Samples were collected at seal haul out aggregations at three locations in the Kazakh part of the Caspian Sea, annually between 2007 and 2017 ( Supplementary Table 1 ). Molting seals (n=46) were sampled on the Durnev Islands (45°30′ N, 52°37′ E) in Komsomolets Bay in spring 2011. Post-molting sampling (n=4) was carried out on the Kendirli islets in Kazakh Bay (42°44′ N, 52°32′ E) in late April and May of 2014-2015. Additional post-molting samples (n=23) were obtained from islets in the vicinity of Prorva (45°53′ N 52°73′ E) in May of 2016. Samples were also taken during the autumn migration of seals at Rybachi sandbank (n=21) to the south of the Kulaly island group (44°76′ N, 50°37′ E) in November of 2007-2008, as well as at Kendirli (n=79) in October and November of 2009-2010, 2012-2013, and 2016-2017.

Live sampling of seals was carried out during satellite tagging studies (Dmitrieva et al., 2016). Seal capture was conducted by deploying seine nets from inflatable boats around haul out sites to entangle seals, or by ‘rush and grab’ of seals on sandbanks with hoop nets. Animals were restrained and handled without the use of chemical sedation, and were released immediately after the completion of sampling and tagging procedures. Animal sampling was regulated by the Resolution of the Government of the Republic of Kazakhstan (Government of the Republic of Kazakhstan, 2004) and the Order of Minister of Agriculture of the Republic of Kazakhstan (Government of the Republic of Kazakhstan, 2004), according to the legislation “Rules for conducting biomedical experiments, preclinical (non-clinical) and clinical studies (№697, 12 November 2007, Republic of Kazakhstan)”, and were approved by the Research and Production Center for Microbiology and Virology (RPCMV) Local Ethics Committee (Approval number: 2015-04-№1-ICPI). All animal handling and sampling procedures were also reviewed by the University of Leeds Animal Welfare Ethical Review Committee.

Body length (from nose tip to tail tip), girth and weight, and sex was recorded for each individual. Seals were categorised as juveniles (<1 year; body length 70–90 cm), sub-adults (older than one year of age; length >91–109 cm), adults (mature, length >110–140 cm). Complete external body examination was made for skin ulcers, parasites, trauma lesions and any visible alterations. Swab samples ( Supplementary Table 2 ) were collected from body orifices (conjunctival, nasal, buccal, rectal, and urogenital) using sterile cotton swabs, which were placed into cryovials with a viral transport medium 199 containing antibiotics (penicillin, 2000 U/ml streptomycin, 2 mg/ml, gentamicin 50 µg/ml nystatin 50 U/ml) and bovine serum albumin at a final concentration of 0.5%. Samples were stored in liquid nitrogen (-196°C) until analyzed. Duplicates of specimens for molecular analyses were preserved in RNAlater.

Blood samples ( Supplementary Table 2 ) were collected from the epidural venous sinuses of the spinal canal (Geraci and Smith, 1975), using the Vacutаiner blood collection system. The venipuncture site was disinfected with 70% ethanol. Depending on the nutritional condition of animals, spinal needles of 18G × 75 mm or 18G × 90 mm size were used for blood drawing. After coagulation, the blood samples were centrifuged (15 min at 3000 rpm), the resulting plasma was transferred into cryogenic polypropylene vials and transported in liquid nitrogen (-196°C).

Swab samples were screened for infection by influenza A and B viruses, morbilliviruses, seal herpesvirus [Phocine herpesviruses] 1 (PhHV-1), and canine coronavirus, parvovirus, adenoviruses, caliciviruses, and Brucella canis at the Laboratory of Viral Ecology of the authors’ institute (RPCMV, Almaty). Animals were visually examined for pox and calicivirus specific skin lesions (vesicles, nodules), seal lice (Echinophthirius horridus) and other ectoparasites. External physical inspection was also conducted for ophthalmic diseases, trauma, clinical signs of infections, and occurrence of aborted fetuses on haul out sites. Presence of IgG antibodies in seal sera was tested for influenza A and B viruses, canine distemper virus (CDV), canine herpes-, corona-, parvo-, adenoviruses, Brucella canis, Leptospira spp. Toxoplasma gondii, and heartworm Acanthocheilonema spirocauda.

