Ex apai Pa e ni y and Ma e ni y in he Th ee-Toed
Woodpecke ,
Picoides idac ylus
: Insigh s om
Mic osa elli e-Based Pa en age Analysis
Meng-Hua Li
1
*, Kaisa Va
¨lima
¨ki
1
, Ma kus Piha
2
, Timo Pakkala
2
, Juha Me ila
¨
1
1Ecological Gene ics Resea ch Uni , Depa men o Biological and En i onmen al Sciences, Uni e si y o Helsinki, Helsinki, Finland, 2Finnish Museum o Na u al His o y,
Uni e si y o Helsinki, Helsinki, Finland
Abs ac
Molecula echniques ha e e ealed ha a ian ma ing sys ems a e mo e di e se and complex han p e iously hough . We
used mic osa elli e ma ke s o de e mine gene ic pa en age, he p e alence o ex apai pa e ni y and quasi-pa asi ism (i.e.
si ua ions whe e a male’s ex apai ma e lay in his nes ) in a socially monogamous popula ion o h ee- oed woodpecke s
(Picoides idac ylus) in sou he n Finland. A o al o 129 adul s and nes lings, ep esen ing 5–9 amilies annually om 2004–
2007, we e geno yped a up o en mic osa elli e loci. The esul s o gene ic assignmen es s con i med ha monogamous
pa en age cha ac e ized he majo i y (84.6%, 22/26) o b oods, and ha mos (93.8%, 75/80) nes lings we e he o sp ing o
hei social pa en s. Two o 80 nes lings (2.5%) in wo o 26 b oods (7.7%) we e si ed by ex apai males and quasi-pa asi ism
occu ed in 3.8% (3/80) o nes lings and 7.7% (2/26) o b oods. Hence, he le els o ex apai pa en age we e low, possibly
because bo h gene ic polygyny and polyand y a e cons ained by he high pa e nal e o equi ed o pa en al ca e. The
co-occu ence o low le els o ex apai pa e ni y and quasi-pa asi ism a e discussed in ligh o ecological and beha iou al
ac o s cha ac e izing he species biology.
Ci a ion: Li M-H, Va
¨lima
¨ki K, Piha M, Pakkala T, Me ila
¨J (2009) Ex apai Pa e ni y and Ma e ni y in he Th ee-Toed Woodpecke , Picoides idac ylus: Insigh s om
Mic osa elli e-Based Pa en age Analysis. PLoS ONE 4(11): e7895. doi:10.1371/jou nal.pone.0007895
Edi o : Tom Pizza i, Uni e si y o Ox o d, Uni ed Kingdom
Recei ed Sep embe 25, 2009; Accep ed Oc obe 23, 2009; Published No embe 18, 2009
Copy igh : ß2009 Li e al. This is an open-access a icle dis ibu ed unde he e ms o he C ea i e Commons A ibu ion License, which pe mi s un es ic ed
use, dis ibu ion, and ep oduc ion in any medium, p o ided he o iginal au ho and sou ce a e c edi ed.
Funding: This wo k was suppo ed by he g an (No.118673) om he Uni e si y o Helsinki (h p://www.helsinki. i/uni e si y). The unde s had no ole in s udy
design, da a collec ion and analysis, decision o publish, o p epa a ion o he manusc ip .
Compe ing In e es s: The au ho s ha e decla ed ha no compe ing in e es s exis .
* E-mail: menghua.li@helsinki. i
In oduc ion
Ex a-pai e iliza ions, which can esul om emales engaging
in copula ions wi h ex a-pai males (ex apai pa e ni y; EPP), o
om males copula ing wi h ex a-pai emales ha lay eggs in he
male’s nes (ex a-pai ma e ni y; EPM o quasi-pa asi ism, QP)
[1], is known om app oxima ely 90% o he a ian species (see
[2]). EPPs a e known o be common in passe ines and less so in
non-passe ines [3]. Despi e i s ubiqui y ac oss a ian species, he
p e alence o EPP a ies conside ably wi hin and among species
(see [2]). In con as o EPP, QP is a e and has been desc ibed in
only a ew bi d species. Howe e , a close examina ion o hese
s udies e ealed ha unequi ocal e idence o QP is slim due o
possibili y o apid ma e-swi ching and/o insu icien molecula
wo k [4]. Hence, i emains unclea whe he QP is gene ally a e,
o whe he i s appa en sca ci y e lec s he di icul y o iden i ying
i when occu ing.
Gene ic pa en age s udies ha e been conduc ed only in ou ