Impacts of Mortality Events on Bottlenose Dolphins in U.S. Atlantic Waters, Study notes of Natural Resources

Download Impacts of Mortality Events on Bottlenose Dolphins in U.S. Atlantic Waters and more Study notes Natural Resources in PDF only on Docsity!

Coastal Stock(s) of

A.tlanticBottlenose Dolphin:

Status Review and Management

Proceedings and Recommendations from

a Workshop held in Beaufort,

North Carolina, 13- 14 September 1993

Cornpiled by

Katherine R. Wang

P. ldichael Paysle

Victoria G. Thayer

U.S. Ilepartment of Cosnmerce National Oceanic and Att~~osphr:ricAdministration Natiotlal Marine Fisheries Service

NOPLA Technical Memoratldutr~NMFS-OPR- October 1994

'I'his report documents, in a timely manner, the comunication of preliminary results that were presented at the workshop. It has not undergone external scientific review. Also, the inlterpretation of the results of these studies, and the information presented at the workshop, represent the opinions and vxews of the participants and not the NMFS. This workshop report is intended to act as a first step towards the clo~npletioriof a Conservation Plan under the MMPA for the coastal st ock(s) of Atlantic bottlenose dolphin that, when approved by the A,ssistant Administrator for Fisheries, will prepresent the official position of the NMFS.

Suggested citation:

Mrang, Katherine R., P. Michael Payne, and Victoria G Thayer. (Cotnpilers)

1994. Coastal Stock(s) of Atlantic Bottlenose Dolphin. Status Review and Nhnagernent. U.S. Dep. Commer., NOAA Tech. Memo. NMFS-OPR-4, 121 p.

Contents

A. Species Desciription.................................... .... I

B. ?lie 1!)8711988 Die-off - What Happened?............................ 4

C. I l i e Coastal Migratory Stock - The Depleted Determination a............. I I

D. Determinatiom of Bottlenose Dolphin Stock Discreteness: Application of a Combined Behavioral and Genetic Approach........... 16

E. Recognizing 'Two Populations of Bottlenose Dolphin Tursiops truncutu.~

off the Atlantic Coast of North America: Morphologic and

IZcologic Considerations................................... 20

S'I'ATUS OF RESEARC:H AND/OR MANAGEMENT PROGRAMS TIIAT MONITOR

COASTAL BO'I'TLENOSE DOLPtIINS

A. Slu~rveysof U.S. East Coast Bottlenose Dolphin Abundance................ 24 B. R.esearch/Management Programs and Sfranding Networks in Coastal Atlantic Waters (by state).....^.^.............^.^..^.^.^.^.^..^.^...^..^......^28

NewJersey....................................... 28

Delaware........................................ 2 8

MIaryland....................................... 30

Maryland Department of Natural Resour.ces

h4ar'i,n%eh4arnmal Stranding Program............. 3!

Virginia.........................................^33

'"ha,[ Do We Know About Coastal Bottlenose Dolphins in Virginia?............................ 34 'I'en~pclraland Geographic Occurrences of' 7irr.siop.s I I. ~ ~ I I (. N I L (. S Strandings in Virginia, 1983- 1993.............. 4 1

North Carolina...................................... 45

Site-Specific Monitoring of Atlantic Coastal t3ottlenose Dolphins in the Beaufort, North Carolina Area....... 45 Marine Mammal Strandings in North Carolina............. 50

South Carolina..................................... 5 3 Marine Mammal Research and Stranding Response in South Carolina......................... 5 3 Population and Behavioral Patterns of Bottlenose Ilolphins in Bull Creek, South Carolina................. 5 0

G~wrgia......................................... 57

Flor~da^ -^..........................................^00

Bc~ttlenosedolphins belong to the Order Cetacea, 12amily Delphinidae and Genus Tursiops.

Currently, all forms of Atlantic bottlenose dolphins are assigned to the species truncatus, although

this statu!i may change as more knowledge is obtained about the different forms now considered to be

separate :fi~rsiopstrurlcatus stocks.

