West Nile virus (WNV), a mosquito-borne flavivirus, was first detected in eastern North America during the summer of 1999 (Asnis et al. 1999, Nash et al. 2001) and spread rapidly across the continent, arriving in California in 2003 (Reisen et al. 2004). As WNV spread, it caused significant morbidity and mortality in naïve native birds, particularly of the families Accipitridae, Strigidae, and Corvidae (Komar 2003, Marra et al. 2004). While free-ranging individuals of some of these species respond to WNV with antibodies (Stout et al. 2005, Hull et al. 2006, Crosbie et al. 2008), WNV infection appears to cause near 100% mortality in captive individuals of several species (McLean et al. 2001, Komar et al. 2003, Gancz et al. 2004). In California, birds' susceptibility to WNV varies by species, with several population declines associated with WNV infection, notably in the family Corvidae (Wheeler et al. 2009, Smallwood and Nakamoto 2009).

In the Sierra Nevada region of California, populations of the Northern Goshawk (Accipiter gentilis), Spotted Owl (Strix occidentalis), and Great Gray Owl (S. nebulosa) are of conservation concern (Winter 1980, Seamans et al. 2001, Boyce et al. 2006), and WNV may pose a significant threat to these populations. Among Northern Goshawks naturally infected with WNV, significant lesions of the heart and central nervous system have been documented on post-mortem examination and histopathology (Wünschmann et al. 2005). Among Great Gray Owls, WNV may cause significant hepatic and splenic necrosis, and infection frequently results in sudden death (Gancz et al. 2004, Lopes et al. 2007).

Our primary objective in this study was to test for immunological exposure to WNV in wild populations of the Northern Goshawk, Spotted Owl, and Great Gray Owl in the Sierra Nevada of California. The samples we collected from Spotted Owls were associated with long-term research to investigate and monitor demographic trends, providing an opportunity for future research into the relationship between WNV prevalence and population trends. The samples from Northern Goshawks and Great Gray Owls were collected as part of short-term projects from 2004 to 2007. Our secondary objective was to compare serology results with WNV mortalities in Sierra Nevada birds documented as part of WNV surveillance.

METHODS

SAMPLE COLLECTION

For testing for antibodies to WNV, we sampled plasma from 62 Northern Goshawks, 209 Spotted Owls, and 22 Great Gray Owls. We captured birds with mist nets, noose poles, and pan traps (Bub 1991) at three study sites in the northern and central Sierra Nevada during the breeding seasons of 2004, 2005, 2006, and 2007 (Table 1). We banded all individuals with U.S. Geological Survey leg bands and released them at the site of capture. We drew approximately 0.5 mL of whole blood via venipuncture of the medial metatarsal vein and separated plasma from whole blood by centrifugation. We stored plasma samples at -20°C before submitting them for testing.

We sampled blood from Northern Goshawks and Spotted Owls at elevations from 800 to 2150 m in the northern Sierra Nevada in the Plumas and Lassen national forests, Plumas and Lassen counties. We sampled Great Gray Owls at elevations from 700 to 2600 m in Yosemite National Park and the Stanislaus National Forest in the central Sierra Nevada, in Calaveras, Tuolumne, and Mariposa counties. In our third and southernmost study area, at elevations 300–3100 m in the Sierra National Forest and Sequoia and King's Canyon national parks, Fresno and Tulare counties, we sampled only Spotted Owls. Vegetation in all three areas is broadly similar, consisting of forest and woodland vegetation types characteristic of the western Sierra Nevada. Lower-elevation habitats consist of oak woodland dominated by Blue Oak (Quercus douglasii), Interior Live Oak (Q. wislizenii), Canyon Live Oak (Q. chrysolepis), and Gray Pine (Pinus sabiniana) ranging into pine—oak woodlands dominated by Ponderosa Pine (Pinus ponderosa) and Black Oak (Quercus kelloggi). Middle elevations are covered with Sierran montane mixed-conifer forest consisting of Ponderosa Pine, Black Oak, White Fir (Abies concolor), Sugar Pine (P. lambertiana), Incense Cedar (Calocedrus decurrens), and Douglas Fir (Pseudostuga menziesii), which grades into upper montane forest consisting of Red Fir (Abies magnifica), Lodgepole Pine (P. conforta), Jeffrey Pine (P. jeffreyi), and Western White Pine (P. monticola). Wet and dry meadows, annual and perennial grasslands, chaparral, and riparian vegetation are interspersed throughout these forests and woodlands (Kuchler 1977).

