Migratory species display a range of migration patterns between irruptive (facultative) to regular (obligate), as a response to different predictability of resources. In the Arctic, snow directly influences resource availability. The causes and consequences of different migration patterns of migratory species as a response to the snow conditions remains however unexplored. Birds migrating to the Arctic are expected to follow the spring snowmelt to optimise their arrival time and select for snow-free areas to maximise prey encounter en-route. Based on large-scale movement data, we compared the migration patterns of three top predator species of the tundra in relation to the spatio-temporal dynamics of snow cover. The snowy owl, an irruptive migrant, the rough-legged buzzard, with an intermediary migration pattern, and the peregrine falcon as a regular migrant, all followed, as expected, the spring snowmelt during their migrations. However, the owl stayed ahead, the buzzard stayed on, and the falcon stayed behind the spatio-temporal peak in snowmelt. Although none of the species avoided snow-covered areas, they presumably used snow presence as a cue to time their arrival at their breeding grounds. We show the importance of environmental cues for species with different migration patterns.
Animals are expected to alter their behaviour as a response to ongoing climate change1. Animal movement is a behaviour that can be strongly influenced by external, biotic and abiotic environmental factors2. Several studies suggested that avian species adjust their movements in response to environmental conditions3,4,5 and consequently, optimise their reproduction and survival6,7. However, more evidence is necessary to fully understand this mechanism. It is thus important to study individual and species-specific movement responses to the environment, and assess whether these responses differ regarding their spatio-temporal scale8.
Avian migration i.e. seasonal movement between breeding and non-breeding areas is in large part driven by the availability of resources9. Furthermore, the predictability of resources is presumed the main stimulus of migration pattern10,11. Ephemeral resources, unpredictable in time and space, can lead to irruptive (facultative) migration, whereas predictable availability of resources is associated with regular (obligate) migration. However, species and populations can display either of these two migration patterns or a mixed migration pattern that combines elements of both behaviours. Depending on the predictability of available resources, arctic-breeding raptor species exhibit contrasting migration patterns, representing a gradient from irruptive, mixed, to regular migration. The snowy owl (Bubo scandiacus) is irruptive in parts of its range with a variable migration schedule, and weak fidelity to breeding sites10,12,13,14. The rough-legged buzzard (Buteo lagopus) exhibits some flexibility in its migration schedule and site fidelity, but it can occasionally show irruptive movements15. The peregrine falcon (Falco peregrinus) is a regular migrant that has a fixed migration schedule in combination with high breeding-site fidelity16,17,18.
Migration patterns in birds are among other factors linked to the predictability of available prey and consequently diet of the species7,19. Because prey availability and abundance fluctuates across space and time and differently among prey species19, diet generalists have a greater chance than specialists to encounter at least some prey types. During the breeding season, the snowy owl is a specialist feeding mostly on microtine rodents20, a highly fluctuating food resource12. The rough-legged buzzard has a mixed diet, specialising on small mammals when available, switching to medium-sized mammals and birds when small mammals are scarce20,21,22. The peregrine falcon is a diet generalist, feeding on medium-sized birds, a non-fluctuating resource, but also on small mammals20,23. Prey availability for the peregrine falcon (a generalist predator) should thus be less variable among years than for the snowy owl (specialist predator) with the rough-legged buzzard in between. While in both irruptive and regular patterns food availability is the ultimate cause for migration, in irruptive migrants it also acts as a proximate stimulus. Irruptive migrants thus respond to local food conditions directly by delaying/advancing migration11, whereas regular migrants should rely less on food conditions but more on proxies such as day length change coupled to a strong endogenous control24. The proximate response is especially important during spring migration for migratory species breeding in the Arctic. The window of opportunity with favourable environmental conditions for successful breeding in the Arctic is very narrow and the start of breeding is restricted to a few days only. Arriving at the breeding grounds at the right time is thus crucial for successful reproduction25,26,27. While early arrival secures a larger time window for breeding, being too early can expose migrants to harsh weather conditions and low food resources. Migrants should thus adjust their movement during spring migration to arrive at the breeding grounds at the optimal time.
Snow cover can limit access to small mammals at the breeding grounds28,29. For this reason, arctic migrants may follow the northern progression of snowmelt during spring migration to optimise arrival time. The snowy owl, being mainly an irruptive migrant and lemming specialist, is expected to closely follow the progression of snowmelt, followed by the rough-legged buzzard and then the peregrine falcon. Arctic migrants may also actively avoid snow-covered areas because of the limited access to small mammals30 when they feed during migration31. The snowy owl, which relies most strictly on small mammals may exhibit the strongest preference towards snow-free areas, somewhat lower preference is expected from the rough-legged buzzard and the lowest preference from the peregrine falcon. The literature on this topic is scarce (but see32,33) and the process of decision-making during migration as a response to snow conditions is largely unexplored. Other environmental factors such as temperature, day length34, productivity (NDVI - normalized difference vegetation index)35 and wind conditions36,37 were previously described influencing arctic migrants. Here we focus on snow cover as an additional factor that was rarely investigated before and might modulate spring migration movements.
