disease transmission may occur when Iiwi move downslope to forage on ephemeral patches of flowering ohia in the nonbreeding season, and thus contact disease- carrying mosquitoes (Ralph and Fancy 1995, p. 741; Fancy and Ralph 1998, p. 3; Guillaumet et al. 2016, p. 192; LaPointe et al. 2015, p. 1). Iiwi have not demonstrably developed resistance to avian malaria, unlike some species of native honeycreepers including Amakihi and Apapane. Having already experienced local extinctions and widespread population declines (see

55

Potential Stressors

Population Status, above), it is possible that the species may not possess sufficient genetic diversity to adapt to these diseases (Atkinson et al. 2009, p. 58).

The natural susceptibility of native forest birds to introduced diseases in combination with the observed restriction of Hawaiian honeycreepers to high elevation forests, led Atkinson et al. (1995, p. S68) to predict two decades ago that a shift in the current mosquito distribution to higher elevations could be “disastrous for those species with already reduced populations.” Thus, climate change has significant implications for the future of Hawaiian forest birds, as predictions suggest increased temperatures may largely eliminate the high-elevation forest barrier currently inhospitable to the transmission of mosquito-borne diseases (Benning et al. 2002, 14247-14249; LaPointe et al. 2012, p. 219; Fortini et al. 2015, p. 9). In short, the viability of forest bird populations, including Iiwi, will be at risk because warming temperatures will facilitate upslope spread of mosquitoes, malaria, and pox. Samuel et al. (2015, p. 15) predict further reductions and extinctions of native Hawaiian birds as a consequence, noting Iiwi are particularly vulnerable due to its high susceptibility to malaria. The impact of this synergistic interaction of avian disease and climate change on Iiwi is discussed in detail under the Climate Change stressor, below.

Predation

• Predation of eggs, nestlings, and incubating adults by nonnative rats is known to significantly impact several species of Hawaiian forest birds with some evidence of impact to the Iiwi as well.

Predation—Rats

Three species of nonnative rats occur in Hawaii. The Polynesian rat (Rattus exulans) likely arrived with the first human settlers in the islands (Tomich 1986, p. 41). The Norway rat (R. norvegicus) and black rat (R. rattus) arrived aboard European ships in the 18th and 19th centuries, respectively (Atkinson 1977, pp. 109, 118; Tomich 1986, pp. 39–40; 41–42). The black rat is the most common in wet and mesic forests (Lindsey et al. 1999, p. 100), and because of its arboreal habits is considered the most significant predator on forest birds (Atkinson 1977, pp. 118, 122; Tomich 1986, p. 39). The introduction of the black rat to Hawaii in the late 1800s may have been a primary contributor to the rapid extinction of 30 species or subspecies of endemic Hawaiian forest birds in the 20 years between 1890 and 1910 (Atkinson 1977, pp. 109– 110). Black rat predation on native Hawaiian birds and their nests is well documented, including predation on the Maui creeper (Alauahio, Paroreomyza montana) (Baker and Baker 2000, p. 10); the Puaiohi or small Kauai thrush (Myadestes palmeri) (Snetsinger et al. 2005, p. 72; Tweed et al. 2006, p. 753); the crested honeycreeper (Akohekohe, Palmeria dolei) (Simon et al. 2001, p. 736); the Oahu Elepaio (Chasiempis ibidis) (VanderWerf and Smith 2002, p. 73; VanderWerf 2009, p. 737); and the Palila (Loxioides bailleui) (Pletschet and Kelly 1990, p. 1,017). Rat predation has had significant negative effects on some of these species, including the Puaiohi (Tweed et al. 2006, p. 753), Palila (Pletschet and Kelly 1990, p. 1,017), and Oahu Elepaio (VanderWerf 2009, p. 737).

56

Potential Stressors

Unfortunately, we lack measured information on the impacts of predation on Iiwi (see for example Lindsey et al. 2009, p. 283 and Cummins et al. 2014, p. 22). However, the anecdotal evidence from those studies and other rat control studies suggest that such impacts occur, and the weight of expert opinion points to rat predation playing some role in the decline of Iiwi, as it has in so many island bird species (VanderWerf 2016, pers. comm.). During a two year (2013-2014) study focusing on forest bird nesting success, Cummins et al. (2014, p. 22) found evidence of a positive relationship in the daily survival rate of Iiwi nests and nest height in the non-stormy year (2013), and a negative relationship in the stormy year (2014). The study hypothesized that the positive relationship in 2013 was driven by rat predation, and the negative relationship in 2014 driven by weather, suggesting that the Iiwi may be under constant pressure to balance nest height selection to counter both predation and stochastic weather events. Predator control has had an apparent positive effect on the species’ demography, with an increased ratio of young to adult Iiwi in an area within Hakalau Forest NWR where rodents were removed compared with an untreated control site, suggesting rodent control contributed to greater numbers of young birds fledged in the treated area (Lindsey et al. 2009, pp. 280–282). Similar results were documented for the Hawaii Elepaio and Apapane. However, perhaps due to the close proximity and small sizes of the control and study areas, no statistically significant differences in nest success were found between treated and untreated areas, and no differences were detected in bird abundance during point-count censuses or mist netting (Lindsey et al. 2009, p. 281–282).

Predation—Feral Cats

• Although a known predator of many Hawaiian bird species, feral cats are not known to prey on Iiwi, although they occur and prey on other forest birds in areas where Iiwi occur. Iiwi nests, typically placed high in the terminal branches of ohia trees, are not likely to be accessible to cats.

Since their introduction to Hawaii in the early 1800s by European whalers, feral cats (Felis catus) have adapted to survive in all Hawaiian ecosystems, and are now abundant on all of the islands, ranging from high-density populations at sea level to scattered individuals in high elevation wet and dry montane forest environments (Scott et al. 1986, p. 363; Hess and Banko 2006, p. 2; USGS 2015). As early as 1840, feral cats were observed in the remote wilderness around Kilauea Volcano on Hawaii Island by explorer William Brackenridge. Mark Twain reported seeing “millions of cats” during a visit to Honolulu in 1866, and by 1903, R.C.L. Perkins had recorded possible observations of forest birds eaten by feral cats on the island of Lanai (Perkins 1903 in Hess and Banko 2006, p. 2).

Feral cats occur in most habitats across the islands. On Kauai, feral cats occur in the Alakai Swamp (Scott et al. 1986, pp. 363–364; Tweed et al. 2006, p. 753), the only place on the island where the Iiwi persists. The State of Hawaii Department of Health estimates that there are over 500,000 cats on the island of Maui, where they inhabit all forest types. In wet montane forest in Hanawi NAR and The Nature Conservancy's Waikamoi Preserve (windward east Maui), researchers have observed cats and cat scat or tracks, and an analysis of cat scats found opportunistically within Hanawi NAR revealed the remains of both nestlings and adults of native and nonnative bird species (Maui Forest Bird Recovery Project (MFBRP) 2015). Both of these conservation areas harbor Iiwi populations. On Hawaii Island, native forest birds are a regular

57

Potential Stressors component in the diets of feral cats in montane wet forest (Smucker et al. 2000, p. 233). An examination of the stomach contents of 118 feral cats captured on the flanks of Mauna Kea (home to a small Iiwi population and a foraging site) found native and introduced birds to be the most common prey item (Banko et al. 2004, p. 162). From studies on the flanks of Mauna Kea, feral cats are known to prey on the Palila in dry subalpine forest and on the Hawaiian petrel (Pterodroma sandwichensis) in their nesting areas on lava flows above treeline (Banko et al. 2004, p. 162; Hess and Banko 2006, p. 2). Within this same area both resident and transient Iiwi forage on flowers of mamane trees (see Species Information section, above), which are much lower in height than ohia trees, and thus possibly less likely to preclude predation by cats.

Competition

• Iiwi may be affected by competition for nectar resources with nonnative birds, and by competition for insects with introduced western yellowjacket wasp colonies.

Competition between introduced insects, birds, and small mammals and native Hawaiian honeycreepers is possible, but not well-documented historically (Lindsey et al. 2009, pp. 286– 287). Competition is difficult to document in the wild, because to do so requires measurements of resource limitation and of reduced fitness resulting from niche overlap with one or more competing species, phenomena that are themselves difficult to measure (Maurer 1984, pp. 385, 386–387; Mooney and Cleland 2001, pp. 5,449–5,450; Davis 2003, p. 481). Introduced species potentially could compete with Iiwi by eating nectar from ohia flowers (the Iiwi’s primary food source), or by consuming insects the Iiwi use as a secondary food source.

Pertaining to nectar competition, researchers first began suggesting in the 1980s that Iiwi may face competition from the nonnative Japanese white-eye (Mountainspring and Scott 1985, p. 219), a malaria-resistant alien species, whose numbers have dramatically increased throughout the Hawaiian Islands, particularly since the 1970s (Foster et al. 2004, p. 716). Negative correlations between Iiwi and Japanese White-eye densities may correlate with competition between the species for limited nectar resources (Fancy and Ralph 1998, p. 7). Based upon studies in Hakalau Forest NWR, Freed et al. (2008, p. 1,018) and Freed and Cann (2014, p. 1) proposed competition between Iiwi and the Japanese-white eye, based upon observations of reduced fitness of Iiwi and other native forest birds. The authors observed niche overlap of Iiwi and Japanese white-eye and provided measures of declining fitness (lower fat scores, higher frequency of feather fault bars and reduced growth rate) of native forest birds over time. However, they did not provide information on resource availability and use to link fitness reduction and food resource limitation.

Using demographic and life history information, Freed and Cann (2015, p. 1) suggested that eight native forest bird species populations at Hakalau Forest NWR have lost an average of 31.5 percent of individuals in an 8,335 ac (3,373 ha) area due to competition in particular with the Japanese white-eye. This purported decline is not supported by other studies (Paxton et al. 2013, p. 1; Camp et al. 2014, p. 97; Camp et al. 2015, pp. 1–2) which have interpreted annual forest bird survey data, to show that populations of forest birds at Hakalau Forest NWR are stable or increasing over the long term.

58

Potential Stressors

Pertaining to competition for insect food resources, the most likely competitor in some Iiwi population sites is the introduced western yellowjacket wasp (Vespula pensylvanica). It is widely recognized for its indirect impacts to native forests including predation upon a variety of native insects including many pollinator species (Gambino and Loope 1992, p. 1; Foote et al. 2011, pp. 6-7; Hanna 2012, p. 1; Cause et al. 2014, p. 1,622). The wasps shift to become voracious feeders of nectar following seasonal changes in colony nutritional needs, and the species is even documented to impact ohia fruit set due to its effective disruption of ohia pollination ecology (Hanna et al. 2014, p. 1,622). Compared to its native range in the northeastern US mainland, which experiences freezing temperatures during the winter, colonies of the western yellowjacket wasps in the tropical climate of Hawaii are able to persist for several seasons and grow to enormous sizes, often with many thousands of individuals, and several studies have documented the ability of these large colonies to consume incredibly large numbers of native insects (Gambino and Loope 1992, p. iv, 36; Hanna et al. 2014, p. 1,622).

Ectoparasites

• The Iiwi population in the North Windward region of the Hakalau Forest NWR may periodically experience reduced fitness from unusually high levels of parasitism by chewing lice.

According to their study, Freed et al. (2008, p. 1009) noted an unprecedented population explosion of an unidentified specie(s) of chewing lice (family Phthiraptera (Insecta)) between the years 1987 to 2006 on 12 species of both endemic and nonnative forest birds within the Hakalau Forest NWR. The study investigated the recorded observations from periodic surveys, including physical examinations of the 12 forest bird species within ohia and koa-dominated forests in the refuge ranging between the elevations of 5,200 and 6,233 ft (1,585 and 1,900 m). During the first 16 years of the study, from 1987 to 2002, only very occasional observations of chewing lice were noted on only two species of birds, followed by a pronounced increase in chewing lice observations between the years 2003 through 2006 across all 12 species. While the authors could not correlate a change in humidity or change in the host birds’ behavior that might explain the increase in chewing lice, they did record reduced fat levels and increases in broken wing and tail feathers preceding the increase, suggesting possibly a prior food limitation (nutritional stress). The study also noted that the increase in chewing lice coincided temporally with detections of the nonnative Japanese bush warbler (Cettia diphone), common at lower elevations and often with high infestations of lice. The authors suggested a host switching hypothesis as one possible explanation for the explosion of lice observed among the forest birds of Hakalau Forest NWR. Based upon observed major fault bars in wing and tail feathers of infected birds, Freed et al. (2008, p. 1009) correlated nutritive stress with the intensity of infection, indicating an indirect cost to the hosts of being parasitized. Additionally, they noted that birds with lice were less likely to be recaptured than birds without lice (which could imply many scenarios, or simply correlate with birds feeling less apt to fly while infested with lice). Freed et al. (2008, p. 1,009) indicated that Iiwi were one of the 12 species impacted by chewing lice, although additional studies have yet to validate their conclusions (Camp et al. 2010, pp. 12-13). See below, Figures 30 and 31 (from Freed et al. 2008), for graphical information on the fault bar impacts and recapture rates for the affected birds, including Iiwi.

59

Potential Stressors

60

Potential Stressors

Figure 30. Recapture of individual birds identified without chewing lice (white bars) and those with chewing lice (black bars). Data come from 4 study sites used during 2003 to mid-2006. (Image from Freed et al. 2008; used with permission).

Figure 31. Comparison of proportion of individuals with major fault bars of each species in relation to infection rate by chewing lice. White bars indicate birds without lice. Black bars indicate birds infected with lice. Data are from 2003–2005. (Image from Freed et al. 2008; used with permission).

61

Potential Stressors

Small Population Size

• Iiwi generally have low genetic diversity compared to other native Hawaiian honeycreepers, including those that show signs of developing malaria resistance. However, Iiwi have maintained similar levels of genetic variability at the major histocompatibility complex, which is associated with immune response. Nevertheless, the Iiwi remains at risk of increased vulnerability to stochastic events, as populations are lost and redundancy decreases across the island chain. It is widely established that small populations of animals are inherently more vulnerable to extinction because of random demographic fluctuations and stochastic environmental events (Mangel and Tier 1994, p. 607; Pimm et al. 1988, p. 757). Furthermore, island populations are more prone to extinction than mainland populations, largely due to anthropogenic effects including habitat loss, nonnative species introductions, over-exploitation, and reduced genetic variation because of founder effects (Gilpin and Soulé 1986, pp. 24-34; Traill et al. 2009, p. 1-2). The problems associated with small population size and vulnerability to random demographic fluctuations or natural catastrophes are further magnified by synergistic interactions with other threats (see discussion below, “Compounded Cumulative Effects”).

Formerly widespread populations that become small and isolated often exhibit reduced levels of genetic variability, which diminishes the species’ capacity to adapt and respond to environmental changes, thereby lessening the probability of long-term persistence (e.g., Barrett and Kohn 1991, p. 4; Newman and Pilson 1997, p. 361). For example, Atkinson et al. (2009, p. 58) note that the ability of native Hawaiian birds to potentially adapt and develop resistance to avian malaria depends upon the retention of sufficient genetic variability within populations. As populations are lost or decrease in size, genetic variability is reduced, thereby restricting the potential evolutionary capacity to respond to novel stressors such as avian malaria.

Studies with mitochondrial DNA and control region sequences have demonstrated little to no genetic variation in Iiwi; Iiwi on all islands have identical haplotypes (Jarvi et al. 2004, Table 4 and references therein, p. 2164; Foster et al. 2007, p. 4,743). Amakihi, by contrast, exhibit substantial levels of genetic variation both within and between populations located at different elevations, which some researchers hypothesize may be indicative of a potential genetic basis for resistance to avian malaria (e.g., Eggert et al. 2008, entire). Jarvi et al. (2004, p. 2,166) suggest that the Iiwi’s lack of variation in mitochondrial DNA tells us that the species experienced a genetic bottleneck on the order of thousands or tens of thousands of years ago, and not likely more recently, because levels of nuclear variation appear to be comparable between Iiwi and Amakihi (Jarvi et al. 2004, p. 2,166). Jarvi et al. (2004, p. 2,166) also found similar levels of variability between Iiwi and Amakihi at some MHC (major histocompatibility complex) sites, and suggest this indicates Iiwi maintained some genetic variation even in the face of a past genetic bottleneck. Eggert et al. (2008, p. 6) reported that in contrast to past observations of low genetic diversity in Iiwi, levels of allelic diversity and heterozygosity at microsatellite loci in Iiwi were comparable to levels observed in Amakihi and Apapane. Furthermore, they found Iiwi exhibited slight but significant differentiation between those sampled at mid-elevations and one high elevation site (Eggert et al. 2008, p. 6). The sample sizes for these sites were relatively small, however, leading the authors to suggest that these results must be interpreted with caution.

