CAST SAS Plan December 2017

Review of the risks and benefits of
National Capital Region CAST SAS Experiment Proposal
APPENDICES
COLLABORATION FOR ATLANTIC SALMON TOMORROW (CAST)

SMOLT-TO-ADULT-SUPPLEMENTATION (SAS) EXPERIMENT
PROPOSAL: PHASE I (2018-2022)
SEEKING ANSWERS TO IDENTIFIED SCIENTIFIC KNOWLEDGE GAPS
REGARDING THE SMOLT-TO-ADULT SUPPLEMENTATION CONSERVATION
STRATEGY
4 December, 2017
Prepared by:
Tommi Linnansaari 1*, R. Allen Curry 2*, Kyle Wellband 3, and Mark Hambrook 4
*
Corresponding authors
1
Canadian Rivers Institute, Department of Biology, University of New Brunswick, NB , Canada.
Email: [email protected]; Ph. 1 506 458 7569
2
Canadian Rivers Institute, Department of Biology, and Faculty of Forestry and Environmental
Management University of New Brunswick, NB , Canada.
Email: [email protected]; Ph. 1 506 452 6208
3
Canadian Rivers Institute, Department of Biology, University of New Brunswick, NB , Canada.
4
Miramichi Salmon Association, South Esk, NB, Canada
18
Table of Contents
1. Background........................................................................................................................20
2. General Description of the SAS Experiment.......................................................................21
2.1. Objectives of the SAS Program...................................................................................21
2.2. Experimental Components – General Description .......................................................22
2.3. Area of Experiment – The Miramichi River ..................................................................23
3. Smolt Capture, Operational Protocols in Captivity, and Status Quo of the Collections to
Date ...................................................................................................................................27
3.1. Smolt collections .........................................................................................................27
3.2. Operational Protocols in Captivity ...............................................................................27
3.3. Status quo (31 October 2017) of the CAST SAS program...........................................30
4. Genetic Monitoring Tools of the Program; General Approach ............................................31
5. Detailed description of proposed studies, hypothesis framework, and knowledge gaps to be
addressed ..........................................................................................................................34
5.1. Program 1: Sub-basin genetic structure of Atlantic salmon on the Miramichi River .....34
5.2. Program 2: Laboratory experiments ............................................................................37
5.3. Program 3: Experimental River ...................................................................................39
5.4. Program 4: SAS impacts on a natural river .................................................................42
6. Considerations related to First Nations perspectives ..........................................................51
7. Literature Cited: .................................................................................................................53
19
1. Background
Canada's Policy for Conservation of Wild Atlantic Salmon (DFO 2009) dictates that
management intervention should increase when Atlantic salmon (Salmo salar) populations
decline below selected benchmark values under which a population is no longer considered to
be healthy. Declines in Atlantic salmon populations below conservation targets have been
observed in many index rivers in Atlantic Canada (e.g., ICES 2017). Supplementation, or
“stocking”, is often seen as a rational and reasonable response to a situation where a population
of salmon is depleted, or perceived to be depleted (IBIS 2013).
Scientific consensus on salmon stocking suggests that assumptions of net benefits of traditional
stocking programs are frequently not valid (IBIS 2013). Indeed, controlled studies have found
that progeny of hatchery fish have decreased fitness in the wild compared to progeny of wild
fish (e.g. Reisenbichler and Rubin 1999).
Management intervention is important because the risks to population perseverance increase as

a population declines below its established conservation target (CAFSAC 1991). These risks
include, but are not limited to accentuation of annual fluctuations in run size, increased
susceptibility to extinction from genetic, demographic, or environmental catastrophes,
decreases in productivity, and permanent changes in demographic characteristics of the
spawning population (CAFSAC 1991).
Smolt-to-adult supplementation (SAS; a term coined by Fraser (2016)) has emerged as a

potential conservation strategy for Atlantic salmon (and other salmonid) rivers where
populations have declined below targeted conservation levels (CSAS 2016). The SAS strategy,
in its simplest form, captures outward migrating, wild Atlantic salmon smolts and rears them to
adults in either fresh- or saltwater (CSAS 2016). Upon maturity, the adults are returned to the
river of their origin to spawn in the wild. Such a strategy is hypothesized to be beneficial in
populations where the population decline is attributable to high at-sea mortality, i.e., SAS
effectively circumvents this at-sea bottleneck for the population. The SAS strategy has benefits
in comparison to traditional juvenile stocking methods based on broodstock collections;
however, the SAS strategy may similarly have risks that are not thoroughly understood (Fraser
2016).
The SAS strategy is not a novel idea. Smolt-to-adult supplementation first emerged in Thomas
(1996) as a conservation strategy for rainbow (steelhead) trout (Oncorhynchus mykiss). Thomas
(1996) outlined the scientific and monitoring needs, potential risks, and potential promise as a
conservation tool. Fraser (2016) provided a detailed outline of the potential risks and benefits
related to the SAS strategy, and recently, a 20-year report of experiences of adult
supplementation in Idaho was compiled and has added significantly to the scientific knowledge-
base regarding the strategy (Kozfkay et al. 2017).
The SAS strategy has already been implemented where wild salmonid populations have
declined below conservation targets (Dempson and Furey 1997; Dempson et al. 1999;
Berejikian et al. 2008; Jones et al. 2014; Venditti et al. 2013; Kozfkay et al. 2017), and ongoing
20
SAS programs for Atlantic salmon exist in New Brunswick in the Saint John River by Fisheries
and Oceans Canada (Outer Bay of Fundy populations; releases ranging from 339 to 1348 adult
SAS spawners between 2003 to 2015 in sympatry with wild salmon; Jones et al. 2015) and in
the Upper Salmon River by Parks Canada (Inner Bay of Fundy populations; releases ranging
from 429 to 965 adult SAS spawners between 2015 to 2017 in allopatry; Corey Clarke, Parks
Canada, unpubl. data). In New Brunswick, the preliminary observations and anecdotal data
suggest successful spawning and population maintenance (Jones et al. 2015; C. Clarke, Parks
Canada, unpubl. data). Still, there are a number of key knowledge gaps and generally, a paucity
of data exists that assess SAS individuals and their progeny against their truly wild counterparts,
particularly in the natural environment (CSAS 2016).
Collaboration for Atlantic Salmon Tomorrow (CAST) was created to address key knowledge
gaps in our understanding of the Atlantic salmon populations of eastern Canada and specifically
those factors hypothesized to be limiting factors for population success. One major component
of CAST is the rigorous assessment of the SAS strategy as a supplementation tool. The
identified knowledge gaps associated with the use of SAS as a management tool are outlined in
CSAS (2016). These questions form the science premise for the SAS studies proposed in the
CAST program. Herein, we identify how the proposed studies address the current knowledge
gaps. To that end, the CAST premise is that the proposed SAS project in the Miramichi River,
New Brunswick, is a science experiment seeking to compare the quantifiable aspects of the
SAS fish to their wild counterparts across a variety of experimental settings and thus truly
understand the merit of this conservation strategy which includes assessing the possible risks
and benefits for a salmon population’s recovery.
This document outlines the proposed CAST SAS studies. This is a novel experiment and thus it
will always be adaptive in structure, i.e., as new knowledge is gained, the experimental
components may require adjustments. In addition, the core science team of CAST has been
the Canadian Rivers Institute (CRI) at the University of New Brunswick (UNB), Universite Laval,
Cooke Aquaculture, and the Miramichi Salmon Association (MSA), but also includes Fisheries
and Oceans Canada (DFO – Gulf Region) and Mi'gmawe'l Tplu'taqnn Incorporated (MTI). Once
all parties are satisfied with the review of the current document, then the proposal will move
from a “draft” to “final” version.
2. General Description of the SAS Experiment

2.1. Objectives of the SAS Program
The main objective of the CAST SAS program is to determine if Smolt-to-Adult Supplementation
is a functional conservation strategy that can be used to supplement Atlantic salmon
populations in situations where conservation targets (as defined by authorities managing
respective populations) are not being met due to high at-sea mortality but before populations
decline into state necessitating at-risk classification. A large number of knowledge gaps
pertaining to SAS strategy have been identified (CSAS 2016; Fraser 2016); the objective of
CAST SAS program is to answer the most critical of the identified knowledge gaps, including the
most important question to be addressed (as per CSAS 2016): quantify and compare the
21
lifetime fitness of SAS progeny versus wild progeny in the natural environment to examine the
extent to which SAS may reduce marine adaptation.
The CAST SAS program is designed as a science experiment aimed at providing answers not
only to the objective above, but also to provide a better understanding of potential phenotypic
and genotypic deviations between SAS and wild Atlantic salmon and consequences of potential
deviations. Specific hypotheses to be addressed by the different SAS studies are detailed in
Section 5. Ultimately, the studies aim to determine if juvenile fish (smolts) collected from the wild
and grown into adulthood in a captive environment (F0 generation) produce progeny in wild? If
yes, then are their progeny (F1 generation) viable and therefore, does the supplementation
strategy fulfill the objectives of producing added, wild-like progeny into the system? There are
several aspects of life history that may be different in SAS vs. wild fish and to this end, the
CAST SAS project will monitor the performance of the SAS fish compared to wild in four levels
of studies.
The conservation targets of the Atlantic salmon population are consistently not being achieved
in the Miramichi River system (see Section 2.3; CSAS 2017a). Based on DFO’s Atlantic salmon
management policy, the SAS strategy is a candidate as a conservation tool in the near future if
the population trajectory remains unaltered in the Miramichi River (CAFSAC 1991; DFO 2009).
Consequently, it is critical to first assess the effectiveness and risks of a SAS conservation
strategy before a full implementation at a scale required to achieve adequate conservation
status in this system (see Size of the experiment in section 5.4). Importantly, the CAST SAS
project is an experiment and not a stocking program.
2.2. Experimental Components – General Description

The CAST SAS program is proposed as a two-phased approach at four scales of study. In
Phase 1, wild smolt collections are proposed to occur for five years (five smolt “cohorts”) during
the period 2016-2020. The first release of maturing salmon is proposed in the autumn of 2018
and continue to 2023 when the last salmon from the 2020 smolt cohort are predicted to be fully
mature (as Multi-Captivity Winter salmon, or MCW,) and be ready for release. While the Phase
1 is characterized by the smolt collections and adult salmon releases, the monitoring begins
already in Phase 1. Phase 2 is monitoring period with no experimental releases, minimally
continuing until 2029 when the first Multi-Sea Winter (MSW) salmon are expected to return from
the last release cohorts (of 2023).
The first level of study will establish the genetic structure among tributaries of the Miramichi
River. It is a fundamental baseline and building block for the other experiments, i.e., it
establishes the foundation for rearing smolts (relevance based on tributary “strains”),
establishes the baseline level of genetic uniqueness between the tributaries and lays the
groundwork for future management planning using supplementation strategies. Data from these
studies will also be used relevant genetic (parentage) tool for monitoring the success of SAS
releases. The second level of study is the necessary laboratory experiments that provide a
fully controlled environment where basic SAS fish characteristics can be assessed, e.g.,
maturation rate and size at maturation, compared to wild fish performance. Fundamental
22
information on fecundity, fertilization success, egg size and fitness relative to wild salmon,
defined as survival to eyed stage, survival to hatch (yolk-sac fry) and survival to swim-up (start
of external feeding) will be assessed. These parameters are a key to development of the rearing
practices to maximize survival and health of SAS fish in captivity.
The third level of study will be a controlled experiment in a small, natural stream to understand
how SAS fish will respond to natural environment and interact with wild salmon. The proposed
stream, Northwest Millstream (NWMS), flows directly to the Miramichi River estuary (Figure 1).
The NWMS is annually blocked with numerous impassable or near impassable beaver dams
and salmon migration to upstream sections where the experiment is proposed is naturally highly
impeded. A barrier fence, with a trap to account for any wild fish that may migrate up to the
fence will be erected in the upstream section of the NWMS and will be maintained during the
autumn of each year of the experiment.
The fourth level of study is an experiment in natural rivers, with controls, where the population
scale assessments of the SAS strategy occur. Monitoring the behaviour and spawning success
of the SAS adults relative to wild counterparts and the subsequent density, survival, growth and
genotypic and phenotypic differences in resulting juveniles will involve a variety of telemetry,
genetic and other field-based methods.
2.3. Area of Experiment – The Miramichi River

