Atlantic salmon {Salmo salar) in Cobourg Brook, Ontario

Survival, growth, and out-migration timing of reintroduced

Atlantic salmon {Salmo salar) in Cobourg Brook, Ontario

A thesis submitted to the Committee on Graduate Studies in partial

fulfillment of the requirements for the Degree of Master of Science in the

Environmental & Life Sciences graduate Program

TRENT UNIVERSITY

Peterborough, Ontario, Canada

© Copyright by Russell Bobrowski 2010

Environmental & Life Sciences Program

September 2010

?F? Library and ArchivesCanada

Published HeritageBranch

395 Wellington StreetOttawa ON K1A0N4Canada

Bibliothèque etArchives Canada

Direction duPatrimoine de l'édition

395, rue WellingtonOttawa ON K1A0N4Canada

Your file Votre référenceISBN: 978-0-494-68234-0Our fìle Notre référenceISBN: 978-0-494-68234-0

NOTICE:

The author has granted a non-exclusive license allowing Library andArchives Canada to reproduce,publish, archive, preserve, conserve,communicate to the public bytelecommunication or on the Internet,loan, distribute and sell thesesworldwide, for commercial or non-commercial purposes, in microform,paper, electronic and/or any otherformats.

The author retains copyrightownership and moral rights in thisthesis. Neither the thesis norsubstantial extracts from it may beprinted or otherwise reproducedwithout the author's permission.

AVIS:

L'auteur a accordé une licence non exclusivepermettant à la Bibliothèque et ArchivesCanada de reproduire, publier, archiver,sauvegarder, conserver, transmettre au publicpar télécommunication ou par l'Internet, prêter,distribuer et vendre des thèses partout dans lemonde, à des fins commerciales ou autres, sursupport microforme, papier, électronique et/ouautres formats.

L'auteur conserve la propriété du droit d'auteuret des droits moraux qui protège cette thèse. Nila thèse ni des extraits substantiels de celle-cine doivent être imprimés ou autrementreproduits sans son autorisation.

In compliance with the CanadianPrivacy Act some supporting formsmay have been removed from thisthesis.

While these forms may be includedin the document page count, theirremoval does not represent any lossof content from the thesis.

Conformément à la loi canadienne sur laprotection de la vie privée, quelquesformulaires secondaires ont été enlevés decette thèse.

Bien que ces formulaires aient inclus dansla pagination, il n'y aura aucun contenumanquant.

¦+¦

Canada

Abstract

Survival, growth, and out-migration timing of reintroduced Atlantic salmon (Salmo

salar) in Cobourg Brook, Ontario

Russell Bobrowski

Lake Ontario once hosted an abundant population of Atlantic salmon (Salmo

salar), which died out prior to the 20th century due to human disturbances. A recovery

effort has recently been established which uses an experimental approach by stocking

three Atlantic salmon life stages into best-bet Lake Ontario tributaries. This thesis

focused on the in-stream phase of the Atlantic salmon life cycle to evaluate the

comparative growth, survival, and out-migration timing of stocked juvenile life stages.

Fry and yearling Atlantic salmon were stocked as a paired-release introduction into

Cobourg Brook, and evaluated for differences in survival, growth, and timing of

migration/movement into Lake Ontario. In addition, fall fingerlings and age-2 Atlantic

salmon were also stocked into the study site, and some insight into their performance was

provided. Stocked yearlings exhibited poorer growth, but produced significantly more

and larger out-migrants than salmon stocked as fry. Atlantic salmon stocked as fry

displayed typical smolt behaviour and out-migrated during the predicted environmental

smolt window for Cobourg Brook, whereas stocked yearlings out-migrated during

atypical and environmentally suboptimal times. Older life stages appear to display greater

short-term survival by avoiding high mortality events in the wild, however, maladaptive

behaviors may develop due to increased time in captivity which may reduce overall

fitness and subsequently their efficacy at re-establishing wild populations. This

i

generalization may be valuable for future native species recovery programs when

choosing a life stage to release into the wild. By integrating the new information provided

in this thesis and re-evaluating information needs, the experimental approach employed

in the current effort to re-establish Atlantic salmon populations in Lake Ontario

tributaries would not only assist the repatriation of an single extirpated species, but also

aid in the restoration of a highly valued ecosystem and help advance our knowledge of

effective methods to reintroduce extirpated wildlife populations.

11

Acknowledgements

I would like to thank my co-advisors, Dr. Chris Wilson and Dr. Nick Jones. Your

guidance and expertise has been much appreciated. On a personal note I would like to

thank my friends and family, particularly my fiancée Meagan. Their support has helped

make this thesis an enjoyable process. I would also like to thank my friends/collogues,

Fritz Fischer, Peter Addison, and Marc Desjardins for their input and suggestions

throughout this process. Peter Amiro, Steven McCormick, and John Casselman provided

insight into their specialized fields of expertise which was very helpful. I would also like

to thank my field assistants, Justin Post and Steve Agaliotis who provided tremendous

help while attending the smolt trap during rain events. I would also like to thank the

many volunteers who assisted with field work.

Many organizations have contributed to make this project possible. Special thanks

to the Ontario Ministry ofNatural Resources and the Ontario Federation of Anglers and

Hunters who provided the original research opportunity, as well as considerable financial

and technical assistance. The Department of Fisheries and Oceans Canada and the

Ganaraska Region Conservation Authority provided additional resources to assist with

data collection. Financial assistance from the National Science and Engineering Research

Council of Canada was also much appreciated.

in

Table of Contents

ABSTRACT IACKNOWLEDGEMENTS ??TABLEOFCONTENTS IVLISTOFTABLES VLISTOFFIGURES viCHAPTER 1: GENERAL INTRODUCTION 1CHAPTER 2: EVALUATING THE IN-STREAM SURVIVAL AND GROWTH OFJUVENILE LIFE STAGES OF ATLANTIC SALMON STOCKED INTO COBOURGBROOK, ONTARIO 22

Abstract 23Introduction 25Methods 29Results 48Discussion 71References: 80

CHAPTER 3: THE OUT-MIGRATION TIMING OF STOCKED FRY ANDYEARLING ATLANTIC SALMON FROM COBOURG BROOK 89

Abstract 90Introduction 91Methods 96Results 101Discussion: HOReferences 117

CHAPTER 4: SUMMARY AND SYNTHESIS 122APPENDIXl 140

IV

List of Tables

Table 2-1: Summary of Atlantic salmon stocked into Cobourg Brook throughout theduration of this study 33

Table 3-1: Summary of published initiating temperature and duration of wild Atlanticsalmon smolt migration 100

Table 3-2: Stream conditions during out-migration of stocked juvenile Atlantic salmonfrom Cobourg Brook 105

Table A-I: Results of analyses investigating possible criteria for differentiating betweenknown and unknown origin Atlantic salmon smolts 148

Table A-2: Potential origin of salmon captured in Cobourg brook from 2007-2008, andvalidation or corroboration techniques suggested 158

?

List of Figures

Figure 1-1: A generalization of the factors influencing the fitness of reintroducedanimals 4

Figure 1-2: Historical distribution of Lake Ontario tributaries used by Atlantic salmonfor spawning(Parsons, 1978) 8

Figure 1-3: Issiltration of the environmental smolt window hypothesis (McCormick etal, 1998) 14

Figure 2-1: Map of the study site 30

Figure 2-2: A timeline of Atlantic salmon stocking and capture events that occurredthroughout the duration of this study 34

Figure 2-3: An illustration of the stream level and temperature regime of Cobourg Brookthroughout the duration of this study 49

Figure 2-4: Spring and summer 2007 daily smolt trap catch of marked (adipose clip orfloy tag), or unmarked juvenile Atlantic salmon 51

Figure 2-5: Spring 2007 weekly smolt trap catch, capture probability, and estimatedpopulation sizes of out-migrating Atlantic salmon 52

Figure 2-6: (A) Densities of stocked Atlantic salmon (ATS), and (B) other salmonids(brown trout, brook trout, rainbow trout, and Chinook salmon) at sites sampled inCobourg Brook during fall 2007 55

Figure 2-7: Summary of Atlantic salmon smolt trap catch during 2008 58

Figure 2-8: Weekly catch and pre-handling mortalities ofjuvenile Atlantic salmonduring 2008 smolt trapping 60

Figure 2-9: (A) Densities of stocked Atlantic salmon (ATS), and (B) other salmonids atsites sampled in Cobourg Brook during fall 2008 using backpack electrofishing 63

Figure 2-10: Fall 2008 length frequency distribution ofjuvenile Atlantic salmon thatwere originally stocked as fry in 2007 or 2008, or yearlings in 2008 64

Figure 2-11: Capture probability of single-pass backpack electrofishing for group-0 (A),and group-1 (B) salmonids in Cobourg Brook 65

Figure 2-12: Length frequency distribution of stocked fry captured in fall 2008 bybackpack electrofishing 68

Vl

Figure 2-13: Spring 2008 CPUE of age-1 out-migrants originally stocked as fry (A) oryearlings (B) in the paired release-recapture experiment 69

Figure 2-14: Condition (weight-at-length) of stocked yearling (A), and fry (B) Atlanticsalmon from the paired release-recapture experiment during spring and fall 2008 70

Figure 3-1: Summary of Cobourg Brook Atlantic salmon smolt trap catch and streamconditions during the spring/summer of 2007 103

Figure 3-2: Cobourg Brook Atlantic salmon smolt trap catch in 2007; (A) expected usinghypothetical smolt windows, (B) observed marked yearlings, and (C) observed unmarkedjuvenile Atlantic salmon 104

Figure 3-3: Summary of Cobourg Brook Atlantic salmon smolt trap catch and streamconditions during the spring/summer of 2008 108

Figure 3-4: Cobourg Brook Atlantic salmon smolt trap catch in 2008; (A) expected usinghypothetical smolt windows, (B) observed stocked yearlings, and (C) observed stockedfry 109

Figure 4-1: A conceptual diagram of the ideal life stage hypothesis 130

Figure A-I: The distribution of the distances to checks greater than 2/9ths in circularextent. Both known (e.g. yearling) and unknown origin Atlantic salmon collected fromCobourg Brook are shown 147

Figure A-2: The number of substantial checks (those 3/9ths or greater in circular extent)observed on scales of known yearling and unknown origin Atlantic salmon smolts fromCobourg Brook, Ontario 149

Figure A-3: Frequency distribution of distance to annuii of known and unknown originsmolts captured from Coboug Brook, Ontario 150

vii

Chapter 1: General Introduction

?

Humans have altered much of Earth's habitats, resulting in numerous species

extinctions and local extirpations (Pimm et al, 1995; Vitousek et al, 1997). The human-

induced decline in global species diversity has received considerable attention, given the

ethical, aesthetic, and monetary benefits associated with natural and diverse ecosystems

(Daily, 1997; Tilman, 2000; Balmford et al, 2002; Hooper et al, 2005). Reintroductions

of native species have become an increasingly popular management technique to regain

the societal (e.g. aesthetics, ethics, monetary) and biological (e.g. biodiversity) value of

damaged landscapes and ecosystems (Fischer and Lindenmayer, 2000; Seddon et al,

2006). In addition, reintroducing native species may promote the re-establishment of

historic ecosystem structure and function (Mittelbach et al, 1995; Simberloff, 1998). and

also test if damaged ecosystems have sufficiently recovered to once again support native

taxa (Dobson et al, 2006; Rooney et al, 2006).

Although native species reintroduction programs have a long history, early

attempts were often unsuccessful, with little broader knowledge gained (Lyles and May,

1987; Griffith et al, 1989; Kleiman, 1989; Seddon, 1999; Fischer and Lindenmayer,

2000). In response, reintroduction biology was established as a scientific discipline, with

an emphasis on employing large-scale experiments with testable a priori hypotheses and

predictions to reduce the uncertainty associated with releasing native species into habitats

where they once died out (Sarrazin and Barbault, 1996; Armstrong and Seddon 2008).

Reintroducing known-origin individuals into novel habitats affords an excellent

opportunity to conduct ecosystem-scale experiments to address both basic and applied

ecological questions (Sarazin and Barbault, 1996). Research questions tend to be applied

in nature during the initial phase of reintroduction programs to improve captive breeding

2

and release methodology (Sarrazin and Barbault, 1 996). However, the diverse array of

factors which could effect population re-establishment may be best approached using key

ecologie concepts from the fields of population biology, genetics, behavioral ecology,and evolution (Sarrazin and Barbault, 1996). The success of animal reintroduction

programs is largely contingent on the availability of suitable habitat (including the biotic

environment), alleviation of the original causes of extirpation (Sarrazin and Barbault,

1996; Seddon et al, 2007), and the ancestry of released animals (Olsson 2007; Meffe

1995; Lacy 1997). However, other secondary factors may also influence the performance

of reintroduced animals such as the quantity, composition (life history stage, size, etc.),

and pre- or post-release management of the release group (Seddon and Armstrong, 2008:

Figure 1-1). By integrating an experimental approach, ongoing reintroduction programscan contribute to the growing general knowledge of factors affecting the success of

reintroductions as well as identifying efficient methodologies for reintroducing extirpatedpopulations.

3

(1) Original cause of extirpation (3) Ancestry (2) Habitat

Pre-release

management

Fitness of reintroduced animals

Rearingconditions

Composition of therelease group (i.e. life

stage, size, age)

Y

(4)

\Post-releasemanagement

Releasemethod

(location,quantity)

Figure 1-1: A generalization of the factors influencing the fitness of reintroduced

animals. The size of the arrow indicates the relative strength of the factor.

4

Many species which have been severely impacted by human disturbance have

specific habitat requirements, low reproductive rate, long generation times, and exist at

low densities, making reintroduction difficult. Examples include the California condor

(Gymnogyps californianus), white-tailed eagle {Haliaeetus albicilla), griffon vulture

{Gyps fulvus), and the white rhinoceros {Ceratotherium simum) (Wilson and Price, 1994;

Griffith et al, 1989; Cade and Temple, 1997). However, most fish species present an

interesting case, as they are typically short lived, highly fecund, and exist at high

densities, suggesting that reintroduction may be feasible if adequate habitat exists, the

cause of extirpation has been eliminated, and released animals are of suitable ancestry.

Although arguments can be made for many fish species, the wide global distribution, rich

cultural history, economic importance, and relatively large amount of research attention

make the Atlantic salmon {Salmo salar) a suitable candidate for local reintroduction

programs.

Throughout their native range, Atlantic salmon populations have declined or been

lost due to human activities such as stream dewatering, damming, and pollution (Parrish

et al., 1998). Efforts to restore or reintroduce Atlantic salmon populations have been

occurring at least since the 1 800s, and have taken place in nearly every country

throughout their range (Crawford, 2001; Bielak and Davidson, 1993; Webb et ai, 2008;

Mills, 1989). Like many other species; however, restoring Atlantic salmon populations

has proven a difficult endeavor (Gephard, 2008; Kennedy, 1988), and relatively little

general knowledge exists on how to successfully re-establish self-sustaining populations

in regions where they have been lost. By applying the key principles of reintroduction

biology such as large-scale experimentation with apriori hypotheses, our understanding

5

of the ecological requirements for successfully reintroducing Atlantic salmon can be

improved.

Atlantic salmon typically exist as either landlocked or anadromous populations

throughout their range. Both types have similar, but complex habitat requirements.

Generalizations of the life history and associated habitat requirements are provided by

Klemetsen et al, (2003) and Webb et al, (2007). Atlantic salmon spawn in streams where

eggs are laid in gravel nests and embryos incubate over winter and hatch in early spring.Juveniles rely on egg-derived nourishment from their yolk sac for a short duration after

hatching, but as endogenous nourishment dwindles they emerge from the gravel as free-

swimming and feeding fry. Depending on the productivity of the stream, juvenile salmon

may spend the next 1-7 years in their natal streams as parr (Webb et al., 2008). Fry and

parr feed opportunistically on available aquatic and terrestrial invertebrates, and may also

become cannibalistic at larger sizes. Parr prefer shallow riffles 10-80 cm sec"1 (Webb et

al,. 2008), and prime feeding conditions occurs at temperatures ranging from 15-19°C,

and mortality can occur when temperatures exceed 27°C (Garside 1973; Elliott 1991 ;

Elliott and Elliott, 1995). Upon attaining a size-related threshold, parr undergo

physiological, morphological, and behavioral changes to become smolts (McCormick et

al, 1998). Smolts out-migrate from their natal streams in response to environmental cues,

and inhabit lakes or oceans until they return to their natal streams to reproduce. Unlike

Pacific salmon (Oncorhynchus sp.), Atlantic salmon are iteroparous spawners, and cansurvive multiple spawning and migration events.

The historical population of Atlantic salmon in Lake Ontario represented the

largest freshwater population across the species range (Blair, 1938; Parsons, 1978;

6

MacCrimmon and Gots, 1979; Webster, 1982). Historically widespread and abundant,

they spawned in nearly every available tributary (Figure 1-2: Parsons, 1978) and over 400

pounds of salmon could be harvested per night in a single spawning stream by two people

with pitchforks in the early 1800s (Dymond, 1966).

Early settlers of the Lake Ontario basin (ca. 1793) quickly cleared land, farmed,

and dammed streams for timber driving, mills, and industry such as tanneries (Parsons,

1978). Salmon were harvested from nearly every tributary by the 1 820s to feed the

growing human population, and were amongst the least expensive food items available

(Parsons, 1978; Dunfield, 1985). The cumulative effects of these stressors had rapid

negative impacts on the salmon population. By the 1830s fewer salmon were taken in

southern regions (Dunfield, 1985), and by 1853 Atlantic salmon had disappeared from

the Don River, which flows through present day Toronto (Huntsman, 1 944). In Canada,

the decline was taken seriously, as fish ladders were constructed in 1860's, restocking

began 1866, and harvest was banned in 1870 (Carcao, 1987). Despite conservation efforts

and the active stocking program, the Lake Ontario Atlantic salmon became extirpated,

and the last salmon was reported in 1898 (Carcao, 1987; Crawford, 2001).

7

N

I ONf**'0 2e

/ ZTZOta /3 5 IO SS 20 25

¿* Jp

/?

ONTARIOm 7 / 28^6 \> CÄKftO* 29

O.S.*»

3t

\34\ F39 3840

35NEW YORK 363?

Figure 1-2: Historical distribution of Lake Ontario tributaries used by Atlantic salmon

for spawning. Spawning streams are numbered, and circled numbers indicate streams

planted with native salmon during the original restoration effort from 1 866-1 881

(Parsons, 1978).

8

Their extirpation has been attributed to the previously mentioned anthropogenic

stressors during early human colonization of the Lake Ontario basin (Parsons, 1978;

Carcao, 1987; Crawford, 2001), although other factors may have also contributed.

Invasion of alewife (Alosa pseudoharengus) into Lake Ontario in the late 1800s and

subsequent consumption by Atlantic salmon may have caused thiamine deficiency, which

can lead to catastrophic mortality of early life stages (Ketola et al, 2000). Poor climatic

conditions may have also stressed the last remaining Lake Ontario Atlantic salmon, as

exceptionally hot, dry summers and cold winters occurred from 1872-1877 (Carcao,

1987).

After the demise of Atlantic salmon the Lake Ontario ecosystem continued to

deteriorate, coming to a ecological low in the early 1 970s when virtually all native fish

stocks were either lost or severely depressed, and only three non-native fish species were

abundant (Christie, 1972). Land management improved through the latter 20th century,benefiting both terrestrial and aquatic habitats, and by the 1990s many native fish stocks

rebounded to historic highs (Mills et al, 2003). A variety of non-native salmonid species

have been introduced and some species (particularly rainbow and brown trout) are now

well-established as naturalized populations which may partially occupy the ecologie

niche previously held by Atlantic salmon (Crawford, 2001).

Attempts to conserve or restore Lake Ontario Atlantic salmon have been tried

repeatedly since their initial decline. In 1 866, Samuel Wilmot began an artificial

propagation program which utilized the native stock of Lake Ontario Atlantic salmon,

(Parsons, 1978). By 1884 however, too few adults could be obtained and the program

was abandoned (MacCrimmon, 1965; Carcao, 1987). In the 1940s the Ontario

9

Department of Lands and Forests attempted to reintroduce Atlantic salmon in Duffins

Creek (Bisset et al., 1995), but failed to re-establish a population due to high mortality of

early life stages that were associated with poor stream conditions (MacCrimmon, 1954).

In 1987 the Ontario Ministry of Natural Resources (OMNR) initiated a program to assess

the feasibility of restoring naturally reproducing Atlantic salmon populations to support a

sport fishery, and although adult returns were lower than expected, findings suggested

that Atlantic salmon reintroduction was possible and a restoration plan was developed in

1995 (Bisset et al., 1995). The restoration plan was implemented through small-scale

experimental stockings in high quality streams which found survival was acceptable up to

the first winter, although competition with naturalized salmonids may present an issue

(Jones and Stanfield, 1993; Stanfield and Jones, 2003). Recent workshops have focused

on identifying key information needs to restore self-sustaining populations (Grieg et al.,

2003).

Restocking efforts to mitigate the original decline of Lake Ontario Atlantic

salmon also occurred along the south shore of Lake Ontario, although to a lesser extent

than the Ontario effort (Crawford, 2001). New York state revisited Atlantic salmon

stocking in 1953, and again in 1983; both attempts resulted in returning adults, but failed

to re-establish a successfully reproducing population (Abraham, 1983; Eckert, 2003).

Efforts were than refocused to establish a hatchery-reliant sport fishery (Eckert, 2003),

which is still in operation today. Although the physical conditions within many south

shore Lake Ontario streams are not suitable for juvenile salmonids, some streams in the

south-eastern region are promising candidates for Atlantic salmon reintroduction

(Coghlan et al, 2007).

10

In 2006 a collaborative partnership between universities, non-government

agencies, and several levels of government was established with the goal of establishingself-sustaining populations of Atlantic salmon in Lake Ontario tributaries within the next

ten to fifteen years (OMNR, 2007). Complying with the general principles of

reintroduction biology, the partnership employed a landscape-scale experiment to

develop an effective recovery strategy. Atlantic salmon brood stocks (adults for juvenileproduction) from Sebago Lake (Maine), Lac St. Jean (Quebec), and LaHave River (NovaScotia) have been acquired for reintroduction purposes (OMNR 2009). Juveniles from

each captive population are planned to be introduced into Lake Ontario streams in hopesof re-establishing populations (Cross et al, 2007; Grieg et al, 2003). Prior to the

availability ofjuveniles from all three source populations in 2012, identifying aneffective life history stage(s) for stocking has been identified as a research priority (Grieget al, 2003).

Ontario hatcheries are currently producing fry, fall fmgerlings, and yearlings for

Atlantic salmon reintroduction in Lake Ontario tributaries. Fry are retained in captivityuntil transfer to exogenous feeding and are stocked out at the beginning of their first

growing season. Fall fmgerlings are stocked out at the end of the their first growingseason, and yearlings over-winter in captivity to be stocked out at the beginning of theirsecond growing season, ideally upon reaching smolt stage (G. Durant, OMNR, per.comm.).

The survival of stocked juveniles until out-migration from recipient watershedsmay be an appropriate metric to gauge the performance of stocked Atlantic salmon life

stages, assuming that the abundance of out-migrants will be proportional to the

11

abundance of returning adults (Jonsson et al, 1998). Maximum survival of stocked

juveniles may be achieved by avoiding high-mortality life stages which occur in the wild

and mitigating maladaptive traits acquired through captive rearing (Coghlan and Ringler,

2004). The juvenile life history of wild Atlantic salmon is marked with many high

mortality stages, particularly during gravel emergence (MacKenzie and Moring, 1988),

transfer from endogenous to exogenous feeding (Balon, 1985), over-wintering (Cunjak

et al, 1998), and smolt migration (Scarnecchia, 1984; Hvidsten and Lund, 1988; Thorpe,

1988). Captive rearing relaxes selective pressures and improves survival through early

life stages which typically show high mortality in the wild (Araki et al, 2008). However,

many critical behaviors such as foraging and predator avoidance can become adapted to

the unnatural conditions within hatchery facilities, causing high mortality after transfer

into the wild and ultimately resulting in reduced overall fitness (Brown and Day, 2002;Araki et al, 2008).

The timing ofjuvenile migration/movement out of nursery streams may also

provide insight into the performance of stocked Atlantic salmon life stages. McCormick

et al, (1998) presented the 'environmental smolt window' hypothesis which suggestsmaximum survival occurs when out-migration overlaps with favorable environmental

conditions such as suitable stream temperatures or levels, presence ofpredators or prey,

or lake/ocean conditions (Figure 1-3). This hypothesis has been supported by the early

work of Hansen (1987), who found atypical out-migration timing ofjuvenile Atlantic

salmon resulted in relatively few returning adults. In addition, the more recent work of

Hansen and Jonsson (1991), Jokikokko and Mantyniemi (2003), and Kallio-Nyberg

12

(2004), found environmental conditions upon entry into the marine environment were

important for subsequent survival of post-smolt Atlantic salmon.

The objective of this thesis was to address a research priority in the current effort

to re-establish Atlantic salmon populations in Lake Ontario tributaries: which life stage(s)

should be stocked to provide the best opportunity for population re-establishment (Grieg

et al, 2006)? Likewise, this thesis also addresses a key question in reintroduction

biology: how is the likelihood of population re-establishment affected by the composition

(e.g. life history stage) of the release group (Armstrong and Seddon, 2008, Robert et al,.

2003; Sarrazin and Legendre, 1999)? Two research projects presented in Chapters 2 and

3 investigated complementary aspects of the juvenile ecology of multiple Atlantic salmon

life history stages stocked into Cobourg Brook, one of three Lake Ontario tributaries

selected for the first phase of the current restoration effort.

13

THE StóOLT "WiNDOW" AND EFFECTOW SURVIVAL

ft¥M3f«¡ESTAI

i S 4

/oß;

Sü»wm.

X*ieaa -?

s i s s 4 s- s ?

Figure 1-3: The environmental smolt window hypothesis (McCormick et al, 1998).

When smolt out-migration timing and optimal environmental conditions coincide (solidlines), adult returns are maximized. When smolt out-migration and environmental

conditions are out of phase (dotted lines), adult returns are reduced.

14

Chapter 2 evaluates the growth and survival of multiple genetically tagged life

stages of Atlantic salmon released into Cobourg Brook by employing a paired-releaseexperiment. Chapter 2 also investigates the influence of the abiotic environment on the

performance of the stocked life stages, as well as providing management implications and

recommendations from supplementary information collected during field surveys.

Chapter 3 complements the growth and survival study by focusing on the behavioral out-

migration timing ofjuvenile Atlantic salmon from Cobourg Brook. Specifically, I testedthe hypothesis that fry would out-migrate within the 'environmental smolt window'

(McCormick et al, 1998), while stocked yearlings would not due to shifts in ecology

caused by longer rearing in unnatural hatchery rearing conditions. Migration out of phase

with the environmental smolt window suggests that survival to adulthood may be

reduced. Chapter 3 also describes the stream conditions and seasonal timing ofjuvenile

Atlantic salmon out-migration from Cobourg Brook, as well as the relationship betweenout-migration timing and stream conditions.

Together, these chapters provide a detailed investigation of the early life history

of Atlantic salmon stocked into Lake Ontario tributaries, providing new information onwhich life history stage(s) should be stocked in the effort to re-establish Atlantic salmon

populations in Lake Ontario. The fourth and final chapter addresses the implications of

this research and provides recommendations to improve the effectiveness of stocking

efforts. The restoration of self-sustaining populations of Atlantic salmon in Lake Ontario

would not only represent the repatriation of an extirpated species, but also aid in the

restoration of a highly valued ecosystem and help advance our knowledge of effective

methods to reintroduce migratory fish populations.

15

References

Abraham, W. J. (1983). A Plan to Determine the Feasibility of Atlantic SalmonRestoration in Lake Ontario. New York Department ofEnvironmental Conservation. 13p.

Armstrong, D. P. & Seddon, P. J. (2008). Directions in reintroduction biology. Trends inEcology & Evolution 23, 20-25.

Armstrong, D.P., Soderquist, T. & Southgate, R. (1994) Designing experimentalreintroductions as experiments. Réintroduction biology ofAustralian and New ZealandFauna fed. M. Serena), pp. 27-29. Surrey Beatty & Sons, Chipping Norton.

Balmford, A., Bruner, ?., Cooper, P., Costanza, R., Farber, S., Green, R. E., Jenkins, M.,Jefferiss, P., Jessamy, V., Madden, J., Munro, K., Myers, N., Naeem, S., Paavola, J.,Rayment, M., Rosendo, S., Roughgarden, J., Trumper, K. & Turner, R. K. (2002).Ecology - Economic reasons for conserving wild nature. Science 297, 950-953.

