This article was downloaded by: [US Fish & Wildlife Service] AR-1256 On: 02 February 2012, At: 07:25 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Transactions of the American Fisheries Society Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/utaf20 A Review of Dissolved Gas Supersaturation Literature Don E. Weitkamp a & Max Katz b a Parametrix, Incorporated, 13020 Northup Way, Suite 8, Bellevue, Washington, 98005, USA b Environmental Information Services Incorporated, 3455 72nd Place Southeast, Washington, 98040, USA Available online: 09 Jan 2011

To cite this article: Don E. Weitkamp & Max Katz (1980): A Review of Dissolved Gas Supersaturation Literature, Transactions of the American Fisheries Society, 109:6, 659-702 To link to this article: http://dx.doi.org/10.1577/1548-8659(1980)109<659:ARODGS>2.0.CO;2

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and­ conditions This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. The publisher does not give any warranty express or implied or make any representation that the contents will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims, proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in connection with or arising out of the use of this material. Transactionsof the American Fisheries Society 109:659-702 @ Copyrightby the AmericanFisheries Society 1980

A Review of Dissolved Gas Supersaturation Literature

DoN E. WEITK^Me Parametrix,Incorporated, 13020 NorthupWay, Suite 8 Bellevue,Washington 98005

MAX KATZ EnvironmentalInfo•ation ServicesIncorporated, 3455 72ndPlace Southeast MercerIsland, Washington98040

Abstract Dissolvedgas supersaturation is a conditionthat resultsfrown natural and human-causedpro- cesses.Supersaturation can result in gas bubble diseasewhich has been describedin a wide varietyof fishesand invertebrates.In recentyears dissolved gas supersaturation resulting from dams and thermal dischargeshas produced mortalitiesof fish in severalcases. This review discussesmost of the available literature dealing with dissolvedgas supersaturationand the recordedcases of gasbubble disease.

Gas bubble disease is a condition that affects conditions. Much remains to be learned about aquatic animals residing in fresh or marine the physiologicalaspects of gasbubble disease. watersthat are supersaturatedwith atmospher- In order to understandgas bubble disease ic gases.Supersaturation, and the gas bubble and its cause,it is necessaryto be familiar with diseasethat may result in aquaticorganisms, the physicallaws governing dissolved gases and are not recent discoveriesnor are they only the factors that determine the level of super- causedby human activities.However, only in saturation.Boyer (1974), Woelke et al. (1974), recent years has supersaturation become a and Harvey (1975) discussedthe solubilitiesof problemof sufficientmagnitude to draw wide- dissolvedgases in water as they relate to gas spreadattention and concern. bubble disease. The majority of research dealing with dis- Harvey (1975) provided an excellentdiscus- solvedgas supersaturation has been stimulated sionof this subjectfor thosenot familiar with by a problem of considerablemagnitude that the physicallaws describingthe solubilitiesof wasobserved in the ColumbiaRiver systembe- gasesin a liquid. The solubilityof atmospheric ginning in the 1960's. More recently, interest gasesin water is determinedby the water'sdis- hasbeen further stimulatedby the discoveryof solved solids content, characteristics of the var- deleteriouseffects of supersaturationresulting ious gases,the total pressure, and the water from thermal effluents. temperature. Although total dissolvedsolids This reviewis an attemptto providea greater canaffect solubility, this is not a significantvari- Downloaded by [US Fish & Wildlife Service] at 07:25 02 February 2012 disseminationof the availableexisting knowl- able in most fresh waters but must be consid- edge regarding dissolvedgas supersaturation ered as a significantvariable in marine waters. and the resultinggas bubbledisease. The re- The atmosphericgases of importanceare ni- view discussesthe causesof supersaturation, trogen, oxygen, and argon. These gasesare the organismsaffected by supersaturation,fac- present in air at partial pressuresof approxi- tors affecting susceptibilityof aquatic organ- mately78% nitrogen,21% oxygen,and 1% ar- ismsto gas bubbledisease, and variousother gon. Nitrogen and argon are normallyconsid- related topics.The k•owledge of this subjectis ered together becauseboth are biologically considerableas evidenced by the length of this inert gaseswhile oxygenis a biologicallyactive review. Many important questionsremain to be gaso answered.This is particularlytrue regarding The solubilityof each gas is determinedby the applicationof laboratoryresults to condi- the massof the individual gas and its partial tionsfaced by aquaticorganisms under natural pressurein the atmosphere.Oxygen (21%) has 659 660 WEITKAMP AND KATZ

onlyone-fourth the partialpressure of nitrogen marine fishes.Bubbles frequently occurred be- (79%) in the atmosphere,but is twiceas soluble hind the cornea and in the loose connective tis- asnitrogen. Therefore, in water,oxygen (35%) suesof the eye, producinga severeexophthal- is one-halfas plentiful as nitrogen (65%) (Har- mia or "pop-eye" condition. Bubbles were vey 1975). found lessfrequently in the gills,lining of the The major environmentalfactors that affect mouth, or along the lateral line of exposedfish. solubilityare pressureand temperature.Ac- Thesebubbles gradually increased in sizeas the cordingto Henry'sLaw, the massof a gasdis- length of exposure to supersaturation in- solvedin a liquid at a constanttemperature is creased.Fish with these signsalso showed loss proportionalto the pressureexerted on the sol- of equilibrium.The time to death varied from vent. Thus, as the pressureon a givenvolume severalhours to severalweeks following the ap- of water increases,the capacityof that volume pearanceof detectablebubbles. of water to hold dissolvedgas also increases. Marsh and Gorham (1905) further described Pressureis increasedin water by hydrostatic internal signsand lesionsof the disease.Free head. Hydrostaticpressure increases rapidly gas bubbles(or emboli) within blood vessels with depth, greatlyincreasing the capacityof were observed. The amount varied from a few deeperwater to hold dissolvedgas as compared scatteredbubbles to complete occlusionand to shallow water. distention of the vessels. The walls of the au- The capacityof water to hold dissolvedgas ricle and ventriclewere often emphysematous. is inverselyrelated to temperature.As the tem- In somefish, the auricle wasfilled with gaseven perature of a volume of water increases,the though it continuedto beat. The main vessels volumeof dissolvedgas it will hold at equilib- of the gills containedgas bubbles.Gas in the rium decreases.Thus, increasingwater tem- gill filamentswas described as the mostconstant peratureswill producesupersaturation in water and significantlesion of gas bubble disease. that is initially saturated. Death of the fish was attributed to stasis of the This brief reviewof the factorsaffecting sol- blood causedby emboli. ubilityonly discussesa few of the majorvari- In addition to observations of fish, Gorham ables. Those interested in a more detailed re- (1901) also reported signs of the diseasein view of the subject should refer to Harvey squid, bivalve mollusks, scallops, hydroids, (1975) or to textsdescribing the gaslaws in de- squid egg-sacs,and green algae. No detailed tail. discussionof the diseasein theseorganisms was History of Gas BubbleDisease given by Gorham. Gorham(1901) experimentallyproduced gas Early Observations bubble disease in fish held in closed containers There are severalnineteenth century records by reducingthe pressure.He alsowas able to of what appearsto have been gasbubble dis- causethe signsto disappearby subjectingfish ease.Hoppe-Seyler (1857), Bert (1873), and to a pressurecomparable to that exerted by a Regnard(1884) recorded external signs in fish water depth of 4.9 m. He concludedthat gas that apparentlyrepresent gas bubbledisease. bubbledisease was caused by a reductionin the The first completedescription of gas bubble pressureto whichfish normally were subjected. Downloaded by [US Fish & Wildlife Service] at 07:25 02 February 2012 diseaseand its causewas provided in a seriesof Marsh and Gorham (1905) later corrected this papersresulting from an air supersaturationmistaken conclusion. problemat the United StatesBureau of Fish- Gorham (1901) reported that whereasfish in eries station at Woods Hole, Massachusetts shallowaquaria developedgas bubble disease, (Gorham 1898, 1901;Marsh 1903, 1910;Marsh fishheld in ponds24 m deepwith waterfrom and Gorham 1905). This seriesof initial papers the same source remained free of it. This was dealtwith mortality,signs, and experimentsto the first indicationof the major differencebe- determine and correct the cause of the disease tween artificiallyshallow water conditionsand in aquarium fish. the more natural situationsthat permit hydro- The first descriptionof the outwardsigns of staticcompensation. gasbubble disease was given by Gorham(1901). Later, Marsh and Gorham (1905) discussed Vesicles(gas bubbles or blisters)were found in the solubilityof gasesin water and the relation- the fins and other external surfaces of several ship of respiratoryprocesses to gasbubble dis- DISSOLVED GAS SUPERSATURATION--A REVIEW 661

ease. They concluded that the diseasewas uration. This situationoccurred during winter causedchiefly, if not solely,by excessivedis- conditionswhere photosyntheticactivity oc- solvednitrogen gas. This conclusionwas based curred under clear ice, producing excessgas on the analysisof bubblesfrom tissuesand that couldnot escapeto the atmosphere.Wiebe bloodvessels of animalswith gasbubble disease. and McGavock(1932) experimentallyexposed Thesebubbles contained 92% to 97% nitrogen, a variety of fishesto dissolvedoxygen concen- the remainder being oxygen.Bubbles formed trations two to three times saturationfor pe- in the supersaturatedwater had nitrogenand riodsof 20 to 50 days.No signsof the disease oxygenin the sameratio asfound in air. Marsh were found in thesefish even though evidence and Gorham concludedhemoglobin has the ca- of it wassought. pacity to modify the effect of oxygen by re- Mrsic (1933) observedthat fish reared in tap movingit from the dissolvedstate. watersuffered gas bubble disease when oxygen Marsh and Gorham (1905) reported several and nitrogen concentrationswere below satu- instancesof naturallyoccurring supersaturated ration but carbon dioxide was supersaturated. fresh waters.Rainbow trout (Salmogairdneri) in Mrsic reported 75% of fish in water with 138 these waters showedthe samesigns of the dis- mg/liter carbon dioxide suffered gas bubble ease as the marine fish from Woods Hole. Ex- diseasewhereas no fishin water having 135 mg/ perimentallythey determinedthat trout and liter carbon dioxide showed signs of it. Al- somecyprinids have nearly equalsusceptibility though Mrsic attributed the diseaseto carbon to gasbubble disease, whereas goldfish (Caras- dioxideit appearsunlikely that a rise of only 3 siusauratus) are not affectedby the samelevels mg/liter at those high concentrationswould of supersaturation. producesuch a high incidenceof the disease. The preventionof the diseaseby removalof Embody(1934) reported gas bubbledisease excessdissolved gases through aeration was dis- in trout fry at the Cornell Universityhatchery. cussedby Marsh and Gorham (1905). The su- The fry were hatchedin water that had been persaturationcondition at WoodsHole wascor- heatedto raisethe temperature5 C. Unhatched rected by replacing an intake pipe that had eggs were apparentlyunaffected whereas the allowedair to be suckedinto the water supply. yolk sacsof newlyhatched fry were distended Marsh (1910) later removed excessgases by by gas bubbles.The diseasewas attributed to trickling water over stacksof perforated shal- excessdissolved nitrogen and was prevented low pans. through aeration,which was accomplishedby In this seriesof papers, Marsh and Gorham passingthe water over a seriesof baffles. establishedthe basicknowledge of gasbubble Woodbury (1941) describeda suddenmor- disease.Most subsequentinvestigations have tality of fish showingsigns characteristic of gas confirmedand expandedon their work. Any- bubble disease in a Wisconsin lake. Bubbles one seriouslyinterested in the problemwould wereobserved in the gillsof thesedying fish as be well advisedto read these early works, in well asbetween fin raysand under scales.The particular,Marsh and Gorham (1905). diseasewas attributed to dissolvedoxygen levels During the 40 years followingthe work by in excessof 300% saturation.This mortalityfol- Marsh and Gorham a few scatteredreports of lowedan extensivealgal bloom during a period Downloaded by [US Fish & Wildlife Service] at 07:25 02 February 2012 gas bubbledisease appeared in the literature. of sunnyweather. Shelford and Allee (1913) encountered it in ex- Rucker and Tuttle (1948) noted a hatchery perimentsto test the reactionof fish to gradi- water supplyat Leavenworth,Washington, was ents of atmosphericgases. Supersaturation was supersaturatedwith nitrogen.Fish subjected to producedby raisingthe temperatureof water this water were reported to suffer gasbubble about 9 C. This studywas designed to achieve diseasealthough few detailsare given. other objectivesand made no new contributions In the late 1940's, the disease was encoun- to the understandingof the disease. tered in larval marine fishat a Swedishhatchery Plehn (1924) reported the occurrenceof gas (Dannevigand Dannevig 1950). Larval herring bubble diseasedue to supersaturationbrought (Clupeaharengus) and flatfish developed bub- aboutby photosyntheticactivity. Dissolved ni- bles in the intestine. The fiatfish larvae were trogen concentrationswere not measuredbut able to passthe bubblesthrough the anuswhile thoseof dissolvedoxygen were three timessat- the herring were unable to do so and suffered 662 WEITKAMP AND KATZ

extensivemortalities. The causeof the super- (1967) described the occurrence of the disease saturationwas air leaksin the watersupply sys- during 1965 in juvenile chinooksalmon held in tem. Henly (1952) describeddifferences be- aquaria at Rocky Reach Dam on the mid-Co- tween physoclistousand physostomousfish lumbiaRiver. These papers do not describethe larvaewith the diseaseat thishatchery. She also sourceof the aquariumwater whichprobably noted that oysterlarvae suffer great mortality wassupersaturated river water. when subjectedto supersaturatedwater. Su- The first indicationthat a serioussupersat- persaturationin the Swedishhatchery water uration problem existedthroughout the Co- supply was corrected by passing the water lumbia River system was provided by Ebel throughtwo sandfilters. (1969) and Meekin (1971). Monitoring of dis- Severalcases of gas bubbledisease were re- solvedgas levelsin the Columbiaand Snake ported in the 1950's.Rucker and Hodgeboom rivers during high-flow periods showed that (1953) describedthe diseasein salmonidyolk levels of 120% to 130% saturation occurred in sacfry reared in a springwater supplywhich 1966 and 1967. Migrating juvenile chinook had oxygenat 70% of saturationand nitrogen salmonwhich were examinedat PriestRapids at 120% of saturation.Supersaturation was re- Dam on the mid-ColumbiaRiver during 1966 ducedby passingthe water through an 18-m- showedsigns of gas bubbledisease. Consider- long agitationweir. able increasesin the migration time of these Matsueet al. (1953) reportedthe diseasein fishresulted in long periodsof exposureto su- fishesheld in supersaturatedwater supplies in persaturation(Raymond 1968, 1969). The in- Japan. Fish mortalitiesoccurred in watershav- creasedlength of exposurecoupled with the ing nitrogenlevels from 117% to 158%of sat- high levelsof supersaturationapparently were uration. Fishreportedly could not live in water responsiblefor the appearanceof the disease with nitrogenlevels above 130% of saturation. in the migratingjuveniles. Live-cagestudies Satomi (1955) indicated there was no fear of withjuveniles at PriestRapids Dam in 1967also gas bubble diseasein adult trout reared in indicatedthe sameproblem. Adult salmonids springwater that had dissolvednitrogen levels were also observedfor signsat severallower under 120% of saturation. Columbia River dams in 1967; a small number Renfro (1963) attributed the death of nu- of them showed indications of the disease. merousmarine fishesin GalvestonBay, Texas, In 1968, signsof gas bubble diseasewere to this disease.The mortality occurredfollow- againobserved in juvenilesalmonids in the low- ing a periodof calmsunny weather and intense er ColumbiaRiver (Beiningenand Ebel 1970). photosynthesis.On the day followingthe mor- High mortalitiesoccurred among juveniles held tality, dissolvedoxygen concentrationswere for inspectionat The Dalles Dam, where, at 250% of saturation. times,approximately half the fishshowed signs These early instancesof gas bubbledisease of the disease. In addition, there were several were very minor in size and duration. In the mortalitiesof adult salmonidsshowing similar Columbia River system,the diseasehas been signsdownstream from the recentlycompleted recognizedas a seriouslong-term problem af- John Day Dam. At John Day Dam in 1968, all fectinga large area and numerousfish.

Downloaded by [US Fish & Wildlife Service] at 07:25 02 February 2012 water passingthe dam traveledover the spill- waybefore the turbineswere installed. This sit- ColumbiaRiver System uationproduced dissolved nitrogen saturations In the mid-1960's,it graduallybecame evi- of 123-143% downstream from the dam. Prob- dent that a seriousdissolved gas problem exist- lemswith fish-passagefacilities further compli- ed in the Columbia River system.Westgard catedthe situation,causing delays of migrating (1964) observed adult chinook salmon (Onco- adult salmonidsin the highly supersaturated rhynchustshawytscha) suffering gas bubble dis- water. It was estimated that over 20,000 sum- easeat the McNarySpawning Channel in 1962. mer chinook salmon were lost in this area dur- In this case,supersaturation was caused by the ing this episode. spawningchannel intake. A dissolvednitrogen Ebel (1971) and Raymond (1970) reported level of 119% of saturation was measured in an mortalitiesand signsof gasbubble disease in area of the channelwhere the fish spentcon- bothjuvenile and adult salmonidsin the Snake siderable time. River during 1970. Accordingto Ebel, dis- Pauleyet al. (1966) and Pauleyand Nakatani solvednitrogen levels in the river waterranged DISSOLVEDGAS SUPERSATURATION--A REVIEW 663

from 120% to 140% of saturation for well over a hatchery on the South Santiam River in Or- a month.Juvenile chinook salmon held in live- egon. As in the CultusLake Hatcheryepisode cagesat Ice Harbor Dam sufferedsevere mor- (Harvey and Smith 1961), supersaturationwas talitiesduring this period. due to conditionsthat permitted air to be Meekin and Allen (1974c) estimatedthat 6% suckedinto the watersupply. Apparently most to 60% of adult salmonidsin the middle region of thesefish died within hours when subjected of the Columbia River died between 1965 and to supersaturationapproaching 150%. 1970. Carcasses of adult salmon were found in Although supersaturationat hydroelectric the river when N2 supersaturationreached projectsnormally resultsfrom spillingwater, 120% or higher. Few carcasseswere found MacDonaldand Hyatt (1973) reported super- when nitrogen saturations did not exceed saturationwas causedby air vented into tur- 112%. bines. Atlantic salmon (Salmosalar) and Amer- A general review of the supersaturation ican eels(Anguilla rostrata) suffered gas bubble problemin the ColumbiaRiver systemthrough diseasebelow the MactaquacDam on the St. 1970 waspresented by the EnvironmentalPro- John River, New Brunswick.An estimated200 tectionAgency (USEPA 1971). That brief re- Atlantic salmon were killed in this incident. view wasprepared prior to completionof the Nitrogen saturation of 118-125% was mea- manyreports available today. sured downstream of the dam. Dissolvedgas levels in the Columbia River Several recent disease incidents have oc- systembetween 1965 and 1969 were given by curred at steam-generatingfacilities. DeMont Beiningen and Ebel (1971). In each of these and Miller (1972) reported mortalities and years,dissolved gas levels in excessof 120% of signsof gasbubble disease in severalspecies of saturation were measured. At times, dissolved fish at the Marshal Steam Station on Lake Nor- nitrogensaturations exceeded 140%. man, North Carolina,during 1970-1971. The May (1973) describedmortalities due to gas diseaseoccurred among fish in the discharge bubbledisease in 1973 below the recentlycon- area during late winter and spring. No dis- structedLibby Dam in the upper reachesof the solved gas measurementswere taken in this ColumbiaRiver system.All mountainwhitefish study,though Adair and Hains (1974) calculat- (Prosopium williamsani) and cutthroattrout (Sal- ed that dissolvedgas levels during thismortality mo clarki) held in shallowlive-cages were dead periodreached a high of 144%of saturationin within 4 days at total dissolvedgas saturations March. The supersaturation resulted from above 130%. When average total gas levels temperatureincrease in coolingwater at this droppedbelow 120%, mortality rates dropped steamstation. Miller (1974) alsoreported signs markedly. Most fish, in an area of the river ex- of gas bubbledisease in fishesat the Marshal ceeding 130% of saturation, showedsigns of SteamStation during 1971-1972, but the inci- the disease,whereas fish collectedfrom a down- dencewas lower than in the previousyear and stream area with saturations of 105-118% maximum dissolvedgas saturation levels of showedno sign of it. 131% were recorded. Raymond (1979) reviewed the history of Marcelloand Strawn(1972) attempted to cul-

Downloaded by [US Fish & Wildlife Service] at 07:25 02 February 2012 salmonmigrations in the Columbiaand Snake ture three fishspecies in the dischargecanal of rivers between 1964 and 1975. This review in- a steam-electricgenerating station at Galveston cludesdescriptions of dissolvedgas supersatu- Bay, Texas.They attributedhigh mortalitiesof ration problemsduring these years and im- thesefish to gasbubble disease. provingdissolved gas conditions by 1975.The Marcello and Fairbanks (1975)discussed a supersaturationproblem in the ColumbiaRiver massmortality of Atlantic menhaden(Brevoor- systemhas since been essentiallyeliminated. tia tyrannus)at the Pilgrim Nuclear PowerSta- Ebel (1979) stated that in the Columbia and tion, Boston,Massachusetts, in April 1973. An Snake rivers "... fishery agenciesbelieve the estimated 43,000 Atlantic menhaden died in problemof supersaturationand corresponding this brief kill. Fish examinedshowed typical lossesof fish to gasbubble disease is solved." signsof gasbubble disease. Other fishobserved near the thermalplume by scubadivers showed Other Recent Observations no indicationof the disease.Total dissolvedgas Wyattand Beiningen(1971) encounteredgas levelswere not determined;however, oxygen bubble diseasein trout and salmonjuveniles at levelswere as high as 142% of saturation. 664 WEITKAMP AND KATZ

Gas supersaturation(up to 110%) of alpine A muchmore commonly reported sign of gas streamsin Oregon,due to geothermalheating, bubble disease is bubbles or blisters under the was reported by Bouck (1976). Fish in the skin, primarily between fin rays. Marsh and streamsshowed no evidenceof gasbubble dis- Gorham (1905) described vesiclesin the fins ease,but fishin a trout hatcherysupplied with and on other external surfaces of several ma- such water died from the disease. rine fishes. The size and location of these bub- Thesecases demonstrate the widespreadoc- bles vary with the severity of the diseaseand currenceof gasbubble disease and the variety the speciesof fish. In addition to locationsin of situationsthat can cause supersaturation. the fins, blistersfrequently occur on the head Eachof thesefactors is discussedseparately, in and in the lining of the mouth(Marsh and Gor- greater detail, below. ham 1905;Weitkamp 1974, 1976;Dawley and Gas Bubble DiseaseSyndrome Ebel 1975; Dell et al. 1975; Dawley et al. 1976;Nebeker and Brett 1976).Photographs of ExternalSigns thesesigns are in reportsby Blahm(1974), Ebel Detectionof gas bubble diseaserequires fa- and Raymond (1976), Nebeker and Brett miliarity with the externalsigns of the disease. (1976), and Weitkamp (1976). These bubbles Thesevary from blatantlyobvious signs to sub- graduallyincrease in size as the length of ex- tie signsthat will be recognizedonly with care- posureto supersaturationincreases (Marsh and ful study. Gorham 1905). The first descriptionsof the outwardsigns of Accordingto Meekin and Turner (1974), the the diseasewere given by Gorham (1901) and cutaneous bubbles were the most common ex- Marshand Gorham(1905). Numerous descrip- ternal signin juvenilechinook salmon that died tionsof similarexternal signs followed for oth- of the diseasein their bioassaystudies. These er speciesof fish (Shelford and Allee 1913; authors and Weitkamp (1976) reported gas- Woodbury 1941; Egusa 1959; Renfro 1963; filled bubbles were most common in the caudal Shirahata 1966; Wood 1968; Beiningen and fin. Bubbleswere less frequent in the paired Ebel 1970; Wyatt and Beiningen1971; DeMont fins,and onlyoccasional in the anal fin. Bubbles and Miller 1972; Elling and Ebel 1973; Dawley on the head, opercles,jaws, and mouth gener- and Ebel 1975; Dell et al. 1975; Dawley et al. ally occurredonly after the appearanceof bub- 1976; Nebeker and Brett 1976; Weitkamp blesin the fins (Weitkamp1976). 1976). These and many other recent papers Hemorrhagesfrequently accompany bubbles describethe signsdiscussed below. in chronic gas bubble disease.Meekin and Exophthalmiaor pop eye is a signcommonly Turner (1974) observedhemorrhages in about associatedwith gasbubble disease. Marsh and 38% of fishin their bioassays,frequently at the Gorham (1905) found vesicles(gas bubbles or baseof the paired fins but seldomat the base blisters)frequently occurred behind the cornea of the caudalfin. Stroudet al. (1975)suggested and in the looseconnective tissue of the eye, hemorrhagemay be a secondaryeffect of em- producing a severeexophthalmia. Failure to boli. The hemorrhagesusually occur following observethis obvioussign should not be consid- the appearanceof bubblesin the fins,and seem

