MORTALITY TRENDS IN NORTHERN SEA OTTERS (ENHYDRA LUTRIS KENYONI) COLLECTED FROM THE COASTS OF WASHINGTON AND OREGON, USA (2002–15) Open Access

Northern sea otter (Enhydra lutris kenyoni) populations in Washington State have recovered from previous declines and occur primarily within the relatively undeveloped Olympic Coast National Marine Sanctuary. The population continues to steadily increase at a rate of 8.6% annually, with the majority of the population (68%) now residing in the south segment (south of La Push, Washington; Jeffries et al. 2016). Human development to the south (coastal Grays Harbor County) and east (e.g., Puget Sound and Strait of Juan de Fuca) of the core range has been moderate to extensive, which can disturb nearshore ecosystems (through dredging, erosion, and sedimentation; Thom and Hallum 1990; Waycott et al. 2009; Gaeckle et al. 2011), directly disturb otters (e.g., human water-based activities), and create potential for increased transmission of terrestrial pollutants or diseases to otters (Jessup et al. 2007). Sea otters are apex predators of nearshore systems, and knowledge of their health and population status is a valuable indicator of coastal system changes and health (Jessup et al. 2004).

We examined demographic, temporal, and geographic trends in causes of mortality from infectious disease, trauma, and other health conditions in beachcast northern sea otters from 2002 to 2015. We also qualitatively compared mortality trends among the northern sea otter population residing in Washington and the southern sea otter population (Enhydra lutris nereis) in California, which inhabits a much more developed coastal environment, to encourage continued examination of the roles of human development and coastal disturbance on disease transmission in marine environments.

Sea otter carcasses were recovered through a stranding network established by the US Fish and Wildlife Service and Washington Department of Fish and Wildlife. These agencies responded to reports of dead and moribund otters along the northern Washington coastline in collaboration with federal, state, and county jurisdictions, tribes, and local volunteers. Stranding location data included the county and description of the area (e.g., name of beach or closest access road). Otters estimated by field biologists to be in good post-mortem condition (dead <4 d) or stranded alive but that died or were euthanatized shortly afterwards during 2002–15 were examined by veterinary pathologists at US Geological Survey–National Wildlife Health Center (NWHC; Madison, Wisconsin, USA).

The estimated age of otters was determined at necropsy by combined assessment of weight, total length, extent of grizzling (hair on head and neck that whitened with age), and teeth condition as previously described (Garshelis 1984). Otters were assigned to age groups: immature (<1 yr), subadult (1–3 yr), or adult (≥4 yr). Complete gross necropsies and histopathologic examinations were conducted on all otters including microscopy of tissue sections from major organs to determine the cause of death (COD). Ancillary tests, including radiographs, parasite identification, and bacterial and fungal cultures were performed as needed. Routine diagnostic testing to identify infectious agents was performed at NWHC using standard culture, histopathology, and molecular identification techniques (i.e., sequencing), except where detailed in the upcoming text. When more than one significant health condition was detected, primary COD was assigned by the examining pathologist based on expert assessment as to which health condition or conditions were sufficiently severe to have resulted in death.

Specific histopathological and laboratory testing criteria were applied to diagnosing specific conditions in sea otters, including protozoal encephalitis caused by Sarcocystis neurona or Toxoplasma gondii, morbilliviral encephalitis, leptospirosis, dilated cardiomyopathy, and trauma (and assignment of likely traumatic origins). Diagnostic criteria for these conditions, which made up a large portion of the CODs recorded, are described here briefly.

Protozoal encephalitis was diagnosed as primary COD using previously described histopathological criteria and immunohistochemistry (IHC; Thomas et al. 2007). Briefly, S. neurona encephalitis was characterized by multifocal to diffuse gliosis concentrated in gray matter and molecular layer of the cerebellum. Toxoplasma gondii encephalitis was characterized by discrete foci of gliosis and malacia usually widely separated and found predominantly in the cerebral cortex. The organisms within or at the edge of a lesion were identified by morphology of the active life stage (merozoites for S. neurona and tachyzoites for T. gondii) and then confirmed by IHC (Dubey et al. 2015), performed at the US Department of Agriculture, Animal Parasitic Diseases Laboratory (Beltsville, Maryland, USA).

