This document describes a study that developed methods to determine the sex of Booted Eagles using morphometric data and molecular techniques. Researchers measured physical characteristics of 100 young and 42 adult Booted Eagles and used blood samples to determine sex via molecular analysis. Discriminant functions were developed that could determine sex with 100% accuracy for adults using mass and forearm length, and 98.8% accuracy for young using additional measurements. The functions provide reliable methods for sexing Booted Eagles in the field without molecular analysis.

1. j RaptorRes.35(1):20-23
¸ 2001 The Raptor ResearchFoundation,Inc.
SEX DETERMINATION IN BOOTED EAGLES (H,RAAETUS
PENNATUS) USING MOLECULAR PROCEDURES AND
DISCRIMINANT FUNCTION ANALYSIS
JAVIERBALBONTiN,MIGUELFERRER,ANDEVACASADO
EstacidnBiol6gicadeDohana(CSIC)Avda.Maria Luisas/n, PabellSndelPerd,41013 Sevilla,Spain
ABSTRACT.--Westudieda breeding population of Booted Eagles(Hieraaetuspennatus)in Dofiana Na-
tional Park (southwesternSpain) to develop a method of determining the sex of an individual based
on the use of discriminant functions. Becausethere are size differencesbetween age classesand sexes
of eagles,we developed two different discriminant functions for each age group. Our discriminant
function method approached100% accuracyin correctlyagingindividualsusingforearm lengthand
bodymassaspredictorvariables.Sexof youngeagleswasalsodeterminedwith 98.8% accuracyusing
forearm, tail, bill, and tarsuslengths.
KEYWORDS: BootedEagle;,Hieraaetuspennams;sexdetermination;morphometrics;molecularsexing.
Determinaci6ndel sexodel figuilacalzadaHieraaetuspennatusutilizandotecnicasde sexadomolecular
y analisisdiscriminates
RESUMEN.--Unapoblaci6nreproductorade figuilacalzadaha sidoesmdiadaen el ParqueNacionalde
Dofiana (Sudoestede Espafia) con el objetivo de obtener un modelo de clasificaci6nde los sexos
basadosen finalisisdiscriminantesapoyadosen procedimientosde sexadomolecular.Existendiferencias
importantesen el tamafio entre figuilasadultasy pollos,por 1oque sehan desarrolladodosfunciones
discriminantesde sexodiferentespara cadaclasede edad. E1sexode losadultossedetermina con una
funci6n discriminanteque clasificabien el 100% de los individuos,utilizandoel antebrazoy el peso
como variablespredictoras.E1sexode los polloses determinado tambi•n correctamentecon una fun-
ci6n discriminateque clasificabien el 98.8% de los individuos,utilizando cuatrovariablespredictoras:
E1antebrazo,la cola, el pico y el tarso.
[Traducci6n de Autores]
Easyand reliablemethodsto identifythe sexof
individualsare usefulfor the studyof manyaspects
of avian biology,including foraging ecology(An-
dersonand Norberg 1981), behavior,evolutionary
ecologyand genetics(Clutton-Brock1986), survi-
vorship (Newton et al. 1983), and dispersionand
conservationgenetics (Griffith and Tiwari 1995).
Sex determination is alsoimportant in conserva-
tion programsthat concern the reintroduction of
endangered birds when a fixed sex ratio is pre-
ferred. Recently,Ellegren (1996) proposedmolec-
ular methods to sex birds based on chromosome
differences but few studies have used this infor-
mation to developadditionalinethodsto sexbirds
basedon biometric data. Field methodsto sexrap-
tors have severaladvantagesover moleculartech-
niques that require time and/or money. Despite
the fact that the majorityof raptorsare highly di-
morphic in size,which shouldallowfor the devel-
opment of sexingmethodsbasedon morphomet-
ric data, only a few specieshave been utilized
(Bortolotti 1984a, 1984b, Garcelon et al. 1985, Ed-
wards and Kochert 1987, Ferrer and De Le Court
1992). The majority of these studieshave been
based on live individuals and museum skins. In
most cases,both adults and immatures have been
studied at museumsor in private collectionsand
few studies have been based on wild individuals.