Downloaded by [US Fish & Wildlife Service] at 07:25 02 February 2012 ered evidencethat gasbubble disease does not to indicatean advancedstage of the disease,at exist,as exophthalmia may be absentor present least in chronic cases. in only a few of the fish sufferingthe disease. Bubblesalong the lateralline are the first ex- Meekin and Turner (1974) reported exo- ternal sign of gasbubble disease to appear in phthalmiain lessthan 5% of deadjuvenile chi- juvenile salmonids(Schiewe and Weber 1976; nook salmonsuffering gasbubble disease.Ex- Weber and Schiewe 1976). These bubbles are ophthalmiaalso may lead to a falsediagnosis of small and not easilyrecognized without expe- the disease,as it canresult from kidneydisease, rience,which mostlikely accountsfor their ab- hypoproteinemia,trauma, or parasitism(Stroud sencein mostdescriptions of the disease.Daw- et al. 1975;Weitkamp 1976).Dukes and Lawlet ley et al. (1976) reported bubblesalong the (1975) describedexophthalmia due to infection lateral lines of 50-100% of juvenile chinook with lymphocystisvirus. Exophthalmia fre- salmon and steelheads(Salmo gairdneri) ex- quently is associatedwith chronic gas bubble posedto supersaturatedwater. These bubbles disease,rather than an acute form causedby progressivelyfill the scalepockets of the lateral very high levelsof supersaturation. line system,reducing the abilityof the sensory DISSOLVED GAS SUPERSATURATION--A REVIEW 665

unitsto respondto stimuli (Schieweand Weber Signsin Eggs 1976). Scatteredbubbles (covering less than The signsof gasbubble disease vary with the 15% of the lateral line) may alsooccur on fish life stagesof the fish.In general,eggs appear not exposedto supersaturatedwater (Dawley to be resistantto the effectsof supersaturation. et al. 1976). This signshould be consideredan Ruckerand Kangas(1974) reported no signsof indicationbut not necessarilyconclusive proof the diseasein chinook salmon eggs held in of gas bubbledisease unless the lateral line is water up to 128% TGP. Meekin and Turner extensivelyinvolved. Bubbles in the lateralline (1974) likewisefound no signsof it in chinook alsoappear to be lostrapidly in deadfish. Pho- salmoneggs but reported heavymortality of tographsof bubblesin the lateral line system steelheadeggs reared in supersaturatedwater, were publishedby Dawleyand Ebel (1975)and althoughthey did not describegas bubble dis- by Weber and Schiewe(1976). easesigns in them. The signmost pertinent to recognitionof the diseaseis probablythe appearanceof bubbles Signsin Larvaeand Fry or emboli in gill blood vessels.Dawley et al. In larval fish, gas bubble diseaseappears (1976) rarely observedemboli in branchialar- quite differently than it does in juvenile and teriesand gill filamentsof live fish being held adultfish. Herring larvaereared in supersatu- in supersaturatedwater, but found them prev- ratedwater developed gas bubbles in their guts, alent in dead fish. This is an indication that which resultedin death (Dannevigand Dan- bubblesin the gills may be a causeof death. nevig 1950; Henly 1952). Similar bubblesde- Wyatt and Beiningen(1971) found that emboli velopedin fiatfishfry but thesefish were able in the gillsmay be only externallyvisible signs to passbubbles through the anusand survive. under acute conditions;gill damageoccurred Gasbubbles may form on the surfaceof newly in all fish checkedduring their acute bioassay hatchedtrout fry reared in supersaturated tests. water, causingthem to rise to the surface,or Abnormal behavior is an obvious but rather bubblesmay form in their mouthscausing suf- nonspecificsign exhibited by fishwith gasbub- focation (Shirahata 1966; Peterson 1971). ble disease (Marsh and Gorham 1905; Bouck Embody (1934), Wood (1968), Rucker and et al. 1970; Coutant 1970). Marsh and Gorham Kangas (1974), and Stroud et al. (1975) de- (1905) describeda lossof equilibriumin fish scribedsimilar signsof gasbubble disease in with the disease.Wyatt and Beiningen(1971) salmonidfry (sacfry). Bubblesformed between found that in a rapidly lethal or acutesituation the yolk sac and the perivitellinemembrane (about 150% total gaspressure or higher),fish causingfry to swimhead up. As the bubbles suddenlylost the ability to swimagainst a cur- expandedand movedposteriorly the fry swam rent, were unable to avoid obstacles,soon lost tail up and eventuallybelly up. Photographsof equilibrium,moved near the surfacewithout an this conditionwere includedby Harvey and apparent senseof direction,and then exhibited Cooper (1962) and by Shirahata(1966). Death violent writhing movementsinterspersed with occurred when the vitelline membrane rup- periodsof inactivity.Dawley and Ebel (1975) tured. Free bubblesoccurred in the digestive

Downloaded by [US Fish & Wildlife Service] at 07:25 02 February 2012 reported behavioral changes and reduced tractsof all fry, includingcontrols. Adams and growthin juvenile chinooksalmon exposed to Towle (1974) describederratic swimming be- 115% total gaspressure (TGP) in shallowwater. haviorof cohosalmon fry (Oncorhynchuskisutch) Feeding responsesbecame lethargic after 20 havinglarge gasbubbles in the yolk sac.No days exposure. Many fish developed spinal damageto gillsor other tissueswas found by flexures,exophthalmia, and large blistersin the theseauthors. Rucker and Kangas(1974) re- buccalcavity, and did not acceptfood. Meekin portedthat fry thatwere able to survivebubbles and Turner (1974) reported northern squaw- in the yolksac later developed signs of the dis- fish (Ptychocheilusoregonensis) in supersaturated easesimilar to thosedescribed above for juve- water failed to feed on juvenile salmonids.In nile salmonids.Zirges and Curtis (1975) re- general, the behaviorof fish with gas bubble ported large gas bubbles formed in the diseaseappears to follow that expected from posteriorportion of the yolk sacof chinook any fish suffering severephysical stress that in- salmonsac fry. Dead fish had frayedfins and terfereswith respiratoryand equilibriumfunc- coagulatedyolks, as previously described (Shir- tions. ahata 1966; Wood 1968). Jones and Lewis 666 WEITKAMP AND KATZ

(1976) found bubblesin the peritonealcavity of ophthalmiawas rare. Emboli were found in the channelcatfish fry (Ictaluruspunctatus) with gas gill archesof all fishkilled in bioassaysand in bubble disease. most survivors of 117% and 120% TGP in shal- low water. As with adult chinook salmon, Signsin AdultSalmonids northernsquawfish remained on the bottomof The external signsof gas bubbledisease in test containers,thereby minimizing their ex- adult salmonids are similar to those described posureto supersaturation. for juvenilesbut generallyoccur with a differ- Jensen(1974) describedexophthalmia and ent frequencyin the varioustissues. Adults fre- bubblesin the finsof warmwaterspecies having quently developexophthalmia and bubblesin gasbubble disease. Emboli occurred in thegills the roof of the mouth near the branchiostegal of someof the fish showingexternal signs of region (Westgard1964; Coutantand Genoway the disease.White bass(Morone chrysops) from 1968; Wood 1968; Beiningenand Ebel 1970). the samearea showeda high incidenceof fun- Gasbubbles of varyingsize may or may not be gusinfections that may have secondarily invad- presentin the fins and other exposedsurface ed the lesions. areas.Nebekera et al. (1976) describedgill dam- The signsand lesionsof gasbubble disease age,hemorrhaging, and bubblesin the oral cav- in adult Atlantic menhadenwere describedby ity and on the body surfaceof adult sockeye Clay et al. (1976), Marcello and Fairbanks salmon (Oncorhynchusnerka) exposed to 110- (1976), and McLeod (1978). Erratic swimming 120%TGP in shallowwater. Exophthalmiadue behavior occurred at 95% TGP (105% N2). At to the diseasemay lead to the developmentof 107% TGP (110% N2) mucussecretion, erratic cataractsaccording to Postonet al. (1973).They swimming,and colorchange, along with bub- found the incidenceof cataractswas highest in blesin the eyes,intestines, pyloric caeca, and fish held under conditionsmost likely to pro- mesenteries,were noted in twofis• thatdied duce supersaturationalthough they did not within 96 hours. All fish tested atoll0% TGP measuredissolved gas levels.Cataracts, how- (120% N•) died within 24 hours, apparently ever, may be due to other causes,as described due to bubblesin the bulbousarteriosus and gill by Hoffert et al. (1971). arterioles. These fish also showed the external Bouck et al. (1976) and Stroud et al. (1975) signsdescribed above. alsodescribed various signs of gasbubble dis- ease in adult salmonids. Bouck et al. (1976) Signsin Invertebrates found adult chinook salmon swam aimlessly, Invertebratesare alsosusceptible to gasbub- were unresponsive,and exhibiteda coughing ble disease.Marsh and Gorham (1905) found syndromeas they approached death. These fish it in American lobsters (Homarus americanus), also tended to remain at the maximum available horseshoecrabs (Limulus sp.), bivalve mollusks, depth during the bioassays,thus utilizingthe hydroids,and seaspiders (Anoplodactylus sp.). available hydrostaticpressure to compensate The American lobsters and horseshoe crabs for supersaturation.Emboli in the gillscaused surviveda long timewith their bloodin a con- discolorationof thistissue in dying fish.Stroud dition resemblingfoam. These crustaceans usu-

Downloaded by [US Fish & Wildlife Service] at 07:25 02 February 2012 and Nebeker(1976) describedin greaterdetail allylived much longer than fish under the same the externalsigns of gasbubble disease in adult conditionsof supersaturation.Legs of seaspi- chinook salmon. At 120% TGP, in shallow ders becamefilled with gasand developeda water, massive blisters occurred in the fins, palecolor. Mollusks, hydroids, and somegreen body surface,mouth, and gills,and behindthe algaedeveloped and emittedbubbles. Hughes eye,as in juvenile salmonids. (1968) alsoreported the diseasein American lobstersbut did not describeany signs. Signsin NonsalmonidFishes Malouf et al. (1972) reportedgas bubble dis- Bentley et al. (1976) describedgas bubble easein oysters( Crassostrea gigas and C. virginica) diseasesigns and lesionsin northernsquawfish: and clams(Mercenaria mercenaria) held in heat- adultscontained gas bubbles between the fin ed water.The oystersfirst developedcrescent- rays and in the subcutaneouslayer of large shapedconchiolin blisters in the shelfborder- areasof the body surface.Hemorrhages in the ing the mantle.Gas bubbles were also observed subcutaneouslayer were alsocommon, but ex- in the gill filamentsand in the outer layersof DISSOLVEDGAS SUPERSATURATION--AREVIEW 667

the mantle. In severecases, the shell cavitybe- men. Focal degenerativechanges occurred in came filled with conchiolin blisters. Clams muscle and nervous tissues, due to ischemia. showedobvious lightening of the gill coloration Pathologyof the gillswas confined to mechan- due to the presenceof bubblesthat prevented ical disruptionand displacementof tissues.Ad- free blood circulation.Henly (1952) indicated ditionaldetails of cellulardamage are described oysterlarvae suffer from the diseasebut gave by Johnson. no descriptionof it in these molluscs. Internal Lesions Evansand Walder (1969) causedgas bubble diseasein shrimp (Crangoncrangon) by means Internal lesionscharacteristic of gas bubble of decompression.Bubbles in the shrimp were diseasecan be noted by necropsyor histopath- clearly visible through the carapace.Bubbles ological examination. Marsh and Gorham commonlyappeared under the dorsal side of (1905) found diseased fish contained notable the carapacebut were also observedat other quantitiesof free gas in blood vessels.The locationssuch as at the baseof a leg. Lightner amountof gasvaried from a few smallbubbles et al. (1974) found the brown shrimp (Penaeus to largequantities that distendedthe heartand aztecus)exhibited erratic and disorientedswim- causedcomplete occlusion of vessels.The gas- ming when affectedby gasbubble disease. This filled auricle continued beating without pro- behaviorwas followed by a stuporand helpless pellingblood. The wallsof the auricleand ven- floating prior to death. Emboli occurred tricle were often emphysematous.Gas bubbles throughout the tissuesof thesebrown shrimp. in the gill filamentswere describedas the most Suppleeand Lightner (1976) reportedit in Cal- constant and significantlesion. Pathological ifornia brown shrimp (Penaeuscaliforniensis) changesassociated with the diseasehave been due to oxygen supersaturation.Onset of the discussedin a number of early papers(Renfro diseasewas marked by erratic swimming be- 1963;Harvey 1964;Pauley and Nakatani1967; havior. Shrimp then floated near the surface Boucket al. 1970;Wyatt and Beiningen1971). with the ventral side of the head higher than Gas bubbles filled the branchial arterioles caus- the tail region. Bubbleswere apparent in the ing degenerationof gill filaments,which be- gill processesand appendagesand over the en- came edematous,developed aneurysms,and tire body surface;however, few shrimp died. sometimes contained hemolyzed red blood Nebeker2et al. (1976) found that gasbubble cells. diseasein daphnia, crayfish, and stone flies Pauley and Nakatani (1967) also described caused bubbles throughout their body fluids degenerativechanges in many other tissuesof and tissues,and general body distentionprior young salmon. Diseased fish had enlarged to death, and they photographed these condi- spleenswith hemolyzedred blood cellsand a tions. Nebeker (1976) reported the daphnia reductionof white pulp. The kidneys,liver, and developedmassive air bubblesin their gut and intestines of those fish showed serious necrotic under the carapacein the brood pouch. Cray- changes.Degenerative changes in the epithe- fish were immobilizedfor many hoursprior to lium of the roof of the mouth were described death, with signsranging from nothing observ- as unique to gasbubble disease.The extensive

Downloaded by [US Fish & Wildlife Service] at 07:25 02 February 2012 ableto massiveblockage of body fluids and em- tissuechanges described by Pauleyand Naka- boli in the gills. The immobilizedcrayfish fre- tani appearinconsistent with the reportedabil- quently became turgid and swollenfrom the ity of fish to recover from this disease(Marsh internal pressureof gasand osmoticimbalance. and Gorham 1905; Rukavina and Varenika Stone flies developed bubbles at the base of 1956; Shirahata 1966; Weitkamp 1976). It is their legs and gills and in other areassuch as possiblethe changesdescribed are indicativeof the mandibles that were visible through the chronicgas bubbledisease or someother dis- body wall. easesyndrome present in thesefish. Johnson (1976) describedthe pathologyof Bouck et al. (1970) studied the histopathol- gasbubble disease in blue crabs(Callinectes sa- ogy of adult sockeyesalmon suffering from gas pidus).Dying crabshad hemal spacesof many bubble disease downstream from John Day gill filaments filled with gas. Live crabs had Dam on the Columbia River in 1968. No sig- grosslyvisible gas emboli in hemalsinuses, large nificant changeswere observedthat could be arteries,heart, and, occasionally,the midgut lu- attributed to the disease other than those as- 668 WEITKAMP AND KATZ

sociatedwith exophthalmia.Strout et al. (1975) other causes,such as predation.Gas bubble dis- reported emphysema (bubbles in tissue) be- easelesions can alsoincrease susceptibility to neaththe peritoneumcovering the kidneyand other diseases.Weitkamp (1976) found fishthat ribs,and occasionallyunder the epicardiumin were not able to recoverfrom gasbubble dis- the region of the bulbusarteriosus, in several easeapparently died due to secondaryfungal speciesof juvenile salmonids.Petechial hem- infection. orrhagesoccurred in muscles,gonads, nares, brain, and other tissues.Edema of the mesen- Recoveryfrom GasBubble D•ease terieswas often accompaniedby ascites(fluid Early in the historyof gasbubble disease in- accumulationin the abdomen).A unique le- vestigationsit was recognizedthat fish can re- sion,muscular emphysema, occurred following cover from the disease.Gorham (1901)caused prolongedexposures (400 hours)to 115% TGP the signsof the diseaseto disappearin marine in shallowwater. Exophthalmia was due to gas fish from the WoodsHole aquarium.He sub- accumulationin the fatty tissueof the orbit. jected small scup(Stenostomus chrysops) showing Bubblesin the choroid plexuscaused detach- exophthalmiaand emboliof the fins and head ment of the retinaand rupture of bloodvessels. to a pressureequal to 4.9 m of water. All signs Beyeret al. (1976a, 1976b)reported similar le- disappeared within 24 hours. Similar results sionsin decompressedjuvenile coho salmon. were found with adult fish.Henly (1952) used Stroud et al. (1975) were unable to detect the hydrostaticpressure to alleviate signsof the tissuechanges reported by Pauleyand Nakatani disease in young cod (species not given). (1967). Stroud and Nebeker (1976) described The fish were held at depthsof 2-5 m in nat- the internal lesionsof gas bubble diseasein ural seawaterto affect recoveryfrom the dis- adult chinooksalmon. At 120%TGP, gas-filled ease. spaces were common in the muscle tissue. Although a report on the pathologyof gas Blindness, which was common in adult salmo- bubble disease indicated that the fish would not nids with the disease,was due to emboli in the survive the associatedtissue damage (Pauley choroidplexus causing detachment of the ret- and Nakatani 1967) there have been numerous ina and hemorrhaginginto the eye. Photo- reportsof fishrecovering from the disease;sev- graphs of internal lesionswere included by eral recent ones have provided some insight D'Aoust and Smith (1974), Woelke et al. (1974), into why recoverydoes or doesnot occur.Tests Stroud et al. (1975), and Beyer et al. (1976b). haveused equilibrated water, hydrostaticpres- sure,and artificiallyproduced pressure to pro- Causeof Death mote recovery. Although gasbubble disease produces a va- Meekin and Turner (1974) demonstratedthe rietyof signsand lesions,the causeof deathhas recoveryof juvenile chinooksalmon following generally been attributed to anoxia resulting exposure to 120% and 135% Ns (110% and from stasis of the blood. Marsh and Gorham 122% TGP). Followingexposure times of 4-67 (1905) found gas to be present in sufficient days,the distressedfish were placedin equili- quantitiesto completelyfill and distendthe bul- brated water (100% Ns). Seven of the 67 dis- tressed fish died within the first 24 hours. The Downloaded by [US Fish & Wildlife Service] at 07:25 02 February 2012 bus of the heart, preventing movement of blood even though the heart continuedto beat. remaining fish were released in apparently Lesseramounts of gas in the vascularsystem healthy conditionafter a 2-week period. The may causeemboli only in the gills(Woodbury sick fish that did not recover included both fish 1941; Renfro 1963; and others); these can with severeexternal signsand fish with no ex- causeblood stasis in the gill arterioles,leading ternal signsof gasbubble disease. to death. Dawleyet al. (1976) found bubblesin Schiewe(1974) found that recoveryof fish the gills of dead fish but seldom in live fish; from sublethal exposure to supersaturation they concluded"these signs are directly asso- may occur rapidly. The swimming perfor- ciated with death." mance of chinook salmon exposedto 120% Stroud et al. (1975) describedsimilar roles of TGP showedno effect due to the supersatu- emboli in death causedby gas bubbledisease ration exposurewhen the fish were allowedto but also mentioned other factors. Sublethal ef- recoverin equilibratedwater for as little as 2 fects such as blindness, stress, and decreased hours. This rapid recoverycorresponds to the lateral line sensitivitycan lead to death from rapid disappearanceof external signsof gas DISSOLVED GAS SUPERSATURATION--A REVIEW 669

bubbledisease described by Dawleyand Ebel proportionsas those found in air, whereasbub- (1975) and Dawleyet al. (1976).These authors bles in blood are depletedin oxygen.On this found that after a recoveryperiod of 2 weeks, basisthey concludedthat hemoglobinhas the mostfish no longershowed any externalsigns capacityto modify the effect of oxygensuper- of disease. saturationby combiningwith dissolvedoxygen Weitkamp(1976) testedthe recoveryof ju- and removingit from the dissolvedstate. The venilechinook salmon by holdingthe sickfish actualrole of this mechanismin gasbubble dis- at increaseddepth to decreasethe level of su- easehas not been fully investigated;however, persaturation through increased hydrostatic severalreports discussed in the followingsec- pressure.Fish with obvioussigns of the disease tion indicateoxygen must reach very high levels followingexposure to 118-126% TGP for 10 to causethe diseaseif nitrogen levelsare low. or 20 dayswere held at a depthof 3-4 m for Beyer et al. (1976b) discussedthis relationship 20 days.Most of thesedistressed fish recovered in great detail. andwere released with no apparentsigns of gas Swimmingperformance following exposure bubble disease; about 10% died. Most of the to supersaturationwas studied by Schiewe fishthat died developed fungal infections of the (1974). He found the time of activeswimming caudalfin, apparentlya secondaryinfection and the totaldistance swum by juvenile chinook due to lesionsin thisregion. Lesions of the cau- salmonwas affectedby exposureto supersat- dal fin were never accompaniedby hemor- uration.At exposuresof 120%TGP in shallow rhagesas were lesionsof other locationson the water,swimming performance was significantly fish.The completelack of hemorrhagesat this lower than in control fish. At 117% and 120% site indicates the circulation to the caudal fin TGP, a significantlygreater percentageof ex- lesionsmay not be adequateto preventover- posedfish wasunresponsive during the swim- whelmingsecondary infections in somecases. ming performancetest. Schiewe suggested that The abnormalfungal infectionsdescribed by reduced swimming performance resulting Jensen(1974) in largemouthbass (Micropterus from exposureto supersaturationwould make salmoides)may have also been due to gasbubble thesefish more susceptibleto predators.His disease lesions. suggestionassumes that the predatorsare not Adamsand Towle(1974) used a recompres- likewise affected. Meekin and Turner (1974) sionchamber to alleviatesymptoms of the dis- and Dawley et al. (1975) found that northern ease.Coho salmonsac fry that showederratic squawfishsubjected to supersaturationshowed behaviorcaused by largebubbles in their yolk no, or reduced,feeding on juvenile salmonids. sacswere subjectedto a pressureof four at- Northern squawfish,in their natural habitat, mospheres..All of the fish then appearednor- may maintain an adequateaverage depth to mal. Decompressionwas not attempteddue to compensatefor the levelsof supersaturation the costsinvolved. This methoddoes not ap- commonlyencountered. pear worth seriousconsideration when similar Beyeret al. (1976a, 1976b)demonstrated the resultshave been obtainedwith only equili- rapid equilibrationof gaspressures that occurs bratedwater (100% TGP or less)or hydrostatic within fish tissues. In these tests, fish were sub-

