j. Field Ornithol.

, 61(2):192-200

REGIONAL SIZE DIFFERENCES AMONG FALL-MIGRANT

ACCIPITERS IN NORTH AMERICA

JEFF P. SMITH•
Departmentof Biology
Utah StateUniversity
Logan,Utah 84322-5305USA
STEPHEN W. HOFFMAN
WesternFoundationfor Raptor Conservation,
Inc.
P.O. Box 35 706
Albuquerque,
New Mexico87176-5706USA

JAMESA. GESSAMAN
Departmentof Biologyand EcologyCenter
Utah State University
Logan, Utah 84322-5305 USA

Abstract.--Measurementsof fall-migrant accipitersfrom four regionsin the United States

werecompared.CoopeftsHawks (Accipitercooperii)from GoshuteMountains,Nevadaand
Marin Headlands,California were significantlysmallerand had longerwingsand tails in
proportionto their weightthan thosefrom Cedar Grove,Wisconsinand Cape May Point,
New Jersey.Northern Goshawks(A. gentilis)from the Goshutesweighedsignificantlyless,
but hadlongerwingsandtailsthan thosefrom CedarGrove.The samewastrue for Goshute
Sharp-shinnedHawks (A. striatus)in comparisonwith thosefrom Cedar Groveand Marin
Headlands, but differenceswere less pronounced.These data demonstratethe need for
regionalidentificationcriteria. Among the four samples,Goshutemigrantsaveragedthe
lowestflight-surfaceloading(i.e., proportionatelylow weightand longwingsand tail) and
inland migrantsaveragedlower flight-surfaceloadingthan coastalmigrants.Low flight-
surfaceloadingmay be adaptivefor inland migrants,whichaveragelongermigrationsand
occupyhabitatswith more open vegetation,and particularly for Goshutemigrantswhich
may dependmore on exploitingthermal updrafts.

DIFERENCIAS REGIONALES EN EL TAMAlqlO DE MIGRATORIOS OTOlqlALES

(ACCIPITRINAE) EN NORTE AMERICA
Sinopsis.--Secompararon lostamafiosde partesde la anatomlade individuosmigratorios
otofialesde Accipiteresde cuatro regionesde los EstadosUnidos. Individuosde Accipiter
cooperiide lasmontafiasGoshute,Nevaday Marin Headlands,Californiaresultaronsignifi-
cativamentemils pequefiosy con alas y rabosmils largosen proporci0na su peso,que
individuosde Cedar Grove, Wisconsiny Cape May Point, New Jersey. Por su parte
individuosde A. gentilisoriginadosde Goshutemostrarontener alas y rabosmils largos,
pero con peso significativamente menor que aves de Cedar Grove. Lo mismo aplica a
individuosde A. striatusde Goshuteen comparasi6ncon otrosde Cedar Grove y Marin
Headlands,aunquela diferenciafue menosmarcada.Los datosde estetrabajodemuestran
la necesidadde criteriospara la identificaci6nde estasaves a nivel regional. Entre las
muestras,losmigrantesdeGoshutepromediaron la menorcargasuperficie-vuelo(Ej. menos
pesoproporcionala alasy raboslargos).Los migrantesde tierra adentropromediaronuna
menorcargasuperficie-vuelo que losmigrantescostaheros.Cargaspequefiasde superficie-
vuelo podrlan ser una adaptaci6npara migrantesde tierra adentro,los cualespromedian
vuelosmigratoriosmils largosy ocupanhabitatsconvegetaci0n mils abierta.Esto aplica

• Currentaddress:Departmentof WildlifeandRangeSciences,
UniversityofFlorida,118Newins-
Ziegler Hall, Gainesville,Florida 32611 USA.

192
Vol.61,No.2 Accipiter
SizeDifferences [193

particularmentea los migrantesde Goshuteque muy bien pudierandependerde explotar

adecuadamente corrientes termales de aire.

