A Comparison of Nonlethal and Destructive Methods for Broad-Based Infectious Agent Screening of Chinook Salmon Using High-Throughput qPCR

Abstract

Molecular tools such as high‐throughput quantitative polymerase chain reaction (HT‐qPCR) are useful for monitoring multiple infectious agents in wild animal populations (i.e., broad‐based screening). If destructive tissue samples cannot be obtained due to experimental design requirements (e.g., bio‐telemetry; holding with repeated biopsy) or the conservation status of host species, non‐lethally sampled tissues can be substituted. However, infection profiles have been found to differ between non‐lethally and destructively sampled tissues. We present a comparative analysis of two non‐lethal (gill, blood) and one destructive (pool of internal and external) approach for broad‐based infectious agent screening of adult Chinook Salmon (Oncorhynchus tshawytscha). Of a possible forty‐seven agents, sixteen were detected overall. Results indicate moderate differences in infection profiles among tissues, with limitations of each tissue type relevant to the ecology of each agent. Gill was the most comprehensive screening tissue, as more infectious agents were detected overall in gill (16) than in blood (12) or multi‐tissue pools (15). Agreement in estimated agent prevalence between tissue types ranged from poor to excellent, while overall agent community structure (combined prevalence of all agents) showed low agreement between tissue types. Two agents occurred at 100% prevalence in all tissue types. Nine agents, including bacteria and gill parasites, had higher prevalence in gill than blood, while five agents, including one virus and several microparasites, were better detected in blood. Future studies should pair microscopy and histopathology with HT‐qPCR to better characterize host health and disease development relative to molecular detection of agents across tissue types. This article is protected by copyright. All rights reserved.

Full text

ARTICLE
A Comparison of Nonlethal and Destructive Methods for Broad-Based
Infectious Agent Screening of Chinook Salmon Using High-Throughput
qPCR
Amy K. Teffer*
Department of Biology, University of Victoria, Post Office Box 1700, Station CSC, Victoria,
British Columbia V8W 2Y2, Canada
Kristina M. Miller
Fisheries and Oceans Canada, Molecular Genetics Section, Pacific Biological Station, 3190 Hammond Bay Road,
Nanaimo, British Columbia V9T 6N7, Canada
Abstract
Molecular tools, such as high-throughput quantitative polymerase chain reaction (HT-qPCR), are useful for moni-
toring multiple infectious agents in wild animal populations (i.e., broad-based screening). If destructive tissue samples
cannot be obtained due to experimental design requirements (e.g., bio-telemetry; holding with repeated biopsy) or the
conservation status of host species, then nonlethally sampled tissues can be substituted. However, infection profiles have
been found to differ between nonlethally and destructively sampled tissues. We present a comparative analysis of non-
lethal (gill and blood) and destructive (pool of internal and external tissue) approaches for broad-based infectious agent
screening of adult Chinook Salmon Oncorhynchus tshawytscha. Of a possible 47 agents, 16 were detected overall by
nonlethal and destructive methods. Our results indicated moderate differences in infection profiles among tissues, with
limitations of each tissue type dependent on the ecology of each agent. The gill was the most comprehensive screening
tissue, as more infectious agents were detected overall in gill (n=16) than in blood (n=12) or multi-tissue pools (n=
15). The agreement in the estimated agent prevalence between tissue types ranged from poor to excellent, while overall
agent community structure (the combined prevalence of all agents) showed low agreement between tissue types. Two
agents occurred at 100% prevalence in all tissue types. Nine agents, including types of bacteria and gill parasites, were
more prevalent in gill than in blood, while five agents, including one virus and several microparasites, were more preva-
lent in blood. Future studies should pair microscopy and histopathology with HT-qPCR to better characterize host
health and disease development relative to molecular detection of agents across tissue types.
A broad array of infectious agents are carried by wild
animals, and the dynamics of this community can strongly
influence a host's physiology and ecology (Altizer et al.
2011; Johnson et al. 2015). Infectious agents can cause
host impairment or disease under suboptimal environmen-
tal conditions; subsequently, unbalanced host-pathogen
relationships can affect the evolution of virulence (Wolin-
ska and King 2009; Engering et al. 2013). Beyond
single-agent effects, disturbances to the infectious agent
community structure within hosts can reduce host survival
odds by disrupting both host-pathogen and pathogen-
pathogen relationships (Sofonea et al. 2017). Multiple
infections are common in wild animals, prompting the
inclusion of multiple infectious agents in monitoring pro-
grams (Bordes and Morand 2011). Early detection of
pathogens (or signs of infection) in wild animals can be
*Corresponding author: akteffer@gmail.com
Received March 20, 2019; accepted July 20, 2019
Journal of Aquatic Animal Health 31:274–289, 2019
©2019 American Fisheries Society
ISSN: 0899-7659 print / 1548-8667 online
DOI: 10.1002/aah.10079
274

predictive of host survival, reproductive success, and dis-
ease development (e.g., Korpimaki et al. 1993; Telfer et al.
2005; Miller et al. 2011; Jeffries et al. 2014; Cikanek et al.
2015). Monitoring programs that include multiple infec-
tious agents can therefore provide vital information to
resource managers to better predict potential disease con-
sequences of increased environmental and anthropogenic
stressors in a multi-agent context.
Although destructive sampling is often used in monitor-
ing programs, nonlethal methods for broad-based
infectious agent screening are increasingly needed. Conser-
vation initiatives directly benefit from the development,
application, and assessment of nonlethal methods by
reducing the need for destructive sampling (e.g., for spe-
cies of conservation concern). Tools that can simultane-
ously screen many hosts for an array of infectious agents
in a cost-effective manner are especially useful to quantify
pathogen dynamics in the wild. The BioMark high-
throughput quantitative polymerase chain reaction (HT-
qPCR) platform (Fluidigm, San Francisco, California),
initially developed for transcriptomic applications, has
been validated against traditional qPCR as a screening
tool to rapidly quantify infectious agents in wild salmon
(Miller et al. 2014, 2016). The nanofluidic circuitry of HT-
qPCR uses very small amounts of tissue, enabling
researchers to measure agent loads in nonlethal biopsies, a
technique that can be integrated into telemetry and labo-
ratory studies of host fate. Researchers in British Colum-
bia have used this tool to describe how infectious agents
influence migration success, stressor resilience, survival,
and immune responses of Pacific salmon (adults: Miller et
al. 2014; Bass et al. 2017; Teffer et al. 2017; juveniles: Jef-
fries et al. 2014; Miller et al. 2014; Furey 2016; Tucker et
al. 2018). The BioMark platform has also been used to
characterize molecular markers of disease states in salmon
that may precede pathogen detection and are agent type-
specific (e.g., viral versus bacterial: Miller et al. 2017).
Although nonlethal sampling can cause physiological
stress to the animal, survival rates of biopsied salmon are
generally similar to those of nonbiopsied fish (Cooke et al.
2005, 2012; Jeffries et al. 2014; Teffer et al. 2018).
Broad-based infectious agent screening using HT-qPCR
is relatively new to wild fish health studies, so our under-
standing of how multi-agent load profiles compare across
tissue types is limited. Studies that have assessed the utility
of nonlethal tissue sampling for monitoring single infectious
agents have found a variation of results depending on the
host tissue examined as well as the agent (e.g., Cornwell et
al. 2013; Burbank et al. 2017; Chiaramonte et al. 2018).
Considering multiple infections, cumulative variation in
results among host tissues could substantially bias survey
conclusions regarding coinfection status and virulence
potential. One recent broad-based screening study (using
HT-qPCR) of adult Sockeye Salmon Oncorhynchus nerka
found variable agreement in multiple infectious agent met-
rics derived from nonlethal (gill tissue) and destructive (vari-
ous pooled tissue) samples (Teffer et al. 2017). As nonlethal
approaches continue to be applied in broad-based infectious
agent screening, the potential biases of using nonlethal sam-
pling relative to destructive methods must be quantified
beyond a single host species. Gill tissue has been the pri-
mary nonlethal tissue used for broad-based infectious agent
screening of wild Pacific salmon (Jeffries et al. 2014; Miller
et al. 2014; Teffer et al. 2017, 2018). An evaluation of other
nonlethally sampled tissues, such as blood, could provide
additional options to researchers. Blood screening may
prove highly effective given that blood is both a target tissue
and a transport means for infectious agents within the body
(e.g., Evelyn and Traxler 1978; Raida and Buchmann 2008;
Bjork and Bartholomew 2010; Finstad et al. 2014). If blood
can also be used for pathogen screening, experimental han-
dling time can be reduced for studies where blood is already
collected for physiological assessment (e.g., Maule et al.
1996; Ortuño et al. 2002; Jeffries et al. 2011; Drenner et al.
2018).
In this study, we used HT-qPCR to characterize how
infectious agent metrics (prevalence and load) differed
between nonlethal samples (gill tissue or blood) and destruc-
tive samples (internal and external organs) from adult Chi-
nook Salmon Oncorhynchus tshawytscha. Adult Chinook
Salmon were used as the host model for this study because
(1) infection burdens that were previously described in gill
tissue and multi-tissue pools from adult Chinook Salmon
showed a broad range of infectious agent taxa (Bass et al.
2017), and (2) declining population productivity of Chinook
Salmon will warrant continued application of nonlethal
sampling to describe fish health and infection status (Riddell
et al. 2013). Destructive sampling for this study consisted of
combined biopsies from five organs (gill, liver, kidney,
heart, and brain) into a multi-tissue pool for each host for
HT-qPCR screening, which is relevant to current broad-
based surveys of wild salmon (Miller et al. 2016). Sources of
disagreement between tissue types were identified, with an
emphasis on agents that are likely to bypass detection by
either nonlethal or destructive sampling methods. Our
results are presented in terms of detection probability of
agents in each tissue and the potential influence of agent
ecology (e.g., target tissue of the host and infection develop-
ment stage) on detection probability. Logistical constraints
and feasibility of each method regarding field and labora-
tory protocols are included to aid researchers in decision-
making for nonlethal sampling.
METHODS
Tissue samples were collected from adult Chinook Sal-
mon returning to the Chilliwack River Hatchery in Chilli-
wack, British Columbia, on October 13, 2016, by the
CHINOOK SALMON qPCR SCREENING 275

