Appendix
Appendix 1: Proving different residency
categories
To prove that individuals remained for different periods throughout the
non-breeding season at our study site – that individuals have different
within-winter residency strategies – we compared the observed
frequencies of the number of visits individuals were detected each year,
with that expected by chance assuming that all individuals were
long-term winter residents. To do this, we first calculated the number
of individuals seen per year and the respective mean number of
visits. We ran 50 iterations to obtain a representative mean of birds
detected each visit, assuming that (1) birds were present in the area
throughout the study, i .e . long-term winter residents, and
(2) that the detection rate per visit was 0.33 (see manuscript).
We then compared these estimates with our observed data using a
two-sample t -test. We expect that if our observed frequencies
match those expected by chance, then all individuals are long-term
winter residents. If frequencies do not match, however, we assume that
individuals have different duration residency periods. All years were
analysed separately. Individuals seen in multiple years were not
excluded from any analyses.
We repeated this same analysis using data from individuals that were
observed at least twice throughout the year to eliminate individuals
that were likely to be simply passage birds. By doing this, we eliminate
individuals that may have been passing by and detected by chance, and
not necessarily utilising resources from the area. If afterwards we
still observe differences in expected and observed frequencies, then we
expect that not all individuals detected at our study site are long-term
winter residents.
We found that the expected frequencies of the number of visits that
individuals were predicted to be detected at was statistically different
from what was observed when analysing both the data set with information
from all individuals and the data set with individuals that were
detected at least twice during the year (Fig. A.1; Table A.1). These
results are similar across years. With this, we confirm that not all
birds seen at our study sites are long-term winter residents.