Introduced Avian Diseases, Climate Change, and the Future of Hawaiian Honeycreepers Author(S): Carter T

Introduced Avian Diseases, Climate Change, and the Future of Hawaiian Honeycreepers Author(s): Carter T. Atkinson MS, PhD and Dennis A. LaPointe MS, PhD Source: Journal of Avian Medicine and Surgery, 23(1):53-63. 2009. Published By: Association of Avian Veterinarians DOI: http://dx.doi.org/10.1647/2008-059.1 URL: http://www.bioone.org/doi/full/10.1647/2008-059.1

BioOne (www.bioone.org) is a nonprofit, online aggregation of core research in the biological, ecological, and environmental sciences. BioOne provides a sustainable online platform for over 170 journals and books published by nonprofit societies, associations, museums, institutions, and presses. Your use of this PDF, the BioOne Web site, and all posted and associated content indicates your acceptance of BioOne’s Terms of Use, available at www.bioone.org/page/terms_of_use. Usage of BioOne content is strictly limited to personal, educational, and non-commercial use. Commercial inquiries or rights and permissions requests should be directed to the individual publisher as copyright holder.

BioOne sees sustainable scholarly publishing as an inherently collaborative enterprise connecting authors, nonprofit publishers, academic institutions, research libraries, and research funders in the common goal of maximizing access to critical research. Journal of Avian Medicine and Surgery 23(1):53–63, 2009 ’ 2009 by the Association of Avian Veterinarians

Historical Perspective

Introduced Avian Diseases, Climate Change, and the Future of Hawaiian Honeycreepers Carter T. Atkinson, MS, PhD, and Dennis A. LaPointe, MS, PhD

Introduction Of 41 species and subspecies of honeycreepers known since historic times,5 17 are thought to be The Hawaiian archipelago is isolated in the extinct and 14 are federally listed as endangered.6 central Pacific and consists of 7 large islands and a Only 3 species and subspecies, Hawaii àmakihi chain of low coral atolls and small rocky islets that (Hemignathus virens virens), Maui àmakihi extend in a long arc from Hawaii Island in the (Hemignathus virens wilsoni), and àpapane (Hi- southeast to Kure Atoll in the northwest. The matione sanguinea) are robust enough in terms of archipelago is the most isolated island system in geographic range and population size to be of the world, separated from the nearest continental minimal concern (Table 1). Reasons for these landmass by more than 2000 miles of ocean.1 The declines are complex and include a suite of islands range in elevation from just above sea level for atolls in the Northwestern Hawaiian Islands to interacting factors, including habitat degradation peaks that exceed 4000 m on Hawaii Island. The and loss from human activities and invasive interaction of extreme topographic relief, trade species, introduced predators, introduced avian winds, and local climatic patterns creates a wide competitors, and introduced avian diseases and diversity of habitats, ranging from alpine deserts disease vectors. on the highest peaks to montane rain forests with Although the interactions among these various precipitation exceeding 7600 mm per year. limiting factors makes it difficult to determine The endemic passerine avifauna of the Hawaiian their relative impacts on native forest bird Islands, particularly the endemic Hawaiian hon- populations, it is clear that the abundance, eycreepers (subfamily Drepanidinae) is often diversity, and geographic distribution of Hawaii’s heralded as an outstanding example of adaptive native birds changed significantly after the arrival 7–10 radiation, equal to Darwin’s finches from the of mosquitoes, avian malaria, and poxvirus. Galapagos Islands in terms of diversity of bill types Other indigenous and introduced pathogens and and number of species that descended from a diseases do not appear to have had the popula- common founder.2 From an initial colonization by tion-level impacts of avian malaria and pox. These only a few individuals of a single ancestral include Toxoplasma gondii from àlalaˆ(Corvus cardueline finch, this group radiated throughout hawaiiensis), neˆneˆ(Branta sandvicensis), and wild 11,12 the diverse habitats on the islands, specializing on game birds ; erysipelas (Erysipelothrix rhusio- 13 a variety of food resources that included nectar, pathiae)fromàlalaˆ ; echinuriasis (Echinuria un- 14 fruits, and insects. Based on recent studies of cinata) from Laysan teal (Anas laysanensis) ;and subfossils, the diversity of this group may have sporadic outbreaks of avian botulism (Clostridium reached 20 genera with more than 50 species prior botulinum) from water birds on Oàhu and Hawaii to human contact with the islands.3 Remarkably, a islands and most recently from Laysan teal on new genus and species of honeycreeper, the po’ouli Midway Island (National Wildlife Health Center, (Melamprosops phaeosoma), was described in the unpublished data, December 2008). 1970s from remote rain forests on Maui.4 Today, the endemic Hawaiian avifauna faces History and Origins of Avian Malaria and Pox one of the highest rates of extinction in the world. in Hawaii

From the US Geological Survey Pacific Island Ecosystems Avian malaria is a disease caused by intracel- Research Center, Kilauea Field Station, PO Box 44, Hawaii lular, mosquito-transmitted protozoan parasites National Park, HI 96718, USA. in the genus Plasmodium. These parasites have a

53 54 JOURNAL OF AVIAN MEDICINE AND SURGERY

Table 1. Status of species and subspecies of Hawaiian honeycreepers (Drepanidinae) known from historic times. Compiled from Scott et al,5 US Fish and Wildlife Service,6 Leonard,72 and the IUCN Red List.73

