Barge Transportation of Juvenile Salmonidson the Columbia and Snake Rivers, 1977

GEORGE T. McCABE, Jr., CLIFFORD W. LONG,and DONN L. PARK

Introduction

Transporting ocean-bound juvenilePacific salmon, Oncorhynchus sp.,and steelhead trout, Sa/rna gairdneri,from upriver collection points to saferelease sites in the lower ColumbiaRiver is one of the major tools beingdeveloped by the National MarineFisheries Service (NMFS) and the U.S.Army Corps of Engineers to reducefish mortality caused directly orindirectly by dams on the Columbiaand Snake Rivers.

Dams are responsible for threemajor causes of fingerling mortality: I)Gas bubble disease caused by supersat­uration of the river with atmosphericgas, 2) passage offish through turbines,and 3) delays in migration throughreservoirs (Ebel and Raymond, 1976;Collins et al., 1975'; Chaney and Perry

'Collins, G.B., W.J. Ebel. G.E. Monan, H.L.Raymond, and G. K. Tanonaka. 1975. TheSnake River salmon and steel head crisis.Northwest and Alaska Fisheries Center,National Marine Fisheries Service, NOAA,2725 Montlake Blvd. East, Seattle, W A 98/12Unpubl. rep., 30 p.

ABSTRACT-During the spring 9f1977 the National Marine FisheriesService, in cooperation with the u.s.Army Corps of Engineers, used two carROtank barRes to transport ju venile salmun­ids down the Snake and Columbia Rivers(through as man)' as eight dams) to saferelease sites in the lower Columbia River.Barge design and performance aredescribed. The transported Pacific salm-

28

1976). Research to solve each problemindividually has been only partiallysuccessful. The most promising re­search approach is the development ofa fish transportation system thatcircumvents the known causes ofmortality associated with both damsand impoundments.

In 1968, a system for collecting thedownstream migrants at the upper­most dams and transporting them bytruck around the remaining dams wasinitiated at Ice Harbor Dam (Ebel etaI., 1973) and is now in use at LittleGoose, Lower Granite, and McNaryDams.

During the winter of 1976-77, itbecame evident that because of lowprecipitation and low snow pack,extremely low river flows were prob­able during the spring migration ofjuvenile salmonids in the Snake and

George T. McCabe, Jr., Clifford W. Long, andDonn I.. Park are with the :\orthwest andAlaska Fisheries Center, National MarineFisheries Service, 'OAA, 2725 MontlakeBlvd. East, Seattle, WA 98112.

un. Oncorhynchus sp., and steelheadtruut, Salmo gairdneri, H'ere continuall)'supplied with fresh river waleI' pumpedIhruugh the harge. Fish were released inIhe main channel approximaleh 9 km(5.6 miles) helow Bunnel'ille Dam, thelowermost dam on the Culumhia River. Atotal of 3,517,242 juvenile Pacific salmonand steel/wad trout were hauled on thebarges.

Columbia Rivers. Faced with theprospects of little or no spill at thedams and based on the high percentageof fish lost during a similar but lesssevere situation in 1973 (Ebel andRaymond, 1976), we believed that loss­es of downstream migrants in 1977would be catastrophic unless a masstransportation program were carriedout. Based on our estimates of thenumbers of fish expected and theirtiming, we predicted that the collectionof fish at Lower Granite Dam duringthe spring of 1977 would exceed ourcapacity to transport them by truck.Rather than return large numbers offish to the river with little chanceof survival, the NMFS and the Corpsmade plans to supplement the trans­portation capability by using barges.

The primary objective of the bargingprogram was to transport fish collectedat Lower Granite Dam to safe releasesites in the Columbia River belowBonneville Dam. A second objectivewas to similarly transport juvenilesalmonids directly from selectedhatcheries (Fig. I). This report containsdetailed descriptions of the design andoperation of the barges and provides ageneral description of the bargingprogram carried out in 1977.

