Influence of Transmitter Attachment Procedures on
Swimming Performance of Wild and Hatchery-Reared
Atlantic Salmon Smolts
Ahsrrut~t.-Effects 01 transn~it~cr
ntlachmcnr proccdurcs on srnolts 01 Allantic salmon S<rlrrlo
III thc Inbornlory wilh hnlchcry-rcared fish. Bicitclcmc~ry
rccnrchcrs must extrapolulr tindirl~sof tnggirlg sludics to wild. migrnling smtills. Hccuusc hutchrry
rcnrinp can uff'ccr snlrnonid dcvtiltipni~~~t
;tnd hrcnuse field tagging proccdurc:, and cnvironmcntal
condition> can differ frcrrrt thosc in the lnhoralory. wc rneasurcd critical swimnlinp spccd o l
hntchery-rcarccl Atlantic salmon ~ m o l t cqu~pprd
i
wilh inlernnl, external, rir p s l r i c dummy radlo
tranhrnitlers in thr labomlor-y. Wc ~hcrlr-rpcn~rdthr cxpcrirncnl with wild, migraling fish I n amhicn~
cnviro~~mcntal
condition.\. Swirnm~np;tbil~ly01 ha~chcryamolth wa\ no1 hipnilicnnlly nl-Ucctrd hy
nny nttachmrnt proordurc, xnd no hifinilicanr difl'crcncrs wcrt. found bc~wcenfish t c s ~ r dI nr I(>
h after lagging. Similarly, wild rnroll~wcrc no1 impaired hy castric ~ransmittcrh:howcvrr, pcrlcrrnurhx of wild fish tec~cdI or 10 h after intrrnal or cxtcrnnl attachmcm wns s~pnificantlylower
than Ihal of wild controls. Thus. rcilction o i h;\lchcry-rcarcd Arlanlic snlnio~ls n ~ o l l I\t r nr~achmcnt
procedures can differ from t h a ~oi'wild nilgrnnls. Biotclemcrry reacnrchcrs h o u l d cxcrcisc cnul~cin
in inlerprcting dnta oht;~incdfrom wild Allantic salmon srrloll.: equipped w ~ t h~rttrrr~al
or cxlcrrrnl
radio tmnarnillcrs dur~ngthc First Itr 11 ;~l'lcrlogping.
s u l ~ r rhnve prilnnrily hcun cxnmirlrd
Reccnl technological advanccs in the design of
radio and acoustic t r a n s m i l t ~ x sh a w allowcd hiot c l r m c l r y studics t o h e performed on incrc;~singly
s n d l c r fish such a s smolts of- Atlantic salrr~onSolmo salnr-. Migratory bchavior 01- lapgcd fish can
represen! lhal of thc untagpcd population if Lransmittcr attachment causes minimal physiological
dislurbuncc and n o impairment of swimming. Currcntly used u a n s m i t k r s arc surgically implanrcd
inlo thc body cavity (intcrnul), insertcd into thc
s ~ o r n a c h(gnstric), or fixed t o thc surfacc of thc
body (external; M e l l a s a n d Hayncs 1985). Studics
that have invesligatcd t h r ~ n c a s u r a b l cc f f c c t ~of
thcsc proccdurcs generally indicalcd that c x ~ c r n a l
tugging could significanlly aller bchavior or swimming ability of juvenilc Atlantic salmon whercas
gastric and intcrnal tagging might not (McCleavc
and Strcd 1975; Grecnstrcet and Morgan 19x9;
M o o r e c t al. 1990).
T h e majority of studies i n v e s t i g a t i n ~tagging cf-
' 10 whom corrcspondrncc should be 8ildrcsccd.
fects on Atlantic salmon smolts has cxarnined lurgc
ha~chery-reared fish in thc l a b o r a ~ o r y .Thy dirccl
application of thcsc rcsults t o biotelcn~etrysrudics
in thc held, howevcr, may no1 b c appropl-ia~cfor
scveral rcasons. First, rcaring conditions within
hatcheries can c a u s c allerations in growth (Bocuf
and Ciaignon 19H9). d c v c l o p m c n ~(Fleming ct al.
1994). and smol~ification (Kazakov and Kozlov
1085: Bjornsson et al. 1989; Farmer c t al. 19X9),
g
all of which may afl'cct s w i ~ n n i i ~ ipcrformancc
(Smith 1 9 x 2 ) o r tag-carrying ability.
