* If individuals were ringed during January or February, they could potentially be long-term winter residents.
All categories could include an unknown number of individuals that may not have migrated beyond our study site but could have settled close by and gone undetected. Note that these categories do not differentiate between departures and mortality.
Table 2. General Linear Model results of site persistence and distance shifted between years predictors. In models, age = adult, sex = female, year = 1, group = A, residency = long-term were the base categories. Significant p values are highlighted in bold and italics. All interactions were NS.