Search

A species’ occurrence can be influenced by natural and anthropogenic factors; disentangling these is a precursor to understanding the mechanisms of distribution. Anthropogenic factors may be especially important at contracting range edges. We test this premise for wolverines (Gulo gulo luscus L., 1758) at the edge of their Rocky Mountain range in Alberta, Canada, a mosaic of natural heterogeneity and extensive landscape development. As wolverines have a suspected negative response to human activity, we hypothesized their occurrence on the Rockies’ slopes is predicted by a combination of natural and anthropogenic features. We surveyed wolverines at 120 sites along a natural and anthropogenic gradient using hair trapping and noninvasive genetic tagging. We used abundance estimation, generalized linear, and hierarchical models to determine whether abundance and occurrence was best predicted by natural land cover, topography, footprint, or a combination. Wolverines were more abundant in rugged areas protected from anthropogenic development. Wolverines were less likely to occur at sites with oil and gas exploration, forest harvest, or burned areas, even after accounting for the effect of topography. The relative paucity of wolverines in human-impacted portions of this range edge suggests that effective conservation requires managing landscape development, and research on the proximal mechanisms behind this relationship.

Fine-scale shape variation and the added effect of partial digestion often limits accurate identification of different teleost prey species in marine diet studies using otoliths. We evaluated the use of fine-scale shape and trace element variation in digested otoliths to identify fish prey species from the diet of predators. Fourier analysis of otolith shape revealed significant variation between red cod ( Pseudophycis bachus ) and bearded rock cod ( Pseudophycis barbata ) otoliths. Incorporating otoliths that had been consumed by Australian fur seals ( Arctocephalus pusillus doriferus ) into a Fourier analysis discriminant model identified 73% of otoliths as those of red cod and 27% as those of bearded rock cod. However, in vitro digestion of red cod and bearded rock cod otoliths resulted in incorrect classification of both cod species otoliths to varying degrees when using Fourier analysis shape descriptors. There was significant variation between red cod and bearded rock cod otolith core chemistry. Incorporating otoliths consumed by the seals into an otolith core chemistry discriminant model identified all otoliths as those of red cod. Using otolith core chemistry to identify prey species was found to be successful, and there is great potential for this technique to have wider applications in investigating ecosystem trophic interactions.

Wolves ( Canis lupus L., 1758) are subject to liberal public harvests throughout most of their range in North America, yet detailed information on populations where sport harvest is the primary source of mortality are limited. We studied a harvested wolf population in west-central Alberta from 2003 to 2008. Demographic data were collected from visits to den sites, 84 collared wolves from 19 packs, and a harvest monitoring program that augmented mandatory reporting for registered traplines. Annual harvest rate of wolves was 0.34, with harvest on registered traplines (0.22 ± 0.03) being twice that of hunters (0.12 ± 0.04). Most wolves harvested (71%) were pre-reproductive. Probability of a pack breeding was 0.83 ± 0.01, litter size averaged 5.6 ±1.4, and these rates and stability of home ranges were unaffected by the number of wolves harvested. Natural mortality (0.04 ± 0.03) and dispersal rates (0.25 ± 0.04) were lower than reported for wolf populations in protected areas. Reproductive rates balanced total wolf mortality, indicating harvest was likely sustainable. We suggest that a high proportion of juveniles harvested and the spatial structure of the registered trapline system contributed to the sustainability of harvests.

Interspecific interactions between sympatric carnivores can be important for the behaviour and demography of involved species. We studied spatial and temporal interactions between an obligate predator, the Eurasian lynx ( Lynx lynx (L., 1758)), and a facultative scavenger, the wolverine ( Gulo gulo (L., 1758)). Wolverines are known to utilize lynx-killed reindeer ( Rangifer tarandus tarandus (L., 1758)) and may benefit from being sympatric with lynx if interference competition is low. We used individual location data from 9 lynx and 17 wolverines to analyse interaction between inter- and intra-specific dyads (n = 195). We found no spatial segregation between lynx and wolverines and we observed no attraction or avoidance between individuals of the two species, independent of proportion of home-range overlap. This opposed our prediction that wolverines will show direct or delayed attraction to lynx. Wolverines may still benefit by scavenging lynx-killed reindeer while avoiding direct encounters with the lynx. Within species, we found attraction between males and females, increasing with proportion of overlap for lynx. Attraction was also found between consexual lynx, while consexual wolverines showed little home-range overlap (7%–9%) and neutral temporal interaction, indicating territoriality. Individual space use may be more influenced by conspecific interactions than by other species.

Because of their wide-ranging habits, conserving large carnivores such as American black bears ( Ursus americanus Pallas, 1780) often depends on understanding habitat needs beyond the boundary of protected areas. We studied habitat selection by black bears in the vicinity of Bruce Peninsula National Park, Ontario — a small, isolated population whose persistence appears dependent on habitat on lands outside the Park. We used an approach based on Euclidean distances to document seasonal habitat selection at two spatial scales and to identify candidate habitat types for protection. Adult females selected dense mixed forests to establish home ranges within the population range, whereas subadults and yearlings selected dense deciduous forests. Within home ranges, adults selected dense mixed forest in spring–summer and dense deciduous forest in late summer – fall. Subadults selected dense deciduous forest, marsh, dense mixed forest, and water during the spring–summer and avoided developed lands and roads. Yearlings selected dense mixed forest, dense deciduous forest, and sparse forests in spring–summer and dense deciduous forest and dense mixed forest in late summer – fall. The selection of dense deciduous and dense mixed forest stands, especially at the broader scale, suggests that strategies to ensure persistence of this isolated population should focus on protecting the integrity of these stands.