To illustrate morphologicaldifferencesamong Sharp-shinnedHawks

(Accipiterstriatus),Cooper's Hawks (A. cooperii),and Northern Gos-
hawks(A. gentilis)we comparedmeasurements
of fall migrantstrapped
at four sitesin the United States:Marin Headlands,California (37ø38'N,
122ø49'W); Goshute Mountains, Nevada (40ø53'N, 114ø28'W); Cedar
Grove,Wisconsin(43ø34'N,87ø49'W);andCapeMay Point,New Jersey
(39ø09'N,74ø46'W).The comparisons
demonstrated
theneedfor regional
speciesidentificationand sexingcriteria for Cooper'sHawks and North-
ern Goshawks;disclosedinterspecificdifferencesin the patternsof size
variationamongNorth Americanaccipiters;and revealedmorphological
differencesbetween eastern and western, and between coastal and inland
migrants.In this paper we discussthe possibleadaptivesignificanceof
thesetrendsand the implicationsfor in-flight identificationof species.
Regionalspeciesidentificationand sexingcriteria are discussedin Hoff-
man et al. (1990).
METHODS

Hoffman et al. (1990) summarized measurementsfrom the Goshute

Mountainsand Mueller et al. (1976, 1979a, 1981a) provideddata from
Cedar Grove. Data from Marin Headlandsand Cape May Point were
obtainedin raw form for analysis.Many individualscollectedthe Marin
Headlandsmeasurements fromAugustthroughDecember1986and 1987.
William S. Clark and his associates collectedthe Cape May Point mea-
surementsduring Octoberfrom 1970-1980. In all caseshawkswere aged
by plumagecriteria describedin Mueller et al. (1976, 1979a, 1981a).
Adult or AHY (after-hatching-year)includedhawksin their secondyear
or older, exceptthat we usedmeansfor the Adult II category(hawksin
at least their third year) of Mueller et al. (1976) to representadult
goshawks from Cedar Grove.Immatureor HY (hatching-year)included
hawks lessthan one year old.
We sexedGoshute migrants by the criteria presentedin Hoffman et
al. (1990). Cooper'sHawks from Cape May Point were sexedusing
criteria now representedin the Canadian Wildlife Service and United
StatesFish and Wildlife Serviceraptor age-sexkey (CWS and USFWS
1980) and in Mueller et al. (1976, 1979a, 1981a). Marin Headlands
migrants also were sexed using the CWS and USFWS age-sex key;
however,accordingto the criteria we developedfrom measurementsof
Goshutemigrants, nine of the Marin Headlands Cooper'sHawks had
been incorrectlysexedas males.Golden Gate Raptor Observatoryper-
sonneldiscussedthis problem in their biannual newsletterThe Pacific
Raptor Report (Winter 1986-1987). For this analysis,we sexedthe
Marin Headlands migrantsby the Goshutecriteria.
We obtainedonly wing chordand weightmeasurements for immature
Cooper'sHawks from Cape May Point. Fewer than 10 adult sharpshins,
of each sex, and 10 adult male Cooper'sHawks were capturedand
194] y. P. Smithetal. J.Field
Ornithol.
Spring 1990

TABLE 1. Regional comparisonsof means with sample sizes (in parentheses)for mea-
surementsfrom fall-migrant Sharp-shinned Hawks. •

Marin Goshute Cedar

Measure Age-sex Headlandsb Mountains Grovec
Weight (g) HY-M 102 (17) NS 96 (714 98 (489)
AHY-M -- 102 (175 * 103 (435)
HY-F 167 (72) 160 (427 166 (522)
AHY-F -- 171 (298 174 (487)
Wing chord (mm) HY-M 170 (17) NS 171 (901 169 (493)
AHY-M -- 174 (264 171 (437)
HY-F 201 (74) 204 (860 200 (544)
AHY-F -- 206 (524 203 (489)
Standardtail length(mm) HY-M 135 (17) NS 136 (475 134 (494)
AHY-M -- 134 (94) 132 (440)
HY-F 159 (74) 161 (537) 158 (548)
AHY-F -- 160 (204) 156 (492)

• All pairwisecomparisons of means,within a sex and age, are significant(t-test, P <

