Comparing timing and routes of migration based on ring encounters and randomization methods

Abstract

A method for comparing two migration routes is introduced. In the method are needed approximations of the average daily positions which are computed based on averages of dates and of positions in a window sliding through the encounters ordered in ascending order by date. The method contains two tests. The first, global, test statistic compares the entire migration routes and is the average of the distances between the daily positions of the two routes to be compared. The second test is used to identify sections during migration where the routes may deviate and is based on consecutive averages of the distances of short time periods between the daily positions of the two routes. A randomization test is used to assess the statistical significance of the test statistics in both components. The methods are applied to artificial and real data. Examples of the use of the method are computed with data sets of ring recoveries of Common terns (Sterna hirundo) and Ospreys (Pandion haliaetus) ringed in Finland in 1930-2002.