Migration is a critical part of the life history strategies of a diverse group of organisms. An optimal strategy may include the need to move from one location to another and back in some systematic fashion. Migration is the large-scale movement between different parts of the home range, with some energy allocation to support movement or time to meet reproductive needs.
Migration underscores an individual’s need for some resource such as food or mates. An individual has a home range that is a function of relative body size and mobility as well as a variety of other factors. In general, large animals need large home ranges. As home range increases in size, an individual experiences variability in environmental conditions. In die marine environment, two parcels of water 10 cm apart are more similar on average than two parcels of water 1000 km apart. Accordingly, waters of higher latitude have generally higher productivity than those of lower latitudes. In addition, a single parcel of water may have similar oceanograpliic conditions 2 days apart but inay differ significantly between winter and summer. Length of die productive season is shorter at higher latitudes due to decreased sunlight in winter, in addition to other oceanograpliic changes. Thus resources are variable in space as well as time. In general, one part of a home range may be veiy different with respect to the availability of resources than another. Disparity in resource availability results in the necessity to move between places in the home range.
Resources for marine mammals include food, mates, and space. Mates may only be seasonally receptive or available, or sexes may have a different spatial distribution in a particular season. Space is a variable resource because not all habitats are suitable. In territorial species, once a territory is occupied, other individuals of the same species are excluded. In nonterritorial species, crowding often occurs, and while habitat is nearby, and apparently suitable, there is a tendency to form a crowd. Space for hauling out may not be available at all times due to covering by tides once or twice a day.
Given that prey resources are variable in space and time, one strategy is to stay in one area and tough out the hard times by somehow reducing the effects of variability. This leads to formation of denser fur, diicker blubber layer, or adopting a new strategy such as hibernation. Female polar bears (Ursus maritirrms) hibernate over winter, giving birth and feeding cubs until emerging from die den in spring. Other segments of polar bear populations do not hibernate, but make large-scale movements in search of food. One benefit of staying in one location is diat an individual does not face any energetic cost of moving or potentially adverse conditions along die way. For cetaceans, the cost of moving long distances while on migration is probably not verv different than moving around in one location. Some pinnipeds may conserve energy by hauling out or lying at die surface for extended periods. The second strategy is to move to another part of their home range where conditions are more favorable, such as a higher resource density, or where environmental conditions are better. For example. Caribbean manatees (Trichechus manatus) move in relation to changing water temperature on a seasonal basis. They prefer waters warmer dian 68°F and move in relation to diese waters. Migration occurs in populations where some parts of a home range may have more and/or better resources oi one type, while another part of the home range has more and/or better resources of another type dian the first. The general pattern is movement between feeding and reproductive grounds or haul-out site.
For marine mammals, migration was assumed to occur based oil the seasonal occurrence of large numbers of a particular species at different locations. However, migration can only be proven by Lagrangian studies, involving marking individuals in some fashion on one migratory destination and recapturing them on another. Whaling provided the first real evidence for migration in large whales. Numbered darts were fired into dorsal blubber and muscle. If that individual was killed during subsequent whaling operations, tagging and killing dates and locations could be compared and some assessment of movements could be made. Some pinnipeds were marked with numbered flipper tags. Censuses were conducted on a number of haul outs. Movement was documented as tagged individuals moved between haul-out sites. Movement and migration patterns have been described in varying levels of detail using photo-identification and satellite telemetry. For example, locations of individual northern elephant seals (Mirounga angustirostris) sent by satellites every 2 days provide valuable information on movement and migration to and from areas of high food productivity, inferred by persistent signals from a relatively confined location, such as over a seamount.
Defining migration is the first step to discussing it because not all large-scale movements are migration. This is not an easy task, as literature on migration does not have a consistent definition. How do we distinguish between daily movements of a seal that returns to the same haul out every low tide from a whale that swims from the Antarctic ice edge to the waters off Brazil? Movement encompasses a hierarchy of displacements ranging from thousands of kilometers, encompassing thousands of surfacings. The following terms describe a variety of types of movement of increasing scale (Fig. 1). A step is a relevant distance moved, such as distance between sites of long dives. Kinesis refers to changes in turning or movement rates. Kinesis can often result in station keeping, where an individual maintains a relative position through relatively frequent turns. Foraging is a search for resources within a patch. A patch is an area within which resources are randomly distributed. Commuting occurs between adjacent patches. After searching patches in a region, animals can then move to another region, which is subject to a different set of local oceanographic conditions. Movement between regions is referred to as ranging. For example, fin whales (Balaenoptera physalus) make an overall east to west displacement through the Gulf of Alaska during a feeding season.
