Opinions expressedherein are those of the individual authors and do not necessar- ily representthe views of the TexasARM UniversitySea Grant College Program or the National SeaGrant Program.While specificproducts have been identified by namein various papers,this doesnot imply endorsementby the publishersor the sponsors.

$20.00 TAMU-SG-89-1 05 Copies available from: 500 August 1989 Sca Grant College Program NA85AA-D-SG128 Texas ARM University A/I-I P.O. Box 1675 Galveston, Tex. 77553-1675 Proceedings of the First International Symposium on Kemp's Ridley Sea Biology, Conservation and Management

~88

Edited and updated by Charles W. Caillouet, Jr. National Marine Fisheries Service and Andre M. Landry, Jr. Texas A&M University at Galveston

NATIONALSEA GRANT DEPOSITORY PELLLIBRARY BUILDING TAMU-~9 I05 URI,NARRAGANSETT BAYCAMPUS August 7989 NARRAGANSETI, R I02882

Publicationof this documentpartially supportedby Institutional GrantNo. NA85AA-D-SGI28to the TexasARM UniversitySea Grant CollegeProgram by the NationalSea Grant Program,National Oceanicand AtmosphericAdministration, Department of Commerce. jbr Carole Hoover Allen and HEART for dedicatedefforts tmuard Kemp'sridley conservation Table of Conteuts

.v Preface CharlesW. Caillouet,jr. and Andre M. Landry, Jr, Acknowledgements. vl SessionI -Historical Perspective, Trends and Opportunities inKemp's Conservation and Management Internationalefforts in theconservation and management of Kemp's ridley sea turtle Lepidochelyskempi! ...... 1 jack B. Woody Mexico'scontribution to Kemp'sridley sea turtle recovery ..4 ReneMarquez Millan, Daniel Rios Olmeda, Jose Manuel Sanchez P.and juan Diaz TheNational Park Service's role in theintroduction of Kemp'sridley sea turtle 7 Milford R, Fletcher TheNational Marine Fisheries Service's Kemp's ridley sea turtle research and management .....,...... 10 plan:progress and needs Charles A. Oravetz Questions and Answers. 14 S'essionII - Statusof Kemp'sRidley Sea Turtle Population Nestingpopulation and production ofhatchlings of Kemp's ridley sea turtle at Rancho Nuevo, ...... . Tamaulipas, Mexico ReneMarquez Millan, Aristoteles Villanueva O. and Patrick M, Burchfield Questions and Answers .. 20 SessionIII - Publicand Private Participation in Kemp'sRidley SeaTurt! e Conservation Promotingconservation ofKemp's ridley sea turtle through pub}ic education . 22 Carole H. Allen and Albert L. Barr Therole of SeaTurtle Incorporated in Kemp's ridley sea turtle conservation and public awareness ..,...... 25 Ila M. Loetscher Fishingindustry perspective on conservationand management of sea Ralph Rayburn TED- TrawlingEfficiency Device Turtle Excluder Device!: promoting its use 30 Wilber R. Seidel and Charles A. Oravetz TrawlingEfficiency Device acceptance anduse by Louisiana commercial shrilnpers 33 Paul D, Coreil Sea-AramaMarineworld and Kemp's ridley sea turtle: a lookinto thefuture 36 JohnM. Kerivan Questionsand Answers ~ """. ".~ ... -.....,.......40 SessionIV - Hazardsto and Strandingsof Kemp'sRidley Sea Turtle Trash,debris and human activities: potential hazards at seaand obstacles to, 42 Kemp'sridley seaturtle nesting AnthonyF. Amos Kemp'sridley sea turtle strandings along the Texas coast, 1983-19S5 .43 Robert G. Whistler Recentstrandings of seaturtles, cetaceans and birds in thevicinity of MustangIsland, Texas,...,.....,... 51 Anthony F. Amos Theeffects of petroleumon sea turtles: applicability to Kemp'sridley. 52 Peter L. Lutz and Molly Lutcavage TheCorpus Christi Bay Landmass Project: an example of NMFS'role in protectingmarine turtles ...... 55 via Section 7 of the Endangered SpeciesAct Paul W. Raymond Questions and Answers ...... .. 65 SessionV - Kemp'sRidley Sea Turtle HeadStart Research Standardoperating procedures for collectingKelnp's ridley sea turtle eggs for theHead Start Project ...... 67 PatrickM. Burchf1eldand F, jamesFoley Beachtemperature versus polystyrene foam box temperature in incubation of Kemp'sridley ...... 71 sea turtle eggs RobertE. King

111 predictedsex ratios from theInternational Kemp's Ridley Sea Turtle Head Start Research project ...... 77 ThaneR. Wibbels,Yuki A. Morris,David W. Owens, Gayle A. Dienberg,Julia Noell, JorgeK. Leong,Robert E. King andRend Marquez Millan An analysisof unhatchedKemp's ridley seaturtle eggs...... 82 DonnaJ. Shaverand Allen H. Chancy A further evaluationof imprinting in Kemp'sridley seaturtle ..,........,....90 Mark A. Grassman and David W. Owens Kemp'sridley seaturtle headstart operations of theNMFS SEFC Galveston Laboratory 96 ClarkT. Fontaine,Theodore D. Williams,Sharon A. Manzellaand Charles W. Caillouet,Jr, Questions and Answers ...,, ...... ...... ,...... ...... ,...... ,111 SessionVI - Trackingand Distribution of Kemp'sRidley Sea Turtle Statusof satellitetracking of Kemp'sridley seaturtles 112 John O. Mysingand Thomas M. Vanselous Distributionof juvenileand subadult Kemp's ridley seaturtles: preliminary results ...... ,....., 1 '16 from the 1984-1987surveys Lang H. Ogren Distribution,growth and survival of headstarted, tagged and released Kemp's ridley seaturtles ...... ,..124 Lepidochelyskempi! from year-classes1978-'1 983 ClarkT. Fontaine, Sharon A. Manzella, Theodore D. William, Richard M. Harrisand William J. Browning Distributionand abundance of Kemp'sridley sea turtle, Lepidochelys kenrpi, in ChesapeakeBay ...... ..... 145 and nearbycoastal waters RichardA. Byles Dermatoglyphicpatterns on Kemp'sridley sea turtle flippers:can they be used to identifyindividuals? ...... 146 CharlesW. Caillouet,Jr,, DickieB. Revera,Marcel j. Duronsletand John Brucks Questionsand Answers 151 SessionVII - Kemp'sRidley Sea Turtle Data Base Management Captive-rearedKemp's ridley seaturtle database management , ,152 Dennis B, Koi Marineturtle data base management: National Marine Fisheries Service - Miami Laboratory ...,...... ,...... .....153 Barbara A, Schroeder SessionVIII - BiologicalInvestigations and Captive Breeding of Kemp'sRidley Sea Turtle Evolutionaryrelationships, osteology, morphology and zoogeography of Kemp's ridley sea turtle ...... 157 Peter C. H. Pritchard Feeding,growth rate and survival of the1984 year-class of Kemp's ridley sea turtles Lepidochelys kempi! ....165 rearedin captivity CharlesW.Caillouet, Jr.,Sharon A,Manzella, Clark T. Fontaine, Theodore D.Williams, Marty G, Tyree and Dennis B. Koi Healthcare and diseases of captive-reared loggerhead and Kemp's ridley sea turtles 178 jorgeK, Leong,David L. Smith,Dickie B, Revera, Lt. John C. Clary I, DonahlH. Lewis,Janis L. Scott and Anthony R. DiNuzzo Carapacialscute variation in Kemp'sridley sea turtle Lepidochelyskempi! hatchlings and juveniles ...... ,202 RodericB. Mast and john L. Carr Morphometryof captive-rearedKemp's ridley seaturtles 220 AndreM, Landry,Jr, A reporton attempts to breed Kemp's ridley sea turtle, Lepidochelys kerupi, in captivity... 232 Steven R. Rabalais, David W. Owens and Peter Thomas Attemptsat breedingKemp's ridley seaturtles at MiamiSeaquarium . 233 Timothy B, Bentley Captiverearing and breeding Kemp's ridley sea turtle at CaymanTurtle Farm 983! Ltd...... 237 JamesR, Wood and Fern E. Wood Questions and Answers 241 SessionIX - TheFuture for Kemp'sRidley Sea Turtle PanelDiscussion 244 Questionsand Answers 251 Appendix. -- -...... 258

Acknow ledgetnents

First International Symposium on Kernp's Ridley SeaTurtle Biology,Conservation and Management

Program; CharlesW. Caiiiouet,Jr. and Andre M. Landry,JrCo-Chairmen Welcome: JaniceR. CoggeshaG,William H. Claytonand Edward F. Klirna Regt'strafionand Mailing: RhondaS. Elizondo,Sonia L. Ra!Is and Kelly Shoup Pubffcif y: JohnD. Merritt, CaroleH. Allen, Dana!d E.Pitts, Jr, and Dickie B. Revera Social Events: Carole H. Allen and Judith Wern Audio-Visual: John D. Mcrritt and JaneScheidler Refreshments: Sonia L. RaGsand JoNell Matthews Transportation: Donald E. Pitts, Jr., Kerry Stanleyand JamesWeikum SeaTurf le Head Start Research Facility Tour: Clark T. Fontaine, Theodore D. Williatr s and Sharon A, Manzella Housing: RandyCarr and JoNellMatthews Phofography: Daniel Patlan and John D. Merritt ProgramPrinting: William Gomez ProceedingsTyping: ClaraJ. Surberand LeslyeK. Vaught ProgramTyping: Beatrice W, Richardson and Rhonda S. Elizondo SessionCon veners: Edward F, Klima I!, JackB. Woody II!, CharlesA. Oravetz III!, Frank W. Judd IV!, PeterC.H. Pritchard V!, David B.Bowman VI!, StevenC. Rabalais VII!, RossWitham VIII! and Edward F. KIima IX! Panel Members: ReneMarq uez Millan, Charles A. Oravetz,Jack B. Woody, Milford R.Fletcher, CaroleH. Allen,Peter C. H. Pritchardanal Ralph Rayburn S yrnposiumSponsors: U.S.Department of Commcrce,National Oceanic and Atmospheric Administration,National Marine Fishcrics Service, Southeast Fisheries Cen- ter,Galveston Laboratory and Texas ARM University at Galveston,Depart- ment of MarineBiology Confribufors HEART HelpEndangered -Riley Turtles!;EXXON Houston,Tex.! and the Harris L. and Elira KempncrFoundation Beveragesand Snacks Mixer: Ila Loetscher,Sea Turtle Incorporated, South Padre Island, Tex. TexasStyle Barbecue and Tour: Sea-AramaMarineworld, Galveston,Tex. Beveragesand SeafoodDinner: EarlWayne Israel, The LittleShrimp Boat Restaurantand Bar, Galveston, Tex. Beverageand Pastry Breaks; HEART, Houston, Tex. within any of the individual agencies,This programhas gone so smoothlybecause we havekept it at the operational level, and we believein what we are doing without regard to turf or other responsibilities,Often the more people involved in a program, the less is accomplished. Cooperation among U.S. agencies would be of little value if Mexico were not cooperative, for the Mexican participantsare key players,not only to the current programbut alsobecause they are responsiblefor the fate of Kemp'sridley. It is Mexicothat permitted the U.S.to initiate the headstart project,and it is Mexicothat will or will not allow it to continue.I want to emphasizethis point, becausethere is a tendencyon the part of some"gringos" to overlook this fact. Fortunately,Mexico has been willing to cooperate,and has graciouslyaccepted on-the-beach assistancefrom the FWSin exchangefor permit ting removal and export of 20 to 30 clutches of eggs 000 to 3+00 eggs! annually and specific joint research efforts. Theobjectives of the Kemp'sridley recoveryprogram for the past11 years are few and simple,They are,in of priority: 1. protect the Rancho Nuevo nesting beach and adult females, and maxim ize the hatch; 2. collectand transfer2,000 to 3,000viable eggsannually to the U.S.for hatchingand imprinting at PINS; 3. head start as many hatchlingsas possible;and 4. conductappropriate research and managementprojects aimed at understandingthe speciesand contributing to improved management leading to population recovery. The goalsare recoveryof the speciesand establishmentof a secondnesting populahon at PINS,These objectives and goalsare fairly clearand straightforward,However, to accomplishthem in the real world is not that simple.It is difficult for outsidersto imaginethe red tape,logistical problems, and humanblunders encountered from year to year,Several examples will serveto enlighten.Our first year,1978, was a politicaldisaster, A U,S,Coast Guard, twin- engine Convair 440 and a largehelicopter were seized,and 15 sailors, a NMFS specialagent, and I were held for a few days in Tampicoby Mexicanmilitary authorities. Evidently this occurredbecause the CoastGuard did not have clearancethrough the appropriateMexican agencies to enterMexico and land suchmilitary-like aircraft there.The CoastGuard wasto fly out of Mexicowith the eggs.This detentionof U,S.a ircraft and personnelcreated somewhat of aninternational incident that involved our StateDepartment and a numberof otherMexican and U,S. agencies, includingtop level administrators in theDepartment of theInterior. On another occasion, there occurred an oversight regardingU,S. endangered permits, and a seniorNMFS biologist at the SoutheastFisheries Center's 5EFC! GalvestonLaboratory was under investigationby NMFSenforcement agents for havingKemp's ridley hatchlings in possessionwithout having appropriatepermits on his person.As a result,he faceddisciplinary or legal actionsfor violationof theEndangered Species Act. The problem in thiscase was not some dastardly deed that adversely affected theridley, but rathera questionof nothaving followed the letter of federalperuuts, Technically there may have been a permitviolation, but thereobviously was no harmful intent. Certainly, whatever was being done by thebiologist representedefforts to benefit Kernp'sridley, A few yearsago the NMFSSEFC's Galveston Laboratory faced termination of fundingand closure,with the consequenceof discontinuing head starting. Frantic efforts were taken to find other agendesor institutions to continuehead starting because the hatchlings were already in theU.S. In government,this is thetype of situationin whichsupporters look at their "holecards" and call in favors.As a result,the news media, selected Congressional membersandanumber ofother entities made their desires known,and the funding wasrestomd. Now, funding appears more secure than ever, Everyyear, we have major and minor problems, which are to beexpected in a programof thisnature, We have a bureaucracyand Mexico has a bureaucracy,Both are fraught with redtape and breakdowns in in ternaland external communications.Unfortunately, Mexico is strugglingwith seriouseconomic problems affecting the wholenation. Needlessto say,the Mexican fisheries people Instituto Nacional de la Pesca,INP! involved in carryingout thefield projectat Rancho Nuevo have had to face and overcome many problems. Rend Mirquez can provide examples of what heand other Mexican participants have had to overcometo keepthis important recovery program going. Thenesting beach at RanchoNuevo and the turtles coming ashore to nestare relatively secure under the present managementprogram run by Mexicowith assistancefrom FWS.I seethis program continuing, and the FWS will continueto provideassistance solong as Mexico wishes. Not onlydoes this help Mexico during a periodof austerity, but it hasprovided an excellentopportunity for non-Mexicansto gain uniqueexperiences in a truly integrated internationalprogram, U,S,agencies involved in this programplan to continuepresent efforts with the concurrenceof Mexico.Interest and supportofprivate individuals and organizations havebeen of majorsignificance in maintainingthese efforts. Without thissupport, we couldnot haveweathered some of therough spots encountered over the years. Wehave no indication that the species has made any real gains in survival,even with thenesting beach secured and experimental efforts in place to establish a second nesting colony through head starting, Although factors that historicallyled to the majordecline of the specieshave been overcome by Mexico,they havebeen replaced by others involving the U.S. to a much greater degree.Little progress hasbeen made toward correcting those clearly identified conditions that continue to limit Kemp's ridley recovery, I could be lesscritical if we were dealing with a speciesthat had a widerrange or one for which we werestill searchingto identify factorslimiting recoveryor determininghow to overcome such factors. That is not the casewith Kemp's ridlcy. Under the cooperative program, FWS and NPS have eachspent close to $600,000and NMFS somewherein excessof $2 million to identify the causesof the continued decline and to reverse it. The primary limiting factor keeping the speciesdepressed in recent years is the incidental captureof turtlesby US, and Mexicanshrimp trawlers,There will be no progresstoward recoveryuntil the problem of incidentaltake by shrimpersis corrected.In fact,we may scefurther decreasesin the numberof turtles nesting annually. The shrimping industry and government agencies are aware of the problem, and corrective mechanisms are available,How fastthe industry will moveto overcomethis problemremains to be demonstrated.The FWS and others werelcd by shrimpindustry representativesto believethat this problemwould be solvedon a voluntaryand timely basis, with the support of government agencies,rather than requiring an adversarial climate. Regardless,significant progress must occur within the very near future if Kemp's ridley is to avoid . There have been a number of researchefforts undertaken in and out of Mexico by the combined Mexican and U.S. parhcipants,apart from the primary managementproject at the nestingbeach, the transferof eggs,and headstarting in the U.S. Some of this work has been somewhat esoteric, such as de[err»ining hatchling scute counts and variations, More practical work has been aimed at ensuring survival and production, such as determining optimum sand moisturelevels for maximum hatch and determiningincubation temperaturesrequired for obtaining desiredsex ratiosin clutchesto be transplanted,The NMFS SEFC Galveston Laboratory has carried out researchon identification and treatmentof diseasesand to solve other problemsassociated with maintaining ridleys in captivity. Initial experimentswith satellitetracking havebeen completed, and we are now in the processof further refinementsin transmitterpackage design in anticipationof expandinga jointU,S.-Mexico field projectduring the nestingseason. Mexicanturtle biologistshave developed a techniquethat apparentlygreat1y simplifies determination of the sexof hatchlings,making unnecessarythe costly and laborioushistological methods previously required.Cooperative verification and comparisonof thesetechniques are underway in Mexico,Canada, and the U.S, Mexico and FWS have encouragedresearch having direct managementbenehts to the species,and FWS will continueto help theseefforts, making its resourcesavailable to the greatestextent possible, Research of lessdirect or obviousbenefit is not discouraged,but we may not alwaysbe able to provide all the support desired,including appropriatepermits, if someof the work must be done in Mexicoby a foreigner. Mexicohas requested that FWSserve as the clearinghousefor all requestsby non-Mexicanswho wish to visit or work on the RanchoNuevo nesting beach. This includesresearchers, news media, and any other entity.Our recommendationson each request are transmitted to Mexicofor a finaldecision, This is donefor a coupleof reasons. Theprimary one is that the turtle camp is finite,and its resources as well as the site itself are not capable of sustaining extrapersonnel without prior notificationand arrangementsby both the FWSand Mexico.Food, water and shelter are often very limited as are transportationand sanitaryfacilities. Fromthe inception of theU.S,-Mexico cooperative Kemp's ridley re overyprogram in 1978through 1987, there havebeen approximately 776500 eggs protected, 486,000 hatchlings relcascd at RanchoNuevo, 12,000eggs transferredto PINS,and 12,400head started turtles released. A numberof ridleysof variousyear-classes also are beingmaintained at marineaquaria in the U,S,In addition,Cayman Turtle Farm CTF!is maintainingabout 30 Kemp'sridleys of the1979 and 1980 yearwlasscs. Two of theseanimals nested at CTFin 1984,but thehatch was poor andno hatchlingssurvived. This was the first timethe species had nested in captivity,In 1986and 1987,nestings amongthe captive Kemp's ridleys at CTFproduced hundreds of viablehatchlings, and about 160 of thelatter year- classwere exported to the NMFSSEFC's Galveston Laboratory for headstarting. These events demonstrate successfulreproduction among these animals under captive conditions, However, it shouldnot be expectedthat captive propagationwill ever becomea significant supplementto the wild population. Various other research projectsare underway or havebeen completed. More will be initiated. Thepast 11-year coopera tive effort nationallyand internationally! is unique in government.Although we cannot point to a population increasein the speciesof concern,I think the efforts of Mexico and the U.S. have been instrumentalin reducingthe rate of furtherdecline in thepopulation. We can now focusattention on recognized limitingfactors and on increasing our efforts to overcomethese. lf thiscan be done, there is still hopetha t theKemp's ridley will begin the recoveryfor which we all havehoped and worked.

of a captivebrood stockof Kemp'sridley, suchas that at theCayman Turtle Farm,Grand Cayman, BWI, hasrecently been forthcoming. Fisheries Administration The following will reviewbriefly the moreimportant actions taken by the Mexicangovernment after discoveryof the RanchoNuevo nestingbeach by Mr. AndresHerrera, an engineerand sportfishermanfrom Tampico,Mexico Carr, 1963;Hildebrand, 1963!: 1. The TablaGeneral de Vedas,published in 1956,decreed general closed seasons for commercialfisheries but madeno specialmention of Kemp'sridley, It includeda prohibitionon harvestingsea turtle eggs, 2. Circular 63, published on October2, 1963,by the DireccionGeneral de Pesca,included an experimental regulationprohibiting harvest of seaturtle eggsbetween October and March and a closedseason on harvesting them betweenApril and September Direccion General de Pesca,1963!. 3. Circular 104,published on November21, 1964,by the sameoffice, added two monthsto the prohibition on harvestof seaturtle eggsfrom beaches i e.,during Marchthrough October!,but allowed open commerce from Novemberthrough February for eggsremoved from femalescaught during thefishing season from September 1 to April 30 DireccionGeneral de Pesca,1964!, There was no referenceto Kemp'sridley. 4. Circular9, dated April 27,1965, included a decreeof total prohibitionon commercializationof eggshem nests or from sacrificed females,and established for the first time a closedseason on catch of Kemp's ridley between May 1 andAugust 31 DireccionGeneral de Pesca,1965!. No size regulationwas included. 5. Communicacion15, "Disposiciones vigentes sobre Vedas y TamahosMinimos de Captura- 1967,"dealt with the closedseason for seaturtle catchand minimum sizelimits for the year 1967 DireccionGeneral de Pesca, 1967!,It confirmedthe former dispositions,and they w~ in forceuntil 1971. 6. A total closedseason for all seaturtles and their parts was declaredfrom June1, 1971,to December31, 1972. 7. Seaturtle catch was reinitia ted on September1,1973,but only for fishing cooperatives,after the "Diario Ofidal" dated July 13, 1973 Departamentode Pesca,1973!, A total prohibition on catch of Kemp's ridley and leatherbacksea turtles Dermochelyscoriacea! was established, which is in forceto this day. A total prohibition on commerceof any kind involving seaturtle eggsalso was established. 8, TheDecree of the "Zonade Refugioy deVeda para la Proteccionde la TortugaLora Marina Lepidochelyskepi" was publishedon July 4, 1977 Departamentode Pesca,1977!. It explainedregulations protecting Kemp's ridley seaturtle on its unique nestingbeach named Rancho Nucvo, municipality of Aldama, in the Stateof Tamaulipas, Mexico. 9. A last complementary disposition was prohibition of trawling offshore of the nesting beachbetween Barra del Tordo and Barra de Ostionales during the nesting seasonfrom April to August. Historical Review of the Kernp's Ridley Turtle Program After discoveryof theRancho Nuevo nes ting beachwas announced by Hildebrand963! andCarr 963!,attention was drawn to this speciesand its conservation, One of the pioneers in Kemp's ridley conservation efforts was Mr, Dearl Adams of Brownsville, Tex, He made several visits to Barra Coma, the most accessible beach near Barra del Tordo's sport fishing camp, and translocated several thousand eggs to Padre Island, Tex. see Table 1 in Mdrquez, Vilianueva and Burchfield, 1989!,but with poor incubation success.These represented the first trials to protect and enhance the population. The first Mexican turtle camp was occupied in 1966by technicians of the Direccion General de Pesca.From that time on, the program has remained in operation despite logistical and economic problems during 1969to 1971.In the first period,1966to 1977,more than 250,000 hatcMings were released,ormore than 22 000 per year. Since 19'78,with the increased effort associatedwith the U.S.-Mexico jo Int working arrangement, more than 600,000 hatchlings have been released at Rancho Nuevo, or around 50,000per year. At the same time, several studies were undertakensuch as taggingof femaleadults, which provided biologicalinformation including fecundity,breeding cycle,total mortality, recruitment,etc. Mhrquez,Villanueva and S8nchez,1982!. A diagnosisof the populationwas approximated from these studies. The Kemp's ridley program at RanchoNuevo is consideredin Mexico as a classicsea turtle researchand conservation work, and it is used as a training camp for national and foreign students. At the time the first Rancho Nuevo turtle camp was established, the condition of the Kemp's ridley population was worse than that of any other seaturtle species,Recruitment to the adult populationhad decreasedto nearzero between 'i950 at least!and 1965 P. C, H, Pritchard,Florida Audubon Society,personal communication; Mirquez, 1984!,and the meanage of the nestingfemales appeared to be more than 15 years, The situation improved after 1966 with beachprotection, but the adultpopulation continued to decreaseuntil new recruits to thenesting beach appeared after reaching the age of maturity.In themeantime, natural mortality, contamination, debris and incidental catch of seaturtles by shrimp trawlersand other fishing gear took their tolls. From 1966 to 1977,additions to theadult population through re- cruitmentwere offsetby mortality from all causesbecause the annual nesting population size had stabilized at a low level.Were there no continued nega tive factors reducing survival of the population, we would expect that the number ofnesters should increase, because the high hatch ra te at Rancho Nuevo has been repeated from 1978 to the present, andhas been reinforced by additions to the population atlarge by head starting and release ofthousands ofyearlings during that time. Proposalsfor Future Action The following actionsare proposedfor the future: I. continuethe joint U.S.-Mexicoefforts for five moreyears after the first 10-yearperiod ending in 1988; 2. reduceincidental catch of Kemp'sridley, especially that associated with shrunp trawling; 3. developinformation ata popularlevel about efforts to reducemortality and mcrease population size of Kemp's ridley and disseminateit widely; 4, continuebiologicalresearch on sex ratios, the relationship between incubation temperature and survival rates, andthe monitoring of temperatureof nestsand the beach during the entire season; 5. continuesupport of imprintingand head starting work at Padre Island and Galveston, Tex., respectively; 6. continuethe education program for students,especially those of theUniversidad del Noreste; 7. conductbathymetric work in front of the RanchoNuevo beach;and 8. if possible,increase the effort to study migration behavior of juveniles and adults through the use of remote sensors. AcknOwledge1nents The authorswish to expresstheir thanksto the Mexicanand U,S.fisheries authorities for their assistanceand supportin thecontinuation of the work, Heartily we acknowledge thefriendship of the people of Rancho Nuevo and itsvalue in helping solve day-to-day problems inthe turtle camp, We also wish to express our deep gratitude toIng. ErnestoCorripio C., Director of the Cen tro Regional de Investigaciones Pesquera deTampico, for his assistance in the logisticalwork, to theDirector and students of theEscue'la de Biologia of theUniversidad del Noreste, and to theU.S. studentvolunteers for theirenthusiastic work on thebeach and at thecamp. And lastbut not least,we thankthe DelegadoFederal de Pesca and his staff, and the Mexican Navy personnel for their vigilance during each season. Literature Cited Carr,A. 1963. Panspecific reproductive convergence in Lr7ridodrelys kenrpi. Ergebnisse derBiologic 26:298-303. DepartamentodePesca, 1973. Acuerdo sobre veda y explotaciondetortugas marinas. Diario Oficial de la Federacion, 13July 1973. . 1977.Acuerdo que establece la zona de refugio y vedapara la proteccionde la tortugaIora. Diario Oficial de la Federacion,4 July1977. DireccionGeneral de Pesca.1956. Tabla general de vedas. 1 p. , 'I963.Circular No. 63,19 October 1963. 2 p. . 1964.Circular No, 104,21 November1964. 2 p. . '1965.Circular No, 9, 27 April 1965.3 p, .1967.Ley de pesca de los Estados Unidos Mexicanos. Comision Naciona! Consultiva de Pesca, 376 p. Hildebrand,H.H. 1963. Hallazgo del area de anidacidn dela tortuga marina "Iora," Lepi«ncireiys kempi Garrnan!, enla costa occidental dcl Golfo de Mexico. Ciencia Mexico! 22!:105-112. MarquezM,R. 1984. 4.3 Kemp's ridiey turtle., 4.3.1 Kemp's ridley turtle overview ofbiology, p.96-100. Err:Bacon, P.,F. Berry, K.Bjorndal, H.Hirth, L. Ogren and M. Weber Editors!, Proceedings ofthe Western Atb n tieTurtle Symposium, Volume I, 306 p. MarquezM.,R., A. Villanueva O.and P. M. Burchfield. 1989, Nesting population andproduction ofhatchlings ofKemp's ridley seaturtle at RanchoNuevo, Tamaulipas, Mexico. This volume! MarquezM., R,, A. Villanueva O. and M. Sanchez P.1982. The population ofthe Kemp's ridley sea turtle in theGulf of Mcxico- Lepidochelyskernpii, p, 159-164. In:Bjorndal, K,A. Editor!,Biology and Conservation ofSea Turtles, Smithsonian Institution Press,Washington, D,C., 583 p. The National Park Service's Role in the Introduction of Kelp's Ridley SeaTurtle Milford R. Fletcher'

TheNational Park Service has been involved since 1978 in a multi-agencyeffortto establish a nesting population ofKemp's ridleysea turtleat Padre Island National Seashore nearCorpus Christi, Tex. The procedures callfor turtle eggs tobe trartsported frorrtRancho Nuevo inMexico toPadre Island, cohere theeggs arehatched inpolystyrene foamboxes. Thehatchlings arethen releasedonthe beach tomake their uray to the surf. Exposure tothe beach and surf is assumed toimprint hatchlings toPadre Island astheir natal beach. Upon enteri ng the surf, hatchlings arecaptured and transferred tothe National Marine Fisheries Service, SoutheastFisheries Center's Laboratory in Galveston,Tex. Imprintingisa phenomenonthathas been documented in birdsand fish, but the process ishypothesized tobe among the mechanismsresponsiblefor seaturtles returning tonest atnatal beaches. Thisisthe11th year ofthe recovery program,and slightly morethan 17,000 hatchling turtles have beert imprinted atPadre Islartd todate. There have been no documented returns ofhead startedturtles to Padreisland National Seashore, TheNational Park Service has been involved since 1978 in a multi-agencyeffor t toestablish a nestingpopulation ofKemp's ridley sea turtle Lepidochelys kernpi!at Padre Island National Seashore PINS! near Corpus Christi, Tex. The project,involving the Fish and Wildlife Service FWS!, the National Marine Fisheries Service NMFS!, theTexas Parks andWildlife Department TPWD!, the Instituto Nacional de la Pesca INP! of Mexico, and the National Park Service NPS!,has as its primary objective theestablishment ofa new nesting colony of Kemp's ridley sea turtles on a protected beachin theUnited States. The history and decline of theKemp's ridley is documentedand discussed elsewhere in theseproceedings, somy remarks will be confined tothe role of the National Park Service atPadre Island. Kemp'sr idley isthe smallest ofthe sea turtles and probably the mos t endangered.Sofar as is known, the only extant nestingconcentration occurs on an approximately 20-km stretch of beach near Rancho Nuevo in Mexico, although Kemp'sridleys have been sporadic nesters on Padre Island and elsewhere for manyyears. Underthe terms of theoriginal agreement, the NPS was to becomeinvolved in four majoractivities: 1. receivingeggs from Mexico under permits from TPWD and RVS; 2. providingpersonnel and facilities to incubatethe eggs until theyhatched; 3. imprintinghatchling turtles to thePadre Island beach; and 4. transferringimprinted hatchlings tothe NMFS Southeast Fisheries Center's SEFC! Galveston Laboratory for head starting. Theprocess isfairly straightforward, First, polystyrene foam boxes are filled with moist sand from Padre Island and transportedto the Rancho Nuevo beach. Eggs are collected from laying females at RanchoNuevo by catchingthem in plasticbags as they are laid. Eggs gathered in thisway are promptly carried to polystyrene foam boxes and packed into PadreIsland sand. Eggs, sand, and boxes are then transferred from Mexico to Padreisland where the eggs are held m a hatcheryuntil they hatch, Thefirst part, receiving the eggs and moving them to Padre Island, has gone smoothly over the years with only a fewlogistical problems. The NPS, TPWD and FWS all have been involved in oneway or another in transportingeggs toPadre Island. Sometimes eggs havebeen flown to the Corpus Christi Naval Air Station; at other times, to the Corpus ChristiAirport; and several times, directly to the Padre Island National Seashore near the hatchery facility. Thesecond part, hatching the eggs, is very labor intensive. Upon amva1 at Padre Island, boxes containing the eggs aretransferred toa speciallyconstructed hatchery facility. Multi-channel, temperature probes record temperatures hourlyat three elevations in the incubation boxes, and large quantities ofdata are generated and synthesized. Sand moistureand presence-absence of molds or fungiare monitored daily. The NPS staff at Padre Island computerizes the datarecords to facilitatetheir evaluation,Evidence is now mountingthat sexof Kemp'sridley hatchlingsis determinedbythe tempera ture at which eggs are held during a certainphase of theirdevelopment Shaver etal., 1988!.By artificially elevating incubation temperatures, weare now producing a great preponderance of females. Implicationsof thisconcept are profound,

U.S.Departrrrent of theinterior, Rational Park Service Thethird part that NPS plays is imprinting hatchlings, Sea turtles, like salmon,are thought to imprinton their natal surroundings.Something in theirphysical environment acts to imprintthe anima 1s. Imprinting is a phenomenonthat isbetter known in birdsand certain fishes than in seaturtles, but theprocess is assumed to be similar. There are thine classicalcomponents of imprinting that are thought to be important in this project, First, it occursat an early age. Precocialbirds, like ducks and chickens, imprint on the largest moving object they see after hatching. This is usually theirmother, but if themo ther is absent and the young are exposed toa movingballoon or a toy train during this critical period,they instead will form a strongattachment tosuch objects and retain that attachment asan adult, Interestingly enough,the more difficult the moving object isto follow, the stronger isimprinting. This may have a parallelin the difficultywith whichhatchling sea turtles dig theirway out of thenest and traverse down the beach to thesea. Thesecond factor of imprinting isthat it occursina relativelyshort time. Mallard ducks, for example, imprint best somewherebetween 13 and I6 hoursafter hatching. Chicks do not follow a movi~gobject when they are a fewhours oldor when they are several days old, but only during a fairlyshort period of time.Gosslings, forexample, when hatchedin anincubator, readily imprint on the first large object they sec, often a human. Thethird factor is that the attachment ispermanent and is retained from the tinic of imprinting to adulthood. Pacific salmoneggs and fry imprint on the chemical smell and taste of the water in which they hatch. After a seajourney of severalyears they mature, and the adult salmon unerringly return to the same stream they hatched inyears earlier, Onecompelling demonstration of this is afforded by tagged salmon that return to a fishhatchery asadults, Anyof these factors could contribute toturtles returning totheir natal beaches tonest, The reasoning isas follows. First,adult female turtles have the ability to find the nesting beach with great accuracy after several years at sea. In thecase of the Rancho Nuevo beach, turtles nest along a stretchapproximately 20km long, Why only this particular stretchofbeach, when there are thousands ofkilometers ofbeach available? Somehow, after more than seven years at sea,females, and evidently males, are able to locatethis unique stretch of beach and reproduce, Second, the experiencewith the natal beach is limited. Embryos develop within approximat'ly 60 days after eggs are laid. After thistime, hatchlings digtheir way out of the nest, scurry down the beach, and enter the surf, thus disappearing into thesea for the next few years. The time from hatching to entering the sea may range from minutes to a fewhours. Somethinghappens tothe embryonic orha tchling turtles during these 60 days or so to permanently imprint them on thegeographic location or other characteristics ofthe beach. We feel that the embryonic turtles imprint to either the sandin whichthey are incubated and hatched or to physicalor chemicalcharacteristics of thebeach itself. A chemical imprintingsystem is postulatedunder which ernbyonic turtles sense, imprint on, and retain at leastsome natal impressionsthey receive during the incuba tion period. If thisis so, then those turtles expected toreturn to Padre Island shouldbe exposed tonothing but Padre Island sand. This, then, is our reason for moving Padre Island sand from the U,S, to Mexico and back. It is impossibleto exactlyrecreate the natural conditions under which Rancl'o Nuevo turtles are incubated and hatched,butinsofar asis possible anattempt ismade. After the eggs hatch, the hatchling turtles are transported a short distancetothat portion ofthe Padre Island beach that is closed tovehicular traNc, and there they are released tomake theirway down to the surf. Activity increases anda massexodus tothe surf begins asthe hatchlings arewarined by themorning sun, They orient themselves toward the surf and morning sun and scurry across the beach. Soine observersspeculate thatthese few minutes ofexposure tothe sun, sand, beach and surf are what imprint a turtleto returnto that beach. Perhaps, likesalmon, hatchling turtles chemically imprint on the particular chemical makeup of thesea water at that particular beach. Although neither ofthese possibilities hasbeen quan tihed beyond doubt, there isevidence that turtles hatched and imprinted at Padre Island later showa preference forPadre Island sea water when offered a choicebetween that and sea water from other sources, Eventhough the eggs are placed directly insand from Padre Island, and hatchling turtles are exposed tothe smell or tasteof thatbeach, this does not guarantee that imprinting to thatbeach has thus occurred. There has been speculationthatthe geographical location ofthe developing eggs may be signii'icant, Tinybipole magnets have been discoveredinorganisms asvaried asbacteria andpigeons. Biomagnetic orientation isbelieved tofacilitate homing movementsin pigeons and may act in concert with other environmental factors to provide the homing mechanism inturtles. Ifthis is so, and the magneticbipoles areoriented during embryonic devclopment, itfollows that developing eggsshould be moved to Padre Island as soon as possible after collection inMexico. Because ofseveral factors, this hasnot alwaysbeen done, Afterimprinting on the beach, the young turtles are quickly gathered from the surf and moved to the NMFS SEFC's GalvestonLaboratory. Thisinitiates thefinal phase ofthe head start project inwhich the turtles are reared incaptivity fora numberofmonths toallow them to reach a sizethat significantly reduces their vulnerability topredation after theirrelease. Hatchlings have been either driven by vehicle or flownby aircraft from Padre Island to Galveston, Throughtheyears, the NPS has provided personnel orfunds for nearly every facet ofthe head start project, inaddition tocontracting several scientific studies related tothe project. Excellent press cover ge has been a partof the entire U S. mcubating-hatching-imprintingphase of theKe tional Seashore ',

f Hatching Number of Number ved percent at hatchlings hatchlin co Padre Island imprinted to NMFS

88.1 1@67 85,7 1,754 84,1 2,487' 83,3 1/87 77.6 1559 12.1 235 90.7 1,783 84.1 1,720' 88.3 1,775 64.3 '1,283 91.6 929

77,1 '17,279 ! andDonna Shaver

TheNational Marine Fisheries Service's NMFS! Kemp's Ridley SeaTurtle Research andManagement Planwas prepared largelybecause ofa recommendation madebya constifuent reviewpane! ata NMFSseaturf leprogram review inMiami, Florida onMay 3-4,19S4. Fiveplan components relatingtoKemp's ridley Lepidochelys kempi!are performed bythe NMFS Southeast RegionalOffice: InformationandEducation Abroad range ofindependent andconperativeactivi tiesinform andeducate thegeneral public andfisherme ofthe need toconserve Kemp'sridley. These activitiesincludedevelopnrent anddisfribution ofposters, video tapes, brochuresand other means of communication. TEDTechnology Transfer TED isan acronym forTrawling Efficiency Device orTurtle Excluder Device. TheTED was developedprincipally toallow escapement of'seafurtles from shrimp trawls, thereby enhancing conservation ofail 1sea turtles includingKemp's ridley, Useof fhe TED provides benefits toshrimpers suchasreductions oftrash and finfish by-catch. The NMFShasencouraged fhevoluntary useoffhe TED,is continuing tof ransfer TEDtechnology foshrinrpers, andisimplementing TED regulations. IncidentalCatch Incidental catchinformation canbeused formanagement purposesand toincrease conservation through awareness.Amendmenfs tothe Endangered SpeciesAct ESA! alloro reporting ofthe incid entail takeof the endangered Kemp's ridley.Anincidental catchreporting schenrehas been explored withvarious components ofthe shrimpi ngindustry. SeaTurtle Recovery Plan A plan for the recovery ofsix species ofmarine turtles wascompleted inSeptember 1984.This plandefines recovery actions forKemp's ridley. Recovery actions such ashead starting, captive propagation, andat-sea monitoringhave been implemented, Section7 Consultations Sectiori7 ofthe ESA requires thatall federal agencies conserve endangered speciesand consult withthe Fish and Wildlife Service and/orthe NMF S tomake surethat they donot jeopardize fheexisf ence ofany listed species. Secfion7 consultations areused fo mitigate adverse impacts toKemp's ridley. TheNational Marine Fisheries Service's ~! Kemp'sRidley Sea Turtle Rcscarch andManagement Planwas developedinAugust 1984, largely inresponse torecommendations madebya constituent reviewpanel ata NMFS SeaTurtle Program review, This panel ofenvironmentalists, scientists,commercial fishingrepresentatives and administratorsassembled inMiami onMay 3-4, 1984, tolearn ofNMFS'activities regardingsea turtles. Thepanel was askedtocritically review NMFS' programs andrecommend changes.Many recommendations weremade andmany appliedspecifically tothe Kemp's ridley Lepidochelys kempi!.Thereappeared tobe a generalconsensus ofthe panel thatmore emphasis onthis critically endangered specieswas needed. AsfoHnw up,the Southeast Regional Office SERO!andSoutheast Fisheries Center SEFC! ofNMFS developed a plarining document tospecify which ofour activitiesrelated tothe Kemp's ridley. The management partof the plan basically consists offive program elements thatare performed bytwo people inthe Protected Species Branch ofthe SERO inSt, Petersburg, Fla. Inforxnation and Education Inmy view, ini'ormation andeducation activities arehighly valuable inenhancing theconservation ofendangered species.Thepublic cannot beexpected toconserve ifthey are not informed, andsometimes theinformation hasto be prettybasic. Ina conversationwitha shrimper aboutthe TED, I was told that he had caught only one dead sea turtle inhis lifetime. Theturtle apparently waspartially decapitated, sohe removed itshead, and the skull now adorns his retailseafood market, Healso told me that he resuscitates comatosesea turtles that he catches, thenpaints thename ofhis seafood company ontheir backs before releasing them,At sometime inthe past, that shrimper learned thatsea turtlescould be resuscitated andknew that recapture datawere important, However, information anded~-cation effortshadfailed toconvince himthat itwas illegal totake and display a sea turtle skull [Editors' note;except for scientificoreducational purposes under appropriate federaland state perinits, orif it can be proven that the artifact wasobtained prior to listing of the sea turtle as threatened orendangered].

' ¹fivnal Marine FisheriesService

10 Informa tionand educa tionactivities areimportant, butitis difficult todetermine what approaches touse and how toevaluate theireffectiveness sowe continue witha shotgunapproach aimedat informing thepublic ofthe need toconserve seaturtles and of ways todo this. This approach hasproduced a variety ofindividual andcooperative informationalandeducational materials. TheKemp's ridley video tape, Heartbreak Turtle,sponsored byKUHT oftelevisionspecific, station,in-depthChannel documentaries 8,Houston, ofaTex., specialandconservation the book The effort.Great I Rid!ayhavehad Re..-eve the opportunity Phillips, 1989! ofextendingare fine examplesKUHT's effortAnbyexample providing ofa copies more generalof the video education tape toeffortseveral isa marine SeaTurtle scienceIdentification centers where Poster.it is Thisshown poster to visitors,was developed cooperativelybyNMFS andthe Center forMarine Conservation CMC!.Atleast 15/60 ofthese posters havebeen distributedworMwide, A Spanish versionhasbeenproducedby CMCandis part ofan educationalpacket forteachers inSpanish-speaking countries.Thepacket has been distributed widelyinPuerto Ricoand the Dominican Republic. NMFSpurchased about300 packets tosupplement privatedistribution ef'ortsin Mexico. Threeother educational projectsspecific toKemp's ridley have been completed orare planned. A Kemp's ridley identificationposterwasprinted inJuly 1985 anddistributed. Anotherspecies identification sheetdepicting thefive seaturtle species inour region wasdistributed, Oneofits main purposes wastogather information fromfishermen whocatch ridleys incidentally, A thirdeffort that has notprogressedbeyond theplanning stageisthe joint production ofa brochureaboutKemp's ridley similar toone developed onthe Hawaiian MonkSeal. Thisis proposed bythe NMFS centraloffice in Washington,D.C. TED TechnologyTransfer TEDisan acronym forTrawling Efflciency Deviceor Turtle Excluder Deviceas the case may be.This piece ofgear wasdeveloped principally toexclude seaturtles from shrimp trawls. Because ofits other benefits suchas trash exclusion,reduction offinfish by-catch, andimproved shrimp catch, NMFS encouraged itsvoluntary use,NMFS' TEDtechnology transferactivities haveinvolved publicity, demonstrations, anddistribution ofsample TEDs,The TEDisthe subject ofanother paperbeing presented atthis symposium, soit is not being discussed indetail here. However,NMFS currently isimplementing TEDregulations Department ofCommerce, 1987!. Incidental Catch Incidentalcatchinformation hasseveral uses, Itcan be used toidentify loca tions ofoccurrence ofsea turtles, their sizes,timeperiods ofabundance, andwho in the fishing industry mostoften encounters seaturtles. Itcan alsobcused tostimulate conservation. Inthe case ofthe Kemp's ridley, weknow that shrimpers takethem asdo recreational fishermenusinghook and line gear, both from boats and from shore. I also read somewhere thatthe Kemp's ridley istaken ingill nets. This isall good information thatcan be used toconcentrate effortstoreduce thecatch, especially if thereisassociated mortality. Incidental catch information provides data for researchers andfor other management applicationssuchasSection 7 consultations underthe Endangered SpeciesAct ESA! of1973. Perhaps moreim- portant,Letme it providecan be used anexamplefor conservation ofwhat exercises.I mean, Several years ago, I was meeting withsome Georgia shrimpers concerninga veryhigh frequency ofstrandings ofsea turtles onGeorgia beaches, Theshrimpers didnot deny that theycaught andaccidentally killedsome seaturtles, butthey letme know inno uncertain termsthat they would not beheld responsible forall the dead sea turtles that washed upon beaches. Theyhinted that if someonewereto try imposingburdensome regulations onthem, wemight see twice asmany dead turtles onthe beaches. Someofthose same shrimpers thatwere so adamantly oppose to imposed turtle conservation atthat time cooperatedina seaturtle tagging project with the University ofGeorgia. Theresults oftheir first year ofeffort were summarizedinMarine Turtle Newslet ter No. 34, p. 1-2,1985!. Briefly, coopera tingshrimpers caught, tagged, measured andphotographed 40sea turtles. Ofthese, 31were loggerheads Carettacarlotta! andnine were Kemp's ridley. In additiontothe valuable scienhfic information thisproject provided, itenhanced seaturtle conservation. I believe that onlytwo of these turtles werebroughtup dead,and 1 amsure that this was unintentional. Moreover, the45otherusers ofthis shrimping cooperative's facilities, andtheir fellow shrimpers inGeorgia, were no doubt influenced toconserve seaItturtles wouldby be thisnice effort.if wecould involve every shrimper in the southeast ina similarproject. We cannot, but we have triedto expand this effort to Texas, incooperation withshrimper groups. Amendments tothe ESA made it easier to issuepermits allowing theinadvertent takingof an endangered species,A condition isthe establishment ofa conservationplan.We have explored thepossibility ofestablishing a scaturtle by-catch reporting scheme with shrimpersandsome data collection hasbegun, A conservation planthat would bedeveloped asa permit condition mightinclude measures suchasproper resuscitation, relocationand release ofsea turtles, limits onthe length oftow time,and even the testing-adoption' ofTEDs, Major advantages couldbe realized bythe shrimpers, whowould be

11 protectedfrom an otherwise prohibited act, and for the sea turtles, because a person who has a definiterole in controllingits life or deathwould be involved in its conservation. A Recovery Plan for Marine Turtles A RecceeryPlanfor Marine Turtles Hopkins and Richardson, 1984!that covers sixspecies ofmarine turtles was preparedbya recovery team and approved bythe NMFS Director inSeptember 1984.The plan sets forth those activitiesthatshould betaken byagencies sothat sea turtles canbe restored tolevels where they no longer need protectionunderthe ESA. That isthe textbook description ofwhat a recovery planshould be,In reality, forthe Kemp's ridley,theplan sets forth those actions that should betaken totry to stop the drastic decline ofthis species, Theactions areprioritized andlead agencies defined. Insome cases, oneagency isident i fied and in other cases several. ForNMFS thereare 10 lead actions for the Kemp's ridley: 1. Regulatethepetrochemical industry.NMFS probably doesnot have authority toregulate thepetrochemical industrydirectly, butcan best achieve thisobjective through Section 7 consultations withthe Department of theInterior's Minerals Management Service. 2. Usehatcheries andhead starting. Thisproject isongoing atour NMFS SEFC's Galveston Laboratory. 3. Maintaina totalban oncommercial,recreational andsubsistence take. I can assureyou wehave every intention of continuingto do this. 4. Establishcaptivebreeding colonies,This isanother activity accomplished throughour Galveston Laboratory. Kemp'sridleys captive-reared atthe Galveston Laboratory havebeen distributed toa number ofcooperating organizations,inpart to provide a reservoir ofanimals forcaptive propagation experiments Caillouet, 1984!. 5. Regulatespoil dumping, seafloor mining and trawl tows. NMFS shares this lead with the Environmental Pro- tectionAgency with regard tothe fUst two items, Our answer toregulating trawl tows, asleast thus far, is TED. 6. Maintainandenforce theban on take throughout therange. This issimilar toitem 3,which weare doing in cooperationwith manyother agencies. 7. Regulateshrimping methods, gear,areas, andseasons inU.S, waters. Steps inthis direction arebeing taken throughtheTED technology transferprogram, through theincidentalcatch reportingsystem andthrough TED regulations. 8, Recommendregulations forshrimping methods, area,year, and season inMexican waters. Ourapproach to thisaction iscooperation withMexico inTED technology transfer, 9. Deterniineunknown mortality factors, ifany, and take appropriate action.The Kemp's ridley pathology pro- jectthat was carried outat our Galveston Laboratory hasbeen discontinued. Onereason isthat mortality dur- inghead starting Kemp's ridley was reduced toless than 15percent through iinprovements inrearing methods. 10.Determine feasibility ofaerial and other means ofat-sea monitoring. NMFShas several projects inprogress or planned.

Section 7 Consultations Section7 ofthe ESA requires several things offederal agencies. Firstof all, it requires thatall federal agencies "...in consultationwithand with the assistance ofthe Secretary [ofComnierce orInterior], utilize their authorities in furtheranceofthe purposes ofthis act by carrying outprograms forthe conservation ofendangered speciesand threatenedspecies..." Italso requires federal agencies to",ensure thatany action authorized, funded, orcarried out bysuch agency isnot likely tojeopardize thecontinued existence ofany endangered speciesor threatened species..." Section7 istruly a powerfultoolthat can be used forthe conservation andmanagement ofendangered species,and itis used tothe extent practicable inthe NMFS Southeas t Region. Wecomplete about 110 such consultations eachyear. Ourjurisdiction covers five endangered whales, onefish theshortnose sturgeon

12 Literature Cited Caillouet,C.W., Jr, 1984. Essa i de prevention de I'extinction de la tortue de Kemp. Les Camels deZoologic Bulletin of i!re Zoological Societyof Quebec!44! 28-34. Departmentof Commerce. 1987, Sea turtle conservation; shrimp trawling requirements; finalrule. National Oceanic and AtmosphericAdministration, Federal Register 52 24!:24244-24262. Hopkins,S.R.and J.I. Richardson Editors!. 1984. Recovery Pbnfor Marine Turtles. National Marine Fisheries Service, 355p. Phillips,P.1989. The Great Ridley Rescue, Mountain Press Publishing Company, Missoula, Montana, 180p.

13 Questions and Answers BlancheLynn: Do the Mexican people stillhave the privilege oftaking a certain number ofridleys forone purpose oranother fortheir own use? People cohocome toour shows tellof one experience afteranother inwhich a rid leyhas been taken forpersonal use. Mkrquez:Youare aware that the Mexican people have many necessities including food. They also need money to survive.This isso in all our States, Notonly does this create a problem forthe Kemp's ridley nestingbeach atRancho Nuevo,but for all our other sea turtle nestmg beaches inMexico, We do not watch the Rancho Nuevo beach atthe beginningofthe nesting season, andpeople goto the beach prior to our arrival and take eggs, and sometimes the turtles.Ifwe are not on the beach atthe beginning ofthe season wecannot stopthese activities, because thepeople aroundthatarea know when our work isgoing onthene. I donot know forhow long a timethey have been doing this, butfor many years there has been consumption ofeggs taken from that beach. The local people remember thattheir grandmotherorsomeone likethis made bread with the turtle eggs, andthen they wish toeat that kind of food again. Thecommercial harvesting therehas been stopped since 1966. Some people goto the beach anyway when we are not patrollingit.We make pa troIs two times each day, and then they go to the beach inthe afternoon. Sometimes one ortwo ridleys appear nesting inthe afternoon, when we are not there. From such causes welose about 10 percent ofthe eggs during a season. Sometimes wecannot tellwhether theeggs are lost to people orto predators. Insome cases,weare not able to collect eggs from the nests when the wind is very strong. But after 50 days or so, sometimes hatchlingsappear onthe beach innatural condition, Therefore, some of the eggs that we miss remain on the beach andhatch out, but some others arestill stolen bypeople, dogs orcoyotes. Usually they do not take the adults, justthe eggs. ICIima:Would you comment onthe educational program youhave started with the people atthe village ofRancho Nuevo? Mirquez:Theturtlecamp was begun in1966, and at that time we had some conflicts with the people inthe village, becausetheythought thatwe were foreigners. Also,they wanted tocontinue taking eggs asusual. Astime passed, thepeoplebegan tounderstand theproblem andhelp us. In1985 weused a new approach toreach thepeople through theprimary school inthe village near the turtle camp. When thehatch occured andthere were many hatchlings on thebeach, weinvited the children inthe school tohelp in releasing a those11 hatchlings from the beach. The children andthe teacher ofthis school participated, anditwas a veryenjoyable timefor everybody. Wehope todo this with otherschools in1986. It also would be good if we were able to invite some people from Brownsville, Tex.,to participateina hatchlingrelease students and teachers tomake the same joint effort at Rancho Nuevo. DavidForcucci: Yousaid that the commercial harvesfofeggs was stopped in1966. From thatpoint on,how much poaching was fhere?Was it prevented?Were the beaches watched from then on? Mkrquez:Theproblem wasnot stopped untilthe joint Mexico-U.S. programwas initiated in1978. After the joint programbegan wehad motorcycle s andit was easier toremove thepoach ers. But before that time,we were confined towalking the32 km beach orsometimes covereditby horseback,and occasionally wehada good lookby jeep. The poachersonthe beach during 1966 to1977 were using horses. Itwas very difficult topursue thepoachers across the marsh,when they were onhorses andwe were ina jeep.Weestimate thatafter 1977, theloss ofeggs averages 10 percentper season, not only frompoaching but fromother natural causes as well. RossWitham: I havesomeconcern abouftherelease ofhead started turtles. There isapossiblitity thatthe imprinting mayoccur duringtheoffshore swim,when fheturtles areswimming awayfrom Padre Island. Thatcould beeither magnetic, orit might be a sonicconditioni ng. Do you have any comment on that? Fletcher:Youare absolutely correct. Itcould well be that the imprinting phase ofthe turtles takes place inthe surf oras they move offshore. Thatcould well be the case. However, thereason thatwe have not tried todo anything with thatisbecause oncethey would get offshore it would betoo late to retrieve them for head starting. Asit is difficult tocatch them in1 foot0 cm! of water inthe surf. Once they hit the surf, they are extremely difficult tocatch, Soif weactually didlet them swim out farther than that, we would lose them, Sowe do not really have any choice about that. Witham:I donof disagree withyou on that, but what I am suggesting isthat the head sfarfed turtles should bereleased from PadreIsland, sothey can swim offshore fromthere. They may benefit from that experience. I thinkthereis some evidence that showsa veryprecise post-hatch learning in somesea turtles, Fletcher:I can say that we certainly will takethat into account in ourdeliberations. MarydeleDonnelly: Whatsort of changes havebeen made inthe nesti ngtemperature? Areyou tryi ng to have more females inyour population, andif so, what sort of sex ratios doyou have forridleys? Wint I am concerned aboutis thaf there tssome indicationwith green sea turtles that there are fhree or four males per female.

14 Fletcher.That question islong and involved. I am sure someone else is going to present data on this. Basically, we havebeen producing a preponderance ofmale Kemp's ridleys up to 70 percent male. We feel now that the sex of turtlesis pretty well determined bythe temperature during the middle third of the incubation period, and yesterday inour intergovernmental deliberations weagreed that we will change our hatchery facilities and hatching regimen toraise the temperature from an average ofaround 28'C or 29'C to somewherebetween 30'Cand 32'C, with the hope of bringingthe male-female ratio nearer to 50:50, Donnelly:But do you know that 50:50 ts correct? Fletcher:No, we do not. The evidence that we have so far is that,in boxturtles for example, the sex ratio is around 60 femalesto 40 males. We do not know what the seaturtle sexratios are in the wild. Klima:You are probably aware of theextreme difficulty there is in ide~ttfyingthe sex of hatchlings,and we have acquiredthe technology todo this just recently. The National Park Service and Instituto Nacional de la Pesca have beenworkin'g on this problem, and we have just recently obtained informa tion on the sex ratios of some of the hatch- lings.This information says basically that we have produced a preponderance ofmales, and just before this sym- posiumwe took action to see that in the future we will have more of a naturalsituation, orone similar to that which occursat Rancho Nuevo. So in thatregard, we assume that we will producea sexratio similar to that which occurs at RanchoNuevo. Whether that is goodor bad,we do not know,beca use sex ratio could be cyclic. There are many questionsrelated to this problem, not only for Kemp's ridley but for other sea turtles as well, DavidBowman: Would you agree that we should point out that the only way the sex of a seaturtle can be determined isif it is dead,and we are not prepared at this time to sacrifice a lotof turtles to deterniinesex ratios? Klima:Yes, this is a problem.The turtle has to be dead, so our samples are from dead turtles. They may no't be representativeof the sexratio of live ones, Burchfield:Another point that we have been looking at is what we now have at theRancho Nuevo natal beach are isolatednests by individual turtles, If welook back at the historical nesting effort at Rancho Nuevo, it is a predator swampingtype of reproductive strategy, in which on one day in 1947there were an estimated 40,000 turtles on the beachwithin a 2-to 3-hour period. We have some preliminary evidence that the thermal dynamics of closelypacked nestsunder such an arribazon situation may be dramatically different from those of isolatedindividuals nests. So even if weunderstand what is presentlytaking place at RanchoNuevo, in termsof thenesting or breedingpopulation, it still maynot be indicativeof whatmade up thenatural population historically. PaulRayntond: Have you considered eliminating the polystyrene foam box hatcheries and going to a natural,on-the-beach hatcheryon PadreIsland? Hetcher:Yes, we have,We have considered that option several times. The problem, and I amsure that one of the speakerswill presentsome data on this, is that the beach temperatures atPadre Island are not the same as at Rancho Nuevo. RaymondWould they not reflect more natural temperature conditions than the foam boxes? Bill Lukens:We have considered moving the eggs to thePadre Island Beach. The problem is linkedto whenwe get theeggs out of Mexico.It isusually past the point in incubationat whichwe feel the eggs can be safely moved to a beach.So we have been putting them in the boxes,to holdand to transportthem. That has been the easiest way. In 1986,we aregoing to maintaintemperature during the incubation period with considerablyless fluctuation, Bob Kingand I havediscussed eliminating the foam boxes and using some different kind of material. TerryCody: You mentioned the incidental take in Georgiaand the study that involved cooperating shrimpers, When you talk ofpermit ting Texas shrimpers, are you suggesting that the fishermen will beable to keepthe turtles in theirpossession? Oravetz: No. Wha t weare trying to dois to getthe shrirnpers permitted for whatis nowtechnically an illegal activity underthe Endangered Species Act, with thehopes that instead of beingfearful of thelaw andthrowing the turtle overboard,the fishermen will takesome steps to resuscitatethe turtle, tag it, andperhaps release it in anarea where it is not subjectto being capturedagain. AndreLandry: What utilization does National Marine Fisheries Service make of county marine extension agents, particularly in workingwith shrimpers? I know the extension agents have very close contact with shrimpers and have some rapport with shrimpersthat NationalMarine Fisheries Service may or maynot have. Oravetz: I should have mentioned this. In our case,there are four of us. We can spread ourselves only so far. We have hada tremendousamount of supportand help from the cooperative marine extension service - mainlythe Sea Grant MarineAdvisory Agents, Specifically, they assist in passingthe word to the shrimpers about the use and availability of theTED. Sea Grant has a networkof MarineAdvisory Agents in all Gulfof Mexicoand South Atlantic coastal states. Theprogram is verygood, particularly in Texas.We utilize the agents extensively to helpget information about turtles and shrimp to the fishermen,

15 NestingPopulation and Production of Hatchlingsof Kemp's Ridley Sea Turtle at RanchoNuevo, Tamaulipas,Mexico Rend Mirquez Mi1lan, Aristoteles Vilfanueva O. and Patrick M. Burchfield'

Nearlyallmal ure females inthe population ofKemp's ridley sea turtle Lepidochelys kempi!nest at Rancho Nuevo, located betweenBarradel Tordo 335'N, 97'453'W! andOstionales 3'24 6'N and 97 45 7'W!,in Tamauli pasState, Mexico, Small groupsalsonest inseveral locations suchas Playa Washington andsouth ofBarra del Tordo inTamaulipas andTecolutla in Veracruz.Thereare solitary nestings atPadre Island, Texas, Cabo Rol'o, Veracruz, andIsla Aguada, Campeche, Mexico. Nevertheless,theannual number ofnests outside ofRancho Nuevo usually isless than 50. Theannual nesting population atRancho Nuevo hasshown a great decline sincethe discovery in1947 of an arribada ofmore than40,000, Nowadays, therearefewer than 800 fe>nales nestingannual!y atRancho Nuevo. Thetotal nesting population can beassessed through life cycle characteristics andfecundity. However, sexratio at sea is unknown. Thequantity ofhatchlings producedhas varied annually, andisrelated notonly tonumber offemales thatnest every year and naturalmeteorological disturbances suchasstorms orfloods but also to the intensity ofour biological workat the Rancho Nuevo TurtleCamp. About 66430 hatchli ngswere released fromthe Rancho Nue»o beai h between1966and 1985. From 1978 to1985 about13/50 hatchlings emerged fromeggs collected atRancho Nuevo andincubated atPadre Island. These hatchlings were imprintedatPadre Island then head started for10 to 11 months inGalvesto~, Text, Head started survivors were released into theGulf of Mexico asyearlings oryounger from sites off Mexico, Texas and Florida TheKemp's ridley sea turtle Lepidochelys kempi!isunique among seaturtles inmany characteristics andbehavior, especiallyinits panspecific adaptation toonly one primary nesting beach inthe Gulf of Mexico Carr, 1963; Hildebrand,1963;Pritchard, 1969;Cases-Andreu, 1971;Pritchard andMirquez, 1973; Chavez andKaufmann, 1974; Mirquez,1976,1978; Mirquez etal., 1989!, andthere are no documenfcM datatothe contrary inhistorical times, This naturaladaptative feature ofone major nesting sitehas been a determinantofKemp's ridley population abundance. Maybeitworks favorably against natural predation, butfor man ithas provided anattractive opportunity foreasy andprofitable exploitation. Theresult has been that over exploitation byman, coupled with predation andother naturalmortality, hasreduced thenesting population to2 percentofthe largest documented arribada recorded onfilm in1947 Hildebrand,1963!.Because ofthis decline, thelargest arribal ofeach season isno more than 200 to300 nesting females,andthe total number ofnesters perseason isno more than 800. Such low numbers emphasize thehigh vulnerabilityofthis endangered speciestoany change inambient physical en'ironment orto increases infishing effortof any kind that might result in the taking of Kemp's ridley. Nestingofnearly alladult females inthe Kemp's ridl eypopulation takesplace principally between thesandbar outletsofOstionales 3'24.6'N,97'45.7'W! tothe north and Barra del Tordo 3 3.5'N,9745.3'W! tothe south on the beachcalled Rancho Nuevo, inTamaulipas State,which hasbeen designated bydecree asa naturalreserve forKemp's ridley Departamentode Pesca,1977!. Asa characteristicbehavior ofthe Lepidochelys, thenesting ismassive andknown asarribazon orarribal. Nevertheless,it isdifficult to consider a nesting group of 200 females disperwM over several km of beach asan arribazon,Thelast small arribal comprising about2,000 females occurred inMay 1968. Ingeneral, Kemp's ridley's arribalesaredissimilar tothose ofthe Pacific orolive ridley L.olivacea! because theyoccur during theday, usually in the morning, Protectionofthe nesting beach toachieve recovery ofKemp's ridley was initiatod in1966 bythe Direccion General dePesca nowSecretaria dePesca!, through itsFisheries Research Institute. Theturtle camps, occupied eachnesting andhatching season, havecontinued without interruption tothis day. The efforts ofMexico were integrated with thoseofa U.S,assistance teamofuniversity students during 1978, andthis joint effort was continued for10 years. Themore important daily work at the turtle camp includes: 1. detectionand translocation ofnests and incubation ofeggs in protected areas; 'AfdrquezandVillanueva - Instituto Nacional dela Pesca, Mexico; Burchfield - Gladys Port~a Zoo

16 2. near the end of the nesting season,polystyrene foam boxes contaimng Padre Island sand are used to hold about three to five percent of the total egg production from Rancho Nuevo, to provide hatchlings for head starting and release;

3, taggingfemale adults, and recordingeach nesting female NESTING LOCALITIES and its nest; and 1 PADRE ISLAND, TEXAS 4. taking temperatureaf incubation of eggsin foamboxes 2 PLAYA LAURO VILLAR, TAMAULIPAS and temperatureprofile of the beachat severaldepths. 3 RANCHO NUEVO, TAMAUI.IPAS 4 CABO ROJO, VERACRUZ The general methodology of work originated at Tortuguero, 2 5 TECOLUTLA-NAUTLA, VERACRUZ CostaRica Marquez,1966!,and it wasmodified during theyears 6 VERACRUZ, VERACRUZ following MArquez,Villanueva and Contreras,1973!. It hasbecn 7 ALVARADO, VERACRUZ 6 MECDACAN-CHILTEPEC, TABASCO used as a model by several beach workers in Mexican turtle 9 ISLA AGUADA, CAMPECHE camps. 10 MAGDALENA, COLOMBIA NOT SNOwN! Nesting Distribution Nesting of Kemp's ridley is concentratedin spaceand time, This evolu tive behavior works as a mechanismregula ting popu- <4 lation size, and it is induced and affected by internal and external factorsthat are not clearly understood.It is not easy'to reacha diagnosis concerning population size,but despite of that, we can saythat suchnes ting behaviorhas been successful thus far,as the 6T 9 4 population still survives. Distributed among the arribalesare solitary nestingsof Kemp'sridley, and it is supposedby Robin- son in press! for olive ridley that such nests have higher survival rates than those from massive arribazones. Figure 1. NestirIglocalities of Ker1Ip'sridley seaturtle Geographical Dispersion 11vhi:>tcs5111all group ttestliFIgs!. Most Kemp's ridley nestingis restrictedto approximately32 km of sandybeach from Barradel Tordo northward Figure 1!. Outside of this area, there are sporadic and dispersed arrI'b~,"5of very few nesters; e.g., at Padre Island, Texas Werler, 1951;Carr, 1961;Adams, 1966,1974; Pritchard and MArquez, 1973;Francis, 1978;Hildebrand, 1982!, at Playa Lauro Villar, Washington Beach, Tamaulipas Mirquez, Villanueva and Sinchez, 1982; Mager, 1985!, in Tecolutla and Alvarado, Veracruz Mkrquez, 1983,1984; MArquez and Fritts, 1987!,in Mecoacan,Tabasco Mkrquez, 1970!, and in Isla Aguada, Campeche R. Gonzalez, Oficina de Pesca,Secretaria de Pesca,Subancuy, Campeche, Mexico, personal communication; MArquez, manuscript!. We are aware of only one record out of range, that being for the Caribbean coast of Magdalena, Colombia Chavez and Kaufmann, 1974;Meylan, 1982!.As far as we havebeen able to determine,there existsno historicaldocumentation of other Kemp's ridley nestingplaces as important as RanchoNuevo anywhere elsein the world. It is possible that nesting in the past, at least in Tecolutla, was more intense than it is now and that nesting at RanchoNuevo covered a wider areato the north and south as compared to the current nesting zone.

RanCho NuevO Beach Spatialdistribution of nestingat RanchoNuevo changes with the weather.When the wind blows strongly,nesting usually is concentrated in a more narrow area than on calm days. In geI;oral, all arribazanesstrike between Barra del Tordo and El Carrizo, but from time to time dispersion is wider, and a few animals lay north or south of this area.More than 80 percentof the arribazo11esstrike betweenBrasil and SanVicente. The mechanism of fixity of the nesting site has not yetbeenclarified. As in other seaturtle species,around 50percent of the individual Kemp's ridley renestbeyond 1.6 kilometers from their former nest during the same season,and less than 40 percent of the individuals renest beyond 1,6 kilometers of their former nest from seasonto season Marquez, manuscript!, The apparently greater fixity in nest site selection by the same individual from season to season,as contrastedwith that within the sameseason, has no clearexplana tion. Population Assessment Evaluation of population status for Kemp's ridley is more reliable than that for other speciesbecause nearly all Kemp's ridley nestingis at RanchoNuevo, and becausenearly all the turtle crawl tracksthat occurthere on a given day areobserved and recorded.The estimatednumber of nestersarId number of nestsrecorded during 1978through

17 d numberof Kemp'sridley seaturtle femalesnesting at RanchoNuevo as dete

Years 1978 1979 1980 1981 1982 1983 1984 1

959 1,013 927 946 821 823 892 723 763 699 713 619 620 672

l, 982! andRene Manluez Millan InstitutoNacional de ia Pesca,personal communicati erof nests per female per season was assume constant at1.326, but it isknown to change fr u script!.

Q 20 gp z

? V I- I- X 0Cl 0 10

0 Z X th 0I- ? 0 Z

0 0 1950 1960 1970 1980 YEAR Figure2,Empirical decline innumbers ofnesting Kemp's ndtey sea turtles long dasbes! and increa'ein numbers ofltetchlings produced and releasedatthe Rancho Nueoo beach ' represents numbers ofnesters estimated byCases-Andreu, 1971!.

1986are shown in Table1, Because the average number of nestslaid per fernale per season is 1,326 Marquez et al., 1982!,annual number of nesting females iseasily obtained. The historical tre~d in numbersof nesters and hatchlings producedfor the sameperiod is presentedin Figure 2, Acknowledgements Theauthors are grateful to theMexican and U.S. fisheries authorities for theirsupport of thework, to the people of RanchoNuevo, to theMexican Marines and turtle campstaff for their work,to theU,S, and Mexican student volunteers,to the Mexican Navy, and to the inspectors of the Delcgarion Federal de Pesca of Tamaulipas.We also expressour gratitude to lng. ErnestoCorripio, Director of theCentro Regional de InvestigacionesPesquera of Tampico.

18 Literature Cited Adains,D.E. 1966, More about the ridley operationin PadreIsland: egg transplanting. International Turtle and Society journal 1 I!:I8-20, 40-42, 45. . 1974.The sagaof a turtle namedAlpha. Tip-0-Texan9!;18-19, Carr, A, 1961.The ridley mysterytoday. Kingdom 64!:7-12. Carr, A, 1963.Panspecific reproductive convergence in Lepidochelyskern pi. Ergebnisseder Biologic 26298-303. Casas-Andreu,G. 1971.National and regionalreport: Mexico,Marine Turtles, Publicationof the InternationalUnion for the Conservationof Natureand Natural Resources,p. 4146,In: New Series,Supplement Paper No. 31,109 p. Chavez,H. andR. Kaufmann, 1974. Inforrnaci6n sobre la tortugamarina Lepidoehelys kempi Garman!, con referencia a unejemplar marcadoen Mexicoy observadoen Colombia.Bulletin of MarineScien< e 24!:372-377. Departamentode Pesca,1977. Acuerdo que establece la zonade refugioy vedapara la protecci6nde la tortugalora. Diario Oficial de la Federacion, 4 July 1977. Francis,K. 1978.Kemp's ridley seaturtle conservationprograms at SouthPadre Island, Texas, and RanchoNuevo, Tamaulipas, Mexico,p. 51-52.In: Henderson,G. E. Editor!, Proceedingsof the Florida and InterregianalCanference on SeaTurtles, Flarida Marine ResourcesPublication No. 33, 66 p. Hildebrand,H.H. 1963.Hallazgo del areade anidacionde la tortuga marina '1ora," Lepidochelyskempi Garman!,en la costa occidental del Golfo de Mexico. Ciencia Mexico! 22!:105-112. Hildebrand, H.H, 1982. A historical review of the status of sea turtle populations in the western Gulf of Mexico, p. 447453. In: Bjorndal,K. A. Editor!, Bt'ologyand Conseroation of Sea Turtles, Smithsonian Institution Press,Washington, D.C., 583 p. Mager,A., Jr. 1985.Five year statusreviews of seaturtles listed under the EndangeredSpecies Act of 1973.NOAA, NMFS,U.S. GovernmentPrinting Office,90 p. MarquezM., R. 1966.La cria artificial de la tortugablanca Cheloniamydas mydas Linnaeus! en Tartuguero, Costa Rica. Instituta Nacianalde InvestigacionesBiol6gica Pesqueras, Publicacion 13, 24 p. . 1970.Las tortugas marinas de Mexico.Mexico-Instituto Nacianal Investigaciones Biol6gico Pesqueras, Secretaria de Pesca unpublished paper!, 180 p. . 1976.Rescrvas naturales para la conservaci6n de las tortugas marinas de Mexico. Instituto Nacional de la Pesca,Serie Infarmativa No, 83,22 p. , 1978. Natural reserves for the conservation af marine turtles of Mexico, p. 56-60. In: Henderson, G. E. Editor! Proceedings of the Florida and Interregional Conferenceon SeaTurtle-., Florida Marine ResourcesPublication No. 33, 66 p. . 1983.Atlantic ridley project,1983: preliminary account. Marine Turtle Newsletter No. 26,p. 34. . 1984.The national report for the country of Mexico Gulf region, p.3i0-321. In: Bacon,P., F. Berry, K. Bjorndal, H. Hirth, L. Ogren and M. Weber Editors!, Proceedingsof the WesternAtlantic Turtle Symposium,Volume 3, 514 p. , Manuscript!,Synapsis of biologicaldata of the Kemp'sridley seaturtle Lepidochelyskempi!. FAO Fisheries Synopsis SAST/125. andT H.Frit ts. 1987.Prospeceion aerea para tortugas marinas en la costamexicana del Golfode Mexicoy Caribe,1982- '1983. Instituto Nacional de la Pesca, Mexico, CRIP-Manzanillo, Baletin Infarmativo 8:22-46. , D,Rios O.,J, M, SanchezP, and J, Diaz. 1989,Mexico's contributian to Kemp's rid icy seaturtle recovery. This volume! A. ViBanueva O. and J.L. Contreras. 1973.Instructivo para la proieccian de las tortugas marinas, Instituto Nacional de las Pesca,Serie Divulgacion 2:1-34. A, Villanueva O. and M. SanchezP. 1982.The populationof the Kemp's ridley seaturtle in the Gulf of Mexico- Lepidochelyskern pii p. 159-164.In: Bjorndal, K.A. Editor!, Biologyand Conseroation of Sea Turtles, Smithsonian Institution Press, Washington, D.C., 583 p. Meylan,A, 1982,Sea turtle migratio~- evidencefrom tag returns,p, 91-100.In: Bjorndal,K,A, Editor!, Biologyand Conservation of SeaTurtles, Smithsonian Institution Press,Washington, D.C., 583 p. Pritchard,P.C.H. 1969. Studies of the systematicsand reproductivecycles of the genusLepidochelys. Ph.D. Dissertation, University of Florida, Cainesville, xxi plus 196 p. Pritchard,P.C.H. and R. MarquezM. 1973.Kemp's ridley turtle or Atlantic ridley, Lepidochelyskempi, International Union for the Conservationof Natureand Natural Resources,Monograph 2, 30 p. Robinson,D. Las grandesarribadas

19 Questions and Answers

EdwardKlima: What do you think are the prospects for thatpopulation of nesters? Mkrquez:That is verydifficult to say.We have doubled the equipment as compared to whatwe had when webegan protectionofthe beach. The turtles take about 10 years to reach maturity, and we spent 10 years asking for equipment, Thenesting population decreased in thelast 10 years,but if weobtain the necessary equipment, we hopeto increase thepopulation. In the1985 season we had about 100 females less than in formeryears. If thatis a cyclicchange in the populationit couldmean the reduction is notdue to our situation.It mightalso mean some other worse situation for thepopulation, we do notknow exactly, But if nextyear the diminution in numberof nesterscontinues, that would meanthat the pressure on the beach and surrounding area and on the population is worsethan at the beginning of our work I really do not know. DavidForcucci: Regarding the decrease in number of adult females nesting on the Rancho Nuevo beach, from 1,000 to 2,000 in 1970down to what it isnow, is it theresult of mor tali ty ofthe adult females orthe hat."hli ngs frotn previous years? To what doyou attribute that decrease? , Mlrquez:I thinkit isa combinationof many things. The turtles take more than 10 years to reach maturity. We had no recruitmentfrom 1966to 1976.Beginning in 1976,we have observed recruitment of new adult femalesto thebeach. Despitethis, the total number of nesters per season continued todecline. Natural mortality and pressure from fishing continued.Since 1978, we have doubled the production of hatchlingsper year comp-red to the period from 1966 to 1977.From 1976 until now, perhaps the recruitment balanced the total mortality and the nesting population did not increase.Maybe after 10 years of thiskind of work, we will havesome improvenient in the size of nesting population, Thefishing pressure and various kinds of pollution remain, During the period when only the Mexicans protected the beach,we stoppedthat part of thedecline of thepopulation due to removalof eggs.After joint Mexican-U.S.work began,we greatly increased the production of hatchlings.We anticipate an increase in thenumber of nestersin the future. Forcucci: Doyou expectto havea big or gradualincrease in thenumber of nesters? Mlrquez:Not a big increase,bu t maybea slightincrease each year until anequilibrium is reached. RichardByles: What is happeningwith enforcement of the trawling ban off the nesting beach? I noticed that trawlers are still outthere during the season. Are there any plansf'or further enforcement ofthat trau ling ban? Mkrquez:The prohibition iswrit ten on paper, but enforcement isnot so easy. In 1986,we hope to stop the trawling in frontof thebeach. The Head of theFisheries Station in Tampicoisin contact with the Commandant ofthe Navy inthe Port of Tampico, and we hope to have some assis tance in en forcement from the Navy, by making patrols of the areaoff the beach. We will askthem to prevent the Mexican shrimp boats from trawling in thearea, at least between April and August when the nestingseason occurs. RodericMast: How many nests get left in situon the beach, and have you noticed any sign'ficant differences in hatch rate or otherdifferences among the in situ nests,the corral nests and the nests that are sent to PadreIsLand? Mkrquez:In 198S,we left 20nests in situ. Hadwe takencare of all 20in situ ne~tswe would have had a lowerhatch ratethan in corrals,but we made two mistakes. We did not avoid the coyotes at thebeginning and the middle incubationtimes. Also, if wehad put some stones around the small enclosures placed around the in situnests, we wouldhave saved a fewmore ha tchlings and perhaps would have shown a littlebetter hatch rate than in thecorrals. Wecan expect to lose more than 80 percent of thein situnests, if left in a naturalcondition without protection by enclosures.There are many coyotes and skunks as well asother small predators on thebeach, and there are cows walkingon the beach. Cows may squash the nests if theyput their feet on them. Then people are able to find the nests moreeasily the next day. But after one day, they will not be able to find the nests. Coyotes will take the nests during thefirst days of incubation, and near the end, because thenests have a strongodor when the eggs are freshly laid and whenthe hatchlings break out of the eggs and release the odor. At those times, coyotes are able to find and attack the nest more easily. BobWhistler: Are the clutches in the corrals monitored? Inother words, doyou have records foreach clutch concerning tohere it wasfoundand also the temperature of that spot? Do you keep such records? Mkrquez:Yes, we have, but someone else might provide an answer about incubation temperature, Burchfield:Each nest in a corralhas its own specific data sheet which stays with it constantlyfrom the pointat which it iscollected. Within the last 3 to4 yearswe have been looking very closely at nest temperatures within the corrals. Allthein situ nests were checked with thermal probes, aswere some of the corral nests, to tryand compare what the

20 two typesof nestswere doing. This is what! alludedto a littlewhile ago comparing thepotential thermal dynamics of closepacked nests resulting from anarribada with thermaldynamics of isolatednests. As bestwe canat theRancho Nuevo field stationwe arelooking at temperatureson naturalnests, in situ nestsand corralnests, as well asthose that areartificiaHy incubated in hatcherybuildings. We are looking at all of thosedifferent aspects of thetemperature effect.

21 to form a HEART Council, receive a certificate fmm HEART and have a larger heart on the walL Currently there are almost I00 HEART Councils. A". The financial support received from individuals, industries, and foundations reflects the tremendous interest in sea turtles and the concernfor their survival. We have had great support from EXXON CompanyU.S.A., Mr. Earl Burke of Pel-TexOil Company,The Harris and ElizaKempner Fund of Galveston,the SER-Jobsfor Progressoffice in Galveston,DeMets Turtle Candy Save the Turtle Fund and Dr. Joe Flanagan and his staff of the Houston Zoo. One of us Al Barr! spenttime at RanchoNuevo during four summers. He has alsobeen in charge of the HEART exhibit at the Marine Education Symposium at Texas Agdvi University. His environmental scienceclasses from Westfieid High School,Spring IndependentSchool District, Harris County,Tex., have assisted with HEART's Annual Open House held on or about St. Valen- tine's Day at the head start facilitiesin Galveston. Oneofthemostimportantjobs we dois publicizing the Kemp's ridley head start project and the recoveryprogram, Publicity increasespublic awarenessand gives us the opportunity to communicateand educate.Our annual Open House brings hundreds of visitors from the Gulf coast area to see the turtles and rearing facilities. It was one of HEART's Open Houses that inspiredPamela Phillips to write her entertainingand informa- tivebook, The Great Ridley Rescue 989!. Copiesof thebook may FigureT. Daphnis Hernandez, 9-yen~ studentofSpring beordered directly from HEART. We constantly seek media Inde@ ndentSchooloistrict,ffarrisCountY,Taxand Junior coverageof major eventssuch as the annual releaseof tagged rnemb~ ofHEART, preparing Raisin' Ridley" cookies noie turtlesinto theGulf of Mexico.Within thelast fewyears, the head start projecthas received publicity on televisionChannels 2, 8, I I, I3 and 26in Houston,and alsoon ABC,CNN and CBSnationally. An articlein thechildren's publication of the NationalWildlife Federationin November1984 brought hundreds of inquiries from students, librarians, teachers and others in 40 states, Canada, Venezuela, India and Singapore.Galveston's Mayor, Janice Coggeshall, pmclaimed the City's first 'Turtle Week"in April 1985,providing more newspapercoverage. HEART committee members give presentationsin the Houstonarea using a slideshow, and we mail copiesof 'The HeartbreakTurtle," a one-hourvideo feature story producedby Houston'spublic televisionstation, on loanout of state.We also work with studentsin thepreparation of reportsabout turtles, We encouragepeople to write their congressmenin Washington, D.C., about the needfor continuedprotection and conservationof seaturtles. For example,the ChicagoHerpetological Society asked its 600members to write their legislatorsto gain support for theKemp's ridley recovery program, HEART worked closely with otherconservation organizationsto lobbyfor recent reauthorization of theEndangered Species Act and against weakening amendments to the act. Aswe look ahead, we see many opportunities for HEART work and hard work. Continuation of theentire Kemp's ridleyrecovery program and expansion of thehead starting phase are needed, Since completion of thenew rearing facilityat Galveston, about 2,100 ridleys can now be raised and released each year, We will continueto stressthat the Kemp'sridley recoveryprogram is an excellentexample of cooperationbetween the United States and its dose neighborMexico. It is ourbelief that your income tax dollars are wc' spentwherever work is going on for thebenefit of ridleys,We support captive breeding of Kemp'sridley, and will workwith zoos,marine aquaria, universities, and otherorganizations and agencies where Kemp's ridleys are being kept for thatpurpose. We urge that strict controls be maintainedto ensurethe welfare of seaturtles being held by suchfacilities around the country and in foreign countries,We recommendfrequent contacts and inspectionsas often as possible. Asmore hatchlings are released and as the turtles reach maturity in thewild, it isimperative that visitors to possible nestingsites along the Gulf coastbewell informedabout ridleys, More clean,quiet, restricted beach areas are needed to supportsea turtle nesting.Turtle Excluder Devices must be usedin the shrimpingindustry so thataccidental catchesand kills arereduced to a minimumas quickly as possible. Thereis muchto be done for Kemp'sridley sea turtles, and HEART pledges continued efforts, We maybe very close to moreexciting accomplishments toward the recovery of ridleys,or wemay need to continueindefinitely to assure their survival. With patience,determination, and increasedpublic support,both hereand in Mexico,we can have success,Let's go for it!

Literature Cited Owens,D,, D, Crowell, G. Dienberg, M. Grassman, S.McCain, Y. Morris, N. Schwantes and T. Wibbels Editors!. 1983. Wester' Gulf ofMexico Sea Turtle Wort shnp proceeÃings. Texas A8zM University, Sea Grant Collie Program,College Station, TX 77843- 3258,TAMU-SG-84-105, 74 p. Phillips,P. 1989. The Great Ridley Rescue. Mountain Press Publishing Company, Mis~oula, Montana, 180 p.

24 The Role of Sea Turtle Incorporated in Kelp's Ridley Sea Turtle Conservation and Public Awareness Ila IVI. I.oetscher~

After the project to estaMisha newnesting population for Kemp'srid leysea turf le Lepidochelys kern pi! wasrelocated to fhe Padre?slandNational Seashorein the mid-1970s,Sea Turtle Incorporated STI! wasable to put greateremphasis on its public awarenessprograms while continuing to assistin conservationmeasures to protectand propagatethe critically endangered Kemp'sridley. Beginning in 1980,this non-profitcorporation inifiafed a breedingprogram for captiveridleys ust'ng ifs female, Little Fox, Twice-monthlyshows plus shows by reservation are given to acquaintfhe pri vale sector wi th theproblems and needs of Kemp's ridleyand of her sea turtles of theGulf of Mexico. Brochures are published by STl, and slide presentations have been produced for schoolgrades 1-12 and for adultaudiences. A demonstrationof the typical turtle showwas given at thesymposium. Attendancerecords at fheshows and expressed public attitudes have demons! rated an overwhelminglysupportive response. This, in turn, hasallowed STl to channelfunds, when needed,foulard the support of on-goingstudies by Dr. David Owens, Departmentof Biology,Texas A&M University,and by Dr, HenryHildebrand, Corpus Christi, Tex. and to support experimental breeding programs. Our workhasreceived vast international and national coveragein newspapers, regularly published magazines and professional publications,

The work of SeaTurtle Incorporated STI! actuallybegan in 1965when I went with Dearl Adams to RanchoNuevo, Mexico,and brought Kemp'sridley seaturtle Lepidochelyskempi! eggs back to SouthPadre Island, Texas. We did not know if seaturtles eggscould be movedand successfullyhatched, but we were determinedto find out. We brought the eggs,including the mucousand sandfrom wherethey had beenlaid, to SouthPadre Island and buried themon our beach, duplicating the natural nest as exactly as we could. The first step worked,the eggs hatched! From 2,000eggs we got 1,102 babies off to sea.We were so excited that we invited the entire community to seethe turtles hatching and crawling off. It was so early in the morning and the lights of the cameramenwere sobright that they seemedto confusethe little ones,so somehatchlings were getting badly disoriented.From this we got the feelingthat it was the stronglight of the sun that setsthe coursefor newly hatchedridleys. In the excitement, two of the14-gram preciousbabies were stepped on. Another was too weak to crawl to sea,These three were given to me to care for and to begin compiling informa tion as to their growth patterns, diet, and habits, We named the turtlesWynken, Blynkenand Nod. This trio so won our hearts that we were caught hook, line, and sinker!We have never regretted that first step,Thebenefits and learning experiences have proven boundless. So very little was known about seaturtles at that time, Dr. Archie Carr's first books on the subject were just being published, The libraries, scientists and teachershad very few facts to share with us, but with three loving turtles to coax us on our way, it was not hard to dedicate our hearts and time to bet ter the situation for the declining populations of sea turtles. Dr. Henry Hildebrand was a frequentvisitor, as were folks who came for fun and sun, Noticing the growing interest asmore and more people stopped by our turtle kraals to seethe turtles fed and to watch their interaction with humans, Dr, Hildebrand suggestedwe organizea programto acquaintthe generalpublic with thesemarine animals. Most of the inland touristshad neverseen a turtlethat did not haveclawed feet and could not pull its headinto its shell.And so our program, "Meet the Turtles," began and has grown and grown and grown, We quickly learnedthat "Meet theTurtles" gavethe schoolsa greatfieM trip. We realizedthat teacherswanted to get over being afraid of the turtles,so we put Atlantic greens Cheloniarnydas!, rldleys, and hawksbiils Eretmochelys imbricata!into rolesof little boys and girls, gavethem names, and dressedthem in original costumesfor our shows. Not even the most timid child could be afraid of Geraldinein flowing chiffon with her wig tipped over her beak. None of them forgot JonathanLivingston Sea Turtle, our TexasA@M "Aggie," wearinghis school'smaroon and white colors and applaudinghis team,To children,the turtlesbecame real people.Geraldine, Jonathan, and many otherscreated

" SeaTurtle Incorporated charactersthe children loved and held dear, Nowadays, we never see a childwho hangs back, afraid to getclose to our flippered crew. To them a seaturtle is a loving friend, "Meet the Turtles" is now scheduled twice eachweek. More than4,400 people attended our showsduring the summer of 1985.Winter attendance is evenhigher. From the first days when we servedpunch and cookiesand madecostumes for the turtles,our involvementhas expanded,During the years,we found that someincome was neededto supportour program,so SeaTurtle Incorporatedwas established as a non-profitcorporation in the Stateof Texas.Once we startedaccepting donations, we foundit waspossible not only to enlargeand improve the turtle housing facilities, but alsoto respondto needs for moniesfor scientific equipmentand research.In addition to other projects,we were involved with artificial inseminationprobes and an electro-ejaculator designed and made by CarolPlatts. This equipment is nowon loanto theGladys Porter Zoo, Brownsville, Tex, We were able, also, to giveHonda ATVs All TerrainVehicles! to theridley recoveryteam for its useon thebeachat Rancho Nuevo. We havebeen able to awardsmall grants to variousscientific programs.It is with deepsatisfaction that we areable to assistin manyways, making possible studies that are daily ensuring the survival of sea turtles. Fromour small beginnings with the local press, we have seen that the story of therid leys and our work has appeared in 56different magazines, including National Geographic, World, Southern Livi»g, People, Texas Parks & Wildlife,Texas Highioays,practically all of themajor newspapers nationwide, and even the National Enquirer. Television has given usfantastic coverage ABC, CBS, and NBC-on "Today," "Tonight," "Real People," "PM Magazine," "Eyes of Texas," "BelieveIt Or Not," andinnumerable other programs. The story has been aired in Japan,France, and Great Britain by theirTV companies.A largequantity of mail is receivedconstantly, all tremendouslysupportive. Fishermenand beachgoers alike are aware of our seaturtle rehabilitation program. When they find a turtlein distress,they make every effort to get the word to us so that we, with the help of theveterinarian atthe Gladys Porter Zoo,can give the turtle every chance to recover. One of ourdisabled turtles has spent his days happily with usfor more thannine years, He is a tremendoushelp in making children aware of the awful consequences suffered by turtles and othermarine animals when people thoughtlessly use the ocean as a garbagedump, Our visitors see turtles that are bobbingalong with only threeflippers, turtles whose growth has been stunted after havingbeen mixed up in "blobs" of tar,and those that are severely handicapped bygenetic defects leaving them in needof lifelongcare in captivity. Throughthis exposure, youngsters and adults aregiven a lastingimpression, a desireto protectcreatures in thewild, anda deeprespect for theseanimals that have been meinbers of earth'slife chainfor solong, In 1980,we initiated an experimental breeding program for captiveridleys, With theadvice of Dr.David Owens, Departmentof Biology,Texas A&M University, and the cooperation and encouragement of John Kerivan, Sea-Arama Marineworld,Galveston, Tex,, we took our then 9-year-old female, Little Fox, to Sea-Arama. The following year, we returnedher to SouthPadre and brought malesfrom Sea-Aramato be with her in SouthPadre, and in 1983sent her to MiamiSeaquarium. She was returned to usin 1986.We have a inale nine years old in 1985!,Dr. Porter,who we are hopingwill bea usefuladdition to ourbrood stock. Using captive-reared animals, we think it maybe possible to "milk" themales and artificially inseminate the females. Our plans are on-going as we pursue the best ways to continuethis experimen talbreeding program, perfecting artificial insemina tion and developing methods to store and shipsea turtle semen. Dr, David Owens, Steve Rabalais of LouisianaUniversit i'm Research Consortium, Chauvin, La,, andTim Bentley ofthe University ofMiami, Roscns tielSchool ofMarine and Atmospheric Science, have been very supportivein thisprogram, To date we have not achieved any of ourgoals, but we are not about to give up WewiII continue to supportthe breeding program, continue our intensive public awareness campaign, continue to giveassistance to the many scientific studies, and always provide a safehaven for disabledor ill seaturtles in need of our help.

26

thatthe endangered Kexnp's ridley sea turtle Lepidochelys kern~!was being bxought toextinction inthe Gulf of Mexico by shrimpers. nmentalists'expressions ofconcern didnot go unheeded bythe shrimp industzy. Shortly after the public hearing,keyrepresentatives ofthe shrimp industry metin Houston, Tex.,to seek a solution tothe problem. Based on thesediscussions, imxnediate actionwas taken tosurvey theshrimp harvesting industxy tolearn the extent ofthe problemandto prepare educational materials fordistribution tocaptains. A survey byDr. Bruce Cox, then a Texas SeaGrant County Marine Extension Agent inthe Brownsviiie-Port Isabelarea and Mr. Bob Mauermann, then ExecutiveDirector ofthe Texas Shrimp Association QSA!, was used asa modelbyother states Cox and Mauermann, 1976!.Itwas agreed thatneutral parties such asSea Grant Marine Extension Agentsbe used toconduct thesurvey. Asfor educational materials, itwas recommended thata smallposter, usingas a guidea memorandum prepaidby EuclidLewis ofGeorgia andentitled "When It'sTurtle Time, beprepared anddistributed toall shrimp vessel captains.Thebottom line objective ofthe shrimp industry atthat time was to make a good faith effort toensune that seaturtles and shrimp harvesting activities couldco-exist. Itappeared thatthe environmental community wasnot tryingtoput the shrimpers outof business,but wasexpressing deepconcern. Onesuch concern wasexpnessed byDr. ArchieCazr, professor ofzoology, University ofFlorida, GainesviQe, whoestimated thatthe nesting amvals of Kemp'sridley at Rancho Nuevo had dropped from at least 40,000 in 1947 to 1~ in 1974. Shrimpindustxy representatives remainedquite active inmoving thebrewing crisis into a problem-solvingmode. Twoissues needed tobe absolved. Thefirst was solving the jurisdictional problem between theNational Marine FisheriesService NMFS! and the Fish and Wildlife Service FWS!. Inthe spring of1977, ofAcials ofthe FWS initiated andlatez' formalized actionsthat gave NMFS, through theDepartxnent ofCommerce, xesponsibiQty forall activities, regulationsandcontrols over sea turtles while they axe in the water. FWS, through theDepartment ofthe Interior, retainedcontrol ofsea turtles while they are on land. The second issue was one of information. Duringthis same period,NMFS, with encourageznent bythe shrixnp industzy wassetting aside xnonies forresearch. Theshximp industrystxongly supported NMFS'efforts towardadvancement oftechnology whilecontinuing toensure thatthose causativeagents more difficult toassess such as habitat destruction, lossesto predators, illegal trade, and interferencebyother human activities would not be ignored. Somemembers ofthe shrimp industxy continued to makelight of the crisis, but this attitude was changing. Theseriousness ofthe situation wasrealized inthe summer of1979 when a numberofmutilated seaturtle carcasses washedashore onthe Padre Island National Seashore nearCorpus Christi, Tex.Immediately, thefinger was pointed atthe shrimp industxy based primarily oncircumstantial evidence.The Director ofNMFS'Southeast Regionmetwith shximpersfromAransas Pass,Tex., and warned themthata continued highrate of turtle mortality mightmean closing Gulfwaters toshrimping. Therewere a few cases inwhich crewmen onboard shrimp vessels wexeaxrested andjailed forillegally taking seaturtles. Word ofthese actions spread across thewaterfront, highlighted bythe large bailbonds set in some cases. Usinginformation supplied byenfon:ement agents,TSA distributed notices tovessel owners whose vessels had beenidentified asworking inthe south Texas area. These notices advised shrimp boat operators ofthe state-of-the- axtprecautions tobe taken inareas through which sea turtles migrate. lazeresponse tothese notices was excellent. 'I1zroughnumerous meetings andother means ofcommunication, variouscourses ofaction were developed via jointefforts bythe industry, theenvironznental communityand government agencies.The shrimp industry wasas activeaspossible inthose areas forwhich itwas zesponsible. Assistance wasgiven tothe development, organization andconduct ofthe World Conference onSea Turtles heldin Washington, D.C.,inNovember 1979.During thissame period,theindustry actively assisted NMFS indevelopment ofa Turtle Excluder Device TED!, later to be called TrawlingEfficiency Device,Regional tradeassociations formeda conduit forfinancial reimbursement toindustzy operativeswhosuffered losses inshrimpcatch duetoutilization ofthe TED. Weencouraged developmentof a NMFS seaturtle poster thatwould allow better identification ofthe different species ofsea turtles byanyone coming in contactwith them. The shrixnp industry also joined inthe fight in Washington, D.C.,to maintain monies forTED. developmentandfor the NMFS Southeast Fisheries Center's Galveston Laboratozy whereKemp's ridley head startingiscarried out. Addihonal money from industry trade associations, theGulf and South Atlantic Fisheries DevelopmentBaundation aprivate corporation established bycommercial fishingmterests inthe southeast region! andNMFS was funneled toward thevarious other objectives ofprotecting seaturtles. As~izznan of the committee toincorporate TEDsinto the trawl fishery, I canreport thattheprocess lsworking. NMFSscientists andtechnicians haveadvanced thetechnology forexcluding turtlesfmm trawls toa pointatwhich, withvarious areawpecific modifications, itis adaptable tothe fishery. Thevoluntary program ofincorporating this technologyhasproven tobethe best course ofaction, inthat human natu' and especially ffshezznen's naturerespond moxefavorably toa want tothan toa havetomode. Also, cows are better educated tothe need forprotection and pxoperresuscitation ofsea turtles they happen tocatch incidentally while trawling ior shrimp. Theshrimp industry has come a longway in the past fourteen years toward meeting thechallenge ofsea turtle protection.Nolonger will laughs be generated among fishermen bymention ofsea turtle protection inthe context of itsrelation tothe livelihood ofthe shrunp industry. It is incumbent upon us aII not only to maintain the xnomentum wecurrently have, but also to focus on other user groups such as xecrea tional Interests andthe offshore petroleum industry,toensure that they too are aware ofthe plight ofsea turtles, sothat they may refrain from harming these Important seacreatures, In finalanalysis, something else equal in importancetothe rehabilitation ofsea turtle stocks has been achieved throughourexperiences withthis conservation effort,Two groups, initially onopposite sides, have found a wayto meetan emotionally charged issue head on and to develop a strategy toward achieving thegoals ofboth groups. This isnot to say that the problem issolved, but that we have made a significantstepin that direction together. Literature Cited Cox,B.A. and R.C. Mauennann. 1976.Incidental catchand disposition ofsea turtles bythe Brownsville - PortIsabel gulf shrimp fleet.Cameron County Extension Service, SanBenito, TX,and Texas Shrimp Association, Brownsville, TX,4 p.