Bielak, A. T. & Davidson, K. (1993). New enhancement strategies in action. In Salmonin the sea and new enhancement strategies, pp. 299-320. Ed. by Derek Mills. FishingNews Books. Blackwell Scientific Publications Ltd.. Oxford. 424 pp.

Bisset, J., Bowlby, J., Jones, M., Marchant, B., Miller-Dodd, L., Orsatti, S. & Stanfield,L. (1995). An Atlantic salmon restoration plan for Lake Ontario. Picton, Ontario: LakeOntario Management Unit

Blair, A. A. (1938). Scales of Lake Ontario salmon indicate landlocked form. Copeia1938, 206.

Cade, T.J. and Temple, S.A. (1995) Management of threatened bird species: anevaluation of the hands-on approach, Ibis 137(Suppl. 1), 161-172 34

Carcao, G. (1987). Atlantic salmon in the Great Lakes basin: a history of its extirpationand attempted restoration. Unpublished manuscript, H.A. Regier. University of Toronto,Department of Zoology.

Christie, W. J. (1972). Lake Ontario: Effects of exploitation, introductions, andeutrophication on the salmonid community. Journal ofthe Fisheries Research Board ofCanada, 29. 913 -929.

Coghlan, S. M., Connerton, M. J., Ringler, N. H., Stewart, D. J. & Mead, J. V. (2007).Survival and growth responses ofjuvenile salmonines stocked in eastern Lake Ontariotributaries. Transactions ofthe American Fisheries Society 136, 56-71.

16

Coghlan, S. M. & Ringler, N. H. (2004). Comparison of Atlantic salmon embryo and frystocking in the Salmon River, New York. North American Journal ofFisheriesManagement 24, 1385-1397.

Cowx, I. G. (1994). Stocking strategies. Fisheries Management and Ecology 1, 15-30.

Crawford, S. S. (2001). Salmonine introductions to the Laurentian Great Lakes: anhistorical review and evaluation ofecological effects. Canadian Special Publication ofFisheries and Aquatic Science. No. 132.

Cross, T. F., McGinnity, P., Coughlan, W., Dillane, E., Ferguson, A., Koljonen, M.-L.,O'Reilly, P. & Vasmagi, A. (2007). Stocking and Ranching. In The Atlantic salmon;Genetics Conservation and Management (Verspoor, E., Stradmeyer, L. & Nielson, J.,eds.), pp. 325-356. Ames Iowa: Blackwell Publishing.

Daily, G., ed. (1997). Nature's services: Societal Dependence on Natural Ecosystems .Washington, DC: Island Press.

Dobson, A., Lodge, D., Alder, J., Cumming, G. S., Keymer, J., McGlade, J., Mooney, H.,Rusak, J. A., Sala, O., Wolters, V., Wall, D., Winfree, R. & Xenopoulos, M. A. (2006).Habitat loss, trophic collapse, and the decline of ecosystem services. Ecology 87, 1915-1924.

Dunfield, R.W. 1985. The Atlantic salmon in the history of North America. Canadianspecial publication of fisheries and aquatic science. 80: 181 p.

Dymond, J. R. (1966). The Lake Ontario Atlantic salmon {Salmo salar). UnpublishedManuscript, p. 248.

Eckert, T. H. (2003). Lake Ontario Fishing Boat Census. In 2002 Annual Report: NewYork State Department of Environmental Conservation Bureau of Fisheries Lake OntarioUnit and St. Lawrence River Unit to the Great Lakes Fishery Commission's LakeOntario Committee.

Fischer, J. & Lindenmayer, D. B. (2000). An assessment of the published results ofanimal relocations. Biological Conservation 96, 1-11.

Gephard, S. (2008). Restoring Atlantic Salmon {Salmo salar) to New England. In SavingBiological Diversity (Askins, R. A., G.D., D., Visgilio, G. R. & Whitelaw, D. M., eds.),pp. 75-85. Boston: Springer Science & Business Media.

Griffith, B., Scott, J. M., Carpenter, J. W. & Reed, C. (1989). Translocation as a speciesconservation tool - status and strategy Science 245, 477-480.

17

Hansen, L. P. (1987). Growth, migration and survival of lake reared juvenile anadromousAtlantic salmon Salmo salar L. Fauna Norvegica, Series A. 8, 29-34.

Hansen, L. P. & Jonsson, B. (1991). The effect of timing of Atlantic salmon smolt andpost-smolt release on the distribution of adult return. Aquaculture 98, 61-67.

Hooper, D. U., Chapin, F. S., Ewel, J. J., Hector, A., Inchausti, P., Lavorel, S., Lawton, J.H., Lodge, D. M., Loreau, M., Naeem, S., Schmid, B., Setala, H., Symstad, A. J.,Vandermeer, J. & Wardle, D. A. (2005). Effects of biodiversity on ecosystemfunctioning: A consensus of current knowledge. Ecological Monographs 75, 3-35.

Huntsman, A. G. (1944). Why did Lake Ontario salmon disappear? Transactions oftheRoyal Society ofCanada 5, 83-102.

Johnson, J. H. (2004). Comparative survival and growth of Atlantic salmon from eggstocking and fry releases. North American Journal ofFisheries Management 24, 1409-1412.

Jokikokko, E. & Jutila, E. (2004). Divergence in smolt production from the stocking of 1 -summer-old and 1 -year-old Atlantic salmon parr in a northern Baltic river. Journal ofApplied Ichthyology 20, 511-516.

Jokikokko, E., Kallio-Nyberg, I., Saloniemi, I. & Jutila, E. (2006). The survival of semi-wild, wild and hatchery-reared Atlantic salmon smolts of the Simojoki River in the BalticSea. Journal ofFish Biology 68, 430-442.

Jokikokko, E. & Mantyniemi, S. (2003). The survival of stocked Atlantic salmon smoltsduring sea run and the timing of migration in the river Simojoki, northern Finland.Aquaculture 219, 431-444.

Kallio-Nyberg, I., Jutila, E., Saloniemi, I. & Jokikokko, E. (2004). Association betweenenvironmental factors, smolt size and the survival of wild and reared Atlantic salmonfrom the Simojoki River in the Baltic Sea. Journal ofFish Biology 65, 122-134.

Kennedy, G. A. (1988). Stock enhancement of Atlantic salmon. In Atlantic Salmon:Planningfor the Future (Mills, D. & Piggins, D., eds.). Portland, Oregon: Timber Press.

Ketola, H. G., Bowser, P. R., Wooster, G. A., Wedge, L. R. & Hurst, S. S. (2000). Effectsof thiamine on reproduction of Atlantic salmon and a new hypothesis for their extirpationin Lake Ontario. Transactions ofthe American Fisheries Society 129, 607-612.

Kleiman, D. G. (1989). Reintroduction of captive animals for conservation. Bioscience39, 152-161.

18

Klemetsen, ?., Amundsen, P. ?., Dempson, J. B., Jonsson, B., Jonsson, N., O'Connell,M. F. & Mortensen, E. (2003). Atlantic salmon Salmo salar L. , brown trout Salmo truttaL. and Arctic charr Salvelinus alpinus (L.): a review of aspects of their life histories.Ecology ofFreshwater Fish 12, 1-59.

Lacy RC (1997) Importance of genetic variation to the viability of mammalianpopulations. Journal of Mammalogy 78, 320-335

Letcher, B. H., Dubreuil, T., O'Donnell, M. J., Obedzinski, M., Griswold, K. & Nislow,K. H. (2004). Long-term consequences of variation in timing and manner of fryintroduction on juvenile Atlantic salmon {Salmo salar) growth, survival, and life-historyexpression, pp. 2288-2301.

Letcher, B. H. & Terrick, T. D. (2001). Effects of developmental stage at stocking ongrowth and survival of Atlantic salmon fry. North American Journal ofFisheriesManagement 21, 102-110.

Locke, A. (1998). Modeling the effects of post-stocking survival rates on the success ofstocking hatchery Atlantic salmon in a New Brunswick River. North American Journal ofFisheries Management, 547-560.

Lyles, A. M. & May, R. M. (1987). Conservation biology - problems in leaving the ark.Nature 326, 245-246.

MacCrimmon, H. (1965). The beginning of salmon culture in Canada. CanadianGeographical Journal 9, 4-1 1 .

MacCrimmon, H. & Gots, B. (1979). World distribution of Atlantic salmon, Salmo salar.Journal ofthe Fisheries Research Board ofCanada, 422-457.

MacCrimmon, H. R. (1954). Stream studies on planted Atlantic salmon. Journal oftheFisheries Research Board ofCanada 11, 362-403.

Meffe GK (1995) Genetic and ecological guidelines for species reintroduction programs:application to Great Lakes fishes. Journal ofGreat Lakes Research 21(Suppl. 1), 3-9

Mills, D. (1989). The Ecology and Management ofAtlantic Salmon. London: Chapman &Hall.

Mittelbach, G. G., Turner, A. M., Hall, D. J., Rettig, J. E. & Osenberg, C. W. (1995).Perturbation and resilience - a long-term, whole lake study of predator extinction andreintroduction Ecology 76, 2347-2360.

19

Nislow, K. H., Einum, S. & FoIt, C. L. (2004). Testing predictions of the critical periodfor survival concept using experiments with stocked Atlantic salmon, pp. 188-200.

Noble, B. F. (2004). Applying adaptive environmental management. In Resource andEnvironmental Management in Canada (Mitchell, B., ed.), pp. 442-466. Toronto: OxfordUniversity Press.

Parrish, D. L., Behnke, R. J., Gephard, S. R., McCormick, S. D. & Reeves, G. H. (1998).Why aren't there more Atlantic salmon {Salmo salar)! Canadian Journal ofFisheriesand Aquatic Sciences 55, 281-287.

Parsons, J. W. (1978). History of salmon in the Great Lakes, 1850-1970. In U.S. BureauofSport Fish and Wildlife Technical Paper.

Pennell, W. & Barton, B., eds. (1996). Principles ofSalmonid Culture. Amsterdam, TheNetherlands: Elsevier Science.

Pimm, S. L., Russell, G. J., Gittleman, J. L. & Brooks, T. M. (1995). The future ofbiodiversity. Science 269, 347-350.

Reisenbichler RR, Rubin SP (1999) Genetic changes from artificial propagation ofPacific salmon affect the productivity and viability of supplemented populations. ICESJournal of Marine Science, 56, 459^466.

Rooney, N., McCann, K., Gellner, G. & Moore, J. C. (2006). Structural asymmetry andthe stability of diverse food webs. Nature 442, 265-269.

Salminen, M., Alapassi, T. & Ikonen, E. (2007). The importance of stocking age in theenhancement of River Kymijoki salmon {Salmo salar). Journal ofApplied Ichthyology23, 46-52.

Sarrazin, F. & Barbault, R. (1996). Reintroduction: challenges and lessons for basicecology. Trends in Ecology & Evolution 11, 474-478.

Seddon, P. J. (1999). Persistence without intervention: assessing success in wildlifereintroductions. Trends in Ecology & Evolution 14, 503-503.

Seddon, P. J., Armstrong, D. P. & Maloney, R. F. (2007). Developing the science ofreintroduction biology. Conservation Biology 21, 303-312.

Seddon, P. J., Soorae, P. S. & Launay, F. (2005). Taxonomic bias in reintroductionprojects. Animal Conservation 8, 51-58.

20

Simberloff D (1997) Flagships, umbrellas, and keystones: Is single-species managementpassé in the landscape era? Biological Conservation 83: 247-257

Smith, S. H. 1970. Species interactions of the alewife in the Great Lakes. Transactions ofthe American Fisheries Society 99:754 -765.

Stanfield, L. & Jones, M. L. (2003). Factors influencing rearing success of Atlanticsalmon stocked as fry and parr in Lake Ontario tributaries. North American Journal ofFisheries Management 23, 1 1 75-1 1 83.

Tilman, D. (2000). Causes, consequences and ethics of biodiversity. Nature 405, 208-211.

Vitousek, P. M., Mooney, H. A., Lubchenco, J. & Melillo, J. M. (1997). Humandomination of Earth's ecosystems. Science 211, 494-499.

Webb, J., Verspoor, E., Aubin-Horth, N., Romaakkaniemi, A. & Amiro, P. (2007). TheAtlantic Salmon Genome. In The Atlantic salmon; Genetics Conservation andManagement (Verspoor, E., Stradmeyer, L. & Nielson, J., eds.), pp. 17-45. Ames, Iowa:Blackwell Publishing.

Webster, D. A. (1982). Early History of Atlantic salmon in New York. New York Fishand Game Journal 29, 26-44.

Wilson, A.C. and Stanley Price, M.R. (1994) Reintroduction as a reason for captivebreeding, in Creative Conservation: interactive Management of Wild and CaptiveAnimals (Olney, P.J. S., Mace, G.M. and Feistner, A.T.C., eds), pp. 243-264, Chapmanand Hall.

21

Chapter 2: Evaluating the in-stream survival and growth of

juvenile life stages of Atlantic salmon stocked into Cobourg

Brook, Ontario.

22

Abstract

A key question in reintroduction efforts is what life stages are best suited for

releasing into the wild in efforts to re-establishing wild populations? In the current effort

to restore Atlantic salmon in Lake Ontario, multiple juvenile life stages have been

released into selected Lake Ontario tributaries. This study used the paired release of

genetically tagged Atlantic salmon in conjunction with seasonal (spring and fall) field

surveys to evaluate the comparative growth and survival of fry and yearling Atlantic

salmon released into Cobourg Brook, a north-shore Lake Ontario tributary. In addition,

fall fmgerlings and age-2 Atlantic salmon were introduced into the study area and some

insight into their survival and age at out-migration was provided. Atlantic salmon stocked

as fry were consistently large and at high densities at the end of the first growing season.

Compared to stocked yearlings; however, relatively few stocked fry were captured as out-

migrants, suggesting that severe over-winter mortality occurred in Cobourg Brook. The

survival of stocked fry until out-migration was significantly less than 9%. The body

condition (relative mass) of stocked yearlings from spring to fall was poor, while no

difference was observed in the condition of stocked fry. The observed differences in

abundance and growth of stocked fry and yearlings may be attributable to the abiotic

conditions within Cobourg Brook. Significant declines from spring 2007 to spring 2008

in the out-migrant production of rainbow trout and brown trout provides further evidence

that severe mortality occurred during the winter of 2007/2008 in Cobourg Brook. If

severe over winter mortality is a common occurrence in Cobourg Brook, then the

likelihood of re-establishing an Atlantic salmon population in this tributary appears low.

The body size, position in the watershed, and density of stocked Atlantic salmon in

23

Cobourg Brook provided additional information regarding the performance of Atlantic

salmon stocked into Cobourg Brook.

24

Introduction

Reintroduction biology has emerged as an active field within conservation

biology that focuses on improving the success rate of native species reintroduction

programs. In addition to applied restoration efforts, reintroduction biology also

contributes to other established fields such as population biology, genetics, behavioral

ecology, and evolution (Armstrong and Seddon, 2008; Seddon et al, 2007; Sarrazin and

Barbault, 1 996; ). One current direction in reintroduction biology is to employ large-

scale a priori hypothetico-deductive experiments to address key questions in both basic

and applied ecology (Armstrong and Seddon, 2008; Seddon et al, 2007; Sarrazin and

Barbault, 1996; Armstrong et al, 1994). One key question in reintroduction biology is: Is

the likelihood of population re-establishment affected by the composition (e.g. life

history stage) of the release group (Armstrong and Seddon 2008; Robert et al,. 2003;Sarrazin and Legendre 2000)?

Populations of Pacific and Atlantic salmon (Oncorhynchus and Salmo spp.) have

been lost throughout much of their range and considerable effort has been allocated to

their restoration (Augerot, 2005; Parrish et al, 1998). Stocking streams with hatchery-

reared juveniles is a recommended approach when natural recolonization is not possible

(Webb et al, 2007). However, hatchery-reared salmon typically display poor survival in

the wild and their role in restoration efforts has been extensively debated (Myers, 2004;

Fleming and Petersson, 2001 ; Ford, 2002; Ryman and Utter, 1987). The adverse effects

of captive rearing may be offset by identifying an effective life history stage for releasinghatchery-reared salmon into the wild.

25

Maximum survival of stocked juveniles may be achieved by avoiding high-

mortality life stages in the wild and mitigating maladaptive behavioral traits acquired

during captive rearing (Coghlan and Ringler, 2004). The juvenile life history of wild

Atlantic salmon is marked with many high mortality events, particularly during gravel

emergence (MacKenzie and Moring, 1988), transfer from endogenous to exogenous

feeding (Balon, 1985), over-wintering (Cunjak et al, 1998), and smolt migration

(Hvidsten and Lund, 1988; Thorpe, 1988). Captive rearing relaxes selective pressures and

improves survival through early life stages which typically show high mortality (Araki et

al, 2008). However, through social learning many critical behaviors such as foraging and

predator avoidance can be adapted to the unnatural conditions within hatchery facilities

after just a short duration in captivity, resulting in high mortality after transfer into the

wild (Brown and Day, 2002). Ideally, captive-reared juveniles would be released after

periods of high mortality in the wild but early enough to mitigate maladaptive behaviors

learned in the hatchery environment.

Lake Ontario historically hosted the largest freshwater population of Atlantic

salmon (Parsons, 1978), which was extirpated in the late 19th century due to direct andindirect anthropogenic pressures (Parsons, 1978; Dunfield, 1985 Ketola et al, 2000).

Conditions within Lake Ontario have improved since the early 1900's, and the watershed

currently hosts populations of many native fishes (Christie, 1972; Mills et al, 2003), as

well as a variety of exotic salmonid species which support a large sport fishery

(Crawford, 2001).

Environmental improvement throughout the Lake Ontario watershed sparked an

interest to reintroduce Atlantic salmon by the Ontario Ministry of Natural Resources

26

(OMNR) in 1 987. Although ultimately unsuccessful, results from the effort suggested

restoration was possible (Bisset et al, 1995). Conditions in many streams along the north

shore of Lake Ontario were found to be suitable for juvenile Atlantic salmon survival up

until the end of their first growing season (Jones and Stanfield, 1993; Stanfield and Jones,

2003). In 2006 a collaborative partnership was established between academic institutions,

non-government agencies, and several levels of government to restore self-sustaining

populations of Atlantic salmon in Lake Ontario within the next ten to fifteen years

(OMNR, 2007).

Complying with the general principles of reintroduction biology, the partnership

has employed a landscape-scale approach to determine an effective stocking strategy.

Atlantic salmon brood stocks from Sebago Lake (Maine), Lac St. Jean (Quebec), and

LaHave River (Nova Scotia) have been acquired for reintroduction purposes (OMNR,

2009). To evaluate the performance of the three source populations, juveniles from each

stock are planned to be introduced into multiple Lake Ontario tributaries which provide

the best chance for Atlantic salmon restoration (Cross et al, 2007; Grieg et al, 2003).

Prior to the availability ofjuveniles from all three populations in 2012, determining an

effective life history stage for stocking has been identified as a research priority (Grieg et

al, 2003).

I used a watershed-scale experiment to test the comparative survival and growth

of multiple Atlantic salmon life stages reintroduced into their native habitat. Specifically,

I used the paired release of genetically tagged fry and yearlings in conjunction with

seasonal (spring and fall) field surveys to determine the abundance of each release group,

their relative and absolute contributions to out-migrants, and to investigate if the survival

27

of stocked fry is 9% from stocking until outmigration. The temperature and flow regime

of Cobourg Brook, as well as the resident populations of other salmonid species were

monitored to investigate the influences on Atlantic salmon growth and survival.

Unmarked Atlantic salmon were also introduced into the study site, and preliminary

information was provided on the survival and age at out-migration of Atlantic salmon

stocked as fall fingerlings or sub-adults (age-2). Stocking juvenile life stages that

improve the number and size of out-migrants may increase the abundance of returning

adults (Jonsson et al, 2008; Behmer et al,. 1993), therefore increasing the likelihood of

population re-establishment.

28

Methods

Study Site

Cobourg Brook (Figure 2-1) is one of three north shore Lake Ontario watersheds

selected for Atlantic salmon reintroduction within the current restoration program

(OMNR, 2007). Located on the northeastern shoreline of Lake Ontario, the watershed is

comprised primarily of forest and agriculture land cover, with a relatively small

catchment area of 123 km2. This study focused on the main stem of Cobourg Brook as

well as its major tributary Baltimore Creek, and two dams prevented upstream movement

of fishes in the study area (Figure 2-1). A variety of salmonine species inhabit Cobourg

Brook including native brook trout {Salvelinus fontinalis), as well as exotic brown trout

{Salmo trutta), Chinook salmon {Oncorhynchus tshawytscha), coho salmon {O. kisutsh)

and rainbow trout {O. mykiss), although Chinook and coho salmon are confined to areas

downstream of the lower-most dam (Figure 2-1). Prior to the beginning of this study in

2006, Atlantic salmon were last stocked into Cobourg Brook in 2002 (OMNR, 2002-

2007).

29

Cobourg Brook Baltimore Creek

Fry stocking

Smolttrapling stocking

^rLevel andtemperaturedata loggers

0 1 2 3 4 5KmI I I I I I

Ontario, Canada

Lake Ontario

New York, USA

Figure 2-1: Map of Cobourg Brook and its major tributary Baltimore Creek, with the

associated location of the smolt trap, stocking areas, data loggers, fall sampling sites (·),

and dams (—).The inset map shows all three north shore Lake Ontario watersheds that

are presently being stocked with Atlantic salmon as part of the current restoration

program, with an arrow indicating Cobourg Brook.

30

Experimental Design

Since 2006, Cobourg Brook has been stocked with a variety of Atlantic salmon

life history stages, which have been reared at several hatchery facilities as part of the

current restoration effort (Table 2-1). The original intent of this study was to test the

comparative growth and survival of three stocked life stages: fry, fall fingerlings, and

yearlings. Life stages were planned to be marked using genetic techniques, stocked into

the wild, and recaptured during spring and fall field surveys. However, a catastrophic

mass mortality of genetically tagged juveniles occurred at one rearing facility, which

necessitated substituting unmarked fall fingerlings and yearlings from a separate captive

population (Sebago Lake, Maine) for the design shortfall (Figure 2-2, Table 2-1).

Resolving the life stage at stocking of unmarked Atlantic salmon was attempted using

scale pattern analysis, but was not successful (Appendix 1). Scale pattern analysis was

abandoned, and only genetic tags were used to resolve the identity of Atlantic salmon in

this study (Appendix 1). As a result, the study objective was modified to evaluate the

comparative growth and survival of only stocked fry and yearlings.

The comparative survival of stocked fry and yearling Atlantic salmon in Cobourg

Brook was evaluated through a paired release-recapture experiment (Skalski et al, 2009).

Fry and yearlings were released in numbers to produce an equal amount of out-migrants

(e.g. smolts), assuming that all stocked individuals out-migrate from Cobourg Brook at

age-1, and the survival of stocked fry and yearlings until out-migration into Lake Ontario

was 9% and 100%, respectively (Table 2-1). Genetic tagging was employed to mark

individuals stocked at each life stage, and the relative abundance of each stocking

treatment (fry vs. yearling) was compared during smolt trapping in the spring of 2008. A

31

watershed-scale backpack electrofishing survey was conducted the following fall to

evaluate the survival, growth, and location of stocked life stages in the Cobourg Brook

watershed, and to test the assumption that 100% of stocked individuals out-migrated at

age-1. Therefore, according to the design of the paired-release experiment, if an equal

number of each stocking treatment were captured during spring 2008 smolt trapping and

zero were captured the following fall, than the survival of stocked fry would be

approximately 9% from stocking until out-migration.

32

Table 2-1: Summary of Atlantic salmon stocked into Cobourg Brook throughout the

duration of this study.

Date Stocked ParentStock

Life Stage Hatchery Amount Size (g) Mark

May 2006October 2006

April 24 2007April 24 2007

April 24 2007April 18 2007April 20 2007*May 25 2007May 25 2007October 312007*April 17 2008April 16 2008May 2008

LaHaveLaHave

LaHaveLaHave

LaHaveLaHaveLaHaveLaHaveLaHave

Sebago

LaHave

SebagoLaHave

May 26 2008 LaHave

Fry Ringwood 60 556 1.3Fall Normandale 25 005 10.1fingerlingSub-adults Codrington 863 51.9Yearling Normandale 2 882 24.3

Yearling Normandale 3 001 24.3Yearling Normandale 5 891 26.1Yearling Fleming 2 096 72.2Fry Normandale 44 972 1.3Fry Fleming 18 100 0.2Fall Normandale 15 441 14.3fingerlingYearling Fleming 5 449 1 54.6Yearling Normandale 6 130 46.5Sub-adults Harwood 238 51.9

Fry Normandale 46 608 0.95

NoneNone

PIT tagFloy tag and finclipFin clipNoneNoneGeneticGeneticNone

GeneticNone

PIT or fin clip

Genetic

* Indicates stocking treatments used to evaluate comparative growth and survival

33

Fry -

Fingerlings —

Yearlings ¦

Sub-Adults —

#??

CZl

^f ?F^f <^?& ?<$?\<&°f

ézsj

CD

CD

CD

"y

0>

CD

CjT

^

^

C=I

^

Figure 2-2: A timeline of Atlantic salmon stocking and capture events that occurred

throughout the duration of this study. Rectangular icons represent stocking events and

correspond to life stages on the y-axis, and dashed lines represent capture events. Shaded

icons represent genetically marked stocking treatments.

34

Beginning in 2006, individuals were genetically tagged using a structured captive

breeding program. All reproductive adults from each brood stock (LaHave River or

Sebago Lake) were individually marked using PIT tags, non-lethally sampled, and their

microsatellite genotypes were determined. Adults were mated using single pair one-

female to one-male crosses, and the resultant progeny were reared communally and

tracked in captivity until stocking at a common life stage. Upon capturing an Atlantic

salmon in Cobourg Brook, a tissue sample was collected to determine its microsatellite

genotype, and genetic parentage analysis was used to determine when that individual was

initially stocked. Both of the stocking treatments used in the paired release-recapture

experiment were progeny of the LaHave River brood stock. Juveniles stocked as fry were

reared at ambient groundwater temperatures until transfer to exogenous feeding when

heated water was supplemented for a brief period and individuals were fed pelleted fry

food. Upon reaching approximately 1 gram in body weight fry were scatter-stocked in a

high gradient portion of Baltimore Creek at a target density of 8 fish/m (Figure 2-1).

Yearlings were reared similarly to fry, although a water recirculation system and

automated pellet feeders were employed. Yearlings were visually confirmed to have

attained smolt stage prior to stocking (lacking parr marks, silver colouration), and were

stocked at low gradient locations along the main stem of Cobourg Brook on April 17

2008 (Figure 2-1).

To make up for the catastrophic loss of genetically marked fall fingerlings and

yearlings, unmarked juveniles from the Sebago Lake captive population were introduced

into Cobourg Brook (Table 2-1). The restoration program acquired the first juvenile year

class of Atlantic salmon from Sebago Lake in 2006, which was produced from 60 wild

35

families. All juveniles were thereafter combined and reared at Normandale FCS using

heated water only to assist transfer to exogenous feeding at the fry stage (G, Durant pers.

comm.). Fall fingerlings were introduced into Cobourg Book in October 2007 and

yearlings in April 2008 (Table 2-1). Although these individuals can be differentiated

from progeny of the LaHave River captive population using genetic population

assignment tests, fall fingerlings and yearlings unfortunately could not be identified to the

life stage they were stocked at using parentage analysis due to the pooling of families

during captive rearing.

A number of other Atlantic salmon life stages were also stocked into Cobourg

Brook during this study (Table 2-1, Figure 2-2). Fry were stocked in 2008 which were

genetically tagged, and rearing and stocking methods followed the same procedures as

fry stocked in 2007, which therefore provided a second stocking treatment of genetically

tagged fry to evaluate the annual variability in survival. Yearlings stocked in 2007 were

reared at Normandale FCS using ambient groundwater temperatures with heated water

supplemented only to aid transfer to exogenous feeding (P. Malcolmson, OMNR, pers.

comm.). Approximately half of all yearlings stocked into Cobourg Brook were marked

with 90mm floy tags, and/or adipose clips on April 4, 2007 and eventually released into

Cobourg Brook three weeks later (Table 2-1). Surplus brood stock was also stocked into

Cobourg Brook at age-2 as sub-adults (Table 2-1). In 2007 all sub-adults stocked into

Cobourg Brook were marked with PIT tags, and in 2008 approximately half were PIT

tagged, and the remaining half were marked with 6mm hole punches in the caudal fin.

Aside from genetically tagged stocking treatments, the remaining Atlantic salmon

life stages released into Cobourg Brook provide little useful information in the context of

36

this study and are hereafter mentioned only briefly. A timeline of stocking and capture

events is provided in Figure 2-2.

