The contamination of the marine environment with atypical microorganisms as a result of anthropogenic pollution poses unknown risks to both marine wildlife and the wider ecosystem (Oates et al., 2012; Baily et al., 2015). Fecal coliforms such as Escherichia coli are used globally as indicators of anthropogenic fecal contamination and as a measure of pollution of waterways and coastal ecosystems (Beversdorf et al., 2007; Schaefer et al., 2011; Ahmed et al., 2016). E. coli is a commensal species colonizing the intestinal tract of birds and mammals (Gordon and Cowling, 2003; Guenther et al., 2011). However, some strains of E. coli have virulence traits that render them opportunistic pathogens that can cause a range of diseases in multiple host species, including urinary tract infections, neonatal meningitis and septicemia (Russo and Johnson, 2000; Mora et al., 2009; Clermont et al., 2011). There are eight E. coli lineages (A, B1, B2, C, D, E, F, clade I), termed phylotypes, that differ in ecological niches, life histories, host affinities and pathogenic abilities; phylotypes B2 and D are commonly isolated from and associated with humans (Gordon and Cowling, 2003; Gordon et al., 2008). Many clinically relevant strains of E. coli that have the potential to cause extraintestinal disease in humans, domestic animals and birds belong to phylotype B2 (Johnson and Russo, 2002; Clermont et al., 2014).

Marine mammals are considered sentinels of ocean health with marine mammal health status providing insights into the health of the wider marine ecosystem (Bossart, 2010). Escherichia coli has been isolated from a number of aquatic mammal species including pinnipeds (Stoddard et al., 2008, 2009; Wallace et al., 2013; Delport et al., 2015; Power et al., 2016; Fulham et al., 2018; Mora et al., 2018), cetaceans (Schaefer et al., 2011; Melendez et al., 2019) and river otters (Oliveira et al., 2018). Despite this, little is known about the host and environmental factors that influence the diversity and prevalence of E. coli in marine mammals. Furthermore, the intestinal population of E. coli in mammals differs between individuals of the same species (Guenther et al., 2011). As a result, studies that focus on one species at a single time point are unlikely to accurately represent the diversity of E. coli within a species (Schierack et al., 2008, 2009; Leser and Mølbak, 2009).

In the Southern hemisphere, E. coli has been isolated from multiple pinniped species including southern elephant seals (Mirounga leonina), Weddell seals (Leptonychotes weddellii), Antarctic fur seals (Arctocephalus gazella) and Australian sea lions (Neophoca cinerea) (Hernandez et al., 2007; Delport et al., 2015; Power et al., 2016; Fulham et al., 2018; Mora et al., 2018). These studies have characterized E. coli isolated from free-ranging individuals with a majority of samples collected from adults and a predominance of phylotype B2 identified; the prevalence of the B2 phylotype ranging from 49% (Fulham et al., 2018) to 65% (Power et al., 2016). Comparisons of E. coli prevalence and phylotype distribution in captive and free-ranging adult N. cinerea (Delport et al., 2015) determined a higher prevalence in captive individuals; however, in both captive and free-ranging animals, the human-associated phylotype B2 was the most frequently identified (Delport et al., 2015). A higher E. coli prevalence and differing trends in phylotype distribution was described in free-ranging N. cinerea pups from two colonies in South Australia (Fulham et al., 2018). Of the two colonies, phylotype B2 was more prevalent in samples from the colony closer to anthropogenic sources of pollution than the colony located further from human habitation (Fulham et al., 2018). It is unknown whether host or environmental factors could be contributing to the differences in phylotype distribution observed in E. coli isolated from free-ranging N. cinerea pups. For this reason, further investigations are required to understand the distribution and diversity of E. coli in marine mammal species in Australian waters, including other pinniped species such as the Australian fur seal (Arctocephalus pusillus doriferus) and long-nosed fur seal (Arctocephalus forsteri) where the prevalence and diversity of E. coli has not been reported previously.