Sccond, ratio of tag m a s s t o fish mass (tag ratio)
can affect the suitiihility of a parlicular atlachmcnt
proccdure (Marty and Summcrfclt 1 9 8 6 ) . Becausc
domestic stock lend l o h e m o r e robusl than fish
from n wild pupulation (Beamish 1978), large
hatchery smolls, having low tag ralios, could possibly carry transmitlcrs bcttcr than wild fish hiiving
higher t a g ratios. Third, published studics h a v c
generally cxatnined tagging cffccts in the slcrilc
and stahlc conditions of thc Iahorntc~ry.Downstream migration of wild smolts often occurs nfter
TARI.~.
I.-Mcun fork Icngth, rnnss. condition f;rcior. range trf tag ratios, and mean critical s p e d ( U c r i ~ of
) Atlnntic
snlmon molts wilhin treatment grc~ups.An aatcrisk ( * ) ~ndicntesn vnluc for hmlmy smults that is signitic;rn~lygrcntel( P c: 0.05) than rhc corresponding value fur wild smolra.
hcavy rains that cause fluctuations in discharge,
turbidity, and water tcmperaturc (Solomon 1978).
Thercfrm, laboratory studies, pcrformcd in the ahsencc of environment. may underestimate responses of wild taggcd fish to attachrncnt proccdures. Finally. previous studics have allowed
taggcd smolts to recover for periods o f 74 h to
several days bcfore tneasuring swimming ability.
In the lield, tish are often released aftcr recovery
from anesthesia hecause tirncly release is particularly important if migration time and transmitter
battery lifc are short.
These considerations suggest that responses by
large, hatchery-reared Atlantic salmon smolts to
tag attachrncnt procedures under laboratory conditions may not he directly applicable to biotelcmetry studies involving wild migrating smolts. To
test this hypothesis. we rncasured critical swimming spccds of large hatchery smolts equipped
with internal, external, and gastric tags in the laboratory. We then repentcd the cxperirrient with
wild fish tagged and testcd under held conditions.
Groups were examined 1 or 1 h h after transrnittcr
attachment to de~ermineif tagged smolts can be
relcused soon after surgery.
Methods
Transmirrrrr.-Dummy transmitters were identical to operational radio tags except that thc battery and clcctronics were rcplaced with lead shot
of cqual mass. Tags were cylindrical. weighed 7.6
g in air, measured 22 mm in Icngth. had a diameter
of 7 mm. and werc equipped with a 20-cm flcxible
wire antenna.
Hatclzrry snro1ts.-Hatchcry-reared
Atlantic
salmon smolts were ohtaincd from thc Alma Aquaculture Station in Alma, Ontario, and all research
on hatchery f sh was conducted in a laboratory at
this facility during the last 3 weeks o f May 1995.
Prior to this study, tish had been held for several
months in a rectangular raceway (10 X 2 X I m)
in which water was circulatcd at a rate of approximately 30 cm/s. Fish were fcd commercial trout
pellets from demand feeders and wcre exposed to
thc ambient photoperiod.
In all, 78 hatchcry smolts were uscd. Thrcc subgroups of 12 tish each were equipped with either
internal, external, or gastric dummy transmitters;
in each group, 6 tish werc tcsted after 1 h, and 6
werc tcsted 16 h after tagging. Three additional
subgroups of 12 tish each were sham-tagged and
tested in the same manner as taggcd fish. A final
group of 6 fish, testcd I h after approxinlatcly 10
s of handling without anesthesia or surgery. served
as controls. Mean lengths and masses of smolts
within each treatment group are given in Tdhle 1.
One day before experimentation. smolts wcre
transferred to a 7,000-L circular holding tank and
not fed, thus ensuring a postabsorptive state. Thc
holding tank was aerated and supplied with freshwater (mean temperature, 12.OUC; range, 1 1.013.0°C). Water was circulated around the tank at
approximately 20 cmls. Fish to be cxamined 1 h
aftcr transmitter attachment werc tagged late in the
morning, plnccd in a 750-L recovery tank (aerated
and supplied with freshwater circulated at npproximately 20 crnis) for I h, and then tcsted. Smolts
S W I M M I N G P E R F O R M A N C E O F A T L A N T I C S A L M O N SMOL'I'S
to bc examined I 6 h after attachment werc tagged
in the afternoon, placed in thc recovery tank for
16 h, and then tested. After experimcntation, fish
werc sacrificed by uncsthetic ovcrdose.
Hatchery snlolts tagged with internal and external procedures wcrc randomly sclected from the
2,000-L holding tank and ancsthctized by immcrsion in a MS-222 (tricaine mcthanesulfonatc) bath
(50 mg/L, buffcrcd). Procedures used wcrc similar
to those of Mellas and Hayncs (1985) with sonir
minor exceptions. First. all fish were taggcd in a
V-shapcd surgical tray. The tray contained freshwater for external tagging, whereas a diluted ancsthctic solution (75 mg/L) was used for internal
tapping. Second, for smolts cquipped with internal
tags, a small puncturc was madc in the body wull
to onc side of the initial incision, and thc transmitter antcnna was thrcuded through the puncture
from inside the lish. All smolts were considered
LO havc recovcrcd from anesthesia whcn [heir
buoyancy was regained and station was taken up
in thc current.
