Pinniped Ecology (marine mammals)

Ecology deals with the interactions between individuals and their environment. In this context, environment is taken broadly to include other organisms and the physical environment. These interactions take place at various spatial and temporal scales and influence both the abundance and the distribution of individuals. However, ecology is also a historical science in that the patterns we see today reflect past events and phylogenetic relationships. Thus, processes acting on both evolutionary and ecological time scales have undoubtedly influenced many of the characteristics of pinniped ecology we see today. Pinnipeds are large, long-lived, aquatic mammals exhibiting delayed sexual maturity and reduced litter size; a single precocial offspring is the norm. As such, they share many of the demographic features of other large mammals. Population numbers do not change dramatically from year to year, and numbers are most sensitive to changes in adult survival, followed by juvenile survival and fecundity (Eberhardt and Siniff, 1977). We assume that these characteristics are under selection and that variability in foraging success affects survival probability and reproductive performance of individuals. Nevertheless, uncertainty about historical condition and influences limits the extent to which present characteristics can be interpreted as evolutionary adaptations. Inevitably, discussions of pinniped ecology and other aspects of pinniped biology will overlap. This article focuses on five aspects of pinniped ecology: abundance, distribution, reproduction, foraging, and the ecological roles of pinnipeds in aquatic ecosystems.

I. Abundance

Despite interest in the ecology of pinnipeds, the abundance of many species is poorly known. The abundance of species of commercial importance (past, northern fur seals, Callorhinus ursinus; or present, harp seals, Pagophilus groenlandicus) is generally better known than for those species that have not been exploited. The accuracy and precision attached to the estimates of abundance vary greatly due to the difficulty in designing good census procedures or to the lack of effort to obtain good estimates. Good estimates of abundance are important because both abundance and trends in abundance are perhaps the most useful indicators of the status of populations.

Commercial exploitation decimated many pinniped species, in some cases to levels nearing extinction (e.g., northern elephant seals, Mirounga angustirostris). Over the past several decades or more, some species have recovered or are continuing to recover. Thus, the present abundance of heavily exploited species may not be a good measure of their preexploitation numbers. Pinniped species range over 4 orders of magnitude in abundance, from the crabeater seals (Lobodon carcinophaga) at about 12 million (probably the most abundant marine mammal in the world) to the Mediterranean monk seal (Monachus monachus) at probably fewer than several hundred individuals (Reijnders et al, 1993). Phocid species are generally more abundant than otariids, with 15 of 18 phocid species numbering greater than 100,000 individuals compared with only 8 of 14 otariid species (Bowen and Siniff, 1999). The reasons for this difference are not entirely clear. Over the past 100 years, both families have been exploited commercially and subjected to other human factors that might have influenced abundance. More likely, the greater abundance of phocids is die result of their greater use of high-productivity areas in temperate and polar waters than is the case in most otariid species. The three most abundant otariids, the northern fur seal, the Antarctic fur seal (Arctocephalus gazella), and the Cape fur seal (A. ptisillus pusillus), all forage in seasonally productive, high-latitude ecosystems, a characteristic shared with the most abundant phocid species (i.e., the ringed seal Pusa hispida, the harp seal, and the crabeater seal).

Abundance is determined by the movement of individuals (in and out of the population), births, and deaths. These processes are influenced by both direct and indirect human activities, as well as by ecological factors such as predation, food supply, breeding habitat, disease, competition with other species, and environmental variability. In the absence of human effects, combinations of these ecological factors are thought to regulate die abundance of a population about a level known as carrying capacity. Although the concept of carrying capacity has a long history in ecology, with the possible exception of the Weddell seal (Leptonychotes weddellii), there is little evidence that populations of pinnipeds fluctuate about some long-term average level. This lack of evidence may be an artifact of the effects of previous exploitation on population trends and the fact that few populations have been surveyed over many decades. For example, a number of pinniped populations recovering from human exploitation (e.g., Antarctic fur seal, gray seal Halichoerus grypus, harbor seal Phoca vitulina) have increased at rates in excess of 12% per year over several decades or more. At Sable Island, Canada, the number of gray seal pups born each year has increased exponentially, with a doubling time of about 6 years, for more than 40 years. It is also clear that pinniped populations may decline rapidly as a result of epizootics, such as the phocine distemper virus that killed large numbers of harbor seals in the North Sea, and during short-term extreme changes in ocean climate, such as El Nino (see later).

Ecosystem structure and function are influenced by both top-down (i.e., consumer driven) and bottom-up (i.e., producer driven) processes. Pinnipeds likely exert top-down control on some ecosystems through predation (see Section IV) and are affected by changes in food available brought about by changes in primary and secondaiy productivity.

An example of a top-down ecosystem perturbation affecting pinniped abundance occurred in the Southern Ocean in the early 1900s. The overexploitation of some species of seals and whales led to an enormous uncontrolled “experiment” in this cold-water ecosystem. A high biomass of Antarctic krill is the cornerstone of the Southern Ocean food web, accounting for half of the total zooplankton biomass. Of the six species of pinniped that inhabit the Southern Ocean, crabeater seals, Antarctic fur seals, and leopard seals {Hydrurga leptanyx) feed mainly on krill, whereas southern elephant seals (Mirounga leonina) and Ross seals (Ommatophoca rossii) consume mainly cephalopods and Weddell seals eat primarily fish. Krill is also the main food resource for the resident large baleen whales [blue (Balaenoptera musculus), sei (B. borealis), minke (B. bonaerensis), humpback (Megaptera novaeanglioe), fin (B. physalus), and southern right (Eubalaena australis)]. As the cetacean biomass declined from exploitation by more than 50% between 1904 and 1973, an estimated 150 million t of krill were released annually to the remaining predators (Laws, 1985). The abundance of krill-eating species of pinniped, such as the crabeater seal and the Antarctic fur seal, increased substantially following the massive cetacean exploitation.

