Infections by Renibacterium salmoninarum and Chapin are associated with reduced growth of juvenile Chinook , tshawytscha (Walbaum), in the Northeast Pacific Ocean

Sandell, T. A., Teel, D. J., Fisher, J., Beckman, B., & Jacobson, K. C. (2015). Infections by Renibacterium salmoninarum and Nanophyetus salmincola Chapin are associated with reduced growth of juvenile , Oncorhynchus tshawytscha (Walbaum), in the Northeast Pacific Ocean. Journal of fish diseases, 38(4), 365-378. doi:10.1111/jfd.12243

10.1111/jfd.12243 John Wiley & Sons Ltd.

Version of Record http://cdss.library.oregonstate.edu/sa-termsofuse Journal of Fish Diseases 2015, 38, 365–378 doi:10.1111/jfd.12243

Infections by Renibacterium salmoninarum and Nanophyetus salmincola Chapin are associated with reduced growth of juvenile Chinook salmon, Oncorhynchus tshawytscha (Walbaum), in the Northeast Pacific Ocean

T A Sandell1, D J Teel2, J Fisher3, B Beckman4 and K C Jacobson5

1 Department of Microbiology, Oregon State University, Corvallis, OR, USA 2 NOAA , Northwest Fisheries Science Center, Manchester Research Laboratory, Manchester, WA, USA 3 Cooperative Institute for Marine Resources Studies, Oregon State University, Newport, OR, USA 4 NOAA Fisheries, Northwest Fisheries Science Center, Seattle, WA, USA 5 NOAA Fisheries, Northwest Fisheries Science Center, Newport, OR, USA

first year at sea has been linked to survival for Abstract some stocks of Chinook salmon, the infections We examined 1454 juvenile Chinook salmon, may therefore play a role in regulating these pop- Oncorhynchus tshawytscha (Walbaum), captured in ulations in the Northeast Pacific Ocean. nearshore waters off the coasts of Washington and Keywords: disease ecology, survival, parasite, Oregon (USA) from 1999 to 2004 for infection pathogen. by Renibacterium salmoninarum, Nanophyetus sal- mincola Chapin and skin metacercariae. The prev- alence and intensities for each of these infections were established for both yearling and subyearling Introduction Chinook salmon. Two metrics of salmon growth, Pathogens and parasites can regulate popula- weight residuals and plasma levels of insulin-like tions through direct mortality, by affecting the host growth factor-1, were determined for salmon organism’s ability to compete intra- or interspecifi- infected with these pathogens/parasites, both indi- cally and by affecting fitness or growth. Once con- vidually and in combination, with uninfected fish sidered to play a minor role in ecological used for comparison. Yearling Chinook salmon communities, parasites have been observed to have infected with R. salmoninarum had significantly density-dependent effects and thus may influence reduced weight residuals. Chinook salmon the density of a species. Aquatic pathogens can reg- infected with skin metacercariae alone did not ulate fish populations via host mortality and have significantly reduced growth metrics. Dual reduced fecundity (Lester 1984), particularly when infections were not associated with significantly pathogen transfer rates are high, as among farmed more severe effects on the growth metrics than or captive fish populations (Sindermann 1987). single infections; the number of triple infections Patterson (1996) modelled effects of the parasite was very low and precluded statistical comparison. Ichthyophonus hoferi Plehn & Mulsow on North Sea Overall, these data suggest that infections by these herring, Clupea harengus Linnaeus, and estimated organisms can be associated with reduced juvenile that the population size was reduced by 10–20% Chinook salmon growth. Because growth in the during epizootics. A field study that analysed the influences of an ectoparasite on the bridled goby, Coryphopterus glaucofraenum Gill, showed that Correspondence T A Sandell, Department of Microbiology, Oregon State University, Corvallis, OR 97331, USA infected individuals had significantly reduced (e-mail: [email protected]) growth, gonad mass and higher instantaneous

Ó 2014 John Wiley & Sons Ltd 365 Journal of Fish Diseases 2015, 38, 365–378 T A Sandell et al. Infections affect salmon growth

mortality (Finley & Forrester 2003). The effect of trematodes are also acquired in fresh water and these infections on growth is of interest because persist for a minimum of several months (Fergu- numerous studies have demonstrated that size can son et al. 2010) to years (Hoffmann & Putz be a critical factor in the survival of juvenile fish in 1965) and may persist for the entire lifespan of their first year at sea (Sogard 1997; Zabel & Achord the infected salmon host. Species identification 2004; Van Doornik et al. 2007; Duffy & Beau- can be difficult because the morphology of the champ 2011). Parasitic infections that reduce metacercariae is variable; ‘Neascus’ is a larval form juvenile growth can reduce survival and thus may common to the families Strigeidae and Diplos- play a role in regulating fish populations, including tomatidae and may refer to several species (Olson Chinook salmon. et al. 2003). Thus, we refer to cases of black spot The pathogenic bacterium Renibacterium salmon- infection simply as ‘skin metacercariae’. Acute, inarum, metacercarial stages of the trematode Nan- heavy infections with A. donicus were shown to ophyetus salmincola Chapin and metacercarial stages cause mortality in small , O. kisutch of the trematodes causing ‘black spot’ disease (either (Walbaum) (6 cm in length) (Niemi & Macy Neascus or Apophallus sp.) have been documented 1974), and Ebersole et al. (2006) noted lower- in many coastal rivers and estuaries of the Pacific than-expected overwintering survival in coho sal- Northwest (Arkoosh et al. 2004). All are acquired mon smolts parasitized by Apophallus spp. These in fresh water and persist in the host after ocean results were confirmed by Ferguson et al. (2010), entry. Renibacterium salmoninarum is the causative who reported dramatic declines in the abundance agent of bacterial kidney disease (BKD) in suscepti- of Apophallus spp. between the parr and smolt ble salmonids. The infection is typically acquired in stages in coho salmon, suggesting parasite-related the fresh water phase of the salmonid life cycle and mortality. is transmitted both vertically (in ovo) and horizon- Few studies have examined the effects of infec- tally (fish-to-fish, via the faecal–oral route) within tions on juvenile salmon growth during their mar- populations (Evelyn, Prosperi-Porta & Ketchenson ine phase. The goal of this study was to determine 1986; Balfry, Albright & Evelyn 1996; McKibben the prevalence (percentage infected) and intensity & Pascho 1999). Studies by Rhodes et al. (2006, (number of parasites) of these common infections 2011) suggest that the infection can also be passed and to determine whether juvenile Chinook salmon horizontally among juvenile salmon in salt water. with single or multiple infections have reduced Infection with R. salmoninarum can cause epizoot- growth early in marine residency. We tested for cor- ics, particularly in hatcheries where fish densities are relations between infection intensity and the growth high (Banks 1994), but more typically leads to metrics and looked for evidence that multiple infec- chronic infections with persistent low-level mortal- tions resulted in increased mortality. This research ity. There is some evidence of recovery from the was part of a larger study of juvenile salmon ecology infection in fresh water (Bruno & Munro 1986; off the Oregon and Washington coast (Peterson Pascho, Elliot & Streufert 1991; Cvitanich 2004; et al. 2010) in areas where juvenile Chinook sal- Turgut et al. 2008). mon are predominately from the Columbia River The trematode N. salmincola infects juvenile sal- Basin (Fisher & Pearcy 1995; Daly et al. 2012). mon as a second intermediate host in fresh water. Our study comprised 6 years of sampling and, in The encysted metacercariae survive for the life of the case of R. salmoninarum, utilized newer and the salmon (Bennington & Pratt 1960; Farrell, more sensitive detection assays, providing the most Lloyd & Earp 1964) or until the fish is eaten by the comprehensive study to date on the role of these parasite’s definitive host, either a mammal (Mille- common infections in juvenile Chinook salmon mann & Knapp 1970) or piscivorous bird (Farrell growth. et al. 1964). Several studies have demonstrated the ability of N. salmincola to increase mortality in infected juvenile salmonids (Baldwin, Milleman & Methods Knapp 1967; Butler & Milleman 1971; Newcomb, Fish capture, storage and classification Snell & Waknitz 1991; Jacobson et al. 2008). Metacercarial stages of the trematodes which Juvenile salmon were captured by surface trawl in cause ‘black spot’ disease encyst in the salmon nearshore marine waters in May, June and Sep- host’s skin or subcutaneous muscle. These tember of 1999–2004. Although variations

