Piscine orthoreovirus‐3 is prevalent in wild seatrout (Salmo trutta L.) in Norway

Abstract In 2017, a PCR‐based survey for Piscine orthoreovirus‐3 (PRV‐3) was conducted in wild anadromous and non‐anadromous salmonids in Norway. In seatrout (anadromous Salmo trutta L.), the virus was present in 16.6% of the fish and in 15 of 21 investigated rivers. Four of 221 (1.8%) Atlantic salmon (Salmo salar L.) from three of 15 rivers were also PCR‐positive, with Ct‐values indicating low amounts of viral RNA. All anadromous Arctic char (Salvelinus alpinus L.) were PCR‐negative. Neither non‐anadromous trout (brown trout) nor landlocked salmon were PRV‐3 positive. Altogether, these findings suggest that in Norway PRV‐3 is more prevalent in the marine environment. In contrast, PRV‐3 is present in areas with intensive inland farming in continental Europe. PRV‐3 genome sequences from Norwegian seatrout grouped together with sequences from rainbow trout (Oncorhynchus mykiss Walbaum) in Norway and Coho salmon (Oncorhynchus kisutch Walbaum) in Chile. At present, the origin of the virus remains unknown. Nevertheless, the study highlights the value of safeguarding native fish by upholding natural and artificial barriers that hinder introduction and spread, on a local or national scale, of alien fish species and their pathogens. Accordingly, further investigations of freshwater reservoirs and interactions with farmed salmonids are warranted.

In 2013, rainbow trout fingerlings from three different hatcheries in Norway showed disease signs and necropsy findings resembling HSMI in Atlantic salmon. Gross necropsy findings were signs of circulatory disturbances including haemorrhages, ascites and anaemia, and histopathological findings comprised inflammation of the heart and red skeletal muscle and necrosis in the liver. Two of the hatcheries reported low to moderate mortality. The third hatchery reported 0.3% mortality in tanks without clinical signs, and up to 21% in tanks with diseased fish.
Disease and related mortality was reported up to four months after sea transfer. Extended investigations, initiated due to the resemblance with HSMI in Atlantic salmon, led to the identification of a virus with gene sequences sharing approximately 85% identity with PRV in Atlantic salmon and closely related a PRV-like virus in Coho salmon. Hence, both viruses probably belong to the same group with the suggested designation PRV-3 (Dhamotharan et al., 2018).
Experimental infection studies have strengthened the association between PRV-3 and the disease signs seen in rainbow trout (Hauge et al., 2017). The infection transmitted to all of rainbow trout cohabitants that also developed disease symptoms. The virus seemed less adapted to Atlantic salmon, as less than 50% of the cohabitants were infected 8-16 weeks post-challenge and experienced only minor lesions.
Epidemiological investigations and PCR-based surveillance programmes show that PRV-3 is present at all levels of the marine rainbow trout production cycle, including broodfish, hatcheries and after sea transfer (Olsen et al., 2015). PRV-3 has also been detected in farmed rainbow trout in Germany (Adamek et al., 2018) and associated with disease in this species in Denmark and Scotland (Dhamotharan et al., 2018). In brown trout, the virus has been detected in Italy (Dhamotharan et al., 2018) and France (Bigarre, 2016).
In the latter case, mortalities due to a PRV-3 and infectious pancreatic necrosis virus (IPNV) co-infection were recorded in juvenile farmed brown trout (Salmo trutta L.). Recently, proliferative darkening syndrome (PDS), a severe disease causing die-off of wild brown trout (Salmo trutta fario) in pre-alpine river systems in Southern Germany, Austria and Switzerland has been associated with PRV-3.
The Norwegian Food Safety Authority annually conducts health monitoring in wild salmonids. In 2017, the objective of the programme was to investigate the occurrence of PRV-3 in wild salmonids. Here we report the first detection of PRV-3 in wild seatrout (Salmo trutta L).

| Study sample
A cross-sectional study designed to investigate the occurrence of PRV-3 in wild adult anadromous and non-anadromous salmonids in Norwegian watercourses was conducted (Table 1 and Figure 1).
The source material comprised wild-caught broodfish from stock enhancement hatcheries and the Genebank for wild Atlantic salmon (http://www.miljodirektoratet.no/old/dirnat/rapporter/673/rapport.pdf). Broodfish from the same river were kept together in freshwater tanks from a few days up to 6-7 weeks before stripping.
Additional samples from seatrout were obtained during rotenone treatment, a measure issued by the Norwegian Environment Agency to control the parasite Gyrodactylus salaris in the rivers Hundøla, Vefsna, Fusta andDrevja in 2011 (Anonymous, 2014). Brown trout were sampled from one lake in the County of Trøndelag (2010) and two lakes in the County of Sogn & Fjordane (2016). All Atlantic salmon were classified as wild based on scale reading (Antere & Ikonen, 1983;Fiske, Lund, Lund, & Hansen, 2004;Lund & Hansen, 1991). Figure 1 displays the different regions of Norway represented in the study.
Samples from 2010 to 2016 consisted of heart and kidney tissue fixed in RNAlater TM and were sent frozen to the Norwegian Veterinary Institute (NVI). Samples collected in 2017 consisted of kidney tissue fixed in RNAlater ™ sent to PatoGen AS for PCR-analyses.