The Miramichi River consists of two major branches: the Southwest Miramichi (approx. 7 700
km2 drainage area) and the Northwest Miramichi (approx. 3 900 km2 drainage area; Figure 1)
(Chaput et al. 2016). The Miramichi River has historically had the largest run of Atlantic salmon
in North America (Chaput et al. 2016). However, its Atlantic salmon population has experienced
a multi-decadal decline and adult returns to the river in 2014 were the lowest in recorded history
meeting only 22 % of the approx. 7300 MSW female salmon conservation requirement (Randall
1985; Douglas et al. 2015), in the Northwest Miramichi River (CSAS 2015).
The Atlantic salmon population in the Miramichi River is managed as a composite, but DFO has
generated branch-specific, stock monitoring for the Southwest (SW) and Northwest (NW)
Miramichi River since 1992 (CSAS 2015). DFO (2006) prescribes management planning
guidelines, including triggers for intervention when a stock’s status declines from “Healthy” to
“Cautious” to “Critical” status. The critical “Cautious” and “Critical” change occurs at 2.4 eggs /
m2 of fluvial, rearing habitat (Elson 1957; Gibson and Claytor 2012) and represents a level
below which it is hypothesized that serious harm occurs to the stock (CAFSAC 1991). Based on
these management thresholds, the overall Miramichi River has achieved “Cautious” status only
three times over the past 20 years, while being in the “Critical” level for the remaining years
(CSAS 2017a).
There is a disparity in the performance of the two main branches regarding conservation targets
(CSAS 2017a). The Northwest Miramichi has met the conservation requirements only three
times in the last two decades and underperforms in relation to the Southwest (CSAS 2017a).
The Northwest Miramichi still meets general standards for a minimum viable population, or a
median estimate of 4 169 individuals based on a meta-analysis of 30 years of published
23
estimates for vertebrates (Traill et al. 2007). The consequence of the depressed Atlantic
salmon population status is the increased vulnerability to inverse density dependence or the
Allee effect (Allee et al. 1949). While effects of inverse density-dependent mechanisms are not
widely studied, and are generally poorly understood for salmon populations, possible
mechanism of the Allee effect for the Atlantic salmon in the Miramichi, is relatively lower survival
because antipredator strategies become inefficient in small groups of prey (Courchamp et al.
1999). Such mechanism may manifest especially if predator-prey dynamics are unbalanced,
e.g., the very abundant Striped Bass population in river’s estuary which is estimated to be
experiencing a >10-fold exceedance of its conservation target (CSAS 2017b).
Supplementation activities in the Miramichi River have a long history. The Miramichi Salmon
Conservation Centre, where proposed SAS fish husbandry will take place, is the oldest fish
hatchery in Canada (est. 1873). Supplementation methods and quantities have varied over the
years including supplementation of first-feeding fry, fall fingerlings, 1+ parr, and smolts (Chaput
et al. 2016). Supplementation activities occur annually in both the Northwest and Southwest
systems. Stocking numbers have ranged from 13 000 to 133 000 fish per year in the Northwest
(excluding Little Southwest, where additional 800 to 106 400 have been stocked annually) and 9
000 to 469 400 in the Southwest Miramichi in the period of 1978 to 2008 (Chaput et al. 2016);
other stocking programs and quantities precede those compiled in Chaput et al. (2016). Since
2010, the supplementation activity has been first-feeding fry stocking (occurring in early
summer) and generally targeting areas where natural production in previous years have been
determined to be low (MSA 2016). In 2016, approximately 91 171 first-feeding Atlantic salmon
fry were stocked into 45 sites; 15 483 and 75 688 fry were stocked in the Northwest and
Southwest Miramichi, respectively (MSA 2016). Unfortunately, there has been very little
assessment of the efficiency of the stocking strategies in the Miramichi system. In a recent
study, Wallace and Curry (2017) determined the effectiveness of juvenile stocking was
undetectable. CAST SAS plan includes monitoring the success of the current broodstock
program in rivers where SAS activity will take place using genetic tools.
The “Experimental Stream” for the controlled studies is a small stream where the history of past
anthropogenic activity and logistics of maintaining a barrier fence during autumn spawning
period make. Northwest Millstream (NWMS) as the ideal candidate (Figure 1). The NWMS was
obstructed by a dam without a fish pass in 1947 (Moore & Chaput 2007). The center spillway of
the dam was opened in 1979 to allow spawner access above the dam and NWMS was stocked
(stock source unknown) in 1970’s and 1980’s. The dam was removed in 2005. Juvenile surveys
(1994 to 2004) upstream of the dam indicate that the areas upstream of the dam were
inaccessible in most years (Moore & Chaput 2007).
A conduit fence that acts as a barrier to adult salmon emigration from the experimental area can
feasibly be maintained 13 km from the stream mouth (M. Hambrook, personal observations; T.
Linnansaari, unpubl. field data from 2017). The experimental area upstream is approximately 19
km in linear stream length consisting of 3rd (10km) and 4th (9km) order streams and associated
1st and 2nd order streams.
24
The population-scale SAS experiments in natural rivers is proposed for the Northwest Miramichi
system. The sub-basins selected for the SAS studies are proposed as:
1) Little Southwest Miramichi River (LSW) upstream from Sillikers (Upper Oxbow; Figure
1); and
2) The main Northwest Miramichi River (NWM) upstream of mouth of Trout Brook (Figure
1).
These are the locations for smolt collections, and adult releases are planned upstream in the
first, suitably-sized holding pools with accessibility using fish transport truck (Figure 1). These
are appropriately sized tributaries (and sub-populations) for assessing a population-scale effect
in the wild. In addition, these sub-basins provide feasible logistics for monitoring in the SAS
experiment, e.g., smolt collections, automating salmon counts using sonar, electrofishing, and
PIT-tag reader deployment (details in Section 5.4).
The natural structure of the overall Miramichi River watershed provides for several control
systems for the SAS experiments where the juvenile abundance, relative to SAS rivers, can be
monitored. Spatially, the Sevogle River is a control population that exists within the Northwest
system and the Main Southwest provides an untreated subwatershed (Figure 1). Juvenile
assessment data exists for the experiment and control rivers since the 1970’s (e.g. Moore &
Chaput 2007), and monitoring is ongoing by the DFO and will be supplemented by CAST in
coordination with DFO. Adult return assessments exist for the Main Southwest and Northwest
composite. Both juvenile and adult data sets are long-term, providing a temporal assessment in
the control set. In addition, there are 10+ rivers where similar data sets exist thus providing
more control sets for the SAS experiment in terms of potential population recovery trends.
Although high natural variability in juvenile density and size data exists, it is hypothesized that
the average density in the SAS intervention rivers will increase above background levels in
proportion to the SAS fish releases while the densities are predicted to remain at background
levels in control rivers as dictated by the returns of only wild salmon (assuming population
trajectories remain similar or decline further). The growth of juvenile salmon is predicted to
remain relatively unaltered relative to control rivers as the effects on juvenile growth are typically
observed via exploitation competition already at low population densities (Grant & Imre 2005;
Imre et al. 2005).
25
National Capital Region Review of the risks and benefits of CAST SAS Experiment Proposal
Figure 1. The Miramichi River in New Brunswick, Canada, highlighting the main sub-watersheds associated with the SAS experiments: the
Experimental Stream (Northwest Millstream), and the two tributaries with proposed SAS intervention (Little Southwest and Northwest). The
orange fish symbols highlight the areas where wild smolts are collected and the red fish symbols indicate the release locations of the mature
adults. The Miramichi Salmon Conservation Centre (“Hatchery”) is where salmon are being raised to maturity.
26
3. Smolt Capture, Operational Protocols in Captivity, and

Status Quo of the Collections to Date
3.1. Smolt collections
Wild Atlantic salmon smolts are annually collected by the Miramichi Salmon Association staff
during the month of May using standard rotary screw traps (RST) on the LSW Miramichi
(Sillikers) and NW Miramichi upstream of Trout Brook (Figure 1). The RSTs are installed as
soon as the water level recedes to operable conditions, typically in the early part of May prior to
start of the smolt migration. Two or three RSTs are operated in each tributary, depending on
availability.
The traps are fished each morning and 25 smolts are hand counted into pails that have a
perforated cover and holes in the upper half of the pail to allow for water movement through the
pail when in the transportation tank, yet can be carried while still retaining ½ pail of water. The
transportation tank can accommodate nine pails, or 225 smolts / lot. The protocol of smolt
transportation by individual pails ensures only one handling of the smolts (from RST to pail) and
the smolts are simply poured into their initial holding tanks at the Miramichi Salmon
Conservation Centre (without additional stress of netting, etc.).
Smolts for the SAS experiments are collected randomly and throughout the full smolt migration
period to avoid unintentional selection as per Fraser (2016). At the start and end of the
migration, smaller quantities of smolts are collected (less than a lot of 225 / day), however,
multiple lots of 225 smolts will be collected during the peak migration days. Based on the
experience from smolt migration in previous years, it is anticipated that approximately 15 % of
the total catch will be collected in the first and last 10 days of migration, and 70 % of the total
will be collected during a week of most intense migration typically in mid-May.
3.2. Operational Protocols in Captivity