The objectiveof this studywasto assessthe differ-
encesbetweenyoungand adult BootedEagles(Hi-
eraaetuspennatus)and to develop predictive dis-
criminant models to determine the sexes of adults
and immaturesof the species.
METHODS
We useda sampleof the breedingpopulationof Boot-
ed Eaglesin Dofiana National Park. The park is located
in southwesternSpain (37øN,6ø30'W).It hasa Mediter-
ranean climate with an Atlantic Ocean influence. Marsh-
es, Mediterranean scrubland mixed with scattered cork
oak (Quercussuber)or stonepine (Pinuspinea),andcostal
20

2. MARCH 2001 SEX DETERMINATION IN BOOTED EAGLES 21
Figure 1. Measurementof forearm length in Booted
Eagle.
sand dunes are the main habitats found in the area. A
more detailed description of this area is presented in
Rogersand Myers (1980).
Six morphometric measurementswere takenfrom wild
adult and immature eagles.To obtain measurements,we
visitednestswhen youngwere 35-45 d old and their skel-
etonswere completelygrownbut their featherswere still
growing.Youngleavethe nestwhen theyare about 55 d
old (Balbontin unpubl. data). A total of 100 young were
measuredbetween 1996-98. Adultswere trapped usinga
2 X 3 m dho gaza net and an unreleasablecaptiveowl
(Bubobubo)lure. Forty-twoadultswere caughtusing this
method, 12 in 1997 and 30 in 1998. We took measure-
mentsofwing,tail, bill withcere,and tarsuslengthsusing
a digital caliper to the nearest0.1 mm and metal rulers
to the nearest 1 mm (Bortolotti 1984). We also measured
the forearm length, or the length from the front of the
folded wrist to the proximal extremity of the ulna using
calipers (Fig. 1) (Ferrer and De Le Court 1992). All the
individualswere weighedwith 1 kg or 2.5 kg Pesolascales
with precisionsof 5 g and 10 g, respectively.
We extracted 2 ml of blood from the brachial vein of
each eagle and stored part of it (50 •xl) in buffer and
kept it refrigeratedfor later analysis.The cellularfraction
wasusedto sexthe eaglesfollowingEllegren (1996). We
used primers 2945F, cfR, and 3224R to amplify the W-
chromosome gene following Ellegren's (1996) recom-
mendations.Using thistechnique,we identified the sexes
of 81 immature (41 females, 40 males) and 41 adult ea-
gles (16 males, 25 females)(Fig. 2). This sample of
known-sex individuals was used to derive the discriminant
function using morphometric data.
Becauseyoung often differ in size from adults (Bor-
tolotti 1984b), we used multivariate analysisof variance
(MANOVA) to check for differences in size between
malesand femalesand young and adult eagles.Six mea-
surementstaken from all age and sex classeswere com-
pared usingunivariateanalysisof variance (ANOVA) and
nonparametric statisticsfor thosevariableswhen homo-
geneity of variancewasnot met. We used the SPSSpro-
gram (Norusis 1992) to do this analysis.We separated
youngfrom adultsto examine differencesbetweensexes.
First, we checked for sexual differences for each of the
six morphological charactersusing t-tests.We derived a
discriminantfunction using DISCRIM procedure of the
SAS Systemprogram (version 6.12). A jackknife proce-
dure wasapplied to testthe efficacyof the estimateddis-
criminant function (Lachenbruch and Mickey 1968).
Each individual in the samplewasclassifiedusing a dis-
criminant function derived from the total sample, ex-
cluding the individual being classified(Chardine and
Morris 1989, Amat et al. 1993). We chose the function
which had the lowest percentage of misclassification
basedon the molecular determination of gender.
RESULTS AND DISCUSSION
Our analysesof the morphometric data showed
that adult Booted Eaglesdiffered significantlyin
sizefrom youngeaglesand that maleswere signif-
icantlysmallerthan females (MANOVA: sex - F =
72.0,df = 6, 111,P < 0.001;age - F = 181.85,df
= 6, 111, P < 0.001). Tail, wing, and culmen mea-
surement showedthe greatestdifference between
age groups,with the featuresof adult individuals
larger than those of immatures (Table 1). There
were no significantage-or sex-specificdifferences
in bone measurements such as tarsus and forearm
F M F M F F F M F M F M F F M M M F F F M F
.. .
Figure 2. Gender identification using PCR test.A multiple amplification with 2945F and cfR specificallyamplify a
210 bp fragment of the W chromosomein females and 2945F q- 3224R that amplifies 630 bp fragments in both
sexes.Femalesare indicated by the arrow.

3. 22 BALBONTiNET AL. VoI•. 35, No. 1
AVVVVV
+1 +1 +1 +1 +1 +1
+1 +1 +1 +1 +1 +1
VVVA
+1 +1 +1 +1 +1 +1
+1 +1 +1 +1 +1 +1
lengthsbut forearm lengthwassignificantlysmall-
er in youngfemaleeagles(Table1). BootedEagles
showhigh sexual dimorphism in size and both
adultsand youngdiffered significantlyin the ma-
jority of the variableswe studied. Adult females
were significantlylarger than males for all mea-
surementstaken,with forearm and bodymassthe
most dimorphic characters(Table 2). Youngfe-
malesare alsolargerthan malesand theyhavealso
longerforearmsandbeaks,but similar-sizedwings
and tails. Our discriminantfunction analysisclas-
sified100%of the adultfemaleand maleeagles
correctlyusingbodymassandforearmaspredictor
variables.The discriminantfunction equationfor
adults was:
D = -178.885 + 0.05613(MASS) +
0.95937(FOREARM)
Youngwereclassifiedmostaccuracyusingthefour
variablesforearm,tail,bill, andtarsusaspredictors
in the model. The discriminant function misclas-
siftedonly one female. The discriminantfunction
for youngwas:
D = -197 + 0.6761(FOREARM) - 0.19286(TAIL)
+ 2.99438(BILL) + 0.5858 (TARSUS)
Valuesof D > 0 representfemalesand valuesof
D < 0 representmales.By deletingtail andwing
measurementswhich are highly variablefrom the
model,youngeagleswere alsoclassifiedwith 84%
accuracyusing only tarsusand forearm measure-
ments in the discriminant function:
D = -33.815 + 0.147(FOREARM) +
0.207(TARSUS)
The equationswe derivedfor sexingBootedEa-
glesshould be usefulfor future work on the biol-
ogyof thisspecies.For immature eagles,measure-
mentsof wingsand tailsshouldbe takencarefully
if theyareusedto discriminategenderbecausethe
feathersof youngbirds keep growingafter they
first take flight. Adultswere correctlyclassifiedto
gender in 100% of casesexaminedby usingthe
twovariables,bodymassand forearm. The latteris
an easymeasurementto take and repeated mea-
surementstakenby different observersshowedlow
variances (Ferrer and De Le Court 1992). Gender
discriminationfor youngeaglesisvalid at 35-45 d
of agewhennestlingshavealmostcompletedtheir
growth.

4. MARCH 2001 SEX DETERMINATION IN BOOTED EAGLES 23
Table 2. Differencesin morphometric measurementsbetweenmale and female young and adult Booted Eagles.