29 TED TrawlingKfficiency Device Turtle ExcluderDevice!: Promoting Its Use WilberR. Seideland Charles A. Oravetz' A TrawlingEfficiencyDevice TED!, or Turtle Excluder Device,hasbeen developed torelease captured seaturtlesfrom shrimp associatedtrawls.Thewithincidental trawling takeisofessentially sea t urtles eliminated.inshrimp trawIsis reduced bymore than 97 percent, sosea turtle mortality otherunse Besidesconserving endangeredand threatened sea turtles, the TED provides benefitstoshrimp fishermen. Unwantedby-catch isdecreased considerably,th us reducing the amount oflabor in sorting thecatch andimproving qualityofthe shrimp harvested. decreaseCannonballtowingj ellyfishtime, Stomolophusare eliminated meleagris!,sothat towswhich of longer sometimes duration occur caninbesuch made, large Finfishnumbers in thethat by-catchthey clog fromshrimp shrimping trawls andare releasedatrates exceeding 50percent atnight and 60 percent duringdaytime, Overall,the reduction ofby-catch usually results ina slightincrease inshrimp catch, Because ofits ability toimprove thequality andefficiency ofshrimp catchin addition to conservingseaturtles, theTED isnow more appropriately describedas the Trawling Efficiency Device. TheNational MarineFisheries Serviceiscond ucting a technology transferprogram aimedateducating shrimpfishermen and shrimpdemonstratingfishermento inthem all shrimpingthe benefits statesthatcanon the be Atlanticrealized byandusingGulfofkfexico the TED. Thiscoasts promotionalof the U.S. activityinan effort hasintroduced to achieve voluntaryTEDsto useofthe TED, TheTED also hasbeen demonstrated inseveral foreign countries, Development workisbeing conducted ona smallerTEDfor shrimp netsnormally usedininshore waters,If it is effective, thesmall TED could playa significant rolein protectingjuvenile Kemp's ridleys Lepidochelys keinpUand other sea turtle species ininshore waters. TheTED Turtle Excluder Device! wasdeveloped bythe NaHonal Marine Fisheries Service NMFS! todecrease mortalityofsea turtles caught inshrimp trawls. Extensive testingin trawls hasdemonstrated thatthe TED releases morethan 97 percent ofthe sea turtles captured incidentally andessentially eliminates trawl-related seaturtle mortality,Atthe same time, the TED does notcause shrimp lossand may infact produce a small increase inthe catch of shrimpAYatson, Mitchell and Shah, 1986!. Inaddition toprotecting endangered andthreatened seaturtles success fully,the TED can provide positive benefits toshrimp fishermen whouse the device, bymaking shrunp trawling moreefficient. TEDscan effectively reducethe by-catchofcannonball jellyfish Stomolophus meIeagris!,sponges, horseshoe crabs Limulus polyphemus!, seaturtles, sharks,raysand bony finfish. Finfish reduction ratesfor trawls withTEDs exceed 50percent atnight and 60percent duringthedaytime Watson etaL, 1986!. Thelarge reduction inbywatch allows longer towing times, reduces labor requiredtosort shrimp from the by~tch, and improves quality ofthe shrimp catch byreducing amounts ofbroken ordamaged shrimp normally assodated withlong tows and by-catch. Shrimpfishermen havebeen urged touse TEDs voluntarily througha technology transfereffortaimed ateducating themand demonstrating tothem the operational andeconomic benchis that can be realized fromTEDs. TED isnow commonlyknown asthe Trawling Efficiency Device because it hasbroad benefits beyond theconservation ofsea turtles. Description of the TED Theoriginal TEDdesign hasbeen improved significantly. Thelatest design islight, easy tohandle andcollapsible. TEDoriginally wasconstructed ofsteel pipe and measured 08m high by0.9 m longby13 m wide.Itweighed 44kg. Theoriginal TED had good performance characteristics, butits weight and handling requirements wereless than desirabletofishermen whotried the device. Datacollected since1978 on the capture ofsea turtles inshrimp trawls wereanalyzed tosee if size ofthe TED could bereduced. The0,9 m widthofthe escape dooris sufficient toprovide a turtlerelease rateapproaching 100percent. However, theanalysis showed further that ifthe width were reduced 15cm, more than 95percent ofthe captured seaturtles stillwould beaHowed toescape because lessthan 5 percent ofturtles captured hada carapace widthgreater than03 m. Even some ofthe turtles witha carapacewidthgreater than0.8m wouldescape ifthey could move andturn inside theTED. With the escape doorreduced toa 08 m width,

'NationalMarine Fisheries Service thepvera5 width of theTED could be reduced to I.i m;The length of theTED was retained at 0.9mb u~ it i necessarytpmaintain the angle of the deflector bars at less than 45 degrees from the hpriapntai1 tpensur ~~ pperatipn- Thesecond major change in TEDdesign to encourage itsacceptance byfishermen was that of making itcoHapsible, ThecoHapsible TED has been thoroughly tested on commercialshrimp boats including those towing four nets or ad rigs,It takesup muchless space when on deck, is considerably easier to handle and is very stable during fishing.' 1viakgtheTEDcoU pslbl rempmmuchpfthest 1st ct ~mat rialf mit f me.Th~chang~~l~i a steelcollapsible model of theTED that weighed only IS kg . Fiberglassand other plastic materials also were investigated todetermine their effectiveness. I|berglass was found tpbe the best pf thesehght and strong materials, and the collapsible TED made of fiberglassweighed only II kg. rgiass,however, seems iobe a littleless durable than steel and its use required more labor in construction p thef ~D problemsa]so were encountered in obtaining a consistent,inexpensive supply of fiberglassrpd material for TED construction.For these reasons, the steel coHapsible model of theTED is recommended. TechnologyTransfer Approach TEDtechnology currently is being transfexxed to the shrimp industry. Through our earlier efforts we attempted to e, on a voluntarybasis, a significantlevel of TEDusage by theshrimp industry. It washoped this couldbe cdby makingshrixnp fishermen aware of directbenefits the use of TEDs can provide thexn. Technology transfer work is directedat demonstratingthe ability of TEDsto removeunwanted parts of the by~tch and to improve efficiency. particularlytroublesome toshrimp fishermen in variousalas of thesoutheast are cannonbaHjeHyfish, smaH finfish andlarge objects such as skates, rays, loggerhead sponges, etc. The spacing between deflector bars in theTED normaHyis15 cm. When heavy concentrations ofjeHyfish, large fish and bulky objects hke sponges are encountered, theycan be ejected by reducing the distancebetween deflector bars, This is usually accomplished byadding an insert thatxeduces this distance to about8 cm.Such objects are then vexy effectively xemoved from theby~tch by theTED. Finfishescapement is achieved by addinga finfishdeflector grid behindthe maindeflector baxs together with openingsin thewebbing around the TED. Webbing panels help guide fish out of thetrawL Educationpxograms, workshops and at-sea demonstrations are being conductedto describeand promotethe benefitsof TED,especiaHy that pf reductionin by-catch.Jellyfish and flnfish frequently axe so abundant on some shrimpinggrounds that shrimping operations either must be severely curtafled or occasipnaHystopped altogether. Redudngby~tch of suchitems in shrimptrawls with TEDs,particularly during periods when they are in heavy concentrations,represents a significanteconomic incentive to fishermenwho canmake longer tows under othmvise adverseconditions. There also is a significantreduction in labor required to separateshrimp from the bymtch. Anotheradvantage of TEDuse usuaHy isa S to7percent increase in shrimpcatch as compaxedto that in a standard txiwL Although we have not beenable to comeup with a completeexplanation for this result, it probablyoccurs becausereduction inby~tch lessensthe weight of total catchin thetrawl, thusaIhwing a greaterspread in themouth of the trawL This benefit,however, is npt highlighted during technologytransfer activities. Many vaxiablesoccur during shrimping that could affect or reducethis benefit, so the increasein shrimp catchis leA for shrimpersto determinefor themselves.However, it is stressedthat if TEDsare usedthey wiH not causeany shrimp loss. Results ok TED Technology Approach Key elementsin the NMFS TED technologytransfer pxogramhave been support of the shximpindustxy and partidpathn of theSea Grant programs in selectiveplacement of TEDson shrimp vesselsin variousshrimping axeas. Someshrimp industryleaders have fuiiy supportedthe useof TED and havestrongly urged shrimpersto useTED. SeaGrant, partidpa ting thxoughits marineadvisory progxam, has publicized and encouraged TEDusein many ways. During I985,SosteelcoHapsible TBDs were given to selectedflshermen in all thesoutheast coastal states of theU9. In eachcase, one or twoTEDs were plaid ona boat,and a demonstrationfishing trip wasconducted to ensurethat the vesselcaptain understood how tp install TEDsin shrimp txawlsand to use them properly.Construction demonstrationsalso were conducted at several net shops to stunulatecommercial manufacture of TEDsso fishexxxien wpuMhave access to placeswhere they couM pun.hase TEDs. At thepresent time, it is diffkuit to state precisely how many TEDs are in usebecause of thedive~. hidependent andmobile nature of shrimpingoperations. The foHowing are brief descriptions of knownTED use-': Domestic Use "OMdesign TEDa96 in number!were constructed and distributedunder contract to 13escoMarine pf St. Augustine,Fla. Their level of usevaried, but theywere used primarily during periods of heavyby~tch

31 concentrations.Themajority ofthese TEDs were distributed onthe Atlantic coast o thef southeastern Ug, 2.Under ajoint Alabama SeaGrant andSaltonstaH-Kennedy project,adozen orso TH3s wexebuiltby twodifferent netmakers inAlabama andwere distributed tonorthern Gulfshrimpers. 3. Duetoworkshops anddirect construction assistance toindividual shrimpers, about45 TEDs have been built. 4.New coHapsible TEDs0in number! wexeconstructed andinstaHed onshrimp boats incoastal statesfmm North Carolinato Texasduring the summer of 1985. 5.A significant numberof TED-like deviceshavebeen builtinsome shrimping areas,principaHy tosolve by-catch problems.Several hundred areused periodically inLouisiana forfinfish by-catch reduction. Mostof these axe designedasjeHyfish eliminators andwill probably ejectsea turtles ifthe eliminators areof sufficient size. ForeignUse 1.A TEDworkshop washeld inJune 1984 inTampico, Mexico. Results werenot very good. Thevessel could not goto sea for a checkoutduringthe workshop, andan old design TED constructed fromaluminum wasused. The TEDsbecame bentduring subsequent use.Aluminum isno longer used asa constructionmaterial. 2.In Indonesia, moxethan 1ANO TEDsare in use inthe western partof the country onjoint-venture Japanesevessels. Indonesiangeartxainees havebeen sent tothe NMFS Harvesting Technology Division,Mississippi Laboratories onthree occasions tohelp them stay cuxrent inTED developments. Thelatest trainee wasin Pascagoula, Miss. betweenJuly1 andSeptember 30,1985, tolearn how to build the collapsible TED. Future Direction Continuedwork on TED will focus on threeareas: 1. continuationoftechnology transfer toindividual shrimpers; 2. testingof a smallmodel TED for use in inshore waters; and 3. potential TED manufacturexs. Thecontinuation ofvessel-based demonstrations, workshopsand publicity activities isnecessary over term.Amajor purpose ofwork onthe smaH TEDis protection ofjuvenile Kemp's ridleys l' epidochelys knnpilan seaturtles ininshore waters. Thesize ofthe small TED was chosen toaccommodate juvenileridleys an compatiblewithsmaHer shrimp trawls used ininshon fisheries. Theeffectiveness ofa smaller TED in red jeHyfish,finfishand otherby-catch hasyet tobe determined. Ifresults are comparable tothose obtained with TED,amajor push wiHbe made tointroduce smaHTEDs. Emphasis onstimulating commercial production isnecessary toensu~hat TEDsaxe readily avaQable ata reasonable price.We in the NMFS feelthat TED tech transferisthe number onepriority itemfor sea turtle conservation inthe southeast xegioxL Literature Cited

Watson,JW,,JS.FisheticsRcmno Mitchell48 !:and 1AK 9. Shah. 1986.Trawlingefficiencydevice: anewconcept forselective shrimptrawl>88ear

32

Port of Cameronin southwesternLouisiana developed a devicesimilar in design to the NMFS TED Figures1-3!. However,aluminum wasused for the framework,and a slit in the webbingof the trawl throat wasutilized to allow escapementof by-catch in lieu of usinga hingeddoor as in theNMFS TED!. The weight of theCameron excluder rangedfrom 8 to11 kg, whichwas considerably lighter than the original NMFS TED. This device, placed into the throat of the trawl, was developedprimarily to excludejellyfish, which are numerousthroughout southwestern Louisiana in late summer, fail and winter Figure 2!. Turtle exclusionis a secondarybenefit of the Cameron Aluminum Excluder CAE!. In no casedid fishermenreport that turtles werecaught while fishing with this device, Most fishermendo not want the hassleof handling large seaturtles on their boatsnor the risk of exposureto large finesor prisonterms associated with thewillful hariningof anendangered species. A smallerCAE design Figure 1!also was developed for use in 5 mtest trawls try nets!and butterfly nets that are used extensively in theLouisiana shrimp fishery. PerformanceComparisons Tocompare the NMFS TED and CAE, field tests were conducted in June1983 on thedouble-rigged, shrimp vessel GAMBLER off the coast of Cameron. Onboard handling and fishing performance of the NMFS TED and CAE were comparedby fishingone of eachtype of excluderdevice on oppositesides of thevessel. Both excluders performed well in excludingjellyfish. However,weight of the NMFSTED seemedto causesome additional handling problems whencompared to thatof theCAE. Through these field tests it wasdetermined by onboardNMFS personnel that the CAE would work well as a turtle excluder in addition to its primary use as a jellyfish excluder, NMFS Collapsible TED In 1984,NMFS introduceda muchlighter collapsibleversion of the TEDthat weighsless than 9 kg,which is a lot less cumbersome than the old steel TED first released for testing in 1983.The additional feature of fish exclusion developedby NMFS was provided to give the shrimp industry an incentiveto use this TED. To date Louisiana shrimpershave not voluntarily usedthe collapsibleTED to any greatextent.

Condusiom Marine advisory agentsthroughout Louisianahave conductedworkshops in severalcoastal communities to highlight advantagesof the NMFScollapsible TED and CAE. Informationon constructionof both excluderdevices hasbeen distributed to thousandsof fishermen,Use of the CAE throughout Louisianahas varied dependingon cannonballjellyfish densities;however, one port had as much as 75 to 80 percentusage during peaksof jellyfish occurrence in 1985 and 1986. As a MarineAdvisory agentworking with commercialshrinipers out of the Port of Cameron,La., I havehad the opportunity to conductshrimp retentiondemonstration on at leastseven shrimp vessels,These vessels range in size from 14m togreater than 60 m, TEDstested included the NMFS TED, the Georgia Jumper, the CAE and the Morrison" Soft TED. In all cases,shrimp loss was experienced, and it increased with increasein by-catch volume. It is my opinion that areaswith greaterby-catch, such as jellyfish, jellyballs,small finfish, blue crabs,etc., will experiencegreater shrimp lossdue to an increasedvolume of by-catchbeing excluded from the trawl. This increasedvolume being excluded apparently carries additional shrimp out of the net and, therefore, increasesshrimp loss.Shrimp loss seemsto increase with decreasein by-catch, and I assume it is becausethe excluder hole or door is left in closed position, reducing the opportunity for shrunp to be lost, Louisianaexperiences times during theyear when by-catch is a tremendousproblem, The greatest problem seems to be with jellyfish and/or jellyballs.During the summerand fall of 1987,a largepopulation of jellyfish wasnoted, One vesseltesting the shrimp retentionof a "Morrison"Soft TED experienced19.7 percent shrimp loss in heavy jellyfish concentrations. This compares to a 4.6percent shrimp loss on another vessel pulling in areas of low jellyfish concentra tions. TED testingin areasof high by-catch,especially jelIyfish, are neededto assurethat commercialfishermen choose a devicethat will causethe least amount of economic hardship. The shrimp fishery in Louisiana is a by-catcharea and will experience shrimp loss with high bycatch using TEDs currently certified. It is my hope that cooperation between the Cooperative Extension Service, SeaGrant, NMFS and the commercial shrimpingindustry in the southeasternU S.can continue, not only in the improvementof trawling efficiencybut also in continuing the searchfor a longterm, mutually agreeablesolution to the seaturtle declinein the Gulf of Mexico and Atlantic states of the U,S, the femalewould be inseminatedarhficially. Sadly,this programproduced no results. Attemptsat captivebreeding at Sea-Aramaended with thetransfer of theadult ridleys to SeaWorld of Texasin December 1988, Husbandry and Health Care We learnedmuch about husbandry and healthcare of theseturtles over theyears. Knowledge of requirementsfor seawaterquality, foods,feeding techniques, and medicaltreatment has been greatly expanded, Seawater System and Tanks Sea-Arama'sseawater system can be describedas either semi-closedor semi-opened.Raw seawateris pumped from the Gulf of Mexicointo a largesettling basin under the main aquariumbuilding where the Kemp'sridleys are housed.After a specifiedtime allowed for settlingof beachsediment particulates, the seawateris pumpedthrough two large anthracitefilters to remove remaining particlesof sedimentbefore entering the main system,Filtered seawateris used mainly for refillingthis system after being used to backwasha 74-metersand and gravel filter. This biologicalfilter is backwashedtwice a week,resulting in the dumping and replacementof 303,000liters of water,or 40 percentof this 757,000-litersystem. The turtles usedonly a fractionof this system,requiring about45,400 liters of seawater,The Kemp'sridleys were housedin eight 1,760-literconcrete holding tanks,one turtle per tank.Initially, two turtles werehoused per tank.Due to the aggressivenature of juvenile Kemp'sridleys, it wasnecessary to separateeach pair. This wasaccornplishedby installingmarine plywood partitionsbetween them. Holes were drilled randomly in eachpartition to allow seawater flow throughout the tank. Seawaterquality parameterswere monitoreddaily. Oncea day is adequate,but due to use of chemicalsin the marine mammalsystems, seawater tests were conductedthree times daily. Thesetests included the portion of the systemhousing sea turtles. No chemicalswere used in thesesystems except soda ash anhydroussodium carbonate! addeddaily for slight adjustmentof pH, SeawaterpH wasmaintained at 7.5to 7,8in this way. Seawatertemperature was maintainedat 29'C during summermonths and not lessthan 13' C during winter. Eachsea turtle tank wascleaned daily. Eachtank wasscrubbed with a stiffbrush, debris wasallowed to settleand then siphonedout. After eachfeeding, any leftover food wasremoved, and the tank wascleaned again. The need for goodmaintenanceofseawatercannotbestressedenough.Thehealthand well beingofanimalscanberelateddirectly to proper seawaterquality, Spatialrequirements for Kernp'sridleys during the rearing processare minimal, Theseturtles canbe housedin relatively confined areas over long periods of time. Initially, yearling turtles were confined two to a tank, allowing 871 liters of seawater within each compartment of dimensions 107 cm x 117cm x 69 cm for each turtle. Turtles were raised to an average weight of 30 kg in such compartments. We then removed the tank dividers, placing one turtle in each tank, allowing each turtle 1,760liters of seawater within tank dimensions of 213 cm x 117 cm x 69 cm. One very important aspect of the housing requirements for juvenile Kemp's ridleys is that these turtles must be separatedfrom one anotherfrom the time of hatching until they are severalyears old. If they are kept in close confinement, this period may extend to sexual maturity and possibly beyond. These animals are tremendously aggressive and, without hesitation, will attack and inflict serious injury upon one another. Adult ridleys do not dis- play asaggressive a behaviorand can be housedtogether if neededand if given enoughroom. The aggressive nature of these turtles is interesting in itself and worthy of further investigation. Foods and Feedmg Over the years, we developed fairly efficient methods of feeding, selected the proper foods, and found a useful feedingtechnique, Working with Kemp'sridleys thataresometimes finicky feeders,we fed themherring Clupeidae!, mackerel Scombridae!,white-bait and silversides Atherinidae!,capelin Ma00tusvillosus!, squid Cephalopoda!, shrimp Penaeidae!, and live crabs Calli~ectes spp.! during the coumeof the project. The basic diet consisted of capelin 8.8 percent protein, 79.0 percent moisture, 2.1 percent fat, 1.1 percent ash! and vegetable matter. We also fed them fresh, leafy spinach. Of course, the favored foods of these turtles are crustaceans,During the crabbing season,we supplementedtheir diet with a weeklyfeeding of live blue crabs Calti~ectessapidus!, In 1985we were feedingthem threetimes each week wi th eachturtle receivingapproximately 0.5 kg of fish at eachfeeding, Such feeding level and frequency might seem adequate only for subsistence,but monthly weighings and carapacemeasurements showed that steady weight gain and growth in length were achieved. Oneproblem we facedin the beginningof the projectwas an acce'cratedweight gain due to overfeeding.This can lead to death in these animals within a short period of time, Feeding frequency and portion size were reduced until we achieveda properly controlledweight gain, as monitored with a daily feedingchart or record for eachturtle. Feedingroutines for individual turtles canoften inform you of problemsthat canbe correctedbefore they become

37 serious.If an animal is off its regular feedingroutine, this often meanstrouble and usually needsto be investigated, However,feeding patterns do changesomewhat due to seasonalfluctuations in seawatertemperature. Anotherfeeding problem we noticed in Kemp'sridlcys was swallowing air duringsurface feeding. This was only anoccasional problem, but its correctioncan some times be rather time~onsuming.When feeding larger turtles, make surethat the foodsyou usesink to thebottomof theenclosure so they canbe eatenwithout thepossibility of the turtles gulpingair. This problem is sometimesreferred to asa floateror gas-in-the-gutproblem. Such animals may swallow somuch air thatthey have a relativcly hard time swimming to thebottom, and usually float back to thesurface. This cansometimes be very stressfulto theanimals, and aftera whilethey tire and fina in an odd planeat the surface,For somereason, certain turtles have difficulty passingthe air out of their systems.If this problemdoes not correctitself in a fewdays, one may be in for a longperiod of handfeeding below the water surface until theair is dissipated. Althoughall foodsfed to theturtles were of thehighest possible quality and nu tritiona1 value, it wasnecessary to supplementtheir diet with a goodvitamin program.Vitamins supplied duringeach feeding included A andD tablets for propershell growth,vitamin E for preventionof steatitis,and SeaTabs distributedby PacificResearch Laboratories,Inc., P.O. Box 1877,El Cajon,Calif. 92022!,a multi-vitaminand mineralsupplement formulated especially for marine animals. Weighing, Measuring and RecordKeeping TheKemp's ridleys were weighed and their carapace measured monthly when Sea-Arama first received them, and lessfrequently after their first year at Sea-Arama.Weighing was accomplished by suspendingthe turtle in a specially madecargo net, The rectangular net was constructed of soft,knotless, nylon webbing with largestainless steel rings oneach corner, The net was spread on the floor, and the turtle placed in its center. The net was pulled up aroundthe turtle to restrain it, then the rings were placedtogether and hung on the hook of an overheadscale. This method workedwell, with no injuries reported during weighing sessions. When using this method,make sure that the flippers arefolded against the body in a naturalposition. Care should bc takento prevcrt webbingfrom contacting the eyes. Straightline carapacelength and width werc measuredwith a yardstickcaliper, Wekeep three individual record systems for theseanimals, The records cnntain daily andmonthly accounts of weightsand measurements, weight gains and losses, seawater quality, general conditions, foods, feedings, general observationsand comments. Needless to say,accurate, up-to-date record systems are essential to maintaininggood health and husbandry. Identification Tags Identificationtags have posed a problem,The metal flipper tags we originally used were soon outgrown, causing sometissue pinching. There also were problems with electrolysisof thcsctags. The tags corroded and disintegrated tothe point of fallingoff. Some older, smaller tags were replaced with new ones, but within a year,we experienced thesame problems, Many identificatio options were discussed with NMFS. Plastic, livestock ear tags have replaced theold metal clip tags.The Allflcx Tag System Allflcx Tag Company, Ltd., P.O. Box 3132, Santa Monica, Calif. 90403! isdescribed as a two-partplastic tag,applied with a handheld applicator. This tag punctures the flipper and is secured on thetop andbottom, allowing free movement. There are two benefits from this type of tag.It allowsunrestricted growthwith no flipperpinching and, after two years,there is no lossdue to electrolyticdeterioration. Health Care Healthrelated problems werc fcw, although shell lesions caused by bacteriaand fungi are a recurringproblem. Althoughwencvcr had serious problems combating these infections, treatments tended tobe time consuming, Ofthe threetreatments we have used,bactericidal ointments, fungicidal scrubs SclemiumSulfide Shampoo, for nonspecific dermatoses, Rugby Laboratories!, and exposure todirect sunlight, it seemed tha t exposuretodirect sunlight yielded thebest results. A lightingsystem over the turtle tanks could come very close to duplicatingnatural sunlight. Amongthe important aspects ofrearing are stress-related problems due to handling. These turtles frighten easily whenhandled and sometimes will injurethemselves by swimminghead-on into tankwalls. Capture leads to their lossof bowel con trol on almost every occasion. We also had turtles refuse to feed a terf being handled. It is my personal opinion that the lessthese turtles arehandled, the fewer will be the problemsrelated to stress. Unfortunately,there werc mortalities within the group of head startedturtles, and three of ten had died as of December1988. One turtle died following la paroscopy inJuly 1988. This death was preceded by two that died within a 48-hourperiod, A thirdturtle became ill atthat same time but it recovered.By far, the most unusual problem with thesedeaths was that there were no indicationsof illnessor othersymptoms associated with theturtles before their deaths.The animals appeared normal the day before they were found dead, The third turtlewas reported dead, but it wasonly moribund.Though presumed dead, the turtle was removed from thewater, placed on a 15.2-cm-thick foammat, and kept dampwith wet towels,After a lengthydiscussion with Dr. JackBrundrett, Sea-Arama's

38 consultingD.Vie., the following drugs wereadministered: Flo-Cillin sterilePenicillin G, Benzathineand Penicillin G Procainein aqueoussuspension!, injectable vitamins SuperB and B, Solu-DeltaCortef prednisolonesodium succinate,Upjohn!, and Depo-Medrol methylprednisoloneacetate, Upjohn!. Treatmentcontinued over a 22-day period, with full recovery within six weeks, The final laboratory conclusion as to cause of illness was clostridial septicemia.

Plans for the Future More than sevenyears ago, preliminary plans were drawn up for what the staff at Sea-Aramahoped would be the largest, possibly one-of-a-kind, seaturtle exhibit and breeding facility to come into existence.Due to lack of funding, the new SeaTurtle ResearchInstitute, as we proposedto nameit did not materialize.However, my ideasof sucha facility are sharedwith you in the contextof what might be the ideal for sucha facility, Upon entering the institute grounds, guests would have the opportunity to view first hand a mixed-speciesexhibit of seaturtles housed in a 227,000-literoutdoor tank, Graphics would describeand identify each species.Guests would enter the main exhibit hall where a numberof wall graphicsand free standingexhibits would explain evolution, biology, life history and importanceof the seaturtles. In the main hall would be various exhibits,including; 1. explanations of biological aspectsof predation; 2, displaysof egg laying and nestchambers; 3, a static exhibit of the leatherback turtle; and 4. sea turtle products and uses prior to the passageof the Endangered SpeciesAct. Adjacentto the exhibithall, theguests would be ableto view a mixed-species,sea turtle and fish exhibit representing a coral reef habitat. A mini-movie theater, adjacent to the main exhibit hall, would have a seating capacity of 150. Films such as Heartbreak Turtle and multi-screen slide presentationswould be offered. The guests also would get a first-hand view of theresearch laboratory and its activitiesthrough gla ss viewing panels.Of majorimportance would be the breeding lagoon and nesting beach. The lagoon would be an average of 1.5 m deep, with a sloping ramp rising to the nesting beach,Actual dimensionswould be at least30,5 m x9.1 rn x 1.5m with a totalcapacity estimated at 378,000liters. The nestingbeach would run the entire length of the lagoo~on one side.The oppositeside would be used for public viewing, giving the guests a first-hand view of breeding activities and actual nestings. Of all Sea-Arama'sprograms, that involving Kemp'sridley hasproven to be the most costlyand labor-intensive, But theexpenses and time contributedto thisproject may, in thelong term,prove tobe worth morethan any monetary mvestmentor physicaleffort we made.The sevensurviving adults are now housedat SeaWorld of Texasin San Antonio, where they will be available for researchby Dr. Owens and others,

39 Questions and Answers

JackWoody: Mr, Rayburn,do you have a timeframe or schedule,as far asTexas is concerned,to getmarfoity use of TEDsby shrimp vesselsand boats? Rayburn:That item hasbeen brought up in our discussionsover the past numberof years.We havebeen asked to give a time frame or date for the shrimpers to have a majority, at least 50 percer t or more, TED adoption. We have nevercome up with one.Perhaps if oneneeds something to includein a report,that is important.We in the shrimping industry take the position that a forcingor a mandatingof regulationsof this sort arenot in the bestinterest of what we are trying to do, becauseof a numberof reasonsthat we could go into. No, we havenot seta deadline.We have not seta time-frame.Our objectiveis to proceedas quickly aspossible to achievewhat we feel is in thebest interest to both groups,and that is the tact we have taken in agreementwithin the TED committeethat I serveon as co- chairman.Basically, that is the mode that we continueto be in, MarydeleDonnelly: Are yougetting any typeof resi-tancefrom thefishermen? Rayburn:The technology is still beingintroduced into thefishery on this sideof theworld. Therewas early resistance to TEDsprimarily during the developmentprocess. This involved the various prototypesthat were developedas turtle excluders.First, theuse o flarge mesh over the m outh of the trawl, thenthe problemof bulkinessof theoriginal TEDapparatus that wasused in the Atlantic statesyears ago. Concerning the first device,there was a negativevisual impactof a nethanging from the webbing,i,e.; an apronover the mouthof thc trawl. Therewas very muchconcern over the possibleloss of shrimp from this large meshbarrier. The bulky structureof the TED that wasdeveloped a few yearsago causeda great deal of concernfor the lives of the crewmanbc "ause of its weight and bulk. The collapsible,light-weight TED that is now being incorporatedis promising.We had at our pastconvention a sample for display. The initial responsewas that the apparatuslooked very complicated.All fishermenhave not had the opportunity to seeit on boarda vessel.They have not testedit on thestyle vessel tha t theyuse or with thetype of trawls they have,That is now 985! beingdone in our particulararea. Fishermen are now getting to seehow it operates.So asfar ashaving fishermen resist the TED, I wouldsay at thistime that there probably is some,primarily because of ignoranceabout just exactly how the new TED looks and what it does.With fisher nen, there is a resistancetoany type of change.To saythere is no resistanceon anythingis just impossib!c.So, if the kinds of resistancewe haveseen so far is typical and I amhoping that it doesnot increase we candeal with it througheducation. If anattempt is madeto overcomeresistance through a regulatoryprocess or to forceimplc",:enation, I think that would really be self-defeatingin this particular case. I reallydid notcome here to try to placeson-e kind ofbarrier between regulation andeducation and those kind of things.I reallyfeel that, in this case,and from my experienceworking with the fishermen,the people I workwith on a day-to-daybasis and talk to ona day-to-daybasis, education is a keyto solving theproblem. I amsure that if youhave spoken with Ila Loetscherhere, she perhaps has commented to you about the responseshe gets down in the Brownsville-PortIsabel area from the shrimpers.The fishermenthere are interested. Theyare people that live off thesea, and they are not interested in goingaround killing seaturtles at all. Theanswer is to givethem a methodof protectingsea turtles that works, and I think theywill adoptit. I reallydo. I think the approachwe are taking is thecorrect one, If not,we are in theposition where we canchange it. SallyMurphy: If theother type fishing gear you mentioned the "quad-rigged" trawls wereadopted within, say, a sixmonth period,or at leastover one season, why do you think it is takingso long to get acceptance of TEDs by the shrirnping industry? Rayburn:The quadruple-rigged trawls represented a technology that had been around for quite a while,Not many fishermenwere involved at first. Adoption of quadruple-riggedtrawl technology took place in a shorttime oncesome fishcrmentested the gear and proved it tobe successful. Then it caughton like wildfire,but it wasnot something all thefishermen jumped into right away, So, the clock for TED adoption, in myopinion, has not necessarily started yet. Weare still showingthe fishermen, through the work of theMarine Advisory agents, the National Marine Fisheries Serviceand others, that these devices can work and be effective in solvingsome of theproblems e.g., by-catch! that they deal with on a dailybasis. AndreLandry: Could Charles Orvatez, or Ralph Rayburn, address the role that economics plays or wi71 play in theacceptance andimplementation ofthe TED? Also, do you envision the need for financial incentives in lieuof regulation to get100 percent usageof TEDs? Oravetz:Well, I certainlythink that if wecould prove conclusively tothe shrimp fishermen that use of TEDs was by farof great economicbenefit it would enhance the rate of TED adoption. We had a preliminaryeconomic analysis of thevery first model TED about four years ago, It essentiallywas favorable. 1am not sure whether or notwe have had anyadditional economic analyses of theTED since then. Perhaps Ralph Rayburn has something to add, Rayburn;The shrimpers do not now have, or in thelast three to fouryears, have they had any excess monies that could beused for experimentationwith TEDs,That represents the negative economic side to theindustry adopting any new technology,The positive side is whatCharles Oravetz referred to. It hasbeen shown to someextent already that there is a positiveeffect of reducedwork from theuse of theTED. But, one has to payfor a TEDfirst, thenhope that one getseconomic returns on that investment.So we areat a pointin theshrimping business at which onehas to dealwith the realities of economics, Frank Judd: How much doesa TED costand how long doesit last? Oravetz:A metalTED without webbingcosts less than $200.A TED with the webbingaround it one readyto be insertedinto and attachedto the shrimp trawl costs about $400to $450,The latter will last at least two years, dependingon how badly it getsbanged up during use e.g,,what kinds of hangsthe fishermenmay run acrosswhile trawling with a TED-equippedtrawl!. David Bowman:Could you repeat exactly how long the TED technology has been available and how many TEDs are being used t'nthe Gulf of Mexicoat this time? Oravetz:We think that thereare around 200to 300TED devices being usedthroughout the shrimp industry in the Atlantic coaststates, probably on a parttimebasis.The technology transfer program essentially hasbeen in effectthere for about two years.But TED s havegone through a metamorphosisduring that time, so even though we were involved in a TED technology transfer program a couple of years ago, it was not the right TED we were testing then. We think we now have overcome the major objections with the TED we now have, based on what we have been told is an acceptableTED, Sothe technologytransfer program involving the new TEDessentially started with thedistribution of 50 of these TEDs in the summer of 1985, Bowman: In other words,is the presentlevel zero? Oravetz: I did not say that, I said the present level is 200 to 300. Bowman: Are most of thesein the Atlantic, or entirely in the Atlantic? Oravetz: Yes, because that is where we first focused the TED effort. We have only begun to move the new TED technology transfer program to the Gulf during the 1985 shrimping season,with the introduction of 50 collapsible TEDs into the fishery, Larry Ogren:ls useof the TED from Cameron Parish, la., spreadingeastward to any other parishes? Why is thatTED so unique to theCameron area, and why has it notbeen adopted sooner, or its adoptionexpanded to theeast side of theMississippi river or even to Terrebonne Parish? Coreil: Fishermenare more or lessport-oriented, and the word on technologydevelopments generally is held among thosewithin a port, and they do not sharevery much information about it outside that port. In our job as Marine Advisory agen'tswe try to spreadinformation to other ports,It is difficult for any oneof us to go from port to port, sowe give our information to otheragents in otherports, The other agents are now distributing theCameron Excluder Devicetechnology to otherports. When you tell a fishermanin anotherport that "this is a CameronExcluder Device," he automaticallysays, "Well, I don't want to do somethingthat Camcrondoes." We must try to overcomethat type of resistance,and we areat that stagenow. In 1985,we areworking to distributeTED sketchesand blueprintsto the fishermenin otherparishes, We areat tempting to getfishermen in eachport to put CameronTEDs in their trawls and spreadthe news about their experiences, If one fishermen, who is well accepted,uses the TED then more will accept it as time goes on. EdwardKlima: Whatabout the other Sea Grant Marine Advisory agentsin Texas, Louisiana or Mississippi?Are you trying to spreadthe news about the NMFS TED or theCameraon Excluder Device to them,and, if so,why? Coreil: We are trying to spread information aboutboth the Cameron Excluder and the NMFS TED,becausethe NMFS TEDis working well too, We alwaysbring a NMFSTED to our workshops,but we alsobring the CameronExcluder to show the fishermenthat they canbuild an excluderdevice themselves, without having to useexact specifications to achievewhat the governmentsays will work, If a fishermanbuilds the devicehimself - for exampleCameron fishermenbuilt the CameronExcluders themselves he will feel a lot better about using it than if told to use one designedby theNational Marine Fisheries Service. It givesthe fishermena greatersense of controlif theybuild excludersthemselves. The CameronExcluder works in the sameway as the NMFS TED. Fishermenmay use the NMFS TED,but they feel better knowing that they candesign and buUd somethingthat will alsowork, Jack Woody:Am I correct,from whatyou said,to assumethat oneof thevariations in this equipmentis in standarduse t'n Cameron Parish now, and on both vesselsand boats? Coreil: Yes,throughout the industry, on shrimpingvessels and boatsalike. Woody: Wehave heard NMFS' estimatesof $200to $400for theNMFS TED. Whatdo thesehomemade excluders cost in Louisiana? Coreil:About $150 each, but thatis withoutthe netting around them. They would probably cost about $200 apiece with netting. Woody:So, they are in thesame ballpark cost-wise as the NMFS TED,

41

1. turtle postersprepared and distributed in Texasto alert the public about who should be contactedabout stranded turtles; 2. newscasts on radio and television; 3. televisioninterviews of network participants; 4. television documentaries; and 5. newspaper coverage. Suchpublicity hasbeen increasing and improving, Publicity has genera ted greater understanding of thepurpose and importanceof theSTSSN. This has improved the quality and quantity of dataobtained, but alsohas created greater demandsfor interpretationof the information with regardto its significanceto seaturtle conservation. TheNPS staff at thePadre Island National Seashore responded to reportsof strandedsea turtles along Padre and MustangIsland beaches from 1976to 1978and documentedthe strandings on NPSforms. These forms were revised andimproved from timeto timeuntil 1982when the NMFS Southeast Fisheries Center SEFC! Miami Laboratory establisheda standardized form FigureI! tobeused throughout the southeast coastal states to documentstrandings. Thesenew formswere used to reportthe data summarized by this paperfor theyears 1983 to 1985.All stranding reportsfrom Texas volunteers were sent to myoffice where they were examined, validated and corrected as necessaqr beforebeing copied and forwarded to STSSNheadquarters at theNMFS SEFC Miami Laboratory. Volunteersrepresent the foundation of theSTSSN. STSSN volunteers in Texaswere screened by theNMFS SEFC and weregiven proper permits, They enabled us to get not only the type of datarequested, but alsoadditional informationrecorded in the commentssection of thedata form, whenthey thoughtthat suchadditional observations about the strandedturtle and its surroundingswere pertinent.Such comments added a newdimension to our data collecting,Stranded sea turtles were measured in centimeters down the length of theircarapace. Included in thesedata aresome measurements that were estimated rather than being accurately measured directly. Networkvolunteers and other individuals also were encouraged to recordobservations about live specimens sighted within Gulf and coastalestuarine waters, I have received information from a number of fishermen and others concerningsightings of turtleswithin theirnatural habitats, These observations include specific areas where turtles weresighted, their observed activities at thetime and their feeding activities. Such observahons represent the least completesegments of our data becausethey do not containmeasurements or correctspecies identification. Nevertheless,they do provideimportant information on number,time of yearand location of seaturtle sightings, Suchin formation can be used to detectchanges in abundanceof seaturtles, Its accuracy will improveas move interest is generated.However, improved methods for speciesidentification of sightedsea turtles must be devised and used to improvethis source of data.Our da ta on sigh tings maybe very helpful, and they will becompiled and summarized in time. Interestedparties have saved the carcasses of a fewstranded sea turtles for necropsy. Many of theturtles found exhibitno external clues to the cause of death,and this represents anopportunity for additional investigation. Only through necropsycan one hope to makea determinationof the causeof death,We at the National Seashorehave neitherthe capability nor resources tohandle this function, but I understandthat others at The University of Texas MarineScience Institute at Port Aransas,and at TexasA&M Universityin CollegeStation and Galvestonare conducting necropsies of stranded carcasses. Dataaccumulated from 1983through 1985 are the source for my resultsand discussion. Results and Discussion Fight.2 showsthe number of eachspecies stranded along the Texas coast during 1983 to 1985.Loggerhead Caret ta caretta!dominated the stranding records, with Kemp'sridley second in numberof reportedstrandings. Figure 3 showsthe number of strandingsreported by speciesand month of theyear, This gives a pictureof theseasonal occur- renceof eachspecies in thestrandings. Total reported strandings were highest in April andMay. Kemp's ridley strandingsoccurred in highestnumber in Apriland June through October Figure 3!. This is similar to thefindings of Rabalais983!. October appeared tobe the secondary peak month for strandings. Rabalais 983! showsNovember andJuly as secondary peaks in strandings,What factors could be at work to provide peak strandings in springand fall? Frequencyof carapace lengths of strandedKemp's ridley sea turtles, grouped in 10-cmsize classes, isdepicted m Figure4. Thesesizes include not only theturtles that weremeasured directly but alsothose for whichsizes weve estimated.Figure 5 showsthe strandings ofsea turtles by species and county. Nueces County had the highest number ofsea turtle strandings followed by Jefferson County, Fritts et al. 983! noted that the majority of Kemp's ridleys were observedoff theeastern and western coasts of Florida,They made no mention of Texas.Sabine Pass and Sea Rim State Parkin jefferson County have provided the greatest number of stranding reports for Kemp's ridleys in Texas. Port SEATURTLE STRANDING AND SALVAGE NETWORK - STRANDINGREPORT PlgaSEPRINT CIEARLT ANDNU IN All APPLICARIF RIANRSU«d b Io al t oy t ~ «ghr' Ic~ IpI d Ihc« ltop *-- 'r As I gthl, h I, I f~ n,l,I l,t,t tl,,v Ip, p*«% g I AV e dthoP d VaIU o «her~ CDCLRTNEUNITSUSED Seedog o ~bsh Plc ~ c aspecks lkono d pto INCWDELATIIVDE ANDtDNDtlUDE Flthh, Sr e-gbe n Ado Afftot n AreaCode Pho c Hyhe Spec~ PrDoy- Itclatllrr*IIDICIRCIE V»c Pobbt P d.t Spce y I dbybsotCo*no .'. 'I rc,tt« S ~rCIRCLEI F I~ AAole V dcr r ed mrNc dere iredr Enrc Coty theat.*,bcpecf. o d ' IM I I r

log! d C*dh* IT«l I *d I TogN abcsl 1ncl de rop e n odd.t«ondd per' elrogr

Re -h,,nol~ I .a orn cled hh«~ oilpeo o det' e to glee I o" Io rllaro*po Iffodo cp papl'c cnep coo. erc ! ~ I e ~ harlI eeet-rr

CODES VREALENTSCIRCLEVk!TS SPECIES CC= lalg. od I 1cog ll CAAO ee DC= teat"tact I Vy«I El= Hakbg LK= Kcp'I «dt y I «Ilh Ure= V'd reld vydrh COI«Dlrlotsot I«alit D Al~ I ~F est dead 2 Ased. Iy deca..pred S c.elyde oa osed Angora ~ «nds, nroTI. gIo e' Foe*IDISPOSITIOlr OFTURTIE I =Pa -: d.Icl o benet 2 R cdor brad ofthooch S =Sor gcd per acn ag ' pert A=egadpo I I I d S*Up ~t d lelrO*bool Porter r O Al efe-d Norg.na 7=AhkNle toehold'Ntfo.I'll

Figure 1. Sea turtle StraFIdingand SalvagenetwOrk stranding report reduced!.

Theback of thestranding report form includesa speciesidentification guide,

45 250

200

150 0 100Z

Z

50

0 I:1 LOGGERHEAD I GREEN t LEATHERBACK KEMP S RIDLEY HA WKSBILL

SP EC IES

Figure2. Seaturtle sfrandingsin Texasby species,1983 to 1985combined.

150

HA WKSBILL

'LEATHERBACK

KEMP'S RIDLEY

> 100 GREEN

LOGGERHEAD I-

0 C

50 Z

0 JAN INAY JUL SEP NOV

MONTH Figure3. Seaturtle strandingsin Texasby speciesand month, 1983 to 1985combined

46 CO

50 0K25 X DZ

0 I I I I I I I I 20 40 60 80 100 CARAPACE LENGTH CM! Figure4. Carapacelength-frequency for Kemp'srtdley sea turtles stranded in Texas,1983 to 1985 combined.

CO

150 5 OTHER Z KEMP'S RIDLEY lO Q LOGGERHEAD I- P 100 LL 0 Z0 O 50 Z K 0 o O co I- to Z LL Lu K 0 Figure5. Sea turf lestrandings in Texas by species and county left to right represrnts north to south along the Texas coast!, 1983 to 1985 cond;ined. ationsconcerning condition of strandedKernp's ridley seaturtles, pro , dispositionof theturtles and notes on theirsurroundings by month

Jan Mar Apr May Jun Jul Aug Sep Oct Nov Dec Total

2 2 2 3 2 2 I 14 2 4 2 I 2 I 3 15 I I I 3 6 I 3 4 2 1 I 2 I 2 I I 11 I 2 I I 3 2 2 I 13 I I 2 I 5 I I 2 I 2 2 3 I I 2 9 5 I I 10 6 8 I I 2 I I 2 I 5 2 3 6 I I 2 I I 17 2 2 11 5 5 6 I 33 I 2 2 I I I 9 2 2 15 5 4 5 2 3 2 4 44 2 9 12 4 I 28 4 5 7 8 4 13 12 8 61

not mutually exclusive. There were no strandings in February,

Aransasand Mustang Island in Nueces County have been second infrequency ofKemp's ridley stranding reports, Otherareas along the Texas coast account forthe rest of the Kemp's ridley strandings. Sightings and strandings of Kemp'sridley have been reported from inshore waters, including Sabine Lake, between the jethes of Sabine Pass, in AransasPass, and in otherbays of the upper coast of Texas, aswell as in IntracoastalCanal and Laguna Madre. Therefore,Kemp's ridley obviously occurs in baysand inlets of theTexas coast. Shownin TableI arefrequency, by month,of 20categories representing the condition of strandedsea turtles and thecircumstances underwhich they were found, aswell as their final disposition. These categories arenot mutually exclusive,asa giventurtle stranding can be assigned tomore than one of the categories. Such data can be used to link strandingsto man'svarious at-sea activities or events,Included are data for deadturtles found fresh and without externalwounds ordamage ofany sort. Clearly, these deaths were not caused bynatural predation ormuhlation, Still otherturtles were found with various wounds onthe carapace andother parts of the body, Additional observations notshown inmy data summaries include unusual wounds or theflippers and other pa,s of the body as a resultof entanglementinplashc rope or netting. Many reports havebecn received ofyourtg ridleyys caught by hook-and-line sportsfishermenusing cut mullet Mugil cephalus! asbait while fishing in the surf adjacent toGulf beaches. Turtles reportedas caught in thisway ranged from 19- to 43-cmcarapace length. TheSTSSN provides anincomplete picture oftotal strandings, andit reflec tsin part the varying degree ofcoverage ofTexas' coastal beaches byvolunteers. Todate, ithas not shown the true total number ofsea turtle strandings. Also, notenough necropsies have been performed todetcrotine the true causes of death, Yet, it is probablethat human activitiesand perhaps pollutants caused the death of stranded seaturtles, either directly or indirectly. Aswe look at the history of developmentof the STSSN compared toits prcscnt status, it is obviousthat it is now providinga much better overall picture ofstrandings inTexas than here-to-fore hasbeen available. I refer not only tothe documented numbers ofsea turtles washing ashore, but also the growing number ofsightings oflive turtles in open waters,observations on the condition of seaturtles found strandedand attemptsto determineindirect or direct causesof strandings. Strandedsea turtles axefound on the Texascoast during every xnonthof the year Figure 3!. Largenumbers of strandedKemp's ridleys are neonates

49 Yetthe proporhon of neonateKemp's ridleys in thestrandings decreased, For example, only six neonate ridleys were reportedin NuecesCounty in 1982by Rabalais 983!. In thisstudy, Jefferson County led in totalnumber of seaturtle strandings, but neonates have continued to decreasein their ratio to total numbers of seaturtles found stranded. There alsohave been decreases in thenumber of areasin whichneona tes are found, and in thetimes of theyear in which neonate strandings occur,

Conclusion It is very important that STSSNactivities continue to assesssea turtle strandings,and to bring them to the attention of thepublic. Crea ter public awareness will stimulateconcern about the problems faced by seaturtles. The wealth of availabledata and the many problems associated with seaturtle conservation provide opportunities for researchby graduatestudents and othersinterested in lending their talentstoward solutionsto the problemsand to the critical situation facing these valuable natural resources. In closing,I will sharewith yousome o fthe questions tha t havebeen asked me while in pursuitof thiswork, and which provide opporturutiesfor research: 1. Are therecertain routes traveled by seaturtles and tune frames in whichcertain age groups of seaturtles are found in certain places? 2. Are therecertain areas bays,inlets, nearshore areas, sargassum weed beds, etc,! that seaturtles seekout for the purpose of finding food or for various other reasons? 3. Whena seaturtle is killedat sea, what happens to its carcass?Does it continueto float?Presumably it will sink until it becomesbloated, What happens if sharksattack it, andparts are eaten or removed?Does it sinkagain? If so, will such a carcassever reachshore? How long does a seaturtle carcassfloat? In other words, do strandingsprovide an accurate picture of all theturtles that are killed in oneway or anotherat sea? 4. Whatcan we sayabout the total sea turtle population based on data from strand ings? What effect does such lossof animals have on sea turtle populations? What percentage ofthe sea turtle population isrepresented by thereported strandings,by species? Are certain age-groups particularly susceptible tovarious impacts leading to strandings, and what are the reasons? 5, Are thereadditional regulations that can be developed to helpprevent strandings of seaturtles? It hasbeen proposedtha t shrimpingoperations be regula ted by requiringuse of a TurtleExch tder Device TED! in shrimp trawls.If suchregulations are implemented, how likely are they to be accepted by theshrimp industry? 6. If thesequestions suggest additional needs for information,what are they, and how shouldthe studiesbe designedto answerthe questions? Perhapsthe STSSNwill help answersome of thesequestions in the future. Literature Cited Anonymous.1982. Sea turtle conservation strategy, p,567-583. In:Bjorndal, K. A. Editor!, Biology and Conservation ofSea Turtles, SmithsonianInstitution Press, Washington, D.C., 583 p. Fritts,T.H., A.B. Irvine, R.D. Jennings, L.A. Collum, W. Hoffmanand M.A. McGehee. 1983. Turtles, birds, and mammals in the northernGulf of Mexico and nearby Atiautic waters: an overview based on aerial surveys of OCSareas, with emphasis on oiland gas effects, U.S. Department ofthe Interior, Fish and Wildlife Service, Division of Biological Services, FWS/OBS-82- 65,455 p. Hildebrand,H.H. 1963. Hsllaxgo del area de atudacion de Ia tortugamarina '1ora," Leputochelys kempi Carman!, en la costa occidental del Golfo de Mexico.Ciencia Mexico! 22!:105-112. Rabalais,S,1983. Sea turtle stranding and salvage research, p.42A3, In; Owens, D., D. Crowell,G, Dienberg, M, Grassman, S. McCain,Y. Morris, N.Schwantes andT. Wibbels Editors!, Western Gulfof AfexicoSea Turtle Workshop, Texas A AMUniversity SeaGrant Program, College Station, Tex,, TAMU-SG-84-105, 74p. Rabalais, SC, andN N. Rabslais.1980. The occurrence of seaturtles on thesouth Texas coast. Contributions in MarineScience 23:123- 129. Witham,R. 1978. Does a problemexist relative to smailsea turtles and oil spills?,p. 630-632 In: Proceedingsofthe 197B Conference onAssessment ofEcological Impacts of Oil Spills,American Petroleum Institute, Washwngton, D.C. Recent Strandings of Sea Turtles, Cetaceans and Birds in the Vicinity of Mustang Island, Texas Anthony F. Amos

Strandingsof sea turtles, small cetaceans and birds are common events along south Texas beaches bordering the Gulf of Mexico. In recentyears, the incidence of suchstrandings has been recorded formally as part of a long-termstudy of bird populations utilizingMustang Island beach near Corpus Christi, Tex. Data have been submitted to theMarine Mammal Stranding Network andSea Turtle Stranding and SalvageNetwork and to theAtlantic and Gulf CoastBeached Bird Survey, One-hundredtwenty sea turtles of five specieshave been reported strandrd frown 1983 to 1985.In orderof abundance they are loggerhead Caretta caretta!, Kemp's ridley Lepidochelyskemp0, hau'ksbill Eretmocheiys iznbricata!, greeii Chelonia mydas!and leatherback Dermochelys coriacea!. Kemp's ridleys were found most frequently following the offshore releases of headstarted yearlings; several found alive were later re-released. The occurrence of 12juvenile hawksbills, inost of themalive, is notableas this species has been virtually unreported for Texaswaters, Peak abundance for loggerheads,almost always found dead, is in Marchand April andappears to coincidewith shrimping activities offshore. Thirty five smalcetaceans have been reported stranded in thesame period. Nine of these have been livestrandings of uncommon torare pelagicspecies,including pygmy killer whale Feresa attenuata! and dwarf sperm whales Kogia simus!, spi nner dolphin Stenellalongirostris!, short-snouted spinner dolphin S. clymeme!,and striped dolphin S,coeruleoalba!, Live strandings haveoccurred in October,December, April and May. Fifteen-hundredbirds havewashed ashore dead from 1980to 1985.Mcrtality of laughinggull Lagus atricilla!, the most commonlyfound dead bird, peaked dramatically following the opening of theGulf to shrirnpingin July,More than 4,000 oiled birds havebeen observed and a periodicityin thefrequency of oiling spri ng and fall! discovered.

~ 7heUniversity of TexasMarine Science Institute

Methods ROUTE In a study fundedby MineralsManagement Service, U.S.Department of the Interior, greenand loggerhead sea turtles were exposed for various periods to weath- ered, South Louisiana Crude Oil. Minimal exposures were used, consistent with levels that might occur in a SKI II natural spill, In all casesin which harmful effectswere seen,the experimentswere terminated,and there was full recoveryof all experimentalanimals. Results LUIIG A preliminaryanalysis of thebehavior data by Dr. D Odell showed no evidence that either green or logger- headsea turtles actively avoidedweathered oil, There- fore, the possibility exists that sea turtles can be in GUT extendedand prolongedcontact with oil in a spill,with- out making an effort to leave the area. The potential effects of oil on sea turtles, as indicated by our studies,are shownin Figure1, To our surpriseall primary contactroutes appear to be affected:

S EIISE 1. Skin Dr. G. Bossart has found evidence of 0 RGAIIS dermatologicaldamage and abnormalcell differ- entiation in the skin. Presumably, these effects would eventually be lethal, but they certainly Figun 1. Potential would open routes of attack for parasitesand infection. 2, Lungs We recently found that seaturtles are amongthose having the most efficient and highly developed lungs Lutcavage, Lutz and Bauer 1987!. Any interference with operation of the lungs would probably reduce a sea turtle's capacity for sustained activity aerobic scope! and its dive time, both effects decreasing the turtle's chance of survival, Marked changesin respiratory patterns were in fact evident in oil- exposed turtles, with an increase in both tidal volume and frequency following exposure and a decreasein oxygenconsumption. It appearsthat oxygenextraction is compromisedin oiled turtles. 3. Gut Studies on the effectof oil on digestiveefficiency are still underway at the time 985! of this writing, but we have found that oil is swallowed by experimental turtles, probably when they surface to breath, and oil has been found in their feces.One of the most striking differences between chronically oiled and control non-oiled! turtles was the hematological response,presumably due to the passageof oil products acrossthe gut lining, Even so,mean red blood cell counts,hematocrits and hemoglobinconcentrations were lower in oiled turtles compared to controls. These results indicated dysfunction of red blood cell-forming tissues.We recently found that the salt gland in sea turtles is a very important organ. In the leatherback it is continually active Hudson and Lutz, 1986a!and in the loggcrhcadit is involved activelyat leastin regulating minor ions suchas Mg++ and Br-aswell asNa+ andCl- Hudsonand Lutz,]986b!,We have indications that theoperation of this organ is disrupted with exposure to oil, but the dL~turbancedoes not appear until several days after exposure. 4, Hormonesand SenseOrgans We havenot yet lookedat hormonebalance, but we observedthat the eyes and nareswere coated with oil that requiredcleansing. Disruption of either sensingsys tem would havehighly deleterious consequences.

Discussion Experimental results indicatethat seaturtles would indeedbeatgreatriskif they were engulfed byan oil spill. This was a surprisingfinding consideringthe high toleranceof seaturtles to severephysical damage. Contact with oil would be particularly harmful if it occurred when adults congregated during the nesting seasonto copulate and lay their eggs or when hatchlings emerged and crawled over the beach in large numbers, For theprecarious Kemp's ridley theseobservations have serious implications, Effective strategies mustbe quickly formulatedto mitigate the damageof oil pollution since,without doubt, we will be facedat sometime in the future

53 with a crisis;e.g., a largeoil slickapproaching a beachwhere rid leys are hatching. It is essentialthat we haveplans to dealwith suchemergencies. One of themost attractive possibilities, for example,is theuse of chemicaldispersants tobreak up theoil slick,but wedo notknow if theuse of dispersantsis anacceptable management option Because seaturtles have been shown to be extraordinarily sensitive to oil, theyalso might be highly affected by dispersants, Theseand similar questions must be answered before we will beable to protect Kemp's ridley from the potentially devastatingeffects of a majoroil spill. Acknowledge1nertts Thisstudy was supported by theU.S. Department of theInterior, Minerals Management Service, Grant USDI MMS «14-12-0001-30063,

Literature Cited Bennett,J.A. and H. Kleerekooper,1978. A preliminaryinvestigation of the effectsof chemicalstimulation on the locomotor behaviorof hatchlinggreen turtles Chelonia mydas!. Florida Marine Resources PubIication No. 33, p. 3-7. Chan,E.L, 1977. Oil pollution and tropical littoral communities: biological effects of the 1975 Florida Keys oil spill,p. 539-542. In; Ludwigson,J O. Editor!, Proceedings ofIhe1977 OiI Spill Conference, American Petroleum Institute, Washington, D C., 640p. Clark,D R.and A J.Kiynitsky. 1980, Organochlorine residues ineggs of loggerhead and green sea turtles nesting at Mcrritt Island, Florida Julyand August, 1976. Pesticides Moriitoring journal 14: 7-10. Delicat,D.S. 1980. IXTOC I oil spilland Atlantic ridlcy survival, p. 312-319. In:Edge, B.L. Editor!, Coastal Zone 'BO, Volume 1, AmericanSociety of Civil Engineers,New York. Diaz-Piferrer,M. 1962.The effects of anoil spillon the shore of Guanica,Puerto Rico, p. 12-13.In; Proceedings, Fourth MeeHng, AssociatedIsland Marine Laboratories, Curacao, University of PuertoRico, Mayaguez, Puerto Rico. Fontaine,C,T,, R.M. Harris, W.J. Browning, T.D. Williams and S.A. Manzella. 1989. Observations ondistribution, growth and survivalof captive-reared,tagged and released Kemp's ridiey sea turtles Lepidoc helys kempi! from year classes 1978-1983. This volume!. Frazier,J.G, 1980. Marine turtles and problems in coastalmanagement, p. 2395-2411. In: Edge, B.L Editor!,Coastal Zone 'BO, AmericanSociety of Civil Engineers,New York. Fritts,T H.and M A.McGehee. 1981. Effects of petroleumon the development and survival of marine turtle embryos. Report toU S. Fishand WddlifeService, Contract No. MOV AA 851MUO-21. Hillestad,H.O., R J.Reimold, R.R. St ickney, H.L. Windom and J.H.Jenkins. 1974. Pesticide., heavy metals and radionuclide uptake in loggerheadsea turtles from South Carolina and Georgia. Herpetological Review 5!:75. Hudson,D, and P.L. Lutz. 1986a. Salt gland function in theleatherback sea turtle Dermochelyscoriacea!. Copeia 1986 !:247-249. Hudson,D.and P.L. Lutz. 1986b. Salt gland function inthe Atlantic green turtle Chelonia mydas mydas!. ThePhysiologist. 28!:343. Light,M. and J,J. Lanier, 1978. Biological effe ts of oil pollution a comprehensive bibliography with abstracts. Report No. CGR/ DC-18/78: USCG-D-75-78. Lutcavage,M.,P.L. Lutz and H, Bauer. 1987. Pulmonary transport inthe loggerhead seaturtle. Journal of Experimental Biology 131:315-372. Moore,S.F. and R L.Dwyer. 1974. Effects of oil onmarine organisms: a critical asses smcnt of publisheddata. Water Resout ~ 9: 819427. Rutzler,K. and W. Sterrer. 1970.Oil pollution: damage observed in tropical communities along the Atlantic seaboard ofPanama. Bioscience20: 222-224. Schwartz,J.H. and W. Flamenbaum. 1976.Heavy-metal induced alterations inion transport byturtle urinary bladder. American Journalof Physiology230!:1582-1589. Stoneburner,D.L., M.N. Nicrora and E,R. Blood. 1980. Heavy metals in loggerheadsea turtle eggs Carelfa caretta!: evidence to supportthe hypothesis that demes exist in thewestern Atlantic population. Journal ofHerpetology 14!: 171-176. Thompson,N.P., P.W. Rankin and D. W. Johnson. 1974, Polychlorinatcd biphenyl and P, p-DDE in greenturtle eggs from AscensionIsland, South Atlantic Ocean. Bulletin of Environmental Contamination andToxicology 11!:399-406. Witharn,R. 1978,Does a problemexist relative tosmall sea turtles and oil spills. p. 629439. In:Proceedings ofthe 197B Conference onAssessment ofEcological Impacts ofOil Spills, American Petroleum Institute, Washington, D.C. Witham,R.1983. A reviewofpetroleum impacts on sea turtles, p. 7-8, In: Keller, C.E. and J.K, Adams Editors!, Proceedings ofa WorkshoponCetaceans andSea Turtles inthe Gulf of Mexico: Study Pla~ning forEffects ofOuter Continenlal Shelf Development, Preparedby the U.S. Fish and Wildlife Service for the Minerals Management Service, Metarie, LA.

THE ENDANGEREDSPECIES ACT SECTION7 CONSULTATION PROCESS

EVD DQ ~VTI CDDEDLTET[DD [DEED'TAT[ID[

[[ID 'IE'I ED[[EL']TIT[D'lg

EED EVD DEVELLTIT [DII EDEEn,TIIT[DEQ

Figure1,Generalized floEochart ofthe Section 7consuitation process ¹te: the process hasbeen modified toincorporate the19B5 proposed regulations!. Section 7 Consultation Procedure Figure 1 is a flow chart depicting procedures and options of the Section 7 process. Amendments to the ESA have modified the processin the past,and proposedregulations undoubtedly will changethe interpretationof specific segmentsin the future. Details of regulations governing interagencyconsultations involving threatenedand endangeredspecies can be found in the Code of FederalRegulations CFR-50, Part 402, or by contactingthe Department of Commerce, National Oceanic and Atmospheric Administration, National Marine Fisheries Service, Habitat and Protected SpeciesDivision, Washington, D.C. 20235.

Background on the Corpus Christi Bay Landmass Project On July 24,1985,the Corps of EngineersGalveston District providedthe NMFS SERO a BiologicalAssessment BA! for a permit activity to constructa 3.2-hectarelandfill structurein CorpusChristi Bay.This activity, known as the CorpusChristi Bay LandmassProject, was proposedby the City of CorpusChristi, and is hereinafterreferred to as the Land~ass Project.It was proposed tobe located within the city's marina for the purpose of supporting a waterfront recreationalcenter and tourism complex,an aquariumcomplex, boat slips,and a breakwater-fishingpier. Basedon dataregarding the presenceof threatenedand endangeredsea turtles in the CorpusChristi Baysystem, the Corps determinedthat the proposedLandmass Project might affect the endangeredKemp's ridley seaturtle Lepidochelyskempt!, Therefore, pursuant to Section7 of the ESA,they requestedinitiation of a formalconsultation. Consequently,the NMFS was requiredto review the availabledata and preparea documentknown asa Biological Opinion, which would determineif the proposeddevelopment was likely to jeopardizethe continuedexistence of the Kemp's ridley. The NMFS was also concernedwith the potential effectsof the proposedactivity on three additional seaturtle species:the endangeredhawksbill Eretrrrochetysimbricata!, and the threatenedgreen Chelorria rtrydas!and loggerhead Carettacaretta!. Proposed initial construction was to use an hydraulic dredge to dig a trench around the perimeter of the landfill site removingapproximately 35,170 cubic meters of bay bottom sedimentand placingit in the centerof the landfill site Corpsof EngineersGalveston District, 1985!. Approximately 3,2 to 3,6hectare of baybottom at theland fill site wouldhave been eliminated by theseconstruction activities. In addition,the proposed activity was to construct425 boat slips and a breakwater-fishingpier. Species Accounts ThoseESA-listed species under the jurisdictionoF the NMFSand ide~tifiedas potentiallyoccurring in the LandmassProject area were Kemp's ridley, green, hawksbill and . All Fourspecies had been documentedas occurring in theCorpus Christi-Nueces Bay system. A briefspecies account for eachof thesesea turtles is provided,with emphasison theiroccurrence in Texasbays. These species accounts were used by NMFSin pre- paring 'the Biological Opinion, Kemp's Ridley Sea Turtle Theprimary range of theKemp's ridley sea turtle is theGulf of Mexico,but juveniles occur along the Atlantic coast of theU S.from Florida to NewEngland Lazell, 1980; Lu tea vage and Mu sick, 1985!. The Kemp's ridley is a turtleof coastalareas. Hildebrand 982! hypothesizedthat the two areas of greatestconcentration of Kernp'sridley were the shallowwater areas of Louisiana,from MarshIsland to theMississippi Delta, and the Tabasco-Campeche area of Mexico.Both areas are thought to be feedinggrounds for subadultand adult Kernp's ridley turtles Chavez,1969; Hildebrand,1982!. These waters are extremely productive and are abundant with decapodcrustaceans, particularly two generaof crabs Ovatipesand Callinectes!,which are favored foods of Kemp'sridley. Other regions of known relativeabundance of Kemp'sridley are the coastline of Texas,the Cedar Key-Crystal River region of WestFlorida, andthe Chesapeake Bay region of Virginia LarryOgren, NMFS SEFC Panama City Laboratory,Panama City, Fla., personalcommunication, July 1985; Carr and Caldwell, 1956; Anonymous, 1984; Lutcavage and Musick, 1985!. The numberof Kemp'sridley strandings on theTexas coast has increased in thepast five years. This is probablydue to a varietyof factorssuch as the betteridentification of speciesand more efficient Sea Turtle Stranding and Salvage Network STSSN!activities. Rabalais and Rabalais 980! documentedonly 15Kemp's ridley strandings adults, 7 subadultsand 6 hatchlings!between September '1976 and October1979, in the areafrom Port Mansfield to Port Aransas,Tex. In contrast,the 1984 total for thecoast of Texaswas 68 stranded Kemp's ridley Anonymous,1984!, but alreadythere have been 34 reportedstrandings of this speciesin the first sevenmonths of 1985. It is becomingincreasingly evident that Kemp'sridleys utilize shallowwater bays throughout their known distribution.Lutcavage and Musick 985! reportedat least43subadirlt Kemp's ridleys found in a threeyear period in ChesapeakeBay. They concluded that the Kernp's ridley distribution is closelyrelated to theabundance of seagrass ~ ~Leidcchelys kempi X Carerca care>is Q Chelcnia ~des V Erecscrbelys iabricata

W W 97 i10 97 30 97 20 97 10 97 00 98 50 28 10 2810 N

N 97 30 97 10

Figure2. Distributionof documented seaturtle strandings in theCorpus Christi Bay region Does not include records for strandingsonbeaches bordering theGulf of Merica or strandings ofhead started Keek p's ridleys; Data supptiedby theSea Turt- leStranding and Salvage network, Nationa! Marine Fisheries Service, Southeast Fisheries Center, Miami Laboratory! bedsand blue crabs Calli neet es sapid us! in thebay. Subadult rid'obeys also have been found in CapeCod Bay in New England Lazell, 1980!. East coast bays appear to providedevelopmental habitat for juvenileridley turtlesin the summer. In the Gulf of Mexico, the west coast of Florida once was described as an area of maximum abundance for Kemp'sridley Carr,'1957!. Kemp's ridley hasbeen documented by Larry Ogren ibid.!in LakePontchartrain Louisiana!,Biloxi Bay Mississippi! and Choctawahatchee andApalachicola Bays Florida!. Historically, reports of Kemp'sridley in bays of the western Gulf of Mexico havebeen scarce. However, there are scattered reports of ridleys in AransasBay, Corpus Christi Bay,Galveston Bay and LagunaMadre, Tex, Hildebrand, 1980!. TheSTSSN data provide evidence that Kemp's ridleys maybe more common near inlets along the Texas coast than earlierbelieved Anonymous, 1984!. Figure 2 illustratesridley stranding locations from Aransas Bay to the north end of LagunaMadre. This map includes only untagged wild ridleystrandings compiled over the past several years. Figure2 alsoillustrates that such ridleys occur occasionally in theCorpus Christi Bay region. Hildebrand 980! suggestedthat, historically, Kemp's ridleys may have been relatively rare in Texasbays, but cautionedthat this impressionmay have resulted from inadequate reporting. Undoubtedly, most of theturtles caught incidentally by theinshore shrimp fishery are not reported.In addition,ridleys often have been misidentified as loggerhead turtles, therebymaking earlier assessments of apparently low ridleyabundance in Texasbay systems highly suspect. 9.5% 7 da

3.2% more the

% 2 mrs!

Figure3. Distribution

59 20' 10 97 so'

20'

20' 97 50' Figure4. Distribution ofrecovery sites for head started Kernp's ridley sea turtles in CorpusChristi Bay and surrounding inlandwaters iData supplied bythe National Marine Fisheries Service, Southwest Fisheries Center, Galveston Laboratory, Galveston, Tax.!.

60 Harris,Conley and Huff, 1984!.Juvenile or subadultgreen turtles carapacelength < 81cm! areknown to inhabit lagoonalwaters and bays along the coasts of Floridaand Texas. In Florida,Sebastian Inlet, Mosquito Lagoon, the FloridaKeys and Cedar Key are areas of knownaggregation. Green turt'es once flourished in Texasbays and were exploitedby the turtle fisheryin the late1800s Hildebrand, 1980 and 1982;Doughty, 1984!. Texas bays contain extensivebeds of turtle grass Thalassiatestudinum! and shoalgrass Halodulewrightii!, which provide importantfood sourcesfor thegreen turtle. Hildebrand's 982! excellentaccount of thehistorical status of greenturtles in Texas statestha t "themost important species in Texaswas and is thegreen turtle, which fed in theseagrass meadows from MatagordaBay to thelower Laguna Madre." Corpus Christi fisher nien captured green turtles primarily in Aransas Bayand Laguna Madre, Apparently those two bayssupported adult as well asjuvenile greens as average weights werereported by Doughty984! as121.5 kg. Texas' green turtle ca tch reached a recordhigh in 1890with 265metric tonsreported for theentire coast of Texas Stevenson, 1893!, By the last half of the1890s, the turtle population could no longersustain such high levelsof take.By 1900the turtle industry had dropped off significantly. Green turtles still inhabit the sameTexas bays today, but in reducednumbers Hildebrand, 1980and 1982!. Strandingand capturerecords are scarce, due partially to incompletereporting, Figure 2 illustratesgreen turtle strandingsin theAransas Bay-Corpus Christi Bay regions. All currentrecords ind ica te thatthe inshore greens are of subadult size < 81 cm!. The distribution of greenturtle observationsresembles the distribution of grassbeds in AransasBay, Nueces Bay, Corpus Christi Bay and Laguna Madre. LoggerheadSea Turtle Theloggerhead is listedas threatened.Like mostother sea turtle species, it inhabitscoastal areas. It frequents temperatewaters of theU.S. continental shelf along the Atlantic Ocean and Gulf of Mexico,foraging around rocky places,coral reefs and shellfish beds. Subadult loggerheads commonly enter bays, lagoons and estuaries Ernst and Barbour,1972;Mendonca and Ehrhart,1982!,The greatest concentration of loggerheadsin theU S.is found in Atlantic coastalwaters. Lesser concentrations are found alongthe west Floridaand Texascoasts. The loggerheadis the most abundantspecies in Texaswaters, and its preferredhabitat is theinner continental shelf rather than the bays Hilde- brand, 1980!,The subadultpopulation of loggerheadshas been under extremestress during the last 10years due to an increasein incidentalmortality in shrimp trawls.In 1984,105 loggerhc'wads were reported stranded along the Texas coast,and many of theseoccurred on Mustangand PadreIslands Anonymous,1984!, Thereare records of sporadicoccurrence of loggerheads,all of them subadults,in Texasbays. Recent strandings within the Corpus Christi Bay systemare illustrated in Figure 2. As with the other speciesdiscussed herein, the depictedstrandings of loggerheadsdo not representtotal strandingson beachesalong the Texascoast,but only those reportedfor inshoreor bay waters i.e.,there are addititional records of strandingson thegulfside beaches of theTexas coastthat are not illustrated in Figure 2!. Hfldebrand980! reportedtwo recordsof loggerheadsin CorpusChristi Bay in the 1970sand one in Rockportin 1980.There also have been numerousreports of subadult loggerheads occurring within the Laguna Madre system. The hawksbill is rare within the continental waters of the U.S. Its distribution is more tropical than that of other speciesas it preferscoastal reef habitats like thosefound in the Caribbeanand CentralAmerica, However, there are accountsof hawksbiIls in south Florida and a surprising number in Texas. Most of the Texasrecords are for very small individuals, perhapsone to two yearsof age.Many of the strandingrecords indicate that thesehawksbills were unhealthy or injured when found Hildebrand, 'I980and 1982!.The absenceof sponge-coveredreefs and the cold Texaswin tersprobably contribute to thelack of hawksbiils abundance in Texas.In 1984,10hawksbill strandings were reported on the Texas coast Anonymous, 1984!. Few hawksbiils have been observed within Texasbays. However, there are two records of hawksbills captured in Nueces Bay Figure 2!. HiMebrand 980! speculated that hawksbills that enterthe shallowwater bays of Texasprobably do not survive the cold winters if they are unableto exit in time to avoid them. Potential Impacts of the Landmass Project on Endangered or Threatened SeaTurtles Themost obvious impact of theLandmass Project on listed scaturtle speciesis thedestruction of 3,2hectare of open

~ Asstated in the following letters: a> 16August 1984, from Southeast RegionaI Director, National Marine Fisheries Service, to Col. A. L. Laubscher,U.S. Army Corpsof Engineers,Galveston District, regardingLandmass Pcrnut App]ication 17069, Corpus Christi Bay. b! 18August1984, from RegionalDirector,U S. Fish and Wildlife Service,to Col.L. Laubscher,US. Army Corpsof Engineers, Galveston District, regarding Public Notice No. 17069,City of Corpus Christi, c! 14February1985, from ExecutiveDirector, Texas Parks and Wildlife Department,to Col. L. Laubscher,US. Army Corps of Engineers, Galveston District, regarding Permit Application No. 17069,City of Corpus Christi.

61 bayhabitat. The proposed disposal site provides nursery and foraging habitat for a varietyof recreationallyand commerciallyimportant estuary-dependent organisms'. These organisms include blue crab, white shrimp Penaerrs setiferus!, southern flounder Paralichf byes Iethostignra! and a varietyof otherfin fishes. Data are lacking on the foraging behaviorof juvenile Kemp's ridleys in theCorpus Christi Bay system. However, it is welldocumented that a favored preyof ridleysisblue crab Pritchardand Marquez, 1973; Hildebrand, 1980; Lutcavage and Musick, '1985!, Productive crabhabitat occurs throughout much of theCorpus Christi Bay system, especially in nearby Nueces Bay and in the shallowwaters of RedfishBay and the East Flats of CorpusChristi Bay K.Mcadnr, Texas Parks and Wildlife Depart- ment,Rockport, Tex,, personal communication, July 1985!. Kemp's ridley habita t also occurs within this bay system. Historically,green turtles were much more abundant within the Corpus Christi Bay system than Kemp's ridley. Beforethe demise of thegreen turtle in thelate 1890s, this species was common and an active feeder on sea grasses in thebay. Hildebrand 980! concluded that, while juvenile green turtles still occur in grassybays and grass flats of thissystem, its numbers are being reduced by deposition of dredge-spoiland by channeling.The bay bottom of the proposed3.2-hectare fill sitecurrently consists of softbottom sediments, apparently without seagrass beds, In this regard,no datawere supplied by theCorps of Engineers'Biological Assessment, but lushseagrass beds exist in portionsof CorpusChristi Bay, Nueces Bay and the adjoining waters of Redfish Bay and Laguna Madre K,Meador, ibid.!.The NMFS assessment concluded that Kemp's ridley, loggerhead, and green sea turtles would suffer some loss of habitatand disturbanceif the proposedproject was permitted. Whileit is unlikelythat direct injury or mortalityof seaturtles would result from the construction activities of the proposedLandmass Project, several potential conflicts were possible, Dredging projects have caused mortalities in bottomdwelling loggerhead sea turtles in Florida'sPort Canaveral Navigation Channel Raymond, 1980; NMFS, 1984!.Mortality occurred because the density of loggerheads inhabiting thebottom of Canaveral Channel was high andmobile hopper dredges have powerful suctions. The Landmass Project plan proposed the use of a cutterhead dredgeto dig a perimetertrench. No dataare available and no studieshave been conducted on theImpacts of cutterheaddredges on sea turtles. No dredge-related seaturtle mortalities have been documented for dredgings of Texasbays in thepast, but theseactivities have never been monitored adequately, Theproposed Landmass Project would construct 425 boat slips and two smallfishing piers, Indirecteffects of additionalboat traffic and increased fishing activities on seaturtles are difficult to predict.Data obtained from the STSSNind ica ted an increase in thenumber of documented cases of boat-tur tie co11 isions in thesoutheastern U.S. B.A, Schroeder,NMFS SEFC Miami Laboratory, Miami, Fla., personal communication, July 1985!. In Virginia,5 percent ofthe documented sea turtle strandings were attributed to propeller impacts Lutcavage and Musik, 1985!. In Florida, seaturtles were found wounded by propellers more frequently ininlet areas than elsewhere L.Ehrhart, University ofCentral Florida, Orlando, Fla., personal communication, July1985!. The impact of boat traffic on turtles along the Texascoast is unknown.No correlationbetween increased incidence of propellerwounds and increase in marina developmentshasbeen reported, but no such correlation analysis of strandingrecords has been done. Reports of entanglementand hook-and-line capture of juvenile sea turtles also have been increasing, Recapture records show that16 headstarted Kemp's ridleys were caught on hook-and-line after release. Some of theserecaptures occurred in theCorpus Christi Bay area. Hook-and-line fishermen often catch sea turtles from piers and jetties Hildebrand, 1980!. Theturtles either swallow the hook or became foul-hooked, Inmost cases, such turtles are released alive, En tangle- mentof sea turtles in discarded monofilament lineand other fishing gear is of greater concern. The line entangles the flippers and head and often causesloss of a limb or death, Recommendations NMFSmade the following recommendations to the Corps to eliminateor reducepotential adverse impacts to threatenedor endangered sea turtles resulting from the proposed Landmass Project activity: 1. Theproposed non-water-dependent facilities should be constructed landward ofCorpus Christi Bay shore- line,as also recommended by NMFS'Environmental Assessment Branch, FWS and Texas Parks and Wildlife Department',In other words, NMFS strongly recommended that the majority of thedredge and fill activities be terminated, 2. TheCorps should initiate a thoroughassessment ofabundance, distribution and frequency ofoccurrence of subadultKernp's ridley and green sea turtles inhabiting the bay system, based on a! interviewswith the inshorecommercial fishermen, b!radio and sonic tagging and tracking ofseveral individuals ofboth species todetermine their movement patterns and/or c! seasonal and systematic trawling surveys combined with taggingand recapturingsea turtles. 3. A surveyof theremaining turtle grass beds should be made to detcrrnine their current use by juvenile green turtles.Specific mapping of seagrasses such as turfle grass, shoalgrass and wigeongrass Ruppia maritima! shouldbe performed.

62 4. Thefood supply of Kemp'sridley within the Corpus Christi/Nueces Baysystem should be identified toassess impactsof futurelandfill projects on the feeding grounds in thatsystem. BiologicalOpinion It wasthe opinion of theNMFS, based on the best available information, that construction of the 3.2-hectare LandmassProject was not likely to jeopardize thecontinued existence ofKemp's ridley, green, loggerhead and hawksbillsea turtles. Although a nojeopardy opinion was issued tothe Corps of Engineers Galveston District, several potentialadverse impacts were identified asdescribed above. Consequently theNMFS recommended thatthe permit toconstruct the Landmass Project be denied, Thus, the hnal decision toissue or deny the permit was left to the Corps. Conversely,hada jeopardyopinion been derived and issued, and had the Corps proceeded topermit the Landmass Project,then the Corps would have run the risk of violation ofSection 7 a!! of the FSA, However, NMFS' opinion wasthat the potential adverse effects ofthe Landmass Project would not reach a levelof"jeopardizing thecontinued existence"oflisted sea turtles. In other words, the strict language ofSection 7 a!! didnot allow the potential adverse effectsto be usedas a basisfor issuinga jeopardyopinion. Althougha no jeopardy ruling was made, the NMFS study did anticipate that there would be some low level of incidentaltake of turtles as a resultof the proposed Landmass Project, were a permitissued. Take, in thiscase, would be in the formsof harassmentand loss of habitat.Therefore, the following mandatory terms and conditions were establishedto minimize potential take if a permitwere issued by theCorps of Engineers: 1. A trawlermust survey the water in andaround the proposed Landmass Project site prior to any construction. 2. Constructionactivities must be halted if anysea turtles are observed in thevieinit. 3. Seaturtle identification posters must be displayed by thecity at appropriatepiers and marinas. As with all Section7 consultations,if new information revealedpotential unpactsnot consideredin the ~' BiologicalOpinion, or if theproject were modiAed in a mannernot previously considered, then the Corps would have to reinitiate the Section 7 consultation,

Literature Cited Anonymous.1984. The Sea Turtle Stranding and Salvage Network, Annual Report 1984. Unpublished Report by National Marine FisheriesService, Southeast Fisheries Center, Miami Laboratory, Miami, Florida, 16 p. Bacon,P.,F. Berry, K. Bjorndal, H. Hirth, L Ogreuand M. Weber. 1984. proceedings ofthe Western Atlantic Turtle Symposium. Vol.1,Intergovernmental Oceanographic Commission Association forthe Caribbean and Adjacent Regions and Food and AgricultureOrganization ofthe United Nations, Center for Environmental Education, Washington, D.C., 306 p. Caillouet,C.W., Jr. 1984. Essai de prevention de I'extinction de la tortuede Kemp. Les Carnet de Zoologic Bulletin of the Zoological Societyof Quebec!44!:28-34. Carr,A. 1957.Notes on the zoogeography ofthe Atlantic sea turtles of the genus Lepidochelys. Revista de Biologia Tropical 5!:45- 61. Carr,A.F. and D.K. CaldwelL1956. The ecology and migrations of seaturtlcs: 1. resultsof fieldwork in Florida,1955. American Musmrn Novitates 1793:1-23. Chavez,H.1969. Tagging and recapture ofthe lorn turtle ~ochelys kepi!. Internaf r'onal Turtle and Tortoise Society journal 3!:14- 19, 32-36. Doughty,R. W. 1984.Sea turtles in Texas: a forgottencommerce. Southwestern Historical Quarterly 88!:43-70, Emst,C.H. andR.W. Barbour, 1972. Turtles of theUnited States. The University Pressof Kentucky,Lexington, Kentucky, 347 p. Fontaine,C.T. and C.W, Caillouet, Jr. 1985. The Kemp's ridley sea turtle head start research project: an annual report for fiscal year 1984.NOAA TechnicalMemorandum NMK-SEFC-152, ii plus13 pp. and3 tables. Fontaine,C.T., RM Harris,W.J. Browning, T.D. Williams and S.A. Manzclla, 1989. Observations ondistribution, growth and survivalof captive-reared, tagged and released Kemp's ridley sea turtles Lepidochelys keinp>! from year-classes 1978 to 1983. This volume!, Harris,B., W, Conleyand A. Huff.1984. The status of Florida's nesting sea turtle populations from 1979 through 1983. Florida Departmentof NaturalResources, St. Petersburg, Florida, 26 p. Hildebrand,H.H. 1980. Report on the incidental capture, harrassmeut aud mortality of sea turtles in Texas, Unpublished Report to NationalMarine Fisheries Service, Southeast Fisheries Center, Pascagoula Laboratory, Pascagoula, Mississippi, 34 p. Hildebrand,H.H. 1982,A historicalreview of thestatus of seaturtle populations in the westeniGulf of Mexico,p. 447-453.In: Bjorndal,K.A, Editor!, Biology and Conservation ofSea Turtles, Smithsonian Institution Press, Washington, D.C., 583 p. Lazell,J.D., Jr, 1980. New England waters: critical habitat for marine turtles. Copeia 1980!:290-295. Lutcavage,M.and J.A. Iviusick. 1985. Aspects ofthe biology ofsea turtles in Virginia.Copeia 1985!;449-456.

63 McVey,J.P. and T. Wibbels.1984. The growth and movcrnents of captive-rearedKemp's ridley sea turtles Lepidocbdyskernpi!, followingtheir release in theGulf of Mexico,NOAA Technical Memorandum NMFS-SEFC-145, 25p, plus3 figuresand 3 tables. Mcndonca,MT. andL.M. Ehrhart. 1982. Activity, population size and structure oF immature Chelonia mydas and Caretta caretta. Copeia1982!:16 I-167, NationalMarine Fisheries Service. 1984, Biological opinion: maintenance dredging of thePort Canaveral Navigation Channel, BrevardCounty, Florida National Marine Fisheries Service, Southeast Regional Office, St. Petersburg, Florida, 72 p, Pritchard,P.C.H. and R. Marquez.1973, Kemp's ridley turtle or Atlanticridley Lepidochelyskemln!. International Union for ConservationofNature and Natural Resources, Monograph Xo. 2: Marine Turtle Series, 30 p. Rabalais, S. C. and N, N, Rabalais. 1980.The occurrence of seaturtles on the south Twas Coast. Contnbrdionsin Atarine Science 23 123-129. Raymond,P.1980. Marine turtle observations aboard dredge LONG ISLAND, Port Canaveral, Florida, 19 July-1 August 1980. UnpublishedReport to theNational Marine Fisheries Service, Southeast Regional Office, St. Petersburg, Florida, 42 p. Stevenson,C,H. 1893. Report on the mast fisheries of Texas,p. 373-420. In:Report ofthe Cornrnissioner, United States Commission of Fishand Fineries, Washington, D.C., 420 p. U9. ArmyCorps of Engineers,Galveston District. 1985. Biological assessment of impacts to endangeredor threatenedspecies for theproposed city of CorpusChristi Landmass! permit activity. Unpublished Report, U.S. Army Corps of Engineers, ' GalvestonDistrict, Galveston,Texas, ]7 p. Questions and Answers SallyMurphy: I noticed thatyour highest frequency ofstra ndings isin A pril.Do you have anycl ue as to what iscausing these Aprilstrandings? Dvthey coincide perhapswith offshore oilcoming in,or do you know what elsemight becausing theApril strandings? Whistler;Wecan only make inferences. Weneed someone tohelp us evaluate suchresults. There tend to be, for example,theeffects ofthe currents atthat time. As Tony Amos hasfound and aswe have found, themost numerous depositsofmaterial anddebris onthe beach occur atthat time. I think there isshrimping atthat time. We need to evaluatethisfurther, I have summarized thedata up to 'i985, hoping that we now can begin tolook a't these data in greaterdetail to evaluate factors that may be causing the strandings. Amos:Wecan say that strandings maycoincide withcurrents, andthey may coincide withoil coming ashore, Butour greatestchallenge isnot to say what may behappening, butto be able tosay with certainty whatare definitely the causesofthe strandings. Thatis a verydifficult thing to do. There obviously areperiods during which we know that shrimpersareout there, Butis anybody confident thatshrimpers willtell them how many turtles they catch outthere? Forexample, I canshow that the number ofdead birds, especially gulls,increases assoon asthe shrimp boats getout there.Now is that because I am prejudiced against the shrimp boats, oris tha t agenuine linkage between shrimping andoccurrence ofdead birds, or is something elsehappening thesame time that we do not unders'tand? Untilwe do somemore research and get some of the shrimpers themselves tohelp us, we will not have the answer. We must convincetheshrimpers thatwe are not trying toblame them for the strandings, butthat instead weare really trying to getto thebottom of theproblem. ThomasRennie: Was the insult that you subjected the turtles to a freshoil? Lutz: No, it wasSouth Louisiana crude oil andit wasweathered. Rennie:How longhad it weathered? Lutz:Twenty-four hours and it wasconcentrated quite a lot. Rennie.Areyou planning tolook atoil that has been weathered fora longer period oftime to try to show what might actually behappening i nthe field d ringu an oil spill event inwhich if might take three tofive days for the oil to approach theshoreline? Wouldthere be a lossof volatiles, aromatics and things like that in sucha case? Lutz:Our results are 24-hour related and that is why the oil wasweathered, Actually we found that within 12 hours therewas only a slightchange in themass of theoil. Rennie:Then a largepercentage ofthe aromatics had already been lost is thatcorrect? Lutz: Yes, within 24 hours. DavidForcucci: Concerning theeffects onthe lung, you said that the turtles increase their inspiration, andthat the ability of thelung to pickup oxygenwas lowered. Lutz: That is correct, ForcucciCould that be due not only to the effect on the lung tissue, but also due to the effect that you found on the hemoglobin? Lutz:That certainly could be a confoundingeffect, but the difference here was that the oxygen profile, the partial pressureofoxygen, was the same. It did not change. Sothere had been no change in the gradient. Forcucci:Do you mean the partial pressure ofoxygen in theblood? Lutz:Yes. There was no change inthe oxygen partial pressure in the blood,but there was a decreaseinhemoglobin. Thecompensation could be due to an increase in blood flow through an increase in heartbeat rate, which is a reasonablysimple one. All theconclusions aretentative, and all the results could have other interpretations, butthe basicinterpretation isas I statedit;i.e. that the increase ininspira tion is necessary toovercome a decease inefficiency ofthe lung in handling oxygen. That explanation isa tentativeone. There are no data to support the idea that the oil has that preciseeffect on the turtle. MilfordFletcher.' Concerning thecontrol ofhabitat loss through Section 7consultations under the Endangered Species Act,if welose eight acres ofprime habitat inone place now, and later another 20acres ofprime habitat, atwhat point does one become concededabout the cumulative long-term environmental impacts ofsuch losses? When the last eight acres of the bay get filled in, thenall wehave is a golf'course! Raymond:Prior to the consultation related to the Corpus Christi Bay Landmass Project, our office had been doing Section7 consultations forseveral years on projects such as maintenance dredging ofthe Corpus Christi Channel and otherlandfill projects like theLand Mass Project. There are many such projects in baysup anddown the Texas coast. TheLand Mass Project consultation was the first time that evidence had come forward, primarily from the NMFS SKFCGalveston Laboratory, indicating that the bays are habitat for Kemp'sridley e.g,,at leastfor headstarted Kemp'sridley!. I agreethat cumulativeeffects should be addressed. We did not decideto addresscumulative effects onthis particular project, because wewanted to consult on this as a separateproject to look at the data supplied by theCorp of Engineerson turtleabundance and habitat use in thebays, before we consultedwith themon thenext project. Amos;Yes, that is a verygood point. Corpus Christi seems tobe particularly vulnerable to thissort of thing.Just last weekone of the largest projects inTexas was announced to dredgeopen a channelbetween Mustang and Padre Islands.That is why I saywe should look at the overall effect, and not just each little bit of habita t that is disappearing.

66 StandardOperating Procedures for CollectingKemp's Ridley SeaTurtle Eggsfor the Head StartProject Patrick M. Burchfield and F. James Foley»

A summaryofKemp's ridley sea turtle Lepidochelys kempU conservation efforts atRancho Nuevo, Tamaulipas, Mexico, during1978 through 1985 was presented. Artificial hatchery techniques using polystyrene foamboxes and facsimile nests onthe beach were described. Eggcollection techniques andthe initial incubation ofeggs earnmrked forfurther incubation andimprinting atthe National ParkService's Padre Island National Seashore near Corpus Christi, Tex., were described. Experiments for the1984 and 1985 nestingseasons werepresented, Theseactivities supplied hatchlings tothe National MarineFisheries Service, Southeast Fisheries Center'sGalveston Laboratory for headstarting. Beachsand temperatures weremonitored atRancho Nuevo at a 30cm depth along a beachtransect perpendicular tothe water' s edge,Temperatures wereread three times daily, Every two weeks a 24-hour study measured beach temperature at2-hour intervals.This was done to re-adjust the times of day at which transect temp erat ures were read to assure that daily high and low temperatureswere recorded throughout the season. Thetemperature in the concrete hatchery building was mechanically regulated tocorrespond asclosely aspossible tothe temperatureregime ofthe natal beach. Moisture content ofthe incubation medium, Padre Island sand, ranged from 1 to29 percent saturationwith a meanof14 percent within the polystyrene foam incubation boxes, each containing one clutch of eggs. Moisture contentofnatural nest cavities ranged from 7 to21 percent saturation, with a meanof14 percent. Incubation temperatures at theRancho Nuevo hatchery ranged from a lowof 26,4'C toa maximumof37.5'C

'GladysPorter Zoo and Potentials, Inc,

67 intermittentexcavations when the turtle pauses in its alternatingexcavations with left andright hind flippers.Once theaccess hole is completed, the worker grasps one of the polyethylenebagsby hand roughly 8 cmdown on the neck andfolds the top portion over the hand in thesame fashion as one would do to prepare toblow up a paper bag by mouth, Whenthe turtle reaches the point where she can no longer excavate sand with her hind flippers fully extended into thenest cavity, she centers herself over the cavity to lay her eggs, At thistime the worker's bagged hand is positioned directlybeneath and encircling the cloacal opening of theturtle, On occasion the first two or threeeggs may drop beforethe bag is in position.These eggs are not retainedfor PadreIsland, but are addedto a corralnest maintained undersupervision of INPbiologists. Once oviposition commences it typically lasts roughly 15 minutes. When the femalefinishes laying she immediately begins to coverthe eggs with sand, so the top of thebag is closedquickly to avoidexposure of theeggs to RanchoNuevo sand. The most expeditious way to accomplishthis is to lift thebackend ofthe turtle with a freehand and pivot her oH the nest cavity. The female seems oblivious to what happens atthis point andoften continues to cover,shell-tamp and obliterate the "nest," even though removed more than 1 meterfrom the actual nest site. Handling of Eggs Afterthe eggs are collected, the bag containing the eggs is placedon the sand while vital data are recorded and the nesteris measured and tagged. However, the hot sand at the surface isbrushed asideby hand or footprior to putting thebag of eggson thesand. The same is donewhen Kemp's ridley eggs are coll or ted for theRancho Nuevo corrals becausethe extremely high temperaturesof thesurface sand may kill eggsin directcontact with it. Eggsfor Padre Island are protected from heat of the sun and taken immedia tcly to camp. This is best accomplished byplacing them in a polystyrenefoam box on a cushionofPadre Island sand, Qn occasion, it is not possible to return immediatelytocamp when there are large numbers ofturtles on the beach and the workload isheavy. In this case the bagof eggs is sealed tightly with data card inside, then placed inside the nest cavity and covered with cool sand. The locationismarked for easy retrieval atthe earIiest convenience, Theentire nesting process requires approximately 45 minutes,and during that time the attending worker is not available for other duties of thecamp's work force, Padre Island Sand Eggsfrom each clutch destined for Padre Island are placed in a separateincubation box containing incubation medium,Padre Island sand, as soon as possible. Dimensions ofthe polystyrer e foariboxes are 19.5 x 27.9cm and 35.6 cmdeep. Each year since 'I978, Padre Island sand has been collected and transported in such boxes from the Padre IslandNational Seashore to theRancho Nuevo sanctuary and back. Moisture Content Usingthe technique described byMcGehee 978!, we determined that 22 ml of distilled water added to 100 g of oven-driedRancho Nuevo sand produced 100 percent saturation, Sand samples collected from Padre Island in 1983, 1984and 1985 were oven-dried and also required 22 ml of distilledwater per 100 g sandto achieve 100 percent saturation.The amounts of added water required were determined from 12 trials each with samples ofRancho Nuevo and Padre Island sand, Ourstudies in 1983,1984 and 1985 at Rancho Nuevo of 21 natural nests at mid-nest depth of 30 cm indicated that femaleKemp's ridleys selected nest sites with moisture contents ranging from 7 to21 percent saturation .5-4.6 ml ofwater per 100 g ofsand! with a meanof 14 percent, Sand used for incubation ofeggs translocated toPadre Island shouldbe of thesame moisture level as that selected byKemp's ridleys nesting at Rancho Nuevo, Therefore, we consider14 to '16percent saturation .1-3,5 ml of waterper 100 g ofsand! as optimum moisture content for Padre Islandsand used for early incubation and embryonic development ofKemp's ridley eggs. Distilled water is added, if necessary,to raisethe moisture content to theproper level, Placing Eggs in Boxes Oncethe eggs have been transferred from the beach to the Rancho Nuevo research station they are placed in the incubationboxesassoon as possible. Sterile surgical gloves are used in handling the eggs, First, a layerofPadre Island sandabout 2.5 cm deep isplaced inthe bottom ofthe box prior to adding the first layer of eggs. The eggs are placed inrows for ease ofcounting, A margin 2.5cm wide is left between theeggs and sidewalls ofthe box, then this space isfilled with Padre Island sand. Next, the remaining eggs are placed in rowson top of thc first layer and cloacal fluids fromthe polyethylene bagare poured over the eggs. A layerof Padre Island sand at least 2.5 cm thick is poured over thetop of the eggs and leveled off. Weight isan important factor in airlifting the boxed eggs to the US., so excess sand is avoided. Aseggs are placed into the box, the worker judges how many layers ofeggs the clutch will form, so a temperature

68 probecan be placedin thecenter of theclutch, The probe is usedto monitorincubation temperatures at thePadre Island hatchery, Rancho Nuevo Hatchery TheRancho Nuevo hatchery is a concretestructure containing shelves to holdboxed eggs during the early stage of incubation,At eachshelf level within the hatchery,ambient air temperatureand temperatureof a boxcontaining onlysand as a control!are monitored, Shelf levels 1, 2 and3 are168 cm, 124 cm and70 cm from the floor, respectively. Thehatchery building is mechanically thermo-regulated toapproximate natural incubation temperatures asclosely aspossible. There is no constantsource of electricityat thecamp, so temperature control by electricalcooling and heatingis notpossible, A thermaltransect established on thenesting beach provides information to guidehatchery thermo-regulation.Bailey BAT-23 digital read-outthermometers are used. Temperatures at 30cm depth are read threetimes daily. In addition,a 24-hour at 2-hourintervals! study is doneevery two weeksto providedata for adjustingtimes of daywhen transect temperatures are read, so that daily high and low temperatubs of thebeach sand can be monitered throughout the season. Transport Ideally,eggs should be exported to Padre Island within the first 24 hours following their collection, This has not been possiblefor severalreasons, First and foremost, export permits are required from Mexico's Instituto Nacional de la Pescaand Mexico'sSecretaria de RelacionesExteriores, and authorizationis required from the Secretariade Aero- nauticalCivil authorizingan U.S. aircraft to exportthe eggs. A MexicanINP inspector must issue a documentcalled a Guiade Pescain effectfor 72hours, whereby Officials of Mexico'sINP donate the eggs, on site,to theU,S, gov- ernment,in compliancewith Mexicanlaw. Theselegal requiremen'ts coupled with the unpredictablearrivals of nesflngturtles and vagaries of wea'theraffecting both vehicular and aircraft transportation 'to and from the site make long-termplanning difficult, if notimpossible. Therefore, it seems best to alloweggs to developfor 2 5weeks or more beforetransporting them to Padre Island to avoidkilling embryos by movingthem too soon. After this initial interval of incubation,eggs can be moved safely and with goodsuccess, barrmg some catastrophic occurrence. A typicalbox loadedwith a clutchof eggsand filled with sandto not more than 2.5cm abovethe uppermostlayer of eggsweighs about 14 kg. Twenty boxes containinga total of 2,000eggs wouM weigh about 0.28metric tons, and 30 boxes containing3/00 eggswould weigh as much as 0,42metric tons, Actual weights could vary aboveor below these averages. Accordingto King, Shaverand Phillips 984!, the incubationboxes had walls 22ml thick and an internalvolume of 23/50 cubiccm. The shapeof boxesoriginally usedwas not suitablefor loading and shipping, sodifferent boxes were selected that contained 16390 cubic cm of interior spaceand were more nearly square. Holes, 6 ml in diameter, were drilled in eachbox; 30 eachin the top and bottom, eight through eachside and six through eachend, We recommendcontinued use of boxesof this sizeand shape,both for convenienceand for continuity in technique. Eachbox containingeggs must be markedwith genus,species and number of eggscontained within to complywith U.S.import regulations,Persons accompanying the shipment must possess a U.S.Fish and Wildlife Servicethreatened andendangered species permit and a CITES Convention on InternationalTrade in EndangeredSpecies of Wild Fauna and Flora! export permit in addition to the requisiteMexican perinits previously mentioned.The shipmentmust depart from the Republicof Mexicofor the U,S,via the MatamorosAirport. To avoid dilemmasthat might occurat the border, the Mexican Consul must be provided copies of all appropriate documents and his assistancerequested in dealing with authorities at the Port of Matamoros. The earthenairstrip at RanchoNuevo is approximately0.3 km long.When the strip is dry and in good condition, a Cessna206 with baHoontires cantransport the pilot, onepassenger and up to 20boxes of eggsand sand.A Cessna 207with balloon tires is necessaryfor 30 boxes. From mid-June through July, the rainy seasonbegins, and the airstrip is usableonly intermittently. In both 1984and 1985,landings and takeoffswere made on the beachbecause the landing strip was inundatedby water. When landings and takeoffson the beachare necessary,flight times for transportof theboxed eggs must coincide with low tides.This can present additional problems in that theMatamoros Airport has no landing lights and closesat night. Thus, egg export flights must occur only during daytime, Oncethe flight leavesMatamoros Airport, it can continueon to PadreIsland NahonalSeashore where boxes are offloadedand transferredto the hatchery,At that time, the National Park Servicetakes over and completesthe incubationand imprinting phasesof headstar ting Kernp'sridleys. It is hoped that the preceedinginformation will serveas a helpfulguide to thosewho in the future areinvolved in the long-termeffort to establisha newnesting colony of Kemp'sridley on PadreIsland through headstarting.