Migration is persistent and more or less rectilinear movement, presumably between two different parts of the home range, each with its own resources or use. For marine mammals, these destinations are areas for feeding/breeding/birtli/ lactation and, in addition, for pinnipeds, molting. General mysticete migration patterns are shown in Fig. 2. This movement occurs on a seasonal basis, with the majority of a population, sex, or age class undertaking the same overall movement pattern. Upon arrival at one destination, behavior changes to maintain relative location. For molting, breeding, and lactating pinnipeds, this is achieved by hauling out on land or, if on feeding grounds, floating in one spot or clusters of deep dives in a limited area.
Figure 1 Different type of movements are described in the text.
Figure 2 Generalized migration patterns of baleen whales in the Northern and Southern Hemispheres. Whales spend summers on productive feeding grounds and then migrate to the winter breeding/calving grounds in warmer waters. Most species of whales are believed to feed little, if at all, while away from their feeding grounds.
For cetaceans, station keeping involves changing the direction of travel relatively frequently.
One characteristic of marine mammal migration is seasonal changes in energy allocation and storage. This results in an increased fat store to support reproduction. Based on stomach contents analysis of whales killed on winter grounds, mysticetes are thought to feed little on migration routes and breeding grounds, living off stored blubber. However, recent evidence suggests that feeding occurs on the winter grounds, at least opportunistically. Some phocids such as elephant seals (Mirounga spp.) migrate long distances. They feed continuously to store energy to support time spent on breeding and molting haul outs. Stored energy is used for lactation in females and for mating and agonistic displays in males. Elephant seals lose much of their fat when hauled out and need to begin storing energy as rapidly as possible upon reentry into the water. Otariids feed daily around haul outs, although males may fast while defending a territory.
In this discussion, migration and movement focus in the horizontal dimensions (x and y). Displacement in the vertical dimension (s) is trivial by comparison, although is not a trivial part of foraging ecology. For marine mammals, migration is movement considerably greater than 2 hr, the maximum dive time recorded for a marine mammal and the maximum depth of a dive.
The following scenarios are not true migrations, but are often labeled as such in literature. Seasonal movement may be a response to changing prey distribution. The occurrence of some groups of killer whales (Orcinus orca) in the inland waters of the Pacific Northwest correlates with the seasonal migration of salmon (Onchyrhtjnchtis spp.). Because these fish ultimately go upstream to die, the whales must find other prey during the winter. It is not clear if they feed on other species during the winter or go offshore to find other salmon schools. Gray seals (Halichoerus grypus) move to distinctly different areas on a seasonal basis to feed in productive areas in the Northwest Atlantic. Polar bears seasonally roam over large home ranges, searching for available food. Other large-scale movements are also not truly migration. For example, movement may be in relation to shifting environmental conditions, such as the seasonal advance or retreat of an ice edge. While geographic location of an individual changes, it is essentially maintaining itself in the same general environment.
Dispersal is not migration, as there is no return to the original area. Dispersal is colonization of new or recolonization of historic breeding habitats. For example, breeding sites of northern elephant seals were historically on islands off the mainland of California, likely due to the presence of terrestrial predators oil the mainland. In association with postexploitation recovery and the decline of terrestrial predators, elephant seals returned to all historic breeding islands and invaded new sites on the mainland. Another example of a dispersal event was observed off California in response to the 1982/1983 El Nino Southern Oscillation event. Common bottlenose dolphins (Tursiops truncatus) moved from southern California to San Francisco with the northward advance of warmer waters. This group of dolphins remained after the warm waters retreated back south.
Tine migrating species have distinct adaptations. Because energy is a limiting resource, migration can be examined from the perspective of energy acquisition (Table I). Fasting species have different physiological and energy intake characteristics than nonfasting species. Fasting during some part of the year requires intense feeding during some other part of the year in order to store energy as well as to meet metabolic needs at the time.
