Results

Detection probabilities

The mean of all seven detection probabilities (manual detections from years 1, 2 and 3, MARK detections from years 1, 2 and 3, and detection from radio-tagged birds in year2) was 0.33 (se = 0.02); the probability of detecting a Whitethroat at our study site when it is present was once every three visits (Fig. 1). Detection probabilities were similar between years when undertaken manually (mean = 0.36, se = 0.02,F (2,48) = 0.13, p = .88) and in MARK (mean = 0.29, se = 0.03, F (2,48) = 1.48, p = .24) and were similar across methods during all three years (year1:t (38)= 0.88, p = .38; year2:F (2,32) = 2.44, p = .10; year3:t (28)= 1.18, p = .25).

Site persistence

Site persistence, defined as the number of days an individual was present and detected in the area, varied widely across individuals, ranging from one day to 165 days (mean = 31 days, se = 3 days, n= 341) but did not seem to differ significantly between years and between adult female and male birds (Table 2). First-years, however, remained for significantly shorter periods when compared to adults (Table 2).

Between-years site fidelity

Return rates

Overall return rates were similar across years but varied between age groups dependent on year and residency category, with more long-term and short-term winter residents returning than passage birds. A similar proportion of individuals returned between years (χ2 = 0.56, df = 1, p = .45): 36/182 (20%) individuals returned from year1 to year2 (group A), and 24/145 (17%) individuals returned from year2 to year 3 (group B). Seven individuals from year1 failed to return in year2 but then returned in year3 (group C). Only 12 individuals were seen during all three fieldwork seasons. In group A, a similar proportion of individuals of adults and first-year birds returned the following year: 13/62 (21%) adults and 22/96 (23%) first-years (χ2 = 0.08, df = 1, p = .77). In group B, however, there were clear differences between individuals of different ages: 20/90 (22%) adults and 3/50 (6%) first-years returned (χ2 = 6.16, df = 1, p = .01). Most individuals from group C were first-year birds in year1. Female and male adults had similar return rates in group A (females = 5/21, 24%, males = 6/28, 21%; χ2 = 0.04, df = 1, p = .84) and in group B (females = 9/35, 26%, males = 8/42, 19%; χ2 = 0.49, df = 1, p = .48). When comparing return rates amongst residency categories in group A, long-term winter residents (14/43, 33%) and short-term residents (2/7, 29%) had higher return rates than passage birds (10/90, 11%) (χ2 = 9.34, df = 2,p = .009). A similar trend was seen in group B (χ2 = 6.98, df = 2, p = .03); 12/31, 39% of long-term winter residents returned; 3/16, 19% short-term residents; 5/40, 13% passage birds.
The distance moved from one year to another varied among individuals (Fig. 2) but, on average, individuals moved less than 300 meters (Fig. 2; Appendix 2). This figure was similar amongst groups A, B, and C (F(2,51) = 0.006, p = .99).
The distance shifted between years did not vary significantly according to previous age (F(1,45) = 2.1, p = .16), sex (F(1,33) = 0.58, p = .45) or previous residency (F(2,47) = 1.61, p = .21; Fig. 3; Appendix 2). Results from the averaging model, however, show that first-years in group A (seen form year1 to year3) shifted longer distances than adults (Table 2). All other variables were NS (Table 2).

Residency repeatability

The degree of residency category repeatability, i .e . whether individuals remained in the same residency category through different years, varied across individuals (Fig. 4). 68% of long-term winter residents remained as such the following year, and 32% remained for similar or shorter periods. Most of the short-term winter residents (66%), when they returned the following year, were categorised as passage birds, 17% remained for similar periods and 17% remained for longer. Half of the passage birds remained as such the following year, while the other half remained for longer periods: 31% were categorised as long-term winter residents and 19% as short-term winter residents (Fig. 4).
When comparing the duration (in days) spent at the site of individuals from one year to another, we found that there was a significant somewhat positive correlation between the duration in year i and the duration in year i +1 (correlation R = 0.32, p = .026): individuals that remained for longer periods in year i remained longer periods in year i +1 but, overall, individuals remained for shorter periods the following year (Fig. 5). The latter is especially true for short-term and long-term winter residents. Passage birds, however, remained longer periods during year i +1 compared to during year i (Fig. 5).

Departure dates

Departure dates for individuals seen between January and April during years 1 and 2 did not vary between years (F(1,179) = 0.02, p = .90), between adults and first-years (F(1,179) = 0.002, p = .89), or between males and females (F(1,137) = 0.03, p = .31). Individuals that were seen during at least two years showed relatively low repeatability values (r = 0.15, Table 3). The difference (in days) between the departure date in year i that of yeari +1 was statistically significant when categorising individuals by their residency at year i (F(2,37) = 4.3,p = .02). This means that long-term birds departed at more similar dates across years compared to passage birds (Table 3). When categorising individuals by their previous age, we found that there was no significant difference in departure dates between adults and first-year birds (F(1,37) = 0.27, p = .61).