WESTEKN BIRDS

Volume 28, Number 1, 1997

THE TIMBERLINE SPARROW, SPIZELLA (BREWERI) TAVERNERI, IN ALASKA, WITH NOTES ON BREEDING HABITAT AND VOCALIZATIONS

TERRYJ. DOYLE, TetlinNational Wildlife Refuge, P.O. Box 779, Tok,Alaska 99780

The TimberlineSparrow was originally described as a fullspecies, $pizella taverneri,by Swarth and Brooks (1925), whodistinguished it from Brewer's Sparrow,Spizella breweri,by itsdarker color, heavier streaking, and larger averagesize but smaller,slenderer, darker bill. Subsequently,ornithologists haveclassified the TimberlineSparrow as a subspeciesof Brewer's,though Sibleyand Monroe(1990) rankedit as a full speciesagain, on the basisof unpublisheddifferences in vocalizationsand ecology. The publishedbreeding range of taverneri extendsfrom southwestern Yukon Territory, northwesternand central British Columbia, and west- central Alberta to southeastern British Columbia and southwestern Alberta (AOU 1957), where it approachesthe north end of the rangeof nominate breweri.The winterrange is poorly known, with mostU.S. specimenstaken during migration, in Arizona (Monson and Phillips 1981), New Mexico (Grinnell1932), Texas (Oberholser1974), and Washington(Jewett et al. 1953). Only one winterspecimen has beenreported, from California(Rea 1967), suggestinga winter rangesouth of the U.S. On 22 June1992, JeffreyJ. Boutonof the NationalPark Servicefound a singingmale TimberlineSparrow in the NutzotinMountains of Alaska, 100 km northwestof the species'closest known location in Yukon(Clarke 1945, PamelaH. Sinclairin litt.).Over the followingfour years I studiedthe occurrenceand statusof this speciesin Alaska.I characterizedits habitat, recordedsongs, and collectedspecimens (University of AlaskaMuseum, Fairbanks)to document its occurrenceand to help shed light on the relationshipbetween taverneri and nominatebreweri.

STUDY AREA

I searcheda 15-km 2 area near Gold Hill from 1993 to 1995 and an 8-km 2 areain the Upper Cheslinadrainage from 1994 to 1996 (Figure1). In addition,

Western Birds 28:1-12, 1997 1 THE TIMBERLINE SPARROW IN ALASKA an independentgroup visited Gold Hill in 1996. Gold Hill is in the Nutzotin Mountains10 km northeastof Chisana,Alaska (62 ø 06' N, 141ø 54' W). The UpperCheslina drainage is in the Mentasta Mountains 80 kmnorthwest of Gold Hill (62ø 39' N, 142ø 44' W). The Mentastaand Nutzotin mountains are part of theAlaska Range, which extends east to theWhite River in Yukon(Orth 1971); the KluaneRange continues east of the WhiteRiver. Habitatswithin the GoldHill and Upper Cheslinaareas are similar(Figure 2) and may be describedwith Kessers(1979) nomenclature.Unvegetated screeslopes of ruggedmountain peaks, elevation 2300-2600 m, descendto rollingfoothills of alpineand subalpine vegetation. Mats of dwarfshrubs, <0.4 m high, in the alpinezone giveway to subalpinehabitat at approximately 1500 m in the GoldHill area and 1300 m in the Upper Cheslina.The alpine zone descendsmuch lower on north-facingslopes. Subalpine vegetation on south-facingslopes is composedof patchyopen and closedthickets of mediumwillows (Salix spp.) and alders (Alnus crispa), 1.2-2.4 m high,with DwarfBirch (Betula nana) a commoncomponent of the understory.Thickets of low shrubs,0.4-1.1 m high,dominate the subalpinezone of north-facing slopes.Dwarf spruce (Picea spp.) woodlands, <5.0 m highand <20% canopy cover,occur at elevationsbelow the subalpinezone, along with closed thickets of tall willowand alder,2.5-4.9 m high.

