reproductive output but a long lifespan as adults. They return to their survival habitats in an early migration as soon as possible after breeding, postponing moult until they are in the survival habitats. In contrast, birds with a surplus of breeding resources, like e.g. many dabbling ducks or songbirds exploiting the huge peak of summer resources at northerly wetlands and forests, start breeding at an early age with high reproductive output during a shorter lifespan. They often moult in the breeding areas and depart late to their survival areas. Migration is a widespread plastic trait among animal groups, showing a very weak phylogenetic signal and becoming rapidly lost and regained among species and populations. Artificial selection has demonstrated that complete migration or residency may develop from partial migration within only a few generations. It is likely that migration in many species has been expressed and suppressed several times in association with the repeated glaciations during the Quaternary period. Glaciation cycles have occurred much more frequently than speciation events among for example birds, so many bird species are likely to have undergone repeated changes in distribution ranges and migration habits. Migration to a large degree builds on the reinforcement of traits that are already used by stationary animals, like locomotion, orientation and energy deposition. This gives regulatory genes a key role for expression/suppression of migration (although nocturnal migratory flights among otherwise diurnal birds may be a specialized circadian behaviour linked to migration). The key genes for regulation of migration remain to be discovered. Given the different migration segments and living stations at different places it is a difficult task to unravel where and when the most limiting conditions occur during the annual cycle. Population studies of migratory songbirds in North America have indicated that most mortality occurs during the migration periods, although these are much shorter than the stationary breeding and winter periods. A similar situation holds for migratory raptors, for which satellite tracking has revealed where and when the birds die (Figure 4). There is great concern that, although migration may often lead to abundance, it may also incur increased vulnerability. If one of the links in the chain of living stations during the annual cycle is damaged, the whole migration system may collapse. Migration is a powerful illustration of the adaptive complexity of life on Earth. Small animals travel tens of thousands of kilometres using a perplexing variety of behavioural, sensory and physiological adaptations to successfully complete their annual circuits. This way of life has consequences not only for the life history and population dynamics of the migratory animals but on a larger scale also for exchange between different ecosystems through transport and trophic effects mediated by the migratory animals. The large amounts of biomass that are moving and being transported in different migration systems, like six million tons of spawning herring in the Norwegian Sea (leaving behind more than a million tons of reproductive output at the coast), three hundred thousand tons of wildebeest in the Serengeti migration system, almost equally much of desert locusts in the locust migration systems of Africa and Asia, and fifty thousand tons of mainly smaller birds in the Palaearctic-African migration system, corroborate the significance of migration in the global ecology. Exploring migration makes us see more clearly the magnificence in biological evolution and what is at risk for the future. FURTHER READING Bauer, S., and Hoye, B.J. (2014). Migratory animals couple biodiversity and ecosystem functioning worldwide. Science 344, 1242552. Chapman, J.W., Reynolds, D.R., and Wilson, K. (2015). Long-range seasonal migration in insects: mechanisms, evolutionary drivers and ecological consequences. Ecol. Lett. 18, 287 302. Dingle, H. (2014). Migration. The Biology of Life on the Move. 2 nd Edition. (Oxford: Oxford University Press). Hansson, L.-A., and Åkesson, S. (eds). (2014). Animal Movement across Scales (Oxford: Oxford University Press). Hays, G.C., and Scott, R. (2013). Global patterns for upper ceilings on migration distance in sea turtles and comparisons with fish, birds and mammals. Funct. Ecol. 27, 748 756. Hein, A.M., Hou, C., and Gillooly, J.F. (2012). Energetic and biomechanical constraints on animal migration distance. Ecol. Lett. 15, 104 110. Milner-Gulland, E.J., Fryxell, J.M., and Sinclair, A.R.E. (2011). Animal Migration a Synthesis (Oxford: Oxford University Press). Newton, I. (2008). The Migration Ecology of Birds (London: Academic Press). Wilcove, D. (2008). No Way Home. The Decline of the World s Great Animal Migrations (Washington DC: Island Press). Zink, R.M. (2011). The evolution of avian migration. Biol. J. Linn. Soc. 104, 237 250. Department of Biology, Lund University, Sweden. *E-mail: Thomas.Alerstam@biol.lu.se Primer Marine migrations Nathan Putman There is great diversity in the animal species that migrate, the biomechanics that propel their locomotion and the ecosystems through which they transit. This diversity, however, is unified by a common condition: the relative suitability of places changes in predictable and cyclical ways. Owing to the periodicity of environmental change (e.g., seasons) and animal life-cycles (e.g., growth and maturation) locations become favorable, lose favorability and become favorable again at somewhat regular intervals. Migratory animals have adapted to these predictable fluctuations by moving among locations, sometimes over extraordinary distances (Figure 1). In marine ecosystems, currents, tides and waves seem to keep the ocean s inhabitants in constant motion. An animal living in the water column must actively orient and swim whether it wants to maintain its position or travel to a foraging area that is a thousand kilometers distant. Given their near-constant susceptibility to displacement, marine animals have evolved diverse adaptations to direct their movements. In contrast, in terrestrial ecosystems animals at rest tend to stay at rest; energy and volition are required for movements. This key difference suggests that the evolutionary pressures acting on migratory tendencies may fundamentally differ between species that move in the ocean and those that travel by land or air. Evolutionary considerations An attractive hypothesis is that the crisscrossing of ocean basins by animals as diverse as whales, tunas, eels and turtles evolved as simple extensions of more routine movements (Figure 1). For instance, Kemp s ridley sea turtle (Lepidochelys kempii) is found throughout the Gulf of Mexico, along the eastern US coast, and occasionally sighted in European waters. Despite this broad distribution, most Kemp s ridley sea turtles nest in the western Gulf of Mexico, with more than 90% along a 100 km stretch R972 28, R952 R1008, September 10, 2018 2018 Elsevier Ltd.
of beach in Tamaulipas, Mexico (Figure 2A). Hatchling turtles dig out from their nests, scramble down the beach and into coastal waters, where little turtles are surrounded by a sea of predators. Turtles swim offshore to escape predation, but this behavior puts them at the mercy of currents and they disperse widely (Figure 2B). Across the ocean, some areas tend to be more favorable to turtles than others. For instance, phytoplankton, the base of marine food webs, is consistently high along oceanic fronts and low at the centers of gyres. However, both fronts and gyre centers are areas where passively drifting organisms would tend to accumulate. During the turtle s dispersive stage, directed swimming, even at seemingly small speeds, can increase the probability of individuals reaching favorable locations and thus increase their fitness. As the turtles grow, predation risk decreases, and moving into higher-productivity coastal regions allows increased growth. Turtles at higher latitudes seasonally migrate offshore or southward to avoid cold temperatures in shallow waters, which can be lethal. Upon reaching maturity, Kemp s ridley shuttle between foraging grounds and their natal site to reproduce, where the adults can be reasonably assured to find mates and that environmental conditions will be suitable for their offspring (Figure 2). Thus, although the simple strategy of hatchlings swimming offshore to avoid predation introduced a variety of other problems, those challenges were met in the same way, through directed swimming that eventually returned individuals to their home beach. Similar behavioral patterns are common among marine migrants (Figure 1) and can largely be explained by a series of ontogenetic shifts in habitat to avoid predation, maximize growth and assure the needs of offspring will be met. These simple movement subroutines result in a complex pattern of migration, the importance of which is clearly shown in the spatial ecology of diverse marine species (Figures 1 and 2). Suboptimal movements As technology has advanced, it has become possible to track the movements