Morbillivirus infection was diagnosed using a combination of histologic and diagnostic assay results. The histopathologic features of the disease were inflammation that was prominent in white as well as gray matter of the brain (encephalitis), particularly affecting the brainstem, cerebellar medulla, and corpus callosum; astrocyte hypertrophy; spongy change in white matter; and the presence of eosinophilic intracytoplasmic and intranuclear inclusion bodies. A pan-morbillivirus real-time PCR followed by restriction endonuclease digestions (Saliki et al. 2002) or sequencing on positive PCR products to determine the specific morbillivirus (Sierra et al. 2014) was performed at Oklahoma Animal Disease Diagnostic Laboratory (Stillwater, Oklahoma, USA) or Athens Veterinary Diagnostic Laboratory (Athens, Georgia, USA). Paraffin block sections (4 μm thick) of brain including cerebellum, brainstem, and cerebrum from each otter were examined for morbillivirus antigens using commercially available anti-canine distemper virus (CDV) monoclonal antibodies at Oklahoma Animal Disease Diagnostic Laboratory (DV2-12; Custom Monoclonals, Inc., West Sacramento, California, USA) or Athens Veterinary Diagnostic Laboratory (CDV-NP, VMRD, Inc., Pullman, Washington, USA). Animals with characteristic brain lesions and brain tissues with positive PCR and IHC tests results were diagnosed with morbilliviral encephalitis.

Leptospirosis was diagnosed histologically when Leptospira sp. spirochetes and interstitial nephritis were observed in the kidney via the use of Steiner and H&E stains, respectively. A leptospirosis infection was confirmed when spirochetes were also detected via a real-time PCR specific for the lipL32 gene of pathogenic Leptospira spp. (Stoddard et al. 2009) performed at Wisconsin Veterinary Diagnostic Laboratory (Madison, Wisconsin, USA).

Streptococcus phocae septicemia and associated suppurative lesions were diagnosed based on bacterial isolation from lesions, liver, or both using culture isolation followed by molecular sequencing. Dilated cardiomyopathy was diagnosed when grossly enlarged, dilated atria and ventricles were observed, often accompanied by chronic passive congestion in the liver. Trauma was attributed to shark attack when wounds sufficient to cause death occurred in a series, arched array, or in opposing positions indicative of shark bite. Trauma was attributed to boat propellers when wounds were long, parallel, consecutive, and deeply lacerating. Trauma was attributed to gunshot when there were wounds indicative of gunshot injury and metal pellets or fragments were found in wounds or seen in radiographs. Death was attributed to drowning from net entanglement when an otter was found in a fishing net and the major observations at necropsy were consistent with drowning, including compressed empty lungs with pulmonary hemorrhage or wet, heavy lungs.

Proportionate mortality, noted as percent, was used to identify major causes of death in northern sea otters during 2002–15. Cause-specific proportionate mortality was summarized and compared for demographic, temporal, and geographic trends. Annual (2002–15) northern sea otter population estimates accessed from Washington Department of Fish and Wildlife's published surveys (2016) were used as denominators to calculate stranding report and necropsy rates to evaluate consistency and representativeness of post-mortem sampling. The distribution of the established population of Washington sea otters ranges from approximately Destruction Island to Makah Bay, Washington. To examine spatial patterns in COD of sea otters, we used La Push, Washington, to divide the population into northern and southern segments. La Push was the 1970 release site for otters translocated to Washington (Lance et al. 2004) and has been used as the geographical dividing point during annual populations surveys to examine changes in the distribution and growth rates of the northern and southern segments of the population (Jeffries and Jameson 2015). Since population estimates were not available for 2009 (Jameson and Jeffries 2009), population size was imputed in R version 3.2.1 (R Core Team 2015) from a logistic model fit to available population estimates from 2002–15 (y=128,259×ln(x)−974,408, R²=0.87). The percentage of the population in the northern and southern segments was imputed for 2009 as the average of the proportions north and south of La Push, Washington, from 2008 and 2010 counts.

The data associated with this publication can be found at https://doi.org/10.5066/F7NK3C6R.

A total of 333 moribund or dead sea otters from the Washington State population were reported during 2002–15, and 93 (28%) of these were recovered and necropsied at NWHC. Otter stranding reports and carcass retrieval occurred along the full extent of the Washington outer coast (see Supplementary Material Fig. S1). Although there are no permanent sea otter residents in Oregon, 11 beachcast otters were reported (of which two were necropsied) on the coast of northern Oregon (Fig. S1).

Although the Washington sea otter population has continued to increase since 2002 (Fig. 1a), reports of moribund and dead sea otters fluctuated annually between 2 and 5 strandings/100 otters in the population (Fig. 1b). The annual proportion of reported strandings recovered for necropsy ranged between 18% and 51% (median=25% or 0.5 post-mortem exams/100 otters in the population). The number of necropsies performed relative to population size was highly variable across years in both the northern and southern segments of the range (Fig. 1c). Examined sea otters were collected in every month of the year, with peak numbers occurring from late spring to early fall (May–September; Fig. S2).

Of the 93 otters examined, the majority were adults (62%), followed by immature (22%) and subadult (16%) otters (Table 1). The majority of otters (70%) were recovered from the southern segment. There were more than four times as many adult otters necropsied from the southern (n=47) compared to the northern (n=11) segments, and the majority of males (80%) were recovered from the southern segment. Beginning in 2009, there was at least one female recovered from the southern segment each year. In the northern segment the number of recovered males (n=13) was similar to females (n=15).

