WESTERN BIRDS Volume 25, Number 3, 1994 STATUS OF THE FORK-TAILED FLYCATCHER (TYRANNUS SAVANA) IN THE UNITED STATES AND CANADA GUY McCASKIE, 954 GroveStreet, Imperial Beach, California 91932 MICHAEL A. PAI-FEN, Departmentof Biology,University of California,Riverside, California92521 andTierra MadreConsultants, 1159 IowaAve., SuiteE, Riverside, California 92507 The Fork-tailedFlycatcher (Tyrannus savana) breeds in Middleand South Americaand occursin North Americaas a vagrant.Monroe and Barron (1980) summarized40 reportsof Fork-tailedFlycatchers in the United Statesand Canadaand concluded that "NorthAmerican vagrants... have been presumedto be vagrantsthat 'overshot'their normal 'wintering' groundsin northernSouth America." While they notedsome movements that did not fit thispattern, they offeredno hypothesesto explainthem. Furthermore,whereas they noted "that most North American reports pertain to the nominatesubspecies" (Bond 1940), they did not discuss ageingof thesevagrants. Our review of American Birds and other sources reveals 63 additional reportsthrough 1992, bringingthe total numberreported in the United Statesand Canadato over 100 (Appendix).The authenticityof someof these reportshas been questioned,leaving 94 recordswe considervalid. Aside from two birds seen togethernear Florida City, Florida, on 15 September1984 (Am. Birds 39:44, Powell1986, B. H. Andersonin litt.), all acceptedrecords involve single individuals, although two were within 15 milesof each other on Cape Cod, Massachusetts,during the fall of 1980 (Am. Birds 35:159). Our analysisof theserecords suggests that the proposalby Monroeand Barron(1980) only partlyexplains the movementof this speciesinto the United States and Canada, as there is a marked contrast between the occurrencepattern along the Gulf of Mexico and that along the Atlantic coast,in the Midwest,and in New England.In attemptingto explainthese movementpatterns, we also examined ageing criteria and subspecific identification,both of whichshed light on the issue. Western Birds 25:113-I27, I994 113 FORK-TAILED FLYCATCHER STATUS PATTERN OF OCCURRENCE Sixty-sixof the Fork-tailedFlycatchers found in the United Statesand Canada were along the Atlantic coastfrom Florida north to Nova Scotia (Figure 1), with more than 85% of these north of the Carolinas.Eleven more were in the interior,west to Dorion, Ontario (Goodwin 1978), Duluth, Minnesota(Svingen 1991), Columbia County, Wisconsin(Freese 1979), and Henderson,Kentucky (Audubon 1834, Mengel 1965). In addition,14 havebeen recorded along the Gulf of Mexicofrom the southwesterntip of Floridato southernTexas, with five of these in Florida and sevenin Texas. In the western United States and Canada, one was seen at Drumheller, Alberta,on 1 June 1988 (Wedgwood1989), one was photographednear Picabo,Idaho, between25 Augustand 7 September1991 (Trost1991), and a third was photographedat Bridgehavennear the mouth of the Russian River on the north-central coast of California between 4 and 8 September 1992 (Am. Birds 47:146, 1993). One reportedlyshot near Santa Monica, California, in the "latter part of the summer" of 1883 (Toppan1884) cannotbe verified,because the specimenwas destroyed by fire in 1896 (Grinnell 1915). Fifty-sixof the 77 Fork-tailedFlycatchers (the singleVermont record cannot be assignedto season)along the Atlantic coast and inland to Ontario, Minnesota,Wisconsin, and Kentuckyhave appearedin fall be- tween 23 August(Meritt 1970) and 1 December(Knight 1910), with the majorityduring September and October(Figure 2). This is the time when the nominatesubspecies should be migratingfrom winteringgrounds in northernSouth America to breedinggrounds in southernSouth America. We proposethat these birds flew a latitudinalmirror-image misoriented migrationroute northwest instead of southwestduring the australspring. TwentyFork-tailed Flycatchers have been found along the Atlanticcoast and inlandto Minnesotaand Wisconsinin springand summerbetween 2 May (Am. Birds 44:398) and 17 July(Kingbird 37:206). We proposethat thesebirds overshot their wintergrounds when migratingnorthward during the australfall migration.One foundnear Rustic,Virginia, on 3 June 1988 remainedthrough 1 August(Williams 1988), duringwhich time it molted. Late June and July birds,such as thosephotographed in easternDouglas County,Wisconsin, on 24 June 1988 (Johnson1989) and on Cape Breton Island,Nova Scotia, on 25 June 1984 (Tufts1986), and one videotapedat LoxahatcheeNational Wildlife Refuge, Florida, on 13 and 14 July 1992 (J. L. Baker in litt., B. H. Andersonin litt.), may havearrived earlier than when found and simplyhave been wanderingin searchof suitable"wintering" localities.One reportedlycollected at Lake Ridge, Michigan,in July 1879 also fits this pattern of occurrence,but