Human-caused landscape change negatively affects the sustainability of many wildlife populations. In Alberta, Canada, grizzly bears ( Ursus arctos L., 1758) live in one of the most populated and heavily exploited landscapes in which the species survives. Long-term physiological stress in individual animals may be the predominant mechanism linking landscape change with impaired wildlife population health. Hair cortisol concentration has been validated as a biomarker of long-term stress in humans and domestic animals; however, limited work has examined factors that may affect its measurement or interpretation. We have measured cortisol in as few as five guard hairs of a grizzly bear and have identified factors influencing hair cortisol concentration in this species. Hair cortisol varies with hair type, body region, and capture method. It is not influenced by colour, age, sex, environmental exposure (18 days), or prolonged laboratory storage (>1 year) and does not vary along the length of the hair shaft. Recommendations for prudent use of hair cortisol analysis in grizzly bears are discussed with implications for the development of hair cortisol concentration as a tool to monitor long-term stress in other wildlife.

Two types of wolves, gray ( Canis lupus L., 1758) and eastern ( Canis lupus lycaon Schreber, 1775 or Canis lycaon ) or Great Lakes wolves, representing Old World (OW) and New World (NW) mitochondrial DNA (mtDNA) haplotypes, have been reported in eastern Canada and the Great Lakes region. Both haplotypes were found in Duck Mountain Provincial Park and Forest, Manitoba. Only OW haplotypes have been reported from the isolated Riding Mountain National Park (RMNP), 30 km to the south. Wolves with NW haplotypes hybridize with C. lupus and coyotes ( Canis latrans Say, 1823) and could mediate gene flow between canids. We examined available data on wolf body mass, skull morphology, and mtDNA from the RMNP region, as well as mtDNA from Manitoba and Saskatchewan, to assess the occurrence of NW haplotypes in wolves and possible canid hybridization. Mean body mass of female (n = 54) and male (n = 42) RMNP wolves during 1985–1987 was higher than that of females (n = 12) and males (n = 8) during 1999–2004. Thirteen skull measures from 29 wolf skulls did not suggest significant differences between RMNP and Duck Mountain wolves. Nineteen of 20 RMNP samples had OW haplotypes, whereas one clustered together with NW haplotypes.

We studied the effect of a recolonizing wolf ( Canis lupus L., 1758) population on a resident lynx ( Lynx lynx (L., 1758)) population in south-central Sweden. Wolf and lynx share the same prey species, western roe deer ( Capreolus capreolus (L., 1758)), and the size difference between the two species suggests a strong potential for interference competition. The spatial distributions of lynx family groups (n = 378) over four winters were not significantly affected by the increase in size and range of the wolf population. Survival of lynx kittens until 9 months of age did not differ significantly inside (54%; n = 37) and outside (62%; n = 42) wolf territories, and female lynx (n = 3) selected natal den sites (n = 19) in the same local area before and after wolf establishment. Furthermore, lynx home-range size (n = 42) did not increase as a result of presence of wolves and space use by female lynx (n = 3) was not affected by wolf establishment. We found no evidence of cleptoparasitism by wolves on roe deer killed by lynx. We conclude that the intensity of interference and exploitation competition between wolves and lynx was low.

Territory quality and site fidelity were assessed for male sea otters ( Enhydra lutris (L., 1758)) in Simpson Bay, Prince William Sound, Alaska, during the summer months (May–August) from 2003 to 2006. Territory quality was coded with a score of 0–2 for four attributes: territory size, shoreline enclosure, territory accessibility, and number of females observed feeding in each territory. Total quality scores ranged from 0.14 to 1.96 (0.9 ± 0.61, mean ± SD), with the highest quality territories having large areas, moderate to little shoreline enclosure, high accessibility, and many foraging females. A total of 13.1 km2 (62.5%) of Simpson Bay was occupied at least once by a territorial male during the 4-year study, and 7.9 km2 (37.5%) was never occupied by a territorial male. There were 7 ± 2.4 (mean ± SD) territorial males in Simpson Bay each year. Five males maintained territories for two consecutive years, with their respective territories having a 16%–34% overlap between years, indicating that repeat territory holders returned to the same area. The number of territorial sea otters, the size of territories, and whether or not a sea otter was a repeat territory holder did not significantly influence the size or quality of male territories.

We examined the pupping phenology and genetic variation between the currently defined stocks of harbor seals, Phoca vitulina richardsi (Gray, 1864), in Washington’s coastal and inland waters and looked in detail at genetic variation within the inland waters of Washington. We analyzed mtDNA variation in 552 harbor seals from nine areas in Washington State and the Canada–US transboundary waters. A total of 73 haplotypes were detected; 37 individuals had unique haplotypes. Pupping phenology and levels of genetic variation between the outer coastal stock (WA Coastal Estuaries, WA North Coast) and the inland waters stock (British Columbia, Boundary Bay, San Juan Islands, Smith/Minor Islands, Dungeness Spit, Hood Canal, Gertrude Island) corroborated the appropriateness of the present stock boundary. However, within the inland waters stock, Hood Canal and Gertrude Island were significantly different from the coastal stock, from the rest of the inland waters stock, and from each other. This indicates a total of four genetically distinct groups in Washington State, suggesting that managing the inland waters as a single stock may be erroneous.