0.001) unlessindicatedby * (significant,P < 0.05) or NS (not significant,P > 0.05); there
are no significantdifferencesbetweenmeansfor Marin and Cedar Grove.
bData providedby Golden Gate Raptor Observatory,California.
c Data from Mueller et al. 1979a.

measured in the Marin Headlands; we excluded these data from com-

parison.Data for goshawkswere unavailablefrom the coastalsites.Stan-
dardmeasurements of weight,unflattenedwing chord,tail length,exposed
culmenlength,andtarsuswidth werecomparedwhenpossible.The t-test
was usedto identify significantlydifferent means(P < 0.05).
RESULTS

Regional size differenceswere least pronouncedin Sharp-shinned

Hawks. Goshutemigrantsweighedlessand had longerwings and tails
than thosefrom the Marin Headlands and Cedar Grove (Table 1), but
sharpshinsfrom the latter two sitesdid not differ significantly.Cooper's
Hawks consistentlyexhibitedthe greatestpercentvariation in size, and
alsogenerallyvaried more in termsof absolutemeasurements(Table 2).
Western Cooper'sHawks were significantlysmaller than easternmi-
grants; however, in proportion to their weight, western migrants had
longerwingsand tails (Table 3). Goshawksshowedan intermediatelevel
of variation;Goshutemigrantsusuallyweighedlessand had consistently
longerwings and tails than thosefrom Cedar Grove (Table 4).
Among Sharp-shinnedand Cooper'shawks, with one exception,mean
weight increasedfrom inland sitesto the nearestcoastalsite, whereas
mean wing chordand tail length decreased(Tables 1-2). Marin Head-
lands migrants also had longer beaks and thicker tarsi than Goshute
migrants (Table 5), which further suggestedthat the coastalmigrants
were larger overall, despitehavingshorterwings and tails.
Measurements revealed smaller differences in size between western
Sharp-shinned
andCooper'shawks,andbetweeneasternCooper'sHawks
Vol.61,No.2 Accipiter
SizeDifferences [195

TABLE2. Regionalcomparisons of meanswith samplesizes(in parentheses)

for mea-
surementsfrom fall-migrant Cooper'sHawks.a

Marin Goshute Cedar Cape May

Measure Age-sex Headlandsb Mountains Grovec Pointd
Weight (g) HY-M 288 (50) 269 (183) 335 (53) 347 (119)
AHY-M -- 281 (177) 349 (51) --

HY-F 417 (117) 399 (310) 499 (58) 518 (44)

AHY-F 443 (18) 439 (416) 529 (57) --

Wing chord(mm) HY-M 219 (61) 224 (317) 234 (52) 232 (121)
AHY-M -- 225 (287) 238 (48) --

HY-F 249 (130) 254 (444) 266 (59) 263 (46)

AHY-F 253 (18)* 256 (545) 270 (56) --

Standardtail length(mm) HY-M 187 (60) 190 (194) 196 (53)

AHY-M -- 181 (128) 191 (51)
HY-F 212 (124) 214 (286) 221 (58)
AHY-F 212 (15) NS 209 (285) 217 (58)

aAll pairwise comparisonsof means,within a sex and age, are significant(t-test, ? <
0.006) unlessindicatedby * (significant,P < 0.05) or NS (not significant,P > 0.05).
bData providedby Golden Gate Raptor Observatory,California.
c Data from Mueller et al. 1981a.
dData providedby Brian Millsap.

and goshawks.Wing chordsand tail lengthsof western Sharp-shinned

and Cooper'shawksoverlappedsignificantly(Hoffman et al. 1990), un-
like in the East (Mueller et al. 1979a, 1981a). Furthermore, differences
betweenmeanweightsof femaleSharp-shinnedand male Cooper'shawks
(within age groups)were consistently lessfor Goshutemigrantsthan for
Cedar Grove migrants (Tables 1-2). In contrast,weight differencesbe-
tweenfemaleCooper'sHawks and male goshawkswere consistently less
for Cedar Grove migrants (Tables 2-3). Similar resultsemergedwhen
we comparedmaximumand minimum valuesfor observedranges(Hoff-
man et al. 1990; Mueller et al. 1976, 1979a, 1981a).