Downloaded by [US Fish & Wildlife Service] at 07:25 02 February 2012 pressureproduced by a headof only 3-4 m. jected to internal and external supersaturation Suppleeand Lighther (1976) were able to re- and a combinationof both. Internal supersat- versegas bubble disease caused by oxygensu- uration was produced by decompressingfish persaturationin shrimp.Supersaturation of the exposedto pressuresgreater than two atmo- water was reducedby vigorouslyjetting fresh spheres. External supersaturationwas pro- seawaterinto the surfaceof raceways.Even the duced by exposure to supersaturatedwater most severely affected shrimp usually re- (150% TGP) at a pressureof one atmosphere. covered within 4-8 hours, and there were few Fish were more sensitive to external than to in- deaths. ternal supersaturation.There apparentlyare slightlydifferent mechanismsrelated to the di- Physiologicaland Behavioral Effects rectionof the supersaturationgradient and the Marsh and Gorham (1905) were the first to total volumeof gasthat accountfor thisdiffer- considerthe physiologicalaspects of gasbubble ence. Equilibrationof the critical tissuesoc- disease.They showedthat bubblesforming in curred in 60-90 minutesat any level of super- supersaturatedwater containgases in the same saturation. At levels less than 150% TGP, 670 WEITKAMP AND KATZ

mortalitycannot be directlyrelated to the equil- exposedto 110% TGP or less.The testfish had ibration time. The mechanismresponsible for a 46% incidenceof gasbubble disease, primar- formation of the emboli that block portionsof ily characterizedby lateral line bubbles. Ac- the circulatorysystem, thereby causing death, cording to Newcomb,the data supportthe the- appearsto be regulatedby somefactor or fac- ory that hypoxiarather than starvationoccurs tors other than equilibrationtime. in gas bubbledisease, although the declinein These studiesalso clearly demonstratethe cholesterolis not explainedby this theory. ability of fish to recoverfrom exposureto su- Changes in hemostatic variables of coho persaturationor to avoid supersaturationby salmon under conditions of decompression sounding.Beyer et al. (1976a) statedthat the were reported by Casillaset al. (1975). Rapid fish will not eliminate the gas in their tissues decompression,producing internal supersatu- when theysound and thusthe tissuesagain will ration, causedfibrinogen concentrationsto de- be supersaturatedon return to the surface. creaseinitially, rise within 10-15 minutes,and This may not be a problem for fish sounding then decreaseto one-half original levels after frequentlyin natural waters.As theseauthors 1 hour. Partial thromboplastintimes increased have shown, the time to formation of emboli after 10-15 minutes. Prothrombin times in- that apparentlycause death or sublethaleffects creased hlmost immediately in juvenile coho is considerablylonger than the 60-90 minutes salmon and after approximately 1 hour in required for equilibrationof the tissues.This adults. Decompressionapparently activatesa may enable a periodically sounding fish to hemostaticmechanism that resultsin consump- spenda significantportion of eachday near the tion coagulopathy. surfacewithout producing the measurableef- Casillaset al. (1976a) reported that numbers fects of gas bubble disease.This may account of thrombocytes(platelets) in the blood of ju- for the greatly reduced evidenceof the disease venile salmondecreased significantly following in fish intermittentlyexposed to supersatura- decompression.This responsewas highly de- tion by daily changesin their depth (Meekin pendenton depth and rate of decompression. and Turner 1974; Weitkamp 1976). Thrombocyte numbers returned to normal Schiewe and Weber (1976) and Weber and within 48 hours in most cases.Erythrocyte Schiewe(1976) describedthe sublethaleffect of levels increasedsignificantly 1 day following bubbles in the lateral lines of juvenile steel- decompression,which suggestshemoconcen- heads. In supersaturatedwater, bubblesmay tration. Leucocytesshowed no significantre- form along the lateral line in the scalepockets, sponse;however, conditionsof the test may where they block activityof the afferent lateral have masked any response. Casillas et al. line nerve fibers, preventing the lateral line (1976b) discusseda possiblerole of the blood from responding to stimuli. The loss of re- clottingmechanism in gasbubble disease. Their sponse,however, was completely reversed after discussionis based on increasedthrombocyte 16-20 hoursin equilibratedwater. The lossof countsthat havebeen found with physicalstress ability to respondto stimuli decreasesa fish's but decreasednumbers of thrombocyteswere capacityto detectand locatepredators or sta- not detected in the decompressionexperi- tionary objects.Although this may be an indi- ments. Downloaded by [US Fish & Wildlife Service] at 07:25 02 February 2012 rect causeof death, as the authorsstated, pred- Bubblesforming in the vascularsystem and ators may be similarlyaffected (Meekin and tissuesof the fish appear to be the mechanism Turner 1974). that causesdeath, so it is important to under- Newcomb (1976) describedchanges in the stand how they form. There is considerablein- blood chemistryof juvenile steelheadsfollow- formationavailable on this subjectbecause bub- ing sublethalexposure to supersaturation.In- ble formation in the vascularsystem has been creasesin bloodpotassium and phosphatecon- studiedextensively in regard to decompression tent over that of control fish were found in fish sicknessof humans. Marsh and Gorham (1905) that were subjectedto swimmingperformance attributed the releaseof gaswithin the vascular testsfollowing exposureto 116% TGP. Con- systemof fish to a rise in blood temperature centrations of albumin, calcium, cholesterol, ( 1-7 C) asit passesfrom the gillsto the systemic alkaline phosphatase,and total protein in the system.Marsh and Gorham also pointed out blood of test fish were decreased. No major the need for gas nuclei for bubble formation. changesoccurred in the bloodchemistry of fish They statedthat the bloodcells provide loci for DISSOLVEDGAS SUPERSATURATION--A REVIEW 671

the separation of gas from supersaturated ported in severalpapers (Englehorn 1943; blood. No evidenceis offered for this assump- Shirahata1966). The compositionof the gases tion. Rulifsonand Abel (1971) briefly described was essentiallythe same as the dissolvedgas current theories of bubble formation in relation contentof water that had been supersaturated to gasbubble disease. with air. The analysisof bubbles described The role of gas nuclei in bubble formation aboveand by Harvey (1961) has been said to with regardto decompressionsickness has been indicatethat nitrogenalone is not responsible thoroughlydiscussed in severalpapers (Harvey for gasemboli characteristic of gasbubble dis- et al. 1944; Harvey 1951; Evans and Walder ease.It shouldbe noted that the analysisof the 1969; Rulifson and Abel 1971). Bubbles will bubbles formed is not evidence that the disease form most easilyat an interface of water and is due either to total gassupersaturation, or to another substancewhere gas nuclei are nor- nitrogen alone. Harris2 et al. (1945) showed mally presentin supersaturatedconditions. Ev- that bubblesequilibrate within a few secondsto ans and Walder (1969) indicated that these nu- the compositionof gasesin the surrounding clei normallyare presentin vivo but that they medium. The methodsused to analyzeemboli can be destroyedby exposureto high pressure, have not been capableof measuringthe gas at least in shrimp. Gas nuclei decreaseeither compositionprior to the rapid equilibration the solubilityor the surfacetension of the su- that takesplace naturally. Nitrogen could cause persaturatedwater, allowingthe excessgas to the formation of emboli which would rapidly comeout of solution.Thus, gasnuclei allow the takeup oxygen,resulting in bubbleswith a gas formation of gas emboli in fish that reside in compositionsimilar to that of the surrounding supersaturatedwater. Hemmingsen (1970) was medium. able to saturate distilled water with oxygen at 140 atmospheresand with nitrogenat 170 at- Nitrogen-Oxygen-TotalGas Pressure mosphereswithout cavitation(bubble forma- A questionthat hasreceived considerable at- tion) when the pressurewas released. tention in recent yearsis the role of nitrogen Bubble formation in animalsis influencedby partialpressures versus total dissolved gas pres- a number of factors such as muscular activity sure (TGP) in causinggas bubble diseasein (Whitaker et al. 1945; Harris• et al. 1945). In aquaticorganisms. Most of the studiesreported fact, muscular activity is essentialfor bubble in the 1960'sand early 1970'sconsidered only formation. At very high levelsof gassupersat- dissolvednitrogen gas as importantin causing uration even the small muscular movements in- the disease. This emphasis apparently was volved in breathing will causebubble forma- basedon the assumptionthat nitrogen,being tion. High concentrationsof metaboliccarbon biologicallyinert, wasthe causativeagent while dioxide have also been implicated in bubble oxygensupersaturation would be regulatedor formation (Harris2 et al. 1945; Whitaker et al. reducedby biologicalprocesses. 1945), but are probably not significantlyin- It has been indicatedin severalreports that volvedin gas bubbledisease problems, as the oxygenalone can producegas bubble disease. concentrationof this gas does not normally Plehn (1924), Woodbury(1941), Alikunhi et al. (1951), Renfro (1963),and Suppleeand Light-

Downloaded by [US Fish & Wildlife Service] at 07:25 02 February 2012 reach supersaturation.It has been suggested that nitrogenmay fulfill a similarrole if present ner (1976) all reportedinstances of the disease at supersaturatedlevels (Berg et al. 1945). Me- causedby very high concentrationsof oxygen. chanicalagitation, such as movementof a bone In thesecases, oxygen generally reached 300% fractured prior to decompression,also is suf- of saturationor higher,but dissolvednitrogen ficient to cause bubble formation. was not measured. Wiebe and McGavock (1932) Bubble formation alsomay be related to the were unableto producegas bubble disease in fat content of an animal. Gersh et al. (1944) severalspecies even thoughoxygen exceeded found the number, size, and location of extra 300% of saturation.Harvey and Cooper (1962) vascularbubbles showed a positivecorrelation pointedout the distinctionbetween physical to fat content.Berg et al. (1945) found high fat (abiogenic)and biological(biogenic) oxygen su- contentper sewas not sufficientto increasesus- persaturation.Abiogenic oxygen supersatura- ceptibilityto bubbleformation. tion occurswith nitrogen supersaturationand Analysisof the gascontent of bubblesin fish the two gasestogether rapidly induce the dis- suffering from gasbubble diseasehas been re- ease.Biogenic oxygen produces deleterious ef- 672 WEITKAMP AND KATZ

fectsonly at very high levelsnear 300% of sat- effect occurredat 119% TGP when the oxygen uration or higher. and nitrogen partial-pressure ratios were There has been one report that CO2 super- changedfrom 159%:109% to 173%:105%. saturationcan causegas bubble disease (Mrsic Nebeker•et al. (1976) investigatedthe effects 1933). The data presented, however, raise of carbondioxide and oxygenon the disease- questionsas to the validityof the conclusion.As producingability of a constanttotal dissolved discussedabove, Mrsic reported a veryhigh in- gas pressure. When carbon dioxide over a cidenceof the diseasewith a very slightincrease range of 2.1-22.0 mg/liter wasused, there was in CO2 from 135 to 138 mg/litey.Other workers no effect on the survivalof fish exposedat a (Weigelt 1885; Winterstein1908; Reuss1910; constanttotal dissolvedgas pressure. Variation Wells 1913; Gutsell 1929) have discussedthe of the oxygen concentration,however, pro- toxicityof carbondioxide to fish but have not duced differences in survival. At 120% TGP, indicatedthat it may be a causeof gasbubble 50% of the test fish died in 65-70 hours when disease. dissolvedoxygen was 11.5-11.8 mg/liter, but The conclusionthat the diseaseis probably only 2% died at 17.2 mg O•/liter. due to total gas pressurerather than just dis- Meekin and Turner (1974) exposedjuvenile solvednitrogen was indicated by Shelfordand chinooksalmon to supersaturatedwell waterin Allee (1913) and Doudoroff (1957). It wasorig- containers10 cm deep and to supersaturated inallypointed out by Marshand Gorham (1905) Columbia River water in containers 61 cm that only two gases,oxygen and nitrogen,are deep. The well water contained 122% Nz and dissolvedin water in importantquantities. The 112% TGP while the river water contained verysmall amount of argondissolved in water 124% Nz and 123% TGP. In the shallower well is frequently reported togetherwith nitrogen watera mortalityof only 2-5% occurredin 5- (nitrogenand argon)as both are inert gases.In 10 days.In the deeper river water with nearly blood, nitrogen is held in simplesolution, while the samenitrogen levelsbut a higher TGP, the oxygenis primarily bound to hemoglobin.It fish suffered92-100% mortalityin 5 to 7 days. has not been determined whether hemoglobin If nitrogenalone were responsiblefor gasbub- doesor doesnot have an effect on the partial ble disease,mortality in the shallowerwell water pressureof oxygenin simplesolution in the shouldhave equalledor exceededthat in the blood under conditionsof total dissolvedgas river water. supersaturation. Dawleyand Ebel (1975) alsoprovide indirect Ruckerand Kangas(1974) exposed salmonid evidenceof the rolesof thesegases in gasbub- fry to constant nitrogen partial pressures ble disease.They divided shallowtroughs into (120%) at different total dissolvedgas pres- upstreamand downstreamsections with differ- sures.The test fish were exposedto 116% and ent groupsof testfish in each.Fish in both sec- 120% TGP with each having 120% nitrogen tions experienced nearly identical nitrogen saturation but different oxygen partial pres- concentrations while oxygen concentrations sures.At 120% TGP, the fry showeda signifi- were 5-10% (of saturation)lower in the down- cant upsurgeof mortalityat 25 dayswhile the stream sections.The TGP was 112.1% in up- sameupsurge did not occuruntil 35 daysof stream sections and 110.0% in downstream Downloaded by [US Fish & Wildlife Service] at 07:25 02 February 2012 exposureat 116%TGP. Theseresults indicate ones. Gas bubble diseasemortality of 50% oc- that the total gas pressureis more important curred in the upstreamsections within 20 days than the nitrogenpartial pressure(Nz) in pro- while no mortalityoccurred in the downstream ducinggas bubble disease. sections. This marked difference indicates the Rucker (1976) also found that variousnitro- role oxygenpartial pressuresmay play in re- genand oxygenratios normally encountered in ducingtotal gaspressure and thusreducing the supersaturatedwaters do not showa significant propensityto producethe disease. difference in their ability to produce the dis- These studiesclearly show the importanceof ease.The effect of a constanttotal gaspressure total gaspressure as opposed to nitrogenpartial is significantlyreduced only when the oxygen pressurein causinggas bubble disease. Consid- partial pressureis increasedconsiderably in re- erationof nitrogenpartial pressures alone rath- lation to the nitrogen partial pressure.Rucker er than total gas pressuremay indicatethat a reported that a drasticdecrease in the lethal water sourceis much more dangerousto the DISSOLVED GAS SUPERSATURATION--A REVIEW 673

healthof aquaticlife than it reallyis. Severalof mortalityestimates for the ColumbiaRiver sys- thesestudies also show that highoxygen partial tem were considerablyhigher than a more re- pressurescan reduce the capacityof a given cent estimate based on additional information. total gaspressure to producethe disease.This Both Ebel (1973) and Bouck(1976) pointedout effectoccurs only when the oxygenrises far the difficultiesin applyingexperimentally de- above the partial pressuresnormally experi- rived data to the natural situation. The differ- enced in supersaturatedwaters (160-175%). ences between information derived from the laboratory and natural situationsshould be se- CriticalLevel of Supersaturation riously consideredin establishingor revising The maximumor criticallevel of supersatu- dissolvedgas standards as well asin estimating ration has been defined or assumed to be the mortality due to gasbubble disease. maximumlevel of supersaturationthat can be permittedto ensuresurvival and propagation Tolerance to Supersaturation of aquaticbiota. A few early studiesindicated In order to prevent or eliminategas bubble that 110% nitrogen saturation was a critical disease, it is desirable to know what levels of levelfor youngsalmonids held in shallowwater. supersaturationcan be toleratedby fish or oth- When supersaturationwas first recognizedas er aquaticorganisms. Frequently the causesof a problem in the Columbia River system,the supersaturationare of sufficienteconomic, so- criticallevel of 110%N2 was adopted as a water cial, and politicalvalue to make their total re- qualitystandard by severalnorthwestern states movalunacceptable. Hydroelectric projects and and the National Academyof Sciences(Water steam power plants are far too valuableto be Quality Criteria 1972). The United StatesEn- eliminated.It is, therefore, necessaryto deter- vironmentalProtection Agency has sinceper- mine how much supersaturationaquatic organ- petuated this critical level as 110% TGP. More isms can tolerate under various conditions. recent studieshave shownthat this value may With this informationmore effectivemanage- be only the minimum level of supersaturation ment and engineeringefforts can be made to that can be safelytolerated by fish confinedto reduce supersaturationto acceptablelevels in shallowwater. This doesnot, however,apply to those situations where it cannot be eliminated. most natural situations. Factorsthat can affect an organism'stoler- Bentley et al. (1976), Meekin and Turner anceto supersaturationinclude life stage,size, (1974), Weitkamp (1976), and others have species,and depth distribution. The various shownthat a varietyof fish, giventhe oppor- major factors that have been studied are re- tunity to sound,can survivefor extendedpe- viewed below. riods of time in deep water that is supersatu- rated at a levelconsiderably higher than 110% Salmonidsat Near-SurfacePressure TGP withouta significantincidence of gasbub- Many of the studies conducted to date, in ble disease or death. Fish in most waters that particular most laboratory studies,have ex- are likely to be supersaturatedassume a depth posedfish to supersaturationin water depths distributionadequate to compensatefor super- of 0.5 m or less.Although theseconditions are saturationwell above110% TGP. Johnsonand seldomencountered in natural watersthey are Downloaded by [US Fish & Wildlife Service] at 07:25 02 February 2012 Dawley (1974) and Weitkamp (1976) have pertinentto fish hatcheryconditions and pro- shownthat an apparent thresholdlevel exists vide a practicalmeans for studyingmany as- near 120-125% TGP for youngsalmon that are pectsof the problem. held in waterof severalmeters depth. Relative- Meekin and Turner (1974) exposedsalmonid ly smallincreases in this range of supersatura- eggs and young salmonids to supersaturated tion produce a marked increasein the inci- water in water depths of 20 cm. At 112% TGP, dence of gas bubble diseaseand death. Below chinooksalmon eggs were not affectedwhile this level, the incidence of the diseaseis low and eyedsteelhead eggs suffered 50% mortality.No few deaths occur. This is an indication that the signsof gas bubble diseasein the eggs were present dissolvedgas standardsare far more described by Meekin and Turner. Minimal restrictivethan necessaryto protectthe fishery mortalities of juvenile chinook salmon, coho resource under natural conditions. salmon, and steelheads occurred at 103% and Ebel et al. (1975) have describedhow early 106% TGP in periodsof 30-60 days.Mortali- 674

tiesof 8-100% occurredamong several groups periods.A mortalityof 50% wasreached in 10 of juvenile chinook salmon at 114% TGP in 6 daysat total dissolvedgas saturationsbetween daysand 64-100% mortality occurred at 124% 118% and 123%. In a third test, when the su- TGP in 5-7 days.Juvenile chinook salmonin persaturationrose to 125% TGP and higher, river water of 124% TGP suffered 92-100% 50% of the test populationdied within 2 days. mortality in 5-7 dayswhen held within 0.6 m This study showed a dramatic increasein gas of the surface. bubble diseasemortality as total gas pressure Dawleyand Ebel (1975) and Ebel (1973) ex- increased. posedjuvenile chinook salmonand steelheads Blahm (1974) and Blahm et al. (1976) de- to supersaturationin 25 cm of water. At 115% scribedthe exposureof juvenile salmonidsand TGP, 7% of the chinook salmon and 50% of other fish to Columbia River water of ambient the steelheadsdied after 35 days'exposure. At dissolvedgas content in tanks 1 m deep. Total supersaturationsof 120% and 125% TGP, chi- gaspressures varied from 110% to 126%. Dur- nook salmon experienceda 50% mortality in ing 55 daysof exposure,the mortalitiesof chi- 27 and 14 hours,respectively, while 50% of the nook salmon, steelheads, and cutthroat trout steelheadsdied in 33 and 114 hours. No gas were 80%, 80%, and 42%, respectively.The bubble disease mortalities occurred at levels of chinook salmon and steelheads suffered about 105% and 110% TGP. Dawleyet al. (1976) dis- 40% mortalities within the first 10 days, with cussedthe long-termeffects of supersaturation essentiallyno deathsoccurring during the next in shallowwater. These papersdiscuss further 30 days' exposure.During the latter 30 days, results of the experiments discussedabove. the supersaturationremained near 118% TGP. Mortalities in fish held at 110% TGP increased Mortalitiesof all three speciesincreased consid- at periodsgreater than 60 days.At 60 daysthe erably during the last 5 daysof the testswhen mortality in 110% TGP was 15%, but increased supersaturationrose to about 123% TGP with to 70% after 125 days'exposure. Disease not a peakof about127% TGP. This shallow-water related to gasbubble diseasewas also involved bioassay also provides an indication of a in mortalitiesafter the first 60 days. marked increasein mortality asTGP risesabout Nebeker and Brett (1976) and Nebeker et al. 120%. (1979) exposedjuvenile sockeyesalmon, coho Bouck et al. (1976) exposedjuvenile and salmon, and steelheads to 110%, 115%, and adult chinook,coho, and sockeyesalmon, steel- 120% TGP in water 0.6 m deep. About 5% of heads, and rainbow trout to various levels of the steelheads and none of the salmon suffered supersaturation up to 130% mean TGP in mortalitiesat 110% TGP during the 26-48-day water 0.65 m deep. All fish tested tolerated tests.The steelheadssuffered a 50% mortality 110% TGP. At 115% TGP and above, the re- in 21 days and in 2 days at 115% and 120% suitswere somewhatvariable even within a giv- TGP, respectively.Sockeye salmon suffered a en age-groupof a singlespecies. Coho salmon 50% mortality in similar times while coho salm- parr suffered 50% mortalitiesfollowing expo- on had only a 10% mortalityin 26 daysat 115% suresof 77 and 44 hoursduring two separate TGP. At 120%, coho salmon reached 50% mor- tests at a mean TGP of 125%. Above 115% tality in 5« days. TGP, suchvariation may be related to peaksof Downloaded by [US Fish & Wildlife Service] at 07:25 02 February 2012 Weitkamp(1974) exposedwild chinooksalm- dissolvedgas levels reached during the tests on smolts to Columbia River water of 100- rather than the mean total dissolvedgas pres- 110% TGP during a simulatedmigration down sures. In over half of the 10- and 14-day tests the lower Snake and lower Columbia rivers. No at 115% TGP, the test fish did not suffer 50% evidenceof gasbubble disease was found in the mortality. At 120% TGP there was 50% mor- fish held within 1 m of the surfaceat total gas tality in 2 to 10 dayswhile at 125% TGP most pressuresnot exceeding110%. The mortalities test populationsreached 50% mortality within that occurred in this studywere due to second- 2 daysin the shallow-watertests. ary fungal infections, following scale loss The above studies indicate that a dramatic causedby screeningand handling. change occursin both the number of deaths In a live-cagestudy in the Columbia River, and the time to death at approximately 120- Weitkamp(1976) held juvenile chinook salmon 125% TGP in shallow water (1 m or less). At within 1 m of the surface for 10- and 20-day gaspressures below this generallevel, a low in- DISSOLVEDGAS SUPERSATURATION--A REVIEW 675

