Mortality of Coho ( kisutch) Associated with Burdens of Multiple Parasite Species

International Journal for Parasitology 2011

Ferguson, Jayde* *Corresponding Author [email protected] Alaska Department of Fish and Game, Commercial Division, Fish Pathology Laboratory, Anchorage, Alaska Koketsu, Wataru Ecology and Civil Engineering Research Institute, Fukushima pref. Japan Ninomiya, Ikuo Laboratory of Forest Resource Biology, Ehime pref. Japan Rossignol, Philippe A. Department of Fisheries and Wildlive, Oregon State University Jacobson, Kym C. NOAA Fisheries, Northwest Fisheries Science Center, Newport, Oregon Kent, Michael L. Department of Microbiology, Oregon State University

This is the authors’ post-peer review version of the final article. The final published version is copyrighted by Elsevier and can be found at: http://www.elsevier.com/wps/find/journaldescription.cws_home/353/description#description

1 1 Mortality of (Oncorhynchus kisutch) associated with burdens of multiple parasite

2 species

3 Jayde A. Fergusona.1,*, Wataru Koketsub, Ikuo Ninomiyac, Philippe A. Rossignold, Kym C. Jacobsone,

4 Michael L. Kenta

5

6 a Department of Microbiology, Oregon State University, 220 Nash Hall, Corvallis, Oregon 97331 USA

7 b Ecology and Civil Engineering Research Institute, Ishibatake275, Miharu-cho, Fukushima pref. 963-

8 7722 Japan

9 c Laboratory of Forest Resource Biology, Faculty of Agriculture, Ehime University, Tarumi 3-5-7,

10 Matsuyama, Ehime pref. 790-8566 Japan

11 d Department of Fisheries and Wildlife, Oregon State University, Nash Hall, Corvallis, Oregon 97331

12 USA

13 e NOAA Fisheries, Northwest Fisheries Science Center, Hatfield Marine Science Center, 2030 South

14 Marine Science Drive, Newport, Oregon 97365 USA

15 1 Alaska Department of Fish and Game, Commercial Fisheries Division, Fish Pathology Laboratory,

16 333 Raspberry Rd., Anchorage, Alaska 99518 USA.

17

18 *Corresponding author. Alaska Department of Fish and Game, Commercial Fisheries Division, Fish

19 Pathology Laboratory, 333 Raspberry Rd., Anchorage, Alaska 99518 USA.

20 Tel.: +1 907 267 2364; fax: +1 907 267 2194.

21 E-mail address: [email protected] (J.A. Ferguson).

22 Note: Supplementary data associated with this article.

23 2

24 ABSTRACT

25 Multiple analytical techniques were used to evaluate the impact of multiple parasite species on

26 the mortality of threatened juvenile coho salmon (Oncorhynchus kisutch) from the West Fork Smith

27 River, Oregon, USA. We also proposed a novel parsimonious mathematical representation of

28 macroparasite distribution, congestion rate, which i) is easier to use than traditional models, and ii) is

29 based on Malthusian parameters rather than probability theory. Heavy infections of Myxobolus

30 insidiosus (Myxozoa) and metacercariae of Nanophyetus salmincola and Apophallus sp. occurred in

31 parr (subyearlings) from the lower mainstem of this river collected in 2007 and 2008. Smolts

32 (yearlings) collected in 2006 - 2009 always harbored fewer Apophallus sp. with host mortality

33 recognized as a function of intensity for this parasite. Mean intensity of Apophallus sp. in lower

34 mainstem parr was 753 per fish in 2007 and 856 per fish in 2008, while parr from the tributaries had a

35 mean of only 37 or 13 parasites per fish, respectively. Mean intensity of this parasite in smolts ranged

36 between 47 - 251 parasites per fish. Over-dispersion (variance to mean ratios) of Apophallus sp. was

37 always lower in smolts compared with all parr combined or lower mainstem parr. Retrospective

38 analysis based on smolt data using both the traditional negative binomial truncation technique and our

39 proposed congestion rate model showed identical results. The estimated threshold level for mortality

40 involving Apophallus sp. was at 400 - 500 parasites per fish using both analytical methods. Unique to

41 this study, we documented the actual existence of these heavy infections prior to the predicted

42 mortality. Most of the lower mainstem parr (approximately 75%) had infections above this level.

43 Heavy infections of Apophallus sp. metacercariae may be an important contributing factor to the high

44 over-wintering mortality previously reported for these fish that grow and develop in this section of the

45 river. Analyses using the same methods for M. insidiosus and N. salmincola generally pointed to

46 minimal parasite-associated mortality. 3 47 Keywords: Multispecies ; Negative Binomial distribution; Truncation; Digenea; Myxozoa

48 4

49 1. Introduction

50 Parasites may be a significant source of mortality in wild fish populations (Dobson and May,

51 1987; Sindermann, 1987; Adlard and Lester, 1994; Bakke and Harris, 1998). Assessing the impact of

52 parasitism on wild populations presents several significant challenges. Specifically for macroparasites,

53 impacts are a function of parasite load rather than prevalence alone (Brass, 1958; Crofton, 1971; May

54 and Anderson, 1979; Dobson, 1988; Burgett et al., 1990; Scott and Smith, 1994; Shaw and Dobson,

55 1995; Galvani, 2003; Holt et al., 2003). Furthermore, estimates of effects are complicated by the

56 aggregated distribution of parasites, as often most hosts harbor few or no parasites (Smith, 1994;

57 Galvani 2003). A corollary is that heavy infections occur in few hosts, many of which may have died

58 and cannot be sampled. Consequently, prevalence of infection yields at best only a weak assessment of

59 macroparasite impact (Smith, 1994) and may be misleading (Dobson and Hudson, 1986).

60 Lester (1984) reviewed the common methods used for estimating parasite-associated mortality

61 in wild fishes, many of which require temporal observations of the same host populations. There are

62 practical limitations involved in the study of hosts in an aquatic environment. For example, fish are

63 often inaccessible and the most impacted fish are likely to die without detection (Bakke and Harris,

64 1998). Nevertheless, there have been several studies reporting that wild fish with higher intensities of

65 trematode metacercariae have a higher mortality rate (Gordon and Rau, 1982; Lemly and Esch, 1984;

66 Lafferty and Morris, 1996; Jacobson et al., 2008).

67 Coho salmon (Oncorhynchus kisutch) from coastal Oregon, USA are listed as threatened under

68 the Endangered Species Act (US National Research Council, 1996). We previously reported on high

69 loads of the digeneans Apophallus sp. (Heterophyidae) and Nanophyetus salmincola (Nanophyetidae),

70 and the myxozoan Myxobolus insidiosus in parr (resident stage subyearlings) from the lower reaches of

71 the West Fork Smith River, Oregon, USA (Rodnick et al., 2008). However, the older smolts (out- 5 72 migrating yearlings) collected downstream in this river had low burdens of these same parasites

73 (Ferguson et al., in press a). Parr from the lower reaches of the river also have greater than expected

74 over-wintering mortality based on fisheries prediction models (Ebersole et al., 2006, 2009). Therefore,

75 we hypothesized that parasites may have a role in over-wintering survival of the threatened coho

76 salmon from this river.

77 Studying host-parasite systems in wild salmon presents two specific challenges: i) many

78 populations are listed as threatened, making it difficult to obtain large samples, and ii) parr grow and

79 develop typically as separate, multiple, sub-populations and migrate to the ocean as a randomly mixed

80 population of smolts, making temporal observations of the same cohort problematic. Hence, while

81 numerous parasites have been described from Pacific salmon species, (Love and Moser, 1983;

82 McDonald and Margolis, 1995; Hoffman, 1999), few studies have evaluated parasite-associated

83 mortality involving these infections in these fish in the wild (e.g., Henricson, 1977; Halvorsen and

84 Andersen, 1984; Vincent, 1996; Kocan et al., 2004; Krkosek et al., 2006; Jacobson et al., 2008).

85 An alternative technique to tracking infections in cohorts over time is to conduct a retrospective

86 analysis by predicting the parasite distribution in host populations based on observed data from lightly

87 infected fish, as originally proposed by Crofton (1971). He demonstrated how analyzing the negative

88 binomial distribution can estimate mortality associated with macroparasitism. Regarding macroparasite

89 infections in wild , there are usually fewer heavily infected hosts than would be predicted. An

90 explanation for this phenomenon is that heavily infected hosts are more predisposed to mortality.

91 Crofton’s technique has become widely accepted and is used extensively in theoretical and empirical

92 models (e.g., May and Anderson, 1979; Lanciani and Boyett, 1980; Anderson and May, 1982; Dobson,

93 1988; Royce and Rossignol, 1990; Scott and Smith, 1994; Galvani, 2003). Crofton’s techniques rely on

94 approximating the distribution. Although the statistical assay has proven reliable as a theory (Dobson 6 95 and Carper, 1992), it is descriptive, having at best indirect biological interpretation and it is also

96 somewhat arduous to perform.

97 Here, we evaluated the impacts of parasites on coho salmon from parr to smolt stage from the

98 West Fork Smith river by comparing parasite burdens of different age classes (parr and smolt) using

99 four analytical techniques: i) comparison of parasite prevalence and intensity between life stages, ii)

100 comparison of parasite over-dispersion (variance to mean ratios) between life stages, iii) a

101 retrospective analysis of smolt data using the negative binomial truncation technique developed by

102 Crofton (1971), and iv) our new parsimonious mathematical representation of macroparasite

103 distribution, which was first reported by one of the current authors (Koketsu, 2004 M.Sc. thesis,

104 Environmental correlates of parasitism in introduced threespine stickleback (Gasterosteus aculeatus)

105 in the Upper Deschutes River Basin, Oregon, Oregon State University, USA). This model is based on

106 the standard growth model that applies to all life rather than probability theory that is used in current

107 models. No such model has been proposed since the probability-based negative binomial model of

108 Crofton (1971). Based on all four analytical techniques, we conclude that parasites, especially

109 Apophallus sp., have an impact on coho salmon freshwater over-winter survival.

110 2. Materials and methods

111 2.1. Sampling fish

112 Coho salmon parr were collected by electrofishing in September, 2007 and October, 2008 from

113 two general locations of the West Fork Smith River: the lower mainstem and the tributaries (see Fig. 1

114 for sample sizes and exact locations). These two sections were chosen as they represent distinctly

115 different habitats. The lower mainstem has been subjected to extensive logging practices that have

116 simplified substrate and removed riparian vegetation, which has caused increased winter flow rates and

117 high summer temperatures (Ebersole et al., 2006). In contrast, the tributaries of this system are much 7 118 cooler during the summer and flow more slowly during the winter. An additional difference between

119 these two sections of the river is parasite burden in the coho salmon, as lower mainstem parr harbor

120 much higher infections than those from the tributaries (Rodnick et al., 2006; Ferguson et al., 2010).