In this study, we assess the impact of snow cover on movements during the spring migration of three top predator species of the tundra. We explore whether they respond differently to the snow cover depending on migration pattern (i.e., irruptive vs regular). First, we investigate whether arctic migrants follow the northern progression of snowmelt and whether the irruptive and regular migrants respond to the snowmelt progression differently. Second, we examine whether they avoid snow-covered areas from alternative options in the environment during the spring migration. We also test whether the avoidance of snow-covered areas differs with the spatio-temporal scale. We predict that (1) irruptive migrants follow the snowmelt the most, regular migrants least, and the effect of snowmelt on migratory movements of mixed migrants should be intermediate; (2) arctic species avoid areas with snow cover, with the irruptive migrants showing the highest and the regular migrants the lowest avoidance, and mixed migrants exhibiting intermediate avoidance.
We used Argos and GPS transmitters deployed from 2001 to 2018 in North America and Eurasia to collect data on three arctic species. We confirm that all experimental protocols were officially approved and all methods were carried out in accordance with the relevant guidelines and regulations. The following boards approved our study: University of Saskatchewan, Animal Protection Committee of Laval University, Animal welfare unit in Norway, Norwegian Environment Agency, USGS Bird Banding Lab, Government of Greenland and The Danish Polar Center. Details about permits, capture methods38, sites and transmitters used are provided in the supplementary material. We removed Argos locations with CLS Argos class lower than 2 (error >500 meters)39. We did not filter GPS locations because of their high accuracy (<100 meters). Since we were specifically interested in spring movement, we extracted spring migration tracks using First Passage Time (see “Determination of spring migration periods” in the supplementary material). After data cleaning, we used 225 individuals that bred in the Arctic for analyses of spring migration, totalling in 245,509 spatial points (Table 1, Fig. 1). Since data were collected as parts of several projects, sampling intervals were not consistent (min = 1 second, median = 6 minutes, max = 39 days). We dealt with these differences by using Step Selection Function (SSF) so that the consecutive positions were separated by a certain threshold value (see below).
We obtained daily snow cover data with a spatial resolution of 500 meters (Global SnowPackMODIS) from the German Aerospace Center (DLR). The obtained snow-covered data matched the extents and dates of spring migration tracks of the three species (see supplementary material under “Determination of spring migration periods”). This product is based on the daily snow cover products MOD10A1 and MYD10A1 (version 5 and 6 as provided by the National Snow and Ice Data Center NSIDC) of the Moderate Resolution Imaging Spectroradiometer (MODIS), which has been reprocessed to remove the effects of cloud cover and polar darkness40. Note that after the interpolation process the information about snow cover fraction is lost as only algorithm for interpolating binary snow cover is currently available. Using the R package “raster”41, we annotated snow cover values to the modelled movement tracks linking individual locations with snow cover data from the same dates. We calculated day length (hours) using the “daylength” function from the “geosphere” R package42. Through the open-source Env-DATA system43 we obtained and annotated the following environmental variables: air temperature (°C) at a resolution of 6 hours and 0.75 degrees (ECMWF interim full daily SFC, 2 meters above ground), vegetation index (NDVI) with a resolution of 0.05 degrees and 16 days (MOD13C1) and wind direction variables (U, the east-west component of the wind; V, the north-south component of the wind) with a resolution of 6 hours and 0.75 degrees (ECMWF interim full daily at surface, 10 meters above ground). Based on the U and V components (meters per second), we calculated wind support (wind in the direction of each bird’s movement) and crosswind (wind perpendicular to the bird’s movement)44, where heading and ground speed were derived from the locations of the individual track.
Step Selection Function
We used Step Selection Function (SSF) to investigate how environmental predictors affect spring movements of arctic raptors45,46,47,48. The method assumes discrete movement decisions, represented by a fixed time step length. At each step, an animal chooses a location out of a set of available alternate locations characterised by differences in environmental conditions fitted to a conditional logistic regression. We modelled SSFs separately for species and step lengths of one day, three days, and five days to investigate whether decision-making during spring migration changes at different temporal scales (“amt” R package49). Step lengths were chosen so that no auto-correlation of environmental data in time is present (see lags for one - five days in Supplementary Fig. S1). We readjusted the data according to the chosen step length so that when two observations were separated by less than the step length value, the second observation was removed, and when two observations were separated by more than the step length value, the later observation was assigned to a different burst. A burst is a segment of a track that includes only locations separated by the chosen step length value. We estimated the distribution of alternative steps using distance and relative angle (angle between the previous and new direction of movement) of the chosen steps within individual migration tracks of each species. Using an exponential distribution, we randomly selected 10 alternative locations for each actual chosen location. These alternative locations together with a chosen location at each step form a stratum. Finally, we annotated chosen and alternative locations with the relevant environmental variables.
Spring movements in relation to the northern progression of snowmelt
We used the chosen locations of individuals at a one-day step length to investigate whether arctic migrants followed the spring snowmelt. We independently tested day length, temperature and NDVI to evaluate whether these factors additionally influence spring movements in arctic migrants. At each location, we extracted snow cover, day length, temperature and NDVI values and compared them between the past, present, and future (from 10 days in the past to 10 days in the future), separately for each species (Fig. 2a). The scale of 10 days was chosen to exclude possible auto-correlation of environmental data in time (Supplementary Fig. S1). Note also that using larger (15 and 20 days) or smaller (5 days) time-windows resulted in the same overall effect (Supplementary Fig. S2). We used generalised linear mixed models (GLMM) implemented in “lme4”50 R packages. In the model, we used snow cover, temperature, day length and NDVI as dependent variables (separately in each model), day as a predictor variable, and year nested in individual as a random effect. We checked the assumptions of the tests and compared models containing all predictors with models leaving out specific predictors to evaluate the effect of each predictor, using Akaike’s information criterion (AIC). We considered models with an ΔAIC > 2 as different where the model with lower AIC has stronger support51.
Movement decisions based on snow cover
We compared environmental conditions between chosen and alternative locations (Fig. 3a) to test whether arctic migrants avoided snow-covered areas when following the receding snow line. In the analysis, we also included wind support and crosswind (only for one-day step length) to test whether these factors would influence the birds’ choice of distance and direction and to evaluate whether snow cover predicted movement decisions irrespective of the wind conditions. We fitted mixed conditional logistic regression models using “coxme” function from “coxme” R package52,53, separately by species and separately for one-, three-, and five-day step lengths. As a dependent variable, we included movement decision (chosen vs alternative), as predictors, snow cover, wind support, and crosswind, and as a random effect stratum nested in individual. We performed Shapiro-Wilk’s tests for multivariate normality (p < 0.001 for each species) and check the test’s assumptions. Correlation coefficients between snow cover, wind support, and crosswind were less than 0.754, thus we included these variables in the same model. Based on the Akaike’s information criterion (AIC), we compared the full models with those without the predictor of interest to evaluate the effect of this predictor and select a better model. We considered models with an ΔAIC > 2 as different, with lower AIC having more support51. We validated the models with Used habitat calibration (UHC) plots55 by comparing the distribution of the observed and predicted values of explanatory variables at the chosen locations with the distribution at alternative locations.
Spring movements in relation to the northern progression of snowmelt
All three species followed the northern progression of snowmelt but responded to snowmelt progression differently. Movements of the rough-legged buzzards were most tightly adjusted to snow progression, those of the snowy owl less, and those of the peregrine falcon the least (see coefficients, SE’s and ΔAIC in Table 2 and in Supplementary Table S1 and difference in the slopes in Fig. 2). Higher temperature, longer days and higher NDVI were positively correlated with the spring movements in all three species (see ΔAIC in Table 2 and Fig. 3). For the owl, temperature was the most important factor, following by snow cover, for the buzzard, snow cover was the most important followed by NDVI and for the falcon, temperature was the most important, followed by day length (see ΔAIC in Table 2). The species also greatly differed in the snow conditions in the areas they used during spring migration. The owl occurred mostly in snow-covered regions, the buzzard used partially snow-covered areas, and the falcon used snow-free areas (see the snow cover values at day zero; Fig. 2). This can be seen also with individual migration profiles where responses to snow cover were detected (Supplementary Fig. S3). The owl experienced little changes between snow-free and snow-covered areas during migration and continued moving north after entering a snow-covered area. The falcon also experienced little changes between snow-covered and snow-free areas but migrated only when the area was free of snow. The buzzard experienced high variability in the snow-covered and snow-free areas and when it reached the snow-covered area during migration, it waited or retreated to lower latitude before continuing its northward migration.