Ó 2014 John Wiley & Sons Ltd 366 Journal of Fish Diseases 2015, 38, 365–378 T A Sandell et al. Infections affect salmon growth

occurred due to daylight and weather, we typically they were stored at 80 °C at the conclusion of sampled three transects in May and eight transects each sampling cruise. in June and September (Fig. 1). These transects ranged from Tatoosh Island, WA (48.337N; Juvenile salmon age class determination and 124.718W) south to Newport, Oregon jack exclusion (44.40N; 124.4W). Along each transect, stations were typically sampled from 1 to a maximum of Because sexually precocious salmon (‘jacks’) are 25 nautical miles offshore or until the continental physiologically different from juveniles, we shelf break was reached. At each station, an 88-m removed them from our data set. Jacks were iden- rope trawl with an 18 9 30 m opening and a tified by a combination of criteria. First, any fish small (0.8-cm)-diameter mesh liner in the cod with high plasma 11-ketotestosterone hormone end was towed at ~6.5 km h 1 at the surface for levels (Campbell & Emlen 1996) or a high gonad 30 min during daylight hours. Juvenile salmon index [testes weight >2% of total body weight; were sorted from the general catch, immediately (Fisher & Pearcy 1995)] were excluded. In June placed on ice, identified to species and measured and September, fork length was used as a third fork length (FL) to the nearest mm while on deck. criterion to exclude any fish with fork lengths that In May and June, blood samples from a subset of exceeded cut-offs at the ~95th percentile the catch were taken and processed for plasma (243 mm in June; 270 mm in September). Occa- insulin-like growth factor-1 (IGF-1) analysis. Each sionally, a single mm separated several fish at the juvenile salmon was assigned a unique identifica- cut-off length; in those cases, the 97% percentile tion number and separately frozen on ice until was used. The remaining juveniles (non-jacks) were assigned to age classes based on fork length at capture date (Fisher & Pearcy 1990). Juvenile Target Station Locations Chinook salmon with a fork length ≥120 mm in May, 140 mm in June and 250 mm in September

Tatoosh Island were classified as subyearlings (individuals migrat- 48° N Washington ing to the ocean in their first year); fish exceeding La Push LaPush these cut-offs were considered yearlings (those that migrate to sea in their second year).

North Region Queets River Catch subsampling, tissue sampling and 47° N pathogen analysis Grays Harbor Of the juvenile salmon captured off the Oregon Willapa Bay and Washington coasts from 1999 to 2004, 712 Columbia River subyearling Chinook salmon and 742 yearling Region Astoria Columbia River Chinook salmon were subsampled for inclusion in 46° N Oregon this study. Catches of subyearling Chinook sal- mon were lower in summer (May and June; n = 211) than in fall (September and October; Cape Meares Tillamook n = 501) due to their later ocean entry; only two South Region subyearlings were captured in the summer of 45° N Cascade Head 1999. Yearling Chinook salmon tend to migrate to the ocean earlier in the year, and this was Newport Newport reflected in our catch (for yearling Chinook sal- = = 125° W 124° W 123° W mon: summer n 669, fall n 73). In the laboratory, juvenile salmon were partially Figure 1 Map showing transects and study regions (bullets thawed, weighed, re-measured for fork length denote individual sampling stations; dotted line is the 180-m (mm), scanned for marks that indicate hatchery isobath). Figure modified from that available at: http://www. origin (fin clips, coded wire tags (CWT), passive cbfwa.org/solicitation/components/forms/Proposal.cfm?PropID= integrated transponder (PIT) or latex tags), and 569#sect10. the presence and number of parasitic skin

Ó 2014 John Wiley & Sons Ltd 367 Journal of Fish Diseases 2015, 38, 365–378 T A Sandell et al. Infections affect salmon growth

metacercariae (‘black spot’) were noted. Each sal- Analysis of insulin-like growth factor-1 mon was then dissected; care was taken to prevent DNA contamination between specimens by soak- Blood was taken from the caudal vein using a ing all dissecting instruments in a 20% bleach heparinized syringe and stored on ice for up to solution for at least 2 min, followed by two water 2 h. Samples were then spun in a microcentrifuge rinses. at 3000 g for 5 min. Blood plasma was collected Renibacterium salmoninarum prevalence was anal- and frozen at 20 °C while at sea (up to ysed by removal of the anterior kidney; these were 10 days) and then transferred and stored at â frozen at 80 °C in Whirlpak tissue bags (Nasco) 80 °C on land. Plasma IGF-1 concentration until processed. DNA extraction was conducted was quantified in acid–ethanol-extracted samples using the Qiagen DNeasy Tissue Kit following the using the radioimmunoassay developed by Shi- manufacturer’s instructions for Gram-positive bac- mizu et al. (2000) using barramundi antibody and teria. Following elution, samples were either run recombinant salmon IGF-1 as label and standard. immediately or stored at 80 °C until analysed by nested polymerase chain reaction (nPCR) (Pascho, Data analyses Chase & McKibben 1998). To optimize results on our thermocycler (MJ research P100), the thermo- Statistical analyses and graphs were created using cycling regime was slightly modified as follows: StatGraphics (2009). Data on fish weight (Wt) 88 °C for 3 min, 94 °C for 2 min, then 30 cycles were natural log (ln)-transformed prior to analysis of 94 °C for 30 s, 60 °C for 1 min and 72 °C for to achieve a normal distribution; differences in 1 min. Reagent concentrations and primers were weights reported in the text refer to untrans- unchanged from Pascho et al. (1998). formed average weights. Data for each age class To quantify the R. salmoninarum infection were screened for outliers from the simple regres- level, nPCR-positive samples were also analysed in sion of ln(Wt) 9 ln(fork length) (FL); those with duplicate wells using the qPCR protocol of studentized residuals >3 times the mean, or those Rhodes et al. (2006), modified by the addition of with leverage exceeding three times the average minor groove-binding (MGB) probes (Applied value, were removed (Zar 2010). Prior to analysis Biosystems, Inc.) to increase sensitivity (Sandell & of the effect of single infections on the growth Jacobson 2011). Although a number of qPCR metrics, we tested for interactions among infec- assays for R. salmoninarum exist, this one was cho- tions by more than one parasite or the bacterium sen because it was the best at detecting those sam- and ln(Wt) by ANOVA and ANCOVA; when ples with medium to high infection levels, had the significant interactions occurred, comparisons were lowest coefficient of variation (for a comparison of limited to fish with single infections only. available R. salmoninarum detection and quantifi- To examine the effects of multiple infections on cation assays, see Elliott et al. 2013). We defined fish weight, MANOVA models of ln(Wt) were ‘severely infected’ samples as those with an ampli- constructed for each age class using all significant < < fication threshold value (Cq score) 20, because (P 0.05) factors (possible factors included year, samples entering exponential amplification below mark status (hatchery vs. unmarked) and region of this cycle number have large numbers of bacterial capture (north of the Columbia River, Columbia genomes present (Sandell & Jacobson 2011). River region or south of the Columbia River To determine the presence and intensity of region; see Fig. 1) because increased time of ocean N. salmincola infection, the posterior half of the residency may result in larger fish). To examine kidney was excised and placed in a 4 oz. Whirl- the effects of single infections on the growth met- â pak tissue bag (Nasco) and refrozen at 20 °C. rics, comparisons among age classes were made At a later date, each sample was thawed, com- using data from all years (1999–2004). Separate pressed between two glass plates, and the number comparisons were made by season (because fish of metacercariae was counted with light micros- caught in the fall are heavier than fish in summer) copy at 1009 magnification. The majority of data and, where possible given the constraints of sam- presented here for N. salmincola from 1999 to ple sizes, by hatchery vs. unmarked (because 2002 are from Jacobson et al. (2008). Some addi- hatchery fish are larger at the time of release than tional archived samples from 2003 to 2004 were their naturally produced conspecifics). For each also analysed for this study. age class, models were constructed by season to