Isolation of RNA from blood or swab samples was carried out using a QIAamp Viral RNA Mini Kit (Qiagen GmbH, Hilden), in accordance with the manufacturer’s recommendations (Qiagen Ltd, 2005). DNA was extracted using a Qiagen DNeasy blood and tissue kit.

Reverse transcriptase polymerase chain reaction (RT-PCR) assays were performed on the basis of one-step protocols (Payungporn et al., 2004) using appropriate RT-PCR kits (Access RT-PCR System, Promega) according to the manufacturers’ instructions. Previously published oligonucleotide primers ( Supplementary Table 3 ) were used for amplification of the targeted genes of viruses with a concentration of 0.5 μM each primer, and 5× AMV/Tfl buffer, 5 U inhibitor of RNase, 5 U AMV reverse transcriptase, 5 U Tn DNA polymerase, 1.0 mM MgSO4 and RNase-free water in a final volume of 25 μl. The reactions were performed in an Eppendorf Gradient thermocycler under the following conditions: reverse transcription at 48°C for 45 min, initial denaturation at 95°C for 2 min and amplification over 40 cycles, including denaturation, primer annealing and elongation as specified in Supplementary Table 4 , followed by final elongation at 72°C for 10 min. PCR for DNA viruses was performed with the Q5^® Hot Start High-Fidelity 2X Master Mix (NEB #M0492S) according to the manufacturer’s protocol with an initial denaturation at 98°C 30 seconds and other cycle details given in Supplementary Table 4 .

Live infectious bronchitis virus (IBV) vaccine (Lohman Animal Health, Germany); PhHV-1 (from the European Virus Archive); local isolates of influenza A and B virus; live vaccine containing attenuated canine distemper virus (CDV; Vanguard Plus^® (Pfizer, New York, NY, USA)); canine adenovirus type 1; canine adenovirus type 2; canine parainfluenza virus; canine coronavirus; canine parvovirus; and Leptospira spp. were used as a positive control for PCR screening of viral and bacterial pathogens, and sterile ultrapure water samples were applied as negative controls.

Horizontal gel electrophoresis was performed in a 2% agarose gel (Sigma, USA) stained with ethidium bromide in tris-acetate buffer at a voltage of 88V (8 volts/cm) on a Biostep apparatus (UK). The gel electrophoresis was visualized and documented in the GelMax^® 125 Imager (Upland, CA, USA). The PCR products were sequenced using the fluorescent dideoxy-terminator method on ABI 3500 Genetic Analyzer (Applied Biosystems).

The quality and quantity of viral RNA samples was determined using a Bioanalyzer (Agilent) with 100 ng used to make individual TruSeq Stranded Total RNA with Ribo-Zero Gold rRNA depletion (Illumina, USA) libraries according to manufacturer’s recommendations. Each library was sequenced on a HiSeq 3000, 150 bp paired end lane producing approximately 30 million read pairs per sample, at the University of Leeds Sequencing Facility. The genomes were then assembled with the CLC Genomics Workbench software (Qiagen) using a CDV reference sequence (AF014953) as a guide.

For herpes virus, DNA extracted from swabs was quantified using a Tapestation and NEBNext Ultra whole genome libraries were generated according to the manufacture’s protocol. Libraries were sequenced on a MiSeq, 150bp paired end lane, at RPCMV, Almaty, and contigs assembled with the CLC Genomics Workbench software (Qiagen).

Consensus sequences for PCR amplicons were assembled in MEGA X (Kumar et al., 2018). Sequence identity was confirmed using the BLAST tool suite at https://blast.ncbi.nlm.nih.gov/. For phylogenetic analyses relevant reference sequences were imported into MEGA X, and sequence alignments were generated using the MUSCLE application within MEGA X. The most suitable sequence evolution model was then evaluated with ModelTest within MEGA X. Maximum Likelihood phylogenies were generated for each data set using the indicated best-fit sequence evolution model, with 1000 bootstrap replicates. Trees were visualized and annotated with MEGA X’s Tree Explorer application.

A total of 177 seals were live captured, sampled, and released 2007-2017 ( Supplementary Table 1 ). In total, we obtained 678 swab and 97 serum samples ( Supplementary Table 2 ).