For gastric tagging, fish wcrc placed in a gastric
insertion chamher. The chamber was constructcd
from polyvinyl chloride (PVC) pipe (38 X 5 cm)
with thr top removed and thc cnds capped. A small
hole was made in the front end cap and thc apparatus was fillcd with freshwater. Fish wcrc iridividually placcd inside thc chamber with the
tnouth against thc front end cap. The transmiller
was partly inserted into one end of a drinking straw
(6-mm diarnetcr), which was dipped into vugrtablc
oil and inserted through the hole in the cnd cap,
into the fish's mouth and into its stomach. A wirc
was pushed through the struw, injecting the dummy transmitter into the stomach. Thc wire and
straw were thcn removed from the fish. The antenna exited via the mouth and was bcnt hackwards
to trail alongside the fish, which was thcn placed
in the 750-L recovery tank.
Wild srno1ts.-Wild Atlantic salmon srnolts werc
collccted from a counting fence in Micmac Brook
on the Indian River watcrshed in nol-th central
Newfoundland, Canada (approximately 49"N.
56"W). The counting fcnce was located approximately 45 km upstream from the cstuary. Expcrirnents occurred during the smolt migration (June
1995) at the Department of Fishcrics and Occans
Indian Rivcr Research Station approximately 5 km
from thc collection sitc.
From thc 48 wild srnolts that were used, three
subgroups of 12 fish cach were equipped with internal, cxtcmal, or gastric dummy transmitters: in
cach group, fish wcrc tested after 1 h, and the
709
remaining fish wcre tested 16 h after tag attachrncnt. Because wild smolts of lengths similar tu
hatchcry fish werc scarce, not cnough wild fish
could bc obtained to allow parallel sham tagging
of similar groups. Instead, 6 fish wcrc anesthetized
for npproximatcly 3 min bcforc being revivcd and
examined I h laler. Anothcr b smolts. tcsled 1 h
aftcr approximately I0 s of handling without anesthesia or surgery. served as controls. Mcan
lengths and masses of fish from cach treatment
group are give11 in Table 1.
Fish to bc cxamined I h after attachment wcrc
captured in the morning, taggcd, placed in a 750-L
recovery tank containing Indian River water (mcan
tcmpcmturc. 15.9"C; range, 14.8-1 8 3 ° C ; circulated ut approximately 20 cmds) for 1 h, and then
testcd. Smolts to be examined 16 h a f ~ c rtransmitter attachnicnt were collccted in thc uitcrnoon,
tagged, placcd in the 750-L tank fbr I 6 h, and thcn
tested. Control and gastric-taggcd smolts wcrc released aftcr experimentation, while in~ernallyand
externally tagpcd lish wcrc sacrificed by uncsthctic
ovcrdose.
Wild fish wcre subjcctrd to the same surgical
procedures dcscribed for halchery smolts except
for ancsthctic used. Wild srnolts were ancsthctized
with 2-phcnoxyethanol dilit~ed10 1 : 2,500 (volu m e : volume) duc to regulations concerning anesthetic use in thc Geld.
Swimming prrformancc.-Swimming
pcrformancc was cvaluatcd by mr.asuring critical swimming spccd, a measure of prolonged swimming
ahilily first dcscribed by Brett (1964). Expcrimcnts
bcpan by placing 3 smolts fi-om thc same treatment
gi-oup in n 120-L Blazku-type swim speed chambcr. Water was pumped Crom thc 750-L rccovcry
tanks, through the chanibcr, and back into thc
tanks. Fish wcrc initially acclimutcd to a water
velocity of 0.50 mls for 2 h, aficr which velocity
was increased hy 0.13 mls evcry 10 min until
smolts werc unable to leave thc downstream retaining scrccn. Final swimming speed, timc of fatigue, water temperature, and mass and fork length
of each fish were recorded. Critical specd (Ucrit)
was determincd by using the formula described by
Bcamish (1978).
One-way analysis 01- variancc (ANOVA), lincar
regression analysis, und Bonfcrroni multiple comparisons wcrc used to statistically examine the
datu. When equal variance tests failed, nonparamctric tests were uscd (Kruskal-Wallis one-way
ANOVA, Dunn's niul~iplecomparison).