Bclttlenose dolphins are distributed throughout U.S. Gulf of Mexico and U.S. Atlantic waters.

In the U S. Atlantic, this species is distributed along the coast from Long Island, New York to the Florida Keys. North of Cape Hatteras, North Carolina, this species demonstrates a disjunct distributioin, with concentrations of animals near-shore (in embayme~ltsand within several kilometers

of the coast) and offshore, near the continental shelf margin, from 60 to 200 km from the coast.

South uf Cape Hatteras, the nearshoreloffshore distribution pattern is less distinct and there appear to be latitudinal clusters of animal concentration rather than the longitudinally discrete concentration areas Fou~ildnorth of Cape IIatteras (Fig. 1).

Sleasonal density distribution patterns have been described for U S. waters north of Cape I-latteras ((XTAP, 1982) and south of Cape Ilalteras (Burn et al., 1987). Durlng summer in the U.S. Atlantic, bottlenose dolphins are distributed along the coast, usually as far north as Long Island, New York, and offshore as far north as Nova Scotia, Canada. The main coastal concentrations of inigralory bottlenose dolphins during the summer occur from North Carolina, northward to New Jersey ( F g 2). During autumn, dens~tydistribution patterns suggest near-shore animals migrate south ailong the coast to Florida. During winter, coastal migratory bottlenose dolphins do not occur north of Cape Hatteras, but rather, are distributed from south of Cape Hatteras to the central Florida coast. 'I'hey are concentrated at the southern end of their range at this time (Fig. 2). Although bottlenose dolphins occur year-round along the southeastern Florida coast, in winter they are only about IlIlOth as abundant as along the central and northeastern Florida coast (Burn et al., 1987).

Du~ringspring, the dolphin distribution again shifts nortnwarti aiong the coast. ' It is unclear if the off:sliore portion of the population follows a similar north-south pattern, or what the actual extent of off:;lhorc: distribution might be, since sampling has generally been limited to areas within 200 km of the coa.st.

1Jigur.e1. L)istributional Knnge of Rotllenose 1)olphins illong the U. S. Atlantic Coilst

Figure 2. Areas of Majior Concentrations of Coastal Migratory Stock(s) of Bottlenose Ilolphins

lJ., The 1198711988 Ilic-off - .%illatIlal>pened?'

In1 1-oduction: I:I orn early June, 1987, unt~lMarch, 1988, unprecedented number \ of bottleriosc. dolph~ns, ~ M P J ~ O~ ~ U ~ Z C U ~ U J , ~ S washed ashore along the Atlafit~ccoast from New Jersey to I;lor~dal. Deta~lsof the ~nitialresponse to the event, subsequent organizat~onof a multi-disciplinary team of ~nvestigators,and scope of the analyses were provided In an unpublished Interim Report subn~ittedto the U.S. Marine Mammal Commission in May 1988. An account of the extent and in~pactof the mortalitly was prepared by Scott et al. (1988)

'I'he event was unparalleled, and therefore demanded a comprehensive investigation of proximate and contributing factors. Routine laboratory protocols were modified to meet rigorous re:;earch st;~ndards.Contributing laboratories with expertise in pathology, biochemistry, rnicrobiolo:gy, virology, conta~ninants,and biotoxins performed analyses on coded samples from the dolphins. Specimens 1For contaminant and biotoxin analysis were mixed with controls from unrelated 7'ursiops and four other cetacean species. At the termination of each study, data were transferred to our laboratory at the [J~~iversityof Guelph, arid integrated with identifying information.

'I'his report describes how the investigative process evolved, and the evidence implicating a biologic:al toxin as the proxi~natecause. The dolphins apparently were poisoned by brevetoxi~i,a neurotc~lxinproduced by the dinoflagellate Ptychodiscus hrevis, Izlorida's red tide organism. The dolphins were eventua:lly infected with a host of bacterial and viral pathogens which produced an array of be,guiling clinical signs.