62

Potential Stressors

Nonetheless, these results, in conjunction with the observation that some Iiwi have recovered from malaria and gone on to breed successfully in subsequent years (Freed et al. 2005, p. 759; Eggert et al. 2008, p. 9), indicate it is possible there is some genetic basis for this resistance.

Analysis of minimum viable population (MVP) results for 102 vertebrate species suggests a median 5,000 breeding individuals is needed to maintain species populations over hundreds of years with > 90 percent probability of survival (Traill et al. 2009, p. 3); this is a general “rule of thumb,” however, and does not take into account the unique characteristics of each species or the present demography of populations. According to Frankham et al. (2014, p. 56), the genetically effective population size (Ne) is the number of individuals for an ideal population in which all members of the population are breeding and contributing alleles required to maintain sufficient genetic variation within the population to prevent genetic drift (loss of adaptive potential) and limit loss in fitness from inbreeding depression to 10 percent over five generations. To maintain evolutionary fitness in perpetuity in such an ideal population, it has been suggested that an effective population size between 100 and 1,000 is≤ needed (Frankham et al. 2014, p. 56). A wild population is not the same as this ideal population, however, and as generally only a small subset of the individuals in a wild population contribute alleles to the next generation, Ne is higher for wild populations, often significantly so.

If not already extirpated, the extremely small Iiwi populations on Oahu, Molokai, and west Maui are vulnerable to extinction because of random demographic fluctuations and stochastic environmental events, as well as the likely effects of inbreeding depression. The small Iiwi population on Kauai, ranging between 1,934 and 3,167 individuals with a mean of 2,551 individuals, may be at or below the threshold at which there is no longer sufficient genetic variation within the population to prevent genetic drift (Traill et al. 2010, p. 3). In addition, as noted above, the maintenance of genetic variability is also important to maintain the evolutionary capacity to potentially adapt to novel stressors, such as avian malaria. As discussed in great detail below (see discussion on Climate Change), researchers expect Iiwi will become increasingly vulnerable to malaria as temperatures are projected to increase and shrink the narrow range of suitable and safe habitat. As the number of populations of Iiwi across the Hawaiian Islands grows smaller, the remaining populations are at increasing risk from stochastic environmental events, such as severe typhoons. Should all of the populations become reduced to inhabiting a single island, for example, the entire species would then be vulnerable to any random environmental event that may affect that island (see, for example, “Increase in Severe Storms” in the Climate Change section, below).

Climate Change

• In evaluating climate change as a stressor to Iiwi, we: (1) briefly discuss documented climate trends in the Hawaiian Islands, including the historical warming trend, and current effects on avian disease and on forest birds, including Iiwi; and (2) summarize three studies that project future change in climate and its impact on avian disease and on forest birds, including Iiwi. Under all scenarios, rising temperature increases the maximum elevation where the malaria parasite and its mosquito vector can survive and reproduce, leading to the decline of high-elevation, disease-free habitat. (These projections hold regardless of variation in projected precipitation trends.)

63

Potential Stressors

Both exposure and transmission of avian malaria in Iiwi and other forest birds is projected to increase significantly by mid to late century due to climate change. Because Iiwi are highly susceptible to avian malaria, experiences a high rate of mortality from the disease, and to date, shows no to minimal indication of developing resistance, we expect that the expansion of disease into most upland forest areas in Hawaii will lead to increased disease mortality and complete loss of at least some existing Iiwi populations (e.g., on Kauai and West Maui) during the 21st century. If climate change results in increased drought and drying trends, which some projections indicate, the effects on the ecology of native ohia forest may be significant, and Iiwi will be indirectly affected.

Climate Change—Historical Record—Temperature

Analysis of the historical record indicates surface temperature in Hawaii has been increasing since the early 1900s (Figure 32). The time-series spans about 85 years and shows relatively rapid warming over the past 30 years, with greater temperature increase at high elevation. The average increase since 1975 has been 0.48 oF (0.27oC) per decade for annual mean temperature at elevations above 2,600 ft (800 m) and 0.16oF (0.09oC) per decade for elevations below 800 m (Giambelluca et al. 2008, pp. 3-4). This pattern of amplified warming with elevation is similar to the general global pattern (Wang et al. 2014, pp. 95, 97). Air temperature in Hawaii closely tracked the pattern of variation in sea surface temperature (SST) until recent decades, when the trend of increase in air temperature has greatly exceeded that in SST, suggesting the increasing influence of global warming (relative to the influence of SST) on air temperature (Giambelluca et al. 2008, pp. 2–3).

64

Potential Stressors

Figure 32. Record of temperature in Hawaii since 1910 showing trend of increase since 1975, especially at elevations above 2,600 ft (800 m). (Image from Giambelluca, T.W., H.F. Diaz, and M.S.A. Luke. 2008. Secular temperature changes in Hawaii. Geophysical Research Letters 35, Issue 12. Available online at: http://onlinelibrary.wiley.com/doi/10.1029/2008GL034377/abstract;jsessionid=668BE2D7E1B5A55B74FA313794 C20EAE.f04t02. Used with permission).

Climate Change – Historical Record – Precipitation

Summer (May to October) rainfall in the Hawaiian Islands is governed primarily by the trade winds, global air circulation (notably a pattern called the Hadley circulation), and the islands’ topography. Warm, moist air brought by prevailing northeast trade winds, rises rapidly, cools, and falls as rain mainly on the windward sides of the islands and their constituent mountains, a process known as orographic rainfall. In conjunction, a large air mass that descends from the upper atmosphere (the descending branch of the Hadley circulation) at roughly Hawaii’s latitude limits the vertical extent of cloud formation and also the upper elevation of orographic rainfall. This limit, known as the trade wind inversion, occurs most consistently during the summer and at between roughly 5,000 and 10,000 ft (1,500 and 3,000 m) above sea level (Juvik and Juvik, 1998, pp. 52–53; Cao et al. 2007, p. 1,154). The trade wind inversion also influences the treeline in Hawaii; conditions above the inversion are typically arid, and mesic species such as ohia cannot survive there. Storm fronts are the main driver of precipitation in the winter (November to March), and on the leeward sides of the islands which are typically more arid overall than windward sides of the islands. Local patterns in precipitation

65

Potential Stressors

(among watersheds, for example) are governed by local topography; one side of a valley may be wetter or drier than an adjacent ridge or valley.

Inter-annual and decadal patterns in Hawaii’s rainfall are driven ultimately by patterns of variation in sea surface temperature (SST), including the El Niño Southern Oscillation (ENSO) and the Pacific Decadal Oscillation (PDO). Despite the shift between positive and negative rainfall anomalies in Hawaii on inter-annual and decadal scales (as influenced by ENSO and PDO patterns, respectively), Hawaii has experienced an overall drying trend since the 1920s, with an average annual decline in precipitation of 1.78 percent (Frazier and Giambelluca 2016, p. 4). In addition, the occurrence of the trade wind inversion has increased significantly since 1979, and today it is present more than 80 percent of the time (Cao et al. 2007, p. 1,156). The base height of the inversion is linked to sea surface temperature and atmospheric circulation, but the relationship between the inversion base height and ENSO is not clear (Cao et al. 2007, p. 1,158). Because the trade wind inversion truncates the vertical development of clouds and restricts the amount and distribution of rainfall, the increased frequency of the trade wind inversion may result in overall decreased precipitation, especially at high elevation (Cao et al. 2007, p. 1,158– 1,159). If the area that receives consistent rainfall were reduced, the area where mesic and wet forest, including ohia forest, can survive would be reduced as well. The effect of the trade wind inversion on the location of the treeline depends upon the altitude of the inversion base, and clear trends in this altitude are not evident (Cao et al. 2007, p. 1,156). Indeed, the exact outcome for local forest ecology is likely quite complex as detailed during a recent study on status of the Haleakala silversword (Argyroxyphium sandwicense). During their study, Krushelnycky et al. (2013, pp. 1) examined historical and recent survey data for the silversword and determined that the species has experienced significant mortality on the lower elevation (and wetter) portion of its range as a result of increased occurrence of the trade wind inversion and an overall drying effect on Haleakala. This counterintuitive observation involving the silversword may be due to species variation in drought tolerance across its range, and highlights the complexities concerning climate change predictions in Hawaiian forest ecology.

66

Potential Stressors

Figure 33. Time series of 11-yr running mean (based on the water year from July to June of the following year) for HRI (Hawaii rainfall index) (closed circle) and PDOI (Pacific decadal oscillation index) (open circle). Note the pattern of negative anomaly, or values below zero, in the HRI, i.e., decreasing rainfall, since about 1970. (Image from Chu, P.S., and Y.R. Chen. 2005. Interannual and interdecadal rainfall variations in the Hawaiian Islands. Journal of Climate 18: 4,796-4,813. Used with permission).

Figure 34. Time series of winter (NDJFM) HRI anomalies. Straight line denotes the linear regression line fitted to the records, and indicates a trend of decline in winter rainfall. (Image from Chu, P.S., and Y.R. Chen. 2005. Interannual and interdecadal rainfall variations in the Hawaiian Islands. Journal of Climate 18: 4,796-4,813. Used with permission).

67

Potential Stressors

Climate Change –Temperature, Precipitation, and Elevation Tied to Avian Malaria Transmission

Disease has long been recognized as a strong contributor to observed population decline in malaria-susceptible Hawaiian forest birds, including Iiwi (see Avian Diseases section, above), and our knowledge of the interaction of climate and elevation with disease has improved in recent years (see for example Ahumada et al. 2004; Samuel et al. 2011, 2015). Models that test the influence of temperature, precipitation, and elevation on the epidemiology of avian malaria— the dynamics of the pathogen, its vector, and its infection of forest birds—accurately predict the observed patterns of disease in forest bird habitat today on Hawaii Island (Samuel et al. 2011, p. 2967–2968) and Kauai (Atkinson et al. 2014, pp. 2434–2435). On Hawaii Island, transmission is high at low elevation where temperature is most consistently the warmest; episodic at mid elevation and variable among seasons and sites in response to mosquito dynamics mediated by precipitation and temperature; and absent from cooler, high elevation forest, except briefly during summer when the climate is suitable for Plasmodium and mosquito development (Samuel et al. 2011, p. 2970). In the high-elevation Alakai Plateau on Kauai, the range of avian malaria prevalence has increased in altitude in the past two decades, as measured by changing infection rates in blood samples from forest birds, concomitant with increasing ambient temperature and changes in streamflow that affect larval habitat for mosquitos (Atkinson et al. 2014, entire). These modeled and documented patterns provide strong evidence that climate conditions drive mosquito populations, development of Plasmodium within mosquitoes, and malaria prevalence in Iiwi and other susceptible forest birds along a gradient of elevation (Samuel et al. 2011, p. 2970).

Climate Change – Impacts to Iiwi with Documented Increase in Elevation of Malaria Transmission

The replication of the avian malaria parasite and reproduction of its mosquito vector are both temperature-limited (Ahumada et al. 2004, pp. 1164–1165; LaPointe et al. 2009, p. 322). Under current conditions, the presence of avian malaria limits the distribution of Iiwi at low and mid elevations where mosquitoes and Plasmodium can survive and reproduce (Paxton et al. 2013, p. 27). Increasing temperature at high elevations in Hawaii with global climate change creates conditions for the disease to move upslope; increase the total area where Iiwi and other forest birds may be exposed; and reduce the area of high-elevation, disease-free habitat. For example, the increasing temperature at high elevation and concomitant increase in the prevalence of avian malaria is well-documented on Kauai (Atkinson et al. 2014b, p. 2,432), where Iiwi are declining rapidly (Paxton et al. 2013, p. 13).

Although the observed prevalence of avian malaria in Iiwi on Kauai has since decreased (Atkinson et al. 2014b, pp. 2,430–2,431, 2434), this may reflect the high susceptibility of the species to the disease and the high mortality rate (93 to 95 percent mortality; Atkinson et al. 1995, p. S63; Samuel et al. 2015, p. 12). In other words, at extremely high mortality rates, such as those observed in Iiwi, few birds would survive as carriers, and of those, even fewer would likely be captured in a sampling effort simply owing to chance. Therefore, unlike the Amakihi, which has developed some resistance (see discussion in Avian Diseases section, above), Iiwi are a poor indicator of the prevalence of avian malaria, because the majority of birds that contract

68

Potential Stressors

the disease die and escape detection. In addition, in a recent paper, scientists report that trends of decline in the abundance and contraction in the range of iiwi and other disease-susceptible forest birds on the island of Kauai have steepened sharply since 2000 (Paxton et al. 2016, p. 2). Kauai may have passed a “tipping point,” where increasing temperature now permits the survival and reproduction of both avian disease and its vector in forest bird habitat at all elevations; birds are exposed to mosquito-borne disease throughout their remaining range on the island (Paxton et al. 2016, pp. 3, 5).

Climate Change – Projected Future Temperature in Hawaii

Regionally downscaled climate models for Hawaii, based on three scenarios of greenhouse gas emissions, project warming across the islands by the end of the century with disproportionately greater warming at high elevation (Fortini et al. 2015, p. 5; Liao et al. 2015, pp. 4,344–4,345). Two of these emissions scenarios, Representative Concentration Pathway (RCP) 8.5 and RCP 4.5, were part of a suite of scenarios developed for the most recent assessment report by the Intergovernmental Panel on Climate Change (AR5; IPCC 2013a, p. 1,461). These scenarios represent continued increase in emissions of greenhouse gasses this century, RCP 8.5 represents “business as usual” with no decrease, while RCP 4.5 represents a gradual decrease of emissions and stabilization by the end of the 21st century (Van Vuuren et al. 2011, pp. 12, 21). These scenarios were modeled for Hawaii using a statistical downscaling method (Timm et al. 2015, pp. 93-99). The third scenario, A1B, from an earlier generation of modeling, represents emissions declining after mid-century (IPCC 2000, pp. 6–9, IPCC 2007, p. 44). The A1B scenario was modeled for Hawaii using a dynamical downscaling method (Zhang et al. 2012, pp. 3,260-3,274). The projected warming at high elevation in Hawaii ranges from 3.9 to 7.7oF (2.2 to 4.3oC), and the projected moisture regime varies (Liao et al. 2015, p. 4344; Tables 1 and 2). The future conditions in Hawaii projected under these three emissions scenarios are characterized as relatively hot and dry (RCP8.5), warm and dry (RCP4.5), and warm and wet (A1B) (Liao et al. 2015, pp. 4344–4345), with much of the difference in future projected precipitation attributable to the downscaling method used, rather than the emissions scenario.

Climate Change – Projected Future Precipitation in Hawaii

Based on available regional climate projections, future patterns of temperature increase in Hawaii are more certain than potential shifts in precipitation patterns (summarized in Loope and Giambelluca 1998, p. 374; Liao et al. 2015, p. 4,345; Table 2). For example, the dynamic downscaled projection shows a general pattern of increased precipitation in windward areas of Hawaii Island and Maui, and slight drying in areas that are already dry today (Zhang et al. 2016 in press; future projections summarized in Fortini et al. 2015, p. 5); the statistical downscaled prjections show a more pronounced relative drying of leeward areas, and a general pattern of drier wet seasons and much drier dry seasons (Timm et al. 2015, pg. 107; Figure 13). Additionally, some future projections suggest that areas that currently are wet (windward sides of islands) will experience greater rainfall and more extreme rainfall events, while currently dry areas (leeward sides and high elevations) will become drier (Zhang et al. 2016, pp. 8,350–8,351). Whether the documented increasing trade wind inversion frequency is related to a warming climate is currently unclear (Cao et al. 2007, p. 1158), and consequently, whether montane forest

69

Potential Stressors

Iiwi habitat will continue to get drier due to reductions in convective rainfall (Cao et al. 2007, pp. 1,158–1,159; Giambelluca et al. 2008, p. 5). To summarize, the results of models that project climate change effects on Hawaiian ecosystems, including avian malaria transmission risk for Iiwi, have greater confidence when based principally on temperature change rather than precipitation change.

Climate Change – Projected Future Impacts to Iiwi Increased Elevation of Malaria Transmission

• Avian malaria transmission rates will increase and move upslope with projected increase in temperature. The negative impacts to Iiwi during this century will likely be significant. Malaria-free areas that currently exist in high-elevation forests are predicted to shrink or disappear by the end of the century.

We examined three studies that projected climate change impacts on avian malaria distribution and prevalence in Hawaii (Benning et al. 2002; Fortini et al. 2015; Liao et al. 2015) to evaluate the likely future impacts of climate change on disease, which we have identified as the primary stressor to Iiwi. All three models agree that rising temperature will result in a continued increase in the elevation where transmission of avian malaria can occur and an accompanying decline in disease-free habitat for Iiwi and other susceptible honeycreepers. Under some scenarios, disease-free habitat is entirely eliminated by the end of the century. The results of each study are summarized below.

In addition, based on statistical modeling to project documented trends of decline and range contraction in iiwi and other forest birds on Kauai (described above), Paxton et al. (2016, pp. 2–3) estimate the latest time to extirpation for iiwi on the island at 2050. The maximum elevation of Kauai is lower than that of either Maui or Hawaii Island, where similar trends of increase in temperature and the elevation of disease transmission are well documented, as discussed above. Iiwi, and other disease-susceptible honeycreepers, only persist in abundance on these higher islands in high-elevation, disease-free habitat that is shrinking with increasing temperature. Disease-free habitat on Maui and Hawaii Island are experiencing similar trends as that on Kauai, but owing their higher total elevation, the ultimate loss of disease-free habitat will take somewhat longer.

Benning et al. 2002 Examining three areas in Hawaii that currently have intact native forest and a high abundance of native birds, Benning et al. (2002) used GIS simulations to map the isotherms (lines connecting points of equal temperature) for three temperatures zones critical for Plasmodium development, including the minimum at 55oF (13oC), transitional development between 55oF (13oC) and 63oF (17oC) and peak plasmodium development above 63oF (17oC), and analyzed projected changes in disease free forest area at the three sites under a temperature increase of 3.6oF (2oC) over the next 100 years. The three areas, all managed for conservation, are Hakalau Forest NWR, located on the windward side of Mauna Kea on Hawaii Island; Hanawi Forest, located on windward east Maui; and the Alakai Swamp, located on the highest plateaus of central Kauai. Projected climate warming nearly eliminated the area of forest with

70

Potential Stressors

low risk of malaria transmission at Hakalau Forest NWR and on Kauai’s Alakai Plateau, and reduces low-risk forest area by nearly half in Hanawi on windward east Maui (Benning et al. 2002, p. 14,247; Figure 35). Hanawi retained some area of low transmission risk because the highest elevation forest still exists on Maui, whereas the forest above Hakalau Forest NWR (> 6,200 ft (1,900 m)) was lost decades ago to logging and grazing, and has since been replaced by invasive nonnative plants.

Even under optimistic assumptions, malaria-susceptible Hawaiian forest birds, including Iiwi, lose most of their disease-free habitat in the three areas considered in this projection of climate change impacts (Benning et al. 2002, p. 14247–14248). Today, the temperature increase of 3.6o F (2oC) projected by Benning et al. is considered optimistic both for Hawaii (e.g., Fortini et al. 2015, p. 5; Liao et al. 2015, pp. 4344–4345) and globally (IPCC 2013a, p. 20; Gütschow et al. 2015). Furthermore, analysis of the historical record and additional modeling efforts point to the complexities of projecting future patterns in the trade wind inversion and in precipitation in Hawaii, as described above. Because the assumption that the temperature increase during this century would be limited to 3.6oF (2oC) is optimistic, and the assumption that rainfall will increase in elevation is highly uncertain, climate change impacts to disease-free habitat could be considerably worse than those projected by Benning et al. (2002).

71

Potential Stressors

Figure 35. Projected changes in forest cover in relation to 17°C (yellow) and 13°C (white) isotherms under current and 2°C warming conditions. Boxes A: Changes are shown for Hanawi Reserve (blue boundary) on the island of Maui; Boxes B: Hakalau Forest NWR (blue boundary) on Hawaii; and Boxes C: the Alakai swamp region on the island of Kauai. Images were created by using a 1995 Satellite Pour l’Observation de la Terre (SPOT) false-color composite image draped over 30-m digital elevation models for each island. Areas below the yellow isotherm are at high risk for malaria transmission, between the yellow and white lines transmission is possible but limited, and areas above the white isotherm are low-risk. (Image from Benning, T.L., D.A. LaPointe, C. T. Atkinson, and P.M. Vitousek. 2002. Interactions of climate change with biological invasions and land use in the Hawaiian Islands: modeling the fate of endemic birds using a geographic information system. Proceedings of the National Academy of Science 99: 14,246- 14,249. Used with permission).

72

Potential Stressors

Fortini et al. 2015 Decades of species surveys, current and projected patterns of temperature and rainfall in Hawaii, and primary habitat availability were used to determine the future distribution of Iiwi and 19 other native forest birds. Projections of temperature increase in Hawaii from the Hawaiian Regional Climate Model (HRCM), based on the A1B emissions scenario, showed an average warming in the islands of 4.5oF (2.5oC) with an increase of 6.1oF (3.4oC) at higher elevations by the end of the century (Fortini et al. 2015, p. 5). Hawaiian forest birds are reliant on vegetation and other biotic aspects of habitat as well as particular climate conditions, so this effort examined climate-based range shifts with respect to the distribution of currently available habitat for each species. All 20 Hawaiian forest bird species were projected to suffer range loss by the end of the century as a likely result of increased transmission of avian malaria at higher elevations with increasing temperature. While the model reliability varied among species, the model for Iiwi was among those with high reliability (Fortini et al. 2015, p. 7–8). Iiwi was predicted to lose approximately 60 percent, or 470 mi2 (1,214 km2) of its current range, and climate conditions where the species can persist would shift up in elevation, including into areas that are not currently forested, such as lava flows and high-elevation grasslands (Fortini et al. 2015, p. 9–10; Figure 36). The existing habitat area available for a range shift upslope is approximately 40 mi2 (105 km2). Thus, the projected area of Iiwi range in 2100 would be approximately 287 mi2 (743 km2), or 40 percent of what it is today, and most of this would occur on Hawaii Island (Fortini et al. 2015, p. 9, Supplement 6).

73

Potential Stressors

Figure 36. This map shows projected climate-based Iiwi range decrease between the early 21st century (1990– 2010) and the later 21st century (2080–2100). The gridded overlay represents the distribution of primary vegetation types associated with the species. The yellow area identified as “gained range” includes ohia scrub and may not be suitable habitat. (Image from Fortini, L.B., A.E. Vorsino, F.A. Amidon, E.H. Paxton, and J.D. Jacobi. 2015. Large- Scale Range Collapse of Hawaiian Forest Birds under Climate Change and the Need for 21st Century Conservation Options. PloS one, 10(10): e0140389. See also Figures A6-A11. Used with permission).

74

Potential Stressors

Liao et al. 2015 Using projected changes in temperature and precipitation from two downscaling techniques for generating localized analyses or perspectives from global models, and three different emissions scenarios frequently used in climate projections, Liao et al. (2015) projected future malaria risk for three species of Hawaiian forest birds. Irrespective of the scenario modeled, by mid-century (roughly 2040), malaria transmission rates and impacts to bird populations were projected to begin increasing at higher elevations. By the end of the 21st century, increasing temperature was projected to result in increasing mosquito abundance and more rapid development of both mosquitoes and Plasmodium at higher elevation compared with current patterns (Liao et al. 2015, p. 4,345), with an annual malaria transmission rate for Iiwi between 50 percent (under RCP4.5) and 90 percent (under A1B or RCP8.5) (Liao et al. 2015, p. 4,348; Figure 37). Under the most optimistic scenario (RCP4.5), by the end of the century the malaria transmission rate at higher elevation would be similar to the current transmission rate at middle elevations. That is, the transmission rate at higher elevations will be similar to the rate in areas from which Iiwi was recently extirpated, such as south Kau on Hawaii Island, and where the species is declining now, such as Kauai and central windward Hawaii Island (Liao et al. 2015, p. 4346). Under these conditions, population decline in Iiwi was projected to reach 70 to 90 percent by 2100, depending on the emissions scenario (Liao et al. 2015, p. 4,347).

75

Potential Stressors

Fig. 37. Predicted 10-year mean malaria prevalence for three Hawaiian honeycreepers [Apapane (a, d), Iiwi (b, e), and Amakihi (c, f, g)] under three climate change projections [RCP8.5 (solid line), A1B (dot line), and RCP4.5 (dash line)] at high and mid-elevations. Because only malaria-tolerant Amakihi (g) exist at warm lowland, only Amakihi are shown in low-elevation forests. (Image from Liao, W., O.E. Timm, C. Zhang, C.T. Atkinson, D.A. LaPointe, and M.D. Samuel. 2015. Will a warmer and wetter future cause extinction of native Hawaiian forest birds? Global Change Biology 21: 4,342–4,352, doi: 10.1111/gcb.13005. Used with permission).

76

Potential Stressors

Climate Change—Projected Increase in Severe Storms

Anomalous increases in sea surface temperatures, such as ENSO events, have increased in frequency (Chu and Wang 1997, p. 2,683) as consequence of warming of ocean temperatures (Smith and Reynolds 2004, p. 2,466) contributing to increased tropical cyclone activity around Hawaii (Murakami et al. 2013, p. 749; see also, Figure 38). Climate models predict that warming of ocean temperatures in the central Pacific basin will contribute to increased frequency of tropical cyclones in the vicinity of the Hawaiian Islands (Clark and Chu 2002, p. 403; Webster et al. 2005, p. 1,844; Chu et al. 2010, p. 4,881; Chan and Chan 2012, p. 811; Murakami et al. 2013, p. 749). Analysis of past frequencies of storm tracks near Hawaii strongly suggests that global climate change can powerfully influence storm frequency in the Pacific irrespective of natural variation in other oceanographic influences, such as the Pacific Decadal Oscillation, the Interdecadal Pacific Oscillation, the Atlantic Multi-decadal Oscillation, and ENSO (Murakami et al. 2015, p. 118). Severe storms can result in defoliation and death of ohia and other native trees over broad areas (Harrington et al. 1997, pp. 539-540). Temporary loss of nectar sources as the result of strong winds from hurricanes has been shown to have negative population effects on nectarivorous species of island birds (Waide 1991, p. 475; Wiley and Wunderle, Jr. 1993, p. 319). In addition, prolonged periods of rain or high winds have resulted in nest failure and reduction in daily survival rate for nests of Hawaiian forest birds (Woodworth and Pratt 2009, p. 211; Chu et al. 2010, p. 4,889; Cummins et al. 2014, p. 17).

Figure 38. The Hawaiian Islands framed by composite infrared images of 15 separate tropical cyclones that entered the Central Pacific Ocean between June 1 and October 20, 2015. (Image by Kevin Kodama, National Weather Service, Honolulu Office. Used with permission)

77

Potential Stressors

Climate Change—Effects on Ohia Forest Habitat

The effects of climate change on the high-elevation, mesic or wet ohia forest on which Iiwi depends are uncertain. Projections that include continued drying as well as increasing temperature may result in the degradation or loss of area currently forested (Cao et al. 2007, p. 1,159). If the increasing trade wind inversion frequency continues into the future, this would limit the establishment of forest above the current treeline, capping the range where Iiwi can persist as avian malaria moves upslope (Paxton et al. 2013, p. 27). Conversely, if trade wind inversion rises in altitude with increasing temperature, rainfall may move upslope, with a potential rise in the treeline (Benning et al. 2002, p. 14247), which could benefit Iiwi as mid- elevation habitat is lost to disease, if “new” forest habitat above the current treeline (which would require many decades to develop) remained above the elevation where malaria transmission occurs. Adding to the complexity, studies in some areas have shown an unexpected outcome where the cloud base moved downslope in response to changes in the trade wind inversion as was demonstrated in the Canary Islands by Sperling et al. (2004, p. 103). Furthermore, climate change may also increase native forests’ vulnerability to invasion by nonnative plant species that modify Hawaiian ecosystems (Loope and Giambelluca 1998, p. 374; Vorsino et al. 2014, p. 14), including species such as strawberry guava (Psidium cattleianum) and miconia (Miconia calvescens) that have degraded ohia forest on a large scale (USDA 2013; Hawaii Invasive Species Council 2015; see also “Stressors – Ohia Forest,” above). Increased vulnerability to invasion could result from long-term changes in temperature and precipitation that are detrimental to native plants, beneficial to invasive nonnative plants, or both, and from an increase in the frequency of severe storms, disturbances that can facilitate invasion through changes in forest structure and through upslope dispersal of propagules (Perlman 1992, pp. 1–9; Kitayama and Mueller-Dombois 1995, p. 671; Businger 1998, pp. 2, 6; Mitchell et al. 2005a, pp. 3–4). Regardless of modeling approach and future temperature scenarios considered, all available studies show the implication of warming can be severe to Iiwi, impacting at least half of it range or more.

Climate Change—Summary

The historical record makes it apparent that climate change in Hawaii is already occurring. Similar to the rest of the planet, the clearest and most consistent signal in Hawaii is a trend of increased temperature, with the greatest increases documented at higher elevations. Temperature increase in Hawaii is associated with an increase in the elevation where the avian malaria parasite, Plasmodium relictum, can survive and replicate, followed by increases in transmission of the disease and the decline in Iiwi and other malaria-susceptible forest birds. The results of several models projecting the impacts of climate change on Hawaiian forest birds agree that the disease-free forest habitat that currently remains at the highest elevations in the islands will be lost or severely restricted during this century. Other aspects of climate change, such as the amount and distribution of precipitation and the frequency and intensity of severe storms, and their impacts on Iiwi and the species’ habitat, are projected with less certainty. However, the potential exists for climate change to exacerbate other stressors to Iiwi in addition to disease, including loss or degradation of native ohia forest and increased nest failure resulting from severe storms. As Iiwi’s numbers and distribution continue to decline, small isolated populations may become increasingly vulnerable to environmental catastrophes and demographic stochasticity, and face increased risk of genetic drift and loss of adaptive potential.

78

Conservation Measures

4. Conservation Measures to Address Stressors to Iiwi

Of the Iiwi stressors identified and considered here, there is abundant evidence to suggest that the impacts from avian disease (currently) and habitat loss (historically) have been the most profound. There is strong evidence to suggest avian disease is largely responsible for driving current Iiwi population declines and range contraction throughout its distribution. Ungulate and rodent control and forest habitat protection, conservation and restoration provide some benefit to Iiwi by protecting and enhancing breeding and foraging areas and reducing mosquito breeding sites. Climate models, however, suggest areas that currently provide high elevation habitat free from avian malaria transmission will no longer serve as a haven for Iiwi from avian diseases by the end of the century, and possibly sooner. Although habitat management may contribute to slowing the spread of avian disease, survival of Iiwi and other Hawaiian honeycreepers will likely ultimately depend on measures that reduce or eliminate disease transmission, and the possibility that native birds may develop disease tolerance provided sufficient time. Here we describe the conservation measures in place that attempt to address the stressors to Iiwi.

Conservation Measures to Address Habitat Loss and Degradation

Impacts to Ohia Trees—Ohia Rust—Research and Conservation Measures

Presently, there is no approved fungicide that can be used in controlling ohia rust. The Hawaii Department of Agriculture recommends good sanitation practices when dealing with infected ohia including removing and bagging and / or destroying infected leaves or other plant parts as soon as symptoms appear. In 2007, to prevent the introduction of additional and possibly more virulent strains of ohia rust into Hawaii, the State passed a new law, Plant Quarantine Interim Rule 07-2, which restricts the importation of all plants and plant parts in the family Myrtaceae from areas known to be infested with ohia rust. South America and the States of Florida and California are three areas identified in the rule as infested with ohia rust (HDOA 2007, pp. 1-2).