Infectious diseases were the primary COD (56%) for otters examined during 2002–15 (Table 1). Infectious diseases accounted for 60% of mortality in the southern segment and 46% in the northern segment (Table 1). The primary infectious disease and leading COD was encephalitis due to S. neurona (30%; Table 1). Cases were detected during late-winter through the summer months (February–August), with the largest number of cases diagnosed during May (n=11). Documentation of S. neurona encephalitis as the primary COD in examined otters peaked during 2011 (53%, 9/17), declined during 2012 (36%, 4/11), and then was not documented again as the primary COD until 2015 (40%, 2/5). Sarcocystis neurona encephalitis was the primary COD for about 30% of examined males and females (Table 1). Sarcocystis neurona encephalitis was the COD for 47% of subadults, 29% of adults, and 25% of immature sea otters (Table 1). Toxoplasma gondii encephalitis was uncommon in all examined otters (3%, 3/93).

Streptococcus phocae was involved in all confirmed cases of bacterial septicemia. Associated suppurative lesions in these cases included subcutaneous, intramuscular, or tooth root abscesses (n=3), embolic-septic pneumonia (n=4), suppurative arthritis (n=1), suppurative meningitis (n=1), infectious endocarditis (n=1), suppurative splenitis (n=1), or suppurative peritonitis (n=1). In two of these otters, Pasteurella multocida was also isolated from the same suppurative lesions and was considered a synergistic pathogen in the septic process.

Morbilliviral encephalitis and leptospirosis were each associated with the death of three adult male and three adult female otters (Table 1). Morbilliviral encephalitis was only documented in otters recovered during 2004–08. Otters diagnosed with leptospirosis were recovered primarily in August 2002, with a single otter also diagnosed in April and July of the same year. Trauma was responsible for 14% of the mortality in all examined otters. Although boat strike, gunshot, shark attack, and bite wounds (potentially intra- or interspecies aggression) were documented, the majority of trauma cases were blunt trauma to the skull (the cause of which could not be determined; Table 1).

Heart conditions were determined to be a primary COD for seven otters: six cases of dilated cardiomyopathy (five adult males, one adult female) and one case of congestive heart failure (Table 1). Myocardial inflammation was not a significant feature of the dilated cardiomyopathy cases.

Reports of dead or moribund otters represent an unknown percentage of total mortality in any population since numerous factors can affect carcass detection, including search effort, behavioral changes such as beaching, COD, distance from shore at time of death, wind and current patterns, and otter age. The presence of coastal towns south of La Push, Washington, likely contributed to the increased otter recovery in this area due to increased shoreline access and visibility. Population surveys suggest that the distribution of otters also shifted around 2003, favoring the southern segment of the population (Jameson and Jeffries 2009), which continues to grow at a faster rate than the northern segment (12% vs. 3.7%, respectively; Jeffries and Jameson 2014). The recovery of twice as many males compared to females in the southern area was also expected given that the majority of the male population resided south of La Push.

Infectious diseases caused the majority of deaths in examined otters. The primary infectious disease was protozoal encephalitis. The primary cause of protozoal encephalitis in northern sea otters was S. neurona (28/31) rather than T. gondii (3/31). Our study included four S. neurona– and one T. gondii–positive cases previously reported from the Washington sea otter population (Thomas et al. 2007; Dubey and Thomas 2011). Since opossums (Didelphis virginiana) are the only known definitive host for S. neurona but are terrestrial, infections in otters are presumably caused by spill-over of sporocysts released into the nearshore environment (Rejmanek et al. 2010). In California otters, seasonal peaks in S. neurona cases (March–May) have been shown to coincide with seasonal sporocyst shedding by opossums (Rickard et al. 2001) and maximal seasonal run-off periods for the west coast (Kreuder et al. 2003). The role of direct runoff in transporting sporocysts to Washington otters may, however, be somewhat less straightforward since opossums may have only recently reached the Washington Olympic National Park (Happe et al. 2016), which encompasses much of the terrestrial area adjacent to this otter population. Genetic typing of S. neurona isolates could also be useful for understanding the epidemiology and severity of infections in the Washington otter population. For example, in other marine mammal species, such as harbor porpoises (Phocoena phocoena), Pacific harbor seals (Phoca vitulina), and Steller sea lions (Eumetopias jubatus), individuals infected with S. neurona genotype XIII are significantly more likely to develop severe encephalitis than those infected with genotype VI (Barbosa et al. 2015).