121 Wild coho salmon smolts were captured in April 2007 - 2010 (corresponding to brood years 2006 -

122 2009, respectively) in a rotary screw trap downstream from the parr collecting sites (Fig. 1) and killed

123 immediately for parasite evaluation. Data from many of these sampled fish have been previously used

124 in our earlier studies involving different types of analyses (see Supplementary Table S1). Formal

125 ethics approval was given by Oregon State University’s (OSU’s) Institutional Animal Care and

126 Use Committee (IACUC) for the work with all animals in the present and past studies.

127 2.2. Parasite evaluation

128 We were particularly interested in muscle parasites due to heavy infections reported in previous

129 studies in coho salmon parr from this river (Rodnick et al., 2008; Ferguson et al., 2010). One fillet was

130 evaluated for each fish. Tissue squashes were prepared by squashing fillets between two 15- x 30 cm

131 plexiglass plates and parasites were identified and enumerated, which were then multiplied by two in

132 order to represent the number of parasites per fish. The posterior half of the kidney was similarly

133 evaluated for metacercariae of N. salmincola, but counts were not multiplied to estimate the entire

134 kidney because this parasite targets the posterior kidney via the renal portal system (Baldwin et al.,

135 1967). Methods for identification of the metacercariae and myxozoans are described in detail in

136 Ferguson et al. (2010), which included excystation of metacercariae. Adult worms of Apophallus sp.

137 were also obtained from chicks that were fed metacercariae from coho salmon from this river. Chicks

138 were cared for and maintained at OSU’s Laboratory Animal Resources Center and formal animal

139 ethics approval for this work was given by OSU’s IACUC. These worms were consistent with

140 Apophallus but did not correspond with any described species (J.A. Ferguson, unpublished data). 8 141 Hence, we denote this worm as Apophallus sp. Prevalence (number of infected animals per total

142 animals examined), mean intensity (average number of parasites per infected animal examined), and

143 mean abundance (average number of parasites per animal examined, including uninfected animals) of

144 infections are reported in accordance with the definitions provided by Bush et al. (1997).

145 2.3. Inferring parasite-associated mortality

146 2.3.1. Comparison of parasite burden

147 Mean intensities of parasites in parr were compared with those of smolts with a non-parametric

148 bootstrap t-test with 100,000 replications, as data were not normally distributed. Fisher’s exact tests

149 were used to test differences in prevalence of parasites between parr and smolts. Data from parr from

150 both river locations were pooled to represent total parr of the river to compare with out-migrating

151 smolts. Samples of parr and smolts for fish from brood years 2007 and 2008 were matched for this

152 analysis. Data from separate parr sub-populations were also compared with smolt data to determine

153 whether the conclusions would change using this approach. All statistical procedures were performed

154 with Quantitative Parasitology (Rózsa et al., 2000), significance was set at P < 0.05 and P-values are

155 two-tailed.

156 2.3.2. Comparison of parasite overdispersion

157 Over-dispersion (variance to mean abundance ratios) of each parasite species was calculated for

158 comparison between the parr (data pooled from both river locations) and smolt coho salmon life stages.

159 Samples of parr and smolts for fish from brood years 2007 and 2008 were matched for this

160 comparison.

161 2.3.3. Crofton’s truncation model of the negative binomial distribution 9 162 Crofton’s model has been widely accepted, so only a brief overview of this technique is

163 provided (see Scott and Smith, 1994). The truncation technique estimates an overall expected host

164 distribution from the low frequency classes, where lethal effects are less significant. The truncated

165 curve will fit the negative binomial distribution better than the observed curve because the observed

166 data will be missing hosts from the high parasite load class owing to parasite-induced mortality, and

167 the difference in fitness is considered to be the parasite-induced host mortality (see Royce and

168 Rossignol, 1990). The analysis of Crofton’s truncation of the negative binomial distribution was

169 performed using the DOS-based software, BASICA (Ludwig, 1988). Briefly, the observed number of

170 hosts was entered for each frequency class of zero to up to 10, which corresponded to different

171 categorized infections (see below). An estimate of k, which is a measure of aggregation, was obtained

172 through an iterative process of balancing both sides of an equation in the BASICA program, then the

173 probability of finding a given number of individuals (i.e., parasites) in a sampling unit (i.e., host) was

174 computed using the probability-based negative binomial equation. Instructions for using the BASICA

175 program, including all equations and DOS codes, can be found in Ludwig (1988). A +1 transformation

176 on all raw data was performed so that categories did not contain zero hosts. Data from smolts of brood

177 years 2006-2009 were used in this analysis, except for data of N. salmincola in smolts from brood year

178 2006 because kidney data were not available. Infections were categorized into groups of 100

179 metacercariae per fish for Apophallus sp., 75 metacercariae per fish for N. salmincola and 200

180 pseudocysts per fish for M. insidiosus to best fit the negative binomial distribution. This resulted in

181 approximately 10 groups (categories of intensities), which was consistent with other studies using this

182 analysis that have high intensity data (Halvorsen and Andersen, 1984; Lemly and Esch, 1984; Kang et

183 al., 1985). The size of categories was chosen for two reasons: i) groups with a smaller range resulted in

184 too few, and often no, data in most of the individual categories, and ii) groups with too large a range 10 185 resulted in too few categories to properly conduct the analysis. The effects of using differently sized

186 categories with each parasite were evaluated and no real difference was found (data not shown).

187 The algorithm used to truncate the negative binomial distribution first estimates the maximum

188 likelihood value of k (aggregation coefficient of the negative binomial distribution). Then, the expected

189 distribution from zero to one of the frequency classes is estimated and compared with the observed

190 distribution using a Chi-square test. If there is density-dependent mortality associated with parasitism,

191 a sudden change in fit will likely occur at some point. The equations for this algorithm are standard

192 and may be found in the literature (see Box 1 in Royce and Rossignol, 1990).

193 2.3.4. Congestion rate model

194 Koketsu (2004 M.Sc. thesis, Environmental correlates of parasitism in introduced threespine

195 stickleback (Gasterosteus aculeatus) in the Upper Deschutes River Basin, Oregon, Oregon State

196 University, USA) developed and validated an alternative model that may be more biologically

197 meaningful because it is analogous to a population growth model. The relationship between parasite

198 load and host frequency is expressed with the exponential function:

199 Y = ae - bx (1)

200 Where, Y = frequency of hosts (observed plots), x = number of parasites per host, a = constant

201 parameter, b = constant parameter (named congestion rate), e = Euler’s constant. The congestion rate

202 b, (the slope of the linearized equation, which is negative) is analogous to the intrinsic rate of growth

203 of a population.

204 Equation (1) was transformed into a differential equation,

dY 205 −= bY (2) dx

206 then divided on both sides by dt and solved for b. 11 1 dY dx 207 − = b (3) Y dt dt

1 dY 208 Equation (3) represents the per capita rate of change in parasitized hosts, − over the rate of Y dt

dx dx 1 dY 209 change in parasite load class, . Given that b is a constant, if increases, then − decreases, dt dt Y dt

210 and vice versa. Furthermore, when both sides are multiplied by x:

1 dY 1 dx 211 − = bx (4) Y dt x dt

1 dY 1 dx 212 Equation (4) represents the ratio of − to in proportion to x, that is, the negative per Y dt x dt

213 capita rate of host frequency change over the per capita rate of parasite load class change. In other

1 dY 1 dx 214 words, when x is large, the influence or regulation of the parasite load class on both − and Y dt x dt

215 becomes strong.

216 The congestion rate formalizes the relationship between parasite load and host distribution in an

217 equation that is analogous to one found in population dynamics. Congestion rate is itself analogous to

218 the intrinsic rate of growth (frequently represented as r) of that equation. We suggest that congestion

219 may be a useful index of regulation between host and macroparasite. The standard index used in

220 studies of macroparasitism has been the aggregation parameter of the discrete negative binomial

221 distribution, k, which is a statistical descriptor difficult to relate to population dynamics. In addition to

222 a greater ease of use, our alternative interpretation can be related to standard life table equations (see

223 Carey, 1993). Macroparasitism involves the interdependence of multiple hosts and different parasite

224 stages, resulting in great complexity and is often counterintuitive. The interpretation of the constant b

225 is therefore more complex than found in a single species population but does provide insight into the

226 dynamics of macroparasitism.

227 Fitting of this model to observed data and comparison of differences from predicted values were

228 performed in Excel (Microsoft Office Corp., 2003) and used data on sampled smolts from brood years 12 229 2006-2009. Categories of parasites were used as above for Crofton’s method. The basic

230 methodology for this model is to graph the observed data and add an exponential trendline (and

231 displayed equation), using the Excel chart wizard. The exponential equation (analogous to equation

232 (1)) can then be used to generate the predicted values by substituting the number of parasites per host

233 (category) into the x variable and then solving for Y (see equation (1)).

234 3. Results

235 3.1. Prevalence

236 Although prevalence may be an incomplete metric of infection for macroparasites, we

237 compared prevalences to provide additional information on parasite burden in host populations. All

238 parr originating from the lower mainstem were infected with metacercariae of Apophallus sp., whereas

239 prevalence was less than 40% in parr from the tributaries in both 2007 and 2008 (Table 1). Almost all

240 smolts were infected with this parasite each year of the study (Table 1). Prevalence was approximately

241 60% when samples from parr from the lower mainstem and tributaries were pooled. Using these

242 pooled data, prevalence in the parr was significantly less (P < 0.01; Fisher’s exact test) than in smolts

243 (Table 1). For N. salmincola, essentially all parr and smolts were infected (Table 1). The prevalence of

244 M. insidiosus ranged from 47% to 97% with no significant differences between fish life stages, except

245 between lower mainstem and tributary parr from brood year 2008 (Table 1).