T A H L?.-The
~
P-values tor lincnr regrcssior~stesting
for correlations. within Ltcntmcnt groups, hctwcrn crilicnl
sw~nuningspeed (Ucrit) o l Atlantic salrnon sntolts md
fork length, mass. and water tcmpcrntum. No significant
cotrelalions wcre found (all Ps .1: 0 . 0 5 ) .
P-values lix:
Fork
Icn~lh
TUXtyp:
Trcatmcnt
on
Ucrit
Mas\ 1111 Tc~rlp?rit[ure
tJcri~
on IUcr~l
Wallis onc-way ANOVA; H = 3.89; P
wcre not signilicqntly different (Table 1).
0.05)
Mean lengths of wild Atlantic salmon smolts,
among initial treatment groups, were not significantly different (one-way ANOVA; F = 0.296; df
= 7, 40; P > 0.05). Similarly, mcan fish mass
(Kruskal-Wallis one-way ANOVA; H 8.8; P >
0.05) and water temperature (Kruskal-Wallis oncway ANOVA; ff = 10.9; P >; 0.05) did not differ
signiticantly arnong treatment groups. No significant dirfcrcnccs wcre found in Ucrits of fish tested
I and 16 h after being cquippcd with gastric tags
0.14; P ::,
(Bonfcrroni ~nultiplccomparison; r
0.05), cxtcrnal tags (Bonfcrroni multiple comparison; I = 1.57: P > 0.05) or internal lags (Bonfcrror~irnultiplc comparison; I = -0.79; P > 0.05).
so thcsc data werc poolcd. Swimming ability of
controls anti anesthetic sham controls was not significantly diffcrcnt (Bonfemoni multiple compar, 0.05), so lhesc data werc also
ison: t = O.h7; P :
pooletl. Liwar regression anulysis revealed that
critical speeds of controls and taggcd fish were not
signiticanlly al'fectcd by fork length. mass, or water tempcraturc (Table 2 ) . Significant differences
werc found in mean critical switnrning speed
among control smolts and [hose with gastric, intcrnal. and extcrnal tags (one-way ANOVA; F =
79.4: Jf = 3, 44: P < 0.05). Controls and gastric;tlly taggcd lish did not diffcr signilicantly (Bankrroni multiple co~npnrison:t = 0.09; P > 0.05).
and tish with intcrnal and external tags did not
differ signiticantly (Bonferroni mul~iplccornparison: r = 0.08: P > 0.05); however. smolts
cquippcd with internal tags had a significantly lowc r mean Ucrit than control fish (Bunferroni ~ n u l tiple comparison: t = 6.72; P < 0.05) and gashically taggcd tish (Bonfcrroni multiplc cornparison: t = 6.63: P < 0.05). Similarly, smolts
equipped with external tags had significuntly lower
mcan critical speeds than gastrically taggcd fish
(Bonferroni multiplc comparison; r = 6.55: P <
0.051 and control fish (Bonfcrroni multiplc cornparison; r = 6.64; P < 0.05) fish (Table I ) .
-
-
Results
No alterations in behavior arnong tagged wild
and hutchcry smolts wcre noticcd and no mortalities occurrcd. Rccovcry of hatchery and wild
smolts from anesthesia. after internal and external
tag placcmcnt, occurred in 2-5 min. Both hatchery
and wild tish cquippcd with gastric tags displayed
a coughing-type behavior for a short timc aftcr
tagging; howcver, this rcaction was not observed
in tish swimming in the flume.
-
Mean fork lcngth (one-way ANOVA; 1; 1.55;
df = I?, 65: P > 0.05) and mass (one-way
ANOVA; F = 1.73; df = 12, 65; P >: 0.05) of
hatchery Atlantic sulmo~ismolts did nut diSfer significantly among initial treatment groups. No significant diffcrenccs in Ucrits occurred within trratment groups of tish tested I and 16 h after surgery
(one-way ANOVA; F = 1.35; df = 11, 60; P >
0.05), so thcse data werc poolcd. Linear rcgression
analysis showed that critical speeds of control fish,
sham-tagged fish, and tish tagged with internal,
cxternal or gnstric transmitters, wcre not signilicantly affectcd by fork lcngth or mass (Table 3).
No significant differences in criticnl swimming
speed were found (Kruskal-Wallis one-wuy ANOVA: H = 8.84: P > 0.05) betwecn tagged and
sham-taggcd groups (Table I ). Finally. mcan critical swimming spceds among uontrds and smolts
with gasuic. internal, and extcrnal tugs (Kruskal-
Hirrclrery versrrs Wild Srnm1r.r
Becausc significant differences in lcngth occurrcd between wild and hatchery srnolts (Table
I ) , trcstlncnt effects on critical swi~nlningspeed
relative to body lcngth werc compared (Figure I ) .