cidenceof gas bubble diseasewill be found in m and 1 m deep.During the 72-daytest period, juvenile salmonidsand deaths will occur at a the supersaturationranged from 120% to low rate. Above 120-125% TGP, mortalitydue 130%. In the 1.0-m-deep water, mortalities to gas bubble disease increasesdramatically. were 98% and 80%, for the two species,re- This apparent criticallevel hasnot been clearly spectively.Mortalities of 50% were reachedin demonstratedbut is indicated by these studies. 50 daysfor chinooksalmon and in 67 daysfor For juvenile salmonidsmaintaining a deeper coho salmonin the 2.5-m-deep water. distribution,the criticallevel would be higher. Blahm (1974) and Blahm et al. (1976) de- scribed further experiments under the same HydrostaticCompensation conditions.During 50-55-day tests,juvenile Marsh and Gorham (1905) recognizedthat chinook salmon and steelheads suffered 11% hydrostaticpressure exerted on a fish provides and 6% mortalities,respectively, in 2.5-m-deep compensationthat limited the effectsof super- tanks comparedto 80% mortalitiesfor both saturation.The total gas pressure,in percent- speciesin 1-m-deep tanks (120% to 130% ageof saturation,experienced by a fishmay be TGP). The majority of the deaths occurred quite different from the level measured and near the end of the test when supersaturation calculatedfor a fish subjectedto near-surface roseto between123% and 127%TGP. Juvenile pressure.Each meter of depth exertsadditional cutthroat trout held under the same conditions pressure that increasesthe solubility of dis- showedfar less depth compensation,with a solvedgases sufficiently to compensatefor ap- 42% mortalityin 1 m and a 27% mortality in proximately10% of saturation.In the range of 2.5 m. depths and supersaturationsnormally of con- A number of studieshave attemptedto sim- cern, the rule of 10% compensationper meter ulate more natural conditionsby placing live- of water depth is a useful approximation.This cagesin supersaturatedriver water. In live-cage meansthat a total gaspressure of 120% of sat- studiesat Priest RapidsDam on the Columbia uration at the surfaceis actuallyonly 110% at River in 1966, Ebel (1969) reported dissolved 1 m and 100% at 2 m, with no changein the nitrogen saturationsranged from 118% to volume of gas dissolvedor in the partial pres- 143%. Juvenile coho salmon were held at sures.Thus, depth is an important factor in depthsof 0.5-1.5 m, 2.5-3.0 m, 2.5-3.5 m, and determiningthe toleranceof fish to supersat- 0-6.0 m for periods of 8-12 days. Fish held uration in natural situations. below 2.5 m suffered lessthan 3% •r•,rt•!ity in Severalstudies have been conductedin deep each test. In the 0-6-m cage, 6% and 16% of tanks to evaluatethe effect of depth compen- the fish died while in the surfacecage, mortal- sation for salmonidsin supersaturatedwater. ities were 100% in the first two tests and 20% Ebel (1973) held juvenile chinooksalmon in in the third test. During the third test (August) 2.4om-deeptanks for 60 days. At 118% TGP, dissolvednitrogen concentrationswere lower, insignificantmortality occurred in the deep reachinga low of 118%. Completedissolved tankscompared to 100% mortalityin 55 hours nitrogen data are not given by Ebel. for fish held in 0.25 m of water. Dawley et al. Ebel (1971) conducted similar tests at Ice (1976) reported further resultsof the same Harbor Dam on the Snake River in 1970 where Downloaded by [US Fish & Wildlife Service] at 07:25 02 February 2012 study. At 124% and 127% TGP, juvenile chi- dissolvedgas levels ranged from about127% to nook salmon suffered 67% and 97% mortali- 132% TGP during the 7-day tests.Juvenile chi- ties, respectively,in the 2.5-m-deep water. In nook salmonheld in a 0--4.5-m-deepcage suf- water 0.25 m deep, mortalitiesin the same fered 45-68% mortality with most survivors range occurredat lower supersaturationlevels showingsigns of gas bubble disease.All fish of 115% and 120% TGP. At 110% TGP in 0.25 held within 1 m of the surfacedied during the m, 15% mortalityoccurred while only 5% mor- four tests. Fish held below 3 m suffered no tality occurred in 2.4-m-deep tanks at 120% deathsattributable to gasbubble disease. TGP. Meekin and Turner (1974) testedjuvenile Several studies have been conducted with chinook and coho salmon and steelheads in flow-through deep tanks and supersaturated live-cagesat WellsDam, suspendedat 0-0.6-m, Columbia River water. Blahm et al. (1973) held 0.9-1.5-m, and 1.5-2.1-m depthsin river water juvenile chinookand coho salmonin tanks 2.5 at 123% TGP. Nearly all fish held at 0-0.6 m 676 WEITKAMP AND KATZ

died within 3 to 7 days during four tests.At drostaticcompensation due to depth, in both 0.9-1.5 m, chinook salmon suffered 4•1•4%, the laboratory and in the field experiments,is and steelheads24%, mortality.At the 1.5-2.1-m as would be predicted by theory. The hydro- depth, chinook salmon had 4-16% mortali- static pressurecompensates for about 10% of ties while steelheadshad 20% mortality in 14 supersaturationfor each 1 m of water depth. days. Chinook salmon and steelheadjuveniles The live-cagestudies also indicatethat given and northern squawfishheld between2.4 and the opportunity,at leastunder protectedcon- 3.1 m for 14 daysin the 123% TGP river water ditions, juvenile salmonidswill remain deep suffered no mortalitiesand showedno signsof enough to compensatefor total gas pressures gasbubble disease. Coho salmonjuveniles held of approximately120-125%. It is necessaryto at RockyReach Dam on the ColumbiaRiver in determine accuratelythe natural depth distri- water of 125% TGP for 7 dayssuffered 100% bution of fish in supersaturatedwaters in order mortalityin the 0-0.6-m cage,19% mortalityat to predict their tolerance of supersaturation 0.9-1.5 m and no mortality at 1.5-2.1 m. under natural conditions.This is an important Meekin and Turner also held juveniles of factor in attemptingto estimateor predict loss- chinook and coho salmon and steelhead in vo- es of fish in various situations. The results of lition cages extending from the surface to laboratoryand field bioassayexperiments must 0.6-m, 2.l-m, and 3.1-m depths.Volition cages be interpreted in terms of all discerniblenatu- permit the fish to occupythe depth of their ral conditionsif they are to provide accurate choosingwithin the confinesof the cages.In predictionsof what will really happen. this test at 126-127% TGP, all fish in the sur- Ebel (1973) madean attemptto estimatethe face cagedied in 3 days,while only 4% of the actual gasbubble diseasemortalities of juvenile coho salmon and about 60% of the chinook salmonids in the Snake and Columbia rivers. salmon and steelheads died in the 0-2.1-m vo- His estimatewas based on only a portionof the lition cage during the 30-day test. In 0-3.1-m information now available, but Ebel's discussion cages,only 3% of the chinooksalmon died, and points out the complexitiesinvolved in evalu- none of the other two speciesdied, during 21 ating the deleteriouseffects of supersaturation days. on naturally migrating populationswhen the Weitkamp(1976) held juvenile chinook salm- available information comes from limited lab- on in supersaturatedriver water for 10 and 20 oratorytests. The applicationof experimentally daysat variousspecific depths and in volition derived information is alsodiscussed by Bouck cagesextending from the surfaceto l-m, 2-m, et al. (1976) who enumeratedmany of the fac- 3-m, and 4-m depths. Supersaturationranged tors that must be taken into consideration. between 118% and 126% TGP. None of these Many of thesefactors have not beenadequately fishdied duringthe 10-daytest. During the 20- consideredin the formulation of existingdis- day test,however, mortalities of 17%,21%, 1%, solvedgas standards. and 1%, occurred in populationsheld at 0-2 Several recent studies have been conducted m, 0-3 m, 1-2 m, and 2-3 m, respectively.Most to provide information concerningthe depth of these deaths occurred when the total gas distributionof the migratingjuvenile salmonids pressureremained near 125%.During a 20-day in the ColumbiaRiver system.The depth dis- Downloaded by [US Fish & Wildlife Service] at 07:25 02 February 2012 live-cagebioassay, when total gaspressures re- tributionis importantto a determinationof the mained near or above 125% for the first 11 hydrostatic compensationnaturally afforded days,mortality in all cagesincreased consider- the fish. Smith (1974) found 58% of juvenile ably. Chinook salmonheld within 2 m of the chinooksalmon and 36% of juvenile steelheads, surface (0-2-m cage) suffered 30% and 61% collectedwith a fixed gill net, were taken in the mortalitieswhile fish permitted accessto 1 m upper 4 m of the water column. In this study greaterdepth (0-3-m cage)had 1% and 7.5% of a reservoirforebay, Smith also reported 46% mortalities.Juvenile chinook salmon held be- of the chinook salmon and 28% of the steel- tweenthe surfaceand 4 m experiencedno mor- heads were collected above 2 m; 19% of the tality during thesethree tests.A few of the fish chinook salmon and 8% of the steelheads were in the 0•l•-m cage had a few bubblesin their collected above 1 m. fins at the end of the tests. In anotherstudy of depth distribution,Weit- These reports indicatethat the effect of hy- kamp (1974) collectedsmall numbers ofjuve- DISSOLVED GAS SUPERSATURATION--A REVIEW 677

nile chinook and coho salmon and steelheads IntermittentExposure in the Columbia and Snake rivers in drifting Intermittent exposuremay increasethe level and fixed gill netsextending to a depth of 5.5 of supersaturationfish are able to tolerate be- m. Less than 5% of the chinook salmon were causeit increasesthe time over whicha specific collected above 2 m in 1974. About 20% of the exposureaccumulates. It alsoprovides an op- coho salmon were collected between the surface portunityfor recoveryto occur,particularly if and 2 m, and about 10% of the steelheads were it is accompaniedby depth compensation.In- collectedabove 2 m. The depthdistribution in- termittent exposuremay occur through either dicatedthat a major portion of the steelheads changesin the concentrationsof dissolvedgases were below the bottom of the 5.5-m net. These or throughchanges in depth of the fish. fish were collectedprimarily in the shallower Alternating exposure to spilled or heated upstreamportions of reservoirs. waters with periods of little or no exposure to Blahm (1974) and Blahm et al. (1976) used supersaturatedwater providesan opportunity a depth sounderto determinethe depthdistri- for fish to reduce internal supersaturation.Al- bution of migratingjuvenile salmonids.An ar- though Beyer et al. (1976b) describedevidence ray of 10 transducersdescribed by Marshall that critical tissues become saturated within 60- (1976) wasplaced on the bottomof the Colum- 90 minutes,it is unlikely that the tissueswould bia River on a gently slopingbeach. Approxi- alsoequilibrate to a reduction of supersatura- mately72% of 776 fish detectedwith this ap- tion in a similar time. Desaturationnormally paratuswere between 0.9 and 2.1 m deep.Two takes much longer than saturation, as evi- beach seine catches containing 37% juvenile dencedby decompressiontables for saturation chinook salmon in this area "approximately diving by humans. No good evidenceis avail- quantified"the speciescomposition, but the able for determining the rate of equilibration depthdistribution of thechinook salmon within to reduced supersaturationfor fish. the total group observedcould not be deter- Fish experiencingintermittent exposureby mined. changingtheir depthwill experiencevery rapid The abovestudies do not provideall the in- changesin internal saturation.The pressure formationrequired to evaluatethe effectsof changewill be immediatelytransmitted to all dissolvedgas on the variousspecies. Collection tissuesthus increasingor reducinginternal su- of fishat a specificdepth does not indicate these persaturation according to the changes in individualsare at this depthfor any significant depth. Weitkamp (1976) found juvenile chi- periodof time. They may be movingup and nook salmonselected depths in 0-4-m volition down in the water column or they may be re- cagessufficient to avoid death of supersatura- mainingat fairly specificdepths for long pe- tions of approximately125% TGP. This study riods. The volitioncage experiments reported was not designedto reveal any ability of fish to by Meekin and Turner (1974) and Weitkamp detect and avoid supersaturation, but other (1976) indicatedthat all of the fish spent suf- studies have indicated fish do not do so. ficient time at depth to avoid the effectsof The effects of intermittent exposure have about 120-125% saturation. This indicates that been examined in experiments with varying

Downloaded by [US Fish & Wildlife Service] at 07:25 02 February 2012 fish under field conditionsare actuallyexperi- levels of supersaturation,Meekin and Turner encingan intermittentexposure to supersatu- (1974) alternately exposedjuvenile chinook rationthrough changes in depth.Fish collected salmon and steelheadsto supersaturatedand nearthe surfacein supersaturatedwater would equilibratedwater. Using short exposures of 4- showa highincidence of gasbubble disease and 16 hours in 17-cm-deepwater, they found ju- death if they remainedcontinuously near the veniles can tolerate 122% TGP for periods of surface. 16 hours if they are returned to saturatedwater Table 1 is a summaryof bioassayexperiments (100% TGP) for 8-hour periods.Blahm et al. on the toleranceof salmonidsto supersatura- (1976) alternatelyexposed fish for 8 hours in tion. The table includestest depths that are supersaturatedwater of 110-130% Ns (? TGP) necessaryif theseresults are to be extrapolated and 16 hours in saturated Columbia River to natural river conditions, or when these stud- water per 24 hours, as well as by the reverse ies are usedto establishor justify dissolvedgas daily schedule.The time to deathof 50% of the standards. testpopulation was closely related to the length 678 WEITKAMP AND KATZ

T^BI• 1--Summaryof dissolvedair supersaturationbioassays of salmonid fishes, compiled from literaturesources. 02: oxygen;N2: nitrogen; TGP: total gas pressure; LE50: lethalexposure to50%fish.

Depth Species Supersaturation Effectobserved (m) Reference Rainbowtrout fry Not determined "Slowlyfatal"4ay to weeks; 0-1.2 Marshand Gorham 1905 gasblisters on head or mucous membranes; "emphysemaof the skin"; death with free gasin heart; emboliin gill filaments

Rainbowtrout 200-300% O2 No deathsin 14 days Aquaria Wiebeand McGavock 12-15 cm 520-580% O2 No deaths,24-h exposure 1932 Brooktrout fry 112%TGP or "Badlyaffected with gas Hatchery Embody1934 greater bubbledisease" troughs Rainbowtrout fry, 115%N2 "Excessive"mortality; Hatchery Ruckerand Hodgeboom Cutthroattrout fry bubblesin fins,under troughs 1953 skin,in vascularsystem, in gills,and in kidneys

Sockeyesalmon alevins 108-120% TGP Gasaccumulated rapidly in Hatchery Harvey and Cooper 1962 yolk sac,20% mortality troughs alevins 106-108% TGP Somesigns of gasbubble disease,2% mortality fry 108-120%TGP Petechialhemorrhages, necrotic areas on fins, exophthalmia

Rainbow trout swimupfry (130% N2 No effect Hatchery Shirahata 1966 153-166% N2 50% mortality troughs, 12 cm 2.4-2.6 cm (120% N2 No effect 148% N2 50% mortalityin 5 days 2.6-2.9 cm (110% N2 No effect 121% N2 50% mortality

Chinook salmon adults 118% N2 Nearly 50% mortalitywithin 0.6 Coutant and Genoway 10 days 1968

Coho salmon juveniles Test times 8-12 days Ebel 1969 --140% TGP 100% mortality 0.5-1.5 5-70% mortality 2.0-3.0 3% mortality 2.5-3.5 Downloaded by [US Fish & Wildlife Service] at 07:25 02 February 2012 18% mortality 0.0-6.0 - 120% TGP 10% mortality 0.5-1.5 3% mortality 2.0-3.0 0% mortality 2.5-3.5 6% mortality 0.0-6.0

Chinook salmon 127-134% N2 7-day tests Ebel 1971 juveniles 100% mortality 0-0.75 100% mortality 0.75-1.0 34-86% mortality 1.5-2.0 2-38% mortality 3.0-4.0 45-68% mortality 0-4.5

Coho and chinook 125-130% N2 LE50 18 hours 0.2 Ebel et al. 1971 salmon, steelhead juveniles DISSOLVED GAS SUPERSATURATION--A REVIEW 679

TABLE 1--Continued.

Depth Species Supersaturation Effect observed (m) Reference

Chinooksalmon 134%N2 5-10% mortality,7.5 hours 0.6 Wyattand Beiningen juveniles 152%N2 100%mortality, 5 hours 1971 Cutthroattrout 119-136%N2 60% mortality,59 days 1.0 Blahmet al. 1973 40% mortality, 11 days 112-130% N2 40% mortality,49 days 1.0 27% mortality,49 days 2.5 Steelhead 112-129% Ne 80% mortality,55 days 1.0 6% mortality, 55 days 2.5 Chinooksalmon 112-129%N• 80% mortality,55 days 1.0 11% mortality, 55 days 2.5 Cutthroattrout 130% N2 for 16 50% mortality,72 hours Rainbowtrout hours/day; 50% mortality,16-70 hours Chinooksalmon 100% N2 for 50% mortality, 120 hours Cohosalmon 8 hours/day 50% mortalitynot reached (192 hours) Cutthroattrout 130% N• for 8 50% mortality,103.5 hours Rainbowtrout hours/day; 50% mortalitynot reached Chinooksalmon 100% N• for (192 hours) Cohosalmon 16 hours/day 50% mortalitynot reached (all fishjuveniles) (192 hours) 50% mortality not reached (192 hours) Cutthroattrout 131-139%TGP 100%mortality, 3.8 days 0.6 May 1973 125-131% TGP 100% mortality,6 days 50% mortality,2.2 days 110-127%TGP 50% mortality,14 days 113-122% TGP No mortality, 12 days 102-128% TGP No mortality, 12 days 25% signsof gasbubble disease

Mountainwhitefish 131-139%TGP 100%mortality, 1.5 days 116-127%TGP 50% mortality,12 days 113-122% TGP 40% mortality, 17 days 107-128%TGP 1 mortality,17 days 75% signsof gasbubble disease Cutthroattrout 131-139%TGP 50% mortality,17 days 3.0 55% mortality,24 days Mountainwhitefish 50% mortality,18 days 67% mortality,24 days Chinooksalmon 122%TGP 32-100% mortalityin 3-8 0.2 Meekinand Turner 1974 juveniles days Downloaded by [US Fish & Wildlife Service] at 07:25 02 February 2012 114% TGP 8-100% mortalityin 6 days 112% TGP 8-75% mortalityin 18-67 days 106% TGP 0-8% mortalityin 18-67 days Cohosalmon 112%TGP 60-100% mortalityin 6-35 juveniles days 106% TGP 0-4% mortality in 28-36 days no signsof gasbubble disease Steelhead 122%TGP 100%mortality in 3 days juveniles 112%TGP 6-30% mortalityin 6-30 days 106% TGP No effect 680 WEITKAMP AND KATZ

TABLE 1--Continued.

Depth Species Supersaturation Effect observed (m) Reference

Chinooksalmon 123-125%TGP 92-100% mortalityin 3-7 0-0.6 juveniles days 4•10% mortalityin 14 0.9-1.5 days 4-16% mortality, 14 1.5-2.1 days No effect 2.4-3.0

Chinooksalmon fry 128%TGP 83% mortality -0.2 Ruckerand Kangas1974 exposurefrom 124%TGP 75% mortality hatchingto 50 120% TGP 68% mortality daysold 116% TGP 16% inortality 112% TGP 12% mortality Chinooksalmon 125% N2 LE50 13.6 days 0.25 Dawleyand Ebel 1975 juveniles 120% N2 LE50 26.9 days 115% Ne LE50 not reached 110% Ns Same as controls Steelhead 125% N• LE50 14.2 days juveniles 120% Ns LE50 33.3 days 115% N• LE50 486 days 110% Ns Same as controls

Chinooksalmon, 28-daytest 2.4 steelhead 128% N• 50% mortality juveniles 125% N2 24% mortality 100-120% Ns No significantmortality Cutthroattrout 112-136% N• 32-50% mortality 1.0 Blahm et al. 1976 37-50% mortality 2.5 Steelhead 112-129% Ns 70% mortality 1.0 0% mortality 2.5 Chinooksalmon 112-129% N2 80% mortality 1.0 (all juvenile fish) 11% mortality 2.5 (all corrected for control mortality)

Chinook salmon adult 130% TGP 8.5-10 hours <1.0 Bouck et al. 1976 Rainbow trout parr 125% TGP 27-35 hours yearling 31 hours Sockeyesalmon parr 40 hours Coho salmon parr 12 hours adult 19-21 hours Chinook salmonparr 18 hours Downloaded by [US Fish & Wildlife Service] at 07:25 02 February 2012 Rainbow trout parr 120% TGP 51 hours adult 79-92 hours Coho salmon adult 45-51 hours Chinook salmon adult 51 hours

Chinook salmon Mortality in 60-day Dawleyet al. 1976 3-5 cm exposure 120% TGP 97% 0.25 115% TGP 80% 110% TGP 15% 105% TGP <5% 127% TGP 80% 2.5 124% TGP 65% 120% TGP <5% 115% TGP <5% 110% TGP <5% DISSOLVED GAS SUPERSATURATION--A REVIEW 681

TABLE 1--Continued.

Depth Species Supersaturation Effect observed (m) Reference

Steelhead Mortalityin 7-day 16.5-19.5 cm exposure 120% TGP 100% (2 days) 0.25 115% TGP 57% 110% TGP <5% 127% TGP 25% 2.5 120% TGP 5% 115% TGP <5%

Chinooksalmon 119-123% Mortalityin 10-day Weitkamp1976 juvenile exposure 53% 0.1 0% 0-2, 3, 4 0% 1-2, 2-3 120-128% TGP Mortality in 20-day exposure 88-100% 0-1 17-61% 0-2 3-7.5% 0-3 0 0-4 1-30% 1-2 1% 2-3 12-70% 16 hours at 0-1 8 hours at 34 4-39% 12 hours at 0-1 12 hours at 34 1-7% 8 hours at 0-1 16 hours at 34

of the exposure to the supersaturatedwater. of gas bubbledisease occurred in the 10-day Less than 50% of chinook and coho salmon, test with saturations between 118% and 123% steelheads, rainbow trout, mountain whitefish, TGP. At 120-126% TGP, mortalities reached and largemouthbass were killed by an 8-hour 1%, 4%, and 12% in 20 daysof 8-, 12-, and 16- exposure per day to supersaturated water hour surfaceexposures, respectively. During (130% N2). Most of these fish, however, suf- the second20-day test, supersaturationrose to

Downloaded by [US Fish & Wildlife Service] at 07:25 02 February 2012 fered 50% mortality in lessthan 24 hours dur- near or above125% TGP for an 11-dayperiod. ing continuous exposure to supersaturation During this time, 50% mortalitywas reached in (130% N2). 5 days with the 16-hour surfaceexposure. In Controlled changesof depth also have been lessthan 48 hoursduring the sameperiod, 50% used to study the effectsof intermittent expo- mortality occurred for fish given continuous sure. Weitkamp (1976) intermittentlyexposed exposureat 0-1 m. At this higher level of su- juvenile chinook salmon to Columbia River persaturation (125% TGP), mortalities in the water by changingthe depth of 1-m-deeplive- 12- and 8-hour exposurecages were 39% and cages on 8-16-, 12-12-, and 16-8-hour sched- 7%, respectively.Dissolved gas levels below ules. The live-cageswere alternatedbetween 125% during the last9 daysapparently enabled depthsof 1-2 m and 3-4 m for one 10-daytest the survivingfish to loseall signsof gasbubble and between0-1 m and 3-4 m for two 20-day disease. testsin an attempt to represent possiblediel These studies indicate that intermittent ex- changesin migratingfish. No deathsor signs posureeither by changesin the supersaturation 682 WEITKAMP AND KATZ

of the ambientwater or by changesin hydro- fish were testedwith water sourceshaving dif- staticpressure will allow fish to tolerate super- ferent temperatures.Differences in the chem- saturationfor a longerperiod. This increasein ical compositionof the water alsomay havein- the exposuretime that is tolerated is greater fluenced the results of these tests, as the two than the sum of the intermittent exposures, levelsof supersaturationwere drawn from dif- which indicatessome recovery occursduring ferent supplies. short periodsof reduced supersaturation. Blahmet al. (1976) reportedtests of juvenile chinook salmon and steelheads in a divided Detection and Avoidance trough having 130% N2 (? TGP) on one side The ability of fish to compensatefor super- and 102% N2 on the other side. Steelheadsdid saturationmay be increasedif the fish are able not avoidthe supersaturation,for they reached to detect and avoid supersaturation.Fish can 50% mortality in about 43 hours. Chinook avoid supersaturationby either refusingto en- salmonapparently avoided the supersaturation ter supersaturatedwater when a choiceexists for theysuffered no deathsduring either of the or by soundingto compensatefor supersatu- 8-day tests.The resultsagree with Meekin and ration at surfacepressures. Turner's (1974) report of avoidanceof super- It has been generally acceptedthat fish are saturationby juvenile chinooksalmon. not able to detect supersaturationand avoid it. Dawley et al. (1976) reported the apparent Severalrecent reports indicatethat this theory detectionand avoidanceof supersaturationin may not,be valid for all conditions.This ques- deep (2.4 m) tanks.The verticaldistribution of tion is of considerable importance as it can juvenile chinooksalmon and steelheadswas ap- greatlyaffect the extrapolationof experimental parently altered after 3 days'exposure to var- data to the conditionsfaced by fish in natural iouslevels of supersaturation.For the first days waters. Fish in natural waters frequently have of exposurethe depth distributionsof various the opportunity to seek hydrostaticcompensa- testgroups were not significantlydifferent. Af- tion or to avoid enteringsupersaturated waters ter 3 days,the meandepth of the groupsin the if they are able to detect supersaturation. supersaturatedwater wasgreater than that of Ebel (1971) found that juvenile chinook fish in saturated water, and mean depths in- salmonheld in 0•L5-m volition cagessuffered creasedwith increasinglevels of supersatura- much highermortality from gasbubble disease tion. than fish forced to remain in deeper water (3- Bentleyet al. (1976) describedthe apparent 4 m). This suggeststhat thesefish were unable avoidance of supersaturation by northern to detect, or were unwilling to avoid, supersat- squawfishbelow Little GooseDam on the Snake uration. River. Catchesof northern squawfishbelow the The ability of juvenile chinook and coho dam were much lower during the period of salmon to detect and to avoid supersaturation higher supersaturationthan they were before when permitted an alternativein shallowwater and after this period. The fish apparently was studiedby Meekin and Turner (1974). A either avoided the area of supersaturationor divided trough with supersaturatedwater at assumeda deeper vertical distributionduring