Diiscussion: B'etween the time the first dolphin stranded in New Jersey in June 1987, and the last o n 1;lorida's east coast eleven months later, c~vcr740 animals died. The exact toll is not known. since almos,t certainly some aninlals were not rec:overed. However, Scott et al. (1988) estimated that 50 percent. or more of the coastal migratory stock between Florida and New Jersey died during this period. Without a guiding precedent to help uncover the cause, it was necessary for the investigation to sweep a broad range of disciplines before settling on the eventual path to the probable solution. The two most likely potential causes for an outbreak of d~ics kind werc csns.idcred to be infectious disease and poisoning. After weighing evidence from 18 n~onthsof field and laboratory analyses, we concluded that brevetoxin, the neurotoxin proc1uc:ed by the dinotlagellate Ptyc-hodiscu.~brevis, probably was thc proxiniate cause of this devastating event.

Early firdings led the investigators away f r o ~ nmicrobial agents as tlne principal cause of death. There was n o s i ~ ~ g l epattern o f illness that could be associated with a known pathogen, tilough it was clear that infectious agents contributed to and sometirncs domin;itetl the clinical picture. The first aiiirnals to conle ashore on Virginia Beach in late sumtner clearly had been ill for some time, with a condition that ultimately affected skin, liver, and lung, and led to the accumulation of fluid i the abdonlinal and thoracic cavities. Meanwhile, in New Jersey, Urs. W. Medway (University of Pennsylvania) and 11. Roscoc (New Jersey Division of Fish, Game and Wildlik) indicated in persorlal communic:ition that carcasses there were in better condition and less affected with secondary bacterial infectio~l. It appeared these differences were regional; dolphins coming ashore o n Virginia Reach died in warrner waters heavily contaminated wit11 opportunistic bacteria. Over 50 percent of the 21

species of potentially pathogenic bacteria isolated from 48 dolphins were of the genus Vibrio. These seernetl to have been a~ssociatedwith some of the problems in skin and blood vessels that ultimately killed rnany of the ariilmals but were not the primary cause of disease. 'The overwhelming nature of some of tht: infections, which probably arose in the lung, may have been related to im~nunc~incompetence,the cause of which cannot be established. The depletion of lymphoid follicles in spleen, lymph nodes, and the intestine supports this suggestion.

Some dolphins also had viral infections. Eight had a skin condition characteristic of dolphin pox (G~eraciet al., 1979), complete with susp~ciousinclusion bodies but in which no virus particles could he detected. In view of public sentiment expressed during the outbreak, it was comforting but not surlprisi~ngto learn that none of the dolphins examined showed evidence of retroviruses, the group of viru1r;es associated with Acquired Immune Ileficiency Syndrome (AIDS) and whose counterparts in animals could have bee11 a cause of reduced ability to fight normally harmless diseases. In any eveni, such viruses have a long latent period, and would not likely culminate in a single outbreak of disease. Dr. K. Sorners is continuing to characterize the reovirus-like particles isolated from an ulcer on the palate of a dolphin. It is premature to comment on the serological tilers to canine distemper virus, a n~orbilliviri~s,in 6 of 13 blood samples. Kennedy et al. (1988) have diagnosed morbillivirus infection and found distemper-like lesions in harbor porpoises, Phocoena phocoena, from the Irish sea. We found no evidence of such infection nor was a rr~orbillivirusdetected using techniques suitable for its propag;ttion. It is possible that the dolphins had been previously infected with a virus that escaped detect~on,or was no longer present at the time of the outbreak. A study must be undertaken to deterln~ne\whether the virus or other antigen responsible for the serological reaction is widespread In dolpliii~sand whether I I ~is a pathogen. This calls for an examination of blood samples from a broad range of cetaceans, and an investigation into the nature of the antigen.

Geographic and temporal patterns of mortality also lacked the hallmark of infectious disease. During August 1987, at least 125 dolphins stranded dead along the Virginia coastline; nearly 50 came

ashore in each of the months before and after. Others, according to fish-spottertpilot Mr. D.