Common name Scientific name ICUN status Federal status La¯naì hookbill Dysmorodrepanis munroi extinct extinct Lesser koa-finch Rhodacanthis faviceps extinct extinct Greater koa-finch Rhodacanthis palmeri extinct extinct Kona grosbeak Chloridops kona extinct extinct Greateràmakihi Hemignathus sagittirostris extinct extinct Lesser àkialoa Hemignathus obscurus extinct extinct Oàhuàkialoa Hemignathus ellisianus ellisianus extinct extinct Lana` i àkialoa Hemignathus ellisianus lanaiensis extinct extinct Oàhu nukupuù Hemignathus lucidus lucidus extinct extinct Oàhuàlauahio Paroreomyza maculata critically endangered extinct Ka¯ka¯wahie Paroreomyza flammea extinct extinct Lanaì àlauahio Paroreomyza montana montana nd extinct Oàhuà¯kepa Loxops coccineus wolstenholmei nd extinct Ula-ài-hawane Ciridops anna extinct extinct Hawaii mamo Drepanis pacifica extinct extinct Black mamo Drepanis funerea extinct extinct Laysan honeycreeper Himatione sanguinea freethii nd extinct Ò¯ ù¯ Psittirostra psittacea critically endangered endangereda Kauaì àkialoa Hemignathus ellisianus procerus extinct endangereda Kauaì nukupuù Hemignathus lucidus hanapepe critically endangered endangereda Maui nukupuù Hemignathus lucidus affinis critically endangered endangereda Maui à¯kepa Loxops coccineus ochraceus endangered endangereda Poòuli Melamprosops phaeosoma critically endangered endangereda Laysan finch Telespiza cantans vulnerable endangered Nihoa finch Telespiza ultima critically endangered endangered Palila Loxioides bailleui endangered endangered Maui parrotbill Pseudonestor xanthophrys critically endangered endangered Àkiapo¯la¯àu Hemignathus munroi endangered endangered Hawaii creeper Oreomystis mana endangered endangered Hawaii à¯kepa Loxops coccineus coccineus endangered endangered À¯ kohekohe Palmeria dolei critically endangered endangered Àkikiki Oreomystis bairdi critically endangered candidate for listing Àkekeè Loxops caeruleirostris critically endangered not listed Hawaii àmakihi Hemignathus virens virens least concern not listed Maui àmakihi Hemignathus virens wilsoni least concern not listed Oàhu àmakihi Hemignathus flavus vulnerable not listed Kauaì àmakihi Hemignathus kauaiensis vulnerable not listed Ànianiau Hemignathus parvus vulnerable not listed Maui àlauahio Paroreomyza montana newtoni endangered not listed Ììwi Vestiaria coccinea vulnerable not listed Àpapane Himatione sanguinea sanguinea least concern not listed Abbreviations: IUCN indicates International Union for Conservation of Nature and Natural Resources; nd, no data reported. a No confirmed sightings in recent years, possibly extinct. worldwide distribution in multiple avian families skin or diphtheritic lesions on the mouth, trachea, but occur primarily in passerine birds.15,16 More and esophagus of infected birds. It can be than 40 species of Plasmodium have been transmitted mechanically through contact with described, but they are rarely associated with infected objects or on the mouth parts of blood- epidemic disease.17,18 Avian pox (Avipoxvirus sucking arthropods. The 13 different species species), by contrast, is a viral infection caused currently recognized are defined by host associ- by a large, double-stranded DNA virus that ations ranging from the genus to the family typically causes tumor-like swellings on exposed level.19,20 Epidemic outbreaks of pox in wild ATKINSON AND LAPOINTE—INTRODUCED AVIAN DISEASES IN HAWAII 55 passerine birds and gallinaceous birds are not Hawaii, with possible origins from birds that uncommon and have also been reported in wild were introduced from the Old World.34 In spite of turkeys (Meleagris gallopavo),21 Galapagos finch- the likely origin of Prelictumfrom one or more of es,22 and endemic birds in the Canary Islands.23 these introduced species, these species no longer The broad host range of both avian pox and appear to be the primary reservoirs of infection. avian malaria makes them particularly well suited Surveys conducted during the past 30 years have for invading new areas. consistently identified the highest prevalence of The dates when avian malaria and pox were infection in native, rather than introduced, forest introduced to the islands remain unknown. birds.8,35–37 However, from the descriptions of tumor-like The first epidemiologic studies of avian malaria swellings on dead or moribund forest birds by late were conducted in lowland habitats near Lıˆhuè, 19th-century naturalists24 and the recent polymer- Kauaì in the 1950s.7 Laysan finches (Telespiza ase chain reaction (PCR) amplification of poxvi- cantans), àpapane, Kauaì àmakihi (Hemi- rus sequences from museum specimens collected gnathus kauaiensis), and ànianiau (Hemignathus during that period,25 Avipoxvirus clearly was well parvus) were exposed to local mosquitoes in established in native forest bird populations by unscreened cages and subsequently died from the late 1800s. Although domestic poultry had fulminating pox and malarial infections. This long been assumed to be the most likely source of simple but convincing demonstration of local the virus in Hawaii, the 2 or more genetically disease transmission and pathogenicity became distinct variants of the virus circulating in native the cornerstone for more detailed ecologic studies forest birds are distinct from fowlpox and have in the 1970s, which documented widespread strong similarities to canarypox.25–27 occurrence of malaria in native forest birds on The most likely opportunity for introduction of the island of Hawaii.8 avian malaria was in the early 20th century, when local bird clubs introduced nonnative passerine The Role of Pox and Malaria in the Decline and birds from around the world to replace low- Extinction of Hawaiian Birds elevation native birds that were vanishing from pox infection and other limiting factors. More There is little direct evidence implicating avian than 100 documented introductions of passerine pox and malaria in the extinction of any birds from Southeast Asia, North and South Hawaiian bird, even though it is often cited as a America, and Africa were made between 1900 primary factor in the demise of the endemic and the 1960s, with more than 50 species becoming avifauna. It has been suggested that the major established here.28 It is unlikely that any of these waves of extinction that occurred in the late 1800s birds were screened for infectious diseases and the and again after 1910 follow the presumed number of avian pathogens that may have been introduction of avian pox and malaria.8,9 How- inadvertently introduced remains unknown.29,30 ever, the anecdotal observations by early natu- Unlike the externally evident lesions of pox ralists and the absence of any comprehensive infection, the more cryptic Plasmodium infections surveys of native birds or avian disease before the were first detected in the 1930s by blood smear in 1970s make this difficult to confirm. a red-billed leiothrix (Leiothrix lutea)anda The most convincing indirect evidence that Japanese white-eye (Zosterops japonica) collected avian pox and malaria have had a major impact at Hawaii Volcanoes National Park.31,32 Consid- on forest bird populations is their strong negative erable initial confusion existed about which correlation with extant native birds. Warner was species of Plasmodium had been introduced to the first to draw a parallel between declines in Hawaii, or whether multiple