246 3.2. Mean intensity

247 Mean intensities of infection in our samples were compared to evaluate differences in burden.

248 The mean intensity of Apophallus sp. in parr was always higher in fish originating from the lower

249 mainstem (Table 1). For each year that combined parr to smolt data were compared, the mean intensity

250 of Apophallus sp. in smolts was approximately three to seven times lower than that of parr (P < 0.01; 13 251 bootstrap t-test). Although we observed some overlap in the range of infection between smolt and

252 parr samples, the few heavy infections in smolts did not even closely approximate those of heavily

253 infected lower mainstem parr (Fig. 2). The mean intensity of N. salmincola differed (P < 0.01;

254 bootstrap t-test) in parr from the lower mainstem and tributaries for brood year 2007, but not 2008

255 (Table 1). Smolts had a higher mean intensity of N. salmincola than parr, which was statistically

256 significant only for brood year 2007 (P < 0.02; bootstrap t-test). The mean intensity of M. insidiosus

257 was higher in parr from the lower mainstem than parr from the tributaries, which was significant only

258 for brood year 2008 (P < 0.03; bootstrap t-test). Smolts had the same infection level as combined parr

259 for both years (Table 1). Overall, the most prominent difference in parasite burden was with

260 Apophallus sp., as smolts always had a lower mean intensity compared with parr, with lower mainstem

261 fish harboring the most metacercariae.

262 3.3. Variance to mean ratio

263 Over-dispersion (variance to mean abundance ratio) was evaluated as another indicator of

264 parasite-associated mortality. The variance to mean abundance ratio of Apophallus sp. in parr was

265 approximately three times higher than that of smolts of the same brood year for both years (Table 1).

266 The over-dispersion was influenced by infections in the lower mainstem parr because they had high

267 mean intensities compared with parr from the tributaries (Table 1), which increased the variance of the

268 data set. Thus the higher over-dispersion of Apophallus sp. in the lower mainstem parr subset

269 compared with smolts was less pronounced. For the second year, the tributary parr had such a low

270 prevalence and intensity of Apophallus sp. infections that it resulted in a low variance to mean

271 abundance ratio. With N. salmincola, the variance to mean abundance ratio was lower in parr

272 compared with smolts. Lastly, the variance to mean abundance ratio of M. insidiosus was similar 14 273 between parr and smolts (Table 1). Of all three parasites, the most dramatic difference between parr

274 and smolts was with Apophallus sp.

275 3.4. Truncation of negative binomial model

276 A retrospective technique to was applied to compare observed and expected parasite distributions

277 and thus estimate a ‘threshold’ of parasite-associated mortality. For this analysis, data was included for

278 all available smolts (i.e., from brood years 2006-2009), which increased our sample size and was

279 permissible because we were not making direct comparisons with parr of the same year classes.

280 Analysis of data from all three parasites predicted that the parasite distribution in smolts was truncated

281 (Table 2; Fig. 3). The truncation was calculated to occur within the first few infection categories (Table

282 2; Fig. 3A, C, E), particularly for N. salmincola and M. insidiosus. However the truncation point was

283 actually more towards the tail end of the distributions in that the majority of fish had an abundance of

284 infection below the truncation point (i.e., to the left; see Fig. 3A, C, E). Regarding parasite

285 aggregation, N. salmincola was much less aggregated (k value above 1.0) than Apophallus sp. and M.

286 insidiosus (k values of approximately 0.2) (Table 2).

287 Analysis of our data fitted to the negative binomial distribution indicated that the threshold for

288 parasite-associated mortality began at approximately 400 and 150 parasites per fish for metacercariae

289 of Apophallus sp. and N. salmincola (Fig. 3A, C), respectively, and 200 parasites per fish for

290 pseudocysts of M. insidiosus (Fig. 3E). Using these threshold data, we determined the percentage of

291 infected fish above this threshold. For Apophallus sp., approximately 10% of smolts had burdens

292 above this level. Data for parr were available, which represented hosts sampled prior to any potential

293 parasite-associated mortality in general. This allowed us to assign estimates to the proportion of

294 heavily infected fish linked to mortality by parasitism. Approximately 75% of lower mainstem parr

295 and only 1% of parr from the tributaries had infections of Apophallus sp. greater than this value. In 15 296 contrast, approximately 31% of smolts had N. salmincola above this threshold, whereas only 14%

297 and 6% of lower mainstem and tributary parr, respectively, were above this value. With M. insidiosus,

298 approximately 18% of smolts were above the threshold and 22% and 10% of lower mainstem and

299 tributary parr, respectively, were above this level. Based on these analyses, Apophallus sp. was the

300 parasite most closely associated with host over-winter mortality, as essentially all of the lower

301 mainstem parr had infection levels above the predicted threshold and those high infections were

302 consistently not detected in smolt samples over a 4 year period.

303 3.5. Truncation of congestion rate model

304 A similar retrospective technique has been included to augment current results and to provide a

305 simpler analysis that may be more biologically meaningful. The newly proposed congestion rate model

306 provided identical results to the negative binomial method regarding the location of the predicated

307 location for truncation (Table 2; Fig. 3).

308 4. Discussion

309 Studying the impacts of parasites in wild populations is difficult due to the nature of chronic

310 infections being sublethal (McCallum and Dobson, 1995). Such studies involving fish populations are

311 often further complicated due to their inaccessibility, especially for migratory species such as salmon

312 (Bakke and Harris, 1998). Prior knowledge of the West Fork Smith River being separated into areas

313 with heavily and lightly infected rearing parr (Rodnick et al., 2008; Ferguson et al., 2010) was a

314 benefit to the current study. Multiple analytical techniques were used to evaluate parasite-associated

315 mortality and results from all of the different analyses indicated that parasites were associated with

316 over-winter mortality of juvenile coho salmon in this river. This was particularly evident with

317 Apophallus sp. infections in lower mainstem fish. The dramatic difference in intensity and over- 16 318 dispersion of Apophallus sp. between parr and smolts indicated that the heavily infected lower

319 mainstem fish did not survive to the smolt stage in either year.

320 The use of a retrospective analytical technique with migratory species, such as salmon, is

321 particularly useful. This technique requires only one sampling time point and thus avoids the inherent

322 difficulties with sampling the same population of salmon throughout different life stages. However, to

323 our knowledge there are few available reports using this approach with salmonid fishes i.e., Arctic

324 char (Salvelinus alpinus) infected with spp. (Henricson, 1977; Halvorsen and

325 Andersen, 1984). The predicted threshold of parasite-associated mortality for Apophallus sp. was

326 approximately 400 metacercariae per fish using Crofton’s method (Crofton, 1971). It was remarkable

327 that our new congestion model provided identical results. These findings, together with our previous

328 comparison using the same data set from Crofton (1971) (Koketsu, 2004 M.Sc. thesis, Environmental

329 correlates of parasitism in introduced threespine stickleback (Gasterosteus aculeatus) in the Upper

330 Deschutes River Basin, Oregon, Oregon State University, USA), further indicate that similar results

331 were obtained with both models.

332 Most studies using retrospective analyses, such as Crofton’s technique, only provide a

333 prediction of the number of heavily infected animals that should have theoretically existed before death

334 (Lanciani and Boyett, 1980; Burgett et al., 1990; Royce and Rossignol, 1990). Thus a unique aspect of

335 our study is that by sampling lower mainstem parr, we actually documented the existence of heavily

336 infected hosts above the threshold predicted with smolt data. Natural, presumably sustainable, wild

337 populations typically have only a small percentage of hosts that occur above the truncation point and

338 hence parasite-associated mortality seldom threatens an entire population. Most of the lower mainstem

339 fish, however, were above this threshold level. Indeed, some mortality may have even occurred earlier

340 for these fish, based on their relatively low variance to mean abundance ratios of Apophallus sp., and it 17 341 would be interesting to evaluate fry (early summer) fish compared with parr (late summer/autumn)

342 fish from this system.

343 The threshold for parasite-associated mortality indicates the level of infection where mortality

344 begins to occur, but due to the dynamic nature of the process not every fish will die from infection at

345 the tested time point. Typically this occurs towards the tail end of the distribution, which was not the

346 case with our analysis using categorized data. However, Kang et al. (1985) obtained similar results

347 using categories of 50 metacercariae per fish. Nevertheless, with almost 75% of the lower mainstem

348 parr having heavy Apophallus sp. infections above this level, we can conclude that most of these fish

349 die during the late fall or winter before smoltification. This is reflected by the absence of heavy

350 infections in smolt stage fish. In comparison, there were very few parr from the tributaries with

351 infections above this level (only one fish from 176 sampled), suggesting that fish surviving to the

352 smolt stage are over-represented by parr originating from the tributaries. Admittedly, combining data

353 from two separate parr subpopulations has an inherent problem i.e. assuming this population provides

354 the actual representation of parr to compare with the mixed smolt population. However, we emphasize

355 that this is a general limitation of evaluating parasite-related mortality in salmonids, which is

356 particularly problematic for endangered or threatened species for which scientific samples are very

357 limited.

358 Myxobolus insidiosus was also indicated to be associated with over-winter mortality but much

359 less so than Apophallus sp. Anderson and May (1979) defined microparasites as organisms that

360 reproduce directly in the invaded individual host, usually induce long lasting immunity to reinfection

361 and typically cause transient infections (e.g., bacteria and viruses). In contrast, macroparasites do not

362 multiply within a host (the load is thus determined by invasion events), generally elicit a short-term

363 immune response and cause chronic infections associated with a long life expectancy of the parasite

364 (e.g., helminths). Myxobolus insidiosus is a metazoan parasite but it has features of a microparasite in 18 365 that asexual reproduction occurs in the fish host. However, this species is also similar to a

366 macroparasite in that plasmodial pseudocysts, which grow in size due to the proliferation of a single

367 progenitor cell, do not increase in number within the host. Also, with histozoic species such as M.