Facilities
The Miramichi Salmon Conservation Centre (MSCC) is a Miramichi Watershed Management
Committee facility and operated long term by the Miramichi Salmon Association. It is located at
South Esk, New Brunswick (Figure 1; “MSA Fish Hatchery”). The MSCC has been used
primarily for juvenile supplementation programs, but many other husbandry operations over the
years, and it is currently being modified for the SAS program-specific needs. Construction of
new facilities is near complete at this time (December 2017), and herein we describe the
facilities as they are planned (predicted completion January 2018). SAS fish are cared for under
the supervision of both the experienced MSCC and Cooke Aquaculture experts.
Each new smolt cohort will first enter a Quarantine Building (QB) for five months, then move to
Big Greenhouse (BG) where each stock-specific smolt cohort will be held in their own tanks until
maturity without the need to transfer fish between tanks thus minimizing transport and handling.
The buildings and their tank systems are equipped with secure, back-up systems ensuring fish
welfare in case of emergencies, i.e., to avoid non-random die-offs that could lead to
unintentional selection in the hatchery environment (Fraser 2016).
27
Quarantine Building ‒ The QB has 10 – 2.5m (diameter) tanks where the new smolts are held
from May until late autumn of each year. The water to QB is from deep wells that is UV-
sterilized, degassed and oxygenated. The well water is heated indirectly by a plate heat
exchanger that extracts heat from the brook water to provide fluctuation to the water
temperature; temperature range is predicted to be 7 – 14 °C (heat exchanger is currently being
installed). Natural light is used in the tank area, however, overhead fluorescent lights are turned
on when staff are working in the building (coinciding with daytime hours). Each tank is
continuously monitored for oxygen level and each tank is equipped with plate diffusers with
back-up oxygen bottles in case of emergencies. This building and the adjacent well pump has a
backed-up, emergency diesel generator and a second emergency generator. Oxygen is
generated by electricity. An alarm system monitors water levels and pressure connected to a
telephone answering service. Access to the building is controlled including personnel wearing
room-specific footwear. Entry to the tank room is via a disinfection station including a footbath
and hand sanitizing station.
Big Greenhouse ‒ This building is undergoing significant changes to accommodate the SAS
program. The tank-space of the building is enclosed by a large (white) industrial tarp where the
fish receive diffused, natural light. The renovations involve dividing the building into two
separate units with a dividing wall and disinfection station between the two sections such that
each section will house a separate cohort of SAS fish where they will live throughout their
captive growth period. Six new glass-coated steel circular tanks 6.8m (diametre) x 2.1m (depth)
have been installed in each end inside existing tanks. Two extra tanks in each side are used for
water treatment equipment. Each tank has a centre drain for waste and a side pod where
clearer water is removed to be treated and recycled. The water from both drains go through
separate drum filters to remove solids and subsequently, a portion is pumped through a UV-
sterilizer to an aeration facility where CO2 is removed and oxygen is added before returning to
the tanks. The new well water is UV-sterilized, heated by using a plate heat exchanger to
capture heat from brook water and then degassed and oxygenated providing fluctuating
temperature. Water temperature range is predicted to fluctuate between 7 and 12 °C in summer,
while the temperature will be maintained at 7 °C in winter to maintain fish growth. Each tank has
continuous oxygen monitoring equipment and if oxygen levels drop, back-up oxygen through
plate diffusers automatically turns on in each tank.
Each section of the BG has its own alarm system that monitors oxygen and water levels to
numerous areas. The alarms are reported through telephone lines or a cellular network to an
answering service. Each tank is equipped with a side window that can be removed to allow
access through the tank wall to remove fish. A new tarp on the building has ports to permit a
pipe to protrude through the wall to a holding tank to facilitate moving the large salmon to the
transport (stocking) truck upon maturity. All the equipment in this facility and two adjacent
production wells are backed-up by two large diesel generators, with one generator backing up
the other in case it fails. Disinfection stations, as described above, are also in use for anyone
working this building.
During the first two years of collections (2016 and 2017), smolts have been held in existing
tanks at the MSCC while planning and construction has been occurring. A smolt cohort
28
collected in 2015 that has been used to start preliminary laboratory analysis and tracking
studies, has been in new BG tanks since March 2017 and the 2016 cohort will go into new tanks
in December 2017. The 2017 cohort entered into the modified new tanks in the refurbished QB
and will be moved into the new large grow-out tanks in BG upon completion in winter 2018.
Fish Husbandry Procedures (dietary and health practices)
Dietary practices are of critical importance to transition the wild smolts onto consume of
commercial fish feed. Upon arrival to the facility, the smolts are initially fed chopped krill for a
week after which a semi-moist food mixture (Cooke Aquaculture) is introduced during a
“transitional period”. The krill has proven to be critical as a transitional diet to ensure smolts shift
from natural food items to commercial feed (M. Hambrook, personal observations). The semi-
moist food is mixed with krill in order to entice the smolts to feed. Chopped krill is continued to
be supplied during the “transitional period” after the fish have been given abundant, semi-moist
food to ensure the transition has the greatest potential for success for each fish. After two
weeks of feeding semi-moist/krill mixture, plain semi-moist food is introduced and fed to
satiation and a semi-moist/krill mixture is fed in addition so that even the fish that are slower in
transitioning will be fed. After two weeks, the smolts are introduced to dry food (Skretting). Dry
pellets are fed first until feeding rate is reduced, and then the dry pellets mixed with krill is fed
until feeding rate is reduced further followed by the semi-moist food to ensure every fish is
feeding. The fish are fed slowly and observed carefully to judge how much food is being
consumed thus minimizing food waste and its accumulation in the tank.
To date, it has been observed that introduction to dry food is the longest transition period. Dry
food is also placed on belt feeders to feed into the evening. By late July, the fish consume dry
food only. Once the fish have been habituated to dry food with automatic feeders, hand feeding
still occurs four times daily to observe fish behaviour. This will continue until salmon are near
ready to be released back into the wild and have stopped feeding.
Fish healthcare is proactive. As a preventative treatment, smolts in the Quarantine Building are
given a salt bath or formaldehyde treatment every week. After moving into the BG, fish are
given a salt bath every two weeks. The bath is a 2% salt mixture that is premixed in a separate
tub and pumped into the tanks. Formaldehyde is mixed at a 1:4,000 ratio and is sprinkled into
the tank. Both treatments are one hour exposures with the water flow turned off with the
exception for a trickle of high oxygenated water going into the tanks.
Daily food allocations are recorded as well as treatments, mortalities, and visitors to the
buildings. Mortalities, when they occur, are removed as soon as noticed and may be kept for
veterinarian/lab analysis or disposed. A dedicated veterinarian from Cooke Aquaculture is on
stand-by on a priority basis and is available should any indication of a disease outbreak be
evident.
A separate handling event for each smolt cohort is planned to inventory the fish numbers,
measure their size, to individually tag each fish with a Passive Integrated Transponder (PIT-tag;
23mm glass encapsulated half-duplex “silver bullet” manufactured by OregonRFID), and to
collect tissue material for genetic analysis (fin clip). PIT-tags will be injected into the flesh under
the dorsal fin; vertical injection is being currently experimented with to optimize detection
29
distance during the in-stream monitoring phase which uses a flat-bed PIT antennas (see
Section 5.4). The genetic material is required for establishing the parentage analysis for
monitoring (see Section 4). The fish are externally tagged using tributary-specific, colored T-
anchor tags and the adipose fin of the fish is removed for additional external identification
purposes.
3.3. Status quo (31 October 2017) of the CAST SAS program
Wild Atlantic salmon smolt collections for the CAST SAS program started in 2015, and have
continued in spring of 2016 and 2017. The initial intention was that the 2015 cohort would be
released in the autumn of 2017, however, the release did not take place due to pending agency
approvals. Certain laboratory studies (see Section 5.2) and radiotracking of 40 SAS individuals
in the Experimental stream (Section 5.3) were authorized.
In 2015, an authorized collection of 1,100 smolts from the LSW Miramichi River (May 17 to 27),
191 smolts from the Sevogle River (May 20 to 24), and 200 smolts from the NW Miramichi River
upstream of Trout Brook (May 16 to May 27) were secured for a total of 1,491 fish introduced
into the MSCC facilities. In 2016, 2500 (May 13 to 27) and 2132 (May 11 to June 3) smolts
were collected from LSWM and NWM, respectively. In 2017, 2500 smolts were collected from
both systems (May 16 to 29)).
The practice of collecting wild smolts and raising them to maturity is new to the MSCC. The first
years of a new husbandry initiative are necessarily challenging as both fish handling and
facilities must be continuously adapted to maximize successful rearing. Unfortunately, in 2015
significant mortalities occurred due to a “failed smolt syndrome” where the new smolts didn’t
accept the food that was offered. Smolts that did start feeding performed very well with minimal
mortalities thereafter. The diet was modified in 2016, i.e., the introduction of the krill and semi-
moist diets with appropriate transition periods, as described above, and the smolts readily
accepted the food. However, late in the summer 2016, a large mortality event occurred as a
result of the protozoa Costia (Ichtyobodo spp) on the gills that came into the facility with the fish.
Again, improving from 2016 by developing appropriate treatment and filtering protocols with on-
demand priority veterinarian service, the 2017 smolt cohort collection is feeding well with a
regular preventative treatment regime in place. The current collections and survival (or %
remaining) of SAS smolts is shown in Table 1.
Table 1. Numbers of wild Atlantic salmon smolts collected from tributaries of the Miramichi River by stock
and surviving numbers in 31 October (2017) at the Miramichi Salmon Conservation Centre. *denotes
numbers where the number of fish is also affected by the use of 2015 mature fish for the laboratory and
tracking purposes, and survival for this cohort of fish is better described as the number of fish remaining
in the facility.
Smolt Initial Survivors (31 Survival
Cohort Stock Collection October 2017) (%)
2015 Northwest 200 148 74.0
2015 Sevogle 191 59* n/a
2015 Little Southwest 1100 362* n/a
2016 Northwest 2132 1710 81.8
30
Smolt Initial Survivors (31 Survival

Cohort Stock Collection October 2017) (%)
2016 Little Southwest 2500 1205 48.3
2017 Northwest 2500 2255 90.3
2017 Little Southwest 2500 2092 83.5
It should also be noted that the collection of 5 000 smolts annually represents a very small risk
to the wild Atlantic Salmon population in the Miramichi River given the assumed very high
mortality rate at sea. Assuming 80:20 grilse:MSW ratio (Chaput et al. 2016) and 1% survival for
grilse and 0.2 % survival for MSW salmon (Chaput et al. 2016), 5 000 smolt translates to ~40
grilse and ~2 MSW salmon lost to the smolt collections.
The 2015 smolt cohort was ready for release back into their tributaries of origin (LSW and NW)
in the autumn 2017, however, the release was delayed due to approvals for authorization. Some
mature fish from 2015 cohort were used in the laboratory experiments (N=20) and radiotracking
project (N=40; additional fish were used as the group originally radiotagged became overripe
while waiting for authorization approvals, and re-tagging of a new group of fish had to take place
twice). The rest of the mature fish, as indicated in Table 1, were handled to remove eggs and
milt, and are being held at MSCC and are being reconditioned for a release in 2018.
Due to the developing program, construction and systems upgrades at the Miramichi Salmon
Conservation Centre, all mature Atlantic salmon consist of MCW component at this time; the
2016 smolt cohort is not expected to mature as grilse due to developing program. A grilse
release component is expected to be produced from the 2017 smolt cohort.
4. Genetic Monitoring Tools of the Program; General

