To the best of our knowledge, only two studies have described Yellow-legged gull movements using mark-recapture modeling; a study focused on gulls of the michahellis subspecies from Croatia[25], and a second on the lusitanius subspecies from Spain[23]. Our capture-mark-recapture study therefore provides new results to evaluate the effect of age on movement in this species. Our results highlight once again the high mobility of juvenile gulls compared to more sedentary older birds. Interestingly, our work also indicates differences in movement patterns depending on the natal regions. Camargue juveniles were much more sedentary than juveniles from Hyères and Marseille, who almost all moved away from their natal colony zone during the non-breeding period. Among migrating individuals, the areas reached were always in Western Europe. For example, coastal areas of the Bay of Biscay were important, but more distant areas in Great Britain and the Netherlands were also sometimes reached. Our results partially contrast those for Croatian YLGs [26]. As for Croatian YLGs, the farthest areas reached by YLGs of southern France during migration were located further north. However, individuals from southern France had a tendency to move northwest, while those studied in Croatia moved more northeast, with only a slight overlap in the areas used by birds of the two groups in the lower Rhine valley.
Increased mobility in young birds is a frequent observation in seabirds and has been found in other gull species, such as the Lesser black-backed gulls (L. fuscus) [42–44], but also in other seabird groups (albatrosses[10–12], eiders[45] and cormorants[46]). However, the pattern is not systematic. For example, no effect of age on migration distance was found in herring gulls (L. argentatus) [47] ; in this study, only migration departure and return dates varied with age. More widespread movements in juveniles could be an adaptation to avoid low food resources during the post-breeding period, shortages which may occur due to temporal changes in availability and/or increased competition around the breeding area. Older, more experienced, individuals are better competitors and should have acquired a repertoire of alternative foraging locations to avoid this problem. Adult gulls can also be under strong competition to secure high quality nest sites, limiting their ability to move too far from the breeding colony [27]. As juvenile birds do not reach sexual maturity until 4 years old, there is no initial constraint for them to remain locally. Indeed, early wide-scale movements allow young birds to prospect in order to find an optimal breeding area. Prospecting in colonies tends to be most intense at the time of fledgling when the overall quality of the local environment can be evaluated by local breeding success[48]. However, no studies to date have evaluated the role of juvenile prospecting in habitat selection in YLGs.
Among-natal region differences in movement behaviors could be linked to several non-mutually exclusive factors such as social transmission, where juveniles follow migrating adults[49, 50] or landscape features that facilitate or impede flying in specific[51]. It might also be associated with the quality of the local environment. Here, we used survival probabilities as an indicator of local conditions in the colony. We found lower survival estimates for both juveniles and immatures/adults coming from colonies in the Hyères and Marseille regions. This suggests that breeding conditions may have been more difficult in these areas, lowering the relative cost of migration/dispersion for these individuals[7, 52–54]. Regional differences in survival could be explained by resource variation around the breeding sites. For example, it has been shown that landfill closures can impact YLG population demographics via their effect on juvenile survival[55]. It is possible that there was a reduction in accessible garbage around Marseille/Hyères during the study period which forced individuals to migrate further, particularly if alternative food sources were not readily available. It is also possible that differences in migration strategies come from temporal differences in food availability. For example, the proximity of a colony to an landfill could ensure a reliable food source throughout the year[56] such that individuals do not need to migrate. Interestingly, near the Camargue, one of the largest open-air dumps in Europe, the Entressen landfill, was still in operation at the time of this study[57] and may explain why a large proportion of juveniles from this region were considered as sedentary by the model. If juveniles from Marseille and Hyères regions also used this dump for feeding, this would also explain why most young from these colonies were found to move to zone 2 (ie, Entressen is more than 50km from the natal colonies of Marseille/Hyères). The Entressen dump closed in 2010. It would thus be pertinent to investigate this hypothesis more fully, by comparing contemporary juvenile movements to the results of the present study. Interestingly, the diversity of resources available to foraging gulls is much higher in the Camargue compared to the Marseille/Hyères region, with more agricultural land and a more natural littoral zone. We therefore might expect that gull survival in the Camargue has remained relative stable over time, regardless of a reduction in available garbage. This is less likely to be the case for gulls living the in the Marseille/ Hyères region.