368 insidiosus, infection persists throughout the juvenile life stages of the fish host (Ferguson et al., 2010).

369 Therefore, the models used to observe a decrease in parasite burden over time to infer parasite-

370 associated mortality were appropriate for the current study. Mean intensity decreased significantly

371 between parr and smolts and over-dispersion was lower in the latter. Truncation models indicated that

372 parasite-associated mortality occurred and approximately 20% of parr were above the predicted

373 threshold level. To our knowledge, this is the first study to evaluate myxozoans with these

374 macroparasite techniques. Nanophyetus salmincola infection was the least linked to parasite-associated

375 mortality, as mean intensity and over-dispersion of this parasite actually increased from parr to smolt

376 stage, and the percentage of parr above the predicted threshold level was low. Jacobson et al. (2008)

377 used the former two techniques and showed that N. salmincola was associated with mortality in coho

378 salmon during early ocean residency.

379 Ebersole et al. (2006) reported poor survival (ca. 2%) for fish from the lower mainstem of this

380 river but concluded this was due to abiotic factors such as poor habitat. However, this could in turn be

381 related to the heavy infection levels associated with mortality in our study, as the production of

382 digenean trematodes and their intermediate hosts increases with elevated temperature (Poulin, 2006).

383 The extensively logged lower mainstem of this river often exceeds 20° C during summer (Ebersole et

384 al., 2006) and Cairns et al. (2005) found a positive correlation between temperature and neascus

385 metacercariae in the skin (black spot) of coho salmon at the same location. The intermediate snail hosts

386 of Apophallus spp. in Oregon are Fluminicola spp. (Niemi and Macy, 1974; Villeneuve et al., 2005)

387 and Apophallus sp. from the West Fork Smith River can utilize snails from the genera Fluminicola and

388 Juga (Ferguson et al., in press b). Perhaps the increased temperature in the lower mainstem of this 19 389 system enhances these snail populations, which would in turn increase parasite transmission to fish

390 at this location.

391 Apophallus sp.-associated mortality is most likely indirect. In support of this, Ferguson et al.

392 (2010) held coho salmon from the lower and upper mainstem of this river in laboratory tanks and only

393 four of approximately 50 lower mainstem fish died early in the study. However, during winter high

394 flow rates occur in the lower mainstem of this river due to simplified substrate, which may displace

395 fish and account for high over-winter mortality of coho salmon from this section of this river (Ebersole

396 et al., 2006, 2009). Certain parasites may decrease the swimming performance of heavily infected fish,

397 which would exacerbate this phenomenon. Apophallus brevis is associated with reduced growth of

398 yellow perch (Perca flavescens) (Johnson and Dick, 2001) and we have found a similar association

399 with Apophallus infections in coho salmon from the West Fork Smith River (Ferguson et al., in press

400 b). Consequently, smaller salmonids have reduced swimming performance (Taylor and McPhail,

401 1985; Ojanguren and Brana, 2003), which in turn may also decrease predator avoidance (Taylor and

402 McPhail, 1985). Heavily Apophallus-infected lower mainstem coho salmon from this river were

403 associated not only with reduced growth but also decreased swimming performance and

404 osmocompetence (Ferguson et al., in press b), both of which could severely impact survival of

405 outmigrating smolts in the estuarine environment. An Apophallus sp. identified as Apophallus donicus

406 experimentally develops in both avain and mammalian hosts (Niemi and Macy, 1974) and naturally

407 infects gulls (Shaw, 1947). West Fork Smith River resides near a national wildlife refuge that could

408 provide a source of various piscivorous birds to prey upon heavily parasitized fish in this river

409 (Ferguson et al., in press b). When taken together, coho salmon from the lower mainstem of this river

410 are subjected to warmer summer temperatures, higher winter flow rates and higher levels of parasitism

411 than fish from the upstream tributaries, all of which may contribute to a poorer over-winter survival of

412 these fish compared with those from the tributaries. 20 413 The ‘source-sink’ is a well recognized paradigm in population ecology, where a certain

414 species may persist in a deficient habitat due to immigration from nearby rich ‘source’ habitats (Holt

415 and Hochberg, 2002). Holt and Hochberg (2002) present theoretical models on how pathogens could

416 drive this relationship. In our coho salmon system, the sink is the lower mainstem, which is driven by

417 heavy parasite burdens, and new fish are derived mostly from brood fish that originally grew as parr in

418 the tributaries of the same river, i.e., the source. Some salmon have developed genetic strains that are

419 resistant or resilient to other parasite infections when those occur in endemic watersheds

420 (Bartholomew, 1998; Gilbey et al., 2006), and this has even been reported with N. salmincola infecting

421 different strains of cutthroat trout (Oncorhynchus clarkii clarkii) (Baldwin et al., 1967). As very few of

422 the coho salmon from the lower mainstem survive to the smolt stage or even to sexual maturity, these

423 fish contribute very little to the genetic structure of this coho salmon population. This would negate the

424 possibility for genetic pressure for the development of resilience to the parasites in this river, which is

425 consistent with the theory that there are constraints to adaptive evolution to sink conditions (Holt and

426 Hochberg, 2002). Most Pacific salmon populations do not have genetic separation between nearby

427 rivers, let alone within a river (Johnson and Bank, 2008). We conclude that the parents of the lower

428 mainstem parr are almost always derived from regions of the river with little genetic pressure to

429 develop resilience to parasitism, thus resulting in a constant pool of susceptible fish in the lower

430 mainstem.

431 In conclusion, we have shown that metazoan parasites, especially Apophallus sp., are

432 associated with over-winter mortality of juvenile coho salmon. Fish from the lower mainstem were

433 heavily parasitized and the level of mortality associated with parasitism indicated in our analysis

434 approximates previous reports of estimated mortality of these fish in this section of the West Fork

435 Smith River. Our results stress the importance of examining parasitism as a potential limiting factor for

436 threatened populations. Several other coastal rivers in the Pacific Northwest also contain Endangered 21 437 Species Act listed coho salmon and our analytical techniques could be applied to these systems.

438 Understanding why certain salmon populations are heavily infected with these parasites, likely due to

439 landscape characteristics, would provide useful data for management or recovery planning. Lastly, our

440 newly proposed model for inferring parasite-associated mortality may be applied to many host-parasite

441 systems and is a significant contribution to the field of parasitology because a biological basis of

442 macroparasitism is more parsimonious and intuitive than a statistical one.

443 Acknowledgements

444 This research was funded, in most part, by an Oregon Department of Fish and Wildlife

445 (ODFW), USA, Fish Health Graduate Research fellowship (agency grant 010-7032-IAA-FISH) to

446 J.A.F. We would also like to thank J. Sanders, C. Ferguson and the research personnel at ODFW for

447 assistance in gathering field samples. The concept of congestion rate was originally developed by I.

448 Ninomiya, Ehime University, Japan. Thanks to K. Lafferty and E. Casillas for manuscript review and

449 comments.

450 22 451 References

452 Adlard, R.D., Lester, R.J.G., 1994. Dynamics of the interaction between the parasitic isopod, Anilocra

453 pomacentri, and the coral reef fish, Chromis nitida. Parasitology 109, 311-324.

454 Anderson, R.M., May, R.M., 1979. Population biology of infectious diseases: Part I. Nature 280, 361-

455 367.

456 Anderson, R.M., May, R.M., 1982. Coevolution of hosts and parasites. Parasitology 85, 411-426.

457 Bakke, T.A., Harris, P.D., 1998. Diseases and parasites in wild ( salar)

458 populations. Can. J. Fish. Aquat. Sci. 55, 247-266.

459 Baldwin, N., Millemann, R., Knapp, S., 1967. " Salmon Poisoning" disease. III. Effect of experimental

460 Nanophyetus salmincola infection on the fish host. J. Parasitol. 53, 556-564.

461 Bartholomew, J., 1998. Host resistance to infection by the myxosporean parasite Ceratomyxa shasta: a

462 review. J. Aquat. Anim. Health 10, 112-120.

463 Brass, W., 1958. Simplied methods of fitting the truncated negative binomial distribution. Biometrika

464 45, 59-68.

465 Burgett, D., Rossignol, P., Kitprasert, C., 1990. A model of dispersion and regulation of brood mite

466 (Tropilaelaps clareae) parasitism on the giant honeybee (Apis dorsata). Can. J. Fish. Aquat. Sci.

467 68, 1423-1427.

468 Bush, A.O., Lafferty, K.D., Lotz, J.M., Shostak, A.W., 1997. Parasitology meets ecology on its own

469 terms: Margolis et al. revisited. J. Parasitol., 575-583.

470 Cairns, M.A., Ebersole, J.L., Baker, J.P., Wigington, P.J. Jr., Lavigne, H.R., Davis, S.M., 2005.

471 Influence of summer stream temperatures on black spot infestation of juvenile coho salmon in the

472 Oregon Coast Range. Trans. Am. Fish. Soc. 134, 1471-147

473 Carey, J.R., 1993. Applied Demography for Biologists. Oxford University Press, New York, New

474 York. 23 475 Crofton, H.D., 1971. A quantitative approach to parasitism. Parasitology 62, 179-193.

476 Dobson, A., 1988. The population biology of parasite-induced changes in host behavior. Q. Rev. Biol.

477 63, 139-165.

478 Dobson, A., Carper, R., 1992. Global warming and potential changes in host-parasite and disease-

479 vector relationships, In: Peters, R., Lovejoy, T. (Eds.), Global warming and biodiversity. Yale

480 University Press, New Haven, Connecticut, pp. 201-217.

481 Dobson, A.P., Hudson, P.J., 1986. Parasites, disease and the structure of ecological communities.

482 Trends Ecol. Evol. 1, 11-15.

483 Dobson, A. P., May, R.M., 1987. The effects of parasites on fish populations-theoretical aspects. Int.

484 J.Parasitol. 17, 363-370.

485 Ebersole, J.L., Colvin, M.E., Wigington, P.J., Leibowitz, S.G., Baker, J.P., Church, M.R., Compton,

486 J.E., Miller, B.A., Cairns, M.A., Hansen, B.P., La Vigne, H.R., 2009. Modeling stream network-

487 scale variation in coho salmon overwinter survival and smolt size. Trans. Am. Fish. Soc. 138, 564-

488 580.