Bottom-up effects on top predators such as pinnipeds also can occur rapidly over the course of months. Perhaps the most dramatic example of this occurs during El Nino. El Nino events occur approximately every 4 years in the eastern tropical Pacific, resulting in reduced upwelling and a decrease in primary and secondary productivity. During a severe El Nino, the effects of reduced food availability on seabirds and marine mammals can be quite pronounced. Galapagos fur seals (Arctocephalus galapagoensis) have an unusually long lactation period of approximately 2 vears that is thought to have evolved to buffer young against minor El Nino events. However, during the severe El Nino event between August 1982 and July 1983, pup production of Galapagos fur seals was only 11% of previous years, and no pups survived past the first 5 months. Adult females responded by increasing the foraging trip length, while most adult males did not appear on the breeding site and were unable to hold territories during the breeding season (Trillmich and Olio, 1991).

Top-down and bottom-up effects at times may work in concert to influence the abundance of pinnipeds. Between the 1950s and early 1970s, intensive harvesting in the Bering Sea and the Gulf of Alaska critically reduced the populations of large whales, flatfishes, herring, and other primary consumers of krill and zooplankton. It is believed that the resulting increase in the availability of krill and other zooplankton, coupled with a regime shift in the ocean climate, favored some species such as walleye pollock and not others. This resulted in a change in the relative abundance oi both fish and invertebrate assemblages. There was also a regime shift in ocean climate during the 1970s that likelv contributed to changes in the biological productivity of these areas. Although the consequence of these changes on marine mammals remains uncertain, both harbor seal and Steller sea lion (Eumetopias jubatus) numbers declined dramatically following these changes, resulting in the Steller sea lion being declared an endangered species. A probable explanation for this cycle of events involves a combination of environmental change affecting producers (bottom-up effect) and human exploitation of predators resulting in changes to the ecosystems that have been detrimental to pinnipeds (Bowen and Siniff, 1999).

II. Distribution

Fundamentally, pinniped distributions reflect the need to give birth on solid substrate (land or ice) and to feed at sea. Within these broad constraints, the distribution of pinnipeds is affected by physical (e.g., ice cover, location of remote islands) and biological (e.g., productivity, abundance of predators) characteristics of habitat, demographic factors (e.g., population size, age, sex, and reproductive status), morphological and physiological constraints, and human effects (e.g., disturbance). Although each of these factors may influence distribution, combinations of factors are generally responsible for the distribution patterns observed. Pinniped distribution is also three dimensional, where the third dimension is water depth and the underlying bathymetry. Although a complete understanding of pinniped distribution must consider this three-dimensional world, this aspect of pinniped behavior is discussed elsewhere.

Pinniped species have a restricted and generally patchy distribution in most aquatic environments: estuaries and continental shelves (e.g., gray seals); tropical seas (e.g., monk seals, Galapagos fur seals); the deep ocean (e.g., elephant seals); both Arctic (e.g., ringed seals) and Antarctic polar seas (e.g., crabeater seals, Antarctic fur seals); and freshwater lakes (e.g., Baikal seals, Pusa sibirica) (King 1983). It is important to note, however, that our understanding of the distribution of most species is based primarily on the location of breeding colonies. We know less about where most species forage at sea such that our view of the overall distribution of most species is incomplete. For example, based on the location of breeding colonies, northern elephant seals range from Baja California to central California. However, satellite telemetry studies show that this species forages over broad areas of the North Pacific Ocean for much of the year.

If we examine the distribution of pinniped breeding colonies, it is clear that we see patterns that reflect the evolutionary history of pinnipeds and the distribution of resources. At large scales, both sea lion and fur seal distributions reflect their origins in the Pacific Ocean. Northern fur seals and Steller sea lions are widely distributed along both sides of the North Pacific Ocean. The California sea lion (Zalophus Californianus) and Japanese sea lion (Z. japonicus) are or were endemic to opposite sides of the North Pacific, and the Galapagos sea lion occurs at the equator. The four other species of sea lions occupy colonies along the west coast of South America, southern Australia, and New Zealand. With the exception of the northern fur seal and Guadalupe fur seal, the other six species of fur seals occur in these tropical or subtropical southern water, but also extend into the cool, nutrient-rich waters of the South Atlantic and Indian Oceans. Sea lion and fur seal breeding colonies are usually located on remote islands near areas of high biological productivity (e.g., northern fur seals, Antarctic fur seals), which provide both protection from mainland predators and nearby food sources. These conditions are particularly important for lactating females.

Species of the family Phocidae are widely distributed in biologically productive temperate and polar seas. Although most abundant in the North Atlantic and Antarctic Oceans, a reflection of their evolutionary origins in the Atlantic basin, phocid species have circumpolar distributions in both the Arctic Ocean (e.g., ringed seal, bearded seal, Erignathus barbatus) and the Antarctic Ocean (e.g., Weddell seal, crabeater seal) as well as a broad distribution in the North Pacific Ocean [e.g., harbor seal, larga seal (Phoca largha), ribbon seal (Histriophoca fasciata)]. Several endangered species also occur in tropical waters (Hawaiian and Mediterranean monk seals, Monachus spp.).