'Thotnp;<on, were reported dead in small clusters at sea 18 miles from Cape May, New Jersey (August 21, 1987). To create such an overall pattern, an infectious agent would have had to be highly virulenl. -- causing acute disease across all ages and both sexes, spreading rapidly over a broaa geographic range, and killing groups of animals without pause. Viruses and some bacteria introduced either I-ly airhorn transmission or through direct contact are capable of producing such havoc. Seals exposed on crowded rookeries have fallen victim to epizootics of influenza (Geraci et al., 1982), morbillivirus, (Mahy e:t al., 1988; Osterhaus and Vedder, 1988) and leptospirosis (Vedros et al., 1971). Yer. there is little to suggest that these or other contagious organisms could spread as explosi~velyamong cetaceans. Dolphins are rnore dispersed in an environment which, unlike air, solid substrale or even a closed body of water, would not readily support the transmission of such agents.

Tht: accumulating evidence led us to consider a point source contaminant as the cause of mortality. 'This was alsc) a subject of public concern, as reflected by a train of media reports that sewagt: and toxic wastes were being discharged in the New York Right and Delaware Bay areas. We approached the Environmental Protection Agency to obtain information on permitted and illegal dumping of' rnunicipal and industrial wastes off the mid-Atlantic states, and sub~rlittedtissues for heavy rnetal and organochlorine contaminant ainalysis.

analyscs were con1plel:ed in January, 1989, Pb'Tx was found to be in the livers of 8 of the 17 beached dolphiris collected during the outbreak. N o toxin was detected in any of the 17 controls, selected from dolphins that died it1 captivity, others in regions or at a time not related to the fatalities under investigation, and three that died during capture in October, I987 (Table I). A greater nurnber of' analyst:^ would have aldded statistical weight to these findings, yet the tests are tirne-consuming, and by this writing, 34 dolphin samples in addition to the fish specimens were all that could be processed. The pattern is nevertheless clear: 47 percent of the 17 diseased animals contained the toxin; all the rest did not.

Levels in dolphin liver ranged between 80-16,000 nglg, and the calculated total amount in that organ was 0.08-14.7 mg. Assuming all the toxln was confined to liver, the total body burden would have tuc-er~2-290pg/kg, comparable to or orders of magnitude higher than the 2.85pgIkg level known to cause illness in man (McFarren et al., 1965). (^) These values are conservative. Standard extraction procedures are only qualitative for one urlaltered form of PbTx. Other fornis that are covalently bound or otherwise modified were not considered, nor is it reasonable to assume that all the toxin was in the liver.

Signs of PbTx poisoning in fish and rnatnmals are related to its action on the nervous system. Mice lose: rnotor control, become paralyzed and die of respiratory arrest (Baden and Mende, 1982). 'I'he site of' action is the voltage-sensitive sodium channel in excitablt: membranes, where the toxin causes increased sodium flux with subsequent depolarizatio~~and persistent activation of excitable cells (Poli t:t all., 1989). Death is rapid, and there are 110 reports of discerliable histopathologic changes in acutely poisoned animals. Might this account for the presence of PbTx it1 a menhaden recently consumed by dolphin KDL 644 (SWF-TT-BS04-B) that showed no evidence of toxin in its liver'?

Most of the clolphins did not die this way. They manifested an array of chronic disorder.? including fibrosis of liver and lung, adhesions of abdominal and thoracic viscera, and secondary microbial infections associated with inlrnune suppression, as evidenccd by histological charlges in lymph nodes. We suggest that sublethal exposure to PbTx precipitated the train of events leading to some or a,ll of these clironic changes. PbTx promotes peripheral vasodilation (Poli et al., 1989) and is card1iioto:sic (Rodgers et al., 1984). As a toxic aerosol, or once absorbed, it disrupts neural control of respiration (Rorison et al., 1980) and induces bronchoconstriction (Baden et al., 1982). Symptoms of poi:;oning in humans reflect the gastrointestinal and neurologic action of the toxin. They include nausea., vomiting, diarrhea, reversal of temperature sensation, ataxia, and numblless ant1 tingling of extreniitics (Ijaden, 1983). A dolphin so affected would likely stop eating, eventually ex1i;iust its blubbe~rreserve, and thcreby lose its passive buoyancy and thermal shield. The stress associated with these changes alorle could set the stage for infection by the ubiquitous opportunistic organisms that were is;olated from tht: affected dolphins. Superimposed on this, any direct neurotoxic efft:ct of Ph'l'x would be r~articularlythreatening to a diving mammal.

