Ludwig Antoinette, Bigras-Poulin Michel, Michel Pascal, Bélanger Denise
Facultéde Médecine Vétérinaire, Universitéde Montréal, 3200 Sicotte CP 5000, Saint Hyacinthe, QC J2S 7C6, Canada.
J Wildl Dis. 2010 Jan;46(1):195-208. doi: 10.7589/0090-3558-46.1.195.
Soon after the appearance of West Nile virus (WNV) in North America, a number of public health authorities designated the American Crow (Corvus brachyrhynchos) a sentinel for WNV detection. Although preliminary studies have suggested a positive association between American Crow mortality and increased risk of WNV infection in humans, we still know little about dynamic variation in American Crow mortality, both baseline levels and mortality associated with WNV. We hypothesized that the complex social behavior of American Crows, which is shaped by age and seasonal factors, influences both baseline mortality and WNV mortality in American Crow populations. We examined American Crow mortality data from Quebec for the 2005 WNV surveillance year, which lasted from 5 June to 17 September 2005. The variables of interest were age, gender, body condition index, time of year, and land cover. We used a log-linear model to examine baseline mortality. Logistic regression and general linear regression models were constructed to examine variables associated with mortality due to WNV. We found that both age and time of year were key variables in explaining baseline mortality. These two variables were also risk factors for WNV mortality. The probability that a carcass tested positive for WNV increased with the age of the dead bird and as summer progressed. WNV-positive carcasses also had a lower body condition index than WNV-negative carcasses. We believe that the first major wave of American Crow mortality observed in the early summer of 2005 was the result of natural mortality among young American Crows. Because this mortality was not linked to WNV, it appears that American Crow may not be a good species for early detection of WNV activity. Our data also suggest that second-year American Crows play a major role in propagating WNV during their movements to urban land covers during midsummer.
西尼罗河病毒（WNV）在北美出现后不久，一些公共卫生当局指定美洲乌鸦（短嘴鸦）作为WNV检测的哨兵。尽管初步研究表明美洲乌鸦死亡率与人类感染WNV风险增加之间存在正相关，但我们对美洲乌鸦死亡率的动态变化，包括基线水平和与WNV相关的死亡率，仍然知之甚少。我们假设，受年龄和季节因素影响的美洲乌鸦复杂社会行为，会影响美洲乌鸦种群的基线死亡率和WNV死亡率。我们研究了2005年WNV监测年度（从2005年6月5日持续到9月17日）来自魁北克的美洲乌鸦死亡率数据。感兴趣的变量包括年龄、性别、身体状况指数、年份时间和土地覆盖类型。我们使用对数线性模型来研究基线死亡率。构建逻辑回归和一般线性回归模型来研究与WNV导致的死亡率相关的变量。我们发现年龄和年份时间都是解释基线死亡率的关键变量。这两个变量也是WNV死亡率的风险因素。尸体检测出WNV呈阳性的概率随着死鸟年龄的增加以及夏季的推进而增加。WNV呈阳性的尸体的身体状况指数也低于WNV呈阴性的尸体。我们认为，2005年初夏观察到的第一波主要的美洲乌鸦死亡是幼年美洲乌鸦自然死亡的结果。由于这种死亡与WNV无关，看来美洲乌鸦可能不是早期检测WNV活动的良好物种。我们的数据还表明，第二年的美洲乌鸦在仲夏向城市土地覆盖区域迁徙期间传播WNV方面发挥了主要作用。