Environmental Data:

Stream temperature and level data for Cobourg Brook was provided from the

Department of Fisheries and Oceans Canada, who used automated thermometers, and

level loggers to collect observations every 30 minutes or 4 hours from the lower reaches

of Cobourg Brook (Figure 2-1). Data loggers were attached to a sea-lamprey barrier near

the streambed throughout the duration of the study and data was downloaded

intermittently.

Field Surveys

Field surveys were carried out in spring and fall of 2007 and 2008. Only the 2008

field surveys were able to capture fry and yearling Atlantic salmon used in the paired

release-recapture experiment (Figure 2-2). However, surveys carried out in 2007

provided an opportunity to develop survey methodology and gather additional

information on the newly re-introduced juvenile Atlantic salmon population in Cobourg

Brook.

Spring smolt trapping

Juvenile salmonids out-migrating from Cobourg Brook were captured each spring

using a partial capture weir, hereafter referred to as a smolt trap. The smolt trap was

located in the lower sections of Cobourg Brook below major stocking and nursery areas

37

(Figure 2-1), and consisted of a fyke net with two 1 5 ft wings and 13 mm mesh

(stretched). A fabricated live box (Milner and Smith, 1985) was attached to the cod end

of the smolt trap to minimize stress. The smolt trap operated continuously and was

checked daily from April 24 to July 20 2007, and from April 16 to August 13 2008. The

smolt trap was not operational during processing time each morning and during

inoperable conditions due to heavy rainfall events or in-stream debris. To ensure that

sampling effort overlapped with the beginning of smolt out-migration in 2008, pre-season

test netting was conducted for 24-hour periods occurring March 26-27, March 30-April 1,

April 4-5, and April 9-10.

Each day the smolt trap was inspected and all captured fish were counted. A

random sample of up to 100 Atlantic salmon were measured for total length (± 1mm) and

mass (± O.lg), and scales and tissue samples were collected for aging and genetic

analyses. Scales were collected from an area between the posterior margin of the dorsal

fin and the anterior margin of the anal fin above the lateral line, and tissue samples were

collected from the from the caudal fin using a 3mm hole punch. A substantial amount of

dead Atlantic salmon were encountered in the smolt trap prior to handling during the

early stages of the 2008 survey. All dead salmon were counted, measured for length and

mass, and scales and tissue samples were collected. Each day, a random sample of up to

100 rainbow trout and brown trout were measured for total length (± 1mm) and mass (±

O.lg).

Atlantic salmon, rainbow trout, and brown trout that were captured in the smolt

trap were also used for mark-recapture population estimation. Each day up to 50

previously unmarked individuals of each species were randomly selected and tagged

38

using week-specific colour and location combinations of visual implant elastomer (VIE;

Northwest Science and Technology). A 3mm hole punch on the lower caudal fin served

as a tissue sample for genetic analysis and also as a second mark to evaluate VIE tag loss.

Marked individuals were held in-stream until dusk and released upstream of the smolt

trap using an automated device (Miller et al., 2000). Recaptured fish, and fish not

subjected to mark-recapture methodology were released downstream of the smolt trap. To

evaluate the possibility of upstream movement and re-entrainment in the smolt trap, fish

released downstream of the trapping location were marked with hole punch in the upper

caudal fin. Handing mortality and tag loss was assessed by regularly inspecting marked

individuals in the release device prior to release. Methodology was consistent among

field seasons, however in 2007 the automated release device and concurrent assessment

of marking mortality and tag loss was not available, and a smaller fyke net was

substituted during the initial two weeks of the survey.

The origin of Atlantic salmon captured in the smolt trap was resolved using clips,

tags, and genetic markers. In 2007, all individuals captured in the smolt trap were

inspected for adipose clips and/or tags to determine the abundance of marked yearlings.

In 2008, the origin of sampled Atlantic salmon was estimated using genetic parentage

analysis, genetic population assignment, or age interpretation via scale inspection. Due to

the high abundance ofjuvenile Atlantic salmon captured in 2008 sub-sampling was

employed. The origin composition of the total daily catch was estimated by multiplying

the frequency of each stocking treatment in the sub-sample to the total daily catch.

39

Fall electrofíshing

Electrofishing surveys were conducted in the fall of 2007 and 2008. Cobourg

Brook was partitioned into three valley segments which were defined by gradient breaks

or tributary confluences and six sampling sites were randomly selected within each valley

segment (Figure 2-1). Each sampling site was approximately 50±10m in length and

bounded by stream crossovers (where the stream thalweg crosses the middle of the

channel). The fish community was sampled using single-pass backpack electrofishing

with 9.5mm mesh (stretched) seine nets secured at the top of each site. The surface area

of each sampling site was measured prior to sampling. Efforts were made to apply an

equal amount of effort (shocking seconds) throughout sampled sites, and also to maintain

equal effort between sites. Each year one site within each valley segment was sampled

using multiple-pass electrofishing to estimate the total abundance of salmonid species

within each site, and to evaluate the efficiency of single-pass electrofishing at capturing

salmonid species. During multiple pass electrofishing block nets were secured at each site

boundary. Sampling occurred at approximately the same sites in both 2007 and 2008. Site

boundaries were re-selected based on stream morphology (e.g. location of thalweg) with

the exception of two sites that could not be sampled in 2008 due to newly deposited large

woody debris. In both cases, alternate sites in close proximity were sampled. Multiple-

pass sites were re-selected within each valley segment annually. Upon capture all fish

were identified to species, counted, and weighed. Caudal tissue and/or scales were

collected in addition to length and mass data from all captured Atlantic salmon. Length

and mass data were also collected from a sub-sample of each age group from other

salmonid species. Individuals were assigned to age groups by examining of the length

40

distribution of captured species in the field. Group-0 referred to individuals that were

age-0, and group- 1 referred to individuals older than age-0. Generally, individuals less

than 120 mm TL were considered as group-0, while individuals larger than 120mm TL

were considered as group- 1 .

Genetic Analyses

DNA was extracted from tissue samples using simple lysis extraction in 96 well

plates using 250 µL lysis buffer (50 mM Tris pH 8, 1000 mM NaCl, 1 mM EDTA, 1%

sodium dodecyl sulphate weight per volume, and 200 µg proteinase K) per well and

incubated for 16 hours at 37°C. DNA was precipitated using 500 µ? of 80% isopropanol

per well and centrifuged at 2000 gravities for 30 minutes. The supernatant was removed

and the remaining pellets were rinsed with ImL of 70% ethanol, followed by re-

centrifugation. DNA pellets were air dried at room temperature for 20 minutes, then

dissolved in 150ul Ix TE (1OmM Tris, 1 mM EDTA). Extraction yields and quantity

were tested using electrophoresis in 1.5% agarose gel alongside a mass ladder (Bioshop,

Burlington, Ontario), and stained with Sybr Green dye (Cedar Lane Laboratories,

Burlington, Ontario).

Samples were amplified at 9 microsatellite loci: Ssa\91, Ssa202, SsaS5 (O'Reilly

et al, 1996); and SSspl605, SSsp220\, SSsp22l3, SSsp22\5, SSsp22\6, SSspG7

(Paterson et al., 2004). Multiplex PCR amplifications were performed in 1OuI reactions

containing the following; approximately 12ng of genomic DNA; Ix PCR buffer

containing 1.5mM MgCl2 (Qiagen, Mississauga, Ontario), 2mM each dNTP (Bioshop,

Burlington, Ontario),0.08-0^M primers, and 0.2mg/ml BSA (Bioshop, Burlington,

41

Ontario), 0.025U Taq DNA polymerase (Qiagen, Mississauga, Ontario). Multiplex

reactions contained the following primer concentrations: Multiplex 1 SSsp\6Q5 0.2 µ?,

SSsp220l 0.3 µ?, ÄSsp2215 0.08µ?, SsaSS 0.2µ?; Multiplex 2 -1 Ssa97 0.1 µ?, Ssa202

0.5µ?, Multiplex 2 -2 SSspGl 0.3µ?, Multiplex 3-1 SSsp22l3 0.2µ?, Multiplex 3-2

SSsp22\6 0.2µ?. PCR cycling was 1 minute activation at 95°C followed by 35 cycles of45 seconds denaturing at 95°C, 1 minute annealing and 1 minute extension at 72°C,concluded by a final extension of 10 minutes at 72°C. Annealing temperatures for eachreaction were as follows: Multiplex 1- 62°C Multiplex 2 -1 - 58°C, Multiplex 2 -2 65°C,Multiplex 3-1 62°C, Multiplex 3-2 62°C. Amplified products for multiplex 2-1 and 2-2were pooled together at a ratio of 2:3 before visualization. Amplified products for

multiplex 3-1 and 3-2 were pooled together at a ratio of 2:1 . Positive controls were

included on all plates for quality assurance, and to ensure scoring consistency.

Amplified products for all samples were run on an AB3730 DNA analysis system,

with LIZ 500 size standard (Applied Biosystems, Foster City, California). Products were

sized using GeneMapper version 3.1 (Applied Biosystems, Foster City, California), and

corroborated using visual inspection.

A sample of individuals were amplified and scored a second time to assess

scoring error and mutations. Of the 94 individuals that were reamplified and scored at 9

loci, 28 alleles differed, resulting in a total error rate of 1 .6%. Errors were associated with

erroneous allele selection at SSspG7 and allelic dropout across some loci. Genotypes of

all parents and wild captured juveniles were refined by rescoring all individuals at

SSspG7, and reamplifying single loci with allelic drop out to produce a substantially less

erroneous genotypic dataset. The revised genotype dataset was used for final analyses.

42

Data analyses

Environmental conditions

The mean daily stream level and temperature and it's associated standard

deviation was calculated for all days where data was available. The duration of growing

and winter seasons was used to describe the temperature regime of Cobourg Brook. The

Atlantic salmon growing season was calculated as the number of days when stream

temperature first surpassed 50C to when it again declined below 5 0C. Similarly, the

winter season was calculated as the number of days when the stream temperature first

declined below 3°C to when it again exceeded 3°C.

Spring smolt trapping

The daily abundance ofjuvenile Atlantic salmon captured in the smolt trap was

used as a measure of relative abundance, or Catch Per Unit Effort (CPUE). A temporally

stratified mark-recapture estimator as presented by Darroch (1961), was employed to

estimate the total abundance of out-migrants passing the smolt trap location each week.

Mark-recapture was also used to investigate the probability that the smolt trap will

capture a salmonid passing the smolt trap location . Mark-recapture strata consisted of the

weekly VIE marking intervals. The estimator considers the probability of a fish being

captured as the joint probability of marked individuals resuming migration, and the

probability of migrants becoming entrained in the smolt trap. The software package

Darroch Analysis with Rank Reduction (DARR) 2.0 (Bjorkstedt, 2005) was employed for

mark-recapture analysis. DARR was also used to combine statistically similar strata to

43

alleviate issues associated low sample sizes. Weeks 12-13 in 2007were pooled a priori

due to low sample sizes. DARR also calculated weekly smolt trap efficiency estimates.

The annual variability in the survival of naturalized salmonid populations within

Cobourg Brook was investigated by investigating if differences existed in the abundance

of brown trout and rainbow trout captured in the smolt trap in 2007 compared to 2008.

Annual differences in smolt trap CPUE for both rainbow trout and brown trout were

tested using the non-parametric Mann-Whitney test (Zar, 1984).

Fall electrofishing surveys

The density offish captured (fish per square meter) using single-pass backpack

electrofishing was used as a measure of CPUE. All salmonid species other than Atlantic

salmon were combined into two age groups, either group-0 (e.g. age-0) other salmonids,

or group- 1 (e.g. >age-0) other salmonids. The classification used to differentiate age

groups in the field was corroborated using length frequency analysis (DeVries and Frie,

1996). The distribution of Atlantic salmon and other salmonid species throughout the

Cobourg Brook watershed was also evaluated. A Chi-squared test with Yates' correction

(Zar, 1984), was used to test if a difference occurred between the observed densities

(CPUE) compared to a uniform distribution. The total density of fishes within each

sampling site was estimated using multiple-pass capture data and a generalized removal

model (Otis et al., 1978; White et al, 1982), employing the computer program Microfish

3.0 (Van Deventer, 2006). The model uses a maximum likelihood estimator that does not

assume equal capture probability between passes. The efficiency of single-pass

electrofishing was evaluated by estimating the proportion of all individuals present in

44

each site that were captured using one electrofishing pass. The efficiency of single-pass

electrofishing was estimated for each salmonid species and age group. Differences in the

capture probability among age groups and years were tested using the Student's T-test.

The sizes of Atlantic salmon that were captured during backpack electrofishing

surveys were used to predict what proportion was expected to out-migrate the following

spring. Previous studies have found Atlantic salmon parr which attain a size related

developmental stage out-migrate as smolts the following spring (Metcalfe, 1 998). In

North American populations, individuals that exceed 95-1 00mm fork length by the end of

the growing season typically out-migrate the following spring (Nicieza et al, 1991 ;

Whitesel, 1993; McCormick et al, 1998; Pearlstein et al, 2007). In this study,

individuals that exceeded 100mm total length in late October were assumed to out-

migrate the following spring.

Annual differences in the survival of Atlantic salmon as well as other salmonid

species were evaluated. Differences between 2007 and 2008 fall CPUE of stocked

Atlantic salmon fry were compared to evaluate annual differences in survival from

stocking until the end of the first growing season. Data could not be transformed to fit a

normal distribution, thus the non-parametric paired Wilcoxon test was used to test for

differences. Differences between 2007 and 2008 fall CPUE of each age group of other

salmonid species were evaluated to investigate annual variation in the survival of

naturalized populations of salmonid species present in Cobourg Brook. Differences are

tested using the Students T-test.

45

Paired release-recapture experiment

The 2008 smolt trap CPUE of each stocking treatment was compared to evaluate

if yearling or fry stocking produced more out-migrants. Differences were tested using the

non-parametric paired Wilcoxon test. The 2008 electrofishing CPUE of each stocking

treatment was compared to evaluate which stocking treatment produced more age-1 parr.

Differences were tested using the non-parametric paired Wilcoxon test (Zar, 1984).

The relative weight (weight-at-length) as well as the total length of stocked fry

and yearlings was evaluated to compare their size and growth. The change in relative

weight from fall 2008 to spring 2008 for each stocking treatment was tested using an

ANCOVA as described by Pope and Kruse (2007). Differences in the total length

between stocking treatments were tested using a Student's T-test with unequal variances.

Origin identification

Genetic parentage analysis was used to determine when genetically tagged

Atlantic salmon were initially stocked. Exclusion-based parentage analysis was run on

all tissue samples collected in the wild using the software package Family Analysis

Program (FAP) (Taggart, 2007). All parents of genetically tagged salmon stocked into

Cobourg Brook were known. Their mating history and genotypes were analyzed to

predict all possible offspring genotypes. Atlantic salmon captured in Cobourg Brook

whose genotype did not match the predicted offspring genotypes were excluded from

further analysis., and the likelihood of assigning a wild-captured salmon to only one

family was determined. Only individuals with at least 98% likelihood were considered

reliably identified to a stocking treatment.

46

The source population (LaHave River or Sebago Lake) of wild captured salmonwas determined using genetic population assignment. Wild captured individuals were

assigned to source populations using a Bayesian exclusion method (Raímala and

Mountain, 1997), as implemented in the software package Geneclass2 (Piry et al, 2004).Individuals were considered reliably assigned to a source population if probability ofcorrect assignment exceeded 95%.

Age interpretation was also used to further resolve the origin of wild captured

salmon not confidently assigned to a stocking treatment using parentage analysis. Scaleswere rolled onto acetate and microscopically inspected using the methodology outlined inCasselman and Scott (2000). Genetically-marked life stages were age-1 during fall andspring 2008 sampling. Therefore, age-2 salmon identified by scale inspection could beexcluded as potentially unidentified stocking treatments. Scales from all individuals

captured during spring or fall 2008 that were assigned to the LaHave River captivepopulation using genetic population assignment, but not to a stocking treatment usingparentage assignment were inspected to determine their age to eliminate the possibility ofunidentified stocking treatments.

47

Results

Environmental conditions

The Atlantic salmon growing season in Cobourg Brook was 234 days long in

2006, 216 days long in 2007, and 21 1 days long in 2008. The winter period was 81 days

long in 2006/2007, and 139 days long in 2007/2008. Mean daily stream temperaturesurpassed 25°C in 2007, but not 2006 or 2008. Summer stream levels were more

dynamic in 2008 compared to 2007, and brief periods of high stream levels were

common during all observed winters (Figure 2-3).

48

• TemperatureLevel

Oo

?

05s.

0)Q.E?f-

2006 May 2007 May SepDate

May

1.5

1.0?>

l· 0.5

0.0

Sep 2009

Figure 2-3: An illustration of the stream level and temperature regime of Cobourg Brook

throughout the duration of this study. Data was collected approximately 1 km upstream

from Lake Ontario. All data points are daily means and standard deviation (error bars) ofmean daily water temperature was included.

49

2007 smolt trapping

The smolt trap was operational every day from April 24 to July 21 2007 with the

exception of May 17 and July 14-17. A total of 818 Atlantic salmon were captured, of

which 271 were marked with an adipose clip or floy tag (186 clip only, 85 floy tag and

clip), and 50 with PIT tags. The remaining 497 individuals were unmarked and therefore

potentially introduced as unmarked yearlings, fry, or fall fingerlings prior to smolt

trapping in 2007 (Figure 2-4; Table 2-1). High smolt trap catches of Atlantic salmon

occurred in three distinct periods with the first occurring in late April after the smolt trap

was first installed, the second in mid-May, and the last from mid-June to mid-July

(Figure 2-4). A total of 477 juvenile Atlantic salmon were marked with VIE, of those 100

were recaptured. Smolt trap capture probability 1 6% during weeks 1 -4, and 27% during

weeks 5-13 (Figure 2-5; Table 2-2). An estimated total of 4,155 ± 431 Atlantic salmon

passed the smolt trap location throughout the duration of 2007 sampling. VIE tag loss

was found to be less than 1% in from time of marking until recapture. Zero individuals

released downstream from the trap were recaptured. In addition to Atlantic salmon, a

total of 339 group- 1 rainbow trout and 37 group- 1 brown trout were also captured during

the 2007 smolt trapping.

50

cd-s

o

I o(O ?-

? reU= (?

Jo< joO) <¦? -o

1|ID

<

*

3

(O?

oo

oCO

O O OCM

ipjeo de¡\ }|ows áübq

C

<(L)

'S(U>

·'—i

?(D

fegO3

dûcd

?—»

O?=!klo

.&¦73?enOOh

-3cd

?(Ul-ccdS

Ch-HO

OH-»cdOOhCO

'3?r-ooCM

Í-H

3cn

?Ccö

OhOO

riO)·—SOX)

S-H

?—»

(?1)

?-»cdo'S

ofc¡cdCcd?Ccd

o-

3?—?

?Gcd

cd

cd?

cd?-?(UOhO

"?Cencd

Ohcd

? ?:e

Sen(U

O

C

cd?3

cdO

e

CO

IO

OO

cu rtíOh

OùC

J2oOen H-i

OOC

cd4)

(+HO

1000 H

g 800HEcoV)o

'•g 600co

O<- 400(D

XSEZ

200 H

i ì Catch• Estimated population size

V.* Capture Probability

Week

~<i —T-^—^r

10 11 12&13 12

0.30

0.28

0.26

0.24 !dco.OO

0.22 Ql?

0.20 o.coO

0.18

0.16

0.14

Figure 2-5: Spring 2007 weekly smolt trap catch, capture probability, and estimated

population sizes of out-migrating Atlantic salmon. Values for catch and estimated

population size are depicted on the left vertical axis, while values for capture probability

are depicted on the right vertical axis.

52

ÖO

&co

e•a?¿4OOV-C

PQVh

O-OOO

o?· ?

C

feb

3OCO

cdco

Ccd

O

cd-4—»cd?

CDVh

-4—»O*edOCDVh

I

ed

?

»s?

CU?«

H-H

OVhU

3

cd

Vhcd>oo?

Vh

OsOO

?CDVhcdSC

H-HOVhCD

ScCD

cdVh

CD

Or^HM-H

CDN

gcd

CD>OOCDVh

CDVhCD>OCDCDVh

Gcd

cd

CDCOcd

"SVh

?CDOUOOcd

OO

aOO

OVhCD

x>

iC

Id+-»O

_2"cdVhOO

_C?

CDOOcdJi13

S' ' CO

cd oco cIl C

ooCN

?CD

¿¿Vhcd

£3CSV.

QJ

¦aOl

ce

Ol

a

d

S

SWi

e"TT

e-

?d~

d

·—-*?ceOl

hj-!¦? O »M TtHJ- ,—I ?—I l/ì

fiVO l-l t? HJ- ID

fi fi TtI/) -H

t? Vi VO t?OnT-I l/ì Wi <S *S

TH Tt

TH i-? VO

VOHJ- VO

£

dddddd ddddd

ON

fiON

Ui

Ul90

OC

«M

loo

?O)·—3

-H»Q.CS

-S(?

!S?a»-ÍU«see

mITi

2007 electrofìshing

During fall 2007, a total of 584 juvenile Atlantic salmon were captured by

backpack electrofìshing. Genetic parentage analysis identified 485 individuals as fry

stocked in spring 2007, and the presence of adipose clips and/or floy tags indicated five

individuals were stocked as yearlings in 2007. The origin of the remaining 94 individuals

was unknown and could have been unmarked yearlings, fry, or fall fingerlings stocked

prior to sampling (Table 2-1). However, the location and age of unmarked individuals

suggests the majority of age-0 salmon of unknown origin salmon were stocked as fry in

2007, and age-1 salmon of unknown origin were fry stocked in 2007. Few salmon of

unknown origin were captured within the areas stocked with yearlings and fall

fingerlings, suggesting relatively few salmon previously stocked as yearlings or fall

fingerlings were captured (Figure 2-6). The estimated total density of Atlantic salmon

ranged from 2.2±1.9 to 138.4±4.6 salmon/1 00m2 across sampled sites, while CPUE

ranged from 0 to 90 fish/1 00m2, respectively (Figure 2-6). Total Atlantic salmon CPUE

was not uniformly distributed among sites (n=18, ?2= 451.77, P<0.001), and clumpedwithin areas stocked with fry (Figure 2-6).

In addition to Atlantic salmon, 3 brook trout, 1 1 Chinook salmon, 947 rainbow

trout, and 199 brown trout were also captured during electrofìshing surveys in 2007.

CPUE of group-0 other salmonids ranged from 2 fish/1 00m to 50 fish/1 00m and was

not uniformly distributed among sites (n=18, ?2= 157, PO.001). CPUE of othersalmonids exceeded 20 fish/1 00m2 only at sites 3, 7, 8, 15 and 16 (Figure 2-6). CPUE of

group-1 was not uniformly distributed among sites (n=18, ?2= 37, P<0.005), andexceeded 15 fish/1 00m2 at sites 2, 4, 5, 7, 8, and sites 13 through 18. (Figure 2-6).

54

1000

100

ui O

Q u>

Hi CPUEESI CPUE

CPUE

Age-1 Unknown ATSAge-0 Unknown ATS2007 Stocked ATS fry*

¦Je Total ATS Density

1 2 3 4 5 6 7 I 8 9 10 I 11 12 13 . 14 15 16 17 1iFIg and YrIg ,-,·,_\ release location/ 1V FrV stocking reach

B

1000

100

S? ionD w

0.1

CPUE: Other salmonids Age-0CPUE: Other Salmonids Age 1 and older

??-I

12 3 4 5 6 7 9 10 11 12 13 14 15 16 17 1£

Site

Figure 2-6: (A) Densities of stocked Atlantic salmon (ATS), and (B) other salmonids

(brown trout, brook trout, rainbow trout, and Chinook salmon) at sites sampled in

Cobourg Brook during fall 2007. Sites are in geographical order, progressing upstream

from Lake Ontario as shown in Figure 2-1. An Asterisk (*) in the legend indicates

stocking treatments used in the paired release-recapture experiment.

55

2008 smolt trapping

During spring 2008, the smolt trap fished daily from April 16 to August 13 with

the exception of May 4-5, June 6-7, June 13, July 22-25, and August 7-8 (Figure 2-7).

During early-season test netting three juvenile Atlantic salmon of unknown origin were

captured on March 27, but catches did not resume again until April 16 2008, the first day

of yearling stocking (Table 2-1: Figure 2-7). Throughout the duration of 2008 smolt

trapping, a total of 2,753 juvenile Atlantic salmon were captured, with an estimated 1,596

and 1 ,079 progeny from the Sebago Lake and LaHave River captive populations,

respectively (Figure 2-7). Captured salmon originating from the Sebago Lake captive

population could have been stocked as either yearlings in 2008 or fall fingerlings in 2007

(Table 2-1 ). Of the individuals derived from the LaHave River captive population 22

were stocked as sub-adults, and 32 were age-2 but lacking PIT tags or caudal punches,

suggesting they were stocked as fry, fall fingerlings, or yearlings prior to sampling (Table

2-1). Atlantic salmon catch peaked shortly after yearling stocking, but gradually subsided

and no salmon were captured beyond July 14 (Figure 2-7). A high abundance of dead

juvenile Atlantic salmon were captured during weeks 1 and 2 (Figure 2-8). Similarly,

51% and 48% of VIE marked Atlantic salmon died within 24 hours in weeks 1 and 2

respectively, after which mortalities were negligible. Genetic population assignment

indicated that 91% of all mortalities were Sebago Lake strain. In total, 326 juvenile

Atlantic salmon were marked with VIE as part of the mark-recapture experiment, and 36

were eventually recaptured. No Atlantic salmon marked with VIE lost their mark within

24 hours, and total tag loss (from mark to eventual recapture) was less than 1 %.

56

Unfortunately, low sample sizes prevented robust estimation of capture probability and

population size in spring 2008 using the Darroch (1961) estimator (Table 2-3).

In addition to Atlantic salmon a total of 48 group- 1 rainbow trout; as well as 21

group- 1 brown trout were captured during 2008 smolt trapping. The combined CPUE of

brown trout and rainbow trout was significantly lower in 2008 (0.6 ± 0.2fish/day)

compared to 2007 smolt trapping (4.3+ 1.3fish/day: Mann Whitney U= 2998, P<0.001).

57

DDIl·" 3

C e

ï

h S1

1

?-I—»

(UQ

oOCO

OO(O

OO

OOCM

<

(D

OMce

X><d

oo

O

1)>

2(D>

Xce

ce

(Zl«4-1Otí

O

"-4—»tí

00OOCN

OJO_tí'Ctí

-ö

o-4—»ceOOhCS

Oe(?

tíO

O

Lpieo de.ii 1|olus ?µßa

ce

feetí

?CUi-SOX)

ce

cetí

_oceS-H(DftO

?—»Otí(?ce

!-H

Og(D

-tíO

ce-O

(Dceo

t3G

GO

tí<üeCW)

ceCO

ce ?—?(? 3

PhO

(Dtí(DOOOO

_tícñtí

T3(Dtí

(D?a?ce

VO

Oh<,

OJOtí

OO

_tí?

(D>>

(+HO

ce-o

(DH-»

(Z)(D

-4—»ceo

*3_tí

ofc¡ce<a

-4—"

?Cce

(ZltíÖßtí<?tíce

ut"CN

?CNCN

3m

(DCtí

>—>

I

(DGtí

OOIT)

OOOO(NÖß

_gCO

öß.g3?J¡¿OO

m

3O

OOgo

_gfe

3O

O

-4—»03

?

3

&o?>-.

I

^a»-É03g

'-4—»o3

(U(U

3H

T3(U

03g

«1-1OS-I(U

g3C

3-4—»o3

i-(U>O?

?(U

o3O

O JU

?3gC3

O?-4(U

X)e3C(U

X!

e-4—403

(UXJgO

<u_N'(??

m —ftgco(/3

3-4—4o3

i*(U>OOD¡-4

?1-4(U>OO(U(-4?Cco

o3¡-?

(Ui—

-o?OOÖD03

en

(?t*—1O

<uX)gC

"03O

g3

-*—»031-4

-4-4(?

_g?(Uoo03(U

*(U

Og

e

1? -HO -H -q-"? Tf — »S — ?» _

O O O

S3

ONin

ü

coO

2500

2000 H

1500

1000 H

500 A

Weekly catch? i Weekly mortality catch

6 7 (May 28)(April 16) (April 23) (April 30) (May 7) (May 14) (May 21) 13 (July 9)

Week(Date)

Figure 2-8: Weekly catch and pre-handling mortalities ofjuvenile Atlantic salmon

during 2008 smolt trapping.