Ó 2014 John Wiley & Sons Ltd 368 Journal of Fish Diseases 2015, 38, 365–378 T A Sandell et al. Infections affect salmon growth

reflect age-related differences in size, where ‘sum- comparisons for proportions procedure were used; mer’ included fish captured in May and June and results from these tests do not provide specific ‘fall’ consisted of fish captured in September. P-values, but do identify results beyond signifi- The model outputs were compared to the sim- cance cut-offs (P < 0.05, P < 0.01, etc.) (Zar ple model of ln(Wt) as predicted by ln(FL) for 2010). Prevalence data were analysed to determine each age class (De Robertis & Williams 2008). whether the number of infected fish deviated from The best model was then used to generate a the expected values, indicating potential pathogen regression line; vertical deviations from this line or parasite-related mortality in groups with low were the weight residuals. Positive residuals indi- coinfection prevalence. For all other statistical cated fish that were heavier than predicted, and tests, significance was defined as P < 0.05. those with negative residuals weighed less than predicted. Group (e.g. infected vs. uninfected) residuals were compared by ANOVA if assump- Results tions of equal variance were met, and by the Krus- Renibacterium salmoninarum: prevalence and kal–Wallis nonparametric ANOVA if the infection severity variances were unequal. Homogeneity of variance between the groups was tested using the F-test Both yearling and subyearling Chinook salmon and the Kolmogorov–Smirnov comparison of dis- showed interannual variability in the prevalence of tributions (Sokal & Rohlf 1987). For post hoc R. salmoninarum infection. The prevalence was analysis of ANOVA data between multiple infec- significantly higher in 2000 (P < 0.0001) than in tion groups, the Bonferroni adjustment was used all other years for both age classes, with 65.2% of to reduce multiple comparisons bias. subyearling and 47.5% of yearling Chinook sal- Data on the levels of IGF-1 were only available mon infected (Fig. 2). The prevalence among sub- in sufficient numbers for yearling Chinook salmon yearlings was significantly lower in both 1999 in the summers of 2000–04 (n = 229). These (14%) and 2004 (5.9%) than in the other years, data were normally distributed after the removal but was similar from 2001 to 2003 (average of of outliers and were analysed using MANOVA 28%). Among yearling Chinook salmon, the prev- models (StatGraphics 2009) that accounted for alence was significantly higher (P < 0.0001) in other factors that may have influenced IGF-1 lev- both 2000 and 2002 (41.8%) and significantly els, such as year, region of capture, hatchery status (marked vs. unmarked) and coinfection with more than one pathogen or parasite. 80 107/139 106/33 124/109 N 66/158 191/177 118/126 For comparisons of mean infection intensity b between age classes or seasons, the nonparametric Subyearling Mann–Whitney U-test was utilized; by definition, 60 Yearling the mean intensity of infection is calculated using y only infected fish (Bush et al. 1997). Comparisons y of infection intensity between the years for a sin- 40 c x gle age class were made using the Kruskal–Wallis c x test on log-transformed data to equalize the vari- c ances. To investigate the effect of infection inten- 20 Prevalence (% infected) Prevalence a z sity on fish weight, regression with a negative z binomial distribution was used due to overdisper- a sion (i.e. a few fish were heavily infected, but 0 most had few parasites). The models included 1999 2000 2001 2002 2003 2004 year, hatchery status, region of capture and infec- Figure 2 Renibacterium salmoninarum prevalence among juve- tion by the bacterium or parasites (individually) as nile Chinook salmon, Oncorhynchus tshawytscha (Walbaum), by possible factors. year and age class. The comparisons shown here were con- v2 Contingency tables for the chi-square ( ) test ducted within (not between) the age classes. Significant differ- were used to compare the binary prevalence data; ences between the years are indicated with letters; for P-values reported are Fisher’s exact P-values. To subyearlings (striped bars; a–c), for yearlings (solid bars, x–z). compare prevalence data by season, the multiple Sample sizes (N) are provided across the top of the figure.

Ó 2014 John Wiley & Sons Ltd 369 Journal of Fish Diseases 2015, 38, 365–378 T A Sandell et al. Infections affect salmon growth

lower in 2001 (12.1%) and 2004 (9.5%), with an Chinook salmon in 2001 and 2004 precluded average of 38.1% in 2002–03 (Fig. 2). comparison between individual years, plasma IGF-1 Only 22.9% of the salmon infected by R. sal- levels were lower in R. salmoninarum-infected moninarum (those samples positive by nPCR) had yearlings than among those that were uninfected detectable levels of infection by qPCR, indicating in all years except 2000 (data not shown). Com- that most (77.1%) of the juvenile Chinook sal- parisons were also limited to fish captured within mon sampled in this study had low infection a single region (Fig. 1) because differences in emi- severities at the time of capture (Table 1). Year- gration timing would have biased the results. ling Chinook salmon had significantly more With all the years combined, the difference in qPCR-positive results (P < 0.01) than subyear- plasma IGF-1 levels between infected and unin- lings. For both age classes, the highest proportion fected fish in the northern region, which had the of qPCR-positive samples occurred in 2000 largest samples (n = 136), was not significant (Table 1), mirroring the higher prevalence by (P = 0.07). Notably, if 2000, the year with the nPCR and the highest proportion of severe infec- highest prevalence and severity of R. salmonina- < tions (Cq 20) in that year. There were very few rum infection, was excluded, the difference was severe infections in the other years (Table 1). significant for yearling Chinook salmon captured in this region (P = 0.04; data not shown). Renibacterium salmoninarum infection: effects on growth metrics Nanophyetus salmincola: prevalence and infection intensity To examine the potential effects of R. salmonina- rum infection on growth, we compared the weight The prevalence of N. salmincola infection varied residuals of infected and uninfected salmon of interannually in both subyearling and yearling both age classes by season, with all years com- Chinook salmon (Fig. 4a,b). Among subyearlings, bined to increase sample sizes. In both summer the prevalence ranged from 36.4% to 54.7% and and fall, catches of uninfected, unmarked yearling was significantly higher (P = 0.03) in 2001 salmon had higher weight residuals than those (54.7%) than in all other years (Fig. 4a). Nano- that were infected, but the difference was only sig- phyetus salmincola prevalence was more variable in nificant in fall (summer P = 0.07; fall P = 0.04) yearling Chinook salmon, ranging from a low of (Fig. 3). Too few infected, marked yearling sal- 16.6% in 1999 to a high of 65.2% in 2001, and mon were sampled to permit comparisons by sea- was significantly higher (P < 0.001) in 2001 and son. No significant differences among weight 2002 than in 1999 and 2004 (Fig. 4b). residuals of uninfected and infected subyearling The mean intensity of N. salmincola in subyear- Chinook salmon split by marked status were ling Chinook salmon was consistently below 46 found in summer or fall (data not shown). metacercariae and was significantly lower in 2000 Plasma IGF-1 data were only collected from a (mean = 31.9) and 2002 (mean = 20.7) than in subset of yearling Chinook salmon captured in 2003 and 2004 (P < 0.001; Fig. 4a). Mean intensi- each region (Fig. 1) in the summers of 2000–04. ties in yearlings were generally similar to subyear- Although small sample sizes of infected yearling lings, although in 2000 exceeded 80 metacercariae; a