A summary of virus detections via RT-PCR and PCR is given in Table 1 . RT-PCR assays for the CDV P gene detected the expected 429 bp product in five of 13 seals sampled in 2008. All the positive samples were derived from animals under two years old. No PCR positive specimens were detected in any other samples for the subsequent 2009-2017 survey period. Further testing of samples returning negative results with the Pan-paramyxovirus primers specific for the L-gene of all Paramyxoviridae family members also gave negative results.

Conventional RT-PCR testing for the influenza A virus M gene revealed positive ocular and buccal swabs for two animals sampled in 2009. qRT-PCR assays for the Influenza A virus NP gene resulted in specific fluorescent amplification signals (Ct values 23-34) in two swabs collected in October 2016. However, no haemagglutination agents could be isolated following inoculation and passage of material from the 2009 and 2016 samples PCR-positive for the presence of influenza A virus nucleic acids in 10-11 day-old chick embryos.

The expected 290 bp PCR product for PhHV-1 gD gene was detected in oral swabs from five animals sampled in May (three) and October (two) 2016. Positives were verified via real time - PCR with primers to the gD gene of PhHV-1 and Sanger sequencing.

Adenovirus DNA was amplified by consensus nested PCR (Wellehan et al., 2004) targeting the pol gene from pooled nasal and buccal swabs of five Caspian seals sampled in May 2016.

PCR screening for influenza B viruses, coronaviruses were negative in all tested samples.

Pox-like lesions of approximately 2-3cm diameter were observed in three animals sampled in 2010 and 2011, with one having bald skin patches on the head and neck. Other vesicle-like skin lesions were seen in one animal in 2015. Ophthalmic pathologies representing keratopathy or cataracts were noted in four seals sampled in 2007, 2010, 2016, and 2017 ( Supplementary Figure 2A ). No visible ectoparasite infestations or their characteristic dermatological lesions were observed among any captured seals. Potential predator (wolf/jackal) bites were observed on single individuals sampled in 2011 and 2016 ( Supplementary Figure 2B ). Monofilament net entanglement injuries and scars were observed on individuals during most sampling expeditions.

A summary of IgG Antibody prevalence from tested samples is given in Table 2 .

The main periods in which high density of aggregations of Caspian seal occur and provide opportunities for pathogen transmission are during the spring molt, and autumn pre-breeding season period. We aggregated spring and autumn samples across years and compared prevalence of each tested pathogen using Fisher’s exact tests ( Supplementary Tables 7 and 8 ). Only the detections of influenza B virus in HI tests varied significantly (P=0.03), with a higher prevalence in autumn samples (33%) compared to spring (10%). However, these comparisons should be treated cautiously due to small sample sizes in some cases and opportunistic sampling. Similarly, we choose not to formally compare prevalence differences between sex and age classes ( Supplementary Table 6 ) due to the biased nature of sampling during satellite telemetry deployments.

Morbilliviruses of the of Paramyxoviridae family are one of the most frequent causes of marine mammal mass mortality events (Jo et al., 2018). Canine Distemper Virus (CDV; otherwise referred to as canine morbillivirus), has caused confirmed mass mortalities of the Baikal seal (Pusa sibirica) in 1987-1988 in Siberia (Grachev et al., 1989), and was the mostly likely agent responsible for a mortality of Antarctic Crabeater seals (Lobodon carcinophagus) in 1955 (Laws and Taylor, 2009). Antibodies to CDV have subsequently been found in recent serological surveys of Arctic and Antarctic pinnipeds (Kennedy et al., 2019). Canine morbillivirus infection in Caspian seals has been reported since 1997 and caused mass mortalities estimated at thousands of seals in 2000 and 2001 (Forsyth et al., 1998; Kennedy et al., 2000; Kuiken et al., 2006). Recent phylogenetic studies indicated that the CDV isolate recovered from the epizootic in the Caspian Sea in 2000 belongs to a previously undetected novel clade designated as “Caspian” (Jo et al., 2019), which is basal to other known CDV strains.