We propose a fundamental geographic distribution for the wolverine ( Gulo gulo (L., 1758)) based on the hypothesis that the occurrence of wolverines is constrained by their obligate association with persistent spring snow cover for successful reproductive denning and by an upper limit of thermoneutrality. To investigate this hypothesis, we developed a composite of MODIS classified satellite images representing persistent snow cover from 24 April to 15 May, which encompasses the end of the wolverine’s reproductive denning period. To investigate the wolverine’s spatial relationship with average maximum August temperatures, we used interpolated temperature maps. We then compared and correlated these climatic factors with spatially referenced data on wolverine den sites and telemetry locations from North America and Fennoscandia, and our contemporary understanding of the wolverine’s circumboreal range. All 562 reproductive dens from Fennoscandia and North America occurred at sites with persistent spring snow cover. Ninety-five percent of summer and 86% of winter telemetry locations were concordant with spring snow coverage. Average maximum August temperature was a less effective predictor of wolverine presence, although wolverines preferred summer temperatures lower than those available. Reductions in spring snow cover associated with climatic warming will likely reduce the extent of wolverine habitat, with an associated loss of connectivity.

The taxonomic identity of the historical and current wolf ( Canis lupus L., 1758 or Canis lycaon Schreber, 1775 or their hybrids) population in Minnesota (MN) and the Great Lakes region has been, and continues to be, controversial. So too does its legal status under the U.S. Endangered Species Act. This review summarizes the morphological and genetic information about that population and concludes that historically the MN population consisted of a gray wolf (C. lupus) in the west and an eastern type ( Canis lupus lycaon or C. lycaon) in the east with intergrades or hybrids between the two in most of the state. After extirpation in much of its original MN range, the now-recovered population was infused with gray wolves from Ontario but still consists of hybrid lycaon × gray wolves, probably with higher content gray wolves in the west and higher content lycaon in the east but with most wolves morphologically appearing to be gray wolves. Because the current Wisconsin and Michigan wolf population was derived from MN wolves, they would be primarily hybrids as well. Future research should seek to relate genetic data with morphological measurements in MN wolves. In addition, attempts to breed coyotes ( Canis latrans Say, 1823) with gray wolves in captivity would shed considerable light on the controversy over the origin and taxonomic identity of the newly proposed C. lycaon.

The Fennoscandian arctic fox ( Vulpes lagopus (L., 1758)) population is critically endangered, possibly because of increased interference competition from red foxes ( Vulpes vulpes (L., 1758)) and fading cycles in microtine rodents, which cause food shortage. It is not known how these factors drive arctic fox population trends. To test their role in arctic fox decline, we developed a spatially explicit and individual-based model that allowed us to simulate fox interactions and food availability in a real landscape. A sensitivity analysis revealed that simulated arctic fox population size and den occupancy were strongly correlated with fecundity and mortality during the microtine crash phase, but also with red fox status. Model simulations suggested that arctic fox population trends depended on microtine cycles and that arctic fox distributions were restricted by red fox presence. We compared the model predictions with field data collected at Vindelfjällen, Sweden. The model recreated the observed arctic fox trend only with the inclusion of arctic fox avoidance of red fox home ranges. The results indicate that avoidance behaviours can affect population trends and hence that relatively small numbers of red foxes can have a strong negative impact on arctic fox population size and distribution.

Parentage analysis data for 583 individuals genotyped at 27 microsatellite loci were used to study the mating system of polar bears ( Ursus maritimus Phipps, 1774) in the Barents Sea area. We discriminated statistically between full and half-siblings identified through only one common parent. We document for the first time multiple paternity in polar bears. We demonstrated for both sexes low fidelity to mating partners over time. We did not detect any significant difference between the age distribution of adult males at capture and the age distribution of males siring cubs. This might indicate that the male’s age and size are less indicative of the reproductive success than previously thought. This is further supported by a rather long mean litter interval of 3.9 years for males siring several litters. The mating system of polar bears in the Barents Sea appears to be promiscuous, usually with a single successful father siring full siblings within a year, but with consecutive litters of a mother being fathered by different males. We discuss how population density, landscape characteristics, and adult sex ratio might influence the mating system of polar bears. This is of particular importance for management decisions such as, e.g., implementing sex ratios in hunting quotas.

Randomization of survey sites is generally desired because of its unbiased approach, but is often abandoned because of logistical constraints. This is true for river otters ( Lontra canadensis (Schreber, 1777)), with bridges commonly determining survey locations. We conducted seasonal sign surveys for river otters on two rivers in southern Missouri, USA, using randomized survey points and fixed bridge-crossing points in 2001–2003. Otter sign was more likely to be detected at randomized sites than at bridge sites in summer (P < 0.0001), with sign being detected on 68% of visits to random sites (n = 348) and on 40% of visits to bridge sites (n = 60). Scat abundance was higher (P = 0.0001) at random sites (8.82 ± 0.6, mean ± SE) than at bridge sites (3.96 ± 1.0) during the summer. Similar but nonsignificant trends were found during the winter. Detection probabilities were significantly higher at random sites than at bridge sites in both seasons. Our results indicate that surveys of bridge sites for river otters may yield inaccurate results for distribution and relative abundance, particularly if conducted during the summer.