60

Fall 2008 Electrofishing

In the fall 2008 a total of 445 juvenile Atlantic salmon were captured during

electrofishing surveys. Genetic parentage analysis identified 47 and 336 individuals as fry

stocked in 2007 and 2008, respectively, as well as 12 individuals as yearlings stocked in

the spring of 2008. The origin of the remaining 50 individuals was unknown and could

have been stocked as unmarked fry, fall fingerlings, or yearlings prior to sampling (Table

2-1). No individuals originated from the Sebago Lake strain were captured. Few

unknown origin salmon were captured within areas stocked with yearlings or fall

fingerlings, suggesting that few individuals which were previously stocked as yearlings

or fall fingerlings were captured during this survey (Figure 2-9). The total density of

juvenile Atlantic salmon was an estimated 42 salmon/1 00m2 at site 18 and CPUE varied

from 0-51.5 salmon/1 00m2 (Figure 2-9). Total Atlantic salmon CPUE was not uniformlydistributed among sites (n=18, ?2= 470.15, P<0.001), and clumped within areas stockedwith Atlantic salmon fry (Figure 2-9). During fall 2008, Atlantic salmon fry stocked in

2007 and 2008 had a mean total length of 153±2.2mm and 10210.7 mm, respectively

(Figure 2-10).

In addition to Atlantic salmon, one brook trout, three Chinook salmon, 600

rainbow trout, and 132 brown trout, and were captured in fall 2008. CPUE of group-0

other salmonids ranged from 3 fish/1 00m2 to 23 fish/1 00m2 and was not uniformlydistributed among sites (n=18, ?2=238, PO.001), and only exceeded 10 fish/lOOm2 at

sites 2, 3, and 14 to 17 (Figure 2-9). CPUE of group- 1 other salmonids ranged from 0

fish/lOOm2 to 8 fish/lOOm2 and was not uniformly distributed among sites (n=18, ?2= 62,PO.001), and only exceeded 5 fish/1 00m2 at site 5 (Figure 2-9).

61

The size and abundance of stocked Atlantic salmon fry differed between fall 2007

and fall 2008. Age-0 stocked fry CPUE was significantly greater in fall 2007 (10±4.9

salmon/ 100m2) than in fall 2008 (5±2.7 salmon/1 00m2: Wilcoxon matched pairs test Z=3.05, P<0.01 : Figure 2-10). Conversely, total length of age-0 stocked fry was less in fall

2007 (99.± 0.6mm) than in fall in 2008 (101.6± 0.6mm: t0.05,(2),750=-2.6, PO.05).

The CPUE of group-0 other salmonids was significantly less in fall 2008 (9.1±1.6

salmon/ 100m2), compared to fall 2007 (16.3 ± 2.9 salmon/1 00m2: to.os,(2),34=-2.1,

P<0.05). However, there was no difference in the CPUE of group- 1 other salmonids

between fall 2008 (3.5+0.7SE salmon/1 00m2), and fall 2007 (5.6 ± 0.8SE salmon/lOOm2:

t0.05,(2),34=-1.9,P>0.05).

During fall 2007 and 2008, six sites were sampled using multiple pass backpack

electrofishing, and three different salmonid species were captured. Capture probability

was greater for group-1 (0.7±0.03) than group-0 stream salmonids (0.5±0.05) (T2o=-3.39,

P<0.01), and no differences were found between years (to.o5,(2),20=-3.39, P=0.16) (Figure

2-11).

62

1000 ?

100 A

tí) oC t-

Q «j10

-fc Total Estimated ATS DensityI I CPUE: 2008 Stocked ATS Yearlings "CSXS CPUE: 2007 Stocked ATS Fry *¦¦ CPUE: Age-2 Unmarked ATSI I CPUE: 2008 Stocked ATS Fry

Jl iIlI i

12 3 4 5 6 7 8 9 10 I 11 12 13FIg and VrIg I

\ release location/

14 15 16 17 18

Fry stocking reach

B

1000

U) Oc r~ID >Q »

Li.

10

0.1

F~~1 CPUE:Other salmonids Age-0HB CPUE: Other Salmonids Age- 1 and Older

¦¦ I

12 3 4 5 6 7 9 10 11 12 13 14 15 16 17 18

Site

Figure 2-9: (A) Densities of stocked Atlantic salmon (ATS), and (B) other salmonids at

sites sampled in Cobourg Brook during fall 2008 using backpack electrofishing. Sites are

in geographical order, progressing upstream from Lake Ontario as shown in Figure 2-1.

An asterisk (*) in the legend indicates a stocking treatment used in the paired release-

recapture experiment.

63

t s«i

OOOO(N

>-iO

oo(N

>-.<-t—?

&ocS?

CD

OO

C'5b¦co<dS-H

-4—»ce

-4—»

CO

e £

Etm

S'

?-t—»C

"?>

OC

_o

8

U. >- Ll.CO OO 1^-Q Ci QO Q QC-J C-J CN

ïs *b

fcyanbiuy

Oc<u

er(U

-U-t—?CJ)CCD

OOOO(N

cePL,

»s(UUS

OOOO(N

Dû_C"¡?03CD>,¦~O

Q. 0.6

RBT-O BRT-O ATS-O

Species-Group

• Site 4-2003¦ Site 9-2008X Site 5-2007? Site 7-2007T Site 15-2007A Site 18-2008

? 0.6

RBT-1 BRT- 1 ATS- 1

Species-Group

Figure 2-11: Capture probability of single-pass backpack electrofishing for group-0 (A),

and group- 1 (B) salmonids in Cobourg Brook. RBT = rainbow trout, BRT= brown trout

and ATS-= Atlantic salmon, Group-0 = age 0, and Group- 1= >age 0 (B). Year of

sampling is provided in legend.

65

Paired Release-Recapture Experiment

During the fall of 2007, 485 stocked fry from the paired release-recapture

experiment were captured. CPUE ranged from 0 salmon/100m2to 84 salmon/1 00m2

(Figure 2-6). The mean total length of stocked fry was 99mm± 0.6SE, and based on body

size the majority of individuals were expected to out-migrate the following spring.

(Figure 2-12).

During spring 2008, 32 fry and 918 yearlings were captured as age-1 out-

migrants. Mean CPUE of yearlings (10.8 salmon/day ± 6.1 SE) was 28 times greater than

fry (0.4 salmon/day ± 0.1 SE), and this difference was highly significant (n=86,

Wilcoxon matched pairs test statistic =2.85, PO.005: Figure 2-13). Mean total length of

Atlantic salmon stocked fry (143±4.1mm) was significantly less than those stocked as

yearlings (237±2.7mm: t0.05,(2),i67=12.95, PO.001). Total length data for both stocking

treatments were either x3 or x2 transformed to become normally distributed (K-S,P>0.05), and variances were homogenous (Levenes F,, 75o=0.07, P=O. 80).

During fall 2008, 47 fry and 12 yearlings were captured as age-1 parr. CPUE of

stocked fry (0.8 salmon/1 00m2 ± 0.3 SE) was not different than the CPUE of stocked

yearlings (0.2 salmon/1 00m2 ± 0.2 SE: Wilcoxon matched pairs test Z= 1.15, P=0.25:

Figure 2-9). Total length of stocked yearlings (237±7.4mm) was again greater than

stocked fry (152.91±2.20mm: to.05,(2),896=-21.84, PO.001: Figure 2-10). Total length data

for either stocking treatment were x3 or x2 transformed to become normally distributed(K-S, P>0.05), however variances were not homogenous (Levenes F,, 1 16=7.36, P<0.01).

The change in relative weight from spring and fall 2008 was different for fry and

yearling Atlantic salmon stocked in the paired release-recapture experiment, although

66

unfortunately direct comparison between the stocking treatments was not possible. The

length-weight regression slopes of stocked fry and yearlings were different during spring

2008 (ANOVA: F1, n5 = 6.17, PO.05), and fall 2008 (ANOVA: F,,=9.35,P<0.01).

However, the change in relative weight from spring 2008 to fall 2008 was analyzed to

evaluate post-stocking growth. The relative weight of stocked yearlings captured during

spring and fall 2008 were log-transformed to become normally distributed (K-S, P>0.05),

and slopes were found to be homogenous (ANOVA,: Fii99 =2.09, P=O.09). Length and

weight of stocked fry captured during spring and fall 2008 were normally distributed (K-

S, P>0.05), but slopes were not homogenous (ANOVA,: F1, 69 =28.88, PO.001). The

relative weight of stocked yearlings was less in fall 2008 than in spring 2008 (ANCOVA:

F,, ioo = 0.02, P<0.001); and by contrast, no such differences were found for stocked fry

(Figure 2-14).

67

too -t

. Xtm ¦»*>O .>C "ß<s C

2OH

SD

imtoo tao 140

Total length (mm)

ISO 180

Figure 2-12: Length frequency distribution of stocked fry captured in fall 2008 by

backpack electrofishing. The dashed line represents the parr-smolt length threshold, and

the white portion of the length-frequency distribution represent individuals expected to

out-migrate the following spring.

68

1000

3-CD

lu 5

oíco

100

10

0.1

May J un Jul

Date

B

1000

100

coUJ 5g O 10o 1

co

1 H

01May Jun Jul

Date

Figure 2-13: Spring 2008 CPUE of age-1 out-migrants originally stocked as fry (A) or

yearlings (B) in the paired release-recapture experiment.

69

250

200 -

150 -

100

5 50 4

20

• Fall 2008O Spring 2008

120 200

Total Length

300

B

150

100

50 ?

JZ

? 25 -

• Fall 2008O Spring 2008

oQO

75 100 200

Total Length

300

Figure 2-14: Condition (weight-at-length) of stocked yearling (A), and fry (B) Atlantic

salmon from the paired release-recapture experiment during spring and fall 2008.

70

Discussion

Paired Release- Recapture Experiment

Stocked fry and yearling Atlantic salmon introduced into Cobourg Brook differed

dramatically in their relative and absolute production of out-migrants. During 2008 smolt

trapping, significantly more stocked yearlings were captured compared to stocked fry.

Therefore, according to the original stipulations of the release-recapture experiment, the

survival of stocked fry to out-migration was less than 9%. Likewise, although this study

was not designed to quantify the absolute survival, yearling stocking resulted in 28 times

more out-migrants compared to fry stocking, suggesting the survival of stocked fry until

out-migration was approximately 0.32%. The annual survival ofjuvenile Atlantic salmon

during in-stream residence is typically 14-19% for hatchery reared juveniles, and 14-53%

for their wild counterparts, (Apprahamian et al, 2004; Cunjak and Therrein, 1998;

Jokikokko and Julita, 2004; McMenemy, 1995; Orciari et al, 1994; Cote and

Pomererleau, 1985; Kennedy and Strange, 1980; Miester, 1962; Egglishaw and Shackley,

1980; Elson, 1957), which indicates the survival of stocked fry in Cobourg Brook was

alarmingly low. Furthermore, if only 0.32% of stocked fry survive to smolt stage and

100% out-migrate at age-1, it would require stocking 1,041,667 fry to produce 200

returning adults, assuming survival from smolt to adult was 6% (Locke, 1998). Therefore,

if the conditions observed in this study prevail, fry stocking is an ineffective strategy to

restore Atlantic salmon populations in Cobourg Brook.

The growth of fry and yearling Atlantic salmon after release into Cobourg Brook

was also considerably different. The relative weight of Atlantic salmon is generally

lowest just prior to smolt stage (McCormick et al,. 1998). However yearlings stocked into

71

Cobourg Brook displayed the opposite, as their relative weight declined substantially

from spring to summer. Therefore, the in-stream growth of stocked yearlings appears to

be exceptionally poor compared to stocked fry, suggesting starvation is of concern and

few likely survived through the winter of 2008. In turn, yearlings should be stocked in a

manor to promote rapid out-migration, as stream conditions do not appear appropriate for

this life stage.

To evaluate the survival of stocking treatments, the paired release-recapture

experiment employed in this study relied on the assumption that all stocked Atlantic

salmon out-migrate at age-1. In fall 2008, few fry stocked in 2007 were captured

(approximately l/8th) relative to fall 2007. Therefore, few if any age-2 out-migrantswould be expected the following year considering the poor production of out-migrants

observed in 2008. Likewise, during 2008 smolt trapping, 918 yearlings from the paired

release were captured, and smolt trap capture efficiency was estimated to be 21 .5% (16-

27% in 2007). Therefore, approximately 4,269 of the 5,400 genetically marked yearlings

out-migrated during spring 2008 at age-1. In turn, it appears nearly all fry and yearling

Atlantic salmon that were released into Cobourg Brook in the paired release-recapture

experiment out-migrated at age-1, supporting the primary assumption of the study that

100% of both stocking treatments would out-migrate at age-1 .

Influences on the survival ofstocked Atlantic salmon in Cobourg Brook

Repeated field sampling provides insight into when mortality occurred. Fall

densities ofjuvenile Atlantic salmon typically range from 30-80 salmon/1 00m in wild

populations (Mills, 1989), and 5 salmon/100 m2 was established as an benchmark for

72

reintroducing Lake Ontario Atlantic salmon (Grieg et al, 2003). In fall 2007, age-0

densities of stocked fry exceeded the restoration benchmark and were within the range of

wild Atlantic salmon populations. Based on their size most were expected to out-migrate

the following spring. However, during 2008 smolt trapping, very few out-migrants

originally stocked as fry were captured, suggesting a survival bottleneck occurred during

the winter of 2007/2008. In addition, the out-migrant production of rainbow trout and

brown trout declined significantly from 2007 compared to 2008, providing further

evidence that severe over-winter mortality occurred during the winter of 2007/2008 in

Cobourg Brook. Over winter mortality in Atlantic salmon populations can occur during

low stream flows, and severe mortality occurs during events of extremely high water

levels (Cunjak et al, 1998). Both of these conditions occurred in Cobourg Brook during

the winter of 2007/2008, therefore the poor survival of fry stocked in 2007 was likely due

to the abiotic conditions within Cobourg Brook.

Although this study provided evidence that mortality was responsible for the

observed losses in stocked fry during the winter of 2007/2008, the possibility of pre-

smolt downstream movement cannot be dismissed. Atlantic salmon pre-smolts generally

remain near nursery areas until smolt migration (McCormick et al, 1998), however pre-

smolt downstream movement from nursery areas does occur in some populations

(Youngson et al, 1983; Riley et al, 2002; Pinder et al, 2007). Although the smolt trap

location was as close to Lake Ontario as logistically possible and netting commenced as

soon as ice conditions permitted, it is possible that stocked fry moved below the capture

site before sampling began. This possibility is further supported by the capture of three

73

juvenile Atlantic salmon, thirty rainbow trout, and eight brown trout during the first day

of test netting on March 27 2008.

Abiotic factors also appeared to affect the survival of young Atlantic salmon in

Cobourg Brook during the summer months as well. The survival of stocked fry through

their first growing season was significantly lower during 2008 compared to 2007. In 2007

and 2008 fry were reared at the same hatchery, stocked at similar sizes, abundances, and

densities, and the growing season was approximately the same length. However the 2008

growing season was marked with many brief periods of high stream level, and large rain

event occurred shortly after fry stocking on June 2 2008, which may have caused the

mass mortality of newly stocked Atlantic salmon fry. Other studies have also observed

this phenomenon. Letcher and Terrick (1998) suggested a large rain event occurring

shortly after Atlantic salmon fry stocking in New England resulted in high mortality.

Similarly, Coghlan and Ringer (2004) suggested a spike in stream flow may have caused

high mortality of eggs planted in the Salmon River, a Lake Ontario tributary in New

York.

Although this study was not intended to evaluate the effects of prédation on the

survival of stocked Atlantic salmon, some anecdotal information was available. Brook

trout (MacCrimmon, 1954; Symons, 1974), brown trout (Brannas 1995), and older

Atlantic salmon parr (Peppar et al., 1995) are all known to prey on young Atlantic

salmon and were common in the study site. Anglers provided a number of floy tags that

reportedly came from the stomachs of brown trout. The recovered tags were originally

applied to yearling Atlantic salmon stocked in 2007. In one case, I received 16 tags that

reportedly came from the stomach of a single brown trout. Therefore, it appears that

74

prédation the resident population of brown trout in Cobourg Brook may prey heavily onstocked yearling Atlantic salmon.

Influences on the growth ofstocked Atlantic salmon in Cobourg Brook

Generally, the growth of stream salmonids is more influenced by density-

independent factors such as food availability and temperature compared to density-

dependant factors such as competition (Jenkins et al, 1999; Imre et al,. 2005), and

likewise this concept may be acting in Cobourg Brook. Yearlings that remained within

Cobourg Brook after stocking in 2008 and were captured in fall 2008 displayed poor

growth and were found predominantly within the lower reaches of Cobourg Brook where

summer temperatures often surpassed 25°C, which exceeds the temperature preference of

juvenile Atlantic salmon (Garsid, 1973; Elliott, 1991; Elliott and Elliott 1995). Therefore,it appears unsuitable temperatures may be causing stress and starvation in stocked

yearlings which do not outmigrate after stocking.

Age at out-migration is predominantly a growth-related life history trait

(McCormick et al, 1998). Greater growth opportunity results in younger age at out-

migration (Wedemeyer et al, 1980; Metcalfe and Thorpe, 1990; Hutchings and Jones,

1998). Unlimited food is available in a hatchery setting and if heated water is available

age-1 smolts can be produced (Pennell and Barton, 1996). In a wild system however,

forage and stream temperatures can both limit Atlantic salmon growth. Therefore the

large proportion of age-1 out-migrants is indicative of the suitable biotic and abiotic

conditions within fry stocking sites.

75

This study found marked differences in the growth and survival of stocked fry and

yearling Atlantic salmon and provided some evidence suggesting aboitic factors were

largely responsible for such differences. However, other factors may have also

contributed. Stocking treatments were reared at separate hatcheries, and hatchery

conditions may have affected the observed differences in performance. In addition, other

factors may have affected the performance of stocked life stages, including biotic

conditions within Cobourg Brook (e.g. competition and prédation), ancestry (e.g.

domestication), release strategy, conditions (both aboitic and biotic) during transportation

from the hatchery and during stocking, and paternity. Without isolating such alternative

factors, this study cannot conclusively determine the most effective life stage to release in

Atlantic salmon reintroduction programs. In addition, this study used the size and

abundance of released individuals as a metric of their performance. Ideally, a more

definitive measure would be used, such as the production of wild offspring. Despite these

limitations, this study has provided considerable information suggesting the life stage of

released animals does affect their likelihood of re-establishing viable populations.

Supplementary Information

During out-migration, stocked yearlings were larger than stocked fry, and

likewise larger than smolts from most wild Canadian Atlantic salmon populations, which

typically range from 1 10-1 90mm TL (Peppar, 1982; Chaput et al, 2002; Clément et al,

2007). Large size at out-migration can result in increased survival to adulthood (Behmer

et al, 1993; Saloniemi et al, 2004). Therefore, because yearling stocking produced more

76

and larger out-migrants than fry stocking, yearling stocking may result in more

reproductive adults.

Smolt out-migration can be interrupted by high-head dams (Manning et al, 2005;

Plumb et al, 2006). If smolt migration is interrupted or delayed, individuals may loose

their physiological and behavioral predisposition to migrate (McCormick et al, 1998,

McCormick et al, 1999), and instead remain in-stream for an additional year. During the

fall of 2008 the vast majority of stocked yearlings were captured at the sampling site in

closest upstream proximity to the lower-most dam. This suggests that downstream out-

migration was likely interrupted upon encountering the dam, causing some individuals to

remain in-stream over the summer.

There was little upstream or downstream movement of fry from stocking sites in

Cobourg Brook. This contrasts with Webb et al, (2007), who describe fry as territorial,

promoting downstream dispersal up to 1.5 km. The downstream dispersal of stocked fry

allows individuals to locate areas of low competition and high forage availability (Webb

et al, 2008). The lack of movement of fry in Cobourg may suggest that space and food

(competition) are not limiting, at least not at current stocking densities. Future fry should

be stocked directly into areas of relatively high quality habitat. While this habitat may

promote strong growth, overwintering habitat may be limited.

This study provided little information on the performance of sub-adult (age-2)

Atlantic salmon introduced into Cobourg Brook. Individuals stocked at this life stage

were not encountered during any fall survey, likely due to the low abundance of

individuals released. However, sub- adults were captured during smolt trapping in 2007

and 2008. Most individuals were captured within 2 weeks of release. Although little is

77

known about the effectiveness of this life stage in restoration efforts, the low abundance

of these individuals released annually suggests their importance as a restoration strategy

in this stream is low.

Of the primary life stages being stocked in the current Lake Ontario Atlantic

salmon restoration program, only fall fmgerlings were not directly assessed in this study.

Fall fmgerlings are typically individuals from the smaller group of the characteristic

bimodal size distribution ofjuvenile Atlantic salmon, and are expected to smolt after at

least 1 year in the wild (e.g. age-2 smolts: Perinei and Barton, 1996). Fall fmgerlings

stocked in 2006 were progeny of the LaHave River captive population but not genetically

tagged. However, very few age-1 Atlantic salmon of unknown origin (i.e possibly

stocked at any life stage) were captured during fall surveys in 2007 or 2008 near fall

fingerling stocking sites, which suggests fall fmgerlings either out-migrated before fall

assessment at age-0 or age-1, or died. The out-migration of fall fmgerlings at age-1 could

not be directly assessed. Fall fmgerlings could have been a component of the unmarked

individuals captured during 2007 smolt trapping, however scale pattern analysis was

unable to differentiate these individuals from stocked fry or unmarked yearlings

(Appendix 1). Fall fmgerlings stocked in 2007 were progeny of the Sebago Lake

broodstock. During 2008 smolt trapping, genetic population assignment identified 1079

individuals as belonging to the Sebago Lake population; however these individuals

cannot be delineated from Sebago strain yearlings stocked earlier that spring. Fall

fingerling stocking was apparently an effective stocking strategy during the early

attempts to conserve the Lake Ontario Atlantic salmon population in the 1800' s (Parsons,

1978), and has been useful in supplementation and estoration efforts in Finland

78

(Salminen, 2007; Jokikokko and Mita, 2004). Therefore, this study did not thoroughly

evaluate the performance of fall fingerlings, and this life stage may provide a viable

stocking strategy in the current effort to re-establish Atlantic salmon in Lake Ontario.

The marked differences in the survival and growth ofjuvenile Atlantic salmon

observed in this study, coupled with clear evidence of environmental and hatchery

influences underscore the need for an experimental approach in restoration efforts. By

refining goals and objectives, the ongoing effort to reintroduce Atlantic salmon

populations in Lake Ontario tributaries may evaluate the relative importance of these and

other factors obstructing the efficient reintroduction of migratory stream salmonids.

79

References:

Aprahamian, M. W., Barnard, S. & Farooqi, M. A. (2004). Survival of stocked Atlanticsalmon and coarse fish and an evaluation of costs. Fisheries Management and Ecology11, 153-163.

Araki, H., Berejikian, B. A., Ford, M. J. & Blouin, M. S. (2008). Fitness of hatchery-reared salmonids in the wild. Evolutionary Applications 1, 342-355.

Armstrong, D.P., Soderquist, T. & Southgate, R. (1994) Designing experimentalreintroductions as experiments. Réintroduction biology ofAustralian and New ZealandFauna (ed. M. Serena), pp. 27-29. Surrey Beatty & Sons, Chipping Norton

Armstrong, D. P. & Seddon, P. J. (2008). Directions in reintroduction biology. Trends inEcology & Evolution 23, 20-25.

Augerot, X. 2005. Atlas ofPacific Salmon: thefirst map-based status assessment ofsalmon in the north Pacific. University of California Press, Berkeley, California.

Balon, E. K. (1985). The theory of saltatory ontogeny and life history models revisited. InEarly Life History ofFishes (Balon, E. K., ed.). Dorderecht, Netherlands: Dr. W. JunkPublishers.

Behmer, D. J., Greil, R. W., Scott, S. J. & Hanna, T. (1993). Harvest and movement ofAtlantic salmon in the St-Marys River, Michigan. Journal ofGreat Lakes Research 19,533-540.

Bisset, J., Bowlby, J., Jones, M., Marchant, B., Miller-Dodd, L., Orsatti, S. & Stanfield,L. (1995). An Atlantic salmon restoration plan for Lake Ontario. Picton, Ontario: LakeOntario Management Unit

Bjorkstedt, E. 2005. DARR 2.0: Updated software for estimating abundance fromstratified mark- recaptured data. NOAA-TM-NMFS-SWFSC 68. 21pp.

Brannas, E. B. (1995). First access to territorial space and exposure to prédation pressure:a conflict in early emerging Atlantic salmon fry. Evolutionary Ecology 9, 41 1-420.

Brown, C. & Day, R. L. (2002). The future of stock enhancements: lessons for hatcherypractice from conservation biology. Fish and Fisheries 3, 79-94.

Carcao, G. (1987). Atlantic salmon in the Great Lakes basin: a history of its extirpationand attempted restoration. Unpublished manuscript, H.A. Regier: University of Toronto,Department of Zoology.

Casselman, J. M. & Scott, K. A. (2000). A general procedures manual for CSAGES-Calcified Structure Age and Growth data Extraction Software (Version 5.2). p. 96p.Picton, Ontario: Glenora Fisheries Research Station.

80

Chaput, G., Hardie, P., Hayward, J., Moore, D., Shaesgreen, J. & NSPA (2002).Migrations and biological characteristics of Atlantic salmon from the NorthwestMiramichi River, 1998 to 2000. Canadian Technical Report ofFisheries and AquaticScience 2415: Department of Fisheries and Oceans.

Christie, W. J. (1972). Lake Ontario: effects of exploitation, introductions, andeutrophication on the salmonid community. Journal ofthe Fisheries Research Board ofCanada, 29 913-929.

Clément, M., Chaput, G. & Leblanc, P. (2007). Atlantic salmon {Salmo salar) smoltsmigration from the Margaree River, 2001 to 2003. Canadian Technical. Report ofFisheries and Aquatic Sciences, 2683. p. 60.

Coghlan, S. M. & Ringler, N. H. (2004). Comparison of Atlantic salmon embryo and frystocking in the salmon river, New York. North American Journal ofFisheriesManagement 24, 1385-1397.

Cote, Y. & Pomerleau, C. (1985). Survival and dispersion of Atlantic salmon fry {Salmosalar) released in the natural environment. Naturaliste Canadien 112, 549-557.

Crawford, S. S. (2001). Salmonine introductions to the Laurentian Great Lakes: anhistorical review and evaluation ofecological effects. Canadian Special Publication ofFisheries and Aquatic Science. No. 132.

Cross, T. F., McGinnity, P., Coughlan, W., Dillane, E., Ferguson, A., Koljonen, M.-L.,O'Reilly, P. & Vasmagi, A. (2007). Stocking and Ranching. In The Atlantic salmon;Genetics Conservation and Management (Verspoor, E., Stradmeyer, L. & Nielson, J.,eds.), pp. 325-356. Ames Iowa: Blackwell Publishing.

Cunjak, R. A., Prowse, T. D. & Parrish, D. L. (1998). Atlantic salmon {Salmo salar) inwinter: "the season of parr discontent"? Canadian Journal ofFisheries and AquaticSciences 55, 161-180.

Darroch, J. N. (1961). The two-sample capture-recapture census when tagging andsampling are stratified. Biometrika 48, 241-260.

DeVries, D. R. & Frie, R. V. (1996). Determination of Age and Growth .In FisheriesTechniques (Wells, B. R. M. a. D. W., ed.), pp. 483-512. Bethesda, Maryland: AmericanFisheries Society.

Dunfield, R.W. 1985. The Atlantic salmon in the history ofNorth America. CanadianSpecial Publication ofFisheries and Aquatic Science.. 80. 181 p.

Egglishaw, H. J. & Shackley, P. E. (1980). Survival and growth of salmon, Salmo salar(L.), planted in a Scottish stream. Journal offish biology 16, 565-584.

Elliott, J. M. (1991). Tolerance and resistance to thermal stress in juvenile Atlantic

81

salmon, Salmo salar. Freshwater Biology 25, 61-70.

Elson, P.F. (1957) The importance of size in the change from parr to smolt in Atlanticsalmon. Canadian Fish Culture, 21: 1-6.

Egglishaw, H. J. & Shackley, P. E. (1980). Survival and growth of salmon, Salmo salar(L.), planted in a Scottish stream. Journal offish biology 16, 565-584.

Elliott, J. M. & Elliott, J. A. (1995). The effect of the rate of temperature increase on thecritical thermal maximum for parr of Atlantic salmon and brown trout. Journal ofFishBiology 47, 917-919.

Elliott, J.M. (1994) Quantitative Ecology and the Brown Trout. Oxford University Press,Oxford.

Fischer, J. & Lindenmayer, D. B. (2000). An assessment of the published results ofanimal relocations. Biological Conservation 96, 1-11.