TABLE3. Regionalcomparisons of wing chord/weightand tail length/weightratios for

fall-migrant Cooper'sHawks.

Marin Goshute Cedar Cape May

Measure Age-sex Headlandsa Mountains Grove• Pointc
Wing chord/weight HY-M 0.76 0.83 0.70 0.67
AHY-M -- 0.80 0.68 --
HY-F 0.60 0.64 0.53 0.51
AHY-F 0.57 0.58 0.51 --

Tail length/weight HY-M 0.65 0.71 0.59 --

AHY-M -- 0.64 0.55 --
HY-F 0.51 0.54 0.44 --
AHY-F 0.48 0.48 0.41 --

Data providedby Golden Gate Raptor Observatory,California.

Data from Mueller et al. 1981a.
Data providedby Brian Millsap.
196] J. P. Smithetal. J.Field
Ornithol.
Spring 1990

TABLE4. Regionalcomparisons of meanswith samplesize(in parentheses)

for measure-
ments from fall-migrant Northern Goshawks?

Goshute
Measure Age-sex Mountains Cedar Groveb
Weight (g) HY-M 748 (26) 808 (105)
AHY-M 797 (8) NS 925 (38)
HY-F 942 (48) 1005 (52)
AHY-F 967 (20) 1152 (59)
Wing chord (mm) HY-M 325 (37) 319 (109)
AHY-M 327 (15) ** 323 (41)
HY-F 358 (57) 346 (52)
AHY-F 357 (31) ** 353 (60)
Standardtail length (mm) HY-M 243 (20) 239 (106)
AHY-M 227 (7) NS 230 (41)
HY-F 281 (40) 272 (53)
AHY-F 269 (12) NS 266 (60)

aAll pairwisecomparisons of meansare significant(t-test,P < 0.005) unlessindicated

by ** (significant,P < 0.01) or NS (not significant,P > 0.05).
bData from Mueller et al. 1976; AHY is equatedto Adult II.

DISCUSSION

Proportionatelylow weight and high wing- and tail-surfacearea in-

dicatelow "flight-surfaceloading" (sensuAmadon 1980, Mueller et al.
1981b;differentfrom "wing-loading"in that it includeslift generatedby
both wing and tail surfaces).Surfacearea is directlyproportionalto the
squareof linear dimensions(Greenewalt 1962). A proportionaldecrease
in wing width couldoffsetan increasein length and causesurfacearea
to remain constant;however,the result would be a higher aspect-ratio
wing which also would reducewing-loading (Saville 1957). Since the
"width" of the tail dependson differential spreadingof the rectrices

TABLE5. Regionalcomparisons of meanculmenlengthand tarsuswidth with samplesize

(in parentheses)
for fall-migrantSharp-shinnedand Cooper'shawks.a
Marin Goshute
Measure Species Age-sex Headlandsb Mountains
Culmen length (mm) SS HY-M 10.0 (16) 9.8 (478)
SS HY-F 12.0 (70) 11.9 (505)
CH HY-M 15.3 (60) 14.8 (246)
CH HY-F 17.8 (124) 17.4 (374)
CH AHY-F 18.6 (18) 18.3 (447)
Tarsus width (mm) SS HY-M 3.8 (12) 3.5 (185)
SS HY-F 4.7 (65) 4.4 (152)
CH HY-M 5.8 (35) 5.5 (90)
CH HY-F 7.1 (76) 6.7 (73)
CH AHY-F 7.1 (7) 6.8 (73)

All pairwisecomparisons
of meansdiffer significantly(t-test,P < 0.001).
Data providedby Golden Gate Raptor Observatory,California.
Vol.61,No.2 Acczpiter
SizeDifferences [197