RESULTS

Status in Alaska TimberlineSparrows were observedin the GoldHill area eachyear from 1992 to 1996. Observationsincluded an adult feedinga recenfiyfledged juvenilein 1993 (Figure3), birdsgathering food in 1996, and up to six singingmales within 8 km2 in 1994 (Table1). A singingmale was in the Upper Cheslinadrainage in 1994 but not in 1995 or 1996. Elsewhere,a singingmale at sealevel in southeastAlaska at Hyder,3-5 June1996, isthe onlyAlaska record away from the breedingground. All but one of the Alaska observationshave been during the breeding season,from mid-Juneto mid-July.A reconnaissanceof the Gold Hill area 3-5 June 1995 failedto disclosea singleTimberline Sparrow even though all of the regularbreeding birds of the areawere present (Upland Sandpiper, Bartrarnia 1ongicauda;Say's Phoebe, $ayornis saya; Horned Lark, Eremophila alpestris;Townsend's Solitaire, Myadestes townsendi; Ameri- can Pipit,A nthus rubescens; Orange-crowned Warbler, Vermivora cela ta; Wilson's Warbler, Wilsonia pusilla; American Tree Sparrow, $pizella arborea;Savannah Sparrow, Passerculus sandwichensis; White-crowned Sparrow,Zonotrichia leucophrys;Smith's Longspur, Calcarius pictus). A later trip, 25-26 June 1995, revealed three Timberline Sparrows in locationswhere they had been presentin prior years.The earliestthat Swarth(1936) foundthe sparrowsin the Atlin regionof BritishColumbia was 29 May, and recentarrivals near Whitehorse,Yukon, have been from 29 to 31 May (Cameron D. Eckert, Pamela H. Sinclair in litt.). The observationat Hyder in earlyJune suggests that the birdsare stillmigrating at that time. It isnot surprisingthat the TimberlineSparrow arrives relatively THE TIMBERLINE SPARROW IN ALASKA

I i 143 ø 142 ø

Teain

A A G T'ah A A Upper

A

• GoldHill

62 ø

:.: , , I A 0 1020 30 mi : AlaskaIYukon 143 ø 142 ø 141 ø 140 ø I I i i Figure1. Locationsof TimberlineSparrow records in Alaskaand adjacent Yukon Territoryare denotedby solidtriangles. latein the GoldHill area,given that the knownbreeding range in Alaskais farther north than that in Yukon and British Columbia. Despitethe sparrows'late arrivalon the breedingground, their singing seemedto diminishby early July. The latestthat they were found singing was THE TIMBERLINE SPARROW IN ALASKA

Figure2. BreedingTimberline Sparrows in Alaskawere foundon south-facingslopes at the transitionof the alpineand shrubbysubalpine zones. The habitatin whichthey were foundalmost exclusively was thickets of shrubbywillows 1.0-1.2 m high with a Dwarf Birchunderstory 0.4-0.5 m high. This habitatdid not occuron north-facing slopeswithin the area studied.as seenon the right sideof this canyon. Photo by Terry J. Doyle

4 July. Three were seen but not heard singingin the Gold Hill area from 2 to 17 July, and the singingmale in the Upper Cheslinadrainage was not heardthree weekslater, on 15 or 16 July. Departure dates are unknown, but Swarth (1936) found Timberline Sparrowson the breedinggrounds near Atlin, BritishColumbia, until early September.

Habitat I found breeding Timberline Sparrows in Alaska to be restrictedto a narrow band at the transition between the subalpine and alpine zones, vegetatedwith low to mediumthickets of willows(Figure 2). Averagecanopy height rangedfrom 1.0 to 1.2 m, canopycover from <25% to 50%. Salix glauca wasthe dominantwillow: S. planifolia and S. brachycarpawere less common. Dwarf Birch was the primary componentof the next tallestshrub layer, with Vaccinium uliginosum and Potentilia fruticosa as subdomi- nants. Average canopy height of this layer was 0.4-0.5 m, and canopy coverwas <25%-75%. Groundcover included Rhododendron !apponicum, Salix reticulata, Poa sp., Carex sp., Mertensia sp., and trace amountsof ferns, mosses, and lichens. THE TIMBERLINE SPARROW IN ALASKA

Figure3. This adultTimberline Sparrow was feedinga recentlyfledged juvenile at Gold Hill on 17 July 1993, providingthe firstdocumented record for Alaska. Photo by David W. Sonneborn