Fleming, I. A., and E. Petersson. 2001. The ability of released hatchery salmonids tobreed and contribute to the natural productivity of wild populations. Nordic Journal ofFreshwater Research 75:71-98. 261.

Ford, M. J. 2002. Selection in captivity during supportive breeding may reduce fitness inthe wild. Conservation Biology 16:815-825.

Garside, E. T. (1973). Ultimate upper lethal temperature of Atlantic salmon (Salmosalar). Canadian Journal ofZoology 51, 598-600.

Grant, J.W.A. & Kramer, D.L. (1990) Territory size as a predictor of the upper limit topopulation density ofjuvenile salmonids in streams. Canadian Journal ofFisheries andAquatic Sciences, 47', 1724-1737.

Grieg, L., Ritchie, L., Carl, L. & Lewis, C. A. (2003). Potential and Strategy forRestoration of Atlantic salmon in Lake Ontario: A Workshop Report, p. 39.Peterborough, ON: Ontario Ministry of Natural Resources, Lake Ontario ManagementUnit.

Griffith, B., Scott, J. M., Carpenter, J. W. & Reed, C. (1989). Translocation as a speciesconservation tool- status and strategy. Science 245, 477-480.

Hansen, L. P., Jonsson, B. & Doving, K. B. (1984). Migration od wild and hatcheryreared smolts of Atlantic salmon, through lakes. Journal ofFish Biology 25, 617-623.

Hesthagen, T. & Garnas, E. (1983). Migration of Atlantic salmon smolts in River Orkla,central Norway in relation to management of a hydroelectric facility. North AmericanJournal ofFisheries Management 6, 376-382.

82

Horton, G. E., Letcher, B.H., Bailey, M. M. & Kinnison, M. T. (2009). Atlantic salmon(Salmo salar) smolt production: the relative importance of survival and body growth.Canadian Journal ofFisheries and Aquatic Sciences 66, 471-483.

Hutchings, J. A. & Jones, M. E. B. (1998). Life history variation and growth ratethresholds for maturity in Atlantic salmon, Salmo salar. Canadian Journal ofFisheriesand Aquatic Sciences 55, 22-47.

Hvidsten, N. A. & Lund, R. A. (1988). Prédation on hatchery reared and wild smolts ofAtlantic salmon (Salmo salar), in the estuary of the River Orkla, Norway Journal ofFishBiology 33, 121-126.

Imre, L, Grant, J. W. A. & Cunjak, R. A. (2005). Density-dependent growth of young-of-the-year atlantic salmon Salmo salar in Catamaran Brook, New Brunswick. Journal ofAnimal Ecology 74, 508-5 1 6.

enkins, T. M., Diehl, S., Kratz, K. W. & Cooper, S. D. (1999). Effects of populationdensity on individual growth of brown trout in streams. Ecology 80, 941-956.Jokikokko, E. & Jutila, E. (2004). Divergence in smolt production from the stocking of 1-summer-old and 1 -year-old Atlantic salmon parr in a northern Baltic river. Journal ofApplied Ichthyology 20, 5 1 1 -5 1 6.

Jones, M. & Stanfield, L. (1993). Effect of exotic juvenile salmonines on growth andsurvival ofjuvenile Atlantic salmon (Salmo salar) in a Lake Ontario tributary. InProduction ofjuvenile Atlantic salmon, Salmo salar, in natural waters (Gibson R J &Cutting, R. E., eds.), pp. 71-79.

Jonsson, N., Jonsson, B. & Hansen, L. P. (2003). The marine survival and growth of wildand hatchery-reared Atlantic salmon. Journal ofApplied Ecology 40, 900-91 1.Kennedy, G. J. A. & Strange, C. D. (1981). Comparative survival form salmon (Salmosalar) stocking with green and eyed ova in an upland stream. Fisheries Management 1243-48.

Ketola, H. G., Bowser, P. R., Wooster, G. A., Wedge, L. R. & Hurst, S. S. (2000). Effectsof thiamine on reproduction of Atlantic salmon and a new hypothesis for their extirpationin Lake Ontario. Transactions ofthe American Fisheries Society 129, 607-612.Keeley, E.R. & Grant, J.W.A. (1995) Allometric and environ- mental correlates ofterritory size in juvenile Atlantic salmon Salmo salar. Canadian Journal ofFisheries andAquatic Sciences, 52, 186-196.

Lackey, R. T., D. Lach, and S. Duncan. 2006. Salmon 2100: The Future of Wild PacificSalmon. American Fisheries Society, Bethesda, MD.

Letcher, B. H., T. Dubreuil, M. J. O'Donnell, M. Obedzinski, K. Griswold, and K. H.Nislow. 2004. Long-term consequences of variation in timing and manner of fry

83

introduction on juvenile Atlantic salmon (Salmo salar) growth, survival, and life-historyexpression. Canadian Journal ofFisheries and Aquatic Sciences 61:2288-2301.

Letcher, B. H. & Terrick, T. D. (2001). Effects of developmental stage at stocking ongrowth and survival of Atlantic salmon fry. North American Journal ofFisheriesManagement 21, 102-110.

Locke, A. (1998). Modeling the effects of poststocking survival rates on the success ofstocking hatchery Atlantic salmon in a New Brunswick River. North American Journal ofFisheries Management, 547-560.

MacKenzie, C. & Moring, J. J. (1988). Estimating survival of Atlantic salmon during theintragravel period. North American Journal ofFisheries Management, 45-49.Manning, D. J., Mann, J. A., White, S. K., Chase, S. D. & Benkert, R. C. (2005).Steelhead emigration in a seasonal impoundment created by an inflatable rubber dam.North American Journal ofFisheries Management 25, 1239-1255.

MacCrimmon, H. R. (1954). Stream studies in planted Atlantic salmon. Journal oftheFisheries Board ofCanada 11, 363-401 .

McCormick, S. D., Hansen, L. P., Quinn, T. P. & Saunders, R. L. (1998). Movement,migration, and smoking of Atlantic salmon {Salmo salar). Canadian Journal ofFisheriesand Aquatic Sciences 55, 77-92.

McMenemy, J. R. (1995). Survival of Atlantic salmon fry stocked at low density in theWest River, Vermont. North American Journal ofFisheries Management 1 5, 336-374.Miester, A. L. (1962). Atlantic salmon production in Cove Brook, Maine. Transactions ofthe American Fisheries Society 91, 208-212.

Metcalfe, N. B. (1998). The interaction of behavior and physiology in determining lifehistory patterns in Atlantic salmon {Salmo salar). Canadian Journal ofFisheries andAquatic Sciences 55, 93-103.

Metcalfe, N. B. & Thorpe, J. E. (1990).Determinants of geographical variation in the ageof seaward migrating salmon {Salmo salar). Journal ofAnimal Ecology 59, 135-145.Miller, B. A., Rodgers, J. D. & S., S. M. (2000). An automated devise to release markedjuvenile fish for measuring trap efficiency. . North American Journal ofFisheriesManagement, 20, 284-287.

Mills, D. (1989). The Ecology and Management ofAtlantic Salmon. London: Chapman &Hall.

Mills, E. L., Casselman, J. M., Dermo«, R., Fitzsimmons, J. D., Gal, G., Holeck, K. T.,Hoyle, J. ?., Johannsson, O. E., Lantry, B. F., Makarewicz, J. C, Millard, E. S.,Munawar, I. F., Munawar, M., O'Gorman, R., Owens, R. W., Rudtsam, L. G., Schaner, T.

84

& Stewart, T. J. (2003). Lake Ontario: food web dynamics in a changing ecosystem(1970-2000)., Canadian Journal ofFisheries and Aquatic Sciences, 60:471-491.

Myers, R. A., Levin, S. A., Lande, R., James, F. C, Murdoch, W. W. & Paine, R. T.(2004). Hatcheries and endangered salmon. Science 303, 1980-1980.

Nicieza, A. G., Brana, F. & Toledo, M. M. (1991). Development of length bimodialityand smoking in wild stocks of Atlantic salmon {Salmo salar), under different growthconditions. Journal ofFish Biology 38, 509-523.

Nislow, K. H., S. Einum, and C. L. FoIt. 2004. Testing predictions of the critical periodfor survival concept using experiments with stocked Atlantic salmon, iournal ofFishBiology 65:188-200.

Obedzinski, M., and Letcher, B. H. 2004. Variation in freshwater growth anddevelopment among five New England Atlantic salmon (Salmo salar) populations rearedin a common environment. Canadian Journal of Fisheries and Aquatic Sciences 61:2314-2328.

Olsson O. 2007. Genetic origin and success of reintroduced white storks. ConservationBiology 21:1 196-1206.

Ontario Ministry of Natural Resources. 2007. Lake Ontario Fish Communities andFisheries: 2006 Annual Report of the Lake Ontario Management Unit. Ontario Ministryof Natural Resources, Picton, Ontario, Canada.

O'Reilly, P., Hamilton, L., McConnell, S. & Wright, J. (1996). Rapid analysis of geneticvariation in Atlantic salmon (Salmo salar) by PCR multiplexing of dinucleotide andtetranucleotide microsatellites. Canadian Journal ofFisheries and Aquatic Science 53,2292-2298.

Orciari, R. D. & Leonard, G. H. (1996). Length characteristics of smolts and timing ofdownstream migration among three strains of Atlantic salmon in a southern NewEngland stream. North American Journal ofFisheries Management 16, 851-860.

Otis, D. L., K.P., B., White, G. C. & Anderson, D. R. (1978). Statistical inference fromcapture data on closed animal populations. Wildlife Monographs 4, 99-102.

Parrish, D. L., Behnke, R. J., Gephard, S. R., McCormick, S. D. & Reeves, G. H. (1998).Why aren't there more Atlantic salmon (Salmo salar)? Canadian Journal ofFisheries andAquatic Sciences 55, 281-287.

Parsons, J. W. (1978). History of salmon in the Great Lakes, 1 850-1970. In U.S. Bureauof Sport Fish and Wildlife Technical Paper.

85

Paterson, S., Piertney, S. B., Knox, D., Gilbey, J. & Verspoor, E. (2004). Characterizationand PCR multiplexing of novel highly variable tetranucleotide Atlantic salmon {Salmosalar L.) microsatellites. Molecular Ecology Notes 4, 160-162.

Pearlstein, J. H., Letcher, B. H. & Obedzinski, M. (2007). Early discrimination ofAtlantic salmon smolt age: Time course of the relative effectiveness of body size andshape. Transactions ofthe American Fisheries Society 136, 1622-1632.

Peppar, J. L. (1982). Atlantic salmon smolt investigations, Restigouche River System,New Brunswick. Department ofFisheries and Oceans. No. 1641

Pinder, A. C, Riley, W. D., Ibbotson, A. T. & Beaumont, W. R. C. (2007). Evidence foran autumn downstream migration and the subsequent estuarine residence of 0+yearjuvenile Atlantic salmon Salmo salar L., in England. Journal ofFish Biology 71, 260-264.

Piry, S., Alapetite, A., Cornuet, J. M., Paetkau, D., Baudouin, L. & Estoup, A. (2004).GENECLAS S2: A software for genetic assignment and first-generation migrantdetection. Journal ofHeredity 95, 536-539.

Plumb, J. M., Perry, R. W., Adams, N. S. & Rondorf, D. W. (2006). The effects of riverimpoundment and hatchery rearing on the migration behavior ofjuvenile steelhead in thelower Snake River, Washington. North American Journal ofFisheries Management 26,438-452.

Pompanon, F., Bonin, A., Bellemain, E. & Taberlet, P. (2005). Genotyping errors:Causes, consequences and solutions. Nature Reviews Genetics 6, 847-859.

Pope, K. L. & Kruse, C. G. (2007). Condition. In Analysis and Interpretation ofFreshwater Fisheries Data (Guy, C. S. & Brown, M. L., eds.), pp. 423-472. Bethesda,Maryland: American Fisheries Society.

Rannala, B. & Mountain, J. L. (1997). Detecting immigration by using multilocusgenotypes. Proceedings ofthe National Academy ofSciences ofthe United States ofAmerica 94, 9197-9201.

Riley, W. D., Eagle, M. O. & Ives, S. J. (2002). The onset of downstream movement ofjuvenile Atlantic salmon, Salmo salar L., in a chalk stream. Fisheries Management andEcology 9, 87-94.

Ryman, N, and F. Utter. 1987. Population Genetics and Fishery Management. Universityof Washington Press, Seattle, WA.

Salminen, M., Alapassi, T. & Ikonen, E. (2007). The importance of stocking age in theenhancement of River Kymijoki salmon {Salmo salar). Journal ofApplied Ichthyology23, 46-52.

86

Saloniemi, L, Jokikokko, E., Kallio-Nyberg, L, Jutila, E. & Pasanen, P. (2004). Survivalof reared and wild Atlantic salmon smolts: size matters more in bad years. Ices Journal ofMarine Science 61, 782-787.

Sarrazin, F. & Barbault, R. (1996). Reintroduction: Challenges and lessons for basicecology. Trends in Ecology & Evolution 11, 474-478.

Sarrazin, F. & Legendre, S. (2000). Demographic approach to releasing adults versusyoung in reintroductions. Conservation Biology 14, 488-500.

Scarnecchia, D. L. (1984). Climatic and oceanic variations affecting yeild of Icelandicstockes of Atlantic salmon {Salmo salar). Canadian Journal ofFisheries and AquaticSciences 41, 917'-935.

Seddon, P. J., Armstrong, D. P. & Maloney, R. F. (2007). Developing the science ofreintroduction biology. Conservation Biology 21, 303-312.

Skalski, J. R., Buchanan, R. A. & Griswold, J. (2009). Review of marking methods andrelease recapture designs for estimating the survival of very small fish: examples fromthe assessment of salmonid fry survival. Reviews in Fisheries Science 17, 391-401.

Solomon, D. J. (1978). Some observations of smolt migration in a chalk stream. Journalof fish biology 12.

Stanfield, L. & Jones, M. L. (2003). Factors influencing rearing success of Atlanticsalmon stocked as fry and parr in Lake Ontario tributaries. North American Journal ofFisheries Management 23, 1175-1183.

Symons, P. E. K. (1974). Territorial behavior ofjuvenile Atlantic salmon reducesprédation by brook trout. Canadian Journal ofZoology 52, 677-679.

Taggart, J. B. (2007). FAP: an exclusion-based parental assignment program withenhanced predictive functions. Molecular Ecology Notes 7, 412-415.

Thorpe, J. E. (1988). Salmon migration. Science Progress 72, 345-370.

Van Deventer, J. S. (2006). MicroFish 3.0. p. Available at http://www.microfish.com.

Webb, J., Verspoor, E., Aubin-Horth, N., Romaakkaniemi, A. & Amiro, P. (2007). TheAtlantic Salmon. In The Atlantic salmon; Genetics Conservation and Management(Verspoor, E., Stradmeyer, L. & Nielson, J., eds.), pp. 17-45. Ames, Iowa: BlackwellPublishing.

Wedemeyer, G. A., Saunders, R. L. & Clarke, C. (1980). Environmental factors affectingsmoltification and early marine survival of anadromous salmonids. Marine FisheriesReview 42, 1-14.

87

Whalen, K. G. & Labar, G. W. (1994). Survival and growth of Atlantic salmon {Salmosalar) survival and growth stocked at varying densities in the West River, Vermont.Canadian Journal ofFisheries and Aquatic Sciences 51, 2164-2169.

Whalen, K. G., Parrish, D. L. & McCormick, S. D. (1999). Migration timing of Atlanticsalmon smolts relative to environmental and physiological factors. Transactions oftheAmerican Fisheries Society 128, 289-301.

White, G. C., Anderson D, R., , B., K.P & Otis, D. L. (1982). Capture-Recapture andremoval methods for sampling closed populations. Los Almos, NM: Los Almos NationalLibrary.

Whitesel, T. A. (1993). Comparison ofjuvenile Atlantic salmon (Salmo salar) reared in ahatchery and introduced into a stream: a two-size-threshold model of smoltification. InProduction ofjuvenile Atlantic salmon. Salmo salar, in natural waters (Gibson, R. J. &Cutting, R. E., eds.), pp. 239-247.

Youngson, A. F., Buck, R. J. G., Simpson, T. H. & Hay, D. W. (1983). The Autoum ansspring emigrations ofjuvenile Atlantic salmon {Salmo salar) from the Girnock Burn,Aberdeedshire, Scotland-environmental release of migration. Journal ofFish Biology 23,625-639.

Zar, J. (1984). Biostatistical Analysis. Toronto: Prentice-Hall Inc.

Zydlewski, G. B., Haro, A. & McCormick, S. D. (2005). Evidence for cumulativetemperature as an initiating and terminating factor in downstream migratory behavior ofAtlantic salmon {Salmo salar) smolts. Canadian Journal ofFisheries and AquaticSciences 62, 68-78

Williams, R. N. 2006. Return to the River: Restoring Salmon to the Columbia River.Elsevier Academic Press, Amsterdam/ Boston, MA.

88

Chapter 3: The out-migration timing of stocked fry and

yearling Atlantic salmon from Cobourg Brook

Abstract

The timing of migration events with respect to environmental conditions plays an

important role in survival of diadromous fishes, and should be considered in stocking

programs. Multiple juvenile life stages of Atlantic salmon have been stocked into

selected Lake Ontario tributaries as part of the effort to re-establish Atlantic salmon

populations. This study evaluated the spring emigration of marked Atlantic salmon

stocked as fry and yearlings into Cobourg Brook with respect to local temperature and

flow conditions. Stocked life stages exhibited significantly different emigration patterns

between life stages and stocking years in comparison with modeled site-specific

environmental smolt window predictions. The majority of captured yearlings (74%) were

caught from June 13 to July 9 2007, when stream temperature approached and exceeded

200C. By contrast, most (88%) of marked yearlings sampled in spring 2008 were

captured within 14 days of stocking, prior to stream temperature attaining 100C. Of all

stocked fry captured during spring 2008, 91 .5% were caught from April 27 to May 16

2008 once stream temperature exceeded 100C. No correlations were found between smolt

trap catch of stocked yearlings or fry and stream temperature, stream level, change in

stream temperature, change in stream level, or cumulative degree-days. Out-migration of

stocked yearlings was out of phase with the environmental smolt window in both 2007

and 2008 and appeared to be linked to rearing conditions, whereas stocked fry out-

migration in 2008 occurred within the predicted smolt window. Rearing practices for

yearling Atlantic salmon stocked into Lake Ontario streams should be refined as out-

migration outside the environmental smolt window may reduce survival to adulthood,

and therefore likelihood for eventual population re-establishment.

90

Introduction

Captive breeding programs are important tools in efforts to re-establish extirpated

populations (Sarrazin and Barbault, 1996; Sarrazin and Legendre, 2000; Robert et al,

2004). Without careful management, however, captive breeding can degrade adaptive

traits important for wild survival (Snyder et al, 1996; Lynch and O'Hely, 2001 ; Araki et

al, 2007). Hatchery propagation is an important tool in fish conservation and fisheries

management, as billions of hatchery-reared fish are released annually to enhance existing

populations or create new ones (Brown and Day, 2002). Modern hatchery practices often

strive to maximize the production ofjuveniles for stocking into the wild, with little

consideration for the behavioral and evolutionary consequences associated with unnatural

captive rearing conditions (Pennell and Barton, 1996; Brown and Day, 2002; Araki et

ai, 2007). Artificial hatchery environments can alter fish behaviors that are based on

prior life experiences, such as predator avoidance, foraging, and homing to spawning

sites (Brown and Day, 2002; Brown and Laland, 2003), and may likewise alter the

migration timing of diadromous fish species.

Smolting is the adaptation for stream out-migration and entry into lakes or oceans,

and is displayed by many migratory salmonid species (Hoar, 1988). Migration of Atlantic

salmon smolts is believed to be cued by environmental 'priming' and 'releasing' factors,

which occur during behavioral and physiological periods of migratory readiness

(McCormick et al., 1998; Zydlewski et al, 2005). Ifjuvenile salmon (parr) attain a size-

related developmental stage by the end of the growing season, they undergo behavioral,

morphological and physiological changes to become adapted for downstream migration

and life within the pelagic environment of lakes or oceans, and are primed for

91

downstream migration the following spring (Skilbrei, 1991; Whitesel, 1993; McCormick

et al, 1998).

Fluctuations in spring temperature and stream level act as the predominant

releasing factors cueing out-migration of Atlantic salmon smolts. Out-migration can

occur once a stream temperature threshold has been exceeded, typically between 5°C and

100C (White, 1939; Solomon, 1978; Whalen et al, 1999; Antonsson and Gudjonsson,

2002; Byrne et al, 2003). Zydlewski et al. (2005), however, found that temperature

thresholds did not influence Atlantic salmon smolt out-migration, rather the cumulative

effects of temperature increasing over time as reflected by cumulative degree-days, were

the primary releasing factor to cue Atlantic salmon smolt out-migration. Increases in

stream flow may also present an important releasing factor as many field studies have

found out-migration to peak during increases in stream flow caused by rain events (Allen,

1944; Solomon, 1978; Hesthagen and Garnas, 1983; Jonsson and Ruud-Hansen, 1985;

Hvidsten and Hansen, 1988; Dempson et al, 1996; Whalen et al, 1999). The timing of

Atlantic salmon smolt migration may also be influenced by genetic factors (Webb et al,

2007), but its relative importance is currently unknown.

Smolt migration timing is important for survival to adulthood. McCormick et al,

(1998) presented the 'environmental smolt window' hypothesis where maximum smolt

survival occurs when migration overlaps with favorable environmental conditions such as

suitable stream temperatures and levels, presence of predators or prey, or lake/ocean

environmental conditions (Figure 1-3). The hypothesis was supported by the previous

work of Hansen (1987), who found that atypical out-migration ofjuvenile Atlantic

salmon resulted in poor survival to adulthood. More recently, multiple studies have

92

shown that environmental conditions upon entry into the marine environment were

important for subsequent survival of post-smolt Atlantic salmon (Hansen and Jonsson,

1991; Friedland et al, 1998; Jokikokko and Mantyniemi, 2003; Kallio-Nyberg et al,

2004; Kallio-Nyberg et al, 2006). To maximize subsequent survival and adult returns,

Atlantic salmon stocking programs should therefore incorporate optimal life stage and

size considerations to maximize out-migration during the local environmental smolt

window of stocked watersheds.

Efforts to restore or reintroduce migratory salmonid populations should also

consider the potential adverse effects of captive rearing on smolt out-migration and

subsequent survival and return. Although nearly all Atlantic salmon life stages have been

stocked into the wild (Kennedy, 1988), many restoration programs often rely on smolt or

fry stocking (Bielak and Davidson, 1993; Gephard, 2008). As juvenile Atlantic salmon

development is primarily controlled by environmental conditions, stocked fry will

develop in response to natural environmental conditions within recipient watersheds,

whereas the development and subsequent smolt out-migration timing of stocked smolts

will be heavily influenced by hatchery rearing conditions (e.g. Duston and Saunders,

1995). Hatcheries can control juvenile development through manipulations of

photoperiod, water temperature, and food rations, allowing the timing of smolt

production to vary in response to hatchery conditions (Pennell and Barton, 1 996). Ideally,

hatchery conditions would be managed to permit out-migration of stocked Atlantic

salmon from recipient watersheds within the environmental smolt window to maximize

the survival to adulthood.

93

Several life stages of Atlantic salmon are currently being stocked in Lake Ontario

tributaries in hopes of population re-establishment. Lake Ontario historically hosted the

largest freshwater population of Atlantic salmon (Parsons 1978), which was extirpated in

the late 19th century due to direct and indirect anthropogenic pressures (Parsons, 1978;

Dunfield, 1985 Ketola et al, 2000). Conditions within Lake Ontario have improved

since the early 1900's, and the watershed currently hosts populations of many native

fishes (Christie, 1972; Mills et al, 2003), as well as a variety of exotic salmonid species

which support a large sport fishery (Crawford, 2001). A reintroduction program was

established in 2006 with the intent of restoring self-sustaining populations of Lake

Ontario Atlantic salmon in Ontario tributaries. Several Atlantic salmon life stages have

been stocked into Lake Ontario tributaries which provide the best chance for Atlantic

salmon restoration, and Chapter 2 tested the juvenile growth and survival of stocked fry

and yearlings to evaluate the performance of each stocking strategy in the effort to

reintroduce Lake Ontario Atlantic salmon.

This study evaluated the out-migration timing of fry and yearling Atlantic salmon

introduced into Cobourg Brook, a Lake Ontario tributary. Specifically, I tested the

hypothesis that stocked fry would out-migrate within the environmental smolt window,

while stocked yearlings would not due to unnatural hatchery rearing conditions. This

study also described the stream conditions and seasonal timing ofjuvenile Atlantic

salmon out-migrating from Cobourg Brook, as well as the association between out-

migration and stream temperature, level, and cumulative degree-days. This study was

conducted in unison with an evaluation of the growth and survival of stocked fry and

yearlings in Cobourg Brook (Chapter 2 of this thesis) to inform future reintroduction

94

efforts, hopefully leading to eventual reestablishment of Atlantic salmon populations

within Lake Ontario tributaries.

95

Methods

This study was implemented on Cobourg Brook, which is one of three Lake

Ontario watersheds selected for the initial stage of the current Lake Ontario Atlantic

salmon restoration effort (Grieg et al, 2003). A description and illustration of the study

site is provided in Chapter 2.

Since 2006, Cobourg Brook has been stocked with a variety of life stages of

Atlantic salmon as part of the current reintroduction program (Table 2-1). Due to the size

and diverse nature of the restoration effort, several hatchery facilities have been used in

the restoration effort (Table 2-1). This study investigated the comparative out-migration

timing of fry (post-feeding fry) and yearlings. Release groups were marked using a

variety of methods (adipose clips, floy tags, and genetic tracking) to enable post-stocking

recognition of stocked life stages (Table 2-1). All stocking treatments investigated in this

study were from the OMNR LaHave River brood stock. Genetic tagging relied on a

structured captive breeding program that employed PIT tagging and microsatellite

genotyping of reproductive adults, followed by 1 :1 mating of females and males. The

resultant offspring that were destined to be stocked at a common life stage were reared

communally and tracked in captivity until stocking.

Yearlings stocked in 2007 were reared at Normandale Fish Culture Station (FCS)

using ambient groundwater temperatures with heated water supplemented only to aid

transfer to exogenous feeding. A subsample of yearlings were marked with 90mm floy

tags, and/or adipose clips on April 4 (n= 2 882 and 3 001, respectively) and stocked on

April 24 2007 (Table 2-1). Prior to stocking in 2007, a subset of yearlings were

transferred to saltwater at Normandale FCS to test smolt status, which suggested smolt

96

status was not attained prior to stocking (G. Durant, OMNR, pers. comm). Yearlings

stocked in 2008 were reared at the Sir Sandford Fleming College (SSFC) aquaculture

facility using a water recirculation system and smolt status was visually confirmed

(lacking parr marks, silver colouration) prior to stocking on April 16 2008 (R. Enslow,

SSFC, pers. comm.). Prior to stocking in 2007 and 2008, yearlings weighed 24.3 and 72.2

grams respectively. In both 2007 and 2008 yearlings were stocked at point locations on

the lower main stem of Cobourg Brook (Figure 2-1), while fry were scatter-stocked in a

high gradient, large substrate portion of Baltimore Creek at a target density of 8 fish-m-2

in late May 2007 (Figure 2-1).

Salmonines out-migrating from Cobourg Brook during the spring of 2007 and

2008 were captured using a smolt trap (modified fyke net) as part of a mark-recapture

experiment (Chapter 2). Only Atlantic salmon captured for the first time in the mark-

recapture experiment were included in the present evaluation of out-migration timing. To

identify the origin of Atlantic salmon out-migrating from Cobourg Brook in 2007, all

individuals captured in the smolt trap were inspected for adipose clips or tags. In 2008,

stocking ages of out-migrating smolts were assessed by age interpretation via scale

pattern analysis and genetic parentage and population assignment, as described in

Chapter 2. Due to the high abundance ofjuvenile Atlantic salmon captured in 2008, daily

smolt trap catches were sub-sampled to determine the origin of out-migrants, and

extrapolated to estimate the total abundance of each life stage captured in the smolt trap

each day.

An automated data logger (Onset) was attached to the smolt trap, and stream level

data was collected using an automated datalogger by the Department of Fisheries on

97

Oceans Canada from a point location within the lower reaches of Cobourg Brook (Figure

2-1). For each Atlantic salmon captured in the smolt trap, daily means and standard

deviation were calculated for stream temperature, change in stream temperature, stream

level, change in stream level, and cumulative degree-days. Cumulative degree-days were

calculated using the method described by Zydlewski et al., (2005), where the mean daily

stream temperature was summed from January 1 until the date of capture. Temperature

data was not available for winter 2007, thus cumulative degree-days were calculated from

time of stocking up until out-migration.