How Were Il~olphinsExposed to the 'Toxin?: Red tides in southeastern l J .S. waters normally originate 20--75 ktn west of the central f'lorida coast in the Gulf of Mexico (Steidinger and Iladdad, 1($81), and generally dissipate. Occasionally, as in 1972, 1977, and I980 (Roberts, 1979; Steidirigel- and Badell, 1984), they can he entrained and transported to the east coast of 1;lorida by the Gulf L,oop Current-l;lorida Current-Gulf Stream system. This happened in the fall of 1987, and resulte:tl in the eventual closure of shellfish beds along the North Carolina coast; there also were reports of' respiratory and eye irritation in fislierrnen and residents (Tester et a ]. , 1989), yet tile toxin

was found in tlie livers of dolpllir~sthat beached irl Virginia three months before that time. 'They rnust havc 4encounterecl the organisms sometinie and sornewliere along their northerly migration route.

In February, 1987, a /'. hr-evis bloom was 25 km from a point where Gulf waters are transported to the east coast. Drift bottle data (Williams et al., 1977) suggest that a fragment could have reached tlie east coast by spring of that year. The possibility exists that blooms had been occurr1i:ng all sulrlnier :in and adjacent to tlie Gulf Stream, and went undetected until a filament reached the North Caroli~lacoast in October, 1987. Such blooms would have been difficult to detect at sea, as they are not easily seen from vessels and tliere would have been little in the way of toxic aerosols, which are generally produced by waves atid surf action i11 shallow waters. Planktivorous fish lnight have cc?ins~uxnedthe cells offshore during their migration northward, and dolphins could have obtained this toxin by eating t11t:se fish or their predators. These conditions would have exposed dolphins both directly in water, and indirectly in food, to PbTx for an extended period, with effects manifested a short time later as they reached tlie mid-Atlantic coast.

Brevetoxin was recovered fro111 three yellowfin menhaden, Brevoortia srnithii, caught off Vero Beach, Florida in late February 1988, and one unidentified menhaden taken from the stomach of a dolphirl that stranded near Cape Canaveral on January 12, 1988. Thc finding of brevetoxirl in fish at that time and place suggests that there was a persistent, undetected bloom that kept the food-web contaminatc;d tlirough the winter. Alternatively, the bloorli that had delivered the filament to North Carolit-~,ain October 15187, had dissipated and left fish contaminated fix at least three months. The first scenario challenges our understanding of the process of P. brevis blooms, the second of the dynamics of brevetoxiri transfer in marine organisms.

In the fall of 1'387, on their southerly migration, dolphins encountered the bloom off North Carolirxi. F'. Tester (MOAA-NMFS Beaufort I,aboratory, personal communication) observed dolphins surfacing in the bloo111!j at that time. Three nio~ithslater, and perhaps all along, they were feeding on containinated fish. We: believe that this second encounter with the toxin -was responsible for the wave of stranded aninials recovered along the Florida coast in the winter of 198711988; three of six dolphins e:rcanlined had F'bTx-2 in their livers.

Lxvels of PbTx. rn the viscera of the live-caught rnenhaden tratislate to 200pg of toxin per 500g fish. Ilsiug this value, a dolphin feeding 011 ri~enhadenat a rate of 10 kg each day would consurne 4 mg of PbTx.. That is below the 6 mglkg LD50 for mice, but if general toxicological dogma is applied, much lower doses would be required to incapacitate an anirlial as large as a dolphin. In fact, only 0.2 tng can cause illness in people.