We used scat analysis to determine diets and relative nutritional values of diets for black bears ( Ursus americanus Pallas, 1780) in Rocky Mountain National Park, Colorado, from 2003 to 2006, and compared foods consumed and nutritional components to identify important sources of fecal gross energy (GE), crude fat (CF), and fecal nitrogen (FN) in annual and seasonal diets. Patterns of use of food classes followed typical seasonal patterns for bears, although use of animal matter was among the highest reported (>49% annually). Use of animal matter increased after spring, although crude protein levels in bear diets were always >25%. GE was typically lowest for grasses and other herbaceous plants and highest for ants and ungulates; FN was strongly positively related to most animal sources, but negatively correlated with vegetative matter; and CF showed the strongest positive relationship with ungulates and berries, with the latter likely influenced by the presence of seeds. Compared with historic data (1984–1991), contemporary diets included substantially greater prevalence of anthropogenic foods, which likely contributed to increases in size, condition, and productivity of the contemporary bear population. Management strategies are needed to increase quantity and quality of natural foods while minimizing dependence on anthropogenic sources.

Postsealing population recovery rates of fur seals and sea lions have differed markedly, perhaps owing to habitat type. Australian fur seals ( Arctocephalus pusillus doriferus Wood Jones, 1925) employ a benthic foraging mode similar to sea lions, and have exhibited similarly slow population recovery. Nonetheless, the population doubled in recent decades, suggesting a recent change in demographic rates. In the present study, the frequency and size of known-age females (n = 297) were used to create body growth and survivorship models. These were compared with models obtained in the 1970s before the recent population increase. Body growth, which is relatively rapid in comparison to other fur seal species, remains unchanged since the 1970s, suggesting that density-dependent effects are absent despite the population increases. Adult survival rates (weighted mean: 0.885) have increased greatly since the 1970s and are the likely mechanism of the recent increases. Total population abundance was estimated to be 4.5 times that of pups. Australian fur seals display high survivorship, rapid body growth, low fecundity, and low population growth rates; all are characteristics typical of benthic foraging sea lions rather than other fur seals.

The population decline of Steller sea lions ( Eumetopias jubatus (Schreber, 1776)) may be linked to a decline in juvenile survivorship. Limitations in prey availability may contribute to the decline, thus it is important to understand fasting capacities of Steller sea lions. For most mammals, fat catabolism is the preferred energetic pathway to ensure that protein is spared. However, marine mammals also have a conflicting requirement to conserve fat because the main site of fat storage is the blubber layer, which is also their primary thermal barrier when at sea. We developed a dynamic state variable model to demonstrate how protein and fat reserve utilization and maximum fasting duration are influenced by body condition and time spent foraging. This model was parameterized with respect to conditions faced by juvenile and subadult Steller sea lions foraging unsuccessfully during a period of reduced prey availability. The model accurately predicted changes in fat and protein mass of juvenile and subadult Steller sea lions fasting in captivity. Furthermore, the model demonstrated that body lipid content, body mass, and the proportion of time spent in water influence energy reserve catabolism and maximum fasting durations. Consequently, small, lean individuals are particularly susceptible to reductions in prey availability.

Abundance, social organization, and population trend of the arctic wolf ( Canis lupus arctos Pocock, 1935) in north and east Greenland, 1978–1998, were determined from 353 sightings of 552 wolves by the Danish military, by expeditions, and from 8 consecutive years (1991–1998) of fieldwork. Available evidence suggested that this wolf population consisted of up to 55 wolves in favorable times. Six core packs were identified. Maximum wolf density was estimated at 1 wolf/3745 km2, which appears to be the lowest wolf density reported, representing 3.5% of maximum late winter wolf density in Denali Park, Alaska, and 4 wolves were rare, constituting 3.8% of early winter sightings. The population increased, on average, 8% per year during the period 1978–1991 and appeared to reach a peak in 1991–1992. These depressed population characteristics are likely the consequence of the lowest ungulate prey availability in North America, e.g., 2.6% of that of wolves in northeastern Minnesota.

The raccoon ( Procyon lotor (L., 1758)), a generalist mesocarnivore, has thrived in anthropogenically fragmented landscapes. In this study we utilized 13 microsatellites to analyze the population genetic structure in raccoons occupying an agriculturally fragmented habitat. Our analyses revealed low, but significant, levels of genetic structure (FST = 0.022, P < 0.001) among 645 raccoons inhabiting 29 spatially distinct habitat patches (neighborhoods). We found no evidence for the presence of either isolation by distance or barriers. However, spatial autocorrelation revealed evidence for social structure at the raccoon-neighborhood scale. About 30% of the variance in levels of social structure among raccoon neighborhoods was explained by effects of habitat-patch size and the percentage of land cover represented by forest around the individual habitat patch. We found strong evidence of sex-biased dispersal using direct (parentage analysis) and indirect (relatedness and assignment indices) genetic estimators. However, our analysis revealed that among dispersing raccoons there was no difference in dispersal distance between sexes and that dispersal events are randomly distributed with respect to distance. Our study provides evidence that demographic and behavioral processes affecting raccoon spatial organization may be most critical at very fine scales of biological organization, such as within individual habitat patches.

Habitat preferences of river otters ( Lontra canadensis (Schreber, 1777)) are well known, but because most studies were conducted in regions with markedly low or high levels of anthropogenic disturbances, it is not well known how their habitat usage is affected by varied anthropogenic disturbances and land-use regimes on a regional scale. We studied habitat use by otters in eastern New Brunswick, Canada, in an area having both protected and disturbed riparian habitats. Using long-range winter riparian transects, we documented activity-sign distribution along riverbanks in relation to 12 habitat factors and 9 categories of anthropogenic disturbances. We documented variables at site with activity signs and at habitat stations along riverbanks at 500 m intervals. We used logistical regressions and Akaike’s information criterion in an information–theoretic approach to compare models and determine the important factors involved. Habitat-related factors were more important than anthropogenic ones in describing habitat use. The best performing models were those incorporating both habitat and anthropogenic factors. Beaver ( Castor canadensis Kuhl, 1820) ponds were the most important habitat factor, while fields were the most important anthropogenic factor. Our results indicate that otters responded mostly to the presence of habitat features they use and secondarily to the presence of anthropogenic structures or activities in an area.

