1

Klamath River Fish Health Studies: Salmon Disease Monitoring and Research

Oregon State University, BOR/USGS Interagency Agreement #R15PG00065

FY2017 April 01, 2017 – March 31, 2018

ANNUAL REPORT

Principal Investigator: Jerri Bartholomew

Co-principal Investigator: Sascha Hallett

Contributing Scientists: Rich Holt, Julie Alexander, Stephen Atkinson, Ryan Craig, Amir Javaheri, Meghna Babbar-Sebens

Four components of OSU’s Klamath River research – Ceratonova shasta actinospores, sampling polychaetes, monitoring sentinel fish post-exposure and the polychaete host.

2

SUMMARY

- the C. shasta-infectious zone in the lower Klamath River retracted in 2017: it spanned I5 Bridge to Seiad Valley with the greatest sentinel fish loss and spore abundance near Beaver Creek; previously, in 2016, the infectious zone was greater in extent, and spanned I5 Bridge downstream to Orleans.

- sentinel exposures of Iron Gate Hatchery Chinook resulted in no mortalities associated with C. shasta at any of the index sites in April (at the 2 lower basin sites), May (2 upper and 4 lower basin sites) or September (3 lower basin sites).

- mortality of sentinel Chinook post-exposure occurred only in June; this was low (< 10%) at all sites below Iron Gate Dam (2.4% at I5 Bridge, 7.5% near Beaver Creek, 2.6% at Seiad Valley and 0% at Orleans). In contrast, in 2016, the June mortality was an order of magnitude higher (26%) at KI5, with the highest (81%) at Orleans.

- we attempted to compare the susceptibility of Iron Gate Hatchery coho with Trinity River Hatchery coho near Beaver Creek and Seiad Valley but there were high losses of the fish during river exposure, due to columnaris disease. At Beaver Creek, only 5 Iron Gate Hatchery coho survived field exposure, then were held at OSU: one died (20%) and was positive for C. shasta. Fifteen Trinity River Hatchery coho survived exposure; 3 of these died during monitoring and were positive (13.3%) for C. shasta. At Seiad Valley, 20 of each coho stock were exposed, survived exposure, and only 1 Trinity River Hatchery coho subsequently died, and was positive for C. shasta. Genotyped C. shasta-infections in coho were all ITS-1 type II.

- high mortality was again observed in susceptible out-of-basin rainbow trout exposed in the upper Klamath basin, despite cessation of stocking this species in 2011.

- similar to 2016, C. shasta ITS-1 genotype 0 was detected in susceptible rainbow trout exposed at Keno Eddy, and despite producing myxospores, the infection was not fatal.

- rainbow trout mortality at Beaver Creek was high in May and June of 2017, which was similar to many of the years prior to 2016 (losses were unusually low in 2016).

- C. shasta-DNA was first detected in water samples in April; initially, levels were highest at KSV (<5 spores/L), then KBC (11 spores/L).

- abundance of waterborne C. shasta was relatively low in 2017; the highest levels detected were 11 spores/L at KBC and KMN in June (in contrast to KOR in 2016).

- ITS-1 genotype I was dominant in water samples; <5 spores/L of type II was detected during salmonid outmigration; type 0 was also detected.

- polychaete densities and prevalence of C. shasta infection were lower than in previous years at all index sites, which we attribute to the combination of high magnitude and sustained duration of peak discharge, and low spring temperatures. Infection assays demonstrated that prevalence of C. shasta infection ranged from 0.03 - 0.06 % which was both lower than levels observed in previous years and corroborates results of water samples and sentinel fish exposures, which all demonstrated that parasite levels and disease risk were low in 2017.

- data were collected to evaluate the performance of the polychaete predictive model in the context of the 2017 peak flow. Fewer than 15% of the nearly 300 samples had polychaetes present. The majority (80%) of the samples with polychaetes present were low density samples (2564 - 5685 individuals per m2). Prevalence of infection was less than 1% in all samples, and the extrapolated densities of infected polychaetes were low in comparison to previous years (e.g., 2014-2015).

- recolonization rates remain low and polychaetes have not been collected from fine sediment habitats since 2015, which we attribute to the magnitude of peak flows in 2016-2017.

- a novel coho-specific genotype qPCR was developed and will be trialed in 2018 alongside the traditional method.

3

Contents SUMMARY ................................................................................................................................................................ 2

RESEARCH OUTCOMES ............................................................................................................................................. 9

Objective 1. Develop a long-term dataset on disease severity for Chinook and coho salmon that encompasses

years differing in the magnitude and timing of flows, temperatures during spring and summer, and adult

returns. ..................................................................................................................................................................... 9

Task 1. The following metrics will be measured at established index locations in the upper and lower Klamath

River during each study year: ...............................................................................................................................9

Task 1.1. Determine infection and disease severity in sentinel Chinook and coho salmon, according to the

following site schedule. ....................................................................................................................................9

Task 1.2. Determine parasite density in water samples to include data collection during spring out-

migration and fall in-migration. Water collection will occur at the following sites according to the following

schedule: ........................................................................................................................................................ 30

Task 1.3. Determine density and infection of the invertebrate (polychaete) host. Sampling will occur

quarterly at the following polychaete index sites: ........................................................................................ 35

Objective 2. Comply with 2013 Biological Opinion for Reclamation’s operation of the Klamath Project to ensure

weekly monitoring of actinospore genotype II concentrations in the mainstem Klamath River immediately

upstream of Beaver Creek mid-April to June, and expedite analysis and data dissemination. ............................. 42

Task 2.1. Genotype water samples collected weekly for 7 weeks from April 14 or a date within 6 days prior

to April 14 through the first full week of June. ............................................................................................. 42

Objective 3. Determine whether a pulse flow can affect C. shasta densities and infection risk in fishes. ............ 42

Task 3. Monitor Ceratonova shasta in association with pulse flow events - before, during and after an event

using the following metrics: infection prevalence and severity in sentinel fish, density of waterborne

parasite, infection prevalence in polychaetes and density of the polychaete host. ......................................... 42

Task 3.1. Collect water samples daily at Beaver Creek and Seiad Valley index sites, beginning three days

prior to the event and ending ~ three days after the event. Samples will be collected every 2 h using

automated samplers, and pooled to make a 6 h composite sample that is assayed using a C. shasta-specific

qPCR. .............................................................................................................................................................. 42

Task 3.2. Expose ~30 sentinel IGH Chinook and coho salmon juveniles at Beaver Creek and Seiad Valley

index sites for 72 h before and during an event. After exposures, fish will be transported to the Salmon

Disease Laboratory, reared at 18 °C water temperature and monitored for C. shasta infections for 60 days

as described in Task 3.1.1., above. ................................................................................................................ 42

Task 3.3. Collect polychaetes from 4 sites before and after a pulse flow. Three Hess samples each will be

collected from a reference site (KN) upstream of Iron Gate Dam, and from both fine and coarse sediments

at long term monitoring sites, Trees of Heaven, Beaver Creek and The Grange. Samples will be preserved

and returned to the laboratory for density and infection assays, as described in Task 1.3., above. ............ 42

Objective 4. Validate the index for predicting disease severity for Chinook and coho salmon by correlating data

on infection prevalence and disease severity in each fish species with genotype-specific spore densities in

4

water collected at each site. .................................................................................................................................. 42

Task 4.1. Develop a method for high throughput genotyping of C. shasta by identifying a genetic locus for

distinguishing among the 4 C. shasta ITS genotypes and between the 2 genotype II biotypes of C. shasta.42

Task 4.2. Use these loci to develop a method for high throughput genotyping of C. shasta. ...................... 42

Task 4.3. Correlate genotype-specific parasite density with infection prevalence and severity in Chinook

and coho salmon. .......................................................................................................................................... 44

Objective 5. Validate and refine the epidemiological model to identify sensitive parameters in the host-parasite

life cycle, simulate the effect of potential management strategies on the different stages of the life cycle, and

predict disease severity in juvenile salmonid populations under different parasite densities, water

temperatures and flows. ........................................................................................................................................ 44

Task 5.1. Investigate data gaps that are defined as the model is further developed. .................................. 44

Objective 6. Investigate the occurrence of C. shasta below the Trinity River confluence and ascertain spore type

of waterborne stages. Tribal biologists will assist with the following tasks. ......................................................... 45

Task 6.1. Conduct sentinel fish exposures when parasite abundance exceeds 10 spores/L but temperatures

are not lethal for salmon. .............................................................................................................................. 45

Task 6.2. Quantify parasite levels in water samples. .................................................................................... 45

Task 6.3. Characterize density and infection of the invertebrate (polychaete) host in the mainstem

downstream from the Trinity River confluence. ........................................................................................... 45

Objective 7. Develop and validate predictive models for polychaete hosts including distribution, density,

infection, and recolonization rates under different peak discharge, water years, and dam removal scenarios. . 45

Task 7.1. Validate and refine the polychaete distribution model for predicting distribution under different

peak discharge, water years, and dam removal scenarios. ........................................................................... 45

Task 7.2. Add polychaete density and infection prevalence data to the predictive model to examine how

flow regimes will affect density and infection in this host. ........................................................................... 48

Task 7.3. Estimate polychaete recolonization rates. Use the physical models to characterize hydraulic

conditions before and after disturbance. Predict polychaete distribution using the refined distribution

model. Validate with empirical data where available. .................................................................................. 49

Objective 8. Develop and synthesize a dataset, encompassing environmental risk factors and their relationship

with polychaete host ecology, to facilitate predictions about how polychaete densities and infection levels may

change under future climate and temperature regimes........................................................................................ 50

Task 8.1. Synthesize an “environmental risk factor” dataset comprised of water quality data and future

predictions for water temperature and discharge for examining correlations with polychaetes. ............... 50

Task 8.2. Examine correlations between environmental risk factors and polychaete host data including

density, population structure, and infection prevalence. ............................................................................. 50

Task 8.3. Construct models for generating predictions about how polychaete densities and infection levels

may change under future climate and temperature regimes, and how these changes in turn may affect

disease risk in salmon hosts. ......................................................................................................................... 50

5

Objective 9. Regularly disseminate research findings to provide stakeholders, managers, researchers and the

general public ready access to current information and historical datasets pertinent to C. shasta in the Klamath

River. ....................................................................................................................................................................... 52

List of Figures and Tables

FIGURE 1.1.1. Klamath River index sites for 2017 with site abbreviations and river kilometers (Rkm). ................................. 10

TABLE 1.1.1. Average Klamath River water temperature (°C) at sentinel sites during the 72-h fish exposures in 2017. ....... 12

FIGURE 1.1.2. Average daily water temperatures during the months of March-September during the years 2008 - 2017 at

the sentinel site near Beaver Creek (KBC) on the Klamath River mainstem. .......................................................................... 12

TABLE 1.1.2. Percent loss associated with infection by Ceratonova shasta by site and fish species in 2017 following a 72-hr

river exposure, except June 19 was 48 hr. Fishes were held at ambient Klamath River temperature at the Aquatic Animal

Health Laboratory and monitored for disease signs for 62 - 72 days post-exposure. Numbers represent total loss, after the

initial 5 days, and are based on either observation of parasite myxospores in wet mounts, or PCR assay of microscopically-

negative fish. ............................................................................................................................................................................ 13

FIGURE 1.1.3. Comparison of percent loss from C. shasta infections in rainbow trout (Rbt) and IGH Chinook salmon (Chf)

exposed in 72 h sentinel studies near Beaver Creek during April 2009 - 2017. ...................................................................... 14

Coho (IGH) were only exposed in 2009 and 2010. .................................................................................................................. 14

FIGURE 1.1.4. Percent C. shasta-associated mortality of sentinel Iron Gate Hatchery juvenile Chinook salmon and Roaring

River Hatchery susceptible rainbow trout (Rbt) exposed in lower Williamson River (WMR), and Klamath mainstem at Keno

Eddy (KED), near the I5 Bridge (KI5), near Beaver Creek (KBC), at Seiad Valley (KSV) and at Orleans (KOR). Sentinel fish at all

sites were exposed for 72 hours May 25 - 28, 2017. Fishes were held at the OSU Aquatic Animal Health Laboratory and

monitored for 72 days post-exposure. Fish in which no myxospores were observed had an intestinal sample PCR-tested. 15

FIGURE 1.1.5 Cumulative percent mortality associated with C. shasta in rainbow trout at six sentinel sites in May 2017. .. 16

FIGURE 1.1.6. Comparison of percent mortality from C. shasta of juvenile IGH Chinook salmon at six index sites in May

2007 - 2017. The Chinook salmon were exposed at most sites and most years, zeros indicate exposure but no loss. .......... 17

FIGURE 1.1.7. Comparison of percent mortality from Ceratonova shasta in juvenile rainbow trout exposed for 72 h during

May 2007 - 2017 at seven different Klamath River sentinel index sites and held for more than 60 days. The Klamathon site

(KKB) was exchanged for a site downstream a small distance to near the Interstate 5 Bridge in 2013. ................................ 18

FIGURE 1.1.8. Percent C. shasta-associated mortality of sentinel Big Creek Hatchery juvenile coho salmon, Trask River

Hatchery coho and Roaring River Hatchery susceptible rainbow trout (Rbt) exposed in the Lower Williamson River (Nature

Conservancy) (WMR), and at Keno Eddy (KED) in the upper Klamath River, for 48 hours June 19 21 2017. Fishes were held

at the OSU Aquatic Animal Health Laboratory and monitored for 65 days post-exposure. Fish in which no myxospores were

observed had an intestinal sample PCR-tested. ...................................................................................................................... 19

FIGURE 1.1.9. Percent C. shasta-associated mortality of sentinel Iron Gate Hatchery juvenile Chinook salmon and Roaring

River Hatchery susceptible Rainbow trout (Rbt) exposed in the upper River in the lower Williamson River (WMR) but not at

Keno Eddy due to high water temperatures, below Iron Gate Dam near the I5 Bridge (KI5), near Beaver Creek (KBC), Seiad

