Migratory Animals Couple Biodiversity and Ecosystem Functioning Worldwide

See allHide authors and affiliations

Science  04 Apr 2014:
Vol. 344, Issue 6179, 1242552
DOI: 10.1126/science.1242552

Structured Abstract


Every year, billions of migratory animals cross the planet in pursuit of increased foraging opportunities, improved safety, and higher reproductive output. In so doing, these migrants transport nutrients, energy, and other organisms (including seeds, mollusks, parasites, and pathogens) between disparate locations. Migrants also forage and are preyed upon throughout their journeys, thereby establishing transport and trophic interactions with resident communities. Migratory animals thus couple ecological communities across the globe and may mediate their diversity and stability. However, as yet, the influence of migrants and their services on these communities is often overlooked, and as a consequence of global changes, migrations are threatened worldwide.

Embedded Image

Migrants change ecology and ecosystems. By transporting energy, nutrients, and other organisms, as well as foraging and becoming prey, migratory animals can substantially alter the dynamics of resident communities that they connect on their journeys across the globe. We illustrate key examples where migratory species profoundly alter food-web dynamics, community processes, and ecosystem functioning, indicating that migrants represent a unique, yet highly influential, component of biodiversity.


We review several examples in which interactions between migratory animals and resident communities have been quantified, illustrating the processes by which migrants may uniquely alter energy flow, food-web topology and stability, trophic cascades, and the structure and dynamics of (meta-)communities. For example, the inputs of nutrients and energy originating from distant localities by migrants can dramatically increase resource availability, with rippling consequences for productivity at various trophic levels and the potential to drive the transition between alternative stable states. Migrant-mediated transport of propagules of other organisms can lead to the establishment of new or lost species, as well as influencing gene flow and genetic mixing among resident populations. Similarly, migrants can alter parasite transmission, diversity, and evolution by harboring a broader range of parasites than residents and by either facilitating or hindering the long-distance dispersal of parasites.

Foraging by migrants can also have profound effects on community processes and ecosystem functions. For instance, grazing by migratory animals can alter nutrient cycling, primary productivity, biomass of edible plants, competitive interactions between plant species, and ultimately, the composition and long-term persistence of the entire plant community. The most striking difference between migrant and resident consumers is, however, the pulsed nature of migrant utilization and the timing of their interactions. Together, these fundamentally define the relationship between migrant abundance and primary production (in the case of migrant herbivores) or the stability of food webs (in the case of migratory predators).


Our Review demonstrates that the highly predictable, seasonally pulsed nature of animal migration, together with the spatial scales at which it operates and the immense number of individuals involved, not only set migration apart from other types of movement, but render it a uniquely potent, yet underappreciated, dimension of biodiversity that is intimately embedded within resident communities. Given the potential for migration to influence ecological networks worldwide, we suggest integrative network approaches, through which studies of community dynamics and ecosystem functioning may explicitly consider animal migrations, understand the ramifications of their declines, and assist in developing effective conservation measures.


Animal migrations span the globe, involving immense numbers of individuals from a wide range of taxa. Migrants transport nutrients, energy, and other organisms as they forage and are preyed upon throughout their journeys. These highly predictable, pulsed movements across large spatial scales render migration a potentially powerful yet underappreciated dimension of biodiversity that is intimately embedded within resident communities. We review examples from across the animal kingdom to distill fundamental processes by which migratory animals influence communities and ecosystems, demonstrating that they can uniquely alter energy flow, food-web topology and stability, trophic cascades, and the structure of metacommunities. Given the potential for migration to alter ecological networks worldwide, we suggest an integrative framework through which community dynamics and ecosystem functioning may explicitly consider animal migrations.

Migration Monitor

Seasonal migrations move large numbers of animals across often vast distances. Such movement shifts large amounts of biomass from one region to another, but, perhaps more importantly, moves animals that eat, excrete, and sometimes die in multiple remote systems. Such movements impact the communities, trophic structure, and function of these ecosystems in often underappreciated ways. Bauer and Hoye (10.1126/science.1242552) review migrations across taxa to identify the key ecological roles these long-distance movements play, and the unique threats the animals face in our increasingly modified world.