Species that fast either spend their entire lives or part of the year in productive polar, subpolar, and temperate waters where they can feed intensely to form a blubber layer. Other than ice-breeding seals, migrating marine mammals have feeding grounds located pole-ward of the breeding areas.
II. Why Migrate?
Any strategy has associated costs and benefits. Natural selection weighs the costs and benefits of migrating; if the benefits are greater than the costs, a species will migrate. Different areas have different proximate currencies, with energy acquisition important in the feeding areas and reproductive success important in the breeding/calving areas. Feeding success, however, lias a direct influence on reproductive success, the ultimate currency, for survival of the genes of an individual. For cetaceans, migration occurs for food and reproduction. Pinnipeds migrate for these reasons as well as for molting.
TABLE I Energy Acquisition
Mysticetes” (except bowhead whales) Elephant seals”
Some male otariids” (on breeding grounds) Harp seals (Pagophilus groenlandicns)” Hooded seals (Cystophora cristata)” Polar bears (females can hibernate) Most phocids
Recently weaned pinnipeds (some exploratory swimming around haul-out site) Nonfasting Odontocetes Female otariids Some phocids Polar bears Sirenians Otters “A true migratory species.
A. Pinniped Patterns
Pinnipeds are central-place foragers, moving from haul-out sites to feeding grounds at varying distances. Large-scale movement awav from a haul out reduces pressure on local resources and has other benefits as well. Because predators ma)’ congregate around haul outs, an individual can reduce its chance of being killed by a predator by reducing the number of entrances to and exits from the water. Dive profiles of elephant seals suggest that while they feed along an entire migration route, there are areas with a higher frequency of deep dives, suggesting intense, localized feeding. Some of these highly productive areas are associated with seamounts. These seals make deep foraging dives for deep-water prey. The areas around their haul outs are seasonally productive and support a number of mysticete species. Therefore, productivity is not the key; rather it is productivity of a certain type of prey that makes migration beneficial. Some otariids make daily excursions from haul-out sites in search of food. For a lactating pinniped, large movements increase the time spent away from a pup, resulting in less rapid pup growth. While increased entrances and exists from the water increase the probability of being killed by a predator, the sheer number of other seals entering and exiting the water reduces the per capita probability of attack.
B. Mysticete Migration
The feeding grounds of baleen whales are in productive cold waters so it is clear why they migrate to these areas. A question remains as to why they must migrate to warm waters for reproduction. Four reasons have been suggested for migration to warm water breeding and calving areas. The first is to minimize thermal stress on calves. This is likely not a problem for a newborn calf. Smaller mammals with less insulation are able to survive in those conditions. Because of its large body size, a calf is not thermally stressed.
The second reason to move to warm waters is resource tracking, i.e.. following prey. By definition, this is not migration. While an individual, school, or population is moving, movement results in reduced resource variability, as prey are always in the vicinity.
Killer whale predation on calves has been a suggested reason for mysticete migration. By migrating to warm, relatively killer whale-free waters to give birth, calf mortality would be reduced.
The final reason to migrate is essentially an evolutionary holdover: individuals migrate because their ancestors did. The evolutionary holdover hypothesis includes feeding and reproduction into a life history strategy. Intense feeding leads to energy storage as the short-term goal that maximizes reproductive success, the ultimate, evolutionary goal. Natural selection favors individuals that are successful at migrating, feeding, and reproducing.
C. Pleistocene Conditions
The productive waters of the higher latitudes have changed significantly over the past 20,000 years. These changes likely had a profound influence on regional productivity, affecting migration routes, destinations, and foraging areas of many cetaceans, pinnipeds, carnivores (bears, otters) and sirenians (Steller’s sea cow, Hydrodamalis gigas). Increased ice extent and land emergence of the Pleistocene made current feeding grounds unavailable to gray whales (Eschrischtius robustus), bowhead whales (Balaena mysticetus), beluga whales (Delphi-naptrus leucas), narwhals (Monodon monocerous), walruses (Odobenus rosmarus). Northern fur seals (Callorhinus ursi-1111s), and other seals. The North Atlantic north of 45° was an ice-bearing polar sea with conditions similar to the Antarctic Convergence, resulting in a larger, more productive sea than at present. The distance between productive cold water and warmer water was much less, resulting in relatively little or no distance between feeding and breeding/calving areas (Fig. 3). In the Southern Ocean, an equator-ward shift in isotherms was in response to a northward extension of ice. The Antarctic convergence was 5° north of its current position.
