Ethics statement

Our research protocol was approved by the research and ethics committee of the Deputy of Natural Environment of Golestan Province, Department of Environment (S1 Certificate, permit number: 125/7894).

Sample collection and clinical examination

Caspian seals accidentally caught in fishing nets by fishermen near the coast of Northern Iran were physically restrained and transported to Seal Treatment and Research Center (Ashooradeh Island, Golestan Province) by suitable vehicle during 2015–2017 [19]. Animals were handled in strict compliance with the European Union legislation on the protection of animals used for scientific purposes (EU Directive 2010/63/EU) [20].

In the rehabilitation center, seals were housed individually in a pool of fresh saltwater from the Caspian Sea and provided with food, water, and any necessary medicines. Animal welfare was monitored daily. After seals were treated and allowed to recover from any injuries and from the stress of capture, blood collection and flipper tagging were performed under physical restraint as described by Lynch and Bodley [19].

Sex, site of sampling, and clinical signs were recorded. Seals were assigned to one of two age classes. The youngest seals captured were less than one year old, but older than 3–6 months (because the first 3–6 months of life are spent in the Northern Caspian Sea) [1]. We refer to this age class as “yearlings.” The “older” age class consisted of those seals greater than 1 year old. The distinction between age classes was made on the basis of size, pelage, and appearance of teeth. Blood samples were collected from the caudal gluteal vein using Venoject samplers (Novingostar Co., Tehran, Iran). Blood samples were allowed to clot and centrifuged for 10 min at 3000 rpm. Sera were removed and stored at -20°C. The period of rehabilitation varied from 1 to 4 weeks, but averaged about 3 weeks. Once rehabilitation was complete, flipper tagging was performed using sheep ear tags (Part Imensill Co., Tehran, Iran) and animals were released on the shore near the location they were captured.

CDV antibody detection

We detected CDV antibody using the microplate virus neutralization test (VNT) as described by Saliki and Lehenbauer [21]. The Onderstepoort vaccine strain of CDV was used as antigen. After dilution of sera starting from 1:2, plates were incubated at 37°C in 5% C0₂ for 3–5 days. A dilution of 1:8 was considered the VNT-positive cut-off point.

L. interrogans antibody detection

Sera were examined for six L. interrogans serotypes: Pomona, Grippotyphosa, Icterohaemorrhagiae, Canicola, Hardjo, and Australis, by Leptospira microscopic agglutination test (LMAT), in the Leptospira Research Laboratory, Faculty of Veterinary Medicine, University of Tehran. The LMAT was run on sera in a dilution series from 1:100 to 1:2800. Positive and negative standard samples were added to two wells of each micro-titration plate and the plates were incubated at 30°C for 2 hr. The serum sample was considered positive for antibody to a L. interrogans serovar if there was ≥ 50% Leptospira agglutination at a serum dilution of ≥ 1:100 when observed by dark-field microscopy [12].

T. gondii antibody detection

To detect T. gondii antibody, we used the Toxoplasma modified agglutination test (TMAT) as validated by Dubey (1997). T. gondii tachyzoites were suspended in mercaptoethanol, and then this suspension was used to dilute serum samples in a 2-fold series from 1:25 to 1:500 [22]. There are no data on the specificity and sensitivity of this test, nor is there an accurate cut-off titer for Caspian seals. So, according to previous surveys by Aguirre et. al., ≥ 1:25 dilution (Titer ≥ 25) was considered T. gondii antibody positive [23].

Statistical analysis

Results were entered into SPSS (v20) software and Chi square tests were used to determine the significance of differences in antibody frequencies between age classes.

Exposure to CDV

Since the late 1980s there has been much evidence of mass mortality in marine mammals due to morbillivirus infection [3, 24]. These viruses have been detected in cetaceans from the North Pacific and many species of marine mammals from the Atlantic Ocean, Mediterranean Sea, and Arctic waters, leading to at least eight epizootics [25–27]. The literature suggests that morbillivirus infection has played a much deadlier role in marine mammal populations than toxoplasmosis or leptospirosis.

CDV has been diagnosed in mass mortalities of Baikal seals and Caspian seals, and there are a few reports of CDV antibody in other marine mammal species throughout the world [3, 28]. Antibody to CDV was detected in 32 of 96 sampled crabeater seals (Lobodon carcinophagus) and two of three leopard seals (Hydrurga leptonyx) from Antarctica [29], and in 12 of 40 harp seals (Pagophilus groenlandicus) from Greenland [29].