69 Literahxre Cited King,R,, D.J. Shaver and K. Phillips. '1984. Padre Island National Seashore 1984 report: Kemp's ridiey sea turtle restotation and enhancementproject, incubation and imprinting phase.U.S. Department ofthe Interior, National Park Service, Pad re Island NationalSeashore, 'l20 p. McGehee,A.1978. Factors affecting thehatching success ofloggerhead seaturtle eggs. American Societyof Biofogis}s Bulletin 25!90,

70 Beach Temperature Versus Polystyrene Foam Box Temperature in 1ncubation of Kemp's Ridley SeaTurtle Eggs Robert E. King'

Center-clutchtemperatures wererecorded hourly in 20polystyrene foam boxes containing sand in whichKemp's ridley sea turtle Lepidochelys kempi! eggs wereincubating fromJune 23,1982, through theonset ofhatching. During this same period, weekly24-hour beach temperature profiles atsand surface and depths of15, 30 and 45 cm were recorded onthe Kemp's ridley intprintingbeach atPadre Island National Seashore near Corpus Christi, Tex. Throughout theentire incubation period, temperatureswere general lylower in thepolystyrene boxes than at any of the three depths onthe impri nting beach. There were alsodifferences between boxes and beach in daily temperature range and in timesof occurrence andduration ofmaximum and minimum temperatures. A direct24-hour co~parison between center-clutch incubation temperatures in a single polystyrene boxand temperatures at theimprinting beach at30 cm depth the Kem p's rid ley "mid-nest" depth on the natural nesting beach atRancho Nuevo, Mexico} wasalso made weekly throughout theincubation period. During the first two trimesters ofincubation, polystyrene boxcenter- clutchincubation temperatures were1'to 3 Ccooler than the temperatures at30-cm depth on the imprinting beach for each 24- hourperiod. Clutch metabolic heating during the last trimester rai~".d the center-clutchincubation temperatures dramatically abovethose recorded onthe imprinting beach atthe 30-cm depth. Temperat uredifferences such as those seen d uring the first two trimesterscan be expected toalter sex ratios in clutchesincubated in boxes as compared to those incubated in situ. Only polystyreneboxincubation which mimics the mid-nest teniperature pro/le of a naturalnesting beach, should beconsidered as a viableoption for seaturtle conservationprogrants. Incubationof seaturtle eggs in polystyrenefoam boxes has played, and will continueto play,an important role in manysea turtle conservation efforts. However, this method places the eggs in anartificial environment in which incubationtemperature may bear little resemblanceto that of thenatural nesting beach. Thediscovery that sex of sea turtles is determined by incubationtemperature emphasizes theneed for understand- inghow natural and artificial incubation of eggsdiffer in theireffects on sexratios of clutches,If boxincubation is to remaina viablepart of seaturtle conservation efforts, these effects must be clearly understood. Theuse of polystyrenefoam boxes containing beach sand for the incubation of seaturtle eggs is among the most widespreadand simplest tools of seaturtle conservation Mrosovsky, 1983!. It hasalso become controversial because it is knownthat incubation temperature determines sex in seaturtles, the cooler temperatures having a masculinizing effect.Polystyrene foam boxes are used for incubatingKemp's rid! ey sea turtle Lepidochelyskernpi! eggs translocated from RanchoNuevo, Mexico to the Padre Island National Seashorenear Corpus Christi, Tex, Therefore,it is imperativeto know how temperaturesin the boxesand on the beachcompare. Thepurposes of thisstudy were two-fold: ! to determinethe typical pattern of temperaturesfrom surface to 45 cmdepth for theKemp's ridley imprinting beach at PadreIsland National Seashore; and ! to determinedifferences, if any,between this imprintingbeach's temperatures at 15,30 and 45 cm depthsand simultaneouslyrecorded temperaturesin polystyrene foamboxes containing incubating Kemp's ridley eggs. This study was conducted under theincubation and imprinting phase of theU.S.-Mexico Cooperative Project designed to establisha breedingcolony of Kemp'sridley seaturtles on PadreIsland, Tex, Methods Sixrows of threethermocouple probes total of 18probes! were established on the Kemp'sridley imprinting beach at the Padre Island National Seashore,from the high tide zone to half-way up the foredunes.Each probe had temperaturesensors at15,30and 45 cm depths. Rows and columns in thearray of probeswere15 m apart,so the array spanneda segmentof beach30 m wideand 75 m long,from the high tide zone to halfwayup theforedunes. One day eachweek, beginning at theend of Mayand continuing through mid-August 1982, temperatures at thesand-surface and at 15,30 and 45 crn depthswere recordedevery two hours over a 24-hourperiod. 'Departmentof theInterior, National Park Service OnJune 23, 20 polystyrene foam boxes containing Padre Island sand and incubating Kemp's ridley eggs were receivedfrom Rancho Nuevo. Each box contained a thermocoupleimplanted within thecenter of theclutch when theeggs were placed in boxes.The boxes were transferred to a screenedshed hatchery!and incubation continued underambient air! temperature conditions. Incubation temperatures were recorded hourly for each clutch one per box!from theirarrival until thecommencement of hatching.

Results Figure1 depictsa typical24-hour cycle in beachtemperature profi/e. Twenty-four hour trends of temperaturesfor all sixrows of thermocoupleprobes were similar in formthroughout the monitoring period although differences betweenrows, within rows and among dates were noted. Surface temperature maxima were much higher for probes locatedin theupper portions of thebeach than for those in thetidal zone, but the beach location of probes from tidal zoneto dune! had a diminishingeffect on temperatures asdepth of the sensors increased, Heat transfer through the sand,from the surface into deeper layers, occurs at a constantrate. Rain of brief duration altered the form of Figure 1 bycausing a sharpdecline in sandsurface temperature, but it affectedtempera tures at various depths only slightly. However,prolonged periods of heavyrain probably would have a pronouncedeffect on subsurface temperatures, but this wasnot encounteredduring the study. Thetemperatures at 15,30 and 45 cm depths show a dielcycle related to surface temperature, but heattransfer from thebeach surface through the sand is delayed Figure 1!. Each weekly, 24-hour beach temperature series for row 3 rowswere numbered 1-6 from tide zone to dunes! at 15, 30 and 45 cm depths was compared totemperatures inbox 645 Figures 2-7!. These comparisons were restricted to a singlerow of teriiperatures and a singleclutch of eggsfor simplicity.Row 3 waschosen because it was located atapproximately mid-beach. Eighty to ninety percent ofKemp's ridleynests are laid fromthe mid-beach to thetop of thefirst foredune at Rancho Nuevo Burchfield et al,, 1984!, correspondingtorows 3 6of this study. Clutch 645 was chosen because it represented a typical and successful clutch 24 eggs,92 percent hatch and incubation period 52 days! and its incubationperiod closely matched the beach temperature monitoring period. Accordingto Burchfield,Foley and Noyes 983!, Bure'h field et al, 984! and Burchfieid and Foley 985!, mid-nest depthfor Kemp's ridley is 30 cm. In theearly stages ofincubation Figures 2-4!, before metabolic heat production occurred,temperatures at30 cm were generally 1 to ' 3'C warmer than incuba tion temperatures inbox 645. In the later stagesof incubation Figures 5-7!, metabolic heat from embryos first raised incubation temperatures to those at 30 cm Figure6} depth and then significantly higher, This temperature relatIonship continued until hatching occurred. Sexdetermination probably occurred during the middle trimester of incubation Mrosovsky, 1983! in box645 Figures 4 and 5!. Peaktemperatures occurred atdifferent times of the day in the incubation box than at the 15, 30 and 45 cm depths on thebeach throughout the study, The daily peak usua lly occurredbetween hours 1800 and 2000 in theboxes and betweenhours 1600 and 1800 at 15cm, hours 2000-2200 at 30 cm, and hours 0000 and 0600 at 45cm on thebeach. In addition,the duration ofthese daily peak temperatures wasdifferent. The peak temperature at45 cm depth on the beachoften lasted four to six hours as contrasted with peak temperatures within the box and at 15 and 30 cm depths, whichseldom lasted longer than two hours, Disparities ofa similarmagnihide were found for the time of occurrence andduration ofminimum temperatures, Theamplitude ofdaily teniperature excursions wasgreatest at15 cm depth 4' to6 2'C!but much less within the box I 1' to 1 7'C!,at 30 cm depth 6' to25'C! onthe beach, and at 45 cm depth 9'to 1.2' C! on the beach, Thus, at mid-nest depth 0 cm!for the Kemp's rid Icy, the time of occurrence and durationofmaximum and minumum temperatures anddaily range of temperatures werenot markedly different fmm those of the incubation box.

Discussion Themost dramatic and probably significant difference between incubation box and beach mid-nest depth temperaturepat ternsin thisproject is the 1 to' 3' C loweringof incubation box tempera tures during the first two tri- mestersofdevelopment, Differences intemperature pattern between incubation boxes and beach atnest depth are notconfined tothe Kemp's ridley project, and may vary with species and hatchery conditions. Processes responsible forsuch differences arecommon toall conservation efforts employing polystyrene foam box incubation ofsea turtle eggsunder ambient conditions. Specifically, incubation temperature inthe boxes fo! lows a cyclesimilar to that of the surroundingambient air! tempera tures. Temperature ranges and times of occurrence anddura hon of peak tem- peraturesvary with nest depth. Thus, sand at greater depths gains heat more slowly and retains heat longer. Ampli- tudeof thetemperature excursion diminishes or dampenswith increase in depthfrom the surface on the beach. 4 33

31

+ 29

25 1200 1600 2000 0000 HOUR Figure1.Typical Padre Island beach sandtemperature profileatsurface andat15, 30 and 45cm depths overa 24-hour period.

32

31

O 30

29

28

26

1 I $200 1600 2000 0000 0400 0800 HOUR Figure2.Padre Island beach sandtemperature at30 cm depth andincubation temperature forct utch 645 ina polystyrenefoamboxover a 24- hourperiod incubation days 10 and 11, june 24-25, 19S2!.

13 3 31

30 I- 2829

26

ss I l . l 92 ~.. I ~ L 1200 1600 2000 0000 0400 0800 HOUR Figure3.Padre island beach sand temperature at30 crn depth and incubation tentperature fordutch 645 i'n a polystyrenefoambox over a24- hourperiod incubation days 17 and 18, July 1-2, 1982!.

32

31

O 30

29

28

27

26

251200 1600 2000 0000 0400 0800 HOUR Figure4.Padre 1sland beach sandtemperature at30 crn depth andincubation temperature fordutch 645 in a polystyrenefoamboxover a24- hourperiod incubation days 24 and 25, July 8-9, 1982!.

74 32

30

29 I-

28

26

ss~ I 1200 1600 2000 0000 0400 0800 HOUR Figure5. Padre island beach sand temperature at30 cm depth and incubation tanpera! ure for clutch 645 in a polystyrenefoamboxover a 24- hourperiod incubation days 31 and 32, july 15-16,1982!.

32

.o 30

29 I-

28 BEACH-30 CENTIMET

INCUBATION BOX I 27

26

25 J 1200 1600 2000 0000 0400 0800 HOUR Figure6. Padre Island beach sand temperature at 30cm depth and incubation temper ature for clutch645 in a polystyrenefoam box over a 24- hourperiod incubation days 3S and 39, july 22-23,19S2!, 34 O32

30

28 I 1200 1600 2000 0000 0400 0800 HOUR Figure7. Padre Island beach sand teroperature at30 centi meterdepth andincubation temperature Jarciutch 645 in a polystyrenefoarnboxover a 24-hourperiod incubation days 45-46, 29-30 July 1982!.

Cortdusion Differencesin incubation temperature between polystyrene foam boxes and natural beaches have unportant implicationstosea turtle conservation programs. Incubation ofKemp's ridley eggs in polystyrene foam boxes under ambientconditions exposes theeggs to cooler tempera tule duringthe first two trimesters ofincubation possibly the periodwhen sex determination takes place!, as compared to temperaturesin sandon thePadre Island beach at mid- nestdepth. Raising the temperaturein the incubationshed could eliminatethe lower-than-naturalincubation temperatureexperienced byeggs in polystyrene foamboxes during the first two trimes ters. Control of air temperature in thehatchery may be the only way to simulatethe temperature pattern of thenatural nesting beach during incubationof eggsin polystyrenefoam boxes. Such control would be critical during the period of sexdetermination in incubating eggs. In seaturtle conservation programs in whichfoam box incubation may be necessary, steps should be taken to controlincubation temperature inways that mimic the tempera ture pa ttern a nestt depth on natural nesting beaches. If thiscannot be done, eggs should not be removed from their natural incubation conditions until the period of sex determinationhas passed.

Literature Cited Burchfield,P.M. and F.J. Foley. 1985. Report ou Republicof Mexico/UrutedStates of Americaconservation effort ou behalfof Kemp'sridley sea turtle at Playade Rancho Nuevo, TamauBpas, Mexico, 1985. 59 p. Burchfield,P.M., F.J. Foley and P.T. Noyes, 1983. Report on Republic of Mexico/UnitedStates of Americaconservation effort on behalfof Kemp'sridley sea turtle at Playa de Rancho Nuevo, Tamauli pas, Mcvtco, 1983. 37 p. Burchfield,P. M.,F. J. Foley, C. Spitzner and S. L. Huntress.1984. Report on Republic of Mexico/UrutedStates of America conservationeffort on behalf of Kemp's ridley sea turtle at Playa de Rancho Muevo, Tamaulipas, Mexico, 1984. 218 p. Mrosovsky,N.1983. Conserving seaturtles. The British Herpetological Society, Landon, England, '176 p.

76 Predicted Sex Ratios From the International Kemp's Ridley SeaTurtle Head Start Research Project Thane R. Wibbels, Yuki A. Morris, David W. Owens, Gayle A. Dienberg, Julia Noell, JorgeK. Leong,Robert E. King and RendMirquez Millan'

Sexratios were determined for groupsof turtlesfrom six year-classes ofKemp's ridley sea turtles Lepidochelyskern pi! from thejoint U,S.-MexicoKemp's Ridley Sea Turtle Head Start Research Project. Gonadal histology, necropsy, laparoscopy, tail lengthevaluation and serum testosterone assays zuere used to determinesex, Turtlesexamined from the 1982 n=92, 2.9M:1F! and 1984 n=159, 2.5M:1F! year-classes were significantly male-dominated, Theturtles examined from th@ 1978 n&2, 1,9M1F! and 1979 n=22, 1,4M:1F! year-classes also were male-dominated, but not significantlyso. Those examined from the 19S3 n=1 2, 1M:1F!year-class had equal representation of the sexes, and those from the1981 n=4,OM:1F! year-class were all females,Data from the1984 year-clos also indicated that sex ratios of clutchesfrom a givenyear-class can differ significantly from one another. Thesignificantly male-dominated groups of turtles from the 19S2 and 1984 year-classes and the significant differences itt sex ratiosamong clutches of the 1984year-class are consistent with thehypothesis that thesex of Ketnp'sridleys is determined environnMntaliy.Because the numberof turtlesexamined from mostyear-classes was small, our interpretationsshould be consideredwith caution, Conservatively, we concluded that relatively large numbers of both male and fenmle Kemp's rid leys have beenproduced by theproject in thepast. However, a significantmasculinixation was evident in two of theyear-classes. The itnplicationsof male-dominancein Kemp's ridley sex ratios are discussed,

All seaturtle speciestested so far have been shown to possessenvironmentally dependent sex differentiation systems Yntema and Mrosovsky,1980, 1982; Miller and Limpus,1981; McCoy, Vogt and Censky,1983; Mrosovsky, Dut ton andWhitmore,1984!. This facthas serious implications for present-dayconservation programs incorporating artificial incubation under which biasing of sex ratios is possible Mrosovsky and Yntema, 1980;Morreale et al., 1982; Mrosovsky, '1983;Dutton, Whitmore and Mrosovsky, '1985!. Somesea turtle conservationprograms use artificial incubationfollowed by short-termcaptive rearing and release of the turtles. Captive rearing is employed to reduce high natural mortality that occurs during the first year of life. Theassumption is that the percentageof hatchlingssurviving to reachsexual maturity could be increasedby such programs, If short-term captive rearing proves to be an effective management technique, then artificial incubation of eggsin suchcases could havea muchgreater eRect on populationsex ratios than when usedin programsthat release hatchlingsimmediately into the wild, Kemp'sridley seaturtle Lepidochelyskempt! is the mostendangered species of seaturtle. Its current conservation program includes the Kemp's Ridley SeaTurtle Head Start ResearchProject that uses artihcial incubation of eggs followedby captiverearing of the turtlesfor approximatelyone year Klima andMcVey, 1982!, Head started Kemp's ridley sex ratios have a probability of being impacted by artificial ina~bation. The experimental head start project also involves imprinting in hopes of establishing a new nesting colnny of Kemp's ridley on the Padre Island National Seashorenear Corpus Christi, Tex. Kemp'sridley apparentlypossesses environmentally dependent sex determination Wibbelset al,, 1985;Shaver et al., 1988!.Therefore, the consequencesof artificial incubationon Kemp'sridley sexratios produced in thehead start projectrequire evaluation.The purposeof this paperis to review and evaluatesex ratio data from six year-classes 978-1979 and 1981-1984!of Kemp's ridleys. Sexing Techniques Varioustechniques have been used by researchersto sexKemp's rid ley. Histologicalexaminationof hatchling and posthatchlinggonads as describedby Yntemaand Mrosovsky,1980! was shown to be definitive in sexingKemp's ridleys that died afterhatching or during headstarting Wibbelset al., 1985!. Direct observation of thegonads during

' Wibbels,Oruens, Dienbcrg, and Noel TexasA&M University;Morris 7heUniversity of Texasat Austin;Leong LettermanArmy Medical Center;Xing NationalPark Service;Marque=-Institulo Nanonal de la Pica

77 necropsyhas been used to sex pos tyearling turtles Wood, 1982!. A testosteronesexing technique described by Owens etal. 978! wasused to predictsex of living Kemp'sridleys that were at leasttwo yearsold Wibbelset al., 1985; Y. A.Morris and D. W. Owens, personal communication!, Thepredicted sexes of 19out of 55turtles sexed by the testosteronemethod were verified via necropsy in thosethat died!, laparoscopy described by Wood et al.,1983! or by tail lengthevaluation Y.A. Morris and D.W, Owens, personal communication!, Results and Discussion A summaryofsex ratiodata pertaining tothe1978-1979 and1981-1984 year-classes ofhead started Kemp's ridleys isshown inTable I, Datafrom the1984 year-class supported themost comprehensive examination ofa Kemp'sridley year-classtodate, with 10 turtles sexed from each of 14clutches, and from one to eight turtles sexed from each of the remainingfive clutches. The aggregate sexratio n=159,2.5M:IF! for the 1984 year-class samples was significantly maledominated P<0 05!. If thissex ratio is representative ofthe entire 1984 year class, it would seem that eggs from the1984 year-class were incubated slightly below the pivotal temperature i.e., that temperature that wouldproduce a IM:IFsex ratio; see Shaver et al1988!. Sex ratios of samplesof turtlesfrom 19 clutches of the1984 year-class, are listedin Table2. A Chi-squarccomparison among the sex rafios of the14 clu tches from which 10 turtles were sexed indicatedthat they were heterogeneous i.e., sex ratios differed significant1y at P<0.05!, Such results could have derivedfrom different incubation temperatures among the clutches, Sex ratin differences among clutches and the significantmale-dominance ofthe aggrega tescx ratio for the 1984 year-class areconsistent with the hypothesis that Kemp'sridley possesses an environmentallydependent sex determination system. Furthermore, these results emphasizethat sex ratios vary significantly among clutches of a year-class,soan aggregate sex ratio based on combinedclutches would not be representative ofall clutches of a yearwlass,and the particular choice of clutchesin theaggregate could influence the sex ratio of theyear-class as a whole. Onealso must be concerned about adequacy ofsample size within the chosen clutches in evaluatingsex ratios and theirrelation to incubationtemperature. Samples from the 1978, 1979, 1981 and 1983 year-classes Table I! arenot adequateforall clutches inthose year-classes. Conservatively, results for those samples suggest that relatively large numbersof bothmales and females were produced within those year-classes and that there was variation in incubationtemperature among clutches within theyearwlasses. Though92 turtles of the 1982 year-class were sexed, all clutches were not represented. Those that were represented hadonly a fewsexed individuals n<5, Wibbels etal 1985!. Fur thermore, ofthe six clu tches wpresented bythe largest samples

78 's ridleysea turtles by year-class.

e, Sex Chi ratio squar

1.9M:IF 3.1250 1.4M;IF 0.7272 OM;IF 4.0000 1.7M:IF 1.6364 2.9M:IF 16.514 IM:IF 0.0000 2,5M 1F 29.9434 ascompared to knownsex as dete turtles from the i 978and 1982year-c

M:IF sex ratio at P<0,05.

rtles of the 1984year-class t

ample size, n

10 10 10 'i04

10 108

10 10 1024

10 10 10 10I

ho Nuevo, Tamaulipas, Mc Spotila985! estimated that 71 percent ofthe hatchling green turtles Chelonia mydas! produced atTortuguero are female.Mrosovsky, Hopkins and Richardson 984! estimated that the sex ratio of hatchling loggerheads Caretta caretta! produced onseveral beaches inGeorgia and North Carolina was close to IM:IF. Mrosovsky, Dutton and Whitmore984! estimated hatchling sex ratios of green turtles and lea therback turtles Dermochelys coriacea! from a beachin Surinameto be 0.9M:IFand IM;1F, respectively, Sexratios have alsobecn estimated for the subadult and adult portion of sea turtle populations. Limpus 985! has founda significantmaledominance inresident loggerheads onHeron Atoll on the Great Barrier Reef, Wibbels etal. 987!found a significantfemale dominance insex ratios of subadultloggerheads captured along the Atlantic coast of the U.S. Studiesmentioned above represent initial attemptsat estimatingsea turtle sexratios in natureand are far from conclusive,They suggest that sexratio varieswidely includingboth significantmale or femaledominances! in naturalsea turtle popuiahons.Thus, it wouldnot be unreasonableto expect that male-dominated sex ratios could occurin naturalpopulations of Kemp'sridley. Nevertheless, them is noreason to believethat male do~inance of sex ratioswould enhance reproductive success of Kemp's ridley or anyother sea turtle species. FemaleKemp's ridleys are known to beannual and multiannual in theirreproductive cycles Pritchard and Mhrquez,1973!, but the reproductive cycle of males isnot known. If males are primarily annual breeders and if only onemale is required to fertilizeone or more females, then sex ratios domina ted by males or with equal numbers of malesand females may represent surpluses ofmales. If this is true, the total egg production ofthe population might beincreasedby shifting the sex ratio toward more females. Obviously, abetter understanding ofreproductive ecology ofboth male and female Kemp's ridley is needed to determine the validity of this hypothesis. Until such information isavailable, a conservative approach might be to produce a IM:1F sex ratio in head started Kemp's ridleys, thereby preventingdepartures from 1M:IF that might decrease reproductive success within the population. A 1M:1Fsex ratio alsowould be consistent with that expected from classic theory on sex allocation Fisher, 1930!, assuming that species withenvironmentally dependent sex determination conform to classictheory. Wealso offer the more controversial opinion that production offemal~ominated sexratios might be advanta- geousfor severalreasons, Any new population established by thehead start project that nests on PadreIsland probablywould be composed predominantly if not totally of turtles from the project. Data reviewed in this paper indicatemale-dominated sexratios in the 1984 and possibly the 1982 year-classes. Therefore, temporary production offemale-dominated sexratios might compensate theimbalance created by past male-dominated sexratios, Even if headstarted turtles nest on a naturalnesting beach such as that at Rancho Nuevo, instead of Padre Island, a female- dominatedsexratio also could compensate forthe possibly highet mortality experienced byadult females inthe past dueto their greater vulnerability tonatural predators and to man when nesting. Finally, regardless ofwhere head startedfemales nest, if Kemp'sridley reproductionis female-limited,an increasein numberof femalesin the populationshould directly increase total egg production. Giventhe various considerations above, we suggest that it maybe advantageous toincrease incubation tempera- turesslightly for future year-classes of Kemp's ridley eggs, assuming a pivotal temperature of30' C Shaveretal., 1988!as a guide.The amount of increase inincubation temperature needed could be based on the specific sex ratio desiredas extrapolated from availablesex ratio data ibid.!, In closing,we hope this discussion with its focus on currently unanswered questions shows how critically importantit is to continue and intensify studies oneffects ofincubation temperature onsex ratio in Kemp's ridley. Ackno~ledgements TheKemp's Ridley Sea Turtle Head Start Research Project ispart of a cooperativeinternational effort to save the endangeredKemp's ridley sea turtle from extinction. A wide variety offederal, state, and local agencies andorga- nizationsparticipate in thisprogram, including the Instituto Nacional de la Pesca of Mexico,National Marine Fish- eriesService, National Park Service, U.S. Fish and Wildlife Service, Texas Parks and Wildlife Department and others. In referenceto the material from which our paper is derived, we acknowledge the efforts of DickieRevera and MarcelDuronslet of the National Marine Fisheries Service forpreparing many of the gonads for histological examination.Major funding for the histology work was provided by National Park Service Contract No. PX7490-5- 0100.Supplemental funding which supported completion ofthe histology, hormone assays and this manuscript was providedbyTexas A&M University Sea Grant College Program Grant No. NA81-GA-C-00039 anda SeaGrant Marine Fellowship to ThaneWibbels!.

Literature Cited Dutton,P.H., C.P. Whitmore and N. Mrosovsky. 1985. Masculiration oflesthcrback turtle, Derrnochetys coriacea,hatchlings from eggsincubated in Styrofoamboxes, Biologicat Conservation 31249-264. Fisher,R.A. 1930. The Genetical Theory ofSelectiort. Ciarertdon Press, Oxford, 272 p, Klima,E.F. and J.P. McVey. 1982. Headstarting theKemp's ridley turtle, Lepidochetys ken

UnhatchedKemp's ridley sea turtle Q,epidochelys kempU eggs ,656! from the 1980 and 1982-1987 year-classes zoere preservedafterfull-term incubation. Eggs were opened, examined todetermi nefertility, and embryos were removed. Embryo developmentalstageszoere compared tothose described byCratz 982! for olive ridley L. olivacea!and were classified accordinglyif they were compatible orwere placed inaddi tional erected stages if they zoere not. Obvious external deformities of embryosandhatchlings were noted and compared. Overall fertility of the eggs from theseven year-cases wasat least 95 percent. Ofall unhatched eggsexami ned, 19.9 percent were infertile, 38.8 percent were fertile but contained noidentifiable embryos, and38.5 percent contained embryos. Ofthe embryos found, 93.9 percent were staged. Most unhatched eggsand embryos ofthe 1982-1984and1986 year ckrsses ceased development during the first trimester ofincubation, while most of the 1980, 1985 and 1987year-classessuccumbed during the last trimester, Deformities wereobserved in83 percent ofthe staged embryos and05 percentofthe hatchlings examined fromthe1982-1987year-classes. Neithertransport byaircraft nor low fertility zueresignificant factorsleading to thehigh mortality observed in eggs from the 1983 year-class. TheKemp's ridley sea turtle Lepidochelyskenqn! is themost endangered species of seaturtle Pritchard,1969; Anonymous,1978!, In aneffort to enhance the population and establish a second breeding colony, an international cooperativeRestoration and Enhancement Program was undertaken in 1978. Participants in the program include the InstitutoNacional de la Pesca INP! of Mexico,the U.S. Fish and Wildlife Service FWS!, the National Park Service NPS!and Na tiona1 Marine Fisheries Service NMFS!. During each summer since then, approxima tely2,000 eggs have beengathered atRancho Nuevo, Tamaulipas, Mexico, packed inPadre Island, Tex,, sand in polystyrene foam boxes oneclutch per box! and transloca tedto the Padre Island Na tionaI Seashore nearCorpus Christi, before hatching. Yearlymean hatch rates for these eggs have ranged from 64 to 91 perce~t, except in 1983when only 12 percent hatched King et al,, 1982 and 1983; King, Shaver and Phillips, 1984; Leong, 1984; King et al., 1985; Shaver etal1986 and1987!. Several hypotheses wereadvanced regarding thecause ofhigh mortality ineggs from the 1983 year-class including: I! microbialinvasion; ! movement of the eggs during a criticalphase of incuba tion; ! non-fertilization ofthe eggs; ! excessivesand moisture; and ! eggcontamination bychemical toxicants Leong, 1984!. It wasdecided that examination ofunhatched eggs might provide some i~sight into the critical periods of embryonicdevelopment andpossible causes ofembryo mortality. Bystaging embryos obtained from these eggs, we couldestimate time of death, and relate it tothe sequence ofactivities involved in theRestoration and Enhancement Program.Results could be used to evaluateprocedures employed in the program, to aid in improvingthese proceduresandto avoid excessively highmortalities such as occurred in1983. Moreover, investigating embryological developmentanddeformities inembryos from unhatched Kemp's ridley eggs could produce information enhancing our understandingof this criticallyendangered species. Theembryological stagesof development ofthe olive ridley L. olivacea!, loggerhead Caret tacaret ta!, green Chelonia mydas!,flatback Chelonia depressa!, hawksbill Eretrrlochelys imbricata! and leatherback Dermochelys coriacea! sea turtleswere examined anddescribed byCratz 982! and Miller 985! who sacrificed embryos systematically throughouttheincubation period, Itwas hoped that developmental stagesdescribed forolive ridley could be used to stageKemp's ridley embryos. However, Kemp's ridley eggs and embryos could not be sacrificedbecause its populationiscritically low. Unhatched eggsprovided thematerial that could otherwise notbe obtained bykilling eggsor embryos. Theobjectives ofthis study were: ! todetermine if embryos could be salvaged and valid information obtained fromunhatched Kemp's ridley eggs that had been retained through full-term incubation; ! toquantify thenumber offertile and infertile eggs; ! todetermine ifembryological stagesof development forolive ridley Cratz,1982! could beapplied tostage Kemp's ridley embryos, andif not,to develop any necessary modifications oradditional stages; ! toquantify embryonic deaths interms of stage ofdevelopment forunhatched eggs salvaged from clutches from eachyear-class; ! toquantify anddescribe embryo deformities andcompare themto deformities ofhatchlings by clutch;and ! todetermine whether mortality and deformity rates were related tothe methods employed foreach year-class.

' ¹tional ParkService and Texas A&1 University

82 chedeggs of Kemp'sridley seaturtle, by year-class.

ed Unha tched eggs Examined' Not Examined No. No.

84.1 450 15.1 24 0, 77,6 105' 5.2 349 17. 12.1 1,629 81.2 135 6, 90.7 184 9.3 0 1 0. 83.6 340~ 16.4 0. 88.3 235 11.7 0 0 0. 64.4 713 35,6 0.

72.3 3,656 24.3 509 3. ty andembryonic stage. fromonly seven of 20clutches received from Rancho Nuevo 1 nest laid on Padre Island National Seashore. al nest laid on Padre Island National Seashore. 1 nest laid on Padre Island National Seashore.

Materials and Methods Outof 15,062eggs from the 1980 and 1982-1987 year-classes, 4,165 7.7 percent!did not hatch Table1!. Of these unhatchedeggs, 3,656 were preserved and examined, They were injected with andstored in F.A.A.solution, except for 600eggs from 1983that werefixed in 10percent buffered formalin. The eggs were fixed and submergedin preservativewithin two daysof hatching o fa givenclu tch, except for 650from the 1980 and 1983 year-class that first had beenfrozen for one to two yearsbefore injection. Unhatchedeggs that were examined were subjected to thesame procedures. Each egg was opened and examined forsigns of yolk absorptionand presence of embryonicmembranes. All embryoswere removed, and the developmen- tal stageof thosenot greatly deteriorated was compared to stagesdescribed for theolive ridley by Cratz982!. If the Kemp'sridley embryostages were comparable to thoseof olive ridley,they were so classified. If not,additional erectedor modified stageswere establishedand assigned.Embryos also were inspectedfor obvious external deformities,then were storedin 45 percentisopropyl alcohol. On thebasis of theseobservations, eggs and embryos were classifiedaccording to the following categories Tables 2 and3!: 1, Eggsin which no embryoswere found a. Infertile eggs showedno signs of yolkabsorption or changein yolk texture,and exhibited no embryonic membranes b. Fertile eggs containedembryonic membranes, showed signs of yolk absorptionand cha~ges in yolk texture,but ceasedembryonic development early in incubationor eggswhose embryos were so deterio- rated that they could not be found c. Rotten eggs that couldnot be classifiedas either fertile or infertilebecause their yolkshad either deteriorated or were consumed or disturbed by maggots 2. Eggs in which embryos were found a. Stagedembryos embryos assigned to one of thedevelopmental stages described by Cratz 982! forolive ridley or described in this study for Kemp'sridley b, Decomposedembryos embryos thatcould not be stagedbecause they had deteriorated

Results and Discussion

General Conditionof someof theeggs and embryos prior to preservationcould have influenced the results. For example, eggsfrom clutches of the 1980year-class were frozen for two yearsand inadvertently thawed twice before preser- vation,hence they generally were in poorcondition when examined. Also, it wassometimes difficult to distinguish

83 2. Categorizationof unhatchedKemp's ridley seaturtle eggsexamined, by year-class.

Fertile Infertile Rotten Total Year- Without class embryos No. 9o No. No, No.

173 38.4 37 8.2 226 50.2 14 3.1 450 47 44.8 25 23.8 28 26.7 5 4.8 105 230 14.1 1,102 67,6 252 15.5 45 2.8 1+29 75 40,8 59 32.1 39 21.2 11 6.0 184 180 52.9 77 22.6 62 '18.2 21 6,2 340 113 48.1 51 21.7 67 28,5 4 1.7 235 591 83.5 67 9,4 54 7.6 I 0.1 713

'1,409 38.5 1,418 38,8 728 19,9 101 2.8 3,656

nditionof embryosin unhatchedKemp's ridley sea turtle eggs examined,by year-class.

Staged Decomposed' Total No. 'Fo No. No.

115 66.5 58 33.5 173 44 93,6 3 6,4 47 213 92,6 17 74 230 72 96.0 3 4.0 75 178s 97 8 4 2.2 182s 116' 98.3 2 1.8 118 595' 100,0 0 0.0 59'

1333' 93.9 87 6.1 1,420' al stagecould not bedetermined, s of twin embryos, s of twin embryos. s oftwin embryos. tsof twin embryos,

infertilityfrom early embryonic death, and early embryos were difficult to find and stage inmany instances because theywere smaller and had been exposed tolonger periods of deterioration prior to preservation. Ofthe 4,165 unhatched eggs available from year-classes 1980and 1982-1987, 3,656 87.8 percent! were examined TableI!. Ofthose examined, 101.8 percent!could not be classified asinfertile orfertile because they were rotten Table2!. The annual percentage ofunclassifiable eggswas relatively low and constant forall seven year-classes, rangingfrom 0.1 percent in'1987 to6.2 percent in1985. The annual percentage ofin''crtile eggs varied widely from 7.6 percentin1987 to 50.2 percent in 1980.However, it is likely that the numbers ofunha tched eggs tha t wereclassified asinfertile among those examined from the 1980 and 1983 year-classes were exaggerated bytheir poor condition, Fertilityof marine turtle eggs typically exceeds 80percent and, on an average, may exceed 95percent Miller, 1985!, Consideringalleggs from all seven study years, including those that hatched, only five percent were found to be infertile,However, one clutch had a 100percent hatch and 24 percent of all 1982-1987clutches examined were found to be 100percent fertile. Theannual percentage ofunhatched fertile eggs that contained noidentifiable embryos ranged from 8.2 percent in1980 to67,6 percent in1983 Table 2!. Of the 1,418 unhatched eggs classihed asfertile but containing noembryos, 1,102were from the 1983 year-class. Overall, 1,409 or38.5 percent ofall eggs examined contained embryos. Two eggs fromthe 1985 year-class, five eggs from the 1986 year-class and four eggs from the 1987 year-class contained twin embryos,sothere were actually 1,420 embryos inthe 1,409 eggs containing embryos. Overall, 1/33 93,9percent! of

84 ~ O 0 O O 48 Ng 58 8 58

Ih Ih 0

g N

0~ V 0 V V N IN N V

N Ih V N N lh ~ N 0 0 0 N VO

0 O OO N 0 VO Ih 0 V OC} 0 v0

CCI0 0 N 0 WO DC} N v 0 VO o 0

V 0 0 N 0 I V

O N INN N 0 N N CI C. 0 OC. 0~ 0 Ch O N 0 IC} 0 0 I V N VO

I 0 VO

us Za}C us z~ z~ z~

v5 V W allembryos were staged, and the other 87 ,1 percent!were so decomposed that staging was impossible Table 3!. Of thelatter, 58 6.7 percent!were 1980 year-class embryos, EmbryologicalStaging Stagingof Kemp'sridley embryos was done according to Cratz 982!, except that four additional stages were erected.Stages 27-28, 28-29, 29-30 and 30-31 were established toclassify Kemp's rid Icy embryos that possessed yolk sacand caiapace measurements intermediate between those given for olive ridley embryo stages 27 through 31 by Cratz982!. Each of the 1333 staged embryos Table 3! was assumed tohave ceased development atthe approximate dayof development corresponding tothe stage into which it wasclassifled, The earliest dead embryo was placed in stage3 correspondingtoapproximately 3 days ofdevelopment andthe latest in stage 30-31, approximately 50-55 days of development Table 4!. Thedata for all fertile eggs Table 2!, including those without identifiable embryos and those with staged embryos, werecombined to analyzeoverall egg and embryo mortality Figure 1!. Those eggs that were fertile but lacked identifiableembryos were included inthe first third of incubation deaths. There was high mortality during early stagesfor all year-classes,especially for 1983.There were few deaths during middle stages, but deathsincreased duringlater stages. Most embryos of the1982-1984 and 1986 year-classes that died did soduring the first third of development,while those of the 1980, 1985 and 1987 yearwlasses succumbed during the last third of development, Of 1,418fertile eggs without embryos and 1,333 embryos that died and could be staged, 57,7 percent ceased developmentduring the first third, 4.2 percent during the middle third and 38.1 percent during the last third of incubation.However, it is emphasized thatthe 1983 yearwlass, most of whose embryos died during early stages of mcubation,and the 1987 year-class, most of whosedead embryos succumbed during late stages of incubation, dominated the results. Ivloreembryo deaths occurred during later stages ofdevelopment in1985 and 1987 than during other years Figure I!. However,87 of 153embryos that ceased development during the last third of incubationin 1985were from two clutches,and 271 of 574 einbryos that ceased development during the last third of incubation in 1987were from three clutches.These were the only clutches inwhich incubation temperatures exceeded 38'C for prolonged periods oftime. Onlytwo of the358 late stage embryos from these five clutches were obviously deformed, It appears that high incubationtemperature may have contributed to theselate stage mortalities Shaver et al., 1987!. However, the relativelyhigh late stage embryonic mortality in almostall of the other 1987 year-class clutches was not attributable toexcessive incubation temperatures since the other clutches remained below 38'C. All 1987clutches may havebeen adverselyaffected bya coldfront or abruptly changing temperatures atRancho Nuevo and possibly a bacterial infectiontransmitted by thenesting females Shaver et al., 1987!. Questionshavebeen raised byLimpus, Baker and Miller 979! as to the risks of air transporting ormoving sea turtle eggsthat are undergoing incubation. Airtransport ofKemp's ridley eggs inour study occurred from incubation day 1 through46for various clutches during the seven-year study,However, after we examined laying dates, transport datesand stages ofdevelopment, airtransport wasruled out as a factorcontributing toegg mortality. Airtransport hadbeen suggested asa possiblecausi tive agent in the high mortality ofthe 1983 year-class clutches. However, 90 percentofthe deaths inthese clutches occurred before transport. Thus, both air transport andnon-viability ofeggs shouldbeeliminated aspossible factors leading tothe exceedingly poorhatching success inthe 1983 yearwIass, Other causes,such as excessive sand moisture, contaminated sand, or contaminated polystyrene foam boxes, must have actedon these early embryos. High incidences offungi and bacteria were found in internal tissues ofeggs from the 1983year-class andit wasconcluded thatFusarium solani may have been responsible forthe high inortality Leong, '1984!,

Deformities Of the1333 embryos that could be staged Table 3!, 110 8.3 percent! obviously were deformed Table 5!. Approximatelyhalfof all clutches examined from the combined year-classes contained deformed embryos. The earlieststage inwhich deformed embryos were noted was stage 10, but most deformed embryos occuned instage 27orlater stages, Jaw deformities occurred most frequently, followed bydeformities involving the eyes, pigmenta- tion,carapace, flippers, twinning, head, skin, snout and other areas Table 6!. Certain deformities appeared more oftenincertain year-classes, suchaslonger lower jaw in the 1980 year-class, skinclosure failure inthe1982 year-class, headmalformations inthe 1984 year&ass, flipper abnormalities inthe 1985 year-class andtwinning inthe 1986 and 1987year-classes.Theeleven sets of twins were obtained from unhatched eggsof the 1985,1986 andI987year-classes, Twinembryos were most frequently classified in different embryological stagesof development andusually possessednoother external deformities. Theearliest twin embryo was stage 11 and the latest stage 29-30. Only39 .5 percent! ofthe 8395 hatchlings ofthe1982-1987 year-classes weredeformed Table 7!.Approximately 18percent ofthe clutches from these year-classes contained oneor more deformed hatchlings, Carapace deformities 100

Cl First Third of

Middle Third o 80 8 Lect Third of

90 Zus Q X Lli ao

20

I I 1980 1982 1993 1994 199S 1999 1987

YEAR Figure1. Frequencydistribution of incubationtrimester in rohicbKemp's ndtey sea turtie embryosceased deodoprnent, by year-ctass.

occurredmost frequentlyin thesehatchlings followed by deformitiesof the jaws,eyes, flippers, neck and plastron. Almost all clutches with deformed hatchlings also contained deformed embryos, but the converse was not true. This could indicate that certain adults possesseda genetic trait that influenced embryonic development in their clutches.Also, because there were relatively more deformed dead embryos than deformedhatchlings, it appearsthat mostof theaberrant Kemp's ridley embryosdiebefore hatching. Varia tiftns in scutepatterns were not includedamong hatchlingand embryodeformities tabulated herein, although such variations frequently occurred, The rather low de- formity rate of 1.2 percent for 11,485hatched and unhatched eggs from the 1982-1987 year-classes that were examined and not rotten would be higher if scutedeformities had beenincluded. In any case,it is higher than the 1 percent deformity rate reportedby McGehee979!, Slanckand Sawyer981! and Miller 985! for other marineturtles.

Con dusions All proceduresused in the internationalprogram to restoreand enhancethe Kemp'sridley seaturtle population must be monitored and evaluatedconstantly because of this species'critically endangeredstatus. One meansof evaluating procedures used in handling and incubating eggsis examination of unhatched eggs for signs of abnormal embryonicdevelopment and mortality. Unhatchedeggs from eachclu tch can be preservedfollowing the incubation period and can yield salvageableembryos. Eggs should not be fmzen,F.A.A. is the preferredpreservative, Embry- ologicalstages of developmentof theolive ridley canbe used to stageKemp's ridley embryos,provided modifications aremade to accountfor slight differencesin morphologyand developmentalrate between the species.Egg handling and incubationprocedures can be evaluatedby comparingtheir phasesby dateof occurrencewith the time phasing of laying dates,transloca tion dates,embryological stages, embryonic death and ha tching da tes. Translocation of eggs by air apparentlyhas not influencedembryonic development and its useshould be continuedas long assimilar pro- ceduresand comparablecare are maintained as in thepast. Incubation temperatures should notbe allowedto exceed 38'C or changeabruptly, Contaminationof sandand incubationboxes, and excessivesand moisture should also be avoided.

87 Frequencyof deformedembryos in unhatchedKemp's ridley seaturtle eggs, by year~ss,

Non-deformed Deformed Total No. No. No.

99 86.1 16 13.9 115 32 72.7 12 27.3 44 207 97.2 6 28 213 61 84,7 11 15,3 72 I55 87.1 23' 12.9 178' 90 77.6 26 22.4 116' 579 97.3 16' 2,7 595

1223 91.7 'I'10' 8.3 1333' s 2sets of twin embryos es5 setsof twin etnbryos s 4sets of twin embryos es 11 setsof twin embryos.

Table 6. Description,frequency and percent %! of deformitiesin 1333stagedKemp's ridley seaturtle em- bryosfrom the 1980and 1982-1987 year-classes combined.

Location Description Frequency No.' Jaws Bothlacking, lower lacking, lower much longer or shorter than upper, split, crossed,both too long or too short 31 Eyes Trueor falsecyclopia, both lacking, both joined, both too largeor toosmall, lacking pupils 24 1.8 CarapaceSaddleback, indented on one or both sides, curved upward, contorted, toosmall, too narrow,posterior portion missing 24 1.8

Pigment Lackingon part or all of body, lighter than normal on part of body 24 1.8 Flippers All lacking,all too short, both front or both rear lacking, both front too narrow or toowide at tips,both rear tooshort, notch in right or left front, rightfront too short, both front with 6 or7 claws,right rear margin indented, right or left rear small and clawless 1.7

Twins Twin embryos 22 1,7 Head Lacking,two-headed, too wide, too large, encephalocoele, pointed or lumped on top 21 1,6

Skin Patchlacking on back of neckor head 0.6 Snout Too short, too long 0.6

Other Plastronincomplete, lump underthroat 0,2 'Outof a totalof 110 deformed embryos. Some of the embryos had more than one type of deformity. s Outof 1,333staged embryos.

88 tion, frequencyand percent %! of deformities 7 year-classescombined.

Description

es curved upward, contorted, very small and strictedor indentedlaterally rossed, too long e or both lacking nt with 6 clawsand very wide at tips h a ventral and dorsal lump ter lacking bone and scutes ings from a naturalnest on Padre'Island Rational Se 9 deformed hatchlings. Some of the hatchtings had

The data gatheredin this study canbe comparedto similar data from future year-classes,both under conditions of artificial incubation and incubation in natural nests, Systematic sacriBcing and staging of embryos throughout incubation might be useful in providing a picture of normal development for comparison, However, special permit provisions will be required for such studies, Nevertheless, such studies could lead to improvements in procedures used to enhanceand restorethe Kemp'sridley population.

Literature Cited Anonymous. 1978.Action plan for restoration and enhancement of Atlantic ridley turtle populations at Playa de Rancho Nuevo, Mexicoand PadreIsland National Seashore, Texas. Mimeograph, Gainesville, Florida, 30 p. Blanck,C.E. and R.H.Sawyer. 1981. Hatchery practices in relationto earlyembryology of the loggerheadsea turtle, Caretta caretta Linne!.Journal of ExperimentalMarine Biology and Ecology 49:163-177. Cratz,F. 1982.Embryological stages of the marineturtle Lepidochetysolivacea Escoholtz!, Reoiero Biologica Tropicuna 30!:113-120. King, R., A. Anderson,D. Shaverand P. Plotkin. 1982.Padre Island National Seashore1982 report: Kemp's ridiey seaturtle restorationand enhancementproject incubation and imprinting phase.NatIonal Park Service, Department of the Interior, 160 p. King, R., A. Anderson,D. Shaverand P. Plotkin. 1983.Padre Island National Seashore1983 report: Kemp's ridley seaturtle restorationand enhancementproject incubation and imprinting phase.National ParkService, Department of the Interior, 32 p. King, R., D. Shaverand K. Phillips. 1984.Padre Island NationalSeashore 1984 report: Kemp's ridley seaturtle restorationand enhancementproject incubation and imprinting phase.National ParkService, Department of the Interior, 120p, King, R.,P. Plotkin, Y. Weickum, D, CIarkandD.Shaver.1985. Padre Island NationalSeashore1985 report: Kemp'srid1ey sea turtle restorationand enhancementproject incubation and imprinting phase.National ParkService, Department of the Interior, 168 p. Leong,J.K 1984.Final report on preBminary investigationof microbialinvasion as possiblecause of deathof Kemp'sridley sea turtle eggsin summer,1983. Unpublished report submitted to National ParkService, Department of the Interior, 18 p. Limpus, C.J.,V. Bakerand J.D.Miller. 1979,Movement induced mortality of loggerheadeggs, Herpetologica 35:335-338. McGehee,M A. 1979,Factors affecting the hatchingsuccess of loggerheadsea turt! e eggs Caretta caret ta caret ta!. Master of Science Thesis, University of Central F1orida,Or!ando. Miller,J D. 1985.Embryology of marineturtles,p.270 238.In: Gans,C., F. Bigctand P F.Maderson eds !, Biology of the Reptilia, vol. 14. Academic Press, New York. Pritchard,P.C.H. 1969. The survival statusof ridley seaturtles in Americanwaters. Biological Conseroation 2!:13-17. Shaver,D., P. Plotkin, A, Neville and K. Brennan.1986. Padre Island National Seashore1986 report: Kemp's ridley seaturtle restorationand enhancementproject incubation and imprinting phase.National ParkService, Department of the Interior, 135 p. Shaver,D., E. Cheescman,K. Brennan,N. Karrakerand J. Bjork. 1987. Padre island NatIonalSeashore 1987 report: Kemp's ridley seaturtle restorationand enhancementproject incubation and imprinting phase,National Park Service, Department of the Interior, 142 p.

89

imprintingturtles in hopesof establishingnew reproductive populations. For example, the Kemp'sRidley RecoveryProgram is possiblythe most extensive sea turtle conservation program ever attempted, and seeks amongother goals to establisha ridleybreeding rookery on North Padre Island, near Corpus Christi, Tex. by transferringeggs from RanchoNuevo, Mexico, hatching them at PadreIsland where the hatchlingsare im- printed,then rearing them in Galveston,Tex. for a yearor lessbefore releasing them into theGulf of Mexico Klimaand McVey, 1982!, This attempt to establisha newridley breeding population is basedon theas yet undemonstratedchemical imprinting hypothesis.Indeed, the authors have argued that the Kemp'sRidley Sea Turtle Head Start ResearchProject Caillouet,1984! constitutes the best designedexperiment of turtle im- printingever attempted Owens ef al,, 1982!. However, even if theproject is successfulin establishinga new ridleybreeding population, the possibility of imprintingto cuesother than chemical cues could not be ruled out. Chemicalimprinting canbe delineatedinto severalcomponents, some of which suggesttests of imprinting Hasler, '1966; Hasler and Scholz, 1983!: 1. Turtle hatchlingsmust imprint to chemicalcues characteristic of their natal environment. 2. At least within the context of migration, but not necessarily exclusive of migration, chemical imprinting cues mustbe relevantonly to turtlesimprinted to themas hatchlings. General turtle a ttractantsor repellentswould be ineffective chemical imprinting cues. 3, Imprinting requiresa long term chemicalmemory. 4. Relevantchemical signals must be sufficiently stableover the turtles' life spans. 5, Adult turtles must be able to detectchemicals relevant to imprinting and use them to facilitatehoming. This final componentcan only be testedby experimentssuch as the Kemp'sRidley Head start ResearchProject. However, it is suggestedthat the secondcomponent item 2 above!can be testedin the laboratoryusing juvenile turtles, The chemicalimprinting experienceof hatchlingsdoes not necessarilyinfluence migratory behavior only, Thus, it was postulatedthat imprinting could be testedby examining the effectsof early chemosensoryenvironment on subsequentbehaviorsrelevant to juvenile turtles.Chemical imprinting, which couldbe importantin thecontext of adult migration,also might affectjuvenile behaviors such as chemosensory choice and orientation behaviors, A demonstated responseby juvenile turtles to chemical cues to which they had been exposed as nestlings, hatchlings, or both, would constitute strong indirect evidence supporting the chemical imprinting hypothesis, An experimentwas conductedto determinewhether or not early chemosensoryenvironment af'fects subsequent choice and orientation behavior in Kemp's ridley Grassman et al., 1984!.Turtles were artificially imprinted to Padre Island in the head start project, The responseof these turtles to Padre Island-treated, Galveston-treated and untreated seawater was compared. Padre island-imprinted turtles exhibited a responseto Padre Island seawater in the form of increased average time spent in that seawateras compared to the other two. It was suggested that this responsewas likely a demonstratedpreference for a chemicalcue affected by the turtles'earlier exposure; i e.,chemical imprinting. Although chemicalimprinting is the most parsimoniousexplanation of this result, the possibility of a general attraction to PadreIsland seawatercould not be eliminated.The possiblity of a generalizedresponse to a chemicalcue can be addressedby comparing the responsesof turtles that were exposed to the chemical cue as neonatesto turtles that were not exposed to the cue as neonates,Thus, two experiments, one involving Kemp's ridley and another using greensea turtles Grassmanand Owens,1987!, were conductedto testjuvenile turtle imprinting.

Materials and Methods An automaticdata acquisition system consisting of a circulartank Figure1! with 16peripheral compartments was used to monitor turtle movements. Synthetic seawater salinity 25 ppt and temperature 26-29 C! flowed from a moat, through siphons and into the back outer end! of each compartment. Each compartment emptied into a central pen area, and the tank drained through a centrally located standpipe to a biological filtering system and reservoir. The centripitallylocated open end of eachcompartment was equipped with a photoelectricgate which waselectronically interfaced to a clock. The basic system was designed to monitor the behavior of fish that are inherently more active than seaturtles. Thus, in the presentcontext, in order to ensuresufficiemt turtle activity in the compartmentsto facilitate statisflcal comparisons, the tank was partitioned into four quarters instead of 16 compartments. This converted the tank into a four-choice, pie-shaped tank and effectively increased the quantity of activity per compartment, Suspected chemical cues used to artificially imprint turtles were pumped into the back of the four compartmentswith a constant-rateinfusion pump, And each test animal's movementsinto and out of each compartment,and henceresponses to chemicalcues, were automatically recorded.

91 standpipe f'1

Figure2. Circulartank used in testingresponses ofsea turtles to chemicaLs. Repeatedtests of an animal's responses tonatural seawater containing potential imprinting cues were conducted byplacing it in the center ofthe tank and allowing it to swim freely for four hours. Seawater solutions were prepared accordingto thefollowing pro!ocol: one kilogram of sandwas washed with oneliter of seawatercollected from the samelocale. Sufficient seawater volume was then added to bring the total volume to four liters. Equal volumes of syntheticseawater were added at concentrations appropriate toequilibrate salinity of the soluhons toeach other and theholding tank Because equal volumes of syntheticseawater were added to eachsample, the concentrations of possiblechemical cues except for ions! in the two samples did not change relative to each other. Synthetic seawater wassiphoned mto each compartment at a rateof 0.72liter perminute, Responsesof24, 4-month~Id ridleys, which had previously been artificially imprinted to either Rancho Nuevo or PadreIsland by exposure tosolutions prepared from sand and seawater from their respective imprinting sites, were compared,Padre Island PI! ridleys were artificially imprinted in thehead start project. The eggs were collected at RanchoNuevo during oviposition andplaced inpolystyrene foam boxes containing Padre Island sand. The eggs were transportedtoPadre Island National Seashore where they were hatched. The hatchlings were released on the Padre Islandbeach near where the sand had been collected originally, and they werc allowed toe@ter thesurf momentarily. Theywere recaptured in dip netsand shipped to theNational Marine Fishcric s Service,Southeast Fisheries Center's GalvestonLaboratory tobe reared head started! incaptivity for not morc than one year. Ridleys also were hatched andreleased tothe beach and surf at Rancho Nuevo RN!, then recaptured using large dip nets. Both groups were rearedat theGalveston Labora tory under similar conditions, Rancho Nuevo sand and seawater, and Padre Island sandand seawater were collected from the imprinting sites, transported tothe laboratory atTexas A&M University inCollege Station and refrigerated forlater use. Padre Island test solution was metered into one compartment and RanchoNuevo test solution was metered into another ata rateof one liter per hour, Two compartments were left untreatedascontrols. Individual turtles were placed into the center of the monitoring tank and presented simulta- neouslywith choices among the four compartments. Responses dependent variables! measured bytotal entries, time

92 spentand average time spent in eachcompartment by eachturtle were automatica'lly recorded for lateranalysis, Standardparameteric statistical analyses were performed.Single-clas~ification analyses of variance ANOVA! with completelyrandomized design i,e,,mdividual turtles considered as random effects!, factorial arrangement of treatments,and nesting subsampling!were followed by the Tukey-Kramercomparisons of means Hicks, 1973!. Expectedmean squares were examined to determinethe appropriate experimental error residual!mean square for testingeach main effectand interaction.Where no exactF-tests were indicatedby the expectedmean squares, "nt" no test! is shown under columns labeled mean square and F-value in Tables 1-3.The effects of greatest interest were testable.Significance of differencesamong means was accepted at P<0,05.Preliminary investigations sugges'ted the possibilitythat turtles~ight be morelikely to respondto chernosensorycues that weremetered into the edge compartments perhapsthey were more likely to encounterthem!. Thus, a "pattern" effect was included in the ANOVA model. Patterns were as follows: I. Padre Island solution metered into an edge compartment and Rancho Nuevo solution metered into a middle compartment. 2. Rancho Nuevo solution metered into an edge compartment and Padre Island solution metered into the middle. 3, Both solutions metered into edge compartments. 4, Both solutions metered into middle compartments.

Results Therewere no significant differencesattributable to imprinting with respectto any of the dependentvariables examined.Thus, only resultsfor averagetime spent ATS! will be dIscussedin detail becauseit was found to be of interest in the previous ridley experimentdescribed above Grassmanet al., 1984!.Analyses of varianceof ATS responsesof ridley turtles to solutions are presentedin Tables 1-3, There were no significant main effectsor interactionswith respectto either PI or RN turtles.Furthermore, the F-val»efor the Group x Waterinteraction was lessthan 1.00indicating that therewere no differencesco~paring the responsesof PI and RN turtles to the solutions presented in the test situation Table 3!. During the course of this experiment it became apparent that turtle activity was low compared to our other experiments using this paradigm, Twenty-four ridlcys entered four compartments 580 times, This was one-third the activity per individual seenin the previousexperiment involving Kemp'sRidlcy HeadStart Research Project turtles Grassman et al., 1984!.

Discussion There were no consistent responsesto chemical cues in the experimental situation that would indicate chemical imprinting, Becausethe behavior monitoring systemdepends on animal activity,low turtle activity suggeststhat the experimentwas likely flawed. Although accuratehourly temperaturerecords were not kept, holding tank and experimental tank temperatures were maintained between 26and 29 C. The low turtle activity was noted early in the experiment and tank temperatures were not low. In 1983,there was a high mortality approximately 90 percent of the eggsdid not hatch!of embryosat the PadreIsland National Seashore, Poor health of the turtles may haveadversely affected the experimental results. However, PI turtles were almost three times as active in experiments as RN turtles 27 and 153 entries, respectively!, Basedon personal observation, the hatching successappeared to be excellent at theRancho Nuevo rookeryduring oneweek in 1983.Finally, this is theonly experimentof the five we haveconducted to date, using the artificial imprinting-behavior monitoring paradigm, that has failed to detect differences in turtle responsesto treated tank water. Previous experimen'tssuggest that juvenile turtles imprint to chemical cucs to which they were exposedas neonates Grassmanet al., 1984;Grassman and Owens,1987!. First, Kemp's ridleys artificially imprinted to PadreIsland demonstrateda preferencefor Padre Island seawatercompared to other seawatersamples to which they were previously unexposed. Second, green turtles exposed either to morpholine or phcnylcthanol as neonates demon- strated responses to these chemicals that were dependent upon their earlier experience. Furthermore, for green turtles,only thosethat wereexposed in their nests,including an approximately5

93 Analysisof variance of average time spent ATS! responses ofPadre Island-imprinted Kemp's ridley sea water treatments.

Degrees of f variation freedom Mean square 353 0.0358 0.53 tern nt nt ividuals within Pattern! 8 nt nt ter 2 6 0.0100 0,22 tern x Water Interaction 0.0458 1.33 ividuals within Pattern! Water Interaction 16 0.0345 0.51 12 0,0673 47

. Analysisof variance of average time spent ATS! responses ofRancho Nuevo-imprinted K les to water treatments.

Degreesof of variation freedom Mean square 35 0,0333 at tern 3 nt dividuals within Pattern! nt nt Water 86 2 0.0007 attern x Water Interaction 0,0264 0 1 dividuals within Pattern! x Water Interaction 16 0.0173 12 0,0218 47 test. 3.Analysis ofvariance ofthe average time spent ATS! for interactions between two groups Rancho Nuevo- fedand Padre Island-imprinted! of Kemp'sridley sea turtles and water treatments, Degrees of of variation freedom Meansquare F

71 0.0342 0.77 Group 1 nt nt at tern within Group! 6 nt nt ndividual Group x PatternInteraction! 16 nt nt Water 2 0.0051 3.89 Group x WaterInteraction 2 0.0056 0.06 attern within Croup! x Interaction 12 0.0361 1.39 ndividual x Wafer Group x Pattern!Interaction 32 0.0259 0.05 24 0.0446 95 populationsusing artificial imprinting protocols. The relative importance ofimprinhng as compared toother possible horningmechanisms isunknown but it appearspossible that social facilitation could be important in speciesthat nest andpossibly migrate en masse such as is the case with ridleys. Furthermore, chemical imprinting might be expected tobe most important as turtles neared the nesting beach. In sucha case,oceanic currents involved in delivering chemicalcues could be significant,With respectto artificiallyimprinting neonates, little is known concerning appropriatechemical stimuli in terms of quantity, quality and the timing of exposuretostimuli relevan't tothe possible turtle imprinting mechanism. Increasingly,behavioral experiments suggest the importance ofsea turtle chemosensory systems Grassman etal., 1984;Manton, 1979; Owens et al.,1982; Grassman and Owens, 1987!. Because chemical senses may be of primary importanceinturtle natural history, the impact of continued chemical pollution of the marine environment may have seriousconsequences for seaturtle conservation and ultimately species survival, Acknowledgemertts Thisresearch was supported by Texas A&M University Sca Grant College Program project 4A79AA-D-00127, This projectrepresents a joint collaborative effort with the Mexico-U.S. Kemp's ridley recovery team, Their assistance and supportwas critical to conducting the experimentation. Wethank Keith Kreider for assisting with animal collecting. The work was conductedunder the U,S.Fish and Wildlife ServicePermit PRT-1770,Instituto Nacionalde la Pesca permit242,2-00026/1147 and Stateof Texaspermit 409. Literature Cited Caillouet,C.W., Jr, 1984. Essai de prevention de I'extinction de la Tortuede Kemp. Les Carnets deZoologic 44!:28-34. Carr,A.F. 1967.So excellent a fishe.The Natural History Press, Garden City, New York,248p. Grassman,M.A. 1984, The chemosensoty behavior of seaturtles: implications forchemical imprinting, Unpublished Ph.D, Dissertation,Texas ARM University,College Station, Texas, 254 p. Grassmsn,M.A. and D.W. Owens. 1987, Chemoscnsory imprinting injuven ile green sea turtles, Chelonia rnydas. Animal Behaviour 35:929-931. Grassman,M.A.,D W.Owens,J.P. McVey and R. Marquez M. '1984,Olfactory-basod orientation inartificially imprinted sea turtles. Science 224:84-84. Hasler,A.D. 2966. Underwater guideposts - homing of salmon.University of WisconsinPress, Madison, Wisconsm, 155 p. Hasler,A.D. and A.T. Scholz. 1983. Olfactory imprinting in salmon.Springer-Verlag, Berlin, Heidelberg and New York, 153 p. Hicks,C.R. 1973. Fundamental concepts m thedesign of experiments. Holt, Rtnehart and Winston, New York, 349 p. Klima,E.F. and J.P. McVey. 1982. Hesdstarting the Kemp's ridley turtle, LepiIfodietys lzmpi, p. 481-487.In; Bjotndal, K.A. Editor!, Biologyand Conservation ofSea Turtles, Smithsonian Instituion Press, Washington, D.C., 583 p. Koch,A,L., A. Carr and D,W, Ehrenfeld. 1969. The problems ofopen-sea migratIon; the migration of the green turtle to Ascension Island.Journal of 77m.'oreticalBiology 22:163-179. Manton,M.L. 1979. Olfaction and behavior, p.289-301. In: Harless, M. andH. Morlock Editors!, Turtles: Perspectives arufResearch, JohnWiley and Sons, New York,Chichester, Brisbane and Toronto, 695 p. Owens,D.W,, M.A. Grassman and J. R. Hendrickson. 1982. The imprinting hypothrsis and sea turtle reproduction. Herpetologica 38 224-135.

95 Kemp'sRidley SeaTurtle HeadStart Operations of the NMFS SEFCGalveston Laboratory Clark T. Fontaine,Theodore D. Williams, SharonA. Manzellaand Charles W. Caillouet,Jr.»

TheKemp's Ridley SeaTurtle Head Start Research Projectis part of the U.S.-Mexico Kemp'sRidley Recovery Program, and hasasits main objective theestablishment ofa newnesting colony atthe Paddle Island National Seashore nearCorpus Christi, Tex.Asof October 1988, 13/72 tagged Kemp's ridley sea turtle Lepidochelys kernpi!juveniles hadbeen released intothe Gulf ofMexico after being reared incaptivity atthe National Mari ne Fisheries Service NMFS!, Southeast Fisheries Center SEFC!, GalvestonLaboratory in Galveston,Tex. Eachyear during the nesting season, approximately 2WOeggs are collected bythe LI.S. Fish and Wildlife Service andits contractor,Gladys Porter Zoo, from the beach atRancho Nuevo, Tamaulipas, Mexico. The eggs are placed in Padre Island sand withinpolystyrene foamboxes forincubation. Onceall the eggs have been collected, theboxes areflown to the National Seashore wheretheeggs arefurther incubated andhatched, andthe hatchlingsimpr i ntedonthe beach andin the surf at Padre Island, under surveillancebyNational Park Service personnel. Theimprinted hatchlings aretransferred tothe NMFS SEFC Galveston Laboratory wherethey are rearedin seawater raceways, inisolation fromeach other inindividual containers, for9 to11 months. Survivors ingood condition aretagged, weighed and measured,andreleased intothe Gulf of Mexico, usually off Padre and Mustang island.-. Feeding,maintenance andhealth care of the turtles, the major activities ofcapti ve-rearing, aredescribed. Other activities describedinclude: ! monthly wet'ghings ofsamples ofturtles toobtain average bodyweight from which feeding ration is determinedasa percentage ofbody weight; ! tagging bythree methods living tags, internal tags and flipper tags!; ! removal andpreservation ofkidneys andgonads fromturtles that die! for sex determination; and!final weighing andmeasuring of eachturtle beforeit is released. Hatchlingsweigharound 14grams when received i njuly and August, Bylate May or early June ofthe following year,the captiveturtles have increased inaverage weight to0.8 kg, at which size they are released intothe Gulf. Kemp'sridley sea turtle Lepidochelyskempi!islisted asendangered underthe U S.Endangered Species Actof 1973. Itsprimary nesting site is a beachbordering the Gulf of Mexico near the village of Rancho Nuevo, in theState of Tamaulipas,Mexico Chavez, Contreras and Hernandez, 1968!. Hildebrand 963! estimated that 40,000 female Kemp'sridleys nested onthis beach onone day in June 1947, but by 1982 the number had declined toabout 1500 hQrquez,Villanueva andSanchez, 1982; Mlrquez, 1983!. The primary cause ofthe population decline has been overexploitationofthe turtles both directed and incidental! andthe eggs by man. Since1966, thebeach near Rancho Nuevohas been protected during the nesting season from April to July by Mexican Marines and personnel ofthe InshtutoNacional dela Pesca INP! of Mexico, with assistance fromthe U.S. Fish and Wildlife Service FWS! and others.This protection has reduced the poaching byman Mdrquez etal., 1982!, Since1978, the National Marine Fisheries Service NMFS!, Southeast Fisheries Center SEFC!, Galveston Labora- toryhas participated inan international program tosave Kernp's ridley from ex tinction Klima and McVey, 1982!. Theprogram isa jointconservation effortamong the INP, FWS, NMFS, Nationa I Park Service NPS! and Texas Parks andWildlife Department TPWD!, withassistance fromGladys Porter Zoo, Brownsville, Tex.and others, Thegoal ofthe Kemp's ridley recovery program isto increase theKemp's ridley population. Theapproach includes protection ofnesting turtles and their eggs onthe beach atRancho Nuevo, prohibitic nson the capture, possession andsale of theturtles, their eggs and turtle products, promotion ofuse of trawling efficiency devices TED! toallow escapement ofturtles captured incidentally inshrimp trawls, andexperimental headstarting ofKemp's ridley incaptivity during itscriticial first year oflife. The head started turtles arereleased intothe Gulf of Mexico inhopes ofestablishing a new nestingcolony atthe Padre Island Nahonal Seashore, nearCorpus Christi, Tex. For the lat ter purpose, most ofthe head startedKemp's ridley are imprinted ashatchlings atPadre Island, but others have been imprinted atRancho Nuevo in hopesof supplementingthat breeding colony as well, Imprintingisdefined asspecies-specific, rapidlearning during a criticaltimeof early life in which social attachment

"Rational Marine FisheriesService

96 and identifica tion are established.One working hypothesis of the headstar t projectis that imprinting occursduring incubationand hatchingof the eggsin beachsand, during exposureof hatchlingsto thebeachand adjacentsurf, or both. Imprinting is assumedto act throughmemory to guidead»lt t»r'lesback to their natalbeach. At RanchoNuevo eachseason, a smallproportion of the eggsis takenfor headstarting Theseeggs are collectedin plasticbags as they arelaid sothey do not touchthe localbeach sand. They are placedin polystyrenefoam boxes containing beach sand from PadreIsland and are flown to the Pads Island National Seashorewhere they are placedin a hatcheryand allowed to continue incubation. NPS biologists carefully monitor environmental conditions in theboxes during the incubation phase that normally takes 43 to 53 days Robert King, NPS, personal communication,July 1984!. Hatchlingsare takento thebeach and allowed to crav 1 acrossthe sandto the surf to enhancetheir opportunity for imprinfing. Thehatchlings arecollected from thesurf, placedin boxeslined with water-saturated,polyurethane foam cushionsand transportedby NPSpersonnel to Gah eston, Head startingundoubtedly increases survival of theyoung Kemp'sridleys during their first year,and their larger sizeupon releaseis thoughtto give them a subsequentsurvival advantage as corn pared to their naturalcounterparts hQrquez,1972; Klima and McVey,1982; Fontaine, Leong and Cailiouet, 1983; Caillouet, 1984; Fontaine et al,, 1985!, Natural survival of this turtle during its critical first yearof life in the wild maybe lessthan 1 percent.Survival during headstarting the1978-1987 year-classes has ranged from 67 8 to98 4 percent.Of thetotal of16 538hatchlings received alive from year-classes1978 to 1987,13372 82.'Ipercent! had beenreared, tagged and releasedas of October1988. This paperdescribes the facilitiesand methodsused to rear,tag and releasehead startedKemp's ridleys seealso Fontaine et al., 1985; Fontaine et al., 1989!.

Facilities

Quonset Huts The sea turtle head start research facilities at the Calveston Laboratory consist of three, 9 x 20 meter, aluminum- framedquonsets huts manufacturedby X. S.Smith, Inc., Red Bank, N.J. Each is coveredby inflated,double-layered white polyethylenesheathing Figurc 1!. The long axis of eachquonset hut is situatedon an northeast-southwest orientation parellel to the coastline so that prevailing winds provide cross-ventilation to cool the interior during summer. Sidesof the quonset huts are equipped with lateral vent-rails located 1.2 meters above ground level. Panels of polyethylene sheathing attached to the rails can be removed during summer to allow ventilation. The quonset huts also are equippedwith ventilation fansat one end. The spacebetween the dou¹ layered,polyethylene covers is in fia te by small blowers to approximately20 cm, This air-spacebei ween the layersprovides insulation. The tops normally last only 12 to 18 months, so they are replaced annually following the hurricane seasonand before winter. Gas-fired, forced air heaters are used during the winter to maintain warm temperatures. Seawater Supply Seawateris pumped throughsubmerged well-points located approximately 200 meters seaward of the surflineof Galveston'sbeach bordering the Gulf of Mexico.The seawateris storedin' fially in an underground,concrete sump 13,460 liters! and thenpumped into two above-ground,fibergla ss-lined redwood storage tanks each94/501iters!. Seawateris then deliveredfrom the redwoodtanks to eight,above-ground, insulated, fiberglass reservoirs near the head start facilities.Figure 2 showsthe four smallerreservoirs, each ~i ith a capacityof 28390liters, and two of the largerones, each with a capacityof 37+50liters. The other two reservn',rs not shown!also have a 37850 liter capacity. During cold weather,seawater is heatedby thermostatically-contrclle3,unmersion heaters Table 1! to maintain temperatures at 25' to 28' C. Raceways The27 fiberglassraceways used to rear the turtles weremanu factured by RedEwald, Inc.,Karnes City, Tex. The racewaysare rectangular Figure3!, measuring6.1 meters long x 1.8meters wide x 0,6meters deep. They arefilled to a depthof 30 cm, providing a seawatervolume of 3,140liters per raceway. Isolation Rearing Containers HatchlingKemp's rid ley are aggressive and will attack and injure each other Klimaand McVey 1982; Bjorndal and Balazs,1983!, Consequently, they cannotbe rearedtogether without high mortality. Therefore,each turtle is reared in isolationeither in plasticbuckets or boxes.Each raceway usually contains 18 rows of sixplastic buckets, 28 cm deep x 22cm insidetop diameter!,suspended with 10-gaugegalvanized wire from 18 5.1cm x 5,1cm woodenpoles, placed acrossthe width of the raceways Figure4!. The 18rows are letteredA-R from southeastto northwest,and the six columnsare numbered 1-6 from northeast to southwest,providing a totalof 108bucketsper raceway. The bottom of eachbucket is drilled with 12 holes,each 2,5cm in diameter,to allow water exchangeand liberation of turtle excrement anduneaten food, A racewaycan contain 16 rows of fiveplastic boxes not shown; see Caillouet et al., 1988! of thekind

97 Figure1. Kemp's ridley sea turtle head start research facilities quonset huts and sea turtle reservoirsin middleground!, NMFS SEFC Galveston Laboratory.

Figure2. insulate,fiberglass seawater reservoirs supported by voirs15cmarethick, ttotsteel-reinforced, shown!.concrete pads four additional reser- Figure 3.Raneu!ays from viewedtheside!used inrearingKemp's

-I yl rid!ey seaturtles.

Figure4, The buckets used for isolation rearing of Kem p's ridley sea turtlesto preventaggressive attacks.

98 tions,uses and suppliers of equipmentand materials.

Use Supplier ys, basins, Holding turtles,treating Red Ewald, In nd seawa ter wastewater, and storing PO Box 519 seawater Karnes City, TX ers, 2,7HP Aerationof digestiontanks Rotron, Inc, ! Mansfield, OH ,2HP Seawater delivery Pump & Powe 800 Harwin, Dr Houston, TX 77 mp pump, 496 Pumpinguntreated wastewater Granger, Inc. m head 7777 Parnell S 650! Houston, TX 0 liter volume Containers for individual Lorna Plastics, diameter, sea turtles Fort Worth, TX

rs Lora tex, Tops for quonsethuts Farms Supply 500 Clarkesvill Cornelia, GA 30

ion heaters Heahngseawater in reservobs Gla-Quartz El 7074-7190Map Mentor, OH 4005 usedto transportplastic, 3.8-liter milk jugs.Dimensions of theseboxes are 33.0 cm long x 31.1cm wide x 30,5cm deep. Eachturtle remainsin its assignedbucket or box throughout thehead start process, unless it diesorbecornes ill. Theraceway and bucketor box locationsprovide codesused as identifiers for individual turtles throughouthead starting. In this way, the a clutch-of-originidentity of eachturtle is maintained,and can be linked through recordsto the femalethat laid the clutch. Turtlesthat outgrow their containersare transferred to plas- tic laundry basketssuspended in a racewayor to fiberglass basins Figure 5!, Thesehemispherical basins are 61 cm in diameterand 25 cm deepand usually filled with 26,5liters of seawater.The seawateris exchangedand thebasins cleaned on a daily basis, L When ill, a turtle is transferred to sick bay Figure 6! where it is treated, then returned to its bucket or box if it is cured, The sickbay Figure6! where sick turtles are isolatedfor observa- tion and treatment contains fiberglass basins similar to those describedin the previousparagraph. When in use,the basins are drained, scrubbedwith a heavybrush, rinsed, and filled with clean seawater on a daily basis. Seawater Exchanges and C1eaning All racewaysare drained and seawaterreplaced three times a week Monday,Wednesday and Friday!. Oncea week, each raceway is thoroughly cleanedby draining the seawater, washingdown the turtles,buckets and racewaywith freshwa- ter tapwater!,rinsing out the racewaywith freshwater, scrub- bing the inside walls of the racewaywith heavy-dutybrushes Fi~re 5. Herrrisphen'cdfiberglass basins . or power sprayersto removeattached algae, rinsing the race- way onceagain with freshwater,and rehliing theraceway with

99 Figure6, Sick-bay for isolation and trectr»ent! containing hemispherical basins. Figure9. Codesystem used for eighteen 2 3 4 5 6 rows getteredA-R! and su columns num- bered1-6! of buckets.X marksbucket loca- Fig»re7. Raceuy standpipesystem that A 000000 tion83-1-G-4 for 19B3 year-class, raceu'ay preventsoverfilling of raceu>aysand al- 1, row G and column 4!. lowsquick drainage. e 000000 CD000000 00 OOOO E 000000 F 000000 000.00 GI 000000 000000 000000 JK000000 L 000000 M 000000 N 000000 0 000000P 00 0000 00 OOOO 0R FigureB.Fiberglass, wastewater-digs" t >'ontank and air-blower house right foreground!. 000000 Theair-blower iselevated topr~ .c» t back-s;"hcn~r;g ofwastewater into the blower. cleanseawater, The tapwa ter is notheated in thewinter time prior to use,but usuallyremains around 21' C, Itsuse hascaused no apparentill effectsto the turtles. Theraceways are fitted on one end with a 10.2cm insidediameter! standpipe drain Figure7! anda 3.8cm inside diameter!drain pipe on thebottom at thesame end as the standpipe. The raceways are drained by rotatingthe 10.2 cmstandpipe downward and by openingthe valve to the3.8 cm bottom drain. The drained seawaterand water from scrubbingsand rinsingsare collectedin concretetroughs Figure 7! that emptyinto 0,9 meter x 1.8meter fiberglass sumps located outside the quonset huts. Solid-waste sump pumps transfer thiswater into two cylindrical, fiberglass, digestion tanks Figure 8!, each L5 metershigh x6 1 metersin diameterand holdinga volumeof 44+70liters, The waste water from the digestion tanks is drainedintermittently into theCity of Galvestondomestic sewerage system. At theend of eachannual head start period, the residual sludge is washed from the digestiontanks into the seweragesystem. Head Start Operations

Hatchlings Beforethe hatchlings are transported from Padre Island National Seashore to Galvestonthey are weighed wei weight! and measured[carapace length and width, as recommendedby Bjorndaland Balazs983!] by NPS personnel.The hatchlings are packed in plastictubs containing polyurethane foam padding soaked with waterto preventdes sica tion of theha tchlings. They arrive at Galveston af ter two to sixhours in transport.NPS personnel pack theboxes in sucha waythat each clutch is keptsegregated from others. Upon arrival at thehead start facilities, the hatchlingsare first rinsedwith seawaterand then inspected closely for abnormalitiesand mortality. There have been few turtles with abnormalities, but the most common abnormalities that have been observed are: "cross-beak" upper and lower jaws grosslymalformed, left eyemissing!; concave plastron plastrongrossly depressed!;curva ture of thespine; shortened spine turtle much greater in widththan in length!;plas tron improperly healed unclosed yolk-sac attachment site! and deformed flippers. Turtles with abnormalitiesare isolated in thesick bay. Turtleswith improperlyhealed plastrons are treatedwith a topicalantibacterial ointment Terramycin, Gentamycin,Neomycin or Furacin!,and in mostcases the plastron eventually heals. Abnormal turtles that survive the usual headstarting of 9 to11 months have been transferred to TexasA&M University, to TheUniversity of Texas Medical Branchin Galvestonor to The University of TexasInstitute of Marine Science,Port Aransasto be usedfor research.Small numbers of thesehave been transferred by TexasA &MUniversity to SeaTurtle, Inc.,directed by Mrs. Ila Loetscher, South Padre Island, Tex. Clutchesof hatchlingsare usually assignedto racewaysmore or lesssequentially from racewayto racewayas they are received from NPS. However, in some cases,the turtles havebeen distributed according to an experimental design e g.,Caillouct et al., 1989!. Clutch identity of eachturtle in a racewayis kept trackof throughthebucket or box location code Figure9!. For example,the bucketmarked with an X in Figure9 islabeled 83-1-G-4, indicating that the turtle in thisbucket was from the 1983year-class and that thebucket was located in raceway'I, in bucketrow G andin bucket column4. Oncean individual turtle is assignedto a conta incr, it usually staysthere throughout head starting so tha t it canbe linked throughrecords to its clutchwf-originand to the femalethat laid its clutch.In any case,a giventurtle canbe linked through its containercode to ~orded detailsabout oviposition,egg collection, incubation, hatching, growth, amountof food fed,health care, etc, Oneexception is thai four hatchlingscan beheld temporarilywithin four smallercontainers plastic flower pots! placedin a plasticbox. As the turtles grow they are redistributedto larger containers.Another exception is when turtles outgrow their containers.In any case,they are tracedaccording to the codefor thecontainer they spendthe most time in, For taggcd turilesreleased, the tag numbercan be linked with the containeridentification code used during headstarting, Computerized data files carry the tag numberand container identification code to track data for each turtle. Foodsand Feeding The food currently in usein the headstart researchproject is a commerciallyprepared, dry, floating,pelleted diet Table2! manufac tured by Purina.In 1978,the first yearof heads tart operations,foods such as let tuce and fiHet of fish were tried McKey,J.P., J.K. Leong, R.S. Wheeier and R.M. Harris, unpublishedmanuscript on culture of Kemp's ridley sea turtle!, but the cost and inconvenienceof using such foods were prohibitive. A dry, pelleted food manufacturedby CentralSoya and Subsidiariesof Fort Wayne,Ind., was usedbetween 1979 and 1984 Caillouetet ail,,1986b!, After encounteringproblems with the pellets they no longer floated!, we substitutedthe Purina diet Caillouet et al., 1989!.The diet prepared by Purina contains 40 pcrcc nt crude protein, 8 percent crude fat, 5 percent crude fiber, and 47 percent other ingredients Table 2!. This diet provides good growth and survival of the turtles. Feedingof hatchlings and juvenilesmust be carefully monitored as overfeeding can result in compactionof thegut and, in extremecases, death. Because sacrificing live, healthyhatchlings is prohibited,no direct way canbe usedto

101 dients' of the dry, floating, pelleteddiet manufacturedby Purina,Richland, Indiana.

w corn Ascorbic acid Biotin al Choline chloride meal Folic acid ne meal Pyridoxine hydrochloride Thiamine Niacin supplement alfalfa meal Calciumphosphate eserved with BHA, Riboflavin supplement Magnesiumoxide ed yeast Copper sulfate hosphate Manganous oxide a preservahve! Calcium ioda te u pplement Ferrous carbonate nima 1 sterol source of vitamin D-3! Calcium carbonate odium bisulfite source of vitamin K- Cobalt carbonate Zinc sulfate me Zinc oxide pplement Copper oxide supplement available.Proprietary information of Purina, determinewhen the yolk has been absorbed asa guideto when feeding should start. However, this can be estimated indirectlywith samples ofhatchlings from each clutch weighed daily. When hatchlings from a clutchstart losing weight,as indicated by cha nges in average weight for the sample, feeding ofthat clutch is commenced. Thetime lapse to initiationof feedingis approximatelyone to two weeksfrom the hatch date. Therate at which head started turtles are fed is based on the average weight of a randomsample ofturtles selected fromeach raceway atmonthly intervals from those surviving during head starting. An adequate sample size was determinedtobe 25 turtles per raceway Caillouet etal., 1986b!, During the weighings, noattempt ismade todry the turtles.After consecutive weighings of threeturtles, the balance pan is driedand re-zeroed. Both mechanical and electronicbalances have been used, The initial rate of feeding for hatchlings isroughly five percent ofbody weight, Thisrate is gradually changed each month until a rateof roughly 1,5 percent body weight isreached foryearlings. Thedaily food ration is usually divided into two equal portions, one fed in theearly morning and the other in late afternoon.The rations of food are distributed tothe turtles using smail plastic cups that hold the measured volume offeed required todeliver a givenweight offeed. Geometric mean weights are preferred toarithmetic mean weights forsetting feeding rates, because it has been found that the variances ofweights ofhead started Kemp's ridleys are heterogeneousanda logarithmictransformation alleviates this problem Caillouet etaL, 1986b; Caillouet et al., 1989!. Health Care Duringhead starting, each turtle receives a precursory examination forevidence ofdisease during the twice daily feeding,Any turtle displaying signs of disease orinjury is isolated atthat tixne in the sick bay. Others with more serious problemsaresubmitted toJoseph Hanagan, DVIVI, Houston Zoo, for clinical diagnosis andtreatment see also Clary and Leong,1984; Leong et al,, 1989!. Necropsy If a turtledies, its kidneys and gonads are removed and preserved in 10 percent buffered formalin for sex deter- mination Wibbels etal., 1989!. If several die at once or from the same raceway the turtles are put on ice and taken to theTexas Veterinary Medical Diagnostic Laboratory Systems inCollege Station for a thoroughnecropsy. Treatments recommendedfor remainingsick turtles can then be implemented, TagglBg Thehead started turtles are tagged by threedifferent methods Fon taine et al., 1989! before being released into the Gulf of Mexico: 1. A living-tagtechnique developed by Dr,John and Mrs. Lupe He»drickson, University of Arizona,Tucson involvessurgical removal of smallpieces of plash on and carapace, interchanging the gra fts, and securing them withhistological glue, As the turtles grow, the lighter colored plastron transplant makes a vividmark on the otherwisedarker background of thecarapace Fi~~ re 10! . Byplacing the tag on a differentscute each year, the year-classcan be determined for turtles later recap t» red or found Caillnuet et al,, 1986a; Fontaine, Williams andCaillouet, 1988!!. We do not expectanyone but thc informedobserver to recognizesuch a tag. 2, An internal,binary coded, magnetic tag, manufactured by NorthwestResearch Technology Corporation, ShawIsland, Wash., is injectedinto thetip of theright fro»tflipper Figure11!. The tags are 2 mmin length. Theyare sterilized and the area of flipper where the tag is tobe i»jccted isswabbed with tincture of iodine, Once in the turtle, the tag is magnetizedby runni»ga magnetover the flipper.The tag canbe detectedby a magnetometer.The flipper can also be X-rayed to deter»~i»c the exact location of thetag, If theturtle is dead whenfound, the tag can then be surgically removed and the year-class detern~ined. We do notexpect anyone but thoseequipped with magnetometeror X-raydevices to detectand locatesuch tags. However,if the carcassesof tagged turtles that are found dead are made available to us,wc cancheck for presenceof thetags. It isalso possible that a hand-heldmagnetometer maybe developed for futureuse in fieldstudies. In anycase, theseinternal tags are expected to belife-time tags and should remain in placefor futurereference. 3. Flippertags Figure12! are the most frequently used tags, and are easily recognizable as tagsby the public,but we do not considerthem to be lifetime tags,Flipper tagsused for headstarted Kemp's ridley areHasco type, style681, self-piercing, self-clinching, ear tags, manufactured from monel or inconelbyNational Band and Tag Company,Newport, Ky. They are inscribed with a sequen!ial letter-numbercode as well as the message "Send NMFSLab, Virginia Key, Miami, FL 33149." Tagging with this tag is usually done abou t 30days prior to release of the turtlesto allow remedialachon in caseof tagloss or infection,and to allow time for tag-relatedmortality, if any,to beobserved. The tags are normally inserted on thetrailing edge of theright frontflippers. The tags arefirst soaked in gasolinefor 24 hours to remove any oil or grease,then in 90percent ethanol for 24 hours and, finally,they are sterilized by autoclavingprior to tagging.The area of taginsertion on theflipper is swabbed with tinctureof iodineprior to tagging,Neosporin, a broad-spectrumanti-bacterial ointment, is appliedto the tip of thesharp clasping device of thetag before the tag is inserted.A cast-iron,tagging tool NationalBand andTag Company! is used to affixthe tag to theflipper, It is sometimesnecessary to recrimpthe tag with pliers to secureit. Carefulobservations of tagcodes, body weight,carapace length and width straightline!, andgross observationsof generalcondition and healthof eachturtle are madeand recordedas each tag is applied,This usuallyis the last time that measurements and weights of thehead started turtles are recorded before the turtles prereleased. Turtles do not activelyfeed for oneto two daysafter tagging, so feeding is discontinuedfor 48 hours,This procedureprevents fouling the water with uneatenfood.

Release Head started Kemp's ridleys are transportedto releasesites in waxwoated,cardboard boxes. The boxesare modifiedby partitioningthem with plywoodto maketwo horizontallayers within eachbox Figure13!, Two 1.3 cm air-holesare drilled at eachend of the box, and the floor of eachlayer is coveredwith a pieceof 1,3 cm thick polyurethanefoam to cushionthe turtles.The foam is moiste»edwith waterto preventdessication of the turtles duringtransit, Eight yearling turtles are transported in eachbox, four turtles to a layer.Turtles are oriented with their headstoward the cornersof thebox to preventthem from biting eachother, and to placetheir headsnearer the air- holes.The lid of thebox is securedwith grayduct tape,one piece completely around the lengthof thebox and another completely around the width. Copiesof federaland statepermits are attachedto eachbox. Normally, a crewof 10persons requires three hours to packand load the yearlingturtles for shipment,Packing and loading shouldbe doneas quickly aspossible and at night to reducethe amountof time that turtlesare held in theboxes and to prevent their becomingover-heated, The turtles remain in the boxes until released from a vessel. Since 1981,the releasesite for the 9- to 11-month old turtles has been offshore of Padre Island Fontaine et al,, 1989!. Thissite was chosen to reinforceany imprinting the turtlesmight havereceived as eggs and hatchlingsat PadreIsland. Transitby truck and vesselto the releasesite takesabout 10hours, Either U S,Coast Guard vs tters or TheUniversity of Texas' R/V LONGHORN have been used in most of the releases. Figure10. Living tagon a 1984year-class Kemp's ridlr'ysea tu> tie. The tag is onlett costalscute 5.

Figure H, Diagramsho binary-coded,ntagnetic ta of a headstarted Kemp's

I04 FigureT2, Flipper tag Hascotype, style 681, selfpiercing, sdf clinching! used on had startedKemp's ridley sea turtles.

7.

t

''.,--,:-:.;j.;g:='=.'. g=:..

;: tr;""-. ''- .-'",<>"

FigureT3. War-coated box used to transporttagged years'ng Kemp's ridley sea turtle~. Usually,eight turtles two horizonta!layers of four each;right photo,cutaway! are transported in this mannerto therelease site. 40

35

30

lO ~ ~ ~ ~ 25 IL 0 0 ~ 20

x U ~ 15

10

0 NOTE: 1 36 OBS HIDDEN J. 500 1500 2500 3500 AGE DAYS! Figure14. Weight kg! versus age days! for 10 head started Kemp's ridley sea turtles of the 1978 year-class held in captivityat Sea-Arama Marineworld. 0

NORChchOZv If! ccI ~ ~ ~ ~ CD v ~ 'CI I C4 8. Ch a0 QDW W Ch Ch QD QD

$'853 ChLP! ch Ch LPi p W Q C3 g g bD~~8 C> $ R W + v C0 P 9 z 8 '8 W Q v ~ u 6 $ O O QDcV ch B g g

QOCh W Ch QD QDCh QDW v IA M QOChIA Wa0 QO aO W W6 Chv ChQD C5R Ch QD

g C ia o 8o! K~P,~~HN$8~K

8 v CVIA g lq v QD V @ W W W Cr>

f4 0 3!8

vccIh v00$000+CD cg Ia g

Za g Ia ~

QOo

ggms~Ã454~~~9 /II I o~ "c DC!e Results As of October31, 1988, 13372 Kemp's ridley seaturtles had beenhead started, tagged and releasedinto the Gulf of Mexico Table3!. This represents82,1 percent of the 1978-1987year-class hatchlings received alive from the NPS, Asof October31, 1988, 589, or 4.3percent, had been recaptured. Distribution, growth and survival of headstarted turtles in the wild weredescribed by Fontaineet al. 989! and Manzella,Caillouet and Fontaine988!. Growth and movementsof headstarted Kemp's ridleys in thewild alsohave been described by McVeyand Wibbels 984!. Since1978, 264 head started Kemp's ridley have been transferred to oceanariafor extended head starting or captive propagation.As of theend of October1988, 100 survivors remained in captivity.Growth in weightof headstarted Kemp'sridley of the1978 yearwlass at Sea-AramaMarineworld, Galvesion, Tex., is shownin Figure14 seealso McVeyand Wibbels,1984; Caillouet et aL,1986b!. A summaryof headstarhng Kemp's ridleys in captivityis shownin Table3. Survivaland growth of headstarted Kemp'sridleys in captivity havebeen described by Caillouet etal. 986b!. Thecombined survival ratefor the seven year-classeswas 84.9 percent Table 3!. Survival and growth of headstarted Kemp's ridleys in thewild appearto be good Fontaineet al., 1989!, but wecannot gauge the success of maturation,copulation and nesting of suchanimals in the wild as noneof the releasedturtles havebeen reported to havenested. In 1984,egg laying was reported in two5-year-old, head started Kernp's ridleys maintained in captivityat Cayman TurtleFarm Woodand Wood, 1984!. According to Wood and Wood 984!, morethan 60 eggs wcrelaid by thesea turtlesin May1984.Th~ of theeggs in oneclutch hatched, indicating that copulation and fertilization had occurred. Unfortunately,the three hatchlings did notlive. Successful reproduction that which produces viable hatchlings! has sinceoccurred in the captive stock of Kernp'sridleys at the turtle farm in 1986,1987 and 1988,and someof the hatchlings were head started in Galveston,These events indica te thatcaptive propagation of Kemp'sis feasible,and could provide a safetynet for the species Caillouet, 1984!. Publicawareness of thehead start project is sogreat that the turtleshavebecorne celebrities on the Texascoast. The facilitiesprovide a commonsite for field trips for many local schools in theHouston&alveston area. A book Phillips, 1989!recently has been published popularizing the project.Annually, thousands of visitorstour the projectsite. HEART Help Endangered Animals -Ridley Turtles!, a specialcommittee ofthe Piney Woods Wildlife Society, North HarrisCounty College, Houston, Tex., sponsors a oneday open house in Februarythat attracts hundreds of visitors to thefacility. The annual release of turtlesalso is a verymuch publicized event. Local and national news teams cover therelease. Such media coverage helps convey the message of the need for sea turtle conservation to the public, Rec0111rnendations Webelieve there are four areas of researchthat should be emphasized in furturework: 1. Tagsand tagging of sea turtles-Noneof the tags presently being used adequately meet the need. A tagor mark isneeded that will last the entire life of thetagged turtle, and that can be easily recognized and identified by whomeverrecaptures a Kemp's ridley. The PIT passiveintegrated transducer! tag shows promise as a permanenttag, and we are testing it, Further,through publicity and through changes in themessage on the flippertags, the finders of a taggedturtle could be encouraged totake the needed observations andreport them in a timelymanner without removing the tag from live animals. Carcasses of tagged animals that die in the captivestock or arefound dead after release should be saved for examination, asthese specimens provide a meansof evalua ting tag recognition and retention, as well as provide valuable information on possible causes of death. 2. Sexdetermination Asimple techniquemustbedeveloped to determine thesex oflive hatchlingsand juveniles withoutinjuring them, Sex determination is essential for propermanagement and conservation of this endangeredspecies, because sex ratios of hatchlingsare affected by incubationtemperature. The hormonal studiesby Dr.David Owens, Texas ASM University, and his students are encouraging in this regard. Laparoscopyalso has been used successfullyby Dr, Owens and his students on large juvenile and adult turtles, butneither technique issimple nor applicable tohatchlings and small juveniles. However, the non-injurous approachused by Demaset al, manuscript!shows promise. 3. Reproductivephysiology and behavior Additional information is neededon reproduction in the wild and in captivity, 4. Prevention,recognition, diagnosis and treatment of diseases These must be improved for captive stocks of Kemp'sridley. The successful captive-rearing of Kemp's ridleys depends toa largeextent upon prevention andcontrol of diseases encountered inthe husbandry ofthis species. Improvements inseawater systems that providehigh quality seawater forcaptive rearing ofsea turtles go far toward reducing disease problems, Other environmentalinfluences e.g., temperature, sunlight, etc.! also may affect the incidence of diseases.Good nutrition is anotherimportant factor. Ackrtowledgem eats Headstarting of Kemp'sridleys was conducted under INP permits, FWS Endangered and Threatened Species Permits,CITES Convention on International Trade in EndangeredSpecies of WildFauna and Flora! permits and scientificpermits issued by TPWD and Florida Department of Natural Resources. Specialthanks are due Rend Mirquez Millan INP, Manzanillo, Colima, Mexico!, Jack Woody and Richard Byles FWS,Albuquerque, N,M.!, Milford Fletcher NPS,SantaFe, N,M,!, Patrick Burchfield GladysPorter Zoo, Brownsville, Tex.!,Donna Shaver, Jenny Bjork and Robert King NPS, Corpus Christi, Tex.!, their staffs and their predecessors, for theirrespective roles in providingKemp's ridley hatchlings for head starting. Dr,James P,McVey, Mr. Ray S. Wheeler, and Mr. Larry Lansford worked diligently to establish the sea turtle head startresearch facilities and initiate the project from 1978 to 19SI, Bonnii Cockrell, Wane Wibbels, Jean McDonald, DianaNunez, Jody Gruber, Joey Cheeks and Cornel!us Mock contributed much to establishment and improvement of systems,methods and equipment. Dr. Jorge K. Leong was instrumental in developing the facilities for and in recommendingtreatments for diseases ofhead started turtles, as we D asproviding health care for the 1978-1981 year- dasses,along with his staff including John Clary, III andDickie Revcra. The administrative guidance and support providedby Drs. Edward F, Klima and Walter Nelson have been both sthnula ting and encouraging, Theassistance of KathyIndelicato, Marcel Duronslet, Joanne Williams and many others is appreciated. A specialthanks goes to Mrs. Carole H. Allen and the many supporters ofHEART whose diligent and conscientious effortshave enhanced public awareness and support for the conserva tion of Kemp's ridley sea turtles, HEART collects donationsfrom children, school groups and interested people, and each quonse't hut has its own HEART wall that displaysred plastic hearts, each marked with the name and address ofthe person who made a donationforthe turtles, HEARTis alsoinvolved in educatingthe public on theplight of theKemp's ridley. Wethank John Kerivan and the staff at Sea-AramaMarineworld, Galveston, Tex., for caring for 10head started Kemp'sridleys since '1979, and for providing data for our paper. Dickie Revera assisted in checking,verifying and validatingsome of thedata used in ourpaper. Dr. John and Mrs. Lupe Hendrickson, University of Arizona,Tucson, demonstratedtheliving-'tag procedure tous, and assisted in experiments involving living-tags. The figures for this paperwere prepared by DanielPa tlan and the first draft was typed by BeatriceRichardson. Literature Cited Bjorndal,K.A. and G.H. Baiazs Editors!. 1983. Manual of sea turtle research and conservation techniques, Second Edition, Center for Environmental Education, Washington, D.C,, 126 p. Csillouet,C W.,Jr.1984. Essai de prevention de yextinction de ia tortue de Kemp. Les Carnets de Zoologic Bulietin of the Zoological Societyof Quebec!44! 28-34. Caillouet,C.W.,Jr., C.T. Fontaine, S.A, MazizeBa, T.D. Williams and D,B. Revera. '1986a Scutes reserved for livingtags. Marine Turtle NerosletterMo. 36, p. 5-6. Caillouet,C.WJr., D.B. Koi, C.T. Foritaine, T.D. Williams, W.J. Browning arid R.M. Harris. 1986b. Growth and survival of Kemp's ridieysea turtle, Lepidochelys kenipi, in captivity.NOAA Technical Memorsnduin WMFS-SEFC-1 86,iii plus 34 p., 12 Tables and 9 Figures. Caillouet,C W.,Jr.,C T.,Fontaine,T D. Williams,S A. Manzella, AM. Landry,Jr., K. L.Indelicato, M J.Duronslet and D 9,Revera. 1988.Can we saveKemp's ridiey seaturtie? Believe it or not!,p. 20-43.In: Peterson,K.H. Editor!,10th international HerpetologicalSymposium, Zoological Consortium Inc., Thurmont, MD, 206p. Caiiiouet,C.W., S.A. MaiizeBa, C.T. Foutaine, T.D. Williams, M.G. Tyree and D,B, Koi. 1989. Feeding, growth rate and survival of the1984 year-class of Kemp'sridiey sea turtles, Lepidochelys kenipi, reared in captivity Thisvolume!. Chavez,H.M., ContrerasG, and T.P.E.Hernandez D. 1968.On the mast of Tamaulipas.Internatioruzl Turtle and Tortoise Society Journal2 0-29, 37 and 2 !:16-19,27-34. Ciary,J.C., III and].K. Leoiig. '1984. Disease studies aid Kemp's ridley sea turtle head start research. Herpetological Reoieto15!:69- 70. Demas,S.,M. Duronslet,S.Wachtel, C. Cailiouetand D. Nakamara. Manuscript!. Male-specific DNA in reptileswith temperature sex deterininstion. Fontainc, C.,J, Lcong and C W.Cai! Iouet. 1983, Head starting Kemp's ridleys 1982, p. 22-29. In: Owens,DD, Croweli,G. Dienberg, M. Grassman,S. McCain,Y, Morris, N. Schwantesand T. Wibbcis Editors!, WesternGulf of MexicoSea Turtle Workshop Proceedings,Texas A4 M University,Sea Grant College Program, College Station, TX, TAMU-SG-84-'105, 74p. Fontaine,C,T., S.A. Manzella, T.D. Williams, R.M. Harris and W,J. Browning. 1989. Observations on distribution,growth and survivalof head started, tagged and released Kemp's rid ley sea turtles Lepidochelys kenipi! from yearwlasses 1978-1983 This volume!. Fontainc, C,TK.T. Marvin, T.D. Williams,W J.Brow rung, RM Harris, K.L.W.Ind el icato, G.A,. Shattuck and R.A.Sad ler. 1985. Thehusbandry of hatchling to yearling Kemp's ridley sea turtles. NOAA Technical Qernorandurn N!VlFS-SEFC-158, iv plus 34 p., 'l0 Tables,22 Figuresand 2 Appendices. Fontaine,C.T., T.D. Williams and C.W. Caillouet, Jr. 1988, Scutes reserved for livingtags: an update. hfarine Turtle Nerosletter No. 43, p. 8-9, Hildebrand,H.H. 1963. Hallazgo del area de anidacion de la tortugamarina "lora", Lepidochelys kernpi Garman!, en la costa occidental del Golfo de Mexico. Crencia Mexico! 22!:105-112. Klima,E.F. and J.P. McVey. 1982. Head starting the Kemp's ridley turtle, Lepidochelys kempi, p.481487. In: Bjorndal, K.A, Editor!, Biologyand Conservation of Sea Turtles, Proceedr'ngs of the World Conference on Sea Turtle Conservation, Smithsonian Institution Press,Washington, D.C., 583 p, Leong,J.K., D,L, Smith,D.B. Revera,J.C. Clary, III, D.H. Lewis,J.L. Scott and A.R.DiNuzzo. 1989.Health careand diseasesof captive-rearedloggerhead and Kemp'sridley seaturtles This volume!, Manzella,S.A., C.W. Caigouet, Jr. and C.T. Fontaine. 1988. Kemps ridley, Lepidochelys kernpi, sea turtle head start tag ~veries: distribution, habitat,and methodof recovery,Atari ne Fisheries Reviero 50!:33-42. Marquez,R. 1972.Resultados preliminares sobre edad y crecimientode la tortugalora, Lepidochelys kernpi Garman!. Afenr. IV CongresoNacional Ocearrografico Mexico!, p. 419-427. Marquez,R. 1983. Current status of the Kemp's ridley population, p.6-11. In: Owens,D., D. Crowell, G. Dienberg, M. Grassrnan, S.McCain, Y. Morris, X. Schwantesand T. Wibbels Editors!, Western Gulf of IWexico Sea Turtle Workshop Proceedings, Texas AS*MUniversity, Sea Grant College Program, College Station, Texas, TAMU-SG-84-105, 74p. MarquezM., R.,A. VillanuevaO. and M. SanchezP. 1982.The population of theKemp's ridley sea turtle in theGulf of Mexico - Lepidochelyskernpii, p. 159-164,In: Bjorndal, K, A. Editor!,Br'ology and Conservatiorr ofSea Turtles, Proceedings ofthe World ConferenceonSea Turtle Conservation, Smithsonian Institution Press, Washington, D.C., 583 p. McVey,J. P. and T, Wibbels. 1984. The growth and movements of captive-reared Kemp's ridley sea turtles, Lepidochelys kenrpi, followingtheir release in the Gulf of Mexico.KOAA Technical bferrrorandurn NlHFS-SEFC-145, 25p., 3 Figuresand 3 Tables. Phillips,P. 1989. The Great Ridley Rescue. Mountain Press Publishing Company, Missoula, Montana, 180 p. Wibbels,T.R., Y.A. Morris, D.W. Owens, G.A. Dienberg, J.Noeil,J.K. Leong, R.E. King and R. Marquez M. 1989. Predicted sex ratios fromthe international Kernp's ridley sea turtle head start research project This volume !. Wood,J.R. and F.E. Wood. 1984. Captive breeding of theKemp's ridley. IV!arine Turtle Nerostet ter30, p. 12. Questions and Answers SallyMurphy: l am not quite sure I understood whatyou said about hatchling emergence thathatchlings zoere emerging in themorning and that this is different from the natural condition. When do they normally emerge? King;Pat Burchfield mentioned that he was watching normal emergence atthe Rancho Nuevo beach just prior to sunrise, Burchfield:Rend M6rquez and I decided,inlight of some of theexperimental imprinting work that is being done, that duringthe 1985 nesting season we wanted to determine whether there were any relationships between hatchling emergencetimes and beach temperatures, incubation and so on. Unfortunately, wedid not have our data analysis completefor this meeting. About 70 percent of the hatchling ridleys that emerged out of 117nests from junc 13, 1985 to theend of theseason would have essentially been in thewater before sunrise, Hatchling emergence started about 2300hours the prccceding evening and continued through 0900 hours the following morning. But the majority of the turtles would have been in the water before sunrise, PeterPritchard: You mentioned that the hatchlings needed tobe heated up a littlebefore they were active enough to walk down thesand, and the general opinion from Mrosovsky's ct al. observations wasthat sea turtles tend to slow down in activitywhen temperaturerisesabove a certain critical level about 29 C,if I recallcorrectly. Are these compatible witheach other? Inother words,at what temperature didyou find they zoere active? How low zoas the temperature when they were too cold to move? King:During the night, when the hatchlings are in the transfer boxes prior to their transfer from the incubation shed tothe beach for imprinting, the temperature drops to 72'-74' F 2'- 23'C!. I wouldimagine that thc hatchlings were atthat same temperature. At 0800 hours, when the sun is rising, the hatchlings are still at the lowered temperature. Onexposure tothe sun, their temperature begins to rise. I wouldguess that by the time they start moving down to thewater, the temperature, in most cases, is somewhere between 75'F and 85' F 4'- 29'C!. Pritchardzooyouthinkyou could lower the peak temperaturein theegg boxesand perhaps makeit occur laterin theday by using thicker boxes or double boxes? King:The holes that must be put through the boxes to ensure good ventiliahon throughout the sand probably would counteractany benefits of thickerwalls. Thickness of thefoam material does not seemto makea difference.For example,we doubled the thickness of thefoam walls changing from the old boxes used in 1982to thenewer boxes usedin 1983,and this did not seemto changethe temperature curve at all. Thischange was from 1/2 inch.3 cm! to 1 inch .5 cm! thick. Caillouet:It is generallyaccepted that incubation temperature affects the scx ratio in seaturtles. However, it seems to methat two quite different results might derive from this, Either sca turtles that become male or femalehave identicalsex genotypes and their phenotypic sex is somehow induced by incubation temperature orthey have male and femalegenotypes that can be overcomeby temperature.If thc latter weretrue, then it followstha t a certain proportionof theanimals would become sex-reversed in the process of incubationat temperaturesconducive to inductionof thesex that is inconsistentwith thesex genotype. We should be concerned about the possible affects of suchsex reversal, For example, it maybe that sex-reversed animals, even though they are pheno typically females, may not havethe same biotic potential or survivabilityas females that are not sex-reversed. JohnCan: Do you have an idea at whichstage of development thecarapacial scutes pattern is determined or finalized? Shaver:I didnot examine the scutes because oftentimes the embryos were in verybad condition. They were frequently deformed,If I recallcorrectly, it is approximatelyat stages25 and 26 that the scutcs bcgine to appear. Carr: Whatis thefull numberof stagesin development? ShaverCratz described 31. We added four morebecausehis stagesleft out a lotof measurements.Applying his stages to our embryoswould have made it impossibleto characterizesome of our cmbyros. Pritchard:How many days would that be, when scutes first canbe seen, more or less? Shaver. It is about 33 days.