D. Present Oceanographic and Migratory Conditions
At the glacial maxima, cold, productive waters were closer to the equator than at present. At these latitudes, the number of hours of sunlight per day were not as seasonally extreme as at the poles. Assuming that mid and lower latitude waters are relatively unchanged from Pleistocene conditions, the only difference would be the retreat of cold waters toward the poles. As cold, productive waters retreated toward the poles, sunlight for photosynthesis became more variable over the course of a year, leading to intense seasonal peaks in production followed by reduced production in winter. Over time, whale distribution followed the pole-ward retreat of fronts of productive oceans.
This explains the summer distribution of baleen whales, but does not address the need to migrate to warm waters for reproductive activities. Many species of marine mammals, and mammals in general, have highly seasonal reproductive strategies in order to time births relative to optimal environmental conditions. Day length is a cue for seasonal breeding in a number of birds and mammals. Photoperiod is an important seasonal cue because it is invariant from year to year. This means that the timing of reproduction and migratory movements can be the same from one year to the next. The advantage is that an individual can maximize its use of seasonal prey resources as well as seasonally available mates.
Both circadian (daily) and circannual (yearly) cycles use light as a cue: however, the specific cues from light, or zeitgc-bers, vary. Circadian signals are dawn and dusk, whereas circannual signals are perceived as the ratio of number of light to dark hours in a 24-h period. Thus both cycles are used for seasonal cues. Other cues may act as secondary synchronizers, although these, such as food availability, are more variable.
The pineal gland is responsible for time keeping in birds and mammals via the production of melatonin, as well as other compounds. The number of hours of darkness per 24 hr is “counted’ by the biosynthesis of melatonin, which is produced more in hours of darkness. At seasonal scales, as winter approaches, the hours of daylight decrease and hours of darkness increase. In a given 24-hr period, the amount of melatonin produced increases, suppressing gonadal activity. In many mammals the breeding season of females corresponds to periods of decreasing daylight per 24 hr. Increasing daylight after the winter solstice is responsible for triggering estrus. Pineal glands are exceptional in size in polar species such as the Weddell seal (Leptonychotes weddcllii), northern fur seal, and walrus, species that live where day length is most variable.
Figure 3 Post-Pleistocene distribution of the polar front in the North Atlantic Ocean. Lines represent the southern extent of sea surface temperature relating to the polar front. Numbers associated with each line are thousands of years before present. Whale movement from warm-water calving/breeding waters to cold productive feeding waters would have been similarly truncated toward the equator in the past. The equator-ward extension of cold water meant that productive seasons were probably longer in the past than at present due to increased hours of daylight.
As productive waters retreated pole-ward, cetacean distribution shifted accordingly, changing the overall lighting regime from more or less equal hours of dark and light to one with more hours of darkness in winter, with resultant gonadal suppression. In addition, production decreases in winter so there is no benefit to stay in colder waters.
Individuals need to move to a lighting regime that switches on the gonads. In migrating animals, a shift in the lighting regime on the feeding grounds may trigger migration toward the equator. As an individual approaches the equator, the hours of daylight per 24 hr increase. In addition, after December 21, the fewest hours of daylight in the Northern Hemisphere, and June 21, the fewest hours of daylight in the Southern Hemisphere, the number of hours of daylight per 24 hr begins to increase. The rate change in the lighting regime reaches its maximum at the equinoxes. Together, these result in reduced melatonin levels and restored gonadal activity, which triggers mating behavior. This may be a triggering mechanism in the evolution of migratory behavior in mysticetes.
