RAPTOR MIGRATION IN AUTUMN THROUGH THE UPPER TANANA RIVER VALLEY,

CAROL L. McINTYRE, National Park Service, 201 1st. Ave., Fairbanks, Alaska 99701 and Departmentof Fisheriesand Wildlife, Oregon State University, Corvallis, Oregon 97331 ROBERT E. (SKIP) AMBROSE, U.S. Fish and Wildlife Service, 101 12th Ave., Fairbanks, Alask•a99701

Raptor migrationin Alaskais not well documented(Swem 1982, Mindell and Mindell1984). The upperTanana River valley, in east-centralAlaska, is a well-knownmigration route for manyspecies of Alaskabirds (Kesse11984, Cooper et al. 1991). Ambrose'scasual observations of migratingraptors from 1976 to 1986 suggestedthat the area also may be a major raptor- migrationcorridor in autumn. We initiatedautumn raptor-migrationcounts and bandingin the upper Tanana River valley in September 1987. We developedthe project in responseto an environmental-impactstatement prepared for the proposed constructionof a receivingsite for an "over-the-horizonbackscatter" radar unitnear Tok, Alaska. Information gathered for the statementsuggested that the radarantennas associated with the unitposed a potentialcollision hazard to migratingbirds, especially large ones such as geese,swans, cranes, and raptors(Cooper et al. 1991). The mainobjective of ourraptor counts was to recordthe numbersand speciesof raptorsmigrating through the upper Tanana River valley in autumn. This paper summarizesresults of raptor countsfrom 1987 to 1994 and discussesthe potentialof raptor-migration countsas tools for monitoringpopulation trends of certainspecies of raptors in Alaska.

STUDY SITE

The countsite, Dry Lake (63ø 40' N, 141 ø 30' W), liesapproximately 74 km southeastof Delta Junctionand just north of milepost1376 alongthe AlaskaHighway. Dry Lake is an intermittentlyflooded 300-acre lake bed, locatedat one of the narrowestpoints of the upperTanana River valley. At an elevationof 350 m, it is within0.5 km of a majorslough of the Tanana River.The AlaskaRange, crestingin this area at 2000 to 3000 m, rises approximately8 km southof Dry Lake and may be a barrierfor migrating birdsflying directly south. The -TananaUplands, a seriesof broad hillsthat crest at approximately750 m elevation,begin 2 km to the northof Dry Lake. Dry Lake is surroundedby forestsof Black Spruce (Picea mariana), White Spruce (P. glauca), and Balsam Poplar (Populus balsamifera).A 500-acre agriculturalarea lies2 km to the west. Temperaturesduring the countperiods ranged from lowsof -18øC in late Septemberand Octoberto highsof 31øC in earlySeptember. Prevailing windsduring the countperiod in all yearswere from the southand southeast. Southand southeastwinds _•30 mph were not uncommonduring the count periods.

Western Birds 30:33-38, 1999 33 RAPTOR MIGRATION IN AUTUMN THROUGH TANANA RIVER VALLEY

METHODS

Our countsextended from lateAugust until mid-October (Table 1). We did not make countsin 1989 or 1990. Observersequipped with binoculars(10 x 40) and a variable-powerspotting scope made observationsfrom or near a trappingblind on the easternside of Dry Lake. The locationprovided unobstructedviews to the southwest,west, and northwest.Counts usually beganapproximately 30 minutesafter sunrise and ended at sunset.We didnot countduring heavy rain or snow.Two observersmade observationsin 1987 and 1988. In all otheryears, one observermade the count>60% of the time. We identifiedthe speciesof migratingraptors as they crossedan imagi- nary north/southplane extending from the observer.Raptors that couldnot be identifiedwere categorizedas large,small, or unidentified.We collected weather data hourly (temperature,surface wind directionand speed, and cloud cover) and recorded all data on standardizedforms.

RESULTS

Raptor migrationin the upper Tanana Rivervalley was well underway by late Augustand earlySeptember each year of the count.The peak flightsof mostspecies, except the Rough-leggedHawk and GoldenEagle, occurred in mid-Septemberin all years. We countedraptors an averageof 9.6 hoursper countday for a total of 1926 hoursover the sixyears. We counteda total of 7407 migratingraptors of 13 speciesfrom 1987 to 1994 (Table2). Poolingdata from all years,we counted 3.85 raptors per count hour and 1234 raptors per count year (Table 2). Migrantraptors passed by Dry Lake over a broadfront approximately8 km wide and were movingsoutheast. Buteos generally migrated along a seriesof disconnectedhilltops and mountainsto the southand north of Dry Lake, usuallyflying at least300 m abovethe ground.Other speciesflew at variousaltitudes. Many Northern Harriersand AmericanKestrels passed by at heightsunder 50 m and often stoppedto hunt in the lake bed. The Northern Harrier was the mostcommonly observed species annually, composing35.3% of the total numberof raptorsobserved in all years.The Rough-leggedHawk wasthe secondcommonest species, composing 18.3%