The three pinniped species, N. cinerea, A. p. doriferus, and A. forsteri, inhabit numerous offshore colonies along the Australian coast from Western Australia to Tasmania (Kirkwood and Goldsworthy, 2013). All three species were subjected to commercial sealing practices during the nineteenth and twentieth centuries which decimated their numbers (Ling, 1999). Population recovery of each species from historical harvesting has differed greatly, with numbers of A. p. doriferus and A. forsteri steadily increasing since the 1960s and 1970s (Shaughnessy et al., 2015; McIntosh et al., 2018). In contrast, N. cinerea have undergone continual decline; the species is listed as endangered on the IUCN Red list (Goldsworthy et al., 2015). The reasons underlying these differing rates of population recovery are likely complex and multi-factorial and include differing life histories, breeding strategies, geographical distribution, extent of fisheries interactions and disease (Goldsworthy et al., 2009; Shaughnessy et al., 2011; Marcus et al., 2015). The geographical range of these species frequently overlap, with some species breeding at the same sites. This geographical proximity provides an unparalleled opportunity for comparative investigations of E. coli to better understand the role of host and environmental factors on E. coli prevalence and diversity.

The objective of this study was to characterize the diversity of E. coli in three species of free-ranging pinniped pups in Southern Australia, establishing base line data for future comparative studies and to assess potential effects of anthropogenic impacts.

This study reports the prevalence and diversity of E. coli isolated from free-ranging N. cinerea, A. p. doriferus, and A. forsteri pups at eight breeding colonies in Australia, finding no significant difference in E. coli prevalence or the distribution of phylotypes across species, colonies or breeding seasons.

This is the first investigation of the comparative presence and distribution of E. coli in free-ranging A. p. doriferus and A. forsteri pups. The prevalence of E. coli was similar across all three species and supports previous reports in Antarctic pinnipeds including Antarctic fur seals, Southern elephant seals and Weddell seals (Power et al., 2016; Mora et al., 2018), and captive adult N. cinerea (Delport et al., 2015). Previous studies have suggested that E. coli is uncommon in marine mammals (Johnson et al., 1998; Hernandez et al., 2007) with a higher prevalence associated with proximity to humans (Stoddard et al., 2008; Delport et al., 2015). In this study the differences in proximity to humans at each breeding colony and the high prevalence of E. coli observed in N. cinerea, A. p. doriferus, and A. forsteri pups could potentially indicate that proximity to humans is not the only factor contributing to E. coli prevalence in marine mammals.

The B2 phylotype was most frequently identified from pups of all three species, with 73.7% of all E. coli isolates assigned to this phylotype. This is the highest frequency of the B2 phylotype identified in pinnipeds in Australia; in free-ranging N. cinerea, B2 made up 67% of all isolates in adults (Delport et al., 2015). It has been suggested that strains belonging to the B2 phylotype are well adapted to the intestinal environment of mammals (Gordon and Cowling, 2003; Nowrouzian et al., 2006), which could be a contributing factor to the high frequency identified in this study. The B2 phylotype was the most frequently isolated phylotype at seven of the eight breeding colonies during every breeding season sampled. It was previously hypothesized that the difference in phylotype distribution observed in free-ranging N. cinerea pups was due to colony location (Fulham et al., 2018), however, the results from this study indicate that geographical location and proximity to humans, is not a contributing factor to E. coli diversity in N. cinerea, A. p. doriferus, or A. forsteri pups.