Testosterone and its metabolites trigger migratory behavior in some animals. Male California sea lions (Zalophus californianus) spend the bulk of the year hauled out in large bachelor groups. By June, they have left their haul-outs in central and northern California, Oregon, Washington, and Alaska and migrated to the Channel Islands off southern California. Here, breeding occurs as males set up territories and defend females against other males with whom they spend most of the year in relatively peaceful coexistence. Once migration back to the feeding grounds is triggered in July, an individual is exposed to increasing hours of daylight per 24 hr as it swims pole-ward. One benefit from such a signal is that it is invariant from year to year.
Curiously, the peak migration of gray whales is variable from year to year. The initiation of the southward migration to the breeding grounds may be linked to a change in foraging success or some other environment factor, such as the formation of sea ice. Individuals may be able to override these signals. Extended feeding seasons for baleen whales occur on occasion as prey are uncharacteristically available in late full. For example, blue whales (Balaenoptera musculus) were observed feeding on euphausiids in late November in Monterey Bay, California. Whales generally leave the area in late September.
In some cases, not all members of a migratory species actually migrate in a given year. For example, in every location where minke whales (B. acutorostrata and B. bonaerensis) feed, individuals are observed in winter.
III. Orientation and Navigation
The mechanisms of orientation, plotting their location at any time, and navigation, directing movement from one location to another, are not known. Individuals are often seen in the same locations from one year to the next. In the interim, they have traveled thousands of kilometers, indicating that marine mammals use some type of cues for orientation and navigation between migratory destinations. Organisms tend to meander if they lack orientation and navigation cues. Therefore, marine mammals must know where they are at a given time (orientation) and where thev are going next (navigation).
At the initiation of migration, a direction must be selected. Advancing ice may simply eliminate certain directions as a choice, displacing individuals toward the equator. In higher latitudes, changes in sea conditions influence prev availability, which may also trigger the migratory response.
Cues may vary over time and the course of migration. For example, once migration is initiated, the only cue necessary is which overall direction to travel: north, south, east, or west. Celestial navigation has been suggested as one mechanism of navigation. In the north/south directions, the relative location of the sun in the skv can be monitored. This may be as simple as “keep sunrise on the left side when migrating to the breeding ground and on the right side when migrating to the feeding ground” or as complicated as estimating latitude as a function of position of the sun. Navigation by star location has also been suggested as a mechanism.
Another possible large-scale cue is the direction of a major current if an animal is moving against it. Near the equator in the Northern Hemisphere, western boundary currents move from south to north, while eastern boundary currents move from north to south. Coastal processes and minor currents cause the deformation of major currents at higher latitudes, resulting in the formation of gyres and eddies. Migrating whales may use these currents for a free ride.
Magnetite in the brains of some species has been implicated as a mechanism by which individuals could track changes in the earth’s magnetic field. Mass strandings often occur at the same location. These locations may have anomalies in the local earth’s magnetic field, which cause whales to become disoriented and strand. A tantalizing example of the possibility of using magnetic cues is seen in humpback whales (Megaptera no-vaeangline) migrating from Hawaii to Southeast Alaska. Tracks were within 1° of magnetic north.
At smaller scales, other cues could be used. For example, while mvsticetes do not have a true sense of smell, they do have a well-developed Jacobsen’s organ. This may allow them to “taste” differences in water mass composition. For example, freshwater from ice melt or riverine input might provide a “taste trail” to a rich feeding ground, as a lens of fresh water floats on denser salt water.
Routes to and from feeding and breeding grounds may be variable or essentially a retracing of the migratory path. Male humpback whales migrate south from the Gulf of Maine relatively far offshore, while the return trip is much closer to shore. Gray whales along the west coast of North America probably migrate between calving lagoons in Baja California and feeding grounds in the Bering, Chukchi, and Beaufort Seas by following contours of the coastline. Gray whales migrate along the same nearshore corridor in both migratorv directions. There is spatial or temporal segregation based on age class or gender. Northern elephant seals disperse from haul-out sites into the Gulf of Alaska: however, males go further north than females.
In cases where a baleen whale species has a sufficiently long lactation period, offspring can learn migration routes and feeding locations from their mothers. In instances where offspring are weaned prior to reaching the feeding grounds, offspring are left to make exploratory migrations in hopes of finding suitable feeding grounds. This may be one reason that populations of some minke whales are segregated according to age class.
IV. Physiology of Migration
Marine mammals have a suite of physiological adaptations for energy allocation and fasting, as well as storage and mobilization of fats to and from blubber stores. Fats are the most important energy source, as lipids hold more energy per unit weight than other forms. Lipids are also less bulky than protein or carbohydrates because no water is required for storage. During feeding on productive feeding grounds, hyperphagia promotes increased lipid synthesis, fat uptake, and rate of fatty acid synthesis. Storage of fats occurs when the supply in blood exceeds metabolic demand. Mobilization of fats occurs when the demand for energy in blood exceeds supply.
However, little is known about hormonal activity in relation to migratorv movements. Prolactin is a hormone responsible for promoting milk production and lactation in mammals. It has the effect of increased fattening in birds. If there were a similar effect on mammals, it would be of importance. For pinnipeds, where lactating females leave pups for one to a few days, increased fat storage for milk is vital. A lactating female mysticete on the feeding ground would store fat relatively faster. This would not only provide for milk for the offspring, but also help in restoring the female’s blubber laver for subsequent migration back to the breeding/calving grounds.
V. Effects of Migration
Marine mammals may have localized seasonal effects on their feeding grounds. For example, nutrient recycling may be enhanced locally in bays and inlets used by groups of feeding humpback whales in Alaska. Benthie feeders such as gray whales and walruses represent a disturbance mechanism to these communities. Disturbance opens areas for colonization and settlement of pelagic larvae of benthie species. Other coadaptations occur in barnacles living on gray whales that have timed spawning activities to coincide to when whales are in the calving lagoons. Whales, barnacles, and thus released gametes of barnacles are concentrated in limited areas.
VI. Metapopulation-Removal Migrations
Stock boundaries are delineated by mark and recapture analysis, genetic, biochemical, demographic differences, or other techniques. These differences give the impression of little or no flux of individuals between stocks. Evidence from photographic identification studies of humpback whales in the North Pacific Ocean suggests that some differences may be artificial as individuals move between stocks. Within an ocean basin, stocks can be viewed as a metapopulation, which is a number of populations connected by the dispersal of individuals between them. If a species is distributed into populations spread over a sufficiently large area, environmental conditions are more or less independent between areas; therefore a catastrophe in one area will not affect other populations. Dispersal reduces the risk of population extinction by minimizing the effects of chance environmental changes or changes in population demography. Further, genetic heterozygosity is maintained and the population is less likely to exhibit genetic problems.
A Steller sea lion (Eumetopias jubatus) metapopulation has been explored in Alaska. In the Aleutian Islands and Gulf of Alaska, these sea lions have declined by 50% since the 1960s. However, Steller sea lions from Southeast Alaska south to Oregon have remained stable or slightly increased during this time. Evidence from Alaska and the Aleutian Islands suggests that fragmentation will occur, with rookeries being reduced in size and eventually becoming extinct. One reason for hope would be if the population in Southeast Alaska became a source of dispersers into the Aleutian population. The Mediterranean monk seal (Monachus monachus) experienced a recent population decline with habitat fragmentation throughout its range. Large expanses of unsuitable habitat separate major pupping sites, with little chance of dispersal between the two remaining large populations, although each separate population may be viable over time.
VII. Migration, Movement, and the Future
One of the main reasons to construct models of movement and migration patterns is to develop descriptive and predictive models to study the effects of changing environmental conditions. Migration in marine mammals evolved within the context of constantly changing environmental conditions. Species had to adapt to deal with novel situations and conditions. A polar front in the Pleistocene as described earlier retreated at a rate that allowed individuals to adapt to its changing distribution. Climatologists predict that global temperatures will increase by as much as 4.5°C in the next century. While it is clear that marine mammals are capable of adapting to changing environments, they might not be able to adapt at a rate commensurate with that of the change in environmental conditions in the near future. The potential implications are profound, and the environmental effects are not entirely clear.
Global warming will likely have variable effects depending on latitude, with polar and temperate regions affected to a greater extent than more tropical areas. Because these areas represent feeding grounds for migratory as well as resident species, understanding these effects is of considerable importance. The effect on marine mammals will likely be through changes in the distribution of resources in space and time. Key to survival will be how individuals respond to changes in resource distribution over space and time, and how this affects reproductive success.