species were present, native bird populations at lower elevations and because the erythrocytic stages of these parasites the presence of mosquitoes and the pathogens have so few distinguishing morphologic features.33 they transmit.7 He suggested that mosquito A detailed study of blood smears from native and populations declined precipitously above 600 m. nonnative species across the state concluded that Later work documented occurrence of the vector only a single species of Plasmodium was present as high as 2500 m, but also supported Warner’s and that it was most similar to Plasmodium basic idea by demonstrating a negative associa- relictum capistranoae.33 More recently, molecular tion between mosquito numbers and native birds work has supported identification of Prelictumas at elevations below 1500 m.8,38,39 the sole species of Plasmodium in Hawaii. Only a The high susceptibility of native honeycreepers single lineage of this parasite is currently in to avian malaria and pox also provides strong 56 JOURNAL OF AVIAN MEDICINE AND SURGERY circumstantial evidence that these diseases have occur almost exclusively in extensively modified had substantial impacts on wild populations. low-elevation forests with high mosquito densi- Native birds with naturally acquired infection ties.44 may have parasitemia involving more than 95% Unlike avian malaria, relatively little is known of circulating erythrocytes, acute anemia with about the pathogenesis of avian poxvirus in packed cell volumes as low as 16%, and rapid Hawaiian forest birds, and the report by Warner deterioration of body condition, with mortality still provides some of the best documentation rates ranging from 40% to almost 100% between about the course of infection and pathogenicity of 5 and 15 days after exposure.7 the virus in native birds.7 Twenty-four Laysan More recent studies have conducted experi- finches were exposed to mosquitoes in downtown mental malarial infections in native and nonna- Honolulu after keeping their cage wrapped in tive birds using single and multiple mosquito bites mosquito-proof cheesecloth for 2 months. Within to mimic natural transmission.40–43 Declines in 2 weeks after the protective cheesecloth was food consumption and loss of body mass began removed, 6 finches developed indurated swellings approximately 1 week after exposure to single on the lores and tarsal and wing joints where infective mosquito bites in ììwi (Vestiaria cocci- mosquitoes had access to exposed skin. The nea),40 Hawaii àmakihi,41 Maui àlauahio (Par- swellings increased in size and then erupted into oreomyza montana newtoni),42 and àpapane.43 granular, tumor-like lesions that became necrotic These declines correlated closely with increases with accompanying secondary bacterial infec- in numbers of erythrocytic parasites in the tions. As tumors progressed in severity they peripheral circulation. Survivorship was relatively tended to bleed, particularly from the foot lesions. high in infected àpapane and Hawaii àmakihi By the end of 1 month of exposure, every finch with up to 30% developing low-intensity chronic had at least one lesion. Death occurred over a infections that stimulated immunity to reinfection variable period of time after birds lost body when challenged with multiple infective mosquito condition and lesions became necrotic. bites. The effects were much more severe inììwi; More recently, molecular characterization and individuals of this species had significantly higher experimental infections have revealed 2 genetical- mortality (90%) and peak parasitemias at death. ly distinct variants of poxvirus that differ Gross and microscopic lesions were also more significantly in virulence.25 Both variants caused severe in ììwi. Results of necropsy and histo- lesions when inoculated into the foot pads of pathologic examination revealed massive enlarge- Hawaii àmakihi, but size and severity of the ment of the liver and spleen with extensive lesions differed. Lesions associated with variant 1 deposition of malarial pigment and diffuse were relatively small and self-limiting, whereas extramedullary erythropoiesis in the liver and those associated with variant 2 were large, kidneys. Mature erythrocytes were replaced al- bloody, proliferative, and ultimately fatal in 4 most entirely by immature erythrocytes and experimentally infected birds. Both variants were precursors that are normally found only in bone isolated from Hawaii àmakihi from Kilauea marrow. By contrast, it was difficult to infect Volcano, but phylogenetic analyses based on nonnative species with single infective mosquito sequence of the virus 4b core protein grouped bites; parasitemias were low and clinical signs of variant 1 closely with canarypox whereas variant infection were not evident. 2 clustered in a different, but closely related clade. One of 2 extant species of native thrush, òˆmaò Hawaii àmakihi that recovered from infection (Myadestes obscurus), also appears to be relatively with variant 1 were not immune to subsequent resistant to malaria. Four birds did not develop challenge with variant 2. This agrees well with clinical signs of disease when exposed to single field observations of recurring pox infections in infectious mosquito bites and birds also recovered some recaptured birds that have healed lesions. from acute infections.42 The native monarch These findings also raise questions about the flycatchers, Hawaii èlepaio (Chasiempis sand- temporal and spatial dynamics of virus transmis- wichensis sandwichensis)andOàhu èlepaio sion in wild birds and suggest that interactions (Chasiempis sandwichensis gayi), also appear to between genetically and biologically distinct have some resistance to malaria, based on their poxviruses, their susceptible hosts, and concur- current altitudinal distribution and occasionally rent malarial infections may play important roles high abundance in some mid- and low-elevation in determining severity of epidemics. habitats where disease prevalence is high. This is Other than these limited experimental studies, particularly true on Oàhu, where Oàhuèlepaio information about the potential impact of poxvi- ATKINSON AND LAPOINTE—INTRODUCED AVIAN DISEASES IN HAWAII 57 rus on Hawaiian forest bird populations is based The presence or absence of mosquitoes and on observations of pox-like lesions on captured their relationships to native bird distribution was wild birds. Although these studies are limited by first quantified by Scott et al38 who found that the presumptive nature of the diagnosis, it has threatened