The out-migration timing of stocked fry and yearling Atlantic salmon was

described in reference to the stream conditions and dates over which the majority of

individuals were encountered. The association between smolt trap catch and stream

temperature, change in stream temperature, stream level, change in stream level, and

cumulative degree-days were assessed for each stocking treatment using Spearman rank

correlations, as data were not normally distributed.

Environmental smolt windows were created for Cobourg Brook in 2007 and 2008

to evaluate if Atlantic salmon out-migration occurred during periods suitable for survival.

Smolt window predictions were constructed assuming wild Atlantic salmon smolt out-

migration typically occurs during suitable environmental conditions for survival (Webb et

al., 2008), and potential for survival is normally distributed (Dempson and Stansbury,

1991; Chaput et al, 2002; Clément et al, 2007). The literature was examined to

determine the typical timing of Atlantic salmon smolt out-migration in regards to

temperature at the onset of out-migration, and the temporal duration (number of days)

over which Atlantic salmon smolts typically out-migrate (Table 3-1).

98

Smolt windows were created for Cobourg Brook in 2007 and 2008 by predicting

the start and end dates of smolt trap catch, which were based values from previous studies

of Atlantic salmon smolt migration timing (Table 3-1). Within these temporal boundaries,

smolt trap catch was predicted to be normally distributed. Therefore, the expected

Atlantic salmon catch was predicted for each day of spring smolt trapping. Smolt trap

catch is described in terms of the proportion of the total annual catch captured each day.

The temporal duration of smolt windows constructed for Cobourg Brook was generous,

as out-migration was predicted to begin once stream temperatures exceeded 2°C

following the spring freshet, the lowest observed value (Table 3-1); and smolt windows

were predicted to last 66 days, a long period relative to published values (Table 3-1).

Smolt trap catch formed a normal distribution over the predicted time period, peaking

during periods of increased stream level, with a mean of 0.08 fish and a standard

deviation 0.3 fish. The constructed smolt windows for Atlantic salmon out-migrating

from Cobourg Brook in 2007 and 2008 are shown in Figures 3-2a and 3-4a respectively.

Statistical differences between the observed Atlantic salmon smolt trap catch from

Cobourg Brook, and catch predictions based on hypothesized smolt windows were

compared using chi-square tests. Predicted catches based on hypothesized smolt windows

were scaled to total observed catch of each stocking treatment to facilitate statistical

comparison. Daily predicted and observed catches were pooled into 7-day strata, and

those with less than 1 expected or observed capture were pooled with neighboring strata.

Yates' correction was used to analyze 2008 data as greater than 20% of observations

contained values less than 5 (Zar, 1984).

99

Table 3-1: Summary of published initiating temperature and duration of wild Atlantic

salmon smolt migration.

Study Initiating Temp Duration (days) LocationJonsson and Ruud- 5.8-11 .2Hansen, 1985Antonsson andGudjonsson, 2002Zydlewski et al, 2005 7.5-12.5Whalen et al, 1999 5Hvidsten étal, 1995 1.7-4.4Clement et al, 2007 7-8Chaputeía/., 2002Bryne et al, 2003 5.1-5.4

30

10 (for 5 days) <10-50

36-4720-404336-443792-103

Norway (wild)

Iceland (wild)

USA (hatchery)USA (hatchery)Norway (wild)Canada (Wild)Canada (wild)Ireland (wild)

100

Results

Spring 2007

In 2007, the smolt trap was continuously operational from April 24 to July 21

with the exception of May 17 and July 14-17 due to extreme weather events (Figure 3-1).

During this interval, the smolt trap captured a total of 524 unmarked juvenile Atlantic

salmon and 271 marked yearlings (Figure 3-1). An overview of the stream conditions

during the expected and observed out-migration of Atlantic salmon from Cobourg Brook

is provided in Table 3-2.

Three periods of high catch occurred during 2007 smolt trapping. The first period

occurred in late April shortly after yearling stocking and consisted mostly of marked

yearlings (Figure 3-1). The second period of high catch occurred in mid-May once stream

temperatures exceeded 1 00C, and with maximum catches coinciding with a rise in stream

level on May 17. Individuals within the second period of high smolt trap catch were

mostly unmarked (Figure 3-1). The final period of high smolt trap catch in 2007 occurred

from mid-June to late-July during declining stream levels once stream temperatures

approached or exceeded 200C. Of all marked yearlings captured in spring 2007, 74%

were captured within this final period of out-migration occurring from June 13 to July 9,2007.

Mean stream temperature and cumulative degree-days during out-migration of

marked yearlings in 2007 were 16.30C ±2.2, and 882. 1± 22.4 cumulative degree-days,

respectively. There were no significant correlations (P<0.05) between smolt trap catch of

either marked yearlings or unmarked individuals and stream level, change in stream level,

stream temperature, change in stream temperature, or cumulative degree-days.

101

The hypothetical Atlantic salmon smolt window for Cobourg Brook in 2007

predicted 99% of total smolt trap catch would occur from April 24 to May 24, peaking on

May 9. Catch of marked yearlings displayed a significantly different distribution than the

predicted smolt window (? =174,169, p<0.001, v=7; Figure 3-2c). Although the second

pulse of emigration coincided with the predicted smolt window, life stage at stocking

could not be determined for the out-migrating smolts, as these fish pre-dated the genetic

tagging program and scale pattern analysis yielded inconclusive results (Appendix 1).

102

(? ) 9jn}ejediu9i lub9jjsoCM

OIO

(LU) |T?T| LUB8JJSinö

oö

CO CU

S < ü> BCO (U. co co

(? ?

W W

O¡-I?BgCO

o<en<?

-4—»

OßC•a???GO

?COO

ce(U?-·

?Cce

O-4—»(?OOh

OB??

CO

cecoO

CCS

MOOVh

CQS-I

O-ßO?<+?O

gg

OO

en

tíU

teo

_gÏO

-oCce

^

3i—>?Ccer^

>¦>ce

ceöo

ü "3CU

O

OC

ce

oS(U

-4—»

O

OOO)

-í—»ce?cu

-*—»ceo

-?

O

oO

ipjeo dBJi nous ä|ibq I

CU

M)

Vhce

r^ooCN

ce<u>>

U-IO

ce-a

? / \ I ? V-V V

Atlantic salmon catchStream level

--------- Stream temp

o.o ?

20

- 15

- 10

5

- 0

- -5

10

B

r 1.5

0.0 «

Apr

». 1 . „ -t. . ? ¿„a, . ?JOlilMIl flllUIMay

L nMjlìfm fl.f yìjJil.ikjiD Ja_

1.5

0.0 ?

-0.5

Apr May Jun

Date

Jul Aug

Figure 3-2: Cobourg Brook Atlantic salmon smolt trap catch in 2007; (A) expected using

hypothetical smolt windows, (B) observed marked yearlings, and (C) observed unmarked

juvenile Atlantic salmon

104

?Ccdxs¡CCd(UgOOe

o

1/3

OO

enOO

o-OOOsO

CO

O-4—»C

<(U

'2(U>

•t—j

?(U

OO

-4—·(?

OCO

'-+—»ceVhOO

3OOO

_g?

1/3CO

?3COO

Co(U

ri?

mCU

3«

CCj(U>>?(U

??

r-??CN

CO

00

3O

«4-1OVhco(U>,

-*-»CcoC

'go?

<U

(U4=H»—»

î-t

?(U?'>OVhC3h(UVhCd(/3CO

COO

(S<uVh

-4—»GO

t/3?3

CCJ"En(S>

£O

>(U

CO?Ccd

OOOO(N00C

fcCSO,

-O(U

OO

OO(N

oo"OO(NUjO

'CQ-c/3

IC/300

_gTce(U

-a(U

OO

OOOOCN

t—~OOCNOO

_gQ-C/3

cdO-l>-OO(N

Vhcd(U

tí(Ugcd<u

Ci

oUecd<UVh

-4—»

(U

(U

O

(U

cd

(+H

?

oO

(UO(U?-??

(U

BO_gTh1cd(U

?>CX)OO(N

00C

-3oO

GO

?(U

-4—»O(UQ-X?

00_gcd?

?^4

(U

r—??(N

(U-4-4O(UOhX?

VO

??r-

?

>?

-?ioOO

On

(NCN-Ht--IO

-H

C—

CN-H

VO

(N-tí

CNOÖ-Hr-tj-IO

citJ-oö+1tj-ooIO

(NCNOO

Ol

IO Tf

(U

cdVh(UQ-

IZ)

cd

coO

VOmoo-HTfOvm

VOClOÖ-Htt(NIO

>

cd<u

cd

CNtJ"OOO-H

XClci

V~~VOOö-H

^ 3;^h oo oö O

¦ -H

XVOOn??

o??Oö-H

??TfOO-Hoo

ClA fiO ID

*§^- Ot?-" -H

C Cicd /—- cd<u rj <u

(U>(U

cd(UVh

(U00Cicd

O

ciVOCNHH

o

-HciIO

»oCNciHH

HHO civo ciR (?-?

VOTj-OÖ-HOn

VOClci-HIOCNTj-

(UVhCi

-4—»cdVh(UQ-g?

gcd<uVh

?00G _cd r ?

(NF--

-tíOnOONCN

CN-HIOCN

OOCN

OOIOCl-HClVOTf'On

-H2 On. OOr- ¦00CN

VOCl

CNCN <U

O .

IOO

00C

CN MOOOO

IO

5 CU 3-H o .SCN C M

Cl -^CN M(N

(/3

cd?

I(U(UVh00(U?

(U>

c¡g3Ocicd(U

Spring 2008

During spring 2008, the smolt trap fished daily from April 16 to August 13 with

the exception of May 4-5, June 6-7, June 13, July 22-25, and August 7-8 due to extreme

weather events (Figure 3-3). During this time a total of 2,753 juvenile Atlantic salmon

were captured, with genetically marked yearlings and fry comprising an estimated 928

and 32 individuals, respectively. During early-season test netting three juvenile Atlantic

salmon of unknown origin were captured on March 27, but catches did not resume until

April 16, 2008; the first day of yearling stocking. After a peak in smolt trap catch shortly

after yearlings stocking, catch gradually subsided and no salmon were captured beyond

July 14 (Figure 3-3).

Similarly to the data from 2007, juveniles that were stocked at different life stages

showed markedly different emigration times. In contrast to the marked yearlings that

were stocked in 2007, 88% of the total catch of yearlings that were stocked in 2008 were

captured within 14 days of stocking (Figure 3-3). This period of stocked yearling out-

migration occurred predominantly prior to temperatures exceeding 100C, and during

falling stream levels. Stocked yearling out-migration occurred during a mean stream

temperature of 8.57±2.27, and 94.63±3.58 cumulative degree-days after January 1 . By

contrast, 91.5% of the captured smolts that were stocked as fry in 2007 were caught

between April 27 and May 16 2008 (Figure 3-3). This period of stocked fry out-migration

occurred during a period of frequent increases in stream level once temperatures

exceeded 100C (Figure 3-3). Stocked fry were captured at mean stream temperature of

10.57±1.76, and 279.09±14.72 cumulative degree-days past January 1. Throughout the

duration of spring 2008 smolt trapping, no significant correlation (P<0.05) was found

106

between smolt trap catch of either stocked yearlings or stocked fry and stream level,

change in stream level, stream temperature, change in stream temperature, and

cumulative degree-days.

The hypothetical Atlantic salmon smolt window for Cobourg Brook in 2008

predicted 99% of total smolt trap catch would occur from April 21 to May 26, with catch

peaking on May 8 (Figure 3-4). Catch of marked yearlings displayed a significantly

different distribution than the expected smolt window (?2=58283.2, p<0.001, v=7; Figure3-4b), whereas no difference was found between catch of fry-stocked smolts and the

hypothetical smolt window (?2=6.83, 0.25>p<0.10, v=5; Figure 3-4c).

107

(O ) 8jn}BJ9dui8} LueajjsoCM O

(lu) |???| Lueajjs

(D F

y y S SCO CO CO W

oID

o oCO

OCN

O

Ljoieo pajewijsa |ejo} j.o %

OOooCNU-IOS-H(US

"SbC

- i-HS-HDh(?

<u^-»

ÖD_g?(?CO

OO

i(U

-secdGO

cden

O?—»Ccd

<4-?O

-ÖO

?—»cdO

OgIZl(U

-4—»1H-HO>>c3S

¿}?

CU

SOJD

Ccd

irf(N

ICNCN>>3?—?

en

I

?e3

cd

s scdC

_oH—»cdÍ-H<UCUO

H—·OCencd

cd?-

?gen(U

-tíH-J

O-S&

_g(?(U

H—·cd

-O(U

H—»cdO

-S

00

OOS-.

03DOS-

O

OU

gOO

OJD_g?

CD>>

Ch-hO>.cd?H-JenS-H

(U

cdo?

Ofc!cd<u

-tí-4—»

?Ccd

3>

OOO

O

8Bo

0.8

h 0.6

0.4

h 0.2

Atlantic salmonCatchStream level

Stream temperature

o., e0,0 <g-0.2 »

-0.4

20

15

10

5

h 0

-10

Oo

fQlese(O?

¡?

Apr May Jun Jul Aug

B

ë 30

0.0 «

'LxAu h

20

15Oo

io if

5 ef

-5

-10

e(Cf

ft

3 15o

0.0 g

- 20

- 15

- 10

- 5

- O

... .5

- -10

O?

f

25

Figure 3-4: Cobourg Brook Atlantic salmon smolt trap catch in 2008; (A) expected using

hypothetical smolt windows, (B) observed stocked yearlings, and (C) observed stocked

fry.

109

Discussion:

Results from this study support the original hypothesis that stocked Atlantic

salmon life stages would differ in out-migration timing and cues, and that hatchery

rearing practices would alter out-migration behavior. As predicted, Atlantic salmon fry

stocked into Cobourg Brook out-migrate within the environmental smolt window, while

stocked yearlings would not. The observed out-migration timing also suggests that

stocked fry respond to wild smolt migration cues, while stocked yearlings do not. This

study provides evidence that yearlings stocked in 2007 did not out-migrate as smolts,

rather they were forced to out-migrate due to unsuitably high temperatures. Atlantic

salmon released as yearlings likely display lower survival during out-migration or upon

entry into Lake Ontario compared to stocked fry because out-migration timing was out of

phase with the local environmental smolt window, and/or smolt status was not attained

(2007).

Environmental Factors and Out-migration Timing

The environmental smolt window hypothesis (McCormick et ai, 1998), suggests

optimal survival occurs when juvenile Atlantic salmon out-migrate during optimal

environmental conditions. For example, prédation may be decreased by out-migrating in

the spring of the year during high flows and turbidity, or the physiological stress

associated with transitioning between stream and lake habitats may be reduced in the

spring due to the relatively low the temperature differential. Therefore, the out-migration

timing observed in this study suggests stocked yearlings likely display lower survival

during, or shortly after out-migration compared to stocked fry.

110

The out-migration of Atlantic salmon originally stocked as fry occurred once

stream temperatures exceed 100C in the spring of 2008, suggesting a temperature

threshold is required to initiate smolt out-migration. Numerous other studies have

observed similar behavior (White, 1939; Solomon, 1978; Whalen et al., 1999; Antonsson

and Gudjonsson, 2002; Bryne et al, 2003), suggesting that fish stocked as fry respond to

the natural out-migration cues in Cobourg Brook. In addition, it appears the group of

unmarked Atlantic salmon which out-migrated from Cobourg Brook in 2007 just as

stream temperature exceeded 1 00C were likely previously stocked at an early life stage

(e.g., fry or fall fingerlings) and were also responding to the natural smolt migration cues

in Cobourg Brook. However, the out-migration of stocked yearlings showed no such

relationship, suggesting these individuals do not respond to the natural smoking cues in

Cobourg Brook, and may not have displayed 'true' smolt migration.

The insignificant correlation between smolt trap catch and stream conditions

observed in this study is quite surprising. Increases in stream temperature and level are

considered releasing cues which trigger Atlantic salmon smolt migration (McCormick et

al, 1998), and numerous studies have found Atlantic salmon smolt migration to be

correlated to increases in stream temperature (White, 1939; Solomon, 1978; Jonsson and

Ruud-Hansen, 1985; Hvidsten et al, 1995; Antonsson and Gudjonsson, 2002; Byrne et

al, 2003), or flow (Allen, 1944; Solomon, 1978; Hesthagen and Garnas, 1983; Jonsson

and Ruud-Hansen, 1985; Hvidsten and Hansen, 1988; Dempson et al, 1996; Whalen et

al, 1999). This study found no such relationships. The insignificant correlation between

stocked yearling out-migration and stream conditions provides further evidence that these

individuals did not display 'true' smolt migration. However, because stocked fry out-

111

migrated during the environmental smolt window one would expect they would likewise

out-migrate in response to typical releasing cues. The insignificant correlation between

fry smolt trap catch and stream conditions could have been caused by either no actual

relationship, or low sample sizes, as only 32 stocked fry were captured in spring 2008.

Smoltification and Out-migration Timing

It appears not all Atlantic salmon stocked into Cobourg Brook out-migrated as

smolts. Atlantic salmon smolt development and migration timing can be generalized into

a three-part process. First individuals must attain a size related development stage

(typically 95-1 00mm fork length) before the end of the natural growing season (Nicieza

et al, 1991; Whitesel, 1993; Pearlstein et al, 2007). Secondly, individuals begin a

physiological, behavioral, and morphological metamorphosis (e.g. smoltification) to

become 'primed' for migration in response to environmental cues (e.g. photoperiod and

temperature) in the spring of the year (Duston and Saunders, 1995; Duncan and Bromage,

1998; McCormick et al, 1998; Berrill et al, 2006). Thirdly, smolts out-migrate in

response to increasing stream temperature (Zydlewski 2005), and flow (White, 1939;

Solomon, 1978; Whalen et al, 1999; Antonsson and Gudjonsson, 2002; Byrne et al,

2003). In turn, the observed out-migration timing of stocked fry from Cobourg Brook

which overlapped with the typical Atlantic salmon smolt out-migration timing suggests

these individuals displayed true smolt migration. Although the out-migration of yearlings

stocked in 2008 was like typical smolt out-migration they had already taken on a smolt-

like appearance (silver) within the SSFC hatchery and out-migrated immediately after

stocking, suggesting these individuals completed the entire 3-part process of smolting in

112

captivity and were in a migratory state upon release. Yearlings stocked in 2007 were of

sufficient size (approximately 125mm) however, the photoperiod conditions (mid

summer), and thermal conditions (882.06± 22.36 cumulative degree days) were atypical

of smolting process, suggesting yearlings stocked in 2007 were not smolts when they out-

migrated.

Smolts are behaviorally, morphologically, and physiologically adapted for

survival during downstream migration and upon entry into the pelagic environment of

lakes and oceans. Such adaptations include negative rheotaxis, schooling, decreased

territorial behaviors, scale silvering, increased buoyancy, increased metabolic rate, and

increased scope for growth (McCormick et al, 1998). Therefore, because smolt status

was not attained during out-migration, yearlings stocked in 2007 likely displayed

relatively poor survival and may be an inferior restoration technique compared to fry

stocked in 2007 and yearlings stocked in 2008.

Although yearlings stocked in 2007 were not prepared for out-migration, high

stream temperatures (daily max 25°C) likely forced them out of Cobourg Brook in mid-

summer. Atlantic salmon parr show signs of thermal stress once temperatures exceed

200C (Garside, 1973; Elliott, 1991; Elliott and Elliott, 1995), and when faced with such

conditions parr may move to seek thermal refugia (e.g. Sauter et al, 2001 ; Sutton et al.,

2007; Tate et al, 2007). Therefore, the mid-summer out-migration of stocked yearlings in

2007 was likely forced due to unsuitably high stream temperatures. In addition to marked

yearlings out-migrating during mid-summer in 2007, unmarked juvenile Atlantic salmon

were also captured. These individuals appeared visually similar to yearlings stocked in

2007 (R.Bobrowski Pers Obs), and likely represented the majority of unmarked yearlings

113

stocked in spring 2007. The mid-summer out-migration of stocked yearlings was very

unusual, and no other studies to my knowledge have reported similar behaviour.

However, out-migration of non-smolt Atlantic salmon has been observed during fall or

winter and may also be caused by unsuitable abiotic conditions during this time of year

(Youngson et al, 1983; Riley et al, 2002; Pinder et al, 2007), and the survival to

adulthood of non-smolt out-migrants appears to be low (Riley et al, 2009).

The observed differences in yearling out-migration timing between 2007 and

2008 were likely due to captive rearing conditions. Yearlings stocked in 2007 were reared

at Normandale FCS at ambient ground water temperatures and released at a mean body

size of 24.3 g (approximately 125mm FL), with out-migration occurring predominantly

from mid-June to mid-July. In contrast, yearlings stocked in 2008 were reared at SSFC

aquaculture centre using a water re-circulation system at relatively warmer temperatures

and were stocked at a body size of 154g (approximately 255mm TL), and out-migration

occurred predominantly within 2 weeks of stocking. The warmer water employed at the

SSFC aquaculture centre may have stimulated growth and therefore attainment of smolt

status prior to stocking, resulting in mass out-migration shortly after stocking. In contrast

the growth opportunity within Normandale FCS may be insufficient to produce spring

age-1 smolts. However, the photoperiod and temperature regime of Normandale FCS

may have also effected when individuals attain smolt status.

Current yearling culture and stocking practices used in the Lake Ontario

restoration effort may inhibit the natural homing instinct which allows adult Atlantic

salmon to return to their natal stream to reproduce. It appears that juveniles imprint the

location of their natal stream based on chemical odors (Hasler and Scholtz, 1983; Hansen

114

et al, 1993), although the precise time when odors are imprinted remains debatable

(McCormick et al, 1998). However, numerous studies support the hypothesis that

olfactory imprinting occurs during the parr-smolt transformation (Morin et al, 1989;,

Morin et al, 1989b; Carlin 1969; Potter and Russell, 1994; Sutterlin et al, 1982). Current

rearing practices aim to stock yearling Atlantic salmon into Lake Ontario streams after

parr-smolt transformation, and therefore yearlings will seek the olfactory cues associated

with the hatchery for reproduction instead of their release site. Further research is

recommended to investigate the homing capability of stocked yearlings as it may severely

reduce the abundance of adults returning to streams selected for restoration efforts.

The cumulative degree-days prior to the out-migration of stocked fry observed in

this study should provide a strong reference of the thermal regime required to attain smolt

status in captivity. This study found that Atlantic salmon that were stocked as fry out-

migrate 279.1 ± 14.72 degree-days after January 1. Fry were stocked at an early life stage

and subsequently grew and developed in response to the environmental conditions within

Cobourg Brook. Cumulative degree-days have been found to be an accurate predictor of

smolt migration timing, even when thermal regimes vary (Zydlewski et al, 2005), thus

the cumulative-degree-days for stocked fry out-migration should be similar to those

occurring in the captive and wild LaHave River population.

This study has identified that yearling Atlantic salmon released into Cobourg

Brook did not out-migrate during the local environmental smolt window, and suggested

that stocked yearling survival during out-migration or upon entry into Lake Ontario may

be poor compared to stocked fry. However, the absolute success of these stocking

methods should be compared by determining the abundance of reproductive adults and

115

production of wild recruits from either stocking strategy. Overall, this study has provided

information to consider when refining the current approach to re-establishing Atlantic

salmon populations in Lake Ontario tributaries.

116

References

Allen, K. R. (1944). Studies on the biology of the early stages of the salmon {Salmosalar). 4. Smolt migration in the Thurso Riverin 1939. Journal ofAnimal Ecology 13,63-85.

Antonsson, T. & Gudjonsson, S. (2002). Variability in timing and characteristics ofAtlantic salmon smolt in Icelandic rivers. Transactions ofthe American Fisheries Society131, 643-655.

Araki, H., Cooper, B. & Blouin, M. S. (2007). Genetic effects of captive breeding cause arapid, cumulative fitness decline in the wild. Science 318, 100-103.

Bielak, A. T. & Davidson, K. (1993). New enhancement strategies, an overview. InSalmon in the sea, and new enhancement strategies, pp. 267-320: Fishing News Books.

Brown, C. & Day, R. L. (2002). The future of stock enhancements: lessons for hatcherypractice from conservation biology. Fish and Fisheries 3, 79-94.

Brown, C. & Laland, K. N. (2003). Social learning in fishes: a review. Fish and Fisheries4, 280-288.

Byrne, C. J., Poole, R., Rogan, G., Dillane, M. & Whelan, K. F. (2003). Temporal andenvironmental influences on the variation in Atlantic salmon smolt migration in theBurrishoole system 1970-2000. Journal ofFish Biology 63, 1552-1564.

Carcao, G. (1987). Atlantic salmon in the Great Lakes basin: a history of its extirpationand attempted restoration. In Unpublished manuscript, H.A. Regier. University ofToronto, Department of Zoology.

Chaput, G., Hardie, P., Hayward, J., Moore, D., Shaesgreen, J. & NSPA (2002).Migrations and biological Characteristics of Atlantic salmon from the NorthwestMiramichi River, 1998 to 2000. In Canadian Technical Report ofFisheries and AquaticScience 2415: Department of Fisheries and Oceans.

Christie, W. J. (1972). Lake Ontario: effects of exploitation, introductions, andeutrophication on the salmonid community. Journal ofthe Fisheries Research Board ofCanada, 913-929.

Clément, M., Chaput, G. & Leblanc, P. (2007). Atlantic salmon {Salmo salar) smoltmigration from the Margaree River, 2001 to 2003. In Canadian Technical.Report ofFisheries, and Aquatic Sciences, p. 60.

Crawford, S. S. (2001). Salmonine introductions to the Laurentian Great Lakes: anhistorical review and evaluation ofecological effects. Canadian Special Publication ofFisheries and Aquatic Science. No. 132.

117

Dempson, J. B., Oconnell, M. F. & Shears, M. (1996). Relative production of Atlanticsalmon from fluvial and lacustrine habitats estimated from analyses of scalecharacteristics. Journal ofFish Biology 48, 329-341.

Dempson, J. B. & Stansbury, D. E. (1991). Using partial counting fenses and a two-sample stratified design for mark-recapture estimation of an Atlantic salmon smoltpopulation. North American Journal ofFisheries Management 11, 27-37.

Duston, J. & Saunders, R. L. (1995). Advancing smolting to autumn in age 0+ Atlanticsalmon by photoperiod, and long-term performance in sea water. Aquaculture 135, 295-309.

Elliott, J. M. (1991). Tolerance and resistance to thermal-stress in juvenile Atlanticsalmon. Freshwater Biology 25, 61-70.

Elliott, J. M. & Elliott, J. A. (1995). The effect of the rate of temperature increase on thecritical thermal maximum for fry of Atlantic salmon and brown trout.. Journal ofFishBiology 47, 917 -919.

Friedland, K. D., Hansen, L. P. & Dunkley, D. A. (1998). Marine temperaturesexperienced by postsmolts and the survival of Atlantic salmon, Salmo salar L., in theNorth Sea area. Fisheries Oceanography 7, 22-34.

Garside, E. T. (1973). Ultimate upper lethal temperature of Atlantic salmon (Salmosalar). Canadian Journal ofZoology 51, 898-900.

Gephard, S. (2008). Restoring Atlantic Salmon (Salmo salar) to New England. In SavingBiological Diversity (Askins, R. A., G.D., D., Visgilio, G. R. & Whitelaw, D. M., eds.),pp. 75-85. Boston: Springer Science & Business Media.

Hansen, L. P. (1987). Growth, migration and survival of lake reared juvenile anadromousAtlantic salmon Salmo salar L. Fauna Norvegica, Series A. 8, 29-34.

Hansen, L. P. & Jonsson, B. (1991). The effect of timing of Atlantic salmon smolt andpost-smolt release on the distribution of adult return. Aquaculture 98, 61-67.

Hesthagen, T. & Gamas, E. (1983). Migration of Atlantic salmon smolts in River Orkla,central Norway in relation to management of a hydroelectric facility. North AmericanJournal ofFisheries Management 6, 376-382.

Hoar, W. S. (1988). The physiology of smolting salmonids. In Fish Physiology (Hoar, W.s. & Randall, D. J., eds.), pp. 275-343. New York: Academic Press.

Hvidsten, N. A. & Hansen, L. P. (1988). Increased recapture rate of adult Atlantic salmon(Salmo salar) stocked as smolts at high water discharge. Journal ofFish Biology 32, 153-154.

118

Hvidsten, N. ?., Jensen, A. J., Vivas, H. & Bakke, O. (1995). Downstream migration ofAtlantic salmon smolts in relation to water flow, water temperature, moon phase, andsocial interaction. Nordic Journal ofFreshwater Research 70, 38-48.

