THE SUBSPECIES OF THE NORTHERN SAW-WHET OWL ON THE QUEEN CHARLOTTE ISLANDS: AN ISLAND ENDEMIC AND A NONBRING vISrrANT

SPENCER G. SEALY, Departmentof Zoology,University of Manitoba,Winnipeg, Manitoba R3T 2N2,

Two stronglydifferentiated subspecies of the Northern Saw-whet Owl (Aegoliusacadicus) are recognizedby the AmericanOrnithologists' Union (1957), A. a. brooksi, restrictedto the Queen Charlotte Islands(known also as ), , and nominate acadicus, widespread elsewhere in North America. I examinedspecimen and photographicrecords of the subspeciesof the Northern Saw-whet Owl from the Queen Charlotte Islandsand nearby BritishColumbia and Alaskato determinethe extentof the rangeof brooksi and determine the seasonaldistribution of nominate acadicus on the Queen Charlotte Islands.Records of acadicus on the Queen Charlotte Islandshave not been compiled,and the inter-islanddistribution of brooksihas not been ascertained.Godfrey (1986:328) stated that Northern Saw-whet Owls (presumablybrooksi) occur in "the northernQueen CharlotteIslands," but the entire archipelagowas shadedon his rangemap. Cowan (1989) stated that brooksiis foundonly on Grahamand Moresbyislands, the two largest islandsof the archipelago. The Queen CharlotteIslands comprise the most isolatedarchipelago off the Pacificcoast of Canada.The islandsare separatedfrom the mainlandof British Columbiaby about 80 km and from the nearestAlaska islands, Forresterand Dall, by about70 km and 50 km, respectively.The isolationof the Queen CharlotteIslands is reflectedin their beingthe home of several endemic subspecies,some stronglydifferentiated. Seven have been de- scribedso far: in addition to A. a. brooksi, ?icoides villosus picoideus (Osgood,1901)(Hairy Woodpecker), Cyanocitta stelleri carlottae Osgood, 1901 (Steller'sJay), Certhia familiaris stewar.tiWebster, 1986 (Brown Creeper),Catharus ustulatusphillipsi Ramos,1991 (Swainson'sThrush), Ixoreus naevius carlotte (Phillips, 1991) (Varied Thrush), and Pinicola enucleatorcarlottae Brooks,1922 (PineGrosbeak). Individuals of someof thesedisperse or migrateto a greateror lesserextent to the mainlandin the winter. Severalother distinctivesubspecies occur on the islandsas well as elsewhere in the humid Pacific Northwest.

TREATMENT OF RECORDS

I sought specimensthrough correspondencewith many museumsin North America and England (see Acknowledgments).I assembled217 specimens:90 were from the Queen CharlotteIslands (84 of brooski,6 of acadicus)and 127 were referencespecimens of nominateacadicus taken in Alaskaand elsewherein BritishColumbia. Documentation of breedingof nominateacadicus in nearbyAlaska is confirmedby specimensof a juvenile

Western Birds 29:21-28, 1998 21 THE SUBSPECIES OF THE NORTHERN SAW-WHET OWL

(MVZ 24409) and incubatingfemale (USNM 241692) taken on Forrester Island.I examinedall specimensexcept for 13 from Alaska(12 identifiedby D. D. Gibson,one by R. C. Banks).Three records from the QueenCharlotte Islands(two of brooksi that were capturedand releasedand one privately held specimenof acadicus)were confirmedfrom photographsdeposited in the Royal BritishColumbia Museum Photo-Duplicate File. The subspeciesare separablein the field (Figures1 and 2). but, unfortu- nately, field observershave not noted the subspecieswhen reporting sightings.probably because they are not aware that acadicusoccurs on the Queen CharlotteIslands in some seasons.I usedsight records,therefore, onlywhen they provideddocumentation of occurrenceon the smallislands of the archipelago.I assumedthese sightings were all of brooksi;they were madeprimarily by seabirdbiologists between April and Junewhen acadicus is not presenton the Queen CharlotteIslands (see later). Sight recordsof Northern Saw-whet Owls in winter on the Queen Charlotte Islands(e.g., Hamel 1983). a time when acadicus has been recorded there, were not included. A. a. brooksi differs strikinglyfrom acadicus in its darker coloration, particularlyon the breas!and facial disk (Figures1 and 2). Everywhere acadicusis white, brooksiis buff(Tawny Olive, Smithe 1975-1981). All of the brownsare muchdeeper and richer in brooksithan in acadicus.Should brooksi occur off the Queen Charlotte Islands, therefore, it should be recognizedreadily. No intergradesbetween the two subspecieshave been