| Real-time RT-PCR and sequencing
At NVI, real-time RT-PCR for detection of a 121 bp fragment of the sigma 3 protein gene of PRV-3 was performed using forward primer

| Real-time RT-PCR and sequencing
The PCR-based screening showed that PRV-3 was present in wild seatrout ( Table 2)

| PRV-3 in wild seatrout
In total, PRV-3 was present in seatrout from 15 of 21 investigated rivers and in both the northern and southern region (Figure 1). This result is independent of the two cut-offs (<40 or <37), since all the rivers had results below Ct 37. The possible influence of the difference in sampling tissues (heart vs. kidney) on the prevalence of PRV-3 is thus far unknown. Overall, the per-river sample sizes in this study were low (mean 12.6, range 1-35). The sample size range of the six virus-negative rivers was 1-14 (mean 5.7). Consequently, the minimum detectable prevalence in each of these rivers is quite high, indicating that the study underestimates the actual occurrence of PRV-3 in seatrout (Cameron & Baldock, 1998). Accordingly, this study strongly suggests that PRV-3 is a common virus in seatrout in both regions.
The prevalence of PRV-3 is different between the two regions, analyses. In that case, the sequence grouped together with PRV-1 from wild and farmed salmon (Garseth et al., 2013). For a specific pathogen and a specific host-population, the prevalence will reflect the efficacy of pathogen transmission, susceptibility of the host and duration of the infection, which in turn depends on the virulence of the pathogen and the immunization rate of the host (Begon, Harper, & Townsend, 1990). Transmission trials have shown that PRV-1 is less adapted to seatrout than to Atlantic salmon (Grefsrud et al., 2018). Consequently, this survey has revealed a much higher prevalence and viral load of PRV-3 in seatrout than previously recorded for PRV-1, and furthermore, that the PRV-3 sequences group together with sequences from rainbow trout and Coho salmon in Chile. According to Dhamotharan et al., 2018, two clades of PRV-3 are present. Unfortunately, we were not able to assign sequences to specific clades.

| PRV-3 in wild Atlantic salmon
Transmission trials have shown that PRV-3 is less adapted to Atlantic salmon than rainbow trout (Hauge et al., 2017). Cohabitation in tanks with infected seatrout could be a bias in the study. However, none of the PRV-3 positive salmon were kept in tanks with PRV-3 positive seatrout.

| Reservoirs and interaction between farmed and wild salmonids
None of the 11 anadromous Arctic char were virus-positive.
However, the sample size was low and only suitable for detecting highly prevalent pathogens (Cameron & Baldock, 1998).
Altogether, 79 brown trout from three non-anadromous lakes were tested without detecting PRV-3. Similarly, the virus was absent in 40 non-anadromous (landlocked) salmon (Table 1). This could mean that salmonids in relatively naïve non-anadromous watercourses are of minor importance as a reservoir. PRV-3 is nevertheless common in seatrout from anadromous watercourses, meaning that rainbow trout farms that use these water sources are at risk.
In Norway, farming of rainbow trout takes place as two nearly separate production lines: the freshwater-based small-scale inland aquaculture (based on of roe from the marine production), and the large-scale marine production along the coast, where juvenile stages are produced in freshwater. In the marine aquaculture of rainbow trout, PRV-3 is present at all stages of the production cycle, including hatcheries that do not use sea water in their production (Olsen et al., 2015).

| CON CLUS IONS
Piscine orthoreovirus subtype 3 (PRV-3) is a common virus in seatrout in Norway. Compared to rainbow trout, Atlantic salmon are less susceptible to the virus, which may explain the low prevalence and viral loads recorded in wild specimen in this study. The absence of PRV-3 in both non-anadromous Salmo trutta (brown trout) and Salmo salar (landlocked salmon) indicates that the virus may be linked to the marine environment in Norway. Olsen et al., (2015) referred to PRV-3 as a rainbow trout associated virus. This study strongly supports an association with seatrout.

ACK N OWLED G EM ENTS
Thanks to the Genebank for wild Atlantic salmon and two stock enhancement hatcheries that provided broodfish samples to the study. Finally, special thanks to the two reviewers that contributed significantly to the quality of the final manuscript.

CO N FLI C T O F I NTE R E S T S
No conflict of interest has been identified for any of the authors.