and endangered passerine birds were been argued that most pox-like lesions are likely less abundant in areas with detectable mosquito to be caused by poxvirus. In one study, virus was populations and reached their greatest densities at recovered from 20 of 22 pox-like lesions from wild altitudes higher than 1500 m. These observations Hawaiian birds.9 Using a presumptive diagnosis are consistent with the uncommon occurrence of based on presence or absence of tumor-like Maui àlauahio and ììwi below elevations of swellings or missing digits, results of another 1500 m; these are 2 species in which mortality study found declines in numbers of some breeding exceeded 75% after exposure to single infective Hawaii èlepaio that were correlated with the mosquito bites.40,42 These findings are also con- occurrence of pox epidemics.45 Preliminary obser- sistent with laboratory studies that demonstrate vations of Oàhu èlepaio suggest that annual thermal constraints on sporogonic development mortality of birds with active lesions is up to 40%, of Prelictumin experimentally infected Culex whereas birds with mild infections involving only quinquefasciatus.47 Sporogonic development of the oneormoretoesfrequentlyrecover.45,46 parasite dramatically slows at 15uC and ceases at Recent work also indicates that prevalence of 13uC. The 15uC isotherm in Hawaii closely malaria in Oàhu èlepaio is extremely high with follows the 1500-m contour line and high- an average infection rate of 87% in birds that elevation refugia may also be maintained by were sampled between 1995 and 2005.44 Pox and thermal limits on malarial development. malaria infection appeared to be independent of At elevations below 1500 m, the key factor each other in this study, whereas in other field driving epizootics of pox and malaria is the studies in Hawaii, birds with pox-like lesions are seasonal and altitudinal distribution and density infected with concurrent malarial infections more of C quinquefasciatus. The important role of this frequently than expected by chance.8,36 Although mosquito species in the epidemiology of malaria this suggests that the 2 diseases interact with each transmission has been supported by identification other, we know little about whether this apparent of infected individuals in the wild during periods interaction is caused by simultaneous transmis- of peak disease transmission and by laboratory sion of the diseases by the vector or by differential experiments that have demonstrated that C mortality among pox-infected, malaria-infected, quiquefasciatus is several orders of magnitude or coinfected birds or whether it is simply a more susceptible to the parasite than other 8,48,49 reflection of the longer exposure that birds with introduced mosquito species. The dynamics chronic malaria would have to potential pox of vector populations on the east flank of Mauna infections. The high frequency of concurrent pox Loa and Kilauea volcanoes has been recently and malaria infections makes it extremely difficult modeled, illustrating the dependence of vector to identify demographic effects of either agent populations on both temperature and rainfall and alone. the seasonal fluctuations of vector populations across altitudinal gradients.50 Peak mosquito Vector Abundance and Transmission Across populations occur at elevations between 1200 Altitudinal Gradients and 1500 m during the warmer fall months of September to December.47 It is during this time of A model has been proposed that depicts the year that seasonal epidemics of malaria and transmission of avian pox and malaria across an pox transmission are common (C. T. A., unpub- altitudinal gradient on Mauna Loa and Kilauea lished data, 1991–1998). volcanoes.8,9 This model places the highest rates Other than the early mosquito exposure exper- of malaria and pox transmission in midelevation iments by Warner,7 little research has been (1200 m) forests where the declining numbers of conducted on the prevalence and transmission mosquitoes between sea level and tree line of these diseases in habitats below 600 m. Shehata overlaps with increasing numbers of native forest et al51 and VanderWerf46 measured prevalence of birds at elevations above 1000 m.8 Disease malarial and pox infections in lowland bird transmission disappears at elevations above communities on Oàhu but did not study vector 1500 m when mosquito populations fall to low populations or measure transmission rates. Prev- levels, creating high elevation ‘‘refugia’’ from alence of malarial infections in Oàhu àmakihi avian malaria. captured in Maˆnoa Valley were surprisingly low, 58 JOURNAL OF AVIAN MEDICINE AND SURGERY with only 1 of 16 (6.3%) birds positive for malaria studies, it seems likely that malaria swept rapidly by serologic methods and none of 42 individuals across all of the lower Hawaiian Islands after it positive by PCR.51 Prevalence in a sample of 268 was introduced, leaving few survivors. The cool, nonnative birds from the same area was 11.6% by high-elevation mountains of Kauaì, Maui, and PCR, suggesting that the Oàhu àmakihi may Hawaii not only provided the only refugia from have some refractoriness to infection. Without the diseases, but also fueled high rates of disease information about vector populations or trans- transmission at middle elevations by providing a mission rates, however, it is difficult to rule out continuous source of highly susceptible native alternative explanations, including whether the birds. Disease transmission at middle elevations is birds were immune survivors of a previous fueled every fall by increasing mosquito popula- infection with parasitemias too low to be detected tions and a seasonal influx of dispersing, suscep- by PCR, or whether they were simply not exposed tible birds from higher elevations. to infective mosquito bites.51,52 The recent emergence of low-elevation àma- On Hawaii Island, recently emergent popula- kihi populations suggest that the system is tions of Hawaii àmakihi have been found in evolving at different rates because of variation remnant patches of òˆhià(Metrosideros polymor- in selective pressure across the altitudinal gradi- pha) forest in the Puna district at elevations below ent.49,54 With transmission occurring year round 300 m.49,53 These populations are expanding in at lower elevations, and low-elevation popula- range at densities that exceed those of Hawaii tions not being continually diluted by emigrating, àmakihi at elevations above 1500 m on Mauna highly susceptible juvenile birds from high eleva- Loa