Environmental Watch Program of the Department of
Fisheries and Oceans Canada (DFO). Twelve fish (male:
n=6, female: n=6) were sacrificed by cerebral concussion
and cervical dislocation. Approximately 2 mL of blood
was collected from the caudal vasculature using a 21-
gauge needle with a lithium heparinized Vacutainer
(Becton-Dickinson, Franklin Lakes, New Jersey). Blood
samples were placed on ice for ≤15 min before a subsam-
ple of 500 μL was transferred using a pipettor into a sterile
vial and frozen (–80°C) in a portable ultra-low freezer
(Stirling Ultracold, Athens, Ohio). Gill tissue (approxi-
mately 0.5 mg consisting of 2–3 gill filament tips) was col-
lected using sterile scissors, simulating nonlethal biopsy
protocols that are used in field and laboratory studies
(e.g., Cooke et al. 2005; Teffer et al. 2017). Gill samples
and tissue samples from liver, head kidney, heart ventricle,
and brain (whole brain from every other individual) were
preserved individually in 1.5 mL of RNAlater solution
(Qiagen, Germantown, Maryland) for genomic analyses
(whole brain was preserved in 3 mL). Separate sterile tools
were used to sample each tissue to prevent cross contami-
nation among organs within and among individuals. Tis-
sue vials were kept at 4°C for 24 h and then at –80°C
until analysis.
Molecular analyses were performed at the DFO Molec-
ular Genetics Laboratory, Pacific Biological Station in
Nanaimo, British Columbia. Each organ tissue sample
was homogenized in 600 μL tri-reagent (Ambion, Austin,
Texas) and 75 μL 1-bromo-3-chloropropane using stainless
steel beads and a mixer mill (Restch, Newtown, Pennsyl-
vania; model MM 301). Because whole brains were larger,
they were quartered and then homogenized in 600 μL tri-
reagent each; aliquots of 150 μL from each brain were
then pooled before the addition of 75 μL 1-bromo-3-chlor-
opropane. Following centrifugation (1,500 ×gfor 6.5
min), 20-μL aliquots of the aqueous phase from each tis-
sue type (gill, liver, head kidney, heart ventricle, and
brain) were combined to create a tissue pool for each fish.
If the brain was missing, it was substituted by RNA/
DNA-free water. RNA was purified from 100 μL of each
multi-tissue aqueous pool following the manufacturer's
protocols for the “spin method”in the Magmax-96 for
Microarrays kit (Ambion), with an additional DNase
treatment after the first wash. Extractions and normaliza-
tions were performed using a Biomek FXP automated liq-
uid handler (Beckman-Coulter, Mississauga, Ontario).
RNA quality (A
260/280
) and quantity (A
260
) were evaluated
using spectrophotometry and each well was normalized to
1μg of RNA. RNA was also extracted from gill alone
(100 μL of the aqueous phase) following the protocols
described above and normalized to 1 μg RNA.
To enhance the effectiveness of the tri-reagent in lysing
the cells in the blood samples and to decrease the likeli-
hood of protein and hemoglobin contamination affecting
the RNA extraction process, a dilution approach was
applied to blood samples undergoing homogenization. To
reduce the ratio of blood to tri-reagent, low-volume (10-
μL) replicates of each blood sample were homogenized
separately and then pooled before extraction. Frozen
blood samples were thawed on ice and four aliquots of
10 μL were homogenized separately in 600 μL each of tri-
reagent with stainless-steel beads using the mixer mill (i.e.,
four replicates per sample). Next, 150-μL aliquots from
each replicate were pooled in sterile microtubes (totaling
600 μL of blood homogenate for each fish) prior to the
addition of 75 μL 1-bromo-3-chloropropane. Samples were
then centrifuged (1,500 ×gfor 6.5 min) and RNA was
purified from 100 μL of the supernatant as described for
other tissues.
cDNA was synthesized using the SuperScript VILO
cDNA Synthesis kit (Invitrogen, Carlsbad, California;
cycling conditions: 25°C for 10 min, 42°C for 60 min, and
85°C for 5 min). Pre-amplification of cDNA was com-
pleted as per manufacturer's recommendations (BioMark)
in a multiplex PCR including 200 nM primers (TaqMan
Preamp Master Mix, Applied Biosystems, Foster City,
California) using the following cycling conditions: 95°C
for 10 min followed by 15 cycles of 95°C for 10 s and
60°C for 4 min. Pre-amplification is necessary to achieve
adequate sensitivity on the nanofluidics platform. Pre-
amplification was immediately followed by ExoSap-it Pro-
duct Clean-up (Affymetrix, Santa Clara, California) using
the following cycling conditions: 37°C for 15 min followed
by 80°C for 15 min. Product then underwent a five-fold
dilution in TEKnova suspension buffer (Hollister, Califor-
nia). A serial dilution of artificial positive constructs (APC
clones) matching the primer-probe sequences of the infec-
tious agents under evaluation was added to the dynamic
array just before qPCR and tagged with a secondary
probe (NED dye) to identify laboratory contamination
(any NED-positive samples were removed from analysis).
Samples (consisting of TaqMan Universal PCR Master
Mix [Life Technologies, Carlsbad, California], GE Sample
Loading Reagent [Fluidigm], and pre-amplified cDNA)
and assays (in duplicate and consisting of 10 μM primers
and 3 μM probes for the TaqMan assays) were loaded
onto dynamic arrays using the integrated fluidics con-
troller HX system (Fluidigm) for qPCR (50°C for 2 min,
95°C for 10 min, 40 cycles of 95°C for 15 s, and 60°C for
1 min; Table 1).
Positive and negative controls were incorporated as
described in Miller et al. (2016), including a serial dilution
of artificial constructs to measure assay efficiency (all
between 90% and 110%). Negative controls were incorpo-
rated (all with reagents) during extraction (without tissue
homogenate), cDNA synthesis (without RNA, without
reverse transcriptase, and without both RNA and reverse
transcriptase), pre-amplification (without cDNA), and
276 TEFFER AND MILLER

TABLE 1. Assay information for 47 infectious agents and 1 host reference gene measured in the tissue of adult Chinook Salmon using high-through-
put qPCR. Assays developed by the Molecular Genetics Laboratory at the DFO Pacific Biological Station are designated “MGL.”
Agent type
Agent
name Assay reference
Accession
number
Forward primer sequence (5′–3′)
reverse primer sequence (5′–3′)
probe sequence (FAM-5′–3′-MGB)
Bacterium Aeromonas
hydrophila
Lee et al. (2006) AY165026 F: ACCGCTGCTCATTACTCTGATG
R: CCAACCCAGACGGGAAGAA
P: TGATGGTGAGCTGGTTG
Bacterium Aeromonas
salmonicida
Modified from
Keeling et al.
(2013)
M64655 F: TAAAGCACTGTCTGTTACC
R: GCTACTTCACCCTGATTGG
P: ACATCAGCAGGCTTCAGAGTCACTG
Bacterium Candidatus
Branchiomonas
cysticola
Mitchell et al.
(2013)
JQ723599 F: AATACATCGGAACGTGTCTAGTG
R: GCCATCAGCCGCTCATGTG
P: CTCGGTCCCAGGCTTTCCTCTCCCA
Bacterium Flavobacterium
psychrophilum
Duesund et al.
(2010)
F: GATCCTTATTCTCACAGTACCGTCAA
R: TGTAAACTGCTTTTGCACAGGAA
P: AAACACTCGGTCGTGACC
Bacterium Gill chlamydia Duesund et al.
(2010)
FJ897519 F: GGGTAGCCCGATATCTTCAAAGT
R: CCCATGAGCCGCTCTCTCT
P: TCCTTCGGGACCTTAC
Bacterium Moritella viscosa Grove et al.
(2008)
EU332345 F: CGTTGCGAATGCAGAGGT
R: AGGCATTGCTTGCTGGTTA
P: TGCAGGCAAGCCAACTTCGACA
Bacterium Piscichlamydia
salmonis
Nylund et al.
(2008)
EU326495 F: TCACCCCCAGGCTGCTT
R: GAATTCCATTTCCCCCTCTTG
P: CAAAACTGCTAGACTAGAGT
Bacterium Piscirickettsia
salmonis
Corbeil et al.
(2003)
U36943 F: TCTGGGAAGTGTGGCGATAGA
R: TCCCGACCTACTCTTGTTTCATC
P: TGATAGCCCCGTACACGAAACGGCATA
Bacterium Renibacterium
salmoninarum
Powell et al.
(2005)
AF123890 F: CAACAGGGTGGTTATTCTGCTTTC
R: CTATAAGAGCCACCAGCTGCAA
P: CTCCAGCGCCGCAGGAGGAC
Bacterium Rickettsia-like
organism
Lloyd et al.
(2011)
EU555284 F: GGCTCAACCCAAGAACTGCTT
R: GTGCAACAGCGTCAGTGACT
P: CCCAGATAACCGCCTTCGCCTCCG
Bacterium Tenacibaculum
maritimum
Fringuelli et al.
(2012b)
NBRC15946T F: TGCCTTCTACAGAGGGATAGCC
R: CTATCGTTGCCATGGTAAGCCG
P: CACTTTGGAATGGCATCG
Bacterium Vibrio
anguillarum
MGL L08012 F: CCGTCATGCTATCTAGAGATGTATTTGA
R: CCATACGCAGCCAAAAATCA
P: TCATTTCGACGAGCGTCTTGTTCAGC
Bacterium Vibrio
salmonicida
MGL AF452135 F: GTGTGATGACCGTTCCATATTT
R: GCTATTGTCATCACTCTGTTTCTT
P: TCGCTTCATGTTGTGTAATTAGGAGCGA
Bacterium Yersinia ruckeri Glenn et al.
(2011)
FJ518778 F: TGCCGCGTGTGTGAAGAA
R: ACGGAGTTAGCCGGTGCTT
P: AATAGCACTGAACATTGAC
Fluke Nanophyetus
salmincola
MGL AY269674 F: GATCTGCATTTGGTTCTGTAACA
R: CCAACGCCACAATGATAGCTATAC
P: TGAGGCGTGTTTTATG
CHINOOK SALMON qPCR SCREENING 277

