MEDICAL SUPPORTCONSIDERATIONS FOR MIXED GAS DIVING AT WARMMINERAL SPRINGSARCHAEOLOGICAL RESEARCH PROJECT

William ikpper, MD. FamilyPractice of Tallahassee,FL 1885 Professional Park Circle, Suite 30 Tallahassee,FLORIDA 32308 U. S.A.

Thispaper will describe the necessary medical support for theunderwater ar- chaeologicalstudies in thedeep portion of WarmMineral Springs, including considerationsfortreatment ofaccidents, prevention ofaccidents bydecornpressirig ongas mixtures other than air, monitoring forpossible venous gas emboli with dopplertechruque and utilization of "Tnmix" air-helium mixed gases for more effectivedeep water diving by decreasing risksof inert gas narcosis aswell as increasingbottom time available. Thisinfonnation maybe employed asa template forother deep water greater than 130 feet! scientific dive projects which can be managedbysimilar support considerations. Itisoffered aspart of a seriesofpapers aboutthe Barm Mineral Springs Archaeological Research Project which in con- tinuitywilt approach the scope of utilizationof mixed gas diving in scientific research

INTRODUCTION For15 years research in the unique environment of Warm Mineral Springs has been providingvaluable information from the archaeological andpaleontological remains that are wellpreserved atmultiple levels of the spring. However, a large part of this valuable material islocated indepths inexcess ofwhat is considered tobe safe for underwater research done by diverson compressed air.In 1987, the oversight ofthe Warm Mineral Springs Archaeological ResearchProject moved from Manatee Community College tothe Department ofAnthropol- ogyat Florida State University andbegan working with the Acadeinic Diving Program ADP! todevelop diving standards and procedures forthis deep site. The already well established linkbetween the ADP Medical Advisor, as well as the ongoing integration with the Dept. of Archaeologyallowed for these resources tobe applied tothe problems of the deeper diving research.What was not readily available atthe ADP systems wassignificant expertise inmixed gas other than air! diving. Therationale for theuse of mixedgas was based upon two factors. The fist wasa concernabout the risk of involved with prolonged hyperbaric exposures onair. Divirrg for Science...1999

Underwaterresearch atthis site required labor-extensive longbottom-times atdepths below 1SOfsw. This resulted inincreasingly longer periods ofdecompression, oftenon exceptional exposuretables, which had an unacceptable riskof decompressionsickness. The philosopIxy of theADP has been conservative withregard to therisk of exposureto .Forexample, forroutine air diving, we recommend using the D.C.I.E.M. Tables which arerecognized as ultra-conservative in the industry. Prior to FSU'sinvolvement in theVAum MineralSprings Archaeological Research Project WMSARP!, shallow water 02 decompres- sionhad been employed todimirush decompression risks.Moving tothe use of fortifiedair! to increase the off-gas gradient atdeeper decompression stopsin conjunctioxx with the shallow02 stopswas an important enhancement we recommendedto lessen decompressionrisks.

Breathingnitrox at deeper depths, however, looses its advantages because ofincreased risksof 02 Central Nervous System toxicity. Yet the air mixture atthese depths includes too muchnitrogen, inducing an unacceptable level of narcosis.Our secondconcern for these deeperdives and to avoid unnecessary andexpense ofheliox!, was resolved by usinga trimixofhelium, nitrogen and oxygen. This would reduce the risk of diving accidents andallow for greater efficiency inthe precise work of . Becauseofthe complexity in dealing with mixed gases, ie.decisions regarding what tablestouse, optimal gas mixtures andthe calculations involved, wereferred the expertise aMI experienceofR. W. Hamilton, Ph.D. He was able to extrapolate data from dives made to the depthsthat our researchers wouldbe making and develop tables with reasonable decompres- siontimes utilizing the trirnix, nitrox, and in water 02 decompression gases.Since decompres- sionsickness hasbeen described as,at least inpart, a statisticalevent, and because ofthe many variablesinvolved, doppler monitoring of ourresearch divers was included as a checkof our proceduresand as an indicator of thereliability of ourtables. Theaddition ofan on site fully operational double lock multiplace recompressioa chamberfor treatment ofpossible diving accident was integral to ourplans to insure diver safetyshould decompression symptoms appear. Inorder to fully support the chamber it was necessaryforthe Program Physician author! tobecome acquainted with the operations of recornpressionchambers andbecome qualified intheir use to treat diving accidents. %here arequite a fewprograms around thecountry where istaught. The bm weekduration program taught by N.O.A.A. and the military programs of Navy & AirForce tendto bethe most inclusive. 'Were are other programs which concentrate onthe use of HyperbaricOxygen forMedical reasons. Wechoose theweek long course taught byformer NOAAinstructor Dick Rutkowski atHyperbaric International, because ofits emphasis with handson use of a fullyoperational double lock multiplace chamber and its on treatingdiving related maladies. Both the Program Physician and the Assistant Director of theADP attended theclass sothat our direction ofthe chamber operator anddive supervisor wouldbe more rneaningfuL Thisalso provided theADP & WMSARPwitha personwhowas capableof operatingthe chamber should that be needed in thefuture.

226 Kepper;Medical Support for MixedGas Diving at SWS

Choiceof a chamberoperator-dive supervisor was done by a combinedeffort of the managementsofWarm Mineral Springs Archaeological Research Project and the ADP. An agreementwasreached aspart of total plan of how the dive supervisor would interact with thechief scientist and his relationship with the ADP! that allowed adequate communication betweenallparties without undue interference withthe research effort. In order to enhance communication,andto develop proper procedures forthe scope of the project, an on-site visit wasmade by the Program Physician tofamiliarize himself with the dive site, support facilities, theoperation of the chamber and to consultwith the chief scientist aswell as the dive supervisor, Partof the purpose ofthe on site visit was to develop a Medical Plan for WMSARP to beused in evaluationand treatment of potentialdiving mishaps. The contents of thestandard ADPfirst aid kit andrelated equipment were augmented with the addition of stethoscopes, sphygmomanometer bloodpressure cuff!, reflex hanirner, and a positivepressure, resus- citator.Attempts were made to enlistthe local E.M.S, as a secondaryassist to our initial handlingofdiving related injury. This would allow further treatment tobe initiated such as intravenousfluid administrationand respiratory assist during recompression of a diver. Furtherrefinement of the medicalplan continued in Tallahasseeover a periodof severalmonths. Decisions regarding access to thechamber by thepublic, procedures for precautionarytreatments, return towork policies, etc.were debated extensively before a final draftwas agreed upon. The final document was drafted by Mrs. Barbara O'Horo Benton, managerof theWMSARP, and is included as appendix 1. An integralpart of theMedical Plan was the development of an evaluation and emergencynetwork system which could be activated bythe . Alldivers undergodoppler evaluations atset intervals after their dives. In an attempt to quantifyany intravascularbubbles that were detected by the Diving Supervisor, we hoped by early treat- rnenteither by 100 % oxygenatthe surface ora precautionarychamber treatment Table 5 USN!to avoid the occurrence ofan injury and delay of on-going research due to the injury. Theplan also included allowing the scientist who ascended without symptoms ofDCS or with complaintoffatigue torequest a precautionary treatment if he/she felt it appropriate00 1o 02 or a runin thechamber!. The 100go 02 wouMhave to becontinued for at least1/2 hour if begun,and would be monitored forsigns and symptoms ofDCS and for bubble formationwith the doppler. If diagnosisofDCS was not made at this time, the treatment is recordedwithout penalizingthe researcher. Themedical plan addressed theintricacies ofimmediately notifying the needed per- sonnelshould an incident occur, in spiteof theremote location of theresearch site. It was decidedthat rapid transmission ofwritten, and printed material to and from the dive site was bestserved by a "facsimile"machine. This permitted dive statistics and the diver's condition, nowreduced toa reportsheet, to be communicated taFSU's ADP where it couldbe walked to Dr. Kepperand other FSU administrative offices! as well transmitted by FAXto our cortsultantDr. Hamilton for his input. The program physicians' horne, office and voice beeper numberwere all includedin themedical plan in orderto beable to expeditegetting the most Divingfor Science...1990

immediatecontact. Dr. James Lowenhertz ofMercy Hospital's Hyperbaric Unit in Mami was contactedand agreedto serveas our backupphysician and referral center.

A decisionwas made by FSU in conjurictionwith the office of EnviromnentalHealth & Safety!to restrict the hyperbaric chamber usage to theneeds of thediving program only. TMsdecision was made in light of the recognition that the chamber facility was expressly designedto carefor medicalproblems that wouldarise as a resultof an accidentor incident atthe dive site. Neither the number ofrequired trained staff nor the equipment and expertise necessaryfor state of theart diagnosticand therapeutic management of all divingrelated incidentscould be supported bythe WMSARP orFSU. It wasfelt that other diving accident victimsshould be referred to oneof thehospital chainbers located in S.Florida FSUADP has taken advantage ofthe expertise generated by this program to enhance otherdiving research projects. %be use of nitrox diving is quite routine now with some of our relativelymedium 0-100 ft! depth diving exposures. The use of dopplermonitoring has becomean important safety measure in use with our prolonged exposure "decompression dives"that are being done to conduct research incave systems and other deeper depth projects 80-130ft.! Thedoppler may help us to identifythose divers who are more prone to intravascularbubbles and possible DCS. We will be able to offer treatment earlier and modify theirdive exposure tocompensate for their potential increased risks.