Predation on Pacific salmon by bears (genus Ursus L., 1758) can be an important ecosystem process because the spatial distribution of carcasses largely determines whether marine-derived nutrients cycle through aquatic or terrestrial pathways. Direct observations on three streams in southeastern Alaska indicated that 49% of the pink ( Oncorhynchus gorbuscha (Walbaum, 1792)) and chum ( Oncorhynchus keta (Walbaum in Artedi, 1792)) salmon killed by bears were carried into the forest. The tendency of bears to transport carcasses was independent of the sex and species of salmon, but unspawned fish were more often transported than fish that had completed spawning. Data on tagged sockeye salmon ( Oncorhynchus nerka (Walbaum in Artedi, 1792)) in one southwestern Alaska stream indicated that 42.6% of the killed salmon were transported, and that higher percentages were transported in years when salmon densities were greater. At six other streams, on average, 68.1% of the sockeye salmon killed were apparently transported away from the stream into the forest. Combining the data from all sites, the proportion of carcasses transported increased with water depth at the site. These results emphasize the role that bears play in mediating the interactions between nutrients from salmon and the terrestrial and aquatic ecosystems, and the variation in carcass distribution among streams and among years.

In this paper we explore the cyclic interactions of prey–predator systems by examining the relationship between cycle lengths of both species and the strength of their interaction. As a probe of interaction strength, we use the degree of counter-clockwise rotation in phase plots with the prey on the x axis and the predator on the y axis. We compare the results from a 25-year time series from the Hudson’s Bay Company data on American mink ( Neovison vison (Schreber, 1777)) and muskrat ( Ondatra zibethicus (L., 1766)) with results from three simulation models. We found that the strength of interaction (rotation range: –0.4 to 1.1 rad/year) was strongest when the two cycle lengths were similar and that it increased with the amplitude of the cycles (cycle range: 4–10 years). The time difference between prey and predator cycles that corresponded to the highest interaction strength was 2–3 years. Similar results were obtained with simulation models; the most complex Hanski model showing the overall best fit with observations. However, none of the models were able to reproduce long ranges of stable cycles by only changing one of their parameters at a time (ranges 2–4 years), whereas the observed range of stable cycles was 4–10 years.

With a changing Arctic climate, it is important to know whether polar bears ( Ursus maritimus Phipps, 1774) can supplement their stored fat reserves by intake of terrestrial berries. Although polar bears are known to consume berries while on land, it has been difficult to quantify their relative dependence on stored adipose tissue and berry carbohydrates to meet their energy needs. We sampled breath CO2 from 300 bears fasting on land in western Hudson Bay during the open-water seasons of 1997 and 1998 and analyzed this breath for stable carbon isotope (δ13C) values. We found no difference in the bear breath δ13C values for bears known to have recently fed on berries and those that had not. The distribution of δ13C values was remarkably tight with average values ranging from –24.2‰ to –24.8‰ with no effect of age, sex, or year of capture. This result was counter to our simple isotopic discrimination model that predicted bears metabolizing adipose tissue derived from seals would have breath δ13C values close to –24.7‰ and those metabolizing berries exclusively would have values close to –32.6‰. If correct, our results suggests that the bears which had fed on berries while fasting on land received an insignificant amount of energetic benefit.

Factors influencing the outcome of interspecific interactions between sympatric carnivores, along with population-level consequences, are not clearly understood. The reintroduction of gray wolves ( Canis lupus L., 1758) to Yellowstone National Park provides a rare opportunity to study interactions with coyotes ( Canis latrans Say, 1823), which had lived in the absence of wolves for >60 years. We evaluated direct interactions between wolves and coyotes to identify factors influencing the outcomes of interspecific interactions and describe the context and degree of competition and coexistence. Using radio-collared wolves, we documented 337 wolf–coyote interactions from 1995 to 2007. The majority (75%) of interactions occurred at ungulate-carcass sites. Wolves initiated the majority of encounters (85%), generally outnumbered coyotes (39%), and dominated (91%) most interactions. Wolves typically (79%) chased coyotes without physical contact; however, 25 interactions (7%) resulted in a coyote death. Interactions decreased over time, suggesting coyote adaptation or a decline in coyote density. In the majority (80%) of fatal interactions, wolves outnumbered coyotes. However, wolves did not outnumber coyotes in interactions (n = 18) where coyotes chased or attacked/harassed wolves. Our results suggest that wolves are the dominant canid, group size may influence the outcome of interactions, and coyotes must benefit from the access to carrion at wolf-killed carcasses.