Mean relative Ucrit was found to differ significantly (Kruskal-Wallis one-way ANOVA; H = 55:
P < 0.05) hetwcen similarly taggcd hatchery and
S W I M M I N G PERFORMANCE OF ATLANTIC S A L M O N SMOLTS
10.0
9.5
i
71
wild tagged or control
ahatchsry conbol
hatchsry taggad
hatchsry sham tagged
control
external
gastric
internal
F r m r ~ tI.--Mean rclntivc critionl speeds (Ilcrits) in body lengths per sccond (hl/s) of halchcry nnrl wild Atlanric
salmon snlolts from the variouh tag Ireatmcnts. An asterisk (+) ~ndicntcsP signifc~nldilfercncc belween group?.
ot similarly lagged wild and hn~cherysmolts. A plus ( + ) indicatcs a signilicnnt diffcrcncc in mcun rclntivr llcrit
hetwccn R group of harchery nr wild fish and that ol 1hc corresponding ha~chcryor wild control group. Vcrlical
linc nbovc bar indicates I SE
wild fish. N o significant diffcrence was found in
relative critical spced hetween hatchery and wild
controls (Dunn's multiple comparison; Q = 0.92;
P > 0.05) ur hetwccn hatchery and wild gastrically
ragged smolts (Dunn's multiple comparison: Q =
0.82; P > 0.05). Rclative critical speed was, however, significantly reduced in wild srnolts taggcd
with cxtcrnal (Dunn's multiple comparisun; Q =
3.19; P < 0.05) and internal (Dunn's multiple comparison; Q = 4.72; P < 0.05) dummy transmitters
relative to sinlilarly tagpcd hatchcry-reared
smolts.
Discussion
In rivers, migrating Atlantic salmon smolts arc
thought to swim against but not faster than thc
flow rate, resulting in downstream movemeni at a
specd slower than that of the currcnt (Solomon
1978; Smith 1982). If thc ability of a smolt to swim
against the currcnt is impaired by a particular
transmitter attachment, thc smolt's downstrcarn
travel rate will bc faster than that of untagged fish,
thus biasing telemetry data (Greenstreet and Morgan 19X9). The present study indicates that, in the
short term, prolonged swimming ability of hatchery-rcmed Atlantic salmon smolts is not affected
by internal, cxternal, or gastric dummy transmitter
attachment. Critical spced of wild smolts is also
not affected by gastric lags; however, wild fish
cquipped with external and intcrnal tags showed
21.9% and 22.7% reductions in pcrformancc, respectivcly, relativc to controls. Therciore, had
,
thesc fish heeri used in n biotclemctry s ~ u d ydara
colleckd for thc first 16 h pcriod might nor have
reflected the behavior of untagged migrants.
Mean Ucrits of wild fish cquipped with extcrnal
and intcrnal dummy transmittcrs did not differ significantly, suggesting that drag due to cxternal tags
did not impair swimming ability. In addition, nci[her internal nor gastric tags contributed to drag;
yet, pcrformancc was impaired by internal transtnittcrs hut not by gastric transmittcrs. Thcsc patterns suggest that thc effect of the intcrnal and
extcrnal tags on swimming pcrformancc of wild
smolts was physiological rather than mcchanical.
Internal and external procedures share onc important characteristic: thcy both involve cutting or
piercing of tissue. Although gastric tagging may
cause fish somc discoml'ort; littlc or no tissue damage occurs. Thus, stress causcd by the surgical
procedures prohably contributcd most to the rcduccd ability of wild smolts to carry intcrnal and
712
PEAKE ET A L .
cxtcrnal transmittcrs. That hatchery smolts were
unaffcctcd suggests thcy were not as stressed by
the proccdurcs or that iheir tolcrancc to stress was
greater than that of wild fish.
Despite our attempts to treat hatchery and wild
lish in as similar a manner as possible, some diFfercnccs in methodology wcrc unavoidablc. Thcse
differences illustrate the need for caution in directly applying the results of laboratory tagging
studies, perfor~ncd on hatchcry smolts, to fieldbascd hiotelcmctry research. First, different anesthetics werc used becausc cnvironrnental regulativns did not permit us to use MS-222 in the
held. Ncverthclcss, we bcljcve that ancsthctic use
did not influctlcc our results hccause no diffcrcnces
in ~rccovcrybctwccn ancsthctixcd wild and hatchery fish were observed. Furthcrmore, swimming
ability of wild lish was not affectcd by the use of
7-phenoxycthanol because no signiticant diffcrcnccs wcrc observcd between anesthetized and
nonancsthctizcd controls; likewise, MS-277 did
not affcct swimming pcrformancc o f hatchery
srnolts in similar comparisons.