6

Valley (KSV) and Orleans (KOR) in the Klamath River for 72 hours June 24 - 27, 2017. Also, coho from Iron Gate Hatchery

and Trinity River Hatchery were held during the same time near Beaver Creek and Seiad Valley. Only rainbow trout were

exposed at the Williamson River Lonesome Duck Resort site. Fishes were held at the OSU Aquatic Animal Health

Laboratory and monitored for 65 days post-exposure. Fish in which no myxospores were observed had an intestinal

sample PCR-tested. .................................................................................................................................................................. 21

FIGURE 1.1.10. Cumulative percent mortality associated with Ceratonova shasta in IGH Chinook salmon exposed for 72 h

near I5 Bridge (KI5), near Beaver Creek (KBC), Seiad Valley (KSV), and Orleans (KOR) in June 2017. ..................................... 22

FIGURE 1.1.11. Cumulative percent mortality associated with Ceratonova shasta in rainbow trout following exposure at six

sentinel sites in June 2017. ...................................................................................................................................................... 22

FIGURE 1.1.12. Comparison of percent mortality from Ceratonova shasta of juvenile IGH Chinook salmon (upper figure)

and coho salmon (lower figure) at six index sites exposed in June of 2007 - 2017. The Chinook salmon were exposed at

most sites and most years, zeros indicate exposure but no loss. The coho salmon were not exposed at all locations each

year and no exposures at KED have ever been done. ............................................................................................................. 24

FIGURE 1.1.13. Comparison of percent Ceratonova shasta-mortality for rainbow trout exposed in June 2007 - 2017 at

seven index sites of the Klamath River basin. The Klamathon site (KKB) was exchanged for a site downstream a small

distance to near the Interstate 5 Bridge in 2013. .................................................................................................................... 25

FIGURE 1.1.14. Percent Ceratonova shasta-associated mortality of rainbow trout (Rbt) and IGH fall Chinook salmon

exposed September 14-17, 2017 at three Klamath River sentinel index sites, near Beaver Creek (KBC), Seiad Valley (KSV)

and Orleans (KOR) then held for 61 days post-exposure at 18 °C. .......................................................................................... 26

FIGURE 1.1.15. Comparison of percent mortality from Ceratonova shasta of juvenile IGH Chinook salmon at six index sites

exposed in September of 2007 - 2017. The Chinook salmon were exposed at most sites in September in 2007, 2008 and

2009 but only near Beaver Creek and Seiad Valley in 2011, 2012, 2013, 2014, 2015 and Orleans was added in 2016 and

2017. Zeros indicate exposure but no loss. ............................................................................................................................. 27

FIGURE 1.1.16. Comparison of Ceratonova shasta mortality of Juvenile IGH fall Chinook and coho salmon exposed in the

Klamath River near Beaver Creek for 72 h in April, May, June and September (when conducted) in years 2007 - 2017. ...... 28

FIGURE 1.1.17. Comparison of C. shasta mortality of Juvenile IGH fall Chinook and coho salmon exposed in the Klamath

River near Seiad Valley for 72 h in April, May, June and September (when conducted) in years 2007 - 2017 ....................... 29

FIGURE 1.2.1. Density of Ceratonova shasta in water samples collected at the Beaver Creek index site (KBC) from 2009 –

2017. Each point is the average of 3 x 1L 24-hr composite water samples. The upper graph is scaled to show the maximum

level in 2015. ............................................................................................................................................................................ 31

FIGURE 1.2.2. Spatial and temporal density of Ceratonova shasta in water samples from two Klamath River mainstem

index sites in 2017 – Beaver Creek (KBC) and Orleans (KOR). Data are mean number of spores in 3 x 1L 24-hr composite

water samples. ......................................................................................................................................................................... 32

FIGURE 1.2.3. Spatial and temporal density of Ceratonova shasta in water samples collected at Klamath River mainstem

index sites in 2017. Each data point is the average of 3 x 1L 24-hr composite water samples. The Arcata Fish and Wildlife

Office Fish and Aquatic Habitat Conservation Program's predictive tool estimated that 80% of the natural juvenile Chinook

Salmon had migrated downstream of the Kinsman trap site on the Klamath River sample week May 7 - 13; therefore,

7

accounting for the court order's seven additional calendar days, the 80% outmigration 2017 date was May 20 (red line).. 32

FIGURE 1.2.4. Average density of Ceratonova shasta in 24-hr composite water samples collected at the Beaver Creek

index site (KBC) in 2017 (blue) compared with 2016 (gray). The Arcata Fish and Wildlife Office Fish and Aquatic Habitat

Conservation Program's predictive tool estimated that 80% of the natural juvenile Chinook Salmon had migrated

downstream of the Kinsman trap site on the Klamath River sample week May 7 - 13; therefore, accounting for the court

order's seven additional calendar days, the 80% outmigration 2017 date was May 20 (red line). ........................................ 33

FIGURE 1.2.5. Density of Ceratonova shasta in water samples from the Orleans index site in 2017 (green) compared with

2016 (grey). Data are mean number of spores in 3 x 1L 24-hr composite water samples. ..................................................... 33

TABLE 1.2.1. ITS-1 genotype of Ceratonova shasta in 1 L 24-hr composite or manual grab (g) water samples at the Beaver

Creek (KBC) index site. Samples collected during Chinook salmon outmigration are before May 20. ................................... 34

TABLE 1.2.2. Density of Ceratonova shasta in 1 L water samples (average of 3 x 1 L) collected manually at the start and end

of sentinel fish exposures. Red numbers indicate high levels of inhibition even when re-run diluted, hence parasite levels

are probably higher. X = no fish at this site / date ................................................................................................................... 35

FIGURE 1.3.1. Locations of monitoring sites from upstream (KED) to downstream (KOR) shown by black circles, USGS

discharge gages (green circles). Sites, in order from upstream to downstream, include KED, KJB in the JC Boyle bypass

reach, KI5 near the I5 overpass, KTH near the Tree of Heaven Campground, KBC near the confluence of Beaver Creek and

the mainstem river (Fisher’s RV park), KSV near Seiad Valley, and KOR at Dolan’s Bar fishing access near Orleans CA. ....... 36

FIGURE 1.3.2. Densities of Manayunkia sp. (circles), the polychaete host of C. shasta at 7 monitoring sites in 2013 - 2017,

including discharge (blue), and water temperature (black lines), from top to bottom monitoring sites include KED (grey

circles) and KJB (black circles) in the upper basin (top plot), KI5 (black circles), KTH (grey circles) and KBC (white circles) in

the mid basin (middle plot), and KSV (black circles) and KOR (grey circles) in the lower basin (lower plot). Discharge (right

plots, denoted in blue) and water temperature (denoted by red solid line) were obtained from USGS gaging stations and

OSU temperature loggers deployed near the sampling sites. ................................................................................................. 38

TABLE 1.3.1. Prevalence of Ceratonova shasta in Manayunkia sp. at 7 monitoring sites in 2017. NS = not sampled due to

unsafe flows. ............................................................................................................................................................................ 39

TABLE 1.3.2. Summary of seasonal prevalence of Ceratonova shasta in Manayunkia sp. at 7 monitoring sites in from 2013 -

2017. NS = not sampled due to unsafe flows. Lower river sites include KOR and KSV, middle sites include KBC, KTH, and

KI5, and upper sites include KED and KJB. ............................................................................................................................... 39

FIGURE 1.3.3. Densities of Manayunkia sp. (circles) infected with Ceratonova shasta at 7 monitoring sites in from 2013 -

2017, including discharge (blue), and water temperature (black lines), from top to bottom monitoring sites include KED

(grey circles) and KJB (black circles) in the upper basin (top plot), KI5 (black circles), KTH (grey circles) and KBC (white

circles) in the mid basin (middle plot), and KSV (black circles) and KOR (grey circles) in the lower basin (lower plot).

Discharge (right plots, denoted in blue) and water temperature (denoted by red solid line) were obtained from USGS

gaging stations and OSU temperature loggers deployed near the sampling sites. ................................................................. 41

FIGURE 7.1.1. Predicted and observed polychaete distributions under the 2017 modeled peak discharge scenario. The

polychaete distribution model predicted low probabilities of occurrence at the majority of locations where we did not

observe polychaetes, and high probabilities of occurrence at the majority of locations where we observed polychaetes.

8

Overall, our observations support the model predictions but still require some refinement, likely due to substrate

mismatch. ................................................................................................................................................................................ 47

FIGURE 7.1.2. Modeled effects of peak discharge on the probability of polychaete presence at x, y locations in the Tree of

Heaven Study reach. Predicted polychaete distributions under two modeled peak discharge scenarios including a dry

water year having a peak discharge of 1,200 cfs out of Iron Gate Dam (left) and a wet water year having a peak discharge

of 7,950 cfs (right). ................................................................................................................................................................... 48

FIGURE 7.3.1. Observations of polychaete assemblages in 2014 (low peak discharge) and 2017 (high peak discharge) on a

georeferenced sampling location at the Tree of Heaven Study reach. The polychaete distribution model predicted low

probabilities of polychaete occurrence at this location in 2017 given the hydraulic conditions during peak discharge, and

our observations support the model predictions. ................................................................................................................... 49

FIGURE 8.3.1. Model ensemble includes outputs from global circulation models, which provide inputs for a basin scaled

model, which provides outputs of future stream temperatures and discharge. These parameters are used as predictions

for 2-D hydraulic models that are paired with a predictive polychaete model and a degree-day model, and these models

provide inputs for the epidemiological model, which is used to quantify changes in risk of disease in the salmon host under

the different water year scenarios. ......................................................................................................................................... 51

9

RESEARCH OUTCOMES

Objective 1. Develop a long-term dataset on disease severity for Chinook and coho salmon that encompasses

years differing in the magnitude and timing of flows, temperatures during spring and summer, and adult

returns.

Task 1. The following metrics will be measured at established index locations in the upper and lower Klamath

River during each study year:

(a) Infection and disease severity in sentinel Chinook and coho salmon (b) Parasite density in water samples (c) Density and infection of the invertebrate (polychaete) host

Task 1.1. Determine infection and disease severity in sentinel Chinook and coho salmon, according to the following site schedule.

Sentinel fish exposures will occur at the following sites:

(1) Lonesome Duck - RKM 14.4 (Williamson River) (2) Williamson River – RKM 441 (3) Keno Eddy - RKM 369 (4) I5 Bridge Fish Trap - RKM 287 (5) above Beaver Creek – RKM 258 (6) Seiad Valley – RKM 207 (7) Orleans – RKM 90 (8) Tully Creek – RKM 62

Sentinel fish exposures occurred according to the following schedule in 2017:

(1) April 20 - 23 at KBC, KSV (2) May 25 - 28 at 6 mainstem sites (3) June 19 - 21 at KED and WMR, June 24 - 27 at 6 mainstem sites (4) September 14 - 17 at KBC, KSV, KOR

Task 1.1. Methods Sentinel fish exposures were conducted according to the sites and schedule above. In 2017, there were four

exposures for 72 h each at 2 - 7 index sites (Figure 1.1.1) in the lower and upper Klamath River mainstem.

Klamath River water temperatures in the upper mainstem at KED, in mid-to-late June approached 22 - 25 °C, so

exposures were conducted June 19 - 21, 5 days earlier than planned, and for only 48 h at both KED and WMR.

As in previous years, known C. shasta-susceptible triploid rainbow trout stock from Roaring River Hatchery

(Oregon Department of Fish and Wildlife) was held at most sites. Klamath River fall Chinook juveniles from Iron

Gate Hatchery (IGH) (California Department of Fish and Wildlife (CDFW)) were held at all sites except June at

WLD and KED, where water temperatures had become too high. A limited number of juvenile coho salmon

from IGH and Trinity River Hatchery (TRH) (CDFW) were available for June sentinel exposures at KBC and KSV.

Because of the limited availability of Klamath River coho, two out-of-basin Oregon coho stocks (Big Creek

Hatchery and Trask River Hatchery) were utilized for exposure at KED and WMR. Generally, the number of fish

of each species held in live cages (other than when noted) was 40 rainbow trout, 40 IGH fall Chinook salmon

and 20 IGH or TRH coho salmon and 40 Big Creek Hatchery or Trask River Hatchery coho salmon. The sentinel

juvenile fish weighed 0.5 - 1.6 g in April, 1 - 3.5 g in May, 1.3 - 6 g in June, and 14 – 20 g in September.

10

FIGURE 1.1.1. Klamath River index sites for 2017 with site abbreviations and river kilometers (Rkm).

Following the river exposure, the fishes were transported to the OSU John L. Fryer Aquatic Animal Health

Laboratory (AAHL), Corvallis, Oregon and held in specific-pathogen-free flow-through well water at a water

temperature similar to the river water temperature during the 72-h exposure. However, even if river water

temperatures averaged > 18 °C, fish were maintained <= 18 °C post-exposure because higher water

temperatures such as 20 - 22 °C made infections of the bacterium, Flavobacterium columnare, the cause of

columnaris disease, difficult to prevent. During the last hour of transport, the fish were given a 1 - 2 µg/mL

Furan 2 (nitrofurazone) bath in their transport containers to prevent columnaris disease. In addition, within 1 -

2 weeks of their arrival at the AAHL, all fish were treated with formalin baths and oxytetracycline (TM 200)

medicated food to prevent external parasites and bacterial infections. Control groups of each fish stock not

exposed at the Klamath River sites were included for each monthly exposure and given the same preventative

treatments as the river exposed fish. All groups of fish were monitored daily for C. shasta clinical disease signs

for 2 months (in accordance with ACUP #5010). Moribund fishes were euthanized and examined

microscopically for C. shasta-infection by wet mounts of lower gut material; if no myxospores were observed

then intestinal samples were collected for C. shasta-PCR testing. A subsample of moribund fish from each

group was also necropsied for other parasite and bacterial infections to eliminate those as causes of loss.