489 Ebersole, J.L., Wigington, P.J.J., Baker, J.P., Cairns, M.A., Church, M.R., 2006. Juvenile coho salmon

490 growth and survival across stream network seasonal habitats. Trans. Am. Fish. Soc., 1681-1697.

491 Ferguson, J.A., Schreck, C.B., Chitwood, R., Kent, M.L., 2010. Persistence of infection by

492 metacercariae of Apophallus sp., Neascus sp., and Nanophyetus salmincola plus two myxozoans

493 (Myxobolus insidiosus and Myxobolus fryeri) in coho salmon Oncorhynchus kisutch. J. Parasitol.

494 96, 340-347.

495 Ferguson, J.A., St-Hilaire, S., Peterson, T., Rodnick, K., Kent, M.L., in press a. Survey of Parasites in

496 Threatened Stocks of Coho Salmon (Oncorhynchus kisutch) in Oregon by Examination of Wet

497 Tissues and Histology. J. Parasitol. 24 498 Ferguson, J.A., Romer, J., Sifneos, J.C., Madsen, L., Schreck, C.B., Glynn, M., Kent, M.L., in press

499 b. Impacts of Multispecies Parasitism on Juvenile Coho Salmon (Oncorhynchus kisutch) in

500 Oregon. Aquaculture.

501 Galvani, A.P., 2003. Immunity, antigenic heterogeneity, and aggregation of helminth parasites. J.

502 Parasitol. 89, 232-241.

503 Gilbey, J., Verspoor, E., Mo, T.A., Sterud, E., Olstad, K., Hytterød, S., Jones, C., Noble, L., 2006.

504 resistance in Atlantic salmon Salmo salar. Dis. Aquat. Organ. 71, 119–129.

505 Gordon, D.M., Rau, M.E., 1982. Possible evidence for mortality induced by the parasite Apatemon

506 gracilis in a population of brook sticklebacks ( Culaea inconstans). Parasitology 84, 41-47.

507 Halvorsen, O., Andersen, K., 1984. The ecological interaction between arctic charr, Salvelinus alpinus

508 (L.), and the plerocercoid stage of Diphyllobothrium ditremum. J. Fish Biol. 25, 305-316.

509 Henricson, J., 1977. The abundance and distribution of Diphyllobothrium dendriticum (Nitzsch) and D.

510 ditremum (Creplin) in the char Salvelinus alpinus (L) in Sweden. J. Fish Biol. 11, 231-248.

511 Hoffman, G., 1999. Parasites of North American freshwater fishes. Comstock Publishing Associates,

512 Ithaca, New York.

513 Holt, R.D., Dobson, A.P., Begon, M., Bowers, R.G., Schauber, E.M., 2003. Parasite establishment in

514 host communities. Ecol. Lett. 6, 837-842.

515 Holt, R.D., Hochberg, M.E., 2002. Virulence on the edge: a source-sink perspective, In: Diekmann, U.,

516 Metz, J.A.J., Sabelis, M.W., Sigmund, K. (Eds.), Adaptive Dynamics of Infectious Diseases: in

517 Pursuit of Virulence Management. Cambridge University Press, Cambridge, UK, pp. 104–120.

518 Jacobson, K.C., Teel, D., Van Doornik, D.M., Castillas, E., 2008. Parasite-associated mortality of

519 juvenile Pacific salmon caused by the trematode Nanophyetus salmincola during early marine

520 residence. Mar. Ecol. Prog. Ser., 235-244. 25 521 Johnson, M. A., Banks, M.A., 2008. Genetic structure, migration, and patterns of allelic richness

522 among coho salmon (Oncorhynchus kisutch) populations of the Oregon coast. Can. J. Fish. Aquat.

523 Sci. 65, 1274-1285.

524 Johnson, M.W., Dick, T.A., 2001. Parasite effects on the survival, growth, and reproductive potential

525 of yellow perch (Perca flavescens Mitchill) in Canadian Shield lakes. Can. J. Fish. Aquat. Sci. 79,

526 1980-1992.

527 Kang S., Kim, S., Cho, S., 1985. Seasonal variations of metacercarial density of Clonorchis sinensis in

528 fish intermediate host, Pseudorasbora parva. Korean J. Parasitol. 23, 87-94.

529 Kocan, R., Hershberger, P., Winton, J., 2004. Ichthyophoniasis: An emerging disease of Chinook

530 salmon in the Yukon River. J. Aquat. Anim. Health 16, 58-72.

531 Krkosek, M., Lewis, M.A., Morton, A., Neil Frazer, L., Volpe, J.P., 2006. Epizootics of wild fish

532 induced by farm fish. P. Natl. Acad. Sci. USA 103, 15506-15510.

533 Lafferty, K.D., Morris, A.K., 1996. Altfered behavior of parasitized killifish increases susceptibility to

534 predation by bird final hosts. Ecology 77, 1390-1397.

535 Lanciani, C.A., Boyett, J.M., 1980. Demonstrating parasitic water mite-induced mortality in natural

536 host populations. Parasitology 81, 465-475.

537 Lemly, A., Esch, G., 1984. Effects of the trematode Uvulifer ambloplitis on juvenile bluegill sunfish,

538 Lepomis macrochirus: ecological implications. J. Parasitol. 70, 475-492.

539 Lester, R., 1984. A review of methods for estimating mortality due to parasites in wild fish

540 populations. Helgol. Mar. Res. 37, 53-64.

541 Love, M.S., Moser, M., 1983. A checklist of parasites of California, Oregon, and Washington marine

542 and estuarine fishes. NOAA Tech. Rep. NMFS SSRF-777.

543 Ludwig, J., Reynolds, J., 1988. Statistical ecology : a primer on methods and computing. J. Wiley and

544 Sons, New York, New York. 26 545 May, R.M., Anderson, R.M., 1979. Population biology of infectious diseases: Part II. Nature 280,

546 455-461.

547 McCallum, H., Dobson, A., 1995. Detecting disease and parasite threats to endangered species and

548 ecosystems. Trends Ecol. Evol. 10, 190-194.

549 McDonald, T.E., Margolis, L., 1995. Synopsis of the parasites of fishes of Canada: Supplement (1978-

550 1993). Can. Spec. Pub. Fish. Aquat. Sci. 122.

551 Niemi, D.R., Macy, R.W., 1974. The life cycle and infectivity to man of Apophallus donicus (Skrjabin

552 and Lindtop, 1919) (: Heterophyidae) in Oregon. Proc. Helminthol. Soc. Wash. 41, 223-

553 229.

554 Ojanguren, A.F., Braña, F., 2003. Effects of size and morphology on swimming performance in

555 juvenile brown trout (Salmo trutta L.). Ecol. Freshwat. Fish 12, 241-246.

556 Poulin, R., 2006. Global warming and temperature-mediated increases in cercarial emergence in

557 trematode parasites. Parasitology 132, 143-151.

558 Rodnick, K.J., St.-Hilaire, S., Battiprolu, P.K., Seiler, S.M., Kent, M.L., Powell, M.S., Ebersole, J.L.,

559 2008. Habitat selection influences sex distribution, morphology, tissue biochemistry, and parasite

560 load of juvenile coho salmon in the West Fork Smith River, Oregon. Trans. Am. Fish. Soc. 137,

561 1571-1590.

562 Royce, L.A., Rossignol, P.A., 1990. Epidemiology of honey bee parasites. Trends Parasitol. (Parasitol.

563 Today) 6, 348-353.

564 Rózsa, L., Reiczigel, J., Majoros, G., 2000. Quantifying parasites in samples of hosts. J. Parasitol. 86,

565 228-232.

566 Scott, M., Smith, G., (eds.) 1994. Parasitic and infectious diseases : epidemiology and ecology.

567 Academic Press, San Diego, California.

568 Shaw, J. N. 1947., Some Parasites of Oregon Wild Life. Oreg., Agric. Exp. Stn., Tech. Bul. 11., p. 16. 27 569 Shaw, D.J., Dobson, A.P., 1995. Patterns of macroparasite abundance and aggregation in wildlife

570 populations: a quantitative review. Parasitology 111, S111-S133.

571 Sindermann, C.J., 1987. Effect of parasites on fish populations: practical considerations. Int. J.

572 Parasitol. 17, 371-382.

573 Smith, G., 1994. Ecological epidemiology and standard measures of disease occurrence, In: Scott, M.,

574 Smith, G. (Eds.), Parasitic and Infectious Diseases: Epidemiology and ecology. Academic Press,

575 San Diego, California, pp. 65-71.

576 Taylor, E., McPhail, J., 1985. Burst swimming and size-related predation of newly emerged coho

577 salmon Oncorhynchus kisutch. Trans. Am. Fish. Soc. 114, 546-551.

578 US National Research Council, 1996. Upstream: Salmon and society in the Pacific Northwest.

579 National Academy Press, Washington D.C.

580 Villeneuve, D.L., Curtis, L.R., Jenkins, J.J., Warner, K.E., Tilton F.A., Kent, M.L., Watral, V.G.,

581 Cunningham, M.E., Markle, D.F., Sethajintanin, D., Krissanakriangkrai, O., Johnson, E.R., Grove,

582 R., Anderson, K.A., 2005. Environmental stresses and skeletal deformities in fish form the

583 Willamette River, Oregon, USA. Environ. Sci, Technol. 39, 3495-3506.

584 Vincent, E., 1996. Whirling disease in wild trout: The Montana experience. Fisheries 21, 32-33. 28 585 Figure legends 586 587 Fig. 1. Map of West Fork Smith River, Oregon, USA showing sampling sites for coho salmon 588 (Oncorhynchus kisutch) parr (indicated by boxes) from brood year (BY) 2007 and 2008 from the 589 tributaries and lower mainstem,and outmigrating smolts (Smolt Trap). Sample sizes of all parr and 590 smolt samples are also shown. 591 592 Fig. 2. Frequency distributions of Apophallus sp. infections in coho salmon (Oncorhynchus kisutch) 593 smolts, lower mainstem parr and tributary parr from West Fork Smith River, Oregon, USA. Data from 594 all years of the study were pooled for each group. Triangle = smolts, square = lower mainstem parr, 595 diamond = tributary parr. 596 597 Fig. 3. Categorized frequency distributions of parasites infecting coho salmon (Oncorhynchus kisutch) 598 smolts from West Fork Smith River, Oregon, USA fitted to either the negative binomial distribution or 599 the congestion rate. Apophallus sp. metacercariae (A, B), Nanophyetus salmincola metacercariae (C, 600 D) and Myxobolus insidiosus pseudocysts (E, F) using the negative binomial distribution truncation 601 technique (A, C, E) compared with the congestion rate model (B, D, F). Triangle = observed, square = 602 predicted, arrow = threshold where parasite-associated mortality is predicted to initially occur as 603 indicated by comparing predicted and truncated observed distributions. Note the different scales for 604 each type of parasite. 1 Table 1. Prevalence (%), mean intensity and over-dispersion (variance to mean abundance ratio; S2/x) of Apophallus sp. in muscle, Nanophyetus salmincola in muscle and kidney, and Myxobolus insidiosus in muscle from coho salmon parr (subyearlings) from the Lower mainstem and Tributary groups, and outmigrant smolts (yearlings) of West Fork Smith River, Oregon, USA.