Second, water tempcrntures during the time of
the wild smolt migration were slightly higher than
those at the hatchery. Although swimming ability
01' wild s~nolts,within treatment groups, was not
altercd .significantly by lluctuatio~lsin water temperature (Table 2 ) , wild lish possibly were affcctcd
by internal and cxtcrnal tagging hccause of thc
higher watcr ternperaturcs. Also, although internal
tags reduced swimming performance of wild tish,
gastric tags did not, cvcn though thcy h ~ ~ werc
th
tested at thc same mcnn temperature (Table 1).
Nevcrthclcss, subscqucnt tagging studies should
he cnrricd out across a runge of watcr temperatures
that include a11 values that may be encountered in
the ticlcl.
Third, hatchcry fish might have bccn less
stressed than wild stock prior to tagging hecause
hatchery tish were probably more accustomed to
handling and conlinement. Capture and confinerncnt of wild fish are inherent to most biotclcmetry
studies, and their effect on physiological condition
cannot be properly simulated with hatchery tish.
Fourth, becausc of clifferenccs in selective prcssure, tish stocks may diffcr in terms of morphology, swimming ability, and stress tolerancc. However, this possibility exists whenever findings from
one tish population arc extrapolated to another.
Finally, domestic tish tend to be more robust
than wild stock (Beamish 1978). In the prcsent
study, mean length of hatchery fish was only
slighrly grcatcr than that of wild smolts: howcver,
hatchery fish were 43-52% heavier ((Tuble I ). This
phenomenon can have practical implications. Ratio of fish mass to transmitter mass, or tag ratio,
has been shown to be an important consideration
for the biotclemetry rcscarcher, influencing the
suitability of external (Grcenstreet and Morgan
1989), internal (Marty and SummcrTclt 1986). and
gastric (Moser et al. 19W) attachments. In general,
low tag ratios arc cvn~ideredto he more suitable
than hjgher values. T~igratios for hatchery smolts,
in the present study, were lower than those of wild
fish (Tublc I ), thus differences in tag-carrying ability [night have resultcd from inconsistencies in tug
ratios.
Thc high condition factor of our hatchery smolts
also suggests that they might have undergone a
process called reverse stnoltification. Normal
smolriliuation involves a reduction in condition
Facror (Hoar IYXX); howcvcr, Woo et al. (1078)
reported that smolts of coho salmon 0ncorhvnchu.s
kisurth, preverited from entcring salt water, rcgained the high tissue h t and glycogen lcvcls charactcrisric of parr. Other biochemical and bchavioral Factors associated with smoltification, such
as salinity prefcrcnce (McIncrney 1964). glomerular liltration rate (Holtncs and Stainer 1966). and
branchial Nak,K'-A'1'Pasc activity (Zaugg and
McLain 1970). have becn shown to be reversible.
Smoltification of our hatchery stock also might
have becn incomplete or altered. Variables inhcrent to hatchery rearing thal have bccn shown to
affect development, smoltification, or levels of
hormones involved in smoltilicatioti includc size
of holding tanks and rate of water flow (Bocuf and
Gaignon 1989). type of artiticial diet (Bcrgstrom
IYW), watcr chemistry and pH (Farmcr et al.
1989). photoperiod and temperaturc (Bjornsson et
al. 1989), and illumination intensity (Kazakov and
Kozlov 1985). If smoltification was altered. swimming performance of the hutchery stock might
have been affected. Graham ct al. (1996) found
that station holding performance of juvenile Atlnntic salmon, relative to body length, irlcrcased
as fish became more silver in color. In addition,
swimming ability among the hatchery and wild
fish might have diffcrcd due to differences in body
form. Taylor and McPhail (1985) found thnt fusiforrn juvenile coho salmon had greater swimming stamina than more robust fish of thc same
spccics. Thus, the possibility that wild-type devclopmcnt and smoltification of juvenile salmon
can be altered or reversed by conditions in a hatchery makes i t nearly impossible for researchers to
ensure thnt morphology. physiology, hehavior, and
S M ' I M ~ I I N GP E R F O K ~ I A N C EO F ATI.ANTIC' S A I . M ~ N S M ~ L T S
713
pcraturc ailccr plasma grow~h I~ornlonr Icvcls.
growth. condltiori lucror. and hypoos~~ioregulntory
nhilily of ju\'cnilc Atlnnric salrr~on.Srrltr~o. ~ l a r .
during pan-smolr trnnslormarion. Aquaculturc S?:
77-9 1
Boeuf. C;.. and J. L. Gaipnon. 1989. Ellocrs or rcaring
condilionk n n growrh nnd thyroid hormonrh durinp
smollificatio~lof Atlnnric sulnlon. Solrrro .v~rliirI,.