Table 1 Renibacterium salmoninarum assay results by Chinook salmon, Oncorhynchus tshawytscha (Walbaum), age class and year, – = = = < 1999 2004 (‘nPCR’ nested PCR; ‘qPCR’ quantitative PCR; ‘severely infected’ Cq score 20)

Age class Assay 1999 2000 2001 2002 2003 2004 Total

Subyearling Total sampled 107 66 106 191 124 118 712 Chinook % nPCR positive 14 65.2 28.3 25.7 31.5 5.9 25.7 % qPCR positive 3.7 7.6 2.8 6.3 2.4 0.85 3.9 % severely infected 0 3 0.9 0.5 0 0 0.6 Yearling Total sampled 139 158 33 177 109 126 742 Chinook % nPCR positive 26.6 47.5 12.1 41.8 32.1 9.5 31.9 % qPCR positive 7.2 19.6 0 9 6.4 3.2 9.2 % severely infected 0 3.2 0 1.1 0.9 0 1.1 Total sampled by year 246 224 139 368 233 244 1454

Ó 2014 John Wiley & Sons Ltd 370 Journal of Fish Diseases 2015, 38, 365–378 T A Sandell et al. Infections affect salmon growth

37 84 N 18 43 N 0.015 0.015 P = 0.07 P = 0.035 0.01 0.01

0.005 0.005

0 0

–0.005 –0.005 Figure 3 Weight residuals of Renibacterium salmoninarum-infected –0.01 –0.01 (striped bars) and uninfected (solid bars) –0.015 –0.015

unmarked yearling Chinook salmon, residuals Weight residuals Weight Oncorhynchus tshawytscha (Walbaum), in –0.02 –0.02 summer (left panel) and fall (right panel), –0.025 –0.025 1999–2004 combined (sample sizes, N, are Summer Fall –0.03 –0.03 provided at the top of the figure; error bars Infected Uninfected Infected Uninfected are 95% confidence intervals).

(a) (b)

Figure 4 Annual prevalence (solid bars) and mean intensities (striped bars) of Nanophyetus salmincola Chapin infections among subyearling (a) and yearling (b) Chinook salmon, Oncorhynchus tshawytscha (Walbaum). Asterisks indicate significant differences in prevalence between the years; significant differences in mean intensities are indicated with letters. Error bars are 95% confidence intervals; sample sizes (N) are provided across the top of each figure.

large confidence interval high may have precluded variables: year, hatchery status, capture region, a statistically significant difference among years R. salmoninarum and skin metacercariae. For (Fig. 4b). The most heavily infected fish sampled weight, the best model was: was a yearling Chinook salmon captured in the Log WtðÞ g summer of 2000, with 961 N. salmincola À ¼ exp 4:566 0:0007122 number of metacercariae. N : salmincola metacercariae 0:1112 Year ¼ 1999 0:0429 Nanophyetus salmincola infection: growth ¼ þ : metrics and regression results Year 2000 0 257 Year ¼ 2001 þ 0:2052 There were no significant differences in the weight ¼ : residuals of uninfected vs. N. salmincola-infected Year 2002 0 1046 subyearling or yearling Chinook salmon in sum- Year ¼ 2003 þ 1:19 mer or fall. There were also no significant differ- Hatchery ¼ Hatchery 0:3626 ences in plasma IGF-1 levels of N. salmincola- Áà sampling region ¼ Columbia River infected vs. uninfected subyearling or yearling , Chinook salmon (data not shown). where ‘sampling region’ is the region of capture Negative binomial regression models were used (see Fig. 1). As the main factor, the number of to examine whether infection intensity was N. salmincola metacercariae was highly significant correlated with fish weights or plasma IGF-1, tak- (P < 0.0001) and was inversely correlated with sum- ing into account the contribution of other mer yearling Chinook salmon weights, explaining

Ó 2014 John Wiley & Sons Ltd 371 Journal of Fish Diseases 2015, 38, 365–378 T A Sandell et al. Infections affect salmon growth

17.8% of the deviance (the equivalent of R2 for = 7.3) (Fig. 6a). Mean intensity among yearling linear regression). In addition, no fish of >120 g Chinook salmon was highest in 1999 (mean = 63; had more than 200 metacercariae (Fig. 5). range: 12–150; 4/139 infected) and 2000 (mean = Marked status (hatchery vs. unmarked fish) was 56.8; range: 2–100; 4/158 infected) (Fig. 6b). In also significant in the model, as was the region of the other years, mean intensities were much lower, capture. The regression of N. salmincola metac- ranging from 8.6 to 19.8. However, there were no ercariae counts was not significantly correlated significant differences in mean intensity between with yearling plasma IGF-1 levels (data not the years due to the small numbers of infected year- shown). Very few N. salmincola-infected yearling ling Chinook salmon and the high variability Chinook salmon were caught in fall (n = 21), encountered in 1999 and 2000 (Fig. 6b). precluding comparison. There were no significant correlations between the number of N. salmincola Skin metacercariae infection: growth metrics and metacercariae and the weights of infected subyear- regression results. Among subyearling and yearling ling Chinook salmon (n = 230). Chinook salmon, there were no significant differ- ences in the weight residuals of infected vs. unin- fected cohorts in any year. Infection by skin Skin metacercariae: prevalence and infection metacercariae was also not associated with signifi- intensity cant differences in IGF-1 levels in any of the Infection by skin metacercariae was less common comparisons. than the other infections. Overall, subyearling The negative binomial regression models for Chinook salmon had a significantly higher preva- skin metacercariae showed no significant correla- lence (32.3%) than yearlings (5.5%) (P < 0.0001; tions between the numbers of skin metacercariae Fig. 6a). The prevalence among subyearlings was and fish weights or plasma IGF-1 levels for subye- significantly higher in 1999 (44.9%) than in 2003 arling Chinook salmon; too few yearlings were (20.6%) (P = 0.002; Fig. 6a). Among yearling infected to permit comparison. Chinook salmon, the prevalence of skin metac- ercariae was significantly higher (P < 0.0001) in Multiple infections 2001 (30.3%) than in the other years, which ran- ged from 2.5% to 6.4% (Fig. 6b). To examine the effects of multiple infections on The mean intensities of infection by skin metac- weight residuals, juvenile salmon were categorized ercariae were consistently low for subyearling Chi- by the number of single, double (‘coinfected’) or nook salmon in all years, although significantly triple infections (simultaneous infection by R. sal- higher in 2001 (mean = 10.6) than in 2002 (mean moninarum, N. salmincola and skin metacercariae). In 1999–2004, the prevalence of dual infections (any combination of two or more infections) was 200 significantly higher (P < 0.001) among subyear- ling Chinook salmon (24.7%) than among year- ling Chinook salmon (11.9%) (Table 2). Dual 150 infection by both skin metacercariae and R. sal- moninarum was uncommon in yearling Chinook 100 salmon (n = 2) and was therefore excluded from statistical comparisons. Due to the low number of Fish weight (g) Fish weight 50 qPCR-positive samples, data on R. salmoninarum infection intensity were not compared statistically with infections by the trematodes. 0 200 400 600 800 Subyearling Chinook salmon had significantly # N. salmincola metacercariae (P < 0.001) more triple infections (4.92%, n = 35) = Figure 5 Summer (May and June) yearling Chinook salmon, than yearling Chinook salmon (1.08%, n 8) Oncorhynchus tshawytscha (Walbaum): negative binomial regres- (Table 2). Triple infections were rare (3% of all sion of Nanophyetus salmincola Chapin metacercariae counts vs. salmon), precluding statistical comparison by sea- fish weight (g). Bracketing lines indicate the 95% confidence son and between hatchery vs. unmarked fish. intervals (N = 192). There were no significant differences in the weight