Movement decisions based on snow cover
The three species did not exhibit any preference or avoidance for snow cover during spring migration when assessing movement decisions at one-day steps (see coefficients, SE’s and ΔAIC values on Table 3, and slopes on Fig. 4). At three- and five-day steps, no response to snow cover remained for the three species (see ΔAIC values in Supplementary Table S1). Wind conditions measured at one-day steps, however, influenced their movement decisions. The owl and the falcon moved towards locations with less wind support and the buzzard towards locations with more wind support and crosswind (see ΔAIC values on Table 3, and slopes on Supplementary Fig. S4). Model validation confirmed the significant effects of wind support and crosswind (see UHC plots in Supplementary Fig. S5).
Consistent with our first prediction, spring migratory movements of all three arctic raptors we studied were synchronised with the northern progression of snowmelt. In addition to temperature, photoperiod and NDVI, the arctic raptors presumably use also snow cover as a cue during spring migration to avoid arriving too early or too late at the breeding grounds. Indeed, snow cover was the most important predictor for the buzzard, less for the owl and least for the falcon when comparing snow cover with temperature, day length and NDVI. Optimising the timing of arrival allows migrants to find territories rich in resources and to start breeding when conditions are favourable56,57. However, despite the overall trend in following snowmelt progression, our species differed in the details of their responses. We predicted that irruptive species (snowy owl) would track snowmelt most closely, mixed migrants (rough-legged buzzard) would have a weaker association, and that regular migrants (peregrine falcon) would show little association with snowmelt. Consistent with this, the owl and the buzzard were more flexible and followed the snowmelt more than the falcon but surprisingly, the owl responded to the progression of snowmelt less than the buzzard. Perhaps the owl followed the progression of snowmelt by moving to different directions rather than with directed movements. Thus, the pattern of following the northern progression of snowmelt was less evident in the owl than in the buzzard. There is indeed evidence that snowy owls visit and search for suitable areas to settle and breed during spring migration12.
While moving to northern breeding sites, the species occupied areas with different snow coverage and thus seemed to follow the northern progression of snowmelt differently. The owl returning to northern breeding areas moved between areas with 75% of snow cover, therefore was ahead of snowmelt. The slow speed of snowmelt in areas the owl passes during the migration can explain our result of an apparent tighter response in the buzzard than in the owl. The buzzard which moved between areas with 35% of snow cover was at the peak of snowmelt, passing at the time when snowmelt was most rapid. Indeed, the melt of snow likely follows a sigmoidal curve, slow at first, rapid at 50% snow cover and slow at the end when only patches of snow remain58. The buzzard also experienced many changes between snow-free and snow-covered areas which confirms the species being on the peak of snowmelt. In addition, the progression of the buzzard’s northern movements was strictly limited by the snow cover in comparison to the other two species which is another reason for the apparent strongest response to snowmelt in this species. The falcon mostly occupied areas with 10% snow cover and was thus behind the peak of snowmelt. This outcome corresponds to the result that the falcon showed the smallest response to the progression of snowmelt.
Contrary to our second prediction, arctic raptors neither showed a preference nor avoidance for snow-covered areas when following the receding snow line. We predicted that arctic species avoid areas with snow cover, with the owl showing the highest avoidance, the buzzard intermediate avoidance, and the falcon the lowest avoidance. One reason for no response to snow cover could be that the snow cover can represent different conditions of food availability depending on the circumstances. As snow cover affects food availability, the snowy owl and the rough-legged buzzard, which rely on small mammals for food20, can find prey on snow-free areas59. Snow-covered areas might be preferable, given that small mammals are known to concentrate under the snow28,29,59,60. As the spring progresses and the snow starts melting, small mammals may not be concentrated under the snow but dispersed, thus the timing in spring could also play a role. Also, physical properties of snow such as depth and density need to be considered when investigating the availability of small mammals in spring28,60. In addition, there might be species-specific differences in the preference or avoidance of snow-covered areas depending on their thermoregulation and camouflage requirements61.
Other environmental factors additionally influenced the movements of arctic raptors. Besides the northern progression of snowmelt, raptors synchronised their spring migration movements with the increasing temperature, longer days and higher productivity. These environmental factors influenced spring migration timing in several northern migrants, for example, the pink-footed goose (Anser brachyrhynchus)34 and the barnacle goose (Branta leucopsis)62. When investigating movement preference at each step during the spring migration, movement decisions based on wind conditions differed between the species. The owl and falcon moved towards areas with less wind support while the buzzard moved towards more wind support and crosswind. These results imply that arctic raptors are not necessarily negatively affected by wind support and crosswind and might compensate for the drift similar to the honey buzzard (Pernis apivorus)63. Wind conditions, thus, seemed to play a secondary role in movement decisions of arctic raptors.