The prevalence of E. coli in free-ranging pinniped pups is likely diet related; E. coli is a lactose fermenter and as the sampled pups were feeding solely on milk, high E. coli prevalence is not unexpected (Fulham et al., 2018). In adult N. cinerea, E. coli prevalence from free-ranging individuals was significantly lower (7.7%) compared to captive individuals (84%) suggesting that E. coli does not occur naturally in free-ranging adult N. cinerea (Delport et al., 2015). Marine wildlife species that forage in coastal or nearshore waters are at greater risk of exposure to pathogens in fecal bacteria as a result of anthropogenic pollution compared to species further removed from anthropogenic influence (Oates et al., 2012). In Antarctic pinnipeds, it was hypothesized that the presence of human-associated E. coli could be due to the geographic mobility of the animals sampled (Mora et al., 2018). Free-ranging N. cinerea, A. p. doriferus, and A. forsteri adults occupy large geographical ranges (Shaughnessy et al., 2011, 2015; McIntosh et al., 2018), encountering a variety of environments and environmental conditions that could influence their exposure to anthropogenic pollution. In other pinniped species, there is evidence of maternal transmission of gut microbes (Nelson et al., 2013b), however, it is unknown whether the acquisition of E. coli in pinniped pups in this study is environmental or through maternal transfer. Understanding the prevalence of E. coli and phylotype diversity in free-ranging adults could provide valuable insights into the factors that contribute to the acquisition of E. coli and the trends in E. coli prevalence and phylotype distribution seen in pups.

Studies investigating the transfer of E. coli between humans and wildlife species are limited to terrestrial environments. Generally, E. coli isolated from wildlife species that share habitats and have higher levels of interactions with humans and livestock were genetically similar to E. coli isolated from humans (Goldberg et al., 2007; Rwego et al., 2008a,b). This genetic similarity suggests E. coli transmission between these species in shared habitats is likely a result of indirect contact through contaminated environmental sources rather than direct contact (Goldberg et al., 2007). For example, E. coli isolated from the feces of banded mongoose (Mungos mungo) was genetically similar to E. coli isolated from human fecal waste in their environment, highlighting the importance of indirect routes of transmission (Pesapane et al., 2013). These results suggest that the acquisition of human-associated bacteria by wildlife species is linked to fecal contamination of the environment.

Coastal environments can be contaminated by fecal pollution through a number of different sources including sewage and storm water runoff from agricultural, urban and commercial land (Crain et al., 2009; Pandey et al., 2014). The population size and density of people in coastal towns will influence the amount of bacterial contamination of the environment from runoff, with higher density resulting in higher levels of fecal bacteria (Pandey et al., 2014). The population size of towns closest to the pinniped breeding colonies sampled in this study varied, however, the prevalence of E. coli did not differ across colonies or seasons, suggesting that the contribution of human population density to trends in E. coli prevalence observed requires further investigation. Previous studies have determined that E. coli isolates from wildlife species occupying habitats in close association with humans or that are exposed to fecal pollution are more likely to belong to phylotype B2 than those isolated from wildlife living in isolation from humans (Gordon and Cowling, 2003). E. coli isolates belonging to B2 and D phylotypes have been found in treated sewage, suggesting a greater ability to survive treatment processes (Anastasi et al., 2010). The survival of these phylotypes in wastewater coupled with increasing pollution of coastal ecosystems presents a higher risk of transfer of human-associated bacteria into the marine environment. The predominance of the B2 phylotype across the three pinniped species studied could therefore suggest that all species are exposed to similar levels of anthropogenic pollution. This could be important for pup health, given that strains belonging to this phylotype are associated with extraintestinal disease in humans (Dale and Woodford, 2015).

