Postbreeding Dispersal and Drift-Net Mortality of Endangered Japanese Murrelets

The Auk 111(4):953-961, 1994

POSTBREEDING DISPERSAL AND DRIFT-NET MORTALITY OF ENDANGERED JAPANESE MURRELETS

JOHN F. PIATTx AND PATRICKJ. GOULD2 •AlaskaFish and WildlifeResearch Center, National Biological Service; and 2MigratoryBird Management, U.S. Fish and WildlifeService, 1011E. TudorRoad, Anchorage, Alaska 99503, USA

ASSTRACT.--Theincidental catch of seabirdsin high-seasdrift nets was recordedin 1990- 1991 by scientificobservers on commercialsquid and large-meshfishery vesselsoperating in the North Pacific Transitional Zone. Twenty-six Synthliboramphusmurrelet deaths were recorded in the months of August through December.All but one were from the Korean squid fishery in a small area bounded by 38ø-44øNand 142ø-157øE.Five specimensof the dead birds were collected and later identified as JapaneseMurrelets (S. wumizusume).As fishing effort was widely distributedover a large area eastof Japan,these data suggestthat postbreedingJapanese Murrelets migrate north to winter in a relatively small area southeast of Hokkaido, where persistenteddies form at the confluenceof the Oyashioand Kuroshio currents.Fronts between cold Oyashiowater and Kuroshiowarm-core eddies promote the aggregationof zooplankton and pelagic fishes,which in turn may attract murrelets during the nonbreedingseason. The estimatedtotal mortality of JapaneseMurrelets in high-seas drift-net fisheriesrepresents a significantproportion of the total world populationof this rare and endangeredspecies. Received 12 October1992, accepted 4 April 1993.