Apophallus sp. N. salmincola M. insidiosus Groups Mean intensity Mean intensity Mean intensity n % S2/x % S2/x % S2/x (95% CI) (95% CI) (95% CI) Brood Year 2006 Smolts 20 95 47 (24-123) 171 100 43 (29-61)d 32 75 18 (11-30) 24 Brood Year 2007 Lower mainstem Parr 58 100a 753 (656-870)a 237 100a 123 (105-146)a 50 84a 487 (320-824)a 1481 Tributary Parr 76 37b 37 (6-110)b 380 100a 52 (46-58)b 14 86a 300 (217-424)a 643

Combined Parr (Lower mainstem + Tributary) 134 64c 520 (423-636)c 647 100a 83 (73-95)c 52 85a 380 (291-535)a 1113 Smolts 29 100a 77 (42-138)b 208 97a 158 (110-220)a 143 79a 409 (210-753)a 1096 Brood Year 2008 Lower mainstem Parr 70 100a 856 (695-1,107)a 881 100a 103 (89-121)a 45 97a 313 (215-494)a 998 Tributary Parr 100 30b 13 (6-25)b 54 99a 124 (106-142)a,b 70 47b 136 (72-287)b 878

Combined Parr (Lower mainstem + Tributary) 170 59c 603 (465-786)a 1367 99a 113 (102-126)a,b 61 68c 241 (176-366)a,b 1049 Smolts 30 93a 191 (115-329)c 415 100a 161 (125-210)b 88 63b,c 221 (108-422)a,b 575 Brood Year 2009 Smolts 31 100 251 (153-393) 467 97 190 (141-255) 141 81 482 (311-774) 777

The prevalence and mean intensity of infection (95% confidence intervals, (95% CI)) were tested among parr (Lower mainstem, Tributary, and Combined) and smolt groups for brood year 2007 and 2008. a-c = different letters represent significant differences (P < 0.05), if any parr group or the smolts from a given brood year share the same letter, then there was no significant difference between those groups. d = only muscle data were available. Table 2. Categorized frequency distribution of parasites infecting coho salmon (Oncorhynchus kisutch) smolts from West Fork Smith River, Oregon, USA fitted to either the negative binomial distribution or the congestion rate model.

No. Parasites Observed Crofton Congestion Predicted Predicted Apophallus sp. 0-100 79 79 13.12 101-200 10 13.99 10.53 201-300 7 7.48 8.45 301-400 7 4.89 6.78 401-500 1 3.49a 5.44 a 501-600 4 2.62 4.37 601-700 2 2.03 3.50 701-800 1 1.6 2.81 kb NA 0.2256 NA Nanophyetus salmincola 0-75 32 32 30.85 76-150 27 22.95 21.21 151-225 8 15.52 a 14.57 a 226-300 10 10.29 10.02 301-375 8 6.76 6.88 376-450 6 4.41 4.73 451-525 5 2.86 3.25 526-600 1 1.85 2.23 601-675 2 1.2 1.54 676-750 1 2.16 1.06 kb NA 1.1275 NA Myxobolus insidiosus 0-200 84 84 14.14 201-400 8 12.61 a 11.35 a 401-600 6 6.68 9.11 601-800 4 4.39 7.31 801-1000 3 3.16 5.87 1001-1200 4 2.4 4.71 1201-1400 3 1.88 3.78 1401-1600 3 1.51 3.03 1601-1800 2 1.23 2.43 1801-2000 2 1.01 1.95 kb NA 0.1697 NA a Threshold where parasite associated mortality is predicted to initially occur as indicated by comparing predicted and continually truncated observed distributions.

bThe k value is an inverse measure of aggregation for the negative binomial distribution.

NA, Not applicable.

Supplementary Table S1. Raw data obtained from coho salmon (Oncorhynchus kisutch) from West Fork Smith River, Oregon, USA used in this study. Parasite counts are shown for Apophallus sp., Nanophyetus salmincola and Myxobolus insidiosus for each fish. Fish are grouped according to brood year (yr.), river site of residence and life stage.

Sample Fish group Apophallus sp. N. salmincola per M. insidiosus per No. (brood yr., river site, life stage) per fish fish fish a 183 2006, Migrant, Smolt 0 4 14 a 184 2006, Migrant, Smolt 68 114 0 a 185 2006, Migrant, Smolt 52 16 0 a 186 2006, Migrant, Smolt 8 12 16 a 187 2006, Migrant, Smolt 14 20 2 a 188 2006, Migrant, Smolt 32 30 0 a 189 2006, Migrant, Smolt 28 36 18 a 190 2006, Migrant, Smolt 10 6 0 a 191 2006, Migrant, Smolt 400 118 60 a 192 2006, Migrant, Smolt 96 56 26 a 193 2006, Migrant, Smolt 16 44 8 a 194 2006, Migrant, Smolt 40 84 0 a 195 2006, Migrant, Smolt 16 20 24 a 196 2006, Migrant, Smolt 42 112 6 a 197 2006, Migrant, Smolt 18 48 8 a 198 2006, Migrant, Smolt 6 26 2 a 199 2006, Migrant, Smolt 2 8 14 a 200 2006, Migrant, Smolt 10 32 4 a 201 2006, Migrant, Smolt 20 60 58 a 202 2006, Migrant, Smolt 10 14 4

b 207 2007, Lower mainstem, Parr 1412 105 94 b 208 2007, Lower mainstem, Parr 912 67 40 b 209 2007, Lower mainstem, Parr 1914 205 218 b 210 2007, Lower mainstem, Parr 1022 128 826 b 211 2007, Lower mainstem, Parr 1300 91 28 b 226 2007, Lower mainstem, Parr 1178 91 16 b 227 2007, Lower mainstem, Parr 692 41 542 b 228 2007, Lower mainstem, Parr 960 63 384 b 229 2007, Lower mainstem, Parr 2068 216 1012 b 230 2007, Lower mainstem, Parr 972 143 258 b 231 2007, Lower mainstem, Parr 802 109 334 b 232 2007, Lower mainstem, Parr 750 124 0 b 234 2007, Lower mainstem, Parr 340 53 2 b 235 2007, Lower mainstem, Parr 832 138 60 b 237 2007, Lower mainstem, Parr 498 56 0 b 248 2007, Lower mainstem, Parr 546 150 97 b 251 2007, Lower mainstem, Parr 406 291 1780 b 253 2007, Lower mainstem, Parr 248 91 124 b 255 2007, Lower mainstem, Parr 540 76 31 b 256 2007, Lower mainstem, Parr 797 83 409 b 259 2007, Lower mainstem, Parr 302 220 0 b 260 2007, Lower mainstem, Parr 558 352 0 b 262 2007, Lower mainstem, Parr 304 175 724 b 265 2007, Lower mainstem, Parr 1406 113 113 b 266 2007, Lower mainstem, Parr 262 59 0 b 269 2007, Lower mainstem, Parr 1005 79 19 b 270 2007, Lower mainstem, Parr 354 128 48 b 272 2007, Lower mainstem, Parr 352 131 70 b 275 2007, Lower mainstem, Parr 1157 101 630 b 279 2007, Lower mainstem, Parr 1814 274 60 b 284 2007, Lower mainstem, Parr 940 212 1714 b 285 2007, Lower mainstem, Parr 1065 81 2479 b 290 2007, Lower mainstem, Parr 610 61 1133 b 291 2007, Lower mainstem, Parr 409 137 1392 b 293 2007, Lower mainstem, Parr 989 41 0 b 300 2007, Lower mainstem, Parr 593 51 391 b 305 2007, Lower mainstem, Parr 430 75 21 b 309 2007, Lower mainstem, Parr 592 75 7 b 312 2007, Lower mainstem, Parr 439 86 188 b 316 2007, Lower mainstem, Parr 475 44 341 b 317 2007, Lower mainstem, Parr 1010 69 88 b 328 2007, Lower mainstem, Parr 806 88 0 b 329 2007, Lower mainstem, Parr 1158 335 51 b 330 2007, Lower mainstem, Parr 373 323 100 b 331 2007, Lower mainstem, Parr 825 53 0 b 332 2007, Lower mainstem, Parr 816 121 331 b 333 2007, Lower mainstem, Parr 760 61 572 b 334 2007, Lower mainstem, Parr 373 104 4856 b 343 2007, Lower mainstem, Parr 1027 237 28 b 344 2007, Lower mainstem, Parr 343 189 39 b 345 2007, Lower mainstem, Parr 490 93 92 b 243 2007, Lower mainstem, Parr 942 118 128 b 241 2007, Lower mainstem, Parr 168 194 106 b 242 2007, Lower mainstem, Parr 803 120 879 b 244 2007, Lower mainstem, Parr 430 90 0 b 245 2007, Lower mainstem, Parr 432 30 258 b 246 2007, Lower mainstem, Parr 200 39 204 b 247 2007, Lower mainstem, Parr 474 57 558