In our investigation, CDV was found by RT-PCR in five individuals out of 11 seals sampled in 2008. All positive specimens came from individuals under two years old, suggesting an age-dependent susceptibility of animals to the virus, consistent with exposure of naïve juveniles, and resistance in older age classes. No further RT-PCR positive individuals were detected in samples from subsequent years (2009-2017). The lack of CDV-positives in RT-PCR-screening after 2008 may be explained by preferential sampling of adult individuals for satellite tag deployments rather than juveniles.

The RT-PCR positive individuals sampled in 2008 were asymptomatic, and appeared otherwise healthy. However, phylogenetic analysis of complete genomes for the 2008 viruses showed they were derived (average 99.73% identity) from the same epizootic causing “Caspian” clade from 2000, suggesting possible adaptation to reduce virulence. Further characterization of the 2008 strain sequence data is required to evaluate this possibility further. The 2000 “Caspian” strain is not closely related to known CDV strains from terrestrial carnivores (Jo et al., 2019), but there has been limited assessment of CDV strains circulating among wild and domestic carnivores within the Caspian region. Available sequence data for small surveys of CDV strains circulating among dogs in northern Iran in the vicinity of the Caspian shore, indicate they are derived from Arctic and European clades (Namroodi et al., 2015). At this time it is not possible to conclusively determine if the 2000 “Caspian” CDV clade derives from a recent spill over from a terrestrial source, or if it is an endemic strain with long term circulation in the seal population.

Our serological findings indicated maintenance of CDV in the Caspian seal population; antibodies to CDV were detected in six out of 74 seal serum samples (8.1%) between 2007 to 2017, with the most recent positives detected in three adult seals out of 18 sampled in 2016. These findings were obtained with a commercially available ELISA test kit based on monoclonal antibodies against a common epitope of CDV, although a microplate virus neutralization test is considered as “gold standard” in serological surveys of morbillivirus antibodies due to their high sensitivity and specificity. Monoclonal antibody-mediated ELISA tests are also recommended as a useful tool for epidemiological surveillance of morbillivirus infections in marine mammal populations for their comparable sensitivity and specificity to virus neutralization tests, while being simpler to perform and less time consuming (Saliki and Lehenbauer, 2001). Additionally, Saliki and Lehenbauer (2001) demonstrated that Pinniped samples achieve similar levels of performance as Canid sera in ELISA. Hence, we assume the frequency of CDV antibody detection in the Caspian seal serum samples accurately reflects the age class profile for CDV seen in the PCR screening.

Cornwell et al. (1992) reported the persistence of PDV antibodies in harbor and grey seals at moderate to high titers in recovered animals for at least 6.5 months. After administration of an inactivated CDV vaccine to harbor seals, CDV antibody levels declined rapidly to very low levels within the same period, while titres in grey seals persisted at detectable levels for at least 12 months (Cornwell et al., 1992).

Namroodi et al. (2018) also recently reported CDV antibody prevalence from virus neutralization tests, with serum from 12/36 (33%) individuals, sampled 2015-2017 after release from accidental entanglement in fishing gear, testing positive, and concluded this provided evidence for persistence of CDV infections in the population. In principle, the current estimated population size for Caspian seals at approximately 168,000 individuals (Goodman and Dmitrieva, 2016), is above the threshold necessary for the virus to persist as endemic infection (Kuiken et al., 2006), but further systematic serological and genetic surveillance is necessary to fully understand CDV dynamics in the Caspian seal population. Comparisons of CDV antibody prevalence between our studies should be treated with caution due to differences in detection methods, sampling locations and periods, and sex-age class composition of the animals tested.

Avian influenza A viruses frequently jump from birds into various mammalian species, including pinnipeds (Repérant et al., 2009; Anthony et al., 2012; Bodewes et al., 2015a), with spill over events being associated with close contact with avian reservoirs. In the case of pinnipeds, this may occur through sharing of haul out sites with infectious birds shedding virus in aerosols and feces.

Four Caspian seals yielded suspected positives for influenza A virus on RT-PCR. Virus cultivation attempts from samples with Ct values <35 for further analysis failed. Similar results with non-cultivable influenza A viruses were described for qRT-PCR positive samples in live-captured North Atlantic gray seals (Puryear et al., 2016), broadly reflecting difficulties growing and sequencing influenza viruses from wild reservoirs.