Approach
The main objective of the CAST SAS program is to examine the generational contribution of
SAS adults to wild adults. Simultaneously, the currently used broodstock collections – juvenile
stocking program can be assessed using same methods. This assessment is best achieved
using parentage analysis approaches based on genetic markers wherein sampled progeny can
be assigned to specific parents and therefore, parent type (CSAS 2016). Also, monitoring of
genetic markers allows for detection of alterations in allele frequencies over time in both the
SAS and control river populations.
Genetic parentage analyses are based on the principles of Mendelian inheritance where
individuals of a sexually reproducing diploid species receive one gene copy (allele) from each of
their parents. By screening offspring and their potential parents for variation in molecular genetic
loci, inference can be made regarding the parents of a given individual. Various analytical
approaches (e.g. exclusion, categorical allocation, sibling-reconstruction) and statistical
frameworks (e.g. Bayesian, maximum likelihood) exist for conducting genetic parentage
analyses (for a review see: Jones et al. 2010). Each of these approaches are suited to different
types of parentage questions, the genetic markers used, available information about the
breeding systems, sampling of the individuals, as well as computing power and time (Jones et
al. 2010). Parentage-based tagging (Anderson and Garza 2006), unlike genetic stock
31
identification, does not rely on genetic differences among populations to assign individuals.
Instead, it relies on being able to unambiguously assign parentage to adults of known identity.
The use of parentage analysis for monitoring the proposed SAS strategy in the Northwest
Miramichi River has been reviewed and simulation work suggests that a parentage-based
tagging approach would be effective (Pavey 2016). Pavey (2016) recommended the use of
single nucleotide polymorphism (SNP) genetic markers, provided that a panel of sufficiently
variable markers is available, due to their flexibility and scalability. Preliminary analysis of ~50K
SNP markers characterized in juvenile Atlantic salmon from 16 sub-basin populations in the
Miramichi River indicate that there are more than 14K SNPs with a minor allele frequency
greater than 0.3 (the level of variation used by Pavey (2016) in simulations) suggesting there
are no limitations to using SNPs for parentage in this system. SNP data come from the 2016
surveys of juveniles (see Section 5.1.). Our intention is to select a panel of approximately 500 of
these highly variable SNPs to design a low density SNP array that will provide discriminatory
power to assign parentage unambiguously to SAS fish. A 500 SNPs with the shown level of
variation will theoretically provide 100% power to assign individuals (Pavey 2016) as we intend
to use more SNPs with greater variation than Pavey’s (2016) simulation study that had 100%
power. The chances of not assigning an individual if the true parent is in the dataset (e.g., not
assigning a SAS offspring; false negative) are negligible to zero (Pavey 2016).
Existing parentage-based tagging programs for monitoring supplemented populations (e.g.,
Steele et al. 2013; Beacham et al. 2017) have demonstrated good success in parentage
assignment using panels of 100-300 SNP markers (many fewer than we have proposed). For
example, Beacham et al. (2017) assigned 92% of 1599 known origin Coho salmon from 15
hatchery populations to the appropriate year-class within broodstock with SNPPIT (Anderson
2010) with 100% accuracy. The lowest proportion assigned for a hatchery population was
72.4%; however, only ~90% of the broodstock for this population was genotyped and SNPPIT is
not capable of making single parent assignments. This highlights the importance of genotyping
every SAS fish to obtain the highest probability of correctly assigning SAS offspring.
Furthermore, when software that can make single parent assignments (COLONY; Jones and
Wang 2010) was used, the overall success rate of assigning known-origin offspring to the
correct year-class within broodstock was 99.9% (1597/1599). Similar results were observed by
Steele et al. (2013) with assignment rates of known origin steelhead generally >95% with no
false positives. It is important to note both of these studies included large numbers of potential
parents in their assignment procedures that could not biologically have be the parents of the
tested offspring. Despite their inclusion, no false positive assignments were made. This is
consistent with modeling simulations that suggest low to no false assignments with the
proposed number of SNP markers (e.g., Pavey 2016).
To monitor fish with SAS parental origin, all SAS fish will be tissue sampled while they are in
captivity, DNA will be extracted from the tissue samples, and their genotypes characterized
using the low-density SNP array (similar sampling will be done to all adult salmon used for the
ongoing broodstock program). These samples provide reference parents for comparing the
genotypes of juvenile collections and eventual adult returns to the river. Beginning in the year
following release of the first SAS adults, and continuing for six years after the last releases of
32
SAS adults, life stages that could be biological descendants of SAS fish will be collected and
non-lethal tissue samples taken using large scale electrofishing and smolt collection surveys
(See section 5.4 for sample sizes). DNA will be extracted from these samples using non-lethal
fin clips (Dietrick and Cunjak 2006) and all individuals will be assayed for their genotypes with
the low-density SNP array.
A hierarchical approach to assigning parentage, such as that presented by Beacham et al.
(2017), that would first use SNPPIT to rapidly identify parent pairs for individuals whose parents
are both in the database and then a computationally more intensive approach (COLONY) to
match unassigned individuals to single SAS parents. This type of approach will facilitate
identification of SAS x SAS offspring as well as SAS x wild offspring in a given set of samples
while potentially reducing processing time. Other new software that uses a similarly efficient
hierarchical approach to assignment (Huisman 2017) may be used to verify parentage and
provide confidence in assignments as recommended by Pavey (2016). The intention of this
parentage analysis is to identify the number of SAS offspring relative to wild offspring and track
the ratio of SAS:wild through each life stage to assess the relative survival of SAS offspring until
they return as adults. We expect the sampling at the smolt life stage to be the least biased
toward sampling of related or otherwise non-randomly mixed individuals and may provide the
most accurate assessment of the survival proportion of SAS offspring during the juvenile
freshwater phase.
In addition to the parentage-based tracking of performance this genotype data will provide, it will
allow us to assess any allele frequency changes from the basin-wide baseline samples to
assess any potential impacts of releasing SAS fish (e.g., domestication effects). While we
expect these effects to be negligible, any detected changes can be followed up by genotyping
samples with a higher density SNP array (50K) to assess the magnitude and risks posed by
these differences. Genetic samples will be collected also in the control rivers (Figure 1),
however, the control river samples will be submitted to genetic analysis only if differences in
allele frequencies are observed in SAS intervention rivers between the SAS progeny and wild
progeny. As the allele frequency changes in the intervention rivers may be due to temporal
variability, the control river samples can then be analysed (using the 50K SNP array) to assess
amount of temporal variability in control baseline in the absence of SAS intervention, and the
amount of temporal variability between control and SAS intervention rivers can be compared.
Parentage-based methods will also be used in the Experimental Stream (Northwest Millstream,
NWMS). In NWMS, all wild adult salmon are also genetically sampled prior to release to the
experimental arena. The parentage-based analysis in the experimental stream is therefore
predicted to result in high degree of analytical power to answer the knowledge gaps regarding
genetic and phenotypic differences between the SAS and wild fish up to the smolt migration
endpoint. However, NWMS experiment is not large enough to be able to assess SAS strategy in
its main objective of producing returning adults from the ocean (See Section 5.3).
33
5. Detailed description of proposed studies, hypothesis

framework, and knowledge gaps to be addressed
The following sections describe the four levels of research discussing the science premise,
hypothesis to be tested, general methods (including sample sizes and duration, as applicable),
knowledge gaps to be answered (as defined in CSAS 2016), and metrics to be measured (as
per Fraser 2016). Triggers and decision rules regarding when the experiment will be altered,
continued or halted is discussed with respect to the studies in natural rivers, where the risk to
wild Atlantic salmon population must be considered. There is currently no perceived risk
mechanism to affect wild Atlantic salmon in the sub-basin genetics study, laboratory
experiments or the Experimental Stream study and therefore, triggers and decision rules to halt
these studies are not considered.
5.1. Program 1: Sub-basin genetic structure of Atlantic salmon

on the Miramichi River
Science Premise:
Foremost in any recovery strategy assessment should be an understanding of the genetic
population structure for the species of concern (Jonsson et al. 1999). Knowledge regarding the
genetic structure, i.e., the uniqueness of each sub-basin in the Miramichi, is necessary as
baseline information to ensure SAS production lines, i.e., the smolt cohort management, are
appropriately created so as to protect the natural populations from potential genetic risks of the
SAS experiments (and as would be required when using SAS as an intervention strategy).
Additionally, understanding the genetic structure prior to initiating SAS experiment will allow the
identification of any potential genetic changes both during and after the cessation of the SAS
experiment.
Hypothesis: Genetic structuring of the Atlantic salmon population occurs and is defined by the
sub-basins in the Miramichi River system
Conducted studies and ongoing activities:
• In 2016, 16 sub-basins (Figure 2) were sampled by electrofishing in collaboration with DFO
and MSA (target sample size of N=50 age 1+, Atlantic salmon parr in each sub-basin); A
total N=774 was collected (Table 2)
• Samples have been genotyped using a 50k SNP chip in winter 2017 (Wellband et al.,
unpublished data); genotyping was achieved for 743 individuals
• Based on preliminary analysis, traditional population genetic statistics identified statistically
significant genetic divergence among most pairwise comparisons of sub-basins. However,
the overall magnitudes of pairwise divergence among rivers were very low given what was
expected to a river system this size (FST = 0.0006 – 0.0147) (K. Wellband et al., Laval/UNB,
unpublished data). Naïve genetic clustering routines (i.e. those with no prior knowledge of
sample grouping) failed to discriminate meaningful clusters of individuals within the
Miramichi River; however, this result is not unexpected due to the reduced power of these
techniques to detect population structure when the magnitudes of divergence are weak (FST
< 0.05).
34
• Development of a lower density (LD; approx. 500) SNP chip for conducting parentage
analysis (see Section 4) in subsequent SAS studies is ongoing (UNB/Laval with Cooke
Aquaculture)
• Sub-basin genetic study provides necessary baseline information to assess SAS impacts on
natural rivers (Section 5.4)
• A second, complete assessment of genetic structuring in five years will be used to evaluate
temporal genotypic stability, which also provides an additional check on genetic effects of
potential straying and reproduction by SAS–origin fish.
Knowledge gap (CSAS 2016) to be addressed:
o Genetic structure among the sub-basins of the Miramichi River system is currently
unknown. This study will assess genetic structure and verify the requirement of
sustaining unique rearing lines for SAS fish production from different sub-basins
separate (ongoing working hypothesis).
o Provide a genetic baseline to assess potential genetic changes as a result of SAS
experiments in natural rivers.
Table 2. The number of genetic tissue samples collected in 16 sub-basins in the Miramichi River in the
autumn 2016 (See also Figure 2)
SubBasin N Notes
Upper NW Miramichi 52
Lower NW Miramichi 63 13 samples from a tributary (Sutherland
Brook) flowing into tidal parts of
Northwest
Sevogle 49 -
Upper LSW Miramichi 40 -
Lower LSW Miramichi 49 -
Lower SW Miramichi 54 -
Middle SW Miramichi 49 -
Upper SW Miramichi 52 -
Renous 52 -
Dungarvon 52 -
Cains 59 -
Taxis 49 -
Burnthill 27 -
Clearwater 23 -
Rocky Bk 51 -
Northwest Millstream 53 -
Total 774 -
35
National Capital Region Review of the risks and benefits of CAST SAS Experiment Proposal
Figure 2. Sub-basins (N=16) of the Miramichi River where sampling for genetic baseline material was undertaken in autumn of 2016 with
respective sample sizes at each location (Total = 774 fish samples).
36
5.2. Program 2: Laboratory experiments

Science Premise:
The laboratory experiments provide a fully controlled environment where fundamental data on
SAS fish characteristics, both phenotypic and genotypic, (e.g., maturation rate, size at
maturation, maturation timing) can be assessed for progeny and adults, including the relative
differences between SAS and wild fish. Fundamental information is collected on fecundity, egg
size, and fitness which is defined here as survival to eyed stage, survival to hatch (yolk-sac fry),
and survival to swim-up (start of external feeding). In addition, findings are expected to lead to
modifications and development of the rearing practices to continually improve the phenotypic
quality of SAS salmon. Importantly, these are controlled situations where comparisons between
groups, i.e., SAS versus wild, is a relative assessment. These results help us to understand the
mechanisms of differences, if they exist, between groups that will (1) help define the model
parameters for full production SAS programs in the future, i.e., the cumulative progeny
production for a SAS adult and its progeny’s survival to its spawning event (1 SAS adult = X
future egg deposition), and (2) will help us to model the predicted SAS contribution to the total
population (numbers), i.e., spawners, juveniles, and smolts in a subbasin.
Hypothesis: A SAS fish will differ in basic biological (phenotypic) traits from wild at the adult
stage. Survival to described end-points is expected to be lower in SAS than wild, however, no
genotypic or phenotypic differences are predicted a priori for the surviving progeny.
Proposed studies; general methods:
[Adults]
• Maturation rate of the adult SAS fish (by sex) is measured for five years of the experiment
• Size-at-maturation by sex is measured
• Maturation timing by sex is assessed
• Fecundity and egg sizes are compared annually; N = 20 (targeted) adult females (each of
wild and SAS; only 10 wild females and males were available in 2017; 10 + 10 SAS salmon
were similarly used in 2017)
• Fertilization success by males; N=10 males of both wild and SAS (N=8 used in 2017 due to
wild availability)
o Repeated for 2 years;
o Fertilization of sub-samples of eggs (N=100) from selected wild females (N=3) by both
wild (N=10; N=8 in 2017) and SAS males (N=10; N=8 in 2017); eggs from each female
randomly selected to two batches of equal size (N=100) and fertilized by milt from SAS
and wild male
o Total 1000 eggs / female / treatment
• Survival studies
o 10 females and 10 males of each wild and SAS to generate wild x wild, SAS x SAS, wild
x SAS and SAS x wild crosses for a common garden experiment resulting in 40 crosses;
3 replicates / each cross
37
o (for 2017; crosses were fertilized 20- 27 October)

o Egg to eyed stage survival
o Egg to hatch survival
o Egg to swim-up survival
o Swim-up to first feeding fry survival and growth
o Surviving SASxSAS, SASxwild, wildxSAS crossed first-feeding fry to be distributed to the
lower sections of Northwest Millstream below the barrier fence (13 km section) unless
otherwise determined; all genotypes are known
o Surviving Wildxwild crosses distributed to tributaries of origin as part of MSA’s normal
supplementation program (genetic identity of parents known, and success of the
surviving fish will be determined as a part of the SAS monitoring program)
• Potentially, examine gene expression and epigenetic differences between the experimental
groups at the swim-up fry stage (contingent on additional, future funding).
• Characteristics of SAS adults relative to wild
• Characteristics of SAS eggs relative to wild
• Survival (fitness) of SAS progeny relative to wild in controlled hatchery environment up to
first-feeding fry stage (Note that spawning of adults and thus mate selection, is not free
unlike in SAS strategy in wild. Similarly, hatching will be in hatchery environment where
selection may occur. Therefore, the experiment is not a true test of SAS strategy as it would
occur in wild)
Metrics (Fraser 2016; Table 2)
• [Adult]: Growth and body size, body shape (a separate undergraduate thesis project
undertaken in autumn 2017 at UNB), maturation rate, reproductive timing, egg size and
fecundity
• [Juvenile]: Survival to described end-points, and size at the final endpoint relative to wild
Deviations in mean and variance in above metrics between SAS and wild fish will be assessed
and statistically compared (Fraser 2016).
Adaptive Planning - Risk assessment: Lowered survival of wild (female) x SAS (male) crosses
to described endpoints could indicate a risk to wild Atlantic salmon if similar result was
applicable in natural environment (i.e. wild eggs would have lowered survival if fertilized by SAS
male). Lowered survival of SAS x SAS crosses and SAS (female) x wild (male) crosses is not a
critical risk, as wild egg production is not affected by lowered survival of SAS progeny (wild
sperm is not a limiting factor; SAS intervention occurs in areas well under the carrying capacity
of juvenile production; see DFO (2017a)). Caveat in the studies in laboratory is that unlike actual
SAS program in the wild, the mate selection is not natural, and differences in laboratory may not
be reflective of SAS program in wild.
38
5.3. Program 3: Experimental River