Pinnipeds must return to a solid substrate (land or ice) to give birth, rear their offspring, and in many species to molt. For most species, these requirements result in seasonal changes in distribution. In the case of species that breed on pack ice, such as harp and hooded (Cystophora cristata) seals and the walrus (Odobenus rosmarus), seasonal changes in ice cover virtually guarantee some change in the distribution of individuals. This may partlv explain why 7 of 13 (54%) species of pinnipeds that give birth on ice (i.e., most phocid seals and the walrus) are migratory compared to only 4 of 20 (20%) species that give birth on land (2 of 6 phocids, 2 of 14 otariids; Bowen and Siniff, 1999). However, this difference also may be partly explained by the variable quality of data on the at-sea distribution of pinnipeds.

Migration appears to be a common feature of the ice-breeding phocid species, but this behavior is perhaps best documented in the northern elephant seal. This land-breeding species shows extreme sexual size dimorphism, with males being about five times heavier than females. Northern elephant seals undertake the longest known migration and some of the deepest dives reported for a mammal (Stewart and DeLong, 1993). Individual elephant seals make two long-distance migrations of 18,000 to 21,000 km between breeding and molting sites in California and pelagic foraging areas in the North Pacific. Using the California current as a corridor to areas further north, northern elephant seals leave the breeding beaches in southern California for northern offshore foraging areas. The first migration occurs following the breeding season, in which adult male and female elephant seals travel an average of 11,967 and 6289 km, respectively, and remain at sea for an average of 124 and 73 days. After the molt, the seals depart on a second migration. Females are at sea for approximately twice as long as males and cover an average distance of 12,264 km compared to an average of 9608 km by males, males migrate farther north than females, with most males traveling as far as the northern Gulf of Alaska and the eastern Aleutian Islands. These sex differences in foraging distribution, and perhaps diet, may have evolved to reduce competition between females and the much larger males of this species.

III. Reproductive Ecology

The reproductive ecology of pinnipeds varies considerably, reflecting differences in body size, geographic distribution, and habitats used by individual species. Despite this diversity, there are common features that reflect their common ancestry as terrestrial carnivores, and their subsequent adaptation to a predominately aquatic lifestyle.

As noted previously, a conserved trait of their terrestrial ancestry is the requirement for all pinniped species to give birth to their offspring on a solid substrate (land or ice). However, pinnipeds must feed at sea, often some distance from the breeding grounds. This spatial and temporal separation of parturition from aquatic foraging is thought to have played a large role in shaping the mating and lactation strategies of pinnipeds. Three general strategies have evolved to deal with the conflict between at-sea foraging and terrestrial parturition (see later); however, the requirement for terrestrial parturition has likely contributed to some common features of pinniped reproduction, such as birth synchrony.

In most pinniped species, reproduction is seasonal and highly synchronous (e.g., harp seals). The evolution of reproductive synchrony is often associated with seasonal resource availability. In ice-breeding species (e.g., harp and hooded seals), the timing of reproduction is linked to the seasonal availability of sea ice. Seasonal changes in prey abundance and environmental conditions can also influence the timing of parturition and mating. The Hawaiian monk seal (Monachus schauinslandi) displays only weak synchrony in reproduction. In this species, births extend over a 6-month period. Given the less variable tropical habitat of this species, reproductive synchrony may not have been under strong selection relative to the species in more variable temperate and polar environments. Subtropical populations of California and Galapagos sea lions and Galapagos fur seals also show slightly less temporal synchrony of reproduction relative to more temperate populations (Boness, 1991).

Other common features of pinniped reproduction include postpartum mating and delayed implantation. These two characteristics of pinniped reproduction also appear to reflect the terrestrial ancestry of the taxa with both features occurring in many modern terrestrial carnivores. However, selection for postpartum mating may have continued as pinnipeds adapted to their aquatic environment. Given the wide-ranging and dispersed distribution of pinniped species during the at-sea foraging season, the aggregation of individuals at pupping colonies may have offered one of the few predictable opportunities for male and females to interact.

Another common feature of pinniped reproduction is the production of a single, precocious offspring; litters of two are rare. Offspring are born with their eyes open and begin to vocalize within minutes of birth. Neonates are also able to move short distances to their mother and to begin suckling shortly after birth. Harbor seal females produce extremely precocial offspring that are capable of swimming and diving with their mothers within an hour of birth (Bowen, 1991).

A. Mating Systems

Within the Pinnipedia, mating systems range from extreme polygyny (e.g., northern fur seals) to sequential defense by males of individual females. The mating system of individual species is closely associated with the dimensionality and stability of the habitat used and the distribution of females at parturition. Broadly speaking, species can be grouped as land-breeding and aquatic-breeding species.

1. Land-Breeding Species Land-breeding pinniped species include all fur seals and sea lions, northern and southern elephant seals, and the gray seal. These species colonize oceanic islands and coastal areas to give birth and mate. The aggregation of individuals during the breeding season has been attributed to the fact that oceanic islands are relatively rare and unevenly dispersed such that the availability of suitable pupping sites may limit the distribution of females (Boness, 1991). Predation may also select for female clustering, with females being less vulnerable to terrestrial predators and/or harassment by conspecific males when in large groups (dilution effect). Aggregation of females within a stable, two-dimensional habitat has led to the evolution of a polygyny in these species, with males defending either resources needed by females (e.g., birth and thermoregulatory sites in otariid species) or the females themselves (e.g., elephant seals and gray seals). By competing with and limiting the access of other males to females, successful males mate with multiple females, thus increasing their reproductive success. The degree of polygyny in land-breeding pinniped species ranges from extreme in the northern fur seal and elephant seals where one male may mate with 16-100 females to moderate (6-15 females) in gray seals, Hooker sea lions (Phocarctos hookeri), and the Galapagos fur seal (Le Boeuf, 1991).