Our study demonstrates that spatio-temporal changes in snow cover affect the movements of arctic raptors depending on the species differently. The study provides evidence that the species breeding in the same environment and exhibiting contrasting migration patterns respond differently to the snow conditions. Rapid environmental change such as a global increase in temperature is expected to shape snow patterns64 resulting in more rapid snowmelt and prolonged snow-free periods65. With earlier greening, some species might have broader windows of opportunity for breeding. However, environmental conditions are forecasted to become less predictable too66 which represent a challenge for these species. Resulting spatio-temporal changes in food availability67,68 could force species to alter migration patterns or even strategies69,70. The ability of a species to adjust movements to environmental conditions likely depends on the degree of phenotypic plasticity71,72. The peregrine falcon, having regular (less flexible) migration would have probably the most difficulty adjusting to changing conditions, which can result in a phenological mismatch73,74 affecting fitness. In contrast, the snowy owl and the rough-legged buzzard which have more flexible migration, are expected to more easily adapt to these changes.
Changing snow cover patterns caused by climate change will likely shape not only migration patterns of species but have effects that cascade through the entire ecosystem and shape a species’ phenology, breeding success, and survival. Thus, at a larger scale, population dynamics, species distribution, and trophic interactions will likely also change75. These changes might occur especially rapidly in the Arctic ecosystem where warming was three times greater than elsewhere over the last three decades66,76. Studying how species track and use the environment is therefore of great importance to predict the impact of future changes. The next step is to model how migration patterns change under different climate scenarios and identify the species in need of management actions. We speculate that our results are applicable to all other irruptive, regular and intermediate migrants in the Arctic but detailed data on the migration patterns of more species are needed to test the applicability of our results. Species might show different responses to the snow cover not only as a result of migration pattern but also due to differences in migration timing and distance, diet, and possibly physical and physiological adaptation to snow. Although we could not tease apart the potential effects of such variables here, we were able to provide one of the most robust comparisons among species breeding in the same environment. Our study is thus an important step towards understanding the proximate cues that species with different migration patterns use in response to environmental conditions.
All datasets used in this study are available upon request from the authors.
Post, E. et al. Ecological consequences of sea-ice decline. Science 341, 519–524 (2013).
Nathan, R. et al. A movement ecology paradigm for unifying organismal movement research. Proc. Natl. Acad. Sci. 105, 19052–19059 (2008).
Mandel, J., Bildstein, K., Bohrer, G. & Winkler, D. Movement ecology of migration in turkey vultures. Proc. Natl. Acad. Sci. 105, 19102–19107 (2008).
Tøttrup, A. P. et al. Avian migrants adjust migration in response to environmental conditions en route. Biol. Lett 4, 685–688 (2008).
Poessel, S. A., Brandt, J., Miller, T. A. & Katzner, T. E. Meteorological and environmental variables affect flight behaviour and decision-making of an obligate soaring bird, the california condor gymnogyps californianus. Ibis 160, 36–53 (2018).
Cox, G. W. The evolution of avian migration systems between temperate and tropical regions of the new world. Am. Nat. 126, 451–474 (1985).
Alerstam, T., Hedenström, A. & Åkesson, S. Long-distance migration: evolution and determinants. Oikos 103, 247–260 (2003).
Harel, R. et al. Decision-making by a soaring bird: time, energy and risk considerations at different spatio-temporal scales. Philos. Trans. R. Soc. Lond. B: Biol. Sci 371, 20150397 (2016).
Newton, I. The migration ecology of birds (Elsevier, 2010).
Newton, I. Advances in the study of irruptive migration. Ardea 94, 433 (2006).
Newton, I. Obligate and facultative migration in birds: ecological aspects. J. Ornithol. 153, 171–180 (2012).
Therrien, J.-F., Gauthier, G., Pinaud, D. & Bêty, J. Irruptive movements and breeding dispersal of snowy owls: a specialized predator exploiting a pulsed resource. J. Avian Biol. 45, 536–544 (2014).
Holt, D., Larson, M. D., Smith, N., Evans, D. & Parmelee, D. F. Snowy owl bubo scandiacus. Birds of North America, http://bna.birds.cornell.edu/bna/species/010 (2015).
Robillard, A., Gauthier, G., Therrien, J.-F. & Bêty, J. Wintering space use and site fidelity in a nomadic species, the snowy owl. J. Avian Biol. 49, jav–01707 (2018).
Beardsell, A., Gauthier, G., Therrien, J.-F. & Bêty, J. Nest site characteristics, patterns of nest reuse, and reproductive output in an arctic-nesting raptor, the rough-legged hawk. Auk 133, 718–732 (2016).
Fuller, M. R., Seegar, W. S. & Schueck, L. S. Routes and travel rates of migrating peregrine falcons falco peregrinus and swainson’s hawks buteo swainsoni in the western hemisphere. J. Avian Biol. 29, 433–440 (1998).