Mortality percentages given in the Results section below represent total fish loss with C. shasta-infections

determined microscopically or by PCR testing from fish that died later than 5 days after they were brought to

the AAHL.

Above Beaver Creek (KBC)

Rkm 258

IRON GATE DAM Rkm 306

Seiad Valley (KSV)

Rkm 207

Orleans (KOR)

Rkm 90

Keno Eddy (KED)

Rkm 369

Williamson River (WMR)

Rkm 441

Kinsman Fish Trap (KMN)

Rkm 235

I5 Fish Trap (KI5)

Rkm 287

Tully Creek (KTC)

Rkm 62

Lonesome Duck (WLD)

Rkm 14.4

Dolan’s Bar (KDB)

Rkm 95

Tree of Heaven Creek (KTH)

Rkm 281

JC Boyle Bypass (KJB)

Rkm 366

11

Methods specific to each of the exposures are listed as follows:

April 20 - 23: In 2016 and 2017, we conducted the exposures during the third week of April because parasite

levels in the river in early April were low. Susceptible rainbow trout and IGH fall Chinook salmon were exposed

at 2 sites in the lower Klamath River mainstem sites: upstream of the Beaver Creek confluence (KBC) and near

Seiad Valley (KSV). River water temperature during exposure was 11 - 12 °C and all fishes were held at 13 °C at

the AAHL post-exposure.

May 25 - 28: We conducted May sentinel exposures at 6 sites: 2 in the upper basin: lower Williamson River

(WMR) and Keno Eddy (KED); and 4 below Iron Gate Dam: near the I5-bridge (KI5), near Beaver Creek (KBC),

Seiad Valley (KSV) and Orleans (KOR). Iron Gate Hatchery Chinook and susceptible rainbow trout were held at

all 6 sites. No juvenile coho from IGH were available for exposures in May. River water temperatures averaged

17 - 19 °C in the upper river and 14 - 18 °C at the lower river sites. All fishes were held 16 °C at the AAHL post-

exposure.

June 19 - 21: The scheduled 72 h June exposures in the upper and lower Klamath River were planned for June

24 - 27, however water temperatures in the upper Klamath especially at Keno Eddy (KED) were reaching >23 °C

so an early exposure was conducted at KED and the lower Williamson River (WMR) June 19 - 21, but still water

temperatures exceeded 23 °C at KED during exposure, so fish exposed at both sites were removed early, after

48 h. For the KED and WMR exposures, we used 40 juvenile coho salmon/cage, from two different Oregon

hatcheries: Big Creek Hatchery, which produce a lower Columbia River coho stock and Trask River Hatchery, a

North Oregon coastal river coho stock. These stocks were included because only a limited number of Klamath

River coho were available for sentinel studies in June. The Big Creek coho were from a tributary to the lower

Columbia River (where C. shasta is endemic), whereas Trask River coastal coho were from a watershed where

C. shasta has not been found. Also, there has been only limited sentinel exposures using coho in the upper

Klamath River in previous years. We included susceptible triploid rainbow trout for comparison at both KED

and WMR.

June 24 - 27: Exposure at the remaining sites proceeded as planned. Fish were placed at 6 sites, including 2

locations on the lower Williamson River (Nature Conservancy at the mouth of the river, WMR and Lonesome

Duck Resort, a few km upriver, WLD), and 4 in the lower Klamath River mainstem, below Iron Gate Dam near

the I5 Bridge (KI5), near Beaver Creek (KBC), Seiad Valley (KSV) and Orleans (KOR). Juvenile Chinook and the

known susceptible rainbow trout were exposed at 5 of these sites but only rainbow trout were held at WLD. To

examine the relative susceptibility of IGH and TRH coho stocks to C. shasta, we also held 20 juvenile coho per

cage of each stock at KBC and KSV. Average water temperatures during exposure were: 19.5 °C in the lower

Williamson River, and 20-22 °C at sites below Iron Gate Dam. After a 72-h exposure, all groups were

transported to the AAHL, reared in 18 °C well water and monitored daily.

September 14 - 17 exposure: We exposed IGH fall Chinook salmon and Roaring River Hatchery

C. shasta-susceptible triploid rainbow trout (40 fish of each species/cage) at 3 sites in September: KBC, KSV and

KOR. KOR was added to the September exposures in 2016 because exposure results for June 2016 showed

greater loss in Chinook salmon with C. shasta infections at the 2 lower river sites, KOR and KSV. River water

temperatures averaged 19 °C. After the 3-day exposure, fishes were reared in well water at 18 °C at the AAHL

and monitored.

12

Task 1.1. Results and Discussion Average water temperatures during the 72-h exposures at all sites and the laboratory post-exposure rearing

temperature are shown in Table 1.1.1. Temperatures ranged from 10 - 12 °C during April exposures to 19 °C in

September. A maximum laboratory rearing water temperature of 18 °C was chosen to avoid loss from F.

columnare. For comparison, Figure 1.1.2 shows the average daily water temperature during the months of April

to September near Beaver Creek for 2008 - 2017. Water temperatures in April and early to mid-May 2017 were

cooler than most previous years but became warmer than most years late May through August.

TABLE 1.1.1. Average Klamath River water temperature (°C) at sentinel sites during the 72-h fish exposures in 2017.

Site April 20 - 23

May 25 - 28

June 19 - 21

June 24 - 27

Sept 14 - 17

Williamson R - WMR 17.5 20.1 19.5

Williamson R - WLD 18.2

Keno Eddy - KED 19.3 20.5

Klamath R I5 - KI5 17.1 20.9

Beaver Creek - KBC 11.1 17.6 21.9 19.4

Seiad Valley - KSV 10.9 14.6 20.3 19.2

Orleans - KOR 13.7 20.6 19.4

Post-exposure rearing at AAHL 13 16.0 18.0 18.0 18.0

FIGURE 1.1.2. Average daily water temperatures during the months of March-September during the years 2008 - 2017

at the sentinel site near Beaver Creek (KBC) on the Klamath River mainstem.

13

Results of the 2017 sentinel exposures in April, May, June, and September are summarized in Table 1.1.2, and

are shown for each month in Figures 1.1.2 - 1.1.15. The percent loss represents fish that were moribund or

dead and were removed from the tanks during the post-exposure rearing, not including any loss that occurred

during the first five days. These fish were regarded as positive for infections of C. shasta, either by microscopic

observation for myxospores in intestinal wet mounts or polymerase chain reaction (PCR) testing of intestinal

tissue. The results for each exposure are discussed after each figure, and compared with previous year’s

results.

TABLE 1.1.2. Percent loss associated with infection by Ceratonova shasta by site and fish species in 2017 following a 72-hr river exposure, except June 19 was 48 hr. Fishes were held at ambient Klamath River temperature at the Aquatic

Animal Health Laboratory and monitored for disease signs for 62 - 72 days post-exposure. Numbers represent total loss, after the initial 5 days, and are based on either observation of parasite myxospores in wet mounts, or PCR assay of

microscopically-negative fish.

Exposure dates

Exposure site and water

temperature

IGH Chinook salmon

IGH coho

TRH coho

Big Cr coho

Trask R. coho

Rainbow trout

April 20-23 KBC 11 °C 0 0

KSV 11 °C 0 0

May 25-28 WMR 17 °C 0 97.6

KED 19 °C 0 3.7

KI5 17 °C 0 45

KBC 18 °C 0 92.3

KSV 15 °C 0 63.4

KOR 14 °C 0 30.8

June 19-21 WMR 20 °C 0 2.5 100

KED 20 °C 0 0 0

June 24-27 WMR 19 °C 0 95.5

WLD 18 °C 97.2

KI5 21 °C 2.4 72.5

KBC 22 °C 7.5 20 13.3 97.3

KSV 20 °C 2.6 0 5 68.6

KOR 21 °C 0 80.5

September 14-17

KBC 19 °C 0 40

KSV 19 °C 0 57.5

KOR 19 °C 0 51.2

April 20 - 23 exposure (Figure 1.1.2): No Chinook salmon exposed at KBC or KSV became moribund, thus all

were terminated after 67 d rearing. For the susceptible rainbow trout, 1 fish died after exposure at KBC, and 3

died after exposure at KSV, but no infections of C. shasta were detected (by microscope or PCR assay). In

summary, in 2017 the late-April 72 hr exposures resulted in no detected infections of C. shasta. For context,

14

compared with previous years (since 2009 when Chinook salmon were exposed late April, or in 2015 in early

April at KBC; Figure 1.1.3): no Chinook salmon died in April exposures 2010-2013, but in 2014 7% died, in 2015

10% died, in 2016 2.6% died, and none in 2017: of nine years of sentinel exposures in April, four years had only

low loss (< 15%) of Chinook salmon with C. shasta infections. Coho juveniles were exposed only in 2009 and

2010, in April, and no loss with C. shasta infections occurred. Rainbow trout exposed in April at KBC have

ranged in mortality with C. shasta-infections from very high (April 2009 and 2014, just under 100%, and about

80% in 2010) to low (fewer than 20% in 2011, 2013 and 2015); 2016 and 2017 were the only years when no

rainbow trout died with C. shasta-infection.

FIGURE 1.1.3. Comparison of percent loss from C. shasta infections in rainbow trout (Rbt) and IGH Chinook salmon (Chf) exposed in 72 h sentinel studies near Beaver Creek during April 2009 - 2017.

Coho (IGH) were only exposed in 2009 and 2010.

May 25-28 exposure (Figures 1.1.4-7): After 72 d of rearing, all groups except rainbow trout exposed at KED

were euthanized. A subset of the rainbow trout exposed at KED was terminated after 72 d, but the remainder

was held alive longer to monitor for C. shasta because previous exposures with rainbow trout in 2015 and 2016

at KED had shown that while these fish often would become infected with C. shasta in the river, they would not

0 0 0 0 00 00

20

40

60

80

100

2009 2010 2011 2012 2013 2014 2015 2016 2017

% C

. sh

ast

a m

ort

alit

y

Rbt Chf coho

0 0

15

die from the infection within the regular monitoring period. The late-May 2017 sentinel exposures of juvenile

Chinook salmon resulted in no losses with C. shasta infections at any of the six index sites (Figure 1.1.4). One

Chinook died within the first 5 d after exposure at KI5, but was infected with F. columnare. Rainbow trout

exposed at WMR and KBC had high losses with C. shasta-associated infections, while moderate losses were

observed at KI5, KSV and KOR. Ceratonova shasta-infections were detected in 3.7 - 97.6% of susceptible

rainbow trout mortalities, depending on site. All but one (97.6%) of rainbow trout exposed at WMR died with

C. shasta-infection. Only three fish (3.7%) died with C. shasta-infection at KED, but 45% (18 fish) from KI5,

92.3% (36 fish) at KBC, 63.4% (26 fish) at KSV and 30.8% (12 fish) at KOR. For the KED rainbow trout, 10

surviving fish were sampled 72 d post-exposure: 7 were infected with C. shasta, and found to be ITS-1

genotype 0 (similar to 2016).

FIGURE 1.1.4. Percent C. shasta-associated mortality of sentinel Iron Gate Hatchery juvenile Chinook salmon and Roaring River Hatchery susceptible rainbow trout (Rbt) exposed in lower Williamson River (WMR), and Klamath

mainstem at Keno Eddy (KED), near the I5 Bridge (KI5), near Beaver Creek (KBC), at Seiad Valley (KSV) and at Orleans (KOR). Sentinel fish at all sites were exposed for 72 hours May 25 - 28, 2017. Fishes were held at the OSU Aquatic

Animal Health Laboratory and monitored for 72 days post-exposure. Fish in which no myxospores were observed had an intestinal sample PCR-tested.

0 0 0 0 0 00

10

20

30

40

50

60

70

80

90

100

KOR KSV KBC KI5 KED WMR

% C

. sh

asta

mo

rtal

ity

Rbt

Chinook

16

Cumulative mortality curves (Fig. 1.1.5) show disease progression and total mortality. Rainbow trout exposed

at two locations in the upper river (WMR & KED) and at 4 locations in the lower river (KI5, KBC, KSV and KOR) in

May 2017 became infected with C. shasta. However, fish exposed at WMR died with C. shasta infections most

rapidly of all 6 sites, followed by those exposed at KBC; most rainbow trout exposed at WMR or KBC had died

by day 35 - 40 (Figure 1.1.5).

FIGURE 1.1.5 Cumulative percent mortality associated with C. shasta in rainbow trout at six sentinel sites in May 2017.

When comparing the lack of any C. shasta-associated infections in the May exposed IGH Chinook salmon at the

upper Klamath River sites 2007 - 2017, consistently no mortalities occurred in the WMR-exposed fish and only

low mortality associated with C. shasta-infection was detected in one year (2010) at KED (Figure 1.1.6). Below

Iron Gate Dam before 2015, the greatest loss of Chinook occurred in 2008 and 2009 at KBC and KSV and both

locations are considered the "hot zone" where more fish typically become infected than elsewhere in the lower

river. In May 2010 - 2013, Chinook salmon exposed for 72 h had losses with C. shasta that were very low, or

none died. In 2014, juvenile Chinook died at a higher level than the previous four years and they were infected

with C. shasta (40.7% at KBC and 32.5% at KSV). In May 2015, Chinook suffered an extremely high loss with C.

shasta infections of 59% at KBC and 90.5% at KSV. In May 2016 and again in May 2017, similar to the May 2010

- 2013 exposure results, Chinook salmon exposed for 72 h had very low losses with C. shasta infections or none

died. Sentinel exposures of the Chinook in the lower river in May 2016 resulted in much lower losses compared

to those in May 2014 and 2015 and in May 2017 no loss occurred.

0

10

20

30

40

50

60

70

80

90

100

0 20 40 60

% C

. sh

ast

a m

ort

alit

y

# days post exposure

WMR

KED

KI5

KBC

KSV

KOR

17

FIGURE 1.1.6. Comparison of percent mortality from C. shasta of juvenile IGH Chinook salmon at six index sites in May 2007 - 2017. The Chinook salmon were exposed at most sites and most years, zeros indicate exposure but no loss.