CONCLUSIONS Whereasit is impossible witha sampleofone or tworesearch divers to verify any set ofexposure tables asbeing safe with any degree ofstatistical significance, theADP has set up at%arm Mineral Springs a mechanism toprovide exceptional safety precautions fordeep researchdiving which may be used as an example ortemplate forother research projects of similarnature with modification beingmade based onthe particular needs ofthe host agency. Thishas the potential toopen up new frontiers for scientific investigation bydiving scientists withthe cooperation oftheir respective institutions andDiving Control Boards. Kepper:Medical Support for M~> gasDiving at WAS

APPENDIX I

WARMMINERAL SPRINGS ARCHAEOLOGICAL RESEARCH PROJECT

MEDICAL MANAGEMENT PLAN

Reviewed version 02- l2-90 ln Effect Until Further Notice G. Stanton

B. O'Horo Benton

W. Kepper

R. Matthews

I. RESPONSIBILITIES TheWhrm Mineral Springs Archaeological Research Project isa Legislativelyfunded researchproject, administered through Florida State University, Department ofAnthropol- ogy.In accordance withFSU diving policy, allWMSARP personnel arerequired tomeet the University'sDiving Control board standards before they are permitted todive. These stand- ardsinvolve both an initial assessment ofwater skills and a determinationofmedical soundness throughanannual medical exam. The WMSARP isresponsible forpayment ofannual physicalsforits personnel. Allpersonnel medicais mustbe reviewed andapproved byDr. WilliamKepper prior to clearancefor diving. TheFSU Diving control Board and its agent, the University's Diving Officer UDO!, areresponsible forthe overall safety of the WMSARP. In hiscapacity asMedical Officer for the Florida State University Diving Control Board,Dr. Wigwam Kepper is also the Medical Officer for the Warm Mineral Springs ArchaeologicalResearch ProjecL Hehas the overall medical responsibility forthe project, includingdetermining medical qualifications ofdivers, medical management ofdiving acci- dents,and recommendation offollowup therapy. Dr. Kepper also has the responsibility of notifyingWMSARp ofthose times when he will not be available tobe on~i formedical emergencies.Inthose instances WMSARP willcontact Dr.Loewenherz inMiami during an emergency. Divingfor Science...1990

All personnelare responsiblefor maintainingtheir own generalhealth.

Diving safetyoperational responsibility rests with the on-siteDive Office. He makes final daily decisionson fltnessto dive anddiving operations,and inanagestreatments a>n- ductedin the WMSARP chamber. As an extensionof Dr. Kepper, he performslocal neurologicalexams and general physical assessments asneeded.

II DECOMPRESSION SICKNESS/EMBOLlSM TREATMENT PROCEDURES

A.%hen to TEmt

The primarydecision to treatsuspected cases of decompressionsickness and embolism restswith the DiveOfficer, ultimately to beconfirmed by the MedicalOfficer or hisdesignee. The Dive Officer will alsodecide if the diver'scondition warrants calling in the local rescue squadat this time, subjectto the reviewof the MedicalOfficer if the decisionis madenot to involvethe local rescue squad. The purpose of the rescuesquad is to assist,if necaaluy,in placing an i,v. tube.!

8. General Procedures

1.Specific procedures for chamberset-up and personnel duties are in placein another document, and are not addressedhere.

2. Forafter work hourstreatment, the diveris to pagethe Dive Officer, who will return the call. If the decisionto examineand possibly treat the diver is made,the Dive OfncerwlLI pagethe reinaining crew, leaving on the digital display 426-9550. This will betheir notification thatthey are required to reportto thedive locker immediately.

Pagenumbers are:

Bob Matthews 813! 952-4582 Jack Lotito 813! 952-4583 Joe Nelson 813! 9524370 3. Forroutine pain-only DCS a neurologicalexam will begiven to the diverbefore treatmentbegins. For serious cases it will be given as soon as practical after the diver is under .

4. Recornpressionon 100% oxygen to 60feet will beginas quickly as possible. 'Ae choiceof treatment table USN1hbles 6 or6A! will be made on a casebycase basis according toUSN treatment procedures. Because Lhble 5 isnot regarded byall as adequate, it shall only beused with permission ofthe Medical Officer, at the time. Additionally, special therapy as directedby the MedicalOfficer is permitted. Kepper:Medical Suppon'for Mixed Gas Diving ar WAS

5.After treatment isunderway, theDive OKicer or his designee will call the WMSARP administrativeoffice 26-9559! and leave instructions that the number isnot to be answered thereand to disconnect theanswering machine. WMSARP personnel will then connect a dive lockertelephone to that line on-site!,which will be used for incomingcalls during the treatment.The Dive Officer will then makethe followingnotifications:

a. Dr. William Kepper 904! 599-7183 Voice Beeper! 904! 877-5143Office 904! 877-3261Office 904! 385-9336Horne ~ HIS ALTERNATE,WHEN DR. KEPPER WILL NOT BE AVAILABLE: regulartelephone call!

Dr. Loewenherz 904! 274-4880

b. WMSARPCoordinator 904! 298-2920 Beeper! If a page beeper! isused, the Dive Officer will leave the number 813! 426-9559 for theso notified parties to returnhis call. Coordinatorwill contactUDO during office hours at 904!644-1439. If UDOis not availablea message will be left. Afterhours call will be made at 904!926-3389 for Gregg S tanton,or 904! 878-3357 toDan Orr, and he will then have the responsibility ofcontacting the Dive Officer. WMSARPCoordinator shall notify the Chairman,Department of Anthropology,andany other University contacts required ofher. The UDO shall make any Universitycontacts required of him. ShouMany media contact press! become involved, the WMSARP Coordinator isthe onlyindividual authorized tomake statements orto manage that situation in anyway. 6. All medicaldecisions during treatment, after contact has been made, are the responsibilityofthe Medical Officer. In those cases where Dr. Kepper is notavailable, the alternateMedical Officer will direct medical treatment. The Medical Officer may, at any time requireadditional on-site medical aid i.e.,rescue squad, local physician!. TheMedical Officer can, at any time require the use of fax to provide him with copies of treatmentunderway, , etc. 7.The Medical Officer has several individuals who are available to offeradvice. See Attachment B. e useof extratreatment time/oxygen modification 's authorized at thediscretion of the Medical Officer and Dive Officer. C. 1hmsportationof'Diver toa MedicalFacility

At the discretionof the MedicalOfficer and the Dive Officer, the diver~ transferredtoa fulltreatment facility. Transportation byambulance or helicopterispr f anda 100%oxygen cylinder and demand mask, as well as any other medical sup h berequired for theparticular incident, will accompany the diver. If at all ~~,bl WMSARperson will accompany thediver. Transfer will be tnade in order of prefere ! to MercyHospital in Miami,or to ShandsHospital in Gainesville, prior to the t ~ be notifiedand anydocuinents they require will be sent hand~ed, atthe discretion of thereceiving facility. ~e diver will not be pr~~d ~ the~emce orPort Charlotte emergency rooms unless directed by the Medical dicer ~ + the localemergency inedical service.

D. Return to%ork

A diverwho has received chamber treatment shall be clearedby the MedicalOfficer andthe Dive Officer before returning to work. 'Ae MedicalOfficer mayrequire emninatioti bya physicianbefore authorizing a returnto work. Additionally,depending upon the circumstances surrounding the injury, diving may be temporarilysuspended for a reevaluationof the divingprocedures, tables, etc. In such instances,anyreevaluation mustbe made on a prioritybasis and as quickly as can be safely donein order to authorize a return towork in a timelymanner.

E. Documentationot'1hmtinent

All recordincording documents docum anddocuinent subtnission requirements arethe responsihih5 of the siteDive Officerand must comply with the requirements setforth in itemV

III. PRECAUTIONARYTREATMENTS A treatmentis' "p" recautio cautionary"whenthe diver has no symptoms ofdecompr sickness,but for reasons of sususpected e table irregularities orany feelings on the part «the ~" ' or supervisora treatment isfelt to be thenthe standard treatment shallto be desirable.used. If there are any signs or symptom 'any~

A. Wbeuto Iiuplement

Inthe event a diverorthe Dive 'veOffi cerdetermines thata divermay»t be " ' n o owingadive,precaution, p ' narytreatment maybe implemented atthe d~ Kepper: Medical Support for MixedGas Diving at WMS eitherthe Dive Officer or the diverin question.If thereare any signs or symptomsthe standard treatmentshall be used. SeeAttachment D for proceduresto be followed when doppler monitoringindicates a causefor concern grade4 bubbles!.

B. Choice of % eatment

Followinga neurologicalexam, precautionary treatment shall consist of either breath- ing100'fo oxygen on the surfacefor a minimumof 30minutes or more,or chambertreatment of a rninimuinof USN Table5, as determinedby the Dive Officer. In the caseof chamber treatinent,follow procedures in itemII aboveand in AttachmentA. Eithertreatment will be precededand followed by serialneurological exains.

C. Notifications Required

If diver is to be treated on surfaceoxygen for 30 minutesor more only and passesa repeatedneurological exam with no symptoms or signs,no notification is requiredto either the Medical Officer or the ADP,except for the standardend of weekreporting which shall reQectthe detailsof the treatmentused. If there is a specificreason for the precautionary treatment such as uncertainty about the table! this shouldbe logged.Notification of the WMSARP Coordinator is required.

If the diver is treatedin the chamber,treatment notification requirements are in effect, asoutlined in itein II andAttachment A, with the explicitinformation that it is a precautionary treatment only.

D. Return to Work

If all neurologicalexams are performed as required and no signsor symptomsdevelop thenthe diver may return iirunediately to workfollowing a precautionarytreatment.

E. Investigation Eachprecautionary treatment shall be investigated bythe Decompression Monitoring Board.

6! TREATMKNT OP NON-WMSARP PERSONNEL Universitypolicy in thismatter shall be strictly adhered to, under any circumstances. At ilo timeare exceptions to be made. University policy is stated in AttaclunentC

V. DOCUMENTATION Divingfor Science...1990

The Dive Of6cer shaHreport routine deep, mixed gas, and any relevant air dive operationsto the ADP on a weeklybasis. A weeklycompilation of reports will occuron the Fridayof everyweek in which suchoperations take place,and theseshall be mailed no later than5:00p.m. on those Fridays, These routine reports will consistof copiesof dailydive logs, copiesof anyDoppler tapes run, a DCBdiving siunnuuy sheet, and a writtennarrative of the week'sdiving activities, including the DiveOf6cer's observations. During monthlyperiods whenshallow and air dives only are made, reports will bemade monthly in accordancewith establishedDCB regulation. A copyof aH these reports shaH be sentat thesame time to the WMSARP Coordinator.

In thosecases when operations are not routine e.g., injury, recompression! the Dive Officer shaHcomplete aH on-site documentation and forward to the ADP within 24 hoursof suchan occurrence. At thediscretion of theADP, the documentation may be forwarded by fax.Copies of thesame wiH be forwarded to the WMSARPCoordinator in like mannerat the saine time.

Tosummarize policy, if a treatmenthas been used and the diver is regardedas "cured" thenno accident report is needed. If thetreatment does not resolve aH symptoms then an accidentreport shallbe filed.

VI. WORKMEN'SCOMPENSATION AND EMERGENCY OVERTIME Universityregulations shaH be strictly adhered toin reporting accident orinjiUy. Any treatmentrequired from outside the University or itsemployees will bepaid for through Workmen'sCompensation. Universityregulation Series P,Section I, Subject 3.1, regarding OPS employee over- time work wiH be followed. Kepper:Medical Support for MixedGas Diving at WAS

ATTACHMENT A

EMERGENCY TREATMENT PROCEDURES OUTLINE

THIS IS A SIMPLE OUTLINE OF VERY BASIC EMERGENCY PROCEDURES. REFER TO PAGES2 THROUGH 5 OF THE WMSARPDIVING MEDICAL MANAGE- MENT PLAN.