Within the predator guild, wolverines ( Gulo gulo (L., 1758)) have evolved as generalist predators and scavengers on prey killed by other predators. The recovery of wolves ( Canis lupus L., 1758) in the boreal forests of southern Norway during the late 1990s may have triggered consequent recolonization by wolverines through increased carcass availability. We investigated winter foraging behavior of wolverines in the boreal forest with regard to wolf, lynx ( Lynx lynx (L., 1758)), and red fox ( Vulpes vulpes (L., 1758)) presence. We followed 55 wolverine tracks in the snow from at least nine individuals for a total of 237 km during the winters of 2003–2004. We documented 19 moose ( Alces alces (L., 1758)) and 4 bird carcasses, and no successful hunts. Wolverines did not follow guild species trails directly to carcasses; however, they did change their movement patterns after red fox encounters. While wolverines were more active at higher elevations, the probability of encountering a wolf was higher at lower elevations, suggesting a spatial separation between wolverines and wolves. Although wolverines seem to depend on wolf for carrion during winter, they did not use wolf trails to find carcasses. This may indicate that wolverines reduce risk of intraguild predation by avoiding direct confrontation with wolves.

Recent genetic studies suggest that in northern Minnesota two species of wolves (Canis lupus L., 1758 or western wolf and Canis lycaon Schreber, 1775 (= Canis rufus Audubon and Bachman, 1851) or eastern wolf) meet and hybridize. However, little morphological information is available about these two types of wolves in Minnesota. We analyzed the mass of 950 female wolves and 1006 males older than 1 year from across northern Minnesota and found that it increased from 26.30 ± 0.56 kg (mean ± SE) for females and 30.60 ± 0.72 kg for males in northeastern Minnesota to 30.01 ± 0.43 kg for females and 35.94 ± 0.45 kg for males in northwestern Minnesota (females: r2= 0.79, P < 0.02; males: r2= 0.63, P = 0.06). These mass differences add morphological information to the identities of eastern and western wolves and support the view that ranges of the two species meet in Minnesota.

Predators commonly select prey of inferior quality compared with the average animals found in the population. Consequently, predation may often be compensatory to other sources of mortality. We tested whether wolves (Canis lupus L., 1758) in Scandinavia selected moose in poor body condition by comparing mandibular marrow fat (MMF) of wolf-killed moose (Alces alces (L., 1758)) with harvested moose. Model selection analyses indicated that MMF levels were lower for wolf-killed moose compared with harvested moose, but an unbalanced sample design between years for the two causes of death (wolves and harvest) may have confounded a clear interpretation of the results. Nevertheless, corrected MMF levels (to 1 April) showed that a significantly (p = 0.006) higher proportion of the wolf-killed calves (18.5%) were below the level often assumed to reflect acute malnutrition (

The surface preparation technique was used to estimate the sensory cell population and density in the organ of Corti of seven harp seals and four ringed seals. The average total of inner hair cells for the harp seal was 3654 (3078–263) as compared with an average total of 3232 (3120–3354) in the ringed seal. The average total number of outer hair cells in the harp seal was 14 318 (12 173 – 15 709) as compared with an average total of 13 497 (12 903 – 14 894).The distribution of outer and inner hair cells showed an increase in density from base to apex. An increase in density of about 21% and 29% was observed in the inner hair cells of the ringed and harp seal. The increase in density for each row of outer hair cells was 21% in the harp seal and 17% in the ringed seal. The density of outer hair cells per unit surface area decreased from a maximum value at the base to about half its value at the apex.The average total sensory cells of seals exceeded the average total sensory cells of both man and dolphin but were within the range of variation of the human.

The respiratory characteristics of mitochondria isolated from the subcutaneous brown adipose tissue of newborn harp seals indicate that the tissue is thermogenically active. Temperature recordings in vivo revealed, in fact, that the tissue was maintained at a temperature close to that of the body core during immersion of the pups in ice-water. Beta-adrenergic blockade markedly increased the cooling rates at both locations in ice-water, while curarization, accompanied by artificial respiration did not. We conclude that nonshivering thermogenesis through activated brown adipose tissue plays a decisive role in the defence against cold in the newborn harp seal.

The normal heart rate of young breathing harbor seals, 130.1 beats/min (S.E.M. = 22, n = 33), decreases by 50% during periods of spontaneous apnea in air and by 69% during 2 min of forced apnea in air. Apneic bradycardia develops five times more slowly than bradycardia observed during immersion. During forced apnea the heart rate drops to 31% of the value observed during periods of breathing in 200 s while during immersion the heart rate drops to this value only after 20 s. Since the bradycardia during apnea has a slower time course, apnea alone cannot account for the bradycardia observed during immersion.In quietly resting seals, the apneustic breathing pattern consisted of periods of breathing (duration of 10–115 s) interrupted by apneic pauses (duration of 19–104 s). During the breathing periods successive breaths tended to be smaller so that at the end of a breathing period the mean amplitude of a respiratory movement was only 57% of the mean amplitude of the first inspiration. Tracheostomy altered the breathing pattern and lowered the mean heart rate during breathing to 43% of the rate observed in intact animals. Under controlled conditions immersion bradycardia was highly reproducible and showed no signs of conditioning to the experimental regime. Data from 56 immersions on three animals illustrated minor individual differences in the time course of the immersion bradycardia and provided the basis for an estimate of the average time course of immersion bradycardia in young harbor seals.

The population and density of the bipolar ganglion cells were determined from serial horizontal sections and graphic reconstructions of the cochleas of five captive harp seals. The [Formula: see text]-turn spiral ganglion forms a continuous ring throughout its course except at the extreme basal end where it is narrowest. The nerve cell body is 25 μm long (16.1–38.8 μm) and 16 μm wide (10–24 μm). The average number of ganglion cells present was 57 185 (46 389 – 70 952), with a corrected total number of 52 000 ganglion cells. Two peaks are present in the density curve of the ganglion cells. The first was at 1–1.5 mm and the second at 20 mm, where 2620 cells/mm2 and 2250 cells/mm2 respectively are present.The ratio of total ganglion cells to total sensory hair cells was about 3:1. This ratio was not uniform throughout the length of the cochlea; it was 6:1 at 2–3 mm from the basal end and declined gradually to 3:1 at the apical end. The average total of ganglion cells in the harp seal exceeded the average value in humans, but did not exceed the values found in dolphins.