Downloaded by [US Fish & Wildlife Service] at 07:25 02 February 2012 110-117% TGP on one side and equilibrated the high supersaturationperiod. Large num- water at 101% TGP on the other side was used bers of northern squawfishwere captured in a to testthe fishes'response. The fish were intro- side arm of the reservoir below Little Goose duced to the trough in the lower mixing zone Dam during this time. These fish may have of 110-113% TGP. Juvenile chinook salmon been avoiding higher dissolvedgas concentra- showeda strongpreference for the equilibrated tions in the main river, although gas pressures water and avoided the supersaturatedwater in the side arm were not measured. when the water supplywas switched from one Stickney (1968) reported Atlantic herring side of the trough to the other. Coho salmon (Clupeaharengus harengus) showed a definite showed no preference for either equilibrated tendencyto avoidsupersaturation. This avoid- or supersaturatedwater. These resultsare not ance occurred only when the supersaturation definitive due to temperaturedifferences be- washigh enoughto producegas bubble disease: tween the two water sourcesused during part 120% N2 and 130% O2 (122% TGP). of the tests.The report doesnot indicatewhich These studiesindicate somefish may be able DISSOLVED GAS SUPERSATURATION--A REVIEW 683

tO detectand avoidsupersaturation and others creasingtolerance with increasingage in juve- maybe unableto detector do not avoidsuper- nile chinook salmon. Mortalities of 100-mm saturation.Reported mortalities caused by su- chinooksalmon were three to four timesgreat- persaturationin thermaldischarges at electric er than thoseof fish under 40 mm. Larger in- generatingstations indicate that somespecies dividuals of coho salmon and steelhead also are either not ableto detectsupersaturation or showeda reducedtolerance to supersaturation. that their attraction to heated water overcomes Rucker(1975) comparedsmall (38 mm and 46 their aversionto supersaturation.It is obvious mm) cohosalmon to larger (100 mm)juveniles that insufficient information is available to draw in 0.2-m-deepwater at 112% TGP. The time to any useful conclusionson this issue. 50% mortalitywas 2.6 and 4.2 daysfor two groupsof the larger fish, 2.7 daysfor the 46- Life Stage mm fishand morethan 30 daysfor the 38-mm The toleranceof fishspecies to dissolvedgas fish. Dawleyet al. (1976) testedseveral sizes of supersaturationis not the sameat all life stages. juvenilechinook salmon in 0.25-m-deepwater As discussedabove, eggs show no signsof gas at 112%TGP. They found fish40 mm longto bubbledisease when held in supersaturatedbe significantlymore tolerant of supersatura- water. They appear to be tolerant of levelsof tion than fish 53 mm and 67 mm long. The supersaturation that affect fish. Marsh and smallerfish suffered less than 10% mortalityin Gorham(1905) and othermore recent reports 45 dayswhile the largerfish reached over 50% indicateeggs do not developthe disease.Meek- mortalityin lessthan 15 days.These studies in andTurner (1974)provide the singleknown indicatea generaldecrease in toleranceof ju- report of gasbubble disease in eggs.Steelhead venilesalmonids with increasingage and size. eggsdeveloped high mortalitiesbut chinook On the other hand, some workers have in- salmoneggs did not when both were hatched dicated that older fish are more tolerant of su- in water having112% TGP. It is possiblethat persaturationthan young fish. Harvey and somethingother than supersaturationwas re- Cooper(1962) reportedsockeye salmon alevins sponsiblefor the steelheadegg mortalities. No (sacfry) appearto be particularlysusceptible to discussionof the egg pathologyis provided. supersaturation. Wood (1968) described the The majorityof evidenceindicates fish eggs are levelsof nitrogen supersaturationdetrimental extremelytolerant of supersaturation. to variouslife stagesof salmonidsas 103-104% In general, the tolerance of different life for fry, 105-112% for youngjuveniles, and stagesappears to follow two consecutivetrends. 118% for adults. No indicationis given by In veryearly life stages,the toleranceto super- Wood as to how this information was derived. saturationdecreases from verygreat tolerance Bouck et al. (1976) reported the resultsof a in the eggto verylow tolerancein olderjuve- numberof bioassaysin 1-m-deepwater using niles.Life stagesfollowing the juvenile stage severallife stagesof salmonidsand otherfishes. appearto increasein toleranceto supersatu- Althoughthere wasconsiderable variation be- ration,with adultsbeing generally the mosttol- tweenindividual tests, they found younger fish erant free-swimminglife stage. were generallyless tolerant of supersaturation

Downloaded by [US Fish & Wildlife Service] at 07:25 02 February 2012 Marsh and Gorham,however, reported At- than older fish. At 115% TGP, the mean times lanticcod (Gadusmorhua) fry to be tolerantof to 20% mortalityfor salmonidswere 125 hours levelsof supersaturationthat produced gas bub- for juveniles, 154 hours for smolts,and 309 bledisease signs in adultfish; however, they only hours for adults. held the fry in this supersaturatedwater for 2 Individualfish at any particularlife stage days.Egusa (1959) found killirish (Oryzias latipes) have shown considerable differences in their fry to beresistant to supersaturationimmediate- toleranceto supersaturation,as indicatedby ly after hatching,apparently due to "elasticitymost of the bioassaysdiscussed above. Becker and tenacityof tissuesof the bodywall." Shira- (1973) discusseda more direct method of mea- hata(1966) reported rainbow trout fry became suringindividual tolerances than differencesin increasinglyless tolerant of supersaturation the time to death. The formation of emboli in with increasingage for the first 2 monthsafter circulatingblood was recorded by a telemeter- hatching. ing flowmetersurgically implanted on the co- Meekinand Turner (1974)also reported de- nus arteriosus of rainbow trout. The results in- 684 WEITKAMP AND KATZ

dicated individual fish of this speciesvaried ing a lower dissolvedgas content (lessthan widely in their toleranceto bubble formation in 110% N2) alsodied, but at a slowerrate. Using water of the samelevel of supersaturation. adult sockeyesailnon, Bouck et al. (1970) found temperature increasesin saturatedwater fol- Heritability lowing exposureto supersaturationcaused an We havefound only one studythat addressed increasedrate of blindness;deaths at higher the heritability of toleranceto gas bubble dis- temperatures (20 and 22.5 C) were related to ease.Cramer and Mcintyre (1975) studiedthe pathogenicbacteria. toleranceto supersaturationof severalstocks of The relationship between temperature and chinook sailnon from Oregon coastalstreams supersaturation for juvenile chinook sailnon and from the lower and upper ColumbiaRiver. was studied by Coutant (1970), but the results Stockswith the longesthistory of exposureto were inconclusive.Ebel et al. (1971) reported supersaturationwere the most tolerant of su- that a stress of 115-120% N2 for 12 hours did persaturation.A comparisonof 80 tank families not greatly affect the temperature tolerance of producedfrom twenty malesand four females juvenile chinooksailnon. When testedin heated indicated the tolerance is inherited. water supersaturated at 125-130% Na, the Cramer and Mcintyre (1975) madeestimates prior stresssignificantly decreased the temper- of increasesin smwivalthat could be expected ature tolerance of these fish. Fish tested at el- with eachgeneration experiencing a givenmor- evatedtemperatures in saturated water showed tality due to gasbubble disease. For mortalities no effect that could be attributed to the prior of 30%, 50%, and 70% in the sailnonpopula- stressof supersaturation. tion, increases in survival of 0.4%, 0.7%, and The National Academyof Sciences/National 1.0% should occur with each succeedinggen- Academyof Engineering(1972) concludedthat eration. These increasesare not great but do supersaturationhas no real effect on thermal indicate an advantageto using tolerant stocks tolerance. Becker (1973), however, reported in waters that may be supersaturated. This that higher lossesof juvenile sahnonids oc- study suggeststhat experimentson fish from curred among thosethat had been subjectedto rarely supersaturatedwaters may indicatemore the stressof supersaturationthan amongthose seriousproblems than are actuallyencountered that had not. He concludedthat the exposure by populations in frequently supersaturated of sahnonidsto the Hanford thermal plume in waters. the Columbia River was too brief to cause mot- talkies. Temperature Fickeisen et al. (1976) studied the tolerance As water temperature affects many activities of blackbullheads (Ictalurus melas) to supersat- of fish, it is important to determine its relation- uration at temperaturesof 8, 12, 16, and 20 C. ship to gas bubble disease.The consideration Temperature effects were very slight within of temperatureeffects is particularlyimportant this range, and not of ecologicalsignificance. in the Columbia and Snake rivers where aver- The TL50's were 126.7% TGP at 8 C and age water temperatureshave been raised by 124.4% TGP at 20 C.

Downloaded by [US Fish & Wildlife Service] at 07:25 02 February 2012 many hydroelectricprojects (Beiningen and Boucket al. (1976) found adult sockeyesahn- Ebel 1970). The effect of temperature on tol- on were considerablymore tolerant to 120% erance to supersaturationis also important in and 125% supersaturationwhen they were ac- dischargesof heated water that may have be- climatedslowly from 10 C up to 18 C than when come supersaturated within thermal power they were permitted little or no acclimation. plants. Young fish showed a variable response;in- A number of studies have been conducted to creased temperatures increased tolerance in determine if a synergisticor additiveeffect ex- one test but decreased it in two others. The ists betweenhigh temperaturesand supersat- effect of temperatureon toleranceto supersat- uration. Coutant and Genoway (1968) found uration appearsto be so slight that it is often that chinook sailnon acclimated in or tested in overshadowedby other factors.In the testsre- supersaturatedwater (greater than 118% N2 ported by Bouck et al. (1976), the acclimation saturation)could not survivea temperature of temperature and period of acclimation ap- 22 C. Fish acclimated and tested in water hav- peared to significantlyaffect the test results. DISSOLVED GAS SUPERSATURATION--A REVIEW 685

Temperature appearsto have little direct in- the abilityof thejuvenile salmonids to undergo fluence on the toleranceof fish to supersatu- the physiologicalchanges required to makethe ration but it doesproduce an indirect effect by transfer from fresh water to salt water. changingthe solubilitiesof dissolvedgases. As Dawleyet al. (1976)described an experiment water temperaturesincrease the solubilityof with chinook salmon and steelheads that had dissolvedgases decreases, thus resulting in survivedsupersaturation bioassays. Surviving greater levelsof supersaturationeven though fish exposedto 110%, 115%, and 120% TGP the dissolvedgas concentrationremains con- for 127 dayswere transferredto 25%0seawater stant. This has long been recognized in re- and held for 13 days. Most of the 50 chinook searchon gasbubble disease. Engelhorn (1943) salmondied in thistest; only eight of the larger useda risein watertemperature to bring about fishsurvived the full testperiod. The majority supersaturationand produce symptomsof gas of the fish may not havereached smolting size bubble disease.Supersaturation as a result of and thusdid not havethe abilityto adaptto salt warminghatchery water wasreported by Ruck- water. Survival was higher among the steel- er and Hodgeboom (1953). Westgard (1964) heads tested; most deaths were of smaller fish. noted this mechanismwas also responsiblefor The authorsconcluded that prior exposureto a minor portionof the supersaturationproblem supersaturationseemed not to affectthe ability at the McNary spawningchannel. DeMont and of steelheadsto adapt to salt water and that Miller (1972), Adair and Hains (1974), Miller their data on chinook salmon were inconclusive (1974), and Marcello and Fairbanks(1976) all with regard to saltwateradaptation. discussedaspects of supersaturationassociated Boucket al. (1976) alsotested the ability of with thermal effluents. varioussalmonids to adapt to salt water follow- Malouf et al. (1972) describedsupersatura- ing exposureto supersaturation.Following ex- tion produced by heating cold seawater in posureto 110%, 115%, and 120% TGP, steel- closed heat exchangers.The supersaturation head and sockeyeand chinooksalmon juveniles was sufficientto causegas bubble diseasein were transferredto gas-equilibratedseawater. three speciesof bivalvemolluscs. Lightner et al. In all tests the transferred fish either survived (1974) describedthe diseasein shrimpresulting for over 5 days,at which time the experiment from supersaturationproduced by heatingsea- was ended, or died from causes unrelated to water in a closed system. Zirges and Curtis supersaturation.Bouck et al. concludedthat no (1975) encountered similar problems after latent or delayed mortalities occur due to gas heating a water supply for chinooksalmon sac bubble disease after salmonid smolts enter sea- fry. The various aspectsof these reports are water. discussedin greater detail in other sectionsof this review. Nonsalmonids Isaacson(1977) questionedthe role of gas Although much of the recent researchand bubble disease in heated effluents. This author publicityon gasbubble disease has concentrat- pointed out that pop eye, a signof this disease, ed on salmonids because of the Columbia River can be causedby cold stresswhich could be con- problem,there hasbeen considerableresearch fused with gas bubble diseasein heated ef- Downloaded by [US Fish & Wildlife Service] at 07:25 02 February 2012 on other species.Gorham (1901) originallyde- fluents. Although this may be true in some scribedthe diseasefrom scup,a marine species. cases,it is unlikely that cold stresswould cause Scup were killed by gas bubbledisease when emboliand other recordedsigns that appearto held in shallowaquaria containingwater with be specificto the disease. 135-145% TGP (basedon the solubilitiesused by Gorham). Marsh and Gorham (1905) re- SaltwaterAdaptation ported the diseasein a varietyof other species, In recent yearsthere has been considerable presumablyat similar dissolvedgas levels,but concernabout the ability of juvenile salmonids few quantitativedetails are provided. to adapt to salt water followingexposure to su- Woodbury (1941) observedgas bubble dis- persaturationduring their downstreammigra- easein a varietyof freshwaterfish as the result tion. As gas bubble diseaseproduces various of oxygensupersaturation caused by photosyn- tissuechanges, described above, it has been theticactivity. Black crappies (Pomoxis nigromac- theorizedthat it might alsosignificantly reduce ulatus), bluegills( Lepomismacrochirus ), northern 686 WEITKAMP AND KATZ

pike (Esoxlucius), and carp (Cyprinuscarpio) all Carolina. Signs occurred primarily in white died at very high oxygenlevels (30-32 ppm). bass,redbreast sunfish(Lepomis auritus), blue- Dannevig and Dannevig (1950) and Henly gills, and threadfin shad (Dorosomapetenense). (1952) discussedgas bubble diseasein artifi- Signswere alsoreported in a few individualsof cially hatchedAtlantic cod, herring, and plaice ten other species.These fish experiencedsu- (Pleuronectesmicrocephalus), but gave no details persaturationas high as 130% TGP that was about dissolvedgas concentrations.Stickney produced by the heating of lake water at the (1968) producedthe diseasein herring held in Marshal SteamGenerating Station. aquaria at 122% TGP. Fickeisenet al. (1976) studied the tolerance Egusa(1959) comparedthe toleranceof five of blackbullheads to supersaturationat several speciesto dissolvedgas supersaturation by de- temperatures.The dissolvedgas level required termining the "detrimental nitrogen limit" to produce a 50% mortality of the test popu- (signsof gasbubble diseasein 50% of test fish lation during 96-hour tests(TL50) was about within 2 weeks)and the "lethalnitrogen limit" 125% with slightdifferences depending on the (50% mortalitywithin 2 weeks).The detrimen- water temperature. tal limits for the fish were 125% for adult gold- Supersaturationbioassays were conductedby fish and eel (Anguillajaponica);130% for young Blahm et al. (1976) on a varietyof species,in- goldfish,young carp, adult killirish,and adult cluding largemouthbass and mountain white- bitterling (Rhodeusacellatus). The lethal limits fish, which were alsotested in comparablein- were 120% for adult carp; 125% for adult termittent-exposuretests. The largemouthbass goldfish;130% for eel, young carp, bitterling, were more tolerant of supersaturationthan the and young goldfish; over 140% for killirish. salmonid specieswhile the tolerance of moun- Supersaturationvalues given by Egusaare per- tain whitefishwas about equal to that of coho cent nitrogen saturation,and would be about salmon and steelheads. This conclusion was 5% of saturationlower if reported as TGP. based on the time to mortality of 50% of the Renfro (1963) reported gas bubble disease test population(LE50) at 130% N2 for 24-, 16-, mortality for a number of marine fishesin Gal- and 8-hour/dayexposures. Bioassays of smelt, veston Bay causedby oxygen saturationsap- crappies, and northern squawfish were also parentlyover 250%: spottedseatrout (Cynoscion conducted by Blahm et al. in river water at am- nebulosus);gulf menhaden(Brevoortia patronus); bient supersaturation ranging from about bay anchovies(Anchoa mitchilli); juvenile Atlan- 113% to 123% TGP. Although the testsmea- tic croakers (Micropogonundulatus); speckled sured different end points, they did indicate worm eels (Myrophispunctatus); longnose gar that the tolerance of smelt was similar to that (Lepisosteus osseus ). of steelheads but less than that of most salmo- Marcello and Fairbanks (1976) described a nids. The crappies and northern squawfish mortalityof Atlantic menhadendue to gasbub- were more tolerant than the salmonids and suf- ble disease.Supersaturation resulted from tem- fered no deaths in 20 days and 35 days, re- perature increasesin the cooling water dis- spectively. charged from the Boston Edison Company's Bouck et al. (1976) reported lessthan 10% mortality of largemouth bass exposedfor 20

Downloaded by [US Fish & Wildlife Service] at 07:25 02 February 2012 Pilgrim Nuclear Power Station. Dissolvedoxy- gen levelsmeasured during the mortalitywere daysto about 125% TGP in 0.65 m of water. frequently between 130% and 140% of satu- The largemouthbass were able to capture and ration. Total dissolvedgas levels were likely in eat juvenile salmonduring this test. Apparent- the samerange or higher. Severalother species ly, supersaturationin the ranges normally ex- of fish and invertebrates were observed in the periencedwould have little effect on predation supersaturatedarea with no evidenceof gas by this species.Bouck et al. (1976) reported shiners(Notropis sp.) and crappies(Pomoxis sp.) bubbledisease. Clay et al. (1976) found Atlantic have a tolerancecomparable to that of salmo- menhadenheld in shallowtanks showed signs nids. Bluegills, northern squawfish,and war- of the disease at 107% TGP within 96 hours. mouth (Lepomisgulosus) were more tolerant to DeMont and Miller (1972) and Miller (1974) supersaturationthan shinersand crappiesbut described the occurrence of the disease in a were lesstolerant than largemouthbass, bull- number of speciesfrom Lake Norman, North heads(Ictalurus sp.), and carp. DISSOLVEDGAS SUPERSATURATION--AREVIEW 687

A number of studies have examined the tol- is not possiblein this area it is probablethat erance of the northern squawfish,a major recruitment of at least some adult fish occurred predator of juvenile salmonidsduring their from Lake Roosevelt, which is behind Grand downstreammigration. Meekin and Turner Coulee Dam. Lake Roosevelt had also been su- (1974) found adult northern squawfishto be persaturated(140% TGP) in the previousyear more tolerant to 111% TGP than steelhead and (Seattle Marine Laboratories1972b). chinooksalmon juveniles, but the adult squaw- The studies discussed above indicate that the fish were of equal or lower toleranceto 122% tolerance of different speciesto supersatura- TGP when testedin shallowwater (20 cm).Ju- tion can vary considerably.Salmonids are venilechinook salmon and steelheadsplaced in amongthe leasttolerant fish but others,such a trough with northern squawfishwere vulner- as Atlanticmenhaden, may be even lesstoler- ableto predationat 100% TGP but were more ant. In natural situations the tolerance of a activethan the predators,and were ignoredby specieswill be affectedby behavioralpatterns them, at 111% TGP. This indicates that suble- suchas depth distributionor attractionto heat- thal supersaturationmay producegreater ef- ed waters. fectson a prindpal predatorof salmonidsthan on the prey. Causesof Supersaturation Bioassaysof northernsquawfish in constantly Water may becomesupersaturated with at- supersaturatedwater were conductedin shal- mosphericgases through any one of several low tanks (0.25 m) by Bentleyet al. (1976). different processes,either causedby humans Twelve-day mortalities were 100% at 126% or nature. These processeseither causean in- TGP; 60% at 120% TGP; 32% at 117% TGP; creasein the amount of air dissolvedor they and 0% at 110, 107, and 100% TGP. At 126% reduce the amount of air water will hold. TGP, all fish died in lessthan 1 day. All survi- Lindroth (1957) discussedfour ways by vors of the 120% and 117% TGP tests, and which water may becomesupersaturated: (1) mostof the fishexposed to 110%TGP, showed water containsdissolved gas coming from a gas signsof gasbubble disease at the end of the 12 mixture containinga higher percentageof that days. gasthan is normallyfound in air; (2) watercon- Bentleyet al. (1976) alsofound that northern tains gas that was dissolvedunder a higher- squawfishcollected from the SnakeRiver dur- than-atmosphericpressure; (3) water contains ing periodsof moderateto high supersatura- gasdissolved at a lower-than-ambienttemper- tion showedless evidence of feeding than the ature; (4) two bodiesof saturatedwater at dif- fishcollected during timeswhen little or no su- ferent temperaturesare mixed. persaturationwas present in the area.Feeding The first mechanismis probablyof little irq- wasdecreased by about 50% at 115% TGP. In potranceas it is likelyto be encounteredonly the bioassays,northern squawfish tended to re- in experimentalsituations. The secondmech- main on the bottom of the test tanks during anismhas been involved in manyof the docu- exposure to supersaturation.This behavior mentedsupersaturation problems, indicated in maybe responsiblefor the reducedfeeding ob- the followingdiscussions of air injectionand servedin nature, and may contributeto better Downloaded by [US Fish & Wildlife Service] at 07:25 02 February 2012 hydroelectricprojects. The secondmechanism survivalthrough depth compensation. is alsoinvolved in the supersaturationof nat- Parametrix (1974) sampledresident species ural springs.The third mechanismhas caused downstream from Grand Coulee Dam in Lake supersaturationin the heatingof watersupplies Rufus Woods,where levelsof supersaturation for fish culture, the cooling waters of power have been as high as 145% TGP in previous generatingfacilities, and in geothermalheating years.During the time of thisparticular survey, of natural waters. The fourth mechanismmay supersaturationreached 110% TGP for a brief causesupersaturation but isunlikely to produce period. The presenceof a varietyof adult fish levelshigh enoughto causegas bubble disease of severalage-classes in Lake RufusWoods was under most circumstances. An additional taken asan indicationthat the populationswere mechanismnot mentionedby Lindroth (1957) survivingthe previousyears' high levelsof su- is photosynthesis.Photosynthetic activity has persaturation.Although recruitment of fish been responsiblefor severalreported casesof from downstream reservoirs or from tributaries the disease. 688 WEITKAMPAND KATZ