Figure 1. Northern Saw-whetOwl (Aegoliusacadicus brooksi) on Langara Island, Queen CharlotteIslands, British Columbia, 5 April 1971. Photo by Spencer O. Sealy

22 THE SUBSPECIES OF THE NORTHERN SAW-WHET OWL

Figure2. NorthernSaw-whet Owls. From left to right,A. a. brooksi(CMN 14053), male,Masset, Queen Charlotte Islands, 5 July1919: A. a. brooksi(QCIM 8.1), male, 15 km south of , Queen Charlotte Islands, December 1978: A. a. acadicus (RBCM 15015), male, Peachland, British Columbia, 7 November 1952: A. a. acadicus (RBCM 13014), male, Metville, British Columbia, 6 November 1953. identified. Females are larger than males in both subspecies,and the subspeciesare aboutthe same size (Cannings1993).

STATUS ON THE QUEEN CHARLOTTE ISLANDS

Graham and MoresbyIslands I examined 84 specimensof brooksi and two photographs(RBCM Photos 1208. 1209) of live brooksi from the Queen Charlotte Islands.All but one of these recordsare from Graham and Moresbyislands. Most individualsin basicplumage were taken from Septemberto March, whereas those in juvenalplumage were collectedbetween late June and September (Figure3). The subspeciesbrooksi was named and describedby Fleming (1916) on the basisof three adultfemales and one juvenile,all takenin 1915 on .No specimenshave been taken off the Queen Charlotte Islands.All subsequentspecimens of this subspeciestaken since1915 were collectedon the two largest islands,Graham and Moresby. This is not surprisingbecause most of the human populationon the islandsis centered

23 THE SUBSPECIES OF THE NORTHERN SAW-WHET OWL

ß brooksi,adults (n = 60) [] brooksi,juveniles (n = 18)

• 8

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

Figure3. Monthlydistribution of datesof specimensand photographs of Northern Saw-whetOwls (A. a. brooksi)taken on the Queen Charlotte Islands.Three newly fledgedjuveniles taken by Patch(1922) from eachof two broodson 5 and 21 July 1919 are consideredas singlerecords. there,so specimens are morelikely to be salvagedthere, and the earlyefforts by ornithologiststo collectSaw-whet Owls were concentratedon thesetwo islands. Records from some of the smaller islands have been made since 1965 and are predominantlysightings made during censuses and studiesof nestingAncient Murrelets (Syn thliboramph us an tiquus). In addition,five specimens(2 females,3 males,all first-year)and one photograph(RBCM Photo 1207) of nominateacadicus have been obtained on Grahamor Moresbyisland. One of these,ironically, is the firstspecimen of a Northern Saw-whetOwl from the Queen CharlotteIslands, a male taken on 12 December1896 (not 19 December,see Deignan 1961) at Massetby Rev. J. H. Keen. On the basis of this single (type) specimen(USNM 168171), Osgood(1901) describeda new subspecies,scotaea. Ridgway (1914) and subsequentauthors treated scotaea as a synonymof nominate acadicus;it remainedfor Fleming(1916) to describethe endemicsubspe- ciescorrectly. The five other records of acadicus from the Queen Charlotte Islandsfall between12 Octoberand 11 January:male (AMNH 754293), 11 January 1915; unsexed(RBCM Photo 1207), January1965, Sandspit;male (UMZM 959), 10 November 1978, Sandspit;female (QCIM 60), December 1979; and female (QCIM B-190), 12 October 1984, Tlell. The period in which these specimenswere collectedis outsidethe breedingseason of both brooksiand acadicusby severalweeks, judging from the datesof recordsof juveniles(Figure 3). Therefore,mainland Saw-whet Owls occasionally reach the Queen CharlotteIslands during post-breeding movements or migration (Brooksand Swarth 1925). Specimenscollected on the BeringSea islands of St. Lawrence and St. Paul in October and November (Kesseland Gibson 1978) demonstrateeven more extremevagrancy of this subspecies. VagrantNorthern Saw-whetOwls are known to crosswide expansesof openocean elsewhere (e.g. Sladen 1966, DeSanteand Ainley 1980, Soucy