Barge Design

Holding Areas

Transporting juvenile salmonids bybarge is not a particularly new concept.Barging juvenile chinook salmon,0. Ishawl'lscha, was first done on the

Marine Fisheries Rel'i Yo'

Wenatchee

River---t

Leavenworth /

National

Fish Hatchery

Spring Creek National

Fish Hatchery

Willard National jFIsh Hatchery Y k R

\.

a ,rna Iver

~D:::eO ~

McNary

~~ (OLUMllIA

I Bonneville The John IZI V LIZDay

Release Dalles

Site

oI

40 80I I

Kilometers

IO(J)I««S e North Fork

Dworshak

Nat ional Fish

Hatchery

Kooskia

d Natlonal

Fish

• Hatchery

Middle

Fork

South

Fork

Figure I.-J uvenile salmonids reared in Idaho streams must pass through eightdams-from Lower Granite to Bonneville-on their journey to the sea. Studymap shows fish loading sites and the release site.

Figure 2.-A fish barge and tugboat on the Columbia River near the release site (photo courtesy of the U.S. Army Corpsof Engineers).

Columbia River by the WashingtonDepartment of Fisheries in a series ofexperiments from 1955 to 1958. How­ever, an open type (screened) barge was

~

JJIr /979

used (Ellis and oble, 1960).In our operation, two barges (modi­

fied for fish hauling) and a tug boatwere employed. Each barge had a steel

cargo tank 33.2 m (109 feet) long and8.5 m (28 feet) wide (Fig. 2, 3). Thecargo tank was divided into eight com­partments by a longitudinal bulkhead

29

+--------1-- -Longitudinal >'

BUlkhead~ n .~

Oxygen

yp

low euppers

~~ //Cylinders

I I :Y D:TII

I" '!': I C3DTank

Floor?Drain --:- ~I I Pump

n .lJ1Jl-~ No 22 H 3 [

1 f--

F=CJt ]][1

H 4

n n ,.,~ ~

IStandby

II II51 Pump

91 em Dia. :Conduit

1~ 4r-1 I F Pump

m I~No 1= 2 3 u

c:::@J

I~ .ulI ' , II I ' , OTII I : I

7.6em~Bass Stern

30.5 em Dia.

Over! S

/

'::-

~+------~

Transverse /BUlkhead·~

WalkWaY~

>r-..... i::~

@ ---305 em Dia. ~:i:Ff- Spray Header ,;;:

i&\ =~r'-.---1.-3-em-D-ia-.---\ " ..U---l

> '<151 PVC Pipe ~~;~ •

~i~..~~.

Bow

Figure 3.-0verhead view of a fish barge (modified from U.S. Army Corps of Engineers' blueprint). Key to valves: I) Seachest, 2) Recirculation, 3) Pump intake, 4) Back flush, and 5) Main discharge.

and three transverse bulkheads.Hinged screens, 2.7 m (9 feet) by 0.9 m(3 feet), were installed in the trans­verse bulkheads on each side of thelongitudinal bulkhead; the six hingedscreens, constructed of perforated steelplate (perforations were 4.8 mm or 0.2inch in diameter), permitted segrega­tion of fish into eight compartments.

Water Supply System

The basic guideline for the bargedesign was to simulate a modernhatchery pond (raceway) environment.The principal mode of operation wasthe continual pumping of river waterthrough the barge to minimize thebuildup of metabolites in the barge andto provide a continual river water

experience for the fish to avoid inter­ference with their natural homingprocess. A recirculation and oxygena­tion system served as a backup if localchemical contaminants or other factorslimiting river water quality wereencountered. The recirculation andoxygenation systems also minimizedthe cost of treating fish with salt ormedication.

Water was supplied to the bargecirculation system by two diesel­powered pumps (a standby pump wasalso available). Each pump had its ownsea chest-a chamber located in thebottom of the barge providing access toan opening in the hull for intake ofriver water. A spoiler, a triangularmetal structure placed forward of the

sea chest intake and extendi~g alongthe sides of the intake, was designed toprevent air bubbles (passing along theflat bottom of the hull while the bargewas underway) from entering theintakes of the pumps. Air drawn intothe circulation system can cause twomajor problems: I) Supersaturation ofthe water with atmospheric gas and 2) areduction in pump efficiency.