Our data indicate that CDV is still present in the Caspian seal population and that it is not 100% lethal. The moderate frequency of CDV antibody in Caspian seals and the currently low population size argue against sustained, long-term, endemic infection in the population. Although Kuiken et al. (2006) suggested that the Caspian seal population itself could be acting as a reservoir for CDV, they based their calculations on a seal population size of 360,000–400,000, and an estimate that a population size of 120,000 was sufficient to maintain infection [30]. The currently estimated population size of 104,000 to 168,000 makes endemic maintenance less likely [5], but the population may still be above the estimated 120,000 necessary to maintain endemic infection. An alternative source of CDV for Caspian seals may be persistent infection in terrestrial carnivores along the Caspian Sea coast [3]. In the end, it is still impossible, at this time, to definitively implicate either endemic infection or spillover.

Exposure to L. interrogans

The 17% antibody frequency in our sample of apparently healthy Caspian seals suggests that seals are commonly exposed to L. interrogans and that the infection is (at least) not 100% fatal. Antibody frequency may actually be higher than 17%, as the LMAT has not been certified for antibody detection in Caspian seals and sensitivity may be less than 100%. It is also possible that antibody may become undetectable over time. Finally, LMAT may not detect some positive animals due to transient nature of bacteria and bacteria may be hiding in the kidney and genital tracts of infected animals [15].

L. interrogans serovars Pomona and Canicola have been reported to infect other species of marine mammals. Antibody to serovars Canicola and Pomona were detected in New Zealand fur seals (Arctocephalus forsteri) [7] and antibody to serovars Hardjo, Bratislava, Grippotyphosa, Cynopteri, Bataviae, Ichterhaemorrhagiae, and Pyrogenes have been described in other marine mammals [7, 31].

Some serologic studies on marine mammals using LMAT showed reaction to more than one serovar. These results could have resulted from mixed infections or from cross reaction among serovars on the LMAT [31]. However, in our study, all positive sera were reactive with only one serovar.

Leptospira serovar dominance and antibody frequency in susceptible animals can vary based on geographic area, susceptible species of terrestrial animals in each region, number of animals tested, and the test used [15]. Because the common serovars in a given region may vary over time, frequent screening and identification of serovars in each region has been suggested to maintain effective control programs [32].

Because leptospires cannot survive in sea water, detection of Leptospira antibody in Caspian seals suggests that the seals might have been in close contact with other infected animal species on rookeries. Transfer of maternal antibody to pups may occur, however [15], and may or may not be protective against infection. Finally, although leptospires do not survive in “sea water”, the salinity of the Caspian Sea water is far lower (about one third) than the earth’s oceans, varying between 1 part per thousand and about 13 parts per thousand. The lowest values are in the northern basin where the Volga and Ural rivers discharge [33]. To our knowledge, it is not known how this low salinity might alter the survival of leptospires, if at all.

Dogs are a common maintenance host of L. interrogans serovar Canicola, and cattle, sheep, and pigs are common maintenance hosts of serovar Pomona [34]. So, these domestic animal species could be sources of exposure to Caspian seals, especially in Northern Iran, where seals are seen in the near-shore waters and even hauled-out near small villages where inhabitants keep ruminants. These shoreline areas are heavily polluted and contaminated by human refuse, roaming dogs and cats, and occasionally free livestock. Currently, these reports are only anecdotal and specific studies are needed. In more remote areas, rodents or wild canids such as wolves or jackals are another possible source. Serovars Canicola and Pomona are now the most common L. interrogans serovars in some parts of Northern Iran [32]. It is also possible that L. interrogans serovars could be maintained within the seal population and exposure to contaminated seal urine/feces on haul-out sites would be a plausible route of intraspecific transmission.

The migratory lifestyle of Caspian seals makes the Leptospira sources and the locations of transmission sites difficult to determine [18]. However, vaccination of susceptible domestic animals, when feasible, and restriction of contact between Caspian seals and other animals at haul-out sites would likely prevent transmission of Leptospira to seals. Since, Leptospira shedding in urine samples has not been investigated in Caspian seals, the possible role of seemingly healthy seals in bacterial shedding remains unclear. Additional studies are needed to determine if Caspian seals can be maintenance hosts for L. interrogans serovars Canicola and Pomona. Nevertheless, humans in contact with Caspian seals should be aware of the potential risk of Leptospira shedding.

Our results provide basic data about the epidemiology of L. interrogans infection in Caspian seals. Because other species of marine mammals have suffered mass die-offs due to Leptospira infection, understanding the consequences of Leptospira infection in Caspian seal individuals is critical to understand the possible impact of infection on their populations [31].