TABLE 1. Continued.
Agent type
Agent
name Assay reference
Accession
number
Forward primer sequence (5′–3′)
reverse primer sequence (5′–3′)
probe sequence (FAM-5′–3′-MGB)
Parasite Ceratomyxa shasta Hallett and
Bartholomew
(2006)
AF001579 F: CCAGCTTGAGATTAGCTCGGTAA
R: CCCCGGAACCCGAAAG
P: CGAGCCAAGTTGGTCTCTCCGTGAAAAC
Parasite Cryptobia
salmositica
MGL F: TCAGTGCCTTTCAGGACATC
R: GAGGCATCCACTCCAATAGAC
P: AGGAGGACATGGCAGCCTTTGTAT
Parasite Dermocystidium
salmonis
MGL U21337 F: CAGCCAATCCTTTCGCTTCT
R: GACGGACGCACACCACAGT
P: AAGCGGCGTGTGCC
Parasite Facilispora
margolisi
MGL HM800849 F: AGGAAGGAGCACGCAAGAAC
R: CGCGTGCAGCCCAGTAC
P: TCAGTGATGCCCTCAGA
Parasite Gyrodactylus
salaris
Collins et al.
(2010)
F: CGATCGTCACTCGGAATCG
R: GGTGGCGCACCTATTCTACA
P: TCTTATTAACCAGTTCTGC
Parasite Ichthyophonus
hoferi
White et al.
(2013)
AF467793 F: GTCTGTACTGGTACGGCAGTTTC
R: TCCCGAACTCAGTAGACACTCAA
P: TAAGAGCACCCACTGCCTTCGAGAAGA
Parasite Ichthyophthirius
multifiliis
MGL IMU17354 F: AAATGGGCATACGTTTGCAAA
R: AACCTGCCTGAAACACTCTAATTTTT
P: ACTCGGCCTTCACTGGTTCGACTTGG
Parasite Kudoa thyrsites Funk et al.
(2007)
AF031412 F: TGGCGGCCAAATCTAGGTT
R: GACCGCACACAAGAAGTTAATCC
P: TATCGCGAGAGCCGC
Parasite Loma salmonae MGL HM626243 F: GGAGTCGCAGCGAAGATAGC
R: CTTTTCCTCCCTTTACTCATATGCTT
P: TGCCTGAAATCACGAGAGTGAGACTACCC
Parasite Myxobolus arcticus MGL HQ113227 F: TGGTAGATACTGAATATCCGGGTTT
R: AACTGCGCGGTCAAAGTTG
P:CGTTGATTGTGAGGTTGG
Parasite Myxobolus
insidiosus
MGL EU346375 F: CCAATTTGGGAGCGTCAAA
R: CGATCGGCAAAGTTATCTAGATTCA
P: CTCTCAAGGCATTTAT
Parasite Neoparamoeba
perurans
Fringuelli et al.
(2012a)
EF216905 F: GTTCTTTCGGGAGCTGGGAG
R: GAACTATCGCCGGCACAAAAG
P: CAATGCCATTCTTTTCGGA
Parasite Nucleospora
salmonis
Foltz et al.
(2009)
AF186009 F: GCCGCAGATCATTACTAAAAACCT
R: CGATCGCCGCATCTAAACA
P: CCCCGCGCATCCAGAAATACGC
Parasite Paranucleospora
theridion (syn.
Desmozoon
lepeophtherii)
Nylund et al.
(2010)
FJ59481 F: CGGACAGGGAGCATGGTATAG
R: GGTCCAGGTTGGGTCTTGAG
P: TTGGCGAAGAATGAAA
Parasite Parvicapsula
kabatai
MGL DQ515821 F: CGACCATCTGCACGGTACTG
R: ACACCACAACTCTGCCTTCCA
P:CTTCGGGTAGGTCCGG
278 TEFFER AND MILLER

TABLE 1. Continued.
Agent type
Agent
name Assay reference
Accession
number
Forward primer sequence (5′–3′)
reverse primer sequence (5′–3′)
probe sequence (FAM-5′–3′-MGB)
Parasite Parvicapsula
minibicornis
Hallett and
Bartholomew
(2009)
AF201375 F: AATAGTTGTTTGTCGTGCACTCTGT
R: CCGATAGGCTATCCAGTACCTAGTAAG
P: TGTCCACCTAGTAAGGC
Parasite Parvicapsula
pseudobranchicola
Jørgensen et al.
(2011)
AY308481 F: CAGCTCCAGTAGTGTATTTCA
R: TTGAGCACTCTGCTTTATTCAA
P: CGTATTGCTGTCTTTGACATGCAGT
Parasite Sphaerothecum
destructuens
MGL AY267346 F: GGGTATCCTTCCTCTCGAAATTG
R: CCCAAACTCGACGCACACT
P: CGTGTGCGCTTAAT
Parasite Spironucleus
salmonicida
MGL AY677182 F: GCAGCCGCGGTAATTCC
R: CGAACTTTTTAACTGCAGCAACA
P: ACACGGAGAGTATTCT
Parasite Tetracapsuloides
bryosalmonae
Bettge et al.
(2009)
AF190669 F: GCGAGATTTGTTGCATTTAAAAAG
R: GCACATGCAGTGTCCAATCG
P: CAAAATTGTGGAACCGTCCGACTACGA
Virus Infectious
hematopoietic
necrosis virus
Purcell et al.
(2013)
NC_001652 F: AGAGCCAAGGCACTGTGCG
R: TTCTTTGCGGCTTGGTTGA
P: TGAGACTGAGCGGGACA
Virus Infectious
pancreatic
necrosis virus
S. Clouthier and
colleagues,
abstract
presented at
the 7th
International
Symposium on
Aquatic
Animal Health,
2014
F: GCAACTTACTTGAGATCCATTATGCT
R: AGACCTCTAAGTTGTATGACGAGGTCTCT
P: CGAGAATGGGCCAGCAAGCA
Virus Infectious salmon
anemia virus
Plarre et al.
(2005)
F: TGGGATCATGTGTTTCCTGCTA
R: GAAAATCCATGTTCTCAGATGCAA
P: CACATGACCCCTCGTC
Virus Infectious salmon
anemia virus
LeBlanc et al.
(2010)
EU118822 F: TGGGCAATGGTGTATGGTATGA
R: GAAGTCGATGAACTGCAGCGA
P: CAGGATGCAGATGTATGC
Virus Pacific salmon
parvovirus
MGL F: CCCTCAGGCTCCGATTTTTAT
R: CGAAGACAACATGGAGGTGACA
P: CAATTGGAGGCAACTGTA
Virus Piscine myocarditis
virus (CMS)
Løvoll et al.
(2010)
HQ401057 F: TTCCAAACAATTCGAGAAGCG
R: ACCTGCCATTTTCCCCTCTT
P: CCGGGTAAAGTATTTGCGTC
Virus Piscine reovirus
(HSMI)
Wiik-Nielsen
et al. (2012)
F: TGCTAACACTCCAGGAGTCATTG
R: TGAATCCGCTGCAGATGAGTA
P: CGCCGGTAGCTCT
Virus Salmon alphavirus
1, 2, and 3 (PD/
SD/HSS)
Hodneland and
Endresen
(2006)
AY604235 F: CCGGCCCTGAACCAGTT
R: GTAGCCAAGTGGGAGAAAGCT
P: TCGAAGTGGTGGCCAG
CHINOOK SALMON qPCR SCREENING 279