(Mueller et al. 1981b), a longertail will alwaysprovidemore available

surfacearea and lower tail-surface loading. Thus, despitethe lack of
actualsurfacearea measurements, we feel confidentthat low weight and
long wings and tail are valid indicatorsof low flight-surfaceloading.
Temple (1972) providesadditionalsupportfor this assumption:a wing-
loadingindex calculatedfor Merlins (Falcocolumbarius) by dividingthe
cuberoot of bodyweightby wing chord"showeda high correlationwith
the actual wing loading value computedas body weight in grams per
total wing area in cm2."
Accordingly,we suggestthat fall-migrant accipitersfrom the Goshute
Mountains averagelower flight-surfaceloading than thosefrom Cedar
Grove; that westernCooper'sHawks, in general,averagelower flight-
surfaceloadingthan easternmigrants;and that coastalmigrantsaverage
higher flight-surfaceloading than inland migrants. Two aberrant re-
suits--Marin versusGoshutetail lengthin AHY femaleCooper'sHawks
(Table 2) and tail lengthin AHY male goshawks(Table 3)--were likely
due to small sample sizes.Small samplesalso may have preventedthe
demonstrationof significantdifferencesin somecases.Proportionately
long wings and low flight-surfaceloadinghave beenconsideredadaptive
for long distancemigration (Averill 1920, Hamilton 1961, Salomonsen
1955, Zink and Remsen 1986) and for exploitingopen countrywhere
strong,sustainedflying is required (Behle 1942, Hamilton 1961, Linsdale
1938, Pitelka 1951, Temple 1972, Wattel 1973, Zink and Remsen1986).
Low flight-surfaceloadingmay alsohelp Goshuteaccipitersexploitstrong
thermal updrafts.
Bandreturnsindicatethat Cape May Point sharpshinsrarely originate
north of southern Ontario or winter farther south than Florida (Clark
1985). In contrast,Cedar Grovemigrantsoriginateasfar north ascentral
Alberta and generallywinter in the south-centralUnited States(Mueller
and Berger1967a).Similarly,Goshutemigrantsroutinelytravelto central
and southernMexico to winter and many originateasfar north ascentral
BritishColumbiaand Alberta (StephenW. Hoffman, unpubl.data). The
few band returns from Marin Headlands suggestthat most west-coast
migrantswinter in California, with somein northwesternMexico (Allen
Fish, pers.comm.).There are no breedingseasonrecoveriesfrom Marin
Headlands accipitersyet; however, a morphologicalstudy of accipiter
museum specimensrevealed that Marin Headlands migrants were most
similar to breeders from the Pacific Coast south of British Columbia
(Smith 1988a). Together thesedata suggestthat inland migrantsmay
have lower flight-surfaceloadingto facilitatelongermigrations.
Proportionatelylong wings and tail, and small overall size are char-
acteristicof breeding accipitersfrom inland and open habitats (Smith
1988a). Similarly,the prairie-parkland(openhabitat) subspecies of Mer-
lin, F. c. richardsonii,has the longestwings and tail and lowest wing-
loadingof the North American subspecies (Temple 1972). Inland mi-
grantsgenerallybreedin and travel throughhabitatsthat are more open
than the mesic,coastalforests.Thus, low flight-surfaceloadingin inland
198] J. P. Smithetal. J.Field
Ornithol.
Spring 1990

accipitersmay be an adaptationto bothopenhabitatsandlongmigrations.