The distributions of alanine aminotransferase (AIAT), aspartate aminotransferase (AspAT), ornithine carbamyl transferase (OCT), and sorbitol dehydrogenase (SDH) were determined in the tissues of ringed seals, Phoca hispida, harp seals, P. groenlandica, and grey seals, Halichoerus grypus. In all species, AIAT, OCT, and SDH were concentrated primarily in the liver, while AspAT had a more widespread distribution. Species differences were noted in the distribution of AspAT. In grey and harp seals cardiac muscle generally contained the greatest concentration of activity while in ringed seals the highest activity was found in the liver. Reference plasma values were determined in six clinically normal captive ringed seals.

Measurements were made of the relation between body temperature, respiration, diving pattern, and water temperature using three adult harp seals (Phoca groenlandica). Core temperature, ventilation, gas exchange, and diving pattern were not affected by water temperature ranging from 1.8 to 28.2 °C. The basal metabolic rate of the seals in this study was not significantly different than that predicted for terrestrial mammals of equivalent size. Expired heat loss and convective heat loss from the body accounted for 16% of the heat loss and were independent of water temperature. The remaining 84% of the heat was lost from the flippers which accounted for less than 14% of the surface area. Calculations of the ratio of convective heat transfer coefficients of the blubber layer indicate that internal conduction rather than external convection governed heat loss.

The upper frequency-discrimination limits of two ringed seals were estimated by examining each seal's ability to differentiate between constant frequency and frequency-modulated pulses. The upper frequency limit was found to be 60 kHz for both seals. As a result, in the frequency range above 60 kHz pitch and spectral content information will not be differentiated.

Three harbor seals were immersed in five different water temperatures ranging from 7 to 37 °C. Unlike the response in men, the bradycardia observed during immersion was not influenced by water temperature. When the faces of four seals were rendered insensible by neurotomy and local anesthetic, the bradycardia during immersion followed the same time course as that observed during periods of forced apnea while the seals were in air. In two seals with insensible faces, which were forcefully respired during immersion, the heart rate did not drop from the rate observed before immersion. The data presented are consistent with the hypothesis that the two principal sensory contributions to the bradycardia of immersion are: (1) a cessation of the movements of the lungs and rib cage; (2) a change in neural activity in receptors located in the face.

Immobilizing darts containing either of the narcotics fentanyl or etorphine were used on 15 adult hooded seals (Cystophora cristata Erxleben). Of these seals, 13 were successfully tagged. The narcotic antagonist naloxone was administered to all seals. No deaths were recorded and seals which were observed up to 2.5 h after darting behaved normally. Based on visual estimates of weight a satisfactory dose of fentanyl for female hooded seals varied from 0.3 to 0.6 mg/kg.

Plasma and tissue levels of dietary radiotracer-labelled tocopherols were examined in the harp seal, Phoca groenlandica. The d-α form showed much higher plasma maxima and retention than d,l-γ-tocopherol. Concentration peaks occurred somewhat later but were greater when the vitamin was fed with Tween 80 than with herring and corn oil respectively. Greater plasma levels occurred with the use of herring oil as compared with corn oil. Seals previously maintained on a vitamin E deficient diet showed greater plasma maxima than seals not so deprived. Plasma and tissues generally showed maximum levels within 24 h postingestion. Tissue levels of labelled dietary α-tocopherol were examined by biopsy and postmortem studies in selected tissues of vitamin E deprived seals.

Throughout its circumpolar range, the polar bear (Ursus maritimus) feeds predominantly on the ringed seal (Phoca hispida). Despite seasonal variation in availability and distribution of seals, polar bears prefer to eat mainly the fat, often leaving substantial portions of seal meat and blubber. In the present study, 12 seals were minced and analyzed for water, fat, protein, and ash contents. The percentage composition varied from 23 to 58% protein, 34 to 76% fat, 2 to 5.5% ash, and 47.4 to 69.5% total body water. The caloric values varied from 2.3 to 5.3 kcal/g wet weight. Relationships between total body water and fat (%), body water (litres) and protein (kg) are presented. These results are discussed in relation to the ecology and hunting behavior of the polar bear.

The retina of the harp seal (Pagophilus groenlandicus) was studied by means of the light microscope. Ganglion cells occupy a single layer. Thinly dispersed throughout this layer are giant ganglion cells. There is no area centralis. The inner nuclear layer consists of large horizontal cell processes with bipolar and amacrine cells between the horizontal cell processes. The outer nuclear layer is the thickest of all retinal layers. Its density is constant in the central and peripheral areas of the retina, similar to that found in the inner nuclear and ganglion layers. Only rod photoreceptors were found; therefore it is presumed that seals have no color vision. The tapetum covers an extensive area and is 32–34 cellular layers thick centrally, diminishing in thickness peripherally. The combination of tapetum and rod receptors makes possible excellent visual sensitivity to dim light.