of small animals through the ocean via satellite and other European eel Kemp s ridley sea turtle Loggerhead sea turtle (Brazil) Loggerhead sea turtle (South Pacific) Humpback whale (Southern Hemisphere) African penguin White shark (Australia) Bluefin tuna (North Pacific) Leatherback sea turtle (Indo-Pacific) Chinook salmon (Oregon/Washington) Figure 1. Migration movements of iconic marine migrants. Arrows depict the movements of well-known marine migrants from reproductive grounds to foraging areas. For many species, the reproductive sites are located where consistent ocean circulation features favor transport to nursery habitat. Upon reaching maturity, these animals return to their natal site to reproduce. It has long been recognized that a map-like sense is required for such homing, but the sensory basis of this ability remains a matter of speculation. For most species, the movement of juveniles to nursery habitat was thought to depend entirely upon ocean currents. More recently, experiments with sea turtles, salmon and eels have shown juveniles possess simple maps, based on Earth s magnetic field, that extend across their oceanic ranges. Despite lacking migratory experience, these naïve navigators appear to use the maps to direct their movement towards favorable habitat. telemetry methods. Unfortunately, most studies do not explicitly consider the contribution of ocean currents to net movement, which clouds their interpretation. However, some generalizations can be made: movement tends to be directional rather than random; ocean currents contribute to but do not fully account for movement; ocean temperature explains relatively little variation in movements; animals often follow consistent routes, despite differences in ocean conditions from one year to the next; and those consistent routes lead to areas that historically have tended to be favorable. Though strictly canalized migration routes aren t the only behaviors displayed in the ocean, the species that forage in ephemeral hotspots of productivity tend to be large-bodied and strong swimmers, such as adult leatherback turtles (Dermochelys coriacea) or basking sharks (Cetorhinus maximus). A reasonable interpretation is that if these animals incorrectly anticipate where the foraging hotspot will be, they have the swimming ability and energy reserves needed to search out another area. In contrast, to take advantage of consistent features of ocean circulation and productivity, weaker swimmers may need to adhere to the migratory route sculpted over evolutionary history. For instance, irrespective of differences in ocean currents and productivity from one year to the next, juvenile Pacific salmon will swim along a similar route from estuaries to ocean foraging areas. The oceanic regions these fish target have on average favorable temperatures and high food availability, but in some years, this behavior results in young salmon steadily swimming past areas of higher productivity than where they are traveling. Years of large salmon runs may be tied to whether, as juveniles, their migration coincided with high productivity waters. Similar suboptimal behavior is seen in penguins. Year after year, juvenile African penguins (Spheniscus demersus) perform a clockwise migration from southern Africa toward Namibia to nursery grounds. Historically, this migration allowed young penguins to forage on sardine and anchovy. However, intense fishing and changing ocean conditions have shifted fish spawning aggregations southeastward. Rather 28, R952 R1008, September 10, 2018 R973
A B 30 N 27 N 24 N 21 N 29 N 27 N 25 N 23 N 21 N 19 N 99 W Climatic suitability of nesting habitat Daily transport distance (km) 5 15 25 35 45 Nesting sites Major Excellent Good Moderate/Marginal Minor 97 W 95 W 93 W 91 W 89 W 18 N 98 W 93 W 88 W 83 W Figure 2. Migration shapes distribution in Kemp s ridley sea turtle. (A) Approximately 94% of all Kemp s ridley nesting occurs along the coast of Tamaulipas, Mexico (white stars), 4% occurs in Veracruz, Mexico (white squares) and 1.5% occurs in Texas (white circles). Climate suitability models indicate that favorable nesting habitat exists across the western Gulf of Mexico and thus does not account for spatial variation in nest abundance. The colored band shows average daily transport distance of virtual particles tracked within the surface layer of the Gulf of Mexico Hybrid Coordinate Ocean Model (1 June 1 August 2003 2010). Nesting is centered on the area where offshore transport is most favored. Given that sea turtles return to their natal site to nest, this implies that population sizes of turtles are closely tied to the initial migration of offspring (climatic suitability metrics modified from Pike (2013) Glob. Ecol. Biogeogr. 