Volcano. These populations are increasing in tions, it might be predicted that disease resistance spite of prevalences of infection that range from would first appear here. This suggests that disease 55% to 83% by serologic methods.49 Unlike the resistance may subsequently spread over the next apparently Plasmodium-refractory Oàhuàmakihi, few decades, with eventual recovery of mideleva- these lowland Hawaii àmakihi appear to be more tion populations of the more resistant species. resistant to the pathologic effects of malarial Although this predicted scenario is encouraging infection and have unique nuclear and mitochon- for the more common native species that current- drial haplotypes that are not found in àmakihi ly have the widest distribution, there is concern from high elevation.54 Analysis of museum spec- that threatened and endangered species may not imens of low-elevationàmakihi collected between have sufficient genetic variability to adapt to these 1898 and 1948, just before the presumed intro- diseases. duction of malaria to Hawaii, identified the same unique haplotypes present today, suggesting that Global Warming and Its Impact on the recent resurgence of these birds originated Hawaiian Disease Ecology from pockets of surviving individuals with some The ecology of pox and malaria transmission in natural disease resistance, rather than recoloniza- Hawaii depends on climatic conditions—primar- tion of the lowlands by high elevation birds.54 ily seasonal changes in temperature and rainfall Low-elevation Hawaii àmakihi have also been that drive vector populations.49 These environ- recognized in areas of East Maui at 300 m mental factors may be responsible for persistence elevation and almost at sea level in Pelekunu of high-elevation refugia on the higher islands. Valley on Molokaì (C. T. A., unpublished data, The effects of a 2 C warming on the altitudinal 2003). Similar to Hawaii Island, prevalence of u location of the 13 Cand17C isotherms have infection in low-elevation àmakihi on Molokaì u u been modeled. These isotherms delineate a high- exceeds 75% (C. T. A., unpublished data, 2003), risk zone for malaria transmission (below the suggesting that selection for disease resistance has 17 C isotherm), a transition zone with seasonal occurred on other islands as well. u epizootic transmission (between the 17uCand 13 C isotherms), and a low-risk zone for trans- Dynamics of Epidemic Outbreaks u mission (above the 13uC isotherm) where thermal When highly virulent pathogens are introduced constraints inhibit complete development of the into naı¨ve populations, epizootics can spread parasite in the mosquito vector. Given habitat rapidly and have major demographic impacts limitations by current land use and following a over wide geographic areas. Based on what we projected 2uC temperature rise, available high- know about the high susceptibility of honeycreep- elevation forest habitat in the low-risk zone is ers to malaria from both laboratory and field predicted to decline by 57% at Hanawıˆ Natural ATKINSON AND LAPOINTE—INTRODUCED AVIAN DISEASES IN HAWAII 59

sides of the high volcanoes on Hawaii and Maui and allowing the lower limits of high-elevation refugia to continue to remain above disease transmission.55 Recent analyses of climatic data from Hawaii indicate that temperatures have shown a significant upward trend over the past 80 years, but disease transmission may be influenced by more than a general warming trend. The tropical inversion layer may play a more significant role than temperature in determining tree line and the upper extent of forest bird habitat in the Hawaiian Islands.56 Often visible as a thin layer of clouds around higher peaks, the inversion layer forms as cool, dry air descends and meets warm, moist air that is driven upward by convection currents, prevailing trade winds, and topographic relief.57 As these air masses meet, they form an inversion layer that caps moisture and cloud development between 1800 and Figure 1. Aerial photograph of the windward, eastern 2400 m. The inversion layer has remained rela- slope of Mauna Kea Volcano on the island of Hawaii. tively stable in height but has increased in Native forest below 600 m has mostly been lost to frequency of occurrence during the past 25 intensive agricultural development. Although extensive years.56,58 Although it is difficult to predict tracts of native forest persist between 600 m and tree whether this trend will continue, it suggests that line at approximately 2000 m, they are threatened by changes in the base height of the inversion layer invasive species. High-elevation disease refugia on may ultimately have a substantial impact on the Mauna Kea are limited on the eastern side of the mountain to a narrow band of forest above 1500 m persistence of high-elevation disease refugia. If a that is capped by both the inversion layer and by stable inversion layer and its effect on rainfall degradation and loss of forest cover by ranching. High- prevents expansion of forest bird habitat into elevation refugia provide a safe haven from disease higher elevations, remaining high-elevation forest transmission for threatened and endangered species bird populations may be squeezed between such as the àkiapo¯la¯àu, Hawaii creeper, and Hawaii expanding disease transmission from lower eleva- à¯kepa, but also provide a source of uninfected juvenile tions and the upper limits of suitable habitat and adult honeycreepers that help to maintain seasonal (Fig 1). These changes would likely push remain- epidemics of malaria and pox at elevations below ing populations of threatened and endangered 1500 m. In a warming climate, high-elevation populations of Hawaii’s most endangered birds may be honeycreepers to extinction, and cause severe threatened by expanding disease transmission, partic- declines in nonendangered species such as ììwi ularly if upward expansion of critical forest habitat is and Maui àlauahio that exhibit high susceptibil- capped by a stable inversion layer or land-use practices ity to avian malaria. such as cattle ranching. Prospects for Intervention Area Reserve on Maui Island to as much as 96% Without question, the one factor that prevented at Hakalau Forest National Wildlife Refuge on widespread and rapid extinction of virtually all of Hawaii Island. The Alakaì Wilderness Preserve Hawaii’s endemic and highly susceptible honey- on Kauaì Island currently has no areas in the creepers after the introduction of these diseases low-risk zone for malaria transmission because was the presence