Not all the dolph~nswere po~sonedby eatlng t15h PbTx was found In the l ~ v e r sof three 11~1141tigcdlve\ I)olpli~tiWAM^205 (l'lble^ I ) ,^ wlth^ the^ h~gliestconcentratton^ of^ Pb'Tx^ In^ Itver,^ wa
e\tiniated to be le5\ thdn 3 nionth\ of dge The toxln had to li,ive been deltvered In thc n i ~ l k , suggesting that llke otlie~11pld5oluble rei~dues,I'bl'x m y be stored In fatty depoti and rnobil~zed ,dong with tat\ as the ,~nlnialdrdw\ 011 t h e lecelve\ There^ I ^ n o^ precedent lor^ the^ f ~ n d ~ n gof^ I'b'l'x In m ~ l k ,nor lids th15 loute of I'b'l'x e l ~ r n ~ n d t ~ o nheen cons~dered

Tlic c ~ r ~ u r n \ t a n t ~ d ievidence suggest5 ~ l i d tI'bTx 15 the rno\t probable c'iu\e for tile rnortallty ('o~ltr~but~ry;to the ult~~r~latedern~jeot the a ~ ~ ~ ~ n d l. iwd\ a ho5t of m~crobtal,~ndcnvrronmental factors 1h1\ I \ unI~E.elyto h'ive heen tile ftr\t tlme that dolphtn\ hnve been cxpo\ed to tlle toxtn I'^ Orevls

Table 11. Results of Elrevetoxin Analysis in Dolphin Samples from the 198711988 Die-off.

---

• ---^ Strandcti, Virginla,^ Aug^ 198'7- WAM 239 t- + + + 93 WAM ;!3 (^1) + + + + 83 WAM (^226) + + + I WAM 2.14 + + + 2 WAM 2.19 i- + + 2 JGM 418 +

Stranded, Virginia, Sept-Ozt - 1987

WAM 2:95 4- + + , + WAM 280 t + + + WAM 296 + + $- + WAM 282 + + + 2 CWP 273 t- Stranded, Florida Jan-Feb 1988 S-88-TI -5 1 + S-88-'IT57 t S-88-'T'I 01 t S-88 T7 I 1 1 K 644 t SS-88-TT-04 t

D ~ e dD ~ i r ~ n gCapture, V l r g ~ n ~ a - Beach, Oct 1987

VB-87-@I04 t VB-87-Cl14 t VB-87-CO9 t

Stranded, Teras 1987-1988 -

C 552 t C 391 -t C 575 t Sti-anded, n~itl-AtlanticCoast. - Aug-Nov 1988 WAM 331 I - WAM 136 WAM 110 WAM 732 WAM 7 3 5 WAM i

I (^) I )eak prcscnt, t)ut did not cornlgratc w~thstar~d;ird 'NO [ ~ ~ i i k! > U ~ ~ : Y Q I V C of Ph'l'x.

C. Tlhe Coastal Migratory Stock - Tlie Depleted Deterniinatior~'

Introduction: In the IJnited States, marine mammal populations are managed under the legislative authority of'the Marine Mammal Protection Act (MMPA) of 1972 (as amended). 'The ~nanage:~nc:ritgoal defined in the MMPA is optimum sustainable population (OSP) level, which has been defined to be population levels at or greater than those that protiuce maximum net productivity (MNP) to the ecosystem carrying capacity. Population stocks outside of OSP (below MNP) are defined as depleted. Ftemoval of animals (incidental bycatch, live-capture, etc.) can be legally authorizecl from non-depleted stocks.

Although there: is considerably more information on bottlen~sedolphins than most other cetacean species, for the most part, information necessary for deterrninatlon of stock status relative to OSP levels is inadequate. Generally, long, conslstmt indlces of popularinn production and abundance are necessary to determine OSP. However, there are cases where catastrophic changes in populations can occur, thereby allowing assessment of the degree of change and status relatlve to OSP. 'The recent masslve die-off of bottlenose dolphlns along the east coast may be such a case

Durlng the surnnner and fall of 1987 and the wlnter of 1988, an apparent dlsease epidemic resulted In the death and stranding of an unusually large number of ktlantlc bottlenose dolphin,

Tursiops IruncatuJ, along the U S east coast from New Jersey to central Florlda In response to this

anomaly, a multi-agency team was formed to Investigate the causes and effects of the rnortallty event T h ~ spaper is directed at the second component of the Investlgatlon assessment of the effects of the mortality event

Population Levels and Indices of Change: At the present there is no comprehensive

estimate of the size of the stocks of bottlenose dolphins in U.S. jurisdictional waters. 'The abundance of bottlenose dolphins in certain "priority" regions has been estimated. Scott, Ilansen and Burn (1988) suitrunarized these estimates and proposed that the number of bottlenose dolphins comprising the numerous stocks throughout both the U.S. Gulf of Mexico and U.S. Atlantic waters prior to 1983 may h;ave ranged to at least 23,000 individuals. Extrapolation of this estimate to existing abundance, however, assumes that: the stocks have been stable ovei.a,period of 15 or more years and that no net migration occurred during the different sampling periods of the studies summarized. The abundance of the stock(s) affected by the apparent disease epidemic was certainly less than the total number of U.S. Gulf of Mexico iind U. S. Atlantic bottlenose dolphins.

I-listorically, about 15,000 animals are tliougllt to have lived in mid-Atlantic near-shore waters, based on North Carolina shore-based fishery catch records from the turn of the century (Mead, 1!)'75). In 195'9-1981, the estimated average mid-Atlantic summer abuntlance of bottlenose dolphins is believed to have ranged from 4,300 to 12,900 anirllals (95 percent confidence region), including both the near-shore and offshore groups (CETAI', 1982). -I'l~ebcst availahlc information suggests that in recent times, coastal North Carolina and Virginia supported 1,200 or more dolphins during part of the spring and summer (Mead, personal communication). This number may have representecl a substantial portion of the mid-Atlantic coastal migratory stock prior to rhc disease epidemic.

--A- - ' ' ~ ' , i k i > r l f r o r n Si:i,t t i.L a1 ( 1 ' i H l i I

observed niortality may have primarily affected the coastal, migratory stock of animals that ranges hetwecn 1:lorida and New Jersey.

7111il: most direct means of assessing the impact of the nlortality on the dolphin populations is by con~~parisonof consistent pre- and post-event population indices. Assessment of impact on the basis of tht: number of dolphins stranded relative to the population at large is dependent on assumptions about the accuracy of abundance estimates and the relationship between the stranded carcass count and the true total mortality (Scott and Burn, 1987). (^) The number of animals observed washirig ashore is likely a fraction of the total ~nortality. There is also some chance that the reporting rate of stranded carcasses differed between the years prior to the stranding anomaly due to increased public awareness in 198711988. In addition, the accuracy of absolute abundance estimates Inay be questionable since the estimates are usual,ly of surface abundance unless there has been an effort to correct for animals sulbmerged at the time of the sample.

For the offshore stock in the mid-Atlant~cregion, a comparison of pre-event (1980-1981)

average population index, and an index based on a sample taken in August 1987, was used to assess the likely range of the impact of the mortality on the stock (Scott and Burn, 1987). The 1987 sample indicaled that the impact through August was most likely small (< 10 percent) relative to the 1980- 1981 summer abundar~celevel. Because the mortality event was not cornplete at the time of the August 1087 sample, arid due to the uncertainty about the population trajectory since 198 1, this result needs further testing.

For the coastal stock of dolphins, there are no consistent pre- arid post-event population survey indices yet available with which to assess potential impact. The pre-event patterns observed in areas :such as the Chesapeake Bay mouth (Keinath and Musick, 1988) and Nassau County, Florida (Valad~z,personal communication), may be confounded by the apparent increase in abundance in 1987 and 1988 Thus, the ]potential i~npacton the coastal stock was estimated by comparison nf the die-off period to prior year average stranding rates Inherent in this estimation is the assumption that stranding rate is a consistent index of the stock mortality rate.

