Predicted and observed mortality from vector-borne disease in wildlife: West Nile virus and small songbirds
Introduction
The impact of disease on wild animal populations has been notoriously difficult to detect and demonstrate, due to problems of attribution and the rapid disappearance of bodies after death (McCallum, 2005, McCallum and Dobson, 1995). The clearest examples of disease-caused impacts on wildlife populations come from epidemics in large abundant animals such as anthrax and Rinderpest in African mammals (Holdo et al., 2009), experimental or purposeful viral introductions such as myxomatosis and Australian rabbits (Ratcliffe et al., 1952), and experimental studies that remove pathogens from hosts through treatment (Hudson et al., 1998). For many other diseases and populations, impacts are inferred from long term monitoring and observations of sudden declines, and in rare cases scientists have been able to observe a wave of mortality as a pathogen arrives (Hochachka and Dhondt, 2000, Kilpatrick et al., 2010, Langwig et al., 2012, Lips et al., 2006, Vredenburg et al., 2010). However, in many cases mortality due to disease is difficult to detect and even striking patterns, such as distributional limits coincident with disease boundaries, required experimental infection studies to confirm impacts of disease (e.g., avian malaria and Hawaiian birds; (Van Riper et al., 1986, Warner, 1968)).
A recent introduction of a pathogen to North America, West Nile virus (WNV; Flaviviridae; Flavivirus) in 1999, was also accompanied by waves of mortality in wild birds, with large numbers of dead American crows and Blue jays testing positive for WNV in the northeast USA (Bernard et al., 2001, Nemeth et al., 2007). A decade later, transmission still occurs annually in many bird communities throughout North and South America (Kilpatrick, 2011). Several retrospective analyses have shown population declines in birds coincident with the arrival of WNV as it spread south and west from New York, with impacts being largest on corvids (Hochachka et al., 2004, LaDeau et al., 2007, Wheeler et al., 2009). Evidence of WNV-caused mortality in corvids was also provided by experimental infection in laboratory studies (Komar et al., 2003, Reisen et al., 2005). However, evidence of WNV mortality in smaller passerines has been far sparser, with relatively few WNV-infected dead birds collected. The extent to which this is due to poor detectability (Ward et al., 2006) or lack of mortality is not clear.
Two families of small passerines that past studies have suggested may suffer population level impacts from WNV are Paridae (chickadees and titmouse) and Troglodytidae (wrens). Multiple studies have observed declines in one or more species in the family Paridae and Troglodytidae coincident with the arrival of WNV (Bonter and Hochachka, 2003, LaDeau et al., 2007), and several other studies have demonstrated feeding on parids and wrens by WNV mosquito vectors (Hamer et al., 2009, Hassan et al., 2003, Kilpatrick et al., 2006a). However, these data are only suggestive and supportive evidence in the form of WNV-infected dead chickadees, titmouse or wrens is mostly lacking.
The gold standard to determine whether a species suffers mortality from a pathogen, part of Koch’s postulates (Koch, 1893), is through experimental infection. There are far too many species of birds in North America to do this for all taxa, and these studies cannot determine whether in fact birds are exposed to a pathogen in nature. For effective conservation planning, there is clearly a need for a framework to determine whether WNV and other vector-borne diseases cause mortality in small avian hosts, and other small wildlife species that are difficult to detect.
Here we describe how one can use field data on the transmission ecology of a vector-borne disease – specifically the feeding patterns of WNV mosquito vectors, avian abundance, and the WNV antibody prevalence of wild-caught birds – to generate hypotheses about differences in mortality from WNV infection between hosts. We illustrate this method with a study on three species of small songbirds, tufted titmouse (Baeolophus bicolor), Carolina wrens (Thryothorus ludovicianus), and northern cardinals (Cardinalis cardinalis). We generated and tested hypotheses about the relative mortality of three species and measured morbidity and mortality following experimental infection with WNV. Our experimental infection studies also provide data on infectiousness for WNV that can be integrated with the aforementioned data on mosquito preferences to determine the role of species in WNV transmission (Hamer et al., 2009, Kilpatrick, 2011, Kilpatrick et al., 2006a).
Section snippets
Framework for predicting relative host mortality from a vector-borne pathogen
This framework generates a prediction about the relative mortality from infection with a vector-borne pathogen between two or more species and can be applied to any vector-borne pathogen and host taxa with the data described.
The seroprevalence, S, or fraction of a population with antibodies against a pathogen at a point in time is equal to the fraction of the population exposed, e, multiplied by the probability of survival (1 − m, where m is the probability of mortality given infection), divided
Results
Tufted titmouse were present at four of seven sites where they made up 2.8% (±1 SE 2.9%) of the avian community, and we identified Culex pipens or Cx. restuans bloodmeals from them at three sites (Table 1). Carolina wrens were present at all seven sites where they made up 2.7% (±1%) of the avian community and we identified bloodmeals from wrens at 5 of 7 sites. Carolina wrens were fed on by mosquitoes significantly more than expected given their availability at two sites, less than expected at
Discussion
Analysis of population trends following the arrival of WNV suggested that tufted titmouse, chickadees, and house wrens were significantly impacted by disease, with the largest drop in mid-Atlantic populations following the intense 2003 WNV epidemic (LaDeau et al., 2007). However, the inference from that study and others (Bonter and Hochachka, 2003, Wheeler et al., 2009) that trends in these species, as well as several other small songbirds, were due to WNV was indirect. Similarly, although
Acknowledgements
We thank Will Janousek for assistance capturing and transporting birds, the Smithsonian’s Neighborhood Nestwatch program and residents, the staff of Rock Creek Park (Meadowside Nature Center), Fort Dupont Park, the Smithsonian National Museum of Natural History, the National Gallery of Art and the Hirshhorn museum for permission to use their property. We are indebted to Montgomery County Parks and the Patuxent National Wildlife Refuge for permission to use birds from these lands for this study.
References (51)
- et al.
Phylogenetic analysis of North American West Nile virus isolates, 2001–2004: Evidence for the emergence of a dominant genotype
Virology
(2005) - et al.
Population dynamics of mosquito-borne disease – effects of flies which bite some people more frequently than others
Trans. Roy. Soc. Trop. Med. Hyg.
(1986) - et al.
The ecology and impact of chytridiomycosis, an emerging disease of amphibians
Trends Ecol. Evol.
(2010) - et al.
Drivers, dynamics, and control of emerging vector-borne zoonotic diseases
Lancet
(2012) - et al.
Detecting disease and parasite threats to endangered species and ecosystems
Trends Ecol. Evol.
(1995) - et al.
Quantitation of flaviviruses by fluorescent focus assay
J. Virol. Methods
(2006) Observation study of behavior – sampling methods
Behaviour
(1974)- et al.
Host-feeding habits of Culex and other mosquitoes (Diptera: Culicidae) in the Borough of Queens in New York city, with characters and techniques for identification of Culex mosquitoes
J. Med. Entomol.
(2002) - et al.
Host feeding patterns of established and potential mosquito vectors of West Nile virus in the eastern United States
Vector-Borne Zoo. Dis.
(2004) - et al.
West Nile virus infection in birds and mosquitoes, New York state, 2000
Emerg. Infect. Dis.
(2001)