Table 1 Raptor Count Periods, Dry Lake, Alaska, 1987-1994

Year CountPeriod CountDays CountHours

1987 7 Sep-18 Oct 38 301 1988 27 Aug-15 Oct 43 460 1991 27 Aug-30 Sep 33 346 1992 27 Aug-30 Sep 32 286 1993 28 Aug-2 Oct 33 332 1994 8 Sep-1 Oct 22 201

34 RAPTOR MIGRATION IN AUTUMN THROUGH TANANA RIVER VALLEY

Table 2 Summary of Raptors Counted During Autumn Migration, Dry Lake, Alaska, 1987-1994

Species 1987 1988 1991 1992 1993 1994 Total

Osprey (Pandion haliaetus) 2 13 3 1 21 13 53 Northern Harrier (Circuscyaneus) 329 802 524 366 435 138 2594 Sharp-shinnedHawk (Accipiterstriatus) 113 114 110 103 185 76 701 Northern Goshawk (Accipitergentilis) 39 38 18 5 12 5 117 Red-tailed Hawk (Buteojamaicensis) 81 153 58 36 213 39 580 Rough-leggedHawk (Buteolagopus) 315 457 79 160 265 277 1553 BaldEagle (Haliaeetus leucocephalus) 28 55 17 35 19 8 162 GoldenEagle (Aquilachrysaetos) 31 21 8 20 10 12 102 American Kestrel (Falcosparverius) 30 82 104 93 53 25 387 Merlin (Falco columbarius) 79 79 74 52 59 13 356 PeregrineFalcon (Falcoperegrinus) 11 21 5 15 3 0 55 Gyrfalcon (Falco rusticolus) 2 0 1 2 0 0 5 Short-eared Owl (Asioflammeus) 8 26 7 2 1 2 46 Largeraptor 12 234 4 23 26 18 317 Smallraptor 15 123 6 34 36 16 230 Unidentifiedraptor 12 73 11 21 38 8 163 Total 1107 2291 1029 0968 1376 0636 7407

of the totalnumber of raptorsseen over all years.We countedmore Rough- leggedHawks in 1987 and 1988 when the countperiod extended into mid- October.In severalyears, we observedsmall movements of Short-earedOwls. Owlswere usuallyseen at duskor dawn and often huntedin the lake bed. We did not identify10% of the totalnumber of migrantsobserved. Most of the unidentifiedraptors were seenat distances>4 km and couldnot be identifiedto species.From their size and behavior we believethat most of the largeraptors were Red-tailedor Rough-leggedHawks and mostof the small raptorswere Sharp-shinnedHawks or American Kestrels.

DISCUSSION

Resultsfrom our countssuggest that many raptorsmigrate through the upper Tanana River valleyin autumn.The numberof migratingraptors

35 RAPTOR MIGRATION IN AUTUMN THROUGH TANANA RIVER VALLEY observedat Dry Lakeare far lessthan those recorded annually at well-known raptormigration sites farther south such as Hawk Mountain,Pennsylvania, Cape May, New Jersey,the GoshuteMountains, Nevada, and GoldenGate NationalRecreation Area, California.We expectfewer migrants to passby Dry Lake than by concentrationpoints farther south becausemigrants observedat Dry Lake are departingtheir northwesternterminus of range;at lowerlatitudes such passage includes not onlydeparture of localbirds but the movementsof thosedrawn from enormous areas farther north (including birds thatmay pass Dry Lake).Interestingly, our annual counts of NorthernHarriers are similarto or higherthan counts of thisspecies made at countsites at lower latitudesin NorthAmerica. Northern Harriers often hunted in Dry Lake;the largenumber of harriersobserved may be a resultof thisbehavior. We probablyunderestimated numbers of raptorsmigrating by Dry Lake each autumn becausecounts were made simultaneouslywith trapping. Insteadof constantlysearching the skiesfor migrants,our efforts were often focusedon trapping,banding, and measuring birds. Additionally, our counts were biasedtoward speciesusing the lake bed for huntingand toward specieswe weretrying to capture.Because of thiswe probablymissed many raptorsmigrating along ridges to the northand southof Dry Lake. Finally, our countperiods were not equalevery year and were not long enoughto includeearly and late migrants.Despite these shortcomings,our counts indicatethat raptors concentratein this area in autumn and that this is an importantmigration corridor for raptorsleaving Alaska. Migrantsobserved at Dry Lake mostlikely come from breedingareas in interior,western, and northern Alaska. This area encompassesmillions of acresof borealforest, taiga, and alpineand arctictundra. Population trends of many raptorsin these areas are unknownbecause of the logistical difficultyand expenseassociated with workingin remoteareas of Alaska. We suggestthat migrationcounts, in associationwith increasedsampling on breedinggrounds, may be a usefultechnique for monitoringpopulation trendsof certain speciesof raptorsin Alaska (Tituset al. 1989). Dry Lake is a logical site to conductraptor countsin eastern Alaska becauseit is locatedat one of the narrowestpoints in the upper Tanana Rivervalley where migrating raptors tend to concentratewithin a relatively narrowarea. Equally important, Dry Lakeis easily accessible by vehiclefrom the , keepinglogistical costs down. Becausedata from migrationcounts maybe useful for monitoringpopulation trends of raptors, we need to continueto searchfor new raptor-concentrationsites in Alaska and other northern areas. These general areas includethe upper valley, the upper Tanana River valley, southeastAlaska (including coastalareas), and westernCanada (Yukon and BritishColumbia). The only other raptor migrationcount in Alaskais in the MatanuskaValley in south- central Alaska (B. Dittrick and T. Swem pers. comm.). Counts at the Matanuskasite are alsomade simultaneously with trapping.The numbersof raptorspassing through the MatanuskaValley are similarto thoseat Dry Lake, but the speciescomposition of the flightis slightlydifferent (T. Swem unpubl.data). Efforts are alsounderway to find monitoringsites for raptors in Yukon, and raptor monitoringis well establishedin Alberta, Canada (Sherrington1998).