Air Injection mallyencountered in plungebasins, the hydro- Any situationthat allowsair to be mixed with static pressureis sufficientto greatly increase water under pressuremuch greater than one the solubilitiesof atmosphericgases. The air atmospherecan producesupersaturation if ad- thus passesinto solutionin sufficientamounts equatevolumes of air are available.The initial to producesupersaturation with respectto sur- descriptionof gasbubble disease and its cause faceor atmosphericpressure. These sources of by Marsh and Gorham (1905) resulted from supersaturationfrequently have the capacityto this mechanism. A leak on the suction side of supersaturatelarge volumesof water and thus the saltwatersupply system for the WoodsHole causea major problem. aquariumsystem permitted air to be drawn in Jarnefelt (1948) recognizedsupersaturation with the water. associatedwith a hydroelectricproject and The supersaturationproblem at Flodevigen measuredoxygen supersaturation as high as marinefish hatchery was reported by Dannevig 127% at a Swedishproject. Total dissolvedgas and Dannevig(1950). In this instance,air was pressureswere likely at leastas high. Lindroth suckedinto the systemat the shaft bearingsof (1957) alsomeasured supersaturation below a the pump. Harvey and Smith(1961) described dam on the Indalsalven River, Sweden. a Canadianhatchery intake systemthat per- High levelsof supersaturationin the Colum- mitted air to be sucked into the intake under bia and Snakerivers were first reported by Ebel various conditionscausing supersaturation. (1969). During the 1966 spill period(July and Wyatt and Beiningen(1969, 1971) found a sim- August),nitrogen levels above 120% saturation ilar situationat an Oregon hatcheryintake. In were measured in the Columbia River. Dis- both casespartial occlusionof the intake per- solvedgas levels remained near saturation dur- mitted sufficientair to be drawn into the sys- ing the remainderof 1966. In 1967, Columbia temsto producesupersaturation and gasbub- River saturation levels were comparableto ble diseasein exposed fish. Hughes (1968) thoseof 1966. Diurnal variationsin nitrogen encountereda supersaturationproblem result- saturations were measured at The Dalles Dam ing from air leaksin a lobsterhatchery water during 1966;dissolved nitrogen concentrations supply. were found to vary only 0.6 mg/literduring a Suchair leakson the low pressureside of a 24-hourperiod of constantspill. water supplysystem can easilycause supersat- The effect of John Day Dam on the dissolved urationin a fishculture facility. This shouldbe nitrogen levelsin the ColumbiaRiver during one of the first sourcesinvestigated when su- 1968 was describedby Beiningen and Ebel persaturationproblems are encounteredin a (1970). The dam wasclosed in April 1968,be- pumpedwater source. fore the generatorswere operational,so that Johnson(1976) reporteda somewhatdiffer- the reservoircould be filled and the fishways ent source of supersaturation.Seawater was put into operationto accommodatethe spring pumpedinto an unusedline that wasfilled with salmonmigrations. This requiredtotal spillof air. The pressurein the line causedthe air to all water passingJohn Day Dam and produced supersaturatethe water sufficientlyto produce dissolvedgas levels from 120% to 145% of sat- gasbubble disease in the blue crabsand several uration. Levels in excess of 125% of saturation Downloaded by [US Fish & Wildlife Service] at 07:25 02 February 2012 fish. This appearsto be an unusualcause of were recordedbelow John Day Dam through supersaturationbut one that couldeasily be en- September1968. Water temperaturesalso in- counteredin any pipedwater supply. creaseddue to the warming of water in the Fast et al. (1975) and Fast (1979) reported newly createdJohn Day Reservoir. dissolvednitrogen supersaturationin New Dissolvedgas levelsand associatedvariables York and California lakes. In both cases,arti- for numerous locations on the Columbia and ficial aeration causeddissolved nitrogen con- Snake rivers are presentedby Beiningenand centrations to reach 140-150% of saturation. Ebel (1970) for the years 1965-1969. Ebel (1971) presentedadditional data for the year HydroelectricProjects 1970. Dissolvedgas levels were generallylower Spillwaysof hydroelectricprojects cause air in 1970than in previousyears for the Columbia and water to be mixed and carried to substan- River but were high in the Snake River due to tial depthsin a plungebasin. At the depthsnor- the spill at Little GooseDam. DISSOLVED GAS SUPERSATURATION--A REVIEW 689

Roesner and Norton (1971) developed a ing May and June. The reports also described mathematical model to evaluate air entrain- changesin dissolvedgas concentrations as the mentin waterpassing over spillways. The mod- river water traveled 110 km downstream from el wasdeveloped from monitoringdata collect- Bonneville Dam. The United States Army ed at three lower Columbia River dams (John Corpsof Engineers(1975, 1977)provided com- Day, The Dalles,and Bonneville). plete reports for dissolved-gasmonitoring data Meekinand Allen (1974a)reported dissolved for the Columbia and lower Snake rivers in gaslevels associated with controlled-flowstud- 1974, 1975, and 1976. Boyer (1974) presented iesat ChiefJoseph Dam on the ColumbiaRiver. a detailed analysisof the damsand the various Widely varyingdissolved gas levels were mea- physicalfactors involved in the supersaturation sured in the upstream reservoir, as well as problemin the ColumbiaRiver system,a valu- downstream, between individual controlled- able resourceto anyonedealing with supersat- flow tests.The reported variationswithin short uration resultingfrom dams. time periods(2 hours or less)have not been Seattle Marine Laboratories (1972a) pre- found in other dissolved-gasmonitoring stud- sentedresults of dissolved-gasmonitoring stud- ies.Although the report indicatesthe variations ies on the middle and upper reachesof the in supersaturationare probablydue to differ- SnakeRiver in Idaho. This major tributary to ent blocks of water, this conclusion does not the Columbia River contains numerous dams. appear to be supportedby the data presented Dissolvedgas levels in the upper reachesof the in the report. river remained below 110% TGP. Water below Meekin and Allen (1974b) presentedthe re- a seriesof three dams in Hells Canyon along suitsof dissolvedgas monitoring in the mid- the middle reach of the river exceeded 120% Columbia River from 1965 and 1971. Super- TGP at times. Downstream from the last dam saturationoccurred throughout this region of in Hells Canyon,dissolved gas levels gradually the Columbia River during high-flow periods. decreased to below 110% TGP at the conflu- Grand CouleeDam producedhigh levelsof su- ence of the Snake and Salmon rivers. Parame- persaturation,and the dams below Grand Cou- trix (1974) again measuredhigh dissolvedni- lee in this region increasedthe levelsof super- trogen concentrationsbelow the Hells Canyon saturationonly slightly over those in water dams; however, total dissolvedgas levels did arriving at eachdam. PriestRapids and Rocky not exceed110% due to low oxygenconcentra- Reach dams actually reduced supersaturation tions. when the water arriving in their forebayswas Dissolvedgas concentrations in the Canadian highly supersaturated. portion of the Columbia River, its tributaries, Meekin and Allen (1974b) alsomonitored the and severalother rivers were reported by Clark river upstreamof Grand CouleeDam in 1965, and Regan (1973), Clark (1974, 1976), and 1970, and 1971. During theseyears, this water Abelson (1975). These rivers frequently ex- wassupersaturated to somedegree. It exceeded ceeded 110% TGP and the Columbia River ex- 120% TGP in 1970 and approachedthis level ceeded 120% TGP at times. in 1971. Seattle Marine Laboratories (1972b, 1974) monitoreddissolved gas levels above and Thermal Increases Downloaded by [US Fish & Wildlife Service] at 07:25 02 February 2012 below Grand Coulee Dam in 1972 and 1973. In recentyears, electrical generating facilities Lake Roosevelt, the reservoir behind Grand have been constructedthat significantlyraise Coulee, containedwater supersaturatedto over the temperature of large volumes of water. 120% TGP. This supersaturationwas present These temperatureincreases frequently have in Columbia River water entering the United been sufficient to causesupersaturation of the States from Canada. Downstream from Grand water. Supersaturationoccurs because the sol- Coulee Dam, the Columbia River water exceed- ubility of a dissolvedgas decreases as the tem- ed 140% TGP in 1972. perature riseswhile the actual volume of gas Blahm (1974) and Blahm et al. (1975) de- dissolved remains the same. scribeddissolved gas monitoring below Bonne- Harvey (1967) reported that water in a Ca- ville Dam on the lower Columbia River in 1974. nadian lake becamenaturally supersaturated Dissolvedgas levels averaged near or above during the late spring and summer of 1961. 120% TGP in this lower reach of the river dur- Solar radiation increasedthe temperature of 690 WEITKAMP AND KATZ

the lake sufficientlyto producesupersaturation the water supply was apparentlynear satura- of 110-120% TGP at the level of the lake's ther- tion prior to being heated. mocline. Erdman (1961) reported the diseasein At- DeMont and Miller (1972), Adair and Hains lantic salmonfry in streamwater heatedmore (1974), Jensen (1974), and Miller (1974) dis- than 2.8 C. The cold stream water was near cussedvarious aspectsof supersaturationof saturationprior to the temperature rise. No coolingwaters from three steamgenerating sta- dissolvedgas levelswere reported for the gas tionsin North Carolina.Supersaturation in the bubble diseaseincident. Stickney (1968) re- heated effluent of the Marshal Steam Station ported that Gulf of Maine waterswere super- exceeded120% TGP and apparentlyreached saturated due to photosynthesisand heating. 130% TGP at times. Two associated steam sta- Nitrogen levelsas high as 128% were appar- tions, Allen and Riverbend,also caused super- ently due to temperatureincreases. Total gas saturationof coolingwaters but a low incidence pressuresas high as 120% apparentlycaused of gasbubble disease. The MarshalSteam Sta- no problemin natural watersbut did causethe tion causedgreater temperatureincreases and diseasein fishwhen the waterwas pumped into producedsupersaturation at high levelsover a the laboratory. longer period of time than the other two sta- Lightnet et al. (1974) encounteredsupersat- tions. The level of supersaturationwas related urationin a heatedseawater supply that caused to the depth of dischargeoutlets and volumes gasbubble disease in brown shrimp.This sea- of flow. water was heated from 22 C to 28.8 C in a Marcello and Fairbanks (1976) described a closed system.The 7 C increasewould have similar problem at the Pilgrim Nuclear Power produced a supersaturationof about 112% Stationon Cape Cod. Althoughonly dissolved TGP if the water were saturatedprior to being oxygenwas measured, these measurements do heated. indicate that total dissolvedgas levels exceed Zirges and Curtis (1975) raised the temper- 140%at timesand werefrequently above 120% ature 4.5 C for the water supplyto a chinook TGP. Experiencehas shown that oxygenlevels salmon hatchingfacility. Supersaturationwas normally are equal to or lower than total dis- high enoughto causegas bubble disease in sac solved gas levels unless supersaturation is fry. Dissolvedgas levelswere reported to be caused primarily by photosyntheticactivity. 106% for oxygenand 190% for nitrogen. The The high levelsof supersaturationat the Pil- 190% nitrogen figure appearsto have been a grim Station were causedby temperature in- misprintas the nitrogenlevel would only have creasesof 13-25 C over ambienttemperatures. increasedto 110% due to the temperaturerise Only slightdecreases occurred in the volumes if the water had been saturatedprior to heat- of dissolvedgases as indicatedby oxygencon- ing. centrationspresented in the report. The episodesof supersaturationreported Supersaturationhas been caused in othersit- above indicate that supersaturationshould be uations where water has been heated for cul- consideredany time aquaticorganisms may be ture purposes.Shelford and Allee (1913)heat- exposedto heatedwater. This includescooling ed water in a closed system, which caused waters of large industrial facilitiesor waters

Downloaded by [US Fish & Wildlife Service] at 07:25 02 February 2012 supersaturation.The temperatureincreases of heatedfor aquaticculture purposes. 8-17 C, without loss of dissolvedgases, pro- duced supersaturationsufficient to causegas Natural Causes bubble disease.These temperature increases Supersaturationis not a new or necessarilya would have producedsupersaturations of 115- human-causedphenomenon. Spring and well 130% TGP if the water had been saturated watersare frequently supersaturatedwith dis- prior to heatingand no gaswas allowed to es- solvednitrogen although oxygen levels are fre- cape. quently low. Marsh and Gorham (1905) dis- Embody (1934) heated water in a closedsys- cussedsuch situations in the well watersupplies tem to hatchtrout eggs.Temperature increases at a Tennessee and a New Hampshire fish over 5 C causedgas bubble disease in hatching hatchery.The aspiratingeffect of waterflowing fry. A temperatureincrease in this range prob- downwardinto aquiferscarries air with it. The ably causedsupersaturation of 112% TGP as air is thus mixed with water under considerable DISSOLVED GAS SUPERSATURATION--A REVIEW 691

pressure in the aquifers. Oxygen concentra- persaturationcaused by photosynthesis.It is tions may be reduced prior to use of these likely temperaturealso played a role in these waters; however, the inert dissolvednitrogen casesas the intensesunlight necessary for high gas remains in solution, causingsupersatura- levelsof oxygenproduction through photosyn- tion problems. thesiswould alsocause significant increases of Marsh (1910) describeda supersaturatedwell the water temperature. supply for a fish culture stationalong the Po- One caseof natural supersaturationcaused tomac River where dissolvednitrogen levelsex- by geothermal heating has been reported ceeded 140% of saturation. Rucker and Tuttle (Bouck1976). Total gaspressures of 107-110% (1948) describeda WashingtonState hatchery were measuredin streamsheated by geother- water supply having 110% TGP (120% nitro- mal actionin Oregon.Although 105% TGP was gen, 80% oxygen). Matsue et al. (1953) dis- sufficientto causehatchery trout fry to develop cusseda variety of artesianwells and springsin gas bubble disease,fish in the natural stream Japan that producedsupersaturated water. Dis- with 107-110% TGP showedno signsof the solved nitrogen saturationsas high as 150- disease. This is a minor but obvious demon- 160% were reported. stration of the differences between natural and Flowingsurface waters may alsobe supersat- artificial situationsand the difficulty that can be urated by natural causes.Jarnefelt (1948) de- encounteredin assumingthe two are equally scribed how rapids can cause air entrainment susceptibleto any perturbation. and supersaturationjust as dam spillwaysdo. Ebeling (1954) and H611(1955) both recorded Solutionsto Supersaturation oxygen supersaturationof streams,apparently Supersaturationof water with dissolvedgas due to falls and rapids. Mortimer (1956) de- results in an unstable condition that tends to scribed how bubbling and violent agitation, as return to a state of equilibrium.The rate at in a turbulent streamflow, can producesuper- which this naturally occursis usuallytoo slow saturation as well as reduce it. Lindroth (1957) to prevent the problemsdiscussed in this re- observedthat fallswith deep plungebasins pro- view. It is important to recognizethat water is duce supersaturationwhile those with shallow supersaturatedonly with respectto atmospher- basinsat their baseproduce little or no super- ic or surface conditions.Hydrostatic pressure saturation.Harvey and Cooper(1962) observed may greatly reduce or eliminate the tendency similar situations in a British Columbia coastal of the large volumeof deeper water to equili- stream. Supersaturationoccurred at high river brate with respectto surfacepressures. Deeper flowswith high fallshaving deep plungebasins. waters in reservoirs, rivers, oceans, et cetera, Parametrix(1974) reportedthe SalmonRiver may actuallyhave no tendencyto losedissolved containeddissolved nitrogen levelsas high as gasesas they are not truly supersaturated.The 121%; however,total dissolvedgas levelssel- hydrostaticpressure that frequentlycauses the dom exceeded 110%. Even these levels of su- air to becomedissolved at high concentrations persaturation are high when it is considered will also maintain the high concentrationsin that river flowswere extremelylow in 1973, the subsurface water. year of monitoring. Supersaturationwas the A variety of solutionsto supersaturationin Downloaded by [US Fish & Wildlife Service] at 07:25 02 February 2012 result of both temperatureincreases and slight artificialwater supplies have been used over the increasesin dissolvednitrogen concentrations years.Marsh (1910) flowedwater throughshal- as the water moved downstream. The increases low troughswith roughbottoms and over stacks in the amount of dissolvednitrogen were the of six perforated pansto reduce supersatura- primary factor producingsupersaturation dur- tion. Both of these systemsworked for small ing the higher flow periods. water flows.Embody (1934) wasable to remove As discussedabove under the topicof oxygen supersaturationby passingthe water over a se- as a causeof gasbubble disease,photosynthesis ries of bafflesplaced at the head of a trough. can lead to high levelsof dissolvedoxygen in Rucker and Tuttle (1948) succeededin reduc- natural waters.Woodbury (1941), Alikunhi et ing dissolvednitrogen from 140% to near sat- al. (1951), Schmassmann(1951), Rukavina and uration by cascadingwater through a seriesof Varenika (1956), Renfro (1963), and Supplee six troughsor shelvesabout 0.25 m apart. Har- and Lightner (1976) all recorded casesof su- vey and Cooper(1962) useda splashtower with 692 WEITKAMP AND KATZ

12 setsof bafflesto reduce nitrogen from near 1972b, 1974).During summermonths, the sur- 120% to near saturation. face water of this reservoirtends to equilibrate Dennison and Marchyshyn (1973) described in the downstreamportion of the reservoir. a smallinexpensive box designedto equilibrate The deeper watersapparently moving through water. The water flows in a thin layer over a the reservoirshow little lossof dissolvedgas. perforated plate through which air is pumped The various mechanical solutions to the to strip excessgases from the water. This device problem of supersaturationwere discussedby reduced oxygen saturations from 118% to Smith (1972). Perforated bulkheads to be 100%.Wold (1973) describedthe useof a large placedin skeletonturbine bayshave been built aerator facility for removal of excessdissolved and tested.These reducedsupersaturation but gasesat the DworshakNational Fish Hatchery were detrimental to fish passing.through the in Idaho. These large agitatorswere ableto re- orifices.Long and Ossiander(1974), Long et al. duce dissolvednitrogen levels from 130% to (1975), and United StatesArmy Corps of En- less than 105%. gineers(1979) describedtests to evaluatethe Removal of excessdissolved gases in large effect of perforated bulkheads on juvenile bodies of water is considerablymore difficult salmon. Perforated bulkheads have also been due to the stabilityof the supersaturatedgases proposedas a devicefor reducingthe effective under many conditions.Harvey (1961) found hydraulic head for operational turbines. Re- that solar heating and mixing causednitrogen ducingthe effectivehead would permit the pas- supersaturationin lake water to a depth of 6- sageof large volumesof water with minimum 15 m. This supersaturated water mass re- power generation.This proceduremay be ef- mained fairly stablein volume and gascontent fective during the spring high-flow period from June through August. Rapidly flowing when power requirementsare about one-half turbulent streams do not necessarilyprovide of the maximumcapacity of the major damson rapid equilibrationof dissolvedgases. As Mor- the Columbia River. timer (1956) pointed out, rapidly flowing Flip lips or spillwaydeflectors offer the po- streams may even produce supersaturation, tential of reducing the level of supersaturation and oncesupersaturated, water reaches equilib- produced.by water passingover a spillway. rium with the atmosphereslowly. These devices are structural modifications to Ebel (1969) reported that no equilibration the downstreamface of a spillwaythat direct occursin the five reservoirsbetween Chief Jo- the spilled flow along the surfaceof the tail- seph and McNary dams (400 km) on the Co- water rather than allowingit to plungeto the lumbia River during the high-flowperiod. Fail- bottom of the stillingbasin. Tervooren (1972, ure of the supersaturatedwater to equilibrate 1973) reported the first resultsof testsper- wasblamed on lack of circulationand warming formed on spillway deflectors installed at of surface water. When water is warmed, the Bonneville Dam. These indicated that, under decreasein the water's capacity to hold dis- varying water conditions,deflectors would re- solvedgas may compensatefor any lossin the duce the normal increase of excess dissolved actual dissolvedgas content and thus a high gasin the spillwaytailwater by 50%. These re- level of supersaturationwill be maintained. suitswere obtainedwhen the forebaygas levels Downloaded by [US Fish & Wildlife Service] at 07:25 02 February 2012 Beiningen and Ebel (1970) also discussed were between 100% and 120% saturation. This equilibrationof supersaturatedColumbia River meansthat the installationof spillwaydeflectors water. They found that supersaturatedwater for all baysat BonnevilleDam would result in equilibratedto a greater degreein John Day downstreamgas levels of 110% of saturationif Reservoirthan in other previouslystudied res- forebay water did not exceed 100% of satura- ervoirsin 1968. This equilibrationwas thought tion. to be due to the reservoir'sgreat length and Spillwaydeflectors are effectivein reducing long retention time as compared to that of supersaturation without causing increased many of the other reservoirs. mortalitiesof juvenile salmonids.Johnson and Lake Rooseveltprovided only slight equili- Dawley(1974) reportedthat the Bonnevillede- bration of the supersaturatedwater that enters flectorsreduced supersaturationby 6-12% with from Canada and travels 225 km to Grand no decreasein the survivalof fish passingover Coulee Dam (Seattle Marine Laboratories them; Monan and Liscom(1975) actuallyfound DISSOLVED GAS SUPERSATURATION--A REVIEW 693

improvedadult survival. Long et al. (1975)also any reasonto becomefamiliar with the tech- reported an increasedsurvival of steelhead niquesfor dissolvednitrogen determinationor smoltspassing over deflectorsat Lower Mon- of total dissolvedgas analysis. This sectionof umental Dam on the lower Snake River. the review describes a few of the more recent The recent statusof spillwaydeflectors has reportsthat discussuseful information dealing been reviewedby United StatesArmy Corpsof with suchanalyses. Engineers(1979). Ebel et al. (1975) estimated The solubilityof the individualatmospheric that the installation of recommended deflectors gasesin water hasbeen described in numerous and turbinesby 1980will eliminatethe problem recent reports. Elmore and Hayes (1960), of supersaturationin the Columbiaand Snake Klotts and Benson (1963), Green and Carritt rivers for all practicalpurposes. Ebel (1979) (1967), Douglas (1964, 1965), Murray et al. concludedthe supersaturationproblem in the (1969), and Tolk et al. (1969) all discussedthe Columbiasystem has been solved. solubilitiesof oxygen,nitrogen, and argon.The A third proposedsolution to the supersatu- solubilitiesmost commonlyused in recent su- ration in the ColumbiaRiver systemis the col- persaturationresearch are those provided by lectionand transportationof juvenile salmonids Weiss(1970). from upstreamdams to the lowerriver (Ebelet There are severaldifferent techniquesavail- al. 1973;Ebel et al. 1975).Although the super- able for dissolved gas analysis. Swinnerton saturationproblem is apparentlybeing solved (1962) describedthe use of gas chromatogra- by other means,this methodis still being uti- phy for thispurpose. Beiningen (1973) provid- lized to reduce the turbine mortalities that are ed a completediscussion of the useof the time- considerablygreater than supersaturationmor- consumingVan Slykeapparatus (Oesting 1934) talities. for determiningthe oxygenand nitrogen-plus- Supersaturationproblems caused by cooling argon concentrationsof water samples.Bein- watersof power generatingfacilities have been ingen discussedin easy-to-followdetail the op- lessof a problem than hydroelectricprojects eration of this apparatusas well as the calcu- and therefore have received far less attention. lationsrequired for the separatedissolved gas Kraback and Marcello (1976) discussedthe fea- analyses.He included sectionson the correct sibilityof removingexcess dissolved gases from proceduresfor field sampling,procurement of the Pilgrim Nuclear Power Station cooling the necessaryapparatus, and the Winkler water. Heated seawaterwould be degassedby methodof oxygendetermination. This manual means of an air bubbler system.Tests in a is extremelyuseful for anyoneusing the Van flume installedin the dischargecanal indicated Slyke method to determine dissolvedgas con- that the systemwould work. centrationsin water samples. Lee and Martin (1975) computedthe capac- Post(1970) describeda simplifiedvolumetric ities of high- and low-velocitydischarges of method for determiningdissolved gas concen- coolingto produce supersaturationproblems. trations.This method apparentlyhas received They concludedthat high-velocitydischarges little use and may provide significanterrors of are unlikelyto causesupersaturation problems 5-7%.