qPCR (with nonpreamplified cDNA and without cDNA).
Positive controls were incorporated during cDNA synthe-
sis (using a multi-host tissue pool), pre-amplification, and
qPCR. Negative controls did not show any evidence of
cross contamination among samples or independent cDNA
synthesis (without reverse transcriptase). Positive controls
produced values consistent with proper processing (e.g.,
nonpreamplified cDNA had higher quantification cycle
[Cq] values than preamplified cDNA). A reference gene
(in-house design) was incorporated into all reactions with
consistent Cq values that were measured across samples.
All assays were run in duplicate and any agents that
were not detected in duplicate were designated as “failed”
and removed from the analysis. The average of duplicate
Cq values was subtracted from 40 (the maximum value)
and is subsequently referred to as “load.”Reliable limits
of detection (LOD) were not applied, as this approach
would remove low-level detections that are relevant to this
analysis (Miller et al. 2016); however, LOD thresholds are
included in our discussion of the results to highlight agents
that may fall outside of the range detectability at 95%
probability by alternate qPCR platforms. Our findings are
discussed in terms of detection probability, which refers to
how frequently a single agent or agent type (e.g., bacteria,
viruses, other microparasite genera) was detected in a tis-
sue relative to detection in any tissue type (gill, blood, or
multi-tissue pool).
All statistics were performed using R statistical software
(R Core Team, Vienna). Cohen's kappa (Κ) was used as a
measure of agreement between tissue types (e.g., Cornwell
et al. 2013); this statistic could not be applied to all agents
due to insufficient detections or negligible variation between
tissue types (e.g., if an agent was 100% prevalent in all tissue
types, no statistical comparison can be made). To compare
host population “community structure”between tissue
types, the cumulative prevalence of all detected agents were
compared between gill, blood, and the multi-tissue pool
using Κ. Infectious agent loads were compared among tissue
types where sample sizes allowed (>2 agent detections in all
tissue types required for inclusion in this analysis) using lin-
ear mixed effects models and analysis of variance in the
“lmerTest”package. Only positive detections were included
in load comparisons (i.e., zeros were excluded).
RESULTS
Of the 47 agents included in the screening, 16 were
detected in 12 adult Chinook Salmon (Table 2; Figure 1).
These agents included one virus (erythrocytic necrosis
virus; ENV); six bacteria (Flavobacterium psychrophilum,
Candidatus Branchiomonas cysticola, Aeromonas salmoni-
cida,Piscirickettsia salmonis,Rickettsia-like organism
[RLO], and Yersinia ruckeri); several microparasites
including five myxozoans (Ceratonova shasta,Parvicapsula
minibicornis,Kudoa thyrsites,Tetracapsuloides bryosalmo-
nae, and Myxobolus arcticus); one microsporidian (Loma
salmonae); one protist (Dermocystidium salmonis);
one flagellate (Cryptobia salmositica); and one ciliate
TABLE 1. Continued.
Agent type
Agent
name Assay reference
Accession
number
Forward primer sequence (5′–3′)
reverse primer sequence (5′–3′)
probe sequence (FAM-5′–3′-MGB)
Virus Salmonid
herpesvirus/
Oncorhynchus
masou herpes
virus
MGL F: GCCTGGACCACAATCTCAATG
R: CGAGACAGTGTGGCAAGACAAC
P: CCAACAGGATGGTCATTA
Virus Viral
encephalopathy
and retinopathy
virus
Korsnes et al.
(2005)
AJ245641 F: TTCCAGCGATACGCTGTTGA
R: CACCGCCCGTGTTTGC
P: AAATTCAGCCAATGTGCCCC
Virus Viral erythrocytic
necrosis virus
J. Winton,
personal
communication
F: CGTAGGGCCCCAATAGTTTCT
R: GGAGGAAATGCAGACAAGATTTG
P: TCTTGCCGTTATTTCCAGCACCCG
Virus Viral hemorrhagic
septicemia virus
Garver et al.
(2011)
Y18263 F: ATGAGGCAGGTGTCGGAGG
R: TGTAGTAGGACTCTCCCAGCATCC
P: TACGCCATCATGATGAGT
Host Reference gene MGL CA056739 F: GTCAAGACTGGAGGCTCAGAG
R: GATCAAGCCCCAGAAGTGTTTG
P: AAGGTGATTCCCTCGCCGTCCGA
280 TEFFER AND MILLER

(Ichthyophthirius multifiliis). The gill was the most compre-
hensive screening tissue at the population level, with 15
infectious agents detected versus 11 in blood and 14 in
multi-tissue pools. Bacteria were detected at the highest
prevalence in gill (the prevalence in gills was greater than
or equal to that in pools), while the prevalence of bacteria
measured in blood was generally lower than that in gills
or multi-tissue pools. Erythrocytic necrosis virus, a blood
pathogen, was detected more frequently in blood (25%)
than in gills or multi-tissue pools (both 17%). The parasite
and protist detection probability varied across agent spe-
cies and tissue types (pools: 8–100%; gills and blood: 0–
100%).
The prevalence of agents and agreement (Κ) between
tissue types are presented in Table 2. Both gill and blood
screening were 100% effective for detecting C. shasta and
TABLE 2. Infectious agents detected in tissues sampled from adult Chinook Salmon using high-throughput qPCR. Prevalence (percent positive in the
sampled population; n=12) is shown by tissue type, which included gill, blood, a multi-tissue pool (gill, liver, head kidney, heart ventricle, and brain
[from every other individual]), and the combination of all approaches. Cohen's kappa (Κ) describes the agreement between tissue types, and it ranges
from negative values (lesser agreement than by chance) to 1 (perfect agreement). Host population-level community structure (the cumulative prevalence
of all agents) was compared between tissue types using Κ. Blanks indicate insufficient data for comparison due to perfect agreement or low prevalence.
Agent type Agent name
Combined
tissue types
Pool
(%)
Gill
(%)
Blood
(%)
Κ
(gill–pool)
Κ
(blood–pool)
Κ
(gill–blood)
Bacteria Flavobacterium psychrophilum 100% (12) 100 100 100
Ca. B. cysticola 100% (12) 100 100 8
Aeromonas salmonicida 83% (10) 42 83 25 0.25 0.64 0.13
Piscirickettsia salmonis 25% (3) 8 25 8 0.43 1.00 0.43
Rickettsia-like organism (RLO) 17% (2) 0 17 0
Yersinia ruckeri 17% (2) 0 17 0
Virus Erythrocytic necrosis virus (VEN) 25% (3) 17 17 25 1.00 0.75 0.75
Parasite Ceratonova shasta 100% (12) 100 100 100
Dermocystidium salmonis 100% (12) 92 100 0
Parvicapsula minibicornis 100% (12) 83 92 75 −0.13 0.75 0.43
Cryptobia salmositica 83% (10) 58 75 83 0.64 0.44 0.75
Ichthyophthirius multifiliis 75% (9) 42 75 0 0.39
Kudoa thyrsites 58% (7) 58 25 50 0.39 0.83 0.50
Loma salmonae 42% (5) 33 8 42 0.31 0.82 0.23
Tetracapsuloides bryosalmonae 25% (3) 8 25 0 0.43
Myxobolus arcticus 17% (2) 17 0 8 0.63
Population-level agent community structure 0.17 0.24 0.03
100
Pool
Gill
Blood
A. salmonicida
Ca. B. cysticola
ENV
RLO
Y. ruckeri
T. bryosalmonae
P. minibicornis
M. arcticus
L. salmonae
D. salmonis
C. shasta
C. salmositica
K. thyrsites
I. multifiliis
P. salmonis
F. psychrophilum
Bacterial and viral agents Other microparasites
% Prevalence
75
50
25
0
100
75
50
25
0
FIGURE 1. The percent prevalence of infectious agents detected using HT-qPCR in gills, blood, and multi-tissue pools (gill, heart, liver, kidney, and
brain [from every other individual]) from adult Chinook Salmon (n=12).
CHINOOK SALMON qPCR SCREENING 281

F. psychrophilum, which were present in all fish (n=12).
Parvicapsula minibicornis was also prevalent in all fish,
and the detections of this pathogen were generally well-
matched between gills and blood (Κ= 0.43) and blood and
multi-tissue pools (Κ= 0.75), but no better than by chance
between gill and multi-tissue pools (Κ=–0.13); two fish
with P. minibicornis gill detections tested negative in
blood, and one multi-tissue pool detection was negative in
blood and gill. Both Ca. B. cysticola and D. salmonis were
100% prevalent in gill but poorly detected in blood (Ca.
B. cysticola was found in one blood sample that exceeded
the LOD [high Cq, low load], and D. salmonis was not
detected in blood). Aeromonas salmonicida was found in
83% of fish examined and all of the positive fish had gill
detections; five A. salmonicida gill detections were negative
in multi-tissue pools and blood, with most detections near
or exceeding the LOD. For A. salmonicida, there was a
higher agreement between blood and multi-tissue pools (Κ
= 0.64) than in other tissue comparisons (gill–blood: Κ=
0.13; gill–pool: Κ= 0.25). Cryptobia salmositica, a blood
pathogen, was detected in 10 fish. Positive sample types
were blood (83% prevalence) and gills (75% prevalence).
One weak detection (just exceeding the LOD) was found
only in the blood of one host. All of the fish that were
positive for I. multifiliis (75%) and RLO (17%) had these
detections in the gills (although several exceeded the
LOD), with no detections in the blood. Kudoa thyrsites, a
nonpathogenic myxozoan parasite, was moderately preva-
lent (58%) and more frequently detected in blood (50%)
than in gills (25%). All of the hosts that were positive for
L. salmonae (42%) had positive blood samples, including
one that was negative in the multi-tissue pool; only one
high-load host tested positive in the gills, and most detec-
tions approached or exceeded the LOD. Low-prevalence
agents included ENV (best detected in blood samples), P.
salmonis,T. bryosalmonae, and Y. ruckeri. The former
three exceeded the LOD, and all were all detected in the
gills of positive hosts. Myxobolus arcticus, a brain para-
site, was detected at a very low prevalence and was only
present in the blood of one fish and two multi-tissue pools.
The cumulative prevalence (i.e., community structure) var-
ied depending on the tissue type, with a low agreement
(Κ≤0.24; Table 2).
Load data are presented relative to the 95% LOD that
was specific to each agent in Figures 2 and 3. Considering
only positive detections (zeros excluded), the loads of eight
infectious agents differed significantly among tissue types:
Ca. B. cysticola (F=226.6; df =2, 11; P<0.001), C.
shasta (F=9.5; df =2, 22; P=0.001), C. salmositica (F=
9.5; df =2, 22; P=0.001), D. salmonis (F=92.5; df =1,
10; P<0.001), F. psychrophilum (F=32.3; df =2, 22; P<
0.001), I. multifiliis (F=106.2; df =1, 4; P=0.001), K.
thyrsites (F=5.6; df =2, 8; P=0.032), and ENV (F=
37.4; df =2, 2; P=0.026). Load data from several agents
including Ca. B. cysticola, D. salmonis, and I. multifiliis
suggested a consistent dilution of gill-resident infections
within multi-tissue pools (Figure 3). Higher gill loads were
observed for two bacterial agents—F. psychrophilum and
A. salmonicida—but load differences between gills and
multi-tissue pools were not consistent; hence, these infec-
tions may have been present in one or more internal
organs. Generally lower loads of A. salmonicida were
detected primarily in gill, which may indicate external
and/or gill-resident infections, while the heaviest infections
were highest in multi-tissue pools and therefore likely
internal. Cryptobia salmositica showed similar loads
between gills and blood, but was lower in multi-tissue
pools. Ceratonova shasta and P. minibicornis showed
inconsistent load patterns among tissues, though C. shasta
loads were generally higher in gills than in blood and mul-
ti-tissue loads fell between gill and blood load levels or
above both. Only the highest L. salmonae loads were
detected in gills, while blood loads were generally similar
to multi-tissue pool loads. Erythrocytic necrosis virus con-
sistently had the highest loads in blood followed by gills
and then multi-tissue pools. Less-prevalent agents had the
highest loads in multi-tissue pools and gills followed by
blood; these included P. salmonis,M. articus, and T. bryo-
salmonae (which had no blood detections). Kudoa thyrsites
was more prevalent in females (100%) than in males (8%),
with maximum loads in multi-tissue pools that were fre-
quently detected in blood (Κ= 0.83) but not gills (Κ=
0.39).
DISCUSSION
This study characterized the effectiveness of two
nonlethal approaches for broad-based infectious agent
screening of Chinook Salmon (gill and blood sampling) rel-
ative to destructive multi-tissue sampling using HT-qPCR.
Sixteen infectious agents were detected, including one virus,
six bacteria, and nine other microparasite species. Many of
these agents have previously been described in this Chi-
nook Salmon population (Bass et al. 2017; Teffer et al.
2018), excluding P. salmonis and Y. ruckeri, which were
not found previously. The prevalence varied depending on
the tissue type and agent species, ranging from absent or
rare (0–8%) to ubiquitous (≤100%). Known epithelial or
gill pathogens (e.g., Ca. B. cysticola and RLO: Nigrelli et
al. 1976; Olson and Holt 1995; Sun et al. 2009) exhibited a
higher prevalence in gills than in multi-tissue pools or
blood. Support for these and several other agents as gill-
focused (e.g., D. salmonis and I. multifiliis) was evident in
the consistent load differences between gill and multi-tissue
pool samples from the same host. Similarly, known blood
pathogens (e.g., ENV and C. salmositica: Evelyn and Trax-
ler 1978; Woo 2003) were most prevalent and generally at
their highest loads in blood. Agents with more complex life
282 TEFFER AND MILLER

