HomeMy WebLinkAboutSUS240C\1
-C\1
~ _co
0
(")
~ ,.....
a: 0
c 0
l!)
l!)
-~"-_(")
(")
1-
~1
1--
This document is copyrighted material.
Permission for online posting was granted to Alaska Resources Library and Information Services
(AR LIS} by the copyright helder.
Permission to post was received via e-mail by Celia Rozen, Collection Development Coordinator,
on July 25, 2013, from Eileen Evans-Nantais, Client Service Representative, NRC Research Press.
This article is identified as SUS 240 in the Arctic Environmental lnstitute Susitna Aquatic Impact
Assessment Project Bibliography (1986), compiled by Arctic Environmentallnformation and
Data Center (AEIDC}.
TK
1425
.58
59
no.240
Carrying Capacity of an
Enhanced Side-Channel for Rearing
Salmon ids
This document is copyrighted material.
Permission for online posting was granted to Alaska Resources Library and Information Services
(ARLIS) by the copyright holder.
Permission to post was received via e-mail by Celia Rozen, Collection Development Coordinator,
on July 25, 2013, from Eileen Evans-Nantais, Client Service Representative, NRC Research Press.
This article is identified as SUS 240 in the Arctic Environmental Institute Susitna Aquatic Impact
Assessment Project Bibliography (1986), compiled by Arctic Environmental Information and
Data Center (AEIDC).
Carrying Capacity of an Enhanced Side&Channel for Rearing Salmonids
J. H. MUNDIE AND R. E. TRABER
Department of Fisheries and Oceans, Fisheries Research Branch, Pacifie Biological Station, Nanaimo, B.C. V9R 5K6
MUNDIE, J.H., AND R. E. TRABER. 1983. The carrying capacity of an enhanced side-channel
for rearing salmonids. Can. J. Fish. Aquat. Sei. 40: 1320-1322.
Two attempts were made to establish the yield of steelhead srnolts (sea-run rainbow
trout) (Salmo gairdneri) from a seminatural side-channel and compare it to that of the parent
river. In the first, 10 000 fry were introduced to the channel which was rnaintained at a
discharge of0.42 rn 3 /s. The fry, however, were largely displaced by extraneous coho salmon
(Oncorhynchus kisutch), appeared unable to withstand the water velocity in winter, and were
greatly reduced by infection from Cryptobia. In the second trial discharge was 0.14 m 3 /s. The
fry tolerated this. The yield (i.e. numbers) per unit area of steelhead smolts, of mean weight
14.5 g, was 31 times that of the river; in terms of biomass it was 10 times. Channel discharge
was 2.6% of the river discharge. Physical and biological factors determining smolt yield from
streams are considered.
MUNDIE, J.H., AND R. K TRABER. 1983. The carrying capacity of an enhanced side-channel
for rearing salmonids. Can. J. Fish. Aquat. Sei. 40: 1320-1322.
Nous avons tenté par deux fois de déterminer le rendement de la truite arc-en-ciel
anadrome (Salmo gairdneri) d'un chenal secondaire semi-naturel et de le comparer à celui de
la rivière d'origine. Au cours du premier essai, 10 000 alevins ont été introduits dans le
chenal, dont le débit a été maintenu à 0,42 m 3 /s. Toutefois, des saumons cohos étrangers ont
déplacé en grande partie les alevins, qui semblaient incapables de résister à la vitesse de l'eau
en hiver et ont été décimés par une infection de Cryptobia. Au cours du second essai, le débit
s'élevait à 0,14 m 3 /s, vitesse bien tolérée par les alevins. Le rendement des saumoneaux de
truite are-en-ciel anadrome, dont le poids moyen s'élevait à 14,5 g, a été de 31 fois plus élevé
que celui de la rivière par superficie unitaire. En terme de biomasse, cela équivalait à 10 fois
celui de la rivière. Le débit du chenal représentait 2,6 % de celui de la rivière. Les auteurs
étudient aussi les facteurs physiques et biologiques qui déterminent le rendement des
saurnoneaux dans les ruisseaux.
Received November 25, 1982
Accepted May 4, 1983
THERE is evidence that side-channels of streams, either single
channels or the multiple channels of braided rivers, can be
more productive of juvenile salmon and trout than the main
stems. For examplc juveniles of co ho salmon ( Oncorhynchus
kisutch) and chinook salmon (0. tshaw.vtscha) in Campbell
River, east Vancouver Island, B .C., occur in conspicuously
higher numbers in a channel that runs a short distance along
the north side of the river than in the main stem (Hamilton and
Bueil 1976). Sidc-channels may also be favored by over-
wintering juveniles, if they have less extreme freshcts than the
main river, contain pools with dean rubble, and have over-
head cover (cf. Bustard and Narver 1975).