The water was pumped through2S.4-cm (lO-inch) and 30.S-cm (12­inch) diameter pipes, and subsequentlyinto a 91-cm (36-inch) diameter pipethat extended the length of the cargospace (Fig. 3, 4). Water exiting fromthe 91-cm (36-inch) diameter pipeflowed into a 30.S-cm (12-inch) diame­ter perforated pipe which served as a

'-

30 Marine Fisheries Rev/}w

Barge Operations

Capacity and Loading­and- Unloading Procedures

We used Cannady's table (Piper,1970) as a guideline for computing thenumber of fish we could safely trans­port. To compute safe loading densitiesbased on gallons per minute of flow,water temperature, elevation above sealevel, species, and size of the fish wereconsidered. We determined maximumloading capacity to be approximately680,000.chinook salmon (20,408 kg or45,000 pounds of fish averaging 15.2cm or 6 inches) or 350,00 steelheadtrout (26,757 kg or 59,000 pQu!}ds offish averaging 20.3 cm or 8 irich.es).

Juvenile salmonickwere loaded into

1.5 m (4.8 feet). The center measure­ment represented an average becauseeach barge had about 25.4 cm (10inches) of trim.

Auxilliary Oxygen System

In addition to being able to pumpwater directly from the river, watercould be recirculated in the barge byclosing the sea chest valves andopening the recirculation valves.During recirculation, oxygen can beadded to the water thrOl:lgh three airstones-5.l cm (2 inches) in diameterand 2.7 m (9 feet) long-installed in theaft end of the 0.9-cm (36-inch) diameterpipe.

Drain Facilities

Two 25.4-cm (lO-inch) diameterfloor drains, located in the stern of thecargo tank, were used to draip fish imdwater from the barge; the fish werereleased underwater, taking advantageof the head of water within the barge.Water drained readily from the \;>argebecause the barges were' double bot­tomed (the bottom of the cargo tankbeing approximately 0.9 m or 3 feetabove the hull) and had 25.4 cm (10inches) of trim caused by the weight ofthe diesel engines, fuel tank, plumbing,and oxygen cylinders located on the aftdeck. As the barge drained, it rosehigher in the water, thereby maintain­ing the head of water above riverlevel.

."..' ......'

1. SEA CHEST2. RECIRCULATION3. PUMP INTAKE4. BACK FLUSH5. M~AIN DISCHARGE

KEY TO VALVES

modified by: I) Operating either one ortwo pumps; 2) partially closing themain discharge valve; or 3) opening the7.6-cm (3-inch) diameter bypass valve(Fig. 3,4). The average outputs of oneand two pumps were 12.9 m]jminute(3,400 gpm) and 20 m3

j minute (5,250gpm), respectively. With two pumpsrunning it took approximately 20minutes for one complete turnover ofwater in the barge.

If the intakes of the sea chests be­came plugged with debris, they couldback flushed by opening the back flushand recirculation valves and closing thesea chest valves, thus allowing water inthe pipe above the pumps to fall backthrough the sea chest (Fig. 4). Thespray bar could be cleaned of debris byopening butterfly valves on each end.

When the water level in the center ofthe barge reached approximately 1.1 m(3.5 feet), water began flowing out fourscreened 30.5-cm (l2-inch) diameteroverflow scuppers located on the sidesof the stern. Maximum water depth inthe center of the cargo area, with twopumps operating, was approximately

FIgure 4.-Diagram of the stern plumbing on the fish barge: Arrows indicatewater flow during normal circulation. Valve handles parallel to the flow indicateopen valves.

spray bar and was positioned acrossthe forward end of the cargo tank. Thespray bar had l2.7-mm (0.5-inch)diameter holes on 50.8-mm (2-inch)centers; the holes were positioned inthe horizontal plane on each side of thebar. A steel baffle, located aft of thebar, and the forward wall of the cargotank helped diffuse the water.

