Avian disease and mosquito vectors in the Kahuku unit of Hawai`i Volcanoes National Park and Ka`u Forest Reserve

Abstract

While avian disease has been well-studied in windward forests of Hawai&lsquo;i Island, there have been few studies in leeward Ka&lsquo;u. We surveyed four altitudinal sites ranging from 1,200 to 2,200 m asl in the Kahuku Unit of Hawai&lsquo;i Volcanoes National Park (Kahuku) and three altitudinal sites ranging from 1,200 to 1,500 m asl in the Ka&lsquo;u Forest Reserve (Ka&lsquo;u) for the prevalence of avian disease and presence of mosquitoes. We collected blood samples from native and non-native forest birds and screened for avian malaria (Plasmodium relictum) using PCR diagnostics. We examined birds for signs of avian pox (Avipoxvirus sp.), knemidokoptic mange (Knemidokoptes jamaicensis) and feather ectoparasites. We also trapped adult mosquitoes (Culex quinquefasciatus and Aedes japonicus japonicus) and surveyed for available larval habitat. Between September, 2012 and October, 2014, we completed 3,219 hours of mist-netting in Kahuku capturing 515 forest birds and 3,103 hours of mist-netting in Ka&lsquo;u capturing 270 forest birds. We screened 750 blood samples for avian malaria. Prevalence of avian malaria in all species was higher in Ka&lsquo;u than Kahuku when all sites were combined for each tract. Prevalence of avian malaria in resident Hawai&lsquo;i &lsquo;amakihi (Chlorodrepanis virens) was greatest at the lowest elevation sites in Kahuku (26%; 1,201 m asl) and Ka&lsquo;u (42%; 1,178 m asl) and in general, prevalence decreased with increasing elevation and geographically from east to west. Significantly higher prevalence was seen in Ka&lsquo;u at comparable low and mid elevation sites but not at comparable high elevation sites. The overall presumptive pox prevalence was 1.7% (13/785) for both tracts, and it was higher in native birds than non-native birds, but it was not significant. Presumptive knemidokoptic mange was detected at two sites in lower elevation Kahuku, with prevalence ranging from 2‒4%. The overall prevalence of ectoparasites (Analges and Proctophyllodes spp.) was 6.7% (53/785). The site with the highest prevalence was Lower Glover in Kahuku (7.2%; 10/138) and Maka&lsquo;alia in Ka&lsquo;u. In general, mosquito larval habitat was more prevalent at lower elevation sites than higher elevation sites within the Kahuku&mdash;Ka&lsquo;u landscape, and more prevalent in Ka&lsquo;u than Kahuku. We observed significantly more available larval mosquito habitat in total belt transect plots in Ka&lsquo;u than Kahuku for both hapu&lsquo;u cavities (&Chi;2 = 47.06, df = 1, p < 0.01) and other habitat types combined (i.e., ground pools, rock holes, tree holes) (&Chi;2 = 104.35, df = 1, p < 0.01). Mosquitoes were most abundant at low elevation Kahuku, but were captured at all sites up to 1,532 m asl in Kahuku. The malarial infection rate of live mosquitoes was 21% (39/186) at Kahuku and 25% (2/8) at Ka&lsquo;u. There were 19 times more larval habitats available in Ka&lsquo;u than Kahuku on survey transects, yet we captured 53 times more C. quinquefasciatus mosquitoes in Kahuku. We captured very few adult A. j. japonicus across the landscape (Ntotal = 6) and no Aedes albopictus were detected in this study. Larval surveys along ranch roads and infrastructure revealed that ground pools along rutted, overgrown ranch roads were the likely source of Kahuku mosquitoes. We did not find mosquito larvae associated with ranching infrastructure. Unlike the low elevation forests on windward Hawai&lsquo;i Island, avian malaria prevalence, mosquito abundance, and the density of available larval habitat in Kahuku and Ka&lsquo;u were relatively low. Although altitudinal variations in climate appear to be the primary factors limiting the distribution of avian disease, habitat type, avian movements, human activity, and feral pig (Sus scrofa) management all may play important roles in determining the prevalence of avian malaria across the Kahuku&mdash;Ka&lsquo;u landscape.