Furthermore,the xeric, centraland southernintermountainwestsupports
sparservegetationthan the moremesicforestsof the Midwest. This may
favor lower flight-surfaceloadingamongGoshutemigrantscomparedto
Cedar Grove migrants.
The Goshutemigrantsoccupya regionwhere strongthermal updrafts
are prevalent.Thermal productionis greatestwhere the air is clear, dry,
and hot, where vegetationand soilorganicmatter are lacking,and where
exposedslopesare inclinedtoward the sun. Suchconditionsare charac-
teristicof the Great Basin and southernRocky Mountain regions.Fur-
thermore,long,leading-lineridges,whichmigratingraptorstendto follow
becausethey producestrongobstruction-current updrafts(Mueller and
Berger 1967b), are more erratically distributedin the West (Hoffman
1985). Consequently,migrating raptors must crossvalleys and plains
betweendiscontinuous ridges.By riding the strongthermalsgeneratedat
the basesof andin betweenridges,the hawkscansail acrosslongexpanses
of fiat, often barren land. Low flight-surfaceloadingwould maximize a
hawk's ability to exploit suchupdrafts.
Someauthorsarguethat accipitersdonot soar(Eckert 1987) and might
thereforediscountthe importanceof thermal updraftsto thesespecies.
However,alongthe Goshuteflyway it is commonto find all three species
of accipitersoaringwith the buteosand eagles.Calm, hot daysproduce
tremendousthermalsin the Great Basin.Often the only placemigrants,
including accipiters,can be spottedis silhouettedagainsta singlecloud
high in the sky.Low flight-surfaceloadingwould certainlyfacilitatesuch
activity.
Thermal productionalsomay be greater in the Midwest than on the
eastcoast.However, the Great Lakes, like the ocean,probably dampen
most thermal production.The lower flight-surfaceloadingof Cooper's
Hawks from Cedar Grove relativeto thosefrom Cape May is therefore
morelikely an adaptationto facilitatelongermigrationsand exploitation
of open habitats.In contrast,all three factors--openhabitat, long mi-
grations,andstrongthermalproduction--mayfavorthe evenlowerflight-
surfaceloading of Goshutemigrants.
We are less able to explain why regional variation should be more
pronouncedin Cooper'sHawks. Additional researchmay clarify the
causes of this trend.
The observedpatternsof regionalsizevariationalsohaveimplications
for field identification.Flight-style is an important characteristicto con-
siderwhen identifyingaccipiters.Differencesin flight-stylerelatedirectly
to differencesin bodysizeandweight(Mueller et al. 1979b,Smith 1988b).
Cooper'sHawks are closerin sizeto goshawksin the East, but closerin
sizeto sharpshinsin the West. Therefore,femaleSharp-shinnedand male
Cooper'shawks in flight are more likely to be confusedin the West,
whereasfemale Cooper'sHawks and male goshawksare more likely to
be misidentified in the East.
Finally, thisanalysisdemonstrated
that accipiterweightandwing chord
Vol.61,No.2 Accipiter
SizeDifferences [ 199

donot followthe samepatternof variationacrossthe United States.Wing

chordis generallyconsidered
a goodindicatorof overallbodysize(James
1970, Lanyon 1960, Selanderand Johnston1967), but if we had only
analyzedwing chord,we would have concludedthat migrant sharpshins
and goshawksfrom the GoshuteMountainswere largerthan thosefrom
Cedar Grove. Instead, becausewe also comparedweight data, we dis-
coveredthat the Goshute migrants were not larger overall, just long
wingedand tailed, a combinationthat reducesflight-surfaceloading.
ACKNOWLEDGMENTS

We extend specialthanks to Allen Fish and the Golden Gate Raptor Observatoryfor
providingus with the Marin Headlandsdata, and to Brian Millsap for providingthe data
from Cape May Point. Keith Dixon, Ivan Palmblad, Donald Sisson,Patricia Kennedy,
Buzz Hull, Allen Fish, Brian Millsap, Ken Meyer, Mike Collopy, SusanJacobson,and
Ruthe Smith reviewedearlier versionsof this manuscript.We thank them for their advice
and assistance.This paper is contribution No. 8 of the Western Foundation for Raptor
Conservation,Inc., and is a redraft of half the first chapterof JPS's MS Thesis.

LITERATURE CITED

AMADON,D. 1980. Varying proportionsbetweenyoung and old raptors. Proc. Pan-Afr.