The juvenile rainbow trout sentinel fish percent loss with C. shasta-infection for seven sites in May 2007 - 2017

in the Klamath River watershed (Figure 1.1.7) has generally been high for most sites. Two exceptions to this

were at Klamathon (KKB) in 2013 (this site has now been shifted to near the Interstate 5 Bridge - KI5) and KED

where percent loss levels varied from low to high, depending on the year. In 2015, rainbow trout loss with C.

shasta was 78.9% at KI5 and 0% at KED and near 100% at all other sites. The exposure of rainbow trout in May

at WLD was only performed in 2012. In 2016, rainbow trout loss with C. shasta at the WMR continued to be

very high (100%) in May but was greatly reduced at all of the other three sites tested. At KI5, rainbow trout loss

with C. shasta was very low (6.7%) similar to May 2010 - 2012. At KBC, a site where rainbow trout exposed in

May in previous years always incurred high losses from C. shasta, losses were only 17% in 2016. In May 2017,

rainbow trout loss with C. shasta was high (97.6%) at WMR, similar to previous years. Also, at KBC 2017 May

mortality exceeded 90%, in contrast to only 17% in 2016, but similar to 2007 - 2015. Loss of rainbow trout at

KED was 3.7%, similar to 2016. At KI5, rainbow trout loss with C. shasta was 45% in 2017 compared to 6.7% in

2016. Loss of rainbow trout at KSV was 63.4%, which was lower than levels observed in 2007 - 2015; no

rainbow trout were exposed at KSV in 2016. At KOR, loss of rainbow trout with C. shasta was 30.8%, similar to

2010, but much less than the > 90% mortality 2012 - 2015; no rainbow trout were tested in 2016.

0 00 0000 0 0 00 0 00 0 0 00 00 0 0 0 0 00

10

20

30

40

50

60

70

80

90

100

KOR KSV KBC KI5 KED WMR

% C

. s

ha

sta

mo

rta

lity

2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017

18

FIGURE 1.1.7. Comparison of percent mortality from Ceratonova shasta in juvenile rainbow trout exposed for 72 h during May 2007 - 2017 at seven different Klamath River sentinel index sites and held for more than 60 days. The

Klamathon site (KKB) was exchanged for a site downstream a small distance to near the Interstate 5 Bridge in 2013.

June 19-21 exposure (Figure 1.1.8): In the upper Klamath River basin, in addition to the usual groups of IGH

Chinook and known susceptible rainbow trout, we planned to include two out-of-basin Oregon hatchery coho

stocks: Big Creek Hatchery coho from a tributary to the lower Columbia River (expected to have some

resistance to C. shasta since the parasite is endemic in the Columbia River), and Trask Hatchery coho, a North

coastal hatchery stock of coho where C. shasta has not been found (thus these coho should be susceptible to

the parasite). Prior to conducting the exposure, we became aware that water temperatures at KED were rising

rapidly, and exceeded 23 °C some days. Because it was anticipated that we would not be able to expose fish at

KED if water temperatures exceeded 25 °C , we placed the two stocks of coho salmon and susceptible rainbow

trout at KED and WMR June 19, five days prior to the scheduled basin-wide exposures. These fish were

removed from these 2 sites on June 21 after 48 h exposure, and brought to the AAHL for rearing at 18 °C. On

August 25 after 65 days of rearing, all groups except the KED rainbow were euthanized with MS222. Similar to

the May exposure, the KED–exposed rainbow trout were enumerated, then held for extended rearing and

observation of C. shasta infections. Ceratonova shasta infections in coho were genotyped.

WMR results: None of the Big Creek Hatchery coho died. One Trask River coho died and had spores of C. shasta

in the intestine (genotype II). This was a small deformed fish and it is not known if it succumbed to C. shasta. All

of the susceptible rainbow trout died and were positive for C. shasta, either with visible myxospores or by PCR.

00

10

20

30

40

50

60

70

80

90

100

KOR KSV KBC KKB KED WMR WLD

% C

. sh

ast

am

ort

alit

y2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017

19

KED results: One coho from each stock died during the post-exposure rearing, but neither fish had myxospores

or were positive for C. shasta by PCR. A sub-sample of 10 coho from each stock were killed and then PCR

tested: 10/10 Big Creek coho were negative, 10/10 Trask coho were positive (100% genotype II). None of the

rainbow trout exposed at KED died during the post-exposure holding; a sub-sample of 10 rainbow trout were

killed and then frozen for later testing for myxospores. Later PCR testing of a subset of these showed 100%

genotype O.

It is interesting that of the 40 exposed Trask River Hatchery coho from an Oregon coastal stream where C.

shasta has not been found, only one fish died and had myxospores (of genotype II). We expected this group of

coho to have a similar susceptibility to the C. shasta IIR genotype as the known susceptible rainbow trout.

Perhaps the genotype of C. shasta that infects and kills susceptible rainbow trout does not affect C. shasta

susceptible stocks of coho or maybe the Trask coho stock has some resistance to this genotype?

FIGURE 1.1.8. Percent C. shasta-associated mortality of sentinel Big Creek Hatchery juvenile coho salmon, Trask River Hatchery coho and Roaring River Hatchery susceptible rainbow trout (Rbt) exposed in the Lower Williamson River

(Nature Conservancy) (WMR), and at Keno Eddy (KED) in the upper Klamath River, for 48 hours June 19 21 2017. Fishes were held at the OSU Aquatic Animal Health Laboratory and monitored for 65 days post-exposure. Fish in which no

myxospores were observed had an intestinal sample PCR-tested.

0 0 000

10

20

30

40

50

60

70

80

90

100

KED WMR

% C

. sh

ast

a m

ort

alit

y

Rbt Big Cr coho Trask coho

20

June 24-27 exposure (Figures 1.1.9-13): During the June 2017 sentinel exposures, river water averaged 18 - 19

°C at the lower Williamson River two sites, and 20 – 22 °C at exposure sites below Iron Gate Dam. At KED the

river water temperatures exceed 25 °C so our planned exposures of IGH Chinook and susceptible rainbow trout

were not done. Fish from several sites died of columnaris disease during the exposure in the cages or soon

after they were brought to the laboratory. One IGH Chinook exposed at WMR died with a F. columnare

infection after being transported to the AAHL. One of the IGH Chinook exposed in the river at KSV and two

from KOR succumbed to F. columnare infections. Both the IGH and TRH coho salmon held in cages at KBC

suffered the most severely from F. columnare infections. At the end of the 72 h exposure, eight of the IGH coho

were dead in the cage and six died shortly after transport to the AAHL, all showing signs of columnaris disease.

The TRH coho cage had two dead after 72 h and two died within the 5 days after transport to the laboratory.

One rainbow trout exposed at KBC and two at KSV died with F. columnare or fungal disease signs after

transport to the AAHL. On August 31, after 65 days of rearing of the June exposure groups at the AAHL, all

groups except the KED rainbow trout were euthanized with MS222 and enumerated. Similar to the May

exposures, the KED-exposed rainbow trout were enumerated then held for extended rearing and observation

of C. shasta infections.

None of the juvenile Chinook salmon exposed in the upper Klamath River basin at WMR died with C. shasta

infections. At lower basin sites, the Chinook experienced very low loss with C. shasta infections (Figure 1.1.9).

The June exposure of IGH Chinook resulted in the greatest loss (7.5%) with C. shasta infections at KBC. Chinook

exposed at KI5 had a 2.4% loss and at KSV 2.6% were C. shasta positive. Only one Chinook exposed at KOR died

but was negative for C. shasta. For IGH and TRH coho exposed at KBC, because of deaths caused by F.

columnare there was only 5 IGH coho and 15 TRH coho survivors in the post-exposure rearing. One of 5 (20%)

IGH coho died after only 6 days with a F. columnare infection; no myxospores were visible, but the gut sample

was PCR-positive for C. shasta. For the TRH coho, 3/15 died during holding and 2 (13.3%) were positive for C.

shasta: 2 TRH fish that died had F. columnare disease signs on day 6 and day 9 of the post-exposure rearing;

the one that died on day 6 was C. shasta PCR-positive, but the day 9 fish was PCR-negative; the third TRH coho

died after 26 days rearing, and it had myxospores in the intestine. In the comparison of IGH and TRH coho

exposed at KSV, of 20 fish of each stock held for post-exposure rearing, only one TRH coho died and that was

negative for myxospores in the intestine but the right eye was slightly cloudy and PCR-positive for a 5%

prevalence. Because there was so few coho held after the loss from F. columnare at KBC, this comparison of

coho stocks will need to be repeated next year. Similar to 2016, C. shasta-infection in coho likely reflected the

low abundance of genotype II in June 2017.

The susceptible rainbow trout were exposed at two upper river sites WLD and WMR and at four sites below

Iron Gate Dam, KI5, KBC, KSV and KOR. Rainbow trout exposed at the two Williamson River sites experienced

losses of 95.5% at WMR and 97.2% at WLD. The rainbow trout exposed at KI5 had a 72.5% loss and all had C.

shasta infections. Rainbow trout exposed at KBC had a loss of 97.3% and all were positive for C. shasta. At KSV,

68.6% of the rainbow trout died with C. shasta while 80.5% of those exposed at KOR died and all were positive.

21

FIGURE 1.1.9. Percent C. shasta-associated mortality of sentinel Iron Gate Hatchery juvenile Chinook salmon and Roaring River Hatchery susceptible Rainbow trout (Rbt) exposed in the upper River in the lower Williamson River (WMR) but not at Keno Eddy due to high water temperatures, below Iron Gate Dam near the I5 Bridge (KI5), near

Beaver Creek (KBC), Seiad Valley (KSV) and Orleans (KOR) in the Klamath River for 72 hours June 24 - 27, 2017. Also, coho from Iron Gate Hatchery and Trinity River Hatchery were held during the same time near Beaver Creek and Seiad Valley. Only rainbow trout were exposed at the Williamson River Lonesome Duck Resort site. Fishes were held at the OSU Aquatic Animal Health Laboratory and monitored for 65 days post-exposure. Fish in which no myxospores were

observed had an intestinal sample PCR-tested.

Cumulative loss of IGH Chinook salmon exposed for 72 h in June for the four exposure sites below Iron Gate Dam is shown in Figure 1.1.10. Chinook died with C. shasta infections at three lower river sites on days 20 - 24 post-exposure. The Chinook exposed at KOR incurred one dead fish at day 57 with no myxospores in the intestine and gut tissue was PCR-negative. In 2016, the farther down-river the exposure sites the more severe the loss and higher infection levels of C. shasta but in 2017 Chinook mortality was low and the infectious zone extended from KI5 (2.4%) only down to KSV (2.6%), which was a reversion to the pattern pre-2016, with greatest infection at KBC (7.5%).

The cumulative loss of rainbow trout exposed at six Klamath River sentinel sites in June 2017 show the most rapid losses at the two Williamson River sites: WMR (95.5%) and WLD (97.2%), followed by lower river mainstem sites KBC (97.3%), KOR (80.5%), KI5 (72.5%) and KSV (68.6%) (Figure 1.1.12). The rainbow trout exposed in the lower Williamson River had begun to die 19 - 20 days after exposure and most mortalities occurred by day 32. In 2016, the rainbow trout loss with C. shasta infections at KI5 was low (6.7%) and unusually low at KBC (25.6%) but in 2017 the loss with C. shasta near KBC had returned to the high levels observed pre-2016.

The Oregon Department of Fish and Wildlife stopped stocking C. shasta-susceptible rainbow trout into Spring Creek, a tributary in the Williamson River watershed, in 2011. However, we have not observed a reduction in C. shasta infection in susceptible rainbow trout exposed at Williamson River sites (WMR, WLD) for 72 h in May or June, 2012 - 2017.

00

10

20

30

40

50

60

70

80

90

100

KOR KSV KBC KI5 KED WMR WLD

% C

. sh

ast

a m

ort

alit

yRbt Chinook IGH coho TRH coho

0 0Not exposed

22

FIGURE 1.1.10. Cumulative percent mortality associated with Ceratonova shasta in IGH Chinook salmon exposed for 72

h near I5 Bridge (KI5), near Beaver Creek (KBC), Seiad Valley (KSV), and Orleans (KOR) in June 2017.

FIGURE 1.1.11. Cumulative percent mortality associated with Ceratonova shasta in rainbow trout following exposure at six sentinel sites in June 2017.

0

10

20

30

40

50

60

70

80

90

100

0 10 20 30 40 50 60

% C

. sh

ast

a m

ort

alit

y

# Days post exposure

KI5 ChF KBC ChF KSV ChF KOR ChF

0

10

20

30

40

50

60

70

80

90

100

0 20 40 60

% C

. sh

ast

a m

ort

alit

y

# days post exposure

WLDWMRKI5KBCKSVKOR

23

During the June sentinel exposures, the greatest loss of Chinook salmon occurred in 2007, 2008 and 2009 at

KBC and KSV and both locations are considered the "hot zone" where more fish become infected than

elsewhere in the lower river (Figure 1.1.12). In June 2010 - 2013, Chinook salmon exposed for 72 h incurred low

losses from C. shasta, i.e. < 20%. In 2014, Chinook loss was > 40% at KBC and KSV. In 2015, Chinook loss was

near 40% at KI5, KBC and KOR. Interestingly, even though water temperatures were very high at KSV, only

7.7% Chinook loss occurred. 2015 was the first year that Chinook loss in June reached 40% at KI5, and thus

expanded the infectious zone from KI5 to KOR, with the exception of KSV. For the June exposure in 2016, loss

of Chinook with C. shasta infections similar to 2015 occurred from KI5 down to KOR showing the same

expansion of the infectious zone, however, percent mortality increased substantially the further downriver

with the highest loss at KOR (81.1%). The 2016 June exposure at the lowermost test site, KOR, resulted in the

highest C. shasta loss for this site seen in all the monitoring years since 2007. In 2017, Chinook loss at all lower

river sites was very low; the highest C. shasta-mortality was only 7.5% at KBC. In 2017, the infectious zone

appears to have reverted to KI5 down to KSV with KBC being the site of highest infection, with no infection

downriver at KOR.