36 RAPTOR MIGRATION IN AUTUMN THROUGH TANANA RIVER VALLEY

We plan to continueour raptor countsat Dry Lake, and we encourage othersto establishnew raptor countsin Alaskaand westernCanada. In the futurewe plan to standardizeannual count periods to providedata consistent fromyear to yearand to runthe countseparately from the trapping.We also planto experimentwith samplingdesigns to provideestimates of numbers of raptors migrating through the upper Tanana River valley in autumn. Finally,we needto assessthe usefulnessof countdata from Dry Lake for monitoringpopulation trends of selectedspecies of raptors (suchas the Northern Harrier) breedingin Alaska.

SUMMARY

We countedmigrating raptors in the upperTanana River valley in eastern central Alaska from 1987 to 1994. Over the study period we counted migratingraptors on 201 days, averaging33.5 count days per year. On averagewe counted1234 migratingraptors each autumn.The Northern Harrierand Rough-leggedHawk werethe mostnumerous species observed duringthe counts.We recordedpeak numbersof mostspecies, except the Rough-leggedHawk and GoldenEagle, from earlyto mid-Septembereach years.Our resultsindicate that the upper TananaRiver valley is an important migrationcorridor for raptorsin autumn.We suggestthat migrationcounts have high potentialas a techniquefor monitoringof populationtrends of certain speciesof raptors in Alaska. However, more such sitesand data collectedon breedinggrounds are needed to test the potential of using raptor countsto monitortrends in Alaskabreeding populations.

ACKNOWLEDGMENTS

The U.S. Fish and WildlifeService and NationalPark Service,Alaska Regional Office, providedfunds for this project. The Hawk MigrationAssociation of North Americaprovided additional support in 1992. We are extremelygrateful to Michelle Ambrose,Terry J. Doyle,and KathyO'Reilly-Doyle for helpingus countraptors, for continuedencouragement, and logisticalsupport, and the many peoplewho helped us with this project,especially Peter J. Bente, Maria Berger, Nan Eagleson,Todd Eskelin,Janey Fadely, Gary Koy, Debbie Nigro, Pat Owen, Craig Perham, Bob Ritchie,and CharlieWhittaker. Mike Brittenhelped with developmentof thisproject and helpedcollect data in 1987 and 1988. Ted Swem providedtechnical assistance and stimulatingdiscussion of raptor migration.Mike Britten, Katy Duffy, Daniel D. Gibson,Patrick Jodice, and Philip Unitt provideduseful comments on earlierversions of the manuscript.

LITERATURE CITED

Cooper,B. A., Ritchie,R. J., Anderson,B. A., andByrne. L. C. 1991. Alaskaover- the-horizonbackscatter radar system: Synthesis of the avianresearch program, 1987-1990. Final report preparedfor Arctic EnvironmentalInformation and Data Center,University of Alaska--Fairbanks,and Department of the Air Force. AlaskaBiological Research, P.O. Box 80410, Fairbanks,AK 99708. Kessel,B. 1984. Migration of SandhillCranes (Grus canadensis)in east-central Alaska,with routesthrough Alaska and Canada.Can. Field-Nat.98:279-292.

37 RAPTOR MIGRATION IN AUTUMN THROUGH TANANA RIVER VALLEY

Mindell, D. P., and Mindell, M. H. 1984. Raptor migrationin northwesternCanada and easternAlaska, spring 1982. Raptor Res. 18:10-15. Sherrington,P. 1998. CanadianRockies and Plains. Hawk MigrationStudies 23:16- 20. Swem,T. 1982. SitkagiBeach Raptor Migration Study, Spring 1982, in Proceedings of the Hawk MigrationConference IV (M. Harwood, ed.), pp. 13-26. Hawk MigrationAssoc. N. Am., Medford,MA. Titus, K., Fuller,M. R., and Ruos,J. L. 1989. Considerationsfor monitoringraptor populationtrends based on countsof migrants,in Raptorsin the ModernWorld, Proceedingof the Third World Conferenceon Birdsof Prey and Owls (B.-U. Meyburgand R. D. Chancellor,eds.), pp. 19-27. WorldWorking Group of Birds of Prey• Berlin.

Accepted 10 August 1998

38