Ornithol. Congr. 4:327-331.
AVERILL,C. K. 1920. Migration and physicalproportions.
A preliminarystudy.Auk 37:
572-579.
BEHLE,W. H. 1942. Distribution and variation of the Horned Larks (Otocorisalpestris)
of western North America. Univ. Calif. Publ. Zool. 46:205-316.
CANADIAN WILDLIFE SERVICE AND UNITED STATES FISH AND WILDLIFE SERVICE. 1980.
North American bird bandingtechniques:introductionto raptor age-sexkeys.Vol. 2,
pt. 6. Canadian Wildlife Service,Ottawa.
CLARK,W.S. 1985. The migrationof Sharp-shinnedHawks at Cape May Point: banding
and recoveryresults.Pp. 137-148, in M. Harwood, ed. Proc. Hawk Migration Confr.
IV. Hawk Migration Assoc.North America.
ECKERT,K. 1987. Hawks: a review. Birding 19:17-19.
GREENEWALT, C.H. 1962. Dimensionalrelationshipsfor flying animals.Smithson.Misc.
Collect. Vol. 144, No. 2.
HAMILTON,T. H. 1961. The adaptivesignificancesof intraspecifictrendsof variation in
wing length and body size amongbird species.Evolution 15:180-195.
HOFFMAN,S.W. 1985. Raptor movementsin inland westernNorth America:a synthesis.
Pp. 325-338, in M. Harwood, ed. Proc. Hawk Migration Confr. IV. Hawk Migration
Assoc. North America.
, J.P. SMITH, ANDJ. A. GESSAMAN.1990. Size of fall-migrant accipitersfrom the
GoshuteMountains of Nevada.J. Field Ornithol. 61:201-211.
JAMES,F. C. 1970. Geographicsize variation in birds and its relationshipto climate.
Ecology51:365-390.
LANYON,W. E. 1960. The Middle American populationsof the Crested Flycatcher,
Myiarchustyrannulus.Condor 62:341-350.
LINSDALE,J. M. 1938. Bird life in Nevada with referenceto modificationsin structure
and behavior. Condor 40:173-180.
MUELLER,H. C., ANDD. D. BERGER.1967a. Fall migrationof Sharp-shinnedHawks.
Wilson Bull. 79:397-415.
, AND-- 1967b. Wind drift, leading lines, and diurnal migration. Wilson
Bull. 79:50-63.
, --, AND G. ALLEZ. 1976. Age and sex variation in the size of Goshawks.
Bird-Banding 47:310-318.
200] d.P. Smithetal. J.Field
Ornithol.
Spring 1990

, --, AND-- 1979a. Age and sex differencesin size of Sharp-shinned

Hawks.
Bird-Banding
50134-44.
--, AND-- 1979b. The identificationof North American accipiters.
An•.Birds33:236-240.
, --, AND . 1981a. Age, sex,and seasonaldifferencesin sizeof Cooper's
Hawks. J. Field Ornithol. 52:112-126.
, --, AND--. 1981b. Age and sexdifferencesin wing loadingand other
aerodynamiccharacteristicsof Sharp-shinnedHawks. Wilson Bull. 93:491-499.
PITELKA,F.A. 1951. Speciationand ecologicaldistributionin Americanjays of the genus
Aphelocoma. Univ. Calif. Publ. Zool. 50:195-464.
SALOMONSEN, F. 1955. The evolutionarysignificance of bird migration.Dan. Biol. Medd.
22:1-62ß
S^VILLE,D. B.O. 1957. Adaptiveevolutionin the avian wing. Evolution11:212-214.
SELANDER, R. K., AND R. F. JOHNSTON. 1967. Evolution in the House Sparrow. I.
Intrapopulationvariation in North America. Gondor69:217-258.
SMITH,J.P. 1988a. Morphometricvariationin accipiterhawkswith emphasison western
North America. Utah State University, M.S. thesis.
ß 1988b. Field identificationof North American accipiters.Utah Birds 3:37-55.
TEMPLE,S.A. 1972. Systematicsand evolutionoœthe North Americanmerlins. Auk 89:
325-338.
W^TTEL, J. 1973. Geographicaldifferentiationin the genusAccipiter.Nuttall Ornithol.
Club Publ. No. 13. 231 pp.
ZINK, R. g., ANDJ. V. REMSEN.1986. Evolutionaryprocesses andpatternsoœgeographic
variationin birds.Pp. 1-69, in R. F. Johnston,ed. Current Ornithology,Vol. 4. Plenum
Press, New York.

Received15 Dec. 1988; accepted24 Oct. 1989.