For coho salmon, the June exposures resulted in the highest percent infections at KBC and KSV in 2007, 2008,

2011 and 2013. In June 2013, coho salmon had a 28.6% C. shasta infection at KBC and 44.8% at KSV that was

greater than coho exposed in June 2010 and 2012. In June, 2014 the coho loss was severe at both KBC and KSV

and was 32.1% at KOR. The sentinel coho were more affected in June 2014 than the Chinook salmon. In June

2015, the yearling coho loss at KBC and KSV was much lower, 11.1 and 25%; respectively but 57% of the

surviving coho exposed at KBC had eye lesions that were likely C. shasta infections. In June 2016, Trinity River

Hatchery juvenile coho were exposed at KBC and KSV and no mortal infections of C. shasta were observed in

these fish. Since this was the first year we used TRH coho in sentinel studies instead of IGH coho either the

genotype of C. shasta that infects coho was present at a very low level in the river water or TRH coho have

greater resistance to C. shasta or both may be the case. In June 2017, we attempted to answer this question of

comparable susceptibility between IGH and TRH coho by exposing groups of each stock at KBC and KSV during

June sentinel exposures. Unfortunately, at KBC losses of IGH coho from F. columnare infections were so severe

only five fish were held during the post-exposure period. Additionaly, 4 of 19 TRH coho succumbed to

columnaris disease leaving only 15 coho in the post-exposure rearing period. Only 1 IGH coho and 2 TRH coho

that died were infected with C. shasta, which equated to 20% mortality of IGH coho and 13.3% mortality of

TRH coho; C. shasta genotype II was detected in these mortalities. At KSV, one of 19 (5%) TRH coho developed

an aberrant C. shasta infection in an eye, and none of 20 IGH coho were infected with C. shasta. Similar to

2016, the genotype that infects Klamath River coho (II) was not present at a high level in concurrent water

samples. Because of the low number of coho able to be held post-exposure for KBC and the low level of

genoptype II in the river water, this comparison of stocks will need to be repeated next year.

The rainbow trout sentinel loss for seven sites in June 2007 - 2015 in the Klamath River watershed (Figure

1.1.13) show greater than 90% mortality at most locations. In some years, rainbow trout exposed at KED and

KI5 there is moderate C. shasta-mortality and other years’ very low mortality. In 2016, for the five sites tested,

at WMR there continued to be a high level of mortality or slightly reduced to 90% instead of the usual 100%, no

loss at KED and low loss at KI5 and at KBC. For June 2016, at KBC, this was an unusual year with much lower

infections of the susceptible rainbow trout. The genotype affecting this stock of rainbow trout must have been

very low in the river. Most previous years have shown 95 - 100% mortality of rainbow trout at KBC. For June

2017, at the seven sites tested, in the Williamson River, high level of mortality continued at WMR (100%) for

24

the June 19 - 21 exposure and 95.5% for the June 24 - 27 exposure; at WLD, mortality was also high at 97.2%.

At KED after the June 19 - 21 exposure, there were no mortalities associated with C. shasta. In the lower river

at KI5, there was 72.5% mortality compared to 97.3% at KBC, 68.6% at KSV and 80.5% at KOR. The genotype

affecting susceptible rainbow trout appears to be returning after the low level of mortality in June 2016 and

becoming more similar to years before 2016.

FIGURE 1.1.12. Comparison of percent mortality from Ceratonova shasta of juvenile IGH Chinook salmon (upper figure) and coho salmon (lower figure) at six index sites exposed in June of 2007 - 2017. The Chinook salmon were exposed at

most sites and most years, zeros indicate exposure but no loss. The coho salmon were not exposed at all locations each year and no exposures at KED have ever been done.

0 00 0 000 00 0 00 0 00 0 00 0 00

20

40

60

80

100

KOR KSV KBC KI5 KED WMR

% C

. sh

ast

a m

ort

alit

y

2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017

00 0 00

20

40

60

80

100

KOR KSV KBC KI5 KED WMR

% C

. sh

ast

a m

ort

alit

y

2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017

IGH Chinook

IGH coho

25

FIGURE 1.1.13. Comparison of percent Ceratonova shasta-mortality for rainbow trout exposed in June 2007 - 2017 at seven index sites of the Klamath River basin. The Klamathon site (KKB) was exchanged for a site downstream a small

distance to near the Interstate 5 Bridge in 2013.

September 14 - 17 exposure (Figures 1.1.14-15). The September exposure of Iron Gate Hatchery Chinook and

susceptible rainbow trout occurred at KBC, KSV and KOR. In 2016, the lower river KOR site was added for the

September exposure because our June 2016 exposures in the lower river resulted in a higher loss of IGH

Chinook at that site. After the 72-h exposure, the groups of fish were transported to the AAHL, held at 18 °C

and observed for infections of C. shasta. The water temperature during exposure averaged 19 °C and F.

columnare, opportunistic bacteria or fungi infections caused loss in Chinook in the first 5 days post-exposure at

KBC (1 dead), and KOR (3 dead). Only 2 Chinook died after the first 5 days of rearing at the laboratory. On day

8, one Chinook exposed at KOR died with a F. columnare infection. Also, on day 47, one Chinook died in the

KBC exposed group. Gut tissue samples from both of these Chinook were negative for C. shasta in the PCR test.

No rainbow trout died from F. columnare during or after the exposure. In the post-exposure rearing, rainbow

trout losses with C. shasta infections were moderately high at all three sites, 40% at KBC, 57.5% at KSV and

51.2% at KOR. On November 17, 2017 61 days post-exposure, all groups were euthanized with MS222 and

sodium bicarbonate and then enumerated. For September 2017, the low presence of the parasite genotype

affecting Chinook in April, May and June continued for September. Susceptible rainbow trout experienced a

greater loss than what occurred to the September 2016 exposure groups but less than many previous years’

exposures in the fall.

0

20

40

60

80

100

KOR KSV KBC KKB KED WMR WLD

% C

. sh

ast

a m

ort

ali

ty2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017

26

FIGURE 1.1.14. Percent Ceratonova shasta-associated

mortality of rainbow trout (Rbt) and IGH fall Chinook

salmon exposed September 14-17, 2017 at three Klamath

River sentinel index sites, near Beaver Creek (KBC),

Seiad Valley (KSV) and Orleans (KOR) then held for 61 days post-exposure at 18

°C.

Comparison of percent C. shasta-infections in Chinook salmon exposed in September of 2007 - 2017 at selected

sites in the Klamath River basin shows that generally C. shasta-infections in this month are very low (Figure

1.1.15). In 5 of the last 11 years when exposures were conducted in September, in 2007, 2008, 2014, 2015 and

2016, sentinel Chinook became infected, and mostly with a low level of mortality. Exposures in other years

were negative for C. shasta-infections. However, in 2016 the greatest loss with C. shasta was at KOR and the

prevalence was much higher at this site than in any previous September exposures. September exposures at

KOR had only previously been conducted in 2007 - 2009. In 2017, only 2 Chinook died during 61 days of post-

exposure rearing and they were PCR negative for C. shasta.

0.0 0.0 0.00

102030405060708090

100

KOR KSV KBC

% C

. sh

ast

a m

ort

alit

y

no coho exposed in September

Rbt Chinook

27

FIGURE 1.1.15. Comparison of percent mortality from Ceratonova shasta of juvenile IGH Chinook salmon at six index sites exposed in September of 2007 - 2017. The Chinook salmon were exposed at most sites in September in 2007, 2008 and 2009 but only near Beaver Creek and Seiad Valley in 2011, 2012, 2013, 2014, 2015 and Orleans was added in 2016

and 2017. Zeros indicate exposure but no loss.

Comparison of sentinel results for the juvenile IGH Chinook and juvenile or yearling coho salmon exposed at

KBC in 2007 - 2017 indicate a shift toward more severe effects of C. shasta on the Chinook than coho 2007 -

2009 (Figure 1.1.16). In 2007, the loss of juvenile coho was very high while the Chinook loss was lower. In 2008,

both species suffered high loss in May and June. In 2009, the greatest loss occurred in May and June in the fall

Chinook. In general, however, losses for both species due to C. shasta have been high in May and June of 2007

- 2009. In contrast, for 2010 - 2013, Chinook incurred decreased infection and mortality associated with C.

shasta-infection. In 2011 and 2013, the coho loss was much higher at KBC. In 2014, loss of both Chinook and

coho at KBC was much greater than 2010 - 2013 with the exception of coho loss in June 2011. The loss in 2014

is similar to the high loss years of 2007 - 2009. In 2014, the greatest infection level was observed in the coho.

But in 2015, the Chinook had the greatest infection level and loss and yearling coho salmon much lower. In

2016, Chinook mortality was none or low in all months except June (29.5%) and coho were exposed only in

June with no mortalities associated with C. shasta-infection. In 2017, Chinook mortality was none or very low in

all months. In June, the only month where some Chinook mortality with C. shasta infections occurred was at

KBC (7.5%), KI5 (2.4%) and KSV (2.4%). Coho mortality associated with C. shasta-infection at KBC was 20% of

the IGH coho stock and 13.3% of the TRH coho stock (but note low overall survival in these groups). Low

0 0 0 00 0 0 000 0 0 0 0 00 00 00 0 0.00 0 00

10

20

30

40

50

60

70

80

90

100

KOR KSV KBC KI5 KED WMR

% C

. sh

ast

a m

ort

alit

y

2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017

IGH Chinook

28

numbers of coho for the exposure and severe columnaris infections require repeating these exposures to

compare relative susceptibility of the two Klamath River coho stocks.

FIGURE 1.1.16. Comparison of Ceratonova shasta mortality of Juvenile IGH fall Chinook and coho salmon exposed in the Klamath River near Beaver Creek for 72 h in April, May, June and September (when conducted) in years 2007 - 2017.

The comparison of C. shasta mortality of juvenile IGH fall Chinook salmon and juvenile or yearling coho salmon

exposed at Seiad Valley for 72 h in May and June of years 2007 - 2017 show similar results as the comparison

for the same years near Beaver Creek (Figure 1.1.17). No comparison can be made for Chinook and coho in

2007 since no coho salmon were exposed at Seiad Valley in that year. The coho salmon loss from C. shasta in

2011 and 2013 was much higher than the Chinook salmon. Also, in 2013 and 2014 there appears to be a slight

downstream shift of greater C. shasta rate at Seiad Valley compared to near Beaver Creek. In 2014, the coho

loss associated with C. shasta was the most severe observed since we began sentinel studies in the Klamath

River. In contrast in 2015, the juvenile Chinook loss associated with C. shasta infections was the greatest we

have observed at least in the month of April since we began sentinel studies and in May since 2009. In 2016,

coho were exposed only in June and no fish died with C. shasta infections while 60% of the Chinook exposed at

Seiad Valley in June died with C. shasta infections. In 2017, Chinook loss was observed at Seiad Valley only in

June and at a very low level of C. shasta infection (2.6%). Coho juveniles from IGH and TRH exposed only in

June had only one of 20 TRH coho that died with C. shasta-infection. No loss occurred in the IGH coho exposed

at the same time.

Three adult, spawned-out coho salmon carcasses were collected by Tony Dennis of the Mid Klamath

Watershed Council, in Seiad Creek (on 12/28 and 1/3), and Horse Creek (1/15)(likely within one week post-

0

10

20

30

40

50

60

70

80

90

100

MJ SMJ SAMJ SOAMJ SAMJ SAMJ SAMJ J SAMJ SAMJ SAMJ SAMJ S

% C

. sh

ast

a m

ort

alit

y Chinook coho

2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017

Ceratonova shasta-mortality of Chinook salmon versus coho 2007 - 2017 at Beaver Creek

29

mortem). Both Seiad Creek carcasses had visible mature C. shasta spores in the intestine, and these were

genotype II (typical for coho). The Horse Creek fish was negative for C. shasta by both microscopy and PCR

analysis. These data indicate that coho can survive to reproduce, despite being infected with C. shasta. It is

unknown if this is possible only at low doses of the parasite. Additionally, sporulation (formation of

myxospores), if not already complete ante-mortem, can occur within one week post-mortem.

FIGURE 1.1.17. Comparison of C. shasta mortality of Juvenile IGH fall Chinook and coho salmon exposed in the Klamath River near Seiad Valley for 72 h in April, May, June and September (when conducted) in years 2007 - 2017

Pulse-Flow sentinel fish exposures: None conducted in 2017.

0 0 000 00 0 0 0.00.00.00 0.0 00

10

20

30

40

50

60

70

80

90

100

MJSMJSAMJSOAMJSAMJSAMJSAMJ JSAMJSAMJSAMJSAMJS

% C

. sh

ast

a m

ort

alit

y

Chinook coho

2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017

Chinook versus coho mortality 2007 - 2017 at Seiad Valley

30

Task 1.2. Determine parasite density in water samples to include data collection during spring out-migration and fall in-migration. Water collection will occur at the following sites according to the following schedule:

(1) All of the fish exposure sites during exposures (2) Weekly all year at Beaver Creek and Seiad Valley (to encompass previous and current spring parasite ‘hot spots’) (3) Weekly from March through October at I5 Fish Trap, Orleans and Tully Creek (4) March through mid-June at Kinsman Fish Trap

Task 1.2. Methods Water samples were collected weekly by ISCOs (automatic samplers) at three Klamath River mainstem sites: I5

fish trap (KI5; starting 6/05), Orleans (KOR; starting 4/10) and Tully Creek (KTC; starting 4/18), through the end

of October. Water samples were collected throughout the year at two other mainstem sites: upstream of the

confluence with Beaver Creek (KBC) and Seiad Valley (KSV) (Figure 1.1.1). An additional ISCO collected water

samples weekly at the Kinsman fish trap (KMN) during the outmigration period, May 2 - June 26 (high flows

delayed the start of the fish trap and associated water collection in 2017). Temperature loggers (Hobos)

attached to each ISCO recorded river temperature every 15 minutes during the sampling periods. A flow meter

at KBC and KSV recorded water velocity daily. The ISCOs were programmed to begin sampling at 8 am and

collected 1 L from the river every 2 h for 24 h, then the total sample was mixed manually and 4 x 1 L sub-

samples taken. These were chilled until filtered, within 24 h of collection, then the filter disks shipped overnight

to OSU for molecular analysis. Karuk tribal biologists retrieved and filtered the 24-h composite samples from all

sites except the lowermost site, KTC, which Yurok tribal biologists had oversight. Expedited processing of water

samples occurred April through June.