Science Premise: To better understand how SAS salmon will respond in the natural
environment relative to wild Atlantic salmon, a controlled experiment is designed to occur in a
small, natural stream. The proposed stream, Northwest Millstream (NWMS), flows directly to
Miramichi estuary (Figure 1) and has a history of stocking of various Atlantic salmon strains into
the system. The NWMS is annually blocked by numerous beaver dams and salmon migration to
upstream section is naturally impeded. The lack of natural connectivity in stream and repeated,
mixed stock introductions are good guarantees that potential effects of SAS introductions on
wild stocks are minimized in the NWMS. A barrier (conduit) fence will be erected in the
upstream section of the NWMS in the autumn of each year and will be maintained through the
spawning period for each year of the experiment (proposed 3 years of SAS introductions). The
section upstream of the barrier will act as the experimental section for SAS vs. wild
comparisons. The experiment will examine 1) SAS and wild adult behaviour, activity levels, and
survival and 2) SAS and wild progeny survival, growth, and behaviour. The wild fish required for
the experiment are proposed to be collected from DFO index trap, supplemented by seining and
angling to collect additional wild fish as required. First release of 20 SAS females and 20 SAS
males took place in late October 2017; no wild fish could be captured in 2017 due to lateness of
getting authorization to collect wild fish for the purpose (permit received October 26).
Hypotheses:
The experiments will test deviations of the SAS phenotype and genotype from the wild in a
controlled natural setting.
At adult stage, the following hypothesis is examined:
1. SAS adult behaviour is similar to the behaviour of wild fish.

Predictions are that:
• SAS fish will spawn in similar areas to the wild;
• SAS adult spawning timing is similar to the wild adults;
• SAS adults find mates and spawn successfully;
• SAS post-spawner survival is similar to wild adults
At juvenile life stage, the following hypothesis is set forward:
2. SAS adults produce progeny that are phenotypically and genotypically no different from
progeny produced by wild fish.
• Eggs from SAS adults emerge successfully in ratio compared to wild as would be
predicted from relative survival studies from the lab studies (i.e., fecundity and the egg
viability may be less (or more) than that of wild).
• SAS juveniles survive successfully in the wild at similar rate as the wild juveniles
• SAS juvenile growth in the wild is similar to the wild juveniles
• SAS juvenile behaviour in the wild is similar to the wild juveniles
39
• Allele frequencies in the examined loci are no different between SAS and wild juveniles
Proposed studies: Monitoring of movements and spawning behaviour of SAS and wild adult fish
will use radiotelemetry; monitoring will be repeated for 3 years. All salmon, SAS and wild, are
genetically sampled (for progeny parentage assignments). The juveniles will be assessed using
electrofishing surveys; the annual electrofishing surveys to collect fin-clips (genetic material) of
progeny for parentage analysis will be carried out for five years. A sub-group of the captured
juveniles will be tracked with Passive Integrated Transponder (PIT) tags to monitor the
behaviour, survival, and growth of known individual juveniles from the two different groups.
Planned Studies in NWMS (others may be added in our adaptive approach based on new
knowledge gained)
• Release of 20 pairs (20 females and 20 males) of both SAS and wild Atlantic salmon to
NWMS upstream of the barrier fence
o All released fish genotyped, radiotagged (Lotek MCFT2 tags) and PIT tagged
o All released fish are externally tagged (T-anchor tags); different colour code for SAS and
wild adults to facilitate behavioural observations using direct streamside observations
o Repeated for 3 years
• Project 1 - Monitoring of movement and behaviour of adults
o Assessing spawner distribution and survival, spawning behaviour, redd locations
o Active monitoring (radiotags) in streamside surveys
o Two passive (radio)monitoring stations installed to monitor movements during spawning;
a number of passive PIT tracking stations will also be used to monitor movements
o One passive monitoring station retained to monitor post-spawner emigration after barrier
fence is removed in late autumn.
o Repeated for 3 years
o Monitoring of survival of the post-spawners to potential repeat-spawning in both SAS
and wild fish (radiotags with multi-year battery life)
• Project 2 - Distribution and survival of SAS progeny
o Annual electrofishing surveys to collect fin-clips (genetic material) of progeny for
parentage (SNP) analysis; annual monitoring for 5 years (assuming majority of smolts
migrate at age 2 or 3).
o Parentage resolved using LD SNP chip developed based on the sub-basin study (see
Section 4 and Section 5.1)
o Inter-stage survival and growth monitored based on repeated electrofishing surveys, and
genetic analysis
o 10 annual electrofishing sites, N= 50 / age group / site (YOY and “parr”); Total annual
genetic sample size N = 500 for 5 years
40
• Project 3 - Movements and behavior of juveniles

o Monitor behaviour and movements of PIT-tagged individuals
o Use of genetics to resolve parentage (post-tagging; the parentage unknown at the time
of tagging, but representatives of both wild and SAS groups are likely to be tagged by
randomizing the tagging efforts and 1:1 ratio of parents released into the system).
o Install passive PIT antenna arrays in strategic locations to monitor movements and
emigration of PIT tagged juveniles throughout their residence in NWMS [will also monitor
adult movements in autumn]
o Inter-stage survival and growth of known individuals of known parentage
o PIT methods to be used are half-duplex Oregon RFID systems with 12mm and 23 mm
“silver bullet” tags with improved detection distance. The reader systems will be novel
satellite-synchronized multiplex systems to maximize detection distance (currently not
available commercially, but will be available for CAST SAS studies)
• Project 4 - Behavioural studies in small stream enclosures
o Direct observation of PIT-tagged individuals to assess aggression / boldness
o Fish will be PIT tagged, and genetically analyzed as 0+ to determine parentage; in a
subsequent year (1+ or 2+ parr), recaptures will be obtained and balanced design of
SASxSAS, wild x wild and crosses will be selected to stream enclosure studies to study
differences in behavioural metrics.
Size of the experiment
The Experimental Stream component of the SAS program is designed to allow high level of
control in a necessarily, small-scale natural environment. This allows the experimenter to
control the ratio of SAS to wild fish in a semi-closed “arena”. Having higher level of control
means, however, that the experimental section is relatively small stream where aspects relating
to generational output of an actual SAS program becomes infeasible, i.e., too few SAS adult
returns can be generated to have statistical power to examine generational contribution of SAS
adults relative to wild adults in producing returning adults.
Knowledge gaps (CSAS 2016) to be addressed:
• Behaviour (migratory rigor, pairing ability, spawning behaviour, post-spawning behaviour)
and interactions of the released SAS adults relative to wild counterparts in the natural
environment
• Competition for mates and disruption of wild spawning
• Spawn timing relative to wild
• Fitness (individual and group) of SAS adults relative to wild counterparts during juvenile
stages
o Relative individual fitness of both SAS and wild Atlantic salmon (relative production of
juveniles by known SAS and wild parents; all parents genetically sampled) during
juvenile to smolt stages
o Relative group fitness of SAS adults relative to wild (ratio of SAS:Wild progeny through
years within cohort) during juvenile to smolt stages
41
• Behaviour and growth and morphology of SAS juveniles relative to wild juveniles
• Adults - Body shape, Reproductive timing, Migratory rigor, Activity levels
• Juveniles - Inter-stage survival (fitness), growth, body shape, Activity levels, movement
behaviour, aggression/boldness
5.4. Program 4: SAS impacts on a natural river

Science Premise: A successful SAS program, i.e., its value as a potential conservation strategy,
requires an understanding of the generational contribution of SAS adults relative to wild
adults, i.e., can the SAS parents produce progeny that migrate to ocean and return back to
rivers to spawn successfully and contribute to the production of successful (fit) juveniles in the
river. This objective can only be achieved at a population scale and in a system (river) where
SAS adult introductions are of sufficient numbers to generate progeny that: 1) can be
quantitatively assessed in river (counted in each age class) and 2) are sufficient to complete the
cycle of becoming adults returning to the river to spawn and be quantitatively assessed. The
SAS origin returning adults have to be of sufficient numbers to invoke a change in the wild
juvenile production that can be quantitatively assessed. Importantly, the study must be
replicated, i.e., two or more systems are required.
Hypotheses:
1) The experiments will test deviations of the SAS phenotype and genotype from the wild in a
natural setting.
At adult stage, the following hypothesis is examined:
1. SAS adult behaviour is similar to the behaviour of wild fish.
• SAS fish will spawn in similar areas to the wild;
• SAS adult spawning timing is similar to the wild adults;
• SAS adults find mates and spawn successfully;
• SAS post-spawner survival is similar to wild adults
For the SAS progeny, the following hypothesis is set forward:

2. SAS adults produce progeny that are phenotypically and genotypically no different from
progeny produced by wild fish.
• Eggs from SAS adults emerge successfully in ratio compared to wild as would be
predicted from relative survival studies from the lab studies (i.e., fecundity and
the egg viability may be less (or more) than that of wild).
42
• SAS juveniles survive successfully in the wild to smolt stage at similar rate as the
wild juveniles
• SAS juveniles growth in the wild is similar to the wild juveniles
• Allele frequencies in the examined loci are no different between SAS and wild
juveniles
• Survival of SAS smolts to adulthood and return back to the river of origin is
similar to the wild fish, i.e. there is no SAS-induced loss of marine adaptation
2) The SAS experiments will result in a demographic increase in juvenile density in the SAS
intervention rivers relative to the background level in the multi-year electrofishing dataset,
whereas similar increase in comparison to background level in control rivers will not be
observed.
Proposed studies: The current plan is to release mature SAS adults to two proposed
intervention rivers (See Section 2.3). A first release is planned in autumn 2018. In general,
releases are 15 months post-capture for majority of males (i.e., grilse), and 27 months for the
majority of females (i.e., Multi-Captivity Winter, MCW). Maturity can be detected late in the
season, thus only a late season release is initially planned. Early diagnostics of maturity using
ultrasound will be explored to enable mid-summer releases. It is anticipated that not all SAS will
mature in these timeframes and the non-mature fish will be held for an additional year at the
MSCC. The females from the smolt cohort collected in 2015 (originally destined for release in
autumn 2017) is anticipated to be fully mature in 2018 and will be released in mid-summer 2018
coinciding with the wild run; the fate of the males from the 2015 cohort is to be discussed in the
CSAS Expert review in January 2018 as the space limitations may not allow keeping them at
MSCC after spring 2018 (potential release during kelt migration in April 2018).
The current proposal is that SAS fish will be released upstream of the smolt collection areas
(Figure 1 - Sillikers in the LSW and Wayerton Bridge at NWM). The release plan may evolve as
new information is learned regarding their behaviour and sub-population genetics during this
experiment. The period of release is 15 September to 15 October corresponding to a natural
migration peak in the Northwest Miramichi (Chaput et al. 2016); a mid-summer releases will be
used if maturity can be predicted reliably using ultrasound methods. Releases will initially occur
daily in lots of ~75-100 / river depending on water conditions, i.e., water temperature and levels
appropriate for adult salmon.
Fifty (50) SAS adults and 50 wild adults will be equipped with radiotags to monitor movements
and behavior, annually for 3 years in each tributary.
The number of released SAS salmon will be explicitly known each year. All SAS fish will be
marked with a PIT-tag, and the rivers where the fish are released will be monitored at selected
locations using multiple large flatbed PIT antennas that will detect if any of the SAS fish leave
the river systems (double antennas will be installed to stream-bed to ensure detection of
movement directionality; Figure 3). The number and sizes of wild Atlantic salmon entering the
SAS intervention rivers will be monitored using the Adaptive Resolution Imaging Sonars (Figure
3) in a parallel CAST project (full study described elsewhere). This allows calculation of a ratio
of wild:SAS fish in the system, and this ratio can be used to further monitor the success of their
juvenile production (see Metrics section below).
43
The juvenile production will be monitored in a large-scale electrofishing program starting 2019,
and continuing throughout the CAST SAS Phase 1 for at least 5 years (it continues in Phase 2
of the monitoring program). The parentage of the juveniles will be resolved using genetic SNP
markers (see Section 4); the premise is that we will know the DNA of all our SAS fish, and
therefore, will be able to identify any juveniles produced by SAS parents (both SASxSAS and
SASxwild families). The objective of the juvenile monitoring is to document if the ratio of the
wild:SAS fish is the same as the ratio of wild:SAS during the adult phase, and whether the ratio
remains similar as the fish age (indicating similar intercohort survival between wild and SAS
progeny). We will additionally sample the broodstock adults used in the ongoing juvenile
stocking program to additionally compare the success of this program relative to wild and SAS
juvenile and smolt production.
Figure 3. Schematic representation of methods to establish the number of adult salmon (wild or SAS)
upstream of original SAS smolt collection locations. Wild salmon will be enumerated by the ARIS sonar
units; SAS salmon will be detected by both ARIS and associated PIT arrays. In addition, a sub-population
of both SAS and wild salmon will be additionally detected by the radiotelemetry, allowing for a method to
test the efficiency of both ARIS and PIT methods. (The schematic is for illustrative purposes only, and
does not accurately represent the proposed study sites).
44
Planned Studies (others may be added in our adaptive approach based on new knowledge
gained):
• All SAS adults are sampled using genetic tools (LD SNP tool) and are PIT tagged (23mm
Half-duplex “silver bullet”). All SAS fish are also externally marked using T-anchor tags so
that the local First Nations communities can positively identify SAS fish, if captured (see
Section 6).
• Project 1 - Monitoring movements and behavior of post-release adults
o 50 of each of SAS and wild adults radiotagged / tributary for 3 years
o Wild salmon sourced by seining / trapping and angling in the intervention tributaries;
additional wild fish may be available from a potential concurrent DFO radiotracking study
o Active tracking and passive monitoring stations at key locations (N = 4 locations in
LSWM; areas locally known as Upper Oxbow; Catamaran Brook; Junction of Lower
North Branch LSWM; and junction of North Pole Brook: N= 3 locations in NW; Wayerton
Bridge area; junction of Little River; third location to be determined)
o Lotek radiotelemetry systems; MCFT2 tags with multi-year batteries (approx. 2 years)
o Fish will be tracked continuously through the spawning season, occasionally in winter as
ice conditions allow (targeting 2 events), and weekly during the kelt outmigration period
to establish the behaviour of SAS adults relative to wild fish. Tracking will be undertaken
using a variety of methods based on area (vehicle; 4-wheeler; foot; aircraft). Data
collected includes:
o Spawning – Establishes that (1) SAS fish spawn and (2) serves as a guide where
potential SAS progeny may be encountered during progeny monitoring efforts.
o Winter behaviour, survival, and movement during kelt migration – Establishes if
SAS fish may become multi-year spawners.
o At-sea survival (SAS post-spawned fish) and among tributary straying
o Repeat-spawning – Information on SAS fish rates of return relative to wild fish
will be collected as the proportion of radiotagged fish returning (multi-year tags)
• Project 2 - Monitoring of post-release using PIT telemetry
o Flat-bed PIT arrays to monitor of potential emigration of SAS adults; SAS adults are
introduced to areas upstream of PIT arrays.
o Half-duplex multiplex systems with satellite synchronization and marker tags to test for
functionality
o Required to establishing how many SAS fish remain in each tributary through the
spawning period, i.e., is required to establish a ratio between adult SAS:wild salmon in
each tributary during spawning time
• Project 3 – Counting returning wild adults
o ARIS 1800 units are used to establish the number of wild salmon, and to assess the size
distribution of returning adults (ARIS study is described elsewhere in full; units have
been operated in 2016 and 2017, and daily fish counts by size category can be observed
at www.castsalmon.com)
45
o PIT arrays are associated with ARIS units to solve whether migrants are wild or SAS (all
SAS fish PIT tagged and detection of fish by PIT array while in sonar indicates a SAS
fish; synchronization of PIT-arrays to be used are new satellite-based systems; antenna
installation methods have been tried in another project (the Upper Salmon River SAS
program) in 2016 and 2017, and have proven reliable)
• Project 4 - Assessing contribution of SAS progeny
o Annual electrofishing surveys of 20 sites / tributary with 25 young-of-the-year and 25
parr collected / site for parentage analysis (i.e., 1000 juveniles / tributary / year for
minimum of 5 years).
o All genetic samples will be small, non-lethal fin-clips, from which juveniles are expected
to rapidly recover (Dietrick and Cunjak 2006). If possible, other concurrent electrofishing
programs may secure samples (e.g., DFO/MSA e-fish programs)
o Genetic material collected from control rivers through the DFO and MSA electrofishing
programs
o Samples serve as baseline and analysed if genetic deviations between wild and
SAS progeny in intervention tributaries are observed
o Resolve parentage using SNP genetic tools (see Section 4)
o Average inter-stage survival and growth monitoring by origin (SAS, wild or crosses,
including broodstock stocking)
o Assessment of changes in allele frequencies
o Annual comparison of samples of wild juvenile Atlantic salmon (as determined by
the parentage analysis, i.e. fish determined to be of non-SAS or non-broodstock
origin) in the SAS intervention areas (Lower and Upper sub-basins in the Little
Southwest Miramichi, and the Upper Northwest Miramichi) to the wild baseline
(2016) genetic information from those areas and assessed for deviations. Are the
allele frequencies changing in wild juveniles before and after SAS intervention?
o Genetic samples from SAS progeny will also be compared to the wild baseline
(pre-intervention, or 2016) data, as well as to the data from wild progeny
collected during intervention. Are the allele frequencies different between SAS
and wild juveniles either before or during SAS intervention?
o If differences are detected between SAS fish and the wild baseline, the tissue
samples collected from the control rivers (Sevogle River, and selected areas in
the Southwest Miramichi River) will be analysed using the SNP tools. These data
will be compared to the baseline data at the same control sites prior to the start
of the SAS intervention. Are the allele frequencies changing in wild juveniles in
control rivers before and after SAS intervention?
o If differences exist in the control areas and the intervention areas, then the data
are indicative of natural temporal variability in allele frequencies within each
tributary irrespective of the SAS intervention. If differences in allele frequencies
pre- and post-intervention are exhibited only in the SAS intervention rivers but
not in control rivers, then halting of further SAS releases into natural rivers is to
be considered pending further analysis. Results will be discussed by the Science
Team (CAST/DFO/MTI) to determine if genetic evidence is sufficient to trigger
the adaptive planning actions, e.g., modification of studies (see below).
46
• Project 5 - Assessing SAS contributions to smolt migrations

o Smolts will be captured annually in the rotary screw traps (MSA operated)
o Fin clip samples (1000 / tributary) collected for genetic analysis to solve parentage (SAS
and wild population assumed well mixed during the smolt migration; sampling throughout
the run)
• Project 6 - Monitoring of returning adults (progeny of potential SAS adults)
o 250 non-SAS adults (no PIT-tag) handled in each of the DFO Cassilis (NWM; branch
with SAS intervention) and Millerton (Southwest Miramichi; control branch; Figure 1) trap
nets will be tissue sampled (caudal fin clip) and analyzed to assess: 1) temporal, genetic
characteristics of returning, wild salmon; and 2) proportion of SAS-derived adults
(estimated to begin being detected at 5 years after first release).
o The proportion sampled between Millerton and Cassilis index traps may change over
time if the results indicate that SAS fish are almost exclusively returning to NW Branch
(as hypothesized, although natural straying between the two branches occur), and thus,
we favour sampling to detect “effect” (ratio of SAS originated fish returning relative to the
wild) rather than document “risk” (monitoring the potential occurrence, i.e. straying, of
SAS originated fish into non-natal branch). However, preliminarily, the sampling program
will use a balanced design.
o Additional adults will be sampled at conservation barriers (e.g. headwaters at NWM) and
fishing camps along the LSW and NWM. N = 100 per tributary is targeted.
Scale of the experiment
The main objectives of the in-situ experiment is to 1) address the largest outstanding scientific
knowledge gap regarding the SAS strategy, i.e., compare the lifetime success of offspring of
SAS adults vs. wild adults in the natural environment; and 2) examine whether the SAS program
will be able to increase juvenile production in the intervention rivers relative to controls. To be
successful, the experiment must be large enough that we achieve appropriate sample sizes of
returning adults whose parents were SAS fish, and that produced quantities of juveniles are not
masked by natural interannual variation in juvenile densities.
The main method of assessing this outcome is to compare the ratio of SAS:wild (and their
crosses) in the randomly sampled, out-migrating, smolt cohorts in each of the two SAS
intervention rivers, to the ratio of SAS:wild (and their crosses) of returning adults in one-to-three
years after each smolt cohort has migrated to ocean. The majority of smolts migrate at
freshwater age 2 or 3, while <3% migrate at age 4 (Chaput et al. 2016). Complete assessment
of the success of each year class requires assessing the smolt production over the three
outmigration seasons and the subsequent returns to MSW stage.
Accurate calculations of required SAS “seeding” quantities can only be estimated based on
available information. This is because the success rate of SAS fish to produce progeny is not
known at the current time, and the freshwater survival of juveniles or the ocean survival of
adults (either wild or SAS) are not explicitly known. It is estimated that 2500 smolts from each
SAS tributary can generate a production of 2,000 SAS adults (assuming 80 % survival to
release in captive environment), 50% of which are predicted to be female (but will be explicitly
know at the time of release).
47
A release of 1000 SAS females / replicate tributary is considered an absolute minimum to