TrtE J•PAN•E (CR•ST•D)MVRR•t,•T (Synthli- apparentlywere killed by introducedrats (Tak- boramphuswumizusume) is the rarest member of eishi 1987, Ono 1993). The estimated total mor- the Alcidae,and its populationsare seriously talitywas 414 birds,and few breedingmurrelets threatened (Collar and Andrew 1988, Gaston are presently found in what was once a very 1992, Oho 1993, Springer et al. 1993). Along largecolony. The frequencyof carcasseson oth- with the endangeredShort-tailed Albatross (Di- er islands (Higuchi 1979) suggeststhat preda- omedeaalbatrus), Japanese Murrelets have been tion of adults is a widespreadproblem. In the declareda National Monumentin Japan(Has- Izu islands, increasing human trash has pro- egawa1984). Little is known of their population motedgrowth in populationsof the JungleCrow status,but numbersappear to be declining(Bra- (Corvusmacrorhynchos), a common predator of zil 1991,Ono 1993).Hasegawa (1984) estimated murrelet eggs (M. Ueta pets. comm.). Human 1,650 birds from colony counts and observa- visitation and the ensuing destructionof eggs tions at sea.Higuchi (1979) and Ono (1993) re- and nest sites also appearsto be an increasing portedconfirmed records of breedingon e'•ght problem at severalcolonies (Higuchi 1979,Bra- islandsor island groups,and at least 19 breed- zil 1991, Oho 1993). ing locationsare presentlyknown or suspected Away from their colonies,Japanese Murrelets (Fig. 1). The largestcolonies were found at Biro are potentially threatenedby oil pollution and Island(2,000 in 1992),Kozu Island(600 in 1991), gill nets(Kazama 1971,DeGange and Day 1991, Mimiana Island (150-200 birds in 1979), and DeGangeet al. 1993).The adverseeffect of these Koya Island (408 birds in 1974, 282 in 1976, 30 mortality factorsmay be underestimated,how- in 1991) [see below]). It is unlikely that all ever, becauseof the difficulty in distinguishing breeding colonieshave been located;however, JapaneseMurrelets from their closely related it is alsounlikely that a large colonyremains congener, the Ancient Murrelet (S. antiquus). to be discovered(Oho 1993).A recent compi- Outside of the breeding season,when Japanese lationof availabledata (Oho 1993)suggests that Murreletspossess distinctive plumes at the back fewerthan 4,000 Japanese Murrelets exist today. of the crown, they differ from Ancient Mur- Disturbance, habitat destruction and intro- relets only in having lesswhite on the sidesof ducedpredators are seriousproblems at many the neckand a slightlylonger bill (Gaston1992). colonysites (Brazil 1991,Springer et al. 1993). JapaneseMurrelets breed only in warm-wa- In 1987, the remains of 145 individual murrelets ter areas of southern Japan (Fig. 1) and tend to were found on Koya Island, where the birds remain closeto their coloniesin the breeding 953 954 PIATT•ND GOULD [Auk,Vol. 111 season(Brazil 1991, Ono 1993). In general, the We extrapolatedthe total number of murreletskilled postbreedingdispersal of JapaneseMurrelets is from numbers observedin nets using a simple ratio poorly known. Postbreedingbirds from the Izu estimate (Cochran 1963). In extrapolating total by- Islandsmay move northward becauseJapanese catch,we used only fishing effort and observations Murrelets are regularly observed off northern in an area bounded by 36ø-45øNand 141ø-147øE(Fig. 1). This limit is based on the approximaterange of Honshu, and rarely off Hokkaido (Kuroda 1955, murrelet bycatch near Japan (Fig. 1). We excluded Brazil 1991).Vagrants have been reportedas far much of the large eastern drift-net fishing area (Fig. north as Sakhalin and the Kuril Islands (Brazil 1) to avoid overestimatingthe bycatchof murrelets. 1991). Ancient Murrelets are abundant and We assumedthat the five murrelets caught between widespread throughout boreal and subarctic 152ø-157øEwere Ancient Murrelets (see Results). Five waters of the North Pacific (Gaston 1992, different drift-net fisheries were monitored during Springer et al. 1993). A few hundred Ancient 1990and 1991,mostly in an area (Fig. 1) bounded by Murrelets may breed around Hokkaido (Brazil 31ø-46øN and 151øW-141øE. 1991, Gaston 1992), and perhaps a thousand 1. Japanesesquid fishery.--The Japanesehigh-seas squid drift-net fishery generally operatesbetween breed along the coastsof mainland China and 170øE-145øWand 35ø-46øNwith designatedtime and Korea (Gaston 1992). Ancient Murrelets are area closuresnorth of 40øN and excluding the U.S. common winter visitors to coastal and offshore 200-mile (320-km) fishery conservationzones. The watersof Japan,particularly in the Seaof Japan fishing seasonlasts from Junethrough December.Ca- (Brazil 1991). nadian, Japanese,and U.S. scientificobservers mon- Concern about the incidental catch of marine itored 114,095km of Japanesesquid drift nets in 1990 organismsin high-seasdrift nets led to an in- and 103,198 km in 1991. ternationaleffort to monitor five major fisheries 2. Japaneselarge-mesh fishery.--Most Japanese large- in the North Pacific:Japanese squid and large- mesh drift-net fishing occursbetween 175øE-145øW mesh drift-net fisheries;Korean squid drift-net and 24ø-41øN, and between 141ø-175øE and 24ø-43øN. fishery; and Taiwanese squid and large-mesh As in the squid fishery,there are designatedtime and area closures.This fishery operatesthroughout the drift-net fisheries (Canadian Department of year but peaks in the winter. Japaneseand U.S. ob- Fisheries and Oceans 1991, International North serversmonitored 25,668 km of Japaneselarge-mesh Pacific Fisheries Commission [INPFC] 1992a). drift-nets set in the North Pacific from September We report here on the bycatch of Synthlibor- 1990 through May 1991. amphusmurrelets in these five drift-net fisher- 3. Taiwanesesquid fishery.--The major Taiwanese ies, and documentthe first confirmedbycatch squid fishing groundsare between 155øE-165øWand of JapaneseMurrelets in drift nets. We assess 38'-44øN.Some fishing, however, occurs south to 33øN, the potential impact of this bycatch on popu- north to 45øN, west to 143øE and east to 146øW. West lations, and considerthe geographicdistribu- of 170øE,fishing is prohibited north of 39øN. Time and area restrictions, and closed areas are similar to tion of bycatchrecords as it relatesto postbreed- those for the Japanesefleet. Squid fishing extends ing dispersalbehavior and regional oceanog- from May through December. Taiwanese and U.S. raphy. observersmonitored 8,521 km of squid drift nets in 1990 and 8,036 km in 1991.

METHODS AND STUDY AREAS 4. Taiwaneselarge-mesh fishery.--The Taiwanese large-mesh fishery operatesmostly in the area be- Scientific observerswere placed aboard high-seas tween 144*E-154øWand 30ø-440N.Fishing occursfrom fishingvessels from Japan,the Republicof Korea,and May through December. Time and area restrictions Taiwan (the Republicof China). Observersrecorded are similar to thosefor the Japanesefleet. Taiwanese the incidental catch of marine birds in drift nets and and U.S. observersmonitored 9,748 km of large-mesh retrieved voucherspecimens for assessingthe impact drift nets in 1990 and 11,876 km in 1991. of these fisherieson marine bird populations(Anon- 5. Republicof Koreasquid fishery.--The Korean squid ymous199 la, b, 1992a,b, Fitzgeraldet al. 1992,INPFC fishing grounds are between 1410E-150owand 30ø- 1991, 1992a, b, c). 46øN.The seasonextends from April'throughDecem- Five specimensof unidentifiedmurrelets were sal- ber. In April and May, fishing is mostlyin the south- vaged in 1990 and 1991 and sent to the Burke Mu- easternpart of this area and moves to the northeast seum,University of Washington,Seattle. Subsequent- in early summer(June-July). In Augustsome fishing ly, they were identified as JapaneseMurrelets. The shifts to western areasand, from Septemberthrough specimenswere preparedfor the museumcollection, December,all fishing occurswest of 170øE.Korean stomachcontents were preserved,and condition of and U.S. observers monitored 33,483 km of Korean the gonads,bursa, and body fat were recorded. squid drift nets in 1990 and 21,126 km in 1991. October1994] EndangeredJapanese Murrelets 955

Sea of Okhotsk

0 200 400 600 800 1000 ?