ADULTS YOUNG
MALE FEMALE MALE FEMALE
(N = 16) (N = 25) (N = 40) (N = 41)
(5• + SD) (5• + SD) t P (5• + SD) (5• + SD) t P
Tarsus 64.1 +_ 2.77 69.4 ___3.23 -5.715 <0.001 64.4 _+ 2.51 69.1 +_ 3.30 -7.15 <0.001
Forearm 132.2 -+ 2.64 143.5 _+ 3.20 -12.40 <0.001 131.5 +_ 4.72 140.0 --- 4.85 -7.95 <0.001
Culmen 31.5 m 1.14 34.8 -+ 1.32 -8.604 <0.001 28.8 +- 1.29 30.9 --- 1.09 -7.93 <0.001
Wing 355.0 ---27.8 389.2 +_9.41 -5.712 <0.001 244.4 +_25.9 244.6 _+28.8 -0.03 0.970
Tail 195.6 _+8.41 208.7 _+9.24 -4.763 <0.001 121.0 --- 18.6 112.9 +- 21.5 1.78 0.078
Mass 690.3 _+ 40.9 973.2 _+ 76.9 -13.46 <0.001 656.3 _+ 68.7 828.7 +- 88.3 -9.81 <0.001
ACKNOWLEDGMENTS
Wewishto thank to H. Le FrancandJ. Cuestafor their
help with the field work.J.M. Arroyo did the lab work
and E Moreno helped with the drawings.We are indebt-
ed to L. Garciafor hishelp assisting,locating,and show-
mg us eagleterritories.We alsowishto thank M. Berto-
lotti for reviewingan early draft of this article.
LITERATURE CITED
AMAT,J.A.,j. VIY4UELA,ANDM. FERPmR.1993.SexingChin-
strapPenguins(Pygoscelisantarctica)bymorphological
measurements. Colonial Waterbirds 16:213-215.
ANDERSON,M. AND R.A. NOPmERG.1981. Evolution of re-
versedsexualsizedimorphismand role partitioning
among predatory birds, with a sizescalingof flight
performance.Biol.J. Linn. Soc.15:105-130.
BORTOLOTTI,G.R. 1984a. Sexual size dimorphism and
age-relatedsizevariation in bald eagles.J. Wildl.Man-
age.48:72-81.
•. 1984b. Age and sex sizevariation in Golden Ea-
gles.J. FieldOrnithol.55:54-66.
CHARDINE,J.W. ANDR.D. MORVdS.1989. Sexual size di-
morphismand assortativemating in the BrownNod-
dy. Condor91:868-874.
CLUTTON-BROCK, T.H. 1986. Sex ratio variation in birds.
Ibis 128:317-329.
EDWARDS,T.C. ANDM.N. KOCHERT.1987. Use of body
weight and length of footpad aspredictorsof sex•n
GoldenEagles.J. FieldOrnithol.58:144-147.
ELLEGREN,H. 1996. First gene on the avian W chromo-
some (CHD) provides a tag for universal sexing of
non-ratite birds. Proc. R. Soc. Lond. B. 263:1635-1641
FERRER,M. AND DE LE COURT. 1992. Sex determination
in the SpanishImperial Eagle.J.FieldOrnithol.62:359-
364.
GARCELON,D.K., M.S. MARTELL, P.T. REDIG, AND L.C. BUD-
EN.1985.Morphometric,karyotipicand laparoscopics
techniquesfor determinig sexin BaldEagles.J. Wildl
Manage.49:593-599.
GRIFFITH, R. AND B. TIWARI. 1995. Sex of the last wfid
Spix'sMacaw.Nature(Loud.).375-454.
LACHENBRUCH,P.A. AND M.R. MICKEY. 1968. Estimation of
error ratesin discriminantanalysis.Technometrics10'
1-11.
NEWTON,I., M. MARQUISS,AND P. ROTHERY.1983. Age
structure and survivalin a Sparrowhawkpopulation.
J. Anita.Ecol.52:591-602.
NORUSlS,M.J. 1992. SPSSfor Windows. Base System
User's Guide. Released5.0, Chicago,IL U.S.A.
ROGERS,P.M. ANDH.M. MYERS.1980. Animal distribution,
landscape classification,and wildlife management,
Coto de Dofiana,Spain.J. Appl.Ecol.17:545-565.
Received 20 March 2000; accepted 22 October 2000