Impacts to Ohia Trees—Rapid Ohia Death—Research and Conservation Measures

In June of 2015, the Hawaii Department of Agriculture (HDOA) agreed to collect samples from landowners that suspect their trees are infected with ROD from any island. A State Department of Agriculture website was created and launched in July 2015 to disseminate information on how to prevent the spread of ROD from one area to the next on Hawaii Island, or to neighboring islands. In August 2015, the Hawaii Board of Agriculture announced a new interim rule (http://hdoa.hawaii.gov/blog/main/ohiaquarantine/) that imposes a quarantine on the intrastate movement of ohia plants and plant parts, including flowers, leaves, seeds, stems, twigs, cuttings, untreated wood, logs, mulch green waste, and frass (sawdust from boring beetles) from Hawaii Island. The interim rule stipulates that transport of such items may be only conducted by first obtaining a permit issued by the HDOA, and will also restrict the movement of soil from Hawaii Island beginning in January 2016. In

79

Conservation Measures

March of 2016, the Hawaii Board of Agriculture approved permit conditions for the movement of all soil from Hawaii Island, and in November 2016, the to the quarantine rule on interisland movement of ohia to prevent the spread of ROD was made permanent. (CTAHR 2016; http://www2.ctahr.hawaii.edu/forestry/disease/ohia_wilt.html).

Conservation Measures to Protect Ohia Forest Habitat

• During the past four decades, large areas of mid- and high-elevation native forest on Kauai, Maui, and Hawaii Islands have been managed with fencing, removal of ungulates, and control of nonnative plants (Pratt et al. 2009, pp. 569-570; Price et al. 2009, pp. 394-395; Camp et al. 2010, p. 196). These actions have reduced mosquito breeding sites and benefited native forest by slowing both spread of nonnative plants and feral ungulates, and several studies have demonstrated a positive correlation between management and benefit to the Iiwi and other native forest birds.

The 38,030 ac (15,390 ha) Hakalau Forest NWR on the windward slope of Mauna Kea Volcano is the largest protected and actively managed area of mid- to high-elevation rain forest on Hawaii Island, and arguably the most studied Iiwi population site. Camp et al. (2010, p. 196) examined long- and short-term trajectories of Hawaiian forest birds at Hakalau from 1987-2007 in reforested pastureland between 5,250 and 6,562 ft (1,650 and 2,000 m), heavily grazed open forest that was recovering between 4,593 and 6,300 (1,400 and 1,920 m), and lightly grazed closed forest that was relatively intact between 4,953 and 5,577 ft (1,400 and 1,700 m) in elevation. According to the study, overall, long-term population trends of Iiwi in Hakalau were stable or increasing, which contrasted with observed declines in most other areas of Hawaii over the same period, indicating long-term broad-scale habitat management is beneficial for Hawaiian forest birds. Camp et al. (2010, p. 201) also noted since 1999, a decline of Iiwi in the open forest study area and a stable to decreasing Iiwi trajectory within the closed forest area. The study suggested that observed declines may be tied to an increase in one or more established stressors including upslope expansion of mosquitoes and avian disease as noted by Freed et al. (2005, pp. 759-760) and Freed and Cann (2013, p. 1); continued seasonal movements of Iiwi to lower elevations and subsequent exposure to avian disease (Kuntz 2008, p. 1; Guillaumet et al. 2016, p. 192; LaPointe et al. 2015, p. 1); or population growth of competing nonnative birds such as the Japanese white eye (Freed et al. 2008, p. 1,018; Freed and Cann 2014, p. 1).

Within the Hakalau Forest NWR closed-forest study area, sampled only after 1999, densities of both the Hawaii creeper (Oreomystis mana) and Hawaii Akepa (Loxops coccineus coccineus) showed evidence of increased populations following management of lower elevation forested areas. Hawaii creeper and Hawaii Akepa are insectivorous and maintain year-round residence in a local area, and unlike Iiwi, do not move seasonally to lower elevations where they might be exposed to mosquitoes. This finding suggests ungulate management (fencing and ungulate removal) in middle elevation forests is important for reducing mosquito breeding sites from which mosquito dispersal into higher elevation forests may impact native forest birds (Aruch et al. 2007, p. 573; LaPointe 2008, p. 600; LaPointe et al. (2009, p. 409). Protection of native ohia forests at middle and lower elevations is important to provide habitat should potentially disease-tolerant Iiwi populations be discovered at these elevations (LaPointe et al. 2009, p. 420; Price et al. 2009, pp. 382-383). Because of known mosquito dispersal distances

80

Conservation Measures

(mean 1 mi (1.6 km)), to maintain disease free habitat for the Iiwi, large contiguous blocks of native forest should be managed optimally, particularly for the effective control of feral pigs. Likewise, small, managed disease-free areas potentially could be created with the same goal, for example in kipukas (islands of forest vegetation) isolated from mosquito breeding sources by large expanses of lava flows.

Protection of High Elevation Habitats

There are no specific regulatory mechanisms governing the protection of high altitude areas as refugia for the Iiwi or other Hawaiian forest birds, to buffer them against the projected upslope advance of avian disease due to rising global temperatures. There are, however, several institutions and agencies responsible for the protection and / or administration of high elevation habitat areas under their ownership and / or management. These lands include both areas currently occupied by Iiwi and areas unoccupied but located at elevations adjacent to current populations. On the island of Kauai, the State of Hawaii owns and manages the Alakai Wilderness Preserve currently occupied by Iiwi. On the island of Maui, portions of both East Maui Iiwi overlap with Haleakalala National Park, the State of Hawaii’s Hanawi Natural Area Reserve (NAR), and The Nature Conservancy’s Waikamoi Preserve. On Hawaii Island, the various Iiwi populations overlap with and occur on lands identified for protection including Hawaii Volcanoes National Park, eleven different State of Hawaii NARS, several Hawaii State Parks, and the Hakalau National Fish and Wildlife Refuge. Last, but not least, the US Army’s Pohakuloa Training Area (PTA), although not specifically managed for conservation (live fire training does occur there as well), is covered by an Integrated Natural Resources Management Plan (INRMP), which provides for the conservation and rehabilitation of natural resources within PTA. See Figure 39, below, which shows Iiwi population site overlap with all the aforementioned protected high elevation forest habitats for the islands of Kauai, Maui, and Hawaii Island.

81

Conservation Measures

Figure 39. Showing Iiwi estimated range and overlap with managed, protected areas on the Kauai, Oahu, Maui, and Hawaii Island. (Image from VanderWerf, E.A. 2012. Hawaiian Bird Conservation Action Plan. Pacific Rim Conservation, Honolulu, HI, 140 pp. Used with permission)

Using the climate projections from Fortini et al. (2015), we estimated the change in protected or managed areas within climatically suitable Iiwi habitat over time. This analysis led to the conclusion that there will be a net loss of approximately 58 percent of managed, protected areas due to climate change across the Iiwi’s range as a result of range shifts during this century. See, Figures A1 to A14, Appendix A, below, for additional maps showing Iiwi range contraction during this century as well as shifts in usage habitat type and by management classification.

82

Conservation Measures

Iiwi Habitat Protection and Conservation – Restoration

One of the success stories of native forest restoration in Hawaii includes the rehabilitation of formerly grazed pastureland into native forest through the outplanting of koa and ohia trees and common understory plants, as has been ongoing at Hakalau Forest NWR since the late 1980s (Camp et al. 2010, p. 196; USFWS 2012, pp. 9, 20, 99, 116; Borneman et al. 2015, p. 1). Studies of these management activities have demonstrated a strong correlation between increased native forest cover in these restored areas and increased observations of Iiwi and other Hawaiian forest birds. These restoration efforts have been lauded for providing habitat at higher elevations where lower temperatures constrain development of Plasmodium in the mosquito vector, a necessary step toward conserving and helping native birds vulnerable to avian malaria and other diseases. This work is particularly important in light of current climate change projections and the need for these birds species to shift upslope (LaPointe 2008, p. 600).

Reforestation efforts in Hakalau Forest NWR and elsewhere have resulted in the establishment of large tracts of young ohia and koa forest up to elevations of 6,562 ft (2,000 m), now being utilized by Iiwi and other native forest birds (Camp et al. 2010, p. 196; Borneman et al. 2015, p. 1). Compared to koa seedlings, one challenge to outplanting ohia involves its slow growth rate, ultimately requiring many years to grow to large-sized trees with substantial production of ohia flowers. Despite this challenge, some recent efforts have focused on forest protection and reforestation of the heavily grazed Kanakaleonui corridor that connects the upper sections of Hakalau Forest NWR and higher elevation mamane forest, which should benefit Iiwi and other Hawaiian forest birds by creating a migration corridor for birds to access high elevation food sources such as the mamane forests found on the lower saddle slopes of Mauna Kea. There may be limits, however, to the elevation at which ohia will grow. Additionally, the trade wind inversion above Hakalau Forest NWR currently limits vertical cloud development at around 7,300 ft (2,225 m) elevation, dramatically limiting rainfall above the trade wind inversion (Cao et al. 2007, pp. 1,158–1,159). Climate change studies have projected possible scenarios including a continuing trend of warming at higher elevations and increased frequency of occurrence and/or lowering of the trade wind inversion layer (Diaz et al. 2011, pp. 7–8). Depending upon the final outcome, for example, if higher elevation temperatures continue to rise and the trade wind conversion layer decreases in altitude and/or increases in frequency, ohia forest growth may be limited, and thus, the potential area of suitable habitat that might be restored for Iiwi at higher elevations.

Conservation Measures to Address Avian Malaria

Existing tools and approaches have proved largely ineffective to control the mosquito vector, Culex quinquefasciatus, given its dispersal distance of at least 1 mi (1.6 km) and the abundance of mosquito breeding sites in most wet native forest habitats (LaPointe et al. 2009, pp. 408; 411-412; LaPointe et al. 2012, p. 215). To be fair no single agency or collaborative effort has ever attempted to control C. quinquefasciatus or other species of mosquitoes in Hawaii on a landscape scale. However, there is progress in some forest areas using traditional methods, including fencing and ungulate removal, and efforts to remove and minimize the creation of standing water sources associated with human activities (LaPointe et al. 2009, p. 412; LaPointe et al. 2012, p. 219). New opportunities are emerging, however, such as large-scale vector

83

Conservation Measures

control using new genetics technologytools, that have the potential to assist Hawaiian forest birds (LaPointe et al. 2009, pp. 416–417; Reeves et al. 2014, p. e97557; Gantz et al. 2015, pp. E6736– E6743; Fischer in press, pp. 1–2). The most promising of these new tools forego chemicals as a means of lethal control and directly manipulate the viability (or fitness) of the mosquitoes and can be grouped into two broad categories: the Sterile Insect Technique (SIT) and the Population Replacement Technique (PRT) (Fischer in press, pp. 1–2). These tools have positive attributes associated that set them apart from traditional mosquito control options. These new approaches have the potential to achieve landscape scale control, are species specific, and are more effective against dispersed, cryptic, and hard to-reach targets such as the Culex mosquitoes that carry avian malaria in Hawaiian forests (Alphey et al. 2010, pp. 297–299). While promising, these new technologies are still in the research and planning stage and have yet to be implemented.

Conservation Measures to Address Predation

Control of rats in the wild has produced benefits for some Hawaiian forest birds, including possibly the Iiwi (see above, “Predation—Rats”). Field studies have shown localized rat control around nests of Puaiohi and Oahu Elepaio improved nest success and reduced numbers of incubating females killed at the nest (VanderWerf and Smith 2002; Tweed et al. 2006). For Oahu Elepaio, predator control improved population demographics and reversed declining population trajectories (VanderWerf and Smith 2002). As described previously, one study of experimental predator control at Hakalau Forest NWR provided anecdotal evidence of enhanced survival of juvenile Iiwi (Lindsey et al. 2009, pp. 280-282). It has also been suggested that predator control could facilitate the evolution of resistance to avian malaria in Hawaiian birds by improving reproductive success and survival of the remaining scattered, small populations of several forest birds, essentially buying additional time to allow disease tolerant populations to expand more quickly (Kilpatrick 2006, p. 475). Several studies (Clout and Russell 2006, entire; Howald et al. 2007, entire; Witmer et al. 2011, entire) have firmly established the effectiveness landscape-scale control of rats; however, the economics and logistics necessary to apply such methodology to large, contiguous areas in the main Hawaiian Islands still remains a challenge. If further study of rat predation impacts upon Iiwi indicates that this stressor is more significant than we currently understand, we would need to be better prepared to determine the correct application and balance of limited funding and other resources to address predation and the other know impacts. Additionally, it is possible that smaller scale applications of rodent and predator control within select conservation areas may prove more beneficial to the Iiwi or other native forest birds as part of an overall, wide- ranging approach to protect the species across its distribution. To improve the future successful implementation of any of these predator control approaches, the U.S. Fish and Wildlfife Service and the Hawaii Department of Land and Natural Resources are currently developing a programmatic environmental impact statement (PEIS) that will describe principles and methods for controlling or eradicating rodents and mongooses in conservation areas to protect native species. While the PEIS would not initiate any specific project, the document will develop a framework, including regulatory, biological, cultural, and political components that should facilitate implementation of efforts by any and all conservation partners in the future. For example, the PEIS will include steps and recommendations to streamline and improve upon the

84

Conservation Measures application process involved with predator control operations, including preparing an environmental impact statement.

Conservation Measures to Address Competition

Freed and Cann (2012, p. 15; 2014, pp. 30-31; 2015, p. 1) suggested that removal of nonnative bird competitors including the Japanese white-eye, (introduced to Hawaii in the 1950s for insect control), might reverse the effects of diffuse competition allegedly observed at Hakalau Forest NWR. However, the evidence presented by the authors for declines of Iiwi and other native birds linked to competition was unsupported by other research (Camp et al. 2010, p. 201; Paxton et al. 2013, p. 1; Camp et al. 2014, p. 97; Camp et al. 2015, p. 2), and to date, no examples exist of feasible and effective removal of avian competitors on a large scale as a management strategy.

Since the early 1990s, studies have investigated various measures for controlling colonies of the western yellowjacket wasp in native Hawaiian forests (Gambino and Loope 1992, p. iv, 37-40). More recent studies, Foote et al. (2011, p. 1, 6-11), Hanna et al. (2011, p. 1,026), and Hanna et al. 2014, pp. 1,623, 1,629) indicate that large scale control of the wasp is seasonally possible with specific applications of fipronil bait.

Conservation Measures to Mitigate Climate Change

Climate Initiatives – United States In 2009, the Environmental Protection Agency (EPA) published final “Endangerment and Cause or Contribute Findings” under Section 202(a) of the Clean Air Act, finding that six key greenhouse gases constitute a threat to public health and welfare, and that the combined emissions from motor vehicles cause and contribute to greenhouse gas pollution (74 FR 66496; December 15, 2009). The EPA’s findings concluded that greenhouse gas pollution threatens Americans' health and welfare by leading to long-lasting changes in our climate that can have a range of negative effects on human health and the environment. Although the findings did not themselves impose any requirements on industry or other entities, this action was a prerequisite for implementing greenhouse gas emissions standards for vehicles, and laid the groundwork for other subsequent regulatory changes as well. One of the first new regulations to address greenhouse gas emissions following this finding resulted in new fuel economy standards under the Energy Policy and Conservation Act for passenger cars, light-duty trucks, and medium-duty passenger vehicles to reduce greenhouse gas emissions (75 FR 25324; May 27, 2010). The EPA has since issued a series of rules under the Clean Air Act to limit greenhouse gas emissions; for example, setting thresholds to define when permits are required for new and existing industrial facilities (75 FR 82254; December 30, 2010).

In 2013, the Executive Office of the President released The President’s Climate Action Plan (June 2013). In addition to outlining steps to prepare for the impact of climate change and to lead international efforts to combat global climate change, the plan outlines specific objectives

85

Conservation Measures for cutting carbon emissions in the United States. The plan was based on a goal of reducing greenhouse gas emissions in the United States by 17 percent below 2005 levels by the year 2020. It included directives to the Environmental Protection Agency (EPA) to complete carbon pollution standards for new and existing power plants, as well as to develop or further build upon standards for greenhouse gas emissions and fuel economy standards for cars and trucks. The plan also enjoins various Federal agencies to commit to increased development of clean and renewable energy sources, increase energy efficiency, reduce emissions of hydrofluorocarbons and methane, and conserve the nation’s forests, which play a critical role in carbon sequestration.

As a result of this plan, some new regulations have already been put in place, and others are still in the planning stage. In August 2015, the EPA issued the Clean Power Plan, which is intended to reduce carbon pollution from power plants while simultaneously advancing the development and deployment of clean energy technologies. The goal of the Clean Power Plan is to reduce carbon pollution from the power sector to 32 percent below 2005 levels by 2030. The Clean Power Plan sets interim and final carbon dioxide emission performance rates as goals, which are then to be met by States, Tribes, and U.S. territories under a partnership created through Section 111(d) of the Clean Air Act. States are expected to develop and implement plans that achieve interim target carbon dioxide emission rates between 2022 and 2029, and final targets for their State by 2030. To help meet performance goals, the plan also allows for emissions ‘trading’. On October 23, 2015, the EPA issued final carbon pollution standards for new, modified and reconstructed power plants under the Clean Air Act (80 FR 64661) and proposed a Federal Plan and model rule to assist states in implementing the Clean Power Plan (80 FR 64966). However, on February 9, 2016, the United States Supreme Court blocked the ability of the United States government to regulate the emissions of greenhouse gases from coal- fired power plants, so the future effectiveness of this regulation is now in question.

Other regulatory initiatives in the United States to reduce greenhouse gas emissions include proposals to reduce methane gas emissions from landfills (80 FR 52099; August 27, 2015) and to reduce emissions of methane and volatile organic compounds from the oil and natural gas industry (80 FR 56593; September 18, 2015). Greenhouse gas emissions and fuel efficiency standards for medium and heavy-duty engines and vehicles have been proposed for the first time ever (80 FR 4013; July 13, 2015).

The United States is a Party to the United Nations Framework Convention on Climate Change (UNFCCC), and submitted its target to cut net greenhouse gas emissions (Intended Nationally Determined Contribution (INDC)) to the UNFCCC in March 2015, in preparation for the twenty-first session of the Conference of the Parties in December 2015 (see below). The United States target is to reduce emissions by 26 to 28 percent below 2005 levels by 2025, and to make best efforts to reduce by 28 percent.

Climate Initiatives – International

The United Nations Framework Convention on Climate Change (UNFCCC) The United Nations Framework Convention on Climate Change (UNFCCC) is an international environmental treaty or multilateral environmental agreement that was adopted at the “Rio Earth Summit” in Rio de Janeiro, Brazil, in 1992. The primary goal of the UNFCCC is

86

Conservation Measures to stabilize greenhouse gas concentrations at a level that would prevent dangerous anthropogenic (human-induced) interference with the climate system. The UNFCCC entered into force in 1994. There are 195 countries that have ratified the convention and are thus “Parties to the Convention.” The United States signed the treaty in June 1992.

The UNFCCC has held a series of conferences aimed at mobilizing the international community to reduce global greenhouse gas emissions. In an agreement made at the Copenhagen Conference in 2009, known as the Copenhagen Accord, it was agreed that at a minimum, global greenhouse gas emissions must be reduced to a level sufficient to limit the increase in global average temperature to no more than 2ºC (3.6ºF) above preindustrial levels by the end of this century in order to avoid the most dangerous and irreversible consequences of climate change. Although participating world governments agreed to targets to reduce greenhouse gas emissions, the Copenhagen Accord was not formally adopted by the Parties to the Conference, but only noted; therefore no legally binding requirements were established. It is important to note here that even limiting the increase in temperature by the year 2100 to 2ºC (3.6ºF), this increase would still result large range contraction for iiwi and other disease- susceptible Hawaiian forest birds (Fortini 2016, pers. comm.).The UNFCCC (known as a Conference of the Parties, or COP) most recently met in December of 2015, in Paris, France, during which every member country was asked to submit proposals in advance that outline their plans to reduce greenhouse gas emissions (INDCs); UNFCCC 2015, p. 17). As of November 2015, 147 Parties representing 146 countries (75 percent of all Parties to the UNFCCC, which account for approximately 86 percent of the world’s global emissions in 2010) had submitted their INDCs, which cover the time period through 2025 or 2030 (UNFCCC 2015, p. 18). The UNFCCC has aggregated all INDCs submitted to date to assess their potential effectiveness. Although the aggregated INDCs indicate significant reductions in emissions and slow future emissions growth, at present they are not sufficient to reverse the upward trend of global emissions. The UNFCCC has indicated that even if all INDCs were fully implemented and targets met, the goal of limiting the increase in global average temperature to 2˚C (3.6˚F) by the year 2100 would not be achieved (UNFCCC 2015, pp. 11, 45). Rather, full implementation of all INDCs submitted to date for 2025 and 2030 are projected to lead to a warming of around 2.7˚C (2.5 to 2.7˚C; medians of low and high end of pledges) by 2100 (Figure 40). Many governments have not yet committed to efforts sufficient to meet their INDCs; if current policies were to continue, warming is projected to be on the order of 3.6˚C (3.3 to 3.8˚C; medians of low and high end of policy projections) by 2100 (Gűtschow et al. 2015, p. 2).

The INDCs submitted indicate a significant increase in the number of countries taking climate action, and all Parties have raised the ambition of their climate action in their INDCs compared with efforts for the pre-2020 period (UNFCCC 2015, pp. 12-13, 49-50). The global temperature at the end of this century will depend on both emissions up to 2030 and emissions in the post-2030 period. Temperature levels by the end of the century will strongly depend on socioeconomic drivers, technology development, and action undertaken by Parties beyond the time frames stated in their INDCs (e.g. beyond 2025 and 2030) (UNFCCC 2015, pp. 12, 45). The UNFCCC concludes that the extent to which efforts to reduce emissions will be sufficient to limit the global average temperature rise to less than 2°C (3.6°F) above pre-industrial levels strongly depends on the long-term changes in the key economic drivers that will be modified by the implementation of the current INDCs, as well as the determination of Parties to increase

87

Conservation Measures levels of ambition before and after 2030, including through the multilateral process (UNFCCC 2015, pp. 48, 51-52).

The Kyoto Protocol The Kyoto Protocol is an international agreement linked to the UNFCC, which commits its parties by setting internationally binding emission reduction targets to address climate change. The Kyoto Protocol was adopted in Kyoto, Japan, in December 1997 and entered into force in February 2005. The detailed rules for the implementation of the Kyoto Protocol were adopted in Marrakesh, Morocco, in 2001, and are referred to as the “Marrakesh Accords.” The first commitment period of the Kyoto Protocol started in 2008 and ended in 2012, during which 37 industrialized countries and the European community committed to reduce greenhouse gas emissions by an average of 5 percent against 1990 levels. According to some reports, the reduction in greenhouse gases exceeded the target by as much as 19 percent; however much of that reduction has been attributed to shifts in types of manufacturing or localities of manufacturing that are likely to have occurred anyway (Morel and Shishlov 2014, p. 1). In 2012, the Doha amendment to the Kyoto Protocol was adopted, which lays out new commitments from 2013 through the year 2020. During the second commitment period, the parties committed to reduced greenhouse gas emissions by at least 18 percent below 1990 levels in the 8-year period from 2013 to 2020. The signatory parties in the second commitment period are different from the first. The United States is not a party to the Kyoto Protocol.

88

Conservation Measures

Figure 40. If current polices continue, the average global temperature is projected to rise by 3.3 to 3.8° Celsius by 2100 (blue); if all presently submitted INDCs are fully implemented, the average global temperature is projected to rise by 2.5 to 2.7° Celsius by 2100 (red). In both cases, efforts are not sufficient to limit the future increase in average global temperature to 2° Celsius, the threshold identified as necessary to avoid the most significant negative impacts of future climate change. (Image from Gütschow, J., L. Jeffrey, and R. Alexander. 2015. INDCs lower projected warming to 2.7˚C: significant progress but still above 2oC. Climate Action Tracker Update. http://climateactiontracker.org/assets/publications/CAT_global_temperature_update_October_2015.pdf. Used with permission. 89

Summary

5. Summary

We have reviewed the best available scientific information regarding Iiwi populations and the stressors that affect the species and its ohia forest habitat. This information includes, notably, the 2013 USGS analysis of Iiwi abundance, distribution and population trends; numerous studies which provide information on vulnerability of Iiwi to avian malaria; and recent models examining the current relationship between climate and malaria and the likely consequences of climate change for Iiwi and other Hawaiian forest birds.

The Iiwi has declined across large portions of its range, has been extirpated from some islands, and many of the few remaining populations are declining. The Iiwi’s range is contracting upslope in most areas, and population declines and range contraction are concurrent with an increasing prevalence of avian malaria. Clear evidence exists that Iiwi are highly susceptible to avian malaria, and that the prevalence of this disease is moving upslope in Hawaiian forests correlated with temperature increases associated with global climate change. The evidence suggests this disease and its trend of increasing prevalence at increasing elevation are the chief drivers of observed Iiwi population declines and range contraction. Avian pox is another mosquito-borne disease that often co-occurs with avian malaria, and can have significant negative effects acting both independently and synergistically when coinfection occurs. Although habitat management to reduce breeding habitat for mosquitoes may have slowed the decline of Iiwi and other forest birds in a few locations, no landscape-scale plans or strategies exist for eradicating mosquitoes or otherwise reducing the risk posed by avian malaria and avian pox to Iiwi and other susceptible Hawaiian bird species.

The documented trend of temperature increase, which is greatest at high elevation, is projected to continue at least through the 21st century. Multiple independent modeling efforts project that the prevalence of avian malaria will continue to increase upslope with increasing temperature, eventually eliminating most or all remaining disease-free habitat in the islands. These models, which incorporate data on the distribution of forest birds and on disease transmission, project moderate to high avian malaria transmission at highest elevations of Iiwi’s current range by the end of this century, and perhaps sooner. As a consequence, significant declines in Iiwi populations are projected, on the order of 70 to 90 percent by 2100, depending on the future climate scenario.

The impacts of other stressors to Iiwi, such as loss or degradation of native forest by nonnative species (disturbance or destruction by feral ungulates; invasion by nonnative plants; impacts from nonnative pathogens such as ROD), predation by rats and other nonnative predators, and small-population stressors such as chance environmental occurrences, demographic stochasticity, and loss of genetic diversity, have not been well documented or quantified. However, any stressors that result in further degradation or fragmentation of the forests on which the Iiwi relies for foraging and nesting, or result in increased mortality or reduced reproductive success, are likely to exacerbate the impacts of disease on the species. The effects of climate change may exacerbate these other stressors to Iiwi as well.

90

References

6. References

Adee, K. and C.E. Conrad. Downloaded 2014. Metrosideros polymorpha Gaud.: ohia lehua. Available online at: Available on online at: http:www.na.fs.fed.us/pbus/ silvics_manual/volume_2/ metrosideros/polymorpha.html.

Agorastos, N.R. 2016. Hawaii Department of Land and Natural Resources (DLNR), Natural Area Reserves System (NARS), Hawaii Island Administrator. ([email protected]). Personal communication regarding informal surveys and observations of Iiwi in the Kohala Mountains, Hawaii Island.

Ahumada, J.A., D.A. LaPointe, and M.D. Samuel. 2004. Modeling the Population Dynamics of Culex quinquefasciatus (Diptera: Culicidae), along an Elevational Gradient in Hawaii. Journal of Medical Entomology 41: 1,157–1,170.

Ahumada, J.A., M.D. Samuel, D.C. Duffy, A.P. Dobson, and P.H.F. Hobbelen. 2009. Modeling the epidemiology of avian malaria and pox. Pages 331-358 in T.K. Pratt, C.T. in T.K. Pratt, C. T. Atkinson, P.C. Banko, J. D. Jacobi, and B.L. Woodworth (Eds.). Conservation biology of Hawaiian forest birds: Implications for island avifauna. Yale University Press, New Haven and London.

Aila, W. 2012, in litt. In Osher, W. 2012, in litt. Axis deer eradication efforts begin on Big Island. http://mauinow.com/2012/04/16/axis-deer-eradication-efforts-begin- on-big-island/, accessed May 27, 2012.

Anderson, S.B. 2002. The conservation and management implications of axis deer (Axis axis) diet and home range (poster abstract). The 2002 Hawaii Conservation Conference. July 18, 2002.

Anderson, R.C. 2012. A baseline analysis of the distribution, host-range, and severity of the rust Puccina psidii in the Hawaiian Islands, 2005 – 2010. Technical Report HCSU-031. USGS, Honolulu, HI, 35 pp.

Anderson, S.J. and C.P. Stone. 1993. Snaring to control feral pigs sus scrofa in a remote Hawaiian rain forest. Biological Conservation 63: 195-201.

American Ornithologists’ Union (AOU). 1998. Check-list of North American Birds, 7th edition. Allen Press, Lawrence, KS. Pp. 523-684.

Aplet, G.H., S.L. Anderson, and C.P. Stone. 1991. Association between feral pig disturbances and the composition of some alien plant assemblages in Hawaii Volcanoes National Park. Vegetation 95: 55-62.

Aruch, S., C.T. Atkinson, A.F. Savage, and D.A. LaPointe. 2007. Prevalence and distribution of pox-like lesions, avian malaria, and mosquito vectors in Kipahulu

91

References

Valley, Haleakala National Park, Hawaii, USA. Journal of Wildlife Diseases 43: 567-575.

Atkinson, I.A.E. 1977. A reassessment of factors, particularly Rattus rattus L., that influenced the decline of endemic forest birds in the Hawaiian Islands. Pacific Science 31: 109-133.

Atkinson, C.T., K.L. Woods, R.J. Dusek, L S. Sileo, and W.M. Iko. 1995. Wildlife disease and conservation in Hawaii: Pathogenicity of avian malaria (Plasmodium relictum) in experimentally infected Iiwi (Vestiaria coccinea). Parasitology 111: S59-S69.

Atkinson, C.T., J.K. Lease, R.J. Dusek, and M.D. Samuel. 2005. Prevalence of pox-like lesions and malaria in forest bird communities on leeward Mona Loa volcano, Hawaii. Condor 107: 537–546.

Atkinson, C.T., and D. A. LaPointe. 2009a. Ecology and pathogenicity of avian malaria and pox. Pages 234-252 in T.K. Pratt, C.T. Atkinson, P.C. Banko, J.D. Jacobi, and B.L. Woodworth (Eds.). Conservation biology of Hawaiian forest birds: Implications for island avifauna. Yale University Press, New Haven and London.

Atkinson, C.T., and D.A. LaPointe. 2009b. Introduced avian diseases, climate change, and the future of Hawaiian honeycreepers. Journal of Avian Medicine and Surgery 23: 53-63.

Atkinson, C.T., K.S. Saili, R.B. Utzurrum and S.I. Jarvi. 2014a. Experimental evidence for evolved tolerance to avian malaria in a wild population of low elevation Hawaii amakihi (Hemignathus virens). EcoHealth 10: 366-375.

Atkinson, C.T., R.B. Utzurrum, D.A. LaPointe, R.J. Camp, L.H. Crampton, J.T. Foster, and T.W. Giambelluca. 2014b. Changing climate and the altitudinal range of avian malaria in the Hawaiian Islands: an ongoing conservation crisis on the island of Kauai. Global Climate Change 20: Pp. 2,426-2,436.

Baldwin, P.H. and G.O. Fagerlund. 1943. The Effect of Cattle Grazing on Koa Reproduction in Hawaii National Park. Ecology, Vol. 24, No. 1. Pp. 118-122.

Banko, W.E. 1981. History of endemic Hawaiian birds, Part 1—Population histories— Species account, forest birds—Vestiaria coccinea, Drepanis funereal, Drepanis pacifica. Honolulu, Hawaii, University of Hawaii at Manoa, Cooperative National Park Resources Study Unit, Avian History Report 11 B.

Banko, W.E. and P.C. Banko. 2009. Historic decline and extinction. Pages 25-58 in T.K. Pratt, C.T. Atkinson, P.C. Banko, J.D. Jacobi, and B.L. Woodworth (Eds.). Conservation biology of Hawaiian forest birds: Implications for island avifauna. Yale University Press, New Haven and London.

92

References

Banko, P., C. Farmer, S. Hess, K. Brinck, G. Beauprez, J. Castner, J. Crummer, R. Danner, B. Hsu, K. Kozar, J. Leialoha, A. Lindo, B. Muffler, C. Murray, D. Nelson, D. Pollock, M. Schwarzfeld, K. Rapozo, and E. Severson. 2004. Palila restoration project 2004 report: summary of results, 1996-2004. U.S. Geological Survey – Biological Resources Division, Pacific Island Ecosystems Research Center, Kilauea Field Station, Hawaii National Park, HI. Unpublished technical report to Federal Highway Administration, Honolulu, HI, 30 September 2004. 443 pp.