McGrady, M. J., Maechtle, T. L., Vargas, J. J., Seegar, W. S. & Porras Peña, M. C. Migration and ranging of peregrine falcons wintering on the Gulf of Mexico coast, Tamaulipas, Mexico. Condor 104, 39–48 (2002).
White, C. M., Clum, N. J., Cade, T. J. & Hunt, W. G. Peregrine falcon falco peregrinus. Birds of North America, https://birdsna.org/Species-Account/bna/species/perfal/introduction (2002).
Robinson, B. G., Franke, A. & Derocher, A. E. The influence of weather and lemmings on spatiotemporal variation in the abundance of multiple avian guilds in the Arctic. PLoS One 9, e101495 (2014).
Ferguson-Lees, J. & Christie, D. A. Raptors of the world (A&C Black, 2001).
Bechard, M. J. & Swem, T. R. Rough-legged hawk buteo lagopus. Birds of North America, https://birdsna.org/Species-Account/bna/species/rolhaw/introduction/ (2002).
Pokrovsky, I. et al. Diet, nesting density, and breeding success of rough-legged buzzards (Buteo lagopus) on the Nenetsky Ridge, Arctic Russia. Polar Biol. 37, 447–457 (2014).
Bradley, M. & Oliphant, L. W. The diet of peregrine falcons in rankin inlet, northwest territories: an unusually high proportion of mammalian prey. Condor 93, 193–197 (1991).
Berthold, P. Bird migration: a general survey (Oxford University Press, 2001).
Bêty, J., Giroux, J.-F. & Gauthier, G. Individual variation in timing of migration: causes and reproductive consequences in greater snow geese (Anser caerulescens atlanticus). Behav. Ecol. Sociobiol. 57, 1–8 (2004).
Madsen, J. et al. Effects of snow cover on the timing and success of reproduction in high-arctic pink-footed geese anser brachyrhynchus. Polar Biol. 30, 1363–1372 (2007).
Lamarre, V., Franke, A., Love, O. P., Legagneux, P. & Bêty, J. Linking pre-laying energy allocation and timing of breeding in a migratory arctic raptor. Oecologia 183, 653–666 (2017).
Reid, D. G. et al. Lemming winter habitat choice: a snow-fencing experiment. Oecologia 168, 935–946 (2012).
Bilodeau, F., Gauthier, G. & Berteaux, D. Effect of snow cover on the vulnerability of lemmings to mammalian predators in the Canadian Arctic. J. Mammal. 94, 813–819 (2013).
Sonerud, G. A. Effect of snow cover on seasonal changes in diet, habitat, and regional distribution of raptors that prey on small mammals in boreal zones of Fennoscandia. Ecography 9, 33–47 (1986).
Bildstein, K. L. Migrating raptors of the world: their ecology & conservation (Cornell University Press, 2006).
Therrien, J.-F. et al. Is pre-breeding prospecting behaviour affected by snow cover in the irruptive snowy owl? A test using state-space modelling and environmental data annotated via Movebank. Mov. Ecol 3, 1 (2015).
Yamaguchi, N. M., Hiraoka, E., Hijikata, N. & Higuchi, H. Migration routes of satellite-tracked Rough-legged Buzzards from Japan: the relationship between movement patterns and snow cover. Ornithol. Sci 16, 33–41 (2017).
Bauer, S., Gienapp, P. & Madsen, J. The relevance of environmental conditions for departure decision changes en route in migrating geese. Ecology 89, 1953–1960 (2008).
Shariatinajafabadi, M. et al. Migratory herbivorous waterfowl track satellite-derived green wave index. PLoS One 9, e108331 (2014).
Klaassen, M., Beekman, J. H., Kontiokorpi, J., Mulder, R. J. & Nolet, B. A. Migrating swans profit from favourable changes in wind conditions at low altitude. J. Ornithol. 145, 142–151 (2004).
Liechti, F. Birds: blowin’ by the wind? J. Ornithol. 147, 202–211 (2006).
Bloom, P., Clark, W. & Kidd, J. Capture techniques. 193–219. DM Bird and KL Bildstein.[E ds.]. Raptor research and management techniques. Hancock House Publishers. Blaine, WA USA (2007).
Douglas, D. C. et al. Moderating Argos location errors in animal tracking data. Methods Ecol. Evol 3, 999–1007 (2012).
Dietz, A. J., Kuenzer, C. & Dech, S. Global snowpack: A new set of snow cover parameters for studying status and dynamics of the planetary snow cover extent. Remote Sens. Lett 6, 844–853 (2015).
Hijmans, R. J. raster: Geographic Data Analysis and Modeling, https://CRAN.R-project.org/package=raster, R package version 2.9-5 (2019).