cycles and multiple developmental stages (e.g., C. shasta
and P. minibicornis: Bartholomew et al. 1997, 2006; Brad-
ford et al. 2010) or those with systemic infection potential
(e.g., F. psychrophilum: Starliper 2011) had inconsistent
load profiles among tissue types. We hypothesize that this
variation may indicate alternate infection development
stages of these agents among hosts. Our results show great
promise for nonlethal tissue sampling to be used for broad-
based infectious agent screening of wild salmon. Future
research should pair results from nonlethal HT-qPCR
infectious agent screening with histological and micro-
scopic analyses of the same fish to characterize how disease
development relates to molecular detection of potential
pathogens across tissues.
Previous studies that have characterized the utility of
using one or more nonlethally sampled tissues or diagnos-
tic approaches (e.g., molecular methods, cell culture) for
infectious agent screening have found inherent variability
in results among tissues, especially if the nonprimary tis-
sues of agents are evaluated (Chiaramonte et al. 2018).
The broad-based screening approach that we applied seeks
to comprise an array of infectious agents, thereby forcing
compromise in the tissues included in the analysis. Com-
bining results from multiple independently screened tissue
FIGURE 2. Infectious agent loads measured in 12 adult Chinook Salmon of an equal sex ratio using HT-qPCR. Loads are shown by tissue type,
including gills (circles), blood (triangles), and multi-tissue pools (gill, heart, kidney, liver, and brain [from every other individual]; squares). The
horizontal dashed black lines indicate the HT-qPCR 95% limit of detection (LOD) for each qPCR assay. Only positive detections (nonzero values) are
included in plots.
CHINOOK SALMON qPCR SCREENING 283

types generally provides a more comprehensive and sensi-
tive assessment of infectious agent burdens than a single
tissue (Cornwell et al. 2013), but can be costly and time-
consuming. For broad-based screening of multiple infec-
tious agents with inherent variation in primary tissues
among agents, pooling multiple tissues prior to analysis
increases the likelihood of including the primary tissues of
more agents. However, this approach simultaneously
reduces the detection probability if the genetic material of
low-load agents in primary tissues is substantially diluted
by additional tissues. For several gill- and blood-specific
agents, the screening of single tissues may be more
effective at capturing the infection prevalence of low-load
agents.
Broad-based screening of nonlethal gill biopsies identi-
fied the greatest richness of bacteria, while blood screening
may have only detected advanced bacterial infections (i.e.,
systemic or internal). For example, the agent of bacterial
cold-water disease, F. psychrophilum, had inconsistent load
differences among tissue types within hosts (i.e., detections
in gills were sometimes higher than in multi-tissue pools,
and were sometimes lower), which supports intermittent
internal F. psychrophilum infections in our study popula-
tion. Alternatively, Ca. B. cysticola was poorly detected in
FIGURE 3. Infectious agent loads measured in three tissue types sampled from adult Chinook Salmon (n=12): gills (circles), blood (triangles), and
multi-tissue pools (black; gill, heart, liver, kidney, and brain [from every other individual]; squares). The horizontal dashed black lines indicate the
HT-qPCR 95% limit of detection (LOD) for each agent. Each x-axis tick is a host; females are shown on the left (dark gray shaded plot area) and
males are shown on the right (light gray shaded plot area).
284 TEFFER AND MILLER

blood but frequently detected in gill, supporting gill as its
primary tissue (Mitchell et al. 2013). Aeromonas salmoni-
cida loads were generally low and primarily in gill sam-
ples; however, heavy gill loads were associated with
greater pool loads, indicating that internal infections may
result from heavy gill or environmental loads. These spec-
ulations warrant further study into how load differences
among tissues correspond to infection status.
Agents that were well-detected in blood samples
included the hemoflaggelate C. salmositica, (Woo 2003),
the iridovirus ENV, which targets erythrocytes (Evelyn
and Traxler 1978), and the microsporidian gill parasite L.
salmonae (Kent and Speare 2005). These agents were
detected more frequently in blood than in gills, but only
marginally (20–90% by gill versus 100% by blood). The
microsporidian gill parasite, L. salmonae, was more fre-
quently detected in blood than in gills, which warrants
further exploration of the etiology of disease outcomes in
Chinook Salmon. Cryptobia salmositica infects its host via
an intermediate freshwater leech host (Woo 2003) and has
been observed previously in Chinook Salmon at this
hatchery (Bass et al. 2017). Kudoa thyrsites is nonpatho-
genic but can have serious economic implications in aqua-
culture by reducing the quality of flesh postmortem
(Moran et al. 1999). This agent was frequently detected in
blood (86% prevalence), offering a nonlethal detection
method for cultured fish to assess the magnitude of its
economic impact. High blood loads of ENV and L. salmo-
nae corresponded with detection in the gills, possibly due
to their presence in blood within biopsied lamellae. Low-
prevalence agents included the brain parasite M. arcticus
(Moles and Heifetz 1998), the kidney parasite T. bryosal-
monae (Longshaw et al. 2002), and the intracellular bac-
terium P. salmonis (Rise et al. 2004), which showed the
highest loads in multi-tissue pools.
For agents that varied inconsistently among tissue types
within hosts, there is potential for future studies to explore
how load differences between tissues may indicate the
developmental stage of infections. Since many of these
agents infect hosts via the gill (Olson and Holt 1995;
Bartholomew et al. 1997, 2006; Nematollahi et al. 2003;
Mitchell et al. 2013), the stages of infection may be indi-
cated by load discrepancies between blood and gill. For
example, some myxozoan parasites (e.g., C. shasta and P.
minibicornis) move through the body via the blood to ter-
minal tissues (C. shasta in the gut [not included here] and
P. minibicornis in the kidney), and may potentially be
detected in other internal organs (e.g., kidneys) via filter-
ing mechanisms (Bjork and Bartholomew 2010; Bradford
et al. 2010; Okamura et al. 2015). For agents that follow
this general trajectory, heavy loads of genetic material in
the gills alone may indicate early infection, while heavy
loads in the gills and blood may indicate later stages. If
heavy loads are detected in the gills, blood, and multi-
tissue pools, this may indicate more advanced infections.
In the present study, HT-qPCR determined that infectious
loads cannot be interpreted as a disease state without a
histological assessment of the affected tissues. However,
nonlethal biopsies can serve as indicators of disease poten-
tial, thus encouraging the initiation of further assessments
that include microscopy and histopathology.
Previously determined infectious agent detection proba-
bilities in the gills and multi-tissue pools sampled from
Sockeye Salmon in the Fraser River, British Columbia
(Teffer et al. 2017) generally align with our findings. In
addition to the biological differences between host species,
the Sockeye Salmon population described by Teffer et al.
(2017) migrates through the Fraser River months earlier
and over a greater distance than the Chilliwack River fall
run Chinook Salmon population that was described in this
study. Therefore, infection burdens are expected to differ
at least marginally between studies; interestingly, many of
the detected agents are shared, suggesting that the envi-
ronment (in the Fraser River watershed) is a strong deter-
minant of infection burdens (Teffer et al. 2017, 2018).
However, some differences in tissue agreement between
studies are apparent. Low agreement between C. shasta
loads in gills and multi-tissue pools from Sockeye Salmon
was described by Teffer et al. (2017), with a lower detec-
tion probability in gills. Chinook Salmon in our study
likely encounter higher C. shasta spore densities, given
that spawning migration occurs during the fall when
spores are generally released by the intermediate poly-
chaete host (Bartholomew et al. 1997). Low C. shasta load
levels in Sockeye Salmon gills may result in greater vari-
ability in the qPCR detection probability (Miller et al.
2016). Some agents detected in multi-tissue pools were
undetected in the gills of Sockeye and Chinook salmon
(e.g., L. salmonae), but only Sockeye Salmon had multi-
tissue F. psychrophilum infections that were undetected in
gills. This finding again supports an influence of preva-
lence and load on the detection probability of nonlethal
tissues, given that Sockeye Salmon had lower F. psy-
chrophilum prevalence than Chinook Salmon. Loma sal-
monae exhibited marginal prevalence in both host species
and was consistently variable across tissues, though it was
better detected in the blood of Chinook Salmon. Overall,
the results from both studies support the use of gills as a
moderate to excellent indicator of multiple infection bur-
dens across tissues as detected by HT-qPCR.
The methodologies applied in this study were relatively
simple to execute in the field and laboratory but can be
time-consuming and logistically difficult depending on the
resources available to the researcher. Gill biopsies of live
fish can be done quickly and effectively (Cooke et al.
2005; Teffer et al. 2017), but fish must be either restrained
manually or anaesthetized to avoid injuring the gill during
biopsy. Blood sampling can be done in a similar fashion
CHINOOK SALMON qPCR SCREENING 285