Kuril Islands '•i•BREEDING RANGE ß COLONYSITE • J•uu B¾C•TC• • tmuu BYC•TCll

Seo of Jopon

Ar*o Us*d To WESTERN DRIFT NET Bycotch FISHING AREA

• IZU Is.

Ryu k..v • Is. ß North Pacific Ocean Bonin Is. Doito Is. I' •2o oo •3o oo- 140 O0 150 oo •60 OO

Fig.1. DistributionofJapanese Murrelet (Synthliboramphus wumizusume) breeding colonies and bycatch in high-seasdrift nets (JAMU = JapaneseMurrelets; UNMU = unidentifiedmurrelets). Western drift-net fishing areaindicated by largeboxed area, and extends eastward to 151000' West longitude.

RESULTS sifted as JapaneseMurrelets at the time of collection,but not substantiatedby voucherspec- Incidentalcatch of murreletsin high-seasdrift- imens. Similarly for the remaining 16 birds at netfisheries.--Twenty-six Synthliboramphus mur- the time of collection, 6 were classified as An- reletswere recorded as bycatch during the high- cient Murrelets, 9 as unidentified murrelets, and seasdrift-net scientificobserver programs in 1 asa RhinocerosAuklet (Cerorhincamonocerata). 1990-1991 (Table 1). One was taken in the Tai- Basedon postcruisedebriefings and examina- wanesesquid fishery and the remainder were tion of photographs,all these specimenswere takenin the Koreansquid fishery. No murrelet later categorizedas unidentified Synthliboram- bycatchwas noted for the Japanesesquid fish- phusmurrelets by P. Gould (Table 1). No An- ery, probably becausethese vesselsdo not set cient Murrelets were verified by voucher spec- nets west of 160øE. Catch rates of marine birds imens as bycatchin squid or large-meshdrift- smallerthan albatrosses in large-meshdrift nets net fisheries. The specimens reported tenta- are considerablylower than ratesin squiddrift tively as Ancient Murrelets in 1990 (Anony- nets because of the differences in mesh size used mous1991b) were misidentificationsby us from (INPFC 1992b).This partly explainsthe absence photographs. That error is corrected here and of murrelet bycatch in the Japaneseand Tai- elsewhere(Anonymous 1992b). wanese large-meshdrift-net fisheries. Impact assessment.--Wecalculate that at least Five murreletswere salvagedand later iden- 98 JapaneseMurrelets were killed in 1990 (Ta- tified as JapaneseMurrelets. Most were adult ble 3). This lower estimate is conservative be- birdsundergoing molt from summerto winter causeit is basedonly on the five voucherspec- plumage (Table 2). Another five birds were clas- imens that were positively identified. Assum- 956 Pr•rr ANDGOULD [Auk,Vol. 111

TABLEI. Recordsof murreletsfrom North Pacifichigh-seas drift-net fisheriesin 1990-1991.Positions sum- marizedby 1ø latitude-longitude blocks (identified by positionof southwestcorner at startof net retrieval). Temperatures(lowest-highest) obtained at start of net set, end of net set, beginning of net retrieval, and end of net retrieval.

Surface Murrelets temper- Murrelet Latitude Kilometers per 1,000 ature species n Date and longitude of net km of net (*C) Japanese I August 1990 40*N, 144øE 246.5 4.1 19.7-20.3 Japanese I August 1990 40*N, 145øE 373.4 2.7 18.6-21.4 Unidentified 2 August 1990 40øN,146'E 126.0 15.8 21.9-22.8 Unidentified 2 August 1990 41*N, 146øE 629.9 3.2 19.3-22.8 Unidentified I August 1990 41*N, 152øE 16.0 64.1 19.5-19.7 Unidentified 1 August 1990 43øN, 154'E 223.6 4.5 15.8-17.9 Unidentified I August 1990 43øN,155'E 131.5 7.6 15.2-18.7 Japanese I September1990 41*N, 144'E 73.1 13.8 17.5-19.9 Unidentified I September1990 43'N, 154øE 357.3 2.8 15.2-16.6 Unidentified I September1990 44øN, 156'E 412.8 2.4 14.0-15.8 Unidentified I October 1990 39'N, 144øE 343.9 2.9 15.2-16.4 Japanese I October 1991 41*N, 143øE 77.5 12.9 14.5-15.0 Unidentified 2 November 1991 40øN, 142'E 36.9 54.2 13.2-13.6 Unidentified 1 November 1990 40*N, 145øE 262.5 3.8 14.9-17.4 Japanese I November 1990 40*N, 145'E 262.5 3.8 16.0-16.8 Unidentified 3 November 1990 40øN, 145øE 262.5 15.2 16.0-16.8 Unidentified I November 1990 41øN, 142'E 90.0 II.I 14.3-14.5 Unidentified 3 December 1990 38'N, 143øE 172.5 17.4 13.8-15.7 Unidentified I December 1990 39'N, 143øE 22.5 44.4 14.3-15.0 ing that all the unidentified birds caught were ulation and the bycatch(Table 2), then the an- JapaneseMurrelets, then as many as 417 may nual mortality of adults (24-250) in drift nets have been killed in 1990. This upper estimate would equal 1.0 to 10.4%of the breeding pop- is probablytoo high becauseit seemslikely that ulation. Annual mortality of adults in a rela- a few of the birds were Ancient Murrelets. Sim- tively undisturbedpopulation of Ancient Mur- ilarly, between 40 and 160 JapaneseMurrelets relets has been estimated at about 20% (Gaston may have been killed in 1991(Table 4). Summed- 1992).If JapaneseMurrelets have similar annual monthly-ratio estimatesyielded slightly lower mortality rates,then drift-net mortality may ac- total bycatchestimates (Tables 3 and 4). We pre- count for a significantproportion of total adult fer the unweighted annual estimates,however, mortality. becausethe low frequencyof bycatchrecords Distnt•utionof murreletbycatch.--The five Jap- weakensthe extrapolationfor individual months aneseMurrelets salvagedfrom the Korean fish- in which an increased observer effort would ery were from an area bounded by 40ø-42øN likely have yielded bycatch records(e.g. Oc- and 143ø-146øE. These were taken in the months tober 1990). of August through November and in relatively If we assumea total population of 4,000birds, warm waters. Sixteen of the unidentified mur- an annual bycatch of 40 to 417 birds, and an relets were taken in a slightly larger area (38ø- adult to immature ratio of 3:2 in both the pop- 45øN and 142ø-147øE)encompassing the Japa-