of significant altitudinal gradi- the highest elevation is less than 1800 m, but it ents on Kauaì, Maui, and Hawaii, where would experience an 85% loss of forested habitat susceptible native birds could maintain high where transmission is highly seasonal to condi- populations in relatively disease-free refugia. tions where transmission could occur throughout While providing havens where rarer species can most of the year. still persist, these high-elevation refugia also set If climatic changes occur slowly, the tree line up conditions for maintaining high rates of and forest bird habitat might be able to keep pace epidemic transmission of the parasite in adjacent with warming temperatures, slowly moving up the midelevation habitats, through seasonal move- 60 JOURNAL OF AVIAN MEDICINE AND SURGERY ment of susceptible, dispersing juvenile birds and Preservation of remaining low-elevation for- through altitudinal movement of susceptible ests may be one of the most cost-effective adults as they follow nectar resources.59 strategies for long-term sustainability of more Given the likelihood of global warming, common endemic species, but this will do little management of midelevation habitats to try to or no good for rare species that may be on the reduce disease transmission is becoming increas- verge of losing uncommon genetic haplotypes ingly important. The best opportunities for doing that may be needed for disease resistance. Given this will probably come through reduction of the threat of global climate change, it will mosquito larval habitat. Feral pigs and other become increasingly important to ensure that ungulates can create larval habitat in forest threatened and endangered species have the habitats and their removal through fencing and inherent genetic variability to allow selection for control programs may significantly reduce mos- disease resistance under natural conditions. This quito habitat, particularly on Kilauea and Mauna requires continued protection and management Loa Volcanoes on the Island of Hawaii, where of existing populations to prevent further loss of volcanic soils are porous and streams and natural rare alleles and use of captive propagation and bodies of water are rare (C. T. A., unpublished translocations to help manage remaining genetic data, 1995).47 Management of man-made water diversity. sources on ranch lands and agricultural opera- The recent increase in watershed partnerships tions and elimination of standing water sources in the Hawaiian Islands to manage and restore that provide habitats for mosquitoes in residential increasingly large areas of mid- and high- areas adjacent to natural areas are also practical elevation habitats for native species,68 the things that can be done to reduce mosquito investment during the past decade in captive numbers.60,61 One key factor to consider when propagation facilities and expertise for some of determining the scale of these efforts is the ability the rarest species in Hawaii,69 the development of C quinquefasciatus to disperse up to 3 km of methods for translocating and releasing through closed-canopy forests,62 making it essen- endemic forest birds,70 and the steady move- tial to control mosquitoes in buffer zones around ment toward development and application of critical mid- and high-elevation habitats. The predator control strategies for large areas of formation of conservation or watershed partner- native forest71 are all important developments ships, such as the Three Mountain Alliance on during the past decade that will help to prevent Hawaii Island63 and the West Maui Mountains further extinctions of Hawaii’s endemic avifauna. Watershed Partnership,64 has brought together tracts of public and private land under a unified Conclusion resource management that may allow for effective landscape level control of avian disease. The effects of avian malaria and poxvirus on There will likely be an important place for native Hawaiian birds have become a classic vaccines and chemotherapy for management of example of the potential impacts of introduced avian pox and malaria during translocation or diseases on naı¨ve wildlife populations. These release of captive birds or management of small diseases continue to be an important example of populations of critically endangered forest birds,65 what can go wrong when pathogens and vectors but there is currently no effective technology for are moved outside of their natural ranges. delivering them to large numbers of wild birds to Although the road to recovery for some species interrupt disease transmission. Although experi- will be long and challenging, some of the more mental DNA vaccines have been developed for common species such as àmakihi and àpapane avian malaria and canarypox vaccines are com- have proven to be remarkably adaptable to mercially available, these tools have not yet been change. Their survival into the next century may applied in Hawaii and their safety and efficacy in ultimately depend on our ability to remove or Hawaiian forest birds need to be determined.66 mitigate introduced threats and restore native 72 Similarly, the recent interest in development of forests from sea level to tree line. transgenic mosquitoes to reduce their competency Acknowledgments: We thank the US Geological as vectors of Plasmodium may eventually have Survey Wildlife and Invasive Species programs, the important applications in Hawaii, but there are US Geological Survey-Natural Resource Protection still significant technical and regulatory issues Program and the National Science Foundation that need to be overcome before this becomes a (DEB0083944) for the financial support that made this reality.67 research possible. The collection of data for various ATKINSON AND LAPOINTE—INTRODUCED AVIAN DISEASES IN HAWAII 61 studies described here would not have been possible 15. Garnham PCC. Malaria Parasites and Other without the assistance and support of numerous Haemosporidia. New York, NY: Blackwell Scien- technical staff, coworkers, and student interns who tific; 1966. spent many long days in the field and laboratory under 16. Valkiu¯nas G. Avian Malarial Parasites and Other often-difficult living and working conditions. Haemosporidia. Boca Raton, FL: CRC Press; 2005. 17. Atkinson CT, van Riper C III. Pathogenicity and References epizootiology of avian haemoatozoa: Plasmodium, Leucocytozoon,andHaemoproteus. In: Loye JE, 1. Juvik JO, Juvik SP, Paradise TR, eds. Atlas of Zuk M, eds. Bird-Parasite Interactions, Ecology, Hawaii. 