Migration is the story of life on the move. Each year, billions of animals—from butterflies weighing less than a gram to 40-ton whales—fly, walk, or swim their way across the planet in pursuit of improved foraging conditions, safety, and reproductive opportunities (1, 2). Migrations are persistent, directional movements from one destination to another, uninterrupted by intervening resources (2). The distances covered are often astounding, yet, the most extraordinary aspects of migration are perhaps the ubiquity of the phenomenon and the abundance of individuals involved. For instance, an estimated 1855 bird species (19% of extant species) are migratory (3). Of the avian species that use terrestrial and freshwater habitats, roughly 45% of those breeding in North America undergo seasonal migration (4), and over 30% breeding in the Palearctic migrate to sub-Saharan Africa (4), with many more migrating within Europe. Although accurate estimates of the number of individuals involved are scant, more than 2 billion passerine birds are found to migrate to sub-Saharan Africa (5), whereas upwards of 3 billion insects migrate over any 1-km stretch of countryside in southern England (6).

Historically, research on animal migration has focused on the migrants themselves: how, when, where, and why animals migrate (7). However, it is increasingly recognized that migration has ecological effects that pervade resident communities. The mere presence of migrants means that they will be preyed upon and forage and, in so doing, acquire nutrients, energy, and hitchhiking organisms that are subsequently transported to future destinations. Although migration patterns vary considerably across species, the large spatial scales traveled, together with their seasonality, which generate pulsed, highly predictable interactions, critically differentiate migration from other types of movement. Moreover, by integrating resource peaks or avoiding periods of heightened mortality risk over time and space, migrants may sustain considerably larger populations than otherwise similar resident species, often by an order of magnitude (8). Collectively, these features suggest that migratory animals have the potential to uniquely alter community structure, dynamics, and ecosystem function along their routes.

Notwithstanding recent research on spatial linkages between habitats and ecosystems (9), little work has been devoted specifically to the role of animal migration in ecological networks. This knowledge gap may, in part, stem from the inherent difficulty of the task, in addition to historical legacies within the disciplines of community ecology and migration ecology. Yet, neglecting migrants in community and ecosystem ecology restricts our understanding by reducing the number of species considered and failing to account for both spatial coupling and temporal dynamics.

Here, we review several case studies in order to define fundamental processes by which the transport and trophic effects of migratory animals influence ecological networks (Fig. 1). These examples illustrate the potential for migrants to uniquely alter energy flow, food-web topology and stability, trophic cascades, and the structure and dynamics of metacommunities (Fig. 2). In the hope of fostering integrative research that incorporates the regular, directed, mass movements of migrants in ecological networks, we synthesize key concepts from community ecology that may be readily adapted to incorporate migrant processes, particularly resource pulses and spatial subsidies. Looking to the future, we discuss novel network approaches for investigating the relative importance of migrants in maintaining ecosystem diversity, resilience, and stability that will also provide valuable insight into the consequences of rapid declines of animal migrations and the cascading effects of management actions along migratory routes.

Fig. 1 Schematic overview of the interactions between migrants and the multiple resident communities they visit during their annual or life-cycles (purple arrows).

These may influence demographic rates of resident populations directly (solid arrows) and indirectly (dashed arrows) through interactions among residents (green arrows).

Fig. 2

Key examples of migrant-induced changes to the dynamics of resident community processes. [Photo credits in order of appearance: J. B. Armstrong, B. J. Hoye, A. K. Glover, M. D. Tuttle (Bat Conservation International), J. Waldenström, B. J. Hoye, bikeriderlondon (, S. Uryadnikov (, J. Brodersen, D. White Jr., and J. Bêty].