and has been broadly applied in the lab and field (Cooke
et al. 2012). Gill tissue can be preserved in solution or
flash frozen. Blood samples must be frozen as soon as pos-
sible in the field and processed quickly after thawing dur-
ing laboratory processing. Blood preservation in solution
(e.g., RNAlater) was not attempted in this study but may
improve the preservation of nucleic acids relative to flash
freezing and thawing. Freezing was chosen for blood
preservation due to concerns regarding the loss of infec-
tious agents through suspension within the total solution;
a study comparing these two preservation techniques
would be useful to quantify potential impacts on results.
The importance of logistical constraints regarding
nonlethal tissue sampling (e.g., the invasiveness of proce-
dures) has been characterized for single agent assessments,
with the benefits of some less invasive techniques (e.g., fin
clipping) providing insufficient information on host health
and infection development (Burbank et al. 2017). In this
study, the processing of blood was more time-consuming
than of gills, taking into account the dilution process to
avoid downstream issues with protein contamination in
purified RNA as well as troubleshooting the clotting issues
following the thawing of samples. Blood samples could be
further diluted to avoid issues with extraction interference,
but detection probabilities for low-load agents may be
reduced. Sampling the blood was effective at detecting
blood pathogens and potentially systemic infections at the
described dilution rate; however, gills were generally effec-
tive at detecting several agents that showed evidence of
internal infection (i.e., greater loads in multi-tissue pools
than in gill alone). Gill biopsies may be preferable to
avoid laboratory logistics of blood or multi-tissue process-
ing. Gill screening also has the benefit of capturing agents
that may occur on the exterior gill surface (e.g., bacteria)
in addition to those present within the lamellae (e.g., I.
multifiliis) and the blood within the biopsied tissue. Fur-
thermore, host gene expression can be measured in gill tis-
sue to characterize host responses concurrent with
infections (Jeffries et al. 2014; Miller et al. 2014; Teffer et
al. 2017). However, blood can also be used to examine
host physiology (via the blood chemistry or gene expres-
sion of nucleated red blood cells) concurrently with blood
infections, so both nonlethal approaches can provide
information about host-pathogen relationships but at dif-
ferent resolutions. If the goal of the screening is to identify
only those agents present within the animal (i.e., not on
the gill surface), then blood would be preferable as a
nonlethal sampling option.
The conservation status of many fish populations calls
for more minimally invasive methods for examining
infection burdens and host-pathogen relationships (Cooke
et al. 2012; Raby et al. 2014). By pairing nonlethal
biopsy with telemetry (Jeffries et al. 2014) or laboratory
experiments (Teffer et al. 2017), we can improve our
knowledge of the infectious agent communities affecting
wild fish populations and associated impacts on host
behavior and survival. As qPCR only provides information
on relative amounts of pathogen and (or) host genetic
material, this tool must be complemented by additional
methods that interpret those loads, which include sentinel
fish exposure to quantify dose thresholds for infection, dis-
ease, and mortality (Hendrickson et al. 1989) and diagnos-
tic approaches like histopathology to describe tissue
changes and assign disease states. Both gill tissue and
blood show promise as nonlethal tissues for broad-based
infectious agent screening using HT-qPCR. The primary
objectives of infectious agent screening studies should be
considered in the design of the study and sampling
approach including key pathologies and infectious agents
relevant to the region, question, or study species. Our
investigation has characterized potential biases of two non-
lethal tissue types relative to destructive multi-tissue sam-
pling. This provides a better understanding of the results of
broad-based infectious agent screening. We suggest that
future studies characterize the mechanisms of observed tis-
sue-specific differences in the prevalence and load of etio-
logical agents.
ACKNOWLEDGMENTS
We are grateful to the Environmental Watch Program
of the Department of Fisheries and Oceans Canada for
assistance in sample collection, and to the Chilliwack
River Hatchery staff for providing the study fish. Mem-
bers of the Molecular Genetics Laboratory at the Pacific
Biological Station provided valuable insight for blood pro-
cessing. Funding was provided by the Pacific Salmon
Foundation, Genome British Columbia, and Fisheries and
Oceans Canada under the Strategic Salmon Health Initia-
tive and Salish Sea Marine Survival programs. There is no
conflict of interest declared in this article.
REFERENCES
Altizer, S., R. Bartel, and B. A. Han. 2011. Animal migration and infec-
tious disease risk. Science 331:296–302.
Bartholomew, J. L., S. D. Atkinson, and S. L. Hallett. 2006. Involvement
of Manayunkia speciosa (Annelida: Polychaeta: Sabellidae) in the life
cycle of Parvicapsula minibicornis, a myxozoan parasite of Pacific sal-
mon. Journal of Parasitology 92:742–748.
Bartholomew, J. L., M. J. Whipple, D. G. Stevens, and J. L. Fryer.
1997. The life cycle of Ceratomyxa shasta, a myxosporean parasite of
salmonids, requires a freshwater polychaete as an alternate host. Jour-
nal of Parasitology 83:859–868.
Bass, A. L., S. G. Hinch, A. K. Teffer, D. A. Patterson, and K. M.
Miller. 2017. A survey of microparasites present in adult migrating
Chinook Salmon (Oncorhynchus tshawytscha) in south-western British
Columbia determined by high-throughput quantitative polymerase
chain reaction. Journal of Fish Diseases 40:453–477.
Bettge, K., T. Wahli, H. Segner, and H. Schmidt-Posthaus. 2009. Prolif-
erative kidney disease in Rainbow Trout: time- and temperature-
286 TEFFER AND MILLER