3rd ed. Honolulu, HI: University of Evolution and Behavior. New York, NY: Oxford Hawaii Press; 1998. University Press; 1991:19–48. 2. Freed LA, Conant S, Fleischer RC. Evolutionary 18. Atkinson CT. Avian malaria. In: Atkinson CT, ecology and radiation of Hawaiian passerine birds. Thomas NJ, Hunter DB, eds. Parasitic Diseases of Trends Ecol Evol. 1987;2:196–203. Wild Birds. Ames, IA: Blackwell Publishing; 3. James HF. The osteology and phylogeny of the 2008:35–53. Hawaiian finch radiation (Fringillidae: Drepani- 19. Tripathy D. Poxviridae. In: McFerran JB, dini), including extinct taxa. Zool J Linn Soc Lond. McNulty MS, eds. Virus Infections of Birds.New 2004;141:207–255. York, NY: Elsevier Science Publishers; 1993:1–15. 4. Casey TLC, Jacobi JD. A new genus and species of 20. van Riper C III, Forrester DJ. Avian pox. In: bird from the island of Maui, Hawaii (Passer- Thomas NJ, Hunter DB, Atkinson CT, eds. iformes: Drepanididae). Occas Pap Bernice P Bishop Infectious Diseases of Wild Birds. Ames, IA: Mus. 1974;24:215–226. Blackwell Publishing; 2007:131–176. 5. Scott JM, Conant S, van Riper C III. Introduction. 21. Forrester DJ. The ecology and epizootiology of In: Scott JM, Conant S, van Riper C III, eds. avian pox and malaria in wild turkeys. Bull Soc Evolution, Ecology, Conservation and Management Vector Ecol. 1991;16:127–148. of Hawaiian Birds: A Vanishing Avifauna. Studies in 22. Kleindorfer S, Dudaniec RY. Increasing preva- Avian Biology No. 22. 2001:1–12.. lence of avian poxvirus in Darwin’s finches and its J Avian Biol 6. US Fish and Wildlife Service. Revised Recovery effect on male pairing success. . 2006;37:69–76. Plan for Hawaiian Forest Birds. Portland, OR: US 23. Smits JE, Tella JL, Carrete M, et al. An epizootic Fish and Wildlife Service Region 1; 2006. of avian pox in endemic short-toed larks (Calan- 7. Warner RE. The role of introduced diseases in the drella refuscens) and Berthelot’s pipits (Anthus extinction of the endemic Hawaiian avifauna. berthelotti) in the Canary Islands, Spain. Vet Condor. 1968;70:101–120. Pathol. 2005;42:59–65. 8. van Riper C III, van Riper SG, Goff ML, Laird M. 24. Henshaw HW. Birds of the Hawaiian Islands; Being The epizootiology and ecological significance of a Complete List of the Birds of the Hawaiian malaria in Hawaiian land birds. Ecol Monogr. Possessions, with Notes on Their Habits. Honolulu, 1986;56:327–344. HI: Thos. G. Thrum; 1902. 9. van Riper C III, van Riper SG, Hanson WR. 25. Jarvi SI, Triglia D, Giannoulis A, et al. Diversity Epizootiology and effect of avian pox on Hawaiian and virulence of Avipoxvirus in Hawaiian forest forest birds. Auk. 2002;119:929–942. birds. Cons Genetics. 2008;9:339–348. 10. Dobson AP, May RM. Patterns of invasions by 26. Tripathy DN, Schnitzlein WM, Morris PJ, et al. pathogens and parasites. In: Mooney HE, Drake Characterization of poxviruses from forest birds in JA, eds. Ecology of Biological Invasions of North Hawaii. J Wildl Dis. 2000;36:225–230. American and Hawaii. New York, NY: Springer- 27. Kim T, Tripathy DN. Evaluation of pathogenicity Verlag; 1986:58–76. of avian poxvirus isolates from endangered Ha- 11. Work TM, Massey JG, Rideout BA, et al. Fatal waiian wild birds in chickens. Avian Dis. toxoplasmosis in free-ranging endangered àlala 2006;50:288–291. from Hawaii. J Wildl Dis. 2000;36:205–212. 28. Moulton MP, Miller KE, Tillman EA. Patterns of 12. Work TM, Massey JG, Lindsay DS, Dubey JP. success among introduced birds in the Hawaiian Toxoplasmosis in three species of native and Islands. In: Scott JM, Conant S, van Riper C III, introduced Hawaiian birds. J Parasitol. 2002;88: eds. Evolution, Ecology, Conservation and Manage- 1040–1042. ment of Hawaiian Birds: A Vanishing Avifauna. 13. Work TM, Ball D, Wolcott M. Erysipelas in a free- Studies in Avian Biology No. 22. 2001:31–46. ranging Hawaiian crow (Corvus hawaiiensis). Avian 29. Alicata JE. Parasitic Infections of Man and Animals Dis. 1999;43:338–341. in Hawaii. Honolulu, HI: Hawaii Agricultural 14. Work TM, Meteyer CU, Cole RA. Mortality in Experiment Station; 1964. Technical Bulletin 61. Laysan ducks (Anas laysanensis) by emaciation 30. van Riper SG, van Riper C III. A summary of complicated by Echinuria uncinata on Laysan known parasites and diseases recorded from the Island, Hawaii, 1993. J Wildl Dis. 2004;40:110–114. avifauna of the Hawaiian Islands. In: Stone CP, 62 JOURNAL OF AVIAN MEDICINE AND SURGERY

Scott JM, eds. Hawai[i’s Terrestrial Ecosystems: experimentally-infected juvenile apapane (Hima- Preservation and Management. Honolulu, HI: tione sanguinea). Auk. 2000;117:731–738. Cooperative National Park Resources Studies 44. VanderWerf EA, Burt MD, Rohrer JL, Mosher Unit, University of Hawaii; 1985:298–371. SM. Distribution and prevalence of mosquito- 31. Baldwin PH. Checklist of the birds of the Hawaii borne disease in Oàhu èlepaio. Condor. National Park, Kilauea-Mauna Loa Section, with 2006;108:770–777. remarks on their present status and a field key for 45. VanderWerf EA. Distribution and potential im- their identification. Hawaii National Park, HI, pacts of avian poxlike lesions in èlepaio at internal report; 1941. Hakalau Forest National Wildlife Refuge. In: 32. Fisher HI, Baldwin PH. Notes on the red-billed Scott JM, Conant S, van Riper C III, eds. leiothrix in Hawaii. Pacific Sci. 1947;1:45–51. Evolution, Ecology, Conservation and Management 33. Laird M, van Riper C III. Questionable reports of of Hawaiian Birds: A Vanishing Avifauna. Studies in Plasmodium from birds in Hawaii with the Avian Biology No. 22. 2001:247–253. recognition of P. relictum ssp. capistranoae (Rus- 46. VanderWerf EA, Cowell A, Rohrer JL. Distribu- sell, 1932) as the avian malaria parasite there. In: tion, abundance, and conservation of Oàhu Canning EU, ed. Parasitological Topics, A Presen- èlepaio in the southern leeward Koòlau range. tation Volume to P.C.C. Garnham, F.R.S. on the Èlepaio. 1997;57:99–105. Occasion of this 80th Birthday. Society of Protozo- 47. LaPointe DA. Avian malaria in Hawai[i: the ologists Special Publication No. 1; Allen Press, distribution, ecology and vector potential of forest- Lawrence, KS; 1981:159–165. dwelling mosquitoes [PhD thesis]. Honolulu, HI: 34. Beadell JS, Ishtiaq F, Covas R, et al. Global University of Hawaii, Ma¯noa; 2000. phylogeographic limits of Hawai[i’s avian malaria. 48. LaPointe DA, Goff ML, Atkinson CT. Compar- Proc Royal Soc, Series B: Biol Sci . 2006;273: ative susceptibility of introduced forest-dwelling 2935–2944. mosquitoes in Hawai[i to avian malaria, Plasmo- 35. Feldman RA, Freed LA, Cann RL. A PCR test for dium relictum. J Parasitol. 