Transport and Trophic Effects of Migrants

Transport Effects: Nutrients, Energy, and Toxicants

Upon arrival at a site, migrants deposit nutrients, energy, and other substances into resident communities and ecosystems via excreta (e.g., feces and urine), reproductive material (such as eggs), or the dead bodies of migrants themselves (Fig. 1). These allochthonous subsidies (resources that originate in one habitat but are moved into another) can result in a net inflow of energy and nutrients (10). Although allochthonous inputs are likely to occur in most (if not all) migratory systems, quantitative estimates of these subsidies—and their consequences—are primarily limited to the transport of nutrients and energy by oceanic migrants returning to coastal ecosystems (Fig. 2). For instance, the migration of Norwegian spring-spawning herring (Clupea harengus) constitutes the world's largest flux of energy effected by a single population, with the transfer of roughly 1.3 × 106 tons of biomass annually (11).

The migration of anadromous fishes, particularly salmon (Salmonidae), to their natal lakes and streams also constitutes a massive transfer of nutrients and energy from the ocean to freshwater ecosystems (12) and the surrounding terrestrial habitats (13). Salmon predictably increase both nitrogen and phosphorus in their spawning habitat, with estimates of 190% and 390% increases, respectively, in some streams (12). However, these inputs do not result in an increase in gross primary productivity; rather, migratory salmon increase ecosystem metabolism by up to threefold and, in so doing, force a switch from a net-autotrophic to a strongly net-heterotrophic metabolic state (12). Salmon also alter the physical properties of the stream, increasing air-water gas exchange by nearly 10-fold during peak spawning. Although nutrient transfer is not entirely unidirectional (as juveniles migrate to the ocean), an estimated 85% of the nutrients from carcasses or eggs are “stored” in the local food web (14), which alters the composition and structure of phytoplankton communities (15), riparian vegetation (including doubling the density of the overstory and halving the density of the understory) (16), and the phenology of stream insects (17).

Migratory seabirds also import vast quantities of nutrients to their terrestrial breeding colonies (18). Because these nutrients are accumulated over relatively small areas, they tend to result in relatively localized ecosystem changes (19). Seabirds may also introduce toxicants, such as organochlorines and heavy metals, which causes them to accumulate locally (20).

Because migratory populations may form enormous aggregations during certain periods, the allochthonous inputs of migrants can result in short pulses of high-intensity inputs (21) that can drive the transition between alternative stable states (12). Migrants can also exert kinetic energy on a resident ecosystem through their grubbing, foraging, or breeding activities that may fundamentally alter the physical properties of their habitat (22).

Transport Effects: Propagule Dispersal

Migratory movements also represent a unique dispersal mechanism for seeds, spores, and (parts of) other organisms (collectively “propagules”) across biogeographic barriers (23) (Fig. 2). In light of the importance of dispersal for population structure, adaptive capabilities, and evolutionary trajectories in theoretical studies (24), such long-distance dispersal events may be highly important for the (re-)colonization of unoccupied habitats, the recovery of lost populations, maintenance of gene flow, and gene mixing in metapopulations, even if they are relatively rare events (25). Yet, despite the potential importance of animal migrations and the wealth of information on local dispersal of propagules, migrant-mediated dispersal has been investigated in a surprisingly limited number of taxa, and rarely from the perspective of the recipient community.

A critical initial step in the dispersal process is the retention of viable propagules in or on migrants. Experimentally derived retention times, combined with field data on speeds of migratory movements, have been used to estimate the distances over which propagules could be transported, with estimates ranging from 200 to 1200 km for migratory water birds (25, 26). However, retention times in free-living animals could be substantially altered by physiological changes during migration (27).

Although these studies reveal the potential for long-distance propagule dispersal by migratory animals, evidence for (or against) migrant-mediated dispersal is lacking. Notably, long-distance movements are typically confined to a limited time period, which must align with the flowering, fruiting, or reproductive time of the dispersed organism(s) for dispersal to take place (28). Moreover, the survival of propagules during passage in the digestive tract (26) and through deposition, germination, and competition in the resident community must also be assessed in order to quantify the consequences of this dispersal for community dynamics (23) (Fig. 3).

Fig. 3 The transport and trophic effects that migratory animals exert on resident communities critically depend on the attributes of communities and ecosystems.

