PropertyValue
is nif:broaderContext of
nif:broaderContext
is schema:hasPart of
schema:isPartOf
nif:isString
  • Twenty-four migratory female blue tits were captured for experiments during the standardized ringing scheme in 2007 at Falsterbo bird observatory [45] located at the south-western tip of Sweden (55° 23’N, 12° 49’E). Swedish blue tits have a south-westerly migratory direction and follow leading coastlines [12, 15]. Due to the location of the ringing station, birds captured at the ringing station are all on migration and about to cross the sea [15], apart from two breeding pairs of local blue tits that are ringed before the migrants start to arrive. Only females were selected for the experiment as this is the only sex with considerable migration in both juveniles and adults. All birds started to feed within an hour after capture and were released with a higher body mass than at capture. All experimental protocols comply with national legislation and this study was specifically approved by the Malmö/Lund Animal Care committee. Eight birds each were taken from the bird observatory at three different times during the autumn migration period; a) during early migration (23rd Sep.), b) during peak migration (3rd and 4th Oct.), and c) during late migration (13th Oct). Four birds were juveniles (< 1 year) and four adults (> 1 year) at each time. Birds were captured between 9.30 AM and 4 PM and immediately ringed and aged [46]. Wing length, fat score (following [46]) and body mass were recorded and birds were then transferred to individual cages (0.45 x 0.30 x 0.48 cm) indoors. Cages consisted of two upper, outer perches and one lower middle perch, a water suspensor and a feeder at the front of the cage. Birds were given a mixture of mealworms, Tenebrio sp. and sunflower seeds. All birds were kept in the same room with natural light patterns. Birds could hear but not see each other. Overall, birds were tested five times and then released. All experiments were video-taped. Birds were tested in three different experiments between 9.00 to 11.00 AM, except for experiment one which started when the bird was released into the cage. Either two or six birds were tested at a time due to the arrangement of the cages and availability of only three cameras. (1) On the day of capture, latency to feed was recorded on release into the cage as an indication of how quickly individuals adapt to their new situation. (2) Either on day 4 or 5 after capture (half of the birds, each), the bird’s neophilia (attraction to novelty) was tested by placing a novel object (red or green pyramid; 5 x 5 x 3.5 cm) on one of the upper perches for 30 minutes and recording the latency to touch the object. The object was placed at a neutral location in the cage that the bird was free to approach or avoid. In this situation, the novel object elicits approach (neophilia) and avoidance (neophobia) but in case the bird approaches, neophilia is stronger and a good indicator for an individual’s interest in the object [47]. (3) Birds tested on day 4 on neophilia were tested on neophobia (avoidance of novelty) on day 5 and vice versa. A novel object (orange or white round cotton mop; 7 cm in diameter) was positioned beside the feeder for 60 minutes and the latency to feed was measured. Additionally, the latency to feed after the same disturbance (starting of the cameras), but without the novel object, was measured on two days (control latency) within three days of the neophobia experiment. In the neophobia experiment, the bird is in a conflict between the motivation to feed and the motivation to avoid the novel object. The difference in time between feeding with and without the novel object reflects neophobia [47–48]. Neophilia and neophobia represent two independent motivations [47, 49–50] and also belong to two different personality dimensions [51–52]. On day 9 and 10, all birds were re-tested on the neophobia and neophilia test using the colour that was not used in the previous set-up (balanced design) to test for consistency of behaviours over time. The same object but a different colour was used to keep objects as similar as possible but not identical to avoid habituation [53] and retain novelty [54]. Five dependent variables were extracted from the experiments; a) latency to feed on the capture day, b) latency to explore (neophilia) on the first trial and c) on the second trial, d) latency to feed beside the novel object (neophobia) on the first trial and e) on the second trial. Neophobia latencies were calculated from the average time to feed without the novel object subtracted from the latency to feed with the novel object reflecting the neophobic reaction. Latency to explore was taken as a proxy for intensity of exploration as these variables were negatively correlated in an earlier study on the same population [35] as was the case in Great tits (Parus major) [55]. Note that a few birds did not feed or explore within the set time limits for experiments, leading to truncation of our data. For latency to feed on the capture day we used ANOVA with fat (0–5 with zero indicating no fat) and wing length as an indicator of size as continuous variables and age (younger or older than one year) and migratory season (early—peak—late migration) as factors. We also included the interaction terms age x wing length (see below) and age x migratory season to test for differences between age classes along the migratory window. No three-way interaction terms were included because of the small sample size. Non-significant terms were removed in a backward elimination process, where main terms were retained if they were included in significant interactions. Predictor variables were un-correlated, with the exception of adult birds having longer wings (rp = 0.51, df = 22, P = 0.010) and a trend for lower fat scores (rp = -0.40, df = 22, P = 0.054) than juvenile birds. Time of the day at capture did not affect latency to feed (rp = 0.30, df = 22, P = 0.148). Wing length, indicating the general size of the birds, did not change over season in our sample (b = 0.2, t = 1.6, P = 0.14). For neophobia and exploration latencies we fitted linear mixed effect models (LMM) to consider individual variation with the same independent variables as before. These models were fitted with REstricted Maximum Likelihood (REML). ID of the bird was included as a random factor. Non-significant terms were removed in a backward elimination process, where main terms were retained if they were included in significant interactions. Interactions were further explored by re-fitting the model and shifting the reference levels of the categorical variables. Fat and wing length were used to indicate how much body reserves the birds had stored in preparation for migration and as an indicator of general body size, respectively. Latency to feed, exploration latencies and neophobia were square-root transformed to approach normality. Because the neophobia data contained negative values (minimum = -149.5), we added 150 to all neophobia measurements. All three variables were also analysed using model selection based on Akaike’s Information Criterion corrected for small sample sizes, AICc [56] (see Tables A-C in S1 File). Furthermore, consistency of neophilia and neophobia reactions was tested by comparing the first and second trial using Pearson’s correlations. Finally, we tested for a correlation between latency to feed on the capture day, exploration and neophobia for possible behavioural syndromes by use of Pearson’s correlations. In all correlations, square-root transformed variables were used. All analyses were conducted in the program R version 3.2.2 [57], with add-on packages ‘nlme’ [58] for mixed effects models. All birds started to feed within an hour after capture and were released with a higher body mass than at capture. Experiments conformed to Swedish regulations and were conducted under permit no M237-07.
rdf:type