Downloaded by [US Fish & Wildlife Service] at 07:25 02 February 2012 becauseof rapid dilution. Localizedsupersat- By far the simplestmethod for determination uration problemsmay occur with low-velocity of total dissolvedgas levels is that providedby discharges. the Weiss saturometer. As far as can be deter- mined, this apparatushas not been described Dissolved Gas Analysis in anyjournal or other widelydistributed pub- The measurementof dissolvedgas levelsis lication.Fickeisen et al. (1975) gavea brief de- the one aspectof supersaturationthat is least scription and picture of a Weisssaturometer. familiar to mostpeople who mustdeal with the A similar deviceis producedcommercially by problem of gasbubble disease. Analysis of the ECO Enterprises,Seattle, Washington, and is levelsof supersaturationis, however,essential now widelyused to measuretotal dissolvedgas for the evaluationand solutionof the super- pressure. saturationproblem. Although most people like- D'Aoust and Smith (1974) and D'Aoustet al. ly to encounter supersaturationare familiar (1976) described a modification (tensionome- with dissolvedoxygen analysis,few have had ter) of the Weiss saturometer. The tensiono- 694 WEITKAMPAND KATZ

meter providesfor sensingof the gaspressure and analysisof dissolvedoxygen samplesat by meansof a solid stateelectronic pressure depth usingan underwaterhabitat (Tektite II). transducerrather than the Bourdentube gauge In marine waters,they found a decrease'of usedin the saturometer.According to D'Aoust about 2-6% in the dissolvedoxygen concentra- et al., the tensionometerhas the advantagesof tion of samplesfixed at the surfaceas compared muchsmaller size, and a responsetime of about to thosefixed at depth.This lossof oxygenwas 8 minutes.This compareswith a responsetime limited to sampleswhose oxygen concentra- of 20-30 minutes for the saturometer. The tions were greater than the surfacesaturation more rapid responsetime of the tensionometer value. The lossof dissolvedoxygen should not is due to the muchsmaller dead space permit- be a problem with supersaturationmonitoring ted with the pressuretransducer. The tension- as the few deep samplescollected are normally ometer also offers the advantage of remote fixed immediatelyafter collection.There is no sensingwhich is requiredwhen dissolvedgas indicationof how dissolvednitrogen concentra- measurementsare made at depth. Agitationis tions vary prior to analysisby Van Slyke or gas still required to remove bubblesfrom the silas- chromatograph. The amount of dissolvedgas tic tubingwhen the wateris truly supersaturat- these samplescan hold is normally increased ed, such as at surface pressuresof about one during storageby reducingthe temperatureof atmosphere.This is not a problem at water the samplesto below4 C. Thus, supersatura- depthsover 3 m where the water is not actually tion in the samplesis reduced or eliminated, supersaturateddue to the increasedpressure preventingloss of dissolvednitrogen. of the hydrostatichead. D'Aoust et al. (1976) provided a completeparts list and instructions Regulation of Supersaturation for building the tensionometer. Since the identificationof dissolvedgas su- The operationof the Weisssaturometer was persaturation as a problem in the Columbia evaluated by Fickeisenet al. (1975). The satu- River systemin the late 1960's,there have been rometer was mechanicallyagitated at rates of criteriaand standardspromulgated by a variety 108, 132, and 168 cyclesper minute. At the two of regulatory entities.The National Academy faster rates, equilibrium was reached in 15-25 of Sciences/NationalAcademy of Engineering minutes. It requires 10-15 minutes longer to (1972), usingavailable data, recommendedthat reach equilibrium at the slower rate of agita- aquatic life will be protected when total dis- tion. Manual operationproduced comparable solvedgas pressure in water is no greaterthan resultswith an experiencedoperator but lower 110%. Subsequentlythe statesof Washington, readingswith a noviceoperator. Idaho, and Oregon promulgated dissolvedgas Jenkins(1976) attemptedto developa meth- standards,initially for dissolvednitrogen and od for unattended monitoring "in situ" of dis- later for total dissolvedgas. The regulations solvedgas concentrations.An incompletesys- specifiedhuman activitiesshould not increase tem was designed to pump the water to be dissolvedgas levels above 110% in Washington sampledand to strip the dissolvedgases from and Idaho and 105% in Oregon. Other states this water by a modified spinningdisc oxygen- have sincepassed similar regulations.

Downloaded by [US Fish & Wildlife Service] at 07:25 02 February 2012 erator. Major problemsremain to be solvedin The water quality standardswere reviewed the detection of the gases once they are in 1975 by a group of four agencyrepresen- stripped from the samplewater. tatives from the United States Environmental The various correctionsthat have been ap- ProtectionAgency, Idaho, Oregon, and Wash- plied to dissolvednitrogen partial pressuredata ington (Rulifsonand Pine 1976). This group were discussedby Boyer (1974). He concluded suggesteda standard of 115% total gas satu- that the sum total of these corrections is less ration for the Columbia-SnakeRiver systemex- than the inherent errors of the samplingand cept during particularlyhigh-flow years. Their analysistechniques. Boyer also discussedthe recommendationwas ignored by the Environ- "true" value of the supersaturationlevel result- mental Protection Agency criterion issued in ing from a dam spillwayand describedwhat is 1976 (USEPA 1976), which again recommend- neededto developan empiricalformula that is ed a criterion of 110% TGP. Recently,Ebel et unique for each spillway. al. (1979) have reviewed the most recent Envi- Cratin et al. (1971) studied the in situ fixation ronmental ProtectionAgency criterion, and in- DISSOLVEDGAS SUPERSATURATION--A REVIEW 695

dicatedefensible dissolved gas criteria could be BEYER,D., B.G. D'AousT, AND L. SMITH. 1976b. established at either 110, 115, or 120%. Decotnpression-inducedbubble fortnation in sal- moniris.Comparison to gasbubble disease. Un- dersea Biomedical Research 3:321-338. References BLAHM,t. H. 1974. Report to Corpsof Engineers, ABLESON,D. M. G. 1975. Report on the resultsof a gassupersaturation research. Prescott Facility-- monitoringprogramme for dissolvedgases in the 1974. National Marine Fisheries Service, North- Fraser River and PeaceRiver systems,1974. Brit- westFisheries Center, Seattle, Washington, USA. ish Columbia Water Resources Service, Pollution BLAHM,T. H., R. J. MCCONNELL,AND G. R. SNYDER. Control Board, Victoria,Canada. (Unpublished.) 1973. Effect of gas supersaturatedColumbia ADAIR,W- D., ANDJ. J. HAINS. 1974. Saturationval- River water on the survivalof juvenile salmonids. uesof dissolvedgases associated with the occur- Final Report--Part 1. NationalMarine Fisheries rence of gas-bubbledisease in fish in a heated Service, Environmental Field Station, Prescott, effluent.Pages 59-78 in J. w. Gibbonsand R. R. Oregon, USA. Sharitz,editors. Thermal ecology.United States BLAHM,T. H., R. J. McCONNELL,AND G. R. SNYDER. AtomicEnergy Commission Contribution 030505, 1975. Effect of gas supersaturatedColumbia Washington,District of Columbia,USA. River water on the survivalof.juvenile chinook ADAMS, E. S., AND F. G. TOWLE. 1974. Use of a re- and coho salmon. NOAA (National Oceano- compressionchamber to alleviategas-bubble dis- graphicand AtmosphericAdministration) Tech- ease in coho sac-fry.The ProgressiveFish-Cul- nical Report NMFS (National Marine Fisheries turist 36:41. Service)SSRF (SpecialScientific Report Fisher- ALIKUNHI, K. H., V. RAMAGHANDRAN,AND H. CHAUD- ies) 688. HARI. 1951. Mortality of carp fry under super- BLAHM,T. H., R. j. MCCONNELL,AND G. R. SNYDER. saturationof dissolvedoxygen in water.Proceed- 1976. Gas supersaturationresearch, National ings of the National Institute of Science,India Marine FisheriesService Prescott Facility. Pages 17:261-264. 11-19 in Fickeisenand Schneider(1976). BECKER,C. D. 1973. Columbia River thermal effects BOUCK,G. R. 1976. Supersaturationand fisheryob- study:reactor effluent problems.Journal of the servationsin selected alpine Oregon streams. Water Pollution Control Federation 45:850-869. Pages37-40 in Fickeisenand Schneider(1976). BEININGEN,K. t. 1973. A manual for measuring BOUCK,G. R., G. A. CHAPMAN,P. W. SCHNEIDER,JR., dissolvedoxygen and nitrogen gas concentra- ANDD.G. STEVENS.1970. Observationson gas tions in water with the Van Slyke-Neillappara- bubble diseasein adult Columbia River sockeye tus. Fish Commissionof Oregon, Portland, Or- saltnon(Oncorhynchus nerka). Pacific Northwest egon, USA. Laboratory,Federal Water Quality Administra- BEININGEN,K. T., ANDW. J. EBEL.1970. Effect of tion, Corvallis,Oregon, USA. John Day Dam on dissolvednitrogen concentra- BOUCK,G. R., A. V. NEBEKER,AND D.G. STEVENS. tions and saltnon in the Columbia River. Trans- 1976. Mortality,saltwater adaptation and repro- actions of the American Fisheries Society duction of fish exposedto gas supersaturated 99:664-671. water. EPA-600/3-76-054, United States Envi- BEININCEN,K. T., ANDW. J. EaEL. 1971. Dissolved ronmentalProtection Agency, Corvallis, Oregon, nitrogen, dissolvedoxygen, and related water USA. temperaturesin the Columbiaand lower Snake BOYER,P. B. 1974. Lower Columbia and lower rivers, 1965-69. Data Report 56, National Ma- Snakerivers, nitrogen (gas) supersaturation and rine FisheriesService, Seattle, Washington, USA. related data analysisand interpretation,March, BENTLEY,W. W., E. M. DAWLEY,AND T. W. NEW- 1974. United StatesArmy Corps of Engineers, COMa.1976. Someeffects of excessdissolved gas North PacificDivision, Portland, Oregon, USA. Downloaded by [US Fish & Wildlife Service] at 07:25 02 February 2012 on squawfish,Ptychocheilus oregonends (Richard- CASILLAS,E., S. E. MILLER, L. S. SMITH, AND B.G. son). Pages41-46 in Fickeisenand Schneider D'AovST. 1975. Changesin hemostaticparam- (1976). etersin fish followingrapid decompression.Un- BERG, W. E., M. HARRIS, D. M. WHITAKER, AND V. C. dersea Biomedical Research 2:267-276. TW•TTY. 1945. Additional mechanisms for the CASILLAS,E., L. S. SMITH,AND B.G. D'AovsT. 1976a. origin of bubblesin animalsdecorepressed to The responseof fish blood cells,particularly simulatedaltitudes. Journal of General Physiol- thrombocytes, to decompression. Undersea ogy 28:253-258. Biomedical Research 3:273-281. BERT,P. 1873. Influencedes hautes pressions sur les CASILLAS,E., L. S. SMITH,AND B.G. D'AousT. 1976b. poissons.Comptes Rendus des Seances de la So- Effects of stress on sahnonid blood clotting ciete de Biologiede Paris5:160-161. mechanisms.Pages 93-95 in Fickeisen and BEYER,D., B.G. D'AouST, AND L. SMITH. 1976a. Schneider (1976). Responsesof cohosaltnon (Oncorhynchus kisutch) CLARK,M. J. R. 1974. A report on resultsof a mon- to supersaturationat one atmosphere.Pages 47- itoring programmefor dissolvedgases in select- 50 in Fickeisenand Schneider (1976). ed waters of British Columbia. British Columbia •9• WEITKAMP AND KATZ

Water Resources Service, Pollution Control DAWLEY,E. M., M. SCHIEWE,AND B. MONK. 1976. Branch, Victoria, Canada. (Unpublished.) Effectsof long-termexposure to supersaturation CLARK,M. J. R. 1976. Water ResourcesService dis- of dissolvedatmospheric gases on juvenile chi- solvedgas study: interim data summary.British nook salmon and steelhead trout in deep and Columbia Water Resources Service, Pollution shallowtank tests.Pages 1-10 in Fickeisenand Control Branch, Victoria, Canada. (Unpub- Schneider (1976). lished.) DELL, M. B., M. W. ERHO, AND B. L. LEMON. 1975. CLARK,M. J. R., ANnL. REGAN.1973. A report on Occurrenceof gasbubble disease symptoms on the resultsof a monitoringprogramme for dis- fish in mid-Columbiareservoirs. Grant, Douglas, solvedgases in selectedwaters of BritishColum- and Chelan County Public Utility Districts, bia, 1972. British Columbia Water Resources Ephrata and Wenatchee,Washington, USA. Service,Pollution Control Branch, Victoria, Can- DEMONT,J. D., ANDR. W. MILLER.First reported ada. (Unpublished.) incidenceof gasbubble disease in the heatedef- CLAY, A., A. BARKER,S. TESTAVERDE,R. MARCELLO, fluent of a steamelectric generating station. Pro- AND G. C. McLEoD. 1976. Observations on the ceedingsof the AnnualConference Southeastern effectsof gasembolism in capturedadult men- Association of Gmne and Fish Commissioners hadcu.Pages 81-84 in Fickeisenand Schneider 25:392-399. (1976). DENNISON,B. A., ANDM. J. MARCHYSHYN.1973. A COUTANT,C. C. 1970. Exploratorystudies of the in- devicefor alleviatingsupersaturation of gasesin teractionsof gas supersaturationand tempera- hatcherywater supplies.The ProgressiveFish- ture on mortalityof juvenilesalmonids. Battelle Culturist 35:55-58. Memorial Institute, Pacific Northwest Laborato- DounoRoFF,P. 1957. Water qualityrequirements of ries, Richland,Washington, USA. fishesand effectsof toxicsubstances. Pages 403- COUTANT,C. C., AND R. g. GENOWAY.1968. Final 430 in M. E. Brown, editor. The physiologyof report on an exploratorystudy of interactionof fishes, volume 2. Academic Press, New York, increasedtemperature and nitrogen supersatu- New York, USA. ration on mortalityof adult salmonids.A report DOUGLAS,E. 1964. Solubilitiesof oxygenand nitro- to the United States Bureau of Commercial Fish- gen in distilledwater. Journal of PhysicalChem- eries,Seattle, Washington. Battelle Memorial In- istry68:169-174. stitute, PacificNorthwest Laboratories,Richland, DOUGkaS,E. 1965. Solubilitiesof argon and nitrogen Washington,USA. in seawater.Journal of PhysicalChemistry 69: CttaMER,S. P., AmnJ.D. MCINTYP,Z. 1975. Heritable 2608-2610. resistanceto gasbubble disease in chinooksalm- DUKES,T. W., AND A. R. LAWLER. 1975. The ocular on. United States National Marine Fisheries Ser- lesions of naturally occurring lymphocystisin viceFishery Bulletin 73:934-938. fish. CanadianJournal of ComparativeMedicine eRATIN,P. D., R.J. DEXTER,AND R. W. CURRY.1971. 39:406-410. Precisein situ measurementsof chemicalparam- EBEL,W.J. 1969. Supersaturationof nitrogen in the etersof seawater:a report of mission1-50. Tek- Columbia River and its effect on salmon and tite II. In J. W. Miller,J. G. VanDerwalker,and steelhead trout. United States National Marine R. A. Waller, editors. Scientists-in-the-sea.VI-15- FisheriesService Fishery Bulletin 68:1-11. VI-21. United StatesDepartment of the Interior, E•EL, W.J. 1971. Dissolvednitrogen concentrations Washington,District of Columbia,USA. in the Columbia and Snake rivers in 1970 and DANNEVIG,A., AND G. DANNEVIG. 1950. Factors af- their effect on chinook salmon and steelhead fectingthe survivalof fish larvae.Journal du ConsellInternational pour l'Explorationde la trout. NOAA (National Oceanographicand At- Mer 15:277-283. mospheric Administration) Technical Report Downloaded by [US Fish & Wildlife Service] at 07:25 02 February 2012 D'AousT, B.G., ANn L. S. SMITI•. 1974. Bends in NMFS (National Marine FisheriesService) SSRF fish. ComparativeBiochemistry and Physiology, (SpecialScientific Report Fisheries)646. A, ComparativePhysiology 49:311-321. E•EL, W.J. 1973. The relationshipbetween fish be- D'AousT, B.G., R. WHITE, AND H. SIEBOLD.1976. havior, bioassayinformation and dissolvedgas An electronictensionometer for monitoringtotal concentrationson survival of juvenile salmon dissolvedgas pressure. Pages 106-110 in Ficke- and steelhead trout in the Snake River. Proceed- isen and Schneider (1976). ings of the 53rd Annual Conference,Western DAWLEY,E., T. BLAHM, G. SNYDER,AND W. EBEL. Association of State Game and Fish Commission- 1975. Final report, studieson effectsof super- ers:516-527. saturationof dissolvedgases on fish. National E•EL, W.J. 1979. Effectsof atmosphericgas super- Marine Fisheries Service, Northwest Fisheries saturation on survival of fish and evaluation of Center, Seattle,Washington, USA. proposed solutions. In United States Army DAWLEY,E. M., ANn W. J. EBEL. 1975. Effectsof Corps of Engineers.Fifth progressreport on variousconcentrations of dissolvedatmospheric fisheriesengineering research program 1973- gason juvenilechinook salmon and steelhead 1978. Portland District, Fish and Wildlife Sec- trout. United States National Marine Fisheries tion, Portland,Oregon, USA. ServiceFishery Bulletin 73:787-796. EBEL,W. J., K. T. BEININGEN,G. R. BOUCK,W. R. DISSOLVED GAS SUPERSATURATION--A REVIEW 697

PENROSE,AND D. E. WEITKAMP. 1979. Gases, to- FICKEISEN,D. H., J. C. MONTGOMERY,AND R. W. tal dissolved.Pages 113-118 in R. V. Thurston, HANF,JR. 1976. Effect of temperatureon tol- R. C. Russo, C. M. Fetterolf, T. A. Edsall, and Y. eranceto dissolvedgas supersaturation of black M. Barber, editors. A review of the EPA red bullhead,Ictalurus melas. Pages 72-74 in Fick- book: quality criteria for water. American Fish- eisen and Schneider (1976). eries Society,Bethesda, Maryland, USA. FICKEISEN,D. H., ANDM.J. SCHNEIDER,editors. 1976. EBEL,W. J., E. M. DAWLEY,AND B. H. MONK. 1971. Gas bubble disease. CONF-741033, Technical Thermal toleranceof juvenile salmonin relation Information Center, Energy Researchand De- to nitrogen supersaturation.United StatesFish velopmentAdministration, Oak Ridge, Tennes- and Wildlife Service Fishery Bulletin 69:833- see, USA. 843. FICKEISEN,D. H., M. J. SCHNEIDER,AND J. C. MONT- EBEL, W. J., R. W. KROMA, AND H. L. RAYMOND. GOMERY.1975. A comparativeevaluation of the 1973. Evaluationof fish protectivefacilities at Weis saturometer. Transactions of the American Little Goose Dam and review of other studies FisheriesSociety 104:816-820. relatingto protectionof juvenile salmonidsin the GERSH, I., G. E. HAWKINSON, AND E. H. RATHBUN. Columbia and Snake rivers, 1973. National Ma- 1944. Tissue and vascular bubbles after de- rine Fisheries Service, Northwest Fisheries Cen- compressionfrom high pressureatmospheres• ter, Seattle,Washington, USA. correlationof specificgravity with morphological EBEL,W. J., ANDH. L. RAYMOND.1976. Effect of changes.Journal of Cellular and Comparative atmosphericgas supersaturation on salmonand Physiology24:35-70. steelhead trout of the Snake and Columbia GORHAM,F. P. 1898. Some physiologicaleffects of rivers. United States National Marine Fisheries reduced pressureon fish..Journalof the Boston Service Marine Fisheries Review 7:1-14. Societyof Medicine 3:250. EBEL,W. J., H. L. RAYMOND,G. E. MONAN,W. E. GORHAM,F. P. 1901. The gasbubble disease of fish FARR, AND G. K. TANONAKA. 1975. Effect of at- and its causes. Bulletin of the United States Fish mosphericgas supersaturation caused by dams Commission (1899) 19:33-37. on salmon and steelhead trout of the Snake and GREEN,E. J., ANDD. E. CARRITT. 1967. New tables Columbia rivers. National Marine Fisheries Ser- for oxygen saturationof sea water. Journal of vice, Northwest Fisheries Center, Seattle, Wash- Marine Research 25:140-147. ington, USA. GUTSELL,J. S. 1929. Influence of certain water EBELING,G. 1954. Sauerstoff-•bers5ttigungin flies- senden und stehenden Gew•issern. Vom Wasser conditions,especially dissolved gases, on trout. 21:84-99. Ecology10:77-96. HARRIS1, M., W. E. BERG, D. M. WHITAKER, AND V. EGUSA,S. 1959. The gasdisease of fish due to excess C. TWITTY. 1945. The relationshipof exercise of nitrogen.Journal of the Facultyof Fisheries to bubbleformation in animalsdecompressed to and Animal Husbandry, Hiroshima University 2:157-182. sealevel from high barometricpressure. Journal of General Physiology28:241-251. ELLING,D. H., ANDW. J. EBEL. 1973. Nitrogen su- persaturationin the Columbiaand Snakerivers, HARRIS2, M., W. E. BERG, D. M. WHITAKER, V. C. a disasteror blessingin disguise.Trout 14(2):5. TWITTY, AND L. R. BLINKS. 1945. Carbon diox- ELMORE,H. L., ANDT. W. HAYES.1960. Solubility ide as a facilitatingagent in the initiation and of atmosphericoxygen in water.Journal of the growthof bubblesin animalsdecompressed to SanitaryEngineering Division, American Society simulatedaltitudes. Journal of GeneralPhysiol- of Civil Engineers86:41-53. ogy 28:225-240. EMBODY,G. C. 1934. Relationof temperatureto the HARVEY,E. N. 1951. Physicalfactors in bubblefor- incubation period of eggs of four speciesof mation. Chapter 4, page 90 in J. F. Fulton, edi- Downloaded by [US Fish & Wildlife Service] at 07:25 02 February 2012 trout. Transactions of the American Fisheries tor. Decompressionsickness. Saunders, Philadel- Society64:281-292. phia, Pennsylvania,USA. ENGELHORN,O. R. 1943. Die Gasblasenkrankheitbei HARVEY, E. N., A. H. WHITELY, W. D. MCELRoY, D. Fischen. Zeitschrift ffir Fischerei 41:297-317. C. PEASE,AND D. K. BARNES. 1944. Bubble for- ERDMAN,F. S. 1961. How a heat pump improved marionin animals.II. Physicalfactors. Journal water conditions at a fish hatchery. ASHRAE of Cellularand ComparativePhysiology 24:23- Journal 3:62-64. 34. EVANS,A., ANDD. N. WALDER.1969. Significanceof HARVEY,H. H. 1961. Analysisof dissolvednitrogen gas micronucleiin the aetiologyof decompres- in fish blood. Canadian Journal of Zoology sion sickness.Nature (London) 222:251-252. 39:905-909. FAST,A. W. 1979. Nitrogen gassupersaturation dur- HARVEY,H. H. 1964. Dissolvednitrogen as a tracer ing artificial aerationof Lake Casitas,California. of fish movements. International Association of The ProgressiveFish-Culturist 41:153-155. Theoreticaland AppliedLimAology Proceedings FAST,A. W., V. A. DORR,AND R. J. ROSEN.1975. A 15:947-951. submerged hypolimnion aerator. Water Re- HARVEY,H. H. 1967. Supersaturationof lake water sources Research 11:287-293. with a precaution to hatchery usage. Transac- 698 WEITKAMP AND KATZ