As in other polygynous species, land-breeding pinniped species are sexually size dimorphic. Males in these species can be much larger than females and often show other secondary sex characteristics. These dimorphic characteristics are the result of sexual selection for traits that increase an individual’s ability to monopolize and defend resources needed by females or females themselves. Large body size, and concomitant large body energy stores in the form of subcutaneous blubber, permits dominant males to fast and thus remain ashore during the period when females become receptive. The most extreme example of sexual size dimorphism in pinnipeds occurs in elephant seals where males are 5-6 times heavier than females in the northern species and up to 10 times heavier than females in the southern species.

2. Aquatic-Breeding Species Walruses and all other phocid seals [Weddell, Ross, crabeater, leopard, bearded (Erignathus barbatus), hooded, ringed, Baikal, Caspian (Pusa caspica), larga, harp and ribbon] give birth on pack ice or fast ice and mate in the water. Although Hawaiian monk seals and harbor seals give birth to their offspring on land, they too mate in the water. In species where pups are born on ice, females tend to be more widely distributed, although access to breathing holes in the ice may promote clumping in some species (e.g., walrus and Weddell seals). This broader distribution of females, on an unstable habitat, limits the number of females a male can monopolize at any given time and as a result these species typically show reduced levels of polygyny (e.g., harbor seals; Colt-man et al, 1999). The fact that mating occurs in the water, a fluid three-dimensional environment, may also limit the ability of males to monopolize females, resulting in reduced levels of polygyny.

Wells et al. (1999) suggested that the mating strategies used by ice-breeding species could be classified as scrambling (males search for receptive females and move on to the next); sequential defense (males sequentially defend single females through mating); and lekking (males aggregate and attract females using displays). At present, there is insufficient information on the breeding behavior of most aquatic breeding species to draw firm conclusions about the type of mating system used. Data on the mating behavior of these species are limited to that which can be observed on ice prior to copulation. For example, observational data suggest that hooded seals utilize a sequential defense form of mating system whereby males compete with one another to defend a single mother and her pup on the ice. The male then remains with the pair until the pup is weaned and then enters the water with the female, presumably to mate. The application of new methods, including genetic paternity assessment and animal-borne video, will be needed to clarify the mating strategies used by these species.

In species that mate aquatically, there may be less selective advantage for males to be larger than females because of the limited ability of males to monopolize females in this environment. As a consequence, in most of these species, male and females are of similar size and in some cases females are larger than males. For example, male Weddell seals are slightly smaller than females and it has been suggested that smaller size makes the male more agile during underwater mating activities (Le Boeuf, 1991). Underwater vocalizations also appear to be an important component of the mating behavior in aquatically mating pinniped species. For example, in Pacific walruses, which exhibit a lekking mating system, males perforin complex underwater visual and vocal displays in small groups next to female haul-out sites to attract females. Male Weddell, harbor, harp, hooded, and bearded seals also produce a range of underwater vocalizations during the breeding season that may be used to attract females or to establish underwater territories or display areas.

B. Lactation Strategies

Male pinnipeds do not participate in the care of offspring. Thus, parental care is the exclusive responsibility of the female. Female care involves the transfer of energy-rich milk to the pup and protection from conspecifics and terrestrial predators (Bowen, 1991). In some species (e.g., the walrus), females may also teach their young to forage, as young accompany mothers on foraging trips during the lactation period. Female pinnipeds have dealt with the temporal and spatial separation of energy acquisition (aquatic foraging) from high levels of energy expenditure (terrestrial lactation) in different ways, resulting in the three basic lactation strategies: a foraging cycle, fasting, and aquatic nursing. Until recently, each of the pinniped families was thought to use only one of these strategies (otariids, the foraging cycle; phocids, fasting; and odobenids, aquatic nursing) and the evolution of these differing strategies was presumed to reflect phylogeny. However, studies have indicated that this traditional view is an oversimplification (Boness and Bowen, 1996; Boyd, 1998). For example, harbor seal females exhibit a foraging cycle strategy previously thought to occur only in otariid species.

1. Foraging Cycle All otariids and some of the smaller phocid species (e.g., harbor seals) exhibit this lactation strategy. Females come ashore for parturition with a moderate level of stored body energy. After giving birth, females remain onshore and fast while attending and nursing their young for a perinatal period ranging from a few days to a week. After this initial provisioning period, females leave their pups and return to sea to feed. These trips range from less than 1 day in some species to as long as 14 days in others, depending on the distance to the foraging location and prey abundance. Females then return to land to nurse their pup, after which they repeat the cycle until the pup is weaned. The lactation period in otariid species is quite long, ranging from 4 months to 3 years (Bowen, 1991). Females of these species are considered income breeders, relying on current food intake to support both their own metabolic needs and the energetic cost of milk production. The milk produced by female otariids is relatively energy dense (i24—409c fat) compared to terrestrial mammalian species. Pup growth rates are rather low, ranging from 0.06 kg/day in Galapagos fur seals to 0.38 kg/day in Steller sea lions (Boness and Bowen, 1996).

Harbor seals, a phocid species, also exhibit this lactation strategy—alternating short foraging trips to sea (7-10 hr) with terrestrial nursing. The harbor seal is a relatively small phocid species, with females weighing approximately 84 kg at parturition. Given the small quantity of body energy that these females are able to store, female harbor seals are forced to make regular foraging trips to acquire sufficient energy to wean their pups successfully. Compared to otariid species, the length of the lactation period in harbor seals is much shorter (24 days) and the milk produced by females has a relatively higher fat content (50%). Consequently, pup growth rate is higher in harbor seals relative to otariid species (0.6 kg/day). Foraging cycles during lactation may also occur in ringed seals and other relatively small phocid species. Evidence shows that the females of two medium size phocids, the Weddell seal and the harp seal, may also forage during the lactation period. However, the intensity’ of foraging and the degree to which successful weaning of offspring relies on these foraging trips are not clearly understood.

2. Fasting Strategy In the larger-bodied phocid species, females fast during lactation. Females arrive at the breeding site with large energy stores in the form of adipose tissue (i.e., blubber). In the western Atlantic, for example, gray seal females arrive at Sable Island weighing an average of 210 kg. Of this body mass, 32% or 67 kg is fat. After parturition, females fast for the entire lactation period (e.g., 16 days in the case of gray seals) using their stored energy to support the energetic cost of milk production and their own maintenance metabolism. For this reason, female phocids are considered capital breeders—having stored energy often months before it is needed. The lactation period in phocids is much shorter than in otariid species, ranging from 4 days in hooded seals to 60 days in Weddell seals. Maternal body size, metabolic rate, and the fasting ability of females may play an influential role in determining the duration of lactation both within and between species (Boness and Bowen, 1996; Boyd, 1998).

Another characteristic feature of the phocid fasting strategy is the production of extremely high-fat milk, ranging from 47% fat in southern elephant seals to 61% fat in hooded seals. This energy-dense milk results in a high rate of offspring growth, ranging from 1.4 kg/day in the Hawaiian monk seal to 7.1 kg/day in the hooded seal (Bowen, 1991). Weaning occurs abruptly when mothers return to the sea to feed. Pups often fast for weeks following weaning, living off their accumulated fat stores before entering the water and beginning to forage independently.

3. Aquatic Nursing The walrus is the only pinniped species that exhibits aquatic nursing. Just prior to parturition, pregnant females separate from the herd and give birth to their offspring alone on pack ice. New mothers remain on the ice fasting for the first few days postpartum, relying on stored body energy accumulated prior to parturition. Subsequently, females and their young return to the herd to forage. Walrus pups suckle in the water for between 2 and 3 years on relatively low-fat milk (24.1%). As with otariids, weaning is gradual. Young walruses begin to feed on benthic organisms as early as 5 months of age and likely gain valuable foraging experience from their mothers over the remainder of lactation. At weaning, female offspring are assimilated into the mothers herd, whereas male offspring join other male groups.

Lactation strategies are often viewed from the female’s perspective. This seems reasonable, but in long-lived species such as pinnipeds, females may trade-off investment in current offspring against investment in future offspring. This may lead to conflicts between females and their offspring over the level of investment received. The transition from nursing pup to nutritionally independent juvenile usually occurs without parental supervision in pinnipeds. This transition is arguably the most important period of a pinnipeds’ young life. As offspring size affects subsequent survival, we should expect that offspring would attempt to obtain as much milk as they can during lactation. Thus, the nutritional requirements and physiological abilities of individual offspring must also play a role in shaping lactation strategies. For example, the fasting ability of offspring constrains the duration of foraging trips by female fur seals and sea lions.

IV. Foraging

Successful foraging is essential for survival and reproduction and is therefore a critical determinant of fitness. Pinnipeds are among the largest vertebrate carnivores in marine ecosystems and yet the foraging behavior of these apex predators is poorly understood. As noted earlier, pinnipeds inhabit diverse environments, consequently they forage at highly varied spatial and temporal scales and in doing so they exploit a wide range of prey.

A. Methods

As pinnipeds generally feed underwater at remote locations, ecologists rely upon indirect methods to gain insight into their foraging behavior and diets. Very high frequency (VHF) radio tags have been used to study the at-sea locations of coastal species such as harbor seals. Acoustic tags have been used to track the underwater movements of gray seals. More recently, microprocessor-based, time-depth recorders (TDRs) have been used to collect information on dive duration, depth, frequency, and temporal distribution and to calculate the at-sea locations of pinnipeds using solar navigation equations. However, the use of TDRs is limited by the need to recover the instrument to retrieve the stored information and therefore only those species that can be reliably recaptured are used in TDR studies. In contrast, satellite-linked, time-depth recorders transmit collected data on diving parameters and surface positions to polar-orbiting satellites operated by Service ARGOS. This technology has broadened the range of species that have been studied, but the expense of using satellite-linked tags often places limits on the number of individuals studied.

Although we have learned a great deal from the use of location telemetry and dive recorders, these studies have provided little insight into the feeding success rate of pinnipeds. Recent work has demonstrated that estimates of feeding success can be determined using stomach temperature telemetry and animal-borne video. The body temperature of a marine prey is often considerably lower than that of its pinniped predator, thus the stomach temperature of the predator should drop following prey ingestion. This approach has been used successfully on free-living harbor seals and several other species. When combined with information on the diving behavior and movement patterns in the same individual, stomach telemetry can provide new insights into the spatial and temporal patterns of foraging success relative to foraging effort. Animal-borne video technology (Marshall, 1998) has taken our understanding of foraging behavior and diet one step further by providing direct observations of the way in which pinnipeds search and capture prey and how foraging behavior changes as a function of prey type. These video images, coupled with data on swim speed, diving characteristics, environmental conditions (such as sea temperature), and energy expenditure, promise to revolutionize our understanding of pinniped foraging ecology.