I. Dive0%cer makesdecision to treat in chamber,gives neurological exam, may call in rescue squad. A. If decisionis madeafter working hours, Dive Officer is to contactcrew by page.

Jack Lotito 813! 9524583 Joe Nelson 813! 9524370

B. Recompressto 60 feet on oxygenas quickly as possible.

1.Treatment on USN Tables6 or 6A are possible-DiveOfficer determines which one.

2. Dive Officermay elect to useextra oxygen.

C. DiveOfficer/designee makes notifications after treatment underway.

1. WMSARP Administrativeoffice notifiednot to answerphone, disconnect answeringmachine 2.A divelocker phone is to beconnected to 9559outlet for iricoming calls to Dive Officer 3.Dive OKcer/designee inakeoutgoing calls on 9550, make notification to the following.'

a. Dr, Kepper 904! 599-7183 Voice Page! 904! 877-5143Office 904! 877-3261Office 904! 385-9336Horne Divingfor Science...1990

If usingpage, leave 813! 426-9559 and brief message

QE ALTERNATE MEDICAL OFHCER Dr.Loewenherz 05! 274-4800 regular telephone call!

b.Barbara Benton 904! 298-2920 Page! Leave 813426-9559 4.Barbara Benton immediately notifies University Diving Officer at 904! 298-3427. D.Medical Officer returns call,assumes responsibihty formedical management of treatment.May require fax, outside medical help, etc. E.Transport to Medical Facility at direction of MedicalOfficer. 1.Ambulance orhelicopter preferred, need 100% 02 and mask, any other re- quiredmedical supplies. 2.Avoid Venice orPort Charlotte ER's unless ordered byMedical Officer. 3.'&ansport toeither Mercy Hospital, Miami, orShands, Gainesville Mercy preferred! 4.Dive profile and chamber treatment documents tobe either faxed or hand- carried-askhospital 5.If atall possible accompany theinjured diver to the hospital. F.Medical Officer and Dive Officer must approve return to work. 1.Medical Officer may require examination byphysician. 2.Diving may be temporarily suspended if conditions warrant. G.Dive Of5cer responsible fordocumentation-see ItemsV and VI inNarrative for details. Kepper:Medical Support for MixedGas Diving at 89fS

ATTACHMENT B

PHONE NUMBERS FOR WMS CALL SHKKT

DAN 916! 684-8111 Ask for Dr. Richard Moon Duke University

USNEDU 904! 234-4361 Ask for Dr. Schwartz H. PanamaCity 904! 234-1841or Chief Medical Officer

USAFHyperbaric Unit 12! 536-3281 USAFSchool of AerospaceMedicine, Brooks AFB, SanAntionio

Decompressionsickness. Dr. Eric Kindwall 14! 259-2060 h. 14! 781-5453 FroedertMemorial Lutheran Hospital, Milwaukee.

Decompressionsickness, . Dr. AlfredA. Bove 15! 221-3346, op.221-2800 h. Departmentof Cardiology,Temple University Health Science Center

Decompressionsickness, Navy and recreational Dr. Tom S. Neuman 19! 294-6643 h. 19! 755-0795 Universityof CaliforniaSan Diego Medical Center

HBOffrauma Dr. Roy A. M. Moyers 01! 528-6152 or 328-! h. 01! 721-1429 MarylandInstitute for EmergencyMedical Services Systems

Decompressionsickness and everythingelse vast experience! Dr. ChristianJ. Lambertsen 15! 898-8692h. 15! Institute for EnvironmentalMedicine, University of Pennsylvania PRELIMINARY REPORT: EFFECTS OF HURRICANE HUGO ON THE BENTHIC CORAL COMMUNITY OF SALT RIVER SUBMARINE CANYON, ST. CROIX, U. S. VIRGIN ISLANDS

Charlotte Ann Xesling P. O. Box 522 WrightsvilleBeach, NORTH CAROLINA 28480 U. S. A.

In February1989, the National Oceanicand AtmosphericAdministration NOAA!and the NationalUndersea Research Center of FairleighDickinson University NURCIFDU! accepted a proposalfrom the U.S.Virgin Islands Governments Department of Planningand Natural Resources and the University of theVirgin Islands to establisha long term environmental ~oring projectto assessthe changes in the benthiccoral reef community in SaltRiver Submarine Canyon,St. Croix, US. Vien Islands.1' AquariusUndersea Habitat and saturationdiving techniques were utilized at thecommencement of this project. Saturationdiving allowed the project participants to maramizebottom time for carefulsite selection, permanently mark study sites, photographically document eachquadrat, and collect data toestablish a baselineforfuture monitoring periods in the Salt River Submarine Canyon area.

OnSeptember 17th and 18th, 1989, Hurricane Hugo, with sustained winds of 140 nulesper hour andgusts over 200 mits per hour,hit St. Croixinflicting maj or darriageto theterrestrial portion and causing significant changes to thesubmerged landssurrounding the island. Duringthe months of Novemberand December of 1989,the permanent quadrats were relocated and photographed, providing data for aninitial compan'sonbetween the pre andpost Hurricane Hugo state of the benthiccoral reefcommunity in SaltRiver Canyon. Thispaper will a%i'ressthe preliminarydata compiledfrom companngthe photographstaken at thestart of theproj ect and again at thefirst sixth ! month monitoringintervaL

INTRODUCTION

With the recent increasein coastaldevelopment throughout the Caribbeanand the world,scientists, resource managers and government oAicials realize the needto establish tnonitoringprograms to recordbaseline data for evaluatingchanges in coastaland marine Divingfor Science...1990 resources.Baseliae data collectionand recordingcan help in assessingchanges that are occt~ng in near-shoremarine communities aad whether the changes are the resultof natural processesor are a direct result of man's intervention.

Ia February1989, the NationalOceanic aad Atmospheric Administration NOAA! and the National UnderseaResearch Center of FairleighDickinson Uruversity NURC-FDU! acceptedand funded a proposalfrom the U.S.Virgin IslandsGovernment's Department of Planningand Natural Resources aad the Universityof theVirgin Islandsto establisha long term environmentalmonitoring project to assessthe changesoccurring in the benthic coral reefcoramunity in SaltRiver SubmarineCanyon, St. Croix,U.S. Virgin Islands.

'Itusproject required two saturation missions 89-3 and 894C! andinvolved using tbe AquariusUndersea Habitat. The Habitat program is sponsored by NOAA'sNational Undersea ResearchProgram and is operatedby NURC-FDU in St.Croix. By utilizing saturation diviag techniques,it waspossible for the projectparticipants to maximizedaily excursionbottom timesfrom the underseahabitat for carefulselection aad permanentmarking of studysites, photographicallydocumenting each permanent quadrat, and to collect data used in estab- lishinga baselinefor futuremonitoring periods in theSalt River Submarine Canyon area.

SaltRiver Canyon provides a uaiquestudy area. The characteristics of the eastslope aadwest wall are dramatically different. The western wall is steep, often vertical, and has many spurand groove formations which sand is transportedto the canyonfloor. In severalinstances, overhaagsand caves are present. The Grst significant groove formation occurs at a pointwhere thewall meets the canyon floor at a depthof 60 ft 0 m!.This area is tbe begianing of station 1.In deeperportions of thecanyon, {90-120 ft 0-40 m!],large portions of thewall have broken off andhave become part of thecanyon M.

Tbeeastern wall, in contrast, ischaracterized byalternating zones of near-verticalrock wall andcobble-fdled side tributaries, at anglesof 15-20degrees. The innermost 65 %00ft 00-250 m! is of tbe latter type.Further seaward the wall becomesvertical. Thecanyon Qoor has a gentlyseaward slope comprised of medium sand to silt Mz25 mm!.The floor is generally iaactive except for the periodic sorting of burrowingorganisms, but canbecome mobile during periods of highwave or currentactivity. Thelip of thecaayon begias at thebarrier reef fronting tbe Salt River estuary. The depthis between 30 aad 50 ft 9-15m! andcontinues downward to a depthof 12,000ft 500 m!where it joinswith the Christiansted Caayon. At thelip of thecaayon there are scattered Acroporapalmata stands and head corals primarily Diploria spp.! Milkpora spp. are also presentin thisarea The canyon walls are dominated by flattened Agaricia spp., Monkzstrea annakuisaad other corals which are tolerant to lower light levels. Gorgonians and sponges areextremely common. The canyon floor has isolated sea grass HalopMa decipiens! and rhizophyticalgae that canbe foundto depthsof 100ft. Kesling:Species ofHurncane Hugoin St. Croir, USVI

METHODS Eightpermanently inarked inonitoring stations were established throughout the Salt RiverCanyon, one at 30 ft 0 m!,four at 60 ft 0 m!,two at 90 ft 0 m!,and 120ft 0 m !, asshown in Figure l. At eachstation, except station number eight, there were six, tenmeter long transects placedalong a depthcontour. At stationeight, there were ten, ten meter long transects placed alonga depthcontour. Two brass stakes marked the ends of each transect. Holes were drilled into the substratewith an underwaterhydraulic drill. Stakeswere placed in eachhole and cementedinto place with underwater epoxy. A numberwas stamped into the top of each stake foridentification purposes. Along each transect, benthic cover was assessed and quantified by usingthe chain line method Rogers et aL,983!. Thistype of measurementgives a three dimensionalview of the coralreef. It involvesplacing a smallchain along the transectwhich is usedas a scalefor the ineasurementof the percentof benthiccover along each line. Fifteen0.5 in 2 quadratswere sampled at the30, 60, and 90 ft stations.At the 120ft site only12quadratswere established, due to timeconstraints. Twocorners of thequadrat weremarked by using four inch cut nails pounded into the substrate. Each nail has a numbered tagattached toit witha plasticcable tie. These tags were placed in theupper right and left handcorners to insureexact photographic replication. An aluminumquadrapod with a quadratsize of approximately05 m2 wasused in a frame with a NIKONOSIII underwatercamera and 15 mm lens with two strobessecurely mountedto thisframe assembly Suchanek et al. 1983!.The quadrapod was positioned by placingthe permanent numbered tags in theupper left andupper right corners of the quadrapodto ensureexact replication. Thismonitoring project also included 8mm video recording of eachtransect line for lateranalysis, water quality testing, queen conch Strombux gigcrx! monitoring, and Acropora cervicornisgrowth measurements. This data can be found by writing the National Undersea ResearchCenter and referring to Aquariusmissions 89-3 and 898-4C. This paper will only focuson a preliminaryanalysis of the quadratmeasurements. OnSeptember 17th and 18th, 1989, Hurricane Hugo, with sustained winds of 140miles perhour and gusts over 200 iniles per hour,hit St.Croix inflicting severe damage to the terrestrialportion of the islandand causing significant changes to subtidalareas. Hugo was a classicalCape Verde hurricane that left a trailof destructionacross the Leeward Islands, U.S. VirginIslands, Puerto Rico, North Carolina and South Carolina Figure 2 !. Theeye of the Hurricane made landfall on the eastend of St.Croix at approximately0230hours on the 18thof Septemberand exited the West end at approximately0400 hours. Minimum surface pressureswere approximately940 mb nearthe centerof the storm.