TABrE2. JapaneseMurrelet specimenscollected during 1990-1991high-seas drift-net observerprograms. Fat levels were "moderate" for all specimens.

Recoverylocation Date Sex Bodymolt Bursa Estimatedage 40.3'N, 145.8'E 13 September1990 M Heavy Fleshy Immature 40.7øN,144.2'E 18 August 1990 M Heavy None Adult 41.1*N, 144.3øE 6 September1990 M Heavy None Adult 40.3'N, 145.3'E 6 November 1990 F None Fleshy Immature 41.0*N, 143.9øE 2 October1991 M Heavy None Adult October1994] EndangeredJapanese Murrelets 957

T.•m,E3. Extrapolatedincidental catch of JapaneseMurrelets (JAMU) and unidentifiedSynthliboramphus murrelets(UNMU) in five large-scaledrift-net fisheries ain North Pacificeast of 147øEfrom March through December 1990.

Bycatch(n) No. 50-m net panels Observed Estimated Month Commerical Observer JAMU UNMU JAMU UNMU March 252 0 .... April 4,902 951 0 0 0 0 May 858 0 .... June 0 • .... July 111,023 0 .... August 1,173,844 74,304 2 4 32 63 September• 405,257c 25,752 1 0 16 0 October • 793,381 1,470 0 0 0 0 November 752,412a 23,262 1 5 32 162 December 101,529 6,729 0 4 0 60 Total 80 285 Grand total 3,226,423 131,517 4 13 98e 319 e

ßIncludes Japanese squid, Japanese large-mesh, Korean squid, Taiwanese large-mesh, and Taiwanesesquid fisheries. bAssumes 720 50-mpanels set per day in Taiwanesesquid fishery. cExcludes two Korean operations(2,203 50-m panels)in areasof 46'N, 142øEand 46øN, 145øE. a Excludesone Korean operation(1,200 50-m panels)in area of 46øN, 144•E. ' Extrapolatedby ratio estimatefrom total bycatchthroughout March to December.

neseMurrelet bycatcharea, and five were taken titled asJapanese Murrelets were from the same in an area(41ø-45øN and 152ø-157øE)well away area.Therefore, it seemslikely that mostor all from the coastof Japan(Table 1 and Fig. 1). The of the remaining 11 birds also were Japanese 16 murrelets recorded west of 147øE were taken Murrelets. Basedon limited knowledge of dis- from August through Decemberin surfacewa- tribution patterns,the other five specimens(i.e. ters ranging from 13.8ø to 22.8øC. thoserecovered south of the Kuril Islands;Fig. The five voucher specimensof Japanese 1) were more likely to have been Ancient Mur- Murrelet were randomlyselected (Table 1) from relets (Brazil 1991, Gaston 1992). Indeed, two of the total of 21 specimensrecorded near Japan; these birds were tentatively identified as An- moreover,the five specimenstentatively iden- cient Murrelets at the time of collection.

TABLE4. Extrapolated incidental catch of JapaneseMurrelets (JAMU) and unidentified Synthliboramphus murrelets(UNMU) in five large-scaledrift-net fisheriesa in North Pacificeast of 147øEfrom April through December 1991.