Banko, P.C., R.J. Camp, C. Farmer, K.W. Brinck, and D.L. Leonard. 2013. Response of palila and other subalpine Hawaiian forest birds species to prolonged drought and habitat degradation by feral ungulates. Biological Conservation 157: 70-77.

Benitez, D.M., T.R. Belfield, R. Loh, L. Pratt, and A.D. Christie. 2008. Inventory of vascular plants of the Kahuku addition, Hawaii Volcanoes National Park. University of Hawaii, Pacific Cooperative Studies Unit, Honolulu. 118 pp.

Benning, T.L., D.A. LaPointe, C.T. Atkinson, and P.M. Vitousek. 2002. Interactions of climate change with biological invasions and land use in the Hawaiian Islands: modeling the fate of endemic birds using a geographic information system. Proceedings of the National Academy of Science 99: 14,246-14,249.

Big Island Invasive Species Committee. 2014. Albizia (Falcataria moluccana). Available online at: http://www.bigislandisc.org

Bio, K. 2011b, pers. comm. Hawaii Island rare plant expert, Kealii Bio, provided comments at the Hawaii Island Plant Extinction Prevention Program (PEPP) meeting with Pacific Islands Fish and Wildlife Service (FWS) regarding rare and endangered Hawaiian plants and Drosophila (Attendees: PEPP-Joan Yoshioko and Kealii Bio; FWS-Carrie Harrington and James Kwon; and Drosophila expert, Karl Magnacca), Nov. 1, 2011.

Borneman, T., R. Camp, S. Kendall, and E. Paxton. 2015. Settlement patterns of forest birds in koa reforested areas of Hakalau Forest NWR. 23rd Annual Hawaii Conservation Conference, August 3rd-6th, University of Hawaii, Hilo, Hawaii, 1 p Abstract (in press).

Brown, A.C. and M. Matsuura. 1941. Black rot of sweet potato. Agricultural Extension Circular #134, University of Hawaii (from Friday et al. 2015).

Burgan, R.E. and R.E Nelson. 1972. Decline of ohia lehua forests in Hawaii. Pacific Sollthwest Forest and Range Experiment Station. United States Department of Agriculture, Forest Service. General Technical Report PSW-3.

93

References

Clarke, G. and L.W. Cuddihy. 1980. A botanical reconnaissance of the Na Pali Coast trail: Kee beach to Kalalau Valley (April 9-11, 1980). Hawaii Department of Land and Natural Resources, Division of Forestry and Wildlife, Hilo. Pp. C-14–C-20.

Cao, G., T.W. Giambelluca. D.E. Stevens, and T.A. Schroeder. 2007. Inversion variability in the Hawaiian trade wind regime. Journal of Climate 20: 1,145- 1,160.

Camp, R.J., T.K. Pratt, P.M. Gorresen, J.J. Jeffrey, and B.L. Woodworth. 2010. Population trends of forest birds at Hakalau Forest National Wildlife Refuge, Hawaii. The Condor 112: 196-212.

Camp, R.J., T.K. Pratt, P.M. Gorresen, B.L. Woodworth, and J.J. Jeffrey. 2014. Hawaiian forest bird trends: using log-linear models to assess long-term trends is supported by model diagnostics and assumptions. Condor 116: 97-101.

Camp, R.J., K.W. Brinck, P.M. Gorresen, and E.H. Paxton. 2015. Evaluating abundance and trends in a Hawaiian avian community using state-space analysis. Bird Conservation International, Available on CJO 2015 doi:10.1017/S0959270915000088

Campbell, J.D. and I.A.E. Atkinson. 1999. Effects of kiore (Rattus exulans) on recruitment of indigenous coastal trees on northern offshore islands of New Zealand. Journal of the Royal Society of New Zealand 29: 265–290.

Cao, G., T.W. Giambelluca. D.E. Stevens, and T.A. Schroeder. 2007. Inversion variability in the Hawaiian trade wind regime. Journal of Climate 20: 1,145- 1,160.

Chan, K.T.F., and J.C.L. Chan. 2012. Size and strength of tropical cyclones as inferred from QuikSCAT data. Monthly Weather Review 140: 811-824.

Chesser, R. T., R.C. Banks, K. J. Burns, C. Cicero, J.L. Dunn, A.W. Kratter, I.J. Lovette, A.G. Navarro-Siguenza, P.C. Rasmussen, J. V. Remsen, Jr., J.D. Rising, D.F. Stotz, and K. Winker. 2015. Fifty-sixth Supplement to the American Ornithologists’ Union, Check-list of North American Birds; Volume 132, pp. 748–764. DOI: 10.1642/AUK-15-73.1

Chu, P.S., and Y.R. Chen. 2005. Interannual and interdecadal rainfall variations in the Hawaiian islands. Journal of Climate 18: 4,796-4,813.

Chu, P.S. and J. Wang. 1997. Tropical cyclone occurrences in the vicinity of Hawaii: are the differences between El Niño and Non-El Niño years significant? Journal of Climate 10: 2,683-2,689.

94

References

Chu, P.S., Chen, Y.R., and T.A. Shroeder. 2010. Changes in precipitation extremes in the Hawaiian Islands in a warming climate. Journal of Climate 23: 4,881-4,900.

Clark, J.D. and P.S. Chu. 2002. Interannual variation of tropical cyclone activity over the Central North Pacific. Journal of the Meteorological Society of Japan 80: 403- 418.

Cuddihy, L. 1984. Effects of cattle grazing on the mountain parkland ecosystem, Mauna Loa, Hawaii. Technical Report 51, Pacific Cooperative Studies Unit, University of Hawaii, Honolulu. 138 pp.

Cuddihy, L.W. and C.P. Stone. 1990. Alteration of native Hawaiian vegetation: effects of humans, their activities and introductions. University of Hawaii, Honolulu. 138 pp.

Cummins, G.C., S.J. Kendall, and E.H. Paxton. 2014. Productivity of forest birds at Hakalau Forest NWR. Hawaii cooperative Studies Unit, University of Hawaii at Hilo, Technical Report HCSU-056. 21 pages.

D'Antonio, C.M. and P.M. Vitousek. 1992. Biological Invasions by Exotic Grasses, the Grass/fire Cycle, and Global Change. Annual Review of Ecology and Systematics 23. Pp. 63–87.

D’Antonio, C.M., R.F. Hughes, and J.T. Tunison. 2011. Long-term impacts of invasive grasses and subsequent fire in seasonally dry Hawaiian woodlands. Ecological Applications 21: 1,617-1,628.

Dawson, J.W., and Stemmermann, R.L. 1990. pg. 967 in: Manual of Flowering Plants of Hawaii, 2 Volumes. Wagner, W.L., D.R. Herbst, and S.H. Sohmer, (Eds.). University of Hawaii and Bishop Museum Press, Honolulu, HI.

Degener, O., I. Degener, K. Sunada, and A. Sunada. 1976. Argyroxiphium kauense, the Kau silversword. Phytologia 33: 173–177.

Diaz, H.F., T.W. Giambelluca, and J.K. Eischeid. 2011. Changes in the vertical profiles of mean temperature and humidity in the Hawaiian Islands. Global Planet Change 77: 21-25.

Eggert, L.S., L.A. Terwilliger, B.L. Woodworth, P.J. Hart, D. Palmer, and R.C. Fleisher. 2008. Genetic structure along an elevational gradient in Hawaiian honeycreepers reveals contrasting evolutionary responses to avian malaria. BMC Evolutionary Biology 8:315. Available online at: http://www.biomedcentral.com/1471- 2148/8/315.

95

References

Fancy, S.G., and C.J. Ralph. 1998. Iiwi (Vestiaria coccinea). In The Birds of North America, No. 327. A. Poole and F. Gill, (Eds.). The Birds of North America, Inc., Philadelphia, PA.

Fischer, J.R. 2007. Cattle in Hawaii: biological and cultural exchange. Pacific Historical Review 76: Pp. 347–372.

Fischer, J. In press. Novel technologies for addressing avian malaria in Hawai’i. Elepaio.

Foote, D., C. Hanna, C. King, and E.B. Spurr. 2011. Efficacy of fipronil for suppression of alien yellowjacket wasps in Hawaii Volcanoes National Park. Hawaii Cooperative Studies Unit Technical Report No. 28. University of Hawaii at Hilo. 19 pp.

Fortini, L.B., A.E. Vorsino, F.A. Amidon, E.H. Paxton, and J.D. Jacobi. 2015. Large- Scale Range Collapse of Hawaiian Forest Birds under Climate Change and the Need for 21st Century Conservation Options. PloS one, 10(10): e0140389.

Foster, J.T. E.J. Tweed, R.J. Camp, B.L. Woodworth, C.D. Adler, and T. Telfer. 2004. Long-Term Population Changes of Native and Introduced Birds in the Alakai Swamp, Kauai. Conservation Biology, 18: 716–725. doi: 10.1111/j.1523- 1739.2004.00030.x

Foster, J.T., B.L. Woodworth, L.E. Eggert, P.J. Hart, D. Palmer, D.C. Duffy, and R.C. Fleischer. 2007. Genetic structure and evolved malaria resistance in Hawaiian honeycreepers. Molecular Ecology 16: 4,738-4,746.

Frankham, R., C.J. Bradshaw, and B.W. Brook. 2014. Genetics in conservation management: revised recommendations for the 50/500 rules, Red List criteria and population viability analyses. Biological Conservation 170: 56-63.

Freed, L.A., R.L. Cann, M.L. Goff, W.A. Kuntz, and G.R. Bodner. 2005. Increase in avian malaria at upper elevation in Hawaii. The Condor 107: 753-764.

Freed, L.A., M.C. Medeiros, G.R. Bodner. 2008. Explosive increase in ectoparasites in Hawaiian forest birds. Journal of Parasitology 94: 1,009-1,021.

Freed, L.A. and R.L. Cann. 2012. Changes in Timing, Duration, and Symmetry of Molt of Hawaiian Forest Birds. PLoS ONE 7(1): e29834. doi: 10.1371/journal.pone.0029834

Freed, L.A. and R.L. Cann. 2013. More misleading trend analysis of Hawaiian forest birds. Condor 115: 442-447.

96

References

Freed, L.A. and R.L. Cann. 2014. Diffuse competition can be reversed: a case history with birds in Hawaii. Ecosphere 5(11): 147. Available online at: http://dx.doi.org/10.1890/ES14-00289.1

Freed, L.A. and R.L. Cann. 2015. Problems with Hawaiian forest birds. Jacobs Journal of Environmental Sciences. 1(2): 007.

Friday, J.B. and D.A. Herbert. 2006. Species Profiles for Pacific Island Agroforestry: Metrosideros polymorpha (ohia lehua). Permanent Agriculture Resources, Holualoa, Hawaii. Retrieved from http://www.agroforestry.net/images/pdfs/Metrosideros-ohia.pdf, June 8, 2015.

Friday, J.B., L. Keith, and F. Hughes. 2015. Rapid Ohia Death (Ceratocystis Wilt of Ohia). Plant Disease. June 2015. PD-107

Fujioka, F.M. and D.M. Fujii. 1980. Physical characteristics of selected fme fuels in Hawai'i -- some refmements on surface area-to-volume calculations. USDA For. Serv. Res. Note PSW-348.

Gagne, W.C. and L.W. Cuddihy. 1999. Vegetation. In Manual of the Flowering Plants of Hawaii, W.L. Wagner, D.R. Herbst, and S.H. Sohmer, Eds., University of Hawaii Press, Bishop Museum Press, Honolulu. Pp. 45–114.

Gantz, V.M., N. Jasinskiene, O. Tatarenkova, A. Fazekas, V.M. Macias, E. Bier, and A.A. James. 2015. Highly efficient Cas9-mediated gene drive for population modification of the malaria vector mosquito Anopheles stephensi. Proceedings of the National Academy of Sciences, 112: pp.E6736-E6743.

Giambelluca, T.W., H.F. Diaz, and M.S.A. Luke. 2008. Secular temperature changes in Hawaii. Geophysical Research Letters 35, Issue 12. Available online at: http://onlinelibrary.wiley.com/doi/10.1029/2008GL034377/abstract;jsessionid=66 8BE2D7E1B5A55B74FA313794C20EAE.f04t02

Giambelluca, T.W., Q. Chen, A.G. Frazier, J.P. Price, Y.-L. Chen, P.-S. Chu, J.K. Eischeid, and D.M. Delparte, 2013. Online Rainfall Atlas of Hawaii. Bull. Amer. Meteor. Soc. 94, 313-316, doi: 10.1175/BAMS-D-11-00228.1.

Guillaumet, A., B.L. Woodworth, R.J. Camp, and E.H. Paxton. 2016. Comparative demographics of a Hawaiian forest bird community. Journal of Avian Biology, 47: 185–196. doi: 10.1111/jav.00756

Gon III, S. and T. Tierney. 1996. In Wagner, W.L., S.G. Weller, and A.K. Sakai. 2005d. Monograph of Schiedea (Caryophyllaceae-Alsinoideae). Systematic Botany Monographs 72: 1–169.

97

References

Gütschow, J., L. Jeffrey, and R. Alexander. 2015. INDCs lower projected warming to 2.7˚C: significant progress but still above 2oC. Climate Action Tracker Update. http://climateactiontracker.org/assets/publications/CAT_global_temperature_upda te_October_2015.pdf Downloaded 24 November 2015.

Hanna, C., D. Foote, and C. Kremen. 2012. Short- and long-term control of Vespula pensylvanica in Hawaii by fipronil baiting. Pest. Manag. Sci. 68: 1,026–1,033

Hanna, C., D. Foote, C. Kremen. 2014. Competitive impacts of an invasive nectar thief on plant-pollinator mutualisms. Ecology 95(6): 1,622-1,632.

Harrington, R.A., J.H. Fownes, P.G. Scowcroft, and C.S. Vann. 1997. Impact of Hurricane Inkiki on native Hawaiian Acacia koa forests: damage and two-year recovery. Journal of Tropical Ecology 13: 539-558.

Hart, P. J., B. L. Woodworth, R. J. Camp, K. Turner, K. McClure, K. Goodall, C. Henneman, C. Spiegel, J. LeBrun, E. Tweed, and M. Samuel. 2011. Temporal variation in bird and resource abundance across an elevational gradient in Hawaii. Auk 128:113-126.

Hawaii Biodiversity and Mapping Program (HBMP). 2007. Hawaiian species database. GIS shapefiles and Access data, unpublished.

Hawaii Biodiversity and Mapping Program (HBMP). 2010a. Hawaiian species database. GIS shapefiles and database information for Pritchardia lanigera, unpublished.

Hawaii Biodiversity and Mapping Program (HBMP). 2010b. Hawaiian species database. GIS shapefiles and database information for Cyanea tritomantha, unpublished.

Hawaii Biodiversity and Mapping Program (HBMP). 2010c. Hawaiian species database. GIS shapefiles and database information for Schiedea diffusa, unpublished.

Hawaii Biodiversity and Mapping Program (HBMP). 2010d. Hawaiian species database. GIS shapefiles and database information for Stenogyne cranwelliae, unpublished.

Hawaii Department of Agriculture. (M. Killgore and R.A. Heu). 2007. Ohia Rust Puccinia psidii New Pest Advisory Factsheet No. 05-04, 2 pp.

Hawaii Division of Forestry and Wildlife (HDOFAW). 2002. Draft management plan for the ahupuaa of Puuwaawaa and the makai lands of Puuanahulu. 129 pp.

Hawaii Department of Land and Natural Resources (DLNR). 2010. Draft environmental assessment: biocontrol of strawberry guava by its natural control agent for preservation of native forests in the Hawaiian Islands. 131 pp.

98

References

Hawaii Department of Land and Natural Resources (DLNR). 2011, in litt., First axis deer hunted down on Hawaii, DLNR provides photo proof. In Big Island Video News, http://www.bigislandvideonews.com/2012/04/13/first-axis-deer-hunted-down-on- hawaii-dl..., accessed May 28, 2015.