Aquaculturc 82:?9- 38.
Bre~l.J. R. 1964. Ttic 1-cspiralnry rncraholisrn and swirrrming pcrforrnancc of young sockeyc salmon. Journal of thc Fishrrics Rcscarch Uoartl of Cnnatla 2 1:
I 183- 1220.
Fnrmrr. G . 1 . . K. L. Sauiiders. T R. Gofl'. C. E. Johnslon.
nnd E. B . Hcnderhon. 1989. Some phys~ological
respcinscs of A~luntics;rlnior~.S ~ l n i n.FN/CII;
cxposcd
to SOCI. acidic w u t r dul-ing snltilring. Aquncullurc
82:229-244.
Flrminp. I. A.. B. Jonssnn. and M . R. Grosc. 1094.
Phcno~ypicdi\rcrrcncc of sca-ranched. l;lrn~cd.iind
wild salmtir~.Canadinrl Journal of Fichrrics and
A q u ~ ~ Sciencc
ic
51 :2808-2E24.
Gruhn~n.R'. D., J . L. Thorpc, antl N . B. Mclcullc. I99h.
Scsson;rl currcnl holding pcrlormancr o l juvcnilc
Allantic salmon in rclntion lo tcmpcralurc and
cmrilting. Canadian Journal ol' Fivhcrics and Aqunlic Scienccs 53:HO-Xh.
Grccnsrret.~.S . 1'. R., and R . I . G . Morpil~i. 1980. The
rrlrct of ul~rasonicl i l r h 011 thc growth Tale*.ol AIInnlic salmon, Solrno JU/OI~.
purr of varying si/c jusl
prior to smelling. Journal ol'b'ish Riiilopy 24:235 244.
How, W S. 1988. Thc phyh~ologyof sniolling salrr~oAcknuwlcdgments
nids. Papcs 275-343 111 W. S . Hoar and L). J . RandWc thapk T. Beddow. D. L. G . N m k e s . G . Power,
all, cdilor*..Fish physiology, volu~neI I B . Acadcrnand G. Anderson for thcir criticul reviews o f the
ic Prrss. Nca Yorb.
manuscript. W e a l s t ~wish to acknowlcdgc the supIlolmcs, W. N.. und 1. M . Stnincr. 19hh. S~udicson Ihc
port provided by Lntek Engineering, Inc., for this
rcnnl cxcrcrion or clcclroly~ehhy [he rrtiut, Sr~ltr~o
~orr-ilr~r,ri,
Journal of Expcrirncn~alUiology 44:33study and the Canada Dcpartrncnl of Fishcrics and
46.
Occans staff at Indian River (P, Ruse, L. Colc. and
T. Nicholls) f o r their technical assistance. Funding Kaznho~.K. V . . and V. V. Kozlov. I9HS. Qu:~nular~vc
e.;~imatronti1 dcfret: [it silvering diqplaycd hy Arwas provided to K.S.M. by thc Dcpartmcnt ol'Fish1;rnlic salmon. Slrlnri) .\irlur~,~uvcnilcq trrig~nnling
crics and Occans a n d the Natul-al Scicnccs and
Crtirr~ n a r ~ ~ r a~mpulu~ions
l
and I~rorn lihll-rcarinp
Engineering Rcsearch C:ouncil o f Canada.
filrn~h.Aquaculturc 44:213-220.
Marly. G . D.. and R . C-. Summerfcll. IOXh. I'nthwayh
References
antl mcclianisms for expulsion d surgically irnplanled dummy rrunsmllrcrs from channel cntl~sh.
Arm.slronf. J. T).. nnd C. E. Kawlingh. 1903. Thc ctfrcl
Transactions o l Ihc Amcricnn Fishcrics Society 1 15:
of intra~uslricLri~nc~nittcr~
on fccdinp hchaviour o f
577--589.
Atlanlic calmon, Sulmo ~ l l o r purr
.
dur~nguulunin.
McClcnvc. I. D.. and K . A. Strcd. 1075. Lfl'ccl o l d u n Journal of Fish B~ology43:046-h4H.
m y tclcmetry tran.;niittcrh tin the slartrina o l Atlantic
Bcnmiqh, F. W. H. 1978. Swimming capacity. Papcs
snlmon. Sirlrno .sulur-.slntilrk. J o ~ ~ r nofthc
ul
F~hhcr~cc
101- 187 in W. S. Hoar, and TI. J . Kandall, cd~tor-s.
Rcscarch Boar-d 01 Canada 32:559-5(i3.