The construction of a side-channel, initially used for high-
density seminatural rearing of coho (Mundie 1980), alongside
the Big Qualicum River on east Vancouver Island, presented
an opportunity to assess the natural carrying capacity and to
compare it with that of the river. The river is exceptional in
having controlled flow (Lister and Walkcr 1966). The channel
also has controlled flow. and in addition has features that are
known to enhance natural production. Our objective was to
Printcd in Canada (J7133)
Imprimé au Canada (17133)
Reçu le 25 novembre 1982
Accepté le 4 mai 1983
measure the difference in annual yield, per unit area, of sal-
monids from the two systems. Steelhead trout (sea-run rain-
bow trout) (Sai mo gairdneri) was the selected species. In the
first year (1980-81) only one age-class was reared, but in the
second (1981-82) two coexisted so the wild state was simu-
lated. Results from the lst year compelled changes in condi-
tions to be made for the second; moreover, unforeseen factors
arose that influenced the outcome of the trials.
Methods -The experimental channel was a diversion,
396 rn long, off the Big Qualicum River about 5 km upstream
of the river mouth. A screened intake and outlet prcvented fish
escaping from the channel. Channel width was 4.57 rn and
area 1810 rn 2
• Discharge was controlled by an intakc valve
and was set initially at 0.42 m 3 /s. The bed was a series of 25
grave! riffles, each 6 rn long and 15 cm deep, alternating with
25 pools 9 rn long and O. 9 rn dccp. The surface velocity of the
riffles was 60 cm/s and of the pools 10 cm/s. Floating ply-
wood on the pools gave cover for the fish and a net canopy
kept out herons and mergansers.
Fertilized eggs from Big Qualicum River 1980 steelhcad
adults were incubatcd, and fry were reared in tanks for 4 mo.
1320
NOTES 1321
In early Octobcr 1980 (first trial), 10365 fry of 5 g mean
weight, tagged with coded wire nose tags and with adipose fin
removcd, were introduced to the channel along its length. No
artificial food was offered (but see Results). In May 1981 the
outlet screcn of the channel was removed and outmigrants
were collected in a trap, counted, and then released. ln
October 1981 (second trial) 1000 fry from 1981 adults, simi-
larly treated, were introduced to the channel at 8 g mean
weight. Outmigrants were again recorded in May and June,
J F M A M
1980
1982. Remaining nonmigrants werc then enumerated.
To maintain the gravel for benthic production, the channel
was deaned in stages in August 1980 and 1981 . Each riffle
was scoured with a jet from a fire hose to a depth of 20 cm.
During this operation discharge was 0.42 rn 3 /s so that dis-
lodged sand could be transported out of the channel.
Results-Monthly mean temperatures of the channel over
the two trials ranged from 4.7°C to 13.0°C as follows:
1 A s 0 N D
12.7 11.8 10.6 7.5
1981 7.2 6.7
1982 5.2 4. 7
7.7 9.0 8.3 11.7 12.5 12.0 11.5 12.3 10.8 7.6
5.2 7.6 10.7 13.0
As the river was lake-fed the temperatures were high
enough to permit feeding by the trout at !east until January.
The diet consisted of ephemeropteran nyrnphs and chirono-
mid larvae, with smaller quantities of trichopteran larvae.
Filamentous algae were also present. This suggested that the
trout were grazing insects off the substrate.
The first trial lasted from Octobcr 1980 to June 1981.
During November and December, dead or dying fish,
amounting to about five per day collected on the downstream
screen. Unrecorded mortality also occurred, as moribund fish
were rcmoved by birds and mink (Mustela vison). Throughout
January and February deaths increased, until in late February
they amounted to 50 per day. The fish were cmaciated and had
difficulty holding position. Sorne of the moribund specimens
were infected with the flagellated protozoan Cryptobia salmo-
sitica, which occurs naturally in S. gairdneri in streams and
hatcheries (Putz 1972). We decided to conclude the experi-
ment and to feed the rernaining fish so that they could provide
a 1 + class for the second trial. Accordingly, fish were fed a
commercial food lightly in March only. ln addition, discharge
was lowered to 0.14 m3 /s until the following summer.
Mortality diminished to a few fish per day.
In mid-May the downstream screens were removed and
outmigrants counted. Most of the trout migrated between
May 20 and May 26. The total numbers at the end of the
migration (June 25) were 99 steelhead smolts (mean weight
19.5 g with a range of 10.5-36.0 g) and 22 parr which were
retumed to the channel. ln addition were 1553 coho smolts of
mean weight 15.0 g. Therefore, the trout had coexisted with
coho which must have entered the channel from the river
through the screen immediately after emergence in May.