TABLE 1.

Numbers of plasma samples tested by year from the Northern Goshawk, Spotted Owl, and Great Gray Owl in the Sierra Nevada, California.

RESULTS

We found no antibodies to WNV among Northern Goshawks, Spotted Owls, or Great Gray Owls over our 4-yr investigation. The California Department of Public Health had records of 89 WNV-positive dead birds of 26 species at elevations >900 m in 10 Sierra Nevada counties reported during the span of our study (Table 2).

DISCUSSION

While WNV infection in Sierra Nevada avifauna was documented during the study period, we found no antibody response in the populations of the Northern Goshawk, Spotted Owl, or Great Gray Owl we studied. These results were somewhat unexpected, as the California Department of Public Health's reports of dead birds testing positive for WNV RNA included a single Northern Goshawk, one Swainson's Hawk (Buteo swainsoni), and four Red-tailed Hawks (B. jamaicensis). WNV RNA in these and 23 other avian species suggests that WNV was present in the Sierra Nevada region during our study.

TABLE 2.

Numbers of dead birds (26 species) that tested positive for West Nile virus (WNV) RNA during the summers of 2003–2007. The presence of viral RNA specific to WNV was tested with a real-time assay using a reverse-transcription polymerase chain reaction. Data are from the California Department of Public Health Dead Bird Surveillance Program.

The primary route of WNV infection is through the bite of infected mosquitoes, particularly of the genus Culex (Bernard and Kramer 2001, Kilpatrick et al. 2006). In California, experimental and field investigations suggest that C. tarsalis, C. stigmatosoma, C. erythrothorax, and C. pipiens are the most efficient vectors (Goddard et al. 2002, 2003). The range of these, and other, mosquito species extends into the Sierra Nevada and overlaps with both the dead birds collected by California Department of Public Health and the antibody-negative populations of raptors we examined (Bohart and Washino 1978). In spite of overlap with vector-competent mosquitoes, we did not detect WNV antibodies in any individuals examined.

Previous research suggests that a secondary route of WNV infection is through consumption of infected prey (Komar et al. 2003) and possibly feces (Kipp et al. 2006), making the absence of an indication of WNV infection in the Northern Goshawk particularly notable. In the Sierra Nevada, Northern Goshawks prey on both small mammals and medium-sized birds; Steller's Jays (Cyanocitta stelleri) and American Robins (Turdus migratorius) are common prey, and both of thse species were commonly found dead and tested positive for WNV RNA. This diet contrasts with that of the Spotted Owl and Great Gray Owl, which feed primarily on mammalian prey (Bull and Duncan 1993, Gutiérrez et al. 1995).

Our not detecting WNV antibodies may indicate that the populations we studied had little or no exposure or that the rate of infection was so low as to be undetected by our sampling scheme. Limited sampling may miss rare infections; however we sampled between 10% and 60% of the population of the three species within our study areas, indicating an absence of WNV infection. Using the same methods, and sampling a much smaller proportion of the total population, Hull et al. (2006) found that 5%–20% of migrating and 15%–58% of wintering Cooper's Hawks (Accipiter cooperii), Red-shouldered Hawks (B. lineatus), and Red-tailed Hawks from the central coast and Central Valley of California tested positive for WNV antibodies during 2004 and 2005 (Hull et al. 2006).