Hijmans, R. J. geosphere: Spherical Trigonometry, https://CRAN.R-project.org/package=geosphere, R package version 1.5-10 (2019).
Dodge, S. et al. The environmental-data automated track annotation (Env-DATA) system: linking animal tracks with environmental data. Mov. Ecol 1, 3 (2013).
Safi, K. et al. Flying with the wind: scale dependency of speed and direction measurements in modelling wind support in avian flight. Mov. Ecol 1, 4 (2013).
Forester, J. D., Im, H. K. & Rathouz, P. J. Accounting for animal movement in estimation of resource selection functions: sampling and data analysis. Ecology 90, 3554–3565 (2009).
Potts, J. R., Mokross, K., Stouffer, P. C. & Lewis, M. A. Step selection techniques uncover the environmental predictors of space use patterns in flocks of Amazonian birds. Ecol. Evol 4, 4578–4588 (2014).
Potts, J. R., Bastille-Rousseau, G., Murray, D. L., Schaefer, J. A. & Lewis, M. A. Predicting local and non-local effects of resources on animal space use using a mechanistic step selection model. Methods Ecol. Evol 5, 253–262 (2014).
Avgar, T., Potts, J. R., Lewis, M. A. & Boyce, M. S. Integrated step selection analysis: bridging the gap between resource selection and animal movement. Methods Ecol. Evol 7, 619–630 (2016).
Signer, J., Fieberg, J. & Avgar, T. Animal movement tools (amt): R package for managing tracking data and conducting habitat selection analyses. Ecol. Evol 9, 880–890 (2019).
Bates, D., Mächler, M., Bolker, B. & Walker, S. Fitting linear mixed-effects models using lme4. J. Stat. Softw. 67, 1–48 (2015).
Burnham, K. P. & Anderson, D. R. Model selection and multimodel inference: A practical information-theoretic approach (Springer, 2002).
Therneau, T. Mixed effects Cox models, https://CRAN.R-project.org/package=coxme, R package version 2.2-14 (2015).
Thurfjell, H., Ciuti, S. & Boyce, M. S. Applications of step-selection functions in ecology and conservation. Mov. Ecol 2, 4 (2014).
Dormann, C. F. et al. Collinearity: a review of methods to deal with it and a simulation study evaluating their performance. Ecography 36, 27–46 (2013).
Fieberg, J. R. et al. Used-habitat calibration plots: a new procedure for validating species distribution, resource selection, and step-selection models. Ecography 41, 737–752 (2018).
McNamara, J. M., Welham, R. K. & Houston, A. I. The timing of migration within the context of an annual routine. J. Avian Biol. 416–423 (1998).
Kokko, H. Competition for early arrival in migratory birds. J. Animal Ecol 68, 940–950 (1999).
Kattelmann, R. Peak flows from snowmelt runoff in the Sierra Nevada, USA. Snow, Hydrology and Forests in High Alpine Areas (Proceedings of the Vienna Symposium) 203–211 (1991).
Duchesne, D., Gauthier, G. & Berteaux, D. Habitat selection, reproduction and predation of wintering lemmings in the Arctic. Oecologia 167, 967–980 (2011).
Domine, F. et al. Snow physical properties may be a significant determinant of lemming population dynamics in the high. Arctic. Arct. Sci 4, 813–826 (2018).
Potapov, E. & Sale, R. The snowy owl (A&C Black, 2013).
Si, Y. et al. Do arctic breeding geese track or overtake a green wave during spring migration? Scientific reports 5, 8749 (2015).
Klaassen, R. H., Hake, M., Strandberg, R. & Alerstam, T. Geographical and temporal flexibility in the response to crosswinds by migrating raptors. Proc. R. Soc. Lond. B: Biol. Sci. 278, 1339–1346 (2010).
Serreze, M. C. & Barry, R. G. Processes and impacts of Arctic amplification: A research synthesis. Glob. Planet. Change 77, 85–96 (2011).
Bhatt, U. S. et al. Circumpolar Arctic tundra vegetation change is linked to sea ice decline. Earth Interact. 14, 1–20 (2010).
Post, E. et al. Ecological dynamics across the Arctic associated with recent climate change. Science 325, 1355–1358 (2009).
Ims, R. A., Henden, J.-A. & Killengreen, S. T. Collapsing population cycles. Trends Ecol. Evol. 23, 79–86 (2008).
Kausrud, K. L. et al. Linking climate change to lemming cycles. Nature 456, 93 (2008).
Cotton, P. A. Avian migration phenology and global climate change. Proc. Natl. Acad. Sci. 100, 12219–12222 (2003).