Ó 2014 John Wiley & Sons Ltd 372 Journal of Fish Diseases 2015, 38, 365–378 T A Sandell et al. Infections affect salmon growth

(a) (b)

Figure 6 Annual prevalence (black bars) and mean intensities (grey bars) of skin metacercarial infections among subyearling (a) and yearling (b) Chinook salmon, Oncorhynchus tshawytscha (Walbaum). Asterisks indicate significant differences in prevalence between the years; significant differences in mean intensities are indicated with letters. Error bars are 95% confidence intervals; sample sizes (N) are provided across the top of each figure.

residuals between the various infection combina- Renibacterium salmoninarum tions for any subyearling or yearling Chinook sal- mon stocks; statistical power was limited due to Overall, R. salmoninarum was detected in 29% of small sample sizes. the subyearling and yearling Chinook salmon analysed. This is similar to the prevalence of 27% in juvenile Chinook salmon from Puget Sound Discussion reported by Rhodes et al. (2006), 31.8% reported Measuring the direct effects of infection on fish in 2011 (Rhodes et al. 2011) and the prevalence hosts in the wild is difficult because infections that reported by Van Gaest et al. (2011) among year- reduce the growth or physiological performance of ling Chinook salmon from Upper Columbia River infected hosts are likely to result in increased pre- tributaries after barging. In contrast, a series of dation or direct mortality, quickly removing these studies on R. salmoninarum conducted in the fish from the population (Gordon & Rau 1982; Columbia River in the 1980s (Pascho, Elliot & McCallum & Dobson 1995; Bakke & Harris Achord 1993; Maule et al. 1996; Elliott et al. 1998). A few studies have examined the effects of 1997) reported much higher prevalence in juvenile dual infections in the laboratory [e.g. (Seppanen Chinook salmon (frequently in excess of 75%) at et al. 2009)], but most of these rely on short-term hatcheries and in-river despite using the less-sensi- survival (Lee et al. 1999) or swimming perfor- tive enzyme-linked immunosorbent assay (ELISA), mance (Blake, Kwok & Chan 2006) as the metric which measures soluble antigen produced by the of parasite/pathogen effect. In addition, the con- bacterium. While broodstock screening for R. sal- centrations typically used in bacterial and viral moninarum at state and federal hatcheries is challenges may not effectively simulate the effects thought to have reduced the prevalence of the of lower, chronic exposure and infection on the bacterium (Pascho et al. 1991, 1993; Maule et al. host in a natural setting. 1996), the dramatically lower prevalence and Our goal was to determine the prevalence and severities reported here for Chinook salmon from severity of three common infections in juvenile the nearshore (using more sensitive assays) again Chinook salmon shortly after marine entry, as well raises the question of whether in-river or estuarine as to determine the effects of these infections, indi- pathogen-related mortality is significant. Mortality vidually and in combination, on short- (plasma associated with R. salmoninarum infection has IGF-1) and long-term (weight residuals) growth been reported to increase among juvenile salmon metrics. Because plasma IGF-1 levels respond smolts during the fresh water–salt water transition rapidly to diet, measuring plasma IGF-1 levels at (Banner et al. 1986; Sanders et al. 1992), but the the time of capture allows for the assessment of use of different detection assays during the differ- how well an individual salmon has been feeding on ent sampling years makes comparison difficult. A a temporal scale of hours to days, depending on paired study with in-river and nearshore sampling temperature (Beckman et al. 2004). during the same years, using identical assays,

Ó 2014 John Wiley & Sons Ltd 373 Journal of Fish Diseases 2015, 38, 365–378 T A Sandell et al. Infections affect salmon growth

Table 2 Multiple infection prevalence and sample sizes of Chinook salmon, Oncorhynchus tshawytscha (Walbaum), by age class, 1999–2004 combined (abbreviations are: N. sal = N. salmincola, R. sal = salmoninarum, SkinMet = Skin metacercariae)

Subyearling Yearling Chinook Subyearling Chinook Yearling salmon Chinook salmon Chinook Infection codes prevalence (%) salmon (N) prevalence (%) salmon (N)

0. None 34.3 244 47.8 355 1. N. sal 15.7 112 16.0 119 2. SkinMet 10.4 74 1.1 8 3. R. sal 10.0 71 22.1 164 4. N.sal/SkinMet 13.9 99 3.1 23 5. N. sal/R. sal 7.7 55 8.5 63 6. SkinMet/R. sal 3.1 22 0.3 2 7. All 3 4.9 35 1.1 8 Totals 712 742

protocols and analysis of stock origin, is needed to Chinook salmon, infected fish had lower weight address this question. residuals than those that were uninfected in all but There were very few severely infected fish in two comparisons. The lack of plasma IGF-1 data either age class (here defined as those with a Cq for subyearling Chinook salmon makes it difficult score <20). This result is similar to that reported to assess the overall effect of R. salmoninarum infec- by Rhodes et al. (2011), who used the quantitative tions on subyearlings. Because the infection (when fluorescent antibody technique to determine that present) has had less time to progress in the younger <1% of the juvenile Chinook salmon they captured age class, a lack of statistically significant effects on had bacterial cell counts >1000 cells per fish. weight and growth in subyearlings compared to Although the assay methods differed, these data yearlings may be due to: (i) lower infection severities suggest that very few juvenile Chinook salmon in subyearlings, (ii) the limited time the infection severely infected with R. salmoninarum survive to has had to affect their physiology or (iii) methodo- reach the early marine phase of the life cycle. The logical difficulties in detecting lower infection inten- survival rate of even moderately infected fish dur- sities with our assays. It is also possible that the ing the marine phase is unknown if R. salmonina- subyearlings sampled in this study differed in their rum continues to multiply during the hosts’ genetic resistance to R. salmoninarum infection remaining time at sea (2–5 years). However, a compared to the yearling Chinook salmon. study by Kent et al. (1998) that included pathogen Of the 6 years we sampled, the prevalence, prevalence among adult Chinook salmon captured severity and number of severe R. salmoninarum off of British Columbia found a R. salmoninarum infections were anomalously high among both su- prevalence of 58.4% when assayed by ELISA, sug- byearling (65%) and yearling (48%) Chinook sal- gesting that infected salmon can survive to the mon in 2000. Infected yearling Chinook salmon adult marine phase of the life cycle. This result had lower weight residuals and plasma IGF-1 than must be interpreted with caution, however, as the uninfected yearlings in all years except 2000; ELISA detects a soluble antigen of R. salmonina- plasma IGF-1 levels were significantly lower in rum that can persist after an active infection is infected than in uninfected yearlings only when cleared (Pascho, Goodrich & McKibben 1997; those captured in 2000 were excluded. The vari- Nance et al. 2010); horizontal transmission during ability among years suggests that environmental the marine phase could also have resulted in the factors favourable for juvenile salmon survival may elevated prevalence (Rhodes et al. 2006, 2011). have allowed more infected juvenile salmon to Compared to subyearling Chinook salmon, year- survive in 2000 despite the high prevalence. lings had a slightly higher prevalence of R. salmon- inarum infection (32% vs. 26% of subyearling, by Nanophyetus salmincola nPCR, with all years combined) and more moder- ately infected salmon (9% vs. 4% of subyearlings, Nanophyetus salmincola was the most common by qPCR). Although there were no statistically sig- infection among subyearlings (42% infected in all nificant differences in the growth metrics of R. sal- years combined). The prevalence peaked in the moninarum-infected vs. uninfected subyearling 2001 with 55% of subyearlings infected, but