b 212 2007, Tributary, Parr 2 55 124 b 213 2007, Tributary, Parr 4 27 110 b 214 2007, Tributary, Parr 6 80 94 b 215 2007, Tributary, Parr 0 49 66 b 216 2007, Tributary, Parr 6 71 136 b 217 2007, Tributary, Parr 6 20 62 b 218 2007, Tributary, Parr 0 120 4 b 219 2007, Tributary, Parr 0 105 268 b 220 2007, Tributary, Parr 2 82 44 b 221 2007, Tributary, Parr 4 62 6 b 222 2007, Tributary, Parr 0 79 46 b 223 2007, Tributary, Parr 342 52 464 b 224 2007, Tributary, Parr 0 72 4 b 225 2007, Tributary, Parr 2 52 104 b 249 2007, Tributary, Parr 0 49 894 b 250 2007, Tributary, Parr 0 34 80 b 252 2007, Tributary, Parr 19 68 728 b 254 2007, Tributary, Parr 19 100 0 b 257 2007, Tributary, Parr 0 29 1084 b 258 2007, Tributary, Parr 0 77 806 b 261 2007, Tributary, Parr 0 56 5 b 263 2007, Tributary, Parr 0 85 1043 b 264 2007, Tributary, Parr 0 58 130 b 267 2007, Tributary, Parr 529 46 172 b 268 2007, Tributary, Parr 0 70 462 b 271 2007, Tributary, Parr 0 53 305 b 273 2007, Tributary, Parr 0 80 0 b 274 2007, Tributary, Parr 0 15 6 b 276 2007, Tributary, Parr 0 28 383 b 277 2007, Tributary, Parr 0 16 495 b 278 2007, Tributary, Parr 0 41 0 b 280 2007, Tributary, Parr 0 43 0 b 281 2007, Tributary, Parr 4 18 0 b 282 2007, Tributary, Parr 0 28 0 b 283 2007, Tributary, Parr 0 21 284 b 286 2007, Tributary, Parr 0 44 349 b 287 2007, Tributary, Parr 6 39 0 b 288 2007, Tributary, Parr 0 39 138 b 289 2007, Tributary, Parr 5 61 1540 b 292 2007, Tributary, Parr 0 36 4 b 294 2007, Tributary, Parr 0 56 35 b 295 2007, Tributary, Parr 0 23 0 b 296 2007, Tributary, Parr 2 104 0 b 297 2007, Tributary, Parr 0 43 76 b 298 2007, Tributary, Parr 2 77 20 b 299 2007, Tributary, Parr 10 55 361 b 301 2007, Tributary, Parr 4 89 20 b 302 2007, Tributary, Parr 2 23 0 b 303 2007, Tributary, Parr 3 21 58 b 304 2007, Tributary, Parr 0 18 76 b 306 2007, Tributary, Parr 7 56 0 b 307 2007, Tributary, Parr 0 99 8 b 308 2007, Tributary, Parr 4 96 15 b 310 2007, Tributary, Parr 0 24 373 b 311 2007, Tributary, Parr 0 68 338 b 313 2007, Tributary, Parr 6 51 96 b 314 2007, Tributary, Parr 0 23 286 b 315 2007, Tributary, Parr 9 67 92 b 318 2007, Tributary, Parr 0 19 29 b 319 2007, Tributary, Parr 0 35 128 b 320 2007, Tributary, Parr 0 32 192 b 321 2007, Tributary, Parr 0 7 17 b 322 2007, Tributary, Parr 0 33 710 b 323 2007, Tributary, Parr 0 66 56 b 324 2007, Tributary, Parr 0 46 24 b 325 2007, Tributary, Parr 0 39 110 b 326 2007, Tributary, Parr 0 60 609 b 327 2007, Tributary, Parr 0 48 2046 b 335 2007, Tributary, Parr 2 14 150 b 336 2007, Tributary, Parr 0 38 15 b 337 2007, Tributary, Parr 0 15 159 b 338 2007, Tributary, Parr 6 20 64 b 339 2007, Tributary, Parr 13 73 1055 b 340 2007, Tributary, Parr 0 45 187 b 341 2007, Tributary, Parr 0 95 1602 b 342 2007, Tributary, Parr 0 98 28

346 2007, Migrant, Smolt 54 117 0 347 2007, Migrant, Smolt 16 35 30 348 2007, Migrant, Smolt 10 19 24 349 2007, Migrant, Smolt 6 4 0 350 2007, Migrant, Smolt 140 367 1610 351 2007, Migrant, Smolt 4 128 216 352 2007, Migrant, Smolt 298 302 2414 353 2007, Migrant, Smolt 510 381 1168 354 2007, Migrant, Smolt 12 224 1358 355 2007, Migrant, Smolt 196 406 56 356 2007, Migrant, Smolt 14 15 28 357 2007, Migrant, Smolt 62 388 2 358 2007, Migrant, Smolt 4 60 0 359 2007, Migrant, Smolt 10 70 772 360 2007, Migrant, Smolt 352 86 90 361 2007, Migrant, Smolt 266 252 162 362 2007, Migrant, Smolt 52 59 0 363 2007, Migrant, Smolt 20 23 8 364 2007, Migrant, Smolt 12 0 4 365 2007, Migrant, Smolt 14 38 30 366 2007, Migrant, Smolt 8 36 324 367 2007, Migrant, Smolt 14 23 4 368 2007, Migrant, Smolt 24 497 334 369 2007, Migrant, Smolt 20 281 12 370 2007, Migrant, Smolt 16 209 678 371 2007, Migrant, Smolt 6 128 0 372 2007, Migrant, Smolt 4 8 8 373 2007, Migrant, Smolt 16 68 68 374 2007, Migrant, Smolt 72 188 0

c 418 2008, Lower mainstem, Parr 4666 25 46 c 420 2008, Lower mainstem, Parr 3984 75 106 c 421 2008, Lower mainstem, Parr 1098 96 334 c 422 2008, Lower mainstem, Parr 2868 46 98 c 427 2008, Lower mainstem, Parr 2472 48 16 c 430 2008, Lower mainstem, Parr 2004 236 30 c 431 2008, Lower mainstem, Parr 2644 147 54 c 432 2008, Lower mainstem, Parr 870 83 958 c 433 2008, Lower mainstem, Parr 296 223 232 c 434 2008, Lower mainstem, Parr 518 74 72 c 435 2008, Lower mainstem, Parr 1918 111 32 c 436 2008, Lower mainstem, Parr 724 117 42 c 437 2008, Lower mainstem, Parr 204 115 40 c 439 2008, Lower mainstem, Parr 1324 212 188 c 446 2008, Lower mainstem, Parr 1322 69 390 c 447 2008, Lower mainstem, Parr 78 40 52 c 448 2008, Lower mainstem, Parr 434 206 98 c 449 2008, Lower mainstem, Parr 420 88 66 c 450 2008, Lower mainstem, Parr 1030 119 0 c 451 2008, Lower mainstem, Parr 522 85 58 c 452 2008, Lower mainstem, Parr 586 74 654 c 453 2008, Lower mainstem, Parr 268 76 210 c 454 2008, Lower mainstem, Parr 652 53 56 c 455 2008, Lower mainstem, Parr 596 65 162 c 456 2008, Lower mainstem, Parr 984 121 1758 c 457 2008, Lower mainstem, Parr 528 66 52 c 458 2008, Lower mainstem, Parr 224 28 58 c 460 2008, Lower mainstem, Parr 1542 81 76 c 462 2008, Lower mainstem, Parr 1282 95 46 c 463 2008, Lower mainstem, Parr 1394 133 538 c 464 2008, Lower mainstem, Parr 248 163 948 c 465 2008, Lower mainstem, Parr 1238 283 78 c 466 2008, Lower mainstem, Parr 76 118 32 c 467 2008, Lower mainstem, Parr 610 278 24 c 468 2008, Lower mainstem, Parr 1858 72 3454 c 469 2008, Lower mainstem, Parr 372 117 1502 c 470 2008, Lower mainstem, Parr 208 64 322 c 471 2008, Lower mainstem, Parr 540 37 88 c 472 2008, Lower mainstem, Parr 792 69 32 c 473 2008, Lower mainstem, Parr 182 37 78 c 474 2008, Lower mainstem, Parr 208 62 8 c 475 2008, Lower mainstem, Parr 412 80 46 c 476 2008, Lower mainstem, Parr 322 85 956 c 477 2008, Lower mainstem, Parr 456 63 0 c 478 2008, Lower mainstem, Parr 1360 120 114 c 479 2008, Lower mainstem, Parr 446 56 734 c 480 2008, Lower mainstem, Parr 426 22 466 c 481 2008, Lower mainstem, Parr 286 93 502 c 482 2008, Lower mainstem, Parr 546 41 40 c 483 2008, Lower mainstem, Parr 122 45 210 c 484 2008, Lower mainstem, Parr 492 54 64 c 486 2008, Lower mainstem, Parr 1736 208 94 c 487 2008, Lower mainstem, Parr 456 113 16 c 488 2008, Lower mainstem, Parr 284 274 116 c 489 2008, Lower mainstem, Parr 468 59 1842 c 490 2008, Lower mainstem, Parr 784 125 56 c 491 2008, Lower mainstem, Parr 304 24 20 c 492 2008, Lower mainstem, Parr 1034 90 252 c 493 2008, Lower mainstem, Parr 682 43 212 c 494 2008, Lower mainstem, Parr 326 146 234 c 495 2008, Lower mainstem, Parr 946 41 26 c 496 2008, Lower mainstem, Parr 202 66 192 c 497 2008, Lower mainstem, Parr 66 327 12 c 498 2008, Lower mainstem, Parr 1462 89 62 c 499 2008, Lower mainstem, Parr 158 47 32 c 500 2008, Lower mainstem, Parr 526 87 1060 c 501 2008, Lower mainstem, Parr 476 167 152 c 502 2008, Lower mainstem, Parr 184 42 76 c 503 2008, Lower mainstem, Parr 558 164 398 c 504 2008, Lower mainstem, Parr 632 118 202