Hawaii Invasive Species Council. 2015. Miconia fact sheet. http://dlnr.hawaii.gov/hisc/info/species/miconia/

Hess, S.C. 2008. Wild sheep and deer in Hawaii–a threat to fragile ecosystems. USGS Fact Sheet 2008-3102, USGS Pacific Islands Ecosystems Research Center, Honolulu. 4 pp.

Hess, S. 2010, pers. comm. Ungulate distribution on Maui Nui. Telephone call to C. Harrington. Dec. 1, 2010.

Hess, S.C. 2014. A Tour de Force by Hawaii’s Invasive Mammals: Establishment, Takeover, and Ecosystem Restoration through Eradication Mammal Study 41(2): 47-60.

Hess, S.C. and P.C. Banko. 2006. Feral cats: Too long a threat to Hawaiian wildlife. USGS Fact Sheet FS-2006-3006. January 2006. http://www.usgs.gov/ecosystems/pierc/assets/cats.pdf

Hess, S.C., P.C. Banko, M.H. Reynolds, G.J. Brenner, L.P. Laniawe, and J.D. Jacobi. 2001. Drepanidine movements in relation to food availability in subalpine woodland on Mauna Kea, Hawaii. In J.M. Scott, S. Conant, & C. van Riper, III (eds.), "Evolution, ecology, conservation and management of Hawaiian birds: a vanishing avifauna." Studies in Avian Biology 22: 154-163.

Hobdy, R. 1993. Lanai–a case study: the loss of biodiversity on a small Hawaiian island. Pacific Science 47: Pp. 201–210.

Hodges, C.S., K.T. Adee, J.D. Stein, H.B. Wood, and R.D. Doty. 1986. Decline of ohia (Metrosideros polymorpha) in Hawaii: a review. Gen. Tech. Rep. PSW-86. Berkeley, CA: Pacific Southwest Forest and Range Experiment Station, Forest Service, U.S. Department of Agriculture; 1986. 22 p.

Ikagawa, M. 2011, in litt. Excerpt from draft MS thesis: Mouflon range and rare plants on Hawaii Island.

Intergovernmental Panel on Climate Change (IPCC). 2000. Nakicenovic, N. and R. Swart (Eds.) Cambridge University Press, UK. pp 570

Intergovernmental Panel on Climate Change (IPCC). 2007. Climate Change 2007: Impacts, Adaptation and Vulnerability. Contribution of Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change,

99

References

Parry, M.L., O.F. Canziani, J.P. Palutikof, P.J. van der Linden, and C.E. Hanson, (Eds.), Cambridge University Press, Cambridge, UK, 976 pp.

Intergovernmental Panel on Climate Change (IPCC). 2013. Summary for Policymakers. In: Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change, Stocker, T.F., D. Qin, G.K. Plattner, M. Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (Eds.). Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA.

Hodel, D.R. and P.R. Weissich. 2012. Trees in the landscape, Part 5: Metrosideros polymorpha. Western Arborist, Fall 2012: 38-51.

Jacobi, J., G. Gerrish, D. Mueller-Dombois. 1983. Ohia dieback in Hawaii: vegetation changes in permanent plots. Pac. Sci. 37: 327-337.

Jarvi, S.I., C.T. Atkinson, and R.C. Fleischer. 2001. Immunogenetics and resistance to avian malaria in Hawaiian honeycreepers (Drepanidinae). Studies in Avian Biolgoy 22: 254-263.

Jarvi, S.I., C.L. Tarr, C.E. McIntosh, C.T. Atknson, and R.C. Fleischer. 2004. Natural selection of the major histocompatibility complex (Mhc) in Hawaiian honeycreepers (Drepanidinae). Molecular Ecology 13: 2,157-2,168.

Jarvi, S.I.,Triglia, D., Giannoulis, A., Farias, M.E.M., Bianchi, K., and Atkinson, C.T. 2008. Diversity, origins and virulence of Avipoxviruses in Hawaiian Forest Birds. Conservation Genetics, 9, 339-348

Judge, S. W., J. Gaudioso, B. Hsu, R. J. Camp, and P. J. Hart. 2011. Pacific Island landbird monitoring annual report, Hawaii Volcanoes National Park, tract group 1 and 2, 2010. Natural Resource Technical Report NPS/PACN/NRTR—2011/486. National Park Service, Fort Collins, Colorado, USA.

Juvik, J.O. and S.P. Juvik. 1984. Mauna Kea and the myth of multiple use, endangered species, and mountain management in Hawaii. Mountain Res. and Devel. 4(3): 191-202.

Juvik, S.P. and J.O. Juvik. 1998. Atlas of Hawaii, 3rd edition. Department of Geography, University of Hawaii Press. Pp. 7, 11, 13‒16.

Keith, L.M., R.F. Hughes, L.S. Sugiyama, W.P. Heller, B.C. Bushe, and J.B. Friday. 2015. First Report of Ceratocystis wilt on Ohia. Plant Disease. http://dx.doi.org/10.1094/PDIS-12-14-1293-PDN

100

References

Kessler, C. 2010, pers. comm. Interview with Fish and Wildlife Service ungulate expert Curt Kessler conducted by Fish and Wildlife Service biologist Carrie Harrington regarding ungulate distribution in the Hawaiian Islands. Dec. 3, 2010.

Kessler, C. 2011, pers. comm. Interview with Fish and Wildlife Service ungulate expert Curt Kessler conducted by Fish and Wildlife Service biologist Carrie Harrington regarding ungulate distribution in the Hawaiian Islands. Feb. 7, 2011.

Kilpatrick, A.M. 2006. Facilitating the evolution of resistance to avian malaria in Hawaiian birds. Biological Conservation 128: 475-485.

Kirch, P.V. 1982. The impact of the prehistoric Polynesians on the Hawaiian ecosystem. Pacific Science 36: 1–14 pp.

Kuntz, W.A. 2008. The importance of individual behavior to life history and conservation: breeding and movement biology of the Iiwi (Vestiaria coccinea). Ph.D. Dissertation, University of Hawaii at Manoa.

LaPointe, D.A. 2006. Feral pigs, introduced mosquitoes, and the decline of Hawaiian birds. U.S. Geological Survey, Fact Sheet 2006–3029.

LaPointe, D.A. 2008. Dispersal of Culex quinquefasciatus (Diptera: Culicidae) in a Hawaiian rain forest. Journal of Medical Entomology 45: 600-609.

LaPointe, D.A., C.T. Atkinson, and S.I. Jarvi. 2009. Managing Disease. Pages 405-424 in T.K. Pratt, C.T. Atkinson, P.C. Banko, J.D. Jacobi, and B.L. Woodworth (Eds.). Conservation biology of Hawaiian forest birds: Implications for island avifauna. Yale University Press, New Haven and London.

LaPointe, D.A., M.L. Goff, and C.T. Atkinson. 2010. Thermal constraints to the sporagonic development and altitudinal distribution of avian malaria Plasmodium relictum in Hawaii. Journal of Parasitology 96: 318-324.

LaPointe, D.A., C.T. Atkinson, and M.D. Samuel. 2012. Ecology and conservation biology of avian malaria. Annals of the New York Academy of Sciences 1,249: 211–226.

LaPointe, D.A., J. Gaudioso, C.T. Atkinson, and A. Egan. 2015. Avian malaria in Laupahoehoe Forest: possible population sink for Hakalau Iiwi? Presentation, 23rd Annual Hawaii Conservation Conference. August 3-6th, University of Hawaii at Hilo, Hawaii.

LaPointe, D.A., J.M. Gaudioso-Levita, C.T. Atkinson, A. Egan, and K. Hayes. 2016. Changes in the prevalence of avian disease and mosquito vectors at Hakalau Forest National Wildlife Refuge: a 14-year perspective and assessment of future risk. USGS Technical Report HCSU-073. 57 pp.

101

References

Liao, W., O.E. Timm, C. Zhang, C.T. Atkinson, D.A. LaPointe, and M.D. Samuel. 2015. Will a warmer and wetter future cause extinction of native Hawaiian forest birds? Global Change Biology 21: 4,342–4,352, doi: 10.1111/gcb.13005

Lindsey, G.D., S.M. Mosher, S.G. Fancy, and T.D. Smucker. 1999. Population structure and movements of introduced rats in an Hawaiian rainforest. Pacific Conservation Biology 5: 94-102.

Lindsey, G.D., S.C. Hess, E.W. Campbell III, and R.T. Sugihara. 2009. Small mammals as predators and competitors. Pages 274-292 in T.K. Pratt, C.T. Atkinson, P.C. Banko, J.D. Jacobi, and B.L. Woodworth (Eds.). Conservation biology of Hawaiian forest birds: Implications for island avifauna. Yale University Press, New Haven and London.

Loope L.L. 1999. Hawaii and the Pacific Islands. In US Geological Survey. Status and Trends of the Nation’s Biological Resources. Washington (DC): US Government Printing Office. (6 April 2003; http://biology.usgs.gov/s+t/SNT/noframe/pi179.htm)

Loope, L.L. and T.W. Giambelluca. 1998. Vulnerability of island tropical montane cloud forests to climate change, with special reference to East Maui, Hawaii. Climatic Change 39: 503-517.

Loope, L.L., O. Hamann, and C.P. Stone. 1988. Comparative conservation biology of oceanic archipelagoes: Hawaii and the Galapagos. Bioscience 38: Pp. 272–282.

Loope, L.L., A.C. Medeiros, and B.H. Gagne. 1991. Recovery of vegetation of a montane bog following protection from feral pig rooting. Technical report 76-78, Cooperative National Park Resources Studies Unit, University of Hawaii at Manoa, Honolulu. 23 pp.

Lorence, D.H. and S. Perlman. 2007. A new species of Cyrtandra (Gesneriaceae) from Hawaii, Hawaiian Islands. Novon 17(3): Pp. 357–361.

Maui Forest Bird Recovery Project (MFBRP). 2015. Ecosystem threats: feral (wild) cats. http://www.mauiforestbirds.org/articles/23 Downloaded 25 November 2015.

McCarthy, J. 2014. Tropic Lightning: Myth or Menace? Hawaii Journal of Medicine and Public Health. 73 (11 Suppl 2): Pp. 44–47.

Medeiros, A. 2010, pers. comm. Telephone interview with USGS Pacific Island Ecosystems biologist, Art Medeiros, conducted by Fish and Wildlife Service biologist Carrie Harrington regarding Axis deer distribution on Maui, Dec. 3, 2010.

102

References

Medeiros, A.C., Jr., L.L. Loope, and R.A. Holt. 1986. Status of native flowering plant species on the south slope of Haleakala, east Maui, Hawaii. Cooperative National Park Resources Studies Unit, University of Hawaii, Honolulu. 230 pp.

Mehrhoff, L. 1993. Recovery plan for the Hawaiian gardenia. U.S. Fish and Wildlife Service. 69 pp.

Mortenson, L.A., R.F. Hughes, J.B. Friday, L.M. Keith, J.M. Barbosa, N.J. Friday, Z. Liu, and T.G. Sowards. 2016. Assessing spatial distribution, stand impacts and rate of Ceratocystis fimbriata induced Ohia (Metrosideros polymorpha) mortality in a tropical wet forest, Hawaii Island, USA. Forest Ecology and Management 377:83-92.

Mountainspring, S. and J.M. Scott. 1985. Interspecific competition among Hawaiian forest birds. Ecol. Monogr. 55: 219-239.

Mountainspring, S. 1986. An ecological model of the effects of exotic factors on limiting Hawaiian honeycreeper populations. Ohio Journal of Science 86: 95-100.

Mueller-Dombois, D. 1981. Fire in tropical ecosystems. Pages 137-176 In Fire Regimes and Ecosystem Properties. H.A. Mooney, T.M. Bonnicksen, N.L. Christensen, J.E. Lotan, and W.A. Reiners, Eds. USDA Forest Service Gen. Tech. Report WO- 26. 594 p.

Mueller-Dombois, D. 1986. Perspectives for an etiology of stand-level dieback. Ann. Rev. Ecol. Syst. 17: 221-243.

Mueller-Dombois, D., J. Jacobi, H.J. Boehmer, and J.P. Price. 2013. Ohia Lehua Rainforest. The story of a dynamic ecosystem with relevance to forests worldwide. –278 p., Amazon Press.

Munro, G. 1944. Birds of Hawaii. Bridgeway Press, Rutland, Vermont, USA. 192 pp.

Murakami, H., B. Wang, T. Li, and A. Kitoh. 2013. Projected increase in tropical cyclones near Hawaii. Nature Climate Change. 3: 749-754.

Murakami, H., G.A. Vecchi, T.L. Delworth, K. Paffendorf , R. Gudgel, L. Jia, and F. Zeng. 2015. Investigating the influence of anthropogenic forcing and natural variability on the 2014 Hawaiian hurricane season. In “Explaining Extremes of 2014 from a Climate Perspective”]. Bull. Amer. Meteor. Soc., 96 (12), S115– S119.

Myers, R. 2000. Fire in Tropical and Subtropical Ecosystems. Pages 161-171 in J.K. Brown and J.K. Smith (Eds.). 2000. Wildland fire in ecosystems: effects of fire on flora. Gen. Tech. Rep. RMRS-GTR-42-vol. 2. Ogden, UT: U.S. Department of Agriculture, Forest Service, Rocky Mountain Research Station. 257 p.

103

References

Nishibayashi, E., in Kubota, G. 2001, in litt. Deer population boom threatens Maui forests, farms. Honolulu Star-Bulletin, 28 AUG 2001. http://archives.starbulletin.com/2001/08/28/news/story8.html, Aug. 28, 2001.

O’Connor, P.J. 1999. Poaceae, grass family (except Panicum). in Manual of Flowering Plants of Hawaii, W.L. Wagner, D.R. Herbst, and S.H. Sohmer, Eds. University of Hawaii Press, Bishop Museum Press, Honolulu. Pp. 1,481–1,603.

Paxton, E.H., R.J. Camp, P.M. Gorresen, L.H. Crampton, D.L. Leonard, Jr., and E.A. VanderWerf. 2016. Collapsing avian community on a Hawaiian island. Science Advances 2:e1600029. DOI: 10.1126/sciadv.1600029.

Paxton E.H., P.M. Gorresen, and R.J. Camp. 2013. Abundance, distribution and population trends of the iconic Hawaiian honeycreeper, Iiwi (Vestiaria coccinea) throughout the Hawaiian islands. U.S. Geological Survey, Open-File Report 2013-11150. 59 pages.

Plant Extinction Prevention Program (PEPP). 2010. Plant Extinction Prevention Program Annual Report Fiscal Year 2010 (July1, 2009-June 30, 2010). 121 pp.

Perlman, S. 2009, in litt. Response from National Tropical Botanical Gardens botanist Steve Perlman regarding the Fish and Wildlife Service candidate notice of review for several Bidens plant species noting the presence of pigs, goats, and deer, at Papalaua Gulch on Maui Island; and that goats and deer eat the foliage and rats eat the seeds of Bidens species.

Perkins, R.C.L. 1903. Vertebrata (Aves). Pages 368-465 in Fauna Hawaiiensis, Vol. 1, part 4. D. Sharp (Ed.). University Press, Cambridge, England.

Pimm S.L., H.L. Jones, and J. Diamond. 1988. On the risk of extinction. The American Naturalist 132: 757–785.

Pletschet, S.M. and J.F. Kelly. 1990. Breeding biology and nesting success of Palila. The Condor 92: 1012-1021.

Pratt, L.W. 2008. In litt. Email correspondence and memorandum from botanist, Linda Pratt, to Pacific Islands Fish and Wildlife Office listing program supervisor, Christa Russell, regarding several candidate plant species and the implication of rat predation as a reason for the plant species’ absence of recruitment.

Pratt, L.W., and J.D. Jacobi. 2009. Loss, degradation, and persistence of habitats. Pages 137-158 in T.K. Pratt, C.T. Atkinson, P.C. Banko, J.D. Jacobi, and B.L. Woodworth (Eds.). Conservation biology of Hawaiian forest birds: Implications for island avifauna. Yale University Press, New Haven and London.

104

References

Pratt, T.K., C.T. Atkinson, P.C. Banko, J.D. Jacobi, and B.L. Woodworth (Eds.). 2009. Conservation biology of Hawaiian forest birds: Implications for island avifauna. Yale University Press, New Haven and London.