Antibody screening using ELISA test kits from two independent manufacturers suggested regular exposure of Caspian seals to influenza A infection. Our results may be downwardly biased as there are no approved ELISA test-kits for monitoring of anti-influenza A antibody levels of marine mammals, although equivalent kits to those used here have previously been used successfully in other pinniped studies (Bodewes et al., 2015c; Puryear et al., 2016).

The results of the commercially available ID Screen^® Influenza A Antibody Competition Multi-species ELISA kit were verified by an IDEXX Influenza A Ab ELISA kit. Differences in antibody prevalence in test kit results may be partially accounted by deterioration of specific IgG after freeze-thawing cycles and the duration of storage. Our HI assay results revealed a narrow range of antibodies against influenza A virus hemagglutinin subtypes including H1, H4 and H10 in ELISA positive sera of Caspian seals. Elevated titers to influenza A (H10N7) were detected in samples collected in 2016. The absence of antibodies to this strain in serum samples from the preceding years may indicate recent exposure of the population, although no reports of elevated morbidity and mortality were recorded among Caspian seals at the time. Increased seropositivity in seal rehabilitation center patients was also reported in the Netherlands in 2015 (Bodewes et al., 2015c) after an influenza A (H10N7) virus outbreak among harbor seals in the North Sea in the spring/summer of 2014 (Bodewes et al., 2015a).

Three influenza A/NP-positive serum samples did not react with any hemagglutinin subtype antigens in HI assays, suggesting that these sera came from Caspian seals which encountered other antigenically different influenza A viruses. Previously a possible transmission of human influenza A and B viruses to Caspian seals was serologically characterized (Ohishi et al., 2002). Serological evidence for influenza viruses infection with avian-associated hemagglutinin subtypes H1, H2, H3, H4, H6, H7, H8, H9, H10, H11, H12, H13, and H16 has been obtained in marine mammals in a number of studies (De Boer et al., 1990; Steuen et al., 1994; Danner and McGregor, 1998; Nielsen et al., 2001; Puryear et al., 2016). The North Caspian region is covered by intersection of the Black Sea-Mediterranean and East African-West Asian migratory flyways, meaning Caspian seals could consistently interface with avian reservoirs of influenza viruses on their haul-out sites. Thus the detection of antibodies against avian-associated influenza A hemagglutinin subtypes, and potential for regular bird-seal interactions in their environment, strongly supports transmission of influenza viruses from avian reservoirs to Caspian seals rather than being derived from mammalian sources.

In contrast to influenza A viruses, which have been isolated from many different species, influenza B viruses are human pathogens which had no or unknown reservoirs in nature until 1999, when it was reported that harbor and grey seals can become infected with this virus (Osterhaus et al., 2000), with Influenza virus B isolated from a juvenile harbor seal with symptoms of respiratory disease. Since the acknowledgment of marine mammals as new hosts, there have been several reports on the detection of antibodies to the influenza B virus in phocid and otariid species (Ohishi et al., 2002; Blanc et al., 2009), with evidence for infection of harbor seals in the Netherlands on a regular basis (Bodewes et al., 2013a).

Our serology results provide evidence for influenza B virus infection in Caspian seals. The highest antibody titers were found against influenza B/Almaty/8/18 strain belonging to B/Victoria-linage viruses, whereas lower antibody titers were detected for B/Florida/4/2006 strain of B/Yamagata-lineage viruses. Antibodies at high titers were found in serum samples from juvenile seals suggesting recent influenza B infection in those animals. The proportion of influenza B positive seal sera ranged from 9.5% up to 100% (in 2010) per sampling time point, indicating regular exposure of Caspian seals. The negative results of influenza B virus PCR screening in Caspian seals probably relates to low viral load, dependence on sampling time, and if infection has a seasonal character. Transmission is most likely to occur when seals are hauled out in high densities. Haul out densities are highest during the spring molt, and to a lesser extent during the autumn. Winter transmission could occur during breeding, but normally there is some separation between females hauled out on the ice sheet during lactation (Wilson et al., 2017a). Resolving whether influenza B is endemic within Caspian seals, or if there is recurrent infection either from a human or other reservoir requires phylogenomic analysis of influenza B viruses from seals and additional serology analysis with a broad panel of epidemic strains.