Bycatch(n) No. 50-m net panels Observed Estimated Month Commercial Observer JAMU UNMU JAMU UNMU April 0 0 .... May 0 0 .... June 0 0 .... July 585,174 8,140 0 0 0 0 August 1,511,849 20,920 0 0 0 0 September 1,258,364 39,374 0 0 0 0 October 2,175,384 58,592 1 1 37 37 November 1,977,301 23,262 0 2 0 77 December 554,500 23,575 0 0 0 0 Total 37 114 Grand total 8,062,572 201,759 1 3 40 • 120•

aIncludes Japanese squid, Japanese large-mesh, Korean squid, Taiwanese large-mesh, and Taiwanesesquid fisheries. Scientific observers were not placedon Japaneselarge-mesh vessels during this period. bExtrapolated by ratio estimatefrom total bycatchthroughout March to December. 958 Pm•T^ND GOULD [Auk,Vol. 111

DISCUSSION 1991).We concludethat JapaneseMurrelets dis- persenorthward in late summerto stagein per- Postbreedingdispersal.--The bycatch of Japa- sistentwarm-core rings southeastof Hokkaido, neseMurrelets in high-seasdrift nets appears where they forageon zooplanktonand pelagic to be a consequenceof their northwardpost- fishes that concentrate in the frontal zones breedingdispersal, which is poorlyunderstood around the rings. This conclusionis consistent (Kuroda1955, Brazil 1991).Japanese Murrelets with previous limited observationsof post- are warmwater seabirds, and breeding is re- breeding dispersalof JapaneseMurrelets (Ku- stricted to southern areas of Japan. The net- roda 1955, Brazil 1991), and parallelsa north- bycatch recordsreveal, however, that post- ward postbreedingdispersal of Xantus'Murre- breeding birds move northward to overwinter let (Synthliboramphushypoleucus) from colonies in a well-defined area southeast of Hokkaido off southernCalifornia to watersoff Washing- (Fig. 1). The region west of 147øE,between 38øN ton and British Columbia (Roberson1980). and 42øN,is part of the Perturbed Area (Kawai Incidentalcatch.--We believe that the mag- 1972).The oceanhere is influencedby the in- nitude of JapaneseMurrelet bycatchreported teraction of three major currents.The warm here underestimatesthe actualbycatch from all Tsugarucurrent flows from the sea of Japan the combined gill-net fisheriesthat operated through the strait between Hokkaido and Hon- near the coastof Japan in 1990-1991. For the shu, and carrieswarm water south along the fisheries we report on here, observer monitor- eastcoast of Honshu. The cold Oyashiocurrent ing eastof 147øEcovered only 4.1%of the com- carrieswater from the Kuril Islandssouth along mercial fishing effort in 1990 and 2.5% in 1991. the east coast of Hokkaido. The warm Kuroshio Althoughwe haveincorporated all monitoring current brings water from equatorialregions effortsin our analyses,the bycatchof Synthli- north along the east coastof Honshu where it boramphusmurrelets was reportedonly by U.S. meetsthe Oyashioand Tsugarucurrents. At this observers.It seems unlikely that some other confluence, currents are deflected eastward and records of "unidentified birds" were murrelets, large-scale(200+ km) anticycloniceddies reg- especiallyin 1990,when Koreanand Taiwanese ularly detach from the KuroshioCurrent (Ka- observerswere first trained and participatedin wai 1972).Warm-core rings form consistently the program.Observers also reported that birds in the area in which murrelets were taken in frequentlydropped out of netsbefore reaching nets (Vastano and Borders 1984, Kawai and Sai- the ship and, therefore, could not be accounted toh 1986,Saitoh et al. 1986,Kawasaki et al. 1991). for in estimatesof bycatchor speciescompo- Saitohet al. (1986)suggested that zooplank- sition. ton, which aggregate in the frontal zone be- Fishing with gill nets is common in coastal tween Kuroshio warm-core rings and cold waters of Japan. Besidesthe large-scaledrift- Oyashiowater (Aoki and Inagaki1992), provide net fisheries reported on here and elsewhere goodfeeding for Pacificsaury (Cololabis saira) in (Anonymous1991a, b, DeGangeand Day 1991, Septemberand Octoberas they migrate south- INPFC 1991, 1992a,b, c), these fisheries include ward from cold northern regions.In turn, this thousandsof small-boatgill-netters in coastal attractssaury predatorssuch as skipjacktuna and nearshoreareas (DeGange et al. 1993). In (Katsuwonuspelamis) and albacore (Thunnusal- addition to nearshore drift-net fisheries for alunga)that typically are found in warm waters salmon(Oncorhynchus spp.), marlin (Tetrapterus suchas those within the rings.Thus, it appears audax),and tuna (Thunnusthynnus), bottom-set that the distributions of cold- and warm-water gill nets are widely used to catch greenling speciesoverlap where they forage in the pro- (Pleurogrammusazonus), bastard halibut (Prali- ductive frontal zone at the edge of the warm- chthysolivaceus), sculpins (Cottidae), and sole core rings. The diet of JapaneseMurrelets is not (Limandaherzenstenei; Ogi and Shiomi 1991). known, but stablenitrogen ratios of muscletis- Bottom-setgill netsare equally efficientas drift suefrom three Japanese Murrelets salvaged from netsfor catchingdiving speciesof seabirds(Piatt nets (Table 2) suggestthat, like Ancient Mur- and Nettleship 1987, Ogi and Shiomi 1991). relets,their diet is a mixtureof euphausiidsand These coastalfisheries, which remain largely small pelagicfishes (N•5/N 14,œ = 13.2 + SD of unstudied,may have a muchgreater effect than 0.8 ppt; Hobson and Piatt unpubl. data, Hobson the large-scaledrift-net fisherieson coastalspe- October1994] EndangeredJapanese Murrelets 959