As of June, 1988, 742 stranded bottlenose dolphins from New Jersey to Florida's east coast were n~portedto the Smithsonian Institution's marine mammal stranding events program for ,.the 1 I- month period from June, I987 through April, 1988 (Mead, personal communication). In the prior^3 years, for the same gt:ographical range arid months, an average of 73.33 dolphins were reported to the str;+ncling network. Thus, the 198711988 anomaly represents an order of magnitude increase (10.11 ti~nes)in reportcd strandings relative to the most recent 3-year t~istoricallevel. Natural mortality rates on the order of 7 to 14 percent per year are believed to encompass the most likely range for bottlenose dolphin populations (Hersh, 1 9 8 7 ~ ). Assuming the stock natural mortality rate ( t n ) to be the lower end of the reported range (7 percent per year, 6.42 percent per 11 months), that the ob:;ervecl mortality wholly affected the near-shore stock, and further assunling that the reported stranding rate is propor.tional to m and consistent between years, then the observed mortality

represent,s an 1 I-n1onl.h m of 64.9 percent. An annual rn would be slightly larger than this value if

only tl-ie long-term av'erage risk of death was applied to the final I-month period.

7 he annual rate of change in the dolphin stock abundance 1s the d~fferencebetween the annual rllortality rate, annual blrth rate, and annual net ~mm~gratlonrate For^ the coastal^ mid-Atlantlc^ stock of dolph~risaffected by the dlsease epidemic, Hlaylock (1984) observed up to 11 5 percent of the

population sanipled were calves, i~nplyingan annual birth rate (6) on that order. Data collected frorn strandeti aniti-ials suggest that calving for this stock occurs in the spring and is not gerierally protracted over the year. Thus, a potential decline for this stock since early 1987 is estimated as > 5 0 percent (17 - 111 := 53.4 percent)^.

Table 3. Parameter Values Used in Simulations of the Dynamics of the Coastal Migratory Stock(s:)of Bottlenose 1)olphins.

Sy~nbol Values

Natural ]Mortality Rate m 0.07, 0.

Mum;m-induced Mortality Rate f 0.0, 0.007 1n=0. 0.0. 0.014 n-i= 0.

Maximurn Net Productivity Rate MNP 0. 0.04. 0.

Maxirnum Net Productivity Level MNPL 0.6, 0. 8

Median ,4ge at Sexual Maturity x 8, 11, 14

Estimates of Human-induced Mortality: The magnitude of annual retnovals from this stock due to i,ncidental catch ,arid other directed human causes is not well documer~ted. Marine mamma: stranding clata provide information useful for estimating an index of human-induced mortality of cetaceans in the U.S. Ghlf of Mexico and along the Atlantic coast. Many of the cetaceans that strand are examined for cause:; of n~ortality. Burn and Scott (1988) examined the stranding data provided by the Smithsonian Illstitution (Mead, personal communication) for evidence of human-induced ~nortality in bottlenose dolphins. A total of 386 bottlenose dolphins stranded from central Florida nor-tl-i along {.heAtlantic coast were reported from January 1982 through May 1987. Of these, 9. 3 percent showed some evidence of human--induced mortality 01). Data from June 1987 to the present were not exa~nineldin this context because of the (then) ongoing stranding anomaly. For tlie range of the disease-affected coastal sl.ock of dolpliins, Table 2, lists the bottlenose dolphin strandings by state and type of purported human-.induced mortality.

,4s:;uming the cllassificatio~lscheme of Burn and Scott (1988) to be accurate and that tlie stranding data indexes I.~urnan-inducednlortality rate in proportion to the natural mortality rate, then t:stirnates of human-induced mortality rates can be derived from these data. Using the range of natural I-nortality rates (,rn = 7 to 14 percent), the additio~lalnlor-tality rate due to hurnan-related activities; (f) can be estimated by these proportions Cf = rn((l-1)) '-1)). Istimates off by state and for the coastal n~igratorystc.)ck range are presented in Table 3. 'I'hese data indicate that added mortality due to human activities n-lay range from 0.7 to 1.4 percent per year for the coastal migratory stock of clolpliins.