Migrants may also play an important role in propagule dispersal within resident communities. For instance, columnar cacti (Pachycereeae) are pollinated by numerous species migrating between western Mexico and southern Arizona, with Lesser long-nosed bats (Leptonycteris yerbabuenae) in particular shown to be responsible for up to 100% of cacti fruit-set, however their relative importance varies along their migratory route and between cacti species (29). Given that cacti represent crucial water and nutrient sources for a variety of desert animals (29), pollination services provided by these migrants facilitate the persistence of a large proportion of the ecosystem. In addition, migratory pollinators also mediate high levels of genetic diversity and gene flow between cacti populations (30).

Transport Effects: Parasite Dispersal

Although all animals can host a variety of parasites, migrants may play a unique role in parasite dynamics, both within and between resident communities [recently reviewed in (31)]. Migrations may facilitate the long-distance dispersal of parasites, as has been widely assumed for several zoonotic pathogens including Ebola virus (by migratory fruit bats), avian influenza viruses (by migratory water birds), and West Nile virus (by migratory songbirds), yet clearly documented examples are exceedingly rare (31). Migrants may also encounter a broader range of parasites than residents (32), which increases the likelihood of transmitting novel parasites to resident species (31). In addition, the intense physiological demands of migration (in some species) may trade off with immune responses, potentially rendering migrants more susceptible to infection. For instance, migratory birds (in captivity) show shifts in measures of innate immune responses during the migratory period (33). Although the link between these immune responses and susceptibility to primary infection is yet to be established, simulated migration can catalyze a relapse of prior infections (34). Many migrants also aggregate in groups many times larger than at other periods during the year at key refueling locations en route, which can enhance parasite transmission and result in prevalences that are as much as 17 times that in other locations (35).

The role of migrants in altering parasite transmission and diversity is, however, complicated by the infection process itself, particularly the timing of infection, the host’s ability to tolerate infection, and the availability of susceptible hosts (Fig. 3). Animals must become infected before departure from a site, migrate successfully while infected, and remain infectious until arriving at a subsequent site. Yet, in one of the few examples to date, monarch butterflies (Danaus plexippus) infected with a protozoan parasite exhibited 19% shorter flight distances and 10 to 16% slower flight speeds, and lost proportionately more body mass (2.5% as compared with 1.6% of starting body mass) for the distance flown (36). Similarly, Bewick’s swans (Cygnus columbianus bewickii) infected with low-pathogenic avian influenza virus delayed departure by up to 1 month and traveled shorter distances during the early stages of spring migration compared with uninfected individuals (37). Crucially, mechanistic models demonstrate that such infection-induced delays may result in spatial separation of infected and susceptible migrants, substantially dampening infection dynamics (38). Thus, although migrants may experience greater exposure to parasites and exhibit potential for long-distance dispersal, migration may also reduce infection risk (31).

Trophic Effects: Herbivory

The foraging of migrants en route establishes consumer-resource interactions across several sites. Similar to resident species, grazing by migratory animals has been shown to alter nutrient cycling, primary productivity, biomass of edible plants, competitive interactions between plant species, and ultimately, the composition and long-term persistence of the entire plant community (22, 39) (Fig. 3). Perhaps the most pervasive migratory herbivores are insects, many of which are the world’s major agricultural pests (2). For example, desert locusts (Schistocerca gregaria) can consume their own weight in vegetation each day, such that one swarm on the horn of Africa was estimated to have consumed, on a daily basis, enough vegetation to feed 400,000 people for an entire year (40). Although such onslaughts show little evidence for resource competition with nomadic livestock, individual localities can be completely devastated, with long-term socioeconomic ramifications in the human population (41).