Jokikokko, E. & Mantyniemi, S. (2003). The survival of stocked Atlantic salmon smoltsduring sea run and the timing of migration in the river Simojoki, northern Finland.Aquaculture 219, 431-444.

Jonsson, N. & Ruud-Hansen, L. (1985). Water temperature as the primary influence ofseaward migrations of Atlantic salmon (Salmo salar)smolts. Canadian Journal ofFisheries and Aquatic Sciences 42, 593-595.

Kallio-Nyberg, I., Jutila, E., Jokikokko, E. & Saloniemi, I. (2006). Survival of rearedAtlantic salmon and sea trout in relation to marine conditions of smolt year in the BalticSea. Fisheries Research 80, 295-304.

Kallio-Nyberg, I., Jutila, E., Saloniemi, I. & Jokikokko, E. (2004). Association betweenenvironmental factors, smolt size and the survival of wild and reared Atlantic salmonfrom the Simojoki River in the Baltic Sea. Journal ofFish Biology 65, 122-134.

Kennedy, G. A. (1988). Stock enhancement of Atlantic salmon. In Atlantic salmon:planningfor thefuture (Mills, D. & Piggins, D., eds.). Portland, Oregon: Timber Press.

Ketola, H. G., Bowser, P. R., Wooster, G. A., Wedge, L. R. & Hurst, S. S. (2000). Effectsof thiamine on reproduction of Atlantic salmon and a new hypothesis for their extirpationin Lake Ontario. Transactions ofthe American Fisheries Society 129, 607-612.

Lynch, M. & O'Hely, M. (2001). Captive breeding and the genetic fitness of naturalpopulations. Conservation Genetics 2, 363-378.

McCormick, S. D., Cunjak, R. A., Dempson, B., O'Dea, M. F. & Carey, J. B. (1999).Temperature-related loss of smolt characteristics in Atlantic salmon {Salmo salar) in thewild. Canadian Journal ofFisheries and Aquatic Sciences 56, 1649-1658.

McCormick, S. D., Hansen, L. P., Quinn, T. P. & Saunders, R. L. (1998). Movement,migration, and smolting of Atlantic salmon {Salmo salar). Canadian Journal ofFisheriesand Aquatic Sciences 55, 77-92.

McCormick, S. D., Shrimpton, J. M. & Zydlewski, J. D. (1996). Temperature effects onosmoregulatory physiology in juvenile fish. In Global warming: implicationsfor marineandfreshwaterfish (Wood, C. M. & McDonald, D. G., eds.), pp. 279-301.

Mills, E. L., Casselman, J. M., Dermott, R., Fitzsimmons, J. D., Gal, G., Holeck, K. T.,Hoyle, J. ?., Johannsson, O. E., Lantry, B. F., Makarewicz, J. C, Millard, E. S.,Munawar, I. F., Munawar, M., O'Gorman, R., Owens, R. W., Rudtsam, L. G., Schaner, T.& Stewart, T. J. (2003). Lake Ontario: food web dynamics in a changing ecosystem(1970-2000). Canadian Journal ofFisheries and Aquatic Sciences, 471-491.

119

Nicieza, A. G., Brana, F. & Toledo, M. M. (1991). Development of length bimodality andsmoking in wild stocks of Atlantic salmon, under different growth conditions. Journal ofFish Biology 38, 509-523.

Pearlstein, J. H., Letcher, B. H. & Obedzinski, M. (2007). Early discrimination ofAtlantic salmon smolt age: Time course of the relative effectiveness of body size andshape. Transactions ofthe American Fisheries Society 136, 1622-1632.

Pennell, W. & Barton, B., eds. (1996). Principles ofSalmonind Culture. Amsterdam, TheNetherlands: Elsevier Science.

Pinder, A. C, Riley, W. D., Ibbotson, A. T. & Beaumont, W. R. C. (2007). Evidence foran autumn downstream migration and the subsequent estuarine residence of 0+yearjuvenile Atlantic salmon Salmo salar L., in England. Journal ofFish Biology 71, 260-264.

Riley, W. D., Eagle, M. O. & Ives, S. J. (2002). The onset of downstream movement ofjuvenile Atlantic salmon, Salmo salar L., in a chalk stream. Fisheries Management andEcology 9, 87-94.

Riley, W. D., Ibbotson, A. T. & Beaumont, W. R. C. (2009). Adult returns from Atlanticsalmon, Salmo salar, parr autumn migrants. Fisheries Management and Ecology 16, 75-76.

Robert, A., Sarrazin, F., Couvet, D. & Legendre, S. (2004). Releasing adults versusyoung in reintroductions: Interactions between demography and genetics. ConservationBiology 18, 1078-1087.

Sarrazin, F. & Barbault, R. (1996). Reintroduction: Challenges and lessons for basicecology. Trends in Ecology & Evolution 11, 474-478.

Sarrazin, F. & Legendre, S. (2000). Demographic approach to releasing adults versusyoung in reintroductions. Conservation Biology 14, 488-500.

Sauter, S. T., Crawshaw, L. I. & Maule, A. G. (2001). Behavioral thermoregulation byjuvenile spring and fall chinook salmon, Oncorhynchus tshawytscha, duringsmoltification. Environmental Biology ofFishes 61, 295-304.

Skilbrei, O. T. (1991). Importance of threshold length and photoperiod for thedevelopment of bimodial length frequency-distribution in Atlantic salmon. CanadianJournal ofFisheries and Aquatic Sciences 48, 2163-2172.

Snyder, N. F. R., Derrickson, S. R., Beissinger, S. R., Wiley, J. W., Smith, T. B., Toone,W. D. & Miller, B. (1996). Limitations of captive breeding in endangered speciesrecovery. Conservation Biology 10, 338-348.

Solomon, D. J. (1978). Some observations of smolt migration in a chalk stream. Journaloffish biology 12.

120

Spicer, A. V., Moring, J. R. & Trial, J. G. (1995). Downstream migratory behaviours ofhatchery-reared, radio tagged Atlantic salmon smolts in the Penobscot River, Maine,VSA.Fisheries Research 23, 255-266.

Sutton, R. J., Deas, M. L., Tanaka, S. K., Soto, T. & Coram, R. A. (2007). Salmonidobservations at a Klamath River thermal refuge under various hydrological andmeteorological conditions. River Research and Applications 23, 775-785.

Tate, K. W., Lancaster, D. L. & Lile, D. F. (2007). Assessment of thermal stratificationwithin stream pools as a mechanism to provide refugia for native trout in hot, aridrangelands. Environmental Monitoring and Assessment 124, 289-300.

Webb, J., Verspoor, E., Aubin-Horth, N., Romaakkaniemi, A. & Amiro, P. (2007). TheAtlantic salmon. In The Atlantic salmon; Genetics Conservation and Management(Verspoor, E., Stradmeyer, L. & Nielson, J., eds.), pp. 17-45. Ames, Iowa: BlackwellPublishing.

Whalen, K. G., Parrish, D. L. & McCormick, S. D. (1999). Migration timing of Atlanticsalmon smolts relative to environmental and physiological factors. Transactions oftheAmerican Fisheries Society 128, 289-301.

White, H. C. (1939). Factors influencing the descent of Atlantic salmon smolts. Journalofthe Fisheries Research Board of Canada 323-327.

Whitesel, T. A. (1993). Comparison ofjuvenile Atlantic salmon (Salmo salar) reared ina hatchery and introduced into a stream: a two-size-threshold model of smoltification. InProduction ofjuvenile Atlantic salmon, Salmo salar, in natural waters (Gibson, R. J. &Cutting, R. E., eds.), pp. 239-247.

Youngson, A. F., Buck, R. J. G., Simpson, T. H. & Hay, D. W. (1983). The autimn andspring emigrations ofjuvenile Atlantic salmon (Salmo salar) from Girnock Burn,Aberdeenshire, Scotland-environmental release of migration. Journal ofFish Biology 23,625-639.

Zarr, J. (1984). Biostatistical Analysis. Toronto: Prentice-Hall Inc.

Zydlewski, G. B., Haro, A. & McCormick, S. D. (2005). Evidence for cumulativetemperature as an initiating and terminating factor in downstream migratory behavior ofAtlantic salmon (Salmo salar) smolts. Canadian Journal ofFisheries and AquaticSciences 62, 68-78.

121

Chapter 4: Summary and Synthesis

122

General Conclusions

The objective of this thesis was to evaluate the performance of Atlantic salmon

life stages being released into Lake Ontario tributaries in the current restoration program.

Fry and yearlings were introduced in a paired release stocking design in Cobourg Brook

and their in-stream growth and survival was compared (Chapter 2), as well as their timing

of migration/movement out of Cobourg Brook. The juvenile survival and out-migration

timing of Atlantic salmon is thought to be an indicator of the performance of reintroduced

fishes, as the abundance of out-migrants is typically proportional to the abundance of

returning adults (Jonsson et al, 1998) and out-migration during the smolt window is

thought to improve survival to adulthood (McCormick et al., 1998).

Chapter 2 showed that stocking yearlings produced significantly more out-

migrants than fry stocking, although the growth of stocked yearlings was poor compared

to stocked fry. Results from Chapter 2 also suggest that the growth and survival of

stocked Atlantic salmon appears to be strongly influenced by the abiotic conditions

within Cobourg Brook. The poor survival of Atlantic salmon and other resident salmonid

species suggests that severe over-winter mortality may present a considerable limitation

to the restoration of Atlantic salmon in Cobourg Brook.

Chapter 3 found that Atlantic salmon stocked at early life stages (e.g. fry)

displayed typical Atlantic salmon smolt behavior and out-migrated during the predicted

environmental smolt window for Cobourg Brook. By contrast, stocked yearlings out-

migrated during atypical times, both before and after the predicted environmental

optimum for out-migration.

123

Results from both chapters suggest that the life stage of reintroduced animals may

have a significant effect on their survival, growth and behavior, all of which may

influence the likelihood of population re-establishment. In addition to the life stage of the

release group, other factors which may influence the likelihood of re-establishing a

population are pre- or post-release management of the release group (Armstrong and

Seddon, 2008), suitable habitat (including the biotic environment), mitigation of the

original causes of extirpation (Sarrazin and Barbault, 1996, Seddon et al., 2007), and

ancestry (Olsson 2007; Meffe 1995; Lacy 1997: Figure 1-1). Determining the degree to

which each of these factors influences the performance of reintroduced individuals in

different situations would be a tremendous asset for planning future reintroduction

programmes, and is viewed as a key question in reintroduction biology (Armstrong and

Seddon, 2008).

Impediments to re-establishing Atlantic salmon populations in Lake Ontario

tributaries

Relative to the number offish stocked, few adult Atlantic salmon have so far

returned to Lake Ontario tributaries selected for restoration. In 2006 and 2007 the Credit

River, which is the largest tributary currently targeted for restoration, was annually

stocked with approximately 28,000 yearlings, 20,000 fall fingerlings, and 130,000 fry

(OMNR 2007; 2008). Growth rates in the Credit River were similar to those reported

here for Cobourg Brook (J. Bowlby, OMNR, pers comm), and the majority of individuals

are thought to out-migrate at age-1. However, adult assessment in 2008 and 2009 only

succeeded at capturing approximately 50 Atlantic salmon each year (C. Robinson,

124

OFAH, pers. comm.). Poor returns of adult Atlantic salmon were similarly reported for

prior Atlantic salmon restoration efforts in Ontario during the mid-1 800's

(MacCrimmon, 1965; Carcao, 1987), and again in the 1940s, and 1980s (Bisset et al,

1995). Low adult returns were also reported for stockings in New York state during the

1950s and 1980s (Abraham, 1983; Eckert, 2003). This low return rate is not surprising

considering the poor survival and out-migration timing of stocked Atlantic salmon

observed in this study. Therefore, it appears the current effort to re-establish Atlantic

salmon populations in Lake Ontario tributaries may hindered by one or many factors.

The results from both chapters suggest that the current life stages of Atlantic

salmon released into Lake Ontario streams may display poor survival to adulthood.

Similar to the current effort, the initial stockings in the mid- 1800s released parr (fry),

fingerlings, and yearlings into Lake Ontario streams, but this effort was not able to

conserve the declining population (Parsons, 1978; Dunfield, 1985). Other stocking efforts

throughout the last century in Ontario have primarily used fed and unfed fry

(MacCrimmon, 1954; Stanfield and Jones, 2003), and recent stockings in New York have

used yearlings (Eckert, 2003). Of the many life stages of Atlantic salmon released into

Lake Ontario streams, none have re-established a population. However, Parsons (1978)

argued that stocking fall fingerlings was the best strategy for restoring adult returns in the

late 1 800s. Although Chapter 2 suggests stocking older life stages may offer a greater

chance for population re-establishment, this finding is contingent on the assumption that

more and larger out-migrants will result in more reproductive adults (Jonsson et al, 1998,

Behmer et a/.2003), which may or may not be the case. Likewise, the conclusion that

older life stages display maladaptive behaviors (Chapter 3) is based on the assumption

125

that out-migration as smolts during the predicted environmental window may also

increase survival to adulthood (McCormick et al, 1998), which again may or may not be

the case. In addition, this study evaluated the performance of only two stocked life stages,

fry and yearlings, while numerous other life stages (e.g. fall fingerlings) were not

evaluated in this study and may present superior options for Atlantic salmon

reintroduction efforts. Given the uncertainty associated with evaluating the performance

of stocked Atlantic salmon life stages in Lake Ontario streams, further research is

recommended to further evaluate which life stage provides the greatest likelihood for re-

establishing self-sustaining populations.

A key finding in Chapter 2 suggested that severe over-winter mortality occurred

during the winter of 2007/2008 in Cobourg Brook, and may have been caused by adverse

abiotic conditions. Similarly, the hot and dry summers and cold winters from 1872-1877

was suggested as the final stressor leading to the demise of the native Lake Ontario

population of Atlantic salmon (Carcao, 1987). If severe over-winter mortality is a

common occurrence in Lake Ontario tributaries than the natural habitat may be

insufficient to allow Atlantic salmon populations to re-establish, and therefore the current

abiotic conditions in nursery streams may be an important factor limiting the likelihood

of restoring Atlantic salmon populations within Lake Ontario tributaries.

Dams were perceived as a major contributing factor in the initial demise of the

Lake Ontario population of Atlantic salmon (Dymond, 1966; Parsons, 1978; Dunfield,

1985). Dams are widespread throughout the Lake Ontario watershed, and their potential

effects on the feasibility of re-establishing Atlantic salmon populations have been

recognized (Greig et al., 2003). Likewise, the ongoing effort to restore lake sturgeon

126

throughout the Great Lakes basin has at least partially been impaired by reduced access to

spawning habitat caused by dams (Holey et al, 2000; Haxton and Findlay, 2008). Efforts

to restore walleye populations in the Great Lakes have encountered similar dam-

associated hindrances (MacDougall et al, 2007; Wilson et al, 2007). Chapter two

provided some indication that dams may be impeding the downstream

migration/movement ofjuvenile Atlantic salmon within Lake Ontario tributaries. Further

studies should investigate the influence dams on the migration/movement ofjuvenile and

adult Atlantic salmon within Lake Ontario streams, as well as the relative importance of

this factor in the effort to restore Atlantic salmon in Lake Ontario tributaries. However,

the risks of dam removal, such as permitting species invasions or damaging important

habitat should also be addressed.

Alewife is currently the primary forage base for large salmonids in Lake Ontario

(OMNR 2009), and consumption by Atlantic salmon may have contributed to the original

demise of the Lake Ontario population (Ketola et al, 2000). Thiaminase deficiency has

been recognized as a potential impediment to the current effort to restore Lake Ontario

Atlantic salmon populations (Ketola et al, 2000; Greig et al, 2003), and efforts are

underway to evaluate the influence of thiamine deficiency on the current effort to restore

Atlantic salmon populations within Lake Ontario streams (J. Fitzsimons, pers comm.).

Atlantic salmon currently stocked as yearlings into Lake Ontario streams may

display poor homing ability, likely causing straying into streams that were not selected

for restoration (Chapter 3). Streams selected for restoration in Ontario have reportedly

received few returning adults and natural reproduction has not yet been documented.

Adults straying into other streams with less suitable habitat for Atlantic salmon may

127

result in relatively poor survival to adulthood, resulting in a less than optimal use of

resources available for reintroduction. Therefore, efforts should focus on concentrating

adult returns into streams selected for restoration to maximize the likelihood of

eventually re-establishing self-sustaining populations in Lake Ontario streams. However,

the mechanism allowing adult Atlantic salmon to locate and spawn in their natal stream is

debatable, and not well known (McCormick et al, 1998), and research addressing adult

Atlantic salmon straying/homing in Lake Ontario tributaries may be particularly

beneficial to the general knowledge of this subject.

Many life history traits are largely regulated by the genetics of an individual, and

the ancestry of released animals may pose a considerable impediment in efforts to

reintroduce native taxa (Olsson 2007; Meffe 1995; Lacy 1997). To address this issue

Atlantic salmon brood stocks have been acquired from Sebago Lake, Lac St. Jean and

LaHave River for stocking into Lake Ontario streams, and future research plans to assess

the importance of ancestry of released individuals in efforts to reintroduce Lake Ontario

Atlantic salmon (Cross et al, 2007; Greig et al, 2003; OMNR 2009).

Future Direction

Together Chapters 2 and 3 evaluated the 'ideal' life stage hypothesis presented by

Coghlan and Ringler (2004), which states that older life stages display greater short-term

survival by avoiding high mortality events in the wild. This higher survival, however,

may come at the cost of maladaptive behaviors that may reduce their overall fitness

(Figure 4-1). Therefore, results from Chapters 2 and 3 supports the ideal life stage

hypothesis, as I found stocked yearlings displayed greater juvenile survival by avoiding

128

high mortality during the winter in Cobourg Brook, but yearlings displayed un-natural

out-migration behaviour. In turn, the ideal Atlantic salmon life stage would be stocked

after the winter survival bottleneck, but early enough to mitigate un-natural out-

migration. In turn, this ideal life stage would essentially be a modification of the

yearlings currently being released into Lake Ontario streams, but rather than stocking

after smoltification in captivity, they would be released after becoming primed for

migration, but prior to the onset of releasing factors in the wild (or captivity) in late

winter, perhaps March.

The ideal life stage hypothesis may provide an important tool for future

reintroduction programs. If the pattern of survival, and the effects of captive rearing on a

population selected for reintroduction can be provided or estimated, then an ideal life

stage for reintroduction can be predicted (Figure 4-1). This general framework for

releasing an appropriate life stage may be wildly applied to animal reintroduction

programs, including high-interest efforts such as the black footed ferret {Mustela

nigripes: Biggins and Godbey, 2003), California condor (Gymnogyps californianus: Utt

et al, 2008), Griffon vulture {Gypsfulvus: Legouar et al, 2008), and the Eurasian lynx

(Lynx lynx: Shadt et al, 2002). By further testing the ideal life stage hypothesis, our

understanding of the other factors which affect the performance of animals reintroduced

into their native habitat may be improved.

129

03

O

The Ideal" •

Smolt

?

?

<CDCTCD3"Q)<?"C-^W

II

Stocked life history stage (fish size)

Figure 4-1: A conceptual diagram of the ideal life stage hypothesis. Once released into

the wild young life stages may display poor survival, but possess a high degree of 'wild'

behaviors, and vice-versa for older life stages. By understanding wild survivorship, and

the behavioral effect of captive rearing, an ideal life stage may be predicted for

reintroduction efforts.

130

This ideal life stage hypothesis should be tested on a landscape-scale, This could

be achieved by replicating the methodology presented in Chapters 2 and 3 but releasing

the predicted ideal life stage in addition to other potentially suitable life stages. In

addition, multiple streams should be employed to provide replication and the abundance

of returning adults or resultant wild offspring should also be measured to provide a more

robust measure of fitness. By integrating other research topics, a project of this nature

could provide considerable information for the current effort to restore the Atlantic

salmon populations in Lake Ontario. Research of this nature could also test the

environmental smolt window hypothesis (Chapter 3: McCormick et al, 1998), as well as

the assumption that abundance of out-migrants will be proportional to the abundance of

returning adults (Chapter 2: Jonsson et al, 1998). In addition, this research could also

evaluate if the time of release (in reference to the parr-smolt transformation) effects the

survival and homing of stocked yearlings back to their (Chapter 3). Through

collaboration, this research could also gauge the effect of the aforementioned

impediments to restoring Atlantic salmon populations in Lake Ontario tributaries.

On a larger scale, many factors aside from the life stage of the release group may

affect the performance of captive bread animals in the wild. Identifying the relative

strength of some of these factors on the fitness of reintroduced animals would be

invaluable for future reintroduction programs. This could be assessed through a long-

term, experimental release of marked Atlantic salmon into Lake Ontario streams and

monitoring adult returns. Such a project would release salmon from multiple hatcheries,

life stages, and source populations (e.g. ancestries) using multiple stocking strategies into

several watersheds. The abundance of returning adults and their wild recruitment would

131

be used as a metric to gauge the relative effects of the aforementioned factors. This type

of experiment has already been informally incorporated into the Atlantic salmon

restoration effort in Lake Ontario (Cross et al, 2007; Greig et al, 2003), although a

formal description of the objectives, questions, and approach to the restoration effort is

lacking.

Efforts to rehabilitate depressed lake trout populations in the Laurentian Great

Lakes have been ongoing since the 1950s, and can provide some insight into the relative

strength of the many factors affecting the wild performance of captive bred fishes. By the

1 960s many lake trout populations in the Great Lakes were extirpated or severely

depressed (Krueger and Ebener, 2000). By refining stocking methodologies (eg. life

stage, ancestry, release method, release location) and employing post-release

management (limiting exploitation and prédation by sea lampreys) stocks of hatchery

origin lake trout have become established in many areas of the Great Lakes although little

natural reproduction has been documented outside of Lake Superior (Cornelius et al,

1995; Elrod et al, 1995; Eschenroder et al, 1995; Holey et al, 1995; Hansen et al,

1995). Krueger and Ebener (2000) provide five plausible causes for the slow recovery of

lake trout in the Great Lakes. Two are linked to the lack of suitable biotic habitat, as the

native prey base has declined and lake trout now consume the non-native alewife which

may result in thiaminaise deficiency and severe mortality in wild larval lake trout.

Likewise, the ultra-abundant, non-native alewife may be consuming young lake trout,

resulting in high mortality during young life stages. Another explanation suggests that

stocked fishes do not spawn in appropriate locations because of unsuitable stocking

locations. Lastly, the ancestry of the parent stock may be inappropriate, as stocked fish

132

may be poorly adapted for colonizing specific habitats. Therefore, the case of lake trout

restoration in the Laurentian Great Lakes suggests that increasing the survival of stocked

fishes to adulthood by refining stocking methodology and employing post-release

management may be relatively unimportant if suitable habitat is not available, or if the

parent stocks are maladapted for the selected habitat (Kreuger and Zimmerman, 2009;

Krueger and Ebener, 1995).

Efforts to restore the migratory ecotype of brook trout in Lake Superior and its

associated tributaries likewise suggest evolutionary (ancestry) or ecological (habitat)

factors may limit successful restoration. Although the potential genetic or ecological

mechanism is not fully understood, coaster brook trout are functionally different than the

typical lacustrine or riverine types (Ridgway, 2008, Wilson et al, 2008). Therefore,

stocking brook trout into historic coaster habitat may not result in the restoration of the

coaster ecotype; rather, it is possible that both genetic and ecologie conditions may need

to be present for successful restoration.

Reintroduction programs should carefully consider all of the above factors when

designing a release strategy or evaluating the feasibility of successful population re-

establishment. Human alteration of worldwide landscapes has resulted in the loss of

many species and populations of plants and animals (Pimm et al, 1995; Vitousek et al,

1997). The societal (e.g. aesthetic, ethical, and monetary) and biological (biodiversity)

value of damaged landscapes and ecosystems may be regained by reintroducing native

species (Fischer and Lindenmayer, 2000; Seddon et al, 2006). In turn, reintroducing

native species may promote the re-establishment of historic ecosystem structure and

function (Mittelbach et al, 1995; Simberloff, 1998), and test if ecological conditions

133

have been sufficiently restored to support historic species (Dobson et al, 2006; Rooney

et al, 2006). By contributing to the growing scientific discipline of reintroduction

biology, the ongoing effort to reintroduce Atlantic salmon populations in Ontario may

help regain the natural benefits of damaged ecosystems, and contribute to our knowledge

of the requirements for reintroducing taxa into their native habitats.

134

References

Abraham, W. J. (1983). A Plan to Determine the Feasibility of Atlantic salmonRestoration in Lake Ontario. In New York Department ofEnvironmental Conservation,lip.

Araki, H., Berejikian, B. A., Ford, M. J. & Blouin, M. S. (2008). Fitness of hatchery-reared salmonids in the wild. Evolutionary Applications 1, 342-355.

Armstrong, D. P. & Seddon, P. J. (2008). Directions in reintroduction biology. Trends inEcology & Evolution 23, 20-25.

Behmer, D. J., Greil, R. W., Scott, S. J. & Hanna, T. (1993). Harvest and movement ofAtlantic salmon in the St-Marys River, Michigan. Journal ofGreat Lakes Research 19,533-540.

Biggins, D. E. & Godbey, J. L. (2003). Challenges to reestablishment of free-rangingpopulations of black-footed ferrets. Comptes Rendus Biologies 326, Sl 04-Sl 11.

Bisset, J., Bowlby, J., Jones, M., Marchant, B., Miller-Dodd, L., Orsatti, S. & Stanfield,L. (1995). An Atlantic salmon restoration plan for Lake Ontario. Ontario Ministry ofNatural Resources, Peterbrough, Ontario

Carcao, G. (1987). Atlantic salmon in the Great Lakes basin: a history of its extirpationand attempted restoration. Unpublished manuscript, H.A. Regier. University of Toronto,Department of Zoology.

Coghlan, S. M. & Ringler, N. H. (2004). Comparison of Atlantic salmon embryo and frystocking in the Salmon River, New York. North American Journal ofFisheriesManagement 24, 1385-1397.

Cornelius, F. C, Muth, K. M. & Kenyon, R. (1995). Lake trout rehabilitation in LakeErie: A case history. Journal ofGreat Lakes Research 21, 65-82.

Cross, T. F., McGinnity, P., Coughlan, W., Dillane, E., Ferguson, A., Koljonen, M.-L.,O'Reilly, P. & Vasmagi, A. (2007). Stocking and Ranching. In The Atlantic salmon;Genetics Conservation and Management (Verspoor, E., Stradmeyer, L. & Nielson, J.,eds.), pp. 325-356. Ames Iowa: Blackwell Publishing.

Dobson, A., Lodge, D., Alder, J., Cumming, G. S., Keymer, J., McGlade, J., Mooney, H.,Rusak, J. A., Sala, O., Wolters, V., Wall, D., Winfree, R. & Xenopoulos, M. A. (2006).Habitat loss, trophic collapse, and the decline of ecosystem services. Ecology 87, 1915-1924.

135

Dunfield, R.W. 1985. The Atlantic salmon in the history ofNorth America. Canadianspecial publication of fisheries and aquatic science. 80: 181 p.

Dymond, J. R. (1966). The Lake Ontario Atlantic salmon {Salmo salar). UnpublishedManuscript, p. 248.

Eckert, T. H. (2003). Lake Ontario Fishing Boat Census. In 2002 Annual Report: NewYork State Department of Environmental Conservation Bureau of Fisheries Lake OntarioUnit and St. Lawrence River Unit to the Great Lakes Fishery Commission's LakeOntario Committee

Elrod, J. H., Ogorman, R., Schneider, C. P., Eckert, T. H., Schaner, T., Bowlby, J. N. &Schleen, L. P. (1995). Lake trout rehabilitation in Lake Ontario. Journal ofGreat LakesResearch 21, 83-107.

Fischer, J. & Lindenmayer, D. B. (2000). An assessment of the published results ofanimal relocations. Biological Conservation 96, 1-11.

Gephard, S. (2008). Restoring Atlantic salmon (Salmo salar) to New England. In SavingBiological Diversity (Askins, R. A., G.D., D., Visgilio, G. R. & Whitelaw, D. M., eds.),pp. 75-85. Boston: Springer Science & Business Media.

Haxton, TJ. and Findlay, CS. 2008. Variation in lake sturgeon (Acipenserfulvescens)abundance and growth among river reaches in a large regulated river. Canadian JournalofFisheries and Aquatic Sciences 65: 645-657

Hansen, M. J., Peck, J. W., Schorfhaar, R. G., Selgeby, J. H., Schreiner, D. R., Schräm, S.T., Swanson, B. L., MacCallum, W. R., BurnhamCurtis, M. K., Curtis, G. L., Heinrich, J.W. & Young, R. J. (1995). Lake trout (Salvelinus namaycush) populations in LakeSuperior and their restoration in 1959-1993. Journal of Great Lakes Research 21, 152-175.