Alternatively, we may have failed to detect antibodies because of extremely high rates of WNV mortality. If mortality rates due to WNV infection are high, few if any individuals would be expected to survive long enough to mount a detectable immune response. Such high rates of WNV-induced mortality could present a serious threat to the persistence of these populations. In C. pipiens and C. tarsalis, however, ambient temperature has been associated with WNV amplification, suggesting the relatively cool temperatures in the Sierra Nevada, as compared with the Central Valley of California, may slow WNV transmission and make epidemic disease outbreaks less likely (Dohn et al. 2002, Reisen et al. 2006). While we found 89 records of birds positive for WNV RNA in the California Department of Public Health data from 2004 to 2007, 9120 birds positive for WNV RNA were reported statewide. The majority of these were reported from low elevations in the Central Valley and southern California. This pattern may indicate a lower incidence of WNV in the Sierra Nevada due to lower ambient temperature, a reporting bias toward populated areas, or a combination of the two factors.

Whether absence of infection or extremely high mortality is the underlying cause of the nonappearance of WNV antibodies in our study cannot be determined with the current WNV antibody data and reinforces the need for additional research in conjunction with monitoring of seroprevalence (Walker et al. 2007, Wheeler et al. 2009). Long-term demographic data from our study area are available for the Spotted Owl and can be examined for evidence of declines in apparent annual survival rates that coincide with the spread of WNV into the Sierra Nevada. There is no current detailed demographic monitoring of the Northern Goshawk or Great Gray Owl with which survival rates could be estimated. During the course of our study, each species' site occupancy remained high, suggesting that if the rate of mortality from WNV is indeed high, the proportion of the population infected annually may be small. Since data on survival rates are lacking, these conclusions should be tempered. In territorial species, continuous occupancy of a site can mask declines in survival if recruitment of floaters is sufficient to replace loss of the territorial individuals, particularly in the short term.

To determine actual rates of WNV infection in these species, future research should focus on estimating survival rates, identifying causes of mortality to determine if WNV infection is resulting in mortality, and whether this potential added source of mortality is a significant threat to a population. Though requiring a consistent annual commitment to logistics and funding, long-term demographic studies provide a framework of baseline information for assessing associations among ecological factors and environmental stressors on the demographic and population trends of focal species of conservation concern. We recommend that demographic studies of the Spotted Owl be continued and that carefully designed demographic studies of the Northern Goshawk and Great Gray Owl be initiated. Long-term demographic studies of these focal species will be of increasing future value for assessing the effects of invasive species (e.g., Barred Owl, Strix varia) and exotic infectious diseases (e.g., WNV), as these types of stressors are likely to increase in the future.

We thank the U.S. Forest Service, Yosemite National Park, U.S. Bureau of Land Management Medford District, California Department of Fish and Game, Lindsay Wildlife Hospital, the University of California, Davis, Raptor Center, the University of California, Davis, Veterinary Teaching Hospital, N. Anderson, Y. Fang, W. Farrier, C. Gallagher, J. Dunk, S. Godwin, D. Hansen, J. Hawley, T. Hull, G. Jehle, E. Jepsen, R. Kussow, J. Maurer, J. Medley, T. Munton, T. Narahashi, B. Ogren, W. Reisen, P. Shaklee, J. Shafer, D. Shaw, B. Stedman, G. Steger, C. Stermer, S. Stock, A. Stutz, L. Tierney, S. Thompson, S. Vigallon, and M. Werner for assistance with sample collection and logistical support. Financial support was provided by the U.S. Department of Agriculture Forest Service, Pacific Southwest Research Station, Sierra Nevada Research Center, Region 5 of the Forest Service, Yosemite Fund, the U.S. Bureau of Land Management, the University of California Genetic Resources Conservation Program, the University of California, Davis, Graduate Group in Ecology, and the Veterinary Genetics Laboratory at the University of California, Davis.