All TimberlineSparrows were found on rather steep (30-40 ø) south-to southeast-facingslopes at 1460 to 1525 m in the GoldHill areaand at 1325 m in the Upper Cheslina.Below this elevation shrubs are tallerand canopies more closed;above. the vegetationquickly changes to matsof dwarfshrubs. Timber- lineSparrows were foundin contiguousbands of willowthickets rather than in smallpatches of willows.Some individualTimberline Sparrows appeared to prefer the micro relief in ravines.which has also been observedin Yukon (CameronD. Eckertpers. comm.). No TimberlineSparrows were found on north-facingslopes, which lack the medium-heightthickets of willows. The habitatin which Timberline Sparrowsoccurred was quite predictable. I flew over the Gold Hill area in 1994 in a fixed-wingaircraft and identified potentialTimberline Sparrow habitaton the basisof the prior year's field work.All potentialTimberline Sparrow habitat, within the 15 km2 surveyed on the ground, had singingmales in 1994. No Timberline Sparrowswere foundoutside the identifiedhabitat. In 19951 identifiedpotential Timberline Sparrowhabitat from a fixed-wingaircraft over a 4600-km2 area in the Mentastaand Nutzotin mountainsfrom northwestof the Upper Cheslina drainage to the Yukon border. Approximately 61 linear kilometers of potentialhabitat were identified,the majoritybetween Gold Hill and the Canadian border. Several areas more extensivethan those investigatedin thisstudy were identified.as were severalsmall patches <1 km long. Nearly all identifiedhabitat was on steep south-facingslopes, consistent with that found at Gold Hill and the Upper Cheslina. THE TIMBERLINE SPARROW IN ALASKA

Table 1 Dates, locations,and observersof TimberlineSparrow sightings in Alaska.

Date Location Observations Oberservers

22 June1992 GoldHill One singingmale JeffreyJ. Bouton 17 July1993 GoldHill Adultfeeding a T.J. Doyle, recentlyfledged PhilipD. Martin, juvenile DavidW. Sonneborn 18-19 June1994 GoldHill Six singingmales T.J. Doyle 24 June 1994a Upper Cheslina One Singingmale T.J. Doyle, Ryan C. Means 25-26 June GoldHill Threesinging males T.J. Doyle, 1995b GaryH. Rosenberg 3-5 June1996 Hyder One singingmale T.J. Doyle 2-4 July 1996 GoldHill Fourbirds, two CariE., KevinH., and singingmales and RobertE. Gill, two gathering ColleenM. Handel, food LisaJ. Oakley, TheodoreG. Tobish, MargaretM. Vacca aNo birdswere foundin the Upper Cheslina15-16 July 1994, 27-29 June 1995, or 26 June 1996. No birdswere found in the Gold Hill Area 3-5 June 1995.

The migrantbird in Hyderwas found in scatteredshrubs and Sitka Spruce (Piceasitchensis), up to 2 m tall, along a gravelcauseway adjacent to a saltgrassmeadow.

Plumageand MorphometricComparison Specimenscollected (UAM 6669, 6670, and 6939) for this studyagree with taverneriand differ from nominate breweri in beingslightly larger with a more slenderbill that is muchdarker (Figure 4). Measurementsof the three Alaskaspecimens of taverneri (all males)were weight 11.8-13.7 g, wing chord64-66.5 mm, and tail length56-64.6 mm, in comparisonto weight 9.8-11.5 g, wing chord60.6-64.5 mm, and tail length58.5-67.0 mm for the 10 male $. b. breweri reported by Swarth and Brooks (1925). The plumageof taverneri is darker overallwith wide dark brown dorsalstreaks on a dark gray ground color, in contrastto lighter and narrower brown streakson a sandybrown ground color in breweri (Figure5); in many plumage characteristicssuch as its coarserdorsal streaking,fairly well definedgrayish nape, suggestionof a medianline throughthe crown,and fairlydistinct superciliary, ta verneri is intermediate between breweri and the Clay-coloredSparrow ($. pallida).

Song The songsof breweri and taverneri are superficiallysimilar. Both are complex songs composedof a variety of trills at different pitches and tempos. Closer examination, however, reveals differences(Figure 6). I THE TIMBERLINE SPARROW IN ALASKA

Figure4. Lateralcomparison of two $. b. brewer/fromUtah and Wyoming and three $. (•.) taverneri from Alaska (left to right, UAM 2300. 2904• 6939, 6669, and 6670), showingtaverneri as larger with a moreslender, darker bill and a fairlydistinct superciliary. Photo by Barry J. McWayne comparedfour songsof an individualtaverneri (Library of NaturalSounds 73301, Cornell University)to three songsof breweri from Wyoming, California,and Oregon(Peterson 1975). I identified10 differentparts in the three songsof breweri. They tend to be more buzzy, and parts of them (Figure6, lb) resemblethe buzzynotes of the Clay-coloredSparrow's song from Peterson(1975). A seriesof descendingsweet notes is the only part repeated (Figure 6, 2) and has no similar counterpartin taverneri. The THE TIMBERLINE SPARROW IN ALASKA