Water samples (4 x 1 L samples) were also collected manually at the start and end of each sentinel fish

exposure: April (20 & 23; 2 sites); May (25 & 28; 6 sites); June (19 & 21, or 24 & 27; 7 sites); and September (14

& 17; 3 sites).

DNA was extracted from 3 of the 4 replicate filtered 1L-samples collected at each site and time point using a

published protocol and a commercial kit (Hallett et al. 2012). We used a duplex C. shasta/IPC assay to enable

simultaneous detection of C. shasta-DNA and assessment of inhibition using the ABI Internal Positive Control.

Each sample was run in duplicate and was re-run if values of these sample well pairs differed by more than 1

Cq ((an inter-well standard deviation of 0.7) and corresponded to ~1 spore per liter for both wells). Controls

were included in each qPCR run: positive (known numbers of spores) and negative (molecular grade water or

the same buffer used for diluting the samples). Each run also included a reference dilution series of C. shasta

DNA, for standard curve calibration and assay efficiency assessment. Samples with inhibition <=2 Cq had their

final C. shasta Cq value adjusted by this level of inhibition; samples with inhibition > 2 Cq were diluted further

and rerun (if greater than ~1 spore per liter, based on values of the replicate 1L-samples).

Results and Discussion Abundance of waterborne C. shasta was relatively low in 2017, and was comparable to pre-drought years (pre-

2014 - 2016)(Figure 1.2.1). In 2016, waterborne levels were highest (>300 spores/L) at KOR compared to <100

spores/L at the traditional ‘hot spot’ KBC, whereas in 2017, levels were generally higher at KBC (Figure 1.2.2).

Ceratonova shasta-DNA was first detected early April and was highest at KSV (for 4 weeks, <1 - <5 spores/L)

(Figure 1.2.3). Parasite abundance increased through May, with fluctuating maximum levels of fewer than 10

31

spores/L through spring. Waterborne levels did not surpass 5 spores/L prior to the 80% outmigration date of

May 20, 2017. Levels did surpass this threshold in June, but did not surpass 10 spores/L until mid-June at KBC

then 1 week later at KMN (11 spores/L at each site); these were the only two occasions when >10 spores/L

were detected throughout 2017. Levels peaked at different times at different index sites. Levels decreased late

summer but after several weeks at <1 spore/L in late summer/early fall (mid to late September), levels

increased slightly (to <2 spores/L) at several sites before decreasing to <1 spore/L for the remainder of the

year.

Compared to 2016, the levels of C. shasta in 2017 were 5 times lower at KBC (11 compared to 55 spores/L;

Figure 1.2.4). The contrast in levels, 2 orders of magnitude difference, detected at Orleans in 2017 (<10

spores/L) compared to 2016 (>200 spores/L) is highlighted in Figure 1.2.5.

FIGURE 1.2.1. Density of Ceratonova shasta in water samples collected at the Beaver Creek index site (KBC) from 2009 – 2017. Each point is the average of 3 x 1L 24-hr composite water samples. The upper graph is scaled to show the

maximum level in 2015.

32

FIGURE 1.2.2. Spatial and temporal density of Ceratonova shasta in water samples from two Klamath River mainstem index sites in 2017 – Beaver Creek (KBC) and Orleans (KOR). Data are mean number of spores in 3 x 1L 24-hr composite

water samples.

FIGURE 1.2.3. Spatial and temporal density of Ceratonova shasta in water samples collected at Klamath River mainstem index sites in 2017. Each data point is the mean of 3 x 1L 24-hr composite water samples. The Arcata Fish and

Wildlife Office Fish and Aquatic Habitat Conservation Program's predictive tool estimated that 80% of the natural juvenile Chinook Salmon had migrated downstream of the Kinsman trap site on the Klamath River sample week May 7 -

13; therefore, accounting for the court order's seven additional calendar days, the 80% outmigration 2017 date was May 20 (red line).

33

FIGURE 1.2.4. Average density of Ceratonova shasta in 24-hr composite water samples collected at the Beaver Creek index site (KBC) in 2017 (blue) compared with 2016 (gray). The Arcata Fish and Wildlife Office Fish and Aquatic Habitat

Conservation Program's predictive tool estimated that 80% of the natural juvenile Chinook Salmon had migrated downstream of the Kinsman trap site on the Klamath River sample week May 7 - 13; therefore, accounting for the court

order's seven additional calendar days, the 80% outmigration 2017 date was May 20 (red line).

FIGURE 1.2.5. Density of Ceratonova shasta in water samples from the Orleans index site in 2017 (green) compared with 2016 (grey). Data are mean number of spores in 3 x 1L 24-hr composite water samples.

34

Previous research by the Bartholomew Lab (Atkinson and Bartholomew 2010a, b) revealed that there are

multiple ITS-1 genotypes of C. shasta simultaneously present in the Klamath River. These genotypes

differentially cause disease in the various salmonid species: Type I causes mortality in Chinook salmon whereas

Type II can be fatal for coho salmon. Also, a 40% mortality threshold is reached for coho with Type II at 5

spores/L and for Chinook with Type I at 10 spores/L (Hallett et al. 2012). This year we published the results of a

study to understand mixed Type II/III infections – we determined that Types II and III should be regarded as a

single genotype because “III” was present as intra-spore variation in the gene, and is not relevant to the fish

disease questions we are analyzing in this project (Atkinson et al., 2018).

In 2017, at KBC waterborne levels of C. shasta were too low (< 1 spore/L) to genotype until mid-May. Genotype

I (Chinook) was dominant throughout May and June (<10 spores/L). Genotype II was also detected at KBC but

levels of Type II did not surpass 5 spores/L in either month. Genotype 0 was detected at KBC in 2 weeks in June

(< 3 spores/L).

TABLE 1.2.1. ITS-1 genotype of Ceratonova shasta in 1 L 24-hr composite or manual grab (g) water samples at the Beaver Creek (KBC) index site. Samples collected during Chinook salmon outmigration are before May 20.

Date

Average spores/L in 3 x 1L water samples;

most are 24-hr composite samples, (g) = manual “grab” sample

ITS-1 genotypes

04/03/2017 0 insufficient to genotype

04/10/2017 < 1 insufficient to genotype

04/17/2017 0 insufficient to genotype

04/24/2017 < 1 insufficient to genotype

05/01/2017 1 insufficient to genotype

05/08/2017 < 1 insufficient to genotype

05/15/2017 1 (g) 100% I

05/22/2017 1 100% I

05/29/2017 3 100% I

06/05/2017 4 70 - 100% I 0-30% II

06/12/2017 11 (g) 80-82% I 10-13% II 5-10% O

06/19/2017 4 (g) 60-70% I 30-33% II 0-7% O

06/26/2017 9 (g) 50-100% I 0-50% II

35

Water samples (4 x 1 L samples) were also collected manually at the start and end of each sentinel fish

exposure. These provide an indication of spore levels at one point in time, rather than over a 24-h period as for

the ISCO-collected samples.

TABLE 1.2.2. Density of Ceratonova shasta in 1 L water samples (average of 3 x 1 L) collected manually at the start and end of sentinel fish exposures. Red numbers indicate high levels of inhibition even when re-run diluted, hence parasite

levels are probably higher. X = no fish at this site / date

APRIL MAY JUNE SEPTEMBER

SITE FISH IN April

FISH OUT April

FISH IN May

FISH OUT May

FISH IN June

FISH OUT June

FISH IN Sept

FISH OUT Sept

WMR x x < 1 5 1 7 x x

WLD x x x x 3 8 x x

KED x x 4 14 50 20 x x

KI5 x x 1 1 2 < 1 x x

KBC 2 6 1 < 1 3 9 < 1 1

KSV 2 4 1 2 5 3 0 1

KOR x x 1 3 < 1 8 2 2

Task 1.3. Determine density and infection of the invertebrate (polychaete) host. Sampling will occur quarterly at the following polychaete index sites:

(1) Keno Eddy - RKM 369 (2) JC Boyle - RKM 366 (3) I5 Bridge Fish Trap - RKM 287 (4) Tree of Heaven – RKM 281 (5) Beaver Creek – RKM 258 (6) Seiad Valley – RKM 207 (7) Orleans Dolan’s Bar River Access – RKM 90

Polychaete collections will occur according to the following schedule:

(1) Winter - prior to peak flow, typically March (2) Spring – after peak flow, typically June (3) Summer – during peak period of polychaete density, July/August (4) Fall – after beginning of adult returns, October/November

Task 1.3. Overview The aim of this task is to describe demography and prevalence of C. shasta infection in polychaete populations

at index sites in the Klamath River during each season of the year. The index sites overlap with water and fish

monitoring sites and are stratified along the discharge gradient. Monitoring populations in the spring and fall is

important because they overlap with peak juvenile salmon outmigration (spring) and adult salmon returns

(fall), winter is important for understanding the dynamics of C. shasta infection in this host and summer is

important for understanding polychaete host population dynamics. Our specific objectives are to describe the

density of Manayunkia sp. populations, to survey populations for prevalence of C. shasta infection, and to

examine relationships among these factors and the environments of seven sites on the Klamath River.

36

Task 1.3. Methods Polychaetes were collected four times per year in winter (mid-March), spring (May/June), summer

(July/August), and fall (October). We added an additional sample period in late March following a high

magnitude flood event because we were unable to sample all index during the first March sampling period due

to high flows. Polychaete samples were collected by targeting previously identified polychaete assemblages at

seven sites; from upstream to downstream these include Keno (KED), the Boyle bypass reach (KJB), I-5 bridge

(KI5), Tree of Heaven Campground (KTH), Fisher’s RV park near Beaver Creek (KBC), Seiad Valley (KSV), and

Dolan’s Bar near Orleans (KDB) (Figure 1.3.1). Three samples were collected at each site with a modified Hess

sampler (a T section of PVC pipe with a base opening of 229 cm2, fitted with an 84 µm collection net) and a

scraping device. Samples were preserved in 70% EtOH in the field and returned to the laboratory (J.L. Fryer

Aquatic Animal Health Laboratory, Oregon State University, Corvallis, OR) for processing. All samples were

subsampled by placing the entire sample into a sorting tray (20 cm x 28 cm, Wildco, FL) and randomly selecting

three 25 cm x 25 cm subsamples. Subsamples were stained (20% Rose Bengal, Fisher Scientific) and all

polychaetes in each subsample were counted using a dissecting microscope (20 - 50 X magnification).

FIGURE 1.3.1. Locations of monitoring sites from upstream (KED) to downstream (KOR) shown by black circles, USGS discharge gages (green circles). Sites, in order from upstream to downstream, include KED, KJB in the JC Boyle bypass reach, KI5 near the I5 overpass, KTH near the Tree of Heaven Campground, KBC near the confluence of Beaver Creek

and the mainstem river (Fisher’s RV park), KSV near Seiad Valley, and KOR at Dolan’s Bar fishing access near Orleans CA.

Polychaete density: Subsample counts were adjusted to account for misidentified specimens and missed

(progeny and immature) polychaetes that were observed in the samples. Adjusted polychaete density was

calculated as [(adjusted count/# subsamples)/(grid cell area)x(tray area)/Hess area] and expressed per m2 for

each sample.

Prevalence of infection and estimated densities of infected polychaetes: Prevalence of C. shasta-infection in

polychaetes was determined using polychaetes collected for density estimates (see above). Up to 400

37

polychaetes per sample, or as many as were available if fewer than 400, were prepared for DNA extraction and

tested for C. shasta infection by qPCR (Hallett and Bartholomew 2006).

Task 1.3.1 Results and Discussion Polychaete density and prevalence of infection (POI): Assays have been completed for all collections. We were

unable to complete winter collections due to a high flow event that occurred throughout the basin. We

sampled as soon as flows came down to safe levels (May for KED, KJB, KI5, KTH, and KBC, and June for KOR and

KSV). Overall, polychaete densities were highest in the upper basin sites (KED and KJB) and lowest in the lower

basin sites (KSV and KOR) (Figure 1.3.2). At sites in the upper basin, maximum densities were similar to

densities in 2016, densities tracked water temperatures, increasing in spring, and maxing out in summer and

fall. This trend contrasts with polychaete density dynamics observed at these sites in previous years (including

2016), which have historically been relatively stable. We suggest the magnitude of the 2017 disturbance may

explain the difference, and could be important for population dynamics downstream. At sites in the middle

basin, including KBC, KTH and KI5, polychaete densities were similar to those observed in 2016, following a flow

event that was similar in magnitude and timing of peak flow but of much lower duration. This result may

suggest that polychaete densities responses are driven by the timing and magnitude rather than duration.

However, prevalence of infection data (see below) suggest flow duration is also important in the context of

disease risk for salmon. At sites in the lower basin, including KSV and KOR, polychaetes were present at some

of the lowest densities in summer and fall that we have observed in the past decade. Polychaete densities at

these sites typically peak in fall, but in 2017, densities peaked in summer and dropped off in fall. We observed

dramatic substrate deposition at these sites following the winter/spring flow event, which may have reduced

polychaete habitat and resulted in lower densities overall in fall as juveniles dispersed if they were unable to

move into suitable habitat.