The striking difference between migrant and resident herbivores, however, is the pulsed nature of migrant utilization, and consequently, the temporary carrying capacity of sites for migrants is larger than it would be for year-round residents (8, 10). Moreover, the timing of grazing by migrants relative to the timing of growth of primary producers fundamentally defines the relation between grazing intensity and primary production (39). For example, herds of up to 1 million migratory wildebeest (Connochaetes gnou) use the highly nutritious forage of the short grass plains for calving during the wet season; outside of this period, the wildebeest migrate to woodlands where plants are less abundant and of lower nutritional quality (42). Theoretical studies demonstrate that year-round grazing of the plains by an equivalent number of residents would result in herbivory-limited plant growth, reduced standing biomass, and a loss of excess nitrogen from the system, resulting in a long-term reduction in aboveground net primary productivity (39). However, because the timing of grazing by migrants is decoupled from the timing of plant growth, increased grazing intensity from migrants results in substantially higher primary productivity than that in an ecosystem devoid of their presence (39).

In addition to interactions with primary producers, migratory herbivores interact with resident herbivores. The outcome of migrant-resident interactions may, however, substantially differ from standard theoretical predictions. For instance, if resident species are limited by seasonal variation in resources, then during the seasons of plenty, resources are conceivably available in excess of the capacity of residents to consume them. As a result, a migratory subordinate may be able to exploit resources shared with an otherwise competitively dominant resident species, as is the case for the migratory ungulates on the plains during the wet season (10). Conversely, grazers that reside year-round in the woodland interact with migrants during the period of food scarcity, and the influence of migratory competitors may have synergistic negative effects (10).

The annual migration of wildebeest not only typifies the concept of spatially coupled food webs (43, 44) but illustrates coupling between communities whose dynamics operate at very different rates and with different energetic efficiencies (45). The short grass plains are ephemeral, boom-and-bust systems that occur for only a restricted period each year, despite providing the majority of the annual protein requirement of the migratory grazers (45). Thus, energy moves through the food web of the short grass plains at a “fast” pace. The woodlands show “slower” energy chains, with plant productivity less variable between seasons and, as a result, resident herbivores extending reproduction throughout a larger portion of the year (46). The migratory herds uniquely couple the fast resource chains of the plains to the slow resource chains of the woodlands (45).

Trophic Effects: Predation

Many seabirds, raptors, cephalopods, fish, and marine mammals are migratory predators, with the potential to exert top-down regulation on prey populations and resident communities through trophic cascades. For instance, numerous species of migratory birds and bats prey on insects, which reduces their local abundance and thereby limits insect outbreaks and damage to agricultural crops, as well as alters forest ecosystem dynamics (41, 47). Interactions between migratory predators and community processes are, however, exemplified in studies of aquatic systems (Fig. 2).

In addition to the iconic salmonid and eel migrations, numerous cyprinid fish undertake seasonal migrations in north-temperate environments—departing from lakes in autumn to spend the winter in streams and wetlands before returning in spring (48). When a large proportion of the fish population migrates, predation pressure on zooplankton is diminished, which facilitates rapid growth in the zooplankton population in spring and results in a high probability of a stabilized clear-water phase (48). Yet, without mass migration, predation pressure on zooplankton remains high during winter, which results in largely unrestricted phytoplankton growth in spring that can prevent the establishment of macrophytes and either destabilize the clear-water phase or catalyze a transition to a turbid state. Thus, these seasonal migrations may drive transitions between alternative stable states in lakes (48).

Migratory predators may also mediate indirect interactions between prey species in physically separated habitats, generating strong spatial patterns in trophic cascades (Figs. 1 and 3). For instance, the availability of prey in one location can, through its effect on predator abundance, indirectly influence the intensity of predation exerted at another location (10). Theoretical studies have demonstrated that the degree of spatial coupling, particularly by large mobile consumers, is overwhelmingly important to the stability of food webs (49). Weak to intermediate spatial coupling in particular—achieved when predators range over several scales of their prey and also switch prey preference in order to overcome spatial and temporal variation in prey abundance—can be potent stabilizers of food webs (49).