c 415 2008, Tributary, Parr 6 230 0 c 416 2008, Tributary, Parr 0 116 400 c 417 2008, Tributary, Parr 0 204 166 c 419 2008, Tributary, Parr 0 0 152 c 423 2008, Tributary, Parr 0 76 0 c 424 2008, Tributary, Parr 0 148 112 c 425 2008, Tributary, Parr 14 88 134 c 426 2008, Tributary, Parr 20 14 458 c 428 2008, Tributary, Parr 6 82 0 c 591 2008, Tributary, Parr 42 94 98 c 592 2008, Tributary, Parr 82 118 204 c 505 2008, Tributary, Parr 0 154 1854 c 506 2008, Tributary, Parr 2 54 1356 c 507 2008, Tributary, Parr 0 44 0 c 508 2008, Tributary, Parr 0 107 0 c 509 2008, Tributary, Parr 0 69 4 c 510 2008, Tributary, Parr 0 70 0 c 511 2008, Tributary, Parr 0 93 22 c 512 2008, Tributary, Parr 4 20 0 c 513 2008, Tributary, Parr 0 31 2 c 514 2008, Tributary, Parr 0 62 12 c 515 2008, Tributary, Parr 0 170 0 c 516 2008, Tributary, Parr 0 201 0 c 517 2008, Tributary, Parr 0 41 14 c 518 2008, Tributary, Parr 0 98 0 c 519 2008, Tributary, Parr 0 76 4 c 520 2008, Tributary, Parr 0 81 0 c 521 2008, Tributary, Parr 0 433 0 c 522 2008, Tributary, Parr 2 141 20 c 523 2008, Tributary, Parr 0 66 0 c 524 2008, Tributary, Parr 0 282 0 c 525 2008, Tributary, Parr 0 78 0 c 526 2008, Tributary, Parr 0 36 0 c 528 2008, Tributary, Parr 0 129 0 c 529 2008, Tributary, Parr 2 179 24 c 530 2008, Tributary, Parr 0 108 0 c 531 2008, Tributary, Parr 0 35 0 c 532 2008, Tributary, Parr 0 95 2 c 533 2008, Tributary, Parr 6 146 0 c 534 2008, Tributary, Parr 0 129 0 c 536 2008, Tributary, Parr 0 67 6 c 537 2008, Tributary, Parr 0 470 12 c 538 2008, Tributary, Parr 0 42 12 c 539 2008, Tributary, Parr 0 49 8 c 540 2008, Tributary, Parr 2 60 12 c 541 2008, Tributary, Parr 2 158 122 c 542 2008, Tributary, Parr 2 117 90 c 543 2008, Tributary, Parr 2 157 2 c 544 2008, Tributary, Parr 0 9 0 c 545 2008, Tributary, Parr 0 16 0 c 546 2008, Tributary, Parr 0 261 0 c 547 2008, Tributary, Parr 0 37 0 c 548 2008, Tributary, Parr 0 146 0 c 549 2008, Tributary, Parr 0 76 0 c 550 2008, Tributary, Parr 0 66 0 c 551 2008, Tributary, Parr 0 152 0 c 552 2008, Tributary, Parr 0 84 4 c 553 2008, Tributary, Parr 0 55 0 c 554 2008, Tributary, Parr 0 584 0 c 555 2008, Tributary, Parr 0 40 30 c 556 2008, Tributary, Parr 0 48 0 c 557 2008, Tributary, Parr 0 50 8 c 558 2008, Tributary, Parr 0 82 0 c 559 2008, Tributary, Parr 0 204 0 c 560 2008, Tributary, Parr 0 168 0 c 561 2008, Tributary, Parr 0 85 0 c 562 2008, Tributary, Parr 0 170 0 c 563 2008, Tributary, Parr 2 114 290 c 564 2008, Tributary, Parr 2 210 0 c 565 2008, Tributary, Parr 0 44 0 c 566 2008, Tributary, Parr 2 128 70 c 567 2008, Tributary, Parr 0 226 2 c 568 2008, Tributary, Parr 0 110 0 c 569 2008, Tributary, Parr 0 90 38 c 570 2008, Tributary, Parr 2 99 0 c 571 2008, Tributary, Parr 2 187 22 c 572 2008, Tributary, Parr 0 123 0 c 573 2008, Tributary, Parr 0 117 2 c 574 2008, Tributary, Parr 2 134 0 c 575 2008, Tributary, Parr 0 143 0 c 576 2008, Tributary, Parr 0 36 20 c 577 2008, Tributary, Parr 4 151 62 c 578 2008, Tributary, Parr 0 156 0 c 579 2008, Tributary, Parr 0 164 0 c 580 2008, Tributary, Parr 0 128 22 c 581 2008, Tributary, Parr 0 25 6 c 582 2008, Tributary, Parr 0 111 0 c 583 2008, Tributary, Parr 0 30 70 c 584 2008, Tributary, Parr 2 135 0 c 585 2008, Tributary, Parr 0 135 68 c 586 2008, Tributary, Parr 0 136 0 c 587 2008, Tributary, Parr 2 93 68 c 588 2008, Tributary, Parr 0 123 58 c 589 2008, Tributary, Parr 4 81 0 c 590 2008, Tributary, Parr 0 56 0 c 441 2008, Tributary, Parr 6 114 6 c 442 2008, Tributary, Parr 62 133 0 c 443 2008, Tributary, Parr 4 58 0 c 444 2008, Tributary, Parr 96 321 258 c 445 2008, Tributary, Parr 2 143 4

620 2008, Migrant, Smolt 48 61 0 621 2008, Migrant, Smolt 820 186 412 622 2008, Migrant, Smolt 30 261 1148 623 2008, Migrant, Smolt 124 146 552 624 2008, Migrant, Smolt 0 95 0 625 2008, Migrant, Smolt 1150 320 358 626 2008, Migrant, Smolt 0 465 20 627 2008, Migrant, Smolt 62 43 0 628 2008, Migrant, Smolt 76 135 14 629 2008, Migrant, Smolt 74 22 4 630 2008, Migrant, Smolt 214 82 0 631 2008, Migrant, Smolt 626 389 0 632 2008, Migrant, Smolt 282 146 4 633 2008, Migrant, Smolt 34 61 34 634 2008, Migrant, Smolt 22 142 0 635 2008, Migrant, Smolt 344 374 44 636 2008, Migrant, Smolt 350 114 80 637 2008, Migrant, Smolt 16 93 0 638 2008, Migrant, Smolt 100 121 6 639 2008, Migrant, Smolt 8 3 4 640 2008, Migrant, Smolt 40 370 296 641 2008, Migrant, Smolt 6 61 80 642 2008, Migrant, Smolt 50 85 0 643 2008, Migrant, Smolt 80 122 0 644 2008, Migrant, Smolt 16 90 0 645 2008, Migrant, Smolt 42 231 46 646 2008, Migrant, Smolt 522 135 44 647 2008, Migrant, Smolt 16 89 0 648 2008, Migrant, Smolt 64 136 120 649 2008, Migrant, Smolt 144 259 940

818 2009, Migrant, Smolt 314 52 84 819 2009, Migrant, Smolt 72 0 38 820 2009, Migrant, Smolt 8 53 538 821 2009, Migrant, Smolt 1082 291 1512 822 2009, Migrant, Smolt 28 111 32 823 2009, Migrant, Smolt 44 276 1972 824 2009, Migrant, Smolt 988 378 522 825 2009, Migrant, Smolt 1004 185 0 826 2009, Migrant, Smolt 102 62 152 827 2009, Migrant, Smolt 4 227 16 828 2009, Migrant, Smolt 130 91 0 829 2009, Migrant, Smolt 502 647 1072 830 2009, Migrant, Smolt 8 216 10 831 2009, Migrant, Smolt 80 461 712 832 2009, Migrant, Smolt 146 367 1360 833 2009, Migrant, Smolt 50 149 542 834 2009, Migrant, Smolt 40 150 178 835 2009, Migrant, Smolt 80 33 32 836 2009, Migrant, Smolt 872 147 1504 837 2009, Migrant, Smolt 40 53 0 838 2009, Migrant, Smolt 100 40 218 839 2009, Migrant, Smolt 302 370 250 840 2009, Migrant, Smolt 52 26 0 841 2009, Migrant, Smolt 134 509 192 842 2009, Migrant, Smolt 290 110 84 843 2009, Migrant, Smolt 110 172 0 844 2009, Migrant, Smolt 14 58 0 845 2009, Migrant, Smolt 24 71 0 846 2009, Migrant, Smolt 232 263 832 847 2009, Migrant, Smolt 2 56 130 848 2009, Migrant, Smolt 922 77 60

For brood year 2006 migrant smolts, N. salmincola counts are derived from only skeletal muscle, whereas counts of the parasite from all other fish groups include counts from both skeletal muscle and posterior kidney.

a Original data taken from Ferguson et al. (in press a). b Original data taken from Ferguson et al. (2010). c Original data taken from Ferguson et al. (in press b).

Ferguson, J.A., St-Hilaire, S., Peterson, T., Rodnick, K., Kent, M.L., in press a. Survey of Parasites in Threatened Stocks of Coho Salmon (Oncorhynchus kisutch) in Oregon by Examination of Wet Tissues and Histology. J. Parasitol.

Ferguson, J.A., Schreck, C.B., Chitwood, R., Kent, M.L., 2010. Persistence of infection by metacercariae of Apophallus sp., Neascus sp., and Nanophyetus salmincola plus two myxozoans (Myxobolus insidiosus and Myxobolus fryeri) in coho salmon Oncorhynchus kisutch. J. Parasitol. 96, 340-347.

Ferguson, J.A., Romer, J., Sifneos, J.C., Madsen, L., Schreck, C.B., Glynn, M., Kent, M.L., in press b. Impacts of Multispecies Parasitism on Juvenile Coho Salmon (Oncorhynchus kisutch) in Oregon. Aquaculture.

Highlights f Parasite-associated mortality of threatened coho salmon is reported to occur. f Parasite infection, overdispersion and distribution are compared among three species. f Apophallus sp. (Digenea) was most commonly associated with coho salmon mortality. f A new alternative to Crofton’s (1971) distribution truncation technique is proposed. f Our Malthusian based congestion rate model does not require statistical inference.