The number and size of the holes inthe spray bar were designed to create aback pressure on the pumps, thusincreasing the water pressure in the 91­cm (36-inch) pipe. The increased waterpressure, approximately 1.1-1.3 kgj cm(16-18 poundsjinch 2

) with two pumpsoperating, facilitated the uptake ofoxygen by the water when the systemwas operating in the recirculatingmode. Pressure gauges on the spraybar and the discharge pipe from thepumps were used to monitor changesin the circulation system. Large pres­sure changes indicated a reduction inwater flow; i.e., an abnormal increasesignaled a plugging of the spray barwith debris.

Water flow to the spray bar could be

JuJy /979 3/

2Park. D. L.. E. Dawley. R. Krcma. C. Long.E. Slatick. J. Smith. and G. Swan. 1976.Evaluation of fish protective facilities at l.ittleGoose and I.ower Granite Dams and review ofother studies relating to protection of juvenilesalmonids in the Columbia and Snake Rivers.1975. Northwest and Alaska Fisheries Center.National Marine Fisheries Service. NOAA.2725 Montlake Blvd. E.. Seattle. WA 9X112.Unpubl. rep., 50 p.

the barge through either ID.2-cm (4­inch) or 15.2-cm (6-inch) diameter

.- flexible PVC hoses. All the sharp edgeson hose couplings were either cham­fered or covered to prevent mechanicalinjury to fish. When fish were unloadedfrom tank trucks onto the barge, weused a 3.8-cm (1.5-inch) diameter firehose connected to the barge's circula­tion system to flush fish from the truckand the loading hoses. The fire hosewas also used during the unloading toflush tJ1e last fish from the barge.

Fish were released in the middle ofthe main channel of the ColumbiaRiver, approximately 9 km (5.6 miles)downstream from Bonneville Dam.Predatory fish, particularly northernsquawfish, Ptychocheilus oregonen­sis, are not as abundant in the fastflowing midchannel regions as they arein shoreline habitats. While fish werebeing discharged, the tugboat wasupstream from the barge backing intothe current, thereby assuring the fishweren't struck by the tug's screws.Under normal circumstances somewater was pumped through the spraybar during the release, so oxygen levelsremained optimal.

On 5 April 1977, two NMFS scubadivers viewed the release from under­neath the barge and reported the fishappeared to be in good condition.Therefore, by direct observation wewere able to determine that fish werereleased in a safe manner.

Chemical Treatments

A limited experiment involvingtransportation of juvenile salmonids insalt water was conducted during thebarging program. Previous work by

M FS biologists indicated highersurvival rates for chinook salmon 2

transported by" truck in a 5%" salinesolution. However, the effect of usingsalt water during transport has not

Loading springchinook salmonfrom KooskiaNational FishHatchery intobarge at Lewis-'ton. Ida. Photo­graphs courtesy

'of the U.S.,Army Corps ofEngineers.

Barge loading site at Lewiston, Idaho. The fish tanker truck is arriving fromKooskia National Fish Hatchery to unload its ca rgo of spring chinook salmoninto the barge. Photograph courtesy of the U.S. Army Corps of Engineers.

32 Marine Fisheries Rel'few

Loading spring chinook salmon from Kooskia National Fish Hatchery into bargeat Lewiston, Idaho. Photographs courtesy of the U.S. Army Corps of Engineers.

been fully evaluated based on returningadults subjected to the treatment asjuveniles. Consequently, only the firstbarge load of chinook salmon andsteelhead trout transported fromLower Granite Dam was subjected toa salt treatment.

The fish were exposed to a 9.5%0salt solution (sodium chloride) for Ihour. The salinity was monitored witha YSI salinometer and was achievedby pumping a concentrated saltsolution, prepared in a brine tank onshore, to the barge. The fish collectedat Lower Granite Dam that weretreated in the salt solution appeared totolerate the increased salinity well.Benefits from the salt treatment willnot be known until adults return in1978-81.

On three occasions we considered itprudent to disinfect the barges becauseof the possible contamination bydiseased hatchery fish that had beentransported. To disinfect, we usedcalcium hypochlorite (65 percentstrength) to achieve a chlorine concen­tration of 18.9 ppm. Water was

'Reference to trade names or commercial firmsdoes not imply endorsement by the NationalMarine Fisheries Service, NOAA.