24 THE SUBSPECIES OF THE NORTHERN SAW-WHET OWL

1982, Amos 1991). Stable-isotopeanalysis of tissuesderived from 16 specimensof brooksiand two of acadicusfrom the QueenCharlotte Islands revealedthat the diet of acadicusshowed little marine protein compared withthat of brooksi.These results suggest that individuals of acadicusdo not remain long on the islands(Hobson and Sealy 1991). The origin of individualsof acadicus on the Queen Charlotte Islands should not be assumedto be from nearby sourcepopulations (see DeSante and Ainley 1980).

Small Islands Northern Saw-whetOwls, assumedto be brooksi, have been seenor mist- nettedon four smallislands and one islandgroup of the QueenCharlotte Islands,from LangaraIsland in the north to KunghitIsland in the south (Figure4). I could not locate one specimentaken on Langara Island (Campbell1969). These recordsadd the followingislands to the species' knownrange on the QueenCharlotte Islands: Limestone, Reef, Lyell, and Kunghit.Additional investigations likely will reveal Saw-whet Owls on other islandsof the archipelagoas well. LangaraIsland.--Noting that the callnotes of brooksiare quitedifferent from those of acadicus,Darcus (1930) statedthat in 1927 he saw brooksi by day in the densewoods on Langaraand Grahamislands and heardit at night. Campbell(1969) consideredSaw-whet Owls to be commonon the grassymeadows at Dadens, and his party collectedone specimenof brooksi. During the early morning of 5 April 1971, I mist-nettedand photographedan individualof brooksialong the upperbeach of Beal Cove (RBCM Photo 1209, Figure 1). Limestoneislands and Reef lsland.•aston (1992, pers.comm.) made numerousreferences to Saw-whet Owls taking Ancient Murrelet chicks duringtheir passagefrom their burrowsto the sea. On 31 May 1996, R. WayneCampbell and a groupof naturalistswatched one Saw-whet Owl pick up and fly awaywith a downymurrelet and observeda secondowl eatinga murrelet chick. Lyell Island.--Blood et al. (1979) heardSaw-whet Owls at two sites,mist- nettedone individual, and observedthem hunting Ancient Murrelet chicks at DodgePoint. Between20 April and 20 May 1982, Rodwayet al. (1988) heard singleSaw-whet Owls callingat night, one aroundtheir camp on a smallpeninsula west of DodgePoint, and one in the AncientMurrelet colony at Dodge Point. Kunghit Island.--On 27 November 1980, Mary Morris (pers. comm.) startleda sleepingSaw-whet Owl on the lower branchesof a sprucetree about1.5 m abovethe ground. A. a. brooskiis unrecordedas a vagranton the mainland.The breadth of prey used by this population(Sealy, unpubl. data), which includes considerablemarine-derived food taken duringthe nonbreedingseason (Hobsonand Sealy1991), possiblyprovides a stablefood supply for these owls and, hence, individualsare not forced to leave the islands(see Korpim•ki 1986). Individualsof brooksi, however,possibly move to the coastlinesof the large islands(see Hobson and Sealy 1991) during the nonbreedingseason as most specimenswere collectedor found dead

25 THE SUBSPECIES OF THE NORTHERN SAW-WHET OWL

• DALL I. FORRESTER I.

/ LANGARAI.

KUNGHIT I.