In addition to the potential impact of food resources on survival and migration probability, the proportion of migratory individuals within a region may also be conditioned by overall local breeding success, which depends in part on the quality of the local environment[58, 59]. Indeed, the studied colonies likely varied in quality in relation to factors such as population density, vegetation cover, predation, human disturbance, pollution, nest parasites or circulating pathogens. Interestingly, no effect of time was found on survival and/or movement in our study, suggesting that the potential impact of environmental stresses did not change during the study period. If the colonies of Marseille/Hyères represent lower quality breeding locations, we could expect a higher decline in population size over time in these areas relative to the Camargue, both due to a reduction in natal recruitment and to lower emigration rates into the area (i.e., colonial seabirds are known to use conspecific reproductive success to selection breeding habitat)[58]. Although population sizes have declined in this region[60], the role of local breeding habitat quality versus active management strategies, both to close open-air landfill sites and to reduce gull population sizes, cannot be disentangled. Future work will need to consider more carefully the role of such factors in order to better understand the origin of among-colony differences in survival and movement.
In our study, resighting data was divided into two six-month time periods, post-breeding and pre-breeding, in order to obtain robust parameter estimates. However, this division limited our ability to examine movement over shorter time intervals. Short-term movements have been observed in Herring gull adults, which have shorter wintering periods than immatures and juveniles [47]. However, to evaluate this issue more completely, mark-recapture data analysed at shorter time intervals or direct data on adult and juvenile movements from bio-logging will be necessary.
In our analyses, we observed recapture heterogeneity among Yellow-legged gull individuals which led us to include two classes of individuals in our models. The direct source of this heterogeneity is unknown, but likely arises from differences in individual behavior that alter resighting probability. For example, foraging behavior in YLGs can differ greatly both among individuals and among colonies[61]. Individual specialization on particular food resources has been previously observed, with some individuals feeding only at landfills, and others only at sea[62, 63]. The probability of resighting an individual is surely much higher for birds that use landfills because these zones are visited by ornithologists wanting to read rings. Few examples exist in the literature of Mark-Recapture datasets in which individual heterogeneity is corrected for directly in the model, adding fundamental interest to our results[41].
The Yellow-legged gull populations studied here are relatively new since this species did not breed on the French coast prior to 1908. The development of centralized open-air landfills and trawling, in combination with high intrinsic vagility, probably played a major role in this expansion and in the strong increase in population densities seen during the last part of the 20st century[34]. Indeed, widespread juvenile movements and partial migration may help these gulls adapt to novel environmental conditions [64]. However, movement and migration can also be costly. Indeed, the energetic costs of movement and the exploration of unknown areas can lead to higher direct mortality[5]. These movements can also influence exposure to parasites and pathogens, homogenizing their distribution in the environment[65, 66]. Young individuals, which move in greater proportion and to more distant locations, have a higher potential to expose themselves to novel parasites and pathogen and to disperse these agents at different spatial scales than adults[1]. These differences in movement may also create disparities in exposure to other types of environmental stresses such as pollutants that, like pathogens, can have direct consequences on seabird population dynamics[59, 67, 68]. As the presence of diverse environmental stressors can directly impact reproductive success, which in turn can motivate movements, a negative feedback loop may exist between these factors. The use of GPS tags to study the continuous movement of individuals of different ages and from different locations will enable us to obtain a more accurate view of the consequences of age specific movements for population and ecosystem dynamics.
CONCLUSIONS
In this study, we found that Yellow-legged gull movements differed according to age and natal colony. These results provide valuable insights into the movement ecology of this species in the western part of its range. This study demonstrates the importance of studying individuals of different ages and from different colonies/locations when trying to understand migration strategies. Indeed, we show that movements can differ at a relatively small spatial scale, here between colonies only 100 km apart. Similarly, the movement patterns described here took into account heterogeneity in recapture probability, a bias rarely corrected for in studies of seabird movements. This bias is likely to be an important source of variation in other model systems, hindering robust parameter estimation. The method used to correct for capture heterogeneity in this study is thus an example that can inspire other studies facing heterogeneity in their data.