22, 555 566). (B) Ocean currents shape the dispersal routes (black arrows) of young Kemp s ridley to nursery habitat (purple shading). Circular histograms show the swimming directions of 20 young turtles tracked by satellite (Putman and Mansfield 2015). East west swimming was detected across the northeastern Gulf of Mexico, concentrating turtles in areas of preferred habitat. Northward swimming in the southeastern Gulf of Mexico counteracts the Loop Current, which would otherwise transport turtles into the Atlantic. Thus, the reproductive sites selected by adults, ocean currents, and swimming behavior of juveniles contribute to the ability of small animals to target favorable habitat. than tracking the change in fish distribution, the young penguins from western populations continue to migrate northward where the fish used to be. In those penguin populations, the number of breeding pairs has precipitously declined over the past decade, likely owing to the mismatch between the juvenile migration and prey availability. Where am I? A key question emerging from these findings is how might these animals follow consistent migration routes through a seemingly featureless ocean? In most cases, the animals are unaccompanied by experienced conspecifics a notable exception being young whales which migrate with their mothers while nursing. The question is related to another posed by ecologists: how do animals achieve reproductively closed populations in the open marine environment, where currents seem to promote homogenization? To achieve these remarkably consistent migratory routes and to return with precision to foraging and reproductive sites animals appear to require a map and compass sense (Figure 1). The map allows an animal to assess where it is relative to a target, while the compass allows it to maintain a heading. The map might also be used to correct for errors that accumulate during the animal s transit, for instance due to unfavorable ocean currents or temporarily fleeing a potential predator, and to tell the animal it has arrived at its destination. While much progress has been made for decades on animal compasses, the basis of the map has been more difficult to determine. While still an area of active research and debate, an accumulating number of marine migrants have been shown to derive map information by detecting certain aspects of Earth s magnetic field (Figure 3). The magnetic map of animals has at times been described as nature s GPS. Though intended as a convenient shorthand, to some this metaphor may imply navigational precision that is incompatible with neural processing abilities of animals and magnetic noise in the environment resulting from ocean currents, atmospheric tides, and crustal iron deposits. The GPS metaphor of magnetic maps might be better viewed with this scenario in mind: a person flew into an unfamiliar city in the middle of the night and rented a car that came with an out-of-date GPS device. The person programmed the location of a hotel into the GPS, kept her eyes on the unfamiliar road and followed the spoken instructions of the GPS. She took a more circuitous route than necessary (a new more direct route opened after the device s software was last updated). She drove through a tunnel and briefly lost satellite reception just before she needed to make a turn. She overshot the hotel, but the GPS regained satellite contact and after circling the block she eventually locates the hotel by noticing its sign. Positional information was essential to this task, but a cognitive representation of where she was or how she d gotten there was not. Of course, assuming this person remained in the city, she might start to pair her new experiences of navigating the area with the old GPS information and build up a more sophisticated and precise way to efficiently make her way through the city. Magnetic maps The (sub)cellular machinery responsible for transducing magnetic information to the nervous system remains enigmatic, but lab-based experiments that elicit behavioral responses are unequivocal. Electric current running through a wire generates a magnetic field proportional to the amperage; careful arrangement of a system of coiled wires allows precise manipulation of uniform magnetic fields around animals (Figure 3E). Animals can be exposed to magnetic displacements whereby the orientation of animals is