ensure a good likelihood for securing adequate sample size of returning adults whose parents
were of SAS origin. While only a very coarse estimate, it is predicted that 1000 females will
produce an approximate egg deposition of 6.2 million (Assumptions: 80 % of the females
mature as MSW at average weight of 5 kg; 20 % mature as grilse at average weight of 2.5 kg,
and an average fecundity of 1385 eggs/ kg; Chaput et al. 2016; Cunjak & Therrien 1998). An
egg-to-smolt survival of 2.5 % (a mid-range estimate from available juvenile salmon studies;
Cunjak & Therrien 1998) would translate into approx. 156 000 smolts being produced.
Estimating an ocean survival of 1 % from smolt to grilse, and 0.2 % for smolt to MSW (lower
range estimates from Miramichi in the 1998 to 2006 time series; Chaput et al. 2016) and 80 %
return rate of smolt class as grilse (Chaput et al 2016), an estimated 1246 grilse and 62 MSW
salmon may be estimated to return (/ 1000 SAS females). Sampling genetic material to confirm
whether SAS:wild ratio is similar to the ratio during smolt migration requires a chance to capture
of a representative number of the returning SAS individuals. Assuming that 10 % of the
returning adult salmon could be sampled by DFO index trap, sample collections from outfitter
lodges, barrier traps, angling and seining salmon pools, a sample of 120 -130 SAS fish /
tributary could therefore be obtained / returning adult cohort, assuming original “seeding” of
1000 females. It is emphasized that the above calculation is only a coarse estimate, however,
provides justification for the size of the SAS experiment in natural streams, in absence of any
other meaningful way to predict various stages of survival or capture rates of salmon as they
return. It is also noted that the production sizes of females (i.e., 2.5 kg and 5 kg) as well as egg-
to-smolt survival of 2.5%, may be “liberate”, meaning that the de facto return may be less than
the above prediction, calling for potentially larger “seeding” of SAS adults. However, the
potential risk to wild Atlantic salmon increases as the SAS “seeding” numbers increase, and
therefore, until better information is available regarding the risks related to the experiment, the
proposal is to compromise with a collection of 2500 smolts / tributary.
Furthermore, in order to test for the potential of SAS intervention to increase juvenile production
in a system where the average inter-annual variation in average densities is approx.. 80% with a
mode of 35 % (DFO e-fish data), the modeled increase in adult recruitment needs to be at least
in the range of 2000 adults / tributary to be able to detect an “effect” (i.e. increase in juvenile
density above background “noise” level; R. A. Curry, unpubl.).
Currently, the CAST program is predicting to start with a release consisting of approximately
1850 and 1550 SAS salmon for the Northwest and the Little Southwest Miramichi, respectively
(Table 1; 2015 and 2016 cohorts with some minor mortality; assuming 50% females). Current
prediction for releases in years 2 to 5 (continuing to year 6 to release all the multi-captivity
winter fish) is to annually release the surviving mature adults (less the adults required for
laboratory experiments, Experimental stream studies and veterinary testing) from the 2500
smolt collections to NW and LSW (in years 2 to 5, or 2019-2022/23). For context, the
conservation requirement for the Northwest Miramichi system is estimated as 7300 large (MSW)
females and a corresponding amount of adult males (Randall 1985).
The return rates of wild adult Atlantic salmon to either Little Southwest Miramichi or Northwest
Miramichi above Trout Brook are not known (CSAS 2017a). In CAST’s salmon counting
48
program (ARIS program), there will be an assessment of adults returns in both the NWM and
LSW. For context, the average of the median adult returns to the Northwest Miramichi system
(i.e. includes both SAS intervention areas and the Sevogle River control, and other small
streams) is approximately 5 000 MSW and 12 000 grilse (CSAS 2017a). Based on CAST’s
ARIS project the return, the return in LSW in 2017 (at the end of October) was 4287 grilse and
1576 MSW salmon.
In a SAS, or any supplementation program, the potential risks may increase when SAS releases
represent an increasing proportion of the total number of wild adults in the population at
spawning time (CSAS 2016). The releases of CAST SAS program of up to approximately 1000
adult females / tributary (in 2019 and onwards) would comprise a small-to-modest quantity
compared to the amount of adult females required to meet the conservation status. While the
wild returns a number of years into the future cannot be predicted with any confidence, a
maximum number of SAS releases in relation to wild adult returns are proposed to be a
maximum of 1:1 until new information is available to assess this ratio and its potential impact on
the wild salmon. The actual ratio will be known at the time of the releases based on the CAST
ARIS program. Similarly, if the ARIS program indicated that the wild salmon return is so strong
that the conservation targets would be exceeded by additional SAS releases, further discussion
regarding the SAS release quantities would have to take place.
• Behaviour (migratory rigor, between tributary straying, pairing ability, spawning behaviour
and timing, post-spawning behaviour and survival) and interactions of the released SAS
adults relative to wild counterparts in the natural environment,
• At-sea survival information of SAS repeat-spawning adults relative to wild repeat-spawning
adults based on redetection of radiotagged and PIT tagged adults in years following
spawning.
• Relative fitness at various life-stages including first generation returns of anadromous adult
stages (i.e. generational contribution of SAS adults relative to wild adults).
o Relative group fitness of SAS adults relative to wild (ratio of SAS:Wild progeny through
years within cohort) during juvenile to smolt stages.
o Relative group fitness of first generation (F1) returning adults of SAS fish relative to wild
adults (ratio of SAS:Wild F1 adults in each SAS tributary relative to the ratio in the
tributary during smolt migration phase).
• Inter-stage survival and growth of juveniles of SAS and SAS-wild hybrid relative to wild
salmon in natural environment.
• Characteristics of smolt migration (timing, size, age at smoltification) of SAS and SAS-wild
hybrid relative to wild salmon.
• Characteristics of anadromous adult progeny of SAS and SAS-wild hybrid to assess cross-
generational differences in genotype or phenotype relative to wild salmon; i.e. assessment
of loss of marine adaptation.
• Adult - Morphology, Reproductive timing, Migratory rigor, Activity levels
49
• Juvenile - Inter-stage survival (fitness), growth, body shape, genetic differences

Adaptive Planning - Triggers and decision rules for altering, continuing or halting the SAS
studies
• Assessment of the SAS program in natural rivers, relative to wild fish, is based on following
the development of SAS:wild ratio. This ratio is established starting from spawner
escapement by 1) knowing how many SAS adults are released in each of the intervention
rivers, 2) monitoring potential emigration of SAS fish from the rivers based on PIT arrays,
and 3) establishing the amount and sizes of wild fish based on ARIS counts.
o If the ratio of SAS:wild fish declines in favour of wild fish when measured in the young-of-
the-year (YOY) life stage, this may indicate that SAS fish are not as effective in
propagating as wild fish. Such outcome does not pose a risk to the wild Atlantic salmon,
but indicates inefficiency of the SAS program (e.g. lower fecundity, lower egg survival).
o If the ratio of SAS:wild fish increases in favour of SAS fish when measured in the young-
of-the-year life stage, it may indicate that SAS fish are outcompeting wild fish or
otherwise negative affecting wild population. Continuing of SAS program would have to
be re-assessed.
o If the ratio of SAS:wild fish continues to significantly change from the ratio observed at
YOY stage, it would indicate a possibility that certain traits, either phenotypic or
genotypic, have been transferred to the juveniles. The changes in ratios would have to
be carefully assessed, and the consistency between the two SAS intervention rivers
would have to be compared to rule out potential effects due to sampling bias.
Continuing of SAS program would have to be re-assessed after such analysis.
o If the ratio of SAS:wild fish would change from smolt stage to returning adult stage, and
the ratio would change to be biased to higher returns of wild fish, the outcome would
indicate loss of marine adaptation. Continuing of SAS program would have to be re-
assessed, although, it is also considered that the SAS-originated salmon not returning
due to the potential loss of marine adaptation are being eliminated by natural selection,
and would therefore not be further propagating in the system, should this scenario be at
play. Therefore, the potential problem of loss of marine adaptation may be considered to
be “self-correcting”.
o It should be noted that assessment of SAS:wild ratio is best assessed in smolt migration
stage when the population is thoroughly mixed. Monitoring of the ratio during the juvenile
stages is a subject to initial spatial distribution that is somewhat dictated by the spawner
distribution. Random sampling across 20 electofishing sites may alleviate chances of
sampling locations where only SAS fish, or wild fish, have spawned, however, the smolt
stage will represent the best assessment time-point in the freshwater (juvenile) life stage.
• If growth metrics (average size-at-age measured in either length or weight) in wild juvenile
salmon drops below values documented in long-term baseline DFO dataset in SAS
intervention rivers, but not in control rivers, SAS program could be having an effect on
growth of wild Atlantic salmon. Continuing of SAS program would have to be re-assessed.
• If significant changes in allele frequencies in SAS juveniles vs wild juveniles would be
observed in SAS intervention rivers, but no changes in wild juveniles in control rivers
50
compared to (wild) baseline would be observed, continuing of SAS program would have to
be re-assessed and potentially halted.
• If the Atlantic salmon population stock status develops in the Northwest Miramichi River
system to a point where adult returns not only meet the conservation targets set forth by
DFO (CSAS 2017a), but increase enough that population status can be considered “healthy”
(See section 2.3), then the juvenile densities across the system may increase to a point that
density-dependent factors through interference competition may start limiting juvenile
production (density-dependent effects on growth occur primarily already at low population
densities: Grant & Imre 2005; Imre et al. 2005). Adding juveniles in such conditions may be
considered a risk to wild Atlantic salmon in case SAS originated juveniles would outcompete
wild fish. If such conditions are observed due to natural increase in adult returns, continuing
the SAS program would be re-assessed. For the current time, however, increases in
spawning escapement result in concomitant increases in juvenile densities (as expected if
the population follow a Beverton-Holt stock-recruitment relationship; Chaput et al. 2016)
indicating juvenile production well below carrying capacity (non-plateau part of the Beverton-
Holt relationship). A prime example of the expectation for juvenile densities following an
increase in adult spawning escapement above conservation target was observed in 2011 in
the Miramichi system, when both the Northwest and Southwest Miramichi branches
exceeded the minimum conservation requirement and the fry and parr indices showed a
corresponding increase in years following the 2011 spawning cohort (Chaput et al. 2016).
Such increases would not be evident in a system where juvenile production was limited by
excessive density-dependent responses that manifest as changes in survival.
6. Considerations related to First Nations perspectives