In its path,it left a trail of destructionestimated to be 2 billiondollars for the U.S. VirginIslands and Puerto Rico. The Federal Emergency Management Agency estimate of Divingfor Science...M6 moneyoutlay is currently 0.731billion for the U.S. Virgin Islandsand Puerto Rico and is subjectto upwardrevision.

BetweenMay andJune of 1989,the first photographsof the quadratswere taken to establisha baselinefor the longterm monitoringproject. wasused initiaIIy to carryout this task BetweenNovember and Decemberof 1989,a resurveyof the quadrats wasaccomplished. A direct comparisoncan be madeof the effectsof the storm on the in SaltRiver Canyonbased on pre andpost Hurricane Hugo sampling.

The 35mmcolor photographicslides were analyzedusing the random point method Eachslide was projected onto a posterboard with a gridbackground scaled for a oneto one reproductionsize. This grid wascoinposed of 231evenly spaced points. The entire framewas analyzedby countinghow many points each material component encompassed within the grid boundaries.

RESULTS

Livingsubstrate.The pre- and post- hurricane substrate counts are illustrated in Figures 4, 5 and6. The significanceof eachchange was tested using the chi-squaretest and a levelof significanceof 0.05.When looking at thesignificance of eachsubstrate as a whole,the amount of significantchange was dramatic, however when viewedon an individual scale,the com- parisonbetween the differentstations was not as drainatic Table 1!. Diploria clivosahad a levelof significantchange overall but only station 4 wasshown to besignificant Table 1!.

The proportionalcoverage of the coralswas determined by dividing the total numbers of pointscounted for thatcoral by the total numberof all coralpoints in all the different stations.Bichocoenia stokesi was most affected by thehurricane, its coverage was reduced by 81%.The othercoral species common on thereef, Diploria clivosa, Colpophyllia natans, Porites poIites,and Porites astreoides were all reducedin coverageby values ranging f'rom 28% to 14% Tible 2!. Overall,the changesto the coralcoverage were minor, however,certain stations receivedmore damage and alteration than others SeeTable 3 for pre- andpost-humcane substratecounts!. Station3, locatedon the 60 ft outerEast Slope, station 4, locatedon the 60ft innerEast Slope, and station 6, located on the 90 ft EastSlope, showed the most significant change.It wasnot possibleto re-surveythe 120ft station,located on the EastSlope, due to the depthand time constraints.A visualobservation was made by a NURC-FDUstaK memberwho reported that on the afternoon ofthe 17thof September, the entire shallow ridge of theEast Slope area of thecanyon had breaking water. One possible reason why the Bast Slopestations suffered the mostdamage was that theEast Slope took the directhit of the waves,thus somewhat reducing the severityof the wavesfor the WestWalL

The storm trackedfrom the SE to the NW acrossSt. Croix. Directional shifts were recordedon a $4current meter deployed in Salt River Canyon at a depthof 60ft0 m!. The Kesling:Egects of Hum'caneHugo in St.Cr., USVI

'Ihble1. Slgnitkanceofchanges la subslrnte per station by chi~ tesLx ~ algnl5eant, o ~ not significant

Station number

2 3 4 5 6 8

Aguricia spp. 0 Coicophygia natans Dichocoenia straltans X X Dichococnia stokcsi Diplona clivosa Diptort'alabprinthiforrncs Diploria strigal Nontastrra anrutluns Nontastnra cavernosa Nycctophytliafcnu %rites asteroidcs 00 00 Poritcsporitcs Sidcrnstrea ~ 00 00 Sidemerea sidaraa Sponge spp. X X Gorgonianspp. Rubble X X Sand Bare rock DCA

DCA - Dead Coveredwith Algae

Table2. Percentageloss of coralapeciea

Pre-hurricane counts

Agancia spp. 3312 10 Coljophytlia natans 249 189 24 Dichocosnia stokcsi 26 5 81 Dipioriu clivosa 109 28 Diplona strrgosa 644 584 9 2 Nontastrea artnttlans 979 956 Nontastrea cavsmosa 945 &41 11 NycctopltyNafsmr 128 113 12 Writes astreoidss 279 14 porites porttss 111 95 145 Sider gstrsa sidcrea 401 380 Stsprrartocoeniamichciinii 31 27 13

243 Divingfor Science...1RN

lbbte 3 Pre and Fest Hurdausc Sabstratc Counts

Station 4

Agareiaspp. 599/567 158/93 149/124 1076/893 310/300 Colpcphylliaaataas 39f34 162/149 34/- Dcadragyfaeylbatricus Dicboxeoia strclbttis 5n Diebococnia stakei Dipltlia c&eaas 39/5 Diploria iablmnthiformcs 5/5 64/56 15/17 Diploria strigosa 45f28 225/186 123/118 38/41 Eutsasiliafastigiata 1/- McamMm mamdritcs 1/2 Mimcporaakixeuis 9/11 3/- 7/I Millepora complaaata 4/4 Montastrea anaskms 196I'196 174/185 150/135 112/105 11/14 293/275 4' Moatastrea cavemasa 122/116 25/14 178/149 250/242 9/11 217/181 144m' Myeetopbylbaferox 14/16 21/13 37/31 17/17 19/12 20J2t Foritcs astreaides 25/23 14/10 82/57 72/78 5/2 64/57 17/0 Poritcsparitcs 16/13 48/43 8/2 20/23 6/4 LVM Sider astfca racism 19/14 30/32 21/22 8/8 258/264 19/20 LVJ5 Siderastrea siderea 121/118 21/27 62/51 94/91 23/23 39/28 41/42 Stephmxoeaia taidtelinii 31/27 %bastrea aafca 3/- Cria aids 2/- Spcegespp. 22f25 195/156 224/136 128/112 217/222 478/472 Goqlctuaaspp. 15V88 56/40 169/95 4Q/321 3/5 250/194 lbkbic -/17 -/396 -/117 -/7 -/56 Sand 181/161 95/34 218/278 6f20 63/134 52/96 Bare rock -/119 Dead Rock 18'i2/1876 1629/1549 1558/1643 Cmercd«fitb Algae 1671/1722

meterwas located near station 1 onthe inner West%hll See Figure 1 forlocation!. After the storm,the meter was found to havemoved laterally 10 ft m! towardthe West%4H ofthe canyon. There was newly exposed substrate along the wall indicating that sand traxtspctrt andscouring had taken place. Me currentineter had dropped in levelabout 3 ft m! indicatinghow much the floor of the canyon was scoured. Further along the canyon to the north,the depth of scouring hadreached twelve 2! ft m!ghylor and &agester, 1990!. Thesponges and gorgonians were affected significantly bythe storm. Sponges decreasedby13 percent and the gorgonians by28 percent. Stations 2 WestWall, 60 ft,3 East Slope,60ft!, 4 EastSlope 60ft!, 6 EastSlope, 90ft!, and 8 WestWall, 30 ft! hadhIgher losses.As discussed earlier, stations 3,4 and6 areon the East Slope and that waves stere Kesling:Egects of HurricaneHeroin St.Croir, USVI breakingoa Ole EastSlope. Station8 is locatedon the WestWalt in 30 ft depth, therefore,it js conceivablethat the waveenergy is greaterthan at a deeperdepth.

Non-livingsubstrates. The amountof rubble, sand,and dead rock covered with algae wasgreater in theresurvey. The classification of rubblefor thispaper is anydead gorgoniaas, spongesor coralthat was placed there after the storm. Sand increased from 615 counts before the storm to 734 counts after the storm. Rubble increasedfrom zero before the storm to 593 countsafter for the greatestsignificant change. Dead rockcovered with algae DCA! increased onlyslightly from 1852to 1876counts Table 3!. Thesegains were at thedirect expense of theliving components of thereef.'Ae iacrease ia sandcoverage may have been directly related to the fact that somuch sand transport occurred ia the canyonitself andthus sandsettled into thelow lying areas. In someareas the canyon floor dropped a maxiinumof twelve2! ft m!. Thechanges ia livingand non-living substrates types are significant on anoverall scale but when viewedstation by station,the changesare significantin only certainareas, predominantlyonthe East Slope. Figures 7 aad8 arephotographs takea before and after the stormof the samequadrat. These pictures are froin station3, 60ft EastSlope. Figure 7 was takenJune, 1989, notice the large tubular sponge. Figure 8 wastaken Noveinber, 1989. The largetubular sponge was missing and there was an increasein gorgoniaarubble and sand. Figures9 aad10 are pictures again taken from station 3. Figure9 wastaken June, 1989 and Figure10 taken November, 1989. When comparing the two, notice the gorgoniaa rubble and coralrubble present and pieces of Acropora cervicornis found in thebottom right of Figure 10. Figures11 and 12 are taken from station 4, 60 ft innerEast Slope. Figure 11 was taken in June, 1989,and Figure 12 taken November, 1989. Ia Figure12 the large ColpophyNa natans was completelymissing from the site and that there was a significantincrease iacoral and gorgoniaa rubble. Thesepictures are the moredramatic photographs taken. Others show very little chaagebetween monitoring intervals, such as an increase in sand settlement on top of the plating corals.

CONCLUSIGNS Thisis a preliminaryreport prepared todescribe techaiques used ia establishinga Long-Termmonitoring program using reproducible photographic documentation aadinitial analysisbycomparison ofthe Salt River Submarine Canyoa, inboth a Preand Post Hurricane Hugostate. Follow-up photographs havealso been taken for the one year sampling interval andfinal analysis isplanned using computer assisted techniques. Divingfor Science...1990

ACKNOWLEDGEM ENTS

Manythanks to SusanRhoades and Bill Clevelandfor their divingassistance. A big thanksto all thestaff of theAquarius for theirkelp and assistance in almost everyaspect of thisproject. Also to DougKesling for his moral, aswell as technicalarid logisticalsupport that he gaveme throughoutthe entire project - ThankYou.