recorded in response to magnetic fields that exist along their oceanic migratory route. Such experiments have shown that lobsters, salmon, trout, eels and sea turtles use map information from Earth s magnetic field to orient. Simulating the observed orientation within an ocean circulation model shows that the directions selected increase the probability of movement toward subsequent developmental habitat by taking advantage of consistent ocean currents. Experiments with salmon and turtles indicate that migratory experience is not needed to extract map information from the magnetic field, implying that the map is largely innate. This stands in stark contrast to R974 28, R952 R1008, September 10, 2018
migratory birds, which appear capable of using magnetic map cues only after their first migration. Combined with these laboratory studies, analyses of field data suggest an important role of magnetic cues as a driver in marine species distributions. For instance, Kemp s ridley has the most restricted nesting distribution of any sea turtle (Figure 2A). The precision by which these turtles return to the main nesting site can be explained by the rate of geomagnetic field drift over the past 400 years and a simple model of magnetic imprinting, i.e., upon reaching maturity returning to approximately the same magnetic field values in which the animal hatched. In the case of loggerhead sea turtles (Caretta caretta), which have a circumglobal nesting range, regional variation in population structure can be predicted by the similarity of magnetic fields among nesting beaches. Along the Florida peninsula, loggerheads nesting on beaches with similar magnetic fields (but large geographic distances between them) are more genetically similar than are those nesting on beaches with more different magnetic fields (but which are separated by short geographic distances). Such patterns are expected if turtles are primarily using magnetic cues to assess location, as homing mistakes would be more likely among sites characterized by similar magnetic fields. Such relationships between distribution and the geomagnetic field are not restricted to turtles. Spatiotemporal variation in migration routes of pink (Oncorhynchus gorbuscha) and sockeye (O. nerka) salmon homing from the North Pacific to the Fraser River in British Columbia, Canada can be predicted by drift of the geomagnetic field. Over >50 years of monitoring, changes in ocean temperature and currents also provided some explanatory power for shifts in the sockeye migratory routes, but the only significant predictor of changes in pink salmon migratory routes was drift of the geomagnetic field. Given the diversity of marine animals that appear to use magnetic cues to derive positional information, careful consideration of this feature of the global environmental might reveal a surprising degree of similarity in the ecological patterns shaped by migration. A D 67 N 66 N 65 N 64 N 63 N 25 W 21 W Horizontal component Total field intensity Inclination angle Surface of the earth Next steps As global climate continues to change and as humans further modify habitats, there is growing need to predict longterm trends in animal movements, distribution and connectivity among ecosystems. Resource managers need robust predictions about how animals respond to changing environmental conditions when making decisions, such as regarding the operation of fisheries, marine energy extraction or shipping, to minimize negative impacts on species. To generate this information, a mechanistic approach that explicitly considers the interaction between the animal and environment, mediated by sensory B 51.6 51.9 52.2 52.5 52.8 17 W 13 W 53.1 53.4 Vertical component C 65 N 45 N 25 N 80 W 60 W 40 W 20 W 0 20 E E 35.0 37.5 40.0 42.5 45.0 47.5 50.0 52.5 55.0 57.5 Figure 3. The map-like characteristics of Earth s magnetic field. (A) Field lines (arrows) intersect Earth s surface forming the inclination angle. At the magnetic equator (curving line across the Earth) the inclination angle is parallel to Earth s surface (0 ) and it becomes steeper towards the magnetic poles (90 ). (B) The inclination angle can be resolved into two vector components: the horizontal field intensity (points toward magnetic north) and the vertical field intensity (points up or down, depending on hemisphere). These two components sum to the total field intensity. (C) Intensity and inclination form a bicoordinate grid that allows animals to distinguish among locations and have an indication of their current position relative to another magnetic target. This map of total field intensity (color-gradient in µt) and inclination (black contours = 5 ) is based on the International Geomagnetic Reference Field 11, which depicts Earth s main-dipole field. (D) Magnetic field parameters across Iceland (light grey outline) based on the Enhanced Magnetic Model show that substantial heterogeneity in map parameters can occur in areas with magnetic anomalies (e.g., iron deposits). Colored bands indicate the complex local gradients of total field intensity (in µt). Black lines indicate the gradient of inclination angle (contours = 0.5 ). Along the northern coast of Iceland, some locations near each other are less magnetically similar to each other than to those further away. This region (and others like it around the world) may provide locations where natural experiments can explore population-level implications of magnetic navigation. (E) A magnetic coil system consisting of wooden frames (~3 m x ~3 m) wrapped with copper wires and connected to two independent DC power supplies that precisely control the horizontal and vertical components of the magnetic field at the central platform. This coil system in Wales, UK demonstrated a magnetic map sense in European eels by exposing young eels to magnetic displacements fields that existed along their North Atlantic migratory route (Figures 1 and 3C). (Photo Jess Stephenson.) (A,B) from Lohmann et al. (2007). processing and behavior, is likely more sensitive than traditional statistical or mapping techniques to predict shifts in distribution and abundance in response to changing environmental conditions. To this end, efforts are needed to identify the hierarchy of cues animals use for movement decisions. Further developments in telemetry instrumentation are needed, whereby measurements of the animal s ground speed are recorded concurrently with surrounding waterflow to derive the behavioral and environmental contributions to net movement (Figure 2B). Tracking experiments with an emphasis on experimental design and hypothesis 28, R952 R1008, September 10, 2018 R975
testing should be prioritized over metanalyses of opportunistic tag deployments. Likewise, moving beyond correlation and identifying the sensory cues used by animals to make orientation decisions must involve wellcontrolled laboratory experiments. Ultimately, to be of practical use, information on animal movement must be linked to population and ecosystem-level assessments. A promising avenue of research is to integrate sensory ecology, environmental models and individualbased models to simulate animal movements. Correlating the simulation s emergent patterns to real-world observations can test hypotheses of how individual movements shape population-level processes. Comparison to passive markers that infer an animal s previous movements, including analyses of stable isotopes, trace elements and genetics are particularly valuable. Those movement models that produce well-supported predictions can serve as inputs for ecosystem models. The ecosystem models can then assess the implications of these movements (shifts in biomass as physiological and behavioral responses to a dynamic environment) by coupling them to a food web and species-specific population dynamics models. Scenarios in which the model environment is disturbed can be used to predict responses to anthropogenic habitat alterations and environmental changes. Though computationally feasible, such an approach is not trivial. Disentangling the use of multiple sensory cues for orientation in different environmental contexts and ontogenetic stages is timeconsuming. However, the potential power to make population and ecosystem-level predictions based on a mechanistic understanding of animal movement argues it is worth the investment. Focusing attention on species with life-stages that are readily studied in the laboratory and field whilst displaying the full repertoire of marine animals movements may be a tractable way forward. Likewise, sensitivity analyses of movement and ecosystem models can determine which data gaps in model inputs result in the greatest uncertainty of predictions and are therefore most important to study. From these efforts, scientificallysound management decisions related to habitat modifications and the prioritization of conservation efforts can be achieved. FURTHER READING Brothers, J.R., and Lohmann, K.J. (2018). Evidence that magnetic navigation and geomagnetic imprinting shape spatial genetic variation in sea turtles. Curr. Biol. 28, 1325 