tions of the American FisheriesSociety 96:194- JOHNSON,P. T. 1976. Gasbubble disease in the blue 201. crab, Callinectessapidus. Journal of Invertebrate HARVEY, H. H. 1975. Gas disease in fishes•a review. Pathology27:247-254. Pages450-485 in W. A. Adams, editor. Chem- JOHNSON,R. C., ANDE. M. DAWLEY.1974. Final re- istry and physicsof aqueousgas solutions. The port. The effect of spillwayflow deflectorsat ElectrochemicalSodety, Princeton,New Jersey, BonnevilleDam on the total gassupersaturation USA. and survivalof juvenile sahnon.National Marine HARVEY,H. H., ANDA. C. COOPER.1962. Origin and Fisheries Service, Seattle, Washington, USA. treatment of a supersaturatedriver water. Pro- (Unpublished.) gress Report 9, International Pacific Sahnon JONES,D., ANDD. H. LEWIS. 1976. Gas bubble dis- Fisheries Commission, Vancouver, Canada. easein fry of channelcatfish (Ictalurus punctatus). HARVEY,H. H., ANDS. B. SMITH. 1961. Supersatu- The ProgressiveFish-Culturist 38:41. ration of the water supplyand occurrenceof gas KLOTTS, C. E., AND B. B. BENSON. 1963. Solubilities bubbledisease at Cultus Lake Trout Hatchery. of nitrogen,oxygen and argonin distilledwater. Canadian Fish Culturist 30:39-46. Journal of Marine Research21:48-56. HEMMINGSEN,E. A. 1970. Supersaturationof gases KRABACH,M. H., ANDR. A. MARCELLO,JR.1976. An in water: absenceof cavitationor decompression analysisof the feasibilityof usingair agitationto from high pressures.Science (Washington, Dis- reduce gas supersaturationin sea water at Pil- trict of Columbia) 167:1493-1494. grim Nuclear Power Station. Boston Edison HENLY,E. 1952. The influenceof the gascontent of Company,Boston, Massachusetts, USA. (Unpub- sea-wateron fish and fish larvae. Rapports et lished.) Proces-verbaux des Reunions, Conseil Interna- LEE,G. F., ANDP. H. MARTIN. 1975. Dissolvedgas tional pour rExploration de la Mer 131(3):24- supersaturationand dilution in thermal plumes 27. froin steam electricgenerating stations. Water HILLS,B. A. 1967. Decompressionsickness: a study Research 9:643-648. of cavitationat the liquid-liquidinterface. Aero- LIGHTNER, D. V., B. R. SALSER,AND R. S. WHEELER. spaceMedicine 38:814-817. 1974. Gas-bubbledisease in the brown shrimp HOFFERT,J. R., M. B. FAIRBANKS,AND P.O. FROMM. (Penaeusaztecus). Aquaculture 4:81-84. 1971. Ocular oxygen concentrationsaccompa- LINDROTH,A. 1957. Abiogenicgas supersaturation nying severe chronic ophthahnic pathologyin of river water. Archiv f/Jr Hydrobiologie53:589- the lake trout (Salvelinusnamaycush). Compara- 597. tive Biochemistryand Physiology,A, Compara- LONG,C. W., AND F. J. OSSIANDER.1974. Interim tive Physiology39:137-145. report on survival of coho salmon fingerlings H6LL, K. 1955. ChemischeUntersuchungen an klei- passingthrough a perforated bulkhead in an nen Fliessgewassern.International Association of empty turbine bay and through flow detectors Theoreticaland Applied LimnologyProceedings 12:360-372. (with and without dentares)on spillwayat Lower Monumental Dam, Snake River. Report to HOPPE-SEYLER,F. 1857. Archiv f/Jr Anatomie,Phy- United StatesArmy Corps of Engineers,April- siologieund wissenschaftlicheMedizin 24:63-73. May, 1973. National Marine Fisheries Service, Cited in L. Hill and M. Greenwood. 1910. On Seattle,Washington, USA. the formation of bubbles in the vessels of animals LONG,C. W., F. J. OSSIANDER,T. E. RUEHLE,AND G. submittedto partialvacuum. Journal of Physiol- M. MATTHEWS. 1975. Survival of coho salmon ogy (London)39:23. fingerlings passingthrough operating turbines HUGHES,J. T. 1968. Grow your own lobsterscom- with and without perforated bulkheadsand of mercially.Ocean Industry 3(2):46-49. ISAACSON,P. A. 1977. Theoretical consideration of steelheadtrout fingerllngspassing through spill- Downloaded by [US Fish & Wildlife Service] at 07:25 02 February 2012 wayswith and without a flow deflector.Final Pro- the causeof gasbubble disease in fishin a heated effluent. New York Fish and Game Journal gressReport to United StatesArmy Corps of 24:95. Engineers. National Marine Fisheries Service, Seattle,Washington, USA. JARNEFELT,V. H. 1948. Der Einflussder Strom- schnellen auf den Sauerstoff und Kohlensaure- MACDONALD,J. R., ANDR. A. HYATT. 1973. Super- gehaltund daspH desWassers im FlusseVuoski. saturationof nitrogen in water during passage InternationalAssociation of Theoreticaland Ap- through hydroelectric turbines at Mactaquac plied LimnologyProceedings 10:210-215. Dam. Journal of the FisheriesResearch Board of Canada 30:1392-1394. JENKINS,T. F. 1976. Continuousmonitoring of total dissolvedgases, a feasibilitystudy. Pages 101-105 MALOUF, R., R. KECK, D. MAURER, AND C. EPIFANIO. in Fickeisenand Schneider (1976). 1972. Occurrenceof gas-bubbledisease in three JENSEN,L. D., editor. 1974. Environmentalre- speciesof bivalveroofloses. Journal of the Fish- sponsesto thermal dischargesfroin Marshal eries Research Board of Canada 29:588-589. Steam Station, Lake Norman, North Carolina. MARCELLO,R. A., JR., AND R. B. FAIRBANKS.1976. Report to ElectricPower ResearchInstitute and Gas bubble disease of Atlantic menhaden, Bre- Duke PowerCompany, Raleigh, North Carolina, voortiatyrannus, at a coastalnuclear power plant. USA. Pages75-80 in Fickeisenand Schneider(1976). DISSOLVED GAS SUPERSATURATION--A REVIEW 699

MARCELLO,R. A., JR., ANDR. K. STRAWN.1972. The R. Sharitz, editors. Thermal ecology.CONF- cage culture of some marine fishesin the intake 730505, United States Environmental Protection and dischargecanals of a steam-electricgener- Agency,Washington, District of Columbia,USA. atingstation, Galveston Bay, Texas. Texas A&M MONAN, G. E., AND K. L. LISCOM. 1975. Radio-track- UniversitySea Grant PublicationTAMU-SG-72- ing studiesto determine the effect of spillway 206, Galveston, Texas, USA. deflectors on adult chinook salmon and steelhead MARSH,M. C. 1903. A fatality amongfishes in water trout at Bonneville Dam, 1974. National Marine containingan excessof dissolvedair. Transac- Fisheries Service, Northwest Fisheries Center, tionsof the AmericanFisheries Society 32:192- Seattle,Washington, USA. 193. MORTIMER,C. H. 1956. The oxygencontent of air- M^Rsn, M. C. 1910. Notes on the dissolved content saturatedfresh waters,and aids in calculating of water in its effect upon fishes.Bulletin of the percentage saturation. International Association United StatesBureau of Fisheries(1908) 28:891- of Theoretical and Applied Limnology Com- 906. munications 6:1-20. MARSH,M. C., ANDF. P. GORHAM.1905. The gas MRslc, W. 1933. Die Gasblasenkrankheit der Fische. diseasein fishes.Report of the United StatesBu- Ursachen,Begleiterscheinungen und Abhilfe. reau of Fisheries(1904):343-376. Zeitschrift ffir Fischerei 31:29-67. MARSHALL,W. 1976. Equipmentand techniquesfor MURRAY,C. N., J.P. RILEY, AND T. R. S. WILSON. monitoring the vertical distributionof fish in 1969. The solubilityof gasesin distilled water shallowwater. Pages20-23 in Fickeisenand and seawater.I. Nitrogen. Deep-SeaResearch Schneider(1976). 16:297-310. MATSUE, Y., S. EGt•SA,AND A. SAEKI. 1953. On ni- NATIONAL ACADEMY OF SCIENCES/NATIONAL ACADEMY trogen gascontents dissolved in flowingwater of OF ENGINEERING.1972. Water quality criteria artesianwells and springs.(Relating to high su- 1972, a report to the Committeeon Water Qual- persaturation inducing the so-called"gas-dis- ity Criteria. EPA-R3-73-033, United StatesEn- ease"upon fishes.)Bulletin of the JapaneseSo- vironmental Protection Agency, Washington, ciety of ScientificFisheries 19:439-444. District of Columbia, USA. MAY,B. 1973. Evaluationon the effectsof gasbub- NEBEKER,A. V. 1976. Survivalof Daphnia,crayfish ble diseaseon fish populationsin the Kootenai and stoneflies in air-supersaturatedwater. Jour- Riverbelow Libby Dam. Proceedingsof the 53rd nal of the Fisheries Research Board of Canada Annual Conference Western Association of State 33:1208-1212. Fish and Game Comnfissioners:525-540. NEBEKER•,A. V., G. R. BOUCK, AND D.G. STEVENS. McLEOD,G. C. 1978. The gasbubble disease of fish. 1976. Oxygenand carbondioxide and oxygen- Pages319-339 in D. I. Mostofsky,editor. The nitrogenratios as factors affecting fish in air-su- behaviorof fish and other aquaticanimals. Ac- persaturated water. Transactions of the Ameri- ademic Press,New York, New York, USA. can FisheriesSociety 105:425-429. MEEKIN,T. K. 1971. Levelsof nitrogensupersatu- NEBEKER,A. V., ANDJ. R. BRETT. 1976. Effects of ration at Chief JosephDam under variousspill air-supersaturatedwater on survival of Pacific conditions,phase 1. Report to United States salmon and steelhead smolts. Transactions of the Army Corps of Engineers,Contract NPSSU-71- AmericanFisheries Society 105:338-342. 796, Washington Department of Fisheries, NEBEKER,A. V., D.G. STEVENS,AND R. J. BAKER. Olympia, Washington,USA. 1979. Survival of salmon smolts in sea water af- MEEKIN, T. K., AND R. L. ALLEN. 1974a. Effects of ter exposure to air-supersaturatedwater. The controlledspilling at Chief JosephDam on dis- ProgressiveFish-Culturist 41:30-32. solvednitrogen supersaturation in the Columbia NEBEKER2,A. V., D.G. STEVENS,AND J. R. BRETT.

Downloaded by [US Fish & Wildlife Service] at 07:25 02 February 2012 River. Washington Department of Fisheries 1976. Effects of gas supersaturatedwater on TechnicalReport 12:1-31. freshwateraquatic invertebrates. Pages 51-65 in MEEKIN,W. K., ANDR. L. ALLEN. 1974b. Nitrogen Fickeisenand Schneider(1976). saturation levels in the mid-Columbia River, NEBEKERa,A. V., a. G. STEVENS,AND R. K. STROUD. 1965-1971. WashingtonDepartment of Fisher- 1976. Effects of air-supersaturatedwater on ies TechnicalReport 12:32-77. adult sockeyesalmon (Oncorhynchus nerka). Jour- MEEKIN, T. K., ^ND R. L. ALLEN. 1975. Summer chi- nal of the Fisheries Research Board of Canada nook and sockeyesalmon mortality in the upper 33:2629-2633. ColumbiaRiver and its relationshipto nitrogen NEWCOMB,T. W. 1976. Changesin bloodchemistry supersaturation. Washington Department of of juvenile steelhead,Salmo gairdneri, following FisheriesTechnical Report 12:127-153. sublethalexposure to nitrogensupersaturation. MEEKIN, T. K., AND B. K. TURNER. 1974. Tolerance Pages96-100 in Fickeisenand Schneider(1976). of salmonideggs, juveniles and squawfishto su- OEST1NG,R. B. 1934. A modifiedVan Slykemethod persaturatednitrogen. WashingtonDepartment for the determinationof dissolvedoxygen and of FisheriesTechnical Report 12:78-126. total carbondioxide in water. PhysiologicalZo- MILLER,R. W. 1974. Incidenceand causeof gasbub- ology7:542-549. ble disease.Pages 79-93 in J. W. Gibbonsand R. P^RAMETRIX.1974. SnakeRiver 1973 dissolvedgas 700 WEITKAMP AND KATZ

studies.Report by Parametrix,Incorporated (D. gengas (Ns) model for the lowerColumbia River. E. Weitkamp) to Idaho Power Company,Boise, Final Reportto the United StatesArmy Corpsof Idaho, USA. Engineers,Portland District. National Marine PAULEY,G. B., M.P. FUJIHARA,AND R. E. NAKATANI. Fisheries Service, Northwest and Alaska Fisher- 1966. Salmon prespawningmortality studiesat ies Center, Seattle,Washington, USA. RockyReach Dam and PriestRapids Dam, Wash- RVCKER,R. R. 1976. Gas bubble disease of salmo- ingtonfrom Septemberto November1965. Con- nids:variation in oxygen-nitrogenratio with con- tract 13565-8, WashingtonDepartment of Fish- stanttotal gas pressure. Pages 85-88 in Fickeisen eries. Biology Department, Battelle Memorial and Schneider (1976). Institute, Pacific Northwest Laboratories, Rich- RvcasR, R. R., AND K. HOD•EBOO•. 1953. Obser- land, Washington,USA. vationson gasbubble disease of fish. The Pro- PAULEY, G. B., AND R. E. NAKATANI. 1967. Histo- gressiveFish-Culturist 15:24-26. pathologyof "gasbubble" disease in salmonfin- RVCKER,R. R., AND P. H. KANtAS. 1974. Effect of gerlings.Journal of the FisheriesResearch Board nitrogensupersaturated water on cohoand chi- of Canada 24:867-871. nook salmon. The ProgressiveFish-Culturist PETERSON,H. 1971. Smolt rearing methods,equip- 36:152-156. ment and techniquesused successfullyin Swe- RVC•ER, R. R., AND E. M. TVTrLE. 1948. Removal den. Atlantic Salmon FoundationSpecial Publi- of excessnitrogen in a hatcherywater supply. cation Series 2(1):32-62. The ProgressiveFish-Culturist 10:88-90. PLEHN, M. 1924. Praktikum der Fischkrankheiten. RUKAVINA,J., ANDD. VARENIKA.1956. Air bubble Pages301 •,79 in Handbuckder Binnenfischerei disease of trout of the source of the river Bosna. mitteleuropas,volume 1. E. Schweizerbartisch Acta Ichthyologica Bosniae et Hercegovinae Verlagsbuchhandlung,Stuttgart, West Germany. 1(7):5-12. Post, G. 1970. Determinationof inert gasesin water. RULIFSON,R. L., ANDG. ABEL. 1971. Nitrogen su- The ProgressiveFish-Culturist 32:170-173. persaturationin the Columbiaand Snakerivers. POSTON,H. A., D. L. LIVINGSTON, AND D. G. HEDRIGK. Working Paper 82, Technical Report TSO-9- 1973. Effectof methodof introducingwell water 208-016.2, United States Environmental Protec- on the growth, body chemistry,and incidenceof tion Agency Region 10, Seattle, Washington, eyeabnormalities of juvenile brown and rainbow USA. trout. The ProgressiveFish-Culturist 35:187- RULIFSON,R. L., ANDR. PINE,chairmen. 1976. Water 190. quality standards.Page 120 in Fickeisenand RAYMOND,H. L. 1968. Migration rates of yearling Schneider (1976). chinook salmon in relation to flows and im- SATOri,Y. 1955. Hydrochemicalobservation of the poundmentsin the Columbiaand Snakerivers. ground water for trout culture around Hino- Transactionsof the American FisheriesSociety town, Tokyo. Bulletin of the FreshwaterFisher- 97:356-359. ies ResearchLaboratory (Tokyo) 5:31-38. RAYMOND,H. L. 1969. Effectof John Day Reservoir SCI•IEWE, M. H. 1974. Influence of dissolved atmo- on the migrationrate of juvenile chinooksalmon sphericgas on swimmingperformance of juve- in the Columbia River. Transactions of the nile chinook salmon. Transactions of the Amer- American FisheriesSociety 98:513-514. ican FisheriesSociety 103:717-721. RAYMOND,H. L. 1970. A summaryof the 1969 and SCI•IEWE, M. H., AND D. D. WEBER. 1976. Effect of 1970 outmigrationsof juvenile chinook salmon gas bubble diseaseon lateral line function in ju- and steelhead trout from the Snake River. Bu- venile steelheadtrout. Pages89-92 in Fickeisen reau of CommercialFisheries Biological Labo- and Schneider (1976). ratory, Seattle,Washington, USA. SCI•ASSMAN,H. 1951. Untersuchungenfiber den RAYMOND, H. L. 1979. Effects of dams and im- Stoffhaushalt GewSsser. Schweizerische Zeit- Downloaded by [US Fish & Wildlife Service] at 07:25 02 February 2012 poundmentson migrationsof juvenile chinook schrift ffir Hydrologie 13:300-335. salmon and steelhead from the Snake River, SEATTLEMARINE LABORATORIES. 1972a. Final report. 1966 to 1975. Transactions of the American 1972nitrogen monitoring studies. Report (D. E. FisheriesSociety 108:505-529. Weitkamp) to Idaho Power Company, Boise, RECNARD,P. 1884. Effect des hautes pressionssur Idaho, USA. (Unpublished.) les animaux marins. Comptes Rendus des Se- SEATTLE MARINE LABORATORIES. 1972b. Dissolved ancesde la Societede Biologie de Paris 36:394- gas monitoring at Grand Coulee Dam and Lake 395. Roosevelt.Report to United StatesDepartment RENFRO,W. C. 1963. Gas-bubblemortality of fishes of the Interior, Bureau of Reclamation, Boise, in Galveston Bay, Texas. Transactions of the Idaho, USA. (Unpublished.) American FisheriesSociety 92:320-322. SEATTLEMARINE LABORATORIES. 1974. Dissolvedgas REUSS,H. 1910. Die Wirkung der Kohlens•iureauf supersaturation,Grand Coulee Dam Project, die Atmung der niederenWirbeltiere, inbeson- 1973. Report (D. E. Weitkamp) to United States dere der Fische.Zeitschrift ffir Biologie53:555- Department of the Interior, Bureau of Recla- 587. mation, Boise, Idaho, USA. ROESNER,L. A., AND W. R. NORTON. 1971. A nitro- SI•ELFORD,V. E., AND W. C. AI•LEE. 1913. The re- DISSOLVED GAS SUPERSATURATION--A REVIEW 701

actionsof fishesto gradientsof dissolvedatmo- TIONAGENCY). 1971. Nitrogen supersaturation sphericgases. Journal of ExperimentalZoology in the Columbia and Snake rivers, summary re- 14:207-266. port. Region 10, Seattle,Washington, USA. Smga•ATA, S. 1966. Experimentson nitrogen gas USEPA (UNITED STATES ENVIRONMENTALPROTEC- diseasewith rainbowtrout fry. Bulletin of the TIONAGENCY). 1976. Qualitycriteria for water. FreshwaterFisheries Research Laboratory (To- Washington,District of Columbia, USA. kyo) 15:197-211. WEBER,D. D., ANDM. H. SCH1EWE.1976. Morphol- SMIT•, H. A.,JR. 1972. N2 threat to Pacificnorthwest ogy and function of the lateral line of juvenile fisheries?Water Spectrum4(2):41-48. steelheadtrout in relation to gasbubble disease. SMIT•,J.R. 1974.Distribution of seaward-migrating Journal of Fish Biology9:217-233. chinook salmon and steelhead trout in the Snake WEIGELT,C. 1885. Schadingungder Fischerei.Hy- River above Lower Monumental Dam. United giene 3:82. States National Marine Fisheries Service Marine WEISS,R. F. 1970. The solubilityof nitrogen, oxy- FisheriesReview 36(8):42-45. gen, and argonin water and seawater.Deep-Sea STICKNEY,A. P. 1968. Supersaturation of atmo- Research 17:721-735. spheric gasesin coastalwaters of the Gulf of WEITKAMP,D. E. 1974. Dissolvedgas supersatura- Maine. United States Fish and Wildlife Service tion in the Columbia River system:salmonid FisheryBulletin 67:117-123. biDassayand depth distributionstudies, 1973 and STROUO, R. K., G. R. BOUCK, AND A. V. NEBEKER. 1974. Parametrix, Incorporated, Report to Util- 1975. Pathologyof acuteand chronicexposure ity Cooperative,care of Idaho Power Company, of salmonidfishes to supersaturatedwater. Pages Boise, Idaho, USA. 435-449 in W. A. Adams,editor. Chemistry and WEITKAMP,D. E. 1976. Dissolvedgas supersatura- physicsof aqueousgas solutions.The Electro- tion: live cage biDassaysat Rock Island Dam, chemicalSociety, Princeton, New Jersey,USA. Washington. Pages 24-36 in Fickeisen and STROUO,R. K., ANDA. V. NEBE}•ER.1976. A study Schneider (1976). of the pathogenesisof gasbubble disease in steel- WELLS, M. M. 1913. The resistance of fishes to dif- headtrout (Salmogairdneri). Pages 66-71 in Fick- ferent concentrationsand combinationsof oxy- eisen and Schneider (1976). gen and carbon dioxide. Biological Bulletin SUPPLEE,V. C., AND D. V. LIGHTNER. 1976. Gas-bub- (Woods Hole) 25:323-347. ble diseasedue to oxygen supersaturationin WESTGARD,R. L. 1964. Physicaland biologicalas- raceway-rearedCalifornia brown shrimp. The pectsof gas-bubbledisease in impounded adult ProgressiveFish-Culturist 28:198-199. chinook salmon at McNary spawningchannel. SwimmERTON,J. W., V. J. LINNENBOM,AND C. H. Transactionsof the American FisheriesSociety C•EE}•. 1962. Determinationof dissolvedgases 93:306-309. and aqueoussolutions by gaschromatography. WHITAKER, D. M., L. R. BLINKS, W. E. BERG, V. C. AnalyticalChemistry 34:483-485. TWITTY, AND M. HARRIS. 1945. Muscular activ- TERVOOREN,H. P. 1972. Bonnevillespillway test de- ity and bubble formation in animals decore- flector installationin bay 18. Preliminaryevalu- pressedto simulatedaltitudes. Journal of Gen- ation report on nitrogen tests. United States eral Physiology28:213-233. Army Corpsof Engineers,Portland District, Hy- WIE•E, A. H., ANDA.M. McGAvocK. 1932. The abil- draulicDesign Section, Portland, Oregon, USA. ity of severalspecies of fish to surviveon pro- TERVOOREN,H. P. 1973. Bonnevillespillway nitro- longedexposure to abnormallyhigh concentra- gen tests, 29-30 October, 1972, test results. tions of dissolvedoxygen. Transactionsof the United StatesArmy Corps of Engineers,Port- American FisheriesSociety 62:267-274. land District, Hydraulic Design Section,Port- WINTERSTEIN,H. 1908. Beitrage zur Kenntnis der land, Oregon, USA. Fischatmung.Pflugers Archiv ffir die gemaNte Downloaded by [US Fish & Wildlife Service] at 07:25 02 February 2012 TOLK, A., W. A. LINGERAK,A. KOUT, AND D. BERGER. Physiologiedes Menschen und der Tiere 125: 1969. Deternfinationof tracesof hydrogen,ni- 73-98. trogen and oxygenin aqueoussolutions by gas WOELKE, R. A., G. R. BOUCK, AND R. K. STROUD. chromatography.Analytica Chimica Acta 45:137. 1974. Gas-bubble disease: a review in relation to UNITED STATES ARMY CORPS OF ENGINEERS. 1975. modernenergy production. Pages 239-266 in S. Dissolvedgas data report 1974. North Pacific B. Saila,editor Fisheriesand energyproduction. Division,Portland, Oregon, USA. Lexington Books, Lexington, Massachusetts, UNITED STATES ARMY CORPS OF ENGINEERS. 1977. USA. Dissolvedgas data report 1975-1976. North Pa- WOLD,E. 1973. Surfaceagitators as a meansto re- cific Division,Portland, Oregon, USA. duce nitrogen gas in a hatcherywater supply. UNITED STATES ARMY CORPS OF ENGINEERS. 1979. The ProgressiveFish-Culturist 35:143-146. Fifth progressreport on fisheriesengineering WOOD,J. W. 1968. Diseasesof Pacificsalmon, their researchprogram 1973-1978. Portland District, preventionand treatment. WashingtonDepart- Fish and Wildlife Section,Portland, Oregon, ment of Fisheries,Hatchery Division,Olympia, USA. Washington,USA. USEPA (UNITED STATESENVIRONMENTAL PROTEC- WOODBURY,L. A. 1941. A suddenmortality of fishes 702 WEITKAMP AND KATZ

accompanyinga supersaturationof oxygen in gas bubble diseaserelated to a hatchery water Lake Waubesa, Wisconsin. Transactions of the supplyfrom the forebay of a high head re-reg- American FisheriesSociety 71:112-117. ulatingdam. ResearchReports of the FishCom- WYATT,E. J., ANDK. T. BEININGEN.1969. A nitro- missionof Oregon 3:3-12. gen gasdisease catastrophe. Proceedings of the Z•RGES,M. H., ^•D L. D. CVRTIS.1975. An experi- 20th Annual Northwest Fish Culture Confer- mental heated water incubation systemfor sal- ence:71-72. monid eggs. The Progressive Fish-Culturist WY^TT, E.J., ^•D K. T. BE•GE•. 1971. Nitrogen 37:217-218. Downloaded by [US Fish & Wildlife Service] at 07:25 02 February 2012