Ó 2014 John Wiley & Sons Ltd 374 Journal of Fish Diseases 2015, 38, 365–378 T A Sandell et al. Infections affect salmon growth

infection severities were low, particularly in com- Multiple infections parison with those found in steelhead trout, O. mykiss (Walbaum), some of which have been The low prevalence of dual, and especially triple, found with over 1000 metacercariae (Jacobson infections with any of our target pathogens sug- K.C., unpublished data). Infection by N. salmin- gests that multiple infections may adversely affect cola was also frequent in yearling Chinook salmon the survival of juvenile salmon. Dual infections but intensities were generally low and there were were present in 21% of all the infected juveniles no significant differences in the weight residuals analysed; only 3% had triple infections. Dual or plasma IGF-1 levels of infected vs. uninfected infection by R. salmoninarum and skin metacerca- yearlings. However, the negative binomial regres- riae was extremely uncommon (0.85%), although sion analysis showed that the weights of summer the low prevalence by skin metacercariae alone yearling Chinook salmon were inversely correlated suggests that this may have been due to lower with the number of N. salmincola metacercariae exposure rates rather than the lethality of the com- (as the main factor, P < 0.0001), indicating an bined infections. More yearling Chinook salmon inverse relationship between infection severity and were uninfected (48%) than subyearlings and fish weight and providing evidence that this para- yearlings also had a significantly lower prevalence site may adversely affect yearling Chinook salmon. of triple infections, but both of these results may Jacobson et al. (2008) reported that infection by be the result of lower parasite exposure because N. salmincola may reduce juvenile coho salmon yearling Chinook salmon had lower prevalence of (O. kisutch) survival, but that subyearling and infection by N. salmincola and skin metacercariae yearling Chinook salmon were not significantly individually. affected, potentially due to lower infection intensi- Multiple infections were not associated with sig- ties. Our current data, however, suggest that nificant reductions in the growth metrics of juve- N. salmincola may also negatively affect yearling nile Chinook salmon in any of the MANOVA Chinook salmon, perhaps as a result of the inclu- comparisons, indicating that infections by more sion of additional samples and/or the use of dif- than one of the pathogens examined here were ferent modelling approaches. not significantly more detrimental to growth than a single infection. This may be due to the low numbers of Chinook salmon with high parasite Skin metacercariae burdens; those with high intensities of infection Infection by skin metacercariae was relatively may not have survived long enough to be cap- uncommon with only 18.6% of juvenile Chinook tured in the nearshore ocean. salmon infected. Subyearling Chinook salmon had Our attempt to detect significant differences in significantly higher prevalence than yearling Chi- infection prevalence or intensities between hatch- nook salmon, but infection intensities were low. ery and unmarked yearling Chinook salmon was Infection by skin metacercariae alone did not impeded by the low catches of unmarked salmon appear to be associated with declines in weight (over 75% of the yearlings analysed in this study residuals or plasma IGF-1, although statistical were marked). The majority of our yearling catch comparisons were hindered by the low prevalence. consisted of fish from the Columbia River, where Note that our method of assessing skin metac- >94% of emigrating Chinook salmon smolts are ercariae infections (enumeration of epidermal cysts of hatchery origin (Weitkamp, Bentley & Litz by visual examination) does not identify the spe- 2012). In addition, while all marked fish are of cific species of trematode present. Work by Lemly hatchery origin and the majority of hatchery fish & Esch (1983) in centrarchid fishes infected with are marked, it was not possible to determine the ‘black spot’ Neascus (Uvulifer ambloplitis) deter- percentage of unmarked fish which were actually mined that up to 30% of the cysts did not contain of hatchery origin. viable metacercariae after 7 months, which may result in an overestimation of infection intensity. Conclusions However, a recent study by Ferguson et al. (2010) with juvenile coho salmon found that all but one Our data show that single or dual infections by of the Neascus-type black spot cysts they examined R. salmoninarum and N. salmincola were associated contained living metacercariae after 7 months. with reductions in the growth metrics of some

Ó 2014 John Wiley & Sons Ltd 375 Journal of Fish Diseases 2015, 38, 365–378 T A Sandell et al. Infections affect salmon growth

groups of juvenile Chinook salmon. Juvenile Chi- of three anonymous reviewers substantially nook salmon are hosts to a much larger suite of improved this manuscript. Funding for this pathogens and parasites than the three infections research was provided by the Bonneville Power that were examined in this study; they also acquire Administration, the Mamie Markham Research marine parasites upon ocean entry. These infec- Award (HSMC, OSU), the John L. Fryer Fellow- tions, whether viral, bacterial, parasitic or fungal – ship (OSU), an Oregon Flyfisher’s Club scholar- either individually or in combination – may affect ship and the Harriet M. Winton Scholarship the growth of juvenile salmonids. Because the Fund (OSU). growth of juvenile Chinook salmon during their first year at sea is critical for survival, particularly Publication History for Upper Columbia River stocks of subyearlings (Miller et al. 2013), this study provides additional Received: 1 November 2013 evidence that pathogens and parasites can play a Revision received: 6 February 2014 Accepted: 9 February 2014 role in regulating Chinook salmon populations in the region. The development of molecular assays This paper was edited and accepted under the Editorship of for the detection of parasites and pathogens makes Professor Ron Roberts. it increasingly feasible to screen fish for a large suite of infections by lowering the time needed to exam- References ine the host and reducing the need to isolate and/or identify multiple species; such studies are needed to Arkoosh M.R., Clemons E., Kagley A., Stafford C., Glass understand the role of multiple infections on the A.C., Jacobson K., Reno P., Myers M.S., Casillas E., Loge F., Johnson L.L. & Collier T.K. (2004) Survey of pathogens growth and survival of infected fish. In addition, in juvenile salmon Oncorhynchus Spp. migrating through more research is needed to understand the factors Pacific northwest estuaries. Journal of Aquatic Health that govern the prevalence and severity of these 16, 186–196. infections, and to investigate regional and temporal Bakke T.A. & Harris P.D. (1998) Diseases and parasites in changes in prevalence and intensity as the ecosys- wild ( salar) populations. Canadian – tems of the Pacific Northwest are altered by large- Journal of Fisheries and Aquatic Sciences 55, 247 266. scale forces (e.g. climate change) and localized Baldwin N.L., Milleman R.E. & Knapp S.E. (1967) Salmon anthropogenic effects such as water pollution and poisoning disease III. Effect of experimental Nanophyetus salmincola infection of fish host. Journal of Parasitology, 53, habitat degradation. 556–564. Balfry S.K., Albright L.J. & Evelyn T.P.T. (1996) Horizontal Acknowledgements transfer of Renibacterium salmoninarum among farmed salmonids via the fecal-oral route. Diseases of Aquatic The authors are grateful to the Captains and Organisms 25,63–69. crews of the vessels used to perform this work, Banks J.L. (1994) Raceway density and water flow as factors including the F/V Frosti, F/V Ocean Harvester, affecting spring Chinook salmon (Oncorhynchus R/V Ricker and the F/V Sea Eagle. Mary Beth tshawytscha) during rearing and after release. Aquaculture 119, 201–217. Rew and Lidia Sandoval ably assisted with the Banner C.R., Long J.J., Fryer J.L. & Rohovec J.S. (1986) laboratory analyses. We are indebted to Cheryl Occurrence of salmonid fish infected with Renibacterium Morgan for assistance with data compilation and salmoninarum in the Pacific Ocean. Journal of Fish Diseases discussions of oceanographic indices, Laurie We- 9, 273–275. itkamp for assistance in compiling hatchery release Beckman B.R., Fairgrieve W., Cooper K.A., Mahnken C.V.W. data and decoding the vast array of acronyms & Beamish R.J. (2004) Evaluation of endocrine indices of used by the agencies involved in salmon manage- growth in individual postsmolt coho salmon. Transactions of – ment and Bob Emmett and Dan Bottom for dis- the American Fisheries Society 133, 1057 1067. cussions on salmon biology and ecology. We also Bennington E. & Pratt I. (1960) The life history of the salmon-poisoning fluke, Nanophyetus salmincola (Chapin). wish to thank Hong Sheng Bi for assistance with Journal of Parasitology 46,91–100. the negative binomial regression modelling, Alex Blake R.W., Kwok P.Y.L. & Chan K.H.S. (2006) Effects of De Robertis for advice on modelling the length- two parasites, solidus () and Bunodera weight data residuals and their interpretation and spp. (), on the escape fast-start performance of the many scientists who went to sea to collect three-spined sticklebacks. Journal of Fish Biology 69, and dissect the fish studied here. The comments 1345–1355.