The fatty acids of phospholipids and of triglycerides from the lungs and heart of the harbor seal Phoca vitulina concolor De Kay show distinctive differences from the depot (blubber) fat. Primarily, this is due to greatly elevated levels of 20:4ω6 in the phospholipids. Although these have about the same total C20 and C22 polyethylenic acids with five and six double bonds as the depot fat, they occur in different proportions emphasizing 20:5ω3 rather than 22:6ω3. The organ triglycerides resemble the phospholipids in saturated fatty acid composition and the depot fats in monoethylenic fatty acid composition, but have remarkably low levels of 20:5ω3, 22:5ω3, and 22:6ω3 compared to the depot fat. Comparisons are made with phospholipids from the lungs of the leatherback turtle Dermochelys coriacea coriacea Linnaeus but the novel trans-6-hexadecenoic acid found in the turtle lipids could not be detected in the seal lipids.

Ectrodactylism in the hind limbs of a pinniped is described and discussed. This malformation is characterized by defective development of the distal phalanges which result in shortened, thick digits terminating in bulbous swellings. It is unlikely that such flippers are damaged as a result of some physical force; rather, the malformation appears to be of a congenital origin. To the authors' knowledge, this is the first report of a congenital limb deformity in a marine mammal.

A free-field air audiogram from 1 to 32 kHz was obtained for a Pagophilus groenlandicus trained to respond to pure tone signals. The lowest threshold was at 4 kHz at a level of 29 db//0.0002 dynes/cm2. The air audiogram was generally flat. The critical ratios at 2 and 4 kHz were 10%. The lumen of the external auditory meatus is probably acoustically blocked. The outer and (or) middle ear structures and their acoustic impedance mismatch with the air are believed responsible for the comparatively irregular and slightly insensitive hearing of the seal in air.

The trachea of the river otter and sea otter are morphologically similar to each other, whereas the trachea of the harp seal is relatively shorter and broader and has noncalcified overlapping cartilaginous rings and a relatively thicker lamina propria. The harp seal has fewer cartilaginous rings, the arms of which overlap dorsally in the cranial two-thirds of the trachea. The bronchial blanching patterns are similar in the river otter and sea otter, with the pennate-like branching of the lobar bronchi corresponding to the lobes of the lungs. In the harp seal there are more dichotomous branchings of the lobar bronchi, with close symmetry between the right and left lungs. Bronchiograms and dissections indicate that the numbers of subsegmental bronchioles increase from the river otter to sea otter to harp seal.

Terrestrial and aquatic locomotion of the river otter (Lutra canadensis), the sea otter (Enhydra lutris), and the harp seal (Pagophilus groenlandicus) are successively described and analyzed in relation with the external characteristics of these species. Terrestrial locomotion, typical of land carnivores in Lutra and Enhydra, is different and extremely awkward in Pagophilus. In the water, only Lutra has to actively maintain itself at the surface; the other two species have been observed at the surface for long periods with no movements. There is an increased tendency from Lutra to Enhydra to Pagophilus toward enlarged highly specialized feet, approximating a perfect lunate border. Tail surface area decreases in the above order. The length and density of foot hairs decrease if compared with midback and tarsal regions and also from Enhydra to Lutra to Pagophilus.

The azygos venous system of one Weddell seal (Leptonychotes weddelli) has been described. Both left and right azygos trunks were present with segmental anastomoses between them. The right azygos trunk was the larger and opened into the anterior vena cava; the left azygos trunk terminated in the left costocervical vein. At each segmental level the azygos trunks communicated with collecting sinuses into which drained blood from intercostal spaces and the extradural intravertebral vein. These collecting sinuses were themselves connected to each other by longitudinal anastomoses. The relationship of this system to blood flow during diving is briefly discussed.

The large, heavy ossicles of Pagophilus groenlandicus are firmly attached to the epitympanic wall by a single fold of epitympanic mucosa. The epithelium of the middle ear mucosa is pseudostratified columnar epithelium. Large cavernous sinuses are present below the dense collagenous fibers of the submucosa. The cavernous sinuses are prominent around the promontory and the medial wall of the tympanic orifice of the auditory tube, and also in the tympanic sulcus of the tympanic membrane. The distension of the cavernous tissue may aid in hearing.The wall of the auditory tube consists of thick collagenous and elastic fibers. The medial wall is a thick fibrous block of connective tissue. Hyaline cartilage forms the central core of the thick fibrous medial wall of the auditory tube. The robust wall of the tube makes it unlikely that the tube wall is collapsed by the increase pressure in the nasopharynx during diving. Pressure equilibration during diving may be achieved by the opening of the nasopharyngeal orifice by muscular contraction. The closing of the tube may be achieved by the elasticity in the tube wall.

Red cell, plasma, and total blood volumes were determined in three ringed seals, Phoca hispida, through simultaneous labelling of both red cells and plasma. Total blood volumes were 139, 140, and 158 ml/kg of whole body weight ([Formula: see text] ml/kg). Lean body weight was determined in one seal; blood volume was calculated as 226 ml/kg lean body weight.

Ringed seals were observed to inhale exhaust air expelled by SCUBA divers who were diving under 1.9-m thick, 1-year-old fast ice in Lancaster Sound during mid-April, 1973, The underside of the sea ice was imperfectly flat so that some of the bubbles of exhaust air coalesced into large lenticular bubbles 1 to 3 cm in depth, and these, on occasions, were partially inhaled by seals. Between dives, separated in time by as little as 3 h, all the under-ice air bubbles, large and small, vanished. It is assumed that the buoyant air migrated upward into the sea ice by the progressive displacement of brine in drainage channels. Although the exhaust air was readily used by ringed seals, it is unlikely that naturally occurring air pockets can persist beneath 1 -year-old sea ice to enable seals to extend their under-ice travels.