LITERATURE CITEB

Manualfor VisitingScientists. 1989. In: NURC-FOUOperations Manual. Compiledby K. Berey. NationalUndersea Research Center of Fairleigh Dickinson University. St. Croix, U.S.VI. Pp. 1-67.

1hylor,Gand 1hgester,J. H. 1990.Directional Wave and Current MeasurementsDuring Hurricane Hugo. Proceedingsof Marine Instruinentation '90. San Diego, CA. November 1990 Kesling:Eff ebs of HurricaneHugo irt St.Croix, USVI

~t,>tlan, « I i Oft. '," est 0,'sll Inner ¹ ' Oft.'A'.-.t 'i all - Outer i Oft Eo-.,tSlope - Outer «4 &Aft Ea.;t 51epe Inner «'5 90ft WYe~tVeal Outer

'fetor

Figurel. Locationof permanentlymarked stations 1-tL SaltRivet Snbmarine Canyon, St. Croix, US.V.I.

247 Divingfor Science...1990

QBIKI1> 'f hlIII ANElk Ci G IA4T

1"

I!

/

L.

Figure2. The pathof HurricaneHugo. Map suppliedby the UA. Department of Commerce 51OAA - Rational Weather Service.

Figure 3. Photo credit D. Kesiing. Kesling:Effects of HumcaneHugo in Sf.Croix, USVI

Iv43IItI 3 r 'VV r35'03 I

rjI3~f Pg434343 ,49!I 50

i I

I 50 1j

1 ' II I 00

I 5 3 4 5 4 ~ I 3

400rlI0'N0 > l00 55I 3 '3500

I a

4J8

' 00I

0 'I l 5

II%I

~ 0543-I003~ I 3

Si03 I03I 500m503

Figure4 Hlstogramsofthe substratcs bycounts pcr stations before and after Hurrkaae Hugo. 1bc numerator ofthe fraction oa each histogram tsthe total pre.hurrlcaae counts; tbc deuomlaator lstbe total post+urricaoc countsfor tbe substrate. An asterisk g tnarksa statlstkallysigni6cant dl6crcncc x test, slgnlilcance kvel of 0.043!bct55ccu pre- and post-hurrkaneCOunts i'or each subatrate.

249 Divingfor Science...l990

2 E t 5 2 5 4 5 5 0

4 4

1 3 5 I 2 S E 1

Sldl10llfllNllbsr

Figare 5. Mstqpams of the sabstratesby countsper stations beforeand after Hurricane Hugo. Vise numeratsr of thekactioa oa each histogram ls thetotal pre-hurricanecounts; tbe denomiaator is thetotal post-hurslcam countst'or the substrate. An Asterislt j marts a statistica!iysignificant difference x test,silnkflcnnce tery of OA5! betweenpre- aad post4arricmm countsfor each substrate. Kesling:Effects of Hum'caneHugo in St.Croix, USVI

fjhrRDhl~II ~ I RID I3 9 I '

D II I '3 4 D

%DD P

RDDDID 94DD~ 0/593 4'I5 I754

I DD I DD

I 7 I

PIh-IhII I 144hh g DDDI~ IIIII iI4444

h IDIIDD IDII'ADD Figure6.Histograms ofthe substrates bycounts perstations before aad after Hurricane Hugo. 'Ihe numerator ofthe fraction oneach histoyum isthe tota} pre-hurricane counts;the denominator isthe tota} post-hurricane countsforthe substrate, Auasterisk ~!marks a statisticaBy significant difference xtest, significance ievel ofOAS! bet3reen preand post-burricane counts foreach substrate. 251 Divingfor Science...1990 Kesling:Effects of Hurricane Hugo in St.Croix, USVI

Figure 9.

Figure 10.

253 Divingfor Science...1990

Figure 11.

Figure 12,

254 QUANTITATIVE TECHNIQUES FOR UNDERWATER VIDEO PHOTGGRAPHY

EdwardJ. NancyJr. JosephAyers Kenneth P. Sebens Jon D. 8 itman Marine Science Center NortheasternUniversity East Point,Nahant, MASSACHUSETIS 01908 U.S.A.

Recentvideo technology has provided the underwaterscientist with a valuable researchtool. Videocameras have become more compact, have improved resolu- tion, arid work under verylow light conditions. Researchapplicatiorts include; measuringwater flow nearsurfaces, sessile benthic and jish surveys,mapping and quadratphotography, monitoring predation, and time lapse studies of organism abundanceand activity.A techniquefor imageand motion analysis from video using microcomputerswN beintroduced.

INTRODUCTION

The useof underwatervideo hasgreatly expanded the abilitiesof the divingscientist to collect data. Traditionallythe bestmeans of recordingevents underwater was by usingstill or complexand bulky movie photography. A compactvideo cainera can be placedin a housing for useunderwater, with recordingcapability up to 120minutes. The newestcameras have high resolutionand work undervery low light conditions.Video cainerasalso can be mounted on ROVs and usedin areasnot accessibleto divers. Manyquantitative research techniques employingstill photographycan be adaptedfor videosystetns. Some advantages of videoover still or moviephotography are; instantresults, longer recordingtimes, continuous recording, and excellent stop-frame resolution.

EQUIPMENT USED

The followingis a listingof the equipmentused by the authors.Appendix 1 listssome of the underwatervideo equipmentmanufacturers also listed in the 1990DEMA Directory!, The camerasused are Sony8 mmformat videocameras. The CCD-m8has a fixedfocus and comeswith a separateplayback deck. The CCD-V9and V99 Hi8 format! havevariable focus Divingfor Science...1 .6

with zoomand macro.'Re V9 featuresa 6xpower zoom lens with f = 12-72mm, F 1.6 with macrowhile the V99features an 8X powerzoom lens with f = 11-88mm, F 1.4with macro. Bothcameras feature on-screentiine/date/counter functions. They both havebuilt in play backdecks which can produce excellent still framesand forward by singleframes. The V99 useshigh resolution Hig! tapeswhich have over 400 lines of resolutionand a minimumlight reqturementof 4 lux.The Sony Hi8 recorderdeck EV-S900!allows single frame advance forwardand backwardand features digital audiocapabilities. The video housingsused are manufacturedby Ikelite, Hypertech,Aqua Vision and Quest along with underwatervideo lightsby Ikelite,Hypertech, and Subatec.

For scaling,two Metrologiclaser lights Model ML811! in a customhousing can be mountedon the video system laser dots at 10mm apart! Caimi andTusting 1986!. Alternate waysof produciaga scaleare by suspending a plasticruler in the field of view, or by follpwiag a transectline or a seriesof quadrats.

Adjustablecamera stands can be madeby cementingtwo piecesof metal or pVC pipp into a concretebase. A slidingtray to hold thecainera can be madeof aluminumangle stock andheld onto the pipe by split ring pipe hangers Witman and Sebens, 1990!. If heightoff the bottomis notcritical, the video camera can be secured to a cinderblock. For longterm time lapsestudies, a cameraholder consisting of a heavyconcrete base and custom mountiog bracketsis required.

TECHNIQUES

MAPPING AND QUADRATPHOTOGRAPHY

Videoenables easy mapping of fairly large areas. By having a scalein the5eld of view, theinvestigator cansweep anarea with the video camera. The area is instantaneously recorded onvideo and can be mapped onthe surface rather than actually drawing a tnapof anarea while underwater.The tape can be played back and still framed as often as needed to producean accuratemap, although problems ofparallax still exist. Using video thus replaces the need to createa photomosaic using still photography. Forquadrat photography, a quadrat canbe placed and recorded onvideo the quadrat sizewill depend onthe level ofresolution orsize and coinposition ofthe community tobe analyzed!.Thequadrat can then be moved across the area or along a transecttobe analyzed whilethe video records continuous images. Alternately, a video catnera can be mounted on a quadrapodandfocused torecord a fixedquadrat area see Witman, 1985 for still camera quadrapod!.Agreat advantage ofthis method versus 35mm photography isthat the number ofphotos isnot as limiting. 1%is allows formore quadrats tobe photographed, increasinII the samplesizegreatly fora givendive. Potential uses ofthis technique include; percent cover analysisofcommunity structure andcompetition ofencrusting organisms onhard substrates, Maneyet aL Techniriwerfor UnderwaterVideo Photography directcounts of organismsin randomlyplaced quadrats Dayton, 1971, 1975; Connell, 1961a, b!, andphotographs of organisrusin permanentquadrats Connell, 1972!.

Percentcover analysis can be performedby freezinga videoframe and digitizing Buss 1980!or byusing random dot patterns still photography:Connell, 1970, Dayton, 1971, Menge, 1976,Lubchenco and Menge,1978; video technique: Sebensand Johnson,1990!. Alterna- tively,a singlevideo frame can be grabbed using a computersystem and areas of distinctcolor can be quantiTiedusing a programsuch as 'Image'.

