Ecology and Diagnosis of Introduced Avian Malaria in Hawaiian Forest Birds

Avian malaria is a disease caused by tible to this disease. Malaria currently species of protozoan parasites (Plasmo- limits the geographic distribution of na- dium) that infect birds. Related species tive species, has population level impacts commonly infect reptiles, birds and mam- on survivorship, and is limiting the recov- �

mals in tropical and temperate regions of ery of threatened and endangered species � ��

the world. Transmitted by mosquitoes, of forest birds. � � �

the parasites spend part of their lives in � � the red blood cells of birds (Figure 1). Altitudinal and Geographic � �� Avian malaria is common in continental Distribution of Avian Malaria areas, but is absent from the most isolated �� Several factors inﬂuence the prevalence of avian malaria across the Ha- � �� waiian archipelago. Hawai‘i has a wide spectrum of climatic zones and habitats �� that differ in rainfall, temperature and Figure 2. Our current understanding of the elevation. These habitats vary in their distribution of avian malaria on Mauna Loa and suitability for species that harbor and K¯ılauea Volcanoes after emergence of dis- transmit malaria. Mosquitoes, for ex- ease-resistant, low elevation Hawai‘i ‘amakihi ample, are more likely to be found in wet, populations. Disease prevalence (abundance) low elevation habitats with temporary or and transmission fall as elevation increases permanent bodies of standing water that both because of thermal limits on development provide habitat for larval development. of the parasite in the vector and because mos- Similarly, infection rates among birds in quitoes decline in numbers and become more the islands also differ due to differences seasonal at higher altitudes. Populations of in susceptibility and how well they over- native birds at high elevation are more diverse and�� include the species that have little or no lap with suitable mosquito habitat. One natural resistance to malaria. The absence of �

of the most important developments since � ��

native� birds at middle elevations is still incom- � the 1970s is the emergence of recovering �

pletely� understood and may reﬂect lower rates

� ��

lowland populations of Hawai‘i ‘amak- �

of� natural selection for disease resistance that ihi (Hemignathus virens) in the Puna occur�� in areas with seasonal malarial trans- District of Hawai‘i Island. These birds mission. A good test of this hypothesis will be

have changed our view of how malaria to�� see� if these areas are recolonized over the � next decade by disease resistant ‘amakihi from is transmitted across the larger landscape �� on the eastern slopes of Kˉılauea Volcano the� lowlands. Figure 1. Blood smear from an ‘apapane infect- �� ed with Plasmodium relictum. The parasites (red arrows) develop within the circulating red blood cells and stain purple. Red blood cell nuclei (black arrows) in parasitized cells are frequently pushed to one end of the cell.

island archipelagos where mosquitoes do �� not naturally occur. More than 40 different species of avian Plasmodium have been described, but only one, P. relictum, has been introduced to the Hawaiian �� Islands. Because they evolved without natural exposure to avian malaria, native Hawai- ian honeycreepers are extremely suscep- �� U.S. Department of the Interior USGS FS 2005-3151 U.S. Geological Survey Printed on recycled paper December 2005 �

� ��

��

� ��

(Figure 2). Their presence here and birds coupled with increasing mosquito damage) in native birds, but with low elsewhere in Hawai‘i indicates that some populations can lead to epidemic out- infectivity and pathogenicity to other native species have the capacity to evolve breaks that may continue to the onset non-native species that became estab- resistance to this disease. of colder winter temperatures in Janu- lished here. Alternatively, the source host Malaria is most common at middle ary (Figure 3). Transmission may occur may be established in the islands, but �� (900–1500 m) elevations on moist, wind- throughout the year at lower elevations if with a range that is more restricted than

ward sides of all of the main Hawaiian � Is- suitable reservoir hosts and susceptible, the current distribution of avian malaria.

� ��

lands where native species still occur, but� uninfected birds are present. We do not know the original host species, � �

has not been reported in the Northwestern� but one likely culprit is the House Spar-

� �� Hawaiian Islands. It is uncommon to � Reservoirs of Infection row (Passer domesticus) which maintains rare in alpine habitats on Mauna Loa, blood stage infections and infectiousness �� Perhaps surprisingly, the highest Mauna Kea, and Haleakalˉa Volcanoes to mosquito vectors for long periods of prevalence of malaria and the high- and is found almost entirely in birds that�� time with no evidence of clinical signs. have moved from lower elevations. Ma� - est intensity of infections now occur in Among native species, ‘apapane (Hima- highly�� susceptible� � native�� species�� (Figure �� laria appears to be abundant in lowland � tione sanguinea) and ‘amakihi (Hemi- 4). Malaria may have been introduced habitats in areas with recovering native �� gnathus spp.) generally have the highest bird populations, but is less common in in a songbird or game bird species to prevalence of infection in the wild. areas that are dominated by non-native which it was closely adapted. After initial spread to native forest birds, the source species. The disease can be found in very Mortality in Native Forest Birds dry habitats, but generally only in areas host may not have become established where mosquito populations are support- in the islands, leaving a parasite behind Experimental studies of the patho- ed by intermittent sources of water. that was highly pathogenic (able to inﬂict genicity of avian malaria have been instrumental in documenting relative �� susceptibility and mortality among native Seasonal Transmission of and non-native species and documenting �

� Avian Malaria ��

� evidence of evolving resistance in some �

� low elevation populations of Hawai‘i � Malaria transmission becomes increas- �

� ingly seasonal as elevation increases, ‘amakihi (Hemignathus virens). Among � �� both because numbers of mosquitoes �� species we have tested, both ‘i‘iwi are very low at higher elevations during (Vestiaria coccinea) and Maui ‘alauahio �� the cooler winter months and because of (Paroreomyza montana) are particularly thermal constraints on development of sensitive to malaria, with mortalities of the� parasite� �in� the mosquito�� vector.� �Ma�� - �� 90% and 75%, respectively, following ex- laria transmission at�� elevations�� between �� posure to single infective mosquito bites. 900 and 1500 m typically occurs during �� Both of these species rarely occur at the warmest time of the year between elevations below 1200 m, suggesting that Figure 4. Prevalence of avian malaria in native disease transmission may be a primary September and December when mos- and non-native species from mid-elevation factor limiting their current distribution. quito populations reach their peak. This habitats on the windward and leeward slopes period follows the nesting season for of Mauna Loa Volcano. A. K¯ılauea Iki Crater, By contrast, ‘ˉoma‘o (Myadestes obscu- most native species and the abundance Hawai‘i Volcanoes National Park. B. Kona Unit, rus) and low elevation populations of of recently ﬂedged, susceptible juvenile Hakalau Forest National Wildlife Refuge Hawai‘i ‘amakihi from the Puna District are more capable of surviving infection, with little or no mortality after exposure to single infective mosquito bites. �� Signs of Acute Malaria �

� ��

� The earliest detectable evidence � �

� �� of malaria infection in highly suscep- � � tible honeycreepers is the appearance �� of parasites in circulating blood cells

�� from 48 to 72 hours after exposure to an � infective mosquito bite. Parasites begin �� rapid multiplication in the blood cells, �� but the ﬁrst overt signs of infection do Figure 3. Percent prevalence of avian malaria in hatch-year ‘apapane at K¯ılauea Iki Crater in not begin until approximately seven days Hawai‘i Volcanoes National Park between 1992 and 1998. Epidemic outbreaks at this mid- eleva- post-infection when declines in food tion (1200 m) site are limited to the warmer months of the year between September and December. consumption and activity levels ﬁrst be- Prevalence varies from year to year for a variety of reasons including variations in rainfall, vector come evident. Among honeycreepers that populations, and numbers of susceptible juvenile birds. eventually recover from infection, peak �� levels of parasites in the blood normally

��

�� A B

C D

Figure 5. Characteristic signs and lesions of acute malaria. A. Red blood cell destruction and anemia is the primary cause of death in acute infections. After tubes of whole blood are spun in a centrifuge, volume of packed red blood cells is significantly lower in a wild ‘apapane with acute malaria (left) than in an uninfected canary. B. Lesions in an ‘i‘iwi with and have a prominent sternum or “keel”. Chronically infected birds may also carry acute malaria. Note enlarged, blackened liver Gross lesions are easy to recognize at genes for disease resistance and may be (arrow). C. Blood smear from an ‘i‘iwi with necropsy and include enlarged, chocolate good candidates for translocation or cap- acute malaria. Mature red blood cells have brown or black liver and spleen and thin, tive propagation for restoring forest bird been destroyed and replaced with immature watery blood (Figure 5). The most likely populations in locations where vector red blood cells and their precursors to com- cause of death is anemia associated with control is not feasible. Accurate identifi- pensate for anemia caused by cell destruction. destruction of red blood cells. cation of these individuals is important D. Hawai‘i ‘amakihi with acute malaria. At the when assessing disease risks for an area. crisis, birds become sedentary, cease feeding, and may be very susceptible to predation. Chronic Malaria and Problems with Disease Diagnosis Sensitivity and Specificity of Different Diagnostic Methods occur about 12 days post-infection when Honeycreepers that survive acute approximately 25% of the circulating red malaria develop chronic, low level infec- Both direct and indirect methods have blood cells are infected. The crisis passes tions that may persist for the lifetime of been developed for diagnosing malaria. as the immune system begins to control the bird. Unlike acute infections, numbers Direct methods that demonstrate the numbers of parasites in the circulation. of parasites in the peripheral circulation parasite or a portion of its DNA sequence During the peak period of parasitemia may be extremely low and difficult to de- include microscopy and polymerase chain or “crisis”, birds become sedentary and tect by microscopy. Available experimen- reaction (PCR) techniques. Microscopy may cease feeding altogether. Among tal evidence indicates that these birds are is still considered the “gold standard” for honeycreepers that succumb to infection, immune to re-infection. They may exhibit malaria diagnosis because parasites are up to 80% of the circulating red blood no outward signs of being infected and actually seen within the blood cells, but it cells may be parasitized. Most deaths may be in excellent body condition, yet can miss more than 70% of chronic infec- occur between 18 and 24 days after are still infectious to mosquitoes and are tions because numbers of parasites are so infection and birds are typically from excellent reservoir hosts for maintaining low. PCR methods that amplify specific 20 to 40% below normal body weight the disease in forest bird populations. portions of parasite DNA are significantly M 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 + 19 20