Jonzén, N. et al. Rapid advance of spring arrival dates in long-distance migratory birds. Science 312, 1959–1961 (2006).
Charmantier, A. et al. Adaptive phenotypic plasticity in response to climate change in a wild bird population. Science 320, 800–803 (2008).
Switzer, P. V. Site fidelity in predictable and unpredictable habitats. Evol. Ecol. 7, 533–555 (1993).
Both, C. et al. Avian population consequences of climate change are most severe for long-distance migrants in seasonal habitats. Proc. R. Soc. Lond. B: Biol. Sci. 277, 1259–1266 (2009).
Saino, N. et al. Climate warming, ecological mismatch at arrival and population decline in migratory birds. Proc. R. Soc. Lond. B: Biol. Sci. 278, 835–842 (2010).
Walther, G.-R. et al. Ecological responses to recent climate change. Nature 416, 389 (2002).
Comiso, J. C. & Hall, D. K. Climate trends in the Arctic as observed from space. Wiley Interdiscip. Rev. Clim. Chang 5, 389–409 (2014).
We thank Risto Karvonen, Arve Østlyngen, Ken Gøran Uglebakken, Dag Gjerstad, Steve Aslaksen, Torkjell Morset, Jostein Sandvik and Ole Andreas Forseth for assistance with catching rough-legged buzzards in Norway. For assistance in the field with catching snowy owls in Norway, we thank Norwegian Nature Inspectorate (SNO) in Finnmark, Fjelltjenesten, Statskog (The State-owned Land and Forest Company) in Troms and Nordland. For helping to trap owls in Canada we thank Dan Zazelenchuk, Marten Stoffel, Mike Blom, Erhard Pletz, Mike Russel and for the help in the eastern part of Canadian Arctic, we thank Audrey Robillard. For assistance in trapping owls in the US we thank Mike Lanzone, Nova Mackentley, Tom McDonald, Chris Neri, Norman Smith, and Matt Solensky. We also thank all Project SNOWstorm (www.projectsnowstorm.org) supporters and collaborators. For fieldwork with rough-legged buzzards and peregrine falcons in Russia, we thank Olga Kulikova. For work on rough-legged buzzards in North America we thank Nina Kidd of Kidd Biological Inc, Travis Booms Alaska Dept of Fish and Game, Bryan Bedrosian of Teton Raptor Center, Neil Paprocki of Univ of Idaho - Moscow, Scott Thomas of Kidd Biological, Inc., Colin Dillingham USFS, Dylan Kesler of Vertebrate Systems LLC and Keith LaSage GeoTrak, Inc. We also thank Anne Scharf and Martina Scacco for the help with data analyses, participants of the IMPRS writing course for revising the manuscript and finally, to anonymous reviewers for constructive comments that substantially improved our manuscript. Funding for the work on snowy owls in Norway was provided by the Norwegian Environment Agency, NOF-BirdLife Norway’s Snowy owl Foundation and the Environment departments at the Office of the County Governors of Troms and Finnmark, Trøndelag, Innlandet, Vestfold and Telemark, Oslo and Viken, Nordland and Vestland. Funding for the snowy owls in the eastern Canadian Arctic was provided by Polar Continental Shelf Project, the Garfield-Weston Foundation, Natural Science and Engineering Research Council of Canada, and Fonds Quebecois de Recherche en Nature et Technologies. The work on snowy owls in western Canada was funded by the Natural Sciences and Engineering Research Council of Canada (grant 203177). Work with snowy owls in the US was funded by the Project SNOWstorm (www.projectsnowstorm.org). Funding for rough-legged buzzard work in North America was provided by Kidd Biological Inc, Alaska Department of Fish and Game and Arctic Raptors Inc and HWA Wildlife Consulting. Work on peregrine falcons in North America was funded by The Offield Family Foundation, Wolf Creek Charitable Foundation and The Peregrine Fund. Work on peregrine falcons in Yamal, Russia was funded by the Environment Agency-Abu Dhabi and coordinated by International Wildlife Consultants (UK) Ltd. The rest of the project was funded by the Max Planck Institute for Animal Behavior, Hawk Mountain Sanctuary, the Canada Research Chair program, and the Slovene Human Resources Development and Scholarship Fund. This study is part of the PhD project of TC, with committee members KS, MW, NL, JFT, and IP.
The authors declare no competing interests.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Curk, T., Pokrovsky, I., Lecomte, N. et al. Arctic avian predators synchronise their spring migration with the northern progression of snowmelt. Sci Rep 10, 7220 (2020). https://doi.org/10.1038/s41598-020-63312-0
This article is cited by
Autumn migration phenology of polar bears (Ursus maritimus) in Hudson Bay, Canada
Polar Biology (2022)
By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.