Price, J.P., J.D. Jacobi, L.W. Pratt, F.R. Warshauer, and C.W. Smith. 2009. Protecting forest bird populations across landscapes. Pages 381-404 in T.K. Pratt, C.T. Atkinson, P.C. Banko, J.D. Jacobi, and B.L. Woodworth (Eds.). Conservation biology of Hawaiian forest birds: Implications for island avifauna. Yale University Press, New Haven and London.

Ralph, C.J, and S.G. Fancy. 1995. Demography and movements of apapane and Iiwi in Hawaii. The Condor 97: 729-742.

Samuel, M.D., P.H.F. Hobbelen, F. DeCastro, J.A. Ahumada, D.A. LaPointe, C.T. Atkinson, B.L. Woodworth, P.J. Hart, and D.C. Duffy. 2011. The dynamics, transmission, and population impacts of avian malaria in native Hawaiian birds: a modeling approach. Ecological Applications 21: 2,960-2,973.

Samuel, M.D., B.L. Woodworth, C.T. Atkinson, P.J. Hart, and D.A. LaPointe. 2015. Avian malaria in Hawaiian forest birds: infection and population impacts across species and elevations. Ecosphere 6: 1-21.

Schulten J.R., T.C. Cole, S. Cordell, K.M. Publico, R. Ostertag, J.H. Enoka, J. Michaud. 2014. Persistence of native trees in an invaded Hawaiian lowland wet forest: experimental evaluation of light and water constraints. Pacific Science 68: Pp. 267–285.

Scott, J.M., S. Mountainspring, F.L. Ramsey, and C.B. Kepler. 1986. Forest bird communities of the Hawaiian Islands: their dynamics, ecology, and conservation. Studies in Avian Biology 9: 1-431

Scowcroft, P.G. and H.F. Sakai. 1983. Impact of feral herbivores on mamane forests of Mauna Kea, Hawaii: bark stripping and diameter class structure. Journal of Range Management 36: Pp. 495–498.

Simpson 2001, in Kubota, 2001, in litt. Deer population boom threatens Maui forests, farms. Honolulu Star-Bulletin, Aug. 28, 2001.

Skolmen, R.G. and D.M. Fujii. 1980. Growth and development of a pure stand of koa (Acacia koa) at Keauhou-Kilauea. Third Conference in Natural Sciences Hawaii Volcanoes National Park, C.W. Smith (Ed.), Cooperative National Park Resources Studies Unit, University of Hawaii, Honolulu. Pp. 301–310.

Smith, C.W. 1985. Impact of alien plants on Hawaii’s native biota. in Hawaii’s Terrestrial Ecosystems: Preservation and Management, C.P. Stone and J.M. Scott

105

References

(Eds.), Cooperative National Park Resources Studies Unit, University of Hawaii, Honolulu. Pp. 180–250.

Smith, C.W. and J.T. Tunison. 1992. Fire and alien plants in Hawaii: research and management implications for native ecosystems. in Alien Plant Invasions in Native Ecosystems of Hawaii, C.P. Stone, C.W. Smith, and J.T. Tunison (Eds.), University of Hawaii Cooperative Park Studies Unit, Honolulu. Pp. 394–408.

Smith, T.M., and R.W. Reynolds. 2004. Improved extended reconstruction of SST (1854- 1997). Journal of Climate 17: 2,466-2,477.

Smucker, T.D., G.D. Lindsey, and S.M. Mosher. 2000. Home range and diet of feral cats in Hawaii forests. Pacific Conservation Biology 6: 229-237.

Snetsinger, T.S., C.M. Herrmann, D.E. Holmes, C.D. Hayward, and S.G. Fancy. 2005. Breeding ecology of the puaiohi (Myadestes palmeri). Wilson Bulletin 117: 72- 84.

Spatz, G. and D. Mueller-Dombois. 1973. The influence of feral goats on koa tree reproduction in Hawaii Volcanoes National Park. Ecology 54: Pp. 870–876.

Sperling F.N, R. Washington, R.J.Whittaker. Future climate change of the subtropical North Atlantic: implications for the cloud forests of Tenerife. Climate Change. 2004 Jul 1; 65(1 – 2): 103-123.

Spiegel, C.S., P.J. Hart, B.L. Woodworth, E.J. Tweed, J.J. LeBrun. 2006. Distribution and abundance of forest birds in low-altitude habitat on Hawaii Island: evidence for range expansion of native species. Bird Conservation International, 16, Pp. 175–185.

Stone, C.P. 1985. Alien animals in Hawaii’s native ecosystems: toward controlling the adverse effects of introduced vertebrates. in Hawaii’s Terrestrial Ecosystems: Preservation and Management, Cooperative National Park Resources Studies Unit, University of Hawaii, Honolulu. Pp. 253, 254–255, 261–264.

Sumida, S., F. Hughes, K. Friday, and J. Leary. 2013. Albizia: the tree that ate Puna. U.S. Forest Service (USFS), Pacific Islands Research Station Factsheet. 2015. 2 pp.

Timm, E., T.W. Giambelluca, and H.F. Diaz. 2015. Statistical downscaling of rainfall changes in Hawaii based on the CMIP5 global model projections, J. Geophys. Res. Atmos., 120, 92–112, doi:10.1002/2014JD022059.

Tomich, P.Q. 1986. Mammals in Hawaii: a synopsis and notational bibliography. Bishop Museum Press, Honolulu. 375 pp.

106

References

Traill, L.W., B.W. Brook, R.R. Frankham, and C.J.A. Bradshaw. 2009. Pragmatic population viability targets in a rapidly changing world. Biological Conservation 143: 28-34.

Tummons, P. 2011. Pigs undermine progress in restoring native bird habitat at Hakalau Refuge. Environment Hawaii 21 (12): 1, 6–7.

Tunison, J.T., C.M. D’Antonio, and R.K. Loh. 2002. Fire and invasive plants in Hawai’i Volcanoes National Park. In Proceedings of the Invasive Species Workshop: the Role of Fire in the Control and Spread of Invasive Species, Fire Conference 2000: the First National Congress on Fire Ecology, Prevention, and Management, K.E.M. Galley and T.P. Wilson (Eds.), Tall Timbers Research Station, Tallahassee. Pp. 122–130.

Tunison, J.T., C.M. D’Antonio, and R.K. Loh. 2011. Fire and invasive plants in Hawaii Volcanoes National Park. Pages 122-131 in K.E.M. and T.P. Wilson (Eds.). Proceedings of the invasive species workshop: the role of fire in the control and spread of invasive species. Fire conference 2000: the first national congress on fire ecology, prevention, and management. Miscellaneous Publication No. 11, Tall Timbers Research Station, Tallahassee, Florida.

Tweed, E.J., J.T. Foster, B.L. Woodworth, W.B. Monahan, J.L. Kellerman, and A. Lieberman. 2006. Breeding biology and success of a reintroduced population of the critically endangered puaiohi. Auk 123:753-763.

U.S. Forest Service (USFS), Pacific Southwest Research Station. 2015. Biological control of strawberry guava in Hawaii. http://www.fs.fed.us/psw/topics/biocontrol/strawberryguava/strawberry_guava.sht ml. (Downloaded 24 November 2015).

United Nations Framework Convention on Climate Change (UNFCCC). 2015. Synthesis report on the aggregate effect of the intended nationally determined contributions. Advance version, 30 October, 2015. 66 pp.

University of Hawaii (UH) College of Tropical Agriculture (CTAHR). 2016. Rapid ohia death fact sheet and website. http://www2.ctahr.hawaii.edu/forestry/disease/ohia_wilt.html

Van Riper, S.G. and C. Van Riper III. 1982. A field guide to the mammals in Hawaii. Oriental, Honolulu, Hawaii, USA, 68 pp.

Van Riper, C., S.G. Van Riper, M.L. Goff, and M. Laird. 1986. The epizootiology and ecological significance of malaria in Hawaiian land birds. Ecological Monographs 56: 327-344.

107

References

Van Riper, C., S.G. Van Riper, and W. Hansen. 2002. The epizootiology and ecological significance of avian pox in Hawaii. Auk 119: 929-942.

Vanderwerf, E.A. 2009. Importance of Nest Predation by Alien Rodents and Avian Poxvirus in Conservation of Oahu Elepaio. The Journal of Wildlife Management, 73: 737–746. doi: 10.2193/2008-284

Vanderwerf, E.A. 2016. Pacific Rim Conservation. https://www.pacificrimconservation.org/contact-us/contact-eric-vanderwerf/. Personal communication regarding surveys for and the status of Iiwi on Oahu.

VanderWerf, E.A., and J.L. Rohrer. 1996. Discovery of an Iiwi population in the Koolau mountains of Oahu. Elepaio 56: 25-28

VanderWerf, E.A., and D.G. Smith. 2002. Effects of alien rodent control on demography of the Oahu elepaio, an endangered Hawaiian forest bird. Pacific Conservation Biology 8: 73-81. van Vuuren, D.P., J. Edmonds, M. Kainuma, K. Riahi, A. Thomson, K. Hibbard, G.C. Hurtt, T. Kram, V. Krey, J.F. Lamarque, T. Masui, M. Meinshausen, N. Nakicenovic, S.J. Smith, and S.K. Rose. 2011. The representative concentration pathways: an overview. Climatic Change 109: 5–31, DOI 10.1007/s10584-011- 0148-z

Vitousek, P.M., C.M. D’Antonio, L.L. Loope, M. Rejmanek, and R. Westerbrooks. 1997. Introduced species: a significant component of human-caused global change. New Zealand Journal of Ecology 21: 1-16.

Vorsino, A.E., L.B. Fortini, F.A. Amidon, S.E. Miller, J.D. Jacobi, J.P. Price, and Koob, G.A. 2014. Modeling Hawaiian ecosystem degradation due to invasive plants under current and future climates. PloS one, 9: p.e95427.

Wagner, W.L., D.R. Herbst, and S.H. Sohmer. 1999. Vegetation, contributed by W.C. Gagne and L.W. Cuddihy. in Manual of Flowering Plants of Hawaii, W.L. Wagner, D.R. Herbst, and S.H. Sohmer (Eds.), University of Hawaii Press, Bishop Museum Press, Honolulu. Pp. 45-114.

Waide, R.B. 1991. The effect of Hurricane Hugo on bird populations in the Luquillo Experimental Forest, Puerto Rico. Biotropica 23: 475-480.

Warner, R.E. 1968. The role of introduced diseases in the extinction of the endemic Hawaiian avifauna. Condor 70: 101-120.

Waring, G.H. 1996, in litt. Preliminary study of the behavior and ecology of Axis deer on Maui, Hawaii. USGS BRD PIERC Haleakala Field Station, Carbondale. 9 pp.

108

References

Webster, P.J., G.J. Holland, J.A. Curry, and H.R. Chang. 2005. Changes in tropical cyclone number, duration, and intensity in a warming environment. Science 309: 1,844-1,846.

Wiley, J.W. and J.M. Wunderle, Jr. 1993. The effects of hurricanes on birds, with special reference to Caribbean islands. Bird Conservation International 3: 319-349.

Winner, W.E., and H.A. Mooney. 1980. Ecology of SO2 resistance: V. effects of volcanic SO2 on native Hawaiian plants. Oecologia 66: 387-393.

Woodworth, B.L., and T.K. Pratt. 2009. Life History and Demography. Pages 194-233 in T.K. Pratt, C.T. Atkinson, P.C. Banko, J.D. Jacobi, and B.L. Woodworth (Eds.). Conservation biology of Hawaiian forest birds: Implications for island avifauna. Yale University Press, New Haven and London.

Yorinks, N. and Atkinson, C.T. 2000. Effects of malaria Plasmodium relictum on activity budgets of experimentally-infected juvenile apapane Himatione sanguinea. Auk 117: 731-738.

Zhang, C., Y. Wang, A. Lauer and K. Hamilton, 2012. Configuration and evaluation of the WRF model for the study of Hawaiian regional climate. Monthly Weather Review, 140, 3,259‐3,277.

Zhang, C., Y. Wang, K. Hamilton, and A. Lauer. 2016. Dynamical Downscaling of the Climate for the Hawaiian Islands. Part I: Present-day. J. Climate. doi: 10.1175/JCLI-D-15-0432.1, in press.

109

Appendix A

APPENDIX A

Figure A1. This map shows Hawaii Island current Iiwi range (white opaque) overlap with public lands (of various levels of management) and projected Iiwi range (red border) by the end of the 21st Century due to climate change temperature increase and a reduction in the disease-free habitat. The map also shows a concurrent net loss of 66,984 ac (57 percent decrease) of disease-free habitat on managed public lands.

110

Appendix A

Figure A2. This map shows western and southern Hawaii Island current Iiwi range (white opaque) overlap with public lands (of various levels of management) and projected Iiwi range (red border) by the end of the 21st Century due to climate change temperature increase and a reduction in the disease-free zone.

111

Appendix A

Figure A3. This map shows eastern Hawaii Island current Iiwi range (white opaque) overlap with public lands (of various levels of management) and projected Iiwi range (red border) by the end of the 21st Century due to climate change temperature increase and a reduction in the disease-free habitat.

112

Appendix A

Figure A4. This map shows Maui current Iiwi range (white opaque) overlap with public lands (of various levels of management) and projected Iiwi range (red border) by the end of the 21st Century due to climate change temperature increase and a reduction in the disease-free zone. The map shows a concurrent net loss of 7,687 ac (54 percent decrease) of disease-free habitat on managed public lands.

113

Appendix A

Figure A5. This map shows Molokai current Iiwi range (white opaque) overlap with public lands (with various levels of management). Because of Molokai’s relatively low elevation, there is a projected absence of disease-free habitat, and thus, a complete loss of Iiwi range on the island by the end of the 21st Century due to climate change temperature increase (Fortini et al. 2015).

114

Appendix A

Figure A6. This map shows Kauai current Iiwi range (white opaque) overlap with public lands (of various levels of management) and projected Iiwi range (red border) by the end of the 21st Century due to climate change temperature increase and a reduction in the disease-free zone. The map also shows a concurrent net loss of 7,687 ac (82 percent decrease) of disease-free habitat on managed public lands.

115

Appendix A

Figure A7. This map shows Hawaii Island current Iiwi range (white opaque) and projected Iiwi range (red border) by the end of the 21st Century due to climate change temperature increase and reduction in disease-free habitat (Fortini et al. 2015). Concurrent with the downslope loss of disease-free habitat, the map also depicts a net loss to the Iiwi of ohia forest habitat (green shading) and a net gain of less ideal ohia scrub forest habitat (gold shading), with ohia scrub comprising an additional 16 percent of Iiwi projected future habitat.

116

Appendix A

Figure A8. This map shows east Hawaii Island current Iiwi range (white opaque) and projected Iiwi range by the end of the 21st Century (red border) due to climate change temperature increase and reduction in disease-free habitat (Fortini et al. 2015).

117

Appendix A

Figure A9. This map shows western and southern Hawaii Island current Iiwi range (white opaque) and projected Iiwi range by the end of the 21st Century (red border) due to climate change temperature increase and reduction in disease-free habitat (Fortini et al. 2015).

118

Appendix A

Figure A10. This map shows Maui current Iiwi range (white opaque) and projected Iiwi range (red border) by the end of the 21st Century due to climate change temperature increase and reduction in disease-free habitat (Fortini et al. 2015). Concurrent with the downslope loss of disease-free habitat, the map also depicts a net loss to the Iiwi of ohia forest habitat (green shading) and a net gain of less ideal ohia scrub forest habitat (gold shading), with ohia scrub comprising an additional 6 percent of Iiwi projected future habitat.

119

Appendix A

Figure A11. This map shows Molokai current Iiwi range (white opaque) as well as current ohia forest (green shading) vs. current ohia scrub habitat (gold shading. Due to the lower elevation of Molokai, it is expected to lose all of remaining disease-free habitat by the end of the 21st Century due to climate change temperature increase. Therefore, Iiwi is projected to lose all of its range on Molokai because there is no higher elevation habitat in which to shift (Fortini et al. 2015), regardless of any possible elevation shift in ohia forest.

120

Appendix A

Figure A12. This map shows Kauai current Iiwi range (white opaque) and projected Iiwi range (red border) by the end of the 21st Century due to climate change temperature increase and reduction in disease-free habitat (Fortini et al. 2015). Concurrent with the downslope loss of disease-free habitat, the map also depicts a net loss to the Iiwi of ohia forest habitat (green shading).

121