Figure5. Dorsalcomparison of two $. b. brewerifrom Utah andWyoming and three S. (b.) taverneri from Alaska (left to right, UAM 2300, 2904, 6939, 6669, and 6670), showingtaverneri as largerwith coarserdorsal streaking, a fairlywell defined grayisharea betweenstreaking of the headand back,and a suggestionof a median line throughthe crown. Photo by Barry J. McWayne

TimberlineSparrow's song contains seven distinct segments. many of which were repeatedduring the same or subsequentbouts. It tendsto have more trillssimilar to thoseof a junco(Figure 6, la), with someas high-pitchedas a waxwing'scall. The trillsof breweriare of a wider frequencyrange (Figure THE TIMBERLINE SPARROW IN ALASKA

6, lb) than those of taverneri, accountingfor their buzziness.The most similarpart of the two forms'songs is a two-tonedtrill that variesin pitchand tempo; that part of the breweri song (Figure6, 3b) fallswithin the range performed by taverneri (Figure6, 3a). Playbackof the taverneri song in the Gold Hill area produceda more aggressiveresponse of Timberline Sparrowsthan did playbackof the breweri song.

DISCUSSION

The TimberlineSparrow appears to be an annualbreeding member of the Alaskaavifauna. It is locallydistributed and uncommonin suitablehabitat. How longit hasoccurred in Alaskais unknown.Given that thereare records from Yukon within 45 km of Alaska, potentialhabitat in adjacentAlaska, and the remotenessof the area, it is most likely that TimberlineSparrows were simplyoverlooked until recently. More extensivebreeding habitat probablyexists away from the areas investigatedin this study. I suspectthat the Timberline Sparrow is more widespreadin Alaskawith breedingpopulations occurring along the Nutzotin Mountainsfrom Gold Hill to Yukon.The breedingrange probably extends northwestof the Gold Hill area into the Mentasta Mountains, perhaps northwestof the Upper Cheslinadrainage, where somelimited habitat has beenidentified. Though the habitatmay be more extensivethan is currently known, it appears restrictedto the edge between alpine and subalpine habitats,on primarilysouth-facing slopes. There are no immediatethreats to this habitat, most of which is locatedon National Park Serviceand National WildlifeRefuge lands. The habitatin whichtaverneri occurs in Alaskadiffers subtly from that in other partsof its range, but dramaticallyfrom that of nominatebreweri. In Alaskathe breedinghabitat of taverneri is fartherabove timberline than in BritishColumbia where the bird was first discovered(Swarth 1926) and in someareas in Yukonwhere it is knowncurrently. At the southernend of its range, taverneri is associatedprimarily with Dwarf Birch (Bruce W. McGillivrayin litt.) ratherthan willows,and occursat higherelevations (Jon C. Barlowpers. comm.).At thesesouthern latitudes, Dwarf Birchat higher elevations is similar in structure to the willows associated with taverneri in Alaska (Jon C. Barlow pers. comm.). The montane habitat in which taverneriis foundis quitedifferent from the sagebrushsteppes (Artemisia spp.)with which breweri is associated(Wiens and Rotenberry1981). Though superficiallysimilar, upon close examination the songs of taverneriand breweriseem quite different and shouldbe easilydistinguish- ablein the field.Only a few individualswere comparedduring this investiga- tion. A more thoroughstudy of the geographicvariation in songmight shed light on the taxonomicstatus of taverneri. In addition,a geneticanalysis of taverneriand nominatebreweri may be helpfulin definingtheir taxonomic status.Preliminary genetic work is currentlybeing conducted(Robert M. Zink in litt.). THE TIMBERLINE SPARROW IN ALASKA