Determining the diet of marine mammals also requires the use of indirect methods. The most common methods rely on the recovery and identification of hard prey structures that are resistant to digestion from the stomach, intestine, or feces of individual animals. Sagittal otoliths, cephalopod beaks, bones, scales, invertebrate exoskeletons, and shells can be used to determine the species consumed and, in some cases, to estimate the size and age of the prey. Fecal samples are increasingly being used for this purpose because they are less expensive to collect, a high proportion of samples contain identifiable prey, and estimates of diet are less affected by differential rates of digestion than estimates from stomach samples (Bowen and Siniff, 1999). Although the use of hard parts to estimate the diet of pinnipeds is common, this method is subject to a number of biases, which may limit the value of results. First, stomach and fecal contents only provide an estimate of the diet near the point of collection, and as a result, offshore diets cannot be sampled easily. This may seriously bias the diet of wide-ranging species such as elephant seals, harp seals, northern fur seals, and Juan Fernandez fur seals (Arctocephalus philippu). Second, hard parts are often eroded during digestion or digested completely such that prey size may be seriously underestimated and prey identification may not be possible. Finally, perhaps the most serious disadvantage is that dietary analysis based on hard parts is strongly biased against soft-bodied or small prey with fragile structures.

Inevitably our understanding of the diets of pinnipeds is tied to the development of new methods. Fatty acid signature analysis is a relatively new method, which has been developed to study marine mammal foraging and diet (Iverson, 1993). Lipids in marine ecosystems are diverse and characterized by long-chain polyunsaturated fatty acids that originate in unicellular phytoplankton. In monogastric carnivores, such as pinnipeds, ingested fatty acids with a carbon chain length greater than 14 are deposited in body tissues in a predictable way. As a result, the fatty acid composition or signature of the predator reflects the fatty acid composition of prey species consumed (Iverson, 1993). By comparing the reference signature of various prey species to the fatty acid signature of the predator (obtained from blubber tissue or milk), diet composition can be estimated both qualitatively and quantitatively. The use of fatty acid signature analysis eliminates the dependence on recovery of hard parts and integrates the diet over a period of weeks to months such that the location of sampling becomes less important.

Stable isotope ratios of carbon and nitrogen found in the muscle, skin, vibrissae, or blood of pinnipeds and other predators are also being used to investigate diet. These ratios reflect a composite of prey species eaten over a broad time scale. By examining the levels of ISN/UN found in body tissues, scientists can determine the trophic level at which the pinnipeds fed. The carbon isotope ratio (I3C/I2C) has been found to vary geographically and thus the level of carbon isotope in the predator’s tissues provides insight into foraging location. Although this technique is useful in determining trophic level and foraging location, it does not permit the specific diet composition of individuals to be assessed.

B. Diet

A large number of prey species have been identified in the diet of various pinniped species, leading to the view that pinnipeds are generalist predators. This is consistent with the expectation that large, wide-ranging predators consume more types of prey, as their environment becomes patchier. However, in most cases a relatively small number of taxa account for the majority of food eaten (Bowen and Siniff, 1999). For example, gray seals on the Scotian Shelf, Canada, consumed 24 different taxa; however, only two to four species accounted for over 80% of the energy consumed depending on the time of year.

Fish and cephalopod species are the main prey types eaten by pinnipeds (Table I). However, crustaceans also appear to account for a substantial portion of prey consumed by some species. Crustaceans are a major prey of harp seals in the North Atlantic and of ringed seals and bearded seals in the Bering Sea. In three Antarctic species, Antarctic fur seals, crabeater seals, and leopard seals, krill accounts for up to 50% of the diet. Unlike most pinnipeds, which generally feed on mobile prey (e.g., fishes, cephalopod molluscs, and crustaceans) in pelagic and benthic habitats, the walrus feeds almost exclusively on sessile benthic invertebrates in soft-bottom sediments.

Several pinniped species are also known to feed on other pinnipeds (Bowen and Siniff, 1999). Male southern fur seals appear to commonly feed on young South American fur seals (A. australis). Steller sea lions are known to prey on a variety of pinniped species, including harbor seals, ringed seals, bearded seals, young northern fur seals, and larga seals. Walruses prey on larga seals, ringed seals, and young bearded seals.

The diet composition and foraging behavior of pinnipeds are influenced by a number of factors. The ecology and behavior of prey species clearly play a role in shaping the foraging strategies of pinnipeds. Research on male harbor seal foraging behavior at Sable Island, Canada, using animal-born video found that prey behavior affected both capture technique and profitability of different prey types. Other studies have shown that between-year differences in the diet composition of harbor seals were correlated with differences in the distribution and abundance of herring and sprat, two important prey species.

Intrinsic factors, such as age and sex, may also play a role in the diet composition of individuals within pinniped species. Given that pinnipeds are long-lived predators, their individual foraging tactics and behavior may change over time to reflect increased physiological capabilities and learning. For example, harbor seal pups feed on pelagic prey such as herring and squid, whereas the diet of adults is dominated by benthic species. Similarly, the contribution of benthic prey (e.g., crabs, clams, and sculpins) to the diet of bearded seals increases with age. Age-specific differences in diet composition have also been found in southern elephant seals and harp seals.

Diet composition may also differ between sexes in pinniped species that exhibit sexual size dimorphism (e.g., northern and southern elephant seals, otariid species). Due to the relationship between basal metabolic rate and mass, larger individuals require more total energy per unit time than smaller individuals. Oxygen storage capacity also increases with body mass due to the larger blood pool in which to store oxygen and the larger muscle (myoglobin) mass. In addition, larger animals have a slower mass-specific metabolic rate such that they utilize their larger oxygen stores at a slower rate relative to smaller individuals. Thus, larger individuals are capable of longer, deeper foraging dives. These physiological attributes may allow, or require, males (the larger sex) to pursue different prey types (potentially higher quality prey) than females. Although theory suggests that the diets of males and females may differ, currently there are few studies to test this hypothesis. However, studies on southern elephant seals do indicate sex differences in diet.