111

RAMSwith severalsignificantadvantages. The first is a short messageformat that is available where "position only" operation is desired. This means that there can be approximately67percent power savingsas comparedto operationwith a RAMStrans- mitter. The second advantage is that there arealways two satellitesin orbit,effectively doubling the probability of receiptby the satellite of a transmission. Also, additional backup satellitesare availablefor deploy- ment in the event that a replacement is needed, This provides assurancethat a newly developedtransmitter can be used for the foreseeable future, A contract was awarded to Wood-Ivey SystemsCorporation WISCO!, Orlando, F'la.to develop an ARGOS transmitter suit- able for a ttachment toand tracking of marine animals. Two prototypes delivered during thethndquarterof1982werefuliycertifiedFcgm1. F~hl oggerheedseaturtle with NIMBUS-6/RAMS transmitter attudred. by ServiceARGOS for satellitebeacon per- formanceand representthe baselinedesign. By compari- son,they are41 percentof the weight in air!, 22 percentof the displacementvolume, and 13percent of the buoyancy of the NIMBUS-6 transmitters, The NIMBUS-type tag weighedabout 3.1kg about3.2 percent of "Diane's"body weight>compared to the ARGOS-typetag weight of1,25 kg about 2.7 percent of the averageadult fernale Kemp's ridley'sbody weight!. The towing resistance drag! of the ARGOS-typetag is lessthan one-halfthat of the NIMBUS- typetag. Extensivelaboratory testing e.g.,shock, pressure, temperatureand vibration! was conductedwith the AR- GOS-typetransmitters. One controlled field test was com- pletedusing a captiveloggerhead turtle. Releaseof a tagged loggerheadinto the Gulf of Mexicowas attempted in 1982, but the transmitter failed for unknown reasons. Rancho Nuevo Study In June 1985, two WISCO-manufactured ARGOS tags were transported to RanchoNuevo for attachmentto Kemp's ridley femalesin coincidencewith the nestingseason. The objectivewas to establishthe development and operational' Figure 2. ARGOStag foreground! andshor-range receiver back- feasibilityof theARGOS tags as a meansof trackingKemp's ground!. ridlcy, The tags also included an auxiliary short-range transmitterto be usedto aid in locationand recoveryof the unit Figure2!. The tag wasattached as shown in Figure 3, Thefirst tag ¹3176; Table 1! was attached on Junc 7 atabout 1530 Central Standard Time CST!to a Kemp'sridley with a 72-cmcarapace length while shewas nesting. Shereturned to the water immediatelyafterward, towing the tagwithout a pparentdifficulty. A satellitelocation was obtained on June 9 at22 77' N, 9769' W, or about40 km south of therelease point and on the18-m depth contour. A 19-minutetransmission on June 23 led to theconclusion that thetransmitter was on thebcachapproximately 40 km southof Tampico,Mexico. Subsequently, a Mexicanfisherman foundthe tag on theshore and returned it. Thesystem was still operationaland showed little or no damageand a minimum of marine growth, A secondARGOS-tagged ¹3177; Table 'I! Kemp'sridley nesterwas released on June10 at RanchoNuevo, and this turtlewas caught by shrimpfishermen off Freeport,Tex., on July 19. The crew removed the tag and released the turtle

113 in apparentlygood condition. The tag was inoperative upon receipt. Testing indicated that the circuitry was functional,andtha t theproblem was related topremature failure of the ba t tery.No satellite positions were received forthis tagged turtle, but it traveled672 km with the tag in 39 days, before the tag was removed. Inthis experiment, observed surfacing time ranged from 40 seconds to320 seconds, Standard data processing techniquesused by Service ARGOS require a minimumspan of 420 second transmissions. Analternate tagdeveloped forNMFS by the Virginia Institute ofMarine Science VMS; Figure 4!uses a transmitter unitproduced byTelonics, Inc.The unit was not available foruse during the 1985 Kemp's ridley nesting season, so it wasevalua ted on a captiveloggerhead turtle a thet VIMS facility, This unit has a somewhat different packaging and antennaconfigura tion, but it meetsthe standard ARGOS specifications. Summaryand Conclusions Thedata obtained from this experiment onKemp's ridley ncsters were severely limited by the brief duration ofthe tests.Nonetheless, messages werc received from the ARGOS transmit ters by the satellite while the turtles were at sea, severalaccura teloca tion fixes were processed Table 1!, and the results can be used to guide further development of satelliteMore developmenttags. work is needed toachieve thegoals of daily position updates and determination ofsurfacing time.Sur facing times and satellite visibility remain asareas ofconcern. Specialized software needs tobe developed Incorporatingposition processing algorithms that trade off accuracy tomaintain tracking continuity ina manner tailoredtothe requirements ofthis type of experiment, Improved transmitter control, through the use of a surface sensingswitch, will also help raise the number ofsuccessful transmissions. Additionally, theradiation pattern ofthe antennaused needs to be verified in themarine environment toassure signal continuity over an extended period of marineTagged growthturtles, andsuch fouling. as the ones used in this study, are vulnerable tothe actions oflocal fishermen. Improved attachmentand betteridentification methods, and increasedpublic awarenessmay alleviatethe unnecessary separation of the tags from the turtles by those who catch sea turtles, but this will probably remain a problem. Rescue of theturtle and tag from threatening situations might be enhanced byincorporation ofsensors that would detect continuedsurfacing and/or removal from the water. Such information can be transmitted through the ARGOS nel loalert the experimenter. Evenfaster response could be achieved byuse of a LocalUser Terminal LUT! for immediate relay of this information, Finally,although the present tagpackage appears tobe small enough, improved and lighter electronics should be incorporatedasthey become available toreduce even further the weight and size of the satellite package, Literature Cited Pritchard,P.C.H. 1969. The survival status of ridleysea turtles tn American waters. Biological Conservation 2!:13-17. I'ritchard,P.C.H. and 14. Marquez M. 1973. Kemp's ridley turtle or Atlantic ridky, Lepidochelys kempi. international Union for the ConservationofNature and Natura/ Resources Monograph No. 2, 30 p.

115 Distributionof Juvenileand Subadult Kemp's Rid1ey Turtles:Preliminary Results from the 1984-1987 Surveys Larry H. Ogren'

Juvenilelifestages ofKemp's ridley turtle Lepidochelys kempU arewidely distributed throughout thecoastal waters ofthe UnitedStates from Texas toNew England. Forthe most part, these individuals could bedescribed aspost-pelagic "yearlings" that haveleft the pelagic habitat forthe nearshore benthic habitat toforage primarily onmotile forms ofcrustaceans suchasport unid crabs.Historical records fromthe turn of the century characterize theridley asa commoninhabitant ofNorth Carolina baysand estuaries.Kemp's ridley was the second mostabundant seaturtlecaughtin theCedar Key, Fla., turtle fisher, but this may reflect fishermanbias for thegreen turtle. Followingthedrastic decline inthe size of the Rancho Nuevo, Mexico, rookery over the past four decades, a similar decrease innumbers ofjuveniles andsubadults inour coastal waters would beexpected. Earlysurveys conducted inthe northern Gulfof Mexicosubstantiated thatthis was the case. However, occurrences ofunusual numbers ofj uvenile Kemp's ridleys captured in relativelyrestricted areas have been recently reported in Louisiana, Alabama and, to some extent, in northwestHorida. In two cases,themost significant biological factor associated withthese frequent captures wasthe abundance ofportunid crabs. Insome cases,cold-stunned turtleswere obtained from both coasts ofFlorida following episodes ofsevere winter temperatures. RecordsofKernp's ridleys collected orobserved, weighed and measured, andtagged and released arepresented. Anecdotal informationfrom various i nformants and miscellaneous observations aresummarized. Theearly life history stages ofKemp's ridley turtle Lepidochelys kempi!have not received asmuch attention by biologistsashave studies ofthe older, reproductively active adults. This was probably due to their small size, reduced slumbersand cryptic habits in theextensive marine environment ofthe Gulf of Mexico and North Atlantic, and was compoundedbytheir being frequently misidentified Brongersma, 1982!.However, adults, aswell as juveniles, were relativelycommon years ago and well known to the trawlermen ofthe Gulf of Mexico Liner, 1954; Carr, 1977!, After thenesting beach atRancho Nuevo, Mexico was discovered byscientists in1961 Hildebrand, 1982!, ridleys became highlyvisible and relatively accessible tobiologists atthe rookery, Tha t wasthe situation until their numbers declined to thelow levels reported today Marquez, Viilanueva and Burchfield, 1989!. Before their numbers decreased to presentlevels, tagging studies conducted byMexican biologists atRancho Nuevo revealed that females returned to crab-richforaging grounds either south ofthe rookery inthe Tabasco-Campeche Bayregion ornorth, primarily off theLouisiana coast, after the nesting season Chavez, 1969; Pritchard andMdrquez, 1973; Mdrquez, 1984!. Itwas also determinedthatthe majority oftag returns came from shrimp fishermen trawling inthose two areas Marquez,1984; National Marine Fisheries Service, 1987!. Nothingisknown about the distribution, oreven the occurrence, ofKemp's ridley hatchlings inthe pelagic stage inthe Gulf of Mexico. A few are sometimes observed swimming inthe surf zone ot'f Padre Island, Tex,, and some have beentossed upon the beaches ofMustang Island, Tex. during storms Anthony Amos, The University ofTexas, and NationalPark Service personnel, Padre Island National Seashore, personal communication!, Ineither case, the littoral zone,with its attendant predators, isclearly not the appropriate habitat ofthese young turtles. Paradoxically, the smallestpost-hatchlings recordedwere two specimens foundin the Atlantic SchmidtandDunn,1917; Deraniyagala, 1939!.One was from the Azores and the other from an unknown locality identified only as "United States." Thejuvenile lifehistory stage, not including thehatchlings orthe post-hatchling pelagicstage Carr, 1986!, isbest describedasthe post-pelagic orcoastal benthic stage that feeds primarily oncrustaceans suchas portunid crabs, bivalves,anda varietyofother invertebrates Dobie,Ogren andFitzpatrick, 1961!.This developmental stageiswidely distributedthroughout theCulf of Mexico and northward along the Atlantic coast from Florida toNew England. Kemp'sridley has also been record ed from Bermuda Mowbray and Caldwell, 1958!, but nowhereelse in the western NorthAtlantic, and that includes the Bahamas and Antilles Carr, 1980!. lnthe eastern Atlantic, Kemp's ridleys include a wide range ofsize classes, fromsmall post-hatchlings tolarger juvenileswith carapace lengths from 10 to 25 cm Carr, 1980!. Inthe western Atlantic only one adult size individual 6 cm!has been captured atCape Canaveral, Fla. Henwood andOgren, 1987!. Other records showed thatKemp's ridleysfrom the western Atlantic ranged insize from 20 to circa 60 cm carapace length Table 1!, Several more adult ¹tionat Marine Fisheries Service

116 Table 1. Sizedistribution of Kemp'sridley turtles from the Atlantic coast. Straight line' carapacelength, centimeters Mean Range nb Locality Date Source 30 27-33 Cape Cod Bay, Mass. Nov. 1978 Lazell 980! 35.6 26.5 - 43 SandyHook Bay/ jun, Nov. Tom Azarovitz New York Bight 1973-1975 unpublished! "41 27-62 21 ChesapeakeBay, Va, May Nov. Lutcavage and 1979-1981 Musick 985!

34,8 20.3 - 57.2 21 S.C. and Ga. coastal jun. - Nov. Hen wood and zones 1978-1983 Ogren 987!

38.6 24.1 - 66 40 Cape Canaveral, Fla. Dcc. - Mar. Henwood and 1978-1984 Ogren 987! ' Exceptas noted by asterisk '! indicatingcurved measurement. Curved carapace length measurements areapproximatcty 2 centimetersgreater than straightline measurements. b Sample size. sizeindividuals havebeen found strandedon Atlantic coastbeaches of SouthCarolina, Georgia and Florida Barbara Schroeder,National Marine Fisheries Service, personal communication!, and Rene Mdrquez lnstitu to Nacionalde laPcsca, personal communication! reported the capture of two adult females offshore the Dry Tortugas and the upper FloridaKeys. With the discovery of therookery at Rancho Nuevo, a postulatedold world origin for these individuals waslaid to rest Brongersma,1972!. However, the fate of theseturtles in theNorth Atlantic remains undetermined. Someobviously die from exposure tolow temperatures if they venture too far north Lazell,1980; Meylan and Sadove, 1986!.Whether all are lost never to enter the breedingpopulation, or whetherthey return to the Gulf of Mexicoto mature, breed or both is not known, Historicalrecords from the turn of thecentury characterize juvenile Kemp's ridleys as common inhabitants of North Carolinabays and sounds Coker, 1906!. Kemp's ridley was the second most abundant sea turtle caught in theCedar Key,Fla., turtle fishery Carrand Caldwell, 1956!. However, this might not reflect the natural conditions concerning thespecies' abundance at thattime because of thefishermen's bias for thegreen turtle Cheloniamydas!, Elsewhere in theUnited States,unusual numbers of smallKemp's ridleys werefound cold-stunnedin VineyardSound, Mass., and werereported to bea commoninhabitant of NewYork Harbor Babcock,1930; Carr, 1980!. Thenesting population of Kemp'sridley at Rancho Nuevo was first observed by biologistsin theearly 1960s when it was determinedthat numbersof nestcrsestimated from a singlearribada in 1947had beenreduced by more than 92percent Ma rqucz, 1984!. A correspondingdecrease in thenumber of hatchlings produced would be expected, and thenumber that survived the pelagic stage and entered the coastal benthic population as juveniles would therefore alsobe lowered. This declinewas the result of systematicand intensiveegg harvestover a period of many years Hildcbrand, 1982!,in addition to other naturalcauses and continuingincidental catch primarily by shrimp trawlers elsewhere in the speciesrange. During the last two decades,increasing beach protection provided by the Mexicangovernment and a very successfulhatchery program operated by Mexicanand American biologists at RanchoNuevo produced an average of 20,000ha tchlings annually the first decadeand 50,000 annually thereafter Mlrqucz, 1984!. This can be expected to haveresulted in a majordemographic change for Kemp'sridley in the Gulf of Mexico,and possiblythe western Atlanticregion as well. However,the adult populationcontinued to declinesteadily during 1978to 1985 Frazer, 1986!,and only about600 females nested at RanchoNuevo in 1986 Richard Byles, U.S. Fish and Wildlife Service, personalcommunication; National Marine Fisheries Service, 1987!. Observations or incidentalcaptures of adult ridleysat seahave become rare events. Conversely, juveniles in thepost-pelagic benthic stages are now commonly found at variouslocalities in coastalwaters from Texasto New England,probably as a resultof theprotection afforded the nestcrs and nests at Rancho Nuevo for two decades. Thefollowing accountpresents capture, tagging and recaptureresults and sizedistribution dataobtained in recent years.These studies are part of the long-termeffort to determinemigratory patterns, seasonal occurrence and distributionand growth of foragingpopulations of juvenilesea turtles in coastalwaters. They provide support for the overview l will give of the early life history of Kemp'sridley.

117 of Mexico.

Date Source

Apr. - Nov. Carr and 1955 Caldwell 956! May Nov. NMFS b 1984 - 1987 Mar. - Jan. Jack Rudloe 1970 - 1985 unpublished! NMFS an. Dcc, Jack Rudloe 985 - 1987 unpublished! NMFS Feb, Oct, Carr 980! @ coastal zone, Ala. 1966-1983 NMFS 23.7 21.6-26.3 Terrebonne and Jun. 1984 NMFS Caillou Bays, La. 32,3 20.3-45.7 61 Sea Rim State Apr. - Nov. Texas Parks & Park, Tex. 1983 1985 Wildlife Dept. STSSN' 31 24.1-39.8 La, coastal zone Mar. Nov. NMFS and eastern Tex, 1978 ' Sample size. s National Marine Fisheries Service, 'Sea Turtle Strandingand SalvageNetwork.

Capture,Tagging and Recap ture At-scacapture ofsea turtles continues tobe the most important and productive source ofinformation from which wea thet National Marine Fisheries Service, Panama City, Fla., Labora tory ascertain the distribution and abundance ofsea turtles incoastal waters. This isespecially truewhen a turtlebiologist ora similarlytrained person isdirectly involvedin collecting orhandling the turtles and in recordingthe species identification andmorphometrics. In addition,refercnccs dealing with significant numbers ofjuveniles turtles and meristic data have been selected from other sourcesto complementour limited data Tables1 and2!. Ingeneral, most of our sea turtle capture efforts inthe southeast region ofthe United States areeither by active fishinggear, such astra wls and strike gill nets, ormore passive methods suchas turtle entanglement nets.The primary objcctivcofcapturing turtles is to tag and release them. In certainareas along the coasts of theGulf of Mexicoand AtlanticOcean, turtles are sometimes captured inpound nets, by hook and line, and fortuitously when they are immobilizedbycold temperatures Ogren and McVea, 1982; Ehrhart, 1983!. Some stranding records also were includedinthe size distribution examples discussed, butmost captures were made by shrimp trawls. JuvenileKemp's ridleys are not the only species collected during our capture efforts. Adult and subadult loggerheads Careffa careffa! dominate thecatch on the Atlantic coast, and juvenile green turtles, although less abundantinour catches than are loggerheads andKemp's rid leys, are found in both areas in theGulf and Atlantic. Otherareas along the Gulf Coast ofFlorida south ofCedar Key are reported tohave significant numbers ofjuvenile greenturtles Paul Raymond, National Marine Fisheries Service, personal communication!. Anexpansion ofour nettingactivities atCedar Kcy is being planned toinclude sampling inthc Homosassa, Fla.,area, New netting methodsarebeing developed forsampling Corrigan and Waccasassa reefs,and sampling areas will be extended to Homosassa,Fla.,where conditions appear tobe more favorable forcapturing subadult green turtles. Mostof the capture and tagging efforts take place along the northwest and east-central coasts ofF1orida, with considerableseasonable summer! effort in Winyah Bay, S.C. The most productive method ofcapture has been with shrimptrawls, but "run-around"gill netsand, more recently, turtle entanglement nets have accounted for an increasingnumber ofcaptures onthe west coast ofFlorida, More than 600 turtles have been captured, tagged and

]18 Table4. Recapturesof Kemp'sridley turtlesin thenorthern Gulf of Mexico. Tag Date Capture Capture Dateof Recapture Recapture Code tagged locality gear recapture locality gear NNJ Nov. 7, Mississippi Sound, Shrimp Sept. 16, 1985 Sabine Pass Jetties, Shrimp 253 1984 Biloxi, Miss. trawl Tex. La. boundary trawl NNZ Jul. 6, Fidlers Point, Gill net Jul. 9, 1985 Fidlers Point, Seine net 674-907 1985 Wakulla County, Fla. Wakulla County, Fla. NNW Nov. 18, 1.6 km offshore Shrunp Apr. 1986 4.8 km offshore Shrimp 701-702 1985 Shell Island, Bay trawl Marsh Island, trawl County, Fla. Iberia Parish, La, NNZ May 31, Mud Cove, Shrimp Jun. 26, 1986 Mud Cove, Shrimp 219-220 1986 Franklin County,Fla. trawl Franklin County, Fla. trawl NNJ Jul. 15, Mississippi Sound Shrimp Aug. 4, 1986 East Deer Island, Shrimp 266 1986 released 40,3 km trawl Mississippi Sound, trawl offshore Horn Miss. Island!, Miss. NNJ Jul. 15, Horn Island, Shrimp Oct. 23, 1986 Dauphin! s land, Shrimp 267 1986 Mississippi Sound, trawl Mississippi Sound, trawl Miss. Ala. NNW Jul. 30, Corrigan Reef, Turtle net May 21, 1987 Corrigan Reef, Turtle net 729-730 1986 Levy County, Fla. Levy County, Fla. NNZ Dcc. 7, St, George Island, Shrimp Apr. 3, 1987 Alligator Point, Shrimp 231-232 1986 Franklin County, Fla. trawl Franklin County, Fla. trawl NNZ Apr. 13, Shell Point, Shrimp Apr. 16, 1987 Shell Point, Shrimp 252-253 1987 Wakulla County, Fla. trawl Wakulla County, F1a. trawl released.Of this total, 376loggerheads, 146 Kemp's ridleys and 22 greenturtles havebeen captured on the Atlantic coast,During 1978 to 1984,approximately 60 juvenile Kemp's ridlcys and 20 green turtles were tagged and released on the Atlantic coast Henwoodand Ogren,1987!, and theseincluded a significantnumber of juvenileand subadult turtles from thecoastal waters of the United Statesoffshore the Indian River estuarinesystem of castwcntralFlorida asreported by Ehrhart983!. Recentrecaptures of Kemp'sridleys on theAtlantic coast are shown in Table3. On the westcoast of Florida,capture efforts have resulted in the taggingof morethan 100 juvenile sea turtles, primarilyKemp's ridleys. The species composition was 110 Kcmp's ridleys, 10 green turtles and 7 loggerheads.The capturemethods included gill andentanglement nets as well as shrimp trawls. Recent and significant recaptures of Kcmp'sridleys in theGulf of Mexicohave been reported Table 4!. One tagged Kemp's ridley was at largeat Cedar Keyfrom July 1986 until May 1987,and another migrated eastward from Biloxi, Miss,, to DauphinIsland, Ala. Two othertagged Kemp's ridleys were recaptured in Louisiana,a considerabledistance westward of theiroriginal capture sites at Panama City and Biloxi, respectively. Figure1 givesthc carapace length-frequency distribution of Kemp'sridleys captured by shrimptra wls during our samplingeffort in the northeastGulf of Mexicofrom l984 to 1986.

Ovexview JuvenileKcmp's ridlcy turtlesare widely distributed throughout U.S. coastal waters from Maineto Texas.The

119 20

20 30 40 50 60 70 80 CARAPACELENGTH CM! Figure1,Straight Iinecarapace lengthfrequency of61Kemp's rutley turtles capturedin shrimptrawlsin thertortheastern GulfofMexico,19B4- 19B6. smallestindividuals, ranging incarapace length from 20 to 25 centimeters, areprobably post-pelagic stageindividuals enteringshallo w coastalzones ofbays, sounds andestuaries Figure I >,In these developmental habi ta ts, their ecologic nichechanges tothat of a benthiccarnivore. The 20 to 25 cm size class isapparently most numerous inthe Gulf of Mexico,giving credence tospeculation thatthe entire life cycle occurs entirely within the Gulf of Mexico for some unknownproportion of thepopulation. Historically,thesmallest post-ha tchling Kemp's ridleys were reported only from the New England states and the easternNorth Atlantic Brongersma, 1972; Carr, 1980!, However, the ridleys that are found in theAtlantic states to thenorth and that have a carapacelength greater than 25 cm are still relatively small individuals whose movements aremos t likely mediated byocean curren ts.Because ofthe greater distance fromtheir natal beach, Kemp's ridleys in NorthAtlantic waters may take a littlelonger tocomplete thepelagic development stagethan their cohorts inthe Gulf ofMexico, and thus they enter the coastal zone at a largersize than those in theGulf. Examinationofpreviously published data on sizes of Kemp's ridleys along the U.S, Atlantic coast Table I! reinforcesearlier comments onthe possibility ofa northtosouth increase insize Carr, 1980!, a phenomenon thathas alsobeen observed forgreen turtles Carr, 1952!. This gives support tothe idea that the smallest Kemp's ridleys transportedoutof the Gulf of Mexico are carried bycurrents northward toNew England and then shoreward across thecontinental shelfvia Gulf Stream gyres Carr, 1980 and 1986!. Carr980! and others Lazell, 1980; Meylan, 1986! haveproposed this,and the data presented herein agree with such an explanation. Someof these Gulf of Mexico expatriatesdonot leave theGulf Stream offNew England, butcontinue theirjourney tonorthern Europe andpoints southinthe eastern North Atlantic Brongersma, 1972and 1982; Pritchard andMJirquez, 1973; Carr, 1980!. Therehas been considerable debateas to whether ornot these Atlantic expa tria tes survive inthe North Atlantic and livelong enough toreturn tothe Gulf of Mexico tobreed, and conversely asto whether ornot they are all doomed waifs,destined either tobecome cold-stunned inthe winter orpermanently isolated inthe Atlantic Carr, 1980!. Historicalrecords andscientific datastrongly suggest thatsome mortality occurs when young Kemp's ridleys are carriedintonorthern latitudes oneither side ofthe Atlantic andare exposed tothe lethal effects ofwinter tempera tures Lazell,1980;Meylan andSadove, 1986!.However, someindividuals apparently manageto survive, eitherby being

120 carried farther downstream to warmer latitudes in the eastern Atlantic, or by some unknown migratory route southward along the coast of the United Statesto Florida, or both, ln the abovecase of Atlantic expatriates,our data support the possibility of a seasonalmigration south from northernlatitudes in the fall, and a subsequentreturn to northernforaging grounds with the warming of the waters in thespring Henwood and Ogren, 1987!. Other workers studying sea turtle popula tions along the Atlantic coast have alsopostulated a seasonalmigration among New York Bight, Chesapeake Bay and Florida Lutcavageand Musick, 1985;Byles, 1989!, A similarmigratory pattern has been suggested for theloggerhead as well. Otherwise, it is difficult to explainthe regular occurrence of significantnumbers of viableand healthy sea turtles in northernwaters in the warmer months,We havea few Kemp'sridley tag returns from specimenstagged in Florida and recapturedas far northas Chesapeake Bay, and vice versa Table 3! Henwoodand Ogren, 1987!. These could be indicativeof what somemembers of thepopulation do to exploit thecrustacean-rich and mollusc-richforaging grounds north of F1orida and to avoid being exposedto the low temperaturesthat occur for part of the year. Seasonalmigrations by other marinepoikilotherms are not unusual, Sizedistribution of Kemp'sridley alongthe northern Gulf of Mexicois moreor lessunremarkable as compared to that alongthe Atlantic Table2!. However,some significance should be givento two areas:the smallest Kemp's ridleys found in Gulf coastal waters occur in I! western Louisiana and eastern Texas and ! Wakulla and Franklin counties in northwestFlorida, These turtles havecarapace lengths 0 to 25cm! that one would expectthem to haveattained during the timebetween that whenthe hatchlingsfirst enterthe pelagic stage offshore of the natalbeach and the time juvenilesenter the coastalbenthic stage. However, their precise age is notknown. Although we have no documented recordsof Kemp'sridleys from thepelagic stage, conversations with tunalong-line fishermen indicate that "baby" turtles do occur in the Gulf. In either case,favorable onshore currents exist tha t could bring them ashore to these two areaseast and west of the Mississippi Delta. Wind

121 Onthe Atlantic coast, coastal habitats as far north as Massachusetts areutilized by Kemp's ridleys during summer,The New England Kemp's ridleys frequently succumb tocold temperatures during November and December.Other Kemp's ridleys occurring south of Cape Cod may survive and some may migrate south to Floridato overwinter Ogren and McVea, 1982; Henwood and Ogren, 1987!, Preliminaryfindings, albeit qualitative, suggest that the northern Gulf coast from Port Aransas, Tex., to Cedar Key,Fla., is the foraging habitat for subadult Kemp's ridleys. Historically, Florida Bay in southeastern Florida wasidentified assubadult Kernp's ridley habitat but recent information islacking inthis regard Carr, 1952!. Onthe southeast coast of theU.S,, Kemp's ridleys are apparently common from Cape Canaveral north to ChesapeakeBay,but they are less abundant than in the Gulf, They are found inshore only during spring, summerand fall north of a lahtudecirca 20'N, During winter they apparently migrate either south or offshore towarmer waters and are apparently much more abundant atCape Canaveral during December toFebruary than at any other time. Turtlesize and wa ter depth relationships wereobserved fora largesample n= 79!ofjuvenile Kemp's ridleys fromnorthwest Florida. Most 91 percent! were captured indepths 6.1meters orless, but all of the Kemp's ridleysless than 25 cm in carapace length, except one, were collected from depths 0.9 meters orless. This supportsthebelief that their exploitable foraging habitat isrestricted toshallow depths because ofthe feeding energeticsofthese small individuals; i.e.,their high specific metabolic rateand an aerobic diving strategy Peter Lutz, University of Miami, personalcommunication>. Kemp'sridley captures were associated witha varietyofsubstrates andbottom types including mud, sand, oystershell and turtle grass Thalassia!. Nopreference wasindicated except when associated withportunid cr'ab distribution see above!. Acknowledgements I thankKaren Bjorndal and Anne Meylan for their constructive comments onthe manuscript, andJeff Schmid, EddieChadwick andSteve Rabalais fortheir able assistance incollecting Kemp's ridleys. Unpublished reports and recordsreceived from Anthony Amos, Tom Azarovitz, Gerald Corcoran, JakeDameron, PaulRaymond, JackRudloe, BarbaraSchroeder and others are gratefully appreciated.

Literature Cited Babcock,H.L. '1930. Carefte kempii from Massachusetts. Copeia 1930!:21. Brongersma,L.D.1972, European Atlantic turtles. Zoologische Verhandelingen, Rijksmuseum vanNatuurlijke Historic No.121, Leiden,The Netherlands, 318 p. Brongersma,L D.1982, Marine turtles inthe eastern Atlantic Ocean, p,407-416. In:Bjorndal, K A. Editor!, Biology andConservation ofSea Turtles, Smithsoruan Institution Press, Washington, D.C., 583 p. Byles,8 A.1989. Distribution andabundance ofKemp's ridley sea turtle, Lepidochelys kempi,inChesapeake Bayand nearby coastal waters This volume!. Carr,A.1952. Handbook ofturtles ofthe United States, Canada andBaja, California, Comstock Publishing Association, Ithaca, N.Y.,542 p. Carr,A.F. 1977. Crisis for the Atlantic ridley. Marine Turf le IVewstefter No.4, p. 2-3. Carr,A, 1980, Some problems ofsea turtle ecology. American Zoologisf 20!:489-498. Carr,A,1986, New pers pectives onthe pelagic stage ofsea turtle development, NOAATechnical Memorandum NMFS-SEFC-190, 36 p. Carr,A.F. and D.K, Caldwell. 1956. The ecology andmigrations ofsea turtles I.Results offield work in Florida, 1955, American MuseumIVordta tes 1793, 23 p. Chavez,H.1969. Tagging andrecapture ofthe lora turtle Lepidochelys kempi!.International Turfleand Tortoise SocietyJournal 3!14- 19, 32-36, Cokcr,R.E.1906. Thecultivation ofthe .IVorth Carolina Geological SurveyBulletin No.14, 69 p. Deraniyagala,P.E,P.1939. The distribution ofthe Mexican loggerhead turtleColpodfefys kempiGarman. Bulletin deI'Institute OceanographiqueNo.772, Monaco, 4 p. Dobie,J.L.,L.H. Ogren andJ.F. Fitzpatrick, Jr.1961, Food notes andrecords ofthe Atlantic ridley turtIe Lepidochelys kempi!from Louisiana.Copeia 1961!:109-110. Ehrhart,LM. 1983. Marine turtles of the Indian River lagoon system. Florida Science 46:337-346. Frazcr,N.B.'f986. Kemp's decline: special alarm orgeneraI concern. Marine Turfle IVewslet terNo. 37, p. 5-7. Hcnwood,T.A.and L.H, Ogren. 1987.Distribution andmigrations ofimmature Kemp's ridley turtle Lepidochelys kempi!andgreen turtle Cheloniamydas! off Florida,Georgia and SouthCarolina. Northeast Gulf Science9!:153-159. Hildebrand,H,H. 1982.A historicalreview of the statusof seaturtle populationsin the westernGulf of Mexico,p. 447453.In: Bjorndal,K.A. Editor!,Biology and Conservation ofSea Turtles, Smithsonian Institution Press, Washington, D.C., 583 p. Lazcll,J.D., Jr. '1980.New Englandwaters; critical habitatfor marineturtles. Copeia 1980!:290-295, Liner,E A.1954. The herpetofauna of Lafayette, Terrebonne and Vermilion Parishes, Louisiana. Proceedings ofthe Louisiana Academy o fSciences 17:65-84. Lutcavagc,M. andJ.A. Musick. 1985. Aspects of thebiology of seaturtles in Virginia.Copeia 1985!:449-456, Marquez,R. 1984. Kemp's ridley turtle overview ofbiology, p. 96-100. In: Bacon, P., F. Berry, K. Bjorndal,H. Hirth,L, Ogrenand M.Weber Editors!, Proceedings ofthe Western Atlantic Turtle Symposiunt, Volume I, Universityof MiamiPress, Miami, Florida, 306 p, MarquezM,, R,, A, VillanuevaO. and P.M. Burchficld. 1989. Nesting population and production of hatchlingsof Kemp'sridley seaturtle at RanchoNuevo, Tamaulipas, Mexico This volume!. May,E.B. 1973. Extensive oxygen depletion in MobileBay, Alabama. Limnology and Oceanography 18!:353-366, Meylan,A.B. 1986.Riddle of the ridleys.Natural History 951!:90-96. Meylan,A.B. and S. Sadove. 1986. Cold-stunning in LongIsland Sound, New York, Marine Turtle Neroslet ter 37:7-8. Mowbray,L.S. and D.K. Caldwell. 1958, First record of the rid ley turtle from Bermuda, with noteson other sca turtles and thc turtle fishery in the islands.Copeia 1958!:147-148. Ogren,L.H. and C. McV ea, Jr. 1982. Apparent hibernation by seaturtles in NorthAmerican waters, p. 127-132.In: Bjorndal,K,A, Editor!,Biology and Conservation ofSea Turtles, Smithsonian Institution Press, Washington, D.C,, 583 p. NationalMarine Fisheries Service. 1987, Final supplement to the final environmentalimpact statement listing and protectingthe greensea turtle, loggerheadsea turtle, and Pacificridley seaturtle underthe EndangeredSpecies Act of 1973.U.S. Departmentof Commerce,National Oceanic and Atmospheric Administration, National Marine Fisheries Service, 32 p. Pritchard,P.C.H. and R. MarquezM. 1973.Kemp's ridley turtle or Atlantic ridley. InternationalUnion for the Conservationof Nature and Natural ResourcesMonograph No. 2 Marine Turtle Series!, 30 p. Renaud,M L. 1986.Detecting and avoidingoxygen deficient sea water by brown shrimp,Penaeus aztecus Ives!, and white shrimp, Penaeussetiferus Linnaeus!. journal of ExperimentalBiology and Ecology 98:283-292. Schmidt, K.P. and E.R. Dunn, 1917,Notes on Colpochelyskernpi Carman. Copeia44:50-52. Distribution,Growth and Survival of HeadStarted, Tagged andReleased Kemp's Ridley Sea Turtles I.epidochelys kempi! from Year-Classes1978-1983 ClarkT. Fontaine, Sharon A. Manzella,Theodore D. Williams,Richard NI. Harris and William J.Browning'

A cooperativeprogrambetween theUnited States andMexico wasinitiated in1978 inan effort tosave fheKemp's ridley sea turtle Lepidochelys kempi!jrom extinction. Participating agenciesinclude thelnstituto Nacional dela Pesca ofMexico, NationalMarine Fisheries Service, U.S.Fish and Wildlife Service, National ParkService andTexas Parks and Wildlife Department. Parkofthe cooperafive program includes anexperiment inhead starfing Kemp's ridleys toincrease fheirsurvival during the firstyear oflifei n captivityandto establish a new nesting colony onPadre Island, near Corpus Christi, Tex. The head start research prol'ectinvolves removing eggsfrom the nesting beachat Rancho Nuevo, Mexico, hat chi ng the eggs andimpri nting the hatchli ngs atPadre Island, rearing thehatch lings toyearlings atGalveston, Tex.,then releasing thetagged year!i ngsi nto the Gulf ofMexico, Thispaper discusses thereleases andrecaptures of'the 1978 to1983 year-classes ofhead started Kemp's ndleys asof September 30, 1985. Bythe end ofSeptember 2985,8,241 head started andfagged Kemp's ridleys hadbeen released,and therehad been 399 reported recoveries.Included were reporfs from the east coast ofMexico, thecoast ofthe northern halfof the Gulf of Mexico, fheeast coast ofthe United States as far norfhas New York, and from France and Morocco. Growthofthe released Kemp's ridleys wasslower thanthaf ofKemp's ridleys ofthe same agesheld incaptivity. Headstarted turtlessurvived andgrew well after release, though fhose recovered inthe Atlantic Ocean grew slower than fhose recovered in theGulf ofMexico. Further, whenrecovery datawere compared tothe historical distribution ofKemp's ridleys asreported inthe literafureandbythe Sea Turtle Stranding andSalvage Network, ArchieCarr's speculation abouta northern WestAfrica nesting site was resurrected. TheKcmp's ridlcy sea turtle Lepidochelys kempi!islisted asendangered underthe United States Endangered Species Actof 1973. In1978, a multiagency, cooperative, conservation effortwas initia ted by the Instituto Nacional dcla Pesca INP!of Mexico, theU.S. Fish and Wildlife Service FWS!, the National Park Service NPS!, the National Marine FisheriesService NMFS! andthe Texas Parks and Wildlife Department TPWD! towork toward Kemp's ridlcy recovery

Materials & Methods Rearing Duringeach nesting season beginning in1978, a small portion

' Nationa! Ivtarine Fishenes Service

124 to PadreIsland, After incubationof the eggsin boxeswithin a hatcheryat theNational Seashore,the hatchlingsthat emergedwere allowed to crawlacross the Padre Island beach to enterthe surf in hopesof strengtheningor reinforcing the imprintingprocess. NPS personnel then collected the hatchlingswith dipnetsand put themin boxesto be transportedby U.S,Navy or CoastGuard aircraft to theNM FS SEFC head start research facilities in Galvestonwhere theywere reared in captivityusually for 10to 11 months before being released usually offshore of PadreIsland. These yearlingshad been tagged in severalways before being released into theGulf of Mexico. Head start facilitiesand husbandrymethods have been described in detail by Fontaineet al. 985! and Fontaine etal. 989!. First-yeargrowth and survival during head starting have been described by Caillouetet al. 986!. Tagging Usuallyabout 30 days before the annually scheduled release of headstarted Kemp's ridleys, all yearlingsin good healthand physical condition are tagged with metalflipper tags. This allows time for observahonstobe made on tag retentionand on possible infection caused by tagging,as well asfor appropriateremedial actions. It alsoallows time for tag-relatedmortality, if any, to occurbefore the release, Flippertags used on head started turtles Table1! werepurchased from the National Band and Tag Company and wercHasco type, style 681, self-piercing, self-clinching tags. All tagswere monel metal except for inconeltags used on 100turtles of the1980 year-class thc latter tags were obtained from George Balazs, NMFS, Southwest Fisheries Center,Honolulu Laboratory!.The G- and F-seriestags used on the 1978year-class, and the K-scricstags used on the 1980year-class, werc obtained from Dr. ArchieCarr, and were inscribed with themessage "Reward Primio Rcmite, SendDept. Biol., U.F., Gainesville, Fla. U.S.A." The NNA-series through NNQ-series tags, used on the1979 to 1983 year-classesand on one turtle of the1978 year-class were inscribed with themessage "Send NMFS Lab, Virginia Key, Miami, FL 33149." The 800-series inconel tags from Hawaii were i~scribed with the message "Write H,I.M.B. University,Hawaii 96744." Two odd tags, No. 13582 one turtle of the1978 year-class! and J0096 one turtle of the1979 year-class!were also used. Hippertags used on the1978 to 1981year classes were first soaked in gasolinefor 24 hours then in 90percent alcohol for 24 hours prior to use, without apparent problems.Tags used with the 1983year-class were sterilized by autoclaving,The tags were applied to eachturtle usinga standardHasco tag applica tor. Thearea of tag insertionwas swabbedwith tinctureof iodine.Ncosporin, a topicalantibacterial ointment, was placed on thesharp point of thetag and the tag was insertedon the trailing edgeof the right front flipper. Someof the turtles were double-taggedwith flippcr tags;i.e., a secondflipper tag wasattached to the left front flipper. The double-taggingwas donein hopesof increasingthe chancesthat oneof the flipper tagswould be retained.However, it wasrecognized that if eachtag had the sameprobability of loss,then, on theaverage, both would be lost aboutthe sametime, For this reason,additional tag types internal magnetictags and living tags! were also used on someyear-classes. However, only recapture resultsfor flipper tagsare reported herein, because recovery reports from the internaltags and living tagsarc indccd rare events, as they require special equipment magnetometer! or expertise to be recognized. Releaseweight of eachhead started turtle was takenat the time of taggingso some growth could havetaken place between the time the turtles were tagged and the 30 or 50 days thereafter when they were released.An 0'haus "Dial- o-gram"beam balance was used for weighingsto the nearest0.1 gram. Mcasurcmcnts of carapacelength and width werealso made and werestraight line measurementsas recommended by Pritchardet al. 983!. Mcasurcmcntswerc made to the nearest 0,1 cm, initially with a hard plastic rule and later with calipers. Data on length and width measurementsare available, but only weights are included in our analyses herein. Release Turtleswere transportedto releasesites Table1! in cardboardboxes described by Fontaineet al. 985, 1989!.All turtles were released in the Gulf of Mexico or in adjacent bays. Most turtles of the 1978and 1979year-classes werc released in Florida waters, 113 of the 1978year-class and six of the 1979year-class were released off Texas and 197 turtlesof the1980 year-class were released from theNOAA researchvessel Oregon II in the Bayof Campeche, Mexico. All the rest werc rcleascd in Texas waters, either offshore or inshore. In 1983,96 of thc 1982 year-class turtles werc releasedin NuecesBay, Tex., part of the CorpusChristi Baysystem near the PadreIsland National Seashore.Other releasesof small groups or individual turtles represented turtles held back from the major releasesfor a variety of reasons e.g., duc to injuries or illnesses, being too small at times of major releases,or because they were used for potentialbrood stock!,Thc total taggedturtles releasedwas 8,241, or 79.4percent of the 10,376imprinted hatchlings of the 1978 to 1983year-classes received alive from NPS for head starting Table 2!. Sources of Recoveries Reportsof head startedKemp's ridleys found in the wild were receivedby telephoneand from correspondence from the public, the Sca Turtle Stranding and SalvageNetwork, Dr. Archie Carr's office at the University of Florida

125 Table1. Summary ofimprint groups, release sites, dates of release, numbers ofturtles released and tag code series usedfor head started Kemp's ridley sea turtles of the1978-1983 year-classes.

Year- Imprint Release da te No. Tag code class ou Release site /Mn/Yr Released seriess 1978 PINS" SandyKey, Fla, 22/02/79 135 PINS East Cape, Fla,' 22/02/79 52 G G PINS East Cape, Fla, 28/02/79 1 13582 PINS East Cape, Fla. 28/02/79 166 PINS Sandy Key, Fla. 05/03/79 172 RN Homosassa, Fla. 08/05/79 751 G ,F PINS Homosassa, Fla.' 08/05/79 628 G ,F PINS Padre Island, Tex. 07/07/79 112 G ,F RN Padre Island, Tex. 07/07/79 I G0985 PINS Homosassa, Ha. 03/06/80 I NNA260 Subtotal 2,019 1979 PINS Homosassa,Fla. offshore!' 03/06/80 665 RN Homosassa, Fla.

1,033 7 208 Total 8P,41 'PINS= imprintedat PadreIsland National Seashore, and RN = imprinted at Rancho Nuevo. Morrelmetal tags, unless noted otherwise. Each dash represents three or four numerical digits from 0-9; actual numerical series arenot given because they were mixed. Details concerning numerical series can be obtained from Sharon A.Manzella, NMFS SEPCGaivestorr Laboratory, 4700 Avenue U, Galveston,Tex., 77551. ' Thisrelease included turtles also tagged with radio-transmitters seeKlima and McVey, 1982; Wibbels, 1984!.

126 and the EndangeredSpecies Program Office, NMFS, SEFC, Iviiami, Fla. Unfortunately, in manycases, complete information was not available for each recovery,

Resu1ts

Recoveries by Year-class Out of399 recoveries Table 2!, 88were excluded from subsequentsummary tables and figures.These 88 turtles had beenrecovered in 10or fewerdays of theirrelease. Therefore we believed such turtles were not representativedue to disorientation,weakness, illness or other problemsat the time of release.The remaining311 recoveries included some turtles recovered more than once. The condition of eachturtle as to whetherlive or dead at recoveryis shownby year-classin Table3. Most 0,8 percent!were reported to havebeen alive and returned to theenvironment. Unfortunately, the flippertags were removedfrom someof the turtles recapturedalive before they were returned to the water, so they could not be recognizedand reported again as tagged turtles if recaptured,For this reason, we recommend that themessage on theflipper tagsbe changed so that the finder clearly understands that it isimportant to leave the tag affixed to the turtle if it is alive. Severalhead started turtles whoseflipper tagswere not removedon first recapturewere recapturedon one or more occasions, Geographic Distribution Reportedrecoveries by year-classand state or countryindicated diverse geographical distribution Table4!. The geographicaldistribution of recoverieswas related in partto therelease location Tables I, 5-7!.For example, of the 89 recoveries from turtles released in southern Florida ,906 from the 1978year-class and 1339 from the 1979year- class!70 8.6 percent!were from the Atlantic Ocean, along the eastern coast of theUnited States as far north as New York, andat leasttwo crossedover to Europeand North Africa. Of the turtlesreleased in Texas,794 from all six year- classes!,208 95.8percent! were recovered from the Gulf of Mexico Table6!. Of the 197turtles of the 1980year-class releasedoff Mexico,four 80.0percent! were recoveredin the Gulf of Mexico,but one was recoveredin New York Table 7!, The wide dispersalof headstarted Kemp's ridleys alsocan be discernedby comparingsame day recoveries.For instance,two turtles wererecaptured on June15, 1983, one from the 1979year-class at CoreSound, N.C. and theother from the 1981year-class within the GalvestonBay near Texas City, Tex.On June4, 1981,a turtle from the 1980year- class was recovered at Galveston and another from the 1978 year-class was recovered at Bradley Beach, N.J. On November1, 1981,a turtle from the 1980year-class that hadbeen released in the Bay of Campechewas recaptured at Morgan City, La., while anotherfrom the 1980year-class that had beenreleased off PadreIsland was recovered southof Tampico,Mexico. On November20,1982, a turtlefrom the1979year-class that hadbeen at largefor 893days wasrecaptured in a lagoonnear El Jadida,Morocco, another from the 1980year-class that had beenat large for 536 dayswas found strandedon thebeach at Panacea,Fla., and a third from the 1981year-class that hadbeen at largefor 171 days was recovered at South Pass,Ala.

127 Table3. Numbers of headstarted Kemp's ridley sea turtles of the1978-1983 year-classes recovered live, dead or in unknown condition'.

Ycar- Percentage class Condition Jan-Mar Apr-Jun Jul-Sep Tota 1 by condition 1978 Live 27 19 60 83.3 Dead 5 0 9 I 3 5 12,5 Unknown 3 0 0 3 4.2 Total 06.9 5 33 24 10 72 Percentage 45.8 33.3 14.0 by season 1979 Live 14 70.0 Dead 5 I 5 I 4 I 3 15.0 Unknown 0 7 2 8 3 15,0 Total 6 22 Percentage 30.0 30,0 40.0 by season 1980 Live 20 IS 48 57.1 Dead 6 7 40 0 4 4 15 17.9 Unknown 9 7 5 21 25.0 Total 33 29 18 84 Percentage 4 4.8 39.3 34.5 21.4 by season 1981 Live 14 7 27 54.0 Dead 2I 3 8 18 36.0 Unknown 4 3 I 40 37 5 10,0 Total 6 23 14 50 Percentage I 2.0 46.0 28.0 14,0 by season 1982 Live 21 10 I 35 44.9 Dead 3I 2 13 15 30 38.5 Unknown 0 3 5 4 13 16.6 Total 39 29 78 Percentage 6 7.7 50.0 37.2 4 5.1 by season '1983 Live 3 71.4 Dead 20 0 28,6 Unknown 0 5 0 Total 2 Percentage 71.4 28.6 by season

Combined Live 14 90 60 25 189 60.8 Dead 5 31 29 12 77 24.8 Unknown 2 18 15 10 45 14.4 Total 21 139 104 47 311 Percentage 6.8 44.7 33.4 15.1 by season ' Asof September 30,1985, and excluding allrecoveries thatoccurred 10days or less after release.

128 Table 4. Geographicaldistribution of recoveriesof head startedKemp's ridley seaturtles of the 1978-1983year- classes'.

State or Year-Class Count 1978 1979 1980 1981 1982 1983 Total Mexico 3 2 5 Texas 54 21 69 154 Louisiana 17 18I 9 49 Mississippi 4I I3 2 5 Alabama 4 Florida 25 13 43 3I I3 9 Georgia S. Carolina 6 9 12 N. Carolma 19 Virginia 2 Maryland 172I I2 2 New Jersey 2 New York 2 France I Morocco I Unknown I Total 72 20 84 50 78 311 ' As of September30, 1985,and excludinga11 recoveries that occurred10 days or lessafter release. s Anindefinite LORAN readingwas the only recapturelocation reported for this turtle from the 1980year-class; therefore it cou!d not be assigned geographically.

Table 6, Relationship between recovery location state or country! and days at large for head started Kemp's ridley sea turtles of the 1978-1983year-classes released in Texas'. Days at State or Country Lar e Mexico Tex. La. Miss. Ala. Fla. Gulf! Fla. Atl.! Ca. S.C. N.C. NRb Total 11-49 40 2 43 50-99 14 I 15 100-149 I 11 I I 14 150-199 2 3 3 I I 103 200-249 2 I 250-299 I 7 2 10 300-349 I 16 9 I 28 350-399 26 13 I 40 400-449 '10 3 I 16 450-499 11 3 I 500-549 I 3 I I 157I 2 550-599 I 600-649 I 650-699 I I 700-749 4 750-799 I 2I I6 800-849 I

1150-1199 1200-1249

Total 5 151 43 3 3 3 2 3 2 1 1 217 Me Table I regarding releases;recoveries as of Sept. 30, 1985,excluding all recoveries that occurred within 10 days or less after release. 'Not reported Ch GC R N 0 Ch N Z 00 Ch 'Z u w ~ % W N r c5 CI5 00 0CJ Ch QJ 0

0

2 C5

Clj QJ III ;a Ci I

6 ItI 0 II5 a 'aV CJ 00 8 g C5

'U N N

0 U ~ tU 00 0 C0 0 0 'V Ig XQl IPj Q QO Ch

A O

III

0 CJ Q 0QP

W o pK fd Q

C ~ "a 000 t R gp CJ Ch M g Ch Ch Ch r N lA lh N Ã, X ma,R R +8 QJ Ch R w w N ch A W W lh R8 BR BR R a~ 8 5 Geographicaldistribution of recoveriesof headstarted turtles wascompared with that of strandedKemp's ridleys in Figure'I. Thestranding data were obtained from monthlyreports published by the SeaTurtle Stranding and SalvageNetwork STSSN!,NMFS Southeast Fisherics Center, Miami Laboratory. The strandingdata in Figure1 do not include strandingsof headstarted turtles. In general,the patternof geographicaldistribution of Kemp'sridley recoveriesand that of strandingsappears to be similar, with somedeparture for Louisiana,Florida, Virginia and Massachusetts.Only thosestates in which both recoveriesof headstarted Kemp's ridleys and strandingsof Kemp's ridleys had beenreported were includedin this comparison,although the other statesin the geographicseries are shown in Figure 1, Monthly frequencydistributions of recoveriesand strandingsalso were similar Figure 2!. Strandings from Massachusettsin Novemberand Decemberof 1981and 1982were excluded from this comparison becausewe believedat the time of writing this paper in 1985that the sporadicoccurrence of Kemp'sridley there representedaberrant behavior, Similarities in distributions of strandingsand recoveriessuggest the head started turtles successfully integrate themselves into the natural Kemp's ridley population after release. When recoveriesare viewed with respectto the relationshipsamong release area Table 1!, stateor country of recovery Table4! and daysat large groupedby 50-dayintervals!, interesting patterns emerge Tables5-7!. As one might expect, dispersion widened as days at large increased. There also was an annual resurgence of recoveries Tables 5-7!, probably reflecting a seasonalphenomenon related to behavior of the turtles, seasonalfishing activity, or recreational pursuits that bring people to the coasts. The Florida releases Table 1! produced recoveries with the widest dispersion and greatest days at large Table 5!, as these were the oldest turtles of all year-classes.Escapement from the Gulf of Mexico to the Atlantic Ocean characterized the Florida releases,whereas the Texasand Mexico releaseswere characterized by a predominance of recoverieswithin the westernGulf of Mexico Tables5-7!, These results became the primary rationalefor discontinu- ation of Florida releasesafter the first two yearsof headstarting in favor of releasesoff Padreand MustangIslands, Tex., in hopes tha t the turtles would rema in and mature in the Gulf of Mexico. Florida releaseshad been urged by some turtle biologistsin theearly years, because turtles of yearlingsize are known to havebeen abundant in Floridawaters in the past, Seasonal Distribution Recoveriesshowed a definite seasonality Table 3!. This may be related to seasonality in commercial and sports fishing activities, environmental conditions and recreational activities along beach fronts, or a combination of these variables. Recoveries by Year-Class

1978 Year-Class The release and subsequent recovery of the 1978 year-class has been discussed by McVey and Wibbels 984!. Becauseturtles from this year-classhave been at large the longest time, we examined recoveries from this yearclass in greatestdetail. Most ,906 turtles, or 94percent! of the turtles releasedfrom this year-classwere releasedin the Gulf of Mexico off Florida Table 1! becausethis area had been identified as a typical habitat of juvenile Kemp's ridleys. Of the 76recoveries from this yearwlass,57 5 percent!were madeon the Atlantic coastof the United States.As of September30,1985, the latestrecovery from this yearclasshad occurredon August18,1983 at Hunting Island,SC,, and the turtle was alive and released. This turtle had been released into the Gulf of Mexico on May 8, 1979 at Homosassa,Fla., and had been at large for 1363 days. Unfortunately, no lengths or weights were reported. Recoveriesof the 1978year-class, as with other year-classes,evidently were seasonal,with most recoveriesreported

131 40

20

20

40

Figure1.Comparison ofgeographic distributions ofrecoveries ofhead started Kemp's ridley sea turtles and Kemp's ridley strandings datafrom theSea Turtle Stranding and Salvage Network! during calender years 1979-1984.

132 30

20 I- X UJ O10

0 JANF M A M J J A S 0 N D MONTHS Figure2. Comparisonof monthlydistributions of recoveries ofhead started Kemp's ridley sea turtles and Kemp's ridley strandings ldata from the SeaTurtle Strandingand SalvageNetwork! during calenderyears 1979-1984, during warm months Table3, Figure3!. The paucity of reportedrecoveries during the cold monthsmay be linked to reduced commercial and recreational fishing and other ou tdoor activities, as well as possible seasonalmovements of sea turtles. Annual recoveriesfrom the 1978and other year-classesdeclined from year to year as would be expecteddue to mortality,but the decline could also have reflected tag loss Figure 3!. Observations on tagged turtles at private marine aquaria Larry Ogren, NMFS, SEFC, Panama City, Fla,, personal communication, October 1985! and from metal flipper taglossstudies Henwood,1986!on loggerheadsea turtles Careltacaretta!indicatehighrates of tagloss during the first two yearsafter tagging,Such loss of flipper tagsmay be due to oneor a combinationof factors: I! improper initial application of the tag, ! deterioration of the tag from the corrosive action of seawater, and ! tag sloughing due to rapid growth of the tur tie in relation to size of the tag, infection leading to necrosis,or both. Nevertheless,two recoveriesof the 1978year-class were reportedin 1983,five yearsafter release Figure 3!, so the metal tagscan be retained in some cases. During 1979,turtles of the 1978year-class were recovered only in Florida and only during the first three quarters of the year. Few recoverieswere reportedfrom October1979 through March 1980,then in spring April-June!and summer July-September!of 1980the numberof recovericsincreased, particularly alongthe eastcoast of the United States. Three recoveries were reported from North Carolina during winter October-December! 1980, and none during spring January-March! 1981.Turtles of the 1978year-class reappeared from Georgia to New Jerseyduring summer-fall1981,a periodof 768to 950days after release and following two wintersat large.Only onerecovery from the 1978year-class was reportedduring 1982,One recoverywas reported from SouthCarolina in 1983after three winters at large, Five turtles from the 1978year-class were recovered more than once, One tag G0104!was recovered along the U.S. east coast on three different occasionsafter its releaseon February 22, 1979,at Sandy Key, Fla., at a weight of 0.7 kg. lt wasrecovered 47 dayslater on April 10near Miami, Fla,,and weighed0,6 kg, 0,1kg lessthan at release.The turtle wasin goodcondition and wasre-released on thesame day. It nextappeared 730 days later on April 9,198! near Ocean City, Md. It againappeared in good healthand was re-releasedon the sameday no weight or measurementswere taken!. The turtle was recovered a third time 56 days later on June 4, 1981at Bradley Beach,N,J. At that time it was reported to weigh 4.5kg, and was releasedon the sameday, This turtle traveled from SandyKey, Fla,,to Bradley Beach,N.J., over an 833 day sojourn spanning two winters.

133 20

10

0

10

v!

10

O 0 20

10

0

20

0

10

Figure3.Seasonal distribution ofrecoveries ofhead started Kemp's rulley sea turtles, byyear-class andcalender year of recovery.

134 Fourturtles of the 1978year-class were recovered twice. One tag G0460!was recovered first on September25, 1979, 209days after release at JekyllIsland, Ga., and was released; 26 days later it wasrecovered at thesame location and weighed3.9 kg. Another tag G0904! was recovered on August25, 1980, 475 days after release, at ColonelIsland, Ga. It was re-releasedand recoveredagain 14days later at the samelocation on September8, 1980.The turtle wasre- releasedin goodcondition, but no measurementsor weight weretaken. A third turtle tag G2385!was recoveredon April 4, 1980,352 days after release, at Holly Beach,La. It wasre-released and again recovered 59 days later on June 22,1980 at JohnsonBayou, La., at which time it weighed2,7 kg. On September2, 1981,the fourth turtle tag G2830! wasrecovered at Edisto Beach, S C.,848days after release.After re releaseon thesame day, it wasfound dead20 days later on September22 on the samebeach and in the samelocation. It weighed4.5 kg when found dead.From the time of its releaseon May 8, 1979until it was found dead,this turtle had survived two winters. 1979 Year-Class As of September30, 1985, 21 .6 percent!recoveries had been reported from the 1979year-class. One recovery tag K1773!was from a turtleheld into its secondyear in captivity beforebeing releasedin 1981with the 1980year-class offshoreof PadreIsland, Tex. By September30, 1985, the mostrecently recovered turtle tag NNAOOI!from the 1979 year-classwas recovered alive and re-released at CoreSound, N.C. on June15, 1983. It originallyhad been released offshore of Homosassa, Fla., on June 5, 1980. Two of the most interestingrecoveries of head startedKemp's ridleys were from the 1979year-class. One tag NNN893! released offshore of Homosassa,Fla,, on June 5, 1980 was recovered 568 days later at Biarritz, France on December 25, 1981 Wibbels, 1983!.It weighed 2.0 kg at the time of recovery. Another tag NNN678! released with the first wasrecovered 893 days later on November20, 1982 at El Jadido,Morocco. This turtle was found in a lagoon, wrappedin a fisherman'snet, and wasmeasured before being released. The turtle reportedlyweighed 20 kilogram s, and was the largest of all head started Kemp's ridley reported as of September 30, 1985. 1980 Year-Class Included amongthe head startedturtles releasedfrom the I980year-class were 197that had beenimprinted as hatchlingson the RanchoNucvo beach.They werereleased in the Bayof Campechenear Arrecifc Alacran,Mexico, on March 3, 1981,from the NOAA researchvessel Oregonli, From this Bay of Campeche release, five .5 percent! recoverieshad beenreported, as of September1985 Table 7!. Two werereported from Louisiana,within 99days after release Table7!, Between200 to 249days at large,onr more wasrecovered from Louisiana.The last two from this grou p were recovered 500 to 599days after release;one of thesewas recovered at Fire Island, N Y. on August 14, 1982. Of the turtles released from the 1980year-class, 1,526 88 6 percent! were released offshore of Padre Island on Junc 2,1981.All but two of the recoveriesreported from this releasewere made in the Gulf of Mexico. One of the exceptions tag K1495! was rccovcrcd alive at Ossabow Island, Ga. on August 19, 1982,443 days after release, and it was re- releasedon the sameday. The other < tagK1625! was recovered alive at Hunting Island, South Carolina on September 9, 1982and was rc-released on the same day. The preponderance of recoveries from the 1980 year-class have been made in Texas and Louisiana Table 4!, although a few migrated to the south and east.The single most frequent number of recoveries 2! taken in one location from this year-classwas near Matagorda, Tex., approximately 161 kilomctcrs cast of the releasesite, between 43 and 454 days after release, 1981 Year-Class The 50 .0 percent of those released! reported recoveries from the 1981year-class were made from the northeast coast of Mexico to South Carolina, but the majority of these,39 8.0 percent of the recoveries!, were made in Texas and Louisiana Table 4!. 1982 Year-Class Most turtles ,159 turtles or 87.5 percent of those released! of the 1982year-class were released offshore of Padte and Mustang Islands into drifting sargassumweed patches, with the intent of providing them with cover and an immediatefood supply Carr, 1986!,However, within 10days of release,77 of theseturtles had beenfound stranded on beacheswithin 32 kilometers of the releasesite. All showed evidence of contact with oil to varying degrees,and 20 were found dead. All of the dead turtles examined appeared to have ingested tar or crude oil. The 56 turtles found alive were cleansedof oil and rehabilitated, either at the Padre Island National Seashoreor The University of Texas Marine Science Institute, Port Aransas, Texas. They were subsequently released offshore of Mustang Island by Anthony Amos. None of the 77 turtles stranded within 10 days of release were included in our data summaries. Out of the total ot'1,159Kemp's ridlcys releasedfrom the 1982ycar-class off Padreand Mustangislands on Junc 7,1983,24 were recovered from theCorpus Christi Bayarea, one from GalvestonBay and onefrom Brownsviile,Tcx. Two of the turtles from this offshore release were recovered virtually at the exact site of the inshore release within NuecesBay Table1!, one at 20 and the other at 28 days after releaseoffshore, This suggestsa selectionof inshore habitat by head started Kemp's ridleys. Seasonaldistribution of recoveriesof the1982 year-class shown in Figure3 ismuch like thatof otheryear-classes. Recoveriesof headstarted turtles from the 1982year-class began to be reportedon the westernGulf coastin March 1984and reporting continued through the warm months of thatyear, This year-class was foremost among the year- classesfor which reportedrecoveries were confined almost entirely to the westernGulf of Mexico Table4!. 1983 Year-Class Therewas a poor hatch 2 percent!in the1983year-class a tthe Padre Island Na tional Seashore, Only 250 hatchlings that had beenimprinted at RanchoNuevo were receivedfor headstarting. Within 10 days of the releaseof 190 tagged survivors on June 5, 1984,four recoveries were made. From that time to September30, 1985,seven morc recoverieswere reportedfrom the 1983year-class Table 4!.

Growth Growth in weight g! was examinedin head started Kemp's ridleys that were recoveredafter release.When carapacelength but not weight was reported,the following weight-lengthequation from McVeyet at. manuscript!, basedon 5,064 pairs of weightand length from head started turtles ranging in sizefrom hatchlings to 20.0kilograms in weight,was used to estimateweights from lengths: W 0 2301 Liar where W = weight in grams L = straight line carapace length in centimeters Weightsof rccovcredturtles are plotted against years at large Figure4!, for two differentgroupings of points,one for turtlesrecovered in the Gulf of Mexicoand the other for turtlesrecovered in the AtlanticOcean. Separate exponentialgrowth curves, one for theGulf andthe other for theAtlantic, were fitted to thedata points using the following model in logarithmic forin!: lnW = lna + bt, where W = weight in grams, t = years at large, and a and b are estimated parameters. Theexponential curves derived therefrom are shown in Figure4. The slopes, b, wercsignificantly different P<0.05!, and growth rate was greater in the Gulf of Mexico than in the Atlantic. Growthin weightof captive-rearedKcmp's ridleys over a six-ycarperiod at Sea-AramaMarineworld in Galveston wasexamined by Caillouctet al. 986!, andthe fitted growth curve for thecaptive-reared turtles is drawnin Figure 4 forcomparisons with curves fitted to recapture weights, Growth in captivitywas more rapid than growth of head started turtles released into the wild. Growthin turtlesfrom differentyear-classes could not be validly compared because it wasconfounded with the effectsof location Gulf of Mexicovs. Atlantic Ocean! and because the growth histories of thedifferent year-classes spanneddifferent years Figure 3!. Conditionof turtlesand location and method of recoveryinformation on whetheror not theanimal was alive or deadat recoverywas lacking in somecases, but themajority were reported to havebeen found alive and released Table 3!, RecoveryIocahon of headstarted turtles was 54.0 percent offshore and 46.0 percent inshore, despite thc fact that 988 percentofthe turtles were released inwaters seaward ofthebamer islands. Only the 96 turtles released inNueces Bayrepresented an inshorerelease Table 1!. This suggests that the turtles seek out theinshore environment, even whenreleased offshore. The dominant methods of recoverywere unknown 3,1 percent!,shrimp trawl 8.9 percent!,stranded dead 5.1 percent!and stranded alive 0.4 percent! Table 8!. Percentagerecovery Table 2! probably reflects some tag loss. Also, all recoveriesare not reported, and reporting ratesfor recoveries are not constant, either spatially and temporally, or amongyear-classes. Wehad no information totest the various effects on recovery rates. In somecases e.g, 1982 year-class! the recovery was high 1.2 percent! becauseofoiling of the turtles shortly after their release. Thus, we can consider the observed recoveries torepresent anunderestimate for headstarted Kemp's ridleys released into the wild. Figure3 depictsseasonal variation in recoveriesas well as the year to year declinein recoveries.

136 20

16

I- UZ 8

0 0

YEARS AT LARGE

Figure4. Scatterdiagram and fitted exponent alrelationships between recapture umght and yearsat largefor headstarted Kemp's ridley sea turtles recaptured in theGulf of Mexico d otsand solid! ine! andAtlantic Ocean circles a nddashed line! ascompared to grotothcurve for Kemp's ridleysin captivity small short-dashedline! af Sea-AramaMarineworld, Galveston,Tex.

Discussion Carr and Caldwell 958! suggestedthat Kcmp's rid icy may bc endemic to the Gulf of Mexico, and that records of its occurrence on the Atlantic coasts of North America and Europe are based on stragglers lost to the Gulf via the F1oridacurrent and Gulf Stream, Two major forage grounds for mature Kemp's ridleys have been identified, one in Campeche, Mexico and the other off western Louisiana Chavez, 1968;Pritchard and Mkrqucz, 1973!.In Massachu- settswaters, Kemp'sridleyappearstobernorecommon than theloggerhcad,and thisalso seemstobe thccasein Nova Scotia Dodge, 1944; Bleakney, 1965!, In Virginia Musick, 1979!, most ridlcys were small <18 kg!, Lazell 976! suggested that Kemp's ridley's occurrenceoff New England may be cyclic and that this may be true for Virginia as well. Carr 963, p. 300! stated: "It is possible,for example, that the European ridlcy records represent waifs permanently lost to the breeding population." Carr 967!, apparently reconsidering the possibility of a nesting location south of European Atlantic waters, stated that: "no ridlcy of any kind is known from Morocco, Spanish Sahara,and the Canary and Cape Verde Islands, or from anywhere in the open ocean from Mauritania up to Azores." Later, Brongersma 972! statedthat: "Kemp's ridlcy has not beenfound in the southernarea of Francebut hasbeen recordedfrom Ireland, Scotland,England, Wales, Channel Islands, the Netherlands,Azores, and Madeira." All specimensfrom Europeanwaters have been classified as being young to abouthalf grown. A specimenrecorded from the Azoresin

137 Table8. Summary of recoverylocation, method, tag status, and turtle condition for head started Kemp's rid 1 eysea turtles from year-classes1978-1983'.

Year-class 1978 1979 1980 ]981 1982 1983 Combined Recovery Location Offshore Beach 9 0 11 101 43 76 Surf 2 8 Open Water 21 2 7 65 Other 164 2 0 152 2 9 Unknown 5 0 10 Inshore Beach 5 2 3 2 197 Surf 50 0 Open Water 10 157 30 Other 168 136 46 Unknown 18 41 Method of Recovery Shrimp trawl 34 241 16 90 Gill net 137 2 0 3 11 By hand 0 0 1 0 2 Hook and line 10 2 18 Stranded dead 37 5 10 7 20 47 Stranded alive 5 2 7 32 Unknown 35 26 14 20 103 Other 3 1 1 8 Tag Status' Tag left on turtle 51 14 44 25 38 178 Tag removed/alive 11 1 12 5 33 Tag removed/dead 0 0 4 30 8 10 25 Double tags 0 0 0 25 25 Unknown 13 6 26 14 16 75 'The listed categoriesare not alwaystnutually exclusive.

1939was estimated tobe 4 monthsold. Mowbrag and Caldwell 958! referred to a Kemp'sridley from Bermuda that wascaptunC on March 31, 1949, and weighed 6.8 kilograms, However, its species identification is in doubt Pritchard, 1969!. Brongersma972! statedthat: "L, kempi breeds in theGulf of Mexicoand, as far as we know, only there. All specimensfound alongthe North Americaeast coast, in EuropeanAtlantic waters,in theAzores and off Madeiramust havecome from the Gulf of Mexico.The fact that in AmericanAtlantic coastal waters L. kempi seems to be more commonthan C. caretta may perhaps be explained by C.caretta being a moreoceanic species, the young of whichat a veryearly age already move out into the open ocean, while L. kempi tends to keep to shallower water, closer to the coast.This may be one of the reasons too that fewer Kemp's ridleys than loggerheads reach European Atlantic waters. Anotherreason may be that the nesting beaches of L, kempi are further away from Europe." Onthe basis of data from tagging, shrimp trawler catches, and extensive observations, Hildebrand 982! suggested thatthe entire life cycle of Kemp's ridley may be completed within the Gulf of Mexico. Carr 980! also thought this tobe a definitepossibility but also that it wasprobable that the entire population was not so contained, Small ridleys occurfrequently onthe coast of New England Lazell, 1979! and Virginia Lu tea vage and Musick, 1985! which might representa regularstation in thedevelopmental ecology of a partof thespecies. Carr 980! andWitham 980! both believedthat the warm wa ters of theGulf Stream supplied the northern waters with rid leys regardless ofwhether the youngturtles emerged from the current farther south and swam north, or were transported into New England coastal watersbylocal eddies of,or separated from, thc Gulf Stream. Unless the northern ones are lost to the population they mustswim to Florida under their own motivation and efforts. Carr 980! theorized: "The alternative possibility, that theynever go back at all, thatall post-Floridaridleys may be lost waifs, is reinforcedby theoccurrence in European watersof juvenileseven smaller than the average for theU.S. Atlantic coast. Most European specimens range in shell lengthfrom 10to 25cm, and they are derived from Old Worldnesting grounds, but L. kempihas none there." Carr 980! further reinforcedhis belief that the Atlantic Kemp'sridlcys arelost waifs neverto return to theGulf of Mexico, andconcluded that if thelife cycleof theKemp's ridley is completedwithin theGulf of Mexico,then "the lost year puzzleis con firmed as the most subs tan tial of all obstaclesto understandingthe ecology of seaturtles." However, Carr 986! relaxedhis positionregarding thc lost year hypothesis after recognizing an association between juvenile sea turtles, sargassumweed and oceanicconvergence zones, Witham976! arguedstrongly for oceancurrent transport of youngsca turtles and suggested that such currents mightalso be important to adultsea turtle dispersal, It hasbeen estimated Hughes, 1974! that hatchling sca turtles enteringthe sea at a Caribbeansite would require four to seven years to circumnavigatethe North Atlantic and return to their natal beach,On the other hand, Witham 976! and Witham and Futch 977! indicated that such circumnavi- gationcould take less than one year in theNorth Atlantic gyre, which includes the Florida Current and Gulf Stream. Obviously,not everyoneagreed with Carr's955! "lostwaif" theory.Pritchard and Marquez 973! postulated thatmost of theseyoung ridleys that leave the Rancho Nuevo ~esting beach and are carried around the Gulf of Mexico to theFlorida Keys then up theU.S. cast coast as far as Canada eventually migrate back to theGulf of Mexico to Rancho Nuevoto reproduce,even though a feware carried by theGulf Stream to Europeanwaters, Lazcll 980! statedthat "NewEngland's sizable aggregations of marineturtles have been known to be regular, healthy migra n ts,rather than waifs,in New Englandwaters since Bleakney 965! first calledattention to their robust,aggressive, and good condition." At the time Carr 975! wrote his paperthe greatestconcentration of positively identified Kemp'sridleys thathehad heard of,other thanthatat thenestingbeachatRancho Nucvo, occurred in Martha'sVinyard, Mass,Lazell 976! notedthat L. kernpiwas a regularmigrant to Massachusettswaters, and Lazell980! wrote: "Thevast majority of ridleys seenby me in New Englandwaters were active,healthy, and fed voraciouslyat ambientwater temperatures.The evidence available for thenatural, normal, and regular presence of ridleysin summerand autumn is inferential, of course. The alternative is to believe the rid leys come here to die, or perhaps hibernate. Recordsfor Kemp'sridleys in theGulf of Maineregion are substantial, With eightsalvaged in 1978alone, the weight of evidence argueswell for thesewaters as a criticalhabitat for this desperatelyendangered species. It is now incumbentupon thosewho disagreewith my interpretationof the data to muster somecontrary evidence."Shoop 980! further reinforcedthe idea that New Englandwaters are potential developmental habitat by statingthat: "Eachyear several juvenileridleys washup on the shoreof CapeCod Bay,apparently drowned a ftcrbecoming too cold. Perhapsthey aretrapped in thegeographical hook of the Capeas they try to migratesouth ward. Our guessis that thesmall rid leys are pushednorthward in the Gulf Stream,from which they maketheir way to shallowbays, sounds, and estuaries where they feedon mussclsand other invertebratesand then move south,perhaps maturing in thc Gulf of Mexico and never returning to northern waters." ChesapeakeBay is a seasonalforaging area for juvenile loggerheads Caret ta caretta! and Kcmp's ridleys By1es, 1982;Lutcavagc and Musick, 1985!.Stranding and sighting reportsdocumented by Lutcavage981! indicatedthat both speciesare presentin the bay from April through Novemberand disappearwith the advent of cold weather. Byles982! observed:"The turtles in theChesapeake Bay and alongthe Virginia coastcannot survive the cold winter temperaturesand so must either migrate to warmer watersor overwinter submergedin the sediment.The latter methodof survival seemsunlikely due to penetrationof cold water -5' C! to the bottom over most of thc bay and nearshore shelf." A declinein turtle abundanceduring winter was apparentin the turtle fishery in WestFlorida Fritts etal., 1983!. It was suggestedthat the turtlesbury themselvesin themud and becomedormant in the winter Carr and Caldwell, 1956;Pritchard and Mdrquez, 1973!.Winter dormacy of Kcmp's ridlcys was evident when turtles were found with mud on their backs Ehrhart, 1980;Carr, Ogren and McVca, 1980!.Kemp's ridlcys have been found with loggerhead turtlesin an apparentstate of winter dormancyin waterof 11 'C in Florida Carr etal., 1980;Ogrcn and McVca,1982!. Dormantperiods in themud may bean adaptivere ponseto avoid cold shock,In water temperaturesof 10' C orless, ridlcys becomestunned and float immobileat the surface Ehrhart, 1980; Schwartz, 1978!. Ridley turtlesdied within 24hours during exposureto water temperaturesof 5-6'C Schwartz,1978!,No evidencehas been found of scaturtles wintering in Virginia waters Musick et ai., 1983;Lutcavagc and Musick, 1985!. ChesapeakeBay Kemp's ridlcys areall small irnmatures, most being between 30 and 45cm. Beginningin 1979,47 ridlcys werc documentedin ChesapeakeBay, 30 dead and 17alive, and nonebctwccn November and May Musick et at., 1983!.One was a headstarted turtle tag G0080!released a tHomosassa, Fla., on February 22,1979.It had traveled 2,277kilometers, moving at an averageof 5.3 kilometers pcr day, to Hampton Roads, Va. where it was captured alive after approximately14 months at large Musick etat., 1983!. Of the threeridleys reportedfrom the Indian River, Fla., Ehrhart,1983!, two werc only slightly shorterthan the

139 sma]lestadult females reported by Pritchard969!, and the third was well within thesize range of adults.According to Ehrhart983! it seemedunlikely that theseindividuals could have returned to RanchoNuevo to breedby the following spring,but a laterreturn to the natalbeach was neither suggested nor denied. Duringthe first WesternAtlantic Turtle Symposium in SanJose, Costa Rica, in 1983,the following comment was madeto thepanel of seaturtle experts by Dr.C. R. Shoop Marquez, 1983!: "Since most Kemp's ridley strandings in the U.S.are along the Atlantic Coast,does the panel actuallybelieve all of the animalsare waifs?The number of juvenilesalong the eastern U.S. is substantial, almost all are very healthy, and the observations have been made every year.Surely, some emphasis in researchon theseeastern U.S. animals is in order."The discussion panel's response Marquez,1983! was that: "The panel does not have data to reachconclusions on thequestion of Kemp'sridleys on theAtlantic coast. Are they, or arethey not waifs? Research is in order."A statementmade by Musickef al., 983! emphasized:"Since there is a greatdifference in habitatpreference and feeding habit betweenthe various sea turtles!, planningshould address the individual species rather than sea turtles as a generalcategory." Thenumber of headstarted Kemp's ridleys found on theeast coast of theUnited States is not surprisingwhen comparedto thedistribution shown by theliterature we reviewed and by recordsfrom the Sea Turtle Stranding and SalvageNetwork. Multiple recoveries of somehead started turtles raise a numberof questions.For instance, one head startedKemp's ridley tagG0045! was recovered on November 25, 1980, at CoreSound, N.C. after 642 days at large. Theturtle was alive and was released in good condition. It was recovered again at Core Sound 208 days later on June 21,1981,What happened to thisturtle during the winter of 1980-1981?Did it migrateto warmerwaters before winter, thenreturn to thesame area when waters warmed, or did it overwinterat Core Sound? Another tag G0104! was recoveredand released on three different occasions, thefirst at Miami, Fla,, on April 9, 1979, the second at Ocean City, Md.,on April 9, 1981,and the third at BradleyBeach, N.J. on June 4, 1981. Where was this turtle during the winters of 1979-1980and 1980-1981? A third head started turtle tagG0914! was recovered at Beaufort, N.C. on August 20, 1980,after 470 days at large. This turtle was recovered again 1,052 days later on July 9, 1983 at Hampstead Bay, N.C. Again,where was this turtle during the intervening four winters? It wasrecovered twice in thesame general area and dates1,052 days apart! The turtles may migrate to warmerareas when colder weather begins to occurin thefall and thenreturn in thespring, or theycould conceivably burrow into the mud of deep bays and sounds to overwinter It is important to conservationof Kemp'sridley to determinewhether one, the other,or both occur. Thehead started turtle tagNNN893! of the1979 yearwlass recovered in Biarritz,France on December 25, 1981, had beenat large 568 days, for one winter and a portionof another. This turtle was small .0 kg!and cold-stunned when found.However, the mos t interestingof all recoveriesoccurred at ElJadida, Morocco on November 20, 1982 a terf the turtlehad been at largefor 893days. This turtle tag codeNNN678! had survived two winters,was alive and in excellentcondition when captured in a fisherman'snetin a lagoon,and was reported to haveweighed 20 kg Unfortunately,the flipper tag was removed and sent with therecovery information to Miami, Anotherhead started turtle from the 1979 year-class, held in captivity atCayman Turtle Farm 983!, Ltd laideggs therein thespring of 1984.This turtle also weighed 20 kg at thetime Woodand Wood, 1984!, The Rancho Nuevo nestingbeach is locatedat 23 11'Nlatitude and 97'46'W longitude. This latitude strikes the northwestern coast of Africajust south of Golfo de Cintra, Spanish Sahara Figure 5!, approximately 1,127 kilometers south of the lagoon wherethe head started turtle tag NNN678! was recovered. Carr 957! hypothesized anOld World colony of "five- scaled"ridleys implying Kemp's ridley! on thenorthern coast of WestAfrica, and made a flightalong the uninhabitatedcoast from Mauretania to Morocco in search of Kemp'sridleys, but to noavail. In hiscontinuing commentaryonKemp's ridley nesting beaches during the1950s, Carr 957! observed that although this species was apparentlyabsent from the Caribbean, there is some unexplored territory there, and it stillseems possible that "part of theAmerican Atlantic ridley population could be derived from WestAfrica eitherfrom somewherewithin the rangeof the form that lives about and below the bulge, and in the Gulf of Guinea and which has predominantly six laterallaminae, orfrom the long extent of shore from Port Etienne toMorocco, where nothing is known of the turtle fauna." Wesuggest that thearea from Port Etienne toMorroco should bereinvestigated asapossible nesting site for Kemp's ridley.Not only is thisarea near the same latitude as that of RanchoNuevo, Mexico, but many of thenesting prerequisiteslisted by Pritchard 969! for Kemp's ridley occur there; e.g,, numerous, shallow, warm water lagoons. Similarhabitat exists along the northern coast of WestAfrica, and the numerous islands of theArquipelago de Madeiraand the Islas Canarias could provide sanctuary for smallturtles. It isclear that young juvenile Kemp's ridley sea turtles can and do exit the Gulf of Mexico and escape into the Atlantic.There they possibly enter the Gulf Stream where they can become entrapped ineddies that spin off from the maincurrent toward estuarine areas along the east coast of the United States Carr, 1986!. Our recovery data for head startedKemp's ridleys, the published distribution records, and strandings reported by the Sea Turtle Stranding and SalvageNetwork all showthat Kemp's ridley is common to theeast coast of theUnited States. We do notbelieve that SINGLE RECOVERY 0 979 YEAR-CLASS

Figure5, Thenorthwest coast of Africashowing the northward location ltatitude33 4' lV! of therecovery of a 1979year-class head started Kemp'sridley sea turtle as compared tolocations of theRancho Nuevo nesting beach latitude 23 '11' ¹, Tamaulipas,lVlezico, and the Cayman Turtle Farm 983! Ltd, latitude19 3'N!, Grand CaymanIsland, B.WI,

'141 smallerKemp's ridleys found on theAtlantic coast are '1ost waifs," but thatthese east coast turtles represent an integralpart of the life cycle of this species. Further, from previous records and ours we know that Kemp's ridley occursin Europeanwaters, and one of our headstarted turtles has been reported from thenorthern coast of West Africaat a reportedsize comparable toone that matured and nested in captivity. Therefore, it is tempting to resurrect Carr's955! speculationof an Old World nestingsite on the northerncoast of WestAfrica, Thesearch should continue. Acknowledgements Thiswork was conducted under permits from Mexico's INP, the FWS and the Convention on InternationalTrade in EndangeredSpecies ofWild Fauna and Flora CITES!. It would not have been possible without the diligent and conscientiousefforts of a numberof agencies,organizations and individuals. Primary among these werc: Rend Marqucz,INP, Mexico; Jack Woody and Dave Bowman, FWS, Albuquerque, N,M.; Milford Retcher, Jim Wood and BobKing, NPS, Santa Fe, N.M. and Padre Island National Seashore, Corpus Christi, Tex.; Edward Klirna, Charles Caillouct,Jim McVey and Ray Wheeler, NMFS, Galveston, Tex.; Carolc Allen, HEART Help Endangered Animals - RidlcyTurtles!, Houston, Tex.; Herb Kumpf and Fred Berry, NMFS, Miami, Fla.; Pat Burchfield, Gladys Porter Zoo, Brownsville,Tex.; Anthony Amos, The University ofTexas, Port Aransa s;Peter Pritchard, Sea Turtle Recovery Team, FloridaAudubon Society, Maitland, Fla,; and Larry Ogren, NMFS, Panama City, Fla. Datamanagement onrecovcrics ofhead started turtles was possible through the cooperative efforts of Bill Pringle andBarbara Schrocder, N1VIFS, Miami, Fla. and Dennis Koi, NMFS, Galveston, Tex. Many other individuals and organizations,too numerous to mention,have contributed to thesuccess of the head start project at Galveston. To all of them, wc owe a heart-felt and sincere thanks. MartyTyrce and Dickie Rcvera assisted in data preparation and review of themanuscript. The figures were preparedby DannyPatlan and the original manuscript was typed by BeatriceRichardson, Literature Cited Anonymous.1978,U.S. Mexico restoration efforts may be only hope for Kem p's rid Icy. U.S. Fish and Wild life Service, Endangered SpeciesTechnical Bulletin 30!:6-8. 81eakncy,S.1965. Reports of marineturtles from New England and eastern Canada. Canadian Field-¹turalists 79!:120-128, Brongersma,L.D.1972. European Atlantic turtles. Zoologische Verhandelingen, Rijksmuseum vanNatuurlijke Historic teLeiden No. 121, 318 p. Brongersma,L D.1982. Marine turtles of the eastern Atlantic Ocean, p.407 416. In: Bjorndal, K,A. Editor!,Biology and Conseroatton ofSea Turtles, Smithsonian Institution Press, Washington, D.C., 583 p. Byles,R.1982. Radio-tracking ofa Kemp's ridIey off the Virginia Coast. Report submitted tothe U.S. Fish and Wildlife Service, Endangered SpeciesOffice, March, l982, Caillouet,Charles W.,Jr. 1984. Essai deprevention deI'extinction dela tortue de Kemp. LesCarnets deZoologic Bulletin ofthe ZoologicalSociety of Quebec!44!:28-34. Caillouet,C.W,, Jr.and D.B. Koi, 1985. Patterns andvariability infirst-year growth inweight ofcaptive-reared Kemp'sridley sea turtles:a graphical analysis. NOAA Technical Memorandum hlMFS-SEFC-164, i plus4 p,and 52 figures. Caillouet,C W., Jr.,D 8.Koi, C T.Fontaine, T D.Williams, W BrowningJ. andR M.Harris. 1986. Growth and survival ofKernp's ridley sca turtles, Lepidochleys kernpi,in captivity. lVOAA Technical Memorandum NMFS-SEFC-786, i Ii plus34 p., 12 Tables and 7 Figures. Carr,A. 1955.The riddle of therid Icy. Animal Kingdom 58!:146-156. Carr,A. 1957. Notes on the zoogeography ofthe Atlantic sea turtles ofthe genus Lepidochelys. Reoistade Biologia Tropical 5!:45- 6I. Carr, A,1963. Panspecific reproductive convergence inIepidochelys kempi. Ergebnisse derBiologic 26:298-303. Carr, A.1967. So excellent a fishe; a natural history ofsea turtles. The NaturaI History Press, Garden City, NY. Carr, A. 1975.The Ascension Island green turtle colony. Copeia 1975;547-555. Carr, A. 1980.Some problems of seaturtle ecology. American Zoologist 20:489-498, Carr, A.1986. Ncw perspectives onthe pelagic stage of sea turtle development. NOAA Technical Memorandum MvIFS-SEFC-190, 36 p. Carr, A,F. and D.K. Caid wcII, 1956.The ecology and migrations ofsea turtles 1. Results offield work in Florida, 1955. American Museum IVooitates 1973:1-23. Carr, A,F,and D,K. Caldwell. 1958. The problem ofthc Atlantic ridley turtle Lepidochelys kempi!in1958. Revista deBiologia Tropical 6!:245-262. Carr, A.F.,L. Ogren and C. McVea. 1980, Apparent hibernation bythe Atlantic Ioggcrhcad turtle, Caretta caretta! offCape

142 Canaveral, Florida. BiologicalConservation 19:7-'14. Chavez,H. '1968,Tagging and recapture of the loraturtle Lepidochelyskern pi!. InternationalTvrtleand Tortoise Society Journal 3! 32- 36. Dodge,E.S. 1944.Studies of the ridley turtle in Massachusettswaters. Copeia 1944!:120-121. Ehrhart,LM. 1980.Threatened and endangered species of the KennedySpace Center: marine turtle studies.h1ASA Contract 163122, KSC TR 51-2, VolumeIV, Part 1, 407 p. Ehrhart, L.M, 1983.Marine turtles of the Indian River lagoon system. Florida Academyof ScienceSymposium 46/4!:337-346. Fontaine,C.T. and C.W. Caillouet, Jr, 1985.The Kcmp's ridley seaturtle headstart research project: an annual report for fiscalyear 1984.!VOAA TechnicalMemorandum NMFS-SEFC-152, ii plus 13 p, and 3 Tables. Fontaine, C.T., K.T. Marvin, T.D. Williams, W J.Browning, R.M. Harris, K.L. lndelicato, G.A, Shattuck, and R.A. Sadler. 1985.The husbandryof hatchlingto yearlingKemp's ridley sea turtles Lepidochelyskempi!. NOAA Technical Memorandum IsMFS- SEFC-15B,ix plus 34 p., 10Tables, 22 Figures, and 2 Appendices. Fontaine, C T., T D. WIlliams, S A. Manzella and CW, Caillouet, Jr. 1989.Kem p's rid ley seaturtle head start operations of the NMFS SEFCGalveston Laboratory. This volume.! Fritts T.H., A.B. Irvine, R.D,Jeenings, L.A. Collum, W. Hoffmanand M.A. McGehee.1983. Turtles, birds, and mammalsin the northernGulf of Mexicoand nearby Atlantic waters. U.S. Department of theInterior, Fish and Wildlife Service,Washington, D.C., FWS/OBS42/65, 455 p, Henwood, T A. 1986.Losses of monel flipper tags from loggerhead seaturtles Caretta caretta!. JournaI ofHerpetology 20! 276 279. Hildebrand, H.H. 1982,A historical review of the status of sea turtle populations in the western Gulf of Mexico, p, 447-453, In; Bjorndal,K. A. Editor!, Biologyand Conservation of SeaTurtles, Smithsonian Institution press,Washington, D.C., 583 p. Hughes,C R.1974. The seaturtles of SouthAfrica II. Biologyof theTongaland loggerhead turtle Caret ta caretla L. with cornrnents on the leatherback turtle Dermochelyscoriacea L. and the green turtle Cheloniamydas L. in the study region. Oceanographic ResearchInstitute, Durban, South Africa, InvestigationalReport 36:1-96. Klirna,E.F. and J.P. McVey. 1982. Headstarting the Kemp'sridlcy turtle Lepidochelyskempi!, p. 481-487.In: Bjorndal,K.A. Editor!, Biologyarui Conservationof SeaTurtles, Smithsonian Institution Press,Washington, D.C., 583 p. Lazell, J.D. 1976.This broken archipelago. A Demeter Press 13ook,New York, 260 p, Lazell,J.D. 1979. Boreal and temperatemigratory regimesof Atlantic marineturtles. AmericanZoologist 19;954, Lazcl!, J.D. 1980. New England waters: critical habitat for marine turtles. Copeia1980!:290-295. Lutcavage, M. 1981.The status of marine turtles in ChesapeakeBay and Virginia coastalwaters. Master of ScienceThesis, Virginia Institute of Marine Science,Gloucester Point, VA. 126 p. Lutcavage, M. and J.A. Musick. 1985, Aspects of the biology of seaturtles in Virginia. Copeia1985!:449-456. Manzella, S.A., C.W. Caillouet, Jr. and C.T. Fontaine. 1988.Kernp's ridley, Lepidochelyskempi, sea turtle head start tag recoveries: distribution, habitat and method of recovery. Marine FisheriesReview 50!:33-42. Marquez M., R. 1983. Kcmp's ridley turtle overview of biology, p, 96-104.In: l3acon,P., F. Berry, K. Bjorndal, H, Hirth, L. Ogrcn and M. Weber Editors!, Proceedingsof the WesternAtlantic Turtle Symposium,Vol. 1, 306 p. McVey, J. P. and T. Wibbels. 1984. The growth and movements of captive-reared Kemp's ridley seaturtles, Lepidochelyskempi, following their release in the Gulf of Mexico. 1VOAATechnical Memorandum NMFS-SEFC-145, 25 p. plus 3 Figures and 3 Tables. McVey, J.P.,J.K. Leong, R.S.Wheeler and R.M. Harris manuscript!. Thc culture of young Kcmp's ridlcy sca turtles Lepidochelys kempi!.National Marine FisheriesService, Southeast Fisheries Center, Galveston Laboratory, Galveston, Texas unpublished manuscript!. Mowbrag, L.S. and D.K. Caldwell. 1958. First record of the ridlcy turtle from Bermuda, with notes on other sea turtles and the turtle fishery in the islands, Copeia1958!:147-148. Musick,J.A. 1979. The rnarineturtles of Virginia: Families Chelonidae and : with notes on identification and natural history. Virginia Institute of Marine Science,Sea Grant EducationalSeries IVo. 24, 17 p. Mu sick, J,A,, R. Byles, R, Klinger and S, Bellmund. 1983, Mortality and behavior of seaturtles in ChesapeakeBay, Summary Report 1979-1983,Contract NA80FAC00004,Virginia Institute of lvlarine Science,School of Marine Science,College of William and Mary, 52 p. Ogrcn, L. and C. McVea, Jr. 1982. Apparent hibernation by seaturt!es in North American waters. In: Bjornda1,K.A. Editor! Biology and Conservationof SeaTurtles, Smithsonian Institution Press,Washington, D.C., 583 p. Pritchard, P C H. '1969.Studies of the systematicsand reproductive cyclesof the genus Lepidochelys.PhD. dissertation, University of Florida, Gainesville, FL, 197 p, Pritchard, P,C.H, and R, Marquez M, 1973,Kemp's ridley turtle or Atlantic ridley, Lepidochelyskempi. International Union for the Conservation of Nature and Natural Resources,Monograph No. 2, Marine Turtle Series,30 p. Pritchard, P.C.H., P.R. Bacon, F.H. Berry, A.F. Carr, J. Fletemeyer, R.M, Gallagher, S,R. Hopkins, R.R. Lankford, R. Marcluez M., L.H. Ogren, W G. Pringle, Jr., H.A. Reichart and R. Witham. 1983. Ivlanual of seaturtle researchand conservation techniques.

143 SecondEdition. Bjorndal, K.A. and G.H. Balazs Editors!, Center for Environtnental Education Washington, D.C., 126 p. Schwartz,F.G. 1978. 8ehavioral and tolerance responses tocold water temperatures bythree species ofsea turtles in North Carolina,p.16-1 8.ln: Henderson, G.E. Editor!, Proceedings ofthe Rorida Interregional Conference onSea Turtles, Florida Marine ResearchPubh'cation No. 33. Shoop,C.R, 1980. Sea turtles inthe northeast Maritimes. University ofRhode Island, Graduate School ofOceanography, p.9-11. Wauer,R. 1978."Head start" for an endangeredturtle. NationalParks Conservation 52;16-20. Wibbels,T, 1983. A trans-Atlanticmovemcnt ofa headstarted Kemp's ridley. Marine Turtle Newsletter No.24, p, 15-16. Witham,P.R. 1976. Evidence forocean-current mediated dispersal inyoung green turtles, Chelonia ntydas Linnaeus!. Master of ScienceThesis, University of Oklahoma, Norman, OK, 26 p. Witham,R. 1980.The "1ost year" question in youngsea turtles. American Zoologist 20:525-530. Witharn,R.and C.R. Futch. 1977. Early growth and oceanic survival ofpcn-reared seaturtles. Herpetologica 33!:404-409. Wood,J.and F. Wood. 1984. Nesting ofa captiveKemp's ridley sea turtle. Marine Turtle Newsletter No.30, p. 12,

144

DermatoglyphicPatterns on Kemp'sRidley SeaTurtle Flippers: CanThey be Usedto Identify Individuals? CharlesW. Caillouet,Jr., Dickie B. Revera,Marcel J. Duronslet and John Brucks '

Derrnatoglyphicpatternsencompass theconfigurations formedby dermal ridges onthe palms, fingers, toes and soles ofall primates.Fingerprints arethe most widely knoum example, Asapplied tosea turtles, dermatoglyphic patterns arethose formed byscales andintervening spaces between themon the tops and bottoms offront and rear flippers. These patterns canbe replicated, examinedand compared as flipper-prints. impressions negatives! offlippers of50yearling Kemp's ridley sea turtles Lepidochelys kempi! of the1984 year-class were madein moist clay, then reproductions

Introduction Dermatoglyphicsrefersto patterns ofthe skin. The word is derived from the greek words "derma," meaning skin, and"glyphe," meaning carved work. It encompassestheconfigurations formed by dermal ridges of thepalms, fingers,toes and soles of all primates Ausherman etal., 1973!. Fingerprints arethe most widely used and familiar examples.As applied to sea turtles, dermatoglyphics refers to flipper-prints. In hopesof developing a lifetime tag for individual Kemp's ridley sea turtles Lepidochelys kempi! we have been investigatingtheuse of dermatoglyphic patterns formed by scales and intervening spaces on the tops and bottoms offront and rear flippers. Replicas plastic casts, photographs andcomputer-digitized images! offlipper derma- toglyphicsprovide permanent records that can be examined in detail and compared. Our paper describes the preparationof such replicas and our plans for theirexamination and comparison. Materials and Methods Clayimpressions weremade offlippers of50 Kemp's ridley yearlings ofthe 1984 year-class. Theywere among the sameturtles designated fortransfer tomarine aquaria and the Cayman Turtle Farm 983! Ltd., for extended head startingand captive propagation Caillouet etal., 1986!. Therefore, captive survivors ofthis group are available for flipper-printingagain in thefuture. The clay was a mixtureofVelvetex and Marblex. Velvetex isa moist,white, smooth-bodiedmodeling clay containing nogrog. I tis formula ted by V. R. Hood, Inc., San Antonio, Tex. for Houston Artsand Crafts, Inc,, Houston, Tex.Marblex isa self-hardeningclaymanufactured byAmerican ArtClay Co Inc., Indianapolis,Ind.The mixture consisted of three parts Velvetex to one part Marblex blended to a smooth, homogeneousconsistency. Blending was facilitated andproper consistency achieved bythe addiflon ofapproxi- mately 130 mm of warm tap water for each kilogram ofclay used. A precisequantity ofwater needed cannot begiven, becausetheamount depends onthe moisture content present inthe individual clays atthe time of mixing. We used a Hobartmodel A/200 mixer to blend the Velvetex and Marblex clays. Mixing can be done by hand, but it isdifficult andtiring. Once mixed, the clay can be used over and over ifkept moist. The clay mixture was stored ina plasticbucket witha layerofplastic film pressed over its surface and with the lid of the bucket firmly secured asan additional safeguardagainst loss of moisture. Ina vaportight container, theclay will remain workable forsix to eight weeks, Claywas divided into 550-600 gram portions, each formed into a ball.Care was taken towork out any trapped air bubbles.Forming the clay into balls was facilitated by having a containerofwater available sothat hands could be keptwet during the forming process, thenin the final step the ball could be smoothed ontoa shinywet surface. The ballwas placed onthe shiny side of a 25x 25 cm square ofheavy freezer wrapping! paper, and another square ofthe samesize was placed ontop of the ball, shiny side against thesurface ofthe ball. The ball was then flat ted by firm but gentlepounding with the bottom ofa 25cm cast-iron skillet Figure 1!.Any heavy, smooth-surfaced objecthaving a diameter5 cm greater than the desired diameter ofthe clay disk to be produced may be used,

' National Marine Fisheries Service

146 Thefla ttening process produced a claydisk of uniform thickness,sandwiched between two layersof freezerpaper. A 550-600g clayball producesa diskapproximately I8 cm in diameterand 7 to 9 mm thick. The freezerpaper was left in place,Two clay diskswere stacked on top of oneanother and slipped into a 38 litercapacity, self-sealing, plastic freezer!bag Figure2!, Thebags were stacked in plasticboxes with tight fitting lids. Thedisks maybe storedin this mannerin an air-conditionedroom for a weekor in a refrigeratorfor two to threeweeks before use, The diskswere allowed to come to room temperature before use. Impressionsof thedorsal and ventral sides of theleft front flipper weremade in clay,To makea flipperimpression, theclay should be firm but notexcessively sticky, Proper consistency maybe tested by forminga smooth,marble-sized ball, then pressingit gently with a clean,dry fingertip. The clay should not stick to the fingertip and the resulting flipperprint shouldbe well defined.At the time a flipperimpression was made on a claydisk, the top sheetof freezer paper was removed,and the clay surfacewas smoothed with a spatula Figure3!. The smoothedsurface was then sprinkledwith pure talcumpowder as a moldrelease Figure 4!, The powder was smoothed with clean,dry fingertips so that the surfacewas evenly covered with a thin film of powder,The flipper from which the impressionwas to be madewas wiped cleanand gentlyblotted dry. A very light coatingof talcumpowder wasapplied to the flipper this was washed off immediately after the impression was made!. The code from the monel flipper tag, which uniquely identified the turtle, was written on a squareof paper thatremained attached to thebottom of the clay disk. The turtle wassupported by oneperson who alsoheld a foldedpaper towel againstthe sideof the headadjacent to the flipper being impressedto prevent possiblebiting while the impressionwas being made.The impressionwas madeby another person who positioned the flipper on the surface of the clay and applied gentle, even, downward pressure with the fingertips Figure 5!. The thicker anterior edge was pressed down first and the fingertips were carefully worked back to the thin posterior edge to avoid distortion of the impression. The soft consistency of the clay allowed the flipper to be pressedinto the disk without applicationof undue pressureon the flipper. The personsupporting the turtle then lifted the animal so that the flipper was lifted vertically from the newly formed mold to avoid smearing or distorting the impression. The designation of the side V=ventral, D=dorsal! of the flipper so molded was written next to the impression negatives!on the freezerpaper extending beyond the perimeterof the disk Figure6!. Theclay disks containingthe impressions negatives! were placedindividually in plasticboxes with tight-fitting lids, and the junction of the lid and box was sealedwith masking tape. Impressions that were tobe cast within 36 hours were stored at room temperature. It is possible to store clay impressions of flippers for 7 to 10 days if they are refrigerated,However, care must be takento preventmoisture condensate from dripping fiom the box lid onto the impressions during refrigerated storage. We did not experiment v. ith casting materials. Instead, a local representative, Ms. Deanna Hunter, of Patty Cakes, Inc,, made casts positives! from the flipper impressions. The Patty Cakes process is patented by Ms. Irene Theis of Denver,Col. It producesa plasticcast of extremelyfine detail without distortion of dermalconfigurations Figure8!, Before the cast was removed from the clay mold, the information written on the attached paper square was duplicated on the back side of the plastic relica, This was accomplished by using a heat-penof the type used in wood or leather burning. A vibrating engraving tool also works well, The hardened plastic casts were removed from the mold and excesspowder was removed from the clay surface with a wet towel. The paper square was removed from the back and the used disks were stored in a plastic bucket to be reworked as needed. The finished casts were stored in envelopesbearing the tag code, flipper and surface designations and da te of the impression. They were filed in order by tag code. Comparison of Flipper-prints Initial inspection of the plastic models revealed easily discernabledifferences in the dermal patterns Figure 8!. Both ventral and dorsal aspectsappeared equally suitablefor comparison,Since it was obvious that obtaining flipper- prints from large animals and from animals in the field would ultimately require dcvclopmcnt of techniques for obtaining sharply-defined scale patterns in black and white photographs, it was decided to use the dorsal aspect of the flipper. Photographing the dorsal side would involve the least amount of effort for the photographer and, more importantly, the least amount of disturbance for the turtle. It was difficult to obtain photographs of the plastic models with sufficient contrast to define the dermal con- figurations. This difficulty was overcome by using a size4X0 Rapidograph pen to ink the indentations outlining the scales.The enhanced plastic models Figure 8! were reproduced on acetatetransparencies using an IBM photocopier Model 20!, The transparencies were superimposed on an overhead projector and projected onto a screen. This procedure allowed rapid visual comparison for differences in size, shape and arrangement of the dermal configura- tions, Initially, 10 prints were compared.Each print appearedto be unique to the individual seaturtle. Figure9 isa photograph of a computer-digitized image of a flipperwast. Futureefforts should be directedtoward finding a castingplastic of equalquality and reproductiveintegrity to that

147 Figurei. Flatteninga clayball with a cask-ironskillet.

Figure4. Applicationoftalcum powder fothe smooth surface ofa clay disk powderon disk and flipper surface seroes as a moldrelease!.

148 Figure5. Matcingan impression of the ventral surface of a left foreflipper.Figure 6,impressions ofventralanddorsal surjacesofthe leftforeflipper of a Kemp'sridley sea turtle flippertag code NNT-140!,

Figuretagcode7. and Reverseletterside of D=Dorsalplastic Vmodels, ¹te flipper =Ventral!.Figure 8.Plastic model positive!ofthedorsalaspectleft foreflipper.Dermal configu-

rationshave been outlined, using a 4XO Rapidographpen, to enhancecontrast for photographyor computer-digitization.

Figure9. Photoof an unenhancedplastic model of a turtleflipperasseen bythe NASA!NSTLIERL EIKONIX Dt'gitizer.Plastic model was made from clay impressionof a foreflipperof a hatchling that was deadon arrivalat theNMFS Galvestonfacilityaftertransportfrom Padrelsland National Seashore age at time of deathvms three days post-hatching!,

149 ofPatty Cakes, Inc. Casts should be lightweight, durable and non-brittle, andthey must harden inmoist clay but not shrinkor expandduring hardening. Futurecomparisons between flipper-prints of anindividual sea turtle taken at differenttimes must take into co~siderationwhether growth in theflippers is isometric orallometric, Milford Fletcher's National Park Service, personalcomrnunication!preliminaryobservationssuggest thatgrowthis allometric. Allometricgrowth willrequire mathematical orgraphical! transformations offlipper images prior to comparisons among prints of flippers of different sizes.

Literature Cited Ausherman,D.A., W.D, Hall, R.E. Moyers, R.C. Fairchild and R.H. Mitchel. 1973, A proposedmethod for the analysis of dermatoglyphicpatterns, p.171-180. in;Herron Editor!, Proceedings ofthe Society ofPhoto-optica/ Instrumentation Engineers Meetingin Quantitative imagery in the Hiornedicai Sciences ii, Volume 40, 229 p. Caillouet,C.W., Jr.,C.T. Fontaine, T.D.Williams, S.A,Manzella, D.B.Revcra, D,B.Koi, K.L. W. Indelicato, M.G,Tyree, J.K. Leong, M.J.Duronslet andK. T. Marvin. 1986. The Kemp's ridley sea turtle head start research project: anannual report for fiscal year1985. NOAA Technical Memorandum NMPS-SEFC-174, 37p., 29 Tables, 1 Figurc and 5 AppendixTables.

150 Questions and Answers

Sally Murphy: Whatkind of verificationsdo you havethat thetransmitter is still attachedto theturtle andis notfloating or beached? Mysing: At present,the only informationthat would be indicativeis movement.The rate of movementof the tagand the amount of surfacing would suggest that the tag is still on the turtle. When the satellite tag was beached, I did not want to believeit, but I knew immediatelybecause I got about 21messages in oneday. Only five messagesper day is more 1ike what one would expect. If the turtle of water, Clark Fontaine:On thetwo turtlesyou numberedone and two, you indicated that they overwintered. Do you know that they overwi nterd? Ogren: No, such data are like your recapture data. We assume the turtles stayed in the Atlantic. We do not suppose they went aroundhere referringto a map!.We showthese kinds of movementshere. It might be a regularthing, but we have no way of knowing for sure what other routes they are traveling. Fontaine: Of the turtles that go out into the North Atlantic, are theylost to thegeneral population in the Gulf of Mexico,never fo return? Ogren: I thinkboth happen. How was I going to state that in my summary here?We have two things going on. Those turtles that end up at Cape Cod Bay are washed ashore cold-stunned, or dead. The onesin Northern Europe arc dead, cold-stunned. Those turt1esthat get caught up in the upper latitudes in wintertime, the ones that do not get stopped in time I believe they are gone. These turtles that get in the Atlantic here and stay in this area manage to survivcby migrating out of the coastal area and into the warmer deeper water moving south. And over on the eastern Atlantic, those turtles that go into the Canary current system and continue on downstream of the Gulf Stream gct around the Azores, Thc loggerheads are out there living well, and there is no doubt that somehow, constantly, some of those turtles will survive also. We do not know how they get back, or which route they take back to the western Atlantic. I think we have a percentageof mortality, called waifs or whatever you want, that also have a percentageof survivors.

151 Captive-RearedKemp's Ridley SeaTurtle Data Base Management Dennis B. Koi'

Effectivecareof Kemp's ridley sea turtles Lepidochelys kempi! held in captivityforhead starting depends in part on collectionandmaintenance ofcomplete andaccurate records covering many aspects ofthe turtles' lt'ves during that period. Growth,environmental andother types ofdata were collected andarchived since the inception offhe head start project. The methodsofcollection ofdata varied with less emphasis placed ondata collection inthe early years ofhead starting compared to recentyears. This was due to emphasis in early years on learning about husbandry ofKemp's ridIcy. Withincreasingopportunities forresearch inrecentyears, datamanagement bycomputer hasplayed anessential role,and fhe efficiencyofcollection, management andsubsequent analyses ofdata has improved dramatically. However, dueto the dynamic needsofresearchers, nocomplete system could be developed tomanage alldata aspects. Themonthly system currently inuse tracks each t urt I e'sstat us, selects random samples forwei ghings, and manages thedata, Frequentreview and modification ofcomputer-based datamanagement systemsare required until such time as the raising ofsea turtles in captivity becomesroutine. Involvedinthe efforts tosave Kemp's ridley sea turtles from extinction arevarious federal and state agencies, andprivate organizations,eachwith its own tasks and data needs. With few exceptions, regular and routine communication andshari ng of dataamong these entities make the best and widest possible useof the information anddata. As these entities' data needs are different,their needs forexchanging datamay vary. The advantages anddisadvantages ofcentralized versus distributed data management were discussed.

'hIationat Marine Fisheries Service

l52 Marine Turtle Data Base Management: Nationa1 Marine Fisheries Service - Miami Laboratory Barbara A. Schroeder'

The Miami Laboratoryof the National Marine FisheriesService NMFS!, SoutheastFisheries Center, is responsiblefor maintainingand managing several large data bases thaf contai n informationon marine turtles. Data applicable to Kemp'sridley turtle Lepidochelyskempi! arefound in threedafa bases: ! stranding;! taggirtg release and recapture!; and ! Western AtlanticTurtle Symposium WATS!. The stranding data base contains records of strandedsea turtles reported through fhe Sea Turfle Strandingand SalvageNetwork, The tagging data base contains release and recapturerecords of seaturtles tagged primarilywith tagssupplied by NMFS.Data included in the WATSdata base were collected from countriesthroughout the Caribbeanbasin and summarized for the WATSland WATSIl symposiaheld in 1983and 1987.Keeping these data bases currenf andmaintaining t heirvalidityis largely dependent upon receiving timely and accurate recordsfrom the many sources that provide data. Statusof the data basesis discussed,

Sea turtle data bases created, maintained and managed within the Miami Laboratory of the National Marine FisheriesService NMFS!, SoutheastFisheries Center SEFC!were established to compile and make available data on all speciesof seaturtles. Availability, statusand potentialuses of thesedata are not limited to onespecies. Hence, it is difficult to address Kemp's ridley turtle Lepidoche!yskempi! specifically when discussing the data bases. For this reason the data basesare discussed herein in general terms, and specific referencesto Kemp's ridley are made when appropriate. Sea Turtle Stranding and Salvage Network The SeaTurtle Stranding and SalvageNetwork STSSN!was informally organized at a specialsession of the World Conference on SeaTurtle Conservation held in November 1979 in Washington, D.C. A coordination and planning meeting was held during the Symposium on Behavioral and Reproductive Biology of SeaTurtles at the American Society of Zoologists meeting held in Tampa, Fla., in December 1979. The STSSN was designed to function as a cooperative effort among the U,S. Fish and Wildlife Service FWS!, the NMFS and state coordinators from various public and private agenciesand universities. The primary purpose for establishing the STSSNwas to coordina te the collection and centralized management of data on seaturtles stranded along the Atlantic and Gulf of Mexico coastsof the United States.A secondary purpose was to compile and periodically distribute summary reports of reported strandings, thus making the information available to researchersand managers upon request, Prior to theestablishment of theSTSSN, stranding records were compiled independently in somestates, so limited data records are available in the data basefor strandings during the late 1970s.In terms of more comprehensive data, 1980was the first functional year for the STSSN,From inception of the network in January 1980until August 1982, the NMFS SEFCwas responsible for data compilation and data basemanagement. In August 1982,under contract to NMFS, Dr.D. K. Odell, University of Miami, RosenstielSchool of Marine and Atmospheric Science RSMAS! took over these responsibilities. Dr. Odell continued to function as the STSSN coordinator, although NMFS discontinued contract funding of the network. In 1985,NMFS again assumed responsibility for STSSNmanagement including all data compilation. The basic flow of information, from stranding event to data requests by users, is depicted in Figure 1. Upon notification of a sea turtle stranding, the network volunteer closest to the stranding is responsible for locating and documenting it. Documentation includes accurately recording all pertinent data on a standardized stranding data report form and disposing or permanent marking of the carcass.All or part of the specimen may be salvaged for further study, if appropriate. Live turtles are handled in a mannerappropriate to the situation, but this should include an attemptto rehabilitatethe turtle for future release,Completed report formsare submittedto theappropriate state coordinatorwho verifiesthe data and periodicallyforwards the report alongwith an accumulationof similar reports to the network coordinator at the Miami Laboratory.

' Florida Departmentof Natural Resources

153 Numerousversions of theSTSSN data rcport form have been used during the years the network has been in existencc,and this has complicated standardization ofdata collection. In1985, wi th input from the state coordinators andother interested parties, I revised the da ta report form based primarily on thc format utilized in Florida and North Carolina.Acceptance andu tilizationoi'this standardized stranding report form grea tly improved thequality ofthe strandingdata. Information collected oneach stranding report includes: observer, stranding date, state/county of occurrence,species, reliability ofspecies identification, sex,sex determination method, stranding location including latitudeand longitude, carapace length and width measurements, condition ofturtle, tag information ifpresent!, disposition of carcassand remarks. Asstranding reports are received at theMiami Laboratory, they are edited, coded and entered into the STSSN data base,and the original reports are filed for future reference, Quarterly stranding summaries, including anannual report,are compiled bythe STSSN coordinator. These reports are distributed toSTSSN participants andother interested parties. TheSTSSN database ismaintained ina BASEd IIIPlusr" format onmicrocomputer. Editingofthe data base began withthe most recent years, progressing backin tune. This back editing includes theassignment ofcounty and remark codestoindividual records, Additionally, statistical zones shrimp statistical subareas! asdefined bythe current TrawlingEfficiency Device TED! regulations have been assigned toeach stranding record. A timeseries ofsea turtle stranding records for the U,S. Atlantic and Culf of Mexico coasts provides a valuable sourceofinformation that offers many realized and potenhal uses. However, certain limitations must be considered whenthc stranding dataare analyzed andinterpreted. Changes infrequency andrange ofbeach survey coverage over timearc difficult to quantify. Coverage ranges from systematic weekly or morc frequent! sampling insome areas to infrequentorno monitoring inother areas. Thus, data contained in the STSSN data base are considered minimum strandingfigures, asthey are reported strandings only,not all stranding events. Additionally, therelationship betweenreported strandings andactual at-sea mortality hasnot been quantified, Preliminary information suggests thata significantpercentage ofthe coastal marine turtle mortality may not be documented bythe STSSN Ulrich,1978; S.Murphy, South Carolina Wildlife and Marine Resources Department, personal communication!, Nevertheless,thedata collected enable determination oftrends based onreported strandings, andprovide a wide varietyofbiological informa tionabout thetur ties them selves. Significant changes infrequency ofreported strand ings oridentification ofareas of high numbers ofstrandings may indicate problems in need of attention, Measurementsofstranded turtles provide information upon which size frequencies canbe compiled, butsuch frequencicsshouldbe interpreted inthe context thatsome mortalities maybe size-selective, thusbiasing size distributiondataderived from reported strandings. Strandings doindicate presence ofturtles inspecific areas, and suchinformation maybc useful inassessing impacts ofman's activities inthe coastal zone as well as natural causes of scaturtIe mortality. Strandingreports may also include information ontagged turtles that become stranded, providing anadditional sourcefordocumenting recaptures, Collection ofsea turtle carcasses orparts! bynetwork volunteers provides specimensforresearch e.g.,necropsies, stomach contents, museum collections, age-growth studies!, andinformation onthc incidence ofsuch things aspropeller wounds, entanglement infishing gear, ingestion ofdebris ortar, external foulingby oil or tar,and fibropapillomas or otherabnormalities. Itis hoped that the continuity ofthe stranding database can be maintained through continued cooperation and participationofthe coastal states. Bothgeneral andspecific summaries ofstranding datacan be obtained byrequest fromthc STSSN coordinator atthe NMFS SEFC Miami Laboratory. CooperativeMarine Turtle Tagging Program Theneed for a centralizedclearinghouse formarine turtle tags was formally stated atthe World Conference onSca TurtleConservation heldinNovember 1979.The recommendation islisted asan action project under management techniquesinthe proceedings ofthe conference andreads: "Encourage theestablishment ofa worldclearinghouse fordata onall tag series used byall investigators throughout theworld. Tofacilitate exchange ofdata on tag returns, andto avoid confusion resulting from duplication oftag numbers, thisinformation should bemade available toall partiesinvolved inturtle research andconservation" Bjorndal, 1982!, Inresponse, theNMFS SEFC Miami Labora- tory,in cooperation withthc Panama City Laboratory, established andmaintains thecooperative marine turtle taggingprogram. Efforts have focused primarily onthose tags carrying a NMFS return address. A tag distribution database ismaintained thatlogs specific tagseries inuse and the researcher towhom those tags were issued. We welcomeandencourage others tosubmit information onall tag series produced anddistributed, tomake this data baseas comprehensive as possible. Twoadditional tagging data bases arecompiled andmaintained atthe Miami Laboratory: I!tag release records onindividual turtles and ! recapturerecords onindividual turtles, These data files primarily contain NMFS tag RAW DATA FORINT FIELD OBSERVER! STRANDED TURTLE STATE CQORD NATOR

NETWORK COORD INATDRS

SUMMARYREPORTS, USER REQUESTS

DATA PERyANENT ENTRY DATABASE RECORD

Figure1. SeaTurtle Strandingand SalvageNetwork flow ofdata from stranding event to requestsby users. records, but we encourage the archival of non-NMFS tag series release and rccapturc records as well. Information contained in the tag releasedata baseincludes: data source,tag number s!, releasedate, state/country, county, release latitudeand longitude,project type, species, scx, condition of turtle, straight!incand curvedcarapace rneasurcrncnts, weight, capture latitude and longitude, coded notes, and other uncoded remarks. Recapture data base records include:tag number s!,species, recapture date, state/country of recapture,recapture county, regionand statistical zone,recapture latitude and longitude,recapture method, condition of turtle, dispositionof turtle, tag status,report category, original releasedate, release location, original tagger and remarks, Uponnotification of a recaptureof a taggedturtle, we searchthe tag releaseda ta base for theoriginal taggingrecord, senda reply letter thanking the finder for reporting the taggedturtle and providing him or her with background informa tion on the recapturedturtle. The tag return data areen tered into the recapturedata base, and the recapture information is forwarded to the researcherwho originally applied the tag. The release and recapture data bases are maintained in a dBASE lil Plus" format. The IVMFS SEFC Galveston Laboratorymaintains independent data bases for releaseand recapturerecords of headstarted Kemp's ridleys. It is hopedthat theexistence and expansion of thesedatabases, especially the tagseries in usedata base, willbring uscloser to thc goals that were set forth at the World Conference in 1979. Western Atlantic Turtle Symposium Dataon manyaspects of marineturtle biology were gathered from 38 countries in theCaribbean basin by wayof their submissionof nationalreports prior to the first WcstcrnAtlantic Turtle Syrnposiurn WATS I! held in SanJose, CostaRica, in July 1983.Data were compiled on suchtopics as coastal shoreline habitat, known nestingbeaches, aerial andground surveys of nestingbeaches, numbers of nestingfemales, numbers of foragingsca turtles, natural mortality rateestimatcs,utilization and socio-economic statistics, and head start and cultureactivitics. These data constitute the WATSdata base and wereutilized to producethe summary tables published in the WATS1 symposiumproceedings. Thedata base was revised in February1984 based on comments and updates resulting from the symposium. A second symposium,WATS Il, was held in Mayaguez,Puerto Rico, in October1987. The WATSdata basewas expandedas new national reports and data were received. Southeast Turtle Surveys In additionto thedata bases discussed previously, information derived from a three-yearpelagic aerial survey programconducted from North Carolina to KeyWest, Fla., is archivedin theSoutheast Turtle Surveys SETS! aerial surveydata base. This data base primarily contains data on loggerhead Caret ta caretta! and leatherback Derrnorhelys coriacea!turtles. Our aerialsurvey data baseis not applicableto Kcmp's r idley, exceptto strengthenthe conclusion that Kemp'sridley is not a goodsubject for pelagicaerial survey work, at leastnot in the southeasternU.S., bccausc of its coloration, small size, and ncarshorc distribution.

155 Summary Developmentofthe various sea turtle data bases maintained bythe NMFS SEFC Miami Laboratory evolved from informationneeds related tosea turtle conservation andmanagement. Thedatabases onsea turtle strandings, tagging releaseand recapture!, biology WATS I andII!, andaerial surveys SETS! can contribute valuable data toward evaluating,conserving and managing sea turtle populations. Weencourage participation inexpansion ofthese data basesto makethem as comprehensive and accurate as possible, Literature Cited Bjorndal,K.A. Editor!. 1982. Biology and Conservation ofSea Turtles, proceedings ofthe World Conference onSea Turtle Conservation, SmithsonianInstitution Press, Washington, D.C., 583 p. U1rich,G.F. 1978. Incidental catch of loggerhead turtles by South Carolina commercial fisheries. Report toNational Marine FisheriesService, Southeast Regional Office, St. Petersburg, FL,under contracts Nns. 03-7-042-35151 and03-7-042-35121 mimeograph!,48 p. Evolutionary Relationships, Osteology, Morphology and Zoogeography of Kemp's Ridley SeaTurtle Peter C.H. Pritchard

Kemp'srid ley {Lepidochelys kempi! is a memberof thesubfamily Caref ti naeof thefamily . The genus i nclu des tzoospecies, L. kernpi andL. olivaceaand the subfamily also includes Ca refta ca retta and certain fossil forms. It shouldprobably alsoi ~eludeEretmochelys imbricata. The two species of Lepidochelysare considered distinct af thespecies level on the basis of merisficdifferences in scutation,proportional differences in componentsof the skull and carapace,and no evidenceof intergradationdespite the presence of bothspecies in theAtlantic system. Osteologically,L. kempi is ofinterestin thatthe carapaceis more completely ossified than that of other living cheloniids,with the intercostalfontanelles closing completelywith maturity and the plastral fontanellesreduced fo a small areaaround the enfoplastron,Furthermore, the peripheral bones near the middle of theseries on each sideare unusually well-developed and wide, sothat the overall carapaceshape is almost circular in manyjuveniles, the carapaceis act ually longer than wide. Zangerl argues that thesmall adult size,well-ossified carapace, and circularform areprimitive characteristics,typical of earlysea turtles lacking advancedmarinespecializations. However, Kemp's ridleyis actuallya highlymigratory species,well adaptedfor marinelife, and the broadand well-ossifiedshell, far from beingprimitive, may bezoidened and reinforcedso that the animal can usethe sidesof theshell for closureof the nest.The vigorous lateral rocking of thefernale during nestclosure is a veryconspicuous trait, and one peculiarto Lepidochelys.L. kempiand its congenershow more neural bones than any otherliving turtles.The primitiz e series ofaround eight neural elements is increased,usually by transversedivision of individual elements, to asmany as 15. Few fossil t urtles show this feature;indeed, it is only knownin theextinct cheloniids,Glyptochelone suyckerbuyki and Vrocolpochelys grandaevafrom the Cretaceousand theMiocene, respectively, and conceivablywas monophyletically derived in all theseforms, Thefunction of this development,if any, is unknown.Possibly the additional neural elements were z'nfroduced fo reinforce the vertebraltubercles typical ofhalf grown Lepidochelys thoughabsent in adults!. Cerfainfossil toxochelyids{eg,, Prionoche- lys and Cfenochelys! are known in which the vertebraltubercles were extremely well developed,and eachbore an additional, seeminglyneomorphic, "epithecal" bone at fhe tip. Another uniquefeature of Lepidochelys is the extraordinarydevelopment of the Rathke'sglands, present in the inguinal regionof manyturtles, but presentas a seriesof glandswhose outlets penetrate each of thefour inframarginal scuteson eachside in Lepidochelys. Thefunction of theseglands is still uncertain, and their product hasnot beenadequately analyzed, but it has beenspeculated that they may play a rolein olfactorydetection of conspecificturtles in the area,which in turn may assistin the coming-fogefherof animalsin theextraordinary simultaneous nesting assemblages or arribadas for zohichthis genusis famous. Kemp'sridley only nestsin thewestern Gulf ofMexico,with virtually all nestingoccurringin centraltosouthern Tamaulipas, Mexico,Adulfs apparentlyspend their livesin the Gulf, butj uvenilesare not infrequentlyencountered in U.S. Atlantic waters, and evenas apparentwaifs in zvesternEurope. Hozoever, the speciesapparently fails fo enter the Caribbean.The congenericL. oliva cea,zoidespread on mainlandcoasf s ofthe Indian and Pacific Oceans, also occursin West Africa and theGuianas, from where occasionalindividuals swim asfar westas the LesserAntilles. No majorgeographic variati on hasbeen documented between L. olivacea populationsin the threeoceans where if occurs,constituting an exanzpleof the curious phenomenonthat somefimes oceansseem to bemore trenchantbarriers than continentsin seaturf le zoogeography.

Keznp's ridley Lepidochelyskempi! was described by Samuel Garman, of the Museum of Comparative Zoology at Harvard,in 1880 Garman, 1880i. He based the new speciesupon two adult specimensfrom Key West,Fla., that had beensent to him by RichardM. Kemp,a manwhose name has been immortalized in the nameof the speciesthat is thesubIecfof this symposium. Garman actually named the new speciesThalassochelys {Co! pochelys! kempi, Thalassochelys being the generic name then in use for the loggerhead turtle, Garman thought that his new speciesmight be distinct at a level above that of the species,which is why he placed it in a separatesubgenus, or in the same paper! even a full genus.Garman noted that it wasa widespreadbelief that this new turtle wascommonly considered a crossbetween a greenturtle and a loggerhead,being known as"bastard" in theGulf of Mexico,an epithetthat persiststo thepresent in the German vernacular for the species Wermuth and Meztens, 1961l. This persistent belief in the hybrid origins of the ridley derivednot just from themorphology of theanimal, which in somerespects might be superficiallyjudged to be intermediate between the two species,but also becauseit was not known to nest, which seemedto support the

"Florida Audubon Society

157 "sterilehybrid" theory. Archie Carr 956!, in his book 'The Wind ward Road," had to report that the nesting habits ofKemp's ridley were still unknown, and the unusual breeding habits of thespecies were indeed not revealed until 1961as reported by Hildebrand 963! and Carr 963!. Garman 880!, however, while reporting the folk beliefs of thehybrid nature of the ridley, did not subscribe tothem himself. Innaming the species, heconsidered it distinct from theloggerhead inhaving a short,round shell, low "humps," marginal "plates," narrow head and swollen jaws, and theseare indeed good characters for recognizingthe species. Loggerheadshave wider, more triangular heads, a humpor thickenedarea above the tail, and lack the distinctive broadeningofthe shell, which in Kemp's rid ley apparently assists inthe nest-closing process, Garman also wrote that thespecies was distinguished from the olive ridley then called Thalassochelys olivacea!by the shape ofthe head, swollenjaws and the carapace scutes, although it is now known that the first two characters onlyseparate theadults ofthe two species. However, thecostal scute count nearly always five on each side in L. kempi, butgenerally sixto nine,and only occasionally five, in L.olivacea will distinguish even the hatchlings, Garman880! correctly recognized allthe species ofsea turtle considered valid today, and recognized none that todayare considered invalid, Indeed, inaddition toL. kempi, hedescribed Chelonia depressa, although hisprofessional interestswere primarily ichthyological. However, hisexample was not followed by many subsequent authors. In particular,the olive ridley and the loggerhead were confused for decades in the late19th and early 20th centuries, a misconceptionthatperhaps reached itszenith in a paperbyGadow 899!, in which the author attempted toexplain theapparent loss of scutes with ontogeny inloggerheads. Actually, his ha tchlings were the mul tiscu tate L. olivacea and his adults the pentecostalCaretta caretta. Today,the two ridleys are considered fullspecies, kempi and olivacea within the genus Lepidochelys, thisgeneric name beingconsiderably seniorto Garman's 880! Colpochelys. Thespecific epithet cannot bedisplaced bythe discovery ofsenior synonyms since it has been officially conserved bythe International Commission forZoological Nomencla- ture Opinion 660, 1963!. Forexample, one such senior synonym, mydas minor Suckow 1798, was thought by Werrnuth 956! to be basedona specimenofKemp's ridley, although Pritchard andTrebbau 984! summarize whyit was probably based, in fact,on a specimenof theolive ridley. Evolutionary Relationships Thegenus Lepidochelys isplaced within the Order Chelonia orTestudines or,increasingly, theChelonii since CheloniiLatreillc, 1800, has priority over Chelonia Macartney, 1802, and Testudinata Oppel, 1811!, Testudines Linnacus,1758, issimply a Latinplural form ot' Test Mdo, and thus lacks suprageneric nomenclatorial rank. Testudines Batsh, '1788, isclearly presented asa familial,notan ordinal, name, and the suborder . The latter allocation couldbechallenged, sincethe cryptodires aredefined onthebasis ofcertain characteristics, suchasthe retractile neck, notshown by seaturtles. Zangerl 969!, for example,placed the Cheloniidae in a Suborderthat he called Mctachelydia,whereas Gaffney 984! proposed a complex new hierarchial classification, inwhich the Cheloniidae wereplaced in theMicroorder Chelonioida, Infraorder Procoelocryptodira, Suborder Polycryptodira, Parvorder Eucryptodira,Hyperorder Daiocryptodira, Megaorder Cryptodira, andGigaorder Casichelydia, of the Order Testudines. Thefamily Cheloniidae isconsidered notto include natural subfamilies bymany authors see Pritchard and Trebbau,1984, fordiscussion!. Thefamily asa wholeincludes 31genera, ofwhich only four have living representa- tives.Some of thefossil genera are known from extremely fragmentary material, and in somecases names were publishedinthe form of preliminary notices e,g., Dollo, 1909! never followed upby a description. Thereissome justification forallocating theliving cheloniid genera Chelonia, Eretmochelys, Carettaand Lepidochelys! totwo subfamilies, theCheloniinae andthe Carettinae, Theformer subfamily would include, asa minimum,Chelonia, andthe latter Caret taand Lepidochelys. Thelatter two genera show a number ofdistinctive common features including thetriangular head, presence oftwo pairs of prefrontal scales, somewhat similar scalation ofthe crown ofthe head, andthe insertion ofan additional pairof costa 1 scutes atthe front ofthe series. Thetwo genera arealso linked byshared behavioraltraits,such asthe alternating terrestrial gait,similar eggsize, and certain stereotyped featuresof the nesting process,suchas the symmetrical, widely-spaced postureof the hindlimbs during oviposition, Theplacement ofthe hawksbill Fretmochelys! however,is less clear. Some authorities placethis genus close toChelonfa largely because of thepresence ofonly four pairs of costal scutes and the clear-cut, non-fragmented headscales. The former character alsocorrelates withcertain other differences, suchas the presence ofadditional anterior peripheral bones inCaretta andLepidochelys, andthe nuchal scute having contact with the first costals. However, asearly as 1942, Carr 942! pointedouta seriesofcharacteristics thatlinked Eretrrtochelys withCaret taand Lepidochel ysrather than with Chelonia, andthis argument remains persuasive see Pritchard and Trcbbau, 1984, for discussion!. Theallocation ofLepidochelys tothe Cheloniidae appears tobe unassailable. Other sea turtle families, such as the

158 Derrnochelyidae,Protostcgidae or Toxochelyidaeshow different kcy characteristics,such as the layer of mosaic carapacialbones of the Dermochclyidae,the highly fontanelledshell, nasal bones and often absentcpiplastra of the ,and the cruciform plastron of the Toxochelyidae.The rounded, thoroughly ossified carapaceof Lepidochelysdoes show strongsimilarity to that of Toxochelysand someother toxochelyidgenera, but this appearsto be a caseof parallelismrather than of evolutionaryrelationship. As hasbeen discussed by Zangerl980!, tracingthe evolutionaryhistory of the Chcloniidaefrom the to the presentleads to the odd conclusionthat the mostadvanced forms occurred early in the fossilrecord, whereas theliving chcloniidsare remarkably generalized in their morphology,and are sccrningly primitive. Thissuggests that the more specializedforms were ill-adapted for changingenvironmental conditions, and that the unspecialized central stem of the evolving cheloniids represented the formula for survival. Moreover, among the living genera, Lepidochelysseems to be the most primitive on morphologicalgrounds, although not on behavioralgrounds. ln this genus,for example,the gigantic size of many of the specializedsca turtles is not expressed,whilst the rounded, ossifiedcarapace and generalized food habits mark thisas an unspecializedgenus. In connectionwith thegeneralized diet, L. olimceaappears to be quite opportunistic in its feeding habits, often being almost exclusively herbivorous in the Indian Ocean Biswas, 1982!,but in the EasternPacific itis generally carnivorous seePritchard and Trebbau, 1984, for review!. However, the fragmentation of the neural bones discussed below is a rare development and almost certainlya specializedra ther than a primitiveone, even though a dominanttheme in vertebrateevolution hasbeen reduction rather than proliferation of discrete bony elements. Few fossil turtles show fragmented ncurals. The two fossil forms with this condition thc chcloniids Clyptochelone suyckerbuykiand Procolpochelysgrandaeva, from the Cretaceousand the Eoceneto ,respectively, of Europe and North America>may or may not be monophyleticwith Lepidochelys. Osteology and Morphology Carapace Thecarapace of L. kewpi,cordiform in the hatchlings, widens dramatically with growth, sometimeseven becoming wider than long in the half-grown, and even in adults the shell width is very close to the length. The mean carapace length and width of 50 hatchling L. kempiwere 42.05 and 33.79mm respectively Pritchard, 1980!, giving a ratio of 1.24:1.The mean carapacelength of nesting females was found to be 64.64cm in 1966,64.52 in 1979and 64.85in 1980 Chavez, Contreras and Hernandez, 1967;Pritchard, 1981!.Pritchard 969! found the shell width of adults to average 0.969of shell length. The overall range of carapacelength for nesting fcmalcs was 59-71cm in 1980 Pritchard, 1981!. The longest carapacelengths recorded for adult females measured 74 to 74.9cm Pritchard, 1969!. Intercostal fontanclles are evident in the shells of subadult L. kempi,but they close as maturity is approach

159 thesingle pygal bone, but with growth it expandsand makes progrcssivcIy increasing sutural contact with the posteriormostpair of peripherals,The space posterior to thelast pair of ribsis thusclosed even when the intercostal fontanellesare still quite open. Functionalinterpretation of the rounded, highly ossified carapace and the unique proliferation of theneural bones remainsconjectural. Perhaps the former characteristics, assuggested above, are merely primitive. Certainly they are sharedby such early sea turtles as the toxochelyids , Thinochelys andToxochelys, aswell as by the cheloniidPappigerus camperi sce Wecms, 1974; Zangerl, 1980!. On the other hand, these two features together create a verystrong and unusuallyextended mid-peripheral section of the carapace,and thosewho havewatched the vigorousnest-closingbehavior ofthis species, inwhich the shell is rocked from side to side so that the lateral margins ofthe shell alternately thump the substrate, might conclude that the development indeed had considerable functional significance,Thesupernumerary ncurals have no obvious function inadult Lepidochelys andmay simply reflect the actionof some harmless gene, perhaps associated with the tendency towards a parallelfragmentation ofscutes and scalesmanifested inthe carapace ofthe congener L.olivacea, orin the head scales ofall the Carettinae. However, it is conceivablethat the additional neurals might play a roleinstabilizing thevertebral tubercles thatare such a striking characteristicofimmature Lepidochleys. Certainfossil toxochelyids e.g.Prionochelys andCtenochelys! areknown in whichthe vertebral tubercles were extremely welldeveloped, andeach bore an additional, seemingly ncomorphic, "epithecal"bone near its tip seeZangerl, 1953 and 1980!, Plastron Likethe carapace, theplastron ofL. kenipi isextensively ossified, sothat in adults the only persistent cartilaginous areasare along the midlines ofthe hyo-, hypo- and xiphiplastra, andeach side of the posterior dagger-like process ofthe en toplastron. Retention ofcartilage inthese areas allows the plash ona modestdegree oflongitudinal flexibility alongthe midline and along the sides where the hyo- and hypoplastra meetthe mid-peripheral bones.The hyo- hypoplastralsutures in adult Lepidochelys develop distally to a degreethat completely eliminates thelateral fon- tanellespresent inthe plastra of almost all other sea turtles. In mature Lepidochelys, theselarge plastral bones lose all ofthe proximal and distal interdigitations thatcharacterize thehyo- and hypoplastra ofmost sea turtles, their flanks beingseparated fromadjacent bones only by a narrowstrip of gelatinous cartilage, lnvisceral view, the xiphipla stra canbe seen todevelop a process that penctra tesdeeply into the posterior margins ofthe hypopla stra. The medialjunc- tionof the epiplastra doesnot ankylose, presumably because ofthe need toretain some midline flexibilit, but this junctionisstabilized byconsiderable thickening, in a visceraldirection, ofthe juxtaposed ends of thesebones. The entoplastronremains thin and dagger-like throughout life, with a medianvisceral ridge in adults and extensive but non-ankylosinganteriorcontact with the inner edge ofthe proximal ends ofthc epiplastra. A single pair of narrow, contiguousmesoplastra hasbeen reported asan anomaly inL, oiivucea Pritchard, 1966!, butthis has not been reported for L. kernpi. Theplastral formula forL. kernpi, defined as100 times the shortest distance across thebridge betwecn theaxillary andinguinal notches! divided by half the width of the plastron, averaged 650 for11 mature females

161 beena resortbeach, asthe tracks are so ephermal and the nesting emergences themselves sobrief that nesting can easilybe overlooked. However, for the time being, we must proceed on the assumption that the Rancho Nuevo site is theonly location where L. kempinests regularly and in significantnumbers. MatureL. kempi have been found only in theGulf of Mexico,The report by Chavezand Kaufmann 974! of an individualnesting in CaribbeanColombia almost certainly represented a misidentified L.olivacea tagged in the Guianas Pritchard and Trebbau, 1984!. Moreover, such adults are not uniformly distributed in Gulfwaters, as outsidethe nesting season they are commonly encountered in biologically productive, estuarine situations such as thecoast ofLouisiana near the mouth of the Mississippi River, or the Laguna del Carmen area of Estado Campeche, asevidenced bytag returns of individualstagged while nesting Chavez, 1968; Pritchard and Mdrquez, 1973>, The immatureanimals, onthe other hand, frequently leave the Gulf of Mexico, and are or were! commonly found on the Atlanticcoast ofFlorid, and in appropriate embayments andprotected coastal habi tats as far north as New England Pritchard,1969!. Bleakney 955! even reported four specimens fromNova Scotia. Occasionally, individuals may reachEurope, Brongersma 972! listed25 records ofL. kempi inEuropean Atlantic waters, and it isof interest that twoof the animals head started inGalveston were picked up in France and Morocco Fontaine etal., I 989!.Brongersma andCarr 983! discussed thesingle known Mediterranean record, a 29.4centimeter specimen from Malta. Therecontinues debate asto whether the animals inU.S. Atlantic waters, especially inthe northern section, are lost tothe population, orare indulging ina normal orat least optional! part of their developmental odyssey. Specimens asfar north as Chesapeake Bayare known totravel out of the Bay to Florida waters aspart of a seasonalmigration to escapewinter temperatures Musick et al., 1983!. Musick et a!. ibid.!identified 47 L. kempi as well as 664 Caretta carettaand eight Dermochelys coriacea! ina study ofsea turtles inChesapeake Bay.Many of these animals they found dead;fewer than 20 percent of themortalities could be attributed positively toentanglement orincidental catch in fishingoperations, although doubtless thisoccurred inmany other cases but proof was lacking. Winter strandings werealmost non-existent, butthere was a minorsurge of strandingsinJunc of each year, which to Musick etal. 983! waspossibly attributable topoor physical condition ofmany turtles resulting from suboptimal conditions during winterfollowed byan arduous migration tothe bay. These authors found no evidence ofturtles overwintering in Virginiawaters, and although winter torpor has now been documented forboth Caretta caretta in Florida!and Cheloniaagassizi inthe Gu]f of California!, theevidence forthis in L, kempi remains anecdotal andunproven. A series of 29Virginia L. kempi showed an average carapace length of 40,0 cm, with 75 percent between 30and 45 cm, Ridleysoccur regularly, although infrequently, inwaters north ofVirginia, Carr 957! reported anextraordinary observationmade byW. Schevill ofWoods Hole Oceanographic Institution, ofseveral dozen yearling ridleys, out of "awhole fleet of such turtles," being stranded onWoods Hole beaches while travelling outof Buzzards Bayinto VineyardSound. Lazell 976! reported 16specimens ofL. kempi from waters ofCape Cod and associated islands, and hccommented thatpresence oflarge numbers ofimmature L,kempi inMassachusetts watersmay have been cyclic, addingthat the last major inundation wasin 1961, a time, hepostulated, thatcorresponded tothe beginnings ofthe d ecIineof the great arribadasat Rancho Nuevo. Nevertheless,suchinvasions maystill occur ona limitedscale. Ananonymous reporter Anonymous, 1985!related thatno fewer than 36 L. kempi aswell as three C.caret taand three Chelonia mydas, were found in Long Island Sound, NewYork, infall of 1985, anddoubtless morewere found during theensuing winter. The largest ridIey was reported asless than 30centimeters in"diameter". Nomention wasmade oftheirbeing tagged, sothey probably didnot derive fromthe Galveston headstart project. Maigret 983! gave a briefdescription ofan enormous flotilla, numbering thousandsofL. kempi each about 30 cm in carapace length, ofwhich 75 were captured forexamination. Thiswas reportedtohave occurred inMay 1982 ata locationdefined bythe coordinates 33'N,74 W, which corresponds tothe northwesternAtlantic, between North Carolina and Bermuda. However, subsequent correspondence betweenDr. Maigretand myself first revealed tome that this concentration ofturtles was on the other side of the Atlantic, off northwesternAfrica as the title of the paper implied!. The given latitude of74 'W had been a misprintfor14'W, derivingperhaps from the European custom ofwriting the number "1" similar in formto the number "7" in U.S, or Britishorthography. Moreover, Dr.Maigret kindly agreed toseek confirmation ofthe iden tity of the turtles, and was ultimatelyable to locate a specimen thathad been kept and preserved, Hesent me a seriesofphotographs, andthe specimenshown was unquestionably a loggerhead Caretta caretta!! WhetherL.kempi isevolutionarily closerto the ancestral generic stock than isL. olivacea isa matter forconjecture. However,while many ofthe differences between the species area reflectionofthe more durophagous dietof L. kern pi, thedifference inthe scu te configurations ismore difficult toexplain. Certainly themultiscutate condition ofL. olivacea appearstobe the derived condition it is indeed unique, both in the high numbers ofscutes especially costais! borne byalmost anyrepresentative ofthis species, andin the extraordinary degreeof intraspecific variability. Thesubject isdiscussed further by Pritchard 969!, Hill 971!, and Mast and Carr 989!. On the other hand, this proliferation ofscutes can occur in other turtle species, including freshwater andland forms, in caseswhere the animals were

162 hatched artificially under stressful conditions of temperature or humidity. Pritchard969! presentedsome conjecture as to how L. kempimay havearisen by isolationof an early Lepidochelys stock in the Gulf of Mexico following the closure of the Isthmus of Tehuantepec. Although the Caribbean route remainedtheoretically open to reinvasion or escape!,this wasnot utilized becauseof thelong reachof inappropriate habitat in the Caribbean,Lepidocheiys flourishing in areaslittle frequentedby Carettaand requiring productive estuarineconditions for optimal feeding,Between the mouth of the Orinoco an important feedingground for L. olivacea!and the Lagunadel Carmenin the Gulf of Mexico,a stretchof thousandsof kilometers,there are no major estuaries, the rivers of the Caribbean Islands being too small to produce them, the mainland shores of western Venezuela, Colombia, and the Yucatan Peninsula too dry, and the watersheds of Caribbean Central America too restricted by the narrowness of the coastal lowlands in these countries. Hendrickson980! presentedan intriguing hypothesisregarding the zoogeographyof L. kernpi,arguing that the specieswas isolated in Gulf watersfollowing the relativelyrecent closure of the marinepassage through Panama, 3.5 to 4 millionyears ago, rather than that of theIsthmus of Tehuantepec.Hendrickson also hypothesized that theclosing of the Panamanianportal may havebeen accompanied by a majorredirection of oceancurrents, the Atlantic North EquatorialCurrent now being denied accessto the Pacificand insteadsweeping northward and clockwisewith greatly i~creasedforce, He further theorizedthat the presenceof this new currentmight haveconstituted a chronic stressupon populationsof L. kempiby sweepingrelatively large numbers of immaturescompletely out of the Gulf of Mexico,to the easternseaboard of the U.S.and ultimately,in somecases, to the BritishIsles and northern Europe. He expresseddoubt that theseindividuals were ever able to make their way back to the Gulf of Mexicoto join in the reproductive effort of the species. The last question remains open. Certainly the adults of L. kempiseem to be able to resist drifting outside the Gulf of Mexico,and not a singleadult hasbeen reliably recordedfrom extra-Gulfwaters. It is possiblethat juvenilesthat are not carried too far north on the Atlantic seaboard are ultimately able to return to the Gulf and reproduce, but perhaps specimens in Canadian or European waters are truly lost. If specimensin U.S, Atlantic waters were able to survive but were not able to return to the Gulf the question arises as to what they would do when they reached maturity, as regardsegg production and oviposition, In other words, why do they not nest alongsideCaretta on southeastern Atlantic beaches?A fair proportion of the specimensof L. kernpifound in Atlantic waters of Florida are healthy animals only a few centimeters short of mature carapacelengths, and it seemsunlikely that such individuals would inevitably die before actually reaching ma turity, having survived all dangers up to tha t poin t. However, since theydo not occurin Atlantic watersas adults, I am inclined to believethat they arecapable of returning to the Gulf to reproduce. Seaturtles have evolved with ocean currents as a constantfactor in their lives ever since their ancestors first took to the sea,and it seemsimprobable that L, kempihas not yet learned to cope with them, Additional efforts to tag Kemp's ridleys on the Atlantic seaboardmay help settle this long-standing issue.

Literature Cited Anonymous. 1985. Turtle. New Yorker,30 December1985, p. 16-18. Biswas,S. 1982. A reporton the olive ridley Lepidochelysolivacea Testudinere Cheloniidae! of Bayof Bengal.Rec. Zool. Surv. India, 79:275-302. Blcakney, S. 1955. Four records of the Atlantic ridley turtle, Lepidochelyskempi, from Nova Scotia. Copeia1955 !:137. Brongersma,L, D. 1972, European Atlantic turtles. Zoologische Verhande! ingen Rijksmuseum van Natuurllkje Historic, Leidcn, The Netherlands,No. 121,318 p. Brongersma,L. D. and A. F. Carr. 1983.Lepidochelys kempi Carman! from Malta. Proceedingsof theKoninktit'ke Nederlandse Akademie van Wetenschappen,Series C, 86!:445%54. Caillouet, C.W.,Jr., D. B. Koi, C.T. Fontaine,T.D. Williams, W.J. Browning and R.M. Harris. 1986.Growth and surviva! of Kemp's ridley seaturtle, Lept'dochetyskempi, in captivity. NOAA TechnicalMemorandum NMFS-SEFC-186, Iii plus 34 p., 12Tab!esand 7 Figures. Carr, A. F. 1942. Notes on sea turtles. Proceedingsof the New EnglandZoology Club, 21:1-16. Carr, A. F. 1956. The Windward Road. Alfred A. Knopf, New York, 258 p. Carr, A. F. 1957. Notes on the zoogeography of the Atlantic ~caturt!cs of the genus Lepidochelys.Revista de BiologiaTropical 5:45- 61. Carr, A. F. 1963. Panspecific reproductive convergence in Lepidochetyskern p. Ergebnisse der Biologic,26:297-303. Chavez, H, 1968, Marcado y recaptura de individuos de tortuga lora, Lepidochelyskempi Garman!. Instituto Nacional de Investigaciones BiologicoPesqueras, 19:1-28. Chavez, H. and R. Kaufmann. 1974.Informacion sobre la tortuga marina Lep>dochelyskern pi Garman!,con referencia a un exemplar marcado en Mexico y observado en Colombia. Bulletin of Marine Science,24:372-377. Chavez,H.,M. Cont reras and E. Hernandez. 1967. Aspectos biologicos y proteccion dela tortuga lora, Lepidochelys kernpi Garman!, enla costa de Tamaulipas, Meico. Instituto Nacional deInvestigaciones Biologico Pesqueras, 17:140. Dollo,L 1909.The fossil vertebrates of Be! gium. Annuals ofthe New York Academy ofScience 19!:99-119. Cadow, H. 1899. Orthogenetic variation inthe shells ofChelonia. Wiley Zoological Results, Cambrid ge University Press, 3;207-222. Gaffney,E.S. 1984. Progress towards a natural hierarchyof turtles. StudiaGeologica Salmanticensia. StudiaPalaeocheloniologica, Volumen, Especial1, p. 125-132. Carman,S.1880. On certain species ofChelonioidae. Bulletin ofthe Museum ofComparative Zoology, 6:123-126. Hendrickson,j. R, 1980.The ecological strategies of sea turtles. American Zoologist 20!:597-608. Hildebrand,H.H. 1963.Hallazgo delarea de anidacion dela tortuga marina "lors," Lepidochelys kempi Garrnan!, enla costa occidentaldel Golfo de Mexico Rept,,Chel.!, Ciencia Mexico! 22!:105-112. Hill,R. L. 1971.Surinam turtle notes -1. Polymorphismofcostal and vertebral laminae inthe sea turtle Lepidochelys olivacea!, S tichting¹tuurbehoud Surinam e STINASU,Mededeling, 2:3-9. Landry,A. M,, Jr. 1989.Morphometry of captive-reared Kernp's ridley sea turtles. This volume.! Lazell,J.D. 1976. Thi s brokenarchipelago. CapeCod and the Islands, Amphibians andReptiles, Demeter Press, 260 p. Mast,R.B. and J. L. Carr, 1989. Carapacial scutevariation inKemp's ridley sea turtle Lepidochelys kempr'!hatchlings andjuveniles This volume!. Maigret,J.1983. Repartitiondestortuesdernersurlescotesouestafricaines, BulletindelaSocicteHerpetologicluedeFrance28:22- 34. Mora,J.M. and D.C. Robinson, 1982. Discovery ofa blindohve ridley turtle Lepidochelys olivacea!nesting atPlaya Ostional, Costa Rica.Revista de Biologia Tropical, 30;178-179. Musick,J.A., R, Byles, R,Klinger and S. Bellmund, 1983. Mortality andbehavior ofsea turtles inthe Chesapeake Bay.Summary report,1979 through 1983. p. i-ii, 1-52, 1-30, Al-A2 and 1-15, respectively, Virginia Institute ofMarine Science, Gloucester ['oint, VA Mimeograph!. I'ritchard,P,C, H. 1966.Occurrence ofmesoplastra ina cryptodiran turtle Lepidochelys olivacea. Nature 210:652. Pritchard,P.C. H, 1980. Report onUnited States/Mexico conservation ofKemp'sridley seaturtle atRancho Nuevo, Tamaulipas, Mexicoin1979. Final Report onU.S. Fish and Wildlife Service Contract No,14-16-0002-80-2I, iii plus49 p. Mimeograph!. Pritchard,P.C. H. 1981. Report onUnited States/Mexico conservation ofKemp's ridley sea turtle atRancho Nuevo, Tamaulipas, Mexico,in1980. Preliminary Reporton U.S. Fish arul Wildlife Service Contract No.14-16-0002-80-21, iiiplus 42 p, Mimeograph!. Pritchard,P.C.H. 1969. Studies ofthe systematics andreproductivecycles ofthegenus Lepidochelys, Ph.D.Dissertation, University of Florida,Gainesville, FL, 197 p. I'ritchard,P.C. H, and R, Marquez M.1973, Kemp's ridley or the Atlantic ridley, Lepidochelys kempi.International Unionfor ConservationofNature and Natural Resources Monograph No.2 MarineTurtle Series!, p.1-30, Pritchard,P,C. H. and P. Trebbau. 1984. The turtles ofVenezuela. Societyfor the Study ofAmphibians andReptiles, 403p. Weems,R.E. 1974. Middle Miocene seaturtles Syllomus, , ! fromthe Calvert Formation. Journalof Paleontology48:278-302. Werrnuth,H. 1956. Versuch der Deutung einiger bisher iibersehener Schildkroten Narnen. Zool. Beitr. 2;399-423. Wcrmuth,H.and R. Mcrtens, 1961. S childkrttten,Krokodile,Bru~sen. Gustav Fischer Verlag, Jena, 472 p. Williams,E,E. 1950. Variation andselection inthe cervical central articulations ofliving turtles. Bulletin ofthe American Museum of NaturalIIistory 94:507-561. Zangerl,R.1953. The vertebrate faunaof the Selma Formation ofAlabama, PartIV, The turtles ofthe family Toxochelyidae. Fieldiana,Geology Memoirs 3 137-274. Zangerl,R.1958. Die oligozanen rncerschildkroten vonGlarus. Schweizerischen Palaontologischen Abhandlungen 73:5-55. Zangerl,R.1969. The , p.311-399. In:Gans, C.,A. d'A, Bellairs andT. S. Parsons Editors!, Biology ofthe Reptilia. AcademicPress, London and New York, Morphology A. Vol.1. Zangerl,R.1980. Patterns ofphylogeneticdifferentiation inthe toxochelyid and cheloniid seaturtles, Americart Zoologist 20:585- 596. Feeding, Growth Rate and Survival of the 1984Year-Class of Kernp's Ridley SeaTurtles I,epidochelyskernpi! Reared in Captivity Charles W. Caillouet, Jr., Sharon A. Manzella, Clark T. Fontaine, Theodore D. Williams, Marty G. Tyree and Dennis B. Koi'

A feedingexperiment was conducted between August 22, 1984 and February 28, 1985 on the1984 year-class ofKemp's ridley seaturtles Lepidochelys kempi!. Hatchlings were assigned tothree treatments in a randomizedblock design. The treatments includedtwo levelsof feeding, high and low, represented byfeeding rate percentageof body weight per turtle perday! and food ration weight of' food per turtle per day!. The high level began at I.3 timesthe lozo level in termsof weight of food per turtle per day,and reached 2.5 times the lozo level by the end of the ex perimen t. At thehi gh level,t zoofeeding freq uencies were tested, once- dailyin themorning versus twice-daily in morning and afternoon, For twice-daily feeding, the daily ration zoas divided into two equalportions. At thelow level, only once-daily feeding was tested. Response variables included feeding rate food ration, survival, grossfood conversion ef'ficiency, zoeight gain and a grozothrate index. Temperature, salinity and pH were monitored d urzng the experiment. Turtlesthat received the most food had the highest zoeight gains and growth rate indices. Hozoever, for thehigh level of feeding, turtlesreceiving the entire daily ration in onefeeding per day had a smallerdaily weight gain than thosein which the daily ration wasdivided into twoseparate feedings. There was no apparent difference in theresponse to once-dailyvs. twice-daily feeding at the high level! as measuredby thegrowth rateindex. Grossfood conversion efficiency zoas either bet ter loweramount of food fed per unit increasein weightof turtle! at thelow level offeeding or did notdiffer from thatat thehigh level of feeding. Theexperimental feeding levels arzd frequencies had no apparent effect on survival, and overall survival during the experiment zoasvery high 95.8percent!.

Approachesused in the Kemp'sridley seaturtle Lepidochelyskempi! recovery program include experimental head starting to establish a new nesting colony of Kemp's ridleys at the Padre Island National Seashorebordering the Gulf of Mexico near Corpus Christi, Tex. KIIma and McVey, 1982!. Head starting involves collecting, incubating and hatching the eggs, imprinting the hatchlings and rearing the turtles from ha tchlings to yearlings 9 to I I months of age>in captivity Mrosovsky,1983; Caillouet, 1984; Fontaine et al., 1985!.Survivors in good conditionand healthare tagged and released into the Gulf at a sizethat, according to the current working hypothesis, improves their chances of survival as compared to that of wild hatchlings. Aznongthe objectivesof experimentalhead starting has been the improvementof captiverearing methods.This paperdescribes a feedingexperiment conducted on Kemp'sridleys of the1984year-class to determinethe effectsof feedinglevel and frequencyon their growth rate,gross food conversionefficiency and survival in captivity. Hatchlings Padre Island-imprin ted Kemp's ridley ha tchlings from 19 clutches of the 1984year class seeCa illoue t et al., 1986a, Tables 3-13 and 15! were transferred from the Padre Island National Seashore to the head start facilities at the Galveston Laboratory from July 24-27,1984. The eggs from which thesehatchlings were obtained had been collected in the usual way from thebeach at Rancho Nuevo. The eggswere packed in polystyrene foam boxes containing sand from the PadreIsland National Seashore, with oneclutch per box.Boxes containing the eggsand sandwere flown by single-engine aircraft to the National Seashorewere they were tended by NPS personnel during incubation. Upon emergence,the hatchlings were imprinted by brief exposure to the Padre Island beach and surf. Clutches of hatchlings were placed in wax-coated, corrugated cardboard boxes for shipment to Galveston. Some of the boxes containing hatchlings were transported by NPS station wagon from the National Seashoreto the U.S. Navy Baseat CorpusChristi, and thenceto Galveston'sSchole's Airfield aboarda U.S.Navy aircraft.The boxes were thentransferred by pick-up truck to thehead start facilities. Other boxes were transported by NPSstation wagon from

' National Marine Fisheries Service BLOCK

RACEWAY1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

SECTIO 1

TR EAT NIENT L 1 H 1 H2 H1 H2 L1

EAST QUONSET HUT WEST QUONSET HUT

Figure1.Physical layoutof the randomized completeblock design forthe feeding experiment onKemp's ridleysea turtles ofthe1984 yearclass threeblocks ofthree raceways withthree treatments I Hl,l l2 and Ll/ randomlyassignedto each; clutch numbers areshown bysection within raceways!, theNational Seashore tothe head start facilities. Allboxes contained a 2.5cm layer ofmoistened polyurethane foam tocushion the hatchlings and prevent their desiccation during transport. Rearing Facilities Hatchlingsinthe experiment werereared inbuckets placed innine rectangular ,8x 6.1 m!, fiberglass raceways Figure1!located intwo polyethylene sheeting-covered quonsethuts Fontaine etal., 1985, 1989!. Each raceway containedapproximately 3,140liters of seawater, Suspended within each raceway werc 108 yellow, plastic buckets 9.5-litercapacity!, Turtles were reared inisolation from each other inthese buckets, oneturtle per bucket, toprevent theirattacking, bitingand injuring oneanother, asthey are very aggressive and KlimaMcVey, 1982; Clary and Leong, 1984!,The bottoms ofthe buckets were perforated with1.3-cm diameter holes, toallow exchange ofseawater and liberationof turtle excrement and uneaten food. Forced-air, gas-fired heaters maintained warm air and water temperatureswithin thequonset huts during the winter. Turtlesina racewaywereall treated similarly with regard tofeeding, cleaning ofthe raceway andmanagement of seawaterandwastewater. Raceways were drained, flushed byhosing with fresh tap! water, and refilled with clean seawaterthree times a week.Once each week, all raceways were drained, scrubbed with brushes and flushed to removeattached algae, uneaten food and accumulated wastematerials. Temperature, salinityand pH were moriitoredin thenine raceways used in theexperiment. Seawaterforthe raceways waspumped from the Gulf of Mexico through well-points buried inthe sand below water atthc Galveston beachfront Fontaine etal., 1985, 1989!, After a periodallowed forsettling ofparticulates, theseawater wasstored infiberglass reservoirs near the quonset huts, and was used as needed. The reservoirs were shrouded with insulationand were outfitted with emersion heaters tokeep the seawater warm during winter. Feeding Experiment ExperimentalDesign Thcexpcruncnt wasconducted inparallel with that of Landry 989!. A randomizedcomplete blockdesign was usedfor the experiment toisolate possible microcnvironmcntal variationamong raceways fromthe treatment effects Table 1, Figurc1!. For this purpose, the nine raceways used in the experiment were divided into three adjacent groups of three racewayseach, and thesegroups were treatedas blocks numbered 1 3. Oneof the threetreatments Table I! was randomly assignedto eachraceway within eachblock. Prior to receiving the hatchlings, we anhcipatcd that we could reduce possible effects of variation in age and other characteristics of clutches on the treatment effects by distributing thc hatchlings in such a way that each raceway contained three different clutches, with the constraint that clutches assigned to a given raceway were as close in age aspossible. Location of clutcheswithin eachraceway was randomized among three sections of 36 bucketseach Figure 1!, Variation among clutches within raceways was treated as a random nested effect within raceways. All clutches except clutches 1 and 18 contributed hatchlings to the experiment. Clutches hatched between July 15-23,1984, so they varied in age only slightly more than a week. Therefore, all clutcheswere considered the sameage, and their agein days wascalculated from the meanhatch date of July 18,To calculate the mean hatch date, each date was weighted according to the number of hatchlings that emerged on that date. The modal and median hatch dates also were 18 July. Sevenclutches -7, 12 and 15! used in the experiment did not contain enough hatchlings to be assigned randomly to more than one block, but the remaining 10 clutches , 8-11, 13, 14, 16, 17 and 19! were large enough tobe divided betweentwo raceways,with eachof the two racewaysbeing in different blocks Figure 1!, However, section3 of raceways6, 9 and14 contained fewer than the full complementof 36 hatchlings, because clutches 2, 12and 17were not large enough to provide full complements of 36 each to these raceways. Foods and Feeding Two commercialdiets wereused in headstarting the 1984year class seeFontaine et al,, 1985; Caillouet et al., 1986a!. Bothwerc dry, floating,pelleted diets. The experiment began on August22, 1984, with a dietmanu factured by Central Soyaand Subsidiaries, Decatur, Ind. Thisdiet had tobereplaced after November 10 because the ncwbatch we received did not have the same floating characteristicsof the earlier batches.We switched to a scaturtle chow a modified trout chow! manufactured by Purina, Richland, End.The latter was the same diet used by the Cayman Turtle Farm 983!, Ltd., Grand Cayman,B,W,I,, for rearinggreen turtles Chelonia~~da,', -:.id it had beenrecommended earlier by thc farm's director, JamesWood personal communication, August 198'.,!. The standard feeding technique developed by Fontaine et al. 985! in head starting Kemp's ridleys cf the1981-1983 year-classeswas to setthe food ration per turtle asa percentageof thearithmetic mean weight per turtle, determined by weighing samples of turtles at roughly monthly intervals during head starting. The procedure was mochficd for the experiment by substituting geometric mean weight for arithmetic mean weight in the calculations of food ration. This was done becausethe variance in weight among head started Kemp's ridleys increaseswith average weight in sucha way that a logrithmic transformationof weightseliminates such heterogeneity of variance Caillouetet al., 1986b!.Once the weight of food pcr turtle was calculated for a given raceway, the food was distributed to each turtle by volumetric measure based on the weight: volume ratio for the food. Through September4, 1984,all hatchlings received the samedaily ration basedon approximately 10percent of their initial arithmeticmean weight seeCaillouet et al1986a,Table 15!. Feeding under the experimentalprotocol Table 1! began on September5, but August 22 was considered the start of the experiment becauseit was the date on which the turtles were first weighed for the experiment Table 2!. Weighings continued at 22-29day intervals Table 2! until February 28, 1985, when the experiment was terminated as a consequenceof some turtles having outgrown their buckets. Two experimental feeding levels were tested in the feeding cxpcrirncnt Tables I and 3!. Fccding level was represented in two ways: ! feeding rate expressedas a percentageof gcomctric mean body weight pcr turtle and ! food ration expressedas the weight of food fed per turtle perday, At! hebeginning of theexperiment, the high feeding level representedapproximately 1.3 times as much food by weight asthe low level,and it reachedapproximately 2,5 times the low level by the cnd of thc experiment, This shift merely reflecte the differences in growth among turtles in the diffcrcnt treatments.Because day to day conditionsin the racewayswere affected by tcrnpcrature, the amount of uneaten food, and the amount of turtle excrement, the experimental daily food rations had to bc altered from time to time. For example, when bloating causedby overfeeding occurred, feeding was interrupted for a day or two to allow the turtles to recover. After the experiment, theactual rations as well as the feeding rates in pcrcentageofbody weight wererecalculated for eachinterval between weightings Table3!, While thc feedingrate was reduced gradually over the period of the experiment,the food ration increasedbecause thc turtles weregrowing. Becausefeeding rates were controlled as a percentageof mean body weight during the experiment Tab]e 3!, for any given feedingrate the actualfood ration receivedby eachturtle in a givenraceway varied dependingupon the weight per turtle in that raceway at the beginning of each interval when feeding rate was adjusted seeFontaine et al., 1985;Caillouet et al., 1986b!.For this reason,analyses of variancewere conducted on daily food ration Table4! and

167 Table3.Daily food ration grams!' and daily feeding rate %! averaged bytreatment andtime interval during the feedingexperiment on the 1984 year-class of Kemp's ridley sea turtles.

Trea tment High level Low level Time Inclusive once/day twice/day once/day in I H L1 Di grams %! grams %! grams %! Aug. 22-Sept.12, 1984 2.0 .3! 2.1 .0! 1.6 ,7! CentralSoya Sept. 13 - Oct. 10 1.9 .3! 2.2 .3! 1.2 .9! Oct. 11 - Nov. 7 4.3 .6! 5.0 .4! 2.6 .9! NA' Nov. 8 - Dec. 5 3,9 ,5! 4.8 .6! 2,1 .6! CentralSoya Purina 5 Dec. 6, 1984 Jan.I, 1985 2.3 ,4! 2.5 .3! 1.2 0.8! Pu rina 2 6 Jan.2- Jan.30 2.9 ,5! 3.1 .3! 1.4 .9! 7 Jan. 31- Feb.27 3.0 ,2! 3.5 .2! 1.3 .7! 'Dailyfood ration was determined foreach raceway because allturtles ina givenraceway received thesame ration! bydividing thetotal grams ofdry tood per turtle per day during a given timeinterval between weighings bythe number ofdays inthe interval. sDailyfeeding rate was determine foreach clutch within a givenraceway byexpressing dailyfood ration for a giventime intervalbetween weighings asa percentage Va!ofthe geometric mean body weight wet! at the beginning ofeach time interval. 'NA= not applicable asa separate phase.This was a transitionperiodin which the d ietwas changed fromCentral Soya toPurina pelletsafter 10 November 1984 due to problemsencountered with the former diet sectext!. Table5. Analysis of varianceof transformed'daily feedingra te' in thefeeding experiment on Kemp'sridley sea turtles of the 1984year-class.

Source of Degreesof Mean varia tion freedom square

Blocks 2 0.05513 0.20 ns' Treatments, T 2 74.79456 271.81 Hl & H2 vs, Ll 0! 48,48420! 539.61 Hl vs. H2 ! .10494! 4.02 ns Time interval, I 6 249.54165 906.86 ' T x 1 interaction 12 0.89448 3.25 ' Experimental error 40 0.27517 Clutches within raceways 126 0.53264 'Angular arcsine!transformation seeSokal and Rohlf,1981, p. 427-428!. sDailyfeeding rate was determined for each clutch within a givenraceway hy expressingdaily foodration for a giventime intervalbetween weigh ings as a percentage 'F»! of thegeometric mean body weight wet! at the beginning of eachtime interval. 'ns = non-significantat P = 0.05. ~»= significantat P 0.05.

169 ontransformed daily feeding rate Table 5!, An angular transformation Sokal and Rohlf, 1981, p.427428> wasused fordaily feeding rate because percentages usually require such transformation tomeet assumptions ofanalysis of variance, TwofecsIing frequencies weretested atthe high feeding level Tables I and 3,Figure I!:once,and twice-daily, oncein the morning andonce inthe afternoon Treatment H2!.For twice-daily feeding,the daily ration was divided into two equal portions, sothe turtles in Treatments HIand H2 received the same amountof foodper day. The low level of feeding was only once-daily inthe morning Treatment LI!. Treatment H2 wasconsidered a control asit representedtheold system offeeding Fon taine et al., 1985; Caillouet etal., 1986b>, The combinationof low feeding level and twice-daily feeding was not tested. Sampling for Weighings Todetermine growth and adjust feeding rates, wet body weights were deterinined fromrandom samples of10 turtlesper clutch in each raceway ateach weighing during the experiment Table 2!. Weighings were made to the nearest0.1 gram on an O'Haus, triple-beam balance. The balance pan was dried and the balance re-zeroed after consecutiveweighings ofthree turtles. The geometric mean weight ofthe combined samples within a racewaywas usedas the basis for adjustingfeeding rate. Survival Survivalfrom the beginning toend of the experiment wascalculated foreach clutch ineach raceway and was expressedin percentage,

Results Feeding Rations and Rates lnthe analysis ofvariance fordaily food ration Table4!, allmain effects and the treatment x time interval interaction wercsignificant rcfers throughout this paper to the critical region of rejectionof null hypotheses atP <0.05>. Orthogonalcontrasts among treatments Table 4!showed that the daily food ration was significantly andsubstan- tiallyhigher atthe high feeding level as expected!, butalso that the ration for twice daily feeding wassignificantly andslightly higher when split into twice-daily feedings thanwhen given inonly one feeding Table 3!,probably an artifactofthc pellet size inrelation tothe use of two different sizes ofvolumetric measures todole out the pellets for thetwo feeding levels. The significant differences infood ration among time intervals andthe significant interaction betweentreatments andtime intervals Table 4!simply reflected theeffects ofdifferential growth ofthe turtles from differentclutches andraceways within treatment-time intervalcombinations, whichin turn influenced theadjust- ment of feedingrates. Asexpected, daily feed ing ra te d ifferedsignificantly between thehigh and low levels of feeding but not between thetwo frequencies offeeding atthe high level Table 5!.As with daily food ration, there werc significant differences infeeding rateamong time intervals anda significantinteraction between treatment andtime interval. Again, this reflecteddifferential growth ofthe turtles which inturn influenced thesubsequent amounts offood they received. Dailyfeeding ratewas reduced overtime, because theturtles require a decreased feedingrate as they grow larger Fontaineetal., 1985!. Thc analysis ofvariance fordaily feeding rate included therandom nested effect ofclutches withinraceways, because feeding ratewas recalculated ona clutch-within-raceway basisretrospectively toassess variabilitygenerated by differencesin growthfrom clutch to clutch. Survival Analysisofvariance oftransformed proportion ofsurvivors detected nosignificant differences inmain effects or interaction Table6!, so the treatments hadno significant effecton survival. Overall survival was95.8 percent forthe expcr iment. GrossFood Conversion Efficiency Grossfood conversion efficiency, C,was calculated foreach clutch ineach raceway over the intervals between wcighings as follows:

C=F/G where F =food ration in grams of dry food fed pcr turtle per day, and G =weight gain wet!pcr turtle pcr day, G wascalculated bydividing thechange ingeometric meanweight pcrturtle between twoconsecutive weighings by thcnumber ofdays inthe interval between weighings. Usually this resulted ina weightgain, but in some cases there

170 wasa weightloss so the weight change was negative. The food ration, F, was calculated by dividingthe sum of the daily quantitiesof foodfed per turtle during an interval between weighings by thenumber of daysin theinterval. Grossfood conversion efficiency does not represent actual food inta ke and assimilation, because some of thefood was not eaten. It is only an index. Cross food conversionefficiency is summarizedin Table 7. Time interval 4 produced aberrant conversion efficienciesbecause of the problemsthat beganto developin time interval 3 with the CentralSoya diet. In interval 4, growthslowed and at thelow levelof feedingsome of theclutches lost weight producing negative efficiencies, Food was in excessbecause the turtles were not utilizing it and were either growing slowly or losing weight. Therefore, conversionefficiencies in interval4 weredisregarded Table 7!, and separateanalyses of variancewere conduced for two phasesof the experiment,the first involving only the CentralSoya diet and the secondinvolving on1ythe I'urina diet Table 8!. In phase1, the gross food conversion efficiency was significantly higher at thehigh level of feedingthan at thelow level of feeding,and it varied significantlyamong time intervals Table8!. No other main effectsnor the interaction werc significant,During phase2, none of the main effectsnor the interactionwere significant,so the differences observedin phaseI apparently had no significantresidual or carry-overinfluence on thoseof phase2. Weight Gain Weightgain, G, is summarizedin Table9. Analysis of variancedetected significantly greater weight gain at thehigh levelof feedingthan at thelow,but in additionthe twice-dailyfeeding produced significantly greater weight gain than the once-dailyfeeding at the high feedinglevel Table10!. There also were significantdifferences in weight gain among time intervals. Interval 4 had the smallest average weight gain, Growth Rate Index Thegrowth rateindex, b, wascalcula tcd by linearregression analysis for eachclu tch in eachraceway. This was done separatelyfor the two phasesof the experiment,so each line wasbased on 40 paired observations0 per weighing for four consecutiveweighings per phase!.Each line wasfit ted to logarithmicallytransformed weights regressed on the square roots of age as follows: In W! = In a! + bT '"

where, W = wet weight in grams, T = agc in days from the mean hatch date, b = slope an index of growth rate!, and a = empirical constant. This growth model is similar to that derived by Caillouet et at. 986b! to describe first-year growth in weight per turtle for year-classes1978-1983 of head started Kcmp's ridlcys. The exponential equivalent of thc model is: 1!2 aebT where, c = base of natural logarithms. Growth rate index is summarized in Table 11, and analysesof variance detected significantly greater growth rates at the high level of feeding than at the low in both phasesof the experiment Table 12!. Again, the transitional and aberrant time interval 4 wasdisregarded in the analyses.There werc no significant differences ingrowth rate between once and twice daily feedings at the high level in either of the two phasesof the experiment. Environmental Variables Average seawater temperature, salinity and pH were calculated for eachof the nine raceways for eachof the seven time intervals, Analyses of variance detected significant differences among time intervals for all three environmental variables Table13!. Variatio~s in temperature,salinity and pH probablyreflected seasonal changes. Treatments had no significant cffcct and there werc no significant interactions between treatments and time intervals for these variables. Time interval 4 had the lowest temperature and salinity in the time serieswhich may have contributed to the aberrant gross tood conversion efficiencies and low or negative weight gains in interval 4,

Discussion There have been numerousstudies of growth of sea turtles of various specieson artificial diets in captivity Hildcbrandand Hastcl,1927; Caid well, 1962;Uchida, 1967; Stickney, White and Pcrlmutter,1973; Ka ufmann, 1975; LcBrun,1975;Witham and Futch,1977; Whitakcr,1979; Witzcll, 1980;Wood and Wood,1981;Nuitja andUchida,1982;

171 . Grossfood conversion efficiency, C; averagedby treatmentand time interval during the feeding ent on Kemp'sridley seaturtles of the 1984year-class,

Treatment High level Low level T1me Inclusive once/day twice/day once/day interval dates H] H2 LI Diet Ca C C Aug, 22 - Sept, 12, 1984 1.8 1.3 1.3 Central Soya Sept. 13 - Oct, 10 1.6 1.5 1.2 Oct. 11 - Nov. 7 1.9 2.1 1.7 Nov. 8 - Dec. 5 16.0 183.7 -0.3 Central Soya-Purina 5 Dec. 6, 1984- Jan. I, 1985 2.3 0.8 3,6 Purina 6 ]an 2- Jan. 30 1,8 1.8 2.2 IJ 7 Jan. 31 Feb. 27 2.6 2.0 2.1 eterminedforeach clutch within a givenraceway bydividing F,the food ration ingrams ofdry food per turtle per day giventime interval between weighings byG, the weight gain wet! per turtle per day during the interval. Daily weight s determinedforeach clutch within a given raceway bydividing thechange i r>geometric meanweight perturtle during imeinterval bythe number ofdays in the interval. Note that averages inthis table are not the same asthose one might v dividingaverage daily food rations inTable 3 bycorresponding average daily weight gains in Table 9, because C was nedfor every clutch in each raceway and for each time interval before being averaged herein. >tapplicable asa separate phase. This was a transitionperiod inwhich the diet was changed frown Central Soya toPurina fter10 November 1984 due to problems encountered with the former diet seetext!.

Table8. Analysis ofvariance ofgross food conversion efficiency ' for the two phases ofthe feeding experiment on Kemp'sridley seaturtles of the 1984year-class.

Phase 1

Source of Degreesof Mean varia tion freedom square Blocks 0.27881 088 ns Treatments, T 1.05733 3.33 ns HI 8z H2 vs. LI .72356! 543 HI vs. H2 .391 I I ! 1.23 ns Time interval, I 2 1.94480 6.13 T x I interaction 4 0.22622 0.71 ns Experimental error 16 0.31723 Clutcheswithin raceways 54 0.27108 Phase 2 Source of Degreesof Mean va ria t1on freedom square Blocks 2 0.34443 0.04 ns Treatments,T 2 8.58744 1.11 ns HI >I>rH2 vs. Ll 0.27210! 1.33 ns HI vs.H2 .90277! 0.89 ns Timeinterval, I 2 0.67243 0.09 ns T xI interaction 4 5.55306 0.72 ns Experimentalerror 16 7.73725 Clutcheswithin raceways 54 4.79946 'Cwas determined foreach clutch within a given raceway bydividing F,the food ration ingrams ofdry food per turtle per day duringa given timeinterval between weighings byG, the weight gain wet!perturtle per day during theinterval. Daily weight gainwas determined foreach clutch withing a given raceway bydividing thechange ingeometric meanweight perturtle du ring a giventime interval by thenumber of daysin theinterval. sr>s= 1>on-significar>tat P =0.05. ~ = significant at P 0.05.

172 Table9. Daily weightgain' averaged by treatmentand time interval during thefeeding experiment on Kcmp's ridley seaturtles of the 1984year-class.

Treatment High level Low level Time Inclusive once/day twice/day once/day Phase interval dates Hl H2 Ll Diet grams grams grams Aug. 22 Sept.12, 1984 1.2 I.6 1.3 Central Soya Sept. 13 Oct. 10 1.2 1.6 1.0 Oct. 11 Nov. 7 2.3 2.8 1.6 NA Nov. 8- Dec. 5 0,4 0.5 0.4 Central Soya- Purina 5 Dec. 6, 1984 - Jan. 1, 1985 1.2 1.8 0.6 Purina 2 6 Jan. 2- Jan. 30 1,9 2.1 0.9 7 Jan. 31 - Feb. 27 2.1 2.5 0,6 'Dailyweight gain was determined for eachclutch within a givenraceway by dividingthe change in geometricmean weight perturtle during a giventime interval by thenumber of daysin theinterval. bNA=- not applicable as a separatephase. This was a transitionperiod in whichthe diet was changed from Central Soya to Purina pelleisafter 10 November 1984 due to problemsencountered with theformer dict seetext!,

Table10. Analysis of varianceof daily weightgain'during the feeding experiment on Kemp'sridley sea turtles of the 1984year-class.

Source of Degreesof Mean variation freedom square Blocks 2 1.04346 1,64 nsb Treatments, T 2 12.96030 20.41 HI & H2 vs. LI ! 2,56014! 35.52 HI vs, H2 ! .36046! 5.29 Time interval, I 6 8,39384 13,22 T x I interaction 12 'I.11699 1.76 ns Experimental error 40 0.63505 Clutches within raceways 126 0.37152 'Daily weight gain was determinedfor eachclutch within a givenraceway by dividing the changein geometricmean weight per turtle during a giventime interval by the numberof days in the interval. ns = non-significant at P = 0.05. '~ = significant at P 0.05.

173 . Analysisofvariance ofgrowth rate index, b', during thc feeding experiment onKemp's ridlcy sea turtles 84 year-class.

Degrees of Mean freedom square

0.000429 0,48 nsb nts, T 2 0.010610 11.84 H2 vs, Ll .017858! 19.93 .H2 .003362! 3.75 ns ental error 4 0.000896 heswithin raceways 18 0.000752

Degreesof Mean freedom square 2 0.003689 nts, T 2 0.021075 H2 vs.LI .042056! . H2 .000093! mentalerror 4 0.001137 heswithin raceways 18 0.000904 peof the regression ofthe natural ilogarithm ofweight grams! onthe square root of age days>. Growth rate index was ed for eachclutch within a givenraceway. n-significant at P = 0.05. '+ = significant at P < 0.05. HadjichristophorouandGrove, 1983; Frazer and Schwartz, 1984;Rajagopalan, 1984;LeGall, 1985; and Lcbeau, 1986!, butfew such studies havcbeen conducted onKemp's ridIcy Caldwell, 1962; Pritchard andMArquez, 1973;Klima and McVey,1982; Caillouct andKoi, 1985; Caillouet etal., 1986b and Landry, 'I989!. Our study on Kemp's ridley was a logicalsequel tothat of Caillouct etal. 986b! because itstressed statistical design and analysis notpossible under theuncontrolled conditions ofthe previous head starting efforts involving this species. Ourresults showed clearly that the Kemp's ridleys receiving morefood exhibited thehighest weight gains and growthrateindices, asmight beexpected. However, forthe high level offeeding, theturtles receiving theentire daily rationinone feeding pcrday had a smallerdaily weight gain than those inwhich the daily ration was divided into twosepara tefeedings. Again, this was not surprising because theturtles fed twice-daily grew faster and therefore receivedmorefood based ona percentageoftheir body weight under the feeding technique developed byFontaine etak 98S>. Also, this may have been due in part to an artifact ofusing different size volumetric measures todole out feedpellets tothe turtles, with resulting difficulty inmeasuring outsmall quantities ofpellets byvolume. While this isa practicaltechnique which saves time in feeding large numbers ofturtles, it is obviously inadequate fordefinitive studiesoffood intake and growth. There was no apparent difference inthe response toonce-daily vs.twice-daily feeding atthe high level> asmeasured bythe growth rate index developed byCaillouet efal. 986b!. This was probablyducto the fact that there were fewer degrees offreedom forthe experimental errormean square inanalyses ofvariance ofthe growth rate index ascompared todcgrccs offreedom forexperimental errorinthe analyses of varianceof weight gain. Grossfood conversion efficiency wasbest lowest amount offood fed pcr unit increase inweight pcr turtle! atthe lowlevel offeeding, butonly during thefirst phase ofthc experiment duringwhich theCentral Soya dict was being used.During the second phase inwhich the Purina diet was used, the effects ofthe two different feeding levels on grossfood conversion efficiency were indistinguishable. Theshift indiet during the experiment wasan un pla nned event, a conscquencc ofunanticipated problems withthe CentralSoya dict that had not occurred inprior years ofits use. Because theshift indiet occurred sequentially, the experimentdidnot provide a comparison ofthc two diets. The main consideration wasthe health and safety ofthe endangeredKemp's ridley turtles, so the experimental protocol had to be subservient. A slowingofgrowth incaptive-reared Kemp'sridleys during winter, associa tcdwith cooling ofthe wa tcr, has been observed Caillouct andKoi, 1985; Caillouet etal., 1986b!. Lowered temperature mayaccount inpart for the slowing ofgrowth and the poorer food conversion efficiencies observed inthe experiment during time interval 4 November 8-December 5,1984! which had thc lowest average tempera tureand salinity ofany of the time intervals, Nonetheless,

174 Table13. Means ' A! andanalyses of variancefor temperature B!, salinity C! andpH D! duringthe feeding experimenton Kemp'sridley seaturtles of the1984 year-class.

A. Means Time Mean Mean Mean interval temperature, 'C salinity, ppt pH I 25.7 29. I 22.7 29.3 7.2 24,1 24,6 7.3 21.1 23.6 7.4 246 357 22.9 27.2 21.9 26.2 23,1 28.8 7.5

B. Temperature, C Source of Degrees of Mean variation freedom s uare Blocks 0.01300 0.23 ns' Treatments, T 0.07889 1.40 ns Hl & H2 vs. Ll ,08770> 1.56 ns Hl vs. H2 ,07008> 1.24 ns Time interval, I 6 19.56146 347.51 T x I interaction 12 0.00354 0.06 ns Experimental error 40 0.05629

C. Salinity, ppt Source of Degreesof Mean variation freedom square Blocks 0.12749 214 ns Treatments, T 0.01892 0.32 ns Hl & H2 vs. Ll .02388! 0.40 ns Hl vs, H2 .01397! 0.23 ns Time interval, I 6 46.91016 786.03 T x I interaction 12 0.00395 007 ns Experimental error 40 0.05968 D. pH Source of Degrees of Mean variation freedom square Blocks 2 0.00330 4.23 Treatments, T 2 0.00046 0.59 ns Hl & H2 vs. Ll Il .00086! 1.10 ns Hl vs. H2 I! .00007! 0.09 ns Time interval, I 3 0.16910 216,79 T x I interaction 6 0.00022 0.28 ns Experimental error 22 0.00078 'Based on observations taken from each raceway, but not necessarily on every day within the time intervals. Averaged over racewaysand daysof observationwithin time intervals, = insufficient number of observations. 'ns = non-significant at P = 0.05. ~' = significant at P < 0.05.

175 temperaturewas better controlled in 1984than in previousyears. Theexperimental feeding IevcIs and frequencies hadno apparent ef'feet onsurvival, and overall survival during theexperiment wasvery high 95,8 percent!. Therefore, theturtles that received theleast food showed nogreater mortality than thosethat receivedthe largeramounts based on our results. Ourresults have considcrabIe practical significance tothe head starting oflarge numbers ,000 to 2,000 pcr year! ofKemp's ridleys in captivity, Twice as much laboris required tofeed the turtles twice per day than to feed them once pcrday. The turtles seem tobe opportunistic feeders and will eat as many pellets asprovided them until they are satiated. Additional food is wasted and causes problems in deterioration ofseawater quality Fontaine etal., 1985; Caillouet etal., 1986b!. Perhaps a singIe feeding perday would be adequate if thefeeding rate were some what higher thanthe low feeding rate but lower than the high feeding rate used in our experiment, Feedingrate expressed asa percentageofbody weight provides a handy rule of thumb for feeding large numbers ofKcmp's ridleys inmass production head starting, However, itsuse results inheavier feedings forfaster-growing turtles.Some clutches exhibit better growth than others, With the typical mixture ofseveral clutches ina single raceway,thepractical question arisesas towhetherornot thefeeding rate should bcadjusted according tothe average weightof a sampleofturtles representing theraceway orsamples from each clutch within the raceway. Our conventionalpractice has been to feedby raceway,basing the feeding rate on the average weight of turtles representingallclutches in the raceway. Todo otherwise would be impractical. However, this restricts growth in fastergrowing clutches andmay result inoverfeeding ofslower growing clutches. Theproblem canbc Iesscned by puttingclutches ofequal age in a raceway,butthisdoes not accommodate genetic differences thatmight affect growth performance of different clutches. Acknowledge1nents Thiswork was conducted under Mexico's lnstituto Nacional dcla Pesca INP! Permit 2534, U.S. Fish and Wildlife Service FWS! Endangered andThreatened Species Permit PRT-676379 andTexas Parks and Wildlife Department TPWD!Scientific Permit SP073. Kcmp's ridlcy eggs were made available tothe National Park Service NPS! through theefforts ofRennes Marquez Millan INP, Mexico!, Jack Woody and David Bowman FWS, Albuquerque, N.M.!,Pat Burchfield Gladys Porter Zoo, Brownsvillc, Tex! andtheir staffs. Robert King NPS,Corpus Christi, Tex ! andhis staff providedthe"imprinted" hatchlings fromthese eggs, The V.S. Navy transported someof the hatchlings toGalveston. Theassistance ofMarcel Duronslet, Dickie Revera, Christine Olguin, julie Rarnas, Denise Daley, David Smith, AusbonBrown, Jr.,Zoula Zein-Eldin, Medardo Olivarez, FredMattes andJohn Von Cannon wasgreatly appreciated. HEART Help Endangered Animals - Ridley Turtles!, a special committee ofthe Pincy Woods Wildlife Society of NorthHarris County College, chaired byMrs, Carolc Allen, Houston, Tex., provided thefood for the1984 year-class of Kemp'sridleys.

Literature Cited Caillouct,C,W., Jr.1984. Essai dcprevention del'extinction dela tortue de Kcmp. Les Carnets deZoologic Bulletin ofthc ZoologicalSociety of Quebec!44!:28-24. Caillouet,CW,,Jr., CT. Fontaine, TD.Williams, SA. Manzella, AM. Landry, Jr.,KL. lndelicato, MJ.Duronslet andDB. Rereva. 1988.Can we save Kemp's ridlcy sea turtle? Believe it or not!, p. 2043. In: Peterson, K.H. Editor!, 10th international HerpetologicalSymposium onCaptive Propagation andHusbandry, Zoological Consortium, lnc.,Thurrnont, MD,206 p. Caillouct,CW.,Jr,, CT, Fontaine, TD,Williams, SA.Manzella, DB.Rcvera, DB.Koi, K LW.Indelicato, MG.Tyrce, JK.,Lcong, M.J,Duronslet andK,T. Marvin. 1986a. TheKemp's ridley sea turtle head start research project: anannual report forfiscal year1985. NOAA Technical Memorandum NMFS-SEFC-174, ii plus37 p,, 29 Tables, 1 Figure and5 Appendices. Caillouct,C.W.,Jr. and D.B. Koi. 1985. Patterns andvariability infirst-year growth inweight ofcaptive-reared Kemp'sridley sea turtles:agraphical analysis. NOAA Technical Memorandum NMFS-SEFC-164, i plus4 p.and 52 Figures. Caillouet, C.W,, Jr., D.B. Koi, C.T, Fonta inc,T.D. Williams, W.J.Browning andR.M. Harris. 1986b. Growth andsurvival ofKemp's ridleysea turtle, Lepidocheiyskempi, incaptivity. NOAA Technical Memorandum NMFS-SEFC-186, liiplus 34 p., 12 Tables and 7 Figures. Caldwell,D.K.1962. Growth measurements ofyoung captive Atlantic seaturtles intemperate waters,Los Angeles County Museum,Contributions in ScienceNo. 50, 8 p. Clary,J C.ll! and J K.Leong 1984. Disease studies aidKemp's ridIcy sea turtle head start research, Herpetological Review15! 69- 70. Fontaine,C.T.,K.T. Marvin, T.D. Williams, W,J. Browning, R,M. Harris, K.L.W.Indelicato, G.A.Shattuck andR.A, Sadlcr, 1985. The husbandryofhatchling toyearling Kemp's ridley sca turtles Lepidochelys kernpi!.NOAA Technical Memorandum NMFS- SEFC-!58, iv plus34p., 10 Tables,22 Figures and 2 Appendices. Fontaine,C T.,T D, Williams, S A.Manzclla, M G.Tyree andC W.Caillouet, Jr.1989. Kcmp's ridley seaturtle head start operations

176 of the NMFS SEFCGalveston Laboratory. This volume.! Frazer,N.B. and F J.Schwartz. '1984. Growth curvesfor captiveloggerhead turtles, Car etta caretta, in North Carolina,USA. Bulletin of Marine Science 34 !:485-489. Hadjichristophorou,M. andD.J. Grove. 1983. A studyof appetite,digestion and growth in juvenilegreen turtle Cheloniatnydas L.! fed on artificial diets. Aquaculture 30:191-201. Hildcbrand,S.F. and C. Hastel.1927, On the growth and behaviorof loggerheadturtles in captivity. Proceedingsof the National Academy of Sciences13:374-377. Kaufrnann,R. 1975.Observaciones sobre el crecimicntode tortugasrnarinas en cautividad.Caldasia 113!:139-'I50. Klima, E F.and J P.McVey. 1982. Head starting the Kemp's ridley turtle, Lepidochelys kernpi, p. 481 487. In: Bjorndal,K A. Editor!, Biologyand Conservation of SeaTurtles, Proceedings of the World Conference on SeaTurtle Conservation, Smithsonian Institution Press,Washington, D.C,, 583 p. Landry, A.M, Jr. 1989.Morphornetry of captive-rearedKcmp's ridley seaturtles. This volume.! Lcbeau,A. 1986.Un essaide grossissementen captivitede la tortue vertc Cheloniamydas realise a Tahiti PolyncsicFrancaisc!. Review Trav. de la Institute Pechcs Maritime 48 and 4!:143-154. LeBrun,G. 1975. Elcvagc a laReunion de juveniles de la tortucvcrtc Chelonia mydas Linnaeus! 1758. Science et Pcchc,Bulletin de la Institute PechesMaritime No. 248. 25 p. LeGall,J.-Y. 1985. Elevage de la tortue vertemarine a I'ilede la Reunion978-1985!, La PccheMaritime, July-August 1985, p. 434- 440. Mrosovsky,N. 1983.Conserving sca turtles. The British Herpetological Society, c/o theZoological Society of London,176 p. Nuitja,I.N.S. and I. Uchida.1982. Preliminary studies on thegrowth and food consumption of thejuvenile loggerhead turtle Caretta carettaL.! in captivity. Aquaculture 27 157-160. Pritchard,P.C.H. and R. MarquezM. 1973,Kemp's ridlcy turtle or Atlanticridley. International Union for theConservation of Nature and Natural ResourcesMonograph No.2, 30 p. Rajagopalan,M. 1984.Studies on the growth of oliveridlcy, Lepidochelys olivacea, in captivity, Central Marine Fisheries Research Institute Bulletin 35, p. 49-54. Stickney,R R,,D B.White and D. Perlrnutter, 1973, Growth of greenand loggerhead sea turtles in Georgiaon natural and artificial diets, Bulletin of the Georgia Academy of Sciences31:37-44. Sokal,R.R. and F.J.Rohlf. 1981.Biomctry, Second Edition, W.H. Freemanand Company,New York, 859p. Uchida,I. 1967.On the growth of the loggerheadturtle, Carettacaretta, under rearingconditions. Bulletin of thc JapancscSociety of Scientific Fisheries 33!:497-507. Whitaker,R. 1979.Captive rearing of marineturtles. Journal of the BombayNatural History Society76!:163-166. Witham, R. and C.R. Futch. 1977.Early growth and oceanic survival of pen-reared seaturtles. Hcrpctologica 33!:404409. Witzell, W.N. 1980.Growth of captive hawksbill turtles, Eretmocheiysimbricata, in westernSamoa. Bulletin of Marine Science 30!:909-912. WoodJ.R. and F.E.Wood. 1981. Growth and digestibilityfor the greenturtle Cheloniamydas! fcd dietscontaining varying protein levels. Aquacultu rc 25:269-274.