Ó 2014 John Wiley & Sons Ltd 376 Journal of Fish Diseases 2015, 38, 365–378 T A Sandell et al. Infections affect salmon growth

Bruno D.W. & Munro A.L.S. (1986) Observations on Neascus sp., and Nanophyetus salmincola plus two Myxozoans Renibacterium salmoninarum and the salmonid egg. Diseases (Myxobolus insidiosus and Myxobolus fryeri) in coho salmon of Aquatic Organisms 1,83–87. (Oncorhynchus kisutch). Journal of Parasitology, 96, 340–347. Bush A.O., Lafferty K.D., Lotz J.M. & Shostak A.W. (1997) Finley R.J. & Forrester G.E. (2003) Impact of ectoparasites on Parasitology meets Ecology on its own terms: Margolis et al. the demography of a small reef fish. Marine Ecology Progress revisited. Journal of Parasitology, 83, 575–583. Series 248, 305–309. Butler J.A. & Milleman R.E. (1971) Effect of the salmon Fisher J.P. & Pearcy W.G. (1990) Spacing of scale circuli poisoning trematode, Nanophyetus salmincola, on swimming versus growth-rate in young coho salmon. Bulletin ability of juvenile salmonid fishes. Journal of Parasitology 57, 88, 637–643. – 860 865. Fisher J.P. & Pearcy W.G. (1995) Distribution, migration, Campbell W.B. & Emlen J.M. (1996) Developmental and growth of juvenile Chinook salmon, Oncorhynchus instability analysis of BKD-infected spring Chinook salmon, tshawytscha, off Oregon and Washington. Fishery Bulletin Oncorhynchus tshawytscha, prior to seawater exposure. Oikos 93, 274–289. – 77, 540 548. Gordon D.M. & Rau M.E. (1982) Possible evidence for Cvitanich J.D. (2004) Renibacterium salmoninarum bar forms: mortality induced by the parasite Apatemon gracilis in a characterization, occurrence, and evidence of a host response population of brook sticklebacks (Culaea inconstans). to a R. salmoninarum infection. Journal of Fish Diseases 27, Parasitology 84,41–47. – 193 211. Hoffmann G.L. & Putz R.E. (1965) The black-spot (Uvulifer Daly E.A., Brodeur R.D., Fisher J.P., Weitkamp L.A., Teel ambloplitis: Trematoda: Strigeoidea) of centrarchid fishes. D.J. & Beckman B.R. (2012) Spatial and trophic overlap of Transactions of the American Fisheries Society 94, 143–151. marked and unmarked Columbia River Basin spring Jacobson K.C., Teel D., Van Doornik D.M. & Casillas E. Chinook salmon during early marine residence with (2008) Parasite-associated mortality of juvenile Pacific implications for competition between hatchery and naturally salmon caused by the trematode Nanophyetus salmincola – produced fish. Environmental Biology of Fishes 94, 117 134. during early marine residence. Marine Ecology-Progress Series De Robertis A. & Williams K. (2008) Weight-length 354, 235–244. relationships in fisheries studies: the standard allometric Kent M.L., Traxler G.S., Kieser D., Richard J., Dawe S.C., model should be applied with caution. Transactions of the Shaw R.W., Prosperi-Porta G., Ketchenson J. & Evelyn – American Fisheries Society 137, 707 719. T.P.T. (1998) Survey of salmonid pathogens in ocean- Duffy E.J. & Beauchamp D.A. (2011) Rapid growth in the caught fishes in British Columbia, Canada. Journal of early marine period improves the marine survival of Aquatic Animal Health, 10, 211–219. Chinook salmon (Oncorhynchus tshawytscha) in Puget Sound, Lee K-K., Yang T-I., Liu P-C., Wu J-L. & Hsu Y-L. (1999) Washington. Canadian Journal of Fisheries and Aquatic Dual challenges of the infectious pancreatic necrosis virus – Sciences 68, 232 240. and Vibrio carchariae in the grouper, Epinephelus sp. Virus Ebersole J.L., Wigington P.J., Baker J.P., Cairns M.A., Church Research, 63, 131–134. M.R., Hansen B.P., Miller B.A., Lavigne H.R., Compton J.E. Lemly A.D. & Esch G.W. (1983) Differential survival of & Leibowitz S.G. (2006) Juvenile coho salmon growth and metacercariae of Uvulifer ambloplitis (Hughes, 1972) in survival across stream network seasonal habitats. Transactions juvenile centrarchids. Journal of Parasitology 69, 746–749. of the American Fisheries Society 135, 1681–1697. Lester R.J.G. (1984) A review of methods for estimating Elliott D.G., Pascho R.J., Jackson L.M., Matthews G.M. & mortality due to parasites in wild fish populations. Diseases Harmon J.R. (1997) Renibacterium salmoninarum in spring- of Marine Organisms 37,53–64. summer Chinook salmon smelts at dams on the Columbia and Snake Rivers. Journal of Aquatic Animal Health 9, 114–126. Maule A.G., Rondorf D.W., Beeman J. & Haner P. (1996) Incidence of Renibacterium salmoninarum infections in Elliott D., Applegate L.J., Murray A.L., Purcell M.K. & juvenile hatchery spring Chinook salmon in the Columbia McKibben C.L. (2013) Bench-top validation testing of and Snake Rivers. Journal of Aquatic Animal Health, 8,37–46. selected immunological and molecular Renibacterium McCallum H. & Dobson A. (1995) Detecting disease and salmoninarum diagnostic assays by comparison with parasite threats to endangered species and ecosystems. Trends quantitative bacteriological culture. Journal of Fish Diseases – 36, 779–809. in Ecology & Evolution 10, 190 194. McKibben C.I. & Pascho R.J. (1999) Shedding of Renibacterium Evelyn T.P.T., Prosperi-Porta G. & Ketchenson J.E. (1986) salmoninarum by infected Chinook salmon Oncorhynchus Experimental intra-ovum infection of salmonid eggs with – Renibacterium salmoninarum and vertical transmission of the tshawytscha. Diseases of Aquatic Organisms 38,75 79. pathogen with such eggs despite their treatment with Millemann R.E. & Knapp S.E. (1970) Pathogenicity of the erythromycin. Diseases of Aquatic Organisms 1, 197–202. “salmon poisoning” trematode, Nanophyetus salmincola,to Farrell R., Lloyd M. & Earp B. (1964) Persistence of fish. In: A Symposium on Diseases of Fishes and Shellfishes (ed. by F. Sniesko), pp. 209–217. American Fisheries Society, Neorickettsiae helminthoeca in an endoparasite of the Pacific salmon. Science 145, 162–163. Bethesda, MD. Ferguson J.A., Schreck C., Chitwood R. & Kent M. (2010) Miller J.A., Teel D.J., Baptista A.M. & Morgan C.A. (2013) Disentangling bottom-up and top-down effects on survival Persistence of infection by metacercariae of Apophallus sp.,