0 50km

SCALE

Figure4. TheQueen Charlotte Islands, showing Graham and Moresby islands and the foursmall islands and one island group on which Saw-whet Owls have been recorded. alongthe coastalhighway then, when suchmovements likely occur. Nesting is somewhatlater on the Queen Charlotte Islandsthan in south- easternAlaska and on the mainlandof BritishColumbia, with juveniles occurringinto Octoberrather than only to September.

SUMMARY

Thereare two field-identifiable subspecies of the NorthernSaw-whet Owl, Aegoliusacadicus acadicus, widespread on the mainland, and the strikingly

26 THE SUBSPECIES OF THE NORTHERN SAW-WHET OWL darker A. a. brooksi, endemic to the Queen Charlotte Islands, British Columbia.The endemicsubspecies occurs on somesmall islands as well as the two largeones, Graham and Moresby,but not off the QueenCharlotte Islands.It feeds heavily on intertidalinvertebrates and Ancient Murrelet chicks.Juveniles of brooksiwere taken into Octoberrather than only to September,as in acadicusin nearbyAlaska and mainlandBritish Columbia. Examinationof 90 spedmensand two photographsof liveSaw-whet Owls from the Queen CharlotteIslands revealed six recordsof vagrantacadicus from 12 Octoberto 11 January.

ACKNOWLEDGMENTS

I am indebtedto a greatmany people and institutions for assistancethroughout this study.Personnel of the Royal British ColumbiaMuseum, especially R. Wayne Campbell,supported me whileon a sabbaticalleave during which time the ideafor this studygerminated. Wayne has encouragedme from the outsetof my work on birds alongthe BritishColumbia coast, and his knowledge of the avifaunaof the areaand the dispositionof many museumspecimens jump-started my questfor informationon NorthernSaw-whet Owl distribution.Trishia and NicholasGessler, former curatorsof the Queen Charlotte Islands (Haida Gwaii) Museum, permitted me to examine specimensunder their care and provided logistical assistance during a trip to the Queen CharlotteIslands in 1986. Nathalie Macfarlane,curator of the Haida Gwaii Museum, permittedme to examine recenfiyacquired specimens under her care. Daniel D. Gibsondetermined the subspeciesof severalspecimens from Alaskaand RichardC. Banksconfirmed the subspeciesof a specimenfrom ForresterIsland. Al Breitkreutz, DonaldA. Blood,Richard J. Cannings,Harry R. Carter,Paul Chytyk, Ian McTaggart Cowan,John D. Fraser,Anthony J. Gaston,Michael J. Gill, CharlesJ. Guiguet,Anne E. Hetherington,Keith A. Hobson, Joe T. Marshall,Jr., Mary Morris, R. Wayne Nelson,Tom E. Reimchen,J. Dan Webster,and Mrs. GeorgeWillett provided general information,sight records, photographs, unpublished reports, or informationon the dispositionof specimens.My appreciationalso goes to the many other peoplewho respondedto my inquiries.Robert W. Dickermanand DavidP. Mindellcommented on an early draft of the manuscript.I thank Philip Unitt for his commentson the manuscript.Paula M. Grieefcompiled records and organizedother information. I am also gratefulto the curatorsof the followinginstitutions for loansof specimensor permissionto examinespecimens under their care: AmericanMuseum of Natural History (AMNH), New York; British Museum (Natural History), Tring; Canadian Museumof Nature(CMN), Ottawa;Charles R. ConnorMuseum, Washington State University,Pullman; Cowan VertebrateMuseum, University of British Columbia, Vancouver;Field Museumof Natural History, Chicago;Queen CharlotteIslands Museum(QCIM), Skidegate; Museum of VertebrateZoology, University of California (MVZ),Berkeley; Museum of ComparativeZoology, Harvard University, Cambridge; NationalMuseum of NaturalHistory (USNM), Washington;Royal British Columbia Museum(RBCM), Victoria; Royal Ontario Museum,Toronto; University of Alaska Museum,Fairbanks; University of ManitobaZoology Museum (UMZM), Winnipeg; and Universityof MichiganMuseum of Zoology,Ann Arbor. Financialsupport was providedby the NaturalSciences and EngineeringResearch Council of Canada.