177 HealthCare and Diseases of Captive-RearedLoggerhead and Kelp's Ridley SeaTurtles JorgeK. Leong,David L. Smith,Dickie 8. Revera,Lt. John C. Clary III, Donald H. Lewis,Janis L. Scottand Anthony R. DiNuzzo»

During1977to1983, Kemp's ridley Lepidochelys kempi! and loggerhead Caretta caretta! sea turtles were reared forone yearor less at the Natt'onal Marine Fisheries Service, Southeast Fisheries Center, Galveston Laboratory, Manyof the diseases, malformationsandinjuries observed in captive-reared seaturtles were named and classified into27 major categories. The categorieswere:sudden hatchling death syndrome, papillary dermatitis, emaciation, focalerosive dermatitis, injuries from aggressivebiting,focaldermal granulosis,scolecobasidiosis,white-suturesyndrome,yolksac mycosis,internalnodularmycosis, hypernecroticwarts,malabsorption ofyolk sac, urolithiasis, duodenal ulceration, hemorrhagic bacteriosis, mycobacterial pneumonia,swo!Len-eye, intussusception, curved-back, softshell, coelomic edema, lung aplasia, flipper malformation, cross-beak, congenitalblindness, intestinal prolapse, andprolapse ofthe urinary bladder. Each category ofailment isdescribed withrespect toetiology,symptomatology,occurrenceandsuggestedremedyifknown. Current levels ofourknowledgeofdiagnosisand control ofdiseases during captive-rearing ofthese two species ofturtle are discussed. Recommendations aremade on perspectives and needsin seaturtle pathologyresearch. Numerouskinds of diseases andphysical injuries occurred inKemp's ridley Lepidochelys kempi!and loggerhead Carettacaret ta! sea turtles reared for one year or lessat theNational Marine Fisheries Service NMFS!, Southeast FisheriesCenters Laboratory inGalveston, Tex,Rearing Kemp's ridleys from hatchlings toyearlings isa feasibility studyof the head starting concept, part of a broaderconservation program aimed at preventing extinction ofthis criticallyendangered species Klima and McVey, 1982!. Informationonnesting sites and population decline inKemp's ridley has been summarized inseveral papers Klimaand McVey, 1982; Mrosovsky, 1983;Caillouet,1984; Caillouet etal., 1986; Fontaine etal.,1985!. Torecapitulate, thereisonly one known primary nesting beach located near the village ofRancho Nuevo, inthe State ofTamaulipas, Mexico,Thenu mber ofnes ters at tha t beachhasdeclined from an estimated 40,000 reported tohave nested ina single daym June1947 tosome 1,200 in1974 and then to 500 or so in 1977, Itis hoped that head starting will prove useful asone among several methods tohelp preserve and augment the population. Inhead starting, the turtles are reared in captivityfrom hatchlings toabout one year or less of age, then tagged and released into the Gulf of Mexicoor adjacent estuaries. Togain experience before attempting to rear Kemp's ridleys, the Galveston Laboratory staff obtained 1,160 loggerheadhatchlings, a species which had not yet been listed as threatened under the Endangered Species Actof 1973.Experimenting firstwith the loggerheads proved beneficial, because almost allof the loggerhead hatchlings becameill at some time during head starting, sothe staff at the Galveston Laboratory wasable to derive important informationonsea turtle pathology andhealth care methods without jeopardizing Kernp's ridleys. Only about 9 percentofthe loggerheads survived after a 10-monthrearing period. The knowledge, experience andinformation gainedonprophylaxes, diagnoses andtherapeutics forloggerheads wereapplicable inpart to Kemp's ridleys in subsequentyears.Pathology studies and observations onloggerheads andKemp's ridleys contributed greatly tothe highsurvival rates ofsix year-classes 978-1983! ofhead started Kemp's ridleys: 68percent, 83percent, 95percent, 88percent, 89 percent and 77 percent, respectively Fontaine et al., 1985!, Thispaper isa reviewofsome ofthe diseases andinjuries suffered bycaptive loggerhead andKemp's ridley sea turtlesatthe Galveston Laboratory, Italso describes remedies orprophylaxes thatwe have found useful inresolving or preventingsome of theproblems. The diagnostic, prophylactic, therapeutic and health-care methods that we adoptedrepresent thebest approaches thatwe could undertake, given the constraints onfunding, personnel and otherresources. Inaddition, wewere constrained byour U.S. Fish and Wildlife Service Permit, which prohibited deliberatesacrificing orinjuring oflive, normal sea turtles. To secure baseline, in-depth pathological information essentialtodevelopment andrefinement ofdiagnostic, prophylactic andthera peutic methods willrequire controlled experimentsnecessitating sacrificing or injuringsome test animals. ' Leong,Letterman ArmyMedical Center; Smith, Revera andCtary, National Marine Fisheries Service;Lewis, Texas A&M University; Scott,Victoria, Tex.; Di Nuzzo, The University of Texas Medical Branch

178 Withfew exceptions, there was a paucityof documentedinformation on diseases and causes ofdeath in seaturtles andefficient remedies, especially those applicable to mass-culturesituations such as ours, Theinformation in this papermay be useful both to thoseinterested in rearingsea turtles from hatchling to yearlingstages and those concernedwith their morbidity and mortality. Materials and Methods Sources of Hatchlings In September1977, 1,160 loggerhead hatchlings were supplied by theFlorida Department of NaturalResources, JensenBeach, Fla. Claryand Leong, 1984!. Approximately 200 loggerhead hatchlings were received from the same sourcein July1978, Kemp's ridley hatchlings were obtained each summer, beginning in 1978,through the joint efforts of the Instituto Nacional de la Pescaof Mexico, U.S. Fish and Wildlife Service,National Park Service,Texas Parks and WildlifeDepartment and Gladys Porter Zoo in Brownsville,Tex. Fontaineet al,, 1985!. The number of Kemp'sridley hatchlingsreceived annually at theGalveston Laboratory for 1978-1983were 3,081, 1,846, 1+22, 1,865, 1,524 and 250, respectively Fontaine et a!., 1985!, Rearing of Turtles Turtleswere reared at the HeadStart Research Project facilities of the GalvestonLaboratory until they were7 to12 monthsold. Theywere then tagged and released offshore at preselectedlocations off Texas,Florida and Mexico Klima and McVey, 1982;McVey and Wibbels,1984; Wibbels, !984; Fontaine et al., 1985!, Two majorconcepts of maintenanceand rearing were tried by theGalveston Laboratory: ! free-contactbygroup- rearingin racewaysor tanks,and ! individual-rearingin isolation Clary and Leong 1984; Caillouet, 1984!. The first conceptwas implemented in 1977and 1978. Hatchlings were divided into groups of about50 to more than 550 turtles, dependingon thesize of rearingtanks, and the hatchlings were allowed to swimfreely in contactwith oneanother. A varietyof holdingvessels was used,including: concrete raceways, fiberglass-lined wooden raceways and fiberglasstanks, Seawater in the concreteraceways was either recycled non-flow-through!to removewastes via a biodisc Mock, Rossand Salser,1977!, or wasconstantly replaced by flow-through.Oyster-shell filters were usedto remove wastesfrom recycled seawaterin the fiberglass-line9,wooden raceways.Static water in the fiberglass racewayswas replacedthree times per week. Initially,the incidence of disease,traumatic injuries and death was very high, indicating inadequacy of thegroup- rearingmethod. Laboratoryand tanksideobservations revealed that hatchlingsfaired better when isolatedfrom one another in small containers and when the culture water was replaced frequently with clean seawater, Not only were traumatic injuriescaused by mutual biting and seratching eliminated, but morbidity and mortality werealso greatly reduced. Basedon thesefindings, isolation-rearingwas implementedin January1979, after the 1978year-class of Kemp's ridleys had been reared for six months. The isolation-rearingsystem Klima and McVey, 1982;Fontaine et al,, l985! was employedin the samekinds of holding tanksalready described. Each hatchling was raisedin isolationfrom othersin a 10-literplastic bucket, with holes in the bottom to allow seawater exchangeand removal of fecesand surplus food. The waste materials would fall through the holes to the tank floor and would be elimina ted when the tanks were drained during routine cleaning operations.The seawater wascompletely changed daily forrearing the1978year-class,but the frequencywasreduced to two to threetimes a weekfor subsequentyear-classes. Raw fish flesh was the primary food used for rearing the 1977year-class of loggerheads.Feeding of the Kemp's ridleys hasbeen described by Fontaineet al. 985! and McVey,Leong, Wheeler and Harris unpublishedmanuscript on the culture of young Kemp's ridley seaturtles!. Kemp's ridleys were fed a synthetic,floating, pelleteddiet Fontaineet al,, 1985!.In general,two mealsper day were provided,one in early morning and one in late afternoon, Thedaily amountof pelletedfeed ranged from roughly five percentof wetbody weight for hatchlingsto roughly 1.5 percentof body weight for yearlings. Maintenance of Sick or injured Turtles In general,sick or injured turtleswere removed from their rearingtanks and maintained separately for observation and medication unfll recovery. Two different systems, collective isolation and individual isolation, were used for holding sick or injured turtles. Collectiveisolation was used for someof the sick loggerheadturtles of the 1977year-class, Two or more turtles having similar signs or symptomsof illnesswere held together in onecontainer. The number of animalsplaced in each containerwas dependent on thesizes of containerand animals. In general,no lessthan 300 square cm of seawater surfacearea were provided for every 100square cm of turtle carapacearea. Water depth was at least twice the t >Ick nessof thebody of theturtle, except in thespecific situation in whichan animal was unable to raiseits head above therva ter due to physicalweakness, In thelatter cases, shallower water was used to preventdrowning, The kinds of

179 containersused for collectiveisolation of sickor injuredturtles included 2,100-liter circular fiberglass tanks .83 m insidediameter, I.D.!, 190-Liter rectangu]ar fiberglass tanks 0.83 m x0,5 m upper-rimdimensions! x 0,5 m deep,and 19-liter circular plastic tubs .46 m I.D.!. In individualisolation, used for both loggerhead and Kemp's ridley turtles, each sick or injuredturtle was held singlyin a separatecontainer, Containers included 4-liter glass jars, 10-liter plastic buckets, 19-liter tubs, and 190-liter rectangulartanks. When there were only a fewsick turtles, the containers were placed on the floor or bench tops, and seawaterwas manually replaced daily. However, within the 1978 yearwlass ofKemp's ridleys, there were routinely from800 to 1,000sick or injuredturtles to be cared for ata time.Under these circumstances, 10-liter buckets with holes in theirbottoms were suspended in a concreteraceway, and each contained a sick or injuredturtle. Each day the seawaterwas drained, the buckets and turtles were washed with fresh water from a hose,the turtles were given the appropriatemedication, and the raceway was refilled with cleanseawater to theappropria te level, This method also reducedthe Labor required to cleanindividual buckets manually, Feedingsick turtles varied accordingto the natureof the illness.The animals were either fed food similar to that givento healthyturtles, or theywere given a specialdiet such as a drybreakfast cereal rice or wheat!or commercial babyfood usedfor humaninfants!, supplemented in some instances with rnultivitamins.When semi-solid food was used,it wasfed to the turtle via intraesophageal intubationby means of plastic tubing at tached to a graduatedmedical syringe.In general,sick animals had reduced appetites. Therefore, the amount of food wasabout one-half to one-third of normalfeeding rate. On manyoccasions, sick turtles were not fedat all becauseof thenature of certainillnesses suchas suspected intestinal perforation. Tosustain the life of non-feeding oranorexic turtles, subcutaneous injection ofa steriledextrose-multivitamin solution 0 percentdextrose+ 10 percent multivitamin infusion concentrate, U.S. Pharmaceutical!wasadministered atthe rate of 025 rnl per 50 rng wet body weight per day. Daily injection continued u~tilthe animal regained its appetite or foras long as seemed necessary. Drug Administration At leastfour methodsof administeringtherapeutic chemicals to sickor injuredturtles were studied as follows: I. Formedicated baths the calculated dose of a water-solubledrug or antibioticwas added to theseawater. The turtlewas then placed in themedicated seawater for a periodof time,either predetermined or onan as-needed basis,depending on the nature of theillness or injury,the progress of recovery,and the kind of medication used.Examples of medicated baths were formalin 0 to100 ppm!, potassium perrnanganate to5 ppm!, znalachitegreen ,1 to 0.2ppm!, Cutrin-Plus Applied Biochemist, Inc,, Wisconsin; 0.3 ppm copper!, minocyclinhydrochloride .3 to0.6 ppm!, erythromycin to 4 ppm!,and furanace Dainippon Pharmaceu- ticalCo., Osaka, japan; 0.05 to 0,1ppm!, 2. Fororal administration a drug was fed to theturtle either through intraesophageal intubation or by incorporatingit into the feed or agar carrier. For the latter, two methods were tried. The first was to spray by meansofan atomizer a predetermined volume ofsolution ofthe drug onto a premeasuredquantity ofdry pelletedfood. The second was to mix the drug with a onepercent solution of heat-meltedBacto agar DIFCO! whichhad cooled to 45-50'C, After solidification, the medicated agar was diced and offered to the sick turtles, It wasnecessary thatthe agar pieces floated on the surface forthe turtle to bite and swallow them. If theydid notfloat, the quantity ofagar in the solution was reduced until the diced pieces would float, In a Liquidmedium ofa givensalinity, diced pieces having a lowerconcentration ofBacto agar will float more readily than those havinga higherconcentration. Nevertheless, both methods of incorporatingdrugs into the food were discontinuedinfavor of intraesophageal intubation, because effectiveness ofdrugs in the feed or agar matrix was unknown, Examplesofdrugs used in oral administration thequantity of active ingredient for each 50 g wetbody weight ofturtle per day is indica ted in parentheses! wereampicillin sodium,0 5 mg!, chloramphenicol palmitate oral suspension,2.5mg!, quinacrine hydrochloride .35 mg!,chloroquine phosphate .5 to3 mg!,minocyclin hydrochloride,2 to 0.4 mg!, neomycin .75 mg!, and ketaconazole mg!. The daily drug dose was divided intotwo equal portions, one given in the morning and the other in the afternoon. Inintraesophageal intubation, thedrug solution was delivered bymeans ofa hypodermicsyringe fitted with plastic tubing inserted into the esophagus of the turtle. 3. Parenterai injection ofsmall loggerhead andKemp's ridley turtles was accomplished bysubcutaneous injec- tionof the drug. Until 1980, the site of injection was the suprafemoral pouch area immediately above the hind femuron theventral side of thetorso. Thereafter, the dorsal side of theneck was favored because it was discoveredthat sodiumdia trizoa te, a commonlyused excretory, urographic, iodina ted contrast material, was absorbedmore rapidly in thedorsal neck than in thesuprafemoral pouch area in subcutaneousinjection McLellanand Leong, 1982!. These findings implied a similardifferential absorption oftherapeutic drugs in thetwo locations. Examples of drugs quantityof activeingredient of drugfor each25 grams wet body weight of turtle per day is indicatedwithin the parentheses!used in parenteraladministration were ampicillin sodium,0 25mg!, chloramphenicol sodium succinate 63-1.25mg!, pennicillin G 00-2500 units!, gentami- cin 06mg!, terramycin .6 mg!,and amphotericin B. Administration of amphotericinB wasstarted at 0 0125 mg per100 g bodyweight per day, with weeklyincrements of 0,0125mg per 100g perday. 4. Topicalapplication of certaindrugs or chemicalsto minorsurface wounds or lesionswas accomplished with sterilecotton swab.Examples were neosporinointment BurroughsWellcome!, fungizone lotion Squibb!, tincture of iodine, merthiolate Lilly> and gentianviolet. In our experience,gentian violet was ineffective againstInany external lesions, and after it leachedinto the seawater it appeared to be toxic to smallha tchlings, especiallywhen applied too often. Neosporin appeared to beeffective against certain presumptive bacterial lesions. The effects of the other topical agents have not been determined, Diseases, Maladies and Injuries Sudden Hatchling Death Syndrome Suddenhatchling death SHD! syndromeinvolves deathof hatchlings,usually overnight, without forewarning signsor symptoms.On rareoccasions, lethargy in hatchlingsup to threeweeks old maybe observed prior to death. This disease,which affectsboth loggerheadand Kemp'sridley hatchlings,appeared in the former aboutone week after they arrived at the GalvestonLaboratory in September1977. The disease was so devastatingthat within four weeksafter the first hatchling died, about40 percentof the entire captivepopulation was lost. Etiology Our currenthypothesis assumes that SHD syndromeis the result of microbial infectionmagnified by poilu ted culture seawa ter, Pollutioncould haveoccurred in theclosed non-flow-through!,recycled scawa ter system due to thegradual build-up of solubleorganics from excessfood and turtle wastes.Such conditions probably favored multiplication of pathogenicmicroorganisms, and at the sametime createdundue stresseson the turtles, thereby lowering their resistanceto infection.Evidence to support the abovehypothesis came from laboratoryanalyses and tankside observations, Bacteria were isolated from blood taken from moribund and freshly dead Kemp's ridley and loggerheadhatchlings, and the incidenceof SHD syndromewas found to declinesharply from about four percent to about 0.3 percent or less per day> and almost immediately after the hatchlings were transferred from polluted seawater to clean seawater. Bacteria isolated from hatchlings showing SHD syndrome were of many varieties. One of them, Clostridium bifermentans,was derived from blood cultures of onemoribund and onefreshly dead loggerhead. This samebacterium has been isolated from casesof gas gangrene in humans Smith, Conant and Overman, 1964!. Another isolate, Vibrio alginolyticus,which is a commonmarine organism frequently associated with diseasesin marineanimals, was also obtained from blood culture of the same freshly dead loggerhead. The presenceof these two bacteria in moribund or freshly dead turtles has suggestedtheir possible role in causing SHD syndrome. Howeve".,more experimental studies will be required to confirm their true pathogenicity. Other bacteria isolated and identified from post-mortem cultures were Achromobacterpinnatum, Ac. delmarvae, Aeromonasformicans, Aer. punctata,Aer. shigelloides,Citrobacter Escherichia!freundii, C. intermedius,Escherichia aero- genes,Pseudomonas sp. a pathogen of the Atlantic croaker, Micropogoniasundulatus!, gamma Streptococcussp. not group A, B or D!, Proteus mirabihs and Vibrio algosus.All of these except Escherichiaaerogenes, C. intermedius, Achromobacterpinnatum, Ac. delmarvaeand Vibrio algosushave been associatedwith infections in humans, terrestrial animals or aquatic animals Smith et al., 1964;Jawetz, Melnick and Adelberg, 1972;Buchanan and Gibbons, 1974!. Moreover, C. freundii has been associatedwith a fetal infection, known as septicemic cutaneous ulceration disease SCUD! in turtles, Kaplan, 1957;Campbell and Busack, 1979!. However, the symptomatology described for SCUD bears little resemblance to SHD syndrome. Determination of pathogenicity of bacterial isolates in relation to SHD syndrome will require more study. At one time during the 1977 epizootic in loggerhead hatchlings, poisoning from the following sources was considered a possible mortality factor due to: remnants of stale, uneaten fish flesh left in the raceways, and dense blooms of a variety of algae in the closed system raceways. Stomachsof dead loggerhead hatchlings were frequently packedwith fish fleshand algae,However, experimental feeding of healthyha tchlings with stalefish fleshand algal massescollected from the racewaysand with food-packedstomachs of dead turtles producedno ill effects, Signs and Symptoms Although there are usually no observable external signs and symptoms associatedwith SHD syndrome, careful observations in an epizootic may show a few very weak hatchlings up to three weeks of age that float passively on the water surface with their front flippers hanging loosely downward. In normal hatchlings, the front flippersare folded backover thecarapace during rest. Lethargicanimals usually die overnight.Slow-dying animalsdisplay no externalabnormal signs, so the lethargiccondition is consideredpart of the SHD syndrome.

181 Necropsyofhatchlings with SHD syndrome frequently reveals a feed-packed stomach and fecal impaction in the lower bowel. Paralyticileus is indicated. Occurrence An epizootic ofSHD syndrome occurred inSeptember 1977 when newly arrived loggerheads were aboutone week old. The loggerheads weremaintained in concrete raceways with recycled non-flow-through! seawa ter. Within four weeks, more than 400 ha tchlings about 40 percen t!had died. Peak mortality reached about four percent per day toward the end of the fourth week, In 1978,Kemp's ridley hatchlings were introduced and reared in fiberglasstanks and basins with daily changes ofseawater non-flow-through!. InJanuary1979, dueto outbreaks ofvarious diseases andbiting injuries, the ridleys werctransferred toindividual buckets suspended inthe tanks, and seawater was completely replaced atleast three timesa week.About two percent of the3,081 ridleys died from SHD syndrome during the 10 months of rearing. In allsubsequent annual ba tches of Kemp's ridleys reared singly in suspended buckets, mortality attributable toSHD syndromewasnegligible, estimated atless than one percent forthe 1979 year-dass andvirtually non-existent inyear- classes 1980-1983. A conditiondesignated as"carly hatchling mortality" in greensea turtle Chelonia mydas! hatchlings isthought to bean early manifestation of the grey-patch disease Haines, Rywlin and Rebel, 1974!. This latter disease has been attributedto a herpesvirus infection Haines et al., 1974; Rebel, Rywlin and Haines, 1975; Koment and Haines, 1977!. Thesudden death nature of "earlyhatchling mortality" appears toresemble that of SHD syndrome. However, we havenot observed signs ofgrey-patch disease inridley or loggerhead hatchlings. Therefore, wedo not consider "early hatchlingmortality" and SHD syndrome to be thesame. Remedy There is no established treatment for SHD syndrome, The key to its prevention seems to be use of non- pollutedculture seawater. Frequent exchange orreplacement withfresh seawater bymanual draining and refilling of tanksis requiredto ensuregood water quality in therearing tank. Experiments have been conducted toward developingflow-through andrecycled water systems, butnone of these has proved practical for rearing Kcmp's ridleys in our situation. Chlorination pprn chlorine! of seawaterin the reservoir tanks before transferring the seawater to the turtle holdingtanks was tried in early 1978 for prophylactic purposes, Theresults were not encouraging. Incidence ofSHD syndromeandmany other kinds of diseases inthe chlorinated seawater didnot appear appreciably different from thosein control non-chlorinated!seawater. Therefore, chlorination is not consideredan effectivemethod of prophylaxis,These observations suggest that the causal organisms ofmany infectious diseases were introduced into theseawater afterit was transferred tothe raceways andto other rearing tanks from the reservoir tanks. Apparently, thepa thogens multiplied under favorable conditions inthe culture seawa ter after the residual chlorine disappeared. Papillary Dermatitis Papillarydermatitis PD!, formerly called papillary eruption Leong, 1979!, affects skin tissues. Lesions occur aroundthe eyes and the anus and on the limbs and the plastron, Although hatchlings ofboth Kemp's ridley and loggerheadcanbe affected, the disease was more prevalent inthe latter species than in the former when reared in our laboratory. Etiology PDis responsiveto antimicrobial chemicals, and is thereforepresumed to bea microbialinfection. Bacteriaisolated from eye and anal lesions have included Aeromonas formicans, Vibrio alginoiyticus, V.algosas and Pseudomonassp. Rebeletal. 975> mentioned thepresence ofpapillar tissues invirus-caused grey-patch disease ofgreen sea turtles. Electronmicroscopic work performed inour laboratory oneye lesions ofPD in loggerhead hatchlings didnot show the presenceof- viruses. Signsand Symptoms Small, papilla-shaped projections thatdevelop around the eyes and anus and on the skin andthe plastronic suture are off-white, yellowish-tan orlight-tan incolor. When occurring around the eye, multiple papillaeusually protrude from underneath theupper and lower eyelids and from the rear corner ofthe eye-socket. Inadvanced cases, thepapillae may fuse into a crustwhich may spread tocover the whole eye, blocking eyesight. At theanus, thearea surrounding theanal aperture usually becomes paleand swollen andassumes theshape ofa ring bearingrows of smallpapillae. In advancedcases, crusts will formin theaffected area. Localized,small papillae also occur singly on the limbs, protruding through gaps between adjacent scales onthe surfaceofthe flippers. These flipper papillae donot form crusts. Similar single papillae can develop onthe ventral sideof the turtle's body, projecting through sutures between plastronic scutes. Currently eye, anal, flipper and plastroniceruptions are collectively considered under the name of onedisease, PD. PDis frequently accompanied byprogressive emaciation ofthe turtle. Since only one or the other condition alone alsois found in someturtles, it is not known yet whether PD and emaciation arerela ted. Emaciation isdiscussed in moredetail as a separatemalady later in this paper.

182 PD progresses slowly, It takes many days for minor eruptions around the eyes and cloaca to become serious, as manifestedby crustiness,Papillae on the flippersand the plastronicsutures usually remainsolitary and non-fusive. Occurrences PD was first observed in one-month-old loggerhead hatchlings in early October 1977after they were transferred from non-flow-through raceways to flow-through concrete tanks at the Galveston Laboratory's East Lagoonfacility on the northeasttip of GalvestonIsland. The diseasespread relatively fast. Within two months, practically every one of the more than 600 juvenile loggerheads held at the East Lagoon facility had contracted the disease.During the same period, 120 similar loggerhead juveniles were maintained individually in the Galveston Laboratoryin 4-liter glassvessels for experimentalpurposes. Despite daily replacementof the culture seawater, approximately 15 percent of them developed PD, In 1978,when new crops of loggerhead and Kemp's ridley hatchlings were group-reared in closed raceways in which seawater was replaced daily, less than two percent of each speciesdeveloped PD, although other diseases flared. Remedy The prognosis in PD is good when not complicated by emaciation. Although PD alone does not seem fatal, the blocking of eye-sight by necrotic tissues crust! may affect the turtle's ability to find food, a point not yet proven. However, when PD is complicated with emaciation, the survival rate for affected young turtles is estimated as 20 percent or less, even with medications. Formalin baths 0 to 100ppm for seven days with daily replacement of medicated seawater! combined with daily subcutaneousinjections of a onepercent w/v! sodium ampicillin solution for 10 to 21 days has proven to be an effectivetreatment for PD,The lesionsusually clearup within four weeksafter medication,For unknown reasons, neither formalin bath nor ampicillin injection alone seemsas effective as when combined in treating PD. Ampicillin for injection is prepared in either sterile distilled water or a sterile solution containing 10 percent v/v! Multiple Vitamin Infusion Concentrate U.S. Pharmaceutical! and 10 percent w/v! dextrose. The latter preparation is intendedprimarily for an emaciationcomplication, The dosage for injectionis 0,25mg ampicillin, or 0,025ml of the ampicillin solution per 25 g wet body weight per day, preferably divided into two equal doses, one given in early morning and the other in late afternoon, The whole daily dose may bc administered in a single injection, although this may be less effective. Ampicillin, if unavailable, may be substituted with the less effective chlorarnphenicol sodium succinate, The recommendeddosage is 1.25mg chloramphenicolper 25 g wet body weight per day. Inconclusiveresults were obtainedwith the following chemicalsand anflbioticsin clinical testsagainst PD in loggerheads:malachite green, methylene blue may be toxic as a bath!, gentian violet may be toxic to very young hatchlings!, penicillin G, oxytetracycline,minocycline, erythromycin, and furacin. The routes of delivery of thesedrugs includedsubcutaneous injection, medicated bath and topical swabs. Emaciation Emaciation is a wasting syndrome with various underlying causes in young Kemp's ridleys and loggerheads. Affectedturtles are weakenedand their healthdeteriorates progressively until death. From the first discerniblesigns of emaciationuntil death,the whole processmay last up to two monthsor more.In general,itappears that theyounger the turtle, the more susceptibleit is to emaciationand the moredifficult is its recovery.Circumstantial evidence has indicatedthat loggerheadhatchlings are morevulnerable to emaciationthan ridley hatchlings. Emaciationin animalsis thenet resultof dehydrationor depletionof body tissuesor both. Thereare many possible causes.Examples are microbial infection,metabolic or catabolicdysfunctions, inappropriate diet and involuntary reductionin intake of food, water or both. The last causecan be dismissedas a factorcontributing to emaciationof turtlesat theGalveston Laboratory, because plentiful food and seawaterwere always available to the animals,but all the other causeshave been considered as potentialcontributers to emaciation. A bacterium,Mycobacterium ruarinum, was isolated from the lungsof two deadand severelyemaciated loggerhead hatchlings,the carcassesof which werekept frozenfor over onemonth at O'C prior to necropsy,It waspossible tha t emaciationin those turtles was a result of lung infection by the bacillus. In necropsy,small fecal impactionwas frequentlyobserved in the lower bowel of emaciatedspecimens indicative of constipationprior to death. Reichenbach-Klinkeand Elkan 965! suggestedthat faulty nourishmentcombined with lack of exercisemay produce constipation in captive turtles. Thus, faulty diet could have been a possible causeof constipationin loggerheadsof the 1978year-class, leading to a form of emaciationwhich gave the hatchlingsa shrivelledbut hardenedor block-likeappearance. These turtle hatchlingswere fed a dietcomposed exclusively of commercialfeed MasterMix S.S. Turtle Feed9349! which might havebeenunsuitable for them,because within five months95 percent of the head startedloggerheads became emaciated and many died, Loggerheadhatchlings of the 1977year-class, which were fed raw fish fles during their early stageof life, did not contracta similarform of emaciation.Neither Kemp'sridleys at the GalvestonLaboratory nor olive ridleys L. oh'vacea! Lyle Kochinsky,Nova University, Dania,

183 Fla.,personal communication!, appeared to havea similarproblem when fed with thesame pelieted food, Lackof exercisewas an unlikely cause of constipationin theGalvesto~ turtles. Fecal impaction of thebowel was observedin necropsywhether the hatchlings had been allowed to swimfreely in largerearing tanks with ample opportunities to exercise or had been confined in individual buckets. Therehave been reports of emaciation and its causes inother turtles and reptiles. An epizootic among 2 000captive- rearedgreen sea turtle hatchlings at CaymanTurtle Farm 983!, LtdGrandCayman, British West Indies, resulted intypically flat, weak and emaciated hatchlings, and was caused by a coccidianparasite, Caryospora sp. Rebel, Rywlin andUlrich,1974!. Marcus 977! reported tha t reptilesinfected with intestinal amoebiasis exhibited non-slxmific signs of decreasedactivity, progressive weight loss and anorexia although they also exhibited diarrhea, vomiting and excretionofbloody stools, Although emaciation has not been mentioned, it is reasonable toexpect that in prolonged amoebiasis,emaciation eventually would follow. No evidenceof similarparasitic infections has been found in the Kcmp'sridleys and loggerheads head started at theGalveston Laboratory. Weightloss and emaciation in turtles at the Penrose Laboratory ofthe Zoological Society of Philadelphia have been attributedto inanition Cowan 1968!, which is physiologicaldysfunction due to insufficiencyof nutritional factors necessaryforhealth and well-being. Berklow 977! classified nutritional deficiency aseither primary or secondary. Primarydeficiency is dueto inadequatenutrient intake, and secondary deficiency is a resultof failureto absorbor utilizenutrients, increased nutritional requirements or excessive excretion. Digestion and absorption may be disturbedby gastrointestinaldisease, and utilization and storage of nutrientsmay be impaired due to endocrine dysfunction,inborn errors of metabolism,severe infection or degenerative disease. Experiments will berequired to determinewhether the turtles reared at the Galveston Laboratory suffered from primary or secondary deficiency or both. CaptiveKemp's ridley hatchlings suffered a fatal form of systemic mycosis with the fungus Paecilomyses isolated asa presumptivecausative organism see Yolk-sac Mycosis!, At first glance,such affected hatchlings appear emaciated,but carefulexamination will showthat they do notpossess the typical lean-and-wrinkled-neck character- isticof theemaciation syndrome as defined in thissection. Instead, the infected turtles usually exhibit taut cervical skinwith goodmuscle tone. Therefore, mycosis is not considereda causeof emaciation. Si~s audSymptoms Sickturtles gradually lose weight and have a sunkenplastron and a wrinkledand lean- lookingneck. In advancedcases, anorexia is common,the vertical plane of thebody trunk becomes thin and looks flatteneddue to severe sinking of the plastron, and the turtle becomes very weak and lethargic. Quite often the pliable skinofthesuprafemoralpouchbetween thebaseofthehind femurandtheshellsagsand failstoassumea taut, domed shapeas in normalturtles. In somecases, the body shell may become softer and more pliable than normal. On other occasions,the whole torso may appear relatively hard, solid and shrivelled as in constipated loggerheads of the 1978 year-class,which were suspected of having been fed a faultydiet. X-radiography shows that in "soft-shelled" individuals,the ribs in thecara pace are poorly developed, Necropsyshows that emaciation could be associatedwith oneor morekinds of internaldisorder such as fecal impactionof bowel most common!, mycobacterial pneumonia, intestinal perforation and peritonitis. In fecal obstructionofthe bowel, the lower colon is packed with hard feces, while the rectum isusually void. Growthofha tchlings after recovery from emaciation isoften stunted, thea nimals retaining a dwarfed physique and havinga weakand unhealthy appearance, A few recovering turtles also mamtain a rather soft and pliable shell, Occurrence Emaciation first appeared inearly October 1977 inone-month-old loggerhead hatch]ings maintained in outdoor,flow-through raceways. The condition was frequently concurrent with papillary dermatitis PD! describedearlier in this paper. Emaciation, either uncomplicated orcomplicated with PD, continued tospread when theloggerheads weretransferred toan indoor, non-flow-through raceway inDecember 1977in anticipation ofcolder weather.The seawater in theraceway was recycled through an oyster-shell filter bed. Emaciation soon reached epizooticscale, affechng more than 300 of about 600 hatchlings within two months, By April 1978, about 80 percent ofthe turtles had exhibited emaciation atone time or another, and many of them died. Meticulous clinical care and thedevelopment of new therapeutic methods helped some about nine percent! emaciated turtles recover. Kemp'sridlcys and loggerheads ofthe 1978 year-class were mass-cultured in non-flow-through raceways. An estimated10 to 1Spercent of theridleys became emaciated, often in associationwith PD,and died within thefirst nine months.The loggerheads didwell in the first five months, but later 95 percent ofthem developed the hardened and "block"form of emaciation with no contemporary PD.Necro psy showed tha t mostof the emaciated loggerheads had hardfeces inthe colon. Nevertheless, thesedata should notbe taken toimply that Kemp's ridleys were less susceptible toemaciation than loggerheads until more conclusive data from controlled experiments are obtained, because the ridleysof the 1978 year-class were reared insolitude inindividual buckets inthe raceways starting from the eighth month,while the loggerheads continued tobe kept under conditions of non-isolated, group rearing. The isolation cultureofsea turtle hatchlings hadbeen found to play an extremely important role in disease prevention, although thereason is unknown. Emaciation occurred only sporadically in Kemp'sridley hatchlings of subsequentyear-classes kept under isolation rearing. Conditionsresembling emaciation have been reported in otherturtles and reptiles. At thePenrose Research Labor- atory,among 1,249 cases of reptilianmortality including 92 turtles and , 60 percent were characterized bya wastedcondition, depletion of fatdepots and fragile tissue and skin Cowan,1968!. In thespring of 1973,2 000captive greenturtle hatchlings a thet Cayman Turtle Farm were ill anddescribed astypically flat, weak and emaciated Rebel et al., 1974!. Althoughno emaciated turtle hatchlings have been noted among those shipped to us, other workers have reported thaton occasion newly acquired turtles may be received ina severelys tarved and dehydrated condition Campbell andBusack, 1979!. The age of these turtles was not specified. The emaciation syndrome encountered in the Galveston Laboratorywas not shipment-related, Remedy Current prognosis in emaciationis grave because of difficultyin diagnosisand lack of fundamental understandingof the etiology. Despite treatment of symptomsand general clinical care given to sickturtles, the recoveryrate hasbeen less than 20 percent, Toa largeextent, emaciation seems preventable by isolation-rearing.Forexample, in Kemp'sridley ha tchlings of the1978 year-class, anestimated 10 to 15 percent of theanimals contracted fatal emaciation while being held in free- swimminggroups in raceways.Since then, solitary rearing has been the standard, and incidence of emaciationhas decreasedcorrespondingly. In the 1980year-class, the incidenceof emaciationwas lessthat two percent.For loggerheads,there have been no comparable data for both isolation and non-isolation rearing within raceways, but in 1977loggerhead hatchlings that were isolated in individualcontainers under laboratory conditions were rarely afflictedwith emaciation,while about80 percent of thosegroup-reared in racewayssuffered from thesyndrome. In clinicalcare, emaciated turtles should be individually isolated as soon as possible. The culture seawater should becompletely replaced every day. When only a fewanimals have to be cared for, and staff is sufficient, it maybe wise to disinfectthe containerwith sodiumhypochlorite every time the usedseawater is replaced.This disinfection proceduremay not be practical, and therefore omitted, when a largenumber of sickturtles is involved.Resources permitting,non-corrosive and autoclavable containers may be used, so that they can be steam-sterilized. It ispreferable tofeed sick hatchlings thin slices of frozen-and-thawed rawfish flesh in lieuof pelleted commercial feed,Freezing and thawing can destroy some potential pathogens, and freezing can preserve the nutrients in thefish fleshin storage.The daily ration should notbe excessive: three percent ofbody weight or anamount which the sick turtlescan totally consume in an eight- hour work day. Animals which do not eat may be fed a semisolidcereal diet bymeans of intraesophageal intubation. Occasional starving of anorexic animals for 24 hours does not appear to harm them.Either independently or in combinationwith intubationfeeding, subcutaneous injection of life-sustainingdoses of a sterile dextrose-multivitamin solution is also recommended, Intubahon feeding, injection or both Inay be continued until the animals regain appetite. In clinicaltests, injections of antibioticssuch as chlorarnphenicol, ampicillin, oxytetracycline and gentamicin did notproduce significant therapeutic effects, nor did bathtreatments with formalin,methylene blue, chloramphenicol, minocycline either alone or in combinationwith erythromycin!or gentamicin.The recovery rate from emaciation was no better in medicated turtles than in non-medicated ones. A numberof catharticsincluding warm water,light mineraloil, milk of magnesiaand Metamucil psyllium hydrophilicmucilloid; Searle Laboratories! have been tested in attemptsto correctconstipation in emaciatedturtles. Thesecarthartics were administered either orally by intraesophagealintubation or asan enema as appropriate. Only milk of magnesiaat 002Sml per100 g bodyweight per day per Os occasionally induced a slightlevel of defecationin two out of sixjuvenile Kemp's ridley patients. There are unconfirmed reports that castor oil appearsto beeffective in inducing defecationin someconstipated Kemp's ridley hatchlings. Focal Erosive Dermatitis Focalerosive derma titis FED!is characterized by theprogressive disintegration of cutaneoustissues and form a iont of shallow,erosive lesions which aresometimes covered with crustsof necrotictissues. The disease may affect the skin or surfacetissues of manybody parts, such as the eyelids blepharal FED!, head cranialFED!, carapace carapacial FED!,flippers flipperFED! and nose nasal FED!. An additionalform which erodes the edge of a flipperis named flipper-edgefocal erosivedermatitis Flipper-edgeFED!. Althoughthe lesions of FEDare sometimes rather unsightly and alarming, there has been no evidencethat the diseaseby itself is fatal. Etiology FEDis presumablycaused by bacterialinfection. Several bacteria have been isolated from lesionson theeyes and carapace of Kemp'sridleys, and these lesions can be cured by theuse of antibacterialdrugs. Injury to theskin is perhapsa predisposingfactor for bacterialinvasion. Skin injuries may be causedby mutual scratchingorbiting among hatchlings that are placed in closecontact group-rearing! with each other. Also, human handlingofthe turtles during routine maintenance ortranshipment maycontribute toskin injuries. The supposition of aninjury-infection is supported by theobservation that there was a rapiddecline in incidenceof FEDwhen the group-rearingapproach was replacedby solitary rearing, Thebacteria isolated from eye and carapace lesions have included Aeromonas formicans, Vibrio anguillarum, V, a ginolyticusandCitrobacter freundii. One or more of these bacteria may be the causahve agent contingents! upon furtherpathogenicity studies. Stickney, White and Perlmutter 973! reported isolation ofan Aeromonas sp.from superficiallesions inthree and a half-month-old loggerhead hatchlings. Inreference tothis findings, Sinderman 977! commentedthatAerornonas spp.were common inhabitants ofaquatic environments, andmight be facultatively pathogenicto animalsliving under conditionsof environmentalstress. Ananaerobic bacterium, Bacteroides sp.,has been considered asthe causal organism ofa skindisorder in one-or two-month-oldloggerhead hatchlings Witham, 1973a!. The disease has been described to producenecrotic, spreading,nonwalled skin lesions causing most of the hatchlings todie within a weekof thefirst appearance ofthe malady.It is dubious that this disease isthe same as FED, since FED does not seem to be fatal by itself. Signsand Symptoms Except inca ra pacial FED, freshly formed lesions have not been observed. Relatively fresh lesionsof carapacialFED have an ivory color due to exposure ofunderlying ground substance of the shell after the dark-coloredepidermis erodes away. Older lesions are greyish-tan oryellowish-tan, usually shallow, localized erosionsofthe skin. In blepharal FED, it appears that theeyelids become puffy and discolored ina pre-erosionphase, but this observation remains to be confirmed. TypicallyFED forms shallow lesions below the skin level due to the loss of surface flssues. Thelesions areusually irregularinshape, spreading outin all directions andproducing anuneven margin. They are variable insize, ranging in diameterfrom about 2 mm,when first noticeable onhatchlings in the raceway, tomore than 1.5 cm. Often as neighboringlesions spread, they become merged, producing a large necrotic patch, Besidesthe typical FED lesions on the skin surface of theflippers, there is a conditionin which the entire thickness ofpart of the edge of a limbis eroded away, giving a worm-eatenappearance. Because thecondition involves the invasionofdeeper rather than just surface tissues, it maybe a differentform of disease. However, until further studies, it isincluded under FED for the purpose ofthis paper, The name flipper-edge FED is given to this abnormal form of FEDto distinguish it from the typical, skin-surface type of flipperFED. Usuallythere islittle if anybuild-up ofcrust in cranial, carapacial andflipper FED lesions orin flipper-edge FED lesions.Inblepharal andnasal FED, crusts may be present. Insevere cases, anoverabundance ofcrusts may cause eithertotal or partialclosure of theeyelids or blockingof thenostrils. Occurrence Inearly October 1977, approximately 600four-week-old loggerhead hatchlings ofthe 1977year-class weretransferred fromclosed concrete raceways toflow-through concrete tanks, Within two weeks, many of the youngturtles began toexhibit signs of cranial FED, Within the following two months, anestimated 80percent ofthe populationcontracted cranial FEDof either severeor limited extent. Duringthe summer of 1978,3,081 newly emerged Kemp's ridlcy hatchlings were acquired and reared in close contactin closed-system,fiberglass raceways. Theseturtles were extremely aggressive, chasing andbiting each other vigorously. Within three months, practically everyturtle was afflicted with eitherone or moreforms of FEDat onetime or another. Immediateisolation ofsick ha chlings into individual vessels coupled with drug medication generated excellent therapeuticeffects. H owevcr, whenabout 2 000 cured turtles were returned tothe mass popula tions inthe raceways, theywere re-infected. Beginning inJanuary 1979, the isolation-rearing method see section Rearing ofTurtles! was implemented,Theincidence offocal erosive dermatitis rapidly fell. To this date, FED no longer constitutes a health threatto captive Kemp's ridleys reared inisolation fiom each other during their first year of life, Remedy Isolation ofindividual sick tur ties and daily replacement ofculture seawater coupled with subcu taneous injectionofampicillin sodium! at0 25mg per 25 g wetbodyweight per day is effective treatment forFED, particularly theblepharal andflipper forms. A seven-day formalin bath 0 to100 ppm! isalso effective against theflipper-edge formof FED. Lesions are usually healed in about four weeks following onset of medication, Thebest prophylaxis against FED is to raise the hatchlings bythe isolation-rearing method asdescribed in the section Rearing of Turtles, Self-limitingorspontaneous healing of lesions frequently occurs in carapacialFED when the afflicted turtles are individuaUyisolated andprovided with clean sea water daily. The healing process maylast up to two months ormore. Administrationofampicillin does not seem to shorten the wound-healing timein carapacialFED, as it doesin blepharaI FED, orhave harm ful effects from ampicillin been observed. Sinceblepharal, flipper and carapacial FED oftenoccur simul taneou slyin a singleKemp's ridicy, both ampicillin andisolation treatments aregiven tosuch turtles having a mixture of different forms of FED.

186 Cranial and nasalFED do not respondto medicationand isolationas well as do blepharaland carapacialFED. Cranial and nasallesions can bc very persistent.Some cases have failed to respondto clinical treatmentsfor four monthsor more.Occasionally, a lesionmay cvcn expandwhile the turtle is under clinical care. A conditionthat appears to resembleFED and that isdesignated as focal necrosis of skinhasbccn reported in young greenand loggerhead sca turtles Witham,1973b!. However, unlike FED, focal necrosis of skinhas been described ascausing death of the affectedanimals unless treated with potassiumpcrrnanganate g per 220liters in bath treatment!.Thcrcforc, FED and focal necrosisof skin are probably diseasesof a different nature. In short-term observations,application of potassiumpermanganate bath doesnot cure FEDin Kemp'sridleys. Aggressive Biting Aggressivebiting is includedin this paperbecause it happensextensively among ha tchlings of Kemp'sridley and frequentlycauses serious physical damage in thebittenturtles. There is no evidence thataggressivebiting contributed directlyto thedeath of a turtleat the Galveston Laboratory,but apparently the inflicted wounds provide a convenient portal of entry for pathogenicmicroorganisms. Further, it maybe reasonedthat if a turtlewith a missingeye or part of a flipperwere released into thesea, chances of survivalwould bereduced, because such a handicappedturtle would beeasy prey for largermarine animals. Therefore, a standardpractice in thehead start project has been not torelease such injured sea turtles. It hasnot beenconclusively established that Kemp'sridleys older thanone year do not bite eachother, but evidence hasindicated that they at leastare not as aggressiveas individuals oneyear or lessin agc. Etiology Aggressivebiting apparentlyis an intrinsic behaviorin young Kemp'sridleys. Why suchaggressive behavior is so i~tense in captive Kcmp's ridlcy hatchlings but not in loggerhead hatchlings under similar environ- mental conditions is unknown. What actually stimulates Kemp's ridley hatchlings and juveniles to bite onc another is also not understood. Color and movement are two presumptive stimuli, but controlled experiments are needed to evaluate this hypothesis. Si~s and Symptoms Almost everypart of the body of a Kemp'sridley turtle that is accessibleto the mouth of anotherKemp's ridley is vulnerableto injury throughbiting. Themost vulnerablesites arc the flippers, the cdgcof the shell, the neck and the head. In serious cases,a large piece of body tissue or a portion of a limb may be tom off, leaving behind large fleshy wounds. Occurrence Aggressive biting began when captive Kcmp's ridlcys were about three weeks old. Among the 3,081 ridlcy ha tchlings acquired in late July 1978,more that 800 7 percent! showed signs of bite wounds by mid-August, Bite-wound incidcncc continued to mount and affected more than 80 percent of the turtles by January 1979. Afterwards, all the hatchlings were individually confined in plastic buckets in isolation-rearing and therefore werc prevented from biting each other. Surprisingly, when solitarily confined ridlcys were returned to free-contact and group-rearingraceways upon recoveryfrom discaseor injury, they showed a higher level of aggressivebiting behavior than before they were isolated. Minor chasing and biting were observed in hatchling loggerheads, but the turtles suffered no apparent injuries. Ha tchlings of the green sca turtle sccm even more docile. In summer 1978,no aggrcssivc biting was observed by the senior author among approximately 100 young green sca turtles averaging 250 g each, held in an outdoor concrete tank at Miami Scaquarium, Miami, Florida. Remedy Individual isolation-rearing is the best protection against aggressivebiting and wounding in Kemp's ridlcy, but it is labor-intensive, especially in a mass-culture situation such as that at the Galveston Laboratory. Unfortunately, no better rearing system has been developed to date for Kcmp's ridleys, Injured hatchlings are immediately isolated in individual vesselsthat receive daily changesof seawater.For wound treatment, a pieceof sterile surgical spongeor a cottonswab is used to absorb the water from the injured area, fallowed by cleansingwith hydrogenperoxide three percent!on a cottonswab, When the wound is dry, it may be treated topically with a general antiseptic such as neosporin ointment or merthiolatc, and the turtle is then rcturncd to the holding vessel. In serious injuries, ampicillin sodium> may be administered subcutaneously at a doseof 0.25 mg per 25 g body weight pcr day. Wounds usually heal in I'our to eight weeks.Me healed surface will resume the normal dark color, but the bulk of thc missing tissue is not replaced. Focal Dermal Granulosis Focal dermal granulosis FDG! produces localized lesions in the form of discolored and of ten grainy patchcs in the skin of young sca turtles. The discaseis not fatal, and the affected turtles are active and continue their normal growth. Etiology A bacterium, Pseudomonasalcnligenes, was isolated from the deeper tissues in focal dermal granulosis lesions which had been surface-sterilized with 70 percent alcohol and tincture of iodine. Further studies are needed to determine the pathogenicity of that organism, Si~s and Symptoms Discolored, localized lesions that have a granular appearancecan bc found on thc surface

187 ofthe shoulders and the axilla of thefront flippers. Lesion color ranges from light grey to gn.yish-tan, often overlaid by a thin network of bright yellow substance. Thelesions are flat and slightly elevated. They are usually more or less oval-shaped in the neck, but irregularly shapedinother affected skin areas. When first noticed on hatchlings in the raceway FDG lesions are usually about 1 cmor more in length or diameter, depending ontheir shape. Such lesions conhnue toexpand in time.An axillary lesionmay spread to coveran irregular,area 2 cmor morein diameter. OccurrenceAn estimated 2 to 3 percentofloggerhead and Kemp's ridley hatchlings contracted FDG in 1977and 1978under group-rearingconditions. However, no deathswere attributed to the disease. Witham973b! reported a fatalcondiflon called focal necrosis ofskin in tank-reared green and loggerhead turtles. SinceFDG is non-fatal,we assumethat it is a differentdisease from focal necrosisof skin. Remedy Staticformalin bath 0 to 100ppm! continued for five to sevendays is effective. The medicated bath wateris replaced completely ona dailybasis. Occasionally, signs of healing appear after only threedays ofmedication. Careshould be taken not to over-extend thetreatment period, Over-treatment willdelay instead ofpromote healing. Potassiumpermanganate bath ppm! was tried, but the resultswere inconclusive, Occasionally,FDG is self-healingwhen the affected turtle is keptisolated in cleanseawater. The FDG lesion sheds a scaband leaves behind a discoloredscaron the skin. The healing and scab-shedding processes arepromoted with formalin bath treatments. Scolecobasidiosis Scolecobasidiosisis an infectioncaused by a fungusbelonging to the genusScolecobasidium. At least two kindsof scolecobasidialinfection have been observed: scolecobasidial pneumonia SP!, an infection of thelung, and scole- cobasidialosteomyelitis SO!, an infecti~n ofthe bone. An infected turtle may live a Iongtime before it dies.!n some casesof SP,progress of thedisease may arrest spontaneously under the right environmental conditions, and the animalcontinues tolive with only one functional lung and with abnormal swimming and floating patterns; i,e. tilted swimmingand side-floating. Both SP and SO have been encountered only in youngKemp's ridleys and not in loggerhead hatchlings. Etiology Thefungus Scolecobasidium constrictum has been observed repea tedly in and isolated from tumor-like orcyst-like spherical bodies inaffected lungs and rear flippers of Kemp's ridleys. The same fungus has been observed orisolated ina limitednumber ofcases ofscoliosis and inflamed shoulder joints, Cultures ofsynovial fluid taken from inflamedshoulders were negaflvefor bacterialinfections. Sometimesother fungi such as Paecilomyces sp.,Pencil lium sp, and Cephalosporium sp,are also recovered in cases ofSP, Their role in such infection has not been determined. Penici ilium and Cephalosporium areprobably contaminants. Paecilomycessp.is also often isolated from smaller tissue nodules asdescribed under Internal Nodular Mycosis. Signsand Symptoms In themajority estimated 97 percent! of SPcases, an infected turtle swims or restson the seawatersurface with its body tilted to one side, This side-floating syndrome isusually the first sign that the animal mayhave contracted SP.Another frequently observed gross sign is bulging ofthe dorsal front half of the carapace of thebuoyant side of a tiltedturtle into a minorhump. Side-floating andhumping are usually not observed until the animalsarefour months old or older, when the disease isin a relativelyadvanced stage. Methods forearly diagnosis arenot known, It should be noted that side-floating alone isnot specific for SP. Turtles having lung aplasia described separatelyinthis paper! also exhibit a similarsign. Jacobson etal. 979! reported isolation ofthree fungi, Sporotrichum sp,,Cladosporium sp.and Paecilomyces sp.,from infected lungs in green turtles which exhibited tilted swimming behavior, Necropsiesofturtles that have died from SP generaly show consolidation oftissues ofone of the lungs into a relativelylarge spherical cyst.This tumor measures asmuch as2 cm or more indiameter andmay be empty orfilled withblood ora clearfluid. Tissue consolidation appears tooccur more frequently atthe frontal portion ofthe lung. Theinterior wall of anempty sphere may be dark-green, but if thesphere is filledwith fluid, the wall is linedwith coagulatedblood. Also residing in thewall are fungal mycelia and spores from which the fungus S. constrictum is invariablyisolated. Another fungus Paeciiomyces sp.is alsosometimes isolated. Generally,only one of thelungs develops a largespherical body cmor morein diameter!.Smaller solid tissue spheresornodules ranging from pin-head size to 1 cmin diameter may also be present inone or both lungs, in the liverand occasionally inthe cardiac auricles. These smaller nodules cannot bedistinguished fromthose caused by otherfungal organisms such as Paecilomyces sp.Therefore, unless they are confirmed as scolecobasidial nodules throughtheactual isola tion of Scolecobasidium theyare categorized asInternal Nodular Mycosis forthe purpose ofthis paper. Imagesofinfected lungs can be seen inX-radiographs, butthe diagnosis isnon-specific, Also,X-ray can mveal that a non-affectedlung may expand over to theother side of thespinal column.

188 SO affectsthe bonesof the hind flippers and perhapsthe vertebraeand the shoulderjoints. Tissuesadjacent to infectedbones in a flipperswell to forma tumor.X-ray shows that the phalanges, the metatarsals, the tarsals and the tibia may be erodedto a greateror lesserextent. There are indications that SO can cause scoliosis of the vertebral column as well as swollen and stiffened shoulder joints,X-ray shows the formation of a curvaturein the spine and sometimes there is erosion in someof thevertebrae. S. constrictumhas been recovered by culturing tissuestaken by biopsyadjacent to the affectedvertebrae. In correspondenceto curvature of theinternal spinal column, the ridge of thecarapace is also curved, thus affording an externalmeans of detectingscoliosis. The specific relationshipbetween scoliosis and scolecobasidiosis remains tobe determined. Whena turtleexperiences an inflamedshoulder joint, the affected front flipper showsstiffness and difficulty in swimming.On closerexamination, the shoulderjoint is swollen,and X-ray may show erosion of the headof the humerousat theglenoid fossa. S, constrict um sometimes can be recovered through culturing the affected tissues as in scoliosis. Occurrence With theexception of oneunconfirmed case in a youngloggerhead in 1981,all casesof scolecoba- sidiosishavebeen observed in Kemp'sridleys. SP was by far themost frequent of thetwo formsof scolecobasidiosis in Kemp'sridleys with 7, 18,4, 23,8 and6 casesin the1978 to 1983year-classes, respectively. Since SP is confirmed onlythrough isolation of thecausal fungus or by actualobservation of thecharacteristic tumefaction of thelung at necropsy,and since we could not performdetailed laboratory analysis on every turtle, there may have been cases of SP that escapedour detection. Therewere threecases of rear-flipperSO and two casesof scoliosisin the 1979year-class of Kemp'sridleys, One of the scolioticcases was associated with arrestedSP in a Kemp'sridley that hasbeen kept in captivity for thamyears becauseof fungalpneumonia, but therewas no signthat the infection had spread to thevertebrae, In theother case of scoliosis,X-ray showed lesions in a fewvertebrae. There were at leasttwo casesof inflammedshoulder joint diagnosedin the 1979year-class of Kemp'sridleys, Casualobservations suggested a direct relationshipbetween the incidenceof SPand lower environmental temperatures,especially when there were frequent fluctuationsbetween cold and warm temperaturesduring winter, Remedy In clinical trials, fungizone Squibb;containing amphotericin 8 asthe active ingredient! was injected subcutaneouslyinto sixyoung ridleys, averaging about 300 g inbody weight and suffering from advanced SP. The startingdosage was 0,0125 mg amphotericin8 per 100g wetbody weight per day, with weeklyincrements of 0.0125 mg per 100g. Four turtles died after one to two monthsof medicationindicating that the therapy was ineffective. Therefore, treatment was terminated. Other formsof chemotherapywere also tried, but found to be ineffective.These methods included formalin bath, malachitegreen bath, and oral ketoconazole dissolved in dimethyl sulfoxide>.In general,evaluation of chemother- apy for SPwas difficult becausethere was lack of specificinternal monitoring and diagnosticmethods, It is possible that someof the antifungal drugs are effectiveagainst scolecobasidiosis if applied at an early stageof infection. Therefore, research should be done to discover and develop early diagnostic and monitoring methods. ln 1978and 1979,most ridley turtlestha t werea fflictedwith SPeventually died. Occasionally, the infectionin some turtles seemedto be arrestedspontaneously without medication,Upon discoveryof turtles with signsof SP,these turtles were isolatedquickly and were given a warm6' to 28' C! cleanseawater environment. In 1982,eight live Kemp'sridleys of the 1981year-class were diagnosedas having presumptiveSP. After receivingmore than three months, non-medication, clinical care, six 5 percent! of the animals survived, and the SP infection in them appeared to havebeenarrested. This 75percent recovery rate wasan exceptionrather than the norm,In our experience,the rate of spontaneousrecovery of SP infection does not exceed20 percent.Nevertheless, in the absenceof effective chemotherapy, the key to spontaneous recovery of SP-strickenKemp's ridleys is immediate isolation and provision of a clean,stable and warm 6 to 28'C! seawater environment. It is not necessaryto reduce the feeding level unless the animal becomes anorexic. The isolation-rearingmethod was also effective against in flamedshoulder joints. After aboutthree months solitary rearingin two turtles with SO,the swellingand stiffnessat the shoulderjoints subsidedand the turtles wereable to use the flippers freely againin swimming. There is no effective treatment for scolecobasidial infection of flipper bones and scoliosis. Surgical removal of in- fected soft tisues in two casesof rear-flipper SO followed by daily oral administration of ketaconazole mg in dimeth- yl sulfoxide per 100 g wet body weight per day! was ineffective. Both turtles died within one month. The two turtles with scoliosis also died. At this point, the best strategy seemsto be prevention. Basedon our experience,it appears that avoidance of sudden changes of ambient temperature from warm to cold would reduce incidence of scolecobasidiosis.

189 White-suture Syndrome White-suture WS! syndrome involves the whitening and broadening intoa narrowribbon shape ofthe suture lines betweenscutes ofthe carapace, There are two ldnds of WS syndrome: ! dull white-suture DWS! syndrome and! shinywhite-suture syndrome.SWS! These two kinds of WS syndrome notonly are morphologically different butalso responddifferently tothe same medication, It is most likely that their etiology, which remains tobe determined, is alsonot the same. In thispaper, they are grouped under the same heading for convenience. Etiology The etiology has not been determined foreither form of WS syndrome. Fusarium-like fungal spores have beenobserved inthe white ribbon on the suture line in the DWS syndrome. Confirmation ofthe pathogenicity ofthis fungus will requirefurther studies, Signsand Symptoms In DWS syndrome, thecarapace sutures widen into a narrowribbon shape and assume a dullwhite or greyish-whitecolor. The edges of the"ribbon" are smooth and non-undulate. Transformation and discolorationofthe sutures are gradual, starting out with a fewsutures and slowly spreading tothe others, In severe cases,all suturesare affected. Microscopic examinations of ribbon materials reveal a mixtureof debris,bacteria, protozoansand Fusarium-likefungal spores. InSWS syndrome, thewhite ribbon over the transformed sutureis shiny orglistening witha bluishtintand appears slimy.The edges ofthe ribbon are uneven and undulate, Microscopically, theribbon material iscomposed ofdebris, bacteriaa ndprotozoans. Occurrence DWS syndrome occurred very commonly inyoung loggerheads ofthe 1977 year-class. Instatic seawaterinraceways, more than 70 percent ofthe turtles contracted DWS syndrome. Inloggerheads andKemp's ridleysofthe 1978 year-class, lessthan 10 percent ofthe young turtles contracted either DWS orSWS syndrome, and sincethen either syndrome was encountered only sporadically, Thedecrease inWS syndrome could have been due tothe progressive modification ofturtle rearing methods, resulting ina significantimprovement ofseawater quality the culture tanks, Remedy DWS syndrome responds well to formalin bath 0-100 ppm! treatments carried out for three to seven dayswith daily replaceme~t ofmedicated seawater. One week or more after termination of the bath treatment, the ribbon-likewhite material, which by then has turned grey, detaches from the affected sutures. The healed sutures on thecarapace resume a normal appearance. Formalin bath is not effective against SWS syndrome. Fromall indications, bothDWS and SWS syndromes seemto be preventable through theprovision ofa sanitary sea wa ter environmen t. Yolk-sac Mycosis Thisdisea seis called yolk-sac mycosis YSM! because thecausal fungus invades theyolk-sac inside thebody cavity ofneonate Kemp's ridleys. Most infected hatchlings dieyoung, usually within the first month after hatching. Etiology A fungus,Paecilomyces sp.,has been frequently observed inor isolated from infected yolk-sac tissues, Thisfungus is a presumptivecausal organism for yolk-sac mycosis. Signsaud Symptoms Hatchlings usually die very young with little external manifestation ofdisease. Therefore, itis very hard todetect infected livehatchlings ina mass-culture situationsuch asthat at the Galveston Laboratory. Atnecropsy, theinfected yolk-sac isusually found to be hardened intoa blockwith the internal yolk material turninginto a friablemass, Vascular congestion isoften prominent onthe yolk sac surface. Microscopic examination ofdisea sedyolk ma terials show thepresence ofabundant fungal hypha e and spores charac teris ticof Paecilomyces. The funguswhich can be isolated from infected tissues willgrow on Sabouraud dextrose agarsupplemented witha three- salt solution at 26 C. Occurrence YSM was first observed inhatchlings ofthe 1980 year-class ofKemp's ridleys. Since then, the disease occurredineveryyear-class through1983. There were 7,59, 5,and 3 casesinthe 1980 to1983 year classes, respectively. Youngerhatchlings appeared to succumbmore readily to thedisease than older ones. Remedy There isno known therapy forYSM, Currently, thedisease isonly discovered atnecropsy, toolate to providea remedy even if one were available. Methods areneeded forearly detection anddiagnosis. Moreover, informationisneeded onthe source andmode ofinfection toguide possible development ofprophylaxis. Internal Nod ular Mycosis Internalnodular mycosis INM! refers toformation ofnodules inan internal organ asa consequenceofmycotic infection.Thenodules aresolid and relatively small incontrast tothe larger, spherical, hollow cysts formed inthe lungsinadvanced stages ofSP. Itis possible thatsome ofthe nodules inINM are early lesions ofSP, and some are causedbyPaeciI omyces infection, Grossly, INM nodules areindistinguishable withrespect toetiological agent. The onlyway to distinguish themis detection ofthe specific fungus inlaboratory analyses suchas culturing thenodular tissuetoisolate the microorganism ordirect microscopic examinations oftissue specimens. Unfortunately, such diagnosticprocedures aretime-consuming andlabor-intensive, especially when the nodules are numerous and severalorgans are infected. Tocompound theproblem, sometimes Scoiecobasidium constrictum, Paecilomyces sp.and perhapsother fungi such asCephalosporium sp.and Penicillium sp.,which are thought tobecontaminants, arecultured fromthe same nodule. Therefore, precise diagnosis is not easy when many turtles are infected and when resources arelimited. In lightof suchdifficulties, we do not attempt at this time to distinguish the different kinds of rnycotic nodules,but groupthem as one disease category, INM. INMappears toprogress slowly in a hatchlinghostafter initial infection, The net result of INM infection isusually death of the host. Etiology Thefungi S. constrictum and Paecilomyces sp.have been frequently isolated from nodular tissue specimens,These two fungi are presumptive causal organisms, Other fungi such as Cephalospori u msp., Pen ciilium sp, andAspergillus sp.are found occasionally. Pending further studies, weregard them as either secondary pathogens or contaminants. Signsand Symptoms Hatchlings afflicted with INM do not always display external manifestations, Those that dousually but not always show an a trophiedbody trunk, which resembles a solid hard block and is thickened atthe horizontalplane i.e., from carapace toplastron!, sothat the whole trunk appears bloated. The neck, however, isnot affectedas in emaciationsyndrome; i.e., it remainsrelatively full with goodmuscle tone and its skinis taut,not wrinkled.An a fflictedhatchling often,but not always, floa ts higher in theseawater than a normalone, and its activity rangesfrom normal tolethargic. Appetite forfood maybe reduced. Infected turtles usually continue tofeed until they becomevery weak or approachdeath. Externalsigns are nonspecific forINM, Definitive diagnosis isdependent upon observing fungus-infested nodules in internalorgans. Unfortunately, techniques forsuch observations have not yet been firmly established forlive hatchlings,X-radiography coupled with in vitroculture and isolation of fungusfrom biopsied tissues provides a promisingapproach to diagnosisof thismycosis. Manymajor visceral organs or tissuescan be afflicted with the mycotic nodules. Nodular formation is most frequently about 85 percent ofthe cases! encountered inthe lungs, It is also found in the kidneys, liver, and rnesentery, andoccasionally inthe alimentary system stomach, intestine and esophagus!, muscle, inner surface ofthe carapace andin theyolk-sac. Inabout 20 percent of thecases, more than one organ or tissuein thesame turtle have nodules on them. Thenodules are solid, spherical objects that usually protrude on the surface of anafflicted organ. In a fewcases, somenodules have been found embedded below the surface of anorgan. The nodules are either white or creamywhite andrange in size from about 1 to8 mmin diameter. Their solid core distinguishes these nodules from the hollow, often fluid-filledcysts in theadvanced stages ofSP, although S.constrict um, which is a presumptivecausal organism ofSP, is alsoone of the fungi isolatedfrom the smaller, solid nodules. Occurrence Based on necropsy records, the annual occurrence of INM in Kemp'sridley hatchlings main ta ined in theGalveston Laboratory was 19, 8, 13, 138, 73 and 23 cases in year-classes1978 to 1983,respectively. Remedy Clinicaltrials with prolonged, daily intraesophageal incubation of potassiumiodide in hatchlingsthat exhibitedoutward signs of INM succeeded inreversing or reducing some of the abnormal signs in a fewturtles. These turtlescontinued to live. In othersimilarly infected hatchlings, potassium iodide treatment was ineffective and the animalsdied, Nevertheless, potassium iodide treatment appears promising and more research should be doneto exploreits usefulnessas an anti-INM agent in seaturtles. Hyperne erotic Warts Hypernecrotiewarts HW! are tumor-like orswollen lesions tha t occuron the skin, front flippers, carapace, plastron andhead and that usually become heavily encrusted with continuousformation of necrotictissues. The disease a ppears to be fatal, although sometimes ana fflictedhatchling can live a relativelylong time eg., two months or more! beforeit dies.Necropsies reveal that hypernecrotic warts are often accompanied by INM, but not necessarily vice versa.Whether death occurs as a directresult of hypernecroticoutgrowth, INM or bothhas not beendetermined. Etiology Twofungi, Scoloecobasidium constrictum and Paecilomyces sp.,have been isolated from hypernecrotic lesions.They are the presumptive pathogens. Although S. constrictum has been cultured from flipper and carapace lesionsand Paecilornyces sp.from cranial and skin nodules, the specific affinity of individualfungi to specific tissue types hasnot beenestablished. Signsand Symptoms Basically there are two forms of HW lesions: a conicalform and a round non-conical! form, Conicallesions appear to be limited to occurrences onsoft skin parts such as the neck and the suprafemoral pouch. Theyare cone-shaped projections with pointed ends distal from the skin. The nodule is about 3 to5 mmhigh and about 2 to4 mmin diameterat thebase. The color of thenodule is greyor dark-grey,similar to thecolor of theskin. Necrotic tissuesthat appear on olderlesions continue to formas laminated layers on thenodule. A roundedHW is basically represented byelevated, localized swollen tissues, The lesion is relatively round or flat at thetop, rather than cone-shaped. Such lesions may be found on a varietyof externalbody parts.

191 Formationofthe lesions differs according tothe site of infection, On the front flipper, a lesionarises from either the elbowor the base of thelateral claw. In the latter cases, the tissues adjacent tothe claw swell to form a localizedand irregularlyoval-shaped lesion visible from both the upper side and underside of the flipper, The lesion, when measuredfrom either face of theflipper, usually grows to about 5 x10 rnm, At a yetundetermined time, necrotic tissuesbegin toappear onthe swollen lesion, and they continue togrow and accumulate inlaminated layers following the contour of the lesion. Onthe head, a roundhypernecrotic lesionbegins asone or more small eleva ted hard nodules orbumps inthe crown areaabove the brain. The nodule continue s! togrow and usually, although not always, the skin of a nodulebreaks opento expose anaccumulation ofnecrotic tissues. These necrotic tissues continue tomultiply gradually toform a necroticmass outside the skull. Surgical excavation shows that the crust can reach the softtissues underneath the skull. A roundHW on the carapace perhaps begins asa smallvesicle orbleb. Later, this bleb breaks open and necrotic tissuesbegin todevelop into a smallnecrotic nodule ofapproximately 3 to4 mmin diameterabove the carapace surface,Inthe limited number ofcases ofcarapacial HWobserved inKemp's ridley hatchlings, allthe lesions were locatedonthe front margin ofthe carapace closeto the neck or shoulder ofthe animal. Round hypernecrotic lesions havealso been encountered in theplastron and the neck, Hatchlingsorjuveniles afflicted with HW usually remain active and eat well. They may become lethargic shortly before they die. Occurrence There were 3,6, 32 and 10 cases ofHW in the 1978, 1979, 1981 and 1982 yearwlasses ofKemp's ridleys, respectively. Remedy There is no established drug therapy for HW. Experimentally, long-term oral administration with a saturatedsoluhon ofpotassium iodide SSPI! plus daily topical application ofthree percent hydrogen peroxide has producedpositive responses from warty outgrowths onthe flippers and carapace, After more than four weeks of treatmentthelesions begin to show signs of regression andone or two weeks later black pigmentation returns, Diarrheacanbe a side-effectofSSPI trea tmen t,When diarrhea occurs, oralSSPI should bediscontinued forafew days thenresumed when diarrhea stops. Thecranial form of HW responded well to sunlight treatment, After about two months ofineffective treatments withformalin andmalachite-green bathsinside the laboratory, theturtles were taken outdoors daily for a 30to60 minuteexposure tosunlight. After about five days, the cranial lesions began toheal. While the turtles were outdoors, thewater temperature inthe holding vessels was carefully monitored toensure that the turtles were not overheated, lfthe water temperature exceeded28 C, the turtles were moved toa shadyplaceor returned tothe laboratory. Sunlight therapyhas not beenused for the other forms of HW. Informalin bath treatment, thenodule ofthe conical form may shed but is regenerated afterthe treatment is terminated. Malabsorption of Yolk Sac Malabsorptionofyolk sac MYS> refers toa conditioninwhich a hatchling, aftera reasonable periodofposthatching time,fails to absorb either the entire orthe bulk of the embryonic yolksac, presumably dueto physiological dysfunction.Thelength ofposthatching periedwithin which a Kemp's ridleyneonate isexpected tocomplete the yolk-sacabsorption process isunknown. According toF. Wood {Cayman TurtleFarm 983!, Ltd., Grand Canyon, BWI,personal communication, January1985], theyolk sac in ha tchling green sea turtles probably hasbeen absorbed bythe time the hatchlings starttoea t,which isabout five to six days after hatching. Observations atnecropsy suggest thatin many Kemp's ridley hatchlings, bythe seventh dayafter hatching, a large portion ofthe yolk sac has been absorbed.Therefore, if a week-old orolder Kemp's ridley isfound tobear a relativelylargeyolk sac, such asone that fillsa largearea ofthe abdominal cavity, itmight be considered a caseof malabsorption. {Editors' note: initiation of feedingofpeiietized dietsto hatchling Kemp's ridleys isnow postponed forone to two weeks after emergence.j Etiology The cause formalabsorption ofyolk sac in Kemp's ridleys isunknown, It isprobably anorganic dysfunction.G.Harwell, Houston Zoological Gardens, Houston, Tex., suggested that certain avian neona teswould diefrom malabsorption ofyolk sac if fed too soon. Malabsorption ofyolk sac has also been observed inyoung alligatorswith undetermined cause E. Jacobson, University ofFIorida, Gaincsville, personal communication, January1985!. Itis uncertain ifresults withbirds and alligators couldbe applicable tosea turtles, butsome biologists believethatsea turtle hatchlings willnot eat until the yolk sac isabsorbed F.E.Wood, Cayman Turtle Farm, Grand Cayman,BWI, personal communication, January 1985!. Signsand Symptoms There areno external signs orsym ptoms inKemp's ridley hatchlings suffering fromMYS. Theonly evidence comes from week-old orolder hatchlings thatdie in captivity, andwhich are shown through necropsytoretain relatively large yolk sacs 15mm or larger indiameter. Inmany instances, theunabsorbed yolksac almostcompletely fillsthat portion ofthe abdominal spacenot occupied byother viscera. Theunabsorbed yolksac is softand creamy-yellow in colorand its surface is frequently lined with congestedblood vessels. The liver is usually pale and mottled, indicatinganemia, Occunence The occurrenceof MYSin the headstarted Kemp's ridley hatchlingswas 1,4, 25and 4 casesin the 1978,1980, 1981 and 1983year-classes, respectively. Remedy There isno known treatment or prophylacticmethod aga inst MYS. If thehypothesis of too-early-feeding isproven as a causeofthis anomally, then the obvious remedy is prevention by postponing initial feeding. The current standardpractice at theGalveston Laboratory is no feeding of newlyhatched Kemp's ridleys until theyare one to two weeks old Fontaine et al., 1989!, Urolithiasis Urolithiasisis the formationof calculior crystalsin the urinarysystem. The disease is fatal in Kemp'sridley hatchlings. Etiology Thecrystalline deposits found in theurinary systems of affectedKemp's ridley hatchlings have been identifiedby X-raycrystallography as struvite or ammoniummagnesium-phosphate E.Czerwinski, The University of TexasMedical Branch, Galveston, Tex,, personal communication, 1980!. The cause of suchcalculus forma tions has notbeendetermined. It may be a consequenceof an infectious disease, because urolithiasis is frequentlyaccompanied by InternalNodular Mycosis INM! of thekidney, the lung, the liver, or a combinationof theseorgans, as shown by necropsy.However, urolithiasis is not alwayspresent in all casesof suchfungal in feetions. Delinea tion of thepossible relationshipbetween urolithiasis and INM will requiremore study. Signsand Symptoms Externally,a Kemp'sridley havingsevere urolithiasis displays a shriveledand stunted trunk, much like that in INM. The animal canbe very weakand may refuseto eat shortlybefore death. At necropsy, crystallinecalculi are present in theurinary bladder and ureters. The bladder may be so packed with crystalsthat it is distendedup to five timesor moreits originalsize and becomes a solidblock. The kidneys are probably also impregnatedwith crystals,because when they are sliced across with a seapel blade a sandytexture in thetissues is sensed, Occurrence Urolithiasiswas first observedin Kemp'srid leysof the 1982year~lass in which about43 cases were recorded.In the 1983year-class, only two caseswere confirmed. Re~edy There is no known treatmentor therapyfor urolithiasis, Duodenal Ulceration Ulcerationof the duodenumcan occur suddenly in otherwisehealthy-looking hatchlings with few warningsigns. Ulcerscan cause perforation of the duodenumand result in deathof the animals. Etiology The causeof duodenalulcers is not known.There has been evidence that it could be food-related.For instance,in a seriousepizootic caused by duodenalulceration in 1979in Kemp'sridley hatchlings,necropsies showed that along with an ulceratedduodenum, the stomachwas packed with undigested,pelleted feed. Most of the hatchlings about 350!had con trac ted signs of duodenalulcera tion. Eitherthe turtleswere being overfed or something waswrong with the pelletedfeed which causedgastrointestinal paralysis. Various modified feedingregimens were testedincluding feedinga totaldafly ration of no more thanabout seven percent of body weight,reduced frequency of feedingfrom four mealsper day to eitherone or two mealsper day, and either partial or completesubstitution of the pelletedfeed with thawedfish flesh which hadbeen frozen, If a turtlebecame sick or stoppedeating, the ration wasgreatly reduced or withheld. Within a month,new incidenceof duodenalulceration dropped to practicallyzero. There appearedto be no differenceamong the three tested feeding regimenswith respectto effectivenessin suppression of the disease. After successfulimplementation of themodified feedingregimens, the manufacturerof the pellctedfeed informed us that analysesof feedsamples showed contamination by a fungaltoxin, afflatoxin JamesMcVey, National Oceanic and AtmosphericAdministration, Sea Grant ProgramOffice, Washington, D,C., personal communication, 1980!. It is possible,though not conclusive,that afflatoxin may havebeen responsible for the ulceration, Signsam/ Symptoms Live Kemp'sridley hatchlingsthat areafflicted with duodenalulceration usually aremuch bloatedin thebody trunk. In severecases, the intestine partially protrudesoutside the body througha suturebetween plastronic scutcs. In dead hatchlings,the stomachis usually distendedby packed pelleted feed, as shown by necropsy,The duodenumbears an ulcer, usually in the portion that is proximal to the pyrolic valve, In somecases out of 13 necropsiesin the1979 year-class!, duodenal ulcers have occurred concomitantly with ulcersin thelarge intestine. The intestineis frequentlyhighly distendeddue to gasformation in the tract,probably the causeofbloating of thebody. Occurrence A major outbreak of duodenal ulceration occurred in the '1979year-class of Kemp's ridley in October 1979.Suddenly, about 350 hatchlings were discovered ill, and many died within one week. Both morbid and dead turtles showed signs and symptoms of duodenal ulceration. Morbidity and mortality continued to rise daily until

193 modifiedfeeding regimens were adopted for the surviving hatchlings in the raceways. About 150 hatchlings died beforethe feeding changed. Abou t 250surviving ha tchlings that were obviously ill withduodenal u]era tion recovered afterthey were isolated from those remaining in the raceways and placed on either a fastingora restraineddiet. Thirty-twocases of duodenalulceration, frequently concomitant with ulceration,or tissuenecrosis, or bothin the stomachand other parts of the intestine, were also recorded atnecropsy inKemp's ridley hatchlings ofthe 1978 year- class.However, unlike the epizootic inthe 1979 year-class during which cases ofduodenal ulceration surged suddenly tomore than 400 within a fewweeks, the 32 cases ofthe 1978 year-class were spread out throughout the 11 months orso of rearing period. Bloating sometimes occurred in the 1978 year-class cases, but not nearly as frequently asin the 1979year-class. ln addition,at necropsy, ulceration or tissue necrosis, orboth have been observed atsites in thegastrointestinal tract withoutinvolving the duodenum. There wereeight andtwo such cases inthe1979and 1981year classes, respectively. Etiologyof thispathological condition is unknownand its relationto duodenalulceration is uncertain. Remedy Morbid turtles with bloating should be isolated and maintained inclean seawater without feeding. The keyfactor insuccessful treatment ofthe illness appears tobe fasting, which probably provides anopportunity forthe affectedturtle to purgestale food from its gastrointestinalsystem. Treatmentofduodenal ulceration with Maalox aluminum and magnesium hydroxides; Rorer, Inc., Fort Washing- ton,Pa.! did notappear to beeffective. Death continued among turtles under such treatment. Parenteral administra- tionof kanamycinalso did nothelp control the disease, Currently, prophylaxis for duodenalulceration involves avoidanceofoverfeeding andproper storage ofthe pelleted feed to prevent growth of microorganisms whichmay releasetoxins into the food. In feedmanufacture, precautionary steps should be taken to preventpotential contaminationof theturtle feedwith toxin-producingmicroogranisms. Hemorrhagic Bacteriosis Hemorrhagicbacteriosis isa bacterialinfection inwhich there issignificant bleeding from tissues and organs ofthe infectedhatchling. The word hemorrhagic isused to distinguish this type of bacterial infection from those in which thereis noprimary, profuse bleeding involved, either externally or internally.Primary bleeding here refers to bleeding directly attributable to the infection. Etiology Withuse of sterile techniques, a bacterium, Vibrio parahemolyticus, wasisolated from the blood of a few recentlydead, 11-month-old Kemp's ridley juveniles of the 1982 yearwlass during an epizootic in May 1983. Septicemiawassuspected. Theou tbreak could have been triggered bystresses caused bytagging with monel flipper tags.Secondary infection through tag wounds was also possible, Onanother occasion, Gram-negative bacteria were observed in blood smears prepared from a five-month~id Kemp's ridley. The bacteria were not identified, Signsand Symptoms Inhemorrhagic bacteriosis attributable toV. parahemo1yticus, liveKemp's ridleys often vomitblood before death. Even after death, blood frequently flows from the mouth. Atnecropsy, thebody cavity is typicallyfilled with bloody fluid, and the lungs and other internal organs, such asthe stomach, liverand kidneys, are eitherhemorrhagic orcongestive anddark-purple incolor, The liver often displays a mottling pattern ina purplecast. Thelungs are usually not inflated. The deflated lungs probably serve to explain why freshly dead animals are frequentlyfound submerged under water on thebottoms of theirbuckets. Besidesvomiting blood, live turtles exhibit few external signs orsymptoms indicative ofthe infection. Occasionally, thepIas tron is more or less depressed scaphoid! andbruise spots maybe observed onit, bu t thesesigns have not been unequivocallycorrela ted with morbidity. Similarly unconfirmed asa signofthis disease isa depressedappearance ofthe carapace oneach side of the carapacial ridge. This condition issometimes observed inpostmortem examination. It wasalmost impossible toidentify infected turtles with hemorrhagic vibriosis through external gross examina- tions,Infected turtles usually looked healthy with perfectly good muscle tone. Attempts torecover bacteria from the bloodofa fewlive Kemp's ridleys taken from the same groups inwhich death occurred inthe epizootic inMay 1983 wereunsuccessful, The blood sampleswere sterile. Occurrence Hemorrhagic bacteriosis appears tobe enzootic insea turtle juveniles, asevidenced byoccasional sporadicincidences. Thedisease assumes epizootic scale probably when the turtles are subjected tostress orwhen theenvironment becomes favorable for spreadingof thedisease. OnMay 7, 1983, 13Kemp's ridleys ofthe1982 year-class suddenly died in the raceways. Another eight turtles died thenext day. The turtles had recently been tagged with monel flipper tags. Postmortem examinations showedsigns andsymptoms of hemorrhagicbacteriosis, and a Gram-negativerod-shaped bacterium, later identified as V. parahemolyticus,wascultured from blood taken from freshly dead hatchlings, Chemotherapy withampicillin sodium!was applied, prior to positive identification ofthe bacterium, tomore than 400 live turtles maintained inthe sameraceways inwhich deaths had occurred and presumed, therefore, tohave been contaminated. ByMay 27, the epizootichad tapered off followingseven more deaths, for a totalof 28killed by thedisease. Remedy Hemorrhagic bacteriosis, apparently an acute infection, is unpredictable,Since it displayslimited outwardsigns until it is toolate, attempts to identifyinfected live turtlesfor isolationand treatment have been unsuccessful,Such lack of diagnosticcriteria to identifyinfected turtles rapidly is a serioushandicap, and leads to emergencies,In an outbreak in whichthe turtles are fast-dying, as in theepizootic of May 1983,infected live turtles cannotbe quickly identified for therapeutictreatments. An alternativeis to medicateall survivors,whether actually infectedor not. However,when antibioticinjection is the methodto beused, and when thereare hundreds of captive juveniles,it is difficult,labor-intensive and unrealistic if not impossibleto completetreatment on all theturtles in a short time. Under thiscondition, those live turtles thatare reared in thesameseawater as those that die shouldbegiven the highest priority for medication. Thekind of drugor antibioticto beused for treatmentdepends on thespecies of bacterium,the sensitivity of the bacteriumto the drug and the tolerance of thejuvenile turtles to thedrug. In anemergency situation, those three factors cannotbe determinedimmediately, so a drugmust be selectedon a best-guessapproach. However, the situation shouldbe closely monitored and drug susceptibility tests conducted on the bacterial isolates if possible.If theresulting informationindicates resistance to the drug in use,then changes should be made.In the hemorrhagicvibriosis outbreakin May1983, ampicillin sodium!was used with success.The dosage was 0,5 mg per25 g wetbody weight perday, for thefirst day,and half of thatamount for eachof thesubsequent six days. The total daily dosagewas dividedinto twoequal doses injected subcutaneously in the neck at about8 a.m.and 3 p.m.The epizootic subsided by the end of the secondweek. Mycobacterial Pneumonia Mycobacteria 1 pneumonia MP! is an infectionof thelungs by bacteriabelonging to thegenus M ycobacteriurn.It is a wastingdisease and probably fatal. Etiology Acid-fast,rod-shaped bacteria were seen in impressionsmears and paraffin sections of lung lesions takenfrom a three-month-oldloggerhead hatchling of the 1977year-class and storednear O'Cfor aboutone month followingthe turtle's death. Similar bacteria were cultured on syntheticmedia from theaffected lung tissues of the same turtle. They were identified as Mycobacteriummari num. Signsand Symptoms Observationson the three-month-oldmorbid loggerhead hatchling showed stunted growth, emaciationof the body and weakenedcondition before it died. Postmortemexamination revealed a 5-mm long,cylindrical, greyish-colored nodule in theleft lung, M. marinumwas isolated from tissuesof theleft lung. Occurrence Only oneconfirmed case of invasionof lungs by M. marinumwas observed as described above. A secondpresumptive casewas observedin an eight-month-oldloggerhead of the sameyear-class. Postmortem examinationof the latter turtle within 16hours after deathshowed numerous greyish-white small nodules mm or smallerin diameter! in the lungs. Unidentified acid-fastbacteria were cultured from the lung tissues.The animal intermittently exhibited a tilted-swimmingbehavior while still alive. Mycobacterial infectionsin turtlesinvolving otherspecies of Mycobacferi urnhave been reported in theliterature.M. chelonei M. friedmannii!has been isolated from extensivelesions in two turtles Stanfordand Beck,1969!. Brock ef al, 976! describedsix casesof tuberculosisattributable to Mycobacteriumavium in captivegreen turtles. Reichenback- Klinke and Elkan 965! discussed occurrenceof tuberculosis causedby mycobacteria in reptiles and suggestedthat "the clinical picture is that of typical tuberculosiswith pulmonary tubercles in tortoisesand turtles! and analogous lesions in skin, liver and spleen in snakesand crocodiles!." Remedy No treatmenthasbeen established for MP in loggerheadturtles. Murphy 975! reportedthat antibiotic treatment for MP has been ineffective, despite recommended use of streptomycin. Swollen-eye Swollen-eye SE! is a chronicinflammation in the eyeof seaturtle hatchlings. Usually only oneeye is afflicted,and the disease can be fatal. Etiology The cause of' SE can be mycotic infection. On one occasion, a fungus, tentatively identified as Paecilomycessp., was cultured from necrotictissues in the swollen eyeof a Kernp'sridley. Whetherall casesof SE involve mycotic invasion is not known. Signs and Symptoms The tissues surrounding the eye are swollen. In serious cases,the eyelids may be totally closed.Swimming and feedingactivities of thehatchlings generally are not a ffected,except in theterminal stage when such activities become very weak. In one case,at necropsy,necrotic tissues were presentadjacent to an eye-glandbehind the lower corner of the swollen eye of a Kemp's ridley. A fungus, Paecilomycessp., was recovered from these tissues. Occurrence SEhas occurred sporadically in captive populations of Kemp's ridley hatchlings. Incidence hasbeen very low, There was one case each in the 1978,1979, 1981 and 1982year-classes.

195 A swolleneye disease has been reported in landturtles Reichenbach-Klinke andElkan, 1965!. Both eyes were a ffectedand the disease was attributed tovitamin deficiency. Thisdisease probably isnot the same asthat in Kemp's ridleys,since in the latter, usually only one eye was affected, It seems reasonable thatif theSE syndrome inKemp's ridleywere caused by vitamin deficiency, then both eyes instead ofonly one eye would have been affected most of thetime, and that more turtles would have been affected since they were all fedwith the same food. Remedy There is no known effective treatment despite various attempts atchemotherapy witha varietyof generalgermicides andantibacterial, antifungal and antiviral drugs. Afflicted turtles may die after a relativelylong periodof bearing an inflamed eye, Spontaneous remission ofswelling of theeye occurred in two five-month-old Kemp'sridleys after they had beenheld in isolationfor about two months. Intussusception Ina fewKemp's ridley ha tchlings ofthe I 978year-class, intussusception wasobserved atnecropsy. One segment ofthe small intestine sloughed into an adjacent distal segment, resulting ina telescopicappearance inthat part of the bowel. Etiology Whatcaused intussusception andwhether this anomaly contributed directly to thedeath of the hatchlingshave not beendetermined. Signsand Symptoms In one hatchling, localized small patches with a charredappearance were present inseveral areasof anintussuscepted intestine. The significance of such blemishes and their possible relationship with intussusception is unknown. Inhuman pathology, intussusception in time may lead to infarction, asmesenteric blood supply becomes pro- gressivelycompressed dueto the entrapment ofthe mesentery in the fold Robinsand Angell, 1981!. Occurrence Rare. Remedy Remedy for intussusception in human infants isthrough corrective surgery, This procedure isnot practicalfor turtlehatchlings because intussusception in them is not detecteduntil afterdeath. Curved-back Lordosis! In curvedback, the carapace is curved with the rear end turning upward. Etiology The cause forcurved-back isunknown. Death has not been reported inhatchlings withthis anomaly, Signsand Symptoms The carapace iscurved with the rear end turning upward. Occurrences Curved-back syndrome lordosis! occurred in about 16 .1 percent!of the 1977 year-class of loggerheadhatchlings. The condition was observed inboth neonates and hatchlings which had been reared for some time.On one occasion, a loggerhead began to develop a curvedback when it wasabout 10 months old. Remedy Unknown. Soft-shell Insoft-shell, theshell especially thatof the carapace! ina loggerhead orKemp's ridley hatchlings maybecome relatively soft, Etiology Softening of the shellocca sionally occurs after the turtle has recov ered from certain kinds of illness such asemaciation syndrome. X-radiographs showthat the ribs in the carapace ofa so ft-shelled turtle are underdeveloped. It is possiblethat soft-shell is theresult of impairedcalcium metabolism, Signsand Symptoms The shell ofthe turtle issoft to the touch. The turtle displays retarded growth, For example, a one-year-oldturtleaffected bysoft-shell maybe only about one-third thenormal size of af turtle of that age. Occurrence Between 0,2and 0.5 percent ofeach year-class ofhatchlings contracted thesoft-shell syndrome. Remedy Unknown. Coelomic Edema Etiofogy A fewKemp's ridley hatchlings ofthe 1978 year-class thatexhibited a tilted-swimming behavior were indicatedbyX-rays to contain fluid-like substances in one side of the pleural cavity. In collaborationwithG. L. McLellan,TheUniversity ofTexas Medical Branch, Galveston, wewithdrew a clear fIuid via the carapace, using thoranthetictechniques, Asmuch as 7 mlof fluid were obtained from a singleturtle. The fluid was sterile with no growthof bacteria when cultured on synthetic media. The cause of thesyndrome isunknown. Signsand Symptoms Coelomic edema isa signofillness rather than a diseaseinitself, Peritonitis andhepatitis havebeen observed at necropsyin hatchlingshaving the edemic condition. Occurrence Only three tilted-swimming ridlcys of the 1978 year-class were examined forand found to have coelomicedema, The other tilted-swimmerswere not examined. Remedy Unknown. Lung Aplasia Lungaplasia wasobserved inKemp's ridley hatchlings ofthe 1978 year-class. Afflicted turtles had one of the lungs missing.When alive, these turtles exhibited a frequenttilting of thebody to oneside during swimming, Etioloy Thecause of lungaplasia is not known.It is probablycongenital. Signsand Symptoms Side-floatingbehavior was not specificto lung aplasia.Other diseases e.g., SP which damagesone of the lungs! can alsomanifest a similarswimming pattern. Occurrence Theturtles with lungaplasia ranged from two to sevenmonths in age.The incidence was about 0,15 percentas determined from dead turtles. Remedy Unknown. Congenital Flipper Malformation Twoconditions of congenitalmalformation of thefront flippers of Kemp'sridleys were observed in the1978 and 1979year-classes. They are hypoplasiaand multibranched-flipper. Etiology Congenitalflipper malformationis a congenitalabnormality, Signsand Symptoms In hypoplasia,one or bothfront flippers are not fully developedat hatching.The animal cannotresurface after it hassubmerged in thewater in theraceway, To preventsuffocation, the turtleneeds to be maintained in shallow water. In multibranched-flipper,the forearm of a frontflipper branches to formone or two additionalarms, The animals otherwise look normal and survive well. Occurrence Lessthan 0.3 percentof the 1978and 1979year-classes of Kemp's ridleys had congenitalflipper mal forma tion. Remedy None. Cross-beak Etiology Cross-beakis a congenitalabnormality. Signsand Symptoms In cross-beak,the front partof thelower beak of themouth is bentsideways. Afflicted hatchlingsusually cannoteat and haveto be fed via intraesophagealintubation to survive, Occurrence Availablerecords show that asmany as 0 2percent of cross-beakoccur in a givenyear-class of Kemp's ridleys.In someyear-classes, such as 1978, 1980 and 1981, there was no incidenceof cross-beak. Remedy None, Congenital Blindness TwoKemp's ridleys of the1978year-class were hatched without eyesight. One died, The survivor was donated in 1980to TexasA &MUniversity for usein research.Since the blind survivor wasable to locateturtle feedin theholding tank, eat and grow, it indicatedthat Kemp'sridleys are ableto find food via the senseof smell alone. Etiology Congenital, Occurrence Two Kemp's ridleys of the 1978year-class were hatched without eyesight, Remedy None. Intestinal Prolapse Etiology Unknown. Signsand Symptoms The lower part of the inteshneprotruded from the anus,and one of two turtles died. The other received corrective surgery from a local veterinarian but died later. Occurrence Intestinal prolapse occurred in two Kemp's ridley hatchlings of the 1981 year-class. Remedy Surgery could be a potentiaHyuseful correctivemeasure, although pastattempts were unsuccessful. Urinary Bladder Prolapse Etiology The urinary bladder of a Kemp'sridley hatchlingof the 1983year-class protruded from the anus,and the turtle died, Necropsy showed that the urinary system of the animal was seriously enlarged due to mycotic infection.The pressure from that swollensystem had forced theurinary bladderto protrude from theanus. While the turtle was still alive, attemptsto restorethe protrudedbladder to its original placeby pushingit back mto thebody with a small lubricated, cotton swab were not successful.The bladder re-evaginated after a short period of time. Signs and Symptoms Protrusion of the urinary bladder and lethargy are the typical signs. Occurrence Lessthan 0.1percent of the 1983year-class contracted urinary bladder prolapse. Remedy Unknown.