Julj' 1979

recirculated for at least 2.5 hours whiledisinfecting; the residual chlorine wasthen neutralized with sodium thiosul­phate before draining the water.

Oxygen andTemperature Monitoring

Oxygen levels and temperatures inthe barge were monitored with a YSImodel 51 B dissolved oxygen meter.Water directly below the spray bar andnear the overflow scuppers in the aftcompartments was checked periodical­ly for dissolved oxygen.

Oxygen levels under the spray barwere typically near 100 percent satura­tion. The lowest oxygen value recordedin the aft compartment was 7.8 ppm at13°C (55°F). A simultaneous readingin the forward compartment under thespray bar was 10 ppm at 13°C (55° F).The barge was loaded with 153,596steelhead trout at the time of the 7.8ppm reading. Oxygen values at differ­ent depths did not vary significantly­surface reading and measurements 0.5m (1.5 feet) abqve the bottom weregenerally within 0.2 ppm. In all trans­ports by barge there was no increase inthe water temperature between theforward and aft compartments.

Potential Problem Areas

The general performance of thebarge was excellent. However, therewere some minor problems involvingthe pumping system and the unloadingprocedures that merit mention. Duringperiods of rough water on the river, theefficiency of the pumps was reduced.When large waves struck the barge theflow to the spray bar was brieflydecreased, indicating that perhaps thepumps were sucking air for a shorttime. These circumstances were rareand had little effect on transportoperations.

On several occasions the holes in thespray bar became partially pluggedwith river debris, primarily smallwooden sticks, and the bar had to becleaned. The plugging did not create aserious problem; however, the need tomonitor pressure change in the pipewas demonstrated to avoid pluggingand serious reduction of flows.

Although most of the fish were easilydischarged from the barge, the lastseveral hundred fish of a load oftenwere reluctant to leave. Sometimes fishwould hide under the 91-cm (36-inch)diameter pipe that extended along thelength of the barge. Fire hoses proveduseful in flushing these remaining fishtoward the aft drains.

The Transport ProgramA total of 3,517,242juvenile salmon­

ids were transported from 5 April to 5June 1977 (Table I); breakdown byspecies was: Coho salmon, 0. kisutch,21,777; chinook salmon, 3, 111,159;and steelhead trout, 384,306. Thelongest hauls were made fromLewiston, Idaho, to below BonnevilleDam - 520 km (323 miles). LowerGranite Dam fish were transported468 km (291 miles). Transport timesfrom Lewiston, Idaho, and fromLower Granite Dam to the releasesites were about 34 and 30 hours, re­spectively.

When hatchery fish were barged, weestimated the mortality associated withloading, transporting, and unloadingto be less than 0.5 percent. Theestimated mortality rates of chinooksalmon and steel head trout fromLower Granite Dam were 3.6-7.5

33

Table 1.-Juvenlle salmonlds transported by barge from various sourcel end loading lites on the Columbia andSnake River. In 1977.

5/26 Chinook salmon 24.498 10.198 Collector atSteelhead trout 36.695 10.097 Lower Granite

Dam

Chinook salmon 103.200 10.510 Collector atSteel head trout 70.118 10.115 Lower Granite

Dam

Steelhead trout 155.148 17.178 oworshakNatIOnal FIshHatchery

homing ability of adult salmonids,captured as juveniles while they wereactually migrating downstream andthen transported In trucks, has notbeen seriously impaired; therefore, it isreasonable to assume that the homingability of barged fish should not bereduced. It is also theorized that thecontinual exposure of the fish to riverwater throughout the downstreambarge journey might ensure homingsuccess with fish transported directlyfrom hatcheries.

The success of the barge as a usefultransport mode will be fully evaluatedwhen ad ults return in 1978-81. Becausereleases of truck-transported fish weremade concurrently with barge releases,a direct comparison between barge andtruck transport systems can be made.Adults returning from all experimentalreleases will be recorded at LowerGranite Dam, at hatcheries, on natalspawning grounds, and in commercialand recreational fisheries.