TABLE I Major Prey of Selected Pinnipeds

C. Foraging and Diving Behavior

The foraging ecology of pinnipeds and other air-breathing vertebrates is constrained by the need to surface for oxygen. Dive duration is constrained by the interplay between the amount of oxygen that can be stored and the rate at which the diver expends oxygen. Thus, it is inevitable that patch use, and the resulting distribution of foraging in time and space, will be influenced by the physiological constraints. Other factors, such as prey density and depth, may play an important role in how pinnipeds forage within these physiological constraints.

Foraging pinnipeds dive repeatedly with relatively short surface intervals between dives; this cluster of dives is called a dive bout. In general, dive bouts are thought to indicate foraging within a prey patch, particularly in otariid species. Theoretically, divers should organize their behavior for optimal patch use. To organize their behavior in this way, divers should optimize both the time budget of the dive cycle (dive duration and surface interval) and the number of dive cycles to repeat. Both of these factors will influence the amount of prey caught and the energy and time consumed during the dive bout. However, there may be a trade-off between prey depth and profitability such that prey items that might be exploited when closer to the surface are less likely to be exploited as the depth of that prey increases.

Empirical tests of optimal foraging theory and optimal patch use in diving pinnipeds are uncommon, largely due to the difficulty and expense of studying these wide-ranging predators and their prey. However, it appears that some otariids feed near the surface on vertically migrating prey, such as krill, to maximize energetic efficiency.

Phocids are generally better suited for deep diving and for longer periods of time than are their otariid and odobenid counterparts. This is largely because phocids have a larger blood volume and larger myoglobin content in the muscles and thus store more oxygen per unit of body mass. Phocids also dive in continuous bouts and are known to spend up to 90% of their time in the water submerged. Thus, unlike otariids and odobenids, phocid seals live at depth, returning periodically to the surface to breathe. Although diving behavior is often considered to be synonymous with foraging in otariids, dive shape analysis in phocids demonstrates that diving may also be used for travel, predator avoidance, and sleep (Wells et al, 1999).

The function of different types of dives has been investigated through the analysis of the two-dimensional profile (time vs depth) of individual dives (i.e., dive shape) and swimming speed during diving. Foraging dives are those in which time spent at the bottom of the dives is a significant fraction of the total dive duration. In northern elephant seals and Weddell seals, these dives are often to similar depths over time, suggesting that the seals are exploiting prey patches that remain at a constant depth and are dense enough to maintain high encounter rates. In contrast, dives characterized by a middle segment of slow downward drift are thought to be associated with the digestion of food in female northern elephant seals. Although dive shape undoubtedly contains information about the behavior of individuals, animal-borne video has revealed that different behaviors can be represented by the same dive shape, thus limiting the inferences that can be drawn from shape alone.

D. Factors Affecting Foraging Ecology

Pinnipeds are important consumers of marine resources; however, for most species little is known about how they interact with the biotic and abiotic features of marine ecosystems. Knowledge of the spatial behavior of pinnipeds is important because spatial patterns can fundamentally affect the nature and dynamics of species interactions. These interactions largely determine the distribution of foraging. Within the ocean, food is distributed in patches and this distribution can be strongly influenced by the physical properties, such as water temperature and the availability of nutrients. For example, the distribution and migratory patterns of northern elephant seals correspond with the location of three dominant water masses of the North Pacific. The localized biological productivity in these water masses and associated fronts result in a high abundance of cephalopods, an important food of this species.

Seasonal changes in prey distribution and abundance can also influence pinniped foraging patterns. Reduced prey availability leads to changes in foraging behavior that include increased trip duration, trip distance, and increased foraging effort. For example, Antarctic fur seals increase their times at sea, northern fur seals increase diving effort, and California sea lions use both tactics during periods of limited prey resources.

E. Spatial and Temporal Scales of Foraging

The relative mobility, range, and body size of an animal affects the resolution at which it recognizes environmental heterogeneity. For example, a relatively small-bodied, central place forager, such as a lactating harbor seal, would identify resource patches at a smaller meso-scale than a highly mobile animal, such as a gray seal. To understand the relationship of an organism to its environment, one must understand the interactions between the intrinsic scales of heterogeneity within the environment and the scales at which the organism can respond to this heterogeneity. Scale issues are critical for effective conservation and management of pinnipeds because of shifts in habitat use and dispersal over ontogeny and a relatively long life span.

A large body size and the capacity for storing large amounts of fat in the form of blubber enable some species of pinnipeds to feed irregularly and thus to exploit distant foraging locations and patchy resources. In contrast, smaller pinnipeds, such as Antarctic fur seals, perceive environmental heterogeneity at a more local scale. For example, fur seals forage at two spatial distributions: (1) fine scale, represented by short (<5 min) travel durations between patches and (2) meso-scale, represented by longer periods of travel (>5 min) (Boyd, 1996). Similarly, based on fatty acid signature analysis, harbor seals appear to demonstrate meso-scale partitioning of their foraging habitat in Prince William Sound, Alaska. Fatty acid signatures obtained from harbor seal blubber biopsies differed within the Sound at a spatial scale of about 40-50 km, and at a smaller scale of 9-25 km, reflecting fine-scale differences in diet between haul-out sites (Iverson et al, 1997).