related renal pathology and parasite distribution. Diseases of Aquatic
Organisms 83:67–76.
Bjork, S. J., and J. L. Bartholomew. 2010. Invasion of Ceratomyxa
shasta (Myxozoa) and comparison of migration to the intestine
between susceptible and resistant fish hosts. International Journal for
Parasitology 40:1087–1095.
Bordes, F., and S. Morand. 2011. The impact of multiple infections on
wild animal hosts: a review. Infection Ecology and Epidemiology
1:7346.
Bradford, M. J., J. Lovy, and D. A. Patterson. 2010. Infection of gill
and kidney of Fraser River Sockeye Salmon, Oncorhynchus nerka
(Walbaum), by Parvicapsula minibicornis and its effect on host physi-
ology. Journal of Fish Diseases 33:769–779.
Burbank, D. R., T. R. Fehringer, and L. V. Chiaramonte. 2017. Com-
parison of selected nonlethal samples from adult steelhead for detec-
tion of infectious hematopoietic necrosis virus using cell culture.
Journal of Aquatic Animal Health 29:67–73.
Chiaramonte, L. V., D. Burbank, R. Scott, and J. T. Trushenski. 2018.
Comparison of sampling and detection methods for Chinook Salmon
and steelhead naturally infected with Myxobolus cerebralis. Journal of
Aquatic Animal Health 30:57–64.
Cikanek, S., N. Mattheus, L. K. Belden, J. B. Walke, A. E. Savage, B.
Gratwicke, K. P. C. Minbiole, C. N. Santiago, R. N. Harris, and M.
H. Becker. 2015. Composition of symbiotic bacteria predicts survival
in Panamanian golden frogs infected with a lethal fungus. Proceedings
of the Royal Society B: Biological Sciences 282:20142881.
Collins, C. M., R. Kerr, R. Mcintosh, and M. Snow. 2010. Development
of a real-time PCR assay for the identification of Gyrodactylus para-
sites infecting salmonids in northern Europe. Diseases of Aquatic
Organisms 90:135–142.
Cooke, S. J., G. T. Crossin, D. A. Patterson, K. K. English, S. G.
Hinch, J. L. Young, R. F. Alexander, M. C. Healey, G. Van Der
Kraak, and A. P. Farrell. 2005. Coupling non-invasive physiologi-
cal assessments with telemetry to understand inter-individual varia-
tion in behaviour and survivorship of Sockeye Salmon:
development and validation of a technique. Journal of Fish Biology
67:1342–1358.
Cooke, S. J., S. G. Hinch, M. R. Donaldson, T. D. Clark, E. J. Eliason,
G. T. Crossin, G. D. Raby, K. M. Jeffries, M. F. Lapointe, K. M.
Miller, D. A. Patterson, and A. P. Farrell. 2012. Conservation physi-
ology in practice: how physiological knowledge has improved our
ability to sustainably manage Pacific salmon during up-river migra-
tion. Philosophical Transactions of the Royal Society of London Ser-
ies B Biological Sciences 367:1757–1769.
Corbeil, S., K. A. McColl, and M. S. J. Crane. 2003. Development of a
TaqMan quantitative PCR assay for the identification of Piscirick-
ettsia salmonis. Bulletin of the European Association of Fish Patholo-
gists 23:95–101.
Cornwell, E. R., C. A. Bellmund, G. H. Groocock, P. T. Wong, K. L.
Hambury, R. G. Getchell, and P. R. Bowser. 2013. Fin and gill biop-
sies are effective nonlethal samples for detection of viral hemorrhagic
septicemia virus genotype IVb. Journal of Veterinary Diagnostic
Investigation 25:203–209.
Drenner, S. M., S. G. Hinch, N. B. Furey, T. D. Clark, S. Li, T. Ming,
K. M. Jeffries, D. A. Patterson, S. J. Cooke, D. Robichaud, D. W.
Welch, A. P. Farrell, and K. M. Miller. 2018. Transcriptome patterns
and blood physiology associated with homing success of Sockeye Sal-
mon during their final stage of marine migration. Canadian Journal
of Fisheries and Aquatic Sciences 75:1511–1524.
Duesund, H., S. Nylund, K. Watanabe, K. F. Ottem, and A. Nylund.
2010. Characterization of a VHS virus genotype III isolated from
Rainbow Trout (Oncorhychus mykiss) at a marine site on the west
coast of Norway. Virology Journal [online serial] 7:19.
Engering, A., L. Hogerwerf, and J. Slingenbergh. 2013. Pathogen-host-
environment interplay and disease emergence. Emerging Microbes
and Infections [online serial] 2(2):e5.
Evelyn, T. P. T., and G. S. Traxler. 1978. Viral erythrocytic necrosis:
natural occurrence in Pacific salmon and experimental transmission.
Journal of the Fisheries Research Board of Canada 35:903–907.
Finstad, Ø. W., M. K. Dahle, T. H. Lindholm, I. B. Nyman, M. Løvoll,
C. Wallace, C. M. Olsen, A. K. Storset, and E. Rimstad. 2014. Pis-
cine orthoreovirus (PRV) infects Atlantic Salmon erythrocytes. Veteri-
nary Research [online serial] 45:35.
Foltz, J. R., K. P. Plant, K. Overturf, K. Clemens, and M. S. Powell.
2009. Detection of Nucleospora salmonis in steelhead trout, Oncor-
hynchus mykiss (Walbaum), using quantitative polymerase chain reac-
tion (qPCR). Journal of Fish Diseases 32:551–555.
Fringuelli, E., A. W. Gordon, H. Rodger, M. D. Welsh, and D. A. Gra-
ham. 2012a. Detection of Neoparamoeba perurans by duplex quantita-
tive taqman real-time PCR in formalin-fixed, paraffin-embedded
Atlantic salmonid gill tissues. Journal of Fish Diseases 35:711–724.
Fringuelli, E., P. D. Savage, A. Gordon, E. J. Baxter, H. D. Rodger,
and D. A. Graham. 2012b. Development of a quantitative real-time
PCR for the detection of Tenacibaculum maritimum and its applica-
tion to field samples. Journal of Fish Diseases 35:579–590.
Funk, V. A., M. Raap, K. Sojonky, S. R. M. Jones, J. Robinson, C.
Falkenberg, and K. M. Miller. 2007. Development and validation of
an RNA- and DNA-based quantitative PCR assay for determination
of Kudoa thyrsites infection levels in Atlantic Salmon Salmo salar.
Diseases of Aquatic Organisms 75:239–249.
Furey, N. B. 2016. Migration ecology of juvenile Pacific salmon smolts:
the role of fish condition and behaviour across landscapes. Doctoral
dissertation. University of British Columbia, Vancouver.
Garver, K. A., L. M. Hawley, C. A. McClure, T. Schroeder, S. Aldous,
F. Doig, M. Snow, S. Edes, C. Baynes, and J. Richard. 2011. Devel-
opment and validation of a reverse transcription quantitative PCR for
universal detection of viral hemorrhagic septicemia virus. Diseases of
Aquatic Organisms 95:97–112.
Glenn, R. A., P. W. Taylor, and K. C. Hanson. 2011. The use of a real-
time PCR primer/probe set to observe infectivity of Yersinia ruckeri
in Chinook Salmon, Oncorhynchus tshawytscha (Walbaum), and steel-
head trout, Oncorhynchus mykiss (Walbaum). Journal of Fish Dis-
eases 34:783–791.
Grove, S., L. J. Reitan, T. Lunder, and D. Colquhoun. 2008. Real-time
PCR detection of Moritella viscosa, the likely causal agent of winter-
ulcer in Atlantic Salmon Salmo salar and Rainbow Trout Oncor-
hynchus mykiss. Diseases of Aquatic Organisms 82:105–109.
Hallett, S. L., and J. L. Bartholomew. 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–
118.
Hallett, S. L., and J. L. Bartholomew. 2009. Development and applica-
tion of a duplex QPCR for river water samples to monitor the myxo-
zoan parasite Parvicapsula minibicornis. Diseases of Aquatic
Organisms 86:39–50.
Hendrickson, G. L., A. Carleton, and D. Manzer. 1989. Geographic and
seasonal distribution of the infective stage of Ceratomyxa shasta
(Myxozoa) in Northern California. Diseases of Aquatic Organisms
7:165–169.
Hodneland, K., and C. Endresen. 2006. Sensitive and specific detection
of salmonid alphavirus using real-time PCR (TaqMan). Journal of
Virological Methods 131:184–192.
Jeffries, K. M., S. G. Hinch, M. R. Donaldson, M. K. Gale, J. M. Burt,
L. A. Thompson, A. P. Farrell, D. A. Patterson, and K. M. Miller.
2011. Temporal changes in blood variables during final maturation
and senescence in male Sockeye Salmon Oncorhynchus nerka: reduced
CHINOOK SALMON qPCR SCREENING 287