cies such as the JapaneseMurrelet (DeGange tremelyvulnerable to chronicand catastrophic and Day 1991,Ogi and Shiomi 1991, DeGange sourcesof mortality.The 38%reduction in catch et al. 1993). rate from 1990to 1991may have been an artifact Becausethe distributionsof Synthliboramphusof sampling,or may reflect between-yearvari- murrelets overlap in Japan,and the speciesare ation in prey distributionand murrelet forag- difficult to distinguish, it is also possiblethat ing effort (Piatt and Nettleship 1987). It also the bycatch of JapaneseMurrelets has been could reflect a steadily decreasingpopulation overlookedpreviously. DeGange and Day (I 99I) of JapaneseMurrelets. estimated that 307 Ancient Murrelets were killed On the positiveside, Japan, Taiwan, and the in the Japaneseland-based salmon drift-net Republic of Korea have agreed to the cessation fishery in 1977 and 116 in 1987, but some of of large-scaledrift-net fisheriesin international those birds could have been misidentified (A. watersof the North Pacificby the end of 1992 DeGange pers. comm.). Ogiet al. (1992) esti- (as outlined in United Nations Resolution 46/ mated that, in 1990, 367 Ancient Murrelets were 215). Although this action should diminish the caught in Japanesesalmon drift-net fisheries bycatch of JapaneseMurrelets, drift-net and operating east of 155øE,and 61 were taken in coastal gill-net fisheries within the 200-mile the nearshore (small-vessel)Japanese salmon (320-kin) Exclusive Economic Zone of Japan drift-net fishery. Given our experience, we likely will continue to kill JapaneseMurrelets speculate that some records of Ancient Mur- throughouttheir breeding and wintering range relets were based on misidentificationsof Jap- (e.g. Ogi and Shiomi 1991).Taken together,all aneseMurrelets, especiallythose taken west of available data point to an urgent need for re- about 150øElongitude. searchon the life history of JapaneseMurre- We recognizethe uncertaintyof our total by- lets--including completesurveys of breeding catch estimate basedas it is on a sample of 26 locations,effective protection of colonies and identified specimens.We know the area also is important wintering areas, predator-control used by wintering Ancient Murrelets (Brazil programs, and monitoring of coastal gill-net 199I) and, in fact,two specimenscollected there fisherieswithin the known range of the species. in November and Decemberwere tentatively identified asAncient Murrelets.In any case,we

have confidencein the overall extrapolationbe- ACKNOWLEDGMENTS cause murrelets were not taken in just a few bycatchepisodes, which can lead to overesti- The National Marine Fisheries Service and U.S. Fish mation of total bycatch in aggregated species and Wildlife Serviceprovided financial and logistical (Piatt and Nettleship 1987).Rather, bycatch oc- supportfor this project.Copies of all drift-net reports curred on at least 18 different occasions over are available from Drift-net Program Coordinator, five months, suggestinga low but persistent NMFS/AFSC, 7600 Sand Pt. Way NE, Seattle,Wash- rate of mortality. ington 98115, USA. We thank all of the people, es- peciallythe scientificobservers, who cooperatedin Gill-net mortality may be contributingto a the 1989-1991 High-Seas Drift-net Observer Pro- decline in the JapaneseMurrelet population. grams in Canada (Terry Gjernes, Skip McKinnel, Besidesthe mortality from gill nets, oil pollu- Howard McElderry), Japan(Hiroshi Hatanaka,Shi- tion that reportedlykills large numbersof An- geoHayase, Jun Ito, Hiroyuki Tanaka,Yoh Watanabe, cient Murrelets (e.g. Kazama1971) also may be Akihiko Yatasu, Kazuhiko Nagao, Shingo Ota, and affecting JapaneseMurrelets. Predation from Haruo Ogi), Taiwan (Sean-YaYeh and JamesSha), introduced predatorsappears to be another se- Korea (Doo Hae An, Yeong Gong, Seon Jae Hwang, rious mortality factorat colonies(Higuchi 1979, YeongSeung Kim, andJoo Suck Park) and the United Takeishi 1987, Oho 1993). As for other seabirds, States(Michael Dahlberg,Shannon Fitzgerald, Linda it is not known whether these unnatural mor- Jones,Greg Morgan, and Jerry Wetherall).Christo- pher Wood and Gary Shugartof the Burke Museum, tality factors are additive to the natural mor- Universityof Washington,identified and examined tality rate, or compensatedfor by density-de- the specimens.We appreciatecomments on this paper pendentpopulation regulation (Piatt et al. 1991). by Harry Carter,Anthony DeGange,Anthony Gas- Becausethe total populationnumbers less than ton, Robert Garrott, Yutaka Nakamura, Koji Oho, John a few thousand,and birds breed at only a few Ramsey,Mutsuyuki Ueta, Gus van Vliet, and Dan locations,Japanese Murrelet populationsare ex- Waldeck. 960 PiArr raNDGOULD [Auk, Vol. 111