QUANTITATIVE SURVEYS

Underwatervideo cameras can be usedto conducttwo typesof surveysto estimatethe abundanceand distribution of animalsand plants living on the seafloor: 1! line transects Bvrnham et aL, 1980!and 2! random,haphazard or andsystematic sarnplirig using quadrats Pielou, 1974!. The primary advantagesof videotapingover still photographyare that large areascan be sampledquickly and the imagescan be viewedimmediately after the survey, eliminating the potential problem of losingdata in the event that filin is lost or ruined in processing. The only disadvantageof usingvideo vesusstill photographyfor quantitative surveysis that the resolutionof a photographtaken with fine grainedNm is stiII higherthan a video image, meaning that small organisms less than 1 cmsize! will be difficult to identify 2. in an area >0.25m '

The procedure of videotaping a line transect is extremely straightforward; a diver simply runs a transecttape acrossthe bottomor attachesit to a permanentposition marker in the caseof repeatedsampling of the samearea, and videotapesa swathof a certainwidth, usuallyfrom an aerialperspective to avoidforeshortening and image distortion. It isnecessary to keepthe camerato subjectdistance constant throughout the lengthof the transectto ensure that the samewidth of bottom alongthe transectis covered.This canbe achievedby anyone of the following four methods1! by swimmingat the samedistance above the bottom by achievingneutral ,2! by attachinga fixedrod to the front of the camerahousing, 3! by suspendinga plumbbob on.a line belowthe camera,4! or by usinga systemof laserdots that convergeat the correct distancefrom the bottom as developedby Harbor Branch OceanographicInstitution for focusing their 80 mm camera Caimi and 11istiug,1986!. Videotapedline transectsenabled Edmunds and Witman990! to surveylarge reef tracts rapidly to assessthe impact of HurricaneHugo on coraldistribution and abundance,In this application,line transects1.0 rn wide werevideotaped with a SonyV99 camera in a Hypertech housingheld aboveand perpendicularto the transect. The percent cover of live and dead coralswas estimated for the entire20 m longtransects by identifying corals under 200randomly placedcircles mmdiameter! on a transparentacetate sheet placed over a "freeze-framed" videoimages. Because the actualsurveyors transect tape only providesa frameof reference, it is not strictly neededas long as the camerato subjectdistance is maintainedby another method. For example,Witman and Sebens989! estimatedthe abundanceof macrobenthic invertebratesat 30 - 70 m depthsin the Gulf of Maine by videotapingthe bottom from the Johnson Sea-Link submersible. A line transectwas not neededbecause a scalewas provided Divingfor Science...1990 bytwo laser dots 10 cm apart, In a similarapproach using SCUBA divers insead of submer- sibles,Sebens and Johnson 990! surveyedcoral abundance along 10 50m depthgradients off thenorth coast of St.Croix with a SonyV9 camerain anIkelite underwater housing. They were able to swim at the samedistance off the bottom alongthe entire transect. Replicate imagesfrom 10 m depthcontours were randomly sampled with the frame counter on the video cassetterecorder and analyzed for coralpercent cover by therandom dot method. Quadratscan be videotapedas individual replicates for abundancesurveys or for monitoringchange over time e.g. fixed or permanent quadrats!. Excellent reviews ofquadrat samplingdesigns are given in Pielou974!, Green979! andKrebs 989! andwill notbe discussedf'urther here. Once the quadrat has been placed on the substratum by the designated samplingdesign, the diver swims over it tovideotape it. A bettermethod is to attacha quadrat frameor "quadrapod" Witman, 1985, plans in Coyerand Witman, 1990! to thefront of the videocamerato maintainexact camera to subjectdistance. Videotaping permanent quadrats hasgreat potential as a methodof detectingthe influences of biologicaland physical factors onbenthic community structure but it is notwidely used, probably because high resolution videocameras have only recentlybecome available.

MEASURING WATER FLOW NEAR SURFACE ! DyeRelease Methods: Dye release can be used to trackwater movement over short periodsof timein subtidalhabitats, either on a relativelylarge scale meters! using wide angle videophotography orvery close to surfaces millimeters! using video in macromode fluores- ceindye is commonlyused!. Video photography of dye streams has been used to trackfiow throughthe prey capturing apparatus ofa suspensionfeeding invertebrate in laboratory flumes Trageret aL 1990!.Similar techniques work well in fieldtrials, as long as the video camera in itshousing can be stabilized on a solidbase that allows the housing to moveback and forth to achievethe bestfocus and framing, One problemwith very close-upphotography is that the camerahousing could affect the flow aroundthe organismor surfaceunder study. Additional focal distancein macromode can be achievedby addingdiopter lenses e.g.P3 diopterfor 35mm cameras! in front of thevideo lens and focusing in macromode. This modificationworks well for particletracking also Sebensand Johnson, 1990!.

In addition to flow visualizationmethods using dye, dye releasecan alsobe usedto estimatemass transport and the dispersion of particlesin a watermass over time diffusivity! in subtidalhabitats. Release of a dyecloud e.g.from a syringewith tubingattached! at a fixed distancefrom the substrate,followed by videophotography of the dyecloud preferablyfrom directlyabove and from the side!,allows calculation of movementof the centerof mass, rotationand spread of thedye over at leastseveral minutes. Principal components analysis fromdigitized video images of thedye cloud is used for thesecalculations. This technique has beentested in rockysubtidal habitats, and computerizedmethods of analysishave been developed T. Powell, M. Denny,M. Koehl,pers. comm.!. Dye release methods of thistype areparticularly useful in studiesof howlarvae or garnetes might disperse as they are released or how other neutrallybuoyant particles behave in flow.

258 Maney et at Techniquesfor UnderwaterVufeo Photography

! PanicleTracking Methods: Flow nearsurfaces, such as around the feedingstructures of sessileinvertebrates, can be measureddirectly usingelectronic flow meters and other sophisticatedinstrumentation in either laboratoryor field situations.However, any video camera system with macro photographic capabilities also can be used to both visualize and measureflow non-intrusivelyon a very small scale. Particlessuch as naturally occurring non-motile! plankton, artificial neutrally bouyant spheres,or hydratedArternia brine shrimp! cysts,can be releasedupstream of the areaunder study. As theseparticles pass throughand aroundthe structuresof interest,a regioncan be photographedperpendicular to the inain axisof flow, with sufficientdistance between the port of the camerahousing and the objectof studysuch that flow is not affectedby the presenceof the housing.Such techniques havebeen used successfully in laboratoryflume studies of crinoid feeding Leonardet aL 1988! andin field studiesof particle captureby scleractiniancorals Sebensand Johnson, 1990!. In field situations, it is easiest to illuminate the area with a slit of light from an underwater video light. The light beamshould be only severalmillimeters wide whereit reachesthe surfaceto be photographed,so that it illuininates only those particles in the focal plane of the cainera,thosetravelling parallel to the major flow axis.If the light is inountedabove or below the subject,such that it doesnot obstructflow, the beaincan be orientedperpendicular to the camera,and thusparallel to the major flow axisand to the focal plane.

Particle movementsare analyzedby replaying the tape and using the stop-frame function. A clear acetate sheet can be taped to the video monitor, and colored markers used to mark the positionsof particlesin focus as the singleframes are displayedsequentially, resultingin a 'track'of particle locationsover 5-10 frames /30 secbetween frames!. Particle direction and speedin that plane two dimensional!are then calculatedusing an appropriate scale in the field of view and the known time between frames. If particles are moving more rapidlythan about 20 cm/sec, they appear as streaks on the stop-fraine image, and the length of each streak is the distance moved in 1/60 of a second. In field situations, flow is often bidirectional and turbulent. Severalminutes of recording,followed by regular or random samplingof flowthroughout the recording, are necessary to give an accurate characterization of fiow for that time period.

TINK IAPSE VIDEO

Timelapse video is especially useful for investigatinganimal behavior and for studying theeffects of short-tennphysical processes eg. currents,wave surge, storms! on benthic communities.We have been using two systems in our researchon the ecologyof offshore benthiccommunities in theGulf of Maine Witmanand Sebens, 1988!, 1! a systemdeveloped for the SonyV9 andV99 cameras by QuestMarine Video, and 2! a variablemtervolometer designedfor thesame video cameras by K. Sebensafter circuits developed by Mims 984!- Thetime lapse system is, of course, limited by battery life, power requirements of thecamera durationof thevideo tape, and the range of timesettings built into the intervolometer circuiL TheQuest Marine Video systems feature a moldedpolycarbonate housing model 4S0! withpower supplied by 10and 20 amp hour Nicad batteries. A Microntadigital timer V9! Divingfor Science...1990 or SonyRM-95 digital controHer has been modified to turnthe camera on and off up to 7 times dailywith the V9 andfour tiinesa daywith the V99. Wehave used these cameras as remote stationsfor bothshort and long term monitoring of fishpredation in coastaland offshore rocky subtidalhabitats of theGulf of Maine Witman and Sebens, 1990!. For long term e.g.months! deploymentof video cameras, it is important to considerthe potential of pressureto deform plexiglaselements of the camerahousing. A wiseprecaution against this is to ensure that latchesclosing the back of thehousing are equally spaced all aroundthe perimeter of theseal.

The variable intervalometer circuit can be constructed from a Radio Shack kit and modifiedto interruptthe powersupply of the camera.We have successfully used a circuit timedfor three secondrecordings at 1 minuteintervals with a SonyV9 camerato document rapidpredation on tethered invertebrate prey by cod andwolffish in the Gulf of Maine Witmanand Sebens 1990!.

MONITORING PREDATION

Video allowspredation to be monitoredwithout the disturbingpresence of divers. It alsoallows predation events to be recordedfor longerperiods of time than would be possible with SCUBAespecially at depth. Preycan be tetheredto two 1m lengthof chain and placed approximately15 m in frontof, the video camera Witman and Sebens,1990!. The camerais placedon a cainerastand!, turned on, andfocused on the chain with the tetheredprey. ~e camerais allowedto run for 90- 120minutes tapelength or batterytime!, then retrieved by divers.If a longerinterval is required, a second video camera could be placeddown to continue the recording or a video with a time lapse circuit could be used. Predator abundance and predationattacks can be countedfrom the tape. Caremust be takenwhen doingcounts to distinguishindividuals and to separatethem from return visits by the same individual.

IMAGE AND MOTION ANALYSIS FROM VIDEO USING MICROCOMPUTERS

At the Marine ScienceCenter in Nahant,we are implementing a BiologicalImage AnalysisCenter under support from the N.S.F.Instrumentation and InstrumentDevelopnMat Program.Fundamental to this projectis the implementationof systemsand softwarewhich allowthe cataloging of videotape data and quantitative analysis of videoimages using digital itnageanalysis techniques. To interfacevideo equipment to laboratory computerswe have utilized the VidClip systemdeveloped by AbbateSystems Norfolk, MA!. VidClip imple- mentsan interface between the Macintosh serial line interface and the Control-S and Con- trol-L editing controller interfacesof the video decks and camcorders. The Control-$ interfaceis unidirectionaland only sendscommands from the coinputer to the video device while the Control-L interfaceboth acceptsconunands as well as returns status and counter informationtothe computer, allowing precise interactive control of taperegistration. VidC4p is availableboth as a Hypercardinterface as well as a libraryof programmingobjects. Maneyet a1:Tecku6paw for UnderwaterVideo Photogrqpby

INDEXING AND CATALOGING VIDEO TAPE DATA

To keep track of video data we utilize the ClipKeeper Hypercardstack which maintainsa recordof the start and stop timesof segmentsof video tape.Basically, the stack maintainsa "card"for eachsegment of datawhich supports a commentfield aswell aschoices buttons! to control the deckand to controlsearches. To record a clip,the usercreates a new card,queues the point of intereston the tapeand clickson a mark buttonwhich acquiresthe counter index over the Control-L interface and enters it into a field on the card, then repeats this for the end of the segmentof interest. Oncethis procedure has been completed, the data segmentcan be repeatedlyplayed by clickingon a "play"button on the card. Oncea tapehas been indexed in this fashion a section of data can be retrieved by merely clicking on a button.