Adenovirus infection is one of the most common diseases in carnivores. There are two types of this disease - infectious hepatitis and infectious laryngotracheitis, which are caused by adenovirus type I and II, respectively. Adenoviral infections with clinical signs of hepatitis and enteritis have been noted in sea lions and fur seals (Goldstein et al., 2011; Inoshima et al., 2013). Phocine adenoviruses PhAdV-1 and PhAdV-2 were isolated from ocular swabs of northern elephant seals and Pacific harbor seals with ophthalmic disorders (Wright et al., 2015).

Antibodies to canine adenovirus type I (CAV-1) were detected in 25% of Caspian seal sera predominantly in adult animals. The anti-adenovirus antibody detection frequency in Caspian seals is comparable with a 29% prevalence reported in Otariid serological surveys for CAV-1 (Burek et al., 2005). The antigenic cross-reactivity of CAV and CAV-related adenovirus of pinnipeds is discussed by Inoshima et al. (2013). Results of viral surveys in Northern elephant seals and Pacific harbor seals (Wright et al., 2015) demonstrate the existence species-specific adenoviruses, Our preliminary sequencing of PCR amplicons from the Caspian seal swab samples of 2016 suggests the existence of a novel adenovirus specific to Caspian seals, related to Lutrine adenovirus type 1, bovine adenovirus 10 and bat mastadenovirus, but requires further full-genome characterization to fully understand its relationship to other adenoviruses.

Coronaviruses are common pathogens of the respiratory system, gastrointestinal tract and cause systemic infections of domestic and wild animals (Buonavoglia et al., 2006). Harbor seal coronavirus (HSCoV) is manifested by hemorrhagic pneumonia and causes epizootics with high mortality (Nollens et al., 2010). Canine respiratory coronavirus is a new coronavirus of dogs, which is widespread in North America, Japan, and several European countries, and is genetically and antigenically distinct from enteric canine coronavirus (Erles and Brownlie, 2008). Modified pan-coronavirus PCR-screening of the Caspian seal swab and post-mortem samples were negative for the poly L gene of this virus. However, Canine Coronavirus (CCV)-specific IgG were found in one seal serum from 2007 and three samples collected in 2016. The CCV ELISA test kit is designed to detect antibodies against CCV proteins (mostly glycoproteins). CCV is a highly infectious but usually self-limiting disease characterized by diarrhea, intestinal disease, weakness and loss of appetite in dogs. Taking account of the cross-species spill over risks associated with coronaviruses, our findings demonstrate a potential exposure risk of Caspian seals to coronaviruses, which should be closely monitored in future surveillance activities.

Antibodies reacting to PhHV-1 were detected in Caspian seal sera from 68 of 72 samples (94.7%) tested, representing all age classes. This suggests persistent, stable circulation of PhHV-1 in the Caspian seal population, with potential recurrent exposure of individuals. Previous studies have demonstrated significant homology of alphaherpesviruses of terrestrial carnivores and phocid-herpesvirus at the antigenic and immunogenic levels (Osterhaus et al., 1985; Lebich et al., 1994; Martina et al., 2003), and found strong correlation between ELISA and serum neutralisation tests for PhHV-1 (Roth et al., 2013). Therefore, we suggest that the canine herpesvirus ELISA employed here can serve as a useful proxy for PhHV-1. Presence of PhHV-1 was confirmed by molecular tests. Analysis of a partial sequence of the Us2 gene suggests the Caspian seal strain is highly similar to Phocine alphaherpesvirus PhHV-1 isolate PB84 derived from a harbor seal, but sequencing of the full genome is necessary to fully characterise the virus and its relationship to other phocine herpes viruses.

Herpesvirus PhHV-1 was first isolated from newborn habor seals with clinical signs of acute pneumonia and hepatitis at a rehabilitation center in the Netherlands in 1985 (Osterhaus et al., 1985). In the Pacific harbor seal (Phoca vitulina richardsi), this virus is associated with focal caogulative necrosis of the adrenal cortex and liver. In phylogenetic studies of the DNA polymerase (gD) gene of PhHV-1, their relationship with herpes viruses of dogs and cats was established (Martina et al., 2003). Genital herpesvirus otarine herpesvirus 1 is strongly associated as trigger of cervical tumors in California sea lions (Zalophus californianus) (Deming et al., 2021).