Discussion and Recommendations Out of necessity,we adopted a two-phasestrategy to addressthe diseases,injuries and associatedmortality problems of head starting Kemp's ridley and loggerhead seaturtles at the Galveston Laboratory. Many animals that died rapidly, especiallyduring the earlyyears of the rearingoperation, were afflicted with a greatvariety of diseases of little known nature,or they sufferedsevere traumatic injuries caused by intraspecificbiting. If thesediseases and

197 injurieshad continued unabated, catastrophic losses ofturtles would have continued tooccur, seriously hampering progress of the head start project. Thefirst phase of our strategy was aimed a stopping,t asquickly as possible, on-going mortality and reversing any observedconditions ofdisease orinjury even though the reversals might have been accomplished through treatment onlyof thesymptoms. Due to the paucity of information on diseases and injuries of seaturtle hatchlings atthat time, weoften had to employ therapeutic methods, skills and knowledge that originated from other scientific disciplines orthat were developed asa resultofour own observations orshort-term experiments. Prescriptions werefrequently, if notalways, based on educated guesses. Our objective was to save as many of the head started turtles as possible, disregardingmeans, sothat the head start project could progress through its rearing, tagging and release stages. This first phaseof our strategyinvolved clinical practices without much baseline information, Phasetwo of our strategy, which was engaged based upon the availability ofresources and only after most of the needsof phase one has been met, was intended toinvolve elaborate and in-depth studies designed tounderstand betterthe fundamental nature and causes ofsea turtle diseases and injuries. The objective was to provide essential baselineinformation forimprovement anddevelopment oftechniques andmethodologies fordiagnosis, therapy and prophylaxis.Such information could lead to more confidence andsurety inclinical practices and to a higherdegree ofpredictability ofhealth cond itions and prognosis. Thenet result would be more and healthier sea turtle yearlings for release. Initially,the objective ofphase one was rapidly achieved through discovery and subsequent implementation ofthe isolation-rearinganduse of clean seawater forculturing the turtles from hatchling toyearling stages. While the latter requirementofgood hygiene and sanitation was an obvious one, the need for isolation rearing was not. These methods andinnovations, coupled with appropriate mcdications onan as-needed basis, quickly brought under control many significantand wide-spread health problems which plagued the earlier year-classes. Examples ofcontrollable health problemswere sudden hatchling death syndrome, papillary dermatitis, focal erosive dermatitis and aggressive biting. Isolation-rearingandseawater cleanliness became cardinal requirements formaintaining health of captive-reared Kemp'sridleys, and laid the cornerstone for subsequentsuccesses in head starting this species. General statements concerningimportance ofculture water to rearing aquatic turtles Pope, 1950! were published before 1977, the year whenthe Galveston Laboratory initiated loggerhead rearing, but there were few specifics described co~cerning the desirablelevel of cleanliness orwater quality, It wasindeed a surprise tous to find that the quality of the recycled seawatersuitable for the culture of marine shrimp Penaues spp.! in closedraceways was unsuitable forrearing loggerheadhatchlings ofthe 1977 year-class. Nearly 40 percent ofthe animals died from sudden hatchling death syndromewithin about four weeks. Biodisc filter systems used to remove wastes from the recycled seawater had maintaineda seawater quality sufficient for intensive shrimp culture in these raceways Mock, Ross and Salser, 1977!. Despitethe considerable progress and success in phase one of ourstrategy, certain earlier maladies such as emaciationsyndrome remain little understood, incurable and unpredictable. Inaddition, other forms of disease and injuriesemerged, and some of them had the potential ofkilling large numbers ofturtles, Examples were the various formsof scolecobasidiosis,internal nodular mycosis, malabsorption ofyolk sac and urolithiasis. Some of thenew maladies eg., duodenal ulceration inOctober 1979 and hemorrhagic bacteriosis inMay 1982!, suddenly erupted into seriousepizootics, crea ting emergencies ofpotentialy alarming proportions. U~der such circumstances, wedevoted ourentire efforts to addressingthese new problems, trying to findimmediate answers to prevent potential catastrophes.Thus,we remained forthe most part engaged inphase one and inevitably had to interrupt allphase two work,either on-going or scheduled,and to postponeor foregoparts of themission-oriented and fundamental pathological studies, Despiteunforeseen difficulties, wehave been able to Iay much groundwork forfuture pursuits ofphase two work. Someofour accomplishments includedidentifying andcategorizing various major forms of disorders afflicting captivesea turtles, isolating and identifying many presumptive pathogens e.g,, Scotecobasidium constrictum and Paecilomycessp.in mycoses, andVibrio parahemolyticus inhemorrhagic bacteriosis!, providing X-radiography asa potentialtool for disease diagnosis inthe laboratory, andobtaining evidence thatpotassium iodide isa potentially effectivedrug against sea turtle mycoses, Pathologyinformation isthe backbone forresearch anddevelopment oftechniques andknowledge essential for thepractice ofgood medicine onany living system, whether itbe man, other animals including seaturtles! orplants. Correctapproaches todiagnosis, therapy and prophylaxis andthe development oftechniques andmethodologies for suchrequire fundamental understanding ofthe nature and cause ofdisease, injury or death. Such basic information canbe acquired through pathological observations andstudies. Once the basic knowledge andskills are established, clinicaland laboratory examinations andtesting can be evolved from these basics and employed toobtain data in specificpatients to effectcorrect medical judgements and decisions. Becauseof practicality, it was inevitable that many minor illnesses in headstarted Kemp's ridleys would have to be handledmedically on a "best-guess"basis without goingthrough extensive and costlypathology research. However,for diseaseswhich potentially can cause serious debilitation or deathin a significantportion of a captive populationand which may lead to catastrophiclosses, reliance on best-guesses alone is risky.The spread of these diseasescould result in lossof theentire captive population, Such a pricecould be high for a seriouslyendangered speciessuch as Kemp'sridley, where each surviving female could provide a substanhalcontribution to the conservationand augmentation of thedeclining wild population,It isfortuitous that we were able to resolve all major diseaseor injury problemsthat aroseduring the headstart project up to the 1983year-class, but therewere no quaranteesthat suchluck would persist. Forpragmatic reasons, we recommend that more pathological studies on Kemp'sridleys be conducted, These studies,should focus on certain fatal diseases which presently are incurable and unpreventable. Examples are ema- ciationsyndrome, scolecobasidiosis, internal nodular mycosis, yolk-sac mycosis, urolithiasis, hemorrhagic bacteri- osisand malabsorption of yolk sac,Research should be directed toward determining causes of theseillnesses e,g., throughpathogenicity experiments!, sources of theproblems, how the diseases are transmitted and spread, the na ture andbiology of theetiological agents, histopathology both basic and clinical!, hematology both basic and clinical hematologyfor diagnosis!,X-radiography and other similar techniques for specificdiagnosis and monitoring of internalailments, and in vitroresponses of pathogensand host turtles to potentialtherapeutic agents and techniques. A goodunderstanding of themicroflora and chemistry of theculture seawater is alsoof paramountimportance, Baselineprofiles of microfloraand chemical constituents in the culture seawater should be determinedand the potentialrole of theseentities in diseasedevelopment should be delineated. With such information on hand, envi- ronmentally-orientedprophylactic measures, which in thelong run are perhaps the least expensive in diseasecontrol, can be developed. Aggressivebiting can be contained through isola tion-rearing, bu t other methods might be developed to accomplish thesame end, Among other things, isolation-rearing is costly in laborand materials and many turtles outgrow their bucketsduring the typical head start period of 9 toI I months.One obvious alternative is to providelarger containers, but theywould take up morespace in a givenraceway or tankand reduce the numbers of turtlesreared per raceway or tank.Along with addingmore raceways or tanksto accommodategreater numbers of turtles,other needs such as morespace, seawater, electric power, labor and supplies will arise.In light of thecosts of isolationrearing, studies are neededto determinethe underlying mechanism of theaggressive behavior in hopesof developingprocedures for alleviating it. Finally,since the methods by whichthe sea turtles are maintained during head starting have great impact on their diseasesand traumatic injuries, pathologistsshould be involved in the designof culture systemsand modesof operation.What may appeal to theculturist as useful or practicalmay turn out to beunacceptable from a turtlehealth pointof view,as was demonstrated in our earlyexperiences with bothloggerheads and Kemp's ridleys reared in groups,Effective compromises can be worked out throughbetter understanding of pathologyin relationto rearing methods.We encouragemore studiesto provide relevantinformation needed to optimizethe culture and operation systemsin relation to health maintenanceand diseasecontrol while assuringtheir cost-effectiveness, Despitethe continuing need for improvement,the culture and healthcare procedures that haveevolved during the rearing of 11 year-classesof Kemp's ridley seaturtles havebeen successful Caillouet, 1984;Fontaine, Leong and Harris, 1984;Fontaine et al,, 1989!.Each year additional knowledgeis gainedand methodsare improved.In all, asof October31, 1988, abou t 14,060Kemp's ridleys of the 1978-1988year-classes had been successfully reared out of 16338 live hatchlingsreceived, representing 85 percent survival during headstarting Fontaineet al., 'l989!. Of these,13372 hadbeen tagged and released into theGulf of Mexico this doesnot includethe1988 year-class!. The remaining normal and healthyturtles were transferedto other locationsfor extendedhead starting and captivepropagation, and a few that wereabnormal, sick or injured were transferredto other laboratorysfor researchor humanelydisposed of. In conclusion,sea turtle pathologyresearch has barely started. Since1978, there have been good successesin containingdiseases and injuries in head startedsea turtles at the GalvestonLaboratory, but somedifficult disease problemsremain to be solved.Carefully plannedand selectedresearch can shed light on someof thesedifficult problemswhich if left unexploredhave the potentialof developinginto uncontrollableepizootics, Research should include in-depth studieson basicand applied pathologyof the host turtle, the etiologicalagents and the culturing environmentas well astheir inter-relationships.Such information will provide theprerequisite data for development of sound and practicalhealth caremethodologies, within the constraintsof establishedguidelines and regulations controlling the possession,handling and care of endangered sea turtles. Acknowledgements We thank Ray Wheeler,Larry Landsford,Clark T. Fontaine,Theodore Williams and their staffswho furnished specimensof sick and dead seaturtles and pertinenthistories for theseanimals. Specialmention should be made of Rend Mlrquez Millan INP, Mexico!, Robert King NPS, Corpus Christi, Tex.!, JamesWoods NPS, Big Thicket, Tex.!, Jack Woody and David Bowman FWS, Albuquerque, N.M,!, Milford Fletcher NPS,Santa Fe, N,M.!, Patrick Burchfield Gladys Porter Zoo, Brownsville, Tex.! and Ross Witham Florida DepartmentofNatural Resources, Jensen Beach, Fla.!, Through their efforts and cooperation, Kemp's ridley and loggerheadhatchlings were provided to the Galveston Laboratory forhead starting. EdwardF.Klima, Director ofthe NMFS Galveston Laboratory, was instrumental inbringing the sea turtle head start projecttothe Galveston Laboratory in1977, James McVey, who managed thehead start operation from 1978 to 1981, wasappreciative of the significance ofsea turtle pathology study. We are grateful for his support, especially during difficulttimes, Charles W.Caillouet succeeded McVey in1981 and continued toprovide administrative support to sea turtle pathology activities. Invaluablehelp in the forms ofprofessional services, technical advice and use of labora tory facilities was generously donatedby MelvinSchreiber, Gary McClellan, Dean Folse, Michael Kelly, Edward Czerwinski, Patrick Davis and AdamEwart University of Texas Medical Branch, Galveston, Tex.!, Gary Harwell

Literature Cited Berklow,R. Ed itor!. 1977. The Merck manual ofdiagnosis andtherapy. 13th Edition, Merc k 8cCo., inc., Rahway, N J., xxii + 2165 P. Brock,J.AR.M. Nakamura, A.Y.Miyahara and E.M.L. Chang, 1976. Tuberculosis inPacific green sea turtles, Chelonia mydas. Transactionsof the American Fisheries Society 105:564-566. Buchanan,RE.,and N E,Gibbons. 1974. Bergey's manual ofdeterminative bacteriology. 8thEdition, The Williams 8zWilkins Co., Baltimore,Ivfaryland. Caillouet,C.W. 1984. Essai de prevention deI'extinction deIa Tortue de Kemp. Les Carnets deZoologie44:28-34. Cai!louet, C.W., Jr.,D.B. Koi, C.T, Fontaine, T.D.Williams, W.J.Browning andR.M. Harris. f986. Growth and survival ofKemp's ridlcysea turtle, Lepidochelys kempi,in captivity. NOAA Technical Memorandum NMFS-SEFC-186, iii pIus34 p., l2 Tables and 7 Figures. Campbell,H.W. and S.D. Busack. 1979. Laboratory maintenance, Chapter 6,p. 109-125. In:Harles, M. and H, Morlock Editors!, Turtles:Perspectives andResearch, John Wiley k Sons,Inc. Clary,John C.III aud Jorge K.Leong. 1984. Disease studies aidKemp's ridley sea turtle head-start research. Herpetologtcal Review 15:69-70. Combs,C.M. 1976. Illustrated medical dictionary. Consolidated Book Publishers, New York, NY. Cowan,D.F. 1968. Diseases ofcaptive reptiles. Journal ofthe American Veterinary Medical Association 53:848-859. Foutaine,C,T.,J.K. Leong and R.M. Harris. 1984. Head starting theKemp's ridley sea turtle, p.258. ln Bacon,P.,F. Berry, K. Bprndal,H. Hirth,L. Ogren and M. Weber Editors!, Proceedings ofthe Western Atlantic Turtle Symposium, Volume 1, Universityof MiamiPress, Miami, Florida, 306 p. Fontaine,C T.,K T.Marvin, T D. Williams, W J.Browning, RM,Harris, KL. W. Indelicato, GA.Shattuck andR A. Sad ler, 1985. Thehusbandry of hatchling toyearling Kemp's ridley sea turtles Lepidochelys kempi!, NOAA Technical Memorandum NMFS-SEFC-158,iv plus 34 p., 10 Tables, 22 Figures and 2 Appendices. Fontaine,C.T.,T,D. Williams, S.A.Manzella andC.W. Cailfouet, Jr.,1989. Kemp's ridley sea turtle head start operations ofthe NMFSSEFC Galveston Laboratory This volume!. Haincs,H G,, A.Rywlin and G. Rebel. 1974. A herpesvirus diseaseof farmed green turtles Cheloniamydas!,p.183 195. In; Avault, J.W.,Jr. Editor!, Proceedings ofthe Fifth Annual Meeting, World Mariculture Society, Louisiana StateUniversity, Division of ContinuingEducation, Baton Rouge, Louisiana,482 p. Hendrickson,J.R.1958. The green seaturtle, Chelonia mydas Linn.! inMalaya andSarawak. Proceedings ofthe Zoological Societyof London 130!:455-535. Jacobson,E.R.,J.M. Gaskin, R.P.Shields andF.H, White. 1979, Mycotic pneumonia inmariculture-reared greensea turtles. JournaI of theAmerican Veterinary Medical Association 175:929-933. Jawetz,E.,J.L. Melnick,and E A.Adclbcrg. 1972.Reviews ofmedical microbiology. TenthEdition, Longe Medical Publishers, Los Altos, California. Kaplan,HM 1957.Scpticemic, cutaneous ulcerative disease ofturtles. Laboratory ofAnimal Care 7:273-277. KlimaE F.and J PlvlcVey. 1982. Headstarting the Kemp's ridley turtle, Lepidocheiys kempi p 481<87.In: BjorndalK A. Editor!, Biologyand Conseroation of Sea Turtles, Smithsonian Institution Press, Washington, D,C., 583 p. Koment,R. W. andH. G,Haines. 1977. A newreptilian herpesvirus isolated from the green sea turtle, Chelonia rnydas, p. 347. In:Proceedingsof the American Society of Microbiologists Abstract!. Leong,J.K. 1979. Hatchling diseases in Atlanticridley turtle Lepidocheiyskempi! and loggerhead turtle Carettacaret ta! in the GalvestonLaboratory, National Marine FisheriesService, American Zoologist 19:659 Abstract!. Leong,J,K,, R.S. Wheeler and L.M. Lansford. 1980. Tolerance and responses of normal and diseased loggerhead turtles Caret ta caretta!to somechemotherapeutics, p. 291-302. In; Avault,J.W., Jr. Editor!,Proceedings of the World Mariculture Society, Volume 11, 611 p. McLellan,G.L. and J.K. Leong. 1982. Radiologic evaluation of thedifferential absorption of diat~oatein marineturtles. NOAA Technical Memorandum NMFS-SEFC-93,'15 p. McVey,J.P. and T. Wibbels.1984. The growth and movements of captive-rearedKemp's ridley sea turtles, Lepidocheiys kempi, followingtheir release in theGulf of Mexico.NOAA Technical Memorandum NMFS-SEFC-145, 25p.,3 Figures, and 3Tables. Marcus,L.C. 1977,Parasitic diseases of captivereptiles, Volume VI, p. 801-806,In; CurrentVeterinary Therapy, W. B. Saunders Company,Philadelphia, PA. Mock,C.R., L.R. Ross, and B.R.Salser. 1977. Design and preliminaryevaluation of a closedsystem for shrimp culture,p. 335-369. In: Avault, J.W.,Jr. Editor!, Proceedingsof the 8th AnnualWorkshop, Wor1d Maricuiture Society, Louisiana State University, BatonRouge, LA, 984p. plus 52 p. of paneldiscussion. lvlrosovsky,N. 1983.Conserving sea turtles. The Zoological Society of London,Regent's Park, London, England, 176 p. Murphy,J.B. '1975. A briefoutline of suggestedtreatments for diseases of captiverept iles. Society for the Study of Amphibiansand Reptiles,Miscellaneous Publication No. 4, ii plus 13p. Pope,C. 1950. Reptiles, Chapter12, p. 299-330. In: Farris, E. Editor!,The Careand Breeding ofLaboratory Animals, John Wiley &Sons, Inc., New York, NY. Rebel,G., A. Rywlin and G, Ulrich. 1974.Coccidiosis in the greenturtle Cheioniamydas! in mariculture, p. 197204. In: Avault,J W., Jr. Editor!,Proceedings of the Fifth Annual Meeting, WorId Mariculture Society, Louisiana State University, Division of ContinuingEducation, Baton Rouge, Louisiana, 482 p. Rebel,G., A, Rywlin and H. Haines, 1975. A herpesvirus-typeagent associated with skinlesions of greensea turtles in aquaculture. AmericanJou ma!of' Veterinary Research36:1221-1224. Reichenbach-Klinke,H. and E. Elkan.1965. The principal diseasesof lowervertebrates. Academic Press, New York, NY. Robins,S.L. and lvl. Angell, 1971.Basic pathology. W.B, Saunders Company, Philadelphia, PA. Sinderman,C.J. 1977. Aeromonas disease of loggerheadturtles, Chapter6.2, p. 292-293.In: Sinderman,C.J. Editor!, Disease Diagnosisand Control in 1VorthAmerican Marine Aquaculture, Elsevier Scientific Publishing Company, New York, NY. Smith,D.T., N.F. Conant and J.R.Overman. 1964. IVIicrobiology, Thirteenth Edition, MeredithPublishing Company, New York, NY, Stancyk,S.E. 1981. Non-human predators of seaturtles and their control, p. 139-152.In: Bjorndal,K.A. Editor!, Biologyand ConserMtionof SeaTurtles, Smithsonian Institution Press,Washington D.C., 583 p. Standford,JL, and A. Beck.1969. Bacteriological and serologicalstudies of fastgrowing mycobacteriaidentified as Mycobacterium friedrnannii,Journal of GeneralMicrobiology 58:99-106. Stickney,RR., DB. Whiteand D. Perlmutter.1973. Growth of greenand loggerheadsea turtles in Georgiaon naturaland artificial diets. Bulletin of the GeorgiaAcademy of Science31:37-44. Wibbels, T.A. 1984. Orientation characteristics of immature Kemp's ridley sea tutles, Lepidochelyskempi. NOAA Tcchnical Memorandum NMFS-SEFC-13'I,iv plus 67 p. Witham, R. 1973a.A bacterial diseaseof hatchling loggerhead sea turtles. Florida Scientist36:226-227. Witharn, R. 1973b.Focal necrosis of the skin in tank-reared sea sea turt! es.Journal of the American Veterinary Medical Association 163:656,

201 atcertainstages of development hasbeen cited as a sourceofincreased scutevariation in olive ridleys Hill,1971!,and hasbeen shown to inducemortality in embryos Limpus, Baker and Miller, 1979!. Thus, there is evidence to suggest that severaldifferent environmental factors may influence scute pattern variation during incubation, Therealso exists a possibilitythat there is sexualdimorphism in scutecounts as suggested by Hill 971!, As most studiesof adult turtlesare based predominantly on observationsof nestingfemales, perhaps our conceptof a "normal"scute pattern is biasedby lackof datafrom adult males. Given the strong influence of incubation temperatureon sexin seaturtles, and the example cited above of temperature-inducedscute pattern variations in gartersnakes, perhaps incubation temperature could be influencing sex and scute pattern in a parallelmanner, However,Fraz ier 984! examineda seriesof oliveridleys from Mexico and concluded that there were no signif ican t differencesbetween the sexes in anyof themeristic characters he compared, including carapacial scutes. Thetypical chelonian carapacial scutation consists of a medianlongitudinal series of unpairedelements the vertebralscutes!, flanked on eachside by a seriesof bilaterallypaired scutes the costals!,which arebordered exteriorlyby anotherseries of bilaterally paired scutes the marginals!. Situated anteriorly between the first pair of marginalsis a singlenuchal. Thus, midline elements consist of a nuchalscute anteriorly, followed posteriorly by a seriesof vertebralscutes. The paired marginal scutes border the periphery of thecarapace except where the nuchal separatesthem anteriorly. The costal scutes occupy the region lateral to thevertebrals and medial to the marginals on eachside. Some workers consider the posterior-most pair of carapacialscutes the supracaudals! apart from the marginals e.g., Frazier, 1984; Pritchard and Trebbau, 1984!, but for the purposes of thisstudy we have considered these scutes as part of the marginal series. Amongthe sea turtles, the most deviance from the scute pattern common to mostindividuals of a speciesas well as the greatestvariability in patternare foundin the genusLepidochelys. The olive ridley possessessuch a high frequencyof scutepattern anomalies that it isconsidered by Pritchard969b! to be"unique among turtles in having a trulypolymorphic carapace" i.e., essentially having no "normal"pattern!, Though it doesnot possesscarapace polymorphismto thedegree of its congener,Kemp's ridley has nevertheless been shown to exhibitsome carapacial scutevariation Chavez,Con treras and Hernandez,1967, 1968; Pritchaxd, 1969b!, The modal carapacialscutation for Kemp's rid ley consists of 13pairs of marginalscutes, five pairs of costa I scutesand five vertebral scutes, with a single nuchal Figure 1!. The intent of our study was to describeand documentscute anomalies in Kemp's ridley from two perspectives: I. meristic variation with eachseries of scutes left marginals,left costals,vertebrals, right costalsand right marginals!, and 2. variation among individual turtles from the modal scutation pattern Figure I!, Further,we attexnptedto relatevariation in Kemp's ridley scutationto the various types of natural and artificial conditionsunder whichour study animalswere incubated and handledand alsoto theontogenetic stages represented in our study animals, Continued researchinto the causesof scute pattern variation in turtles is necessaryto elucidate possible relationshipsof suchvariation to viability, An improved understandingof theserelationships could greatly assist managementand conservationof endangeredand threatenedspecies of turtles.Zangerl 969! hascommented that "the morphogeneticcontrols that determine the bone and shield patternsof the turtle shell are unquestionably complex."It is hoped that the presentstudy will enhancethe knowledgeof thesecomplex processes.

Materials artd Methods The carapacialscutes of 5,919specimens of Kemp's ridley hatchlingsand juvenileswere examined.The study animalswere grouped both by ontogeneticstage and conditions under which theywere incubated and handled Table I!. During the 1981nesting season at the Kemp'sridley rookerynear Rancho Nuevo, Tamaulipas, Mexico, 4,114 live hatchlingswere examinedand later released.Another 298hatchlings and embryos,either dead in the egg,dead in the nest,or alivebut severelydeformed or underdevelopedand unableto rupture the egg,were examined at Rancho Nuevo during the samenesting season. Those that wereunable to rupture the eggwere included as dead hatchlings becausethey would havedied naturally had they not beenused in our study, Deadhatchlings and embryoswere considereda youngerontogenetic stage than live hatchlings.The remaining 1,507were juvenilesexamined at the National Marine Fisheries Service NMFS! Southeast Fisheries Center SEFC! Galveston Laboratory during head startingof the 1980year-class. The specimensexamined at RanchoNuevo camefrom neststhat hadbeen handled or incubated, or both, in four different ways: The largest group of hatchlings came from 37 nests incubated in corrals at Rancho Nuevo. This category wifl be referred to hereafter as the corral nest category. These eggs were dug by hand from their natural nests on the beach, then transported in bags to the corrals where they were reburied in holes carefully fashioned to

203 closely resemblethe depth and shapeof natural nests.The corralswere located high on the dune to N lessenthe risk of inundationduring spring tides M and werefenced to protectthe eggsfrom largepre- datorsssuch as coyotes.Due to the large numberof turtles nesting during arribadas,eggs replanted in this fashionoften remained in situ for up to seven hours before replantation, These nests were some- timessubjected to bouncingand heat during trans- portation from their natural nests to the corrals, and were at times left out of the substrate for several hoursprior toreplantation.Weauthors consider the eggs replanted to the corral in this manner to have receivedthe roughestor leastcareful handling of any of the incubation-handlinggroups. 2. The secondlargest group of hatchlingscame from 11nests replanted using the specialhandling tech- niques of the international effort to establish a new nestingcolony on PadreIsland, Tex., through head starting.Padre Island boxes will be the category used in this paper to refer to thesenests. Padre Islandeggs were collected in cleanplastic bags as they droppedfrom the cloacasof ovipositingfe- males,thus avoiding contactwith RanchoNuevo sand. They were immediately transported to a shadedarea where they were placed by handinto polystyrenefoam boxes filled with sandcollected Figurel. Modalearapaeial scutearraiigemeiif forKemp's rid+sea fr om P a dre re IsI lad n, Th ese bo sxes co containing I n n P a dre re turfle ind&ing 1 3pairs of marginals M! 5 pairsof costs sC! 5 Island sandand the eggswere placedon elevated shelvesin a concreteblock building at the Rancho Nuevoturtle camp where they were carefully mon- itoredto prevent their dessication or infestationof anykind. The in tendeddestination of these11 nests was the Padreisland National Seashore, where the eggs were to be incubated bythe National Park Service toprovide hatrhlingsforthe Kemp's Ridley Head start Research Project atthe NMFS SEFC Galveston Laboratory, Due tonumerous delays, these eggs hatched at RanchoNuevo before they could be flown to the National Seashore. Thehatchlings usedin our study were released offshore ofRancho Nuevo. The eggs that produced thisgroup of894 hatchlings were exposed tovirtually no rough handling and were expeditiously transplanted. 3. Sevennests were incubated in RanchoNuevo sand in polystyrenefoam boxes in thesame concrete block building,concurrently andu nderthe same conditions asthe 11 originally destined for Padre Island. However, theseRancho Nuevo nests were transplanted asmuch as 24 hours after oviposition. Thisgroup will be referred toas the Rancho Nuevo boxes category. Inall but one of the Rancho Nuevo nests, special handling techniques wereused to avoid changing theorientationof theeggs asthey were removed from the natural nest. These eggs produced549 hatchlings used in our study, 4. The151 hatchlings thatrepresented a sample offive natural nests atRancho Nuevo were discovered asthey wereentering the sea. We categorized thisgroup as natural nests. The exact number of hatchlings from each nesttherefore was unknown, One of the five nests was severely infested with ants and contained many dead hatchlingsthatinightotherwisehaveemerged alive.Thismayhavebiased thatportionof thesampleof natural nestsrepresented bydead ha ichlings; i e., some of the ha tchlings that died due to attack by ants may have had scutationcharacteristics nearerthose ofthelive hatchlings thanthose that died from other causes. Though there wercfewer hatchlings from natural nests than in anyof the other three categories from eggs incubated at RanchoNuevo, we included the data for naturally incubated nests due to the paucity ofsuch information in theliterature for Kemp's ridley and because ofits value in comparativestudies. 5. Inaddition tothe four categories ofhatchlings examined atRancho Nuevo, 1 >07 living juvenile Kernp's ridleys ofthe 1980 year-class wereexamined inApril 1981 atthe NMFS SEFC's Galveston Laboratory. Theywere categorizedasGalveston head starts. These juveniles were presumably handled and incubated in thesame manneras described in item 2 abovefor thePadre Island sample incubated at RanchoNuevo, except that the boxedeggs from whichthe 1980yearwlass hatchlings were produced were transported from Mexicoto the PadreIsland National Seashore where they were incubated, and where the hatchlings were imprinted before being transferred to Galveston. Temperatureregime, including mean temperature and temperature range, differed between the corrals and the concreteblock building at theRancho Nuevo turtle camp. Levels of humidityand their fluctuation also doubtlessly varied betweenthese two environments,as the nestsincubated in polystyrenefoam boxes were shelteredfrom precipitation,receiving only occasional and intentional sprinklings to preventdessication of theeggs.Becauseof these humidityand temperature differences, as well as differences in internalnest conditionsbrought about by incubation in boxes,the environmentalconditions acting on theeggs and embryosincubated for variousperiods in theconcrete block house i.e., PadreIsland boxes, Rancho Nuevo boxes and Galvestonhead starts! were substantiallydifferent from those for the beach-incubated nests i,e,, corral nests and natural nests!. For each turtle, the number of carapacialscutes in eachseries was recordedfrom left to right, including left marginals, le f tcosta ls, vertebrals, rig h costat ls andrigh t marginals e.g., 13-5-5-5-13!, Also recorded for each turtle was whetheror not it had the normalsingle nuchal scute. Each five-element array constituteda scutepattern or scuta tion. Frequenciesofeach pattern variation were also recordedby nest. Variation in frequencieswithin each scute series left marginals,left costals, vertebrals, right costals and right marginals! as well as nuchals, was examined separately from that of scutepattern, This approachto analysisof scutevariations did not recognizeabnormal seam placement unless it alsoentailed a changein thenumber of scutesin a particularseries. For instance, a specimenmay havehad theusual five vertebrals,but with the seamspositioned in sucha wayas to clearlyindicate a non-normalarrangement. In our analysis,such abnormality would not be recognized. In thissense our dataare conservative with regardto levelsof variability,because a specimensuch as that described above was recorded as normal with regardto vertebrals. Statisticalprocedures generally follow the recommendations and procedures of Sokaland Rohlf 9IiI!, Becauseof the disparity in magnitude betweenmarginal counts and costalor vertebral counts,direct comparisonsamong coefficientsof variation for thesedifferent seriesare misleading.While the significanceof eachunit deviationfrom the mode is no different for marginalsthan for costalsor vertebrals,the coefficientsof variation for marginalcounts are smaller as a result of the larger denominator i.e., 13 instead of 5!, Therefore, the standard deviation is the best indicator of relative variation amongscute series in a givenincubation-handling category, In comparingdifferent incubation-handling categories with regardto a givenscute series, the coefficie t ofvariation was a valuableindicator. The magnitude of the standard deviation for a given scute series within an incubation-handling category was an indicatorof relativevariability in that scuteseries because of thediscrete nature of thedata. Paired comparison t-tests were used to test for asymmetryin countsof costalsand marginals.G-tests of independencewere used to test the goodnessof fit of cell frequenciesto expectationsin examiningthe frequenciesof individuals exhibiting the modal scute pattern in the various incubation-handling categories and ontogenetic stages.

Results

Variation in the Nuchal Scute Series Nuchal scute variation from the normal single nuchal was confined to a small number of turtles in which this scute had either split to form a double nuchal or was fused with adjacent scutes from the marginal or vertebral series.The double nuchal was the more common of thesetwo departures from the norm, However, such variations in the nuchal were rare, because99,5 percent of the 5,919 turtles possesseda single nuchal scute. Variation in the Vertebral Scute Series The number of scutesin the vertebral seriesranged from 3 to 9 with a modeof five Table 2!. The standard deviation and coefficient of variation were higher for the vertebral series than for any other scute series. The magnitude of variation in vertebral counts, as indicated by standard deviations and coefficients of variation, ranked highest in the corral category, followed by the Galveston head starts, Padre Island boxes, Rancho Nuevo boxes and natural nests Table 3!. Rangesin vertebral scute counts in live turtles show similar rankings with the widest range for corral nests,followed by the Galveston head starts, Padre Island boxes, RanchoNuevo boxes and natural nests Table 4!. A similar ranking by range is mirrored in the dead turtles except that no dead Galveston head start turtles were examined. The proportion of turtles possessingthe modal count of five vertebrals is another criterion useful in comparing incubation-handling categories.For live turtles, Galveston head starts exhibited the smallest proportion 2.4 percent! with five vertebrais followed by corral nests 7.4 percent!, Padre Island boxes 85.9 percent!, Rancho Nuevo boxes 90.7 percent! and natural nests 97.5 percent!. In all categories, except Galveston head starts which had no dead turtles, the frequency of the modal vertebral count was greater for live turtles than for dead.

205 Table3. Summary statistics forcarapacial scute counts ofKemp's ridley hatchlings andjuveniles, byincubation- handlingcategories and scuteseries. Incubation-handling category and Standard Coefficient of scute series Range Minimum Maximum Mode Mean deviation variation %! Corral nests n=2,818! Left marginals 7 16 13 1326 0,532 4.01 Left costals 5 5 5.16 0.426 8.27 Vertebrals 6 5.41 0.682 Right costals 4 8 9 12,61 5 5.13 0.383 Right marginals 7 7.46 16 13 13.28 0.554 4.17 Padre Island boxes n=894! Left marginals 5 13 13.23 0.473 3.58 Left costals 3 104 157 5.10 0.317 6.22 Vertebrals 3 5 5.16 0,416 Right costals I 5 5 8 6 8,07 5 5.08 0.265 5.23 Right marginals 5 10 15 13 13.27 0.482 3.63 Rancho Nuevo boxes n=549! Left marginals 2 12 14 13 13.17 0,393 2.98 Left costals I 5.05 0227 Vertebrals 3 6 7 4.50 53 4 5.09 0.320 Right costals 3 5 5 629 6 5.04 0.21 9 4.34 Right marginals 3 Il 14 13 13.12 0,530 4.04 Natural nests n=151! Left marginals I 13 14 13 13.17 0.393 Left costals 2.98 5 5.05 0.225 Vertebral s I 4.45 5 5 5,07 0.261 Right costals I 6 6 5 5 5.14 5.04 0.196 3.89 Right marginals 2 I3 N3 15 13,25 0.450 3.40 Galveston head starts n=l /07! Left marginals 5 11 13 13.37 0.532 Left costa ls 3 167 3,98 4 5,12 0.379 Vertebrals 5 7.39 5,46 0.680 Right costals 3 4 4 9 7 5 5 12.45 0.320 Right marginals 4 6.29 12 16 13 13.35 0.513 3.84 Table4. Frequency distribution of vertebral scute counts fordead and live Kernp's ridley hatchlings and juveniles by incubation-handlingca tegory. Incubation-handling category Number of Corral Padre Island Rancho Nuevo Natural Galveston vertebral nests boxes boxes nests head starts Total scutes frequency frequency frequency % frequency frequency % frequency Dead Turtles 3 1 0.7 0 0 0 0 I 0.3 4 4 2.8 0 1.5 5 1.7 5 65 46.1 13 65.0 59 89.4 87.3 ]99 66.8 6 47 33.3 5 25.0 6.1 12.7 65 21,8 20 14.2 2 10.0 3.0 6290 0 24 8.1 7 8 2.1 0 40 20 3 'I 1.0 9 3 I 0,7 0 0.3 Subtotal 141 20 66 71 298 Live Turtles 4 17 0.6 0 I 02 0 5 0.3 23 0.4 1,804 67.4 751 85.9 438 90.7 97.5 941 62.4 4,012 71.4 669 25.0 114 13,0 43 18.9 2.5 440 292 1+68 157 5.9 6 0.7 0.2 7820 0 105 7.0 269 4.8 27 1.0 3 03 0 15 1,0 45 0.8 3 0.'I 0 0 I 0.1 4 0.1 Subtotal 2,677 874 483 80 I/07 5,621 Combined I 0.0 0 2 0 0 0 5 I 0.0 21 0.7 0 0 0,4 0.3 28 0.5 1,869 66.3 764 85.5 497 90.5 140 92.7 941 62.4 4711 71.1 716 25.4 119 13.3 47 8.6 11 7.3 440 29.2 1,333 22.5 3579177 4686.3 8 0.9 0.5 105 7.0 293 5.0 30 1.1 3 0,3 30 0 0 0 15 1.0 48 0.8 4 0.1 0 I 0,1 5 0.1 Total 2+18 894 549 151 1+07 5,919

Table5. Frequencydistribution of left costalscute counts for liveand dead Kemp's ridley hatchlings and juveniles j by incubation-handling category, Incubation-handling category Number of Corral Padre Island Rancho Nuevo Natural Galveston left costal nests boxes boxes nests head starts Total scutes frequency frequency frequency % frequency frequency % frequency Dead 3 I 0.7 0 0 I 9 0,3 4 8 5.7 0 1 5.0 0 0 3.0 5 97 68.8 17 85.0 61 92,4 88.7 238 79.9 6 30 21.3 2 10.0 7.6 638011,3 0 45 15.1 7 4 2.8 50 0 4 I 1,3 8 I 0.7 0 0 0.3 Subtotal 141 20 66 71 298 Live 4 25 0.9 0 0 0 15 1.0 40 0.7 5 2/38 83.6 794 90.8 458 94.8 80 100,0 1,301 86.3 4,871 86.7 6 392 14.6 75 8.6 25 0 180 11.9 672 12.0 7 20 0.7 5 0.6 0 0 11 0.7 36 0.6 8 2 00 0 0 0 0 2 0.0 Subtotal 2,677 874 483 80 I+07 5,621 Combined 3 1 0.0 0 0 0 1 0.0 4 33 1.2 0 I 0.1 0 0 15 1.0 49 0.8 5 2/35 82.9 811 90.7 519 94.5 143 94.7 1,301 86.3 5,109 86.3 6 422 15.0 77 8.6 30 5.5 5.3 180 11.9 717 12.1 7 24 0.9 5 0.6 0 0 80 0 ll 07 40 0,7 8 3 0.1 0 0 3 0.1 Total 2+18 894 549 151 1,507 5,919

207 Table6, Frequency distribution ofright costal scute counts for dead and live Kemp's ridley hatchlings and juveniles,by incubation-handlingcategory.

Incubation-handlingcategory Number of Corral Padre Island Rancho Nuevo Natural Galveston right costal nests boxes boxes nests head starts Total scutes frequency frequency% frequency Vo frequency frequency% frequency Dead 3 0 0 I 1.5 0 I 0,3 7 5.0 0 0 0 7 2.3 5 105 74.5 18 90.0 60 90,9 65 91.5 248 832 6 26 18.4 2 10.0 5 7.6 6 8.5 39 13.1 7 3 2.1 0 0 0 3 I,O Subtotal 141 20 66 71 298 Live 4 19 0,7 0 0 0 15 1.0 34 0,6 5 2,307 86.2 808 92,4 465 96.3 80 100.0 1,351 89.6 5,011 89.1 6 337 12.6 66 7.6 18 3.7 0 138 92 559 9.9 7 13 05 0 0 0 3 0.2 16 0.3 8 I 0.0 0 0 0 0 I 0.0 Subtotal 2,677 874 483 80 1,507 5,621 Combined 3 0 0 I 02 0 0 I 0.0 4 26 0.9 0 0 0 15 1.0 41 0.7 5 2,412 85.6 826 92,4 525 95.6 145 96,0 1,351 89.6 5~9 88.8 6 363 12.9 68 7.6 23 4,2 6 40 138 9.2 598 IO.I 7 16 06 0 0 3 0.2 19 0,3 8 0,0 0 0 0 0 0 I 0.0 Total 2+1 8 894 549 151 1,507 5,919 Table7.Frequency distribution ofleft marginal scute counts for dead and live Kemp's ridley hatchlings and juveniles,by incubation-handlingcategory. Incubation-handlingcategory Number of Corral Padre Island Rancho Nuevo Natural left marginal nests Galveston boxes boxes nests head starts Total scutes frequency 9o frequency% frequency S'o frequency frequency% frequency Dead 9 2 1.4 0 0 0 0.7 10 0 I 5.0 0 0 2 I 0,3 Il 2 1.4 I 5.0 0 0 1.0 12 5 3.5 I 5.0 3 4.5 0 3 9 3.0 13 86 61.0 13 65.0 55 83.3 63 88.7 217 72,8 14 3'7 26.2 3 15.0 8 12,1 8 11,3 56 18.8 15 8 5.7 I 5.0 0 0 16 3.0 I 0.7 0 0 0 9 I Subtotal 14'I 0.3 20 66 71 Live 298 11 0 0 0 I 0.1 I O.O 12 29 1.1 3 0.3 0 0 0,2 4 0.3 37 0,7 13 1,988 74.3 669 76.5 394 81.6 60 75.0 975 64.7 4,086 72.7 14 606 22.6 195 22.3 88 18.2 20 25.0 498 33.0 1,407 25.0 15 52 1.9 7 08 0 28 1,9 87 1.5 16 2 0.0 0 0 0 0 I 0.1 3 Subtotal 2,677 0.1 874 483 80 1,507 Combined 5,621 9 2 OI 0 0 0 I 0,1 2 0.0 10 0 I 01 0 0 4 0.3 I 0.0 11 2 0.1 1 0.1 0 975 64.7 4 0.1 12 34 1.2 4 0.4 4 0 0 0.7 498 33.0 46 0.8 13 2,074 73.6 682 76.3 449 81.8 123 81.5 28 1.9 4,303 72.7 14 643 22.8 198 22,1 96 17.5 28 18,5 I 0,1 'I,463 24.7 15 60 2.1 8 0.9 0 0 96 1.6 16 3 0.1 0 0 0 0 0 4 0.1 Total 2+IS 894 549 15i I+07 5,919 Table 8. Frequencydistribution of right marginal scutecounts for dead and live Kemp's ridley hatchlingsand juveniles,by incubation-handling category.

Nurnbcr of Incubation-handling category right Corral Padre Island Rancho Nuevo Natu ral Galveston marginal nests boxes boxes nests head starts Total sc utes frequency frequency % frequency % frequency frequency % frequency Dead 9 2 1.4 0 2 0,7 10 1 0.7 1 5.0 2 0,7 0 0 01 01.5 0 0 1 0,3 11 12 6 4.3 0 1 1.5 7 2,3 13 89 63,] 17 85.0 57 86.4 58 81.7 221 74,2 14 39 27.7 2 10.0 7 106 13 18.3 61 20,5 15 4 2.8 0 0 0 4 1.3 Subtotal 141 20 66 71 298 Live 9 1 0.0 0 0 0,0 10 0 0 0 1 0 11 1 0,0 0 0303 0 0 0 08 0.0 12 25 0.9 44 9.1 0 0.5 80 1.4 13 1,923 71.8 628 7'1.9 333 68.9 56 70,0 972 64.5 3,912 69.6 14 658 24.6 237 27.1 '106 21.9 23 28.8 515 34.2 1/39 27.4 15 67 2.5 6 0.7 0 1 1.2 9 3 0,6 83 1.5 16 2 0.1 0 0 0 0.2 5 0,1 Subtotal 2,677 874 483 80 1P07 5,621 Combined 9 3 0.'I 0 0 0.1 10 l 0.0 0 1 01 0 0 32 20.0 1 0.0 1 0.2 0 0 08 0.0 11 12 31 1.1 0 3 03 45 8.2 0 05 87 1.5 13 2,012 71.4 645 72.1 390 71.0 114 75.5 972 64.5 4,133 69,8 14 697 24.7 239 26.7 113 20.6 36 23.8 515 34.2 1,600 27.0 15 71 2.5 6 0,7 0 1 07 9 3 06 87 1.5 16 2 0,'1 0 0 0 0.2 5 0.1 Total 2,818 894 549 151 1,507 5,919

Table 9. Mean values of the differences between paired left and right counts for costal and marginal scute series of Kemp's ridley hatchlings and juveniles, by incubation-handling category.

Scute series and incubation-handling Standard cate o iMean error P' Iviarginal Scutes Dead ' 298 0.017 0.038 0.656 Live 5,621 -0.015 0.007 0.039' Corral nests 2,677 -0.030 0.011 0,00' Padre Island boxes 874 -0.046 0.018 0,0 I 2' Rancho Nuevo boxes 483 0,052 0.024 0.035' Natural nests 80 O.063 0.060 0.300 Galveston head starts 1~7 0.011 0.012 0.349 Total 5,919 <.013 0.007 0.062 Costal Scutes Dcads 298 0.030 0.031 0,335 Live 5,621 0.027 0.005 0.001' Corral nests 2,677 0,025 0,008 0.002" Padre Island boxes 874 0.022 0.009 0.013* Rancho Nuevo boxes 483 0.014 0.010 0.162 Natural nests 80 0.000 0.000 1.000 Galveston head starts 1,507 0.038 0.010 0.001' Total 5,91 9 0.027 0.005 0.001' 'Probability that the observed mean difference between paired left and right scute counts was significantly different from zero; i.e.,that therewas significant departure from bilateral symmetry in a matchedpairs t-test.'Indicates significance at alpha 05, 'Not subdivided by incubation-handling category becauseof smail samp]e size. Table10. Summary statistics for ca ra pacial scute counts of dead and live Kemp's ridley ha tchlings from 37 corral nests.

Scute Standard Coefficient of series Ran e Minimum Maximum Mode Mean deviation variation %! Dead n=141! Left marginals 16 13 1328 0.895 6.74 Left costals 522 0,656 1236 Vertebrals 5.67 0,908 16.02 Right costals 87 9 5 5 5,18 0.538 10.40 Right marginals 15 13 1321 0.826 625 Live n=2,677! Left marginais 16 13 1326 0506 3.82 Left costals 5.15 0.411 7.97 Verteb rais 124 49 8 9 5.40 0.666 12.34 Right costals 5 5 5.13 0.373 727 Right marginals 16 13 13.29 0.536 4.04

Table11. Summary statistics for carapacial scute counts of dead and live Kemp's ridley hatchlings from 11 Padre Isla~d boxes.

Scute Standard Coefficient of series Range Minimum Maximum Mode Mean deviation variation %! Dead n=20! Left marginals 13 12.95 l.050 8.11 Left costals 1045 5 156 7 5.05 0,394 7.80 Vertebrals 5.45 0,686 Right costals 5 5 12.59 6 5.10 0.308 6.04 Right marginals 10 14 13 12.95 0.759 Live n=874! Left marginals 12 13 13.24 0.451 3.41 Left costals 157 5.10 0.315 Vertebral s 6.18 5 5 5,15 0.406 Right costals 8 6 5 5 7.88 5.08 0264 5.21 Right rnarginais 12 15 13 13.28 0.472 3,56

Table12. Summary statistics forcarapacial scutecounts ofdead and live Kemp's ridley hatchlings from seven Rancho Nuevo boxes.

Scute Standard series Coefficient of Range Minimum Maximum Mode Mean deviation variation %! Dead n =66! Left rnarginals 14 l3 13.08 0.404 Left costals 3.09 1253 4 6 7 5.08 0,267 5.25 Vertebrals 5.11 0.434 Right costals 5 5 850 6 5.05 0.369 7.32 Right marginals Il 14 13 13.06 0,425 3.26 Live n = 483! Left marginals 12 14 13 1,3.18 0.390 Left costals 2.96 5,05 0.222 Vertebra! s 5 4 6 7 4.39 5 5 0.302 5.93 Right costals 5 6 5.04 0.190 3.76 Right marginals 12 14 13 13.13 0.543 4.13

210 Table13. Summary statistics for carapacialscute counts of deadand live Kemp'sridley hatchlingsfrom five natural nests.

Scute Standard Coefficient of series Range Minimum Maximum Mode Mean deviation var iation %! Dead n =71! Left marginais 13 13,'! I 0.318 2.43 Left costals 146 6 135 5 5.11 0.318 Vertebrals 5 5 5.13 0.335 6.54 Right costals 5.08 0,280 5.51 Right marginals 13 14 13 13.18 0.390 2.95 Live n=80! Left marginals 13 13.25 0.436 3.29 Left costals 145 6 135 5 5.00 0,000 0.00 Vertebrals 5 5.03 0,157 3.13 Right costals 5.00 0.000 0.00 Right marginals 13 15 13 13.31 0.493 3.70

Table 14.Frequencies of the modal scutenumber for Kemp'sridley seaturtles,

13 left 5 left 5 5 right 13 right marginals costals vertebra ls costals marginals Group frequency frequency % frequency % frequency % frequency Hatchlings' Corral nests n=2,677! 1,988 74.3 2~8 83.6 1,804 67.4 2,307 862 1,923 71.8

Padre 1 sland boxes n =874! 669 76.5 794 90.8 751 85.9 808 92.4 628 71.9

Rancho Nuevo boxes n =483! 394 81.6 458 94.8 438 90.7 465 96.3 333 68.9

Natural nests n =80! 60 75.0 80 100.0 78 97.5 80 100.0 56 70.0

Juveniles'

Galveston head starts n=l f07! 975 64.7 1POI 86.3 941 62.4 I +51 89,6 972 64.5

Subadultss n =96! 95 99.0 96 100.0

Adult Females' n=154! 89 57.8 151 98.1 135 87.7 148 96.1 123 79.9

'This study; live turtles only. From Carr and Caldwell 956!. 'From Chavez ct al. 967, 1968!.

211 Clearly,the most interesting observation regarding variation invertebral scute count was its apparent relation to incubation-handlingtechnique. Various aspects of the data suggested greatest departure from normal vertebral countsin thecorral nests and Galveston head starts, least in naturally incuba ted turtles, and intermediate in the Padre Islandand Rancho Nuevo boxes, This same trend recurs inother scute series, aswell as in overall scute pattern. Variation in the Costal Scute Series Patternsofvariation inleft and right cosial scute series were very similar Tables 2 and 3!, reflecting their bilateral symmetry.Among the 5,919 specimens, leftand right costals exhibited the same range !, minimum!, maximum 8!and mode !. Thetwo costal series exhibited the least variation incount among the carapacial scute series with theexception of the nuchal>, asindicated by the low ranges and standard deviations Tables 2 and3!. The means and modesfor each costal series were closer to one another in allincubation-handling categories than the means and modesofany other scute series, further depicting the stability of costal scute counts as compared tocounts inthe other scute series. Variabilityin scute counts within both costal series was greatest in the corral nest category, followed bythe Galvestonhead starts, Padre Island boxes, Rancho Nuevo boxes and natural nests as shown by thestandard deviationsandranges similarinPadre Island and Rancho Nuevo boxes, Table 3!.There was greater variation among deadturtles than among live turtles Tables 5 and6!. The proportion of turtlesexhibiting the modal count of five costals,whether left or right, was lower for dead turtles than for live ones in all incubation-handling categories. This suggestedgreater stability incostal scute count inlive turtles than in those that died as ha tchlings orembryos. Variation in the Marginal ScuteSeries Bilateralsymmetrybetween countsof left and right marginals wasevidenced bytheir identical ranges !, minima 9!,maxima 6! andmodes 3! amongthe 5,919 turtles Table 2!.The grea test range inmarginal scute counts occurred in thecorral category, followed in descendingorder by Galveston head starts, Padre Island boxes, Rancho Nuevo boxesand natural nests, Mean scute numbers and standard deviations indicated that marginal scutes were the most variablescute series in PadreIsland box, Rancho Nuevo box and natural nest categories, and were exceeded in variabilityonly by the vertebral series in thecorral nest and Galveston head start categories Table 3!. Aswas the case with ver tebrals and costals, thecorral nestand Galveston headstart categories exhibited thegrea test variationincounts of marginals ascontrasted with natural nests which exhibited the least. However, the differences invariability o marginalf coun tsamong theincubation-handling categories were slight, suggesting mores tabilityin marginalscute counts among the five incubation-handling categories than for the other scute series. Deadversus live comparisons formarginal count variability were not as clearly diHerenhated aswere thecases for thecostal and vertebral series. Marginal scute counts showed generally greater variability indead than in live turtles, butthe percentages ofturtles exhibiting themodal count of 13 marginals leftor right! followed a different rank order amongincubation-handling categories Tables 7 and8! thanfor other scute series Tables 4-6!. Tests af Symmetry in Paired Counts of Scute Series A seriesoft-tests was performed onmean differences between leftand right counts ofcostals and marginals forall 5,919turtles, and for various subgroupings Table9!. Tested was the null hypothesis ofbilateral symmetry inleft and rightcounts. The difference was obtained foreach specimen bysubtracting theright scute count from the left scute count,for marginaland costal series separately. Formarginals, themean difference forthe entire sample of5,919 turtles was not significantly different from zero, suggestingbilateral symmetry inscute counts, though certain groupings exhibited significant asymmetry Table 9!. Forexample, marginal counts were significantly asymmetric in live turtles, but the dead specimens showed no significantasymmetry. The live turtles from corral nests, Padre Island boxes and Rancho Nuevo boxes had significantlyasymmetric marginal counts, even though Galveston head starts and turtles from natural nests did not. Incontrast, leftand right costals forthe entire sample of5,919 turtles were significantly asymmetric Table 9!.Yet, onlythe live turtles exhibited significant asymmetry incostal counts, while the dead turtles did not. Substantial differencesinsample size between thelive and dead subgroupings mayhave contributed tothe difference incostal scuteresults for these two subgroupings. Thedead subgroup was the smaller of thetwo, so its t-test of themean differencebetween counts ofleft and right costal scutes was more conservative thanthat for the live subgroup. Forcostals and marginals, thenatural nest and Rancho Nuevo box categories didnot have significant mean differencesbetween leftand right counts, whereas corral nest, Padre Island box and Galveston headstart categories exhibitedsignificantly asymmetric counts.In those incubation-handling categoriesexhibiting costalcount asymme- try,the greater number ofscutes occurred inthe left series Table 9!.Among theincubation-handling categories d isplayingasymmetrical marginal counts, corral nest and Pad reIsland box categories hada greaternumber ofscutes inthe right series, while the Rancho Nuevo category hadthe greater number inthe left series. Itis tempting tosuggest, basedonthe corral nest and Padre Island box categories, thatthere isa relationshipbetweenthe increase inright

212 marginalsand left costals.This may reflect some developmental mechanism that balances combined costal and marginalscute counts on left and right sides of thecarapace such that overall symmetry in leftand right counts is maintained. However, this balance was not evident in the Rancho Nuevo box category. Variation in Scute Series by Incubation-Handling Category CorralNests Severalof thepreceeding analyses indicated that hatchlings from corral nests exhibited a higher degreeof variabilitythan those from any of theother incubation-handling category. Corral nest hatchlings had the greatestranges, standard deviations with one exception: left marginals for Galveston head starts!, and coefficients of variation for scutecounts in eachof the scuteseries Table 3!. However, corral nestsrepresented the largest incubation-handlingcategory ,818 turtles!,or 48percent of all turtlesexamined. Also, greater variability in dead than in live turtles from corral nests for all scute series Tables 4 and 10! contributed to the overall variability in the corralnest category. Within the corral nest category, the most variable scute series was the vertebral, followed by the marginal, then the costal. GalvestonHead Starts Juvenileturtles from the Galvestonhead start categorywere secondto the corral nest categoryin degreeof variability within scute series, Based on ranges, standard dcvia tions and coefficients ofvariation for eachscute series Table3!, Galvestonhead starts had slightly lower scutevariability than hatchlingsin thecorral nestcategory, though consistently higher variability than Rancho Nucvo box, Padre Island box and natural nest categories.Galveston turtles, like hatchlingsfrom corralnests, showed greatest variatio~ in the vertebralseries, followed by the marginal then thc costalseries. Thereader is cautionedabout drawing broad conclusions based on our observationsof the 1980year-class of Galvestonhead start juveniles as compared to hatchlingsof the1981 year-class. We only witnessedthe incubation- handlingtechniques used in 1981,and not thoseused for the 1980year-class of Kemp'sridlcys head started at the GalvestonLaboratory. We had no data to compareenvironmental conditions between the two years, thus raising the possibilitythat the observed differences between the Galveston head starts from the 1980year-class and the other incubation-handlingcategories from the 1981 year-class might be due to differences in environmentalvariables other thanincubation-handling techniques. Also, we hadno deadsubgroup to compareto the live subgroupfrom the Galvestonhead start category, thus there was no opportunityto investigatechanges during ontogeny within this category.Finally, we had no wayof determiningwhether incubation-handling was consistent from year to yearfor thosenests destined to provide hatchlingsfor headstarting. Padre Island Boxes Hatchlingsfrom the Padre Island boxeswere consistentlyintermediate in scuteseries variabilitybetween thc two extremesrepresented by the highly variablecorral nestand Galvesto~head start categorics and the slightly variable Rancho Nucvo box and natural nest categories. This is shownby ranges,standard deviations and coefficients of variation for the scute series Table 3!. Costals were thc least variable of the scute series for both live and deadhatchlings in thePadre Island box category,with marginalsshowing greatest variability in the deadsubgroup, and vcrtebralsshowed greatest variability in the live subgroup. The meanmarginal scute counts of deadhatchlings in the PadreIsland box categorywere lower than thosefor live turtles in that category Table11!. This tendencytoward lower countsof marginalsin dead PadreIsland hatchlings Table11! was unique among all live-deadcontrasts, scute series and incubation-handlingcategories Tables 4-8 and 10-13!.However, only 20 dead specimensin the PadreIsland box categorywere examined. RanchoNaevo Boxes In general,both live and deadturtles in theRancho Nuevo box categoryexhibited the least amountsof variability in scutecounts Table12! as judged by ranges,standard deviations and coefficientsof variation in scutecounts for all incubation-handlingcategories except the natural nests.The dead turtles showedgreatest variabilityin thevertebral series, while live turtlesshow greatest variability within themarginal series. Costals were the most stable series for both live and dead hatchlings in thc Rancho Nuevo box category. Natural Nests The least overall variation in scute counts was observed in the natural nest category Table 13!, no doubt due in part to the small samplesize for this category,Indeed, in nearlyall scuteseries, the turtles from natural nestsshowed the lowest variation of all amongthe incubation-handlingcategories. The greatest variation for natural nestturtles wasin the marginalseries, followed by thevertebral and costalseries which exhibited100 percent modal counts in the live subsample!. Departures from Modal Scute Counts Table14 provides for comparisonsamong scute series and incubation-handlingcategories of our study,as well as within sub-adult Carr and Caldwell, 1956! and adult female Kemp's ridleys Chavez et a/., 1967,1968! from other studies.Only live turtleswere used in thiscomparison because the other studies were based on live specimens. Costalsshowed less departure from the mode than vertebralsor marginals,the latter of which had the lowest percentageof turtles with the modal count, An exceptionto this was that the corral nestand Galvestonhead start categoriesexhibited a highpercentage of turtleswith modalcounts for costals,and percentages for modalcounts of

213 vertebralssimilar to those for marginals. Ingeneral, corral nest and Galveston head start categories exhibited greater departuresfrom modal counts for all scuteseries than did the other categories. Adultfemale Kemp's ridleys Chavez efal,, 1967, 1968! also exhibited higher proportions ofmodal scute counts for costalsand vertebrals than for marginals. The percentages ofadult females exhibiting modal counts for vertebrals and costalsfell within therange of comparablepercentages for theRancho Nuevo box, Padre Island box and na tural nest categories.Inthe adult females, the proportion with modal counts for left marginals was considerably lower than that forthe right marginals. It was also the lowest of the percentages forany group of turtles in Table 14. On the other hand, theproportion with modal counts for the right marginal series was higher in the adult female than for any of the categoriesinour study. Only costa l scute counts were available for subadult Kemp's ridleys Carr and Caid well, '1956!, andthe proportions of turtleswith modal counts were on the high end of therange, similar to thoseof theRancho Nuevobox and natural nest categories of our study. Variation in Scutation Amongthe 5 919turtles examined inour study, 264 different scute pa t terns or scutations were recorded: 77patterns amongthe dead turtles n = 298!and 233 patterns among the live turtles n = 5,621!.Included among the 77 scute patternsfound in the dead specimens were 31 pa t ternstha t didnot occur in live turtles. The modal scu te pa ttern 3- 5-5-5-13!was observed in2,646 4,7 percent!turtles. The next most common pat terns were: 13-5-5-5-14 in444 turtles or7.5 percent!, 13-5-6-5-13 in 438 turtles or 7,4 percent!, 14-5-5-5-14 in384 turtles or 6.5 percent! and 14-5-5-5-13 in 373turtles or 6.3 percent!. It is interestingthat the most common of theaberrant scute patterns involved additional marginalorvertebral scutes. A scutecount lower than the mode was a muchrarer occurrence thar, one higher than themode for any scute series. In casesin whichthere were above-modal vertebral and costal scute counts, the additionalscu tes most often appeared atthe posterior end of thescute series as a splitfifth costalor fifthvertebral scute,or as a su pernurncrary scute between the fourth and fifth vertebrals. Extra marginal scutes, on the other hand, nearlyalways appeared atthe anterior end of theseries as a splitsecond, third or fourth marginal scute. Themean number ofdifferent scute patterns pernest i,ep r clutch!within each incubation-handling category providedanother measure ofrelative variability in scuta tion. The mean number ofscute patterns per clutch was highestinthe corral nest category 0,8 patterns!, followed by the Padre Island box 1,2 patterns! and Rancho Nuevo box 9.1patterns! categories. The small amount of informationavailable for natural nests suggested even fewer variantsper nest approximately4.5 patterns!. We were unable to makesuch calculations for theGalveston head starts,because wehad no data on how many clutches were represented inour sample. Departures from Modal Scute Patterns Theproportion ofturtles exhibiting themodal scute pattern 3-5-5-5-13! wasnot significantly different inlive and deadsubgroupings in the 5,919 specimens inour study Table 15!. In that regard, the only significant difference occurredforthe corral nests inwhich more of the live turtles 1.7 percent! exhibited themodal scute pattern than did thedead turtles 6.2 percent!, asshown inTable 15and Figure 2.The greatest percentage ofindividuals showing the modalscu te pattern was in theRancho Nuevo box category, followed closely by natural nests, with the Padre IsIand boxcategory being about 10percent less.Live turtles from corral nest and Galveston headstart categories hadnearly thesame proportions about 42 percent! ofindividuals with modal scute patterns,

Discussion Whilemostearlier researchers examined carapacial scutevariation inKemp's ridley only with regard toindividual scuteseries, we expanded our examinations to include variation in frequenciesof scute patterns as a whole, Comparisonsamongthe five different incubation-handling categories ofour study, each of which was exposed to differentenvironmental conditions during ontogeny, provided testsof the working hypothesis thatscute variability mightbc related to environmental circumstances during ontogeny. Todetermine if there were differences in scute variationwithin a givenincubation-handling categoryduring development, weexamined different ontogenetic stagesrepresented bydead younger hatchlings andembryos! andlive specimens laterstage!. One might expect increasedvariability tobe evidenced inthe dead turtles relative tothe live ones if selection were acting toremove extremephenotypes from the papula tion, Evidence insupport of this hypothesis would indicate that scu te variation maybe related tothe survivability ofthe individual. Thiscould act through underlying problems ofviability associatedwith somescute pattern variants Zangerl, 1969; Hill, 1971!. Wehave no reason tobelieve that scute pattern itself is a selectivefactor. Pritchard 969a! commented that the hydrodynamicsofa turtle's shell would be little a ffectedbythe arrangement orquan tity of scu te seams and the overall appearanceofa tur tie's shell islittle affected byscute variation, Therefore, it seems more likely that an abnormal scute patternisan external phenotypic expression ofsome underlying physiological ormorphological problemwhich may affecttheviability ofthe turtle. The senior author has noted, onseveral occasions, thatextreme deformity andalbinism

214 Table15. G-tests of thenull hypothesisof independencebetween scute pattern and incubation-handling category for Kemp'sridley hatchlingsand juveniles,

Pairwise contrasts between incubation-handlingcategories Ga Dead vs, Lives Corral nests 13.9956» Padre Island boxes 0.5418 Rancho Nuevo boxes ] .6555 Natural nests 0.0424 Combined total! 0.1838 Live only Corral nests vs. Padre Island boxes 11.4840» Corral nests vs. Rancho Nuevo boxes o5 3000» Corral nests vs. Natural nests 7,7996' Corral nests vs. Galveston head starts 0.1186 Padre Island boxes vs. Rancho Nuevo boxes 17.3342' Padre Island boxes vs. Natural nests 2.4970 Padre Island boxes vs. Galveston head starts 8.0904' Rancho Nuevo boxes vs. Natural nests 0.1830 Rancho Nuevo boxes vs. Galveston head starts 46.4818» Natural nests vs. Galveston head starts 7.1012* " indicatessignificance at alpha 0.05,leading to rejectionof thenull hypothesis;i.c., concluding that the two groupsbeing comparedare significantlydifferent. There were no dead turtles in the Galveston head start category, so no tests were possible with that category.

70

60

50

g 40

30

20

10

0 ALL CORRAL PADRE RANCHO NATURAL ALVESTON CATEGORIES NESTS ISLAND NUEVO NESTS HEAD COMBINED BOXES BOXES STARTS

INCUBATION-HANDLING CATEGORY Figure2, Proportionsof live and deadKemp's ridley hatchlings and juveniles that exhibitedthe modalscute pattern 3-5-5-5-7.3h by incubation-handlingcategory sample size is shownabove the vertical bars for eachcategory!. 215 indead hatchling sea turtles is nearly always accompanied byaberrant scute pattern, yet the abnormal scute pattern itselfwas certainly not the cause of death.Certain abnormal scute patterns were never observed in liveturtles, but onlyin dead specimens, There were 31 of these"lethal scute patterns" observed in this study. A highdegree of carapacialscute variation was encountered in the young Kemp's ridleys in ourstudy, despite variousstatements in the literature regarding the great stability of scutation inKemp's ridley as compared toolive ridlcy Pritchard and Mirquez, 1973; Frazier, 1984!, Though the most common carapacial scute pattern we observed in Kemp'sridley was13-5-5-5-13, it occured in only44 percent of the5,919 turtles we examined, thus it constituted lessthan half of our relativelylarge sample. Differencesin amounts ofcarapacial scute variation among the five incubation-handling groupswe examined werenotable. Based on our qualitative judgements in ranking the five incubation-handling categories, themost carefullyhandled and undisturbed eggs Padre Island boxes and natural nests! produced turtles with less carapacial scutevariability, and the most roughly handled eggs Corral nests! produced turtles with higher levels of scute variation.Even if the1980 year-class represented bythe Galveston head starts is ignored,the trend from higher variabilityincarapacial scutes ofturtles from roughly handled eggs to lower variability inturtles from carefully handledeggs or unhandled eggs appears innearly all analyses ofthe data. For the Galveston head starts we possessed theleast complete information regarding actual incubation conditions andno dead subgroup forthis sample. Indeed, largeand significant differences inamount ofvariability were present among thegroups ofturtles derived from eggs incubatedorhandled indifferent ways. These results suggest that environmental circumstances during incubation affectscute variability of theresultant hatchlings. Severalenvironmental factors have been implicated asagents causing scute pattern varia tion see review by Ewert, 1979!,all of them acting during the embryonic phase of the life cycle, Handling ismerely one of these factors Hill, 1971!,so we have considered both incubation and handling together rather than handling alone. Incubation temperaturesand humidity regimes varied greatly among our five incubation-handling categories, with the widest differencesinthis regard being between theconcrete block house and the natural beach. There was no great disparity in carapacialscute variability within the 1981 year-class hatchlings derived from the two incubation-handling categoriesincubated inthe concrete block house viz, Rancho Nuevo and Padre Island!. Both of these categories were consideredtobe carefully handled eggs. The greatest disparity in carapacescute va ria t ionoccurred between the two groupsof eggsincubated onthe Rancho Nuevo beach, with roughly handled eggs represented bycorral nests producingturtles with the grea test a mount of scute varia tion and untouched eggs represented byna tural nests which producedturtles with far less scute variability, Though there were numerous factors, liketemperature andhumidity, thatmay have affected embryonic development ofthe turtles from our five categories, wehad no quantitative measurementsofthese potentially significant factors. Instead our conclusions arebased on qualitative judgements ofthe environmental differences andobvious distinctiveness amongthe five incubation-handling categories. While nocause and effect relationship hasbeen demonstrated, webelieve that the differences in the degrees of scute variabilityamong the categories areattributable, atleast in part, to the ways in which the eggs were treated ornot treatedafter oviposition. Comparisonsofcarapacial scute variability between dead and live individuals were made among the four incubation-handlingcategories examined inMexico. On the whole, dead turtles exhibited somewhat more variation thanlive turtles, and correspondingly lowerrelative frequencies percentages! inthe modal scute class for a given series!than live turtles. This isparticularly notable inthe corral nest category, asthis was the only category witha large sampleofdead individuals. Ourresults suggest thatdead animals theyoungest ontogenetic stagein our study! had morevariability in carapacial scutes than live animals, indicating that selection was acting to remove extreme phenotypesfromthe younger ageclasses. Nevertheless, theseresults areno more than suggestive. Comparisons betweencategories ofKemp's ridley hatchlings inour study and categories represented byjuveniles Carr and Caldwell,1956! and adult females Chavez etal., 1967 and 1968! revealed nomarked differences. Confirmation that selectionactson the extremes ofcara pacia 1scu tation phenotypes inKemp's ridIeys awaits further study. Interestin possible significance ofcarapacial scute variation asan index of the effects ofdifferences inincubation- handlingtechniques mayincrease nowthat we have demonstrated considerable variation incarapacial scutecounts amongturtles obtained from eggs mcubated andhandled ina varietyofways, The process inwhich phenotypic variationoccurs asa consequenceofphysical environmental factorsacting on sensitive portions ofa developing organismwasreferred toby Smith-Gill 983! as phenotypic modulation. Scutes orscales arephenotypic variables thatare easily quantified and which havebeen presumed toreflect, atleast inunderlying genetic variation Fox, 1975!. Severalstudies ofsquamate reptiles have examined theevolutionary significance ofscale count variation atdifferent ontogeneticstages see review by Sirnbotwe, 1981!. However, the carapacial scutes of turtlesdo not seem to have

216 receivedsuch attention, possibly due to therelatively low frequenciesof individualvariations, and to thephylogen- etic stabilityof the overallscute pattern Zangerland Johnson,1957; Zangerl, 1969!. The large extent to which phenotypicmodulation appears to determinethe degreeof variabilityin carapacialscutes of turtlesbrings into questionthe role that heritabilityplays in the observedvariation. It seemslikely that the extentof phenotypic modulationpresent is itself a heritablefeature. Nevertheless, each generation may be subject anew to selectionon the varietyof carapacialscute phenotypes produced by phenotypicmodulation. As notedby Smith-Gill983!, this variationmay not necessarilybe adaptive,or it maybe non-adaptive, Hill 971! foundthat adult oliveridleys L. olivacea!in Surinam were less variable in scutation than hatchlings, and he implied that increased variation was not desirable.A studyof greenturtles Cheloniamydas! by LeToquin, Gamel and Trotignon 980! alsoimplied that the scutesof adult seaturtles were less variable than those of hatchlings. Theseprevious studies support the hypothesis that selectionacts against extremes in carapacialscutation phenotypes, Conclusions Implicationsof our findingsto theconservation of endangeredsea turtles such as Kemp's ridley are numerous. We havepresented evidence confirming that the manner in whicheggs of Kemp'sridley arehandled after oviposition hasa markedeffect on carapacialscu te variability. Further,it isapparen t that increasedvariability is undesirable,with moreextreme var iants being lessviable and selectivelyremoved from the populationover the courseof ontogenetic development.Therefore, transplantation, translocation and artificial incubation of seaturtle eggs should be evaluated with concernfor their possibleeffects on viability of hatchlings. When the mechanismsof scuteabnormality are better understood,scute patterns may becomeuseful external indicatorsof phenotypicor genotypicdeficiencies related to futureviability. Many selective pressures are removed by artificialincubation but arereplaced by others,so any external indicator of phenotypicor genotypicdeficiency would be a usefultool in determiningthe long-termeffects and suitability of artifical incubationand variousother techniquesinvolving theincubation and handling of eggs,For example,culling o fabnormal individuals based on the presenceof certainknown, lethal scutepatterns might be an effectivemeans of maintaininghealthy stocks. Developmentalplasticity is theprocess whereby environmental input altersthe phenotype of an organism Smith- Gill, 1983!.Environmentally determined sex in turtles is a well documentedexample of this sort of alteration Bull, 1980,1983!. The possible consequences of artificially alteringsea turtle sexratios through control of conditionsduring incubationof eggshave been widely publicized Mrosovskyand Yntema,1980; Yntema and Mrosovsky1980, 1982; M orreale elal., 1982; Mrosovsky, 1982!, Smith-Gill 983! reviewedthe concept of critical periodduring development with regard to phenotypicplasticity. It holds that organsare most susceptibleto environmentalinfluences during periodswhen they aremost actively differentiating. The critical periodduring which temperatureinfluences gonadal differentiation in turtles is the middle third of incubation Bull and Vogt, 1981; Yntema and Mrosovsky, 1982!. Similarly, the carapacialscute pattern must alsohave a criticalperiod for differentiation Ewert, 1979!,Research to determine when this critical period occurs could aid conservationists in making wise choices about when to move eggs,if at all, and would doubtlesshelp in determininglinkages between aberrant scute patterns and other organ systemabnormalities which are more directly relatedto long-termviability. Acknowledgements We thank Mexico's Instituto Nacional de la Pescaand U.S. Fish and Wildlife Service for the opportunity to work at Rancho Nuevo on the Kemp's ridley recovery effort. We also acknowledge the assistance of Dr. George Zug, Division of Amphibians and Reptiles,U.S. National Museum of Natural History, the Departmentof Zoology at SouthernIllinois University and the Departmentof Wildlife and FisheriesSciences at TexasAt|rM University which provided technicalsupport. We are particularly indebted to Mr. RobertTruland of the CheloniaInstitute for his generous support of our work, and to Angela Mast and Christine Tweet for their support. We are grateful to the World Wildlife Fund and Dr, JamesI, Richardson for their contributions to the study.

Literature Brongersma, L.D. 1968.Notes upon some turtles from Surinam. Koninklijke NederlandseAkadernie van Wetenschappen,Proceedings, SeriesC, Biologicaland Medical Sciences712!:114-127. Bull, J.J.1980. Sex determirtation in reptiles. Quarterly Reviewof Biology55:3-2L Bull, J.J. 1983. Evolution of sex determining mechanisms, The Bertjamirt/Cummings Publishing Company, inc., Menlow Park, California, xx plus 316 pp. Bull, J.J.and R.C. Vogt. 1979.Temperature dependent sex determination in turtles. Science206:1186-1188. Bul1, J.J. and R.C.Vogt. 1981.Temperature-sensitive periods of sexdetermination in emydid turtles.journal of ExperimentalZoology 218:435-440.

217 Carr,A.F. and D.K, Caldwell. 1956. The ecology and migrations of seaturtles; 1. results of fieldworkin Florida,1955. American MuseumNovitates 1793, 23p. Chavez,H.,M. Contreras G,and T.P.E, Hernandez D.1967. Aspectos biol6gicos y proteccidn dela tortuga lora Lepidochelys kempi Garman!,en la costa de Tarnaulipas, Mexico. Instituto IVanonal deInvestigaciones Bioldgicas Pesquera, Publicacion Numero 17, 40 p. Chavez,H.,M. Contreras G,and T.P.E. Hernandez D.1968. On the coast ofTamaulipas. International Turtle& TortoiseSociety Journal 2!;20-29, 37 and 2!:16-19, 27-34. Coker,R.E. 1905a. Diversity in thescutes and bony plates of Chelonia.Science 21:384-385. Coker,R.E. 1905b. Gadow's hypothesis of"orthogenetic variation" inChelonia. Johns Hopkins University Circular 17B, p, 9-24. Coker,R.E. 1905c. Orthogenetic variation? Science22:873-875. Coker,R.E. 1910. Diversity in scutesof Chelonia.Journal ofIVIorphology 21:1-75. Deraniyagala,P.E. P. 1939. The tetrapad reptiles of Ceylon, I. Testudinates andCrocodilians. Colombo Museum, Colombo, Ceylon,xxxii plus 412pp. E wert,M A.1979. The embryo and its egg: development andnatural history, p.333413. In; Harless, M.and H. Morlock Editors!, Turtles;Perspectives and Research, John Wiley and Sons, N.Y., xvi plus 695 p. Fox,W. 1948. Effect ottemperature ondevelopment ofscutellat ionin the garter snake, Tham nophis elegans atrat us. Copeia 1948: 252- 262. Fox,S.F. 1975. Natural selection on morphologicalphenotypes of thelizard Uta stansbuariana. Evolution 29:95-107. Fox,W., C. Gordon and M H.Fox. 1961. Morphological effectof low temperature during the embryonic development ofthe garter snake,Thamnophis elegans, Zoologica 46:57-71. Frazier,J.C, 1984. Analisis estadistico dela tortuga golfina Lepidochelys olivacea Escholtz! deOaxaca, Mexico. Ciencia Pesquera 4:49-75. Gadow,H. 1899.Orthogenetic variation inthe shells ofChelonia, p.207-222, Part3. In: Willey, A. Editor!Zoological Results Basedon Material from New Britain, New Guinea, Loyalty Islands and Elsewhere, Collected During the Years1895,1896, and1897. Cambridge University Press, Cambridge, England. Hildebrand,S.F,1930, Duplicity andother abnormalities indiamond-back terrapins. Journal ofthe Elisha Ivfitchell Scientific Society 46;41-53. Hildebrand,S.F, 1938. Twinning in turtles.Journal ofHeredity 29:243-253. Hill,R,L. 1971. Surinam turtle notes - I.Polymorphism ofcostal and vertebral laminae inthe sea turtle Lepidochelys olivacea.Stichting NatuurbehoudSuriname STIlVASU!, Atededelingen 2:1-9. LeToquin, A., E. Gamel and J. Trotignon. 1980. Morphologic, croissance individuelle etdynamique despopulations dela tortue verte Chelonia mydas L !au Banc D'Arguin Republiquelslamiquede Mauritanie!, Revued'Ecologic LaTerre et la Vie! 34 271- 302. Limpus,C.J., V. Baker and J.D, Miger, 1979. Movement induced mortality ofloggerhead eggs. Herpetologica 35:335-338. Lynn,W.G. and M.C, Ullrich. 1950. Experimental production ofshell abnormalities inturtles. Copeia 1950:253-262. Morrealc,S J,,G J.Ruiz,J.R.Spotila andE A.Stand ora. 1982. Temperaturedependent sexdetermination: currentpractices threaten conservation of sea turtles. Science 216:1245-1247. Mrosovsky,N.1982. Sex ratio bias in hatchhng seaturtles from artificially incubated eggs. Biological Conservation 23:309-314. Mrosovsky,N.and C.L. Yntema 1980. Temperature dependence ofsexual differentiation in sea turtles: implications for conservationpractices. Biological Conservation 18:271-280. Newman,H.H. 1906. The significance ofscute and plate "abnormalities" inChelonia. Biologt'cal Bulletin ]0:68-114. Parker,G.H. 1901. Correlated abnormalities inthe scutes and bony plates ofthe carapace ofthe sculptured tortoise. American Nat uralist 35:17-24. Pritchard,P.C.H. 1969a. Sea turtles of theCuianas. Bulletin of the Florida State Afuseum 13:85-140. Pritchard,P.C.H.1969b. Studies ofthe systematics andreproductive cycleofthe genus Lepidochelys. Ph.DDissertation, University of Florida,Gainesville, xii plus196 p, Pritchard,I'.C.H. and R. Marquez M.1973, Kemp's ridley turtle or Atlantic ridley Lepidochelys ketnpi.International Unionfor the ConservationofNature and Natural Resources Monograph IVo.2: 30p. Pritchard,P.C.H. and P. Trebbau. 1984. The turtles ofVenezuela. Contributions toHerpetology 2:1-403. Sirnbotwe,M.P.1981. Natural selection inthe lizard Eumeces obsoletus Lacertilia: Scincidae!. Amphibia-Reptilia 2:143-151, Smith-Gill,S J.1983. Developmental plasticity:developmental conversion versusphenotypic modulation. American Zoologist 23: 47-55. Sokal,R.and F J.Rohlf. 1981. Biometry. Second Edition, W.H. Freeman andCompany, SanFrancisco, California, 849p. Wandolleck,B.1904, Eine bucklige Testudo graeca L. Zoologische Jahrbucher, Abtheilung furSystematik 20:151-166.

218 Yntema,C.L. 1976. Effects of incubationtemperatures on sexualdifferentiation in the turtle Ctu.'Iydra serpentina. Journal of lvforphoiogy150:453-462, Ynterna,C.L. and N. Mrosovsky.1980. Sexual differentiation in hatchling loggerhead diana, Geology 10:341-362.

21.9 Morphometry of Captive-RearedKemp's Ridley Sea Turtle Andri M. Landry, Jr. '

Morphometryandgroivth were characterized forhead started Kemp's ridley sea turtles <'Lepidochelys kernpi! from the 19B4 yearclass. Straight-line carapace andplasl ron length and width measurements weretaken biweekly from 100 hatchhngs in nine clutchesduring late August 19B4 through early May 1985. Growth rate and anatomical growth pattern were described through analysisofmorphometric data.Carapace andplastron dimensions wereanalyzed byclutch, feeding rate, raceway location and hydrologicalparameters, Regression equations were developed forcarapace andplastron length-width relationships. Aninternational recovery program designed to save the endangered Kemp's ridley Lepidochelyskempi! from extinctionhas incorporated, among other things, a headstart project inwhich hatchlings arereared incaptivity at NationalMarine Fisheries Service's Galveston Laboratory for 10to 11months. Head starting provides care, maintenanceandhusbandry ofKemp's ridleys to increase their survival and optimize growth during their first year, Projectssuch as this provide opportunities forconducting research ongrowth of captive species such as the Kemp's ridley. Growthstudies of captive-rearedKemp's ridleys conducted byCaldwell <19621, M6rquez 972!, Pritchard and Mlrquez973!, Klima and McVey 982!, McVey and Wibbels 984!, Caillouet and Koi 985!, Fontaine etal. 985! andCaillouet etal. 986! have defined growth interms ofbody weight, The same isgenerally true for growth studies ofcaptive-reared species other than Kemp's ridleys Vchida, 1967; Shckney, White and Perlmutter, 1973; LeBrun, 1975;Witham and Futch, 1977; Wood and Wood, 1981; Frazer and Schwartz, 1982; Hirth, 1982; Nuitja and Uchida, 1982;and Rajagopalan, 1984!. Studiesdefining growth of captive-reared Kemp's ridlcys in terms of carapace and/or plastron measurements are fewand incomplete. Data on morphometric growth of head started ridleys mainly include regression formulas definingtherelationshipof carapace length and width Caillouetet al.,1986! and growth incarapace length following releasefrom captivity McVey and Wibbels, 1984!. Nostudies have systematically examined morphometric growth ofcaptive-reared ridleysor the relationship ofgrowth rate to constitutent variables ofa headstart project such as clutch,feeding rate, rearing location and hydrological factors. Thestudy reported herein was designed to: 1. characterizemorphometry andgrowth of the 1984 head start year-class ofKemp's ridley sea turtles; 2. determinetheeffects ofclutch, feeding rate, raceway location and hydrological effects on growth rate; and 3. developregression statistics describing carapace and plastron length/width relationships. Materials artd Methods Morphometryandgrowth measun ments were taken on 100 head started Kemp's ridleys from August 30, 1984, 1 to46 days after hatching! through May 13, 1985, 97 to302 days old and 8 daysprior to release inthe Gulf of Mexico!.Measurements weretaken biweekly during this period to produce19data sets. Four straight line measurements carapace length, carapace width, plastron length and plastron width werethe basis for all inorphometryand growth characterizations. Measurement protocol followed that outlined in Baconetal. 983! whereby:carapace length

' TexasA&M Universityat Galveston FACILITY A FACILITY B

RACEWAY 2 RACEWAY 3 RACEWAY 4

CLUTCH CLUTCH CLUTCH CLUTCH I 2 3 4 5 6 I 2 3 4 5 6 2 3 4 5 6 2 3 4 5 6 A AB A B AB B C 9 13 0 CF D E EF EF F G GH GH GH H I 8 10 11 J K L K KL K M N MN M MP N N 16 0P 17 Q QR Q R Q R R

FEEDING RATE % BODY WEIGHT!

FREQUENCY I /DAY I /DAY 2/DAY I/DAY Figurei. Allocationofntne clutches ofKernp's ridtey sea turtle hatchings ofthe 1984 year-class across four racer/Jays andthree feeding regimes. FacilityA andeach contained three unique clutches with 10turtles apiece. The fourth raceway, 14, was located in rearingFacility B andcontained only oneclutch, 17, whose constitutents 0 turtles!were partitioned between this racewayand raceway 4. Survivalamong experimental turtles during thc studywas 96 percent. Experimentalturtles were fed a dry,high-protein pelleted diet. Thc ration manufacturedby CentralSoya and Subsidiaries!used for sixprevious year-classes 978-1983! was initially fedthe experimental turtles bu t didnot float asrequired for surfacefeeding hatchlings. This ration was subsequently doubled in orderto obtainenough floating pellet»for adequatefeeding through November 10. Thereafter, a floating,modified trout chowmanufactured by Purinawas fed to turtles.The quantity of pelletcdfood given to eachraceway/clutch combination each day was based ona percentageof the monthly geometric mean weight per turtle in a clutch.Three feeding regimes were provided turtlesthrough February 1985. These regimes included: 1! raceway2 turtles 3 percentofbody weight fed oncedaily; 2! raceway3 and14 turtles 5 percentof bodyweight fed once daily; 3! raceway4 turtles 5 percentof bodyweight fedtwice daily Figure1!. All turtleswere fed a uniformrate 2 percentof bodyweight! twice daily after February. Seawaterin racewayshousing experimental turtles was pumpedfrom the Culf of Mexicoand storedtemporarily in reservoirsheated only during the winter. Racewaywater temperaturevaried as did air temperatureof Facilities A and B, which were ventilatedby exhaustfans during summerand heatedby forced-airheaters in winter. Water temperatureand salinity were monitored daily in each raceway with a mercury laboratory thermometerand refractometer,respectively. Seawater in racewayswas replacedcompletely three times a week, and raceways scrubbed once weekly to remove unconsumed food, algae and excrement, Morphometryand growth wereassessed across four experimentalvariables which included: 1! clutch 9 experimentalclutches; 2! feedingrate 3 feedingrates through February!;3! location 4 raceways in Facility A and 1 in Facility B!; and 4! hydrologicalcomponents water temperatureand salinity, Carapaceand plastronlength-width measurements were logarithmically transformed and subjectedto analysisof varianceto detectdifferences in morphometryand growth acrossclutch and feedingregime. A Bartlett's test of homogcncityof varianceswas conducted to testequality of samplevariances Sokal and Rohlf, 1981!. Regression analysiswas used to determinelength-width relationships for carapaceand plastron measurements across clutch, Functionallength-width relationshipsdescribed by regressionequations were comparedby testsfor homogeneity of slopes Sokal and Rohlf, 1981!.

Results Racewaywater temperatures exhibited a seasonalpattern Figure 2!. Average monthly levels peaked between 26' and27'C in August,declined to fluctuatinglows near21'and 22'C during Novemberthrough January, and gradually

221 29

28

27

I- 26 K CL25 24

I- 23 Z 22

20

AUG SEP OCT NOV DEC JAN FEB MAR APR MAY Figure2.Mean monthly water temperature C!of racetoays housing Kemp's ridleys used in morphometricgrowth studies. 32

31

30

29

28

27 I- 26 ZCo 25 Z UJ24 23

22

21

20

AUG SEP OCT NOV DEC JAN FEB MAR APR MAY Figure3,Mean monthly salinity S~o! ofraceways housing Kemp's ridleys used inmorphometric growthstudies. returnedtonear 26'C in April and May. Temperatures acrossall raceways paralleled oneanother through December. Thereafter,temperatures inraceway 14 the only raceway inFacility B!typically differed from those ofraceways in Facility A by 1.0' C. Salinityexhibited a precipitous decline from mean values of30%a inAugust toa lowof22%c byNovember Figure 3!.Thereafter, salinities rosegradually tofluctuatebetween 26and 29%o. Aswith water temperature, salinity trends acrossraceways were nearly iden tie a through1 December, after which time those in raceway14were 0 5to 1.0%. lower thanthose of otherraceways. MEANCARAPACE LENGTH mm! MEANCARAPACE WIDTH mm! m a > co O M 4 ol a o o o o o o o o o o o o o

0 0 0 0

?C 0 0 CI m 0 m 0

m o mGl

Figure4. Monthlymean carapace length mm! values from all turtles Figure5. Monthlymean carapace width mm!values from all regardlessofclutch! in raceways2,3,4and14. Arrows denote thechange turtles regardless ofclutch! in raceways2,3,4and 14, See Figure fromsinking to floating chow and initiation of a uniformfeeding rate. 4 for meaningofarrows.

160 180

160 160

140 140

120 120

100 100

80 80

80 60

40 40 v z UJ20 20

IU 0 OD. AUGSEP OCT NOV DEC JAN FEB MARAPR MAY AILIGSEP OCT NOVDEC JAN FEB MARAPR MAY 160 160 K 140 140 Z IU120 120

100 100

80 80

60 60

40 40

20 20

0 0 AUGSEP OCT 14OVDEC JAN FEB MARAPR MAY AUGSEP OCTNOV DEC JAN FEBMAR APR MAY Figure6. Monthlymean carapace length mm!values for individualclutches in race a ys2, 3, 4 and14. See Figure 4 for meaningof arrows. RACEWAY 2 RACEWAY 4 28

24 12 20 10 10 16

12 Q2 4 Z 2 m m z 0 O UJ SEPOCT NOV DEC JAN FESMAR APR MAY SEPOCT NOV DEC JAN FES NIARAPR MAY Z UJ CJ IE 0 37 Z Z RACEWAY 3 RACEWAY 14 m 28 IO 14 14 24 C U 12 Cl 10 10 ib 8 16

12 8 12

2 4 2

0 0 0 SEPOCT NOV DEC JAN FEB MAR APR MAY SEPOCT NOV DEC JAN FEB MAR APR MAV Figure7.Comparison ofmonthly groToth increments incarapacelength andmean feeding ratefor turtles int'ndividual clutches acrossraceroays 2,3,4 and14. See Figure 4 for meaningof arrotos.

Morphometry and Growth Carapacelength and width growth patterns were similar Figures 4 and 5!. Growth incarapace length was most rapidand relatively comparable among raceways/feeding regimeduring August through October Figure 4!.This growthslowed a ppreciably through December asraceway water tempera ture and salinity Figures 2 and 3!exhibited strongfluctuations andthe pelletcd diet was changed from a sinkingtofloating chow. This period also marked the beginningofa 6-month spanin which carapace lengthfor turtles given the highest feeding rate raceway4 -5percent bodyweight fed twice daily! diverged more and more from those fed the lowest rate raceway 2 3 percentbody weightfed once daily!. Clutches inraceway 4 consistently produced significantly greater carapace length statistics I'<0.05!during every measurement setin thisperiod Figure 6!, Despitedifferences ingrowth rate, carapace length statistics among all raceways exhibited noticeable increases fromFebruary through May Figures 4 and 6!. This upswing ingrowth coincided with sizeable increases inwater temperature Figure 2!and the initiation ofa uniformfeeding rate .2 percentbody weight fed twice daily! for all turtlesinlate February. Monthly growth rate increases weremost consistent among clutches inraceway 2,which priorto March were fed at the lowest feeding regime Figures 4 and 6!, Raceways initially fed higher feeding rates exhibiteda modestdecline in growthrate after April Figure4!. FebruarytoMay carapace length measurements revealeda pattern Figures 4 and 6!of increasing convergence in growthofturtles inraceways and 14! fed the intermediate feeding rate percentbody weight once daily!. This patternofconverging growth rate differed sharply from that exhibited bythese raceways during October through January Figure 4!. Overallmean increase incarapace length during the study ranged from 8.2 mm per month inraceway 2 to10.0 mm permonth inraceways 3 and 14to a highof12.0 rnm per month inraceway 4.Turtles fed the highest rate exhibited a 428 percent greater increase incarapace length when compared tothose fed 3 percentoncea day.Growth increment forcarapace length fluctuated across raceways, feeding regimes andmo~ths Figure 7!.All raceways exhibited a generalpattern ofdeclining monthly growth increments fromSeptember through December, a noticeable growth spurtin January followed bya similarslowing inFebruary andupswings through April. Raceway 2 turtles were the onlyclutches not to exhibitsizeable declines ingrowth increments during May. Comparisonofmonthly carapace length increments withmonthly feeding totals mean number ofgrams fed to eachturtle per day! revealed intriguing patterns Figure 7!,Monthly length increments acrossall raceways during rapacelength CL!and width CW!across raceways and feeding regimes.

tial CL Final CL CL Gain Initial CW Final CW CW Gain mrn! mm! '/! min! mm! '7

58.5 13'3.0 127,4 52 9 I 21.0 128.7 58.9 150,8 156.0 52.6 136.7 159.9 60.8 164,3 I 70.2 54.3 150,2 I 76.6 14 60.7 153.0 152.1 54.5 138.8 I 54.7 ugust to Late February 2 58.5 100.6 72.0 52.9 91,8 73.5 3 58.9 119,1 102.2 52.6 105.5 100,6 4 60.8 130.4 114,5 54.3 116,9 115.3 14 60.7 123,2 103.0 54.5 109.9 101.7 ate February to Mid-May 100.6 133.0 32.2 91,8 121.0 31.8 3 119.1 150.8 26.6 I 05.5 136.7 29.6 4 130.4 164.3 26.0 116.9 150.2 28.5 14 123.2 153,0 4. 2 109.9 138.5 26.0

Septemberthrough November declined despite increasing volumes of foodfed turtles. This three-month decline in growthincrement also coincided with sharptcmpcrature and salinity reductions Figures 2 and3!. After November, fluctuation»in monthlygrowth increments generally followed those for feedinglevels Figure7!. Thistrend was particularlytrue for the latter part of the study February-April!when temperatureand salinity levelsrose significantlyand dict volumeincreased dramatically, Likewise, declines in theamount of foodfcd raceways3, 4 and I4 duringMay produced noticeable reductions in growthincrement among constitutent clutches. Growthin carapacelength across raceways/feeding regimes exhibited different trends across time and change in feedingrate TableI!. Overthe entire study period, raceway 4 clutchesexhibited an average gain in carapacelength of 170.2percent as compared to 127.4,156,0 and 152.1 percent, respectively, for raceways2, 3 and14. Partitioning carapacelength data into two groupsbased on the late August to lateFebruary period of differentfeeding regimes andthe late February to mid Mayperiod of uniformfeeding rate yields different growth patterns. Carapace length growthpatterns during the August to Februaryfeeding experiment parallel those outlined above for the entire study, However,these growth trendswere reversed during late Februarytn mid May when all clutcheswere fed the same rate.2 percentof body weight fed twice daily!. Average percent gain in carapacelength over this 74-day period was significantlygreatest P<0.05! in raceway2 clutches2.2 percent!and lowest in raceway14 clutches 4.2 percent!, Raceways3 6.6 percent!and 4 6 percent!also yielded significantly smaller gains when compared to that of raceway2. These trends coincided with divergingpa'tterns in theamount of foodfed to raceway2 clutchesand those in the other three raceways Figure 7!. Monthlygrowth patterns and feedingrate relationships recorded for carapacewidth Figures5 and8, TableI! generallyparalled those exhibited by carapacelength Figures4 and 6, TableI!, Meancarapace width was significantlygreater P<005!for raceway4 turtlesthan that ofother raceways.Within racewaycomparisons detected no statisticaldifference in carapacewidth amongconstitutcnt clutches, except in raceway3 whereclutch 2 turtles exhibited significantly P.05! slower growth Figure 8!. Regressionof carapacelength on carapace width forall clutch/racewaycombinations produced fitted lines whose slopeswerc nearly identical Figure 9!. Highcoefficients of determination r' typically> 0.99!indicated exccllcnt fits of regression lines. Plastronlength and width statisticsdisplayed the sametrends evident for carapacemeasuremcnts Figures 10-13, Table 2!. Thesetrends indicated that plastron growth was largestamong raceway4 clutches,intcrmediatc and convergingto similarityamong raceway 3 and14 clutches, and significantlysmallest among raceway 2 cohorts. Within racewaycomparisons of plastronlength and width measurementsacross clutches detected no divergence from thosepresented for carapacestatistics Figures6, 8, 12and 13!. Regressionanalyses for plastronlength and width yieldedstrong correlations bctwecn the two parametersand fitted lines whoseslopes were very similar acrossall clutches/feedingregimes Figure 14!.

225 160 160

140 140

120 120

100 100

eo 80 60 E 80 E 40 Z 2O 20 ZUJ o Z 0 0 AUGSEP OCT NOVDEC JAN FEBMAR APR MAY AUGSEP OCT NOVOEC JAN FEB MARAPR MAY IZ I- 160140 160 140

Z 120 120 UJ oo 100

80 80

60 60

40 40

20 20

AUGSEP OCT NOVDEC JAN FEBMAR APR MAY DEC JAN FEB MAR APRMAY Figure12,Monthly meanplast ron length

140 140

120 120

100 100

80 80

eo eo

E 40 40 Z 20 0 2D

Z0 0 ALIGSEP OCT NOV DEC JAN FEB MAR APR MAY AUGSEP OCT NOVDEC JAN FEBMAR APRMAY ID 140 14D

120 120

ZUJ 100 100

80 80

60 60

40 40

20 20 0 0 AUGSEP OCTNOV DEC JAN FEB MARAPR MAY AUGSEP OCT NOVCEC JAN FEBMAR APRMAY Figure13,Monthly meanplastron width mm! values forindividual clutchesin raceways 2,3, 4 and74. See Figure 4 formeaning ofarrows. EWAY 4 160 RACEWAY 2 160

120 120

80 go

E E 40 40

U 0 40 80 120 160 200 0 40 80 120 160 200 UJ

WAY 14 0 160 EWAY 3 160 Z 00 120 120

80 80

40 40

0 40 80 120 180 200 0 40 80 120 180 200 PLASTRON WIDTH mm! Fi~re24. Rectilinear regressions ofplastron length Y! on plastron width X! for individual clutches across raceways 2,3, 4 and 14.

Discussion Considerablesimilarity existed between the growth patterns and growth rate curves recorded for all carapaceand plastronparameters, This consistency in growth,regardless of morphologicalparameter, justifies the selectionof onlyone meristic character upon which to discussresults of thisstudy. Carapace length will bethe primary meristic discussedherein since turtle researcherstend to use this variable more frequently than other growth variables involving length and/or width. Feedingregime, clutch and microenvironmental effects all appearedto influencegrowth of thehead started, 1984 year classof Kemp'sr idleys.Greatest impact on growth resulted from feeding regime. Feeding experiments during Augustthrough late February clearly proved that growthrate was dependent upon feedingrate and frequency, Turtlesfed 5 percentof bodyweight twice daily experienced a 30.2 percent greater carapace length than cohorts fed 3 percentonce daily during the six-month experiment. Similarly, turtles fed at thesame rate percent!but at different frequenciesexhibited average carapace lengths that differed by 12.3percent. Comparisonof growthpatterns for turtlesfed the same feeding regime but housedin differentlocations appeared to indicatefeeding regime exerted greater influence on growth than did location. Overall percentgain in carapace length for raceway3 and 14 turtles housedin FacilitiesA and B, respectively!was 156.0and 152.1percent, respectively.More importantly, carapace length gain for thesetwo racewaysduring the feedingexperiments was almostidentical 02.2 percentraceway 3; 103.0percent raceway 14!. Theeffect feeding regime had on growth also became more apparent during the 74 days that all racewayswere fed thesame feeding regime .2 percentof bodyweight twice daily!. Average monthly increment in carapacelength growthwas 10.8, 10.6, 11,3 and 9.9 mm for raceways2, 3, 4 and14, respectively. The similarity among these values appearsindicative of fairly uniform growth acrossturtles fed the samerate. Growth curvesrevealed a patternof relativelygradual increase in carapacelength acrossmonths forall raceways. Thesesmooth growth curves tended to maskthe sharpfluctuations in monthlygrowth rate within respective raceway/clutchcombinations, Monthly growth rate dipped noticeablyacross all racewaysduring September throughDecember. These declines in growthincrement occurred despite increased feeding rates through November. I tis unlikely tha t mandatedfeeding rates during this period of rapidgrowth among rela tively small-sized tur ties may haveresulted in foodnot being eaten, Instead, growth rate reductions were probably the result of lower,fluctuating watertemperatures through November. This trend in slower growth among captive-reared Kemp's ridleys also has beenreported by Caillouet and Koi 985!. These workers found turtles in yearclasses 1978-1983 exhibited slowed growthnear the middle ofthe head siarflng period and attributed this trend to temperature reductions during winter. However,monthly fluctuations in raceway temperature andgrowth rates during this four-month period were generallyopposite one another and, as such, cast doubt on temperature acting alone to producereduced growth incremen ts. Sizeableincreases in monthly growth increment across raceways after January were due to a combinationof increasedwater temperatures andamount offood fed turtles. This was particularly true through Aprik Thereafter, raceway temperature dipped slightly, amount of foodfed dropped drasflcally in raceways 3,4 and14, and growth incrementsamong constitutent turtles plummeted. Raceway 2 turtles failed to exhibit the same sizeable reductions ingrowth rate observed across the other raceways. These turtles continued toreceive anincreased feeding rate which bufferedtheimpact ofreduced water temperatures andkept growth increments nearthose recorded inApril. Littledifference could be detected ingrowth among clutches. Detection ofclutch variation was, in part, masked byan experimental design incorporating different feeding rates. Other workers Caillouet and Koi, 1985; Caillouet etal., 1986! have reported considerable variation ingrowth as weight! among individual turtles, clutches andyear classes.However, their studies were based on uniform feeding rates whereby growth responses ofdifferent clutches tothe same treatmentcould becharacterized. Onlyoneclutch in raceway 3!exhibited signiflcantly different growth ratesduring the present study's within raceway/feeding regime comparisons. Thisclutch was among the oldest groupofhatchlings andone of the first to be hatched. Nevertheless, thisclutch exhibited thesmallest iniflal carapace lengthmeasurements recorded among the nine clutches, NMFS' head start personnel have found that clutches hatchingearliest frequently exhibit poorest survival and growth Clark T. Fontaine, NMFS, personal comrnunica- Cion!. Similarityofgrowth among clutches also was evident from regression analysis. Within raceway comparisons of regressionformulas forcarapace length and width yielded nearly identical slopes across clutches. Slope values for clutchl 7 turtles exposed todifferent feeding regimes also were nearly identical. Regression analysis also confirmed Carr's967! contention thatthe carapace ofKemp's ridleys isalmost aswide as it islong, even atan early age. In summary,growth of captive-rearedridleys can be manipulatedby feedingregime and environmental conditions.Higher feeding rate and feeding frequencies produced faster growth over the six-month feeding experiment.Rate of growthduring these months was influenced byfluctuations in environmental parameters, particularlywater temperature and,secondarily, salinity. Similar growth patterns exist for other captive-reared turtles.Growth rates ofcaptive loggerheads Carettacaretta! aredependent uponthe amount andquality percent protein!offood provided Stickney etal., 1973; Nuitja and Uchida, 1982!. Furthermore, ithas been found that captive greenturtles Chelonia rrrydas! achieve optimal growth ator above 25'C Stickney, 1979!, Findings ofthese latter studies alsowere pertinent toraceway 2 turtles inthe present study. These turtles exhibited thehighest percent gain in carapacelength among cohorts inother raceways whenthey were changed tothe higher feeding regime .2percent twicedaily as opposed to3 percentonce daily! in late February. This higher feeding rate and the increased water temperatureduring March toMay enabled raceway 2 turtles toexhibit the second highest overall gain in carapace length2.4 mm as opposed to33.9 mm in raceway 4!recorded among raceways. Literature Cited Bacon,P.,F. Berry, K.Bjornd al,H. H irth,L.Ogren and M. Weber Editors!, 1983. Symposium onsea turfle research ofthe western Atlantic populations andsocioeconomics!. Proceedingsof the Western Atlantic Turtle Symposium, 17-22July 1983, San Jose, CostaRica, Center for Environmental Education, Washington, D.C, In three volumes. Caillouet,C.W.,Jr. and D.B. Koi. 1985. Patterns andvariability infirst-year growth inweight ofcaptive-reared Kernp'sridley sea turtle:a graphicalanalysis. NOAA Technical Memorandum NMFS-SEFC-164, i plus4 p.and 52 figures. Caillouet,C.W.,Jr., D,B. Koi, C.T. Fontaine, T.D.Williams, W.J.Browning andR.M. Harris. 1986. Growth andsurvival ofKemp's ridieysea turtle, Ieputochelys kernpi,in captivity. NOAA Technical Memorandum NMFS-SEFC-186, iii plus34 p., 12 Tables and 7 Figures. Caldwell,D.K.]962. Growth measurements ofyoung captive Atlantic seaturtles intemperate waters. LosAngeles County Museum,Contributions in Science No. 50, 8 p. Fontaine,C.T., K.T. Marvin, T.D. Williams, W.J. Browning, R.M. Harris, K.L.W. Indelicato, G.A. Shattuck and R.A. Sadler. 1985. Thehusband ryofhatchling toyearling Kemp's ridley sea turtles, Lepidochelys kempi.NOAA Technical Memorandum NMFS- SEFC-158,iii plus 34 p., '10 Tables, 22 Figures and 2 Appendices. Frazer,N.B.and F Schwartz.J. 1984.Growth curves forcaptive loggerhead turtles,Caretta curetta, inNorth Carolina, Bulletin of Marine Science 34!:485489, Hirth,H.F. 1982. Weight and length relationships ofsome adult marine turtles. Bulletin ofMarine Science 32!:336-341. Klima,E F.and J.P. McVey. 1982, Headstarting the Kemp's rid!ey turtle, Lepidochelys kempi, p. 481-487. In: Bjorndal, K. A. Editor!, Biologyand Conservation of SeaTurtles, Smithsonian Institution Press,Washington, D.C,, 583 p. LeBrun,G. 1975. Elevage a lareunion de juvenilesde la tortueverte Chelonia rnydas Li nnaeus!1758. Science etPeche, Bulleti n Institute Pechesmaritime. No. 248, 25 p, Marquez,R. 1972.Resultados preliminares sobre edad y crecimientode la tortugalora, Lepidochelys kempi Garman!. Mem. IV Congreso Nacional Oceangrafico Mexico!, p. 419-427, McVey,J.P. and T, Wibbels.1984. The growth and tnovements of captive-rearedKemp's ridley seaturtles, Lepidochelys kempi, followingtheir release in theGulf of Mexico.NOAA Technical Memorandum NMFS-SEFC-145, 25p. plus3 Figuresand 3 tables. Nuitja,I.N.S. and I. Uchida.1982. Preliminary studies on thegrowth and food consumption of the juvenileloggerhead turtle Carettacaretta L.! in captivity. Aquaculture27;157-160. Pritchard,P.C.H. and R.Marquez M. 1973.Kemp's ridley turtle or Atlantic rid ley. InternationalUnion for Conservationof Nature and Natural Resources,Monograph No. 2, 30 p. Rajagopalan,M. 1984.Studies on the growth of olive ridley, Lepidochelysohoacea, in captivity. CentralMarine Fisheries Research Institute Bulletin 35:49-54.Indian Council of Agricultural Research,Cochin, India. Sokal,R.R. and F.J.Rohlf. 1981.Siometry. Second Edition, W.H. Freemanand Company,New York, 859p. Stickney,R.R., D.B. White and D. Perlmutter,1973. Growth of greenand loggerheadsea turtles in Georgiaon natural and artificial diets.Bulletin of theGeorgia Academy of Spence31:37-44. Stickney,R.R. 1979. Principles of warmwater aquaculture. John Wiley and Sons,New York. 375p, Uchida,I. 1967.On the growth of the loggerheadturtle, Carettacaret ta, under rearingconditions, Bulletin of theJapanese Society of Scientificfisheries 33!:497-507. Witham,R. and C.R.Futch. 1977. Early growth and oceanicsurvival of pen-rearedsea turtles. Herpetologica 33!:404-409. Wood,J.R. and F.E.Wood. 1981. Growth and digestibility for the greenturtle Cheloniamydas! fed diets containing varying protein levels. Aquaculture25:269-274.

231 A Reporten Attemptsto BreedKemp's Ridley Sea Turtle, Lepidochelyskepi, in Captivity Steven R. Rabalais, David W. Owens and Peter Thomas '

Four two male and two female! adult Kemp's ridley sea turtles Lepidochelys kemp0 were kept under controlled laboratory conditionsforone year at The University ofTexas Marine Science Institute. During this time, the animal serum androgen testosterone!andestrogen estradioO levels and breeding behavior zoere monitored. Asz'gnificant increase inserum estradiol zoas notedin one female, simultaneous tothe males'at tempts atbreeding zoith this i ndividual. An increasein serum testosterone levels ofthe males prior to the increase in female serum estradiol levels was noted. The significance ofthese changes relative fothe breedingbiology ofthis species wasdiscussed. Thebreeding biology ofthis species coincides withpublished accounts ofturtles breedingin the wild. A videorecord of the apparent unsuccessful breeding activity zoas presented.

' LouisianaUniversities Marine Consortium; Texas A&M University; TheUniversity ofTexas atAustin Marine Science lnstitute

232

andha wksbills Eretmochelys imbricata!, After this initial period, two pens were constructed inthe channel byrunning wiremesh fences across the channel. The downstream pen holding the male was about 7 x7 meters,while the upstreampen holding the three females was about 11 x 9meters and had an open side leading to the nestingbeach, A gatethat could be opened from thewalkway connected the two pens. Theturtles were under observa tion during the day several times a weekfor periods of one to three hours a at time, Specificbehavior patterns for each turtle were recorded and additional written notes were made. Still photographs weretaken using a 35mm camera, and a videotapewasmade of some of the behaviors. The timing of the observation periodsduring theday wasvaried to allow a morecomplete descripflon of theanimals' behaviors. Themale was introduced into the pen containing females atvarious times from March through May 1984 by openmgthe gatebet ween the pens. The behavior ofthe animals was followed closely toobserve any breeding activity. If eitherthe male or all of thefemales were disinterested in breeding, the male was returned to hispen and the experimentrepeated ata laterda te. When no matings occurred among these four turtles, two additional male Kemp's ridleysfrom a stockof headstarted individuals of the1978 year-class were introduced to thethree adult females in hopesthat a largernumber of individualsmight stimulate reproduction. Initially, these two additional males were keptin thepen with the older male and introduced oneat a timeto the females. Later, all three males were placed with all threefemales. Early in 1985,female SB died, The remaining two adultfemales LF and P! and male S aswell asseveral head started males were segregated and then introduced during 1985 as in 1984.