VIDEO TAPE MOTION ANALYSIS SYSTEM

Softwaresupport for our video imageanalysis systems is basedon extensionsto the NIH Imageimage analysis program developed by Wayne Rasband at NIMH andthe VidClip Control-L and Control-S serial line interface developedby Mark Abbate Abbate,Inc., Norton,MA!. Our video-basedmotion analysis systein consists of threecomponents ! A MacintoshII with DataTranslationframe grabberwhich supportsa generalpurpose color imageanalysis program ColorImage!.!. Hi8 video camcorders Sony V9, V99 andCCD- V5000!and Decks Sony ECV3, ECV900, SLV757! which are used for theacquisition of both video motion!and analog sensor! data and are controlled on a frameby framebasis by the MacintoshII for videomotion analysis. Our software platform, ColorIrnage, Think Pascal SourceCode!, developed by J. Ayers and G. Fletcher Fletcher Applied Sciences, Mason, NH!, supportsboth the standardineasurements used in NIH Imagesuch as area, mean density, centerof gravity,and angle of orientationof a userdefined region of interest,as well as segmentationof objects on the basis of colorand time-based measurements from sequential video frames. Measurementresults can be calibratedand exporteddirectly asspreadsheet files. At present,ColorImage supports the following capabilities froin video data.

! ColorImage Acquisition. We have developed color-based extensions toMHImage whichallow us to acquirecolor images from RGB video sources. For laboratory acquisition andrnacrophotography, weuse a RGBCamera Javelin JE3462RGB, 480 lines Horizontal Resolution!or a SonyProMavica electronic still video system which supports NTSC to RGB conversion.The various Sony video decks differ in theformat of stopframe images making acquisitionof stopped-frameimages somewhat inore difficult. The CCD-V500 generates a trueNTSC frame signal in stopframe while all the other decks display only the odd 6eld lines. Forthis reasonit is necessaryto interpolatethe evenfield lines. For Example,the Sony EVS-900Big deckdisplays only the odd field in stopframe mode, so we have implemented inColorImage! a two pass filter which first interpolates theeven field lines and then performs a medianfilter using3x3 convolutions to restorethe full frame.

261 Divingfor Scierrce...1990

Todigitize color digital images from video, we utilize either a RGBsource directly or convertthe NTSC camcorder! data to RGBusing the ProMavica or a TrueVisionVide box to performa NTSCto RGBconversion. By connecting the three inputs of theframe grabber to the red, greenand blue outputsof the RGB source,we sequentiallyacquire 8-bit monochromeimages by switching between the red, green and blue inputs to generateseparate red,green and blue plane files which are then analyzed internally in ColorIrnage.To generate colorfiles, we performa 24bit to 8 bit conversionthrough a mediancut procedurewhich establishesa color lookup table of the256 most common colors in theimage and then generate an 8-bit indexedcomposite color image.

! CoLorfinite Segmgnatioii.A set of 8 and24 bit segmentationalgorithms allows us to segmentobjects organisms,cells, etc! from the digitizedvideo imageson the basisof their color.8-bit segmentation operates on color look up table LUT! valuesand allows us to ! selectranges of gray-scaleor hue,! segmentout the LUT entriescorresponding to pixelsin arbitrarilysegmented regions or ! segment3D rangesof RGB space.The 24 bit algorithm segmentsregion-based objects into color segments at muchhigher resolution allowing multi- ple color segmentsto be separatedsimultaneously. Using these algorithmswe can pick individualorganisms or cells out of compleximages. For example, orangeseastars can be pickedout of an imageby selectinga rangeof "orange"pixels resulting in a bitmapof black starfish or anyother blob! on a whitebackground.

! Birdy ObjectAnalysis. An edgetracing algorithm is utilized to trace the outline of the "blob" and convertit to an object. Eachobject is then quantified in terms of its area, peripherylength, x andy coordinates,optical density, principal components, orientation and the computedobject parametersare storedin a tab-delimited"spreadsheet" file for later statisticalanalysis.

! Dgital MotionAnalysis from Video.An importantfunction for manyaf our experi- mentsis to be able to simultaneouslyand continuouslyacquire both kinematic and analog sensordata for longperiods of time.We nowuse the CCD-VS000which supportsboth high resolution >400 horizontallines! video andstereo digital analog0-40000 hz! recording. Colorlmagesupports two modesof movingimage acquisition. A steppingmovie option opens a window,grabs a videoframe, steps the deckto the nextframe and repeatsthis cycleuntil a movieof the specifiednumber of framesis acquired.In this option, the interframe interval is an integermultiple of the videoframing interval 3 msec!or 30 frames/second.The CCD- V5000 is a shuttercamera which stores the stoppedfraine in a digital buffer and therefore providesfull frame output in stop mode aswell as a time basecorrector wbich provides extremelystable stills and resultant kame grabs at highresolution. In the inoving movieoption, theprogram opens the requestednumber of windows,rewinds the deck,starts it in play mode and grabsframes on-the-fly for the requestednumber of frames.This option permits time; lapsevideo digitization of slowerphenomena with the inter-frame interval speciled as a variableprior to digitization. Maneyet al: Techniquesfor UnderwaterVideo Photography

ACKNOWLEDGEMENTS

Developmentof ColorIrnagewas supported in part by NIH Grant DIR-8917532. Videoequipment and funds for researchwere provided by N.S.F.grants OCE-8900144 and OCE-8800640to K. Sebensand J. Witman,and by N.O.A.A.'sNational Undersea Research Center Univ. of Conn.Avery Point!.

LITERATURE CITED

Ayers,J. andG. Fletcher.1990. Color-based segmentation and quantification of objectsin biologicalimages. manuscript!. Burnham,K. P.,D. R. Anderson,and J. L Laake. 1980. Estimationof densityfrom line transectsampling of biologicalpopulations. Wildlife Monographs.72: 1-202.

Buss,L W. 1980. Competitiveintransitivity and size-frequencydistributions of interacting populations.Proceedings of the NationalAcademy of Sciences.77: 5355-5359.

Cainu, F. M. and R. F. Tusting. 1986. Application of lasersto oceanresearch and image recordingsystems. Proc. Lasers '86. Orlando Florida.

Connell,J. H. 1961a.Effects of competition,predation by Thais lapillusand other factorson natural populationsof the barnacleBalanus balanoides. Ecological Monographs 31: 61-104.

Connell, J, H, 1961b. The influenceof interspecificcompetition and other factorson the distribution of the barnacleCtharnalus stellatus. Ecology 42:710-723.

Connell, J. H. 1970. A predator-prey systemin the marine intertidal region. I. Balanus glandulaand several species of Thais. EcologicalMonographs 40: 49-78.

Connell, J. H. 1972. Communityinteractions on marine rocky intertidal shores.Annual Reviewof Ecologyand Systematics3: 169-192.

Coyer, J. A. and J. D. Witman. 1990. The underwatercatalog: a guide to methodsin underwaterresearch. Shoals Marine Laboratory/N.Y.Sea Grant, Ithaca, N.Y. 72 pp.

Dayton,P. K. 1971. Competition,disturbance, and communityorganization: the provision and subsequentutilization of spacein a rocky intertidal community. Ecological Monographs 41: 351-389.

Dayton,P. K. 1975. Experimentalevaluation of ecologicaldominance in a rockyintertidal cornrnunity.Ecological Monographs 45: 135-159.

Edrnunds,P. J. andJ. D. Witrnan. 1990.The impactof HurricaneHugo on a Caribbeanreef. Nature suhmitted!.

263 Divingfor Science.,19Ã

Green,R.H. 1979.Sampling design and statistical methods forenvironmental biologists. Wiley Publishers,N.Y., N.Y. Krebs,C.J. 1989.Ecological methodology. Harper and Row Publishers, N.Y., N.Y. Leonard,A.B., J.R. Strickler andN, D. Holland. 1988. Effects ofcurrent speed onfiltration duringsuspension feeding in Qligometra serrripina Echinodermata: Crinaidea!. Marine Biology9: 111-125. Lubchenco,J.and B. A. Menge. 1978. Cornrnunity development andpersistence ia a low rockyintertidal zone. Ecological Monographs 48:67-94. Menge,B.A. 1976. Organization ofthe New England rocky intertidal community: Roleof predation,competition, andenvirotunental heterogeneity. Ecological Monographs 46: 355-393. Mims,Forrest M.,IH. 1984. Engineer's Mini-Notebook, 555Circuits. A Siliconcepts Book Radio-Shack!:1-31. Pielou,E.C. 1974. Population andconununity ecology. Gordon and Breach Science Pubbsh- ers Inc., N.Y., N.Y. 424pp. Sebens,K.P. and A. J. Johnson. 1990. The effect of water movement onprey capture and distributionof reef corals. Proc. of the Fifth Internation Conference on Coelettterate Biology.Southhampton, England. Hydrobiologia. in press!. llager, G. C., J,-S. Hwang and J. R. Strickler. 1990. Barnacle suspension-feeding invariable flow. Marine Biology105: 117-127. Witman,J.D. 1985.Refuges, Biological Disturbance, and Rocky Subtidal Comttnunity Structurein NewEngland. Ecological Monographs 55: 421-445. Witman,J.D. and K. P. Sebens. 1988. Benthic community structure ata subtidalrock pinnacle in thecentral Gulf of Maine. In. I. Babb and M. DeLucaeds., Benthic Productivity and MarineResources of the Gulf of Maine. National

UnderseaResearch Program research report 88-3: 67-104. Witman,J. D. andK. P.Sebens. 1990. Mesoscale variation in fish predationinteagity- historicalperspective in the Gulf of Maine.Ecology submitted!. Maneyet aL. Techniques for UnderwaterVideo Photography

AppendixI

Underwatervideo andaccessory dealers and manufacturers listed in the 1990DEMA Directory.