Holey, M., E. A. Baker, T. Thuemler, and R. F. Elliott. 2000. Research and assessmentneeds to restore lake sturgeon in the Great Lakes. Great Lakes Fishery Trust, Ann Arbor,Michigan Available:www.glft.org/resourcelibrary/attachments/PROJECTS-T4WebFilel33- Sturgworkshop00.pdf. (May 2010).

Horton, G. E., Letcher, B.H., Bailey, M. M. & Kinnison, M. T. (2009). Atlantic salmon(Salmo salar) smolt production: the relative importance of survival and body growth.Canadian Journal ofFisheries and Aquatic Sciences 66, 471-483.

Johnson, J. H. (2004). Comparative survival and growth of Atlantic salmon from eggstocking and fry releases. North American Journal ofFisheries Management 24, 1409-1412.

136

Jokikokko, E. & Jutila, E. (2004). Divergence in smolt production from the stocking of 1 -summer-old and 1 -year-old Atlantic salmon parr in a northern Baltic river. Journal ofApplied Ichthyology 20, 511-51 6.

Jokikokko, E., Kallio-Nyberg, I., Saloniemi, I. & Jutila, E. (2006). The survival of semi-wild, wild and hatchery-reared Atlantic salmon smolts of the Simojoki River in the BalticSea. Journal ofFish Biology 68, 430-442.

Jonsson, N., Jonsson, B. & Hansen, L. P. (1998). The relative role of density-dependentand density-independent survival in the life cycle of Atlantic salmon Salmo salar.Journal ofAnimal Ecology 67, 751-762.

Ketola, H. G., Bowser, P. R., Wooster, G. A., Wedge, L. R. & Hurst, S. S. (2000). Effectsof thiamine on reproduction of Atlantic salmon and a new hypothesis for their extirpationin Lake Ontario. Transactions ofthe American Fisheries Society 129, 607-612.

Lacy RC (1997) Importance of genetic variation to the viability of mammalianpopulations. Journal of Mammalalogy 78(2):320-335

Le Gouar, P., Rigai, F., Boisselier-Dubayle, M. C, Sarrazin, F., Arthur, C, Choisy, J. P.,Hatzofe, O., Henriquet, S., Lecuyer, P., Tessier, C, Susie, G. & Samadi, S. (2008).Genetic variation in a network of natural and reintroduced populations of Griffon vulture(Gyps fulvus) in Europe. Conservation Genetics 9, 349-359.

McCrimmon, H. R. (1954). Stream studies on planted Atlantic salmon. Journal oftheFisheries Research Board ofCanada 11, 362-403.

MacCrimmon, H. (1965). The beginning of salmon culture in Canada. CanadianGeographical Journal 9, 4-1 1 .

MacDougall, T. M., Wilson, C. C, Richardson, L. M., Lavender, M. & Ryan, P. A.(2007). Walleye in the Grand River, Ontario: An overview of rehabilitation efforts, theireffectiveness, and implications for eastern Lake Erie fisheries. Journal ofGreat LakesResearch 33, 103-117.

Maynard, D. J., Flagg, T. ?., and Mahnken, C. V. W. 1995. A review of seminaturalculture strategies for enhancing the postrelease survival of anadromous salmonids.American Fisheries Society Symposium, 15: 307-314.

Meffe GK (1995) Genetic and ecological guidelines for species reintroduction programs:application to Great Lakes fishes. Journal ofGreat Lakes Research 21(Suppl. 1):3—9

McCormick, S. D., Hansen, L. P., Quinn, T. P. & Saunders, R. L. (1998). Movement,migration, and smolting of Atlantic salmon (Salmo salar). Canadian Journal ofFisheriesand Aquatic Sciences 55, 77-92.

137

McMenemy, J. R. (1995). Survival of Atlantic salmon fry stocked at low density in theWest River, Vermont. North American Journal ofFisheries Management 15, 336-374.

Meffe, G. K. (1995). Genetic and ecological guidelines for species reintroductionprograms: application to Great Lakes fishes. Journal ofGreat Lakes Research 21, 3-9.

Mittelbach, G. G., Turner, A. M., Hall, D. J., Rettig, J. E. & Osenberg, C. W. (1995).Perturbation and resilience - a long-term, whole lake study of predator extinction andreintroduction Ecology 76, 2347-2360.

Ontario Ministry of Natural Resources. 2007. Lake Ontario Fish Communities andFisheries: 2006 Annual Report of the Lake Ontario Management Unit. Ontario MinistryofNatural Resources, Picton, Ontario, Canada.

Ontario Ministry of Natural Resources. 2008. Lake Ontario Fish Communities andFisheries: 2007 Annual Report of the Lake Ontario Management Unit. Ontario Ministryof Natural Resources, Picton, Ontario, Canada.

Ontario Ministry of Natural Resources. 2009. Lake Ontario Fish Communities andFisheries: 2008 Annual Report of the Lake Ontario Management Unit. Ontario Ministryof Natural Resources, Picton, Ontario, Canada.

Parsons, J. W. (1978). History of salmon in the Great Lakes, 1850-1970. U.S. Bureau ofSport Fish and Wildlife Technical Paper No. 68

Pimm, S. L., Russell, G. J., Gittleman, J. L. & Brooks, T. M. (1995). The future ofbiodiversity. Science 269, 347-350.

Rader, R. B., BeIk, M. C, Shiozawa, D. K. & Crandall, K. A. (2005). Empirical tests forecological exchangeability. Animal Conservation 8, 239-247.

Ridgway, M. S. (2008). A roadmap for coasters: Landscapes, life histories, and theconservation of brook trout. Transactions ofthe American Fisheries Society 137, 1 179-1191.

Riley, W. D., Eagle, M. O. & Ives, S. J. (2002). The onset of downstream movement ofjuvenile Atlantic salmon, Salmo salar L., in a chalk stream, fisheries Management andEcology 9, 87-94.

Rooney, N., McCann, K., Gellner, G. & Moore, J. C. (2006). Structural asymmetry andthe stability of diverse food webs. Nature 442, 265-269.

Sarrazin, F. & Barbault, R. (1996). Reintroduction: challenges and lessons for basicecology. Trends in Ecology & Evolution 11, 474-478.

138

Seddon, P. J., Armstrong, D. P. & Maloney, R. F. (2007). Developing the science ofreintroduction biology. Conservation Biology 21, 303-312.

Shadt, S., Revilla, E., Wiegand, T., Knauer, F., Kaczensky, P., Breitenmoser, U., Bufka,L., Cerveny, J., Koubek, P., Huber, T., Stanisa, C. & Trepl, L. (2002). Assessing thesuitability of central European landscapes for the reintroduction of Eurasian lynx. JournalofApplied Ecology 39, 189-203.

Simberloff D (1997) Flagships, umbrellas, and keystones: Is single-species managementpasse in the landscape era? Biological Conservation 83: 247-257

Stanfield, L. & Jones, M. L. (2003). Factors influencing rearing success of Atlanticsalmon stocked as fry and parr in Lake Ontario tributaries. North American Journal ofFisheries Management 23, 1 1 75-1 1 83.

Vitousek, P. M., Mooney, H. A., Lubchenco, J. & Melillo, J. M. (1997). Humandomination of Earth's ecosystems. Science 277, 494-499.

Utt, A. C, Harvey, N. C, Hayes, W. K. & Carter, R. L. (2008). The effects of rearingmethod on social behaviors of mentored, captive-reared juvenile California condors. ZooBiology 27, 1-18.

Zimmerman, M. S. & Krueger, C. C. (2009). An Ecosystem Perspective on Re-establishing Native Deepwater Fishes in the Laurentian Great Lakes. North AmericanJournal ofFisheries Management 29, 1352-1371.

139

Appendix 1

Evaluating the utility of stocking checks and other scale charactersto discriminate the origin of Atlantic salmon {Salmo salar) smoltsfrom Cobourg Brook, Ontario.

140

Introduction:

Research and assessment of fish populations often relies on partitioning the catch

of a single species into discrete populations, stocks, or research treatments (Ihssen et al.

1981). The characteristics or patterns on scales have become useful for identifying the

origin of fishes. For example, previous studies have been able to differentiate between

Atlantic salmon originating from lakes or streams (Dempson et al, 1996), captivity or

wild (Stokesbury et al, 2001), and between geographical stream reaches (Haas-Castro et

al, 2006), based on scale characteristics.

Fish scales provide a record ofbody growth, and can become useful for

distinguishing between fish with different growth histories. Scales are composed of

numerous concentric rings known as circuii, and changes in body growth are often

registered as breaks or changes in the spacing of the circuii. Disturbances in the pattern of

circuii are known as 'checks'. Environmental differences between captive and wild

environments should affect the growth of newly stocked fishes, and result in the

deposition of a 'stocking check' on their scales. Prior studies have found scale checks

useful for differentiating between wild and stocked fish (Casselman 1986; Ogle and

Spangler 1996; Taylor and Piola 2008).

The present study looks to evaluate the utility of using stocking checks to identify

when a unmarked Atlantic salmon was released into Lake Ontario tributaries.

Specifically, we investigate if checks along the longitudinal gradient of the scale are

indicative of the life stage at the time of stocking. Hypothetically, the youngest life stage

should possess a stocking check closest to the origin of the scale, while older life stages

should deposit a stocking check further from the scale origin. In addition, we investigate

141

the possibility of employing other scale characteristics to identify the origin ofjuvenile

Atlantic salmon in Lake Ontario streams.

142

Methods:

A total of 97 335 fry, fall fingerling, and yearling Atlantic salmon were

introduced throughout Cobourg Brook from the spring of 2006 to the spring of 2007. All

salmon used for stocking were from a single yearclass, and produced from the OMNR' s

LaHave River captive population. Approximately half of stocked yearlings were marked

with a fioy tag or a clipped adipose fin and all other individuals received no tag or clip.

After yearlings were stocked in spring 2007, a 1 5-foot fyke net was installed in

Cobourg Brook 4.2 kilometres upstream from Lake Ontario. Atlantic salmon captured in

the fyke net were presumably a mixture offish stocked as fry, fingerlings and yearlings.

Scales were collected from a sample of all Atlantic salmon captured in the fyke net, and

one scale from each individual was microscopically inspected. Scales were selected for

inspection if they did not appear to have been regenerated (i.e had small distinct origins),

were undamaged, and had clear circuii. Areas of condensed, branching, or broken circuii

were interpreted as checks, and checks associated with the cessation of annual growth

were classified as an annulus (Casselman, 1987). The distance from the scale origin to the

location to checks (mm) was measured. In addition, the circular extent of checks was also

recorded by partitioning the scale into nine equal segments, and recording the number of

segments which contained the check. For example, checks extending the entire

circumference of the scale were given a score of 9, while those extending 1/3 of the scale

were given a score of 3. If checks were at least 3/91 s in extent they were classified as

substantial, and checks which were less than 3/9ths were classified as minor. Checks and

annuii were interpreted along a radial line approximately 20° offset from the longest axis

143

of the scale, and all associated data was recorded using Calcified Structures Age and

Growth Data Extraction Software (CSAGES) ( Cassleman and Scott 2000).

The location of substantial checks was used investigate if the life stage of stocked

Atlantic salmon could be differentiated based on stocking checks. The presence of

adipose clips or tags was used to confirm the origin of Atlantic salmon smolts captured in

the fyke net. Hereafter, all yearlings marked with tags or clips are referred to as 'known

origin smolts', and all unmarked Atlantic salmon are referred to as 'unknown origin

smolts'. The distance to all substantial checks of known and unknown origin smolts were

plotted as a frequency histogram, hypothetically we would expect a trimodal distribution

of substantial checks, the earliest being associated with stocked fry, the middle with

stocked fall fingerlings, and the last associated with stocked yearlings. Differences in the

location of substantial checks between unknown and known origin smolts was tested

using the Mann-Whitney U-Test due to the non-normality of the data. In order to explore

if other scale characters can be used to differentiate origin ofjuvenile Atlantic salmon the

following additional criteria were investigated: abundance of substantial checks,

abundance of minor checks, and distance to annulus. All distances were recorded in terms

of the percent of the scale where the check was located. For example, if the check was

halfway between the origin and the edge, than it would be located at 0.50. The

abundance of substantial checks and the abundance of minor checks were not normally

distributed, requiring a Mann-Whitney U-Test to test for differences between the groups.

Distances to annuii were found to be normally distributed, thus differences between

known and unknown origin smolts were analyzed using the Students T-test.

144

Using the aforementioned criteria to differentiate between known and unknown

origin smolts is undoubtedly confounded by the presence of non-marked yearlings in the

unknown origin group. Thus, frequency distributions for criteria with substantial

(although at times not significant) differences between known and unknown groups were

plotted to estimate the abundance of individuals outside the rejection region of the T-

distribution of known origin smolts, indicating non-yearling origin based on the

respective criterion.

145

Results:

Distance to substantial checks from known and unknown origin smolts is

illustrated in Figure A-I . I found no difference in the distance to substantial checks

between known yearlings and unknown origin smolts (Mann-Whitney U= 3712, P=

0.60).

Results from other criteria Other used to investigated between known and

unknown origin smolts are provided in Table A-I . The only criterion that demonstrated a

significant difference was number of substantial checks (Mann-Whitney U= 20084.5, P=

0.044).

To determine the abundance of non-yearlings suggested by numbers of substantial

checks a frequency distribution was plotted (Figure A-2). Figure A-2 illustrates that

although there is a difference in the number of substantial checks between the groups, the

range of values is identical, preventing an estimate of abundance of non-yearlings based

on the T-distribution of the number of substantial checks.

Another scale character which demonstrated a substantial (but not significant)

difference between known yearling and unknown origin smolts was distance to annulus

(to.os(i), 363=1 .45, p=0. 149). A frequency distribution was plotted to determine the

abundance of captured non-yearlings based on distance to annulus (Figure A-3). Figure

A-3 illustrates the mean distance to annulus for known origin smolts were greater than

that of unknown origin smolts. Employing a one-tailed 95%, and 90% confidence limits

superimposed on the T-distribution, 14 and 24 individuals were identified as non-

yearlings based on distance to annulus, respectively (Figure A-3).

146

35

30

25

20

15

10

5

Distance to Substantial Checks Registered on Atlantic salmon Smolt Scales

ffl

"HNumberof known origin (yearling) smolts "

" il Num ber of unknown origin (fry. fingetling or "yearling) smolts

JLLj JJt, _ia, ? , ti _¦_0 0,05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 Q.S5 0.6 0.65 07 0.75 0.8 0.85 0.9 0,95 1 1.05

Distance to substantial check [% of scale)

Figure A-I: The distribution of the distances to checks greater than 2/9ths in circular

extent. Both known (e.g. yearling) and unknown origin Atlantic salmon collected from

Cobourg Brook are shown.

147

Table A-I: Results of analyses investigating

and unknown origin Atlantic salmon smolts.

Criterion

1) Number of substantial checks

2) Number of minor checks

3) Distance to substantial check(s)

4) Distance to annulus (proportional)

5) Distance to annulus (absolute)

criteria for differentiating between known

Significant Difference

Yes (Mann-Whitney U= 20084.5, P= 0.044),

No (Mann-Whitney U= 21480.5, P= 0.44)

No (Mann-Whitney U= 3712, P= 0.60)

No (teoso), 363=1-45, p=0. 149)

No(t0.o5(i),363=-0.54,p=0.60)

148

Presence of Substantial Checks on Atlantic salmon Smolt Scales

160

I Number of unknown origin (fryfingerling oryearling) smolts

¡Number of known origin (yearlings) smolts

Number of Substantial Checks

Figure A-2: The number of substantial checks (those 3/9lhsor greater in circular extent) observed

on scales of known yearling and unknown origin Atlantic salmon smolts from Cobourg Brook,

Ontario.

149

30

25

20

15

10

5

Distance to Annulus, Known vs. Unknown Origin Atlantic salmon Smoits

>^M=0.89

W«s0.84

¦ Unknown Origin

D Known fifeariing) Origin

0 0.125 0.25 0.375 3.5 0.525 0.75 0.875 1 1.125 1.25 1.375 1.5 0.1 0.225 0.35

Figure A-3: Frequency distribution of distance to annuii of known and unknown origin smoits

captured from Coboug Brook, Ontario.

150

Discussion

Identifying the origin ofAtlantic salmon based on location ofa stocking check

Analyses indicate there was no difference in the distances to substantial checks

between known and unknown origin smolts. This suggests that either (1) all unknown

smolts were stocked as yearlings, or (2) stocked life stages are depositing similar patterns

of substantial checks on their scales. In my opinion I believe the option (2) is occurring.

In the field I observed marked differences between marked (stocked as yearlings), and

some unmarked salmon (possibly stocked as fry or fall fingerlings).Differences were

associated with morphology (degree of parr marking, scale silvering, and fin wear). I

collected scales from what I believe to be early stocked life stages (either fry or fall

fingerlings), however distance to substantial checks were not different between the

groups, suggesting that distance to substantial checks is a poor predictor of origin.

Figure A-I indicates that a single, substantial check is not being deposited upon

transition from captive to wild environments as hypothesized. However, it is possible that

a stocking check is forming, but similar checks are also developing which hinder

detection of a stocking check. In addition, it is possible that the equipment used to locate

scale characters may be imprecise, making the pattern difficult to detect. Regardless of

situation, I believe that stocking checks are not clearly forming on the scales of Atlantic

salmon stocked into Lake Ontario streams, and future analysis of this kind will most

likely be unable to resolve when unmarked salmon were released.

Using other scale characters to identify origin ofjuvenile Atlantic salmon

Analyses indicated that only 1 of 5 other criteria displayed a significant difference

between known yearling and unknown origin smolts. This illustrates the highly variable

151

nature of check and annuii deposition on juvenile Atlantic salmon scales, suggesting that

these characters are not good predictors of origin. Although these scale characters are not

capturing a difference between stocked life stages, I believe I have observed dissimilar

scale patterns among stocked life stages. Yearlings are registering wide, uniform circular

spacing, with no deposition of an annulus shortly after stocking. As time progresses after

stocking, some yearlings are developing an annulus (or stocking check), than registering

growth beyond the annulus. Stocked parr (fry or fingerlings) are registering typical

growth in a wild environment: wide circular spacing indicating spring-summer growth,

narrow circular spacing indicating fall-winter growth, an annulus, than wide circular

spacing typical of pre-smolt growth. Distance to annulus shows potential to differentiate

origin ofjuvenile Atlantic salmon. The aforementioned scale patterns of stocked yearling

and parr Atlantic salmon suggest that there may be a biological reason for this. Because

stocked yearlings are not registering an annulus (in captivity) and stocked parr are

depositing an annulus (in the wild), an individual stocked prior to the winter (fry or

fingerlings) will have an annulus closer to the scale origin that those stocked as yearlings.

The abundance of non-yearlings inferred from the frequency distribution of distance to

annulus is similar to the abundance of non-yearlings I believe I sampled in the field.

Although distance to annulus shown promise to distinguish origin ofjuvenile

Atlantic salmon, I believe there are issues confounding the method. Many authors have

addressed the subjective nature of annulus classification. Pseudoannuli, checks, and other

scale characters often prevent clear demarcation of annual growth increments. In this

case, annuii classification of stocked yearlings is highly subjective as annuii are not

clearly displayed and checks are often located near the scale edge (Figure A-I). An

152

option to alleviate this issue is to abandon annulus classification by assigning all breaks

in circuii as checks and developing a criterion describing the magnitude of checking (e.g.

greatest rank indicating a large amount of broken circuii).

Even with the abandonment of annulus classification, the large temporal and

concurrent growth period over which samples are compared further confounds

implementation of distance to annuii as a criterion to discriminate the origin ofjuvenile

Atlantic salmon. The observed difference in scale growth between stocked life stages

illustrates that at a given time, stocked yearlings will have a greater distance to annulus

than stocked parr (fry or fingerlings). However, this pattern is complicated by the large

temporal extent of the capture period. For instance, dissimilar annulus distances (from

origin) at a given point can appear similar if compared from opposite ends of the capture

period. For example, a salmon stocked as a fry which was captured early in the smolt run

may have a similar annulus location as salmon stocked as a yearling caught at the end of

the run (Figure A-I). Investigating capture date of non-yearlings determined from Figure

1 revealed that most individuals were not captured during the early portion of the smolt

run which, according to my field experience, was the period in which non-yearlings

emigrated. This is likely caused by the large temporal extent of the comparison period.

One option which would alleviate this issue it to temporally stratify my sample period

(e.g. compare samples caught the same week), but unfortunately I do not have sufficient

sample sizes for such analyses. In order to continue using this criterion for origin

differentiation I need to standardize for time. Employing daily growth increments or

other similar methods may be appropriate and requires further investigation.

153

Another growth related issue confounding the use of annuii distances to

discriminate origin ofjuvenile salmon is my observation that yearlings were not

registering an annulus shortly after stocking. During 2007 scale interpretations 1 8% of

individuals did not register an annulus, which is likely attributed to a genuine lack of

annulus, or the aforementioned issue of subjective annulus-check classification.

Regardless of the cause, the substantial number of individuals interpreted without an

annulus requires attention if this method is to be considered fro future analyses.

An additional issue complicating the use of scales to discriminate origin of

juvenile Atlantic salmon is the poor quality of scale samples. Regenerated or damaged

scales were avoided for interpretation, however they were quite prevalent as 13.5%

(70/515) of individuals were not analyzed this reason. Image quality was also an issue,

poor circular differentiation that may impair check or annulus identification occurred in

32% (144/444) of interpreted samples.

Future analysis suggestion:

As my findings illustrate, analyzing the location of hypothetical stocking checks

does not appear to be a suitable method for differentiating the origin ofjuvenile Atlantic

salmon. A previously published method for identifying the origin ofjuvenile Atlantic

salmon which is applicable to my situation does not appear to exist. Thus, I propose a

method utilizing the synergy of (1) distance to annuii, and (2) intercircular spacing about

the 1st annulus to discriminate origin ofjuvenile Atlantic salmon in Cobourg Brook

(Seelbach and Whelan, 1988). Discussion of each criterion, and validation/corroboration

techniques are provided below.

154

Criteria:

(1) Distance to annuii: My preliminary scale interpretations of Lake Ontario Atlantic

salmon smolts indicates a longitudinal gradient of distance to annuii should exist among

stocked life stages, with fry having the nearest annulus to scale origin, than fmgerlings,

than yearlings.

Issues:

• Yearlings not depositing an annulus. I suggest assuming scale edge as annulus

location for those (yearlings) that have not deposited an annulus at their 1st springcapture.

• Annulus classification. I need credible corroboration to justify my annuii

interpretations. I suggest having a second reader to corroborate my annuii

interpretations.

• Image quality. A greater resolution microscope may resolve some image quality

issues. Regenerated and degraded scales cannot be avoided.

• Accuracy and precision. I suggest multiple observations for each scale.

• Time/growth correction: I need to research potential solutions.

(2) Intercircular spacing: Preliminary scale interpretations suggest intercircular spacing

inside the 1st annulus should be smaller for fish residing in the wild prior to winter (fry)than those in captivity (fall fmgerlings and yearlings). Intercircular spacing outside the 1st

annulus should be less for individuals recently introduced (yearlings) to the wild than

those introduced as early life stages (fry and fmgerlings).Issues:

• Annulus classification: see distance to annuii.

155

• Accuracy and precision: see distance to annuii.

• Equipment. Dr. Jones' new microscope may be appropriate for recording

intercircular spacing. He recently purchased a dissection scope with a

magnification range of 5-1 8Ox, and has a spreadsheet interface for recording

objects, locations and distances.

I am sure there are other intricacies involved in the method. I intend to contact persons

who have employed the technique in the past.

Hypothetical criteria differences between origins ofjuvenile Atlantic salmon:

Using the synergy of distance to 1st annulus, and spacing of 3 circuii about the 1st

annulus, fry fmgerlings and yearlings should be distinguishable as:

• Fry should have a nearer annulus, smaller inter-annulus circuii spacing, and

greater outer-annulus circular spacing than yearlings and fmgerlings.

• Fingerlings should have an annulus location intermediate of fry and yearlings,

smaller inter-annulus circuii spacing than fry, and larger outer-annulus spacing

than yearlings.

• Yearlings should have an annulus location further than fry and fmgerlings, and

inter-circular spacing greater than fry, and outer-circular spacing less than fry or

fmgerlings.

In addition to presence of tags and clips mentioned in this study, my experimental

design has incorporated genetic parentage analysis for origin identification, allowing

validation/corroboration of other origin identification techniques. Unfortunately, only a

few treatments (life stages at stocking) can be identified using parentage analysis (Table

156

A-2). Thus, I require scale pattern analysis to expand my treatment size. Origin

identification of all life stages, except fall fingerlings, can be validated/corroborated

using parentage analyses if growth patterns are similar between years. In order to

corroborate distance to annulus, and inter-annulus spacing of fall fingerlings, scales were

collected from a sample of fall fingerlings shortly after stocking in 2008. Proposed

methods for origin identification and corroboration for each survey are presented in Table

A-2.

157

Table A-2: Potential origin of salmon captured in Cobourg brook from 2007-2008, and

identification validation or corroboration techniques suggested.

Origin of salmon captured(year and life stage atstocking)

Method to identify origin Validation/Corroborationmethod

2006 fry Scale pattern analysis 2007 fry parentageanalysis

2007 fry Parentage analysis 2006 fry scale patternanalysis

2008 fry Parentage analysis 2006 fry scale patternanalysis

2007 fall fmgerling Scale pattern analysis 2008 fmgerling (knownorigin) scale patternanalysis

2006 fall fmgerling Scale pattern analysis 2008 fmgerling (knownorigin) scale patternanalysis

2007 yearling Scale pattern analysis 2007 yearling: presence ofclips or tags.2008 yearling: parentageanalysis

2008 yearling LaHave- parentageanalysisSebago-scale patternanalysis

2007 yearling: presence ofclips or tags2008 yearling: parentageanalysis

158

REFERENCES

Casselman, J. 1986. Scale, otolith and growth characteristics ofjuvenile lake trout-criteria for discriminating between indigenous and hatchery fish from the naturalenvironment. Great Lakes Fisheries Commission, Project Completion Report,Ann Arbour, Michigan.

Casselman, J. M. 1990. Growth and reltive size of calcified structures in fish.Transactions ofthe American Fisheries Society 1 19(4):673-688.

Dempson, J. B., M. F. Oconnell, and M. Shears. 1996. Relative production of Atlanticsalmon from fluvial and lacustrine habitats estimated from analyses of scalecharacteristics. Journal of Fish Biology 48(3):329-341.

DeVries, D.R., and Frie, R.V. 1996. Determination of age and growth. In Fisheriestechniques. 2nd ed. Edited by B.R. Murphy and D.W. Willis. American FisheriesSociety, Bethesda, Md. pp. 483-513.

Greig, L., B. Ritchie, L.Carl and CA. Lewis. 2003. Potential and Strategy for AtlanticSalmon Restoration in Lake Ontario: A Workshop Report. Prepared by ESSATechnologies Ltd. Toronto On. Ontario Ministry of Natural Resources. LakeOntario Management Unit. Peterborough ON. 39pp.

Haas-Castro, R. E., T. F. Sheehan, S. X. Cadrin, and J. G. Trial. 2006. Scale patternanalysis discriminates Atlantic salmon by river reach rearing origin. NorthAmerican Journal ofFisheries Management 26(2):289-300.

Ihssen, P. E., and coauthors. 1981. Stock Identification: Materials and Methods.Canadian Journal ofFisheries and Aquatic Sciences 38:1838-1855.

Casselman, J.M., and K. A. Scott. 2000. A general procedures manual for CSAGES-Calcified Structure Age and Growth data Extraction Software (Version 5.2).Glenora Fisheries Research Station, Picton, Ontario.

Ogle, D., and G. Spangler. 1996. Check Formation on the Scales of Hatchery-RearedLake Trout Prior to and Soon after Release into Lake Superior. North AmericanJournal ofFisheries Management. 16: 896-904

Seelbach, P. W., and G. E. Whelan. 1988. Identification and contribution of wild andhatcher y steelhead stocks in Lake Michigan tributaries. Transactions of theAmerican Fisheries Society 1 1 7:444 - 45 1 .

Stokesbury, M. J. W., G. L. Lacroix, E. L. Price, D. Knox, and M. J. Dadswell. 2001.Identification by scale analysis of farmed Atlantic salmon juveniles insouthwestern New Brunswick Rivers. Transactions ofthe American FisheriesSociety \30(5):S, 15-822.

159

Taylor, M. D., and R. F. Piola. 2008. Scale stocking checks to differentiate betweenhatchery-reared and wild mulloway Argyrosomus japonicus. FisheriesManagement and Ecology 15(3):21 1-216.

160