Results Thebehaviors shown by theKemp's ridleys while they were together with other species in thechannel involved verylittle intraspecificor interspecificinteraction, On twooccasions in early March 1984, male S wasobserved followingeither female SB or LF and placing a flipperover her back or driving her before him, Both times the female swamoff, and the interactions were over in several minutes, Interspecific interaction was also limited, primarily consistingof femaleLF biting the gree~ and hawksbill turtles and subsequently being bitten herself, She was bitten bya largegreen turtle and sustained aninjury to her head and jaw that required her isolation until the middle of June. Theother females P and SB! that had been residents in the channel for years appeared to have established territoriesor selectedfavored habitats, SB was consistently found resting on thebottom of thenorth side of the channel,andP onthe sou th side. Male S appearedtobe more active and less attached toone site since heconstantly travelledaround the entire tank. Sedentary habits ofthe females ascompared tothe increased movement shown by themale may indicate a sex-specificbehavior, but the small number of animalsinvolved prevent firm conclusions being drawn. Theturtles werc segregated bysex and held in the smaller pens from early March 1984 on, during which time they wereobserved forany behavior indicating readiness formating. None ofthe turtles showed anybehavior indicating a desiretobe together, and in fact they stayed atopposite ends of their pens possibly because they were attracted to thehigh activity level of thegreen turtles outside the pens. Whenthe male did show Interest inthe females byswimming along the edge of their pen, he was transferred to theirpen where he would approach one of the females and swim behind her. He would drive her before him and attempttomount her. After trying unsuccessfully tomount her several times and to grip her with his foreflippers, hewn»Id turn his attention toward the other female where upon the second female would begin to avoid him. Occasionallyhewas successful inmounting a female and hooking his front claws over her carapace, butthis lasted nomore than one to two minutes before she would dive or turn away, The male often bit the hind flippers of the females,attimes becoming veryaggressive, andthis occasionally induced thefemales tobite each other. Generally, after10 to 30 minutes themale apparently lostinterest, orthe females became more determined intheir resistance, eitherby swimming away rapidly or by biting at him. If hisbreeding behavior oid not resume, the male was removed fromthe female pen and returned tohis own. This pattern ofbehavior continued during each introduction ofthe male into thefemale pen during April andMay. Bythe end of May 1984, theinterest male S showedinthe females haddeclined somuch that he no longer actively pursuedthem, although occasionally heplaced his head under the rear margin oftheir carapaces. Thetwo young malesfrom the head started 1978 year-class wereplaced inthe male pen with S inthe hopes that their presence would inducea competitivedrivein S orthat thev might mate with the females themselves. Thega te between themale and femalepens was left open, but none of the males showed any interest in thefemales. Headstarted males were removed near the end of June 1984, and female LFwas reintroduced intothe male pen. Thethree females andmale S wereheldtogether,but theyremained verypassive, resting onthebottom andshowing nointeraction, OnJune 28, tracks on the beach indicated that one turtle had crawled across the nesting beach and back tothe water. One of two adult crocodiles Crocodylas acutus! kept in the channel also had crawled onto the beach and followedthesame track as the turtle part of the way. The following morning only female LFwas present inthe pen, so apparentlyshe had beenthe turtle making the crawl whetheras a falsenesting crawl, an exploratorycrawl, or to escapethe crocodile that had not previously been seen in the pen area was not determined. Three females, male S and the two head started males were held together in the rebuilt pen in July 1984.All the turtles remained passive and did not interact during the two weeks they were together, FemalesLF and P werc held togetherin the femalepen and on July 30a turtlecrawled from this pen onto the nestingbeach. No nestwas dug, so the crawl apparently was exploratory. No further significant behavior was observed so observations were concluded in early August, and the pens were removed from the channel. Seawatertemperature in the channelwas monitored throughout the period of observations,and despite the relatively large volume of seawaterand a significantturnover from the bay, the temperaturecould changeseveral degreesin only oneor two days.The temperature was as low as 18' C in Februaryand roseduring March and April to 23-24'C,During May,June and July the temperaturevaried between 27'C and30 'C.Most intraspecificinteraction and the breeding behavior shown by male S occurredwhile the water temperature was 20-24'C, although it continued even after the water had warmed to 29' C. During 1985,when maleS washeld with the females,no significant baling behavior occurred,so the two 7-year- old head started males were introduced into the pen with the older turtles, again in hopes of stimulating reproduction. No mating was observed, but on June6 acrawl track was observed on the nesting beach,and several eggswere found in the water of the female pen. Seawater temperature was 30 C at the time. BetweenJune 6 and 13, 1985,approximately 35 eggs were found in the water, and although most of theseeggs were found in the female pen, some of them were in the channel outside the wall of the pen. Approximately 15 clean, intact eggs were buried in two spots on the nesting beach, and those eggs that were tom or appeared to have been in the water for an extended period were frozen for later identification. Only the two ridley females LF and P were in the pen, but numerous adult female green and hawksbill turtles were present outside the pens, A sample of 22 eggs averaged 40.1mm in diameter range 36-44mm!, a sizerange appropriate for Kemp's ridley or hawksbill, though too small for green turtle Groombridge, 1982!.The crawl made by a ridley and the presenceof the majority of the eggs in the Kemp's ridley pen indicated that eggs probably were laid by a ridley, but it remains possible, due to presence of eggs and female green and hawksbill turtles outside the ridley pen, that the eggs were laid either by green oI hawksbill turtles and were carried into the ridley pen by the water. None of the eggs buried on the nesting beach had hatched as of August 16, 1985,suggesting that they had been in the water too long before burial or that they were infertile. The two females LF and P were X-rayed on August 9 in an attempt to determinetheir reproductivestatus. A malewas also X-rayed for comparison.No developedeggs were visible in either female although LF contained what appeared to be follicles 4 to 10 mm in diameter.

Discussion Although no nestingand probablyno successfulbreeding occurred during theperiod of observation,progress was made in the attempt to establish a captive breeding project for captive ridley turtles at Miami Scaquarium. Much informationon the turtles' behaviorwas gathered, initial ideason induction of breedingwere tried and refined,and the habitat requirements were investigated. TheScaquarium's channel habitat appeared to be a verysuitable location for this projectto be continued.Seawater quality, salinity, temperature,shade level andbottom topographyall seemedappropriate for maintainingsea turtles in goodhealth. Since little is known of theenvironmental requirements for breedingin ridleys,it canonly beassumed that if the animalsappear healthy and well adaptedto their confines,the habitatis fulfilling their needs.The nesting beachhas been used by numerousloggerhead and green turtles for nestingin pastyears, and it seemslikely that this beachcould fill the requirementsfor theridleys too,since it hasbeen reported that thereis little differencein themajor aspectsof nestingamong the various seaturtle species Hendrickson, 1982!, The beachused by Kemp'sridleys in Tamaulipas,Mexico is a broad,white sandbeach with no particulardistinguishing features Pritchardand Mkrquez, 1973!,and therefore is similar to the nesting beach at the Seaquarium. One characteristicthat may be of someconcern regarding nesting behavior of Kemp's ridleys in captivity as contrastedwith that shown in Mexicois the distancetravelled from the seato the nestingsite. In one study, most turtles crawled more than 40 metersand up to 60 metersbefore nesting Pritchard and M5rquez,1973!. Since the Seaquarium beach was only about 15meters long, it is possible that its limited area could have deterred the turtles fromnesting, However, the wild turtlesare probably looking for sufficientelevation above the water and a clearspot freeof rocksor vegetation,and since all of theSeaquarium beach fills theserequirements, it seems unlikely tha t the small sizeof the beachwould inhibit nesting. A secondcharacteristic of the Seaquariumhabitat that may havedeterred nesting was the passageof peopleon the boardwalkduring the daytimewhen the Kemp'sridley turtleswould be expectedto nest.The turtIes generally appearedundisturbed by people,but thepresence of peoplemay havebeen sufficient to keepthe turtlesfrom nesting, Unfortunately,the commercialnature of the Seaquariumprecluded closing this area to visitors,

235 Adcnowledgements I thankthe Miami Seaquarium andits conscientious employees who assisted inthe project, Mrs. Ila Loetscher and othersofSea Turtle, Inc. were most supportive ofthe project, providing both a Kemp'sridley turtle and funding for theproject, I appreciate the invaluable efforts of Mr,Suresh Ramanathan and Mr. Pedro Gonzalez who assisted in theproject in 1984and 1985, respechvely. Direction and advice were provided by Mr. Fred Berry of theNMFS SoutheastFisheries Center, Miami, Fla. and Dr, David Owens of TexasA8rM University, College Station, Tex. Fundingfor the1984 portion of theproject came from theNMFS under Contract No. NA-84-CEAOOI48.

Literature Cited Ernst,C.H. and R. W. Barbour. 1972. Turtles ofthe United States. University ofKentucky Press, Lexington, Kentucky, 347p. Ehrhart,L.M, 1982. A reviewofsea turtle reproduction, p.29 38, ln: Bjrondal, K.A, Editor!,Biology andConservation ofSea Turtles, SmithsonianInstitution Press, Washington, D.C., 583 p. Croombridge,B. Compiler!. 1982.The IUCN Amph ibia-Reptilia reddata book, Part I,Testudines, Crocodylia, Rhynchocepha! a. InternationalUnion for the Conservation ofNature and Natural Resources, Gland, Switzerland, 426p. Hendrickson,J.R. 1982. Nesting behavior ofsea turtles with emphasis onphysics! and behavioral determinants ofnesting success orfailure, p.53-57. In: Bjorndal,K.A. Editor!,Biology and Conservation ofSea Turtles, Smithsonian Institution Press, Washington,D.C., 583 p, Kiirna,E.F and J,P McVey. 1982. Headstarting theKernp's ridleyturtle, Lepidochelyskernpi, p 481-487.In: Bjorndal, K.A. Editor!, Biologyand Conservation ofSea Turtles, Smithsonian Institution Press, Washington, D.C., 583 p. Pritchard,P.C. H. and R. Marciuez lvl.1973. Kernp's ridley turtle or Atlantic ridley Lepidochleys kernpi.International Union for the ConservationofNature and Natural Resources Monograph No.2 MarineTurtle Series!, 30p. Captive Rearing and Breeding Kemp's Ridley SeaTurtle at Cayman Turtle Farm 983! Ltd. James R. Wood and Fern E. Wood'

CaymanTurtle Farm 983! Ltd.received 96yearling Kemp's ridley sea turtles Lepidochelys kempi! of the 1979 year-class in July1980 from Mexico. Fed on rations ranging from 35 percent to 45 percent protein, the turtlesincreased in weight from 098 kilogramsto 23,6kilograms in 62months. In May 1984at leastthree of the5-year-old females became sexually mature, and two of themnested. Three hatchlings were produced, but survived only threeto four days.This was the earliest documented age of nestingin Kemp'sridleys.

In July 1980,Cayman Turtle Farm 983! Ltd. received96 yearling and 67 hatchling Kemp's ridley seaturtles Lepidochelyskempi! from Mexico Wood,1982>, The yearlings had beenraised by the U.S.National Marine Fisheries ServiceGalveston Laboratory at Galveston,Tex. These had beenreturned to the Mexicangovernment and thence, alongwith the67 hatchlings from RanchoNucvo, were transferred to GrandCayman, B.W I. All turtleswerc obtained througha cooperativeKemp's ridley programbetween the Cayman Islands and Mexicowith the long term objective of establishmentof a captivebreeding herd of Kemp'sridley. Observationsand data collectedduring the first five yearsof this projectare discussedbelow. Growth and Nutrition During the first three monthsfollowing receiptof the turtles, all were fed ad libidumwith Biodiet,a semi-moist ration producedby Bioproducts,Inc., Warrenton,Ore,, containing 38 percentprotein. Useof the semi-moistpellet reducesthe incidenceof gut compactionassociated with drier rations.For thenext threemonths, the turtlesreceived a hatchlingfood manufacturedby CentralSoya, Fort Wayne,Ind., containing45 percentprotein. When Central Soya ceasedmanufacturing this diet, the turtles were fed a 45percent protein hatchling ration prepared by Ralston Purina, St. Louis, Mo. The hatchlings remained on the Purina ration until September 1982 when they began receiving a modified Purina Trout Chow containing 42 percent protein. At 20 months of age February 1981! the older ridleys began receiving Purina turtle chow containing 35 percent protein. During the next six months, their growth rate declined. Therefore, beginning in August 1981, the larger turtles were split into two groups, one receiving the Purina Chow containing 35 percent protein and the other fed the Purina Chow containing 45 percent protein. Figure 1 il]u strates the results of this experiment. Rate of growth for the five-month period was three times greater in turtles receiving thc 45 percent protein ration than in those receiving the 35 percent protein ration. Following termination of this experiment, all the turtles were maintained on the Purina ration containing 45 percent protein until the modified Purina trout chow becameavailable in September 1982.Turtles remained on the modified trout chow until transferred in May 1984to the breeding pond. At this time, it became impractical to feed them modified trout chow, since thc ridlcys were mixed with green turtles Cheloniamydas! in the breeding pond. In July 1985the breeding pond was modified so that the ridleys could be confined in a separatearea, and they were again fcd trout chow. Growth of the ridleys received as yearlings is shown by Figure 2. Growth was probably adversely affected by the use of the Purina 45 percent protein ration. Green turtle hatchlings fed the Purina 45 percent protein ration certainly have a slower rate of gain and higher mortality than when fed the Purina Trout Chow. Growth was also adversely affected between May 1984 and July 1985 when the ridleys were fed the 35 percent protein ration but improved following feeding of the Purina Trout Chow. Mortality and Disease Of the 96 yearling s receivedin 1980,33 4 4 percent! remained alive after five years at Cayman Turtle Farm 983! Ltd. As shown by Figure 3, 54 percent of the mortality occurred in these turtles during their first year at the farm. In general, mortality appeared to result from secondary infections of lesions causedby biting. By the cnd of their second year at the farm, when thc turtles werc three years old, their aggressivebehavior had virtually disappeared, During

' Cayman Turtle Farm 983! 4td.

237 Ft'ps re 1. Ratev growthof Kemp's ridley sea turtles fnt rationcontaining 35 percent and 45 percent protein.

2 3 MONTHS A VERAGE WEIGHT KILOGRAMS!

0 UI 0

Pigure2. Growthof Kemp's ridley sea turtles reared at CaymanTurtle Farm 983! Ltd. m

Pl

238 90

80

I 70

cp 60

> r0

cr. 40 02030 2 34 5

YEARS Figure3. Survivabilityof Ketttp'sridley sea turtles received as yearlings, thenext three years, ten of theridleys died and one escaped. One mortality occurred following surgical removal of a largeabcess from the ventral surface of a frontflipper. Cause of deathwas not determinedfor othermortalities. Reproduction As previouslyreported Wood and Wood, 1984!,in May 1984,eggs were observedin one of the fiberglasstanks housinga portionof the Kempsridleys, A1143ridleys, consisting of 39five-year-old and four-year-oldanirr als,were transferredto thebreeding pond. During transfer,eggs were found in thecloaca of onefemale tag 1336!.This female weighed24.5 kg with a curvedcarapace length of 53.3cm. On 20May anotherfemale tag 1359!laid 62 eggsat 0010 hour. Thatfemale weighed 20.0 kg andhad a curvedcarapce length of 48.3cm. Female tag 1336! laid seveneggs while crawling on the artificial beachon June9 at2325 hours. Kemp'sridley matingbehavior had beenobserved in the fiberglasstanks during April 1984by someof the farm's staff.The seven eggs laid by thefirst female tag 1336!did not develop.However, several of the eggslaid by thesecond female tag 1359!began development, and after 62days three ha tchlings were produced.Of the 59 eggsthat did not hatch, 32 were undeveloped, 20 suffered early embryonic death and seven died during late embryonic stages.The threehatchlings produced had hard yolk which they wereunable to absorb,and they died after three to four days. Thefemale with tag 1359spent seven hours on thebeachtwo daysafter nestingand spentfour hourson thebeach during thenight of Junc19, but did not nestagain. The female with tag1336 crawled 12 more times in Juneafter having laid the seveneggs on June9, spendingapproximately 5.5 hours in total on the beach.Eight other femalesspent five hoursor moreon thebeach. One of these,tag 1321,was found deadon thebeach after spendingseveral hours during the day on the beach. Shehad dug a shallowbody pit but no egg chamber. Necropsyrevealed that her ovaries containeda numberof partially developedeggs, Another female with tag 1340crawled several times in July,digging at leastone egg chamber, but shedid not nest.This turtle had a verylarge abcesson her right front flipper. Shewas takenfrom the breedingpond on July 22,and the abcesswas removed by two University of Florida veterinarians, Although the turtle appearedin good health following surgery, she was found dead the following morning. Examination showed that one oviduct contained seven shelled eggs, and the ovaries contained a number of well- developedfollicles. During the1984season, five malesalso crawled on the beach.Three of thesemales dug nestsas if preparing to lay. Much of the time spenton the beachby both femalesand maleswas spentresting or sleeping,usually coveredat leastpartially with sand.This behavior is differentthan thatobserved for our captivecolony of greenturtles, For some of the females,the length of time spenton the beachmay havereflected some difficulty they had in laying their first

239 clutchof eggs. Another possibility isthat the animals were attempting torid themselvesof parasites. Earlier in the year,an overnight pump failure resulted in thedeath of all the fish in the breeding pond. Normally these fish clean theturtles of external parasites. During the 1984 season all theturtles in thebreeding pond had extensive infection ofthe fish leech. The greens would get in very shallow water and allow Ruddy Turnstones topick off the leeches, but thisbehavior was not observedfor theridleys. Ninecopulatory mountings lasting five minutes ormore were recorded among the Kemp's ridleys between April 19and June 15, 1985. The combined duration ofthese observed mountings was427 minutes, with the longest single mountinglasting 180 minutes. Beach crawls were limited tofour known females and one male, One female tag 1339! remainedmotionless onthe beach for260 minutes, butaII other crawls were less than 12 minutes, Nonest digging was observed in 1985, Elevenmating behavior interactions between ridleys and green sea turtles were observed. These consisted of four instancesinwhich aridley male attempted tomount a femalegreen, one attempted mounting ofa maleridley by a malegreen, one attempted mounting ofa femaleridley by a malegreen, two instances ofa female ridley chasing a femalegreen, one instance ofa femalegreen chasing a female ridley, a maleridley biting a hybridmale, and a green femalebiting a ridley. Sincethe initial nesting in1984 and the presentation ofthis paper inOctober 1985! the Kemp's ridleys atCayman TurtleFarm have successfully rcprod uced and produced viable hatchlings in1986, 1987 and 1988 Wood and Wood, 1988!.Some ofthe ha tchlings from the 1987 and 1988 year-classes weretransferred tothe Galveston Laboratory for headstarting as part of a Memorandumof Understanding between the Cayman Islands and the United States governmentssigned in early 1985. Other Behavioral Observations Kemp'sridley sea turtles have frequently beenobserved burying themselves inthe sandy bottom ofthe breeding pond.They placed thetwo front flippers together above thehead and pushed them back through thesand, With only a fewstrokes a turtle can be a 1most completely covered andalmost invisible. This could be a typeoffeeding beha vior withthe objective ofuncovering food items in the sand. However, theturtles did not appear tomove their heads or tolook for food. Since seawater surrounding theCayman Islands ismuch clearer than that found inthe normal ridIcy habitat,this burrowing behavior might be a meansofprotection from ultraviolet radiation. However, these animals werealso occasionally seenburied at night. This behavior perhaps ismore likely a methodof hiding in an environmentthat normally offers few naturalfeatures under which to hide. Literature Cited Wood,J.R. 1982. Captive rearing ofAtlantic ridleys atCayman Turtle Farm Ltd. Maririe Turtle Newsletter No.20, p. 7-9, Wood,J.R. and F.E. Wood. 1984. Captive breeding ofthe Kemp's ridley. Marine Turtle iVearslet terNo. 20, p. 12. Wood,J.R. and F.E, Wood. 1988. Captive reproduction ofKcmp's ridley Lepidochelys kepi.Herpetological Journal1:247-249.

240 Questions and Answers

Nat Frazer: Peter, in thegeneral subphylum of thevertebrates as a whole,what we see as we move from fish to amphibiansto reptilesto mammalsis a reductionin the numberof skull bones,Given that, could you not turn your proliferation of scutation patternaround and say that theproliferate scutation pattern is primitiveand that what we get as we move from Lepidochelys to Caretta to Eretmochelys to Chelonia is a reductionin the numberof scutes? Pritchard: I agree that there is that tendencythroughout vertebrateevolutio~ of elimination of bony elements, particularly in the skull, But, looking not at vertebrates as a whole, but turtles as a whole, it is true that some of the earliest turtles like Proganochelysfrom the had a few extra scutesand a few extra bones.But, by the time turtles really became established at all, certainly by the Cretaceous,the standard turtle patterns were incredibly fixed. The five vertebrals and four pairs of costals pattern is fixed acrossboth suborders and every family and with an amazing degreeof fixity for hundredsof millions of years.And only a few turtles today diverge from that. And they arenot ones that seem to be primitive in other respects, ReneMarquez Millan. Do you suspectwhy the ridley is nearlyabsent in the CaribbeanSea? Pritchard: The ridley is indeed nearly absentin the Caribbean. I had tha t on my notes to talk about if time had allowed, I will steal a few secondshere. It is found around Trinidad and Isla Margarita in the bottom right hand corner of the Caribbean but does not reach much farther. That corner of the Caribbean is characterized by being influenced by Orinoco estuary. The Orinoco estuary is an enormous thing, If you fly over it in a light aircraft it takes about an hour to get from one side to the other of that river mouth. So that influences the environment tremendously. Basically you haveestuarine conditions there and in the Guianas,with high productivity and good feedingfor ridleys.Caribbean waters, on the other hand, partake more of the nature of mid-Pacific waters relatively blue, clear, corra line waters, low productivity, not really ridley habitat. Ridleys have had ample opportunity to get into the Caribbean. They live on both sides of it in the Gulf of Mexico and the South Atlantic, and geologically speaking they could have come in before Panama closed from the Pacific. But, I believe that the strays that may from time to time get there simply do not find it appropriate habitat, and are unable to make a living or to reproduce there for that reason. There was one record of a ridley nesting on the Caribbean coast of Columbia reported to Reinhart Kaufmann some years ago, and I believe that you, Rend, received that record. This was published as a Lepidochelyskempi. I believe that it was more likely to bc an L. olivaceain that the same tag number was used on a L. olivaceain Suriname and that the dimensions of the animal were more closelysimilar to that of the known animal from Surinamethan the animal from Rancho Nucvo that corresponded to that tag number, Robert King. How do you explain thediscrepancy in your asymptotic kilogram! limitation in your equationas opposedto what is actually nesting on the beachdown in RanchoNuevo? Caillouet: Well, I think the turtles grow more rapidly in captivity, and as a result they may mature earlier, and therefore grow slower thereafter. That is just a speculation, but that was my interpretation. Jill Cl arridge: It is interestingabout the Vibrio diseasesof theturtles. Theyhave shown that in the bluecrabs, 95 percentof the stressedcrabs have Vibrio spp,in their hemolymph,and in the unstressedor healthycrabs still 75 percentof themhave a hemolymph thathas Vibrio spp.in it. I wonderif youhave any baseline data for healthycrabs whether they have Vibrio spp. Lewis: I am awareof Rita Colwell's work that you are referring to. But, I am not awareof therebeing any baseline data on that. However,I think if you comparethe two, the turtles havea fairly advancedimmunologic system that neitherblue crabs nor otherdecapod crustaceans possess. And soit maybethat that is just part of thenormal existence of a crab and is sort of abnormal existence of a turtle. Clarridge:And also there are some Vibrio spp.that have this very virulent toxin which could account for thatvery quick demise that you noticed. Lewis; Exactly. You are right, In fact, Vibrio parahemolyticusnot only has a very potent toxin, but it is one of the organismsthat has the fastestreplication time of any bacteria,A replicationof about eight minutes. Clarridge:That is right. Someof theothers, vulniflcus anddamsella, also have these very strong toxins, Witharn As a comment,this matingbehaviorcan sometimesget pretty violent, A fewyearsago therewas a casein OceanWorld in FortLauderdale, where a maleloggerhead got soaggressive that he simply bit the female'shead, crushed it and killed her. Caillouet:It hasbeen our experience,with someof theyearling Kemp's ridley turtles and perhaps some of theolder ones that areintroduced either in groupswith oneanother or with loggerheads,that theywill pesterone another. They will injureone another.But also,when they are put in with loggerheadsthey will irritate theloggerheads until theloggerheads kill them.Do

241 youthink that putting even large ridleys in withother species is sort ofjeopardizing them? Bentley:In general,no. I thinkthey get used to being with other turtles, at least it wouldappear to be the case in the Seaquarium,"Little Fox" seems tobe particularly pugnacious despite being bitten that severely. When she was put backin with theother turtles she did notseem to have learned her lesson, and she continued to chasethe other turtles. Theother three adults did not seem to have very much trouble with the other greens or hawksbills,and they did not seemto fightwith each other or aggressagainst each other. The head started animals were held rather closely in captivityfor a numberofyears, and I thinkthey had some trouble with aggression when they were very little, maybe oneor twoyears old, but theyseem to haveoutgrown that for themost part. They can be held, all sevenof them,within thatone tank, 7meters by 7 meters,and there seems tobe relatively little interaction. They will just stack one next to the other. Forcucci:Do you have any explanation why fhose ridleys nested at night? Wood:No, I donot have any idea. I wasactually thinking that, although ridleys normally nest during the day, it was notthat uncommon forthem occasionally tonest at night, but it apparentlyis so. I reallydo not know. All ofour greens,ofcourse, only nest a night,t and we did have them up crawling. There isactually less interference during the daythan there isat night, because a nightt all the greens areup, and we have our watch people collecting eggs as they arelaid, walking around with flashlights and so forth thatsort of thing,So I reallycould not say, I wouldhave thoughtit wouldhave been quieter and everything during the day. Caillouet:Wouldit befair to ask you whether ornot you are exploring thisrnatter ofsex-reversal asit mighf relate tofertilify ratesin greensand maybe later in Kemp'sridleys? Wood: As far as having genotypicsex determination? Caillouet: Yes,with temperaturecausing sex reversal. Wood:We are looking into it. From what we can learn, the technique fordoing that, the antigen technique, seems to berather complex, and also there seems tobe some question asto the validi ty of the test, and so our deba te is whether itis worthwhile getting involved inthe hassle oftrying to do that. Ifsomeone elsehad a setuptodo it, we would surely liketo get involved init, and I thinkthere issuch a fellowsomewhere inTennessee. Wewould like maybe totry to workwith him, provide samples, andget someone elsewho knows how to do it todo it, That sounds like a betterway to do it, Carr:How closely relafed are the individuals which you have in captivity? Wood; I do not have a clue. The 96 we have came from Galveston. Carr: Youdo not knowif theyare from thesame clutches or not? Wood: I do not know about it. Carr: Couldit beinbreeding depression? Wood:The ques tion of whether the 96 animals tha t wehave are all from the same clu tch, or from ten dif feren t clutches, or whatever, I just do not know. Caillouet:The record keeping atthat time was not very good. My guess isthat it wasprobably a number ofclutches, butI donot know the clutch histories ofthose particular ones, We did not really start keeping track of clutches until I becameinvolved in the projectin Qctober19S1. Carr:Do you have any idea why it isthat the behavior exhibited bythe ridley seems tobe rather aberrant incomparison tothe greensyou havein captivity? Wood: How is that? Carr:Like fhe nesfing at nightand the males coming out. Wood:Well, we have had males corning out before and dig nest holes. Male greens have come out in Suriname and dugnest holes before. That in itself is not all that aberrant. Why they spend so much time sitting up on the beach, I donot know, other than the hypothesis weput forth that it wastheir reaction toa parasiteproblem. We have not seemedtohave quite that much problem other than that one female this year came oui and sat there for hours. But otherthan that, I donot know whatthey are doing, MarydeleDonnelly: l would like to ask you fwo questions abouf density atwhich these animals are kepf. Number one, when youfirst got fhe hafchlings andthe yearlings, hadthey not already found fhat there were aggression problems incaptivify, and thaffhey should be kept more separated than perhaps you had them at the farm? Wood; It was obvious that there were aggressionproblems. I meanpeople knew there were aggressio~problems, Fromour point of view,trying to dothis particular project at CaymanTurtle Farm was simply to seeif it werepossible to raisethem in a differentway. Galveston does a goodjob on theway they do it, but it isvery labor-intensive. We weretrying to seewhether it wouldbe possible to do it ona largerscale with lesslabor, That is basically why wedid it. Lateron, we did havevery highdensities of turtlesin a tank,You know, we wouldhave ten of theseturtles in a singlefiberglass tank. After having them two years, their aggressive behavior was completely gone. They did notbite each other at all. Donnelly: After how may years? Wood:They were a yearold whenwe got them.By the time they reached three years of agethey were totally non- aggressive toward each other. Donnelly:Itjust seems asthough it washigh mortality considering how precious theseanirnals are.Secondly, now that you have themin thelarge pond with the greens and hawksbill-green mix,is there any life-threatening aggressiort betzoeen thedifferent species,or betweenridleys? Wood:No, As I thoughtI hadpointed out, we now havethem in a separatepond. They are in thebig pond,but they are completelyseparated. That is how we can feed them the high protein diet now. Donnelly:Alright. It is myunderstanding also that you had better fertility results with greens once you reduced their diet somewhat.You had some greens that were considered fat, and they were breeding better afterwards. Do you think perhaps that thefact that your ridleys have bred, as opposed to the ones that were in theother facilities, is relatedto thediet regimen that i'hey are on? Wood:Well, thereare a lotof differences.I donot really know, Our turtles havegrown relativelyslowly, and we have notparticularly taken pains not to try to getthem fat, or feedthem too fast, But, other than that, I donot really know. Donnelly: Alright. Thankyou, Witham: I havejust a bit of information.A numberof yearsago when I raiseda pair of greenturtles in captivity, it wasa yearafter mating beforeshe laid any eggs.It just may be a point of interest. Pritchard: A pointyou did not reallybring up, firn,is thatI thinkthese r~'dleys are showing reproductive maturity at a smaller sizethan they are known to breedin the wild, Is that true? Wood:Yes, as far as I know, they were very small for matureanimals. Pritchard: Doyou notice this with yourcaptive-raised breeding green turtles, that theystart laying eggs at a smallersize than in the wild? Wood: Well, we had a couple that have, but generally I would say that our captive turtles are bigger than wild turtles. Our greensare bigger than wild turtles,and probablyreflect this more in the senseof being heavier,maybe fatter, than wild greens.We do almostall of our stuff on thebasis of weight rather that length so,I would saybasically our greensare probablybigger than the averagewild turtle when they comeup, whereasthe ridleys were smaller. Pritchard: Now theridleys showed a spurtin growthat theend of your curve,corresponding to a newdiet, am I correct? Wood; Yes, we did go back to the higher protein diet. Pri tchard: I see.That correspondedto thetime when these turtles that you would haveconsidered subad ult alsostarted producing sonteeggs. Is that correct? Wood: Well, no. It has been only the last two months or so that we have had them on the high protein diet. What happenedis that,once they werein thepond receivingthe 35percent protein diet, their conditionseemed to go down. By May of this year 985!, we pickedup a few animalsthat were weak and so forth, We had to pull them out and treat them,because they were losing fitness,and we could not seemto correctit by feedingmore of the 35 percent ration. Also, we deliberatelygot them wherewe could isolatethem and get them on the higher protein ration, and they madea reallygood recovery.The only thing I will have to considernow is where I want to only occasionally, you know, like two daysa weekor somethinglike that, feedthe high protein ration, so that they do not get too fat.

243 Panel Discussion Panel Members EdwardKlima, Director, National Marine Fisheries Service, Galveston Laboratory, Galveston, Tex., Panel Convener ChuckOravetz, Chief, Protected Species Branch, National Marine Fisheries Service, St.Petersburg, Fla. JackWoody, National Sca Turtle Coordinator, U.S. Fish and Wildlife Service, Albuquerque, N.M. PeterPritchard, Florida Audubon Society; Sca Turtle Recovery Team. CaroleAllen, Chairman, Help EndangeredAnimals RidleyTurtles, Houston, Tex. MilfordFletcher, Chief of Natural Resources Management, Southwest Region, National Park Service, Santa Fe, N.M. RendMArquez Millan, Jcfcdcl ProgramaNacional de Investigaci<5ndeTortugas Marinas, Instituto Nacional dc la Pcsca, Mexico. RalphRayburn, Executive Director, Texas Shrimp Association, Opertiag Remarks Klima I wouldlike to focus this discussion, if a tallpossible, onone major subject, and that is the fact that the Kemp's ridleypopulation isnot increasing and may, in fact, be decreasing. I am talking about the nesting population where wedo have some information. And I asksome questions: What isbasically known tn date? What do we need to know formanagement ofthis species, and what should wedo about it, given the fact that the population has been stable, or morcor lessstable for thelast 10 years at a verylow level? We all seemto takethat as an accepted fact. I would liketo ask Rend to give us some comments onwhat we know and what we need to know about this population to start this discussion.

13iscussion Mkrquez I thinkwe know very little, but we have some information, especially from the land-based pari of the life cycleof this species. Weknow roughly the size of the nesting population. Weknow some parts of the life cycle and fccdingbehavior. Wchave some information onadult migrations. Weknow that some part of the female population, afterthe end of the nesting season, goes to the north and some to the south in the Gulf of Mexico, split more or less in 50:50proportion. We do nothave adult recoveries outside of theGulf of Mexico.We suppose that all theadult population is inside the Gulf of Mexico. Weknow that juveniles goout through thc northern Gulf of Mexico, through the Strait of Florida, along all of the U.S.east coast to Europe,and some of themappear especially during winter in Europeanwaters. We have some recordsfrom the north coast ofFrance. Wehave another report from Morocco thatyou heard about yesterday. That isconclusive information, because these reports came from animals that were tagged, and we know how long it took themto go from the point of releaseto recapture.We do not know if thisrepresents the final travels of theseanimals. Wchope that all thoseanimals, or at least a goodpart of thispopulation, are able to come back to the Gulf ot' Mexico and contributeto the nestingpopulation. A pointthat is very important and one that we do not know about there was sIx~lation yesterday about this- isthe proportion ofmales to females necessary tohave a healthypopulation and good fertility, At leastuntil now at RanchoNuevo, the eggs have had good fertility. We have more than 90 percent, maybe more than 95 percent atleast, fertilityin thenatural population. That means that the health of the population ismore or less good. Because the fertilityhas not dropped yct, that means that the population cansustain itself for some years more. Butthe population will notsustain itself if wedo not make any progress against the decline in nestersand if we donot make progress instopping problems likeoil pollution and shrimp trawling. The problem ofshrimp trawling isvery important inthe U.S. and in Mexico. We do not know how many turtles are caught per year in the Gulf of Mex- ico.We suppose, forthe Mexican fleet of shrimp trawlers, that they are catching atfeast 500 per year. This is only sta- tisticallinformation. It is not exact, because fishcrmen donot always report when they catch the turtles. I do not know if thisis the case in the U.S, Maybe they do, maybe they do not. But in Mexico, they clearly do not rcport, because it isagainst thelaw to catch turtles, and thc fishermen want to have some meat tobring to their families when the trip isfinished. We very often saw flippers in the holds of the vessels. Usually we would scc flippcrs and meat, but some ofthem would also have eggs inplastic bags. When we saw only these things we did not know which species was caught,because theofficiaIs inthe ports cannot distinguish between species from only the meat and flippers. Some- timesthey rcport a tagwhen the animal has a tag.Usually they make the reports through the underground, because theydo not want to be prosecuted forpoaching, Maybe wedo not have information onall the animals that had tags. Anotherthing thatwe do not knowexactly or evenapproximately is theage of maturity.It is easyenough to speculateabout the age of maturityafter the nesting of two femalesat GrandCayman. If wedevelop a growthcurve from thecaptive animals and wc apply this curveto the wild population,we will be ableto moreor lessapproximately estimatethe age of maturityin natura1 conditions. We have the size a mat turity andthe weigh t at matur tyi forfemales in wild or natural conditions.And we havethese same figures for femalesin captivity, I havethe measurementsfor about 10 males, all of them mature. And we know that the males reach maturity at slightly smaller size than the females.We could makemore speculationsabout the life cycle,but I think it will be necessaryto have additional scientificinvestigations at the nestingbeach and in captivity so we will havesome more data from captivity that are possible to apply to the wild population. It is necessaryto know moreabout how we shouldmanage the seaturtles. We now havethe refuge at Rancho Nuevo, so at least we will be able to save the nesting beach from construction and tourism. It is the most important nestingbeach for this species.But we haveinformation that this animal nestselsewhere under naturalconditions in smallgroups. For example, at Tecolutla we have historical information from Archie Carr, from Caldwell and from the fisheriesinspector there that everyyear more or less20 or 25rid leyscome in smallgroups during the season.We have someother locationsof nesting,but they arenot many.They arejust solitarynestings, as in lslaguadain Campechc. Maybethis is the easternmost point of nestingnow. And in Veracruz,there are some locations of nesting. We flew over that areasome years ago and saw turtle tracksof this speciesat severalpoints during May and June. Klima RenCis addressingthe major problem,which is that we do not really know whetherwe haveproblems of recruitmentinto the nestingpopulation, or whetherthe nestingpopulation is declining fasterthan is normal,or if it is a combinationof both. Rendraised a numberof questionshere. One is the unknown catchrates in Mexican waters by hshcrmen.It is estimatedat about500. We also do not havethat kind of informationfor theU.S. except through the strandingnetwork that BobWhistler and BarbaraSchroeder are working on,which givesus an inclinationof what is happening, plus a tagging program that was conducted off of Georgia. Chuck,would you want to addressways of trying to get valuableinformation on what the catchrates of turtles are? Of ways to do that? Oravetz Very simply, I think it would be nice if we knew exactlywhat all the sourcesof mortality and incidental take were, I think we have some information, certainly, that shrimp trawlers do take ridley turtles and account for somelevel of mortality. I think that we know theyare taken by sportsfishermen, hook-and-line, from piersand boats, etc.,and somewhatin gill nets.Ed mentionedthat we havesome stranding data, too, that give us someindications, but I think that it is often difficult to make sound management decisions based on indications. I do think thereare certain ways and stepsthat we cantake to gatheradditional information.One of the thingswc talked aboutearlier this week is to try to institute somepermitting system,primarily with the shrimp fishery which we think is a major taker of the species,and to legally permit certain sections of this industry to take and report on their incidentaltake of ridley seaturtles. We think this will provide additional information on what the takesmay be and where and when they are being taken. I think it will also stimulate some additional conservation measures by the shrimp fishery. We have had some preliminary discussionof instituting such permit systemsthrough associationslike the Texas Shrimp Association and have indications that they are willing to do this. So over the next several months we will be helping them develop applications and conservation plans that may allow this permit system to be introduced, which will help our information base. Klima Ralph, would you comment on how receptive that would bc based on your knowledge of the shrimp industry? Rayburn I have not really discussed with the Board of Directors whether or not they are oriented in this direction to take on the permit type of thing. Basically, it hasbeen an information discussion among Chuck, me and others about the possibility of doing this. I would think, though,that they would be responsiveas they havebeen in other areas to providing the necessaryinformation. Basically, we have, in the past, conducted surveys through the Association to determine the level of catchesof seaturtles. I guessthe question would be, if we did get involved in this program, what would be theresponse from theenvironmentaI community as far asthe credibility of any informationwe would get? Is it worth the hassle?If it does show that the catch rates are a lot lower than what is projected now by the environmental community, would that hold any weight? Or is it basically an exercisein fu tility? If it shows high catch rates, I amsure tha t will be usedby thecommunity, but if it showslow catch rates,will it have any credibility among you all? That would be my question, whether it is worth the hassle or not? Klima Peter, would you like to comment on that? Fritchard Your remarkssuggested there is a certainlevel of antagonism,I guess,between the shrimpingcommunity and the environmental community that I do not feel is that bad, We agreed several years ago to try the whole TED thing on a voluntary basis, so there was a lot of good faith, a lot of trust in that. We have not abandoned good faith and trust. I mean, we are disappointed at the rateat which TEDs have been implemented, and weare looking for ways

245 thatwill workto get TEDs implemented more. But, we are not looking for waysof nailingthe industry to thewall orcausing it any more grief or inconvenience than is absolutely necessary tosave the Kemp's rfdley, And, if themeans thatare necessary toprevent the vessels capturing ridleys can be provided without any expense tothe industry, if outsidefunds can be found to buy TEDs and this sort of thing,we would be delighted, In otherwords, we do not regardyou as the enemy. We regard you as a partnerin thiseffort. So, we would not automatically disbelieve data thatcarne from the shrimpers, I mean, if it lookedas if thedata were very different from data that carne from other sources,then we would wonder why the discrepancy occurred. You do bring up a goodpoint. If thedata revealed thatan awful lot of ridleys were being caught, you would probably have some comments within the industry about notrevealing this. Or some people might say "look, we are going to give people the rope to hang us if wereveal this data."And that would bc a dilemma.So, I seethe worry that inspires you to makethese points, but I donot think thatyouneed feel that we are looking to make trouble. We would rather simply find ways of letting everyone continue todo his thing, but keep the Kemp's ridley population vigorous, I have been a littlegeneral, I am afraid, but that is how I feel. Rayburn Well, I appreciatethat and certainly that is one of our reasons to stayinvolved, because I think that has beenthe attitude. ! couldnot really sense from this meeting whether that attitude had varied much or not.But, certainlyit would be my intent to get the type of information needed, For instance, Rend was pointing out earlier that muchof theimpact may be outside where most of mymembership fishes. It maybe closer inshore; there has been a rapidincrease in theinshore fisheries for shrimpover the last four or fiveyears. If wecould find thatthere are critical areaswhere at certain periods of timewe do have a largeincidental take of ridleys, I think that would help us as we try to solve this problem. Kiima Carole, would you like to comment on that. Allen Yes,I would, This must be the conservation block right here. I appreciate knowing your feelings representing theshrimp industry, and I seethat it isvery important that HEART and conservation groups such as ourselves know exactlywhat you are doing and how hard you are working to solve this problem. Because people contact us and we haveinterviews, there are questions asked, and we should have total information sothat we are not blaming anyone prematurely.Wecan say the shrimp industry knows the problem and you are working on it andthese steps havebeen taken.Maybe this is a pivotalpoint. Going on from here there will need to be morc emphasis, But I think,as far as HEARTis concernedand what we are doing, it isvery important that we be in touchwith you and have good communicationwith you, which we have not been or hadbefore, So this is a goodstep. Pritchard Ralph, in ourdiscussions of the best way to solvethis joint problem, I would ask you to giveopen considerationtothis issue of mandatory TED at a certainpoint in the relatively near future. Ibelieve that the objections toit area psychologicalrod-and-reel. Thedevice now is inexpensive, Without webbing, it is a coupleofhundred dollars;with webbing, it is, what $300,$400? The daily costs ofrunning a shrimp boat are high. Even if the shrimp industryis in economicdistress at the moment, which it is,I donot believe tha t suchcost would be the make or break ofthe industry oreven of the individual vessel, especially since the statistics show that one can anticipa teon average abouta 7percent improvement inshrimp catch. And you would have the data, I think, tocalculate how long it would takefor that 7 percentincrease inshrimp catch to pay for a $200TED frame. I do not think it wouldbe very long. I thinkwhat we are mainly doing is asking a rather conservative sectorof the community todo things slightly differently,butnotdrastically differently. And I thinkif theygot these things, within a weekortwothey would regard themas standard equipment. Theywould have gotten over that psychological hurdle,and would regard shrimping assomething youdo with a TED,asstandard practice, with no more worry, no more complaints andprobably, once thething was amortized over a monthor two, moreprofits. Rayburn Wc11, I do not think we want to get into a debateover TED. Let me explain one thing that I thinkmight behelpful toyour unders t and ing, There are dif feren t numbers oftra wls used by the in du s try.The bay boa tstha t you seeon your way to this facility are generally pulling one main net and a try-netduring this time of year; one main netthat can range up to 95 feet measured from the tip end of one door around the head line to the tip end of the other d oor,plus a try-net,Normally, thatis a dayfishery, soif theytear up their nets they will just come back home. Ifyou lookout in theGulf and you see the larger vessels there are some smaller nearshore vessels those further out are pullingfour nets. They also have a co~pietesetof spare nets onboard their vessels, because thepotential forhang- upsis fairly enormous when you consider allthe oil exploration activities being done in the Gulf. They cannot afford, at20 gallons offuel an hour, to run six or eight hours back home to get a newset of nets, and possibly lose a crew, thengo back. So, they carry spare nets. So, you are not talking about buying one TED. You are talking about buying atleast eight TEDs. Not only for the four nets they are pulling now but also a try-net,Let us say five nets, and an extra fivenets; so you are talking about ten TEDs onboard the vessel. So, you are not talking about $300- $400. You are talkingabout $4 000. You are also talking abou t additional costs because, while it wassaid these things might last two years,most of thetrawls that are used in thefleet do not last more than six months. The reason is thewear and tear, butmorc importantly the potential for hang-ups wiping out the entire net. The TED is going to create even more

246 vulnerabilitywithin thetrawl. So it is nota simpleeconomics of $200.Also we might look into the 7 percentincrease in catch.Unless it hasbeen changed, as I understandit, thatis thedifference in catchinga smallnumber of shrimp in onenet compared to theother. I mean,you are talking about seven shrimp on one side and five shrimp on the other, That7 percentisa verysoft number, am I notcorrect on that, Chuck? The 7 percentissomething that is being used. Yes,we put it out asa 7percent increase in catch,and it soundsgood, but if youlook at whatthat is based on it isnot anything tha t I,with any sense of comfort, could sell to our people the idea tha t youcould pay for these TEDs because youare getting a 7percent increase intake, It isnot true, It isnot true, as I understandit,unless there is new information on that. Klima The information I haveis, of course,tha t theTED's main advantageis that it eliminatesa largeportion of the by-catch.This is very, very enticing to Louisiana fishermen where they catch 20 pounds of finfishper one pound of shrimp.This is with thefish excludingoption on theTED. As a resultof that reductionin the finfishcatch, it also reducesyour towing drag, which then reduces your fuel consumption, which is animportant aspect. And I would tendto agree,Ralph, that the increase in shrimpcatch is not a hardnumber, but thatthe decrease in thefinfish catch andthe resulting gain that you obtain from the fuel economy is somethingthat we need to lookinto. We need to have shrimpersaware of theseproblems. And I thinkin factChuck is startingto address many of thesethings at thispoint in time.Would you capsulizesome of thethings that you aredoing with theindustry, Chuck, in this area? Oravetz Let mefirst touchon themagnitude of theproblem, if you will. Someonehanded me a notethat I thinkis a goodquestion and maybe an example to try to useto sayhow we could solve it in oneway and maybe not the other. Someoneasked how wasthis same problem solved in thetuna fishery. Those of you thatfollow thatfishery would knowthat several years ago the tuna fishery had a majorproblem with theincidental catch of porpoises.And there wereregulations imposed that required, basically I think,some gear changes, some sort of a panelthat was used in thosetuna purse seines, as well as some additional procedures tha t wereemployed to releasethe porpoises from the seine.That problem was solved with thoseprocedures and with thetechnology, and it wassolved with a hundred tunaboats. TED is againthe solution to theincidental catch problem in theshrimp fishery, but insteadof a hundred owneroperators, you have 6 000big offshorevessels and another 30 or 40thousand inshore boats. Let me go one step further.Ralph Rayburn>isthe focal point of muchof thediscussion abou t whythe shrimp industry does this or that. I thinkhe is due a lotof creditfor being a leaderin theshrimp industry to try towork ou t acompromise and a solution, Whatyou need to realize is thathe may have a problemmaking commitments on behalf of anindustry -he represents only 400vessels out of a 40,000-boatfleet. So, you needto beaware that when you askfor a commitmentfrom the man,he will giveyou thecommitment to theextent he can, I gotoff on a littleside-track there. But anyway, we are continuingto do it, Thereis a solution.I think the basic problem is themagnitude of it, It isnot as visible and as critical andas potentiaHy diverse as the tuna problem, but it is just thesheer magnitude of it thatwe arefumbling around with. Mkrquez I havea question,I do not know if it will be necessaryto useTEDs all aroundMexico and theU,S. I think, at least in Mexico, we have two critical areas,These are the Campeche Bay near to the coast and in front of Rancho Nuevo, I think it is very important to use TEDs there during the time when the ridleys are there. Also, it is not necessaryto uscthem all yeararound. For example, at RanchoNuevo it is veryimportant to useTEDs from March until August or September,Wc know, from recoveriesshowing the areaof dispersionof the adults, which are the dates,the more important times,to protectthe area.Maybe that will reducethe cost for the operation. RossWitham This is just directedto Ralph,We havebeen talking about theabsolute cost of the TED unit, What is the costof the TED unit in relationshipto the net that it goesinto? If the netslast only six monthsand the TEDslast two years,what is the relationshipof thosecosts? Rayburn I was trying to figure that up. I think the rig is about $1,500.That is with the doorson, as I recall.So the webbingprobably is around $800pcr nct. Now, the reasonI saysix monthslife on a netis not so much that the net wears out, There is a lot of abrasion, of course,as it drags acrossthe bottom. But it is lost or just destroyed in one way or another.So, if you wouldlose a net,even though you wouldhave the two-ycar life on theTED, because the TED would be a partof the net I would assumethat it would be lost with the nct. So,you are talking about,on a per-net basis,a fairlyconsiderable increase in theoverallcost of thatnet whenyou includea TEDin it. Thoseare rough figures, Ross. Witham I think that is somethingwe need to look at. When you losea net,do you lose the boards?Do you lose everything?Or do you just losethe net? Rayburn Yes,well, of courseit dependson the size,the type hang,etc. In the majority of thecases, the net is ripped up beyondrepair, and in someof themore severe cases, the whoic cable will snapand you lose the entire system, the doorsas well as the trawl itself. In the more frequentcases there is a tear-upof the nct that canbe donein dragging through someheavy mud lumps or somethinglike that just weighting it down so much or actually hitting an obstructionof somekind- a pipeline,debris that is thrown overby an oil servicevessel, a capoff and stuff like that,

247 SallyMurphy Whilewe are talking about the cost of things thisis something tha t Ralphand I havediscussed before thereis a costthat seems tobe forgotten inall of this, the cost of cleaning up the carcasses onresort beaches upand downthe coast, I do not think anybody has figured what it coststhe municipalities along the coast to get crews out thereeach day to pick up carcasses andmove them and bury thetn or take them to the dump or whatever. Another unknowncost, and it issort of a philosophicalone,is the cost of one adult female Kemp's. Can you put a pricetag on that animal? Also,we started a discussion ona topic that Chuck broughtup about permitting shrimpers tocatch Kemp's ridleys inorder to get information onthem. I think we have two goals that are in opposi tion to each other. You have the goal togain information onKcmp's ridlcys and you have the goal to save Kemp's ridleys. Ifyou want to get TEDs on boats, youare not going to gct information onturtles because they are not going to catch any, If youare going to try to get informationonturtles, you are not going to get TEDs on boats. This is something wediscussed inour Section 6 project wedrew up. As long as the boats are not going to use TEDs, we ought to try toget the information from them. But to mc the principal goal shouldbe to get TEDson the boats. Klima Goingto a littlemore biological problem, yesterday Larry Ogren brought up some information ontag returns alongthe Atlantic coast that I foundvery interesting, Tocapture this very quickly, asI understandit,Kemp's ridleys aremoving along the Atlantic coast northward during the summer months, and du ring the winter or fall months they migratesouth. Also, Larry said that if theyget caught ina coldspell in thc Cape Cod area or in Europe they are lost tothe population. But, by and large, the data indicate that there is a migrationnorthward in the spring and summer anda southwardmigration in thefall, whichwould indicate that these turtles could contribute to theoverall population.Rcn6 also said that he believed that only adult turtles were in theGulf of Mexico and subadults are in the Atlantic.I wonderif anyonewould like to comment on whether these turtles in theAtlantic are still felt to be lost or do they contributeto the population. Pritchard I think Richard Byles made a goodpoint when he said that the Chesapeake Bayridleys that he handles ona regularbasis simply have the appearance of vigorous, well-fed, healthy animals. I think the occurrence of juvenileridlcys on the Atlantic side of the United States issufficiently abundant, orat least historically hasbeen sufficientlyabundant, and even today is sufficientlyhigh in termsof howmany ridleys there are in theoverall population,that it ishard to believe that these are all waifs. I believe that the ones that show up in France and Ireland andso on, historically have been waifs. Whether the head started one that ended up in Morocco would have survived is it still at large or was it caught? Klima It was released, was it not? Fontaine It was a live turtle and was released, Caillouet It wasjust as large as the ones that nested at Cayman Fa>m. Pritchard Really? It would be very interesting toknow if severalof themended up in Morocco,reached sexual maturityand started to breed if theyhappened to find each other, It bringsup theissue of whetherthere is the possibilityofnatural seeding ofnew breeding grounds forKemp's ridley. The species isa lotolder than any given beachorany given nesting grounds. And, for sea turtles asa whole, we tend to think that what istrue today has been trueforever, barring human exploitation andso on, But the fact is, beaches shift,islands come and go, orbeaches may notso much physically move but simply become unsuitable fornesting. And yet we are talking about species with 50or 100 million year histories. Sothere must bc some potential forridleys toform new nesting grounds, and we do notknow whether weare wa tching thenatural decline ofthe Rancho Nuevo nesting ground, orwhether itis entirely a resultofhuman caused predation oneggs and turtles that has caused it todrop to one percent ofwhat itwas only 40years ago. In 1947, byall agrccd-upon estimates, thatfilm shows 40000 turtles inone day, And even if we assume thatwas all the nesting thattook place that year, that is 40,000 nesting animals, Now we have perhaps 400nesting animals, or one percent of that. Klima In the whole season? Pritchard Yes, in the whole season. Woody One hundred and seventeen was the largest arribada this year 985! for comparison. Pritchard Well, I thinkwhen thc turtles get very rare the arribadas break down, And it couldhave been there was moreintegrity ofarribadas when there were tens of thousands ofturtles. And even from thc conservative viewpoint, lctus just say that with all the breeders thatcame outthat day in 1947 itmight not have been. Wehave lost 99 percent ofthem, and that would make you very pessimistic aboutthe future ofKcmp's ridicy. And yet, there isstill mystery in my mindas to how welost thatmany, that fast. I havesome black and white photos inmy files, taken I think by Antonio Montoya in1968, showing what I think wasthe last big arribada atRancho Nucvo, These arerather grainy black and white photos, butthey show thousands ofKernp's ridleys onthc nesting beach. Itis a pictureasimpressive asthe density you scc at the surviving arribadas

248 orLepidochefys oiivacea. And this was about three years after the Mexican effort started to protect that beach. And this wasnot just token protection. This was marines; this was fisheries inspectors; this was corrals with hatcheries; 25to 50thousand hatchlings released per year. And sincethat year, the species has been very well protectedat thenesting beachend, and there has not beensignificant directed take anywhere as far asI cantell, And yet it hascollapsed. I do not knowwhy. This is oneof thebiggest mysteries I think as to why it hasgot downto thislow a number.And I think we needto keepour eyesopen for anypossibility of newnesting grounds forming somewhere completed remote from Rancho Nuevo, but I do not know where they are, Klima I thinkyour comments are very interesting, The only correlation that we can look at from1947 to thepresent is thatman, in theUnited States at least,is certainlyusing the coastal areas much more than we havein thepast. Pesticidesincreased dramatically. Our utilization of baysand estuaries is phenomenal,and with theinformation that we aregaining from LarryOgren, head start, from strandings we find an enormousamount of Kemp'sin the shallowareas. Richard's Chesapeake Bay work andthe stuff thatwe find at BridgeCity, Tex., show that theyare caughtby a hostof means by man, Hook-and-line, gill nets,shrimp trawls, whatever, and most of them appear in the bays,We see an awful lot of this.I amnot trying to jumpto conclusions,but certainlyit is a speciesthat is very vulnerableto manbecause of its closeproximity and contact, at leastduring the summer months. And I donot think thatthere is anyblack and white correlation, but thereis certainlysuspicion. Ogren I think you hada question,Ed, about the presence or absenceof adultridleys in theAtlantic. Klima Yes. Ogren Theyare there. Given that the capture data we get is somefraction of thetotal population, we had one adult size going by Rene'sand Pritchard's size class minimum size of a femaleat RanchoNuevo at 59 cm at Cape Canaveral,and Lew Ehrharthad one from the Ski Lagoon.We both had about four othersthat werejust shy of that figure.I think Molly Lutcavagehad one dead adult shefound. There are adults in theAtlantic. Pritchard Theseare interestingdata. And I was awarethat you got thesespecimens just a few centimetersshort of maturityon theAtlantic side of Florida.Now, if ananimal meets the minimum size at RanchoNuevo, should we callit adultor not?Some animals may mature at 61cm, some at 66cm, some at 69cm, which seems to bethe pattern for seaturtles. The range of adult sizereflects more the different sizes at whichmaturity is reachedthan it does progressivegrowth after maturity is reached.I would be convinced if I heardof a bigadult outside the Gulf. If I hear of marginaladults outside the Gulf, the question remains as to whetheror not thatanimal had yet reachedmaturity even though it had reached minimum mature size. Ogren I think two of them are in the 60-cmclass, so they are well into the maturecategory, Pritchard Alright, 65 cm is around modal length for adults. Klima Milford, we havenot talkedabout Padre Island and what your plansare when we get someturtles moving up on thatbeach, and how you are going to dealwith thatproblem of peopleand turtles, and what kind of impact thatmight have, not only on the seashore,but what kind of impactpeople would have on turtlesas they are crawling up the beach. Fletcher Well, the obvious answer, of course, is to close Padre Island National Seashoreto all visitation, say the monthsof April, May,June, July and August, for a coupleof monthsof nestingactivities and a coupleof monthsof incubationand then a monthleeway. Unfortunately, I do not think we are going to be able to do that, unilaterally, right off the bat. We do have two of three questionsthat we need to answer first. We have the incubation questionsand the temperatureregime. So far, the preliminarydata that we areseeing lead us to believethat we areproducing a preponderanceof male tur ties because of thetemperatures at whichthe eggs are being incuba ted, We are going to, this comingseason, either put eachstyrofoam box in a separateincubator or put togethera large,walk-in type incubatorso that we canmitigate these oscillations that comefrom ambienttemperature swings during theday. The problemwith thatis, if and thereare a wholebunch of ifs sexis determinedduring that secondtrimester, and if sexis temperature-related,wehave to getthose eggs at Padreand in theincubators before the sex is determined, and lots of times we do not. The sexof thoseturtles is determinedbefore they ever crossthe border.They sit around in Mexicolong enough so that when we get them it is toolate for us to reallydo anythingof influenceas far asthe sex goes.So, that is onething we haveto do. Wehave to get thoseturtle eggs into Padreso thatwe canwork on this temperatureregime. The second thing is that if BobKing's data, which he presentedyesterday, about the beach profilesa tPadre are correct, the temperature at 45cm never reached more that 29.5 C duringthe season. If thatis so, thenwe need to startlooking for anotherchunk of beach,So this coming summer, we will setup anotherfive sets of temperaturemeasurements up anddown the 60 miles of PadreIsland to look at somemore sets of beachprofiles, [Editors'note mean nestdepth for Kemp'sridley is 30 cm, not 45 cm.] Someoneelse has suggested that our imprinting site may not be the best. We imprint these turtles right out behind

249 theranger station where we get the sand and so on. The question has been "Is that where you want them to come back?"The answer is"Lord, no!" We would rather have them come back 40 miles down the beach, So, probably next year,when we do the imprin ting, we will hatch the turtles out at the ranger sta tion at the hatching facilities, put them in a helicopter,fly them 40 miles down the beach to a suitablesite a beachthat we have predetermined will be satisfactory imprint them there, let them run down the beach into the surf, pick them up, put them back in the helicopterand Qy them to Galveston. Those kinds of things are all within the realm of possibility. Asfar as what we are going to do when all these turtles come back we have three or four things that we are going todo immediately, First, we are going to create and present to the public a brochureonsea turtles, in addition to the interpretiveefforts that are already ongoing at PadreIsland. This is whata seaturtle looks like; this is whata crawl lookslike; this is why they are important; here is who you let know when you see one. We are going to try to raise publicawareness onthat beach and see if wecan get more and more people looking for crawls and turtles and that sortof thing and reporting them, The second thing that we are going to do is we are going to put together a visitor usesurvey this coming year, and we are going to sample the visitors at Padre Island. We are going to sample them ana wholebunch of stuff, Where are you from? What is your average income? Weare trying to get a demographic profileof who uses Padre Island. Along with this visitor use survey instrument, we are going to aska numberof questionsabout rid ley turtles. Do you know what a turtleis? Do you like turtles'? Would you be willing to see the beach closedone day a weekforturtles? How about one week a monthforturtles? How about one month a yearfor turtles? Howabout six months a yearfor turtles? Let us get the public involved with this and first raise publica wareness about turtles,and secondly weare going to have to have public opinion behind us before we are going to be able to go in andreally actively manage what is going on at Padre. Aslong as the Texans think it istheir God-given right to drive theirfour-wheel drives up and down the beach, weare going tohave problems withsea turtles nesting, There isjust noway around it. If westartgetting turtles, say100 nesting, or200 nesting, I firmlybelieve that through organizations likeCarole Allen's HEART and the Corpus Christi Audubon Society and the "Fund for Aged Ox Cart Drivers" and thosesorts of things, I firmly believe that we are going to be able to put enough people on that beach volunteers- toreally be able to protect sea turtle nests and sea turtles. Now when I saw that film the other night again, and I looked outthere and saw 40,000 turtles crawling up that beach, I thought 'My God! That is going to be really difficult." I suspectif we ever get those kinds of returns it will be a mootpoint. I donot think that is going to be that big of a problem.Theidea, though, ofclosing Padre Island isjust about the same kind of inflammatory talkas shutting down thefish industry. You are talking about the sports fisheries organizations ofAmerica, the four-wheel drivers, the sportsfishermen, therecreation types. That is like playing with matches ina roomfull of gasoline. Wejust would reallyrather not talk about closing the National Seashore forseveral months of theyear until we have exhausted a number of other kinds of alternatives, Sowe are going tostart an increased levelof surveillance. Weare going torun more patrols down the beach, along withour program toremove hazardous wastes from the beach. We have an ongoing program toremove barrels and otherkinds ofhazardous wastes. Some Superfund monies have been plowed into that, Along with that program, we willincrease oursurveillance: wewill have a leastonegood observa tiontrip a weekforsea turtles up and down tha t 60-mileisland. And that is a 14-hourtripif youare in a bighurry to do it inone day. So those are the things that we aregoing todo, We are going toincrease thevisitor understanding bya brochure thatwill be handed outto everyone thatcomes onto the beach. Fortunately, atPadre wehave a captivesituation. Theyall enter the beach inone place, andthey all leavethe beach in oneplace. Secondthing, We are going tostraighten outour hatching facilities. Weare going toincrease thatsort of activity. Andlet us not kid ourselves, Wehave already decided that if weare turning out a preponderanceofmales, then we wantto raise the temperature. So,we have already decided, ina tacitagreement here,if we are going to raise the temperatureofthe incubation ofcouple ofdegrees andhold it between30and 32'C, then we have tacitly decided thatwhat we are going totry and do is produce a 50:50 splitof males and females, Nobody has written itdown, but thatis what it means. Sowe have already agreed that we are going totry to raise that. And we are going toincrease our beachpatrols. Questions and Answers

CharlesCaillouet: About the ridley that nested on PadreIsland this year- you collected the eggs, incubated them and sent them upto Galveston forhead starting, Is thisgoing to be sort of a standardoperating procedure forall ridleys that nest on Padre, or is theresome number after which the nests will beleft in situ. FletChe:I wish I had an answerto that. The reasonthat particular nestwas dug up was that it was right squarein themiddle of a four-wheeldrive track.And sowe knew that thenext four wheeldrive vehicledown the beachwould crushthat nest, We had no option. Certainly, the idea is to leavethose nests in situand hatch turtles at Padre,Right now,we are going to dig up everynest that comes along because of thepredation problems that occur with eggs.We feelthose nests that arenatural nests on Padre and I know it soundsarrogant are too importantto be left to chance. Weknow that turtle's eggs came from a femalethat found the beach and went up thereand laid hereggs. Those are very,very important eggs, And so, for the foreseeable future, suppose that I imaginethere is a breakingpoint we gotsay 25 or 30 turtles, we would probably pickup IO nests and probably put corrals around the nests and put some kind of surveillanceon them to seeif we could keepthe crabsand coyotesand other sortsof two and four legged crittersfrom bothering them. That of courseis our intent,to eventuallyhave that nesting situation as an in situsite. Is that evasive enough? LarryOgren: I hateto compound the problem with how to anticipate the females that might nest in thefuture, but I understand youare going to the trouble to define a differenttemperature regime, and you will dothat. I waswondering if you had also consideredchecking offshore to seeif thecurrent regime is favorable for thedispersal ofthe hatchlings out to thepelagic zone? Fletcher:So far Bob Whistler,correct me if I am wrong we havenot done any work with the currents.That will have to be looked at. Klima: I am glad you askedthe question. Fletcher:There is somethingthough. Bob Whistler tells me that this coming year he and his crews will bemapping seagrassbeds in boththe Laguna Madre and the ocean side of PadreIsland. So at leastwe will havesome better idea about seagrasshabitat areas, Klima: One of the fruitful thingsof this conferenceis that JackWoody and I wereable to talk to Tony Amosabout currents,and RenChas been interested in obtaining currentsoff RanchoNuevo for severalyears, We havea very tentative,soft agreement at thistime that The University of Texas'R/V LONGHORNmaybe able to gointo Mexico to look at currents,releasing drift bottlesand placingsome long-term current meters offshore of RanchoNuevo, pro- vidingwe havea permitfrom Mexico, we get a freecontribution from The University of Texasand providing a few otherthings. At thesame time, we maybe able to do somecurrent work off of PadreIsland, because this ties in with someother work that the GalvestonLaboratory is doing, We may be able to havea currentregime monitored from Aransasall the way down to RanchoNuevo, We hope that this will work out this year. SallyMurphy: In lightof the statement that was proferred yesterday by RodMast and John Carr, why do you not take a more conservativelook at movingnests and move only those that are in absolutejeopardy,and if theyare in a perfectlygood site and canbe protected by othermeans, then not movingthem might j ustbe a betterstrategy? FletChe:Yes, well certainlythat is a thought.Right now, though,what weare seeing is, if therest of thebeachis similar to theparts of thebeach that we havemeasured, the maximum temperature of thatbeach at 45cm is 295'C, So right now,if everythingis like wesaid it is,we wouldturn out males,entirely males, out of thosenests anyway, if therest of the beach is similar to what we have looked at. I hope it is not. Earl Possardt:Milford, if'the beach temperature profiles that you lookat this nextsummer all turn out to besimilar to theones youalready looked at, and you do not have a temperaturethat can theoretically turn out females, is the Park Service prepared to giveup theirridley program down there? Has that beenthought about? Fletcher:Unequivocably, no! We are into this thing,and whereelse are you goingto put them?I mean,if you aregoing to put them in the U.S.,and if the rest of the temperaturesparallel what we havenow, we are not ready to give up. We are looking at a biologicalpopulation. Therefore, I assumewe are looking a ta somewhatnormal distribution, which meansthat throughthe middle of thebell-shaped curve of normaldistribuhon, we aregoing to get themajority of thepopulation that will probablyhave a 50:50scx determination at around30; 3I; 32'C.But there are, no doubt, thoseon both endsof the curve.And what we may end up doing, assomeone suggested here yesterday, is selecting for anotherpopulation with different kinds of geneticcharacteristics, different kinds of sexdetermination tempera- turesand perhapsa differentlyskewed sort of male:femalerelationship in the population.We are a longway from giving up on this. We arejust starting to figure out what the questionsare. Woody: I think wc needmuch further discussionon that question,Fletch. We recognizethat is the ParksService's position,I assume.But, we needa lot of discussionon that. I do not agreewith you.

25I Fletcher;Are you intimating that the Fish and Wildlife Service is goingto pushthe Park Service out of this? Woody:Weare not goingto pushyou out of anything.But if Padreis not suitable,and we are not ableto establish whatwe would hope is a naturalpopulation, I think we would have to seriously question theartificial extent we go toand the resources we use to try to keep going with a veryartificial situation, thinking of whereour resources could beput elsewhere where they are so badly needed. That is not to say that Rancho Nuevo must not go on. It isthe key, andwe must initiate all management practices possible tomaintain the grea test number ofturtles throughout theGulf of Mexico, whether it is in Mexican or United States waters, A questioncame up before on whether you catch a lotof rid leys or whether you are catching a fcw rid leys. As far aswe are concerned, weare dealing with an endangered species. There are not a lotof ridlcys to catch. One ridley, fromour standpoint, under the Endangered Species Act,is one rldlcy too many. I deal with many, many, many species.The turtle is one with which I deal.Most of the species that Ideal with tha t arelis ted as endangered have more individualsin the reproducing population than we have in thatof theKemp's ridley. I willnot tolerate the loss of a pairofbald eagles, And we are not going to tolerate, ifwe can help itin any way, the loss of a singleendangered species if thereis a wayto prevent it. Now, I wantto make that very clear. Also, and I donot want to steal Rene's thunder, tagreturns based on RcnCs work show tha t between40to 45 percent of the adult tag returns come out of United States' watersand, for all practical purposes, 100percent arecoming out of the Gulf of Mexico. The other 50-odd percent are coming out of Mexican waters. DavidBowman: Perhaps Fletch orjack can help me. If I recallcorrectly, thei'emperature datathat have been obtained forPadre Islandhave been obtained priorto and up to the time that the hatchlings hatchat Padre Island inthe Styrofoam boxes.And they areonly collected in thearea of the egg house or relativelynear there. Is thatcorrect? Fletcher:What temperature dataare you talking about? Inthe Styrofoam boxes, orthe beach profiles? Bowman:The beach profiles. Fletcher:The beach profile temperatures, all right. Bowman: Am I correct? Fletcher:Ycs. There have been people wandering around poking thermometers inthe sand all over the place. But for allpractical purposes, thcreal research thathas been done the temperature atthree depths, every two hours, seven daysa weekforthree months has only been conducted behind the ranger station out at what we call our imprinting beach. Bowman:All right,so /ack is saying that if theturtles are only going to nest from April to july near the Turtle House, then he isopposed tocontinui ngfhe project atPadre Island. What I would suggest isthat we are far from having enough data to make anykind of a decisionabout the future of the head start and the imprinting project onPadre Island. And I amnot sure that we cansay that the turtles can only nest, that Kemp's ridleys can only nest successfully, fromApril to july. Fletcher:Yes,you could be entirely correct, inthat there isspeculation thatdown at Rancho Nuevo the early nests thatare laid when the beach is coldmay produce a preponderance ofmales. Later in thesummer when the beach warmsupthere may be a preponderanceoffemales produced. What wc may cnd up with is a populationthatnests in July or later in thc year, Marquez:I would like to make a recommendation,AtRancho Nuevo, thenests are not more than 35 cm deep. That means,perhaps 1'C morethan your profile. And that is very important inthis situation. So,you must examine the temperature around 35 cm for the nest. RossWitham: I hate tothro~ an extra variable inhere, but I thinkthere isan overemphasis onincubation temperature alone. Hereisa paperthatIj ustgot this morni ngthat says, fora different genus ofturtle, that the hydric environment influences the sexratio, Sofhe moist u recontent thatyou have gotin there, atcertain temperatures thatare supposed toproduce specific sexratios, will influencethe sex ratios or theapparent sex ratios of those turtles, Klima:Itshows you that we are still learning, andwe are learning very quickly. The program hasbeen going on for eightyears and we just recently have the inforina tion that we are producing a preponderance ofmales, and we are takingaction onthis. Jack may want tocomment, notonly on this, but also on what our plans are for next year on the beachand the number of eggsand so forth, Woody:I agree with Ross. I think certainly there isa relationshipbetween themoisture andthe temperature. They gohand inhand. Regarding plansfor next year, we recognize thereare two problems asfar as this scx thing goes. Wchave got a problemwithPadre Island and the way we handle it.We also have probably a bigger problem atRancho Nuevo.Thereason I say that is that a Ranchot Nuevo wegenerally havethose eggs inour crew's possession during thatsecond trimester critical period, Padre Island gets the majority oftheir eggs after it has already been determined whattheir gender isgoing tobe. And of course atRancho Nuevo, wcdo not have facilities tomanipulate temperature verymuch. We are going todo what we can. I think we can do something alongthese lines. We have otheroptions.

252 lt is a logisticalnightmare, getting eggsgathered at a certaintime or getting them out on a numberof flights or by boator howeverwe axegoing to do it, I do not know if that is feasible.Because of Galveston'sacquiring a newturtle house,we will requestof Mexicoan additional1,000eggs as we havein somepast years for a total of 3 000.Now what Mexicowill do is they wiII give us the 2,000and will put a stipulationon the permit that will not allow us to takethis additional 1,000unfil there are probably 60~ eggsin the corral a tRancho Nuevo. And I think this is a correctdecision on thepart of Mexico.So, we will have,theoretically, authority to takean extra 1,000 this coming year, but thatwill not occuruntil we,as I say,have a minimumof 60,000.Now, I amusing 60,000. That has been a figurein thepast. Mexicocould increase that number. They could decrease that number, But, from pastexperience, I would say it is probablyin that neighborhood. Klima:Thank you Jack.I wouldlike to just makea commenthere. Our datafrom the headstarting, I amfirmly convinced,indicate that this program is beyond the experimental stage, When we started the program, we were very carefulto identifythis as an experimental program, that being that we do notknow whether head starting was really goingto workand increase the total number of nesters.The indications to dateshow that obviously we canculture theturtles fairly well.When they are released into thewild, theyare found in areasthat we expect them to be found. Theyare growing very well. We do not haveinformation on theirsurvival rates because the tags drop off. However, wedo find thatwe haveturtles that are out for threeand four years. They have grown very well, and this leads to a conclusion,without proof, that this program is workingand that in thenear future these turtles are going to nest someplace.That is a hope,But it alsois based on some speculation of data,If theyonly nestat PadreIsland or if they nest at Rancho Nuevo or both, any of those combinations are successful. If that does occur, then we have a new managementtool available.And, howeverwe usethat tool we needto useit judiciously,because one of thekey questionswhich we aretrying to addressis, "Is therea lackof recruitmentinto thenesting population?" And that is trying to solve that portion of the problem. I know Jack has some co~ments here. Woody:All I wantedto sayis, I feelthis is still experimental.To me, in all duerespect to you,Ed and Ross, it is still an experimentalprogram. We will not havereached that point until, asyou saidthere at theend, you havegot identifiableanimals coming up at Padreor RanchoNuevo or someother area. I thinkone of my concernshere is I thinkit is goingto work,and I think thoseanimals are going to showup on thebeach that it is a veryexpensive proposition,Extremely expensive. It canalso be expensivein numbersof animalslost. This has not occurredhere, but it couldbe, and I guessI ama littlebit afraidof sellingthis as the thing to do whereit couldbe adopted in areas eitherin or out of theUnited States and not handledproperly because the resources are not thereto supportit, It is an experimentaltechnique, as far asI amconcerned, all theway, Klima:I agreewith you.We need to havecare with thistype of anoperation. I amjust saying that the data indicate thatit is goingto be successful.It is a Iongtime in gettingto the final product.The final productis, of course,the numberof turtles on the beach which we can identify as head started. This is going tobe virtually impossible toprove if theygo back to Rancho Nuevo because we wiII not be able to identify them because they will nothave tags, at least thosefrom the first portion of theprogram. If theygo up on Padre Island, and if weget say four or fivenext year withouttags, we would assume from that type of datathat those are head started turtles, because historically only oneor two Kemp'severnest in anyyear on Padre, So, again it isquestionable because of thetag problem that we have. Murphy;I think that Peter will verify that the recovery team agreed that it wouldbe experimental, andconsidered experimental, notonly until a taggedhead started female came up and nested, but as George Balaxs said andhe is nothere now in his commentson thereview copy of our recovery plan, we should say it is experimentaluntil headstarted turtles that nest exceed thatof the natural population. In other words, if you are trying to get recruitment into the natural population, head starting can onlybe considered a success if head started turtles exceed that of whatyou get from your natural population. Is that not what we said, Peter? Pritchard:Yes. And therewas an exchange about that in theMarine Turtle Newsletter, along those same lines the criteriafor evaluating head starting, The softest criterion might be if a headstarted turtle is recaptured and it isbigger than when you releasedit, thereforeit hassettled down. But the mostconservative criterion would be that thehead startedturtles are appearing on the nesting beaches in greater numbers or greater proportions than the wild turtles. Andeven more conservative than that would be that the head starting was the most cost-effective way. You invested probably$100 or $200in eachhead started turtle. Could you have achieved the same results by a differentform of managementforless money? So these are some pretty tough criteria at the right hand side of thespectrum. I have one ortwo theoretical musings about ridleys tha t dobear on the Padre Island situation. One thing is to contemplate why thisturtle nests so early in theseason, They are coming out in Apriland in Mayin thenorthern hemisphere when the wateris still cold, And I thinkyou might postulate that, evolutionarily, they have been pushed into that nesting season because,inthe northern hemisphere seaturtles cannot nest in the winter, very early spring or late fall. It issimply too coldfor eggs to hatch, When there were lots of turtles of all species occupying most of the availablebeaches, thegreen turtles,loggerheads and possibly the leatherbacks commandeered themid-summer slot If youhave ever seen a leatherbacknesting on a ridleybeach and seen five or sixridley nests thrown aII over the place by theleatherback diggingsitsbody pit, you can see who wins out in that particular coznpetition. TheridIey is the one that has toavoid thetimes orplaces wherethe big turtles arenesting. Itseems tohave done thatby coming injust about asearly asit can,Sothe eggs laid,say in April, willbe out byJune. Andwhen ihebig turtles arepeaking aroundJuly orearly August thelittle ridieys arealready out.This then constrains themtonest incertain places,You would notzeally wantto nest inA prilinVirginia orthe Carolinas. Itwould just be too cold, Soithas pushed theminto this semi-tropical nesting zone,andthe possible scenario isthat the Padre Island nesters aremale-producing, thatthis iswheze themales are made,orat least some ofthem. They arebalanced byfemales thatmight bemade inCampeche andinsouthern Veracruz,bythe stragglers therethat may approximately balanceinnumbers thenorthern PadreIsland nesters. The onesatRancho Nuevo are producing males early in the season andfemales later in the season. So,we are not looking ata series ofdisczete populations. Weare looking atone big population forminga sortof bell-shapeddistribution curvewith Padre Island onone side, Campeche orsouthern Veracruz onthe other orTabasco onthe other, andwith Rancho Nuevoat the peak. Then, thissuggests thatPadre Island isa male-producing areawhich hascertain implications, Itdoes notmean youabandon theeffort onPadre Island. Thereare important administrative reasonsforkeeping PadreIsland efforts going,and thereisnothing wrong withproducingmales, Malesareimportant too.But, itcould wellbe that we might want toadjust thetemperatures inthe egg house forthe artificially incubated increasingeggsthatare number brought ofnests toPadreon Padre Island Island sothatthey wereproduce producingall females. a disproportionateAnd then thisnumber will beof males.partly Youbalanced can get byvery an philosophicalandcome up with some interesting theoriesabout this whole sexdetermination thing,andI cannot reallygetinto them here.But itmay wellbe that you havegot thislatitudinal spread,one population, male-producing atthe colder endand female-producing atthe wanner end.Selection workingon individuals tendstokeep theoverall populationsexratio about 50:50, because assoon as you get a seriousdeviation from50:50, thena female thatputs feedbackhereggs inpositiona placeorcalculateda negative tofeedbackproduce positionthe minority where sexyou willwill tend zero toinbeon selected50 50, Butfor, thatAnd might this notwillbe be what an automaticyou want todo asa manager,Evolutionworks onindividuals, managerswork onpopulations. Andformanagement forrapid recoveryofpopulations, youmay want togo 60, 70or 80 percent femaleand get a morerapidrecovery, a greater achieveproduction the ofdesiredeggs andresult. so on, And you can do that by manipu1ations thatmay beunnatural andyet calculated to Carole,Klima:howLetmedo change you see theimprovingsubject justthebrieflypublic awarenessheze. Public for awarenessthe Kemp's isanridley important program? partofthis whole program. Allen:Well,first of all, I havezeally absorbed a lotuphere. Thecenter ofthe table hasfallen silent aswell asthe end. Wehave a gentleman herefrom the Houston Chronide, andI am very curious asto what I am going tosee inthe thatpaperthistomorrowis an extremely aboutthehealthy future exchangeand what ofisgoingideas, on.thatBut theze I wantis soto muchpoint yetout,toand do. IThis am sureisnot that whathe sees I wasitthis going way, to answer,soI am clear offthe subject. Weput a lotofmoney intoit, and we put a lotoftime into it, and here wesee thatrunningwehave, through this thisbasic entirethread group, ofdesire from tothe save philosophy an endangered thatSallyspecies. touched I thinkon thatto theis wherevery deepwe arescientihc all coming problemsfrom andI could notencourage allof you more, Eventhough youflgh t and scrap andhave toreally work things out,and continue,itis hard work. I like Peter's optimism aboutwhat could behappening here.The work atPadre Island needs to Weneed tokeep doing whatwe are doing inpublic awazeness. I seeonebig value thatwe have. Letme tell you abouta11,asIthe saidfact in thatmy comments we are losingthe aother lotofday, endangered people arespecies,extzemely Weareinterested in a taxinsqueeze, turtles. butThey I arebelieve extremely the majority concerned ofus feelthat wewant tohang inthere withthe tax money forprojects likethis. So, if HEART candoanything letus know. Nowweare sznall, butI like tofeel that wecando a kwthings. Wekind ofput energy whereit will pay offthe most. Weprimarily areconcerned withkeeping headstarting going,because wefeel until ailthese questions areanswered manythattheother best areasthing thatwe canwe dowant is to to keepgo into. raising theselittle turtlesand putting themout into the populaflon. Wehave Renbhasbeen very quiet, butI seea tremendous possibilityofthe American population reachingout to those in Mexicoforyou towhocontinue have forwhatso longyou workedha ve been so doing,hard, I because donot know long howbefore many we gotofyou into there this, aze,butyouweze we downwould there 1ikewonderingto do moze whatwasgoing tohappen tothese turtles. Sowe appzecia tetha t.Many, manypeople do,and wewant toput a lot moreenergy into that part of it- We have got some plans that we have already worked out. I wouldliketosuggest, inthe publicbrochuze thatyou NPS! are going todo, that you print them bythe millions, becausethereattheone Easter group season.thatwe So havewe aretoreach not only isaU goingthe collegeto be givingkids thatbrochures come down,inTexas, andbutthey we are are going going totohit be the sending beach Theythemtohave every got collegeto know campusabout inthe the turtle United program. States.It is very popular andyes,we have gottoface it,the kids arecoming. WeaIso want toorganize, Weare always looking forpeople, andI can guarantee execuflvesalaries of twice what I get,which is zero. But we need people in theCorpus Christi area. We see developments coming. We see the turtles comingin to nest,We seea Navybase coming, and we needpeople down therethat will do somework for us.That will backup all you scientists,so you canspend your time working out all thesevery deepproblems, and we cankeep the heatoff of you while you havethe time to do it, Soit is our privilege to do what we aredoing and we encourage you, RobertWhistler: I certainlysupport Carole IQO percent, and I think this is an important ingredient.In fact,one of the majoringredients in the programis the public awareness,I would like to suggestthat perhapsa considerationin the Padre Island situation is that we have just gotten an indication of what is going on. I would suggest that perhaps we needa lot more study in so far asbeach profiles are concerned, in different areasof thebeach, and perhapsskewed somewhatin so far as time is concerned,so that we get duplication, substantiationof data. I think this is very important,because I think we just have an indication. We do not havethe full understanding,So, I thinkreally we still havea lot of questionsto answerbefore we canreaoy say that that is the case. Barbara Schroeder:I zoouldlike to makea recommendation.This is somethingRend touched on the other day in the earlier meeting,There is somuch we do not know about temperatures. Wecould sit anddiscuss it for a longtime howto manipulate temperaturesin the egg houses or possiblydown in thecorrals of Rancho Nuevo. We know that, at leastat onetime, the Rancho Nuevobeachitself worked to producehatchlings that came backand nested. And 1know that there was at leasta littlebit ofzoork donelast year to protect nests, to useChuck's favorite work, in situ,on the beach. And I thinkthat maybe a littlemore effort should bedirected toward leaving a greaterproportion of thosenests- that are put into thecorrals now onthe beach at RanchoNuevo. Wehave good ways of keeping the predators out,at least on the east coast. I knoarthere are some different problems down in Mexico, but I thinkareshould look into keepingsome more of thosenests in thenatural situatr'on on the beach. We know it workedbefore, andare do not know all of thechanges that werrrrght be introducing zohen we move nests. We know some of themand there are probablya hostof'others that complicate what happens to hatchlingsafter they leave an artificialnest, even if it is in theground in another location, MkrquezrWe haveanother speculation. Maybe if we movesome eggs, or themajority of thoseeggs, if we put all those nestsin corrals,we areduplicating the situationofarrrbadas, themore natural condition of that population.When we put the eggsin the nestsin the corrals,we increase,a little bit, the temperatureinside the corralsas compared to the solitarynest. We do not know exactly.We are working with a wild population,and we do not know exactlywhat will be happeningwhen thoseanimals reach maturity. We do not know exactlyhow many malesand femaleswe are releasingnow from RanchoNuevo. But I do not understandexactly the question,but if we left more nestsin situ we know that the waves or storms and those sorts of problems would erode many parts of the beach where the nestsare laid by the feznales.So we needto move the nests,otherwise we will losemany nestsif we do not move them to a protected area. Schroeder:Well, I amnot sayingmove all the nests,but certainly theremust bea portionof neststhat arelaid high on the beach. 1 meanyou did say that last year or thispast nesting season you did cover some nests and leave them on the beach. Is thatcorrect? 1 amjust sayingthat maybewe should try to dothat with morenests. I donot knowhow many nests you did thatwith, butif it wassay ten, maybe are should try leaving25 or 30 neststhat are laid high up on thebeach, to let themincubate naturally and to let the hatchlirrgsgo down the beachnaturally, Mkrquez: Yes, that is a good point, But, we will need more people working there. For example, we need one student for every four in sit unests,Because otherwise they will lose the nest to predation. They must keep watching the nests, becauseof coyotes. It we left the solitary nests, far away from the camp, it is a very high possibility that we would losethe eggsto coyotes.I think we will be ableto protectthe nests in other waysto avoid coyotesor someother types of predation, but that would increase the cost of the project, Marydele DonneHy: I haveheard a lot of very informativeand interesting thi ngsin the lastfew days.I think that onething we cannotlose sight ofis thefact that this yearin 1985for thefirst timein apparently10years, there has been a declinein thenesting fernalepopulation. Jack has already said that theloss of anyadult female is very important,and hedoes not like to everseeit happen. But weare losing adult females in additionto smalleranimaLs. Now, Rene and Chuck have both touched on thefact tlratthere hasbeen a lossof Kemp'srid leys to shrimping boats,And I thi nk that we aregoing to haveto look very seriouslyat somesort of time table,whether it is voluntary or mandatory,about doing somethingconcerni ng this problem.For how many yearscan wesee a declineof, whatis it, 20 percerrtthis year? Woody: We are down approximately 200 nests out of, what Rend,620 some odd?We have 600 something nests this year.If we follow the normal trend with the stateof stability that we haveseen over the last sevenyears, we should have had approximately 200 more. The data are still rough. We do not have the final polished figures yet. Donnelly:I knowthat there is certainlyno easy solution to this,1 thinkit issomething that are should be looking at veryhard, and thefact is that the nestingbeach temperature profiles and everythingelse are not going to do usany goodif you arenot getting femalesin to nest. Henry Hildebrand: Well, I want to say that I agree100 percent with RendMdrquez. There is nothing naturalabout thatbeach atRancho Nuevo. Theyhave opened a road intoit. The first time I went inthere I had togo by horseback. I debatedwhether I should publish thediscovery ofthat nesting siteor not. You have allsorts ofpredators, bothtwo- leggedandfour-legged predators, there.You have thenatural conditions inwhich thewaves come in,so you have tomove thenests. There isno other way toprotect them, And the Mexican government is100 percent rightthere, NatFrazer: I seetwo disturbing trendsin today's discussion. Onthe one hand, itappears thatwe are making decisions basedon"biological information" thatwe do not really have, That is,in terms oftemperature andincubation, wedo notknow what the pivotal temperature forKemp's ridley is.In terms ofimprinting, wedo not know that that is occurring,butwe assume itis, and I agree itprobably is.But we are making decisions basedon this information which reallyisnot information. Onthe other hand, I seea trend inwhich weare not making thebest useof the biological informationthatwe do have. For instance, I didnot see anyone inthe audience oron the panel react toRend's suggestionthat,interms ofTEDs, wemight consider looking atthose areas inwhich we know ridleys occur and makingthemTED zones. Noone even reacted tothat, But we are willing totalk about putting a TED oneve~ shrimp boatinthe Gulf, which isprobably veryunrealistic. Interms ofnot making thebest use ofbiological informa tiontha t wedo have, I detect sortof a "blindmanlooking atthe elephant" kindofphenomenon, andI would suggest toyou thatifwe say things often enough insea turtle work they become true,I will begin saying thatuntil you take the biologicalinformation thateach ofus has and put it togetherina model, youwill not be able topredict the ramificationsofwhat wedo know, piece bypiece. Itis only when youput itall together thatyou can assess howfuzzy ourinformation is.That is, how important isit to be exactly rightabout ageat maturity? Orcan we afford a lit tie bit ofslop inthe data? Only byputting thisinformation togetherwith other information ina model willyou be able to assesswherewe need tofine-tune ourinformation andwhere wecan afford tohave ourinformation ratherfuzzy. Fletcher:I wouldlike to address thatfor just a moment,Certainly youare correct about theutilization ofthe best data. Butthere isa sayingthatyou "never cannot not do nothing," andby not making decisions, wemake decisions. Ifwe aregoing todrag eggs uphere from Mexico, andwe are going tohatch them, they are going iobe at a temperature. So,what wedo is make anhypothesis andthen we test it,And then wehave change andwe have tokeep adjusting thisas we go. For example, dowe know that we are producing a.pteponderance ofmales atPadre? Thatis what the datasay. But the data came from dead turtles. Arethey representative ofthe population? No!They aredead! The othersarealive. So,how can one say? Ifthere were indeed truths that were hidden inthe books, then we should be tryingtofind them. But there arenot truths that we know of.So we just move alittle bit here, and when werealize thatwe are going inthe wrong direction, wefallback andwe move a little bitthere, and itdoes look like blind men examininganelephant. ButI submit toyou, that insome ways that isprecisely whatscience is,lt is try this, ifit does notwork, try to figure out why, then go try something else. Klima:Thank youvery much. I think wehave hada veryinformative discussion. I wouldlike to thank thepanel membersverymuch fortheir time and their very thoughtful comments, andI would liketo thank theaudience as well,I think thishas been a fruitful meeting, I thinkwe have had exchange ofinformation hem,I think Nat's comments werevery appropria te,and I wouldlike to thank everybody forattending thismeeting, Appendix Welcoming Speakers

JaniceR. Coggeshall, Mayor Edward F, Klima, Director City of Galveston National Marine Fisheries Service City Hall Southeast Fisheries Center~lveston Lab. 823 Rosenberg 4700 Avenue U Galveston, Texas 77553 Galveston, Texas 77551 William H. Clayton, Former President Texas A&M University at Galveston Mitchell Campus Galveston, Texas 77553

Registrants Carole H. Allen Richard A. Byles HEART Help EndangeredAnimals-Ridley Turtles> U S Fish and Wildlife Service P.O. Box 681231 P.O. Box 1306 Houston, Texas 77628-1231 Albuquerque, New Mexico 87103

Anthony F. Amos Charles W. Caillouet, Jr, Marine Science Institute National Marine Fisheries Service University oFTexas Southeast Fisheries Center~lveston Lab Port Aransas, Texas 78373 4700 Avenue U Galveston, Texas 77551-5997 George H, Balazs National Marine Fisheries Service PaulD. Carangelo Southwest Fisheries Center - Honolulu Lab. Island Botanics 2570 Dole St. Environmentaland EngineeringConsultants Honolulu, Hawaii 96822-2396 714 Don Patricio Corpus Christi, Texas 78418 Robert Barber TexasA&M University at Galveston John L. Carr Departmentof Marine Biology Departmentof Zoology P. O. Box 1675 SouthernIllinois University Galveston, Texas 77553 Carbondale, Illinois 62901-6501

Albert Barr Robin Carter HEART TexasA&M Universityat Galveston 22206Fallengate Court P. O. Box 1981,Drop 302 Spring, Texas 77373 Galveston, Texas 77553

Pam Barrick Allan H. Chancy TexasA&M Universityat Galveston TexasA&I University P.O.Box 1981,Drop 209 Box 158 Galveston, Texas 77553 Kingsville, Texas 78363

Tim Bentley Terry J. Cody Max PlanckInst. for ExperimentalMedicine TexasParks and Wildlife Department Departmentof Physiology P. O. Box 1717 Hermann Rein Strasse 3 Rockport, Texas 78382 D. 3400Got tin gen WEST GERMANY Paul Coreil Louisiana Cooperative Extension Service David Bowman LouisianaState University Agric. Center US, Fish and Wildlife Service P.O. Drawer H P,O. Box 1306 Cameron, Louisiana 70631 Albuquerque,New Mexico87103 Ing. ErnestoCorripio Cadena Patrick M, Burchfield DirectorCentro Regional de InvestigacionsePesqueras Gladys Porter Zoo Instituto Nacional de la Pesca 500Ringgold St, Tampico,Tamaulipas 89000 Brownsville, Texas 78520 MEXICO

257 TimothyPatrick Delaney TexasA&M Universityat Galveston Phil Glass P. O. Box1981, Drop 331 US.Army Corps of Engineers P.O. Box 1229 Galveston, Texas 77553 Galveston, Texas 77553 Matthew Dickinson TexasA&M University at Galveston GregoryGray P.O. Box 2487 HoustonMuseum of NaturalScience Galveston, Texas 77553 One Hermann Circle Dr, Houston, Texas 77030 MaiydeleDonnelly Center for Marine Conservation TerryHenwood 1725DeSales St., N.W. SoutheastRegional Office Suite 500 ProtectedSpecies Management Branch Washington,D,C. 20036 9450Koger Blvd, St.Petersburg, Florida 33702-2496 Ben Drucker National Marine FisheriesService Kathy Indelicato Officeof Researchand EnvironmentalInfo. 13111West Markham ff102 Prediction,Analysis and Monitoring Div. Little Rock,Arkansas 72211 Washington,D.C. 20235 Luis Innes Marcel Duronslet Moya de Contreras4i120 NationalMarine Fisheries Service Col. LomasVirreyes SoutheastFisheries Center - GalvestonLab. MEXICO D.F. MEXICO 11000 4700 Avenue U Galveston,Texas 77551-S997 BruceJaildagian GreenpeaceInternational P.O. Box 384 Dennis B. Fenn Rational Park Service New SmyrnaBeach, Florida 32070 Departinentof Recreationand Parks TexasA&M University Frank W. Judd CoastalStudies Laboratory College Station, Texas 77843-2261 PanAmerican University RobertFigler P.O. Box 2591 Departmentof Biology SouthPadre Island, Texas 78597 TexasA&M University CollegeStation, Texas 77843 John M. Keri van Sea World of Texas Michele Finn 10500Sea World Drive TexasA&M University at Galveston San Antonio, Texas78251 P.O.Box 1981, Drop 37 Galveston,Texas 77553 RobertKing EnvironmentalProtection Agency Milford Fletcher Off, of Marineand Estuarine Protection WH-556F National Park Service Box 728 401 NSt,,SW SantaFe, New Mexico 87501 Washington,D.C, 20640 Edward F, Klima Clark T. Fontaine NationalMarine Fisheries Service NationalMarine Fisheries Service SoutheastFisheries Center~lveston Lab. SoutheastFisheries Center~lveston Lab. 4700 Avenue U 4700 Avenue U Galveston,Texas 77551-5997 Galveston,Texas 77551-5997 Dennis B. Koi David Forcucci RationalMarine Fisheries Service NationalMarine Fisheries Service SoutheastFisheries Center~veston Lab. SoutheastFisheries Center~iveston Lab, 4700 Avenue U 4700 Avenue U Galveston,Texas 77551-5997 Galveston,Texas 77551-5997 Miriam F, Korshak Nat B, Frazer KUHT-TV DepartmentofBiology 5327lmogene MercerUniversity Houston,Texas 77004 Macon,Georgia 31207 Andre M, Landry, Jr. Todd Nelson Departmentof Marine Biology TexasARM Universityat Galveston TexasARM University at Galveston P.O.Box '1981, Drop 379 P.O. Box 1675 Galveston, Texas 77553 Galveston, Texas 77553-1675 Larry Ogren Tammy Lobaugh National Marine Fisheries Service TexasAkM University at Galveston SoutheastFisheries Center-Panama City Lab P.O. Box 1981,Drop 209 3500 Delwood Beach Rd, Galveston, Texas 77553 PanamaCity, Florida 32407 Ila Loetscher Tom Olson SeaTurtle Incorporated P.O. Box 2575 HEART South Padre Island, Texas 78597 945 Highland Houston, Texas 77009 Peter Lutz University of Miami Charles Oravetz Rosenstiel School of Marine k Atmos. Sci. National Marine Fisheries Service Divisionof Biologyand LivingResources SoutheastRegional Office Miami, Florida 33149 9450 Koger Blvd, St. Petersburg,Florida 33702 BlancheLynn SeaTurtle Incorporated Cathy Lee PattesonPalmer P,O. Box 2478 5635 Clearwood South Padre Island, Texas 78597 San Antonio, Texas 78233

Sharon Manzella Donald E. Pitts, Jr. National Marine Fisheries Service EnvironmentalEngineering Southeast Fisheries Center~iveston Lab. TexasA8r M University 4700 Avenue U 700 Strand, Suite 206 Galveston, Texas 77551-5997 Galveston, Texas 77550

ReneMarquez Millan Earl E. Possardt Instituto Nacional de la Pesca U.S. Fish and Wildlife Service CentroRegional de lnvestigacionPesquera 2747 Art Museum Dr, Apaitado Postal591 Jacksonville, Florida 32207 Manzanillo, Colima 28200 MEXICO Peter C.H, Pritchard JulieMassey FloridaAudubon Society US. Fish and Wildlife Service 1101 Audubon Way 17629 El Camino Real, Suite 211 Maitland, Florida 32751 Houston, Texas 77058 Steve Rabalais Roderic B. Mast LUMCON MarineLaboratory World Wildlife Fund Star Route Box 541 1250 24th St., N. W, Chauvin, Louisiana 70344 Washington,D,C. 20037 Ralph Rayburn J. D. Miller TexasShriinp Association Departmentof Zoology 403 Vaughn Building TheUniversity of New England Austin, Texas 78701 Armidale, NS.W, 2351, AUSTRALIA Paul W. Raymond Pat Montanio National Marine Fisheries Service National Marine Fisheries Service Office of Law Enforcement ProtectedSpecies Management Division, F/PP2 P.O. Box 2564 Washington,D.C. 20235 Titusville, Florida 32781 Thomas H, Rennie Sally Murphy U.S, Army Corps of Engineers SCWildlife 8s lvlarine Resources Department Galveston District P.O. Box 12559 P.O. Box 1229 Charleston, South Carolina 29412 Galveston, Texas 77553

JohnMysing Dickie Revera National Marine Fisheries Service National Marine Fisheries Service SoutheastFisheries Center-Mississippi Labs. Southeast Fisheries Center~lveston Lab. Bldg.1100, National Space Technology Labs. 4700 Avenue U NSTL, Mississippi39529 Galveston, Texas 77551-5997 Kevin D. Richard 1719Cherry Bend KerryStanley Houston, Texas 77027 TexasA&M University at Galveston 1822 25th St, Galveston, Texas 77550 James I, Richardson Instituteof Ecology Universityof Georgia Harold Stone TexasA&M University atGalveston Athens,Georgia 30602 P,O.Box 1981, Drop 254 Darn..II Robertson Galveston,Texas 77553 TexasA&M University at Galveston P.O.Box 1981, Drop 79 Deborah Tarver Galveston, Texas 77553 National Marine FisheriesService SoutheastFisheries Center~lveston Lab. 4700 Avenue V Steve Robertson DallasAquarium Galveston,Texas 77551-5997 1stand Martin Luther King Fairpark! P.O, Box 26193 ChuckTurley BassPro Shops Dallas, Texas 75226 SouthwestMissouri State University Anne-Marie Sarutto 1935South Campbell TexasA&M University at Galveston Springfield,Missouri 65714 P.O.Box 1981, Drop 108 Galveston,Texas 77553 Kenneth M. Usob TexasA&M University at Galveston Donna Schaeffer P.O,Box 1981, Drop 66 HEART Galveston,Texas 77553 3707 Stillview Houston, Texas 77068 JamesW. Webb,Jr. TexasA&M University at Galveston Jane Scheidler DepartmentofMarine Biology TexasA&M University at Galveston/HEART P.O. Box 1675 2286 Shadowdale Galveston,Texas 77553 Houston, Texas 77043 Robert G. Whistler Barbara Schroeder National Park Service FloridaDept. of NaturalResources Padre Island National Seashore Stuart Field Station 9405 SouthPadre Island Drive P,O. Box 941 CorpusChristi, Texas 78418 JensenBeach, Florida 33458 Ted Williams Nancy Schwantes NationalMarine Fisheries Service Departmentof Biology SoutheastFisheries Center-Galveston Lab. TexasA&M University 4700 Avenue U CollegeStation, Texas 77843 Galveston,Texas 77551-5997

Elizabeth Shaffer Ross Witharn TexasA&M University at Galveston 1457 NW Lake Point P.O.Box 1981, Drop 201 Stuart, FL 33494 Galveston,Texas 77553 Jim Wood Donna J.Shaver CaymanTurtle Farm 983! Ltd. P.O. Box 645 National Park Service PadreIsland National Seashore GrandCayman, BWI 9405 S. Padre Island Dr. CorpusChristi, Texas78418 JackWoody US. Fishand Wildlife Service Jan Lee Shea P.O. Box 1306 TexasA&M University at Galveston Albuquerque,New Mexico87103 1405 25th St. Galveston, Texas 77550

Erich Kurt Stabenau TexasA&M University at Galveston 5800 Seawall Blvd, tf20 Galveston, Texas 77550