1329. Burke, B.J., Anderson, J.J., Miller, J.A., Tomaro, L., Teel, D.J., Banas, N.S., and Baptista, A.M. (2016). Estimating behavior in a black box: how coastal oceanographic dynamics influence yearling Chinook salmon marine growth and migration behaviors. Environ. Biol. Fishes 99, 671 686. Gaspar, P., Georges, J.Y., Fossette, S., Lenoble, A., Ferraroli, S., and Le Maho, Y. (2006). Marine animal behaviour: neglecting ocean currents can lead us up the wrong track. Proc. Biol. Sci. 273, 2697 2702. Grimm, V., Ayllón, D., and Railsback, S.F. (2017). Next-generation individual-based models integrate biodiversity and ecosystems: yes we can, and yes we must. Ecosystems 20, 229 236. Harden-Jones, F.R. (1968). Fish Migration (London: Edward Arnold). Hays, G.C., Ferreira, L.C., Sequeira A.M.M., Meekan, M.G., Duarte, C.M., Bailey, H., Bailleul, F., Bowen, W.D., Caley, M.J., Costa, D.P., et al. (2016). Key questions in marine megafauna movement ecology. Trends Ecol. Evol. 31, 463 475. Hays, G.C. (2017). Ocean currents and marine life. Curr. Biol. 27, R470-R473. Hazen, E.L., Jorgensen, S., Rykaczewski R.R., Bograd, S.J., Foley, D.G., Jonsen, I.D., Shaffer, S.A., Dunne, J.P., Costa, D.P., Crowder, L.B., et al. (2013). Predicted habitat shifts of Pacific top predators in a changing climate. Nat. Clim. Change 3, 234. Lohmann, K.J., Lohmann, C.M., and Putman, N.F. (2007). Magnetic maps in animals: nature s GPS. J. Exp. Biol. 210, 3697 3705. Naisbett-Jones, L.C., Putman, N.F., Stephenson, J.F., Ladak, S., and Young, K.A. (2017). A magnetic map leads juvenile European eels to the Gulf Stream. Curr. Biol. 27, 1236 1240. Putman, N.F., and Lohmann, K.J. (2008). Compatibility of magnetic imprinting and secular variation. Curr. Biol. 18, R596 R597. Putman, N.F., and Mansfield, K.L. (2015). Direct evidence of swimming demonstrates active dispersal in the sea turtle lost years. Curr. Biol. 25, 1221 1227. Putman, N.F., Jenkins, E.S., Michielsens, C.G., and Noakes, D.L. (2014). Geomagnetic imprinting predicts spatio-temporal variation in homing migration of pink and sockeye salmon. J. R. Soc. Interface 11, 20140542. Putman, N.F. (2015). Inherited magnetic maps in salmon and the role of geomagnetic change. Integr. Comp. Biol. 55, 396 405. Secor, D.H. (2015). Migration Ecology of Marine Fishes (Baltimore: Johns Hopkins University Press). Sherley, R.B., Ludynia, K., Dyer, B.M., Lamont, T., Makhado, A.B., Roux, J.-P., Scales, K.L., Underhill, L.G., and Votier S.C. (2017). Metapopulation tracking juvenile penguins reveals an ecosystem-wide ecological trap. Curr. Biol. 27, 563 568. LGL Ecological Research Associates, Inc., Bryan, TX 77802, USA. E-mail: nathan.putman@gmail.com Primer Collective animal migration Iain D. Couzin Migratory movement is a strategy employed by a broad range of taxa as a response to temporally and spatially varying environmental conditions. Multiple factors can drive animal migration, including: movement to hospitable environments when local conditions become unfavourable (such as to reduce nutritional and thermoregulatory stress); movement to find mates and/or breeding sites; and movement to minimise competition, predation, infection or parasitism. Migrating animals can often be seen to move together (Figure 1), sometimes in vast numbers. Despite this, the social aspects of migration have, to date, received very limited attention. Synchronisation of migratory behaviour among organisms does not itself imply that migrants utilise social information: synchrony is inevitable if there are relatively short windows of opportunity in which to move, or if there exist sudden environmental changes to which a response is necessary. However, as will be outlined here in this Primer, there is growing evidence that many migratory animals do utilise social cues and that collective factors could shape migration in a variety of important ways. Information and collective migration Genetic and sensory information It is well-established that migratory timing (i.e. when to move), and at least the initial direction of travel, are in part genetically based in many seasonal migrants. In some species, there is evidence that individuals possess what is effectively a map-like sense, from which the target area of migration can be inferred (even from magnetic cues alone), and a compass sense that is employed to steer in the appropriate direction en route. In addition, swimming or flying organisms exhibit mechanisms to exploit, and where necessary compensate for, the complex movement of the medium through which they move. R976 28, R952 R1008, September 10, 2018 2018 Elsevier Ltd.