Acknowledgments

We thank personnel from the U.S.Army Corps of Engineers for theirmajor roles in the preparation of thefish barges and the loading site atLower Granite Dam. The project wasfinanced by the U.S. Army Corps ofEngineers. Financial assistance wasalso provided by Chelan, Douglas, andGrant County Public Utility Districtsfor the barge operation from Richland,Wash., on I June 1977.

All fish. except lastload of 28.200 remamedIn Clearwater River for12 hours.

Ongmally, fish camefrom Chelan Hatchery

Remarks

An additional 110adult steel head(spawned out) werealso transported.

Loaded barge remainedIn Clearwater Riverfor 14 hours.

Lewiston,Idaho

Lewiston,Idaho

Lower GranIteDam

Lower GraniteDam

LoadIngsite

Spring CreekNational FishHatchery

Spring CreekNatIOnal FIshHatchery

orano Lake,Wash

Spring CreekNational FishHatchery

Priest RapIds North RIchland,spawning channel Wash.LeavenworthNational FishHatcheryWells spawningchannel

Source

Spring CreekNational FishHatchery

Spring CreekNational FishHatchery

Collector al Lower GraniteLower Granite DamDam

21.777 WillardNational FishHatchery

Number

50.160

Spring Creek935.939 76.057 National Fish

Hatchery

999.575

Unmarked MarkedSpecies

Fall chinooksalmon

Fall chinooksalmon

Fall chmooksalmon

COhO salmon

3.161.504 355.738

Fall chmooksalmon 241.000

Steelhead trout 48.455

Spring chmook KooskIasalmon 329.430 31.200 National Fish

Hatchery

Su mmer chinooksalmon 133.876 99.113

Chinook salmon 55.483 10.920Steelhead trout 26.382 10.118

5/5

DateloadingstartedMo/day

5/4

6/1

4/19

4/11

4/12

4/22

6/2

4/5

Literature Cited

percent and 2.2-5.3 percent, respective­ly. The apparent reason for theincreased mortality among fish collect­ed at Lower Granite Dam was that fisharriving at the dam were observed to bedebilitated, probably because of lowstream flows which caused a laboriousmigration.

No sea gulls took any fish out of thebarge while enroute to the release site,although sea gulls (10-20 birds) weresometimes noted feeding near therelease point.

Fish barging should not be viewed asa replacement to the trucking of fish,but as a complementary method offish

transport. Barging is best used whenlarge numbers of fish (i.e., greater than100,000) have to be transported fromcollection facilities. Moreover, bargingcan be efficiently employed in trans­porting large numbers of fish fromhatcheries. Trucks are necessaryduring the time periods (often severalweeks) before and after the peak of thejuvenile salmonid migrations. Tankertrucks carry a small fraction of abarge's capacity; therefore, they arewell suited for use during the pre- andpost-peak migrations of juveniles.

Biologists of the NMFS (Ebel et aI.,1973) have demonstrated that the

Chaney, E., and L. E. Perry. [1976.] ColumbiaBasin salmon and steelhead analysis: Sum­mary report I September, 1976. Pac. North­west Reg. Comm.. Vancouver. Wash., 74 p.

Ebel, W. J., D. L. Park, and R. C. Johnsen.1973. Effects of transportation on survivaland homing of Snake River chinook salmonand steelhead trout. Fish. Bull., U.S. 71 :549­563.

Ebel. W. J., and H. L. Raymond. 1976. Effectof atmospheric gas supersaturation on salm­on and steelhead trout of the Snake andColumbia Rivers. Mar. Fish. Rev. 38(7):1- 14.

Ellis, C. H., and R. E. Noble. [1960.] Bargingand hauling experiments with fall chinooksalmon on the Klickitat River to test effectson survivals. Wash. Dep. Fish., 70th Annu.Rep., p. 57-7 I.

Piper, R. G. 1970. Know the proper carryingcapacities of your farm. Am. Fishes U.S.Trout News 15( 1)4-6.30.

34 Marine Fisheries ReI'lell'