2005;91:843–849. avian malaria in Hawaiian birds. Mol Ecol. 49. Woodworth BL, Atkinson CT, LaPointe DA, et al. 1995;4:663–673. Host population persistence in the face of intro- 36. Atkinson CT, Lease JK, Dusek RJ, Samuel MD. duced vector-borne diseases: Hawaii amakihi and Prevalence of pox-like lesions and malaria in forest avian malaria. Proc Natl Acad Sci U S A. 2005; bird communities on leeward Mauna Loa Volcano, 102:1531–1536. Hawaii. Condor. 2005;107:537–546. 50. Ahumada JA, LaPointe DA, Samuel MD. Mod- 37. Aruch S, Atkinson CT, Savage AF, LaPointe DA. eling the population dynamics of Culex quinque- Prevalence and distribution of pox-like lesions, fasciatus (Dipteria: Culicidae) along an elevational avian malaria and mosquito vectors in Kipahulu gradient in Hawaii. J Med Entomol. 2004;41: Valley, Haleakala National Park, Hawaii, USA. J Wildl Dis. 2007;43:567–575. 1157–1170. 38. Scott JM, Mountainspring S, Ramsey FL, Kepler 51. Shehata CL, Freed L, Cann RL. Changes in native CB. Forest bird communities of the Hawaiian and introduced bird populations on Oàhu: infec- Islands: their dynamics, ecology, and conservation. tious diseases and species replacement. In: Scott Studies in Avian Biology No. 9. 1986. JM, Conant S, van Riper C III, eds. Evolution, 39. Goff ML, van Riper C III. Distribution of Ecology, Conservation and Management of Hawai- mosquitoes (Diptera: Culicidae) on the east flank ian Birds: A Vanishing Avifauna. Studies in Avian of Mauna Loa Volcano, Hawaii. Pacific Insects Biology No. 22. 2001:264–273. 1980;22:178–188. 52. Jarvi SI, Schultz JJ, Atkinson CT. PCR diagnostics 40. Atkinson CT, Woods KL, Dusek RJ, et al. Wildlife underestimate the prevalence of avian malaria disease and conservation in Hawaii: pathogenicity (Plasmodium relictum) in experimentally-infected of avian malaria (Plasmodium relictum) in experi- passerines. J Parasitol. 2002;88:153–158. mentally infected iiwi (Vestiaria coccinea). Parasi- 53. Spiegel CS, Hart PJ, Woodworth BL, et al. tology. 1995;111:S59–S69. Distribution and abundance of native forest birds 41. Atkinson CT, Dusek RJ, Woods KL, Iko WM. in low-elevation areas on Hawai[i Island: evidence Pathogenicity of avian malaria in experimentally of range expansion. Bird Conserv Int. 2006;16: infected Hawaii amakihi. J Wildl Dis. 2000;36: 175–185. 197–204. 54. Foster JT, Woodworth BL, Eggert LE, et al. 42. Atkinson CT, Lease JK, Drake BM, Shema NP. Genetic structure and evolved malaria resistance in Pathogenicity, serological responses, and diagnosis Hawaiian honeycreepers. Mol Ecol. 2007;16:4738– of experimental and natural malarial infections in 4746. native Hawaiian thrushes. Condor. 2001;103: 55. Benning TL, LaPointe D, Atkinson CT, Vitousek 209–218. PM. Interactions of climate change with biological 43. Yorinks N, Atkinson CT. Effects of malaria invasions and land use in the Hawaiian Islands: (Plasmodium relictum) on activity budgets of modeling the fate of endemic birds using a ATKINSON AND LAPOINTE—INTRODUCED AVIAN DISEASES IN HAWAII 63

geographic information system. Proc Natl Acad 66. Grim KC, McCutchan T, Li J, et al. Preliminary Sci U S A. 2002;99:14246–14249. results of an anticircumsporozoite DNA vaccine 56. Giambelluca TW, Luke MSA. Climate change in trial for protection against avian malaria in captive Hawai[i’s mountains. Mountain Views. 2007;1: African black-footed penguins (Spheniscus demer- 13–18. sus). J Zoo Wildl Med. 2004;35:154–161. 57. Giambelluca TW, Schroeder TA. Climate. In: 67. Knols BGJ, Bossin HC, Mukabana WR, Robinson Juvik SP, Juvik JO, Paradise TR, eds. Atlas of AS. Transgenic mosquitoes and the fight against Hawaii. 3rd ed. Honolulu, HI: University of malaria: managing technology push in a turbulent Hawaii Press; 1998:49–59. GMO world. Am J Trop Med Hyg. 2007;77(Suppl. 58. Cao G, Giambelluca TW, Stevens D, Schroeder T. 6):232–242. Inversion variability in the Hawaiian trade wind 68. Price JD, Jacobi JD, Pratt LW, et al. Protecting regime. JClim. 2007;20:1145–1160. forest bird populations at the landscape level. In: 59. Ralph CJ, Fancy SG. Demography and move- Pratt TK, Atkinson CT, Banko PC, et al, eds. ments of apapane and iiwi in Hawaii. Condor. Conservation Biology of Hawaiian Forest Birds: 1995;97:729–742. Implications for Island Avifauna. New Haven, CT: 60. Reiter ME, LaPointe DA. Landscape factors Yale University Press. In press. influencing the spatial distribution and abundance 69. Lieberman AA, Kuehler CM. Captive propagation of mosquito vector Culex quinquefasciatus (Dip- as a conservation strategy for the recovery of tera: Culicidae) in a mixed residential/agricultural endangered Hawaiian forest. In: Pratt TK, Atkin- community in Hawaii. J Med Entomol. 2007; son CT, Banko PC, et al, eds. Conservation Bio- 44:861–868. logy of Hawaiian Forest Birds: Implications for 61. LaPointe DA, Atkinson CT, Jarvi SI. Management Island Avifauna. New Haven, CT: Yale University of mosquito-borne disease in Hawaiian forest bird Press. In press. populations. In: Pratt TK, Atkinson CT, Banko 70. Groombridge JJ, Massey JG, Bruch JC, et al. An PC, et al, eds. Conservation Biology of Hawaiian attempt to recover the po`ouli by translocation Forest Birds: Implications for Island Avifauna.New and an appraisal of recovery strategy for bird Haven, CT: Yale University Press. In press. species of extreme rarity. Biol Conserv. 2004;118: 62. LaPointe DA. Dispersal of Culex quinquefasciatus 365–375. (Diptera: Culicidae) in a Hawaiian rainforest. 71. Hess SC, Swift CE, Campbell EW III, et al. J Med Entomol. 2008;45:600–609. History and development of techniques for con- 63. Three Mountain Alliance. http://hawaii.gov/dlnr/ trolling small mammals in Hawai[i. In: Pratt TK, dofaw/wpp/watershed-partnerships/tma. Accessed Atkinson CT, Banko PC, et al, eds. Conservation February 5, 2009. Biology of Hawaiian Forest Birds: Implications for 64. West Maui Mountains Watershed Partnership. http:// Island Avifauna. New Haven, CT: Yale University hawaii.gov/dlnr/dofaw/wpp/watershed-partnerships/ Press; 2009. wmmwp. Accessed February 5, 2009. 72. Leonard DL Jr. Recovery expenditures for birds 65. Massey JG, Graczyk TK, Cranfield MR. Charac- listed under the Endangered Species Act: the teristics of naturally acquired Plasmodium relictum disparity between mainland and Hawaiian taxa. capistranoae infections in native Hawaiian crows Biol Conserv. 2008;141:2054–2061. (Corvus hawaiiensis) in Hawaii. J Parasitol. 73. IUCN Red List. http://www.iucnredlist.org/. Ac- 1996;82:182–185. cessed February 5, 2009.