Misleading correlations: The case of the Canada warbler and spruce budworm

In a recently published article Sleep et al. (2009) suggested that 30-yr declines in Canada warbler (19752005) based on Breeding Bird Survey (BBS) data could be attributed to 30-yr declines in spruce budworm through a series of analyses based principally on correlations. We demonstrate that the relationship does not hold when a longer-term data set (19682008) is used for the analysis. Sleep et al. (2009) also demonstrated a positive relationship between correlations of Canada warbler abundance and time and spruce budworm defoliation and time using provincial data sets. We examined the underlying BBS data and found that there were insufficient observations of Canada warbler in the western provinces (Alberta and Manitoba) to support the conclusion because most Canada warbler observations occurred far (>100?km) from any budworm defoliation. As well, we used a density-dependent stochastic population growth model as proposed by Sleep et al. (2009) and found only 2 significant relationships (Nova Scotia, Alberta) between Canada warbler population growth rate and budworm defoliation at the provincial scale. We conclude that little analytical evidence exists to support the idea that Canada warbler decline is a function of spruce budworm decline. (C) 2011 The Wildlife Society.