Results of infection assays (Table 1.3.1): Overall, prevalence of infection was low in 2017: 0.125% of all

polychaetes processed tested positive for C. shasta. Prevalence of infection was highest at KTH in spring (0.1%),

KI5 in summer (0.7%) and KOR in fall (0.6%). In previous years, we have always detected C. shasta in >1% of the

population (Table 1.3.2). When prevalence data were applied to polychaete densities to extrapolate densities

of infected polychaetes (Figure 1.3.3), we observed low densities of infected polychaetes at all sites, with the

exception of KI5 in summer (after the majority of salmonids had out-migrated). Interestingly, we also observed

low densities of infected polychaetes at KOR and KSV in fall compared with previous years. This suggests i) low

numbers of myxospores were available for infection in summer and fall, and ii) low numbers of polychaetes

were releasing actinospores in fall.

38

FIGURE 1.3.2. Densities of Manayunkia sp. (circles), the polychaete host of C. shasta at 7 monitoring sites in 2013 - 2017, including discharge (blue), and water temperature (black lines), from top to bottom monitoring sites include KED (grey circles) and KJB (black circles) in the upper basin (top plot), KI5 (black circles), KTH (grey circles) and KBC (white circles) in the mid basin (middle plot), and KSV (black circles) and KOR (grey circles) in the lower basin (lower plot). Discharge (right plots, denoted in blue) and water temperature (denoted by red solid line) were obtained from USGS gaging stations and OSU temperature loggers deployed near the sampling sites.

39

TABLE 1.3.1. Prevalence of Ceratonova shasta in Manayunkia sp. at 7 monitoring sites in 2017. NS = not sampled due to unsafe flows.

Prevalence (%) of C. shasta infection in polychaetes by site and month

Site Early March May Mid-June Late July/early

August Mid-October

KOR NS NS 0 0 0.6

KSV NS NS 0 0.3 0

KBC NS 0 0 0.3 0

KTH NS 0 0.1 0.4 0

KI5 NS 0 0 0.7 0

KJB NS 0 0 0.3 0

KED NS 0 0.3 0 0

TABLE 1.3.2. Summary of seasonal prevalence of Ceratonova shasta in Manayunkia sp. at 7 monitoring sites in from 2013 - 2017. NS = not sampled due to unsafe flows. Lower river sites include KOR and KSV, middle sites include KBC,

KTH, and KI5, and upper sites include KED and KJB.

Year Season Lower Middle Upper

2013 winter 5.1 0 0.2

spring 0.8 0 0

summer 1.0 0.3 0.6

fall 0.9 1.6 0.3

2014 winter 4.4 6.2 2.3

spring 0.8 1.1 0

summer 0.7 1.5 0.5

fall 2.4 1.1 0

2015 winter 1.3 2.0 0.2

spring 0.7 1.5 0.3

summer 0.2 0.8 0.2

fall 3.7 2.6 0.3

2016 Winter-mid 0 2.8 0.1

Winter-late 0 0.3 0

spring 1.1 1.0 0

summer 0.9 0.4 0

fall 2.6 0.4 0.3

2017 winter ns 0.5 0.1

Spring-mid 0 0 0.1

Spring-late 0 0.1 0.2

summer 0.2 0.5 0.2

fall 0.3 0 0

40

Density, prevalence of infection and the environments of monitoring sites: In contrast to 2015, when infected

polychaetes were detected at all but three site sampling period combinations [including KOR (June), KSV (late

July), and KED (Feb)], our molecular assays show we detected infected polychaetes less frequently in 2016

(Table 1.3.1). In 2016, prevalence of infection was generally detected at levels < 1.0%, which also contrasts with

results in 2015, when infection prevalence was generally > 1.0%. The 2016 results are similar to several

previous years when disease risk to fish was lower. For example, from 2010 - 2013 the distribution of infected

polychaetes was limited and infected polychaetes were detected at few sites and were characterized by low

infection prevalence. In contrast, in 2014 - 2015, when disease risk to fish was high, infected polychaetes were

detected at most sites and in most sampling months. Densities of infected polychaetes varied among sites and

sampling periods. Prior to the high flow event in March (“late winter”), densities of infected polychaetes were

highest at KI5 (mean of 31,539 individuals m-2, Figure 1.3.3). Following the flow event, densities of infected

polychaetes were low at all sites until spring (June), when the highest densities of infected polychaetes were

detected at KTH and KOR (mean densities ranged from 180 - 226 individuals m-2). The highest densities of

infected polychaetes were thereafter detected at KOR (over 5,000 individuals m-2 in summer and > 35,000

individuals m-2 in fall). This result indicates a significant shift in the distribution of infected polychaetes

compared to data collected in previous years. In addition, the result corroborates water sampling data from

2016, which demonstrated high levels of C. shasta DNA proximal to Orleans (KOR). Analyses examining

relationships among water year, fish disease risk, and polychaete data (density, prevalence of infection, and

density of infected polychaetes) from 2010 - 2016 are in progress. We have built preliminary predictive models

with data collected from 2010 - 2015, and plan to test the models using data we collected from 2006 - 2009

when possible, and 2016 - 2018.

41

FIGURE 1.3.3. Densities of Manayunkia sp. (circles) infected with Ceratonova shasta at 7 monitoring sites in from 2013 - 2017, including discharge (blue), and water temperature (black lines), from top to bottom monitoring sites include KED

(grey circles) and KJB (black circles) in the upper basin (top plot), KI5 (black circles), KTH (grey circles) and KBC (white circles) in the mid basin (middle plot), and KSV (black circles) and KOR (grey circles) in the lower basin (lower plot).

Discharge (right plots, denoted in blue) and water temperature (denoted by red solid line) were obtained from USGS gaging stations and OSU temperature loggers deployed near the sampling sites.

42

Objective 2. Comply with 2013 Biological Opinion for Reclamation’s operation of the Klamath Project to

ensure weekly monitoring of actinospore genotype II concentrations in the mainstem Klamath River

immediately upstream of Beaver Creek mid-April to June, and expedite analysis and data dissemination.

Task 2.1. Genotype water samples collected weekly for 7 weeks from April 14 or a date within 6 days prior to April 14 through the first full week of June.

Data are shared under Task 1.2.

Objective 3. Determine whether a pulse flow can affect C. shasta densities and infection risk in fishes.

Task 3. Monitor Ceratonova shasta in association with pulse flow events - before, during and after an event

using the following metrics: infection prevalence and severity in sentinel fish, density of waterborne parasite,

infection prevalence in polychaetes and density of the polychaete host.

Task 3.1. Collect water samples daily at Beaver Creek and Seiad Valley index sites, beginning three days prior to the event and ending ~ three days after the event. Samples will be collected every 2 h using automated samplers, and pooled to make a 6 h composite sample that is assayed using a C. shasta-specific qPCR.

Task 3.2. Expose ~30 sentinel IGH Chinook and coho salmon juveniles at Beaver Creek and Seiad Valley index sites for 72 h before and during an event. After exposures, fish will be transported to the Salmon Disease Laboratory, reared at 18 °C water temperature and monitored for C. shasta infections for 60 days as described in Task 3.1.1., above.

Task 3.3. Collect polychaetes from 4 sites before and after a pulse flow. Three Hess samples each will be collected from a reference site (KN) upstream of Iron Gate Dam, and from both fine and coarse sediments at long term monitoring sites, Trees of Heaven, Beaver Creek and The Grange. Samples will be preserved and returned to the laboratory for density and infection assays, as described in Task 1.3., above.

No pulse flow event occurred in 2017.

Objective 4. Validate the index for predicting disease severity for Chinook and coho salmon by correlating

data on infection prevalence and disease severity in each fish species with genotype-specific spore densities

in water collected at each site.

Task 4.1. Develop a method for high throughput genotyping of C. shasta by identifying a genetic locus for distinguishing among the 4 C. shasta ITS genotypes and between the 2 genotype II biotypes of C. shasta.

and

Task 4.2. Use these loci to develop a method for high throughput genotyping of C. shasta.

43

Tasks 4.1 and 4.2 Methods, Results and Discussion Using the PCR primers that we developed in 2016 to target the C. shasta mitochondrial COX locus, we assayed

a test panel of fish and water samples that represented genotypes O, I, II, with different “sub-genotypes” of

each (where the gene sequences have variations, outside the primary ATC genotyping region; Atkinson &

Bartholomew 2010b). Our initial goals were to determine if the COX locus could not only resolve the same

genotypes as the existing ITS-1 assay, but also could give us a molecular signature of II from coho that we could

base the high-throughput assay on. The results from several trials with two different sets of COX primers that

we developed were conclusive: the C. shasta sequence in this region is too variable to be used for the

development of a Coho-specific assay, and no single set of COX primers was able to amplify all reference

samples. These results are in concordance with our most recent analyses of the genome and transcriptome

data for other genes (i.e. that there is much genetic variation in C. shasta).

Therefore, we tried a different approach: we collated the ~60 coho-specific sequence data that we have

generated to date (not water or rainbow trout data, which added significant ‘noise’ to the dataset), then

aligned them with each other, and assessed how much variation was present in coho alone. There was very

little. This meant that we could choose genetic regions that were the same for all coho, but differed from both

genotype I from Chinook and genotype O from steelhead. We designed qPCR primers and a novel qPCR assay

using SYTO9 as the fluorescent reporter. We tested this assay on samples from Chinook, coho and steelhead,

and only the C. shasta-genotype in coho was detected (Figure 4.1.1). We then tested samples where we

artificially mimicked the mix of genotypes typically present in the lower Klamath River: with up to 10 times as

much Chinook genotype I relative to the coho genotype II; and again could successfully detect the coho signal

with this new assay.

We plan to test this assay’s ability to quantify genotype II in the 2018 water samples, in parallel with our

traditional approach to amplify and sequence C. shasta to identify genotype. This new assay holds promise as a

way forward for rapid and sensitive detection of genotype II in Klamath River water samples (in the presence of

the other genotypes that can episodically ‘drown out’ the coho-relevant signal).

FIGURE 4.1.1 Photograph of an agarose gel showing results of

genotyping polymerase chain reaction (PCR) tests. A range of fish

and water samples with different Ceratonova shasta genotypes

(marked) were PCR-tested by general C. shasta genotyping primers

(top) and new coho (II)-specific primers (bottom). Positive samples

are indicated by bright bands in the gel images. Results show that

whereas the general assay amplifies all C. shasta genotypes (as

intended), the new primers are specific to genotype II in fish and

water samples.

44

Publication update: We published our finding that genotypes "II" and "III" are actually just genetic variants of

the same strain of C. shasta (and should therefore no longer be regarded as distinct genotypes; Atkinson et al.,

2018).

Task 4.3. Correlate genotype-specific parasite density with infection prevalence and severity in Chinook and coho salmon.

This analysis is underway.

Objective 5. Validate and refine the epidemiological model to identify sensitive parameters in the host-

parasite life cycle, simulate the effect of potential management strategies on the different stages of the life

cycle, and predict disease severity in juvenile salmonid populations under different parasite densities, water

temperatures and flows.

Task 5.1. Investigate data gaps that are defined as the model is further developed.

Task 5.1.1. Epidemiological model development and progress

We continued work on the epidemiological model in 2017. We completed experiments on effects of water

temperature on C. shasta infection prevalence in polychaetes and these data are being used to parameterize

the epidemiological model. We have several experiments planned for 2018 to help us to determine if we can

continue to use a general model for all C. shasta genotypes and the spatial extent, or if models will need to be

developed for each genotype and site.

Task 5.1.2. Spore transport and water temperature models

Flow modification has the potential to affect C. shasta because of impacts on discharge and temperature.

The parasite is fully aquatic, and has non-motile, waterborne infectious stages (myxospores and actinospores)

that alternate between an invertebrate host (polychaete) and salmonids. Annual prevalence of C. shasta-

infection in out-migrating juvenile Chinook salmon has been estimated at up to 91 percent in some years (True

et al., 2016). High mortality in juvenile salmonids has been observed following low magnitude peak winter and

spring flows when river temperatures reach 15 - 18 °C in late spring (Bartholomew and Foott, 2010; Hallett et

al., 2012). In contrast, reduced C. shasta infection and related mortality in salmonids have been observed

following high magnitude winter and spring discharge (True et al., 2011; Hallett et al., 2012). The addition of

cool, parasite-free water from Iron Gate Dam (the lowermost dam) to the mainstem Klamath River may be an

effective management action when disease risk to fish is high. Water temperature is correlated with disease

progression and mortality in infected fish hosts (Ray et al., 2012). Parasite density is also correlated with

mortality; 5 and 10 spores per L of river water cause > 40% mortality in Klamath River Coho and Chinook,

respectively (Hallett et al., 2012). Thus, managed water releases could reduce disease risk to fish by

reducing downstream river water temperature or by diluting waterborne parasite stages.

Knowledge of water age, or the travel time required for a parcel of water released from the dam to reach

specific location, is necessary for predicting how the timing and volume of water released will affect parasite

density and water temperature within high risk for fish zones located downstream. In addition, water

age is relevant for (i) determining timing and volume of water to be released to achieve decreases in

parasite density or water temperature, and (ii) designing monitoring plans to measure effects on disease

45

risk for fish (e.g. measure parasite density before, during and after a managed water release event). Thus, a

model predicting water temperature and age would be useful for managers considering scenarios of flow

manipulation strategies.

The following manuscript has been published in the Journal of Hydrology

(https://www.sciencedirect.com/science/article/pii/S0022169418301203):

Javaheri, A., Babbar-Sebens, M., Alexander, J., Bartholomew, J., and S. Hallett (2018). Global sensitivity analysis

of water age and temperature for informing salmonid disease management. Journal of Hydrology 561: 89-97.

Objective 6. Investigate the occurrence of C. shasta below the Trinity River confluence and ascertain spore

type of waterborne stages. Tribal biologists will assist with the following tasks.

