A quasi-experimental approach using telemetry to assess migration-strategy-specific differences in the decision-making processes at stopover

Abstract

Background: Migrant birds travel between their breeding areas and wintering grounds by alternating energetically and physiologically demanding flights with periods of rest and fuelling, so-called stopovers. An important intrinsic factor influencing the decision to resume migration is the amount of energy stores available for the next flight. Correlative studies with free-flying birds and experimental studies with caged birds have shown that the amount of energy stores affects the day-to-day, within-day and the directional decision of departure. The methodological advantages of both the correlative and experimental approach are combined when radio-tagging many individuals on the same day and subsequently determining the departure decisions at a high spatiotemporal resolution. Making use of such a quasi-experimental approach with an automated radio-tracking system at stopover, we studied the effect of energy stores on departure decisions and whether they vary between species of different migration strategies experiencing contrasting time constraints. For this, we chose a long-distance migrant, the common redstart (Phoenicurus phoenicurus), and a medium-distance migrant, the European robin (Erithacus rubecula), because the former has to travel at relatively higher speed to reach its wintering ground in a reasonable time at the expense of relatively higher energetic costs for travelling than the latter.

Results: Common redstarts with higher energy stores were more likely to resume migration than their conspecifics with lower energy stores, whereas this pattern was absent in the European robins. The amount of energy stores significantly affected the timing of departure within the day, with large energy stores yielding early departures in both species. Departure directions from the stopover site during the first night after capture were oriented towards the seasonally appropriate direction but were not affected by variation in energy stores.

Conclusions: We demonstrate the importance of variation in energy stores on the departure decisions and that it may affect species with different migration strategies dissimilarly in autumn. Nevertheless, knowledge of other intrinsic factors, such as feeding conditions, health status and physiological consequences of previous flights, is additionally required to better understand the departure decisions of migrants, as this is the key to providing an overall assessment of the decision-making process.

Keywords: Bird; Common redstart; Departure decision; Energy stores; European robin; Migration; Radio-tracking; Stopover; Strategy.

Fig. 1

Energy stores of common redstarts (grey) and European robins (orange) at the time of capture. Box plots show the 5th, 25th, 50th, 75th, 95th percentile and one outlier (filled dot). Raw data are presented as open circles indicating individuals departing on the night after capture or stars depicting individuals staying more than 1 day on Helgoland. Numbers on the left side of the box represent sample sizes

Fig. 2

Variation in minimum stopover duration in common redstarts (grey) and European robins (orange) on Helgoland during autumn migration. Box plots show the 5th, 25th, 50th, 75th, 95th percentile and outliers (filled dots). Numbers on the left side of the box represent sample sizes. Raw data are presented as open circles

Fig. 3

Energy stores (scaled per species) of common redstarts (grey) and European robins (orange) that departed during the first night capture (“departing”, lower panel) and that stayed longer at the stopover site (“staying”, upper panel). Box plots show the 5th, 25th, 50th, 75th and 95th percentile. Numbers on the left side of the box represent sample sizes. Raw data are presented as open circles. See Fig. 2 for the original values of energy stores per species

Fig. 4

Variation in departure timing of common redstarts (grey) and European robins (orange) from Helgoland during autumn migration. Departure timing is given as (a, c) proportion of night at departure and (b, d) departure time in minutes after sunset for all birds (a, b) and for birds departing during the first night after capture only (c, d). Box plots show the 5th, 25th, 50th, 75th, 95th percentile and outliers (filled dots). Numbers at the bottom of the boxes represent sample sizes

Fig. 5

Effect of energy stores on nocturnal departure timing (proportion of night at departure) in common redstarts (grey, n = 6) and European robins (orange, n = 7) departing during the first night after capture. Depicted is the model estimate (solid black line) from the corresponding beta regression model, cf. Table 1b

Fig. 6

Departure directions of common redstarts (grey, n = 14) and European robins (orange, n = 21) from Helgoland during autumn migration. The six common redstarts and the seven European robins leaving Helgoland during the first night after capture are indicated by the black circles encompassing the corresponding filled dots. If departure directions were not uniformly distributed within a group and if the resultant mean vector length was significant, the mean direction of each group was represented by an arrow (thick arrow for all individuals of a species; thin arrow only for individuals departing during the first night after capture), whose length is drawn relative to the radius of the circle (= 1)

Fig. 7

Location of the study site Helgoland in the German Bight. From there about 50 km are to be covered across the North Sea Bering Strait to reach the mainland. a The automated digital radio-telemetry system on Helgoland consists of twelve antennas at three sites (A, B, C). Coloured bars represent the different antennas and correspond to those given in (b). b Two nocturnal departure events as recorded by the system showing raw signal strength data over time (Coordinated Universal Time: UTC). The time of departure (take-off) defined as the time of highest signal strength and the estimated departure direction based on signals from the second half of the departure event are given. Colours denote signals received by different antennas aligned to directions given in the legend. This figure was created by using R, ver. 3.5.3 [76]