The second trial lasted from October 1981 to June 1982.
Discharge was maintained at 0.14 m 3 /s and the 1 + steelhead
parr coexisted with the 0+ fry.
In the win ter of 1981 -82 recorded deaths of trout of both
year-classes amounted to 2300. Death was attributed to infec-
tion by Cryptobia as evidenced by swollen eyes in the fish
recovered. The disease also affected the Big Qualicum hatch-
ery stock of coho. The trout were thin, but not emaciated; nor
bad they difficulty holding position at the lower discharge. In
January the ratio of the 0+ and 1 + fish was 6:1 on the basis
of tags extracted from dead fish. By April it was 9:1.
The outmigration of trout in spring 1982 amounted to 1200
smolts of mean weight 14.5 g, and range 7.5-46 g; 7% were
over 20 g. Half the run left between May 19 and 22, and the
run continued until June 15. (The natural run from the river
usually peaks between late May and early June.) In addition
15 coho smolts migrated. The small numbers of coho are
attributed to the fact that any newly emerged coho entering the
channel in the spring of 1981 would be eaten by the trout;
alevins were found in trout stomachs at that time.
Between June 15 and 17 the remaining nonmigrant popu-
lation was assessed. Discharge was lowered to 0.014 m3 /s so
the riffles were exposed and the fish were confined to the
pools. Each pool was seined and a total of 2598 parr of mean
size 19 g was obtained. Each pool was then electrofished and
a further 350 parr were collected. ln total, then, the yield of
migrant and nonmigrant trout by mid-June was 4248.
The possible categories of trout that migrated or remained
in the channel were srnolts, parr that migrated (parr com-
monly disperse in the spring), parr that remained but would
become smolts within the next few weeks (sorne parr were
taken that were losing their parr marks and becoming sil-
vered), parr that would smoltify next spring, and finally resi-
dent trout that would not leave. The bulk of the trout
migrating had the appearance of smolts; they had lost their
parr marks and were silvered, but in view of their small size
there was sorne question as to their readiness to live in salt
water. Several (identified by their marks) were taken in a trap
for migrants near the mouth of the Big Qualicum River at the
peak of the run, indicating migration from the river. ln addi-
tion, a sample of 24 migrants of mean weight 18 g demon-
strated excellent seawater adaptability in a 24-h seawater chal-
lenge test on May 31 . ln a second challenge test on June 15,
nine nonmigrants (mean weight 13 g) suffered very high
blood sodium levels in seawater while three out of four
migrants (mean weight 22 g) again showed excellent seawater
adaptability typical of smolts; there was no ovcrlap of the two
groups in the test (Dr C. Clarke, Fisheries and Oceans
Canada, persona! communication). Nearly ail the nonmi-
grants had distinct parr marks. Their mean weight in June was
19.9 g~ 4% were l + fish of mean weight 30 g.
Discussion-The first trial was an instructive failure. Wc
1322 CAN. J. FISH. AQUAT. SCI., VOL. 40, 1983
deduccd from it that the dischargc of 0.42 m'/s submitted the
fish to excessive velocities in winter. Protection from high
flows in win ter appears necessary. For age 0 steelhcad in
natural strcams this is usually found in shallow water among
rocks; age l + steelhcad seek caver undcr boulders in the main
stem of rivers (Bustard and Narver 1975). Habitat diversity is
therefore a rcquiremcnt of survival. Crowding and discasc
also contributed to !osses.
In the second trial the small size of the migrant smolts was
attributable to thcir density in the channel; presumably sur-
viving numbcrs were determined largely by the available
food, bence much of what was consumed wcnt into mainte-
nance rather than growth. Small srnolt size is likcly to result
in low marine survival and consequently very low retums of
adults. W e wonder if a similar biomass of trout could be
produced with say, a third of the numbers. The thermal
regime of the channel would have perrnitted the trout in the
second trial (mid-October to mid-May) to reach a mean size
of 45 g if fed to satiation ( calculated from the madel of I warna
and Tautz 1981). The mean weight of steelhcad smolts
leaving the river in years prior to flow control (1961/62,
n = 50) was 35 g; after flow control (197 3, n = 303) it was
55 g, but the age composition of these catches is not recordcd
(Fisheries and Oceans, unpublishcd data).