Although the patch structure of an environment is expressed in both space and time, temporal variation in predator behavior is likely to provide an insight into the spatial distribution of a highly dynamic prey source that may be difficult to track in other ways. For example, in the Antarctic Ocean, krill is distributed patchily and is the major prey resource of lactating Antarctic fur seals. By using the diving behavior of females obtained from TDR records, it is possible to track the way in which fur seals respond to within season and interannual variation in prey patchi-ness and abundance. Over a 5-year period, changes in the distribution of travel durations between diving bouts suggested that the spatial distribution of krill swarms varied between years. Although their foraging behavior did not indicate that there was a reduction in the number of krill patches, reduced pup growth rates suggested that patches were of poorer quality, and thus the females had difficulty meeting lactation needs. To compensate, females spent a greater amount of time at each patch, thereby maximizing their average rate of energy intake (Boyd, 1996).

To maximize fitness during years of reduced prey abundance, pinnipeds must be sufficiently plastic in their foraging strategies to compensate for added foraging costs. To determine the temporal scales at which predators make these behavioral decisions, Boyd and colleagues simulated increased foraging costs in Antarctic fur seals by adding an extra drag to lactating females, thereby increasing energy expenditure. At the scales of individual dives, the treatment group made shorter, shallower dives than the control (no extra drag added) seals. It appeared that diving behavior was adjusted to maximize the proportion of time spent at the bottom of dives. At the scale of diving bouts, there was no variation between the two groups in bout frequency and duration, or the time spent diving. However, at the scale of complete foraging cycles, the time spent at sea was significantly longer in the treatment group, yet there was no difference in pup growth rate between control and treatment groups.

In contrast to otariids, most phocid seals are able to fast throughout much of the breeding season due to their large body size and corresponding energy stores. As a result, behavioral responses of phocids to changes in food availability between years may be more flexible, resulting in less severe effects on their population dynamics. Still, a change to a less profitable prey or increased foraging effort may still have energetic consequences that result in impacts at the population level. In the Moray Firth, Scotland, clupeid fishes are the dominant prey of harbor seals. In years when clupeids are absent from inshore waters, seals travel further to feed and use alternative prey. As a consequence, the seals showed evidence of reduced body condition, suggesting that there were energetic consequences to this change in diet. Between-year differences in survival rates suggest that temporal variation in prey abundance and resulting diets also have consequences for the dynamics of phocid populations (Thompson et al, 1996).

V. Role of Pinnipeds in Aquatic Ecosystems

Although pinnipeds are one of the more visible components of the marine ecosystems, our understanding of their ecological roles is surprisingly limited. Still, there is some evidence that pinnipeds may have important effects on the structure and function of some ecosystems (Bowen, 1997). Given that pinnipeds are large, long-lived animals that are often present in considerable numbers, we might expect some species to exert top-down control on ecosystems. However, conclusive studies are lacking, largely due to the difficulty of conducting manipulative experiments in the ocean, the fact that interactions occur at quite different spatial and temporal scales, and the inherent indeterminancy in the behavior of complex marine systems.

Despite the long-standing debate over the ecological interactions between pinnipeds and commercial fisheries, there is little understanding of the impact of pinniped predation in these situations. For example, the recent collapse of the Atlantic cod stock on the eastern Scotian Shelf has fueled debate over the impact of gray seal predation, both in causing the decline and in preventing early recovery. Model results indicated that gray seal predation accounted for only 10 to 20% of the estimated mortality caused by the fishery and therefore was unlikely to have played an important role in the decline.

One example of top-down control exerted by a pinniped species comes from a study of lakes in northern Quebec. Lower Seal Lake has a population of land-locked harbor seals and, compared to nine neighboring lakes without seals, supports a different fish community. The relative abundance of lake trout (Salvelinus fontinalis) was greater in the nine lakes without seals, whereas brook trout (S. namaijcush) was the dominant species in Lower Seal lake. Compared to lake trout in neighboring lakes, those in Lower Seal Lake were on average smaller, younger, grew more rapidly, and matured earlier, all of which represent life history characteristics that are associated with heavy exploitation. Although based on strong inference rather than direct empirical evidence, it appears that seal predation was responsible for both the changes in community stnicture and the life history traits of fish species in Lower Seal Lake (reviewed in Bowen. 1997).

Pinnipeds may also play a role in structuring benthic communities. Walruses disturb bottom sediments during feeding. By selectively feeding on older individuals of a few species of bivalve mollusks, walruses may be responsible for structuring the benthic fauna. Ingestion and defecation by walruses may result in substantial redistribution of bottom sediments, which may favor colonization of some species. In addition, during the process of feeding, walruses produce many pits and furrows in the soft sediments. Thus, walrus feeding appears to affect community structure in three ways: by providing food for scavengers such as sea stars and brittle stars, by providing habitat under discarded bivalve shells, and by reducing the abundance of macroinvertebrates in feeding pits compared to surrounding sediments. Nonetheless, the effects of walrus feeding behavior on macrobenthic assemblages over periods greater than a few months and at larger spatial scales remain greatly unknown.

VI. Conclusions

Our understanding of the ecology of pinnipeds has increased dramatically over the past several decades, but advances have been rather uneven. For example, the lactation strategies of pinnipeds are reasonably well understood, but many aspects of foraging ecology and the ecological role of pinnipeds in aquatic ecosystems remain elusive. As in all areas of science, our ability to measure the system under study influences the rate of progress profoundly. New types of data loggers, telemetry, and methods to estimates the diet of free-ranging pinnipeds will undoubtedly play a prominent role in advancing our understanding. However, we should not underestimate the importance of collaborative research involving ecologists, oceanogra-phers, and population and ecosystem modelers.