Ocean currents and tides are also potential factors contributing to the high prevalence of E. coli in pinnipeds. The main source of ocean pollution is of terrestrial origin (Robinson et al., 2017) and as a consequence, coastal areas in close proximity to populated areas are exposed to higher levels of anthropogenic pollution (Partelow et al., 2015). The marine environment, being connected over longer timescales (Jönsson and Watson, 2016) has greater connectivity compared to terrestrial environments, with fewer physical barriers between areas. Consequently, the pathways of water that flow along coastlines can influence the dispersal of anthropogenic pollutants. Wastewater effluent is a known source of human-associated E. coli phylotypes (de Stoppe et al., 2017) and E. coli that originate from wastewater have an enhanced capacity to survive in the marine environment, able to survive for several days outside of a host (Rozen and Belkin, 2001). In addition, E. coli can attach to particles in the water column, facilitating movement in the marine environment (Mallin et al., 2000). This enhanced survival and attachment to particles in the water column could result in prolonged environmental persistence of E. coli, facilitating acquisition by marine wildlife. The similarity in E. coli distribution across species and breeding colonies in this study could be indicative of the high degree of connectivity of marine ecosystems (Crain et al., 2009; Robinson et al., 2017) or similar anthropogenic pollution across this large geographical range.

The identification of B2 sub-types I, II and IX are of particular interest given that clonal complexes that belong to these sub-types are commonly associated with disease in avian species and humans (Clermont et al., 2014; Riley, 2014). ST131 and ST95 strains belonging to sub-types I and IX, respectively, are associated with disease caused by extraintestinal pathogenic E. coli (ExPEC) (Dale and Woodford, 2015). These sub-types or strains have previously been identified in Antarctic pinnipeds (Power et al., 2016; Mora et al., 2018). Assignment of E. coli isolates to strains was not undertaken in this study, however, given the presence of B2 sub-types associated with pathogenic strains, further analysis of B2 isolates will assist our understanding of the potential risks posed to pup health by their presence.

It is also important to consider that E. coli is a very small component of the intestinal microbiota in pinnipeds (Nelson et al., 2013a,b; Delport et al., 2016). However, the ease with which E. coli can be cultured and characterized has resulted in E. coli being commonly used for monitoring fecal contamination of marine environments (Beversdorf et al., 2007). For this reason, it can be a useful indicator of the diversity of E. coli phylotypes present in marine mammals, and a relatively high prevalence may suggest greater exposure to anthropogenic pollution. Sampling of substrate and water surrounding breeding colonies and comparing those E. coli phylotypes with diversity seen in pinniped pups could provide useful insights into potential sources of environmental contamination. Investigating the presence of specific markers through microbial source tracking could also be utilized to determine the origin of E. coli found in both wildlife species and contaminated environments.

The presence of bacteria and protozoa that are associated with humans should be explored to further understand potential pathogen transmission from anthropogenic sources into the marine environment. Toxoplasma gondii, Campylobacter spp. and Salmonella spp. have been isolated from marine mammals inhabiting coastal environments, including gray seals (Halichoerus grypus), Californian sea otters (Enhydra lutris nereis) and northern elephant seals (Stoddard et al., 2005; Oates et al., 2012; Shapiro et al., 2012; Baily et al., 2015). Campylobacter spp. isolated from H. grypus were genetically similar to isolates commonly found in agricultural and human sources, demonstrating the dissemination of a human pathogen into the marine environment (Baily et al., 2015). Similar investigations in Australian pinniped species could enable the identification of the source and dissemination of anthropogenic microbial pollution into the Australian marine environment.

In this study, the predominance of the human associated B2 phylotype and similarity of E. coli prevalence and phylotype diversity seen across species, colonies and breeding seasons could indicate that all colonies are exposed to similar levels of anthropogenic pollution. This widespread occurrence of human associated phylotypes highlights the need for ongoing monitoring and surveillance of microbes in both the marine environment and sentinel species, particularly those with potential pathogenicity for marine mammals. Future investigations should focus on whether E. coli is an atypical bacterium in these pinniped species and determine its reliability as an indicator of marine pollution.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The animal study was reviewed and approved by the Animal Ethics Committee at the University of Sydney and Phillip Island Nature Parks.

MF, RG, and MP contributed to conceptualization and design of the study. MF and RG collected samples from pinniped pups. MF completed laboratory analysis of samples and data analysis. MF drafted the manuscript. RG and MP participated in revising the manuscript. All authors contributed to the article and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.