The numbers and biomass of trout smolts Ieaving in the
spring of 1982 amounted to 0.663 smolts/m2 and 9.94 g/m2
•
The Big Qualicum River produces on average 0.021
smolts/m2 and 0.94 g/m2 (Federal/Provincial unpublished
data; a commonly acceptcd figure for steelhead smolt yield
in west-coast streams is 0.025/m2
). The channel, therefore,
produccd per unit arca 31 times as many smolts as the river
and 10 times the biomass. The residual parr amounted to
1.684/m2 or 25.26 g/m2
•
Apart from flushing flows (which may reach 24 rn 3/s) in
September, the river is usually run at about 5.6 m 3 /s in the
fall, 8.5 m3 /s in January and February, 4m3 /sin April and
May, and between 1.3 and 2.8 rn:>./s in the summer months
(Minaker et al. 1979). This gives a mean discharge of about
5.4 m3 /s in the period October-May, which is 38 times the
discharge of the channel, i.e. the channel discharge was 2.6%
of the river discharge.
For coho, small streams in B.C. yield about 0.73
smolts/m2 or 4.6 g; large ones yicld about 0.24 smolts at
3.5 g/m 2
• The Big Qualicum yields 0.17 smolts/m2 at 2.0 g
(D. E. Marshall, Fisheries and Oceans, unpublished data).
The data on coho smolt yields for the channel in 1981 reprc-
sented a yield of 0.85 coho/m2
, or 12.8 g/rn2
•
In addition to having a discharge more beneficiai. bath for
summer residence and for overwintering, than that of the
river, the channel had features that presumably contributed
to high survi val, i.e. the canopy, floating cover, a 1 / l . 5
riffle/pool ratio, and a uniform grave! that promoted ben-
thic food organisms. lt is not possible to assess the rela-
tive importance of these features. It is apparent, however,
that there are contrasting rcquirements of the fry feeding
from spring to wintcr, and of the overwintcring fingerlings.
Whereas the latter require low-velocity water and submergcd
and overhead cover, the former need higher discharges to
promo te food production and to main tain gravel quai ity. In
the channel, gravcl quality was maintained by an "artificial"
frcshet whcn the channel was clcaned.
From the yicld of steelhead srnolts and residual parr it
appcars that a commonly employcd biologists' guideline of a
twofold difference between side-channel and mainstcm juve-
nile carrying capacity is conservative. Finally, it is evident
that advanccs in undcrstanding the factors determining sa! mo-
nid production in strcams are most likcly to be made in field
situations whcrc discharge can be controllcd.
Acknowledgments-We thank G. R. Ladouceur and staff of the
Big Qualicum hatchcry for providing steel head fry. staff of the B .C.
Fish and Wildlife Branch for tagging fish, Dr C. Clarke for tcsting
readincss to accept seawatcr. D. E. Marshall and P. A. Slaney for
unpublished data, G. E. Hoskins for fish health diagnosis, L. L. Zado
for decoding tags, and A. R. Rccalma for field assistance. The project
was funded in part by the Federal/Provincial Salmonid Enhanccment
Program. Messrs H. T. Bilton, R. A. Barns, and P. A. Slaney kindly
made constructive comments on the manuscript.
BUSTARD, D. R., AND D. W. NARVER. 1975. Aspects of the winter
ecology of juvenile coho salmon (Oncorhynchus kisutch) and
steelhead trout (Sa/mo gairdneri). J. Fish. Res. Board Can.
32: 667-680.
HAMILTON, R., AND J. W. BUELL. 1976. Effects of modified hydrol-
ogy on Campbell River salmonids. Fish. Mar. Serv. Habitat
Prot. Dir. Tech. Rep. PAC/T-76-20. 156 + 21 p.
IWAMA, G. K., AND A. F. TAUTZ. 1981. A simple growth model for
salmonids in hatcheries. Can. J. Fish. Aquat. Sei. 38: 649-656.
LISTER, D. B., AND C.E. WALKER. 1966. The effect of flow control
on frcshwater survival of chu rn, coho, and chinook salmon in the
Big Qualicum River. Can. Fish. Cult. 37: 3-25.
MINAKER, B. A., F. K. SANDERCOCK, AND L. L BALMER. 1979. Big
Qualicum RiverProjcct, 1974-1975. Fish. Mar. Serv. MS Rep.
1528: 131 p.
MuNDIE, J.H. 1980. Intensive use of streams. ln J. E. Thorpe [ed.]
Salmon ranching. Academie Press, London. 441 p.
PUTZ, R. E. 1972. Biological studies on the hemoflagellates
Cryptobia cataractae and Cryptobia salmositica. Tech. Pap.
Bur. Sport Fish. Wildl. (U.S.) 63: 1-25.