The protozoan T. gondii was diagnosed as the primary COD for only three (3%) otters examined in this study, whereas Kreuder et al. (2003) reported T. gondii encephalitis as the primary COD in 16.2% of California otters examined during 1998–2001. Domestic feral cats (Felis catus), bobcats (Lynx rufus), and mountain lions (Puma concolor) have been proposed as the source of T. gondii oocysts infecting California otters (Conrad et al. 2005; Lafferty 2015). The smaller number of primary T. gondii infections in Washington otters could be due to lower densities or infection rates of definitive hosts in Washington as well as differences in topography that would influence terrestrial runoff patterns. The smaller number could also be the result of pathogen strain virulence or otter diets (certain invertebrate prey concentrates protozoal parasites; Conrad et al. 2005).

Streptococcus phocae has previously been identified as an opportunistic pathogen of California sea otters that typically invades following skin traumas (Bartlett et al. 2016). In four of our seven S. phocae cases, disruption of the cutaneous barrier (skin lacerations or subcutaneous abscesses), mucous membrane disruption (tooth root abscess), or gastrointestinal mucosa damage (hemorrhagic enteritis caused by Vibrio parahemolyticus) were identified as possible entry points for S. phocae. Streptococcus phocae was also isolated from an unhealed fracture and from the liver of another otter where COD was classified as gunshot. The opportunistic pathogen P. multocida also played a synergistic role in the course of the septic processes observed in two otters with S. phocae septicemia.

Other infectious diseases sporadically appeared in the Washington otter population. Leptospirosis was the COD for Washington sea otters examined in 2002. In 2001 leptospirosis exposure was detected via serology in a single Washington otter (Brancato et al. 2009) but was not detected via serology (n=30) or culture (n=6) when animals were tested in 2011 (White et al. 2013). Leptospira sp. are spread in the urine of infected reservoirs such as rodents and can remain infective in contaminated soil and water for weeks (Johnson 1996). Since the bacterium was not cultured, the serovar for these cases was not determined resulting in the inability to determine the source of exposure in 2002. Leptospirosis exposure has been previously reported in free-ranging apparently healthy California sea otters, but leptospirosis-associated mortality has not been reported in this population despite regular outbreaks of leptospirosis in sympatric California sea lions since 1970 (Vedros et al. 1971; Zuerner et al. 2009).

Cases of morbilliviral encephalitis were documented only during 2004–08 despite consistent testing for this pathogen throughout the study period. Morbillivirus is suspected to have played a role in a 2000 mortality event in this population, because, even though most of the carcasses were too decomposed for necropsy, morbillivirus antibodies were detected in 80% of the live otters tested during 2001–02 (Brancato et al. 2009). In 2011 the live otter population had only a 10% seroprevalence of anti-morbillivirus antibodies (White et al. 2013), which also supported our lack of documented morbillivirus cases in this population after 2008. The molecular results from our study demonstrated that this population was exposed to CDV and is in agreement with the conclusions of White et al. (2013), who used morbillivirus titer differentials to conclude that the live otters had also been most likely exposed to CDV.

Despite the presence of multiple pathogens, the Washington population of sea otters has continued to grow steadily, suggesting that this population is tolerating current mortality levels. In comparison, the California population of southern sea otters is exposed to many of the same pathogens, but has experienced more variable growth rates and continues to be listed as federally threatened (Hatfield and Tinker 2015). Differences in causes of death exist between the otter populations. For example, mortality caused by white shark (Carcharodon carcharias) bites has been considered a major limiting factor for the California population in recent years (Tinker et al. 2016). Only two otters from the recovered carcasses of the Washington state population died from shark bite trauma during the 13 yr of our study, and both were recovered in the southern extent of our study area on the northern Oregon coast where warmer waters may support higher shark densities. Even though many factors including differences in carrying capacity, range expansion, and habitat structure can affect otter population growth rates (US Fish and Wildlife Service 2015), comparison of causes of mortality among populations can be useful in developing hypotheses about pathogen transmission mechanisms as well as factors contributing to unexpected population perturbations. The resiliency of this otter population is likely a reflection of the habitat in which they reside, which is adjacent to a largely undeveloped coastline. This contrasts with the California sea otter population where the coastline is more heavily impacted by human use, representing an additional stressor and a potential source of infectious disease exposure. Continued monitoring of the Washington otter population may provide important comparative data for understanding disease ecology in other nearshore marine ecosystems experiencing greater pressure from anthropogenic disturbance and development.

Many of the specimens included in this dataset were reported by members of the public, and we thank them for their time and efforts. We also thank the staff of the Washington office of the US Fish and Wildlife Service, Makah Tribe, Quileute Tribe, Quinault Indian Nation, and state partners for assistance with retrieval and submission of carcasses and the technicians and diagnosticians at the US Geological Survey–National Wildlife Health Center and outside laboratories, particularly O. Kwok, for contributions to the protozoal diagnostics. Use of trade, product, or firm names is for descriptive purposes only and does not imply endorsement by the US Government.

Supplementary material for this article is online at http://dx.doi.org/10.7589/2017-05-122.