Fish phyh~ology,vtilumc 7. Acadcmrc Prcsk. Ncw
Mclncrncy. J . E. 1964. Salinily ~prelcrcncc:an oricnYork.
ration mcchanisn~in si~lrnonrnigrntion. Journal o l
Bcrgslrom, E. 1089. Effecl o l natural and arr~ticiald i c ~ s
thc F~shcrirsRCLCHI-ch
Board (if Canada 2 1 : W 5 on seasonal change3 in fatty acid compo.;lrlon nnrl
1018.
total body l ~ p i dcontcnt of wild nnd halchcry-rcared
Mcllas, E. J., and 1. M . lfayncs. 1085. Swimrnin~perAtlanlic hulmon. Sulrno wlirr-, prirr-smolt. Aqualormancc and hehaviour of rainhow troul. Srrlmn
culture R2:2US-2 17.
guiriit~r~ri,
and whitc pcrch, Moronr umcrrcurro: cfHJO~~SSR
O .~. H
, . Thorarcnscn. T Hirano, T Ognsawaru.
lccts of ~ r ~ a c l i i nrclcmt~ry
g
transmiltcrs. Canad~an
and J. R. Kri51insson. 1989. Pho~ripcr~od
nnd lemswimming capability of hatchcry-reared fish. nr
any given timc or place. match those of wild
smnlts.
In conclusion, we rcport that, for a varicty of
possible rcasons. wild Atlantic salmon smolts did
not react to trnnsniittcr attachment procedures it)
t h r s a m e way a s hatchery-rcarcd fish. G a s ~ r i utagging of wild and hatchery-reared Atlantic salmon
smolts d i d not impair swimming ability and may
bc an appropriate attachment procedure in the
short term. Problems associated with fccding.
howcver, h a v c been reported (Armstrong and Rawlings 1993) and m a y affect suitability for longertcrm studies. Unlikc halchery fish. wild fish demonstrated signilicantly reduced prolonged swimming ability when tagged with cxtcrnal or interval
trunsmitters, nnd pcrformancc did nor irtiprovc a[tcr Ih h of rccovery. Thus, rcsults of laboratory
tagging studies that use hatchery-rcarcd smolts
should be cautiously cxtrapolatcd to wild migrants.
In addition, researchers should critically examine
tclcmetry data from internally or externally tugged
wild smolts if conditions such as watur tcmperaturc or tag ratio differ from thosc in published tag
studies. T h i s cautionary action can bc cspccially
important if downstream migration ratu is ohscrved t o b c clnsc to !he spccd or thc currcnl.
714
P E A K E ET AL.
Taylor. E. B . . and J . D. McPhail. 1985. Vunation in
lournnl of Fihheries and Aquatic Sciences 423488burst and prolonged swimming pcrforrnance among
493.
Mourc, A , , I. C. Russell, and E. C. E. Potter. 1990. Thc
British Columbia p o p u l ~ t i u n sof coho snlmun. OnefFcct.i of in~rapcritoneally implnntcd dummy
cmrhynchus kisurch. Canadian Journal of Fisheries
ncouslic trnnsmitters on the bchnviour and biology
nnd Aquntic Scicnccs 42:2029-2033.
of juvenile Atlantic salmon, Salnro .rolar. Journal of Woo. N. Y. S., H. A. Bern, and N. S. Nishioka. 1978.
Fish Riology 37:713-721.
Changcs in body composition associnted with smolMoscr. M. L.. A. D. Olson. and T, P. Quinn. 1990. Eftilicntion and prcmaturc transler to senwutcr in coho
durnrny ultrasonic [rnnsmitters on juvenilc
fccts
salmon, Oncorhynchw kisurch, and king salmon,
coho salmon. Pages 353-356 in N. C. Pnrker and
Oncorhvnchus rshuwy~scha.Journal of Fish Biology
five coeditors. Fish-marking techr~iqucs.Amcrican
13:420-428.
Fisheries Sociery, Symposium 7. Bcthesda, Mary- Znugg. W S . . and L. R. McLain. 1970. Adcnosinc triland.
phosphntase activity in gills o r snlmonids: seasonal
Sinith, I-. S . 1982. Decreased swimming performance
vnriation and call water influcncc in coho salmun,
as a necessary cumponeni uf the srnolt migration in
Oncorhvnchus kisurch. Compnrntive Biochemistry
.salmon in the Colurnhia River. Aqunculture 28: 153nnd Physiology 35:5117-596.
Ihl.
Solornon, D. J. 1978. Some ohaervatiun.i on salrnoi~
hrnolt migration in a chalk strcnm. Jnurnnl of Fish
Received Ilcccmhcr 29, 1995
Acceprcd Dccernbcr Ih, IYYh
Biology 12:57 1-574.