Exposure to T. gondii

The high frequency of T. gondii antibody in sampled Caspian seals can be explained by contamination of the marine environment by freshwater runoff containing oocysts from felids and the dissemination of these oocysts by currents. These oocysts can then become sporulated and may remain infectious in seawater for as much as two years [35]. Similarly, Alvarado-Esquivel et al. [36] reported T. gondii infection in 87% of captive Atlantic bottlenose dolphins (Tursiops truncates gillii) in Mexico. Conversely, Aguirre [23] reported a low prevalence of T. gondii antibody in Hawaiian monk seals (Monachus schauinslandi) in the northwestern Hawaiian Islands. Bearded seals (Erignathus barbatus), walruses (Odobenus rosmarus), California sea lions (Zalophus californianus), ringed seals (Phoca hispida), and spotted seals (Phoca largha) have also shown evidence of T. gondii infection [37].

Despite the absence of Caspian seals with clinical signs of toxoplasmosis in this study, the high frequency of T. gondii antibody raises suspicion that T. gondii infection may be, or may become, a health problem for this species, especially in immunosuppressed individuals [10]. After all, seriously diseased animals might quickly be removed from the sampling pool leading to findings of lower T. gondii antibody frequencies in sampled seals. Management directed at minimizing the exposure of Caspian seals to T. gondii oocysts, including restriction of seal contact with felids would help prevent infection.

Frequencies of exposure to all three pathogens were lower in yearlings than older Caspian seals. This is expected and is reflective of the very limited opportunities for exposure for seals during the first year of life, half of which is spent in the northern Caspian Sea. It may be possible however that some infected yearlings may have died and thus were eliminated from the sampling population. Results of some studies have differed. For example, Alvarado-Esquivel [36] reported no significant variation in T. gondii antibody prevalence among age classes of marine mammals. Nielsen et al. [38] found that PDV antibody prevalences in Atlantic walrus (Odobenus rosmarus rosmarus) were not associated with sex or age and that CDV antibody prevalences increased with age in crabeater and leopard seals. Colagross-Schouten et al. [12] reported greater L. interrogans serovar Pomona exposure in adult and subadult sea lions than in juveniles.

Without much more extensive sampling, and an understanding of any pathogenic effects of these pathogens on Caspian seals, it is impossible to predict the epizootic or endemic infection of Caspian seals by these three pathogens, or the effects of such infection on the population. The circulation of enzootic, host-adapted Leptospira serovars in California sea lions, for example, has been cited to explain the high serum antibody prevalence, low antibody titers, and absence of clinical signs of leptospirosis in pups [11]. In our case, the low titers for CDV (8–16), T. gondii (25–50) and L. interrogans (100–400) in apparently healthy Caspian seals could possibly indicate chronic exposure to these agents or simply decreasing titers over time following previous exposure. Sample handling and an imperfect cold chain could also have contributed to the low titers observed.

Climatic conditions along the coasts of countries bordering the Caspian Sea are mostly appropriate for the survival of the three pathogens we studied. For example, 35% of rice farms and water canals, and 30% of rivers in Gilan Province, Iran, which borders the Caspian Sea, are contaminated with Leptospira spp. [39]. However, seasonal movement of Caspian seals shows the need for molecular and phylogenetic analyses on isolated pathogens in all countries bordering the Caspian Sea to detect potential sources and regions of CDV, T. gondii, and L. interrogans exposure to Caspian seals.

Our results provide basic data about the natural threats that may impact the Caspian seal population. However, understanding the importance of antibody-positive animals without clinical signs or isolation of the pathogens begs the important question of whether T. gondii and Leptospira can lead to deadly disease. Because Caspian seals are highly susceptible to CDV infection, emphasis should be placed on minimizing the risk of a new epidemic. On the basis of our data we cannot predict if CDV will cause future outbreaks in Caspian seals. Systematic screening for CDV (especially), L. interrogans, and T. gondii in surviving Caspian seals using molecular techniques is also needed.

Although a Caspian Seal Conservation Action Plan developed by the United Nations Development Program, Caspian Environment Program (http://projects.inweh.unu.edu/inweh/display.php?ID=1057)) recommends “periodic health surveys” for Caspian seals, the plan lacks specifics. It does not identify potentially important diseases, and has not been fully implemented. We strongly encourage the careful planning and vigorous implementation of a plan that would establish long-term monitoring of the general health of Caspian seals in the remaining population, identify the sources of CDV, L. interrogans, T. gondii, and other pathogens affecting Caspian seals, and take steps to eliminate or reduce exposure of seals to these pathogens.