LrlTa•tURE CITED status,ecology and conservationof marine birds of the North Pacific(K. Vermeer,K. T. Briggs,K. ANONYMOUS.1991a. Final report of 1990 observa- H. Morgan, and D. Siegel-Causey,Eds.). Spec. tions of Taiwanese high-seasdrift-net fisheries Publ. Can. Wildl. Serv., Ottawa. in the North Pacific Ocean. Joint report of The FITZGERALD,S., H. MCELDERRY,H. HATANAKA, Y. WA- Republic of China Council of Agriculture, the TANABE,J. S. PARK, Y. GONG, AND S. YEH. 1992. U.S. National Marine Fisheries Service, and the 1990-1991 North Pacific high-seasdrift-net sci- U.S. Fish and Wildlife Service. National Marine entific observerprograms. Pages 77-90 in Drift FisheriesService, Seattle, Washington. net fisheries of the North Pacific Ocean: Sym- ANONYMOUS.1991b. Final report of 1990 observa- posium on biology, distribution, and stock as- tionsof the Koreanhigh-seas squid drift-net fish- sessmentof speciescaught in the high-seasdrift- ery in the North Pacific Ocean. Joint report of net fisheries in the North Pacific Ocean (J. Ito, The Republic of Korea National Fisheries Re- W. Shaw, and R. L. Burgner, Eds.).International searchand DevelopmentAgency, the U.S. Na- North Pacific Fisheries Commission Bull. 53, tional Marine Fisheries Service, and the U.S. Fish vol. 1. and Wildlife Service. National Marine Fisheries GASTON,A.J. 1992. The Ancient Murrelet: A natural Service,Seattle, Washington. history in the Queen Charlotte Islands.T. & A. ANONYMOUS.1992a. Final report of 1991 observa- D. Poyser,London. tions of Taiwanese high-seasdrift-net fisheries HASEGAWA,H. 1984. Status and conservation of sea- in the North Pacific Ocean. Joint report of The birds in Japan,with specialattention to the Short- Republicof China Council of Agriculture,the tailedAlbatross. Pages 487-500 in Statusand con- U.S. National Marine Fisheries Service, and the servation of the world's seabirds (J.P. Croxall, P. U.S. Fish and Wildlife Service. National Marine G. H. Evans, and R. W. Schreiber, Eds.). Inter- FisheriesService, Seattle, Washington. national Council for Bird Preservation Tech. Publ. ANONYMOUS.1992b. Final report of 1991 observa- 2, Cambridge. tionsof the Koreanhigh-seas squid drift-net fish- HIGUCHI,Y. 1979. Breedingbiology and distribution ery in the North Pacific Ocean. Joint report of of JapaneseCrested Auklet. Kaiuyo to Seibutsu The Republic of Korea National Fisheries Re- (Aquatic Biology) 1(3):20-24 (Canadian Wildlife searchand DevelopmentAgency, the U.S. Na- Service translation, 1988). tional Marine Fisheries Service, and the U.S. Fish HOBSON,K.A. 1991. Use of stable carbon and nitro- and Wildlife Service. National Marine Fisheries gen isotope analysis in seabird dietary studies. Service,Seattle, Washington. Ph.D. thesis, Univ. Saskatchewan, Saskatoon. AOKI, I., AND T. INAGAKI. 1992. Acoustic observations INTERNATIONAL NORTH PACIFIC FISHERIESCOMMISSION. of fishschools and scatteringlayers in a Kuroshio 1991. Final report of 1990 observationsof the warm-core ring and its environs. Fish. Oceanogr. Japanesehigh-seas squid drift-net fishery in the 1:137-142. North PacificOcean. Joint report by the national BRAZIL,M.A. 1991. The birds of Japan.Smithsonian sectionsof Canada, Japanand the United States. Institution Press,Washington, D.C. National Marine Fisheries Service, Seattle, Wash- CANADIAN DEPARTMENT OF FISHERIES AND OCEANS. ington. 1991. Scientificreview of North Pacifichigh-seas INTERNATIONAL NORTH PACIFIC FISHERIESCOMMISSION. drift-net fisheries,Sidney, B.C.,June 11-14, 1991. 1992a. Symposiumon the biology,distribution, Report for presentation to the United Nations and stock assessmentof speciescaught in the pursuant to resolutions44/225 and 45/197. Can- high-seasdrift-net fisheriesin the North Pacific ada, Department of Fisheries and Oceans,Insti- Ocean,1991, Tokyo, Japan. National Marine Fish- tute of Ocean Sciences, P.O. Box 6000, 9860 West eries Service,Seattle, Washington. SaanichRoad, Sidney, British Columbia V8L 4B2, INTERNATIONAL NORTH PACIFIC FISHERIES COMMISSION. Canada. 1992b. Final report of observationsof the Japa- COCHRAN,W.G. 1963. Samplingtechniques, 2nd ed. nesehigh-seas large-mesh drift-net fisheryin the John Wiley, New York. North Pacific Ocean. Joint report by the national COLLAR,N.J., AND P. ANDREW. 1988. Birds to watch: sectionsof Canada,Japan and the United States. The ICBP world checklist of threatened birds. National Marine Fisheries Service, Seattle, Wash- International Council for Bird Preservation Tech. ington. PubL 8, Cambridge. INTERNATIONAL NORTH PACIFIC FISHERIES COMMISSION. DEGANGE,A. R., AND R. H. DAY. 1991. Mortality of 1992c. Final report of 1991 observationsof the seabirdsin the Japaneseland-based gillnet fish- Japanesehigh-seas squid drift-net fishery in the ery for salmon. Condor 93:251-258. North Pacific Ocean. Joint report by the national DEGANGE,A. R., R. H. DAY, J. A. TAKEKAWA,AND V. sectionsof Canada, Japanand the United States. M. MENDENHALL.1993. Lossesof seabirdsin gill National Marine Fisheries Service, Seattle, Wash- nets in the North Pacific.Pages 204-211 in The ington. October1994] EndangeredJapanese Murrelets 961