A B

Figure 6. Examples of different diagnostic tests for avian malaria. A. ELISA (Enzyme-Linked Immunosorbant Assay) consists of a 96 well plate coated with malarial antigen and incubated sequentially with test plasma, enzyme-labeled antibodies and substrate. A positive test is indicated by the development of yellow color. Samples are typically run in triplicate. B. PCR (Polymerase Chain Reaction). Primers specific to portions of malarial ribosomal DNA are used to amplify short segments of DNA. These are separated on an agarose gel and stained with a dye that is visible under ultraviolet light. A positive test is indicated by presence of a band of the correct size. Lanes 1–17 are from sequential blood samples collected from the same bird over a two year period. Lanes +, 19 and 20 are positive and negative controls. C. Western blotting. Malarial proteins are separated into bands of different molecular weight on a polyacryl- amide gel, transferred to a membrane and then incubated with test plasma, enzyme-labeled secondary antibodies and substrate. A positive test is indicated by the development of dark bands that correspond to malarial proteins of specific molecular weight. Each lane (1–5) was incubated with plasma from a different bird with known malarial infections. Lane C is a negative control. C more sensitive than microscopy, but are are detected and can be used to obtain Available experimental evidence indi- more expensive, time consuming and information on intensity of infection and cates that PCR is not quite as sensitive may require additional sequencing steps host cellular responses. Serological meth- as serological methods for detecting to confirm that the products originated ods are extremely sensitive for detecting chronic infections, but it is much more from parasite DNA. older infections in recovered birds be- sensitive than microscopy. When used in Indirect methods for demonstrating cause persistent infections stimulate anti- combination, these diagnostic methods infection with malaria include serological body production and cellular immunity to can complement each other and provide (blood) screening for the presence of anti- the parasite. They provide no information critical information to resource managers bodies to the parasite. Two methods have about parasite intensity or morphology, about the prevalence and distribution of been applied to diagnosing avian malaria however, and are also likely to miss very avian malaria in habitats that are being in Hawai‘i: Enzyme Linked Immunosor- early acute infections, before the host is considered for restoration of threatened bant Assay (ELISA) and Western Blotting able to produce antibodies to the parasite. and endangered forest birds. (Figure 6) . ELISA is more sensitive than Western Blotting, but may require additional tests by Western Blotting to verify Recommended Reading van Riper III, C., S.G. van Riper, M.L. - positive results. Western Blotting is the Goff and M. Laird. 1986. The epizo otiology and ecological significance most specific serological test available for Atkinson, C.T., K.L. Woods, R.J. Dusek, avian malaria and can be used to verify L.S. Sileo and W.M. Iko. 1995. of malaria in Hawaiian landbirds. both ELISA and PCR tests. Wildlife disease and conservation Ecological Monographs 56: 327–344. Microscopy, serology and PCR can all in Hawaii: pathogenicity of avian Woodworth, B.L., C.T. Atkinson, D.A. play important roles in providing accurate malaria (Plasmodium relictum) in LaPointe, P.J. Hart, C.S. Spiegel, E.J. diagnostic information about malarial experimentally infected Iiwi (Ves- Tweed, C. Henneman, J. LeBrun, infections in Hawaiian birds. When used tiaria coccinea). Parasitology 111: T. Denette, R. DeMots, K.L. Kozar, alone, the tests vary in their ability to S59–S69. D. Triglia, D. Lease, A. Gregor, T. accurately detect chronic malarial infec- Atkinson, C.T., R.J. Dusek and J.K. Smith, and D. Duffy. 2005. Host tions with low intensities. Microscopy is Lease. 2001. Serological responses population persistence in the face of most likely to miss chronic infections, but and immunity to superinfection with introduced vector-borne diseases: it is extremely accurate when parasites avian malaria in experimentally-in- Hawaii ‘amakihi and avian malaria. fected Hawaii Amakihi. Journal of Proceedings of the National Acad- Wildlife Diseases 37: 20–27. emy of Sciences, 102: 1531–1536. Jarvi, S.I., J.J. Schultz and C.T. Atkinson. 2002. PCR diagnostics underesti- For more information contact: mate the prevalence of avian malaria Carter T. Atkinson (Plasmodium relictum) in experi- Phone: 808-967-8119, ext. 271 mentally-infected passerines. Journal Email: [email protected] of Parasitology 88: 153–158. Photo credits: All photos were taken by USGS researchers.