Ó 2014 John Wiley & Sons Ltd 377 Journal of Fish Diseases 2015, 38, 365–378 T A Sandell et al. Infections affect salmon growth

during early ocean residence in a population of Chinook ecosystem predictors of a bacterial infection in juvenile salmon (Oncorhynchus tshawytscha). Canadian Journal of Chinook salmon. Marine Ecology Progress Series 432, Fisheries and Aquatic Sciences 70, 617–629. 161–172. Nance S.L., Riederer M., Zubkowski T., Trudel M. & Rhodes Sandell T.A. & Jacobson K.C. (2011) Comparison and L.D. (2010) Interpreting dual ELISA and qPCR data for evaluation of Renibacterium salmoninarum quantitative PCR bacterial kidney disease of salmonids. Diseases of Aquatic diagnostic assays using field samples of Chinook and coho Organisms 91, 113–119. salmon. Diseases of Aquatic Organisms 93, 129–139. Newcomb T.W., Snell G.A. & Waknitz F.W. (1991) Sanders J.E., Long J.J., Arakawa C.K., Bartholomew J.L. & Survival of coho salmon smolts in salt water following Rohovec J.S. (1992) Prevalence of Renibacterium exposure to cercariae of the salmon poisoning fluke, salmoninarum among downstream-migrating salmonids in the Nanophyetus salmincola. In: Northwest Fish Culture Columbia River. Journal of Aquatic Animal Health, 4,72–75. Conference, Redding, CA. Seppanen E., Kuukka H., Voutilainen A., Huuskonen H. & Niemi D.R. & Macy R.W. (1974) The life cycle and infectivity Peuhkuri N. (2009) Metabolic depression and spleen and to man of Apophallus donicus in Oregon. Proceedings of the liver enlargement in juvenile Arctic char Salvelinus alpinus Helminthological Society of Washington 41, 223–229. exposed to chronic parasite infection. Journal of Fish Biology – Olson P.D., Cribb T.H., Tkach V.V., Bray R.A. & Littlewood 74, 553 561. D.T.J. (2003) Phylogeny and classification of the Digenea Shimizu M., Swanson P., Fukada H., Hara A. & Dickhoff (Platyhelminthes: Trematoda). International Journal for W.W. (2000) Comparison of extraction methods and assay Parasitology 33, 733–755. validation for salmon insulin-like growth factor-I using Pascho R.J., Elliot D.G. & Streufert J.M. (1991) Brood stock commercially available components. General and – segregation of spring Chinook salmon Oncorhynchus Comparative Endocrinology 119,26 36. tshawytscha by use of the ELISA assay and the fluorescent Sindermann C.J. (1987) Effects of parasites on fish antibody technique (FAT) affects the prevalence and levels populations: practical considerations. International Journal of Renibacterium salmoninarum infection in progeny. Diseases for Parasitology 17, 371–382. – of Aquatic Organisms 12,25 40. Sogard S.M. (1997) Size-selected mortality in the juvenile stage Pascho R.J., Elliot D.G. & Achord S. (1993) Monitoring of of Teleost fishes: a review. Bulletin of Marine Science 60, the in-river migration of smolts from two groups of spring 1129–1157. Chinook salmon, Oncorhynchus tshawytscha (Walbaum), with Sokal R.R. & Rohlf F.J. (1987) Introduction to Biostatistics. different profiles of Renibacterium salmoninarum infection. W.H. Freeman and Company, New York. Aquaculture and Fisheries Management 24, 163–169. Statpoint Technologies, Inc. (2009) StatGraphics Centurion Pascho R.J., Goodrich T.D. & McKibben C.L. (1997) XVI. Statpoint Technologies, Inc., Warrenton, VA. Evaluation by enzyme-linked immunosorbent assay (ELISA) of Renibacterium salmoninarum bacterins affected by Turgut E., Thompson K.D., Ellis A.E. & Adams A. (2008) Persistence of Renibacterium salmoninarum in persistence of bacterial antigens. Journal of Aquatic Animal Health 9,99–107. experimentally infected (Oncorhynchus mykiss). Bulletin of the European Association of Fish Pascho R.J., Chase D.M. & McKibben C.I. (1998) Pathologists 28,66–76. Comparison of the membrane-filtration fluorescent antibody test, the enzyme linked immunosorbent assay, and the Van Doornik D.M., Teel D.J., Kuligowski D.R., Morgan C.A. polymerase chain reaction to detect Renibacterium & Casillas E. (2007) Genetic analyses provide insight into salmoninarum in salmonid ovarian fluid. Journal of the early ocean stock distribution and survival of juvenile coho salmon off the coasts of Washington and Oregon. Veterinary Diagnostic Investigation 10,60–66. North American Journal of Fisheries Management 27, Patterson K.R. (1996) Modeling the impact of disease-induced 220–237. mortality in an exploited population: the outbreak of the fungal parasite Icthyophonus hoferi in the North Sea herring Van Gaest A.L., Dietrich J.P., Thompson D.E., Boylen D.A., (Clupea harengus). Canadian Journal of Fisheries and Aquatic Strickland S.A., Collier T.K., Loge F.J. & Arkoosh M.R. Sciences 53, 2870–2887. (2011) Survey of pathogens in hatchery Chinook salmon with different out-migration histories through the Snake and Peterson W.T., Morgan C.A., Fisher J.P. & Casillas E. (2010) Columbia Rivers. Journal of Aquatic Animal Health, 23, Ocean distribution and habitat associations of yearling coho 62–77. (Oncorhynchus kisutch) and Chinook (O. tshawytscha) salmon Weitkamp L.A., Bentley P.J. & Litz M.N.C. (2012) Seasonal in the northern California Current. Fisheries Oceanography 19, 508–525. and interannual variation in juvenile salmonids and associated fish assemblage in open waters of the lower Columbia River Rhodes L.D., Durkin C., Nance S.L. & Rice C.A. (2006) estuary. Fishery Bulletin 110, 426–450. Prevalence and analysis of Renibacterium salmoninarum infection among juvenile Chinook salmon Oncorhynchus Zabel R.W. & Achord S. (2004) Relating size of juveniles to tshawytscha in North Puget Sound. Diseases of Aquatic survival within and among populations of Chinook salmon. Ecology 85, 795–806. Organisms 71, 179–190. Rhodes L.D., Rice C.A., Greene C.M., Teel D.J., Nance S.L., Zar J.H. (2010) Biostatistical Analysis. Pearson Prentice Hall, Moran P., Durkin C.A. & Gezhegne S.B. (2011) Nearshore Upper Saddle River, NJ.

Ó 2014 John Wiley & Sons Ltd 378