KAWAI,H. 1972. Hydrography of the Kuroshio ex- Ehime University (Univ. California Davis trans- tension. Pages235-352 in Kuroshio,physical as- lation, 1993). pects of the Japan Current (H. Stommel and K. PIATT, J. F., AND D. N. NETTLESHIP. 1987. Incidental Yoshida, Eds.). Univ. Washington Press,Seattle. catch of marine birds and mammalsin fishing KAWAI,H., AND $. $AITOH. 1986. Secondaryfronts, nets off Newfoundland, Canada. Mar. Poll. Bull. warm tongues,and warm streamersof the Kur- 18:344-349. shio extension system. Deep-Sea Res. 33:1487- P•ATT,J. F., H. R. CARTER,AND D. N. NETTLESHIP. 1991. 1507. Effectsof oil pollution on marine bird popula- KAWASAKI,Y., t. KOMO, M. KASHIWAI, AND K. OKUDA. tions. Pages125-141 in The effectsof oil on wild- 1991. Dynamic structureof the Oyashio. I. Net life (J. White, Ed.). Sheridan Press, Hanover, geostrophictransport balance and flow path of Pennsylvania. the Oyashiowater during summer in 1989.Bull. ROBERSON,D. 1980. Rare birds of the west coast of Hokkaido Natl. Fish. Res. Inst. 55:79-82. North America. Woodcock Publ., Pacific Grove, KAZAMA,T. 1971. Mass destruction of Synthlibor- California. amphusantiquus by oil pollution of JapanSea. Misc. SAlTOH, $., $. KOSAKA,AND J. hSAKA. 1986. Satellite Reps. Yamashina Inst. Ornithol. 6:389-398. infrared observations of Kuroshio warm-core KURODA,N. 1955. Observationson pelagicbirds of rings and their application to study of Pacific the Northwest Pacific. Condor 57:290-300. saury migration. Deep-SeaRes. 33:1601-1615. OGI, H., ANDK. $HIOMI. 1991. Diet of murrescaught SPRINGER,A. M., A. Y. KONDRATYEV,H. OGI, Y. V. incidentallyduring winter in northernJapan. Auk $HIBAEV,AND G. g. VAN VLIET. 1993. Status, ecol- 108:184-185. ogy, and conservationof Synthliboramphusmur- OGI, H., A. YATSU,H. HATANAKA, AND A. NITTA. 1992. relets and auklets. Pages 187-201 in The status, The mortality of seabirdsby drift-net fisheriesin ecology, and conservationof marine birds of the the North Pacific.Pages 499-518 in Drift net fish- North Pacific (K. Vermeer, K. T. Briggs, K. H. eries of the North PacificOcean: Symposium on Morgan, and D. Siegel-Causey,Eds.). Spec. Publ. biology, distribution, and stock assessmentof Can. Wildl. Serv., Ottawa. speciescaught in the high-seasdrift-net fisheries TAKEISHI,M. 1987. The massmortality of Japanese in the North Pacific Ocean (J. Ito, W. Shaw, and Murrelet Synthliboramphuswumizusume on the R. L. Burgner, Eds.). International North Pacific KoyashimaIslet in Fukuoka. Bull. Kitakyushu Fisheries Commission Bull. No. 53, vol. 3. Mus. Nat. Hist. 7:121-131. ONo,K.(Ed.). 1993. TheJapanese Murrelet: Its status VASTANO, g. C., AND $. E. BORDERS. 1984. Sea surface andconservation. Symposium abstracts of the 1993 motionover an anticycloniceddy on the Oyashio meetingof the JapanOrnithological Society at Front. Remote Sens. Environ. 16:87-90.