LITERATURE CITED

AmericanOrnithologists' Union. 1957. Check-listof North AmericanBirds, 5th ed. Am. Ornithol. Union, Baltimore. Amos, E. J. R. 1991. A Guideto the Birdsof Bermuda.Corncrake, Warwick, Bermuda.

27 THE SUBSPECIES OF THE NORTHERN SAW-WHET OWL

Blood, D. A., Anweiler,G. G., and Cutter,M. J. 1979. Surveyof AncientMurrelet colonyat DodgePoint, Lyell Island, May 1979. Unpublishedreport to Rayonier Canada (B.C.) Ltd. (Order from Wilson OrnithologicalLibrary, Museumof Zoology,University of Michigan,Ann Arbor,M! 48109-1079). Brooks, A., and Swarth, H.S. 1925. A distributionallist of the birds of British Columbia. Pac. Coast Avifauna 17:1-158. Campbell,R. W. 1969. Springbird observations on LangaraIsland, British Columbia. BlueJay 27:155-159. Cannings,R. J. 1993. NorthernSaw-whet Owl (Aegoliusacadicus), in The Birdsof North America, No. 42 (A. Poole and E Gill, eds.).Acad. Nat. Sci., Philadelphia. Cowan, I. M. 1989. Birdsand mammalson the Queen CharlotteIslands, in The Outer Shores(G. G. E. Scudderand N. Gessler,eds.), pp. 175-186. Queen Charlotte IslandsMuseum Press, Skidegate, B.C. Darcus,S. J. 1930. Notes on birdson the northern part of the Queen Charlotte Islands in 1927. Can. Field-Nat. 44:45-49. Deignan, H. G. 1961. Type specimensof birds in the United States National Museum. U. S. Natl. Mus. Bull. 221. DeSante,D. E, andAinley, D. G. 1980. The avifaunaof the SouthFarallon Islands, California. Studies Avian Biol. 4:1-104. Fleming,J. H. 1916. The Saw-whetOwl of the Queen CharlotteIslands. Auk 33:420-423. Gaston,A. J. 1992. The AncientMurrelet: A NaturalHistory in the QueenCharlotte Islands.T. and A.D. Poyser,London. Godfrey,W. E. 1986. The Birdsof Canada.Natl. Mus. Nat. Sci., Ottawa. Hamel, P. (compiler).1983. The eighty-thirdAudubon Christmas bird count, Masset, British Columbia. Am. Birds 37:441-442. Hobson,K. A., and Sealy,S. G. 1991. Marine proteincontributions to the diet of Northern Saw-whet Owls on the Queen Charlotte Islands:A stable-isotope approach. Auk 108:437-440. Kessel,B., and Gibson,D. D. 1978. Statusand distributionof Alaskabirds. Studies Avian Biol. 1:1-100. Korpim•ki, E. 1986. Gradientsin populationfluctuations of Tengmalm'sOwl Aegoliusfunereus in Europe.Oecologia 69:195-201. Osgood,W. H. 1901. Natural historyof the Queen Charlotte Islands,British Columbia. N. Am. Fauna 21:1-50. Patch, C. A. 1922. A biologicalreconnaisance on Graham Islandof the Queen CharlotteGroup. Can. Field-Nat.36:133-136. Ridgway,R. 1914. The birdsof Northand Middle America. U.S. Natl. Mus.Bull. 50, part 6. Rodway,M.S., Lemon, M. J. F., and Kaiser,G. W. 1988. BritishColumbia seabird colonyinventory: Report 1--East coastMoresby Island. Can. WildlifeServ., Pacificand YukonRegion Tech. Rep. Ser. 50. Sladen,W. J. L. 1966. Additionsto the avifaunaof the PribilofIslands, Alaska, includingfive speciesnew to North America.Auk 83:130-135. Smithe, F.B. 1975-1981. Naturalist'sColor Guide. Am. Mus. Nat. Hist., New York. Soucy,L.J., Jr. 1982. Saw-whetOwls at sea. New JerseyAudubon 8:20.

Accepted 19 August 1997

28