osmoregulatory ability can predict mortality. Journal of Fish Biology
79:449–465.
Jeffries, K. M., S. G. Hinch, M. K. Gale, T. D. Clark, A. G. Lotto, M.
T. Casselman, S. Li, E. L. Rechisky, A. D. Porter, D. W. Welch, and
K. M. Miller. 2014. Immune response genes and pathogen presence
predict migration survival in wild salmon smolts. Molecular Ecology
23:5803–5815.
Johnson, P. T. J., R. S. Ostfeld, and F. Keesing. 2015. Frontiers in
research on biodiversity and disease. Ecology Letters 18:1119–1133.
Jørgensen, A., A. Nylund, V. Nikolaisen, S. Alexandersen, and E. Karls-
bakk. 2011. Real-time PCR detection of Parvicapsula pseudo-
branchicola (Myxozoa: Myxosporea) in wild salmonids in Norway.
Journal of Fish Diseases 34:365–371.
Keeling, S. E., C. L. Brosnahan, C. Johnston, R. Wallis, N. Gudkovs,
and W. L. McDonald. 2013. Development and validation of a real-
time PCR assay for the detection of Aeromonas salmonicida. Journal
of Fish Diseases 36:495–503.
Kent, M. L., and D. J. Speare. 2005. Review of the sequential develop-
ment of Loma salmonae (Microsporidia) based on experimental infec-
tions of Rainbow Trout (Oncorhynchus mykiss) and Chinook Salmon
(O. tshawytscha). Folia Parasitologica 52:63–68.
Korpimaki, A. E., H. Hakkarainen, and G. F. Bennett. 1993. Blood par-
asites and reproductive success of Tengmalm's owls: detrimental
effects on females but not on males? Functional Ecology 7:420–426.
Korsnes, K., M. Devold, A. H. Nerland, and A. Nylund. 2005. Viral
encephalopathy and retinopathy (VER) in Atlantic Salmon Salmo
salar after intraperitoneal challenge with a nodavirus from Atlantic
Halibut Hippoglossus hippoglossus. Diseases of Aquatic Organisms
68:7–15.
LeBlanc, F., M. Laflamme, and N. Gagné. 2010. Genetic markers of the
immune response of Atlantic Salmon (Salmo salar) to infectious salmon
anemia virus (ISAV). Fish and Shellfish Immunology 29:217–232.
Lee, D.-Y., K. Shannon, and L. A. Beaudette. 2006. Detection of bacte-
rial pathogens in municipal wastewater using an oligonucleotide
microarray and real-time quantitative PCR. Journal of Microbiologi-
cal Methods 65:453–467.
Lloyd, S. J., S. E. LaPatra, K. R. Snekvik, K. D. Cain, and D. R. Call.
2011. Quantitative PCR demonstrates a positive correlation between
aRickettsia-like organism and severity of strawberry disease lesions
in Rainbow Trout, Oncorhynchus mykiss (Walbaum). Journal of Fish
Diseases 34:701–709.
Longshaw, M., R.-M. Le Deuff, A. F. Harris, and S. W. Feist. 2002.
Development of proliferative kidney disease in Rainbow Trout,
Oncorhynchus mykiss (Walbaum), following short-term exposure to
Tetracapsula bryosalmonae infected bryozoans. Journal of Fish Dis-
eases 25:443–449.
Løvoll, M., J. Wiik-Nielsen, S. Grove, C. R. Wiik-Nielsen, A. B. Kristof-
fersen, R. Faller, T. Poppe, J. Jung, C. S. Pedamallu, A. J. Neder-
bragt, M. Meyerson, E. Rimstad, and T. Tengs. 2010. A novel
totivirus and piscine reovirus (PRV) in Atlantic Salmon (Salmo salar)
with cardiomyopathy syndrome (CMS). Virology Journal [online
serial] 7(1):309.
Maule, A. G., R. Schrock, C. H. Slater, M. S. Fitzpatrick, and C. B.
Schreck. 1996. Immune and endocrine responses of adult Chinook
Salmon during freshwater immigration and sexual maturation. Fish
and Shellfish Immunology 6:221–233.
Miller, K. M., I. A. Gardner, R. Vanderstichel, T. Burnley, D. Angela,
S. Li, A. Tabata, K. H. Kaukinen, T. J. Ming, and N. G. Ginther.
2016. Report on the performance evaluation of the Fluidigm Bio-
Mark platform for high-throughput microbe monitoring in salmon.
Canadian Science Advisory Secretariat Research Document 2016/038.
Miller, K. M., O. P. Günther, S. Li, K. H. Kaukinen, and T. J. Ming.
2017. Molecular indices of viral disease development in wild migrat-
ing salmon. Conservation Physiology 5:cox036.
Miller, K. M., S. Li, K. H. Kaukinen, N. G. Ginther, E. Hammill, J. M.
R. Curtis, D. A. Patterson, T. Sierocinski, L. Donnison, P. Pavlidis, S.
G. Hinch, K. A. Hruska, S. J. Cooke, K. K. English, and A. P. Farrell.
2011. Genomic signatures predict migration and spawning failure in
wild Canadian salmon. Science 331:214–217.
Miller, K. M., A. K. Teffer, S. Tucker, S. Li, A. D. Schulze, M. Trudel,
F. Juanes, A. Tabata, K. H. Kaukinen, N. G. Ginther, T. J. Ming, S.
J. Cooke, J. M. Hipfner, D. A. Patterson, and S. G. Hinch. 2014.
Infectious disease, shifting climates, and opportunistic predators:
cumulative factors potentially impacting wild salmon declines. Evolu-
tionary Applications 7:812–855.
Mitchell, S. O., T. M. Steinum, E. R. Toenshoff, A. Kvellestad, K. Falk,
M. Horn, and D. J. Colquhoun. 2013. Candidatus Branchiomonas
cysticola is a common agent of epitheliocysts in seawater-farmed
Atlantic Salmon (Salmo salar) in Norway and Ireland. Diseases of
Aquatic Organisms 103:35–43.
Moles, A., and J. Heifetz. 1998. Effects of the brain parasite Myxobolus
arcticus on Sockeye Salmon. Journal of Fish Biology 52:146–151.
Moran, J. D. W., L. Margolis, J. M. Webster, and M. L. Kent. 1999.
Development of Kudoa thyrsites (Myxozoa: Myxosporea) in netpen-
reared Atlantic Salmon determined by light microscopy and a poly-
merase chain reaction test. Diseases of Aquatic Organisms 37:185–193.
Nematollahi, A., A. Decostere, F. Pasmans, and F. Haesebrouck. 2003.
Flavobacterium psychrophilum infections in salmonid fish. Journal of
Fish Diseases 26:563–574.
Nigrelli, R. F., K. S. Pokorny, and G. D. Ruggieri. 1976. Notes on
Ichthyophthirius multifiliis, a ciliate parasitic on fresh-water fishes,
with some remarks on possible physiological races and species. Trans-
actions of the American Microscopical Society 95:607–613.
Nylund, A., K. Watanabe, S. Nylund, M. Karlsen, P. A. Sæther, C. E.
Arnesen, and E. Karlsbakk. 2008. Morphogenesis of salmonid gill
poxvirus associated with proliferative gill disease in farmed Atlantic
Salmon (Salmo salar) in Norway. Archives of Virology 153:1299–
1309.
Nylund, S., A. Nylund, K. Watanabe, C. E. Arnesen, and E. Karlsbakk.
2010. Paranucleospora theridion (Microsporidia, Enterocytozoonidae)
with a life cycle in the salmon louse (Lepeophtheirus salmonis, Cope-
poda) and Atlantic Salmon (Salmo salar). Journal of Eukaryotic
Microbiology 57:95–114.
Okamura, B., A. Gruhl, and J. L. Bartholomew, editors. 2015. Myxo-
zoan evolution, ecology and development. Springer, Berlin.
Olson, R. E., and R. A. Holt. 1995. The gill pathogen Dermocystidium
salmonis in Oregon salmonids. Journal of Aquatic Animal Health
7:111–117.
Ortuño, J., M. A. Esteban, and J. Meseguer. 2002. Lack of effect of
combining different stressors on innate immune responses of seabream
(Sparus aurata L.). Veterinary Immunology and Immunopathology
84:17–27.
Plarre, H., M. Devold, M. Snow, and A. Nylund. 2005. Prevalence of
infectious salmon anaemia virus (ISAV) in wild salmonids in western
Norway. Diseases of Aquatic Organisms 66:71–79.
Powell, M., K. Overturf, C. Hogge, and K. Johnson. 2005. Detection of
Renibacterium salmoninarum in Chinook Salmon, Oncorhynchus tsha-
wytscha (Walbaum), using quantitative PCR. Journal of Fish Diseases
28:615–622.
Purcell, M. K., R. L. Thompson, K. A. Garver, L. M. Hawley, W. N.
Batts, L. Sprague, C. Sampson, and J. R. Winton. 2013. Universal
reverse-transcriptase real-time PCR for infectious hematopoietic
necrosis virus (IHNV). Diseases of Aquatic Organisms 106:103–115.
Raby, G. D., M. R. Donaldson, V. M. Nguyen, M. K. Taylor, N. M.
Sopinka, K. V. Cook, D. A. Patterson, D. Robichaud, S. G. Hinch,
and S. J. Cooke. 2014. Bycatch mortality of endangered Coho Sal-
mon: impacts, solutions, and aboriginal perspectives. Ecological
Applications 24:1803–1819.
288 TEFFER AND MILLER

Raida, M. K., and K. Buchmann. 2008. Development of adaptive immunity
in Rainbow Trout, Oncorhynchus mykiss (Walbaum) surviving an infec-
tion with Yersinia ruckeri. Fish and Shellfish Immunology 25:533–541.
Riddell, B., M. J. Bradford, R. Carmichael, D. Hankin, R. M. Peterman, and
A. Wertheimer. 2013. Assessment for the status and factors for decline of
southern B.C. Chinook Salmon: independent panel's report. Fisheries and
Oceans Canada, Final Report, Vancouver.
Rise, M. L., S. R. M. Jones, G. D. Brown, K. R. von Schalburg, W. S.
Davidson, and B. F. Koop. 2004. Microarray analyses identify molec-
ular biomarkers of Atlantic Salmon macrophage and hematopoietic
kidney response to Piscirickettsia salmonis infection. Physiological
Genomics 20:21–35.
Sofonea, M. T., S. Alizon, and Y. Michalakis. 2017. Exposing the diver-
sity of multiple infection patterns. Journal of Theoretical Biology
419:278–289.
Starliper, C. E. 2011. Bacterial coldwater disease of fishes caused by
Flavobacterium psychrophilum. Journal of Advanced Research 2:97–108.
Sun, H. Y., J. G. Noe, J. Barber, R. S. Coyne, D. Cassidy-Hanley, T. G.
Clark, R. C. Findly, and H. W. Dickerson. 2009. Endosymbiotic bac-
teria in the parasitic ciliate Ichthyophthirius multifiliis. Applied and
Environmental Microbiology 75:7445–7452.
Teffer, A. K., A. L. Bass, K. M. Miller, D. A. Patterson, F. Juanes, and
S. G. Hinch. 2018. Infections, fisheries capture, temperature and host
responses: multi-stressor influences on survival and behaviour of adult
Chinook Salmon. Canadian Journal of Fisheries and Aquatic Sciences
75:2069–2083.
Teffer, A. K., S. G. Hinch, K. M. Miller, D. A. Patterson, A. P. Farrell,
S. J. Cooke, A. L. Bass, P. Szekeres, and F. Juanes. 2017. Capture
severity, infectious disease processes and sex influence post-release
mortality of Sockeye Salmon bycatch. Conservation Physiology 5(1):
cox017.
Telfer, S., M. Bennett, K. Bown, D. Carslake, R. Cavanagh, S. Hazel, T.
Jones, and M. Begon. 2005. Infection with cowpox virus decreases
female maturation rates in wild populations of woodland rodents.
Oikos 109:317–322.
Tucker, S., S. Li, K. H. Kaukinen, D. A. Patterson, and K. M.
Miller. 2018. Distinct seasonal infectious agent profiles in life-his-
tory variants of juvenile Fraser River Chinook Salmon: an appli-
cation of high-throughput genomic screening. PLoS ONE [online
serial] 13(4):e0195472.
White, V. C., J. F. Morado, L. M. Crosson, B. Vadopalas, and C. S.
Friedman. 2013. Development and validation of a quantitative PCR
assay for Ichthyophonus spp. Diseases of Aquatic Organisms 104:69–
81.
Wiik-Nielsen, C. R., P. M. R. Ski, A. Aunsmo, and M. Løvoll. 2012.
Prevalence of viral RNA from piscine reovirus and piscine myocardi-
tis virus in Atlantic Salmon, Salmo salar L., broodfish and progeny.
Journal of Fish Diseases 35:169–171.
Wolinska, J., and K. C. King. 2009. Environment can alter selection in
host-parasite interactions. Trends in Parasitology 25:236–244.
Woo, P. T. K. 2003. Cryptobia (Trypanoplasma) salmositica and salmo-
nid cryptobiosis. Journal of Fish Diseases 26:627–646.
CHINOOK SALMON qPCR SCREENING 289