Serological studies have shown widespread infection of PhHV-1 or other closely related α-herpes viruses in seal populations in the North Sea, as well as in the waters of the Antarctic and the North Pacific (Harder et al., 1991; Have et al., 1991; Zarnke et al., 1997). The serological prevalence of 94% in our study is similar to PhHV-1 antibody prevalence in harbor seals from Svalbard (Norway) where detection frequency ranged from 72 to 100% in each study year (Roth et al., 2013). The annual variation in seroprevalence in harbor seals was concluded as reflecting different frequencies of reactivation of the latent virus, influenced by factors impacting on stress and immune-competence of the animals (Roth et al., 2013). PhHV-1 infection in Caspian seals likely represents a consistent viral burden on health which may be exacerbated by triggering factors. The pathological implications, of PhHV-1 virus infection in Caspian seals are likely to be similar as for other phocid species, but further research is needed for these to be established.

Overall, the risks, profile and prevalence of pathogens in healthy, live, Caspian seals appears comparable to other wild phocid seal populations (Simeone et al., 2015). Where sequence data are available (e.g. for CDV, herpes, and adenoviruses) they suggest that most strains circulating are unique to Caspian seals. There are limited surveys of domestic and wild carnivore populations around the Caspian Sea, and so it is presently not possible to definitively say whether most pathogens in Caspian seals are endemic or spill over from terrestrial sources. However, the current Caspian seal population size is likely to be sufficient to sustain endemic infections for at least some viral pathogens (Kuiken et al., 2006). For influenza A, Caspian seals likely have frequent exposure to migratory birds at haul out sites, providing opportunities for spill over. This presents a future epizootic risk, should a strain highly pathogenic for seals arise.

Caspian seals came to global attention due to the mass mortalities from CDV 1997-2001 (Kuiken et al., 2006), but there have been no confirmed further CDV outbreaks since then despite evidence for its continuing circulation from this and other recent studies (Namroodi et al., 2018). Our observation of apparently asymptomatic infection in juveniles from 2008 suggest potential development of reduced virulence, which could partially account for the absence of recurrent large outbreaks. However, more research is needed to further characterise currently circulating CDV strains and understand future risks.

Mass strandings of Caspian seals, involving hundreds or thousands of individuals are a frequent occurrence (Wilson et al., 2014). For example, in December 2020 an estimated 300 to 2000 carcasses stranded along the coast of Dagestan, Russia News18.com, 2020. Such strandings are rarely investigated in depth due to difficulties in accessing remote locations, carcasses being in advanced states of decomposition, lack of resources, and there being few researchers with expertise in marine mammal pathology in the region (Authors’ observations). Where investigation of carcasses has taken place, the role of disease has often proven inconclusive, but evidence of fisheries related mortality is sometimes recorded (Dmitrieva et al., 2013). Stranding rates increase in the spring and autumn, which could be accounted for by natural mortality due to post-breeding/molting stress, increased disease transmission on haul out sites, and are also coincident peaks in fishing activity (Wilson et al., 2014).

Future research priorities should focus on determining the potential for the pathogens identified here to cause mortality and morbidity in the Caspian seal population, and how this might influence resilience to other conservation threats. The sources of future epizootic risks should be better quantified, and how these relate to anthropogenic environmental pressures such as human population density, pollution (sewage, agricultural and industrial effluent, heavy metals, hydrocarbons and persistent organic pollutants), habitat loss and degradation, and other potential sources of stress arising from human activities. Climate change and decline in the Caspian Sea sea level have potential to increase disease exposure or increase stress and reduce resilience to disease.

Building on this work, it is important to expand routine disease surveillance for Caspian seals, to facilitate both pre-emptive and reactive management of disease risks. At present there are no well-funded, routine, coordinated stranding investigation programs in any Caspian countries. It will be important to provide resources to support stranding investigations, and build capacity for veterinary pathology. This has potential to strengthen overall monitoring and understanding of ecosystem health for the Caspian Sea, and identify and respond to emerging conservation issues.