Migration is fundamental in the life of many birds and entails significant energetic and time investments. Given the importance of arrival time in the breeding area and the relatively short period available to reproduce (particularly at high latitudes), it is expected that birds reduce spring migration duration to a greater extent than autumn migration, assuming that pressure to arrive into the wintering area might be relaxed. This has previously been shown for several avian groups, but recent evidence from four tracked Icelandic Whimbrels (Numenius phaeopus islandicus), a long distance migratory wader, suggests that this subspecies tends to migrate faster in autumn than in spring. Here, we (1) investigate differences in seasonal migration duration, migration speed and ground speed of Whimbrels using 56 migrations from 19 individuals tracked with geolocators and (2) map the migration routes, wintering and stopover areas for this population. Tracking methods only provide temporal information on the migration period between departure and arrival. However, migration starts with the fuelling that takes place ahead of departure. Here we estimate the period of first fuelling using published fuel deposition rates and thus explore migration speed using tracking data. We found that migration duration was shorter in autumn than in spring. Migration speed was higher in autumn, with all individuals undertaking a direct flight to the wintering areas, while in spring most made a stopover. Wind patterns could drive Whimbrels to stop in spring, but be more favourable during autumn migration and allow a direct flight. Additionally, the stopover might allow the appraisal of weather conditions closer to the breeding areas and/or improve body condition in order to arrive at the breeding sites with reserves.

raw_positionsTime stamps from the light data defined by the threshold method and derived locations for each bird. We appreciate that anyone interested in using these data contacts the authors.temp_cond_wetsTemperature, conductivity and wet data for each individual. We appreciate that anyone interested in using these data contacts the authors.