o

(D

ZH• ZHH

10 THE TIMBERLINE SPARROW IN ALASKA

SUMMARY

One of the most recent additionsto Alaska'sbreeding avifauna is the TimberlineSparrow, Spiz½lla(breweri) taverncri, first discoveredin 1992 in the Nutzotin Mountains, 100 km northwestof the nearestknown location in the YukonTerritow. Follow-up surveys revealed the TimberlineSparrow to be uncommonand local.Up to six individuals,including an adultfeeding a fledgling,were within 15 km•' each year from 1992 to 1996. An additional singingmale was found 80 km northwestof the NutzotinMountain location, and a singingmale was foundin migrationat Hyder.Timberline Sparrows were limitedto a narrowband of open thicketsof low to mediumshrubs at the transitionof the alpineand subalpinezones, quite different from the arid lowland sagebrushhabitat with which breedingbrcweri is most often associated.The habitat,although fairly limited, appears to be quitepredict- able. The songof taverncri differsfrom that of brcweri in beinghigher pitched,having less buzzy trills, and lackingdescending sweet notes.

ACKNOWLEDGMENTS

I acknowledgethe NationalPark Serviceand U. S. Fish and WildlifeService for providingthe fundingnecessary to conductthis work; the Wrangell-St.Elias National Park and Preserve,in particularKarin D. Kozie, were especiallysupportive. I was assistedin the fieldby PhilipD. Martin,Ryan C. Means,Kathleen M. O'Reilly-Doyle, GaryH. Rosenberg,and David W. Sonneborn.David W. Sonnebornprovided the first definitivedocumentation of the TimberlineSparrow in Alaskawith his excellent photographs;we wouldnot havemade the 1993 trip withouthis dogged persistence. WilliamR. Maynardfirst suggested recording songs, and Gary H. Rosenbergmade the best recordings.Leonard J. Peyton gave generouslyof his time in producingthe sonograms,and BetsyA. Sturmprepared the figures.Cameron D. Eckert,Wendy A. Nixon, and PamelaH. Sinclairprovided valuable information on Yukonobservations. Karin D. Kozie,Amadeo M. Rea, Gary H. Rosenberg,Theodore G. Tobish,Gus B. van Vliet, andespecially Philip Unitt providedmany useful comments on earlierdrafts of thispaper. I especiallythank Daniel D. Gibson,of the Universityof AlaskaMuseum, for hisencouragement during the preparationof thismanuscript and for hisassistance at all stages.

LITERATURE CITED

AmericanOrnithologists' Union. 1957. Check-listof North AmericanBirds, 5th ed. Am. Ornithol. Union, Baltimore, MD. Clarke, C. H. D. 1945. Some bird recordsfrom YukonTerritory. Can. Field-Nat. 59:65. Grinnell,J. 1932. An United Statesrecord of the TimberlineSparrow. Condor 34:231-232. Jewett, S. G., Taylor, W. P., Shaw, W. T., and Aldrich, J. W. 1953. Birds of WashingtonState. Univ. of Wash. Press,Seattle. Kessel,B. 1979. Avian habitatclassification for Alaska.Murrelet 60:86-94. Monson,G., and Phillips,A. R. 1981. AnnotatedChecklist of the Birdsof Arizona, 2nd ed. Univ. of Ariz. Press,Tucson. Oberholser,H. C. 1974. The Bird Life of Texas.Univ. of TexasPress, Austin.

11 THE TIMBERLINE SPARROW IN ALASKA

Orth, D. J. 1971. Dictionaryof AlaskaPlace Names. U.S. Geol.Survey Prof. Paper 567. Govt. PrintingOffice, Washington, D.C. Peterson,R. T. 1975. A FieldGuide to WesternBird Songs.Lab. Ornithol.,Cornell Univ., Ithaca, NY. Rea, A.M. 1967. Some bird recordsfrom San Diego County,California. Condor 69:316-318. Sibley,C. G., andMonroe, B. L., Jr. 1990. Distributionand Taxonomy of Birdsof the World. Yale Univ. Press,New Haven, CT. Swarth, H. S. 1926. Report on a collectionof birdsand mammalsfrom the Aftin region,northern British Columbia. Univ. Calif. Publ.Zool. 30:51-162. Swarth,H. S. 1936. A listof the birdsof the Aftin region,British Columbia. Proc. Calif. Acad. Sci. 23:35-58. Swarth,H. S., and Brooks,A. 1925. The TimberlineSparrow, a new speciesfrom northwestern Canada. Condor 27:67-69. Wiens, J. A., and Rotenberry,J. T. 1981. Habitat associationsand community structureof birdsin shrubsteppeenvironments. Ecol. Monogr. 51:2 I-4 I.

Accepted29 October 1996

12