Atlantic salmon plays vital importance for the livelihoods and culture of the First Nations
communities that have lived along the Miramichi River since time immemorial. Three Mi’kmaq
First Nations live in immediate proximity of the Miramichi River; The Eel Ground, the
Metepenagiag and the Esgenoôpetitj First Nations. All three rightsholder communities have
indicated the importance and continued resilience on Atlantic salmon for their community
members, and have indicated that it is important that any Atlantic salmon activity in the area is
brought into attention to the communities so that the concerns and comments, as they
specifically relate to the First Nations perspectives, are heard, understood, and taken into
account.
To this end, CAST members have met a number of times with a group representing the interest
of the aforementioned rightsholder communities (Mi'gmawe'l Tplu'taqnn Inc; MTI). Through
these meetings, it has been collectively identified that exchange of knowledge and direct
involvement in the on–the-ground work are important ingredients to ensure the First Nations
interests are not overlooked. It is recognized that the exchange of knowledge will be best in
multiple formats, and is a two-way dialogue involving both the “Western science” (as described
in this proposal with regard to SAS studies) and an Indigenous Knowledge regarding Atlantic
salmon in the area of concern. While CAST is currently working directly with MTI to better
understand the potential scope of an Indigenous Knowledge Study, it is also recognized that the
exchange of knowledge, in two-way manner, is a required component for the proposed CAST
SAS work so that the concerns and comments from the rightsholders can be appropriately
considered. CAST, MTI and DFO discusses the CAST-scope of work (all CAST projects, but to
51
include SAS activity when it occurs) on a monthly information sessions via teleconference.
While the teleconferencing is able to inform the fisheries management staff of the MTI, it is not
sufficient to cover the wider membership of the First Nations. Therefore, community outreach is
required, and will take place in the form of community sessions that are to be arranged, as
needed, in all aforementioned Mi’kmaq communities in collaboration with the MTI. Such
sessions will be arranged prior to releases of fish into rivers in an effort to disseminate the
required information regarding the CAST SAS program to the community members. In addition,
similar sessions may be required as data from the program becomes available to inform
community members regarding the success of the project; such sessions would be arranged, as
necessary, with MTI staff.
With regard to technical aspects of the proposed work, certain components are in place directly
to facilitate First Nations concerns. It is important that SAS adults are identifiable following a
release such that the members of the First Nations communities are able to recognize the
background of a potentially captured adult salmon as a SAS fish. All adult SAS fish are
externally identifiable via adipose fin clipping, and T-anchor tags. Also, as explained previously,
SAS salmon are identifiable by internal PIT-tag, and a hand-held PIT-tag reader can be
provided to each community, as required. As the SAS adults will be identifiable, community
members can decide whether to release a SAS fish upon potential capture (preferred), or to use
it for sustenance. The meat quality of the SAS fish in comparison to the meat quality of wild fish
has not been assessed, but can be done if it is deemed necessary by the rightsholders (the
SAS program objective is to produce spawners to increase juvenile production, not fish for
human consumption).
It is also considered that the removal of the proposed (up to) 5000 wild smolts / year may result
only in a marginal reduction in the ability for First Nation membership to capture returning adults
in subsequent years; this is due to the estimated high at-sea mortality (estimates for smolt-to-
adult survival range 0.3 % to 3.3% for grilse, and 0.2 % to 2.2 % for MSW between 2006 -2010
(Chaput et al. 2016), but with estimated subsequent decline in sea-survival in recent years) and
low-to-moderate trapnet efficiency in First Nations food fisheries (4.1% to 17.5 % for grilse;
4.5% to 14.2% for MSW; Chaput et al. 2001).
52
7. Literature Cited:
Allee, W.C., Park, O., Emerson, A.E., Park, T., and Schmidt, K.P. (1949) Principles of Animal
Ecology. W.B. Saunders Company, Philadelphia.
Anderson, E.C. (2010) Computational algorithms and user-friendly software for parentage-
based tagging of Pacific salmonids. Final report submitted to the Pacific Salmon
Commission’s Chinook Technical Committee, 1–46.
Anderson, E.C., and Garza, J.C. (2006) The power of single-nucleotide polymorphisms for
large-scale parentage inference. Genetics 172, 2567–2582.
Beacham, T.D., Wallace, C., MacConnachie, C., Jonsen, K., McIntosh, B., Candy, J.R., Devlin,
R.H., and Withler, R.E. (2017) Population and individual identification of coho salmon in
British Columbia through parentage-based tagging and genetic stock identification: an
alternative to coded-wire tags. Canadian Journal of Fisheries and Aquatic Sciences.
Berejikian, B.A., Johnson, T., Endicott, R.S., and Lee-Waltermire, J. (2008) Increases in
steelhead (Oncorhynchus mykiss) redd abundance resulting from two conservation hatchery
strategies in the Hamma Hamma River, Washington. Canadian Journal of Fisheries and
Aquatic Sciences 65, 754-764.
CAFSAC [Canadian Atlantic Fisheries Scientific Advisory Committee]. 1991. Definition of
conservation for Atlantic salmon. CAFSAC Advisory Document 91/15.
Chaput, G., Moore, D., Hayward, J., Sheasgreen, J., Dubee, B. (2001). Stock status of Atlantic
salmon (Salmo salar) in the Miramichi River, 2000. Canadian Science Advisory Secretariat
Research Document 2001/008.
Chaput, G., Douglas, S., and Hayward, J. (2016) Biological characteristics and freshwater
population dynamics of Atlantic salmon (Salmo salar) from the Miramichi River, New
Brunswick, Canada. Canadian Science Advisory Secretariat Research Document 2016/029.
Courchamp, F., Clutton-Brock, T., and Grenfell, B. (1999) Inverse density dependence and the
Allee effect. Tree 14, 405-410.
CSAS [Canadian Science Advisory Secretariat]. (2017a) Update of indicators of Atlantic salmon
(Salmo salar) in DFO Gulf Region salmon fishing areas 15 - 18 for 2016. CSAS Science
Response 2017/013.
CSAS [Canadian Science Advisory Secretariat]. (2017b) Spawner abundance and biological
characteristics of striped bass (Morone saxatilis) in the Southern Gulf of St. Lawrence in
2016. CSAS Science response 2017/012.
CSAS [Canadian Science Advisory Secretariat]. (2016) Risks and benefits of juvenile to adult
captive-reared supplementation activities to fitness of wild Atlantic salmon (Salmo salar).
CSAS Science Advisory Report 2016/017.
CSAS [Canadian Science Advisory Secretariat]. (2015) Update of stock status of Atlantic
salmon (Salmo salar) in DFO Gulf Region (Salmon fishing areas 15 to 18) for 2014. CSAS
Science Response 2015/008.
Cunjak, R.A., and Therrien, J. (1998) Inter-stage survival of wild juvenile Atlantic salmon, Salmo
salar L. Fisheries Management and Ecology 5, 209-223.
53
Dempson, J.B., and Furey, G. (1997) Stock status of Atlantic salmon from Conner River, SFA
11, Newfoundland, 1996. Department of Fisheries and Oceans Canadian Stock Assessment
Secretariat Research Document 97/34.
Dempson, J.B., Pepper, V.A., Furey, G., Bloom, M., Nicholls, T., and Hoskins, G. (1999)
Evaluation of an alternative strategy to enhance salmon populations: Cage rearing wild
smolts from Conne River, Newfoundland. ICES Journal of Marine Sciences 56, 422-432.
DFO [Fisheries and Oceans Canada]. 2009. Canada’s policy for conservation of wild Atlantic
salmon.
Dietrich, J.P., and Cunjak, R.A. (2006) Evaluation of the impacts of Carlin-tags, fin-clips and
Panjet tattoos on juvenile Atlantic salmon. North American Journal for Fisheries
Management 26, 163-169.
Douglas, S.G., Chaput, G., Hayward, J., and Sheasgreen, J. (2015) Assessment of Atlantic
salmon (Salmo salar) in salmon fishing areas 16 of the southern Gulf of St Lawrence to
2013. Canadian Science Advisory Secretariat Research Document 2015/049.
Elson, P.F. 1957. Number of salmon needed to maintain stocks. Canadian Fish Culturist 21, 19-
23.
Fraser, D.J. (2016) Risks and benefits of mitigating low marine survival in wild salmon using
smolt-to-adult captive-reared supplementation. Canadian Science Advisory Secretariat
Gibson, A.J.F, and Claytor, R.R. (2012) What is 2.4? Placing Atlantic salmon conservation
requirements in the context of the precautionary approach to fisheries management in the
Maritimes region. Canadian Science Advisory Secretariat Research Document 2012/043.
Grant, J.W.A, and Imre, I. Patterns of density-dependent growth in juvenile stream-dwelling
salmonids. Journal of Fish Biology 67 (Supplement B), 100-110.
Huisman J (2017) Pedigree reconstruction from SNP data: Parentage assignment, sibship
clustering, and beyond. Molecular Ecology Resources 38, 42–49.
IBIS [Integrated Aquatic Resources Management between Ireland, Northern Ireland and
Scotland]. (2013) To Stock or Not to Stock? Support documentation arising from the IBIS /
AST Conference on Restocking. Boosting salmon numbers: is stocking the answer or the
problem? 27‐28 November 2013, Glasgow, Scotland.
ICES [International Council for the Exploration of the Sea]. (2017) Report of the Working Group
on North Atlantic Salmon (WGNAS), 29 March–7 April 2017, Copenhagen, Denmark. ICES
CM 2017/ACOM:20. 296 pp.
Imre, I., Grant, J.W.A., and Cunjak R.A. (2005) Density-dependent growth of young-of-the-year
Atlantic salmon Salmo salar in Catamaran Brook, New Brunswick. Journal of Animal
Ecology 74, 508-516.
Jones, A.G., Small, C.M., Paczolt, K.A., and Ratterman, N.L. (2010) A practical guide to
methods of parentage analysis. Molecular Ecology Resources 10, 6–30.
Jones, O.R., and, Wang, J. (2010) COLONY: a program for parentage and sibship inference
from multilocus genotype data. Molecular Ecology Resources 10, 551–555.
Jones, R.A., Ratelle, S., Anderson, L., and Whitelaw, J. (2015) Captive-reared adult release
program upriver of Mactaquac Dam with an emphasis on Tobique River. A presentation
54
delivered in Canadian Science Advisory Secretariat meeting to review Miramichi Captive-

rearing proposal, December 14-16, 2015. Moncton, NB, Canada.
Jones, R.A., Anderson, L., and Clarke, C.N. (2014) Assessment of the Recovery Potential for
the Outer Bay of Fundy Population of Atlantic Salmon (Salmo salar): Status, Trends,
Distribution, Life History Characteristics and Recovery Targets. Canadian Science Advisory
Secretariat Research Document 2014/008.
Jonsson, B., Waples, R.S., and Friedland, K.D. (1999) Extinction considerations for diadromous
fishes. ICES Journal of Marine Science 56, 405-409.
Kozfkay, C.C., Stark, E.J., Baker, D.J., Ayers, B.S., Venditti, D.A., McCormick, J., Kline, P.
(2017). Captive rearing program for Salmon River Chinook salmon. Completion Report
January 1, 1995—December 31, 2014. Idaho Department of Fish and Game Report Number
16-11. Xiii + 119p.
Moore, D., and Chaput, G. (2007). Juvenile Atlantic salmon (Salmo salar) surveys in the
Miramichi River Watershed from 1970-2004. Canadian Data. Report of Fisheries and
Aquatic Sciences 1188. 117p.
MSA [Miramichi Salmon Association]. (2016) Miramichi salmon and trout restoration – Stocking
2016. Accessed 1 July, 2017 at http://www.miramichisalmon.ca/wp-
content/uploads/2016/02/Stocking-Report-2016-1.pdf
Pavey, S.A. (2016) Molecular techniques for parentage analysis to assess supplementation
effectiveness for Atlantic Salmon (Salmo salar) on the Miramichi River. Canadian Science
Advisory Secretariat Research Document 2016/031.
Randall, R.G. (1985) Spawning potential and spawning requirements of Atlantic salmon in the
Miramichi River, New Brunswick. Canadian Atlantic Fisheries Scientific Advisory Committee
Reisenbichler, R. R., and Rubin, S. P. (1999) Genetic changes from artificial propagation of
Pacific salmon affect the productivity and viability of supplemented populations. ICES
Journal of Marine Science 56, 459–466.
Steele, C.A., Anderson, E.C., Ackerman, M.W. et al. (2013) A validation of parentage-based
tagging using hatchery steelhead in the Snake River basin. Canadian Journal of Fisheries
and Aquatic Sciences 70, 1046–1054.
Thomas, R. L. (1996). Enhancing threatened salmonid populations: a better way. Fisheries, 21,
12–14.
Traill, L.W., Bradshaw, J.A., Brook, B.W. (2007) Minimum viable population size: A meta-
analysis of 30 years of published estimates. Biological Conservation 139, 159–166.
Venditti, D.A., James, C., and Kline, P. (2013) Reproductive behavior and success of captive-
reared Chinook salmon spawning under natural conditions. North American Journal for
Fisheries Management 33, 97–107.
Wallace, B., and Curry, R.A. (2017). Assessing the outcomes of stocking hatchery juveniles in
the presence of wild Atlantic Salmon. Environmental Biology of Fishes. [pagination not final]
DOI 10.1007/s10641-017-0613-2
55

CAST SAS Plan December 2017

Uploaded by

Copyright:

Available Formats

CAST SAS Plan December 2017

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

CAST SAS Plan December 2017

Uploaded by

Copyright:

Available Formats

Review of the risks and benefits of

National Capital Region CAST SAS Experiment Proposal

COLLABORATION FOR ATLANTIC SALMON TOMORROW (CAST)

Management intervention is important because the risks to population perseverance increase as

Smolt-to-adult supplementation (SAS; a term coined by Fraser (2016)) has emerged as a

2. General Description of the SAS Experiment

2.2. Experimental Components – General Description

2.3. Area of Experiment – The Miramichi River

3. Smolt Capture, Operational Protocols in Captivity, and

3.2. Operational Protocols in Captivity

Smolt Initial Survivors (31 Survival

4. Genetic Monitoring Tools of the Program; General

5. Detailed description of proposed studies, hypothesis

5.1. Program 1: Sub-basin genetic structure of Atlantic salmon

5.2. Program 2: Laboratory experiments

o (for 2017; crosses were fertilized 20- 27 October)

5.3. Program 3: Experimental River

1. SAS adult behaviour is similar to the behaviour of wild fish.

• Project 3 - Movements and behavior of juveniles

5.4. Program 4: SAS impacts on a natural river

For the SAS progeny, the following hypothesis is set forward:

• Project 5 - Assessing SAS contributions to smolt migrations

A release of 1000 SAS females / replicate tributary is considered an absolute minimum to

• Juvenile - Inter-stage survival (fitness), growth, body shape, genetic differences

6. Considerations related to First Nations perspectives

delivered in Canadian Science Advisory Secretariat meeting to review Miramichi Captive-

You might also like