the specimenis no longer extant, and the recordhas been questioned(Barrows 1912). Althoughthe samplesize is small(n = 3), the seasonaloccurrence of Fork-tailedFlycatchers in the westernUnited Statesand Canada appears similar to that for the Atlantic coast, with two recordsfor fall and one in early June. The species'seasonal occurrence along the Gulfof Mexico,however, is quitedifferent, with eightindividuals found in spring/summerbut onlytwo 114 FORK-TAILED FLYCATCHER STATUS o o •, • . e ß O• o 115 FORK-TAILED FLYCATCHER STATUS in fall(Figure 2). Springdates of occurrenceextend from 20 April (Bremser and Duncan 1992) to 15 July (Am. Birds 28:895), whereasthe fall dates are 8 October (Oberholser1974) and 17 October (Am. Birds 31:168). In addition, four individualshave been found in Texas in winter between 4 December(Am. Birds 44:293) and 4 February(James 1963). A reportfor March was not acceptedby the Texas Bird RecordsCommittee (G. W. Lasleyin litt.). SUBSPECIES Four subspeciesof the Fork-tailedFlycatcher have been described (Zimmer 1937) and are currentlyrecognized (Traylor 1979). The highly migratorynominate race breeds as far southas the southerntip of Argen- ,41 12 • 10 3 8 z 6 Half-month of initial occurrence. AtlanticCoast, New England, • Gulf Coast and Midwest Figure2. Seasonaldistribution of datedrecords of the Fork-tailedFlycatcher, by half- month, along the Arianticcoast, in New England,and in the Midwestcompared to seasonaloccurrence along the Gulf of Mexico. 116 FORK-TAILED FLYCATCHER STATUS tina and the Falkland Islands,withdrawing to extreme northern South Americanduring the australwinter. The partiallymigratory northern race monachus breedsfrom southernMexico southto north-centralBrazil, with somebirds withdrawing southward from the northernpart of this range in our winter. The other two racesare reported as resident,sanctaemartae alongthe Caribbeancoast of Colombiaand extremenorthwestern Venezu- ela and circumdatusin the Amazon basinof Brazil(Traylor 1979). Although Zimmer stated that monachusis "readily separated"from nominatesavana by its "muchpaler gray" back, this distinction is oftennot apparent without birds in hand (Hilty and Brown 1986). Back color is greatlyaffected by wear, however,with worn monachusbeing as dull as freshsavana (J. V. Remsenin litt.).Zimmer also stated monachus typically showsa "more whitish collar," but Patten's examination of 84 skinsat the NaturalHistory Museum of Los AngelesCounty (LACM; n = 74) and the WesternFoundation of VertebrateZoology (WFVZ; n = 10) showeda great dealof variabilityin thisfeature (seasonal wear wastaken into account),with one nominatebird having a nearlycomplete white collar.The visibilityof a collar dependson specimenpreparation, however, making this feature difficultto judge on skins. In addition, Zimmer stated that monachus generallylacks the grayflanks that contrast with the whitebreast on savana. The most usefulfeature, however,is the shape of the two outermost primarieson adultmale savana,which are more deeplyemarginated than those on adult male monachus. Also, the 8th primary on adult male monachusshows almost no emargination,whereas this feather is asdeeply notched as the outermost two on savana. Both sanctaemartae and circumdatus are similar to monachus in back coloration. Adult males of sanctaemartaehave almostno emarginationon the outer four primaries, whereasadult males of circumdatushave the deepestemargination on the outer two primariesof any of the four races.Because emargination is next to impossibleto ascertainin the field, identificationto subspecieson this basiswithout the birdin hand or clearin-flight photographs (e.g., Am. Birds 47:146), is questionable. Most Fork-tailedFlycatchers reaching the United Statesand Canadaare believedto be of the highlymigratory nominate subspecies. Specimens (see Appendix for abbreviations)from Maine (life mount; BMS X03.163), Pennsylvania(FMNH 304056), and New Jersey(specimen lost) have been identifiedas savana (Bond 1940). Birds from Michigan(UMMZ 206495) and Texas (LSUMZ 25116) have also been identified as the nominate subspecies.The bird in Virginiaduring the summerof 1988 underwenta completemolt duringits stay(Am. Birds 43:79), stronglysuggesting that it alsowas of the nominaterace, because June and July is the time of the year this subspeciesmolts. Monroeand Barron(1980) suggestedthat an adultmale (ANSP 35423) labeled"New Jersey,"with no date or specificlocality, to be the specimen taken at Trenton,New Jersey,in fall 1900 (Babson1901). This specimen wasidentified as sanctaemartaeby JohnT. Zimmer(Bond 1940). Monroe and Barron (1980) questionedthis identification,but apparentlydid not examinethe skin.Mark B. Robbinsexamined this specimenat our request. Robbinsstated (in litt.)that its