Trophic Effects: Migrants as Prey

Migrants not only use, but can also constitute resource peaks that may be exploited by resident predators along migration routes (10). For instance, migratory prey are an unmistakably important resource for lions (Panthera leo) in the Serengeti, who time their reproduction to coincide with wildebeest migration so faithfully that lions in the dry season range breed at a completely different time of year to lions in the wet season range (10) (Fig. 4). Similarly, reproduction in Eleonora’s falcons (Falco eleonorae) in the Mediterranean is coincident with the autumn migration of song birds, and Arctic foxes (Vulpes lagopus) experience higher reproductive rates as a result of the highly predictable subsidies from migratory geese, such that the probability of fox reproduction declined by more than 70% for every 10 km between the fox den and a goose breeding site (50). Yet, migratory prey need not be large-bodied to have a significant influence on resident ecosystems. Larvae of army cutworm moths (Euxoa auxiliaris) hibernate on the Great Plains, then migrate to the Rocky Mountains as newly emerged adults in spring, where they have been shown to form a substantive food source for grizzly bears (Ursus arctos horribilis) inhabiting talus slopes, both in terms of proportion of prey items (>95% of their ingested volume) and energy density (51).

Fig. 4 Migratory herbivores furnish a smorgasbord for resident lions at both ends of their migratory range in the Serengeti.


As migratory populations are larger than “equivalent” resident populations (8), they may represent a double-edged sword for resident species in terms of shared predators. Subsidized by migrants, resident predators may become more abundant than would otherwise be the case (52) and thus increase predation risk to residents when migrants are no longer present, as seen in lions and hyenas (Crocuta crocuta) that switch to resident species when the herds of migratory ungulates depart (46). Yet the presence of migrants may also provide residents with a temporal refuge from predation. The temporary presence of migrants, coupled with their influence on predator populations, therefore acts to sequentially enhance then reduce indirect competition with similar resident species (45).

Synthesizing Migrant Influences on Ecological Processes

As the foregoing examples illustrate, migrations are not simply the movement of animals. They are ecological processes embedded in, and interacting with, a complex web of resident communities. Resident communities both shape migration routes and phenology (1) and are, in turn, modified by the presence of migrants, the carryover effects of previous sites on migrants, and the legacy effects of migrant-resident interactions that persist long after migrants have departed. Systems that play host to migratory species may, therefore, profoundly differ from those that do not because of at least three features: (i) the timing, frequency, and predictability of migrations; (ii) the intensity of interactions; and (iii) the spatial scales over which they connect communities and ecosystems.

The timing of migration relative to resident phenology is fundamental to the strength and direction of migrant-resident interactions (53). For instance, migrants can only be important pollinators if their visits coincide with peak flowering (29). Similarly, if parasite prevalence shows a marked seasonal dynamics, transmission may be restricted to sites where high prevalence and migrant visitation coincide (54). The synergy between migrant and resident phenologies also exists when visitation and resource growth are out of phase, as evidenced by increased primary productivity of savannah woodland habitats through the grazing of migratory ungulates in the dry season.

The frequency of migrations and the immense number of individuals involved often mean that migrant inputs constitute “resource pulses,” defined as occasional, intense, brief episodes of increased resource availability that can profoundly alter demographic rates and abundances of interacting populations (55, 56). In systems where the frequency of these events is longer than resident breeding cycles, resource pulses may result in a consumer population overshooting its carrying capacity once the resource pulse has finished, potentially leading to a population crash, and even (local) extinction (55). The effects of resource pulses can therefore persist long after the pulse itself is extinguished and may fundamentally alter community structure via extinction, invasion, or transition to alternative stable states (55). Critically, the effect of a resource pulse is intensified when the pulse is either more temporally concentrated or allochthonous in origin (57), both of which are expected for communities visited by migratory animals (Fig. 3). Moreover, migrations are often highly predictable events and of similar periodicity to resident breeding cycles, such that migrations could be expected to enhance the demographic rates of residents without the dramatic population crashes associated with resource pulses with a periodicity of multiple years [e.g., seed mast (56)]. However, such predictions have yet to be explicitly tested using the available theoretical frameworks.

The strength of interspecific interactions may be of paramount importance to the influence of migratory species on resident communities. In general, weak interspecies interactions (i.e., low instantaneous rate of per capita population change in one species as a result of changes in the other) have been shown to dampen oscillations in the abundance of resources and consumers, which makes them potent stabilizers in complex food webs (43, 58). Several characteristics indicate that migrants may maintain many weak interactions. The presence of migrants in a resident community is pulsed, and although interactions may be of high intensity while migrants are present, their visits are both infrequent (usually once or twice per year) and short in duration (only a fraction of the annual cycle), such that they rarely constitute the sole consumers of local resources. Likewise, resident predators rarely rely solely on migrants as resource. Thus, not only are weak interactions generated between migrants and resident predators but, through prey-switching, there is a concomitant weakening of the interactions between resident predators and resident prey. It iis noteworthy that the strength of interactions maintained by migrants is likely to be predicated on both the migrants themselves (frequency and duration of presence, resources consumed) and the recipient community, particularly the abundance and diversity of other, ecologically similar species (Fig. 3). On a global scale, the importance of migrants in resident communities may therefore be expected to increase with increasing latitude, altitude, and aridity.

Finally, animal migrations often span distances that are orders of magnitude greater than foraging or dispersal movements, such that interactions between several disparate communities may be mediated by migratory species. For instance, spatial nesting—in which resident primary producers are increasingly coupled across several spatial scales by large mobile consumers that adaptively follow resource asynchronies across spatially distinct regions—is thought to be central to the stability of complex food webs (43, 49). This is particularly evident when the component resident systems operate at significantly different rates (i.e., coupling fast and slow energy chains) (44). Synchronized resource dynamics, on the other hand, would degenerate and possibly remove this spatial coupling. Although some aspects of spatial nesting are evident in the migration of ungulates across the African savannah and the migration of cyprinid fish between lakes and streams (45, 48), many of the predators of these migrants remain resident year-round, in contrast to the increased mobility of predators that is anticipated in current food-web studies. Food-web theory may therefore benefit from explicit consideration of migratory movements in coupling asynchronous resource dynamics.

Such “mobile links” have also been shown to increase ecosystem resilience by providing sources for reorganization after disturbance [i.e., ecological memory (9)]. Indeed, it is well established that dispersal between heterogeneous sites can influence patterns of diversity in metacommunities (24). Although studies of metacommunity structure generally assume that dispersal is constant through time, recent theoretical consideration of stochasticity in dispersal rates suggests that punctuated dispersal events lead to increased heterogeneity among communities (59). The predictability of migration in space and time may therefore render migrant-mediated dispersal fundamental to both the structuring metacommunities and the maintenance of ecological memory.

Outlook and Future Challenges

The breadth of mechanisms outlined above, combined with the sheer abundance of migratory taxa worldwide, suggests that migrants may represent profound but as-yet-underappreciated dimensions of biodiversity that are intimately embedded within local communities. For the handful of species whose interactions have been quantified, it is evident that migrants may substantially alter community structure and ecosystem processes across broad spatial scales; however, it would be premature to assume that this is the case for all migrant-resident interactions. Several synthetic questions therefore pose exciting challenges for future research, including how the effects of migrants differ between sites along migratory routes; how traits of migrants (taxon, trophic position, mode of locomotion, and so on) influence their role in linking communities; and how community attributes (diversity, stability, phenology, resource availability, size, isolation, and so on) modify the relative importance of migrant effects. Collectively, these synthetic questions inform the crucial question: What are the global consequences of the rapid decline and disappearance of many animal migrations?

Across the globe, migration is an increasingly threatened phenomenon as a consequence of habitat destruction, creation of barriers, overexploitation, and climate change (60). The loss of migrants and migratory behavior also entails the loss of their ecosystem services—the manifold transport and trophic effects outlined above. Management strategies must therefore be designed to conserve not only migratory species but also their ecosystem functions. Yet, the conservation of migrants poses exceptional scientific and societal challenges (60), as events at each stage of the migratory cycle affect behavior and demographic rates (61) and ecological interactions at other stages (Fig. 1). Management actions therefore require detailed understanding of the cascading influence of multiple stages across multiple locations.

A promising framework for understanding complex systems and developing effective conservation strategies comes from network theory (62). Network approaches detail the pattern of interactions (“edges”) between entities (“nodes”: species, populations, and individuals), and have been increasingly used to identify structural properties (“topology”) that confer stability and resilience in ecological systems (63, 64). The majority of such network approaches have considered single (resident) communities, in static (time-aggregated) networks, focusing on a single interaction type (e.g., trophic or mutualistic interactions) (65). Given that animal migrations link several disparate ecological communities, introduce dynamics by their temporary presence, and interact through several simultaneous processes, three recent advances in network theory would greatly enhance our ability to assess the ecological importance of animal migrations, the consequences of their loss, and the effectiveness of potential management measures: networks-of-networks, dynamic networks, and multiplex networks (Fig. 5).

Fig. 5 Network approaches are ideally suited to investigating the role of migratory animals (orange nodes) and may assist in designing effective management actions.

(A) Network-of-networks—several resident networks linked by migrants; (B) dynamic networks—nodes and edges change with the passage of migrants; (C) multiplex networks—nodes interact through several interdependent networks, e.g., trophic, mutualistic, and parasite transmission.

The conservation of migratory species is often inherently transnational, and as a result, targeted management actions that consider the entire migratory range and identify the optimal collaboration between jurisdictions are required. To this end, the analysis of network-of-networks (Fig. 5A), through explicit consideration of migrant-mediated interactions between resident networks, facilitates a mechanistic understanding of the cascading consequences of losing migrants, as well as identification of the nodes and edges most influenced by such a loss and those with the greatest influence on the persistence of migrants and their migrations. Networks-of-networks therefore allow detailed assessment of how perturbations and management actions in one resident network cascade through other resident networks. Furthermore, because the arrival and departure of migrants represents the addition and removal of nodes and edges in a network (Fig. 5B), standard time-aggregated networks may not accurately depict the gross topology (the arrangement of nodes and edges) and flow (how resources, information, and such are propagated between nodes) of the entire network, including properties related to network stability [reviewed by (65)]. By examining changes in network topology (“rewiring”) and flow over time, and their mutually dependent feedbacks, using several time-ordered interaction matrices, dynamic network approaches are ideally suited to assessing the speed, extent, and magnitude of migrant-induced rewiring and, hence, the consequences of disappearing animal migrations for network flow and stability. Finally, migrants interact with resident communities through several processes simultaneously (Fig. 1), by forming a network of several interdependent networks (Fig. 5C) (66). So-called multiplex networks are increasingly recognized as vital to understanding the behavior and stability of complex systems because most systems are inherently comprised of coupled networks (67); the properties promoting stability and resilience can be vastly different from those identified for networks of single-interaction types (66); and multiplex networks are far more susceptible to catastrophic failure, which may even follow the removal of a single node (66, 67). Although multiplex networks are still rare in ecological studies (68, 69), coextinction cascades can be dramatically accelerated when feedbacks between multiple networks are included (68). Because each of these network approaches match scales of action to the scales of problem (70), the efficiency of various targeted interventions, in terms of diffusion through the entire network-of-networks, can be assessed (71).

Understanding the causes and consequences of community structure and dynamics remains a major frontier in ecology, particularly in light of the unprecedented rate of species extinctions facing the planet (43). Given the ubiquity of migrants and their potential to link disparate communities the world over, ecological processes in one location cannot be viewed in isolation. Greater understanding of how migratory patterns uniquely influence ecological networks is therefore of paramount importance to the conservation and management of migratory and resident species alike.

References and Notes

  1. Acknowledgments: We are grateful to V. Grimm, C. van Leeuwen, G. Hays, and K. Medley for valuable comments and J. Armstrong, D. White Jr., J. Waldenström, J. Brodersen, J. Bêty, and Bat Conservation International for kindly providing images. B.J.H. was supported by the Netherlands Organization for Scientific Research (825.11.036). This is publication 5591 of the NIOO-KNAW.

Stay Connected to Science

Navigate This Article