What was done
Using data collected at Long Point Bird Observatory on the north shore of Lake Erie in Ontario, Canada over the period 1975-2000, the author investigated whether there had been changes in the timing of spring and autumn migrations of 13 species of birds that might be viewed as ecological responses to global warming.  Specifically, he analyzed for trends in the day of first arrival in the spring, and in the days of first, second and third quartile arrivals in the spring, as well as like departures in the autumn.

What was learned
All thirteen species exhibited earlier trends in the date of first arrival in the spring.  However, when subjected to stringent thresholds of statistical analysis, only one species was considered to exhibit a definite trend.  Spring quartile analyses were not nearly as universal.  Only nine of the thirteen species trended toward an earlier arrival date, while four trended toward a later arrival date; and of the nine that trended toward an earlier arrival date, only two were statistically significant, leaving the author to conclude that "most individuals of most species cannot conclusively be said to have migrated in 2000 earlier than their 1975 ancestors."  Furthermore, analysis of the quartile means revealed no significant phenological change by bird sex - male or female.  With respect to autumn trends, five species showed delayed migration, two showed advanced migration and six exhibited no trend.

What it means
In pondering the results of his analysis, Mills writes that "caution should be exercised in drawing broad conclusions about changes in migration phenology" with respect to global warming.  He notes, for example, that using first arrival dates to characterize migration systems "can be problematic because they are data from one tail of a distribution, they comprise a mostly male population and they may not correlate well with the balance of the migration period."  He also notes that "changes do not appear to be universal in spring, and change in that season is (a) considerably less than that suggested by [first arrival date] analysis and (b) both less complex and less common than that exhibited during autumn migration."  Consequently, global warming - if it is even occurring at all - seems not to be causing major alterations in bird migration dates on the north shore of Lake Erie.