Task 6.1. Conduct sentinel fish exposures when parasite abundance exceeds 10 spores/L but temperatures are not lethal for salmon.

Water temperature was too high for sentinel fish exposures.

Task 6.2. Quantify parasite levels in water samples. Details of this task were shared earlier, under Task 1.2.

Task 6.3. Characterize density and infection of the invertebrate (polychaete) host in the mainstem downstream from the Trinity River confluence.

This aspect of the research requires that we collaborate with the Yurok tribe for access to relevant river sections. We planned this research for August 2016, but high water temperatures resulting in an Ich outbreak prevented the Yuroks from assisting during this period. We subsequently planned this research for August 2017, but high flows during the winter and spring resulted in high amounts of substrate mobilization and habitat change in this reach. We are working with the Yurok tribal biologists to determine an appropriate time to sample this reach for polychaetes in 2018 to ensure it occurs on schedule.

Objective 7. Develop and validate predictive models for polychaete hosts including distribution, density,

infection, and recolonization rates under different peak discharge, water years, and dam removal scenarios.

Task 7.1. Validate and refine the polychaete distribution model for predicting distribution under different peak discharge, water years, and dam removal scenarios.

We developed and published a tandem modeling approach to predict the distribution of Manayunkia sp. (Alexander et al. 2016, Freshwater Science) using data collected in 2012-2013. Samples collected in 2012 were used to build the statistical (predictive) model estimating the relationship between physical habitat characteristics and the distribution of Manayunkia sp. under the context of peak discharge. We have subsequently collected data to refine the predictive model to improve predictions under different peak discharge scenarios including i) low peak. We plan to complete data collection again in 2018 to further refine the model and improve predictive accuracy. The data collection planned for 2018 is particularly important because it will allow us to disentangle flow legacy, discharge duration, and discharge magnitude and evaluate

46

their effects on polychaete host distribution. Samples collected in 2014-2017 are used for model validation. The 2016 and 2017 datasets demonstrate model predictions match the majority of our observations ((Figure 7.1.1) but fail to predict well on fine substrates in high peak flow water years. We attribute this to substrate mismatch and are working to update the substrate map in 2018. Consequently, we have limited our predictions to a narrow range of peak discharges: We predicted the distribution of Manayunkia sp. under alternate flow scenarios, 1,200cfs and 7,950 cfs, to simulate dry and wet water years, respectively. Results suggest that manipulating the hydrograph could influence distribution of polychaete hosts (Figure 7.1.2).

47

FIGURE 7.1.1. Predicted and observed polychaete distributions under the 2017 modeled peak discharge scenario. The polychaete distribution model predicted low probabilities of occurrence at the majority of locations where we did not observe polychaetes, and high probabilities of occurrence at the majority of locations where we observed polychaetes.

Overall, our observations support the model predictions but still require some refinement, likely due to substrate mismatch.

48

Task 7.2. Add polychaete density and infection prevalence data to the predictive model to examine how flow regimes will affect density and infection in this host.

We developed and published (Alexander et al. 2016, Freshwater Science) a tandem modeling approach to predict the distribution of Manayunkia sp. and evaluate the effects of discharge scenarios in sections of the Klamath River. Two-dimensional hydraulic models (2DHM) were built for three river sections using topographic survey data, water surface elevation profiles, stage-discharge relationships, and spatial maps of substrate (Wright 2014). The 2DHMs were used to describe hydraulic variation and stratify sampling locations across depth velocity gradients within substrate classes. Benthic samples were collected in July 2012, 2013, 2014, 2016 and 2017. Validation of the model predictions using real data collected at both the low and high peak discharge scenarios are in progress, and are needed before we can evaluate whether manipulation of the hydrograph may in turn influence prevalence of C. shasta and disease in salmonids, but the preliminary results are very exciting. We ran simulations of the peak flood event that occurred in March 2016 (>10,000 cfs from IGD) and examined predictions against empirical data. Although the 2D models had difficulty running at discharge >10,000, the predictive model performed well, predicting within 90% accuracy among sites. We completed assaying 3,500 polychaetes from each year of model validation (2014, 2016, 2017) for infection with C .shasta. Prevalence varied from 1.9% (2014) to 0.3% (2017). Analyses to examine relationships between spatial distribution of infected polychaetes and peak discharge are ongoing. Models will be refined in 2018, and submitted for publication once complete.

FIGURE 7.1.2. Modeled effects of peak discharge on the probability of polychaete presence at x, y locations in the Tree of Heaven Study reach. Predicted polychaete distributions under two modeled peak discharge scenarios including a dry

water year having a peak discharge of 1,200 cfs out of Iron Gate Dam (left) and a wet water year having a peak discharge of 7,950 cfs (right).

49

Task 7.3. Estimate polychaete recolonization rates. Use the physical models to characterize hydraulic conditions before and after disturbance. Predict polychaete distribution using the refined distribution model. Validate with empirical data where available.

The polychaete distribution model developed and validated for predicting distribution and density (7.1 and 7.2)

and four locations we have monitored for recolonization since 2010 (“recolonization monitoring sites”) are

providing data for this task. We completed data collection in 2017 and found model predictions to be highly

accurate (Figure 7.3.1), and plan to collect data again in 2018 to assess the effects of flow legacy. Polychaete

tubes were observed at the reach 1 pool and eddy sites and the reach 2 pool and eddy sites in 2015 but not in

2016 or 2017. These habitats will continue to be monitored in 2018 to determine if recolonization occurs, and

on what timeframe.

FIGURE 7.3.1. Observations of polychaete assemblages in 2014 (low peak discharge) and 2017 (high peak discharge) on a georeferenced sampling location at the Tree of Heaven Study reach. The polychaete distribution model predicted low probabilities of polychaete occurrence at this location in 2017 given the hydraulic conditions during peak discharge, and

our observations support the model predictions.

50

Objective 8. Develop and synthesize a dataset, encompassing environmental risk factors and their

relationship with polychaete host ecology, to facilitate predictions about how polychaete densities and

infection levels may change under future climate and temperature regimes.

Task 8.1. Synthesize an “environmental risk factor” dataset comprised of water quality data and future predictions for water temperature and discharge for examining correlations with polychaetes.

We have completed examining correlations between water temperature, discharge, and polychaete risk factors, and have developed preliminary predictive models. The results are being used as inputs for the epidemiological model and to guide laboratory experiments to address outstanding questions relevant to polychaetes.

Task 8.2. Examine correlations between environmental risk factors and polychaete host data including density, population structure, and infection prevalence.

Preliminary correlations have been completed for 2010-2017. We plan to validate these relationships using data collected 2006-2009.

Task 8.3. Construct models for generating predictions about how polychaete densities and infection levels may change under future climate and temperature regimes, and how these changes in turn may affect disease risk in salmon hosts.

We have linked a series of Klamath River models including i) a fine-scale climate change model to predict future

stream temperatures and discharge (Perry et al. 2011), ii) a 2-D hydraulic model coupled with a statistical

model to predict changes in polychaete populations under different river discharge scenarios (Task 7, above),

iii) a degree-day model to predict the potential number of generations per year under different thermal

regimes (Chiaramonte 2013), and iv) the epidemiological model to quantify the risk of disease in the salmon

host under the different climate scenarios (Ray 2013, Task 5) (Figure 8.3.1).

The models and their outputs predict changes in disease severity in salmon as a result of C. shasta infection

under different climate scenarios. We are examining predictions as a function of change to a baseline (2010)

using past years (hindcasting) and long-term data for on the intensity and distribution of C. shasta infections in

juvenile salmon. The next steps will include using future climate predictions to predict the effects of climate

change on temperature and precipitation on the phases of the C. shasta life cycle involving Manayunkia sp.

51

FIGURE 8.3.1. Model ensemble includes outputs from global circulation models, which provide inputs for a basin scaled model, which provides outputs of future stream temperatures and discharge. These parameters are used as predictions for 2-D hydraulic models that are paired with a predictive polychaete model and a degree-day model, and these models

provide inputs for the epidemiological model, which is used to quantify changes in risk of disease in the salmon host under the different water year scenarios.

52

Objective 9. Regularly disseminate research findings to provide stakeholders, managers, researchers and the

general public ready access to current information and historical datasets pertinent to C. shasta in the

Klamath River.

(a) Preliminary Result Summaries: The contractor will provide brief preliminary summary information to Reclamation on a monthly basis each field season or as-requested by Reclamation. Additionally, preliminary findings may be available in the form of a professional presentation at a meeting with Reclamation and other state, federal, and tribal agencies.

(b) Annual Reports: The contractor will provide Reclamation an annual report of research for this study, due March 31 each contract year. This report will include a description of the study questions, methods of data collection and analyses, results of data analyses, and a discussion of the significance of the data. Draft copies of the annual report of research will be distributed to Reclamation and other interested parties for review before the report is finalized.

(c) Website: maintained by the contractor for dissemination of results and project information to the public.

(d) Annual Klamath River Fish Health Workshop: public forum to review results of disease research, and will be coordinated by the contractor.

(e) Annual project coordination meeting: with project collaborators; coordinated by the contractor. (f) Publications: submit findings for publication in peer-reviewed scientific journals. (g) Final report: the final report that summarizes and synthesizes the multi-year findings will be submitted

by Dec 31 2019. Research summaries were provided to a broad range of audiences as requested, at professional meetings, during conference calls and online. This document serves as the 2017 annual report. Data are posted on the following website weekly during salmonid outmigration: http://microbiology.science.oregonstate.edu/content/monitoring-studies. The annual Klamath River Fish Health Workshop and management meetings were scheduled for April 4, 2018, but were rescheduled due to a conflict with a court case. The meetings are planned to occur in summer 2018. The following articles were published:

Atkinson S.D., Hallett S.L., Bartholomew J.L. (2018) Genotyping of individual Ceratonova shasta (Cnidaria: Myxosporea) myxospores reveals intra-parasite ITS-1 variation and invalidates the distinction of genotypes II and III. Parasitology https://doi.org/10.1017/S0031182018000422 Javaheri A., Babbar-Sebens M., Alexander J.A., Bartholomew J.L., Hallett S.L. (2018) Global sensitivity analysis of water age and temperature for informing salmonid disease management. Journal of Hydrology 561: 89–97. https://doi.org/10.1016/j.jhydrol.2018.02.053

Acknowledgements The Karuk and Yurok tribal biologists assisted with water sample collection and filtration. Jamie Graen, Emily Lawrence and Sophia Jadzak (OSU) assisted with molecular procedures. California Department of Fish and Game (Keith Pomeroy and crew at the Iron Gate Hatchery) provided Klamath River fall Chinook. The Trinity River Hatchery provided coho salmon juveniles for our sentinel studies. Roaring River Hatchery, Oregon Department of Fish and Wildlife, Scio, provided susceptible rainbow trout. Demian Ebert (Pacific Power – Hydro Resources) and Tony Dennis (Mid Klamath Watershed Council) helped source adult coho. We are grateful to land owners who allowed us access to conduct the sentinel studies and water sampling: The Nature Conservancy (Klamath Falls) and Lonesome Duck Resort (Chiloquin), OR; The Sportsman’s Park Club near Keno, OR; Fisher’s RV Park at Klamath River, CA; Wally Johnson, Seiad Valley; Sandy Bar Resort, Orleans, CA.

53

References

Alexander, J. D., J. L. Bartholomew, K. A. Wright, N. A. Som, and N. J. Hetrick. 2016. Integrating models to predict distribution of the invertebrate host of myxosporean parasites. Freshwater Science 35: 1263-1275. doi: 10.1086/688342.

Atkinson and Bartholomew (2010a) Disparate infection patterns of Ceratomyxa shasta (Myxozoa) in rainbow trout Oncorhynchus mykiss and Chinook salmon Oncorhynchus tshawytscha correlate with ITS­1 sequence variation in the parasite. International Journal for Parasitology, 40(5):599­604

Atkinson and Bartholomew (2010b) Spatial, temporal and host factors structure the Ceratomyxa shasta (Myxozoa) population in the Klamath River Basin. Infection, Genetics and Evolution 10:1019­1026

Atkinson SD, Hallett SL, Bartholomew JL (2018) Genotyping of individual Ceratonova shasta (Cnidaria: Myxosporea) myxospores reveals intra-parasite ITS-1 variation and invalidates the distinction of genotypes II and III. Parasitology. (In press March 2018)

Chiaramonte LV (2013) Climate warming effects on the life cycle of the parasite Ceratomyxa shasta in salmon of the Pacific Northwest. Master’s thesis Oregon State University.

Hallett SL, Bartholomew JL (2006) Application of a real-time PCR assay to detect and quantify the myxozoan parasite Ceratomyxa shasta in river water samples. Diseases of Aquatic Organisms 71:109.

Hallett SL, Ray RA, Hurst CN, et al. (2012) Density of the waterborne parasite Ceratomyxa shasta and its biological effects on salmon. Applied and Environmental Microbiology 78:3724–3731.

Perry RW, Risley JC, Brewer SJ, et al. (2011) Simulating daily water temperatures of the Klamath River under dam removal and climate change scenarios. U. S. Geological Survey.

Ray RA, Bartholomew JL (2013) Estimation of transmission dynamics of the Ceratomyxa shasta actinospore to the salmonid host. Parasitology 140:907-916.

Wright KA, Goodman, DH, Som NA, and Hardy TB (2014) Development of two-dimensional hydraulic models to predict distribution of Manayunkia sp. in the Klamath River. U. S. Fish and Wildlife Service. Arcata Fish and Wildlife Office, Arcata Fisheries Technical Report Number TP 2014-19, Arcata, CA.

This publication was funded, in part or whole, by the Bureau of Reclamation (Reclamation), U.S. Department of Interior. Funding was provided by Reclamation as part of it mission to manage, develop, and protect water and related resources in an environmentally and economically sound manner in the interest of the American public. Funding was provided through Interagency Agreement # R15PG00065. The views in this report are the authors’ and do not necessarily represent the views of Reclamation.