Amphibico,Inc. OceanTechnology Systems 9563 Cote de Liesse Rd. 2610Croddy %by, Unit H Dorval, PQ H9P1A3 Santa Ana, CA 927G4 Canada

AquaVideo Inc. Quest Marine Video 5055 NW 159th St. 505 Calle Sorpreso Miami, FL 33014 SanClemente, CA 92672

AquaVision Systems, Inc. Sea& Sea/GMI Photographic 804 Deslauriers St. 1776New Highway Montreal, PQ H4N1X1 PO. Drawer U Canada Farmingdale,NY 11735

Bennett Marine Video Signal3 Corporation 730 WashingtonSt. 1400 26th St. Marina Del Rey, CA 90292 Vero Beach, FL 32960

EquinoxUnderwater Video Housings SonyCorporation of America 101G1 Shaver Rd. SonyDrive MD 1-31! Kalamazoo,MI 49015 ParkRidge, NJ 07656

Helix Catnera And Video Underwater Kinetics 310 South Racine Ave. 1020 Linda Vista Drive Chicago, IL 60607 San Marcos, CA 92069

HYPERTECH Inc. UPI Inc. 750East Sample Rd. P.O. Box 19851 PompanoBeach, FL 33064 Houston,TX 77224

Ikelite UnderwaterSystems UIVPro International 50 West 33rd St. P.O. Box 455 P.O. Box 8100 Napervilie,IL 60566 Indianapolis,IN 46208 DESCRIPTION OF A LOW-COST, SHALLOW-WATER, SURFACE- SUPPLIED DIVING SYSTEM

Dan C. Marelli WakerJaap Florida Departmentof Natural Resources Marine Research Institute 100 8th Avenue Southeast St. Petersburg,FLORK!A 33701-5095 U.S.A.

Researchand maintenancediving tasks that ale conductedin shallow water, particularly those involving extendedperiods of time, are cumbersomeusing SCUBA. Consequently,these tasks are often conductedusing gasoline-powered small air compressors.?use air sourceshave several drawbacks which may compromisethe health and safetyof diversor supportpersonneL A relatively low-cost $200-225!alternative surface-supplied system is describedthat performs as well asgasoline-powered systems, and yetimproves the safetyof divingopera- tionsand is alsomore reliable and requires much Less maintenance. Airis supplied easilyand economically fmmicuge volume SCUBA cylindess 80-100 fs !. llus systemshould be usefulto researchersas well as to personsinvolvedin urukrwater inspectionand maintenanceoperations, and maybe of interestto shallowwater fisheriessuch as spongeand clamindustries.

INTRP DUCTION

Manyresearch and maintenance diving tasks are conducted in shallowwater -30 ft!. Thesetasks often involveextended periods of time underwaterand, because of thedepths and bottom time involved,are cumbersomeusing traditional SCUBA gear. Consequentlymany diversinvolved in shallowwater surveys,periodic maintenanceof experiments,hull cleaning etc.,use small gasoline-powered low-pressure compressors which feed one or twolow pressure hosesand secondstage demand regulators. This systemallows a diver to be relatively unencumbered,but presentsother inherent problems.

Paramountare potentialhealth and safetyrisks to both the diver and the support personnel.The air quality of gasoline-powered "hookah" systems is questionableand unpre- dictable.Most models simply draw air througha valvein the headof the compressor.Since the compressoris driven by a shaftfrom the gasolineengine, this placesthe air intakein dose proximity to engineexhaust. Unpredictablewind conditionsmay causecarbon monoxide, carbondioxide, and hydrocarbonemissions to enter the compressor.Since these low pressure Diving for Science...1990 compressorsareoil free,they deliver unfiltered air directly into the low-pressure hose. Other problemsassociated with the air dehverysystems include water vapor and . Becausethe air is unfiltered,water vapor is alsocarried into the low-pressurehose in appreciableamounts, and the teinperature oftbe air leaving the compressor may reach 88 C 90 F!. In addition,engine failure may place a diverat risk, although the use of anin-line reservetank inay help nutigate the teinperature, moisture, and engine failure drawbacks. Finally,gasoline used to power the engine isa firehazard, the noise of theengine running as wellas its emissions area healthhazard tothe tender, and the engine and compressor require signiTicantmaintenance.

An alternativesurface-supplied diving system exists that performs as well asthe gasoline-poweredhookah, and yet improves the safety of diving operations, ismore reliable, andrequires much less maintenance. In this paper we describe the system, its components andassembly, andprovide examples ofits use in research diving tasks undertaken byFMRI scientists and technicians.

MATERIALS AND METHODS Our systemdiffers from most traditional low-pressure hookah systeins in thatit providesintermediate pressure air to thesecond stage, not low pressure air. Thiseliminates theneed to have the second stage modiTied toaccept low-pressure air.The heart of the system isa conventionalsinglehose SCUBA demand regulator coinplete with a pressuregauge. The firststage mounts on a conventionalSCUBA valve which is threaded into a high-volume cylinder,80-100 ft,3 althoughany compressed aircylinder isacceptable. Thesecond stage is connectedto the first stage by 100 ft. of Synflex 3600-06 light intermediate pressure hose working pressure 250 p.s.i.! with an in-line non-return orcheck valve. This quantity of hoseweighs only 10 lb. and floats above the diver's bead during diving operations. Based on theequation of Soiners972! for surface-supplied requireinents, this hose is morethan adequatefor delivering air at shallow depths. In fact if oneconservatively assumes that the firststage provides 110 p.s.i., then this type of hose will adequately supply a diverto a depth of 100 ft.

Bothends of the hose are flitted with 3/8" female hose fittings. In orderto matethe Synflexhose with the regulator first stage a modifiedintermediate-pressure hosemust be fabricated;this involves removing the female fitting from the distal end of theintermediate- pressurehose and replacing it with a 3/8"male fitting. Thecheck valve is mountedto a surface-supplyharness with a fabricatedaluminum bracket and plate, and the distal end of the longhose threads into this valve. On the distal end of thecheck valve a swivelcross is fitted tofacilitate the connection ofthree intermediate pressure hoses, one for the regulator second stage,and two for inflatorhoses. The longintermediate pressure hose is connectedto the harnessby a snapshackle so thatstrain on thedistal fitting is reduced,and all threaded connectionsare wrapped with Teflontape Fig.1!. For depthsgreater than 15-20fta Marelli & Jaap:Loiv-Cost, Shalloiv-Water, Surface- SuppliedDiving System redundantair sourceis utilized. This consistsof a 14ft 3 cylinderand a singlehose demand regulatorstrapped to thediver's harness, with the valve in a positionwhich is easy for thediver to reach.The costof thissurface-supplied system, without the redundantair supply,is one regulatorand approximately $200-225, With the the costis tworegulators and approximately$325-350. All equipmentnecessary to fabricatethis system is availablefroin anycommercial diving supply finn. Air hosefabrication can be performed by any pneumatic or hydraulicsupply finn.

RESULTS AND DISCUSSION

Thesurface-supplied system is operated in thefollowing manner: The harness is simply strappedon overthe diver'senviroiunental suit asthe lastpiece of equipmentdonned. In shallowwater less than 15feet! we do not weara buoyancycontrol devicebecause the diver caneasily swim to the surfaceand thenditch the weightbelt if the situationarises. The tender assiststhe diver in dressing,especially if a drysuit is being worn, monitors the diver's air supply and tendsthe intermediatepressure hose. In depthsexceeding 15 feet both a buoyancy compensatorand a redundantair supplyis worn. Wemost commonly use the systemwith 100 ft cylinders,and regularly experience bottom times of 3P hourson onecylinder. We find greatadvantage in this systeinwhen multiple stationsmust be divedbecause a divercan enter andexit the water with ease,greatly reducing fatigue and greatly increasing the efficiency of thediver. As many as 25-30 stations can be sampled in oneday by one diver using this system.

Mostof the divesusing this surface-suppliedsystem have been in associationwith the hard clam project in the Indian River lagoonand Horida's eastcoast. We haveused it to surveydensities of adultclams, and in a varietyof field- monitoringand inanipulative experiments.A good dealof our researchhas involved sampling for juvenileclams in the lagoonalsubstrate. This involves placing a circularaluminum 0.25 m 2 quadratinto the substrateand reinoving the substrate to a depthof 5-10cm using a venturi-poweredsuction dredge.The surface- supplied system allows an unencumbered diver to wrestlearound with thequadrat and the suctiori dredge underwater without great difficulty, however experience hasdemonstrated that it is desirableto be overweightedwhile suction dredging in shallow water.This same technique has been used by us to surveyareas of theLooe Key National MarineSanctuary reef flat for spiny lobster prey items, It worksequally well in both soft-sedi- mentsand in seagrassmeadows. Other projects at the FloridaMarine Research Institute utilizethe surface-supplied system tosurvey seagrass beds, to monitor the progress ofrestored seagrasshabitat, and to inspectand clean fish ponds at theredfish and snook stock enhance- mentfacility in PortManatee, Florida. In addition,inembers of DNR'sDivision of Marine ResourceRegulation and Management usethe surface-supplied system tosurvey oyster reefs in ApalachicolaBay.

Thissystem could be utilized by any project which involves long hours of underwater surveyor observation work, especially where tasks involve a lotof manipulating ofcollecting

269 Divingfor Science...1990 gearor requiredivers to performmaintenance tasks. It is limited by the lengthof the intermediatepressure hose to perhaps30 ft, but is extreinelyuseful in depthsof lessthan 20 ft. Other possibleuses for the systemwould be in conunercialfisheries such as clammingor spongingwhere a greatdeal of divingis performedin veryshallow water, and a largetiumber of operationsare currently using gasoline-powered compressors. Additionally, hull cleaning andinspection services will find this a safeand useful alternative to the traditionallow-pressure compressors.

Maintenanceof this systemis no inore difficult than maintaining a conventional SCUBAregulator. Freshwater rinses are all that are routinely needed,and the entire system shouldbe overhauled by a certifiedregulator repair technician at leastonce a year,With 4 systemscurrently in operationand over 300 hours of divingto date,we have experienced only one free-flowingsecond stage and oneworn o-ring, in both casesrequiring minimal repair. Weconsider this system to be safe,easy to use,and economical,and recommend it to all scientificdiving prograins that perform diving tasks similar to thosedescribed in thispaper.

ACKNOWLKDGEMENTS

Wewish to thankGregg Stanton of the FloridaState Academic Diving Programfor supplyingthe check valve for ourprototype system, and the University of SouthFlorida Marine Sciencernachine shop for fabricatingthe bracketto mountthe checkvalve to the diver's harness.

LITERATURE CITED

Somers,L 1972. Researchdiver's manual,revised edition. Sea Grant ProgramTechnical Report16. University of Michigan,Ann Arbor, Michigan.

270 Marelli & Jaap:Low-Cost, Shallow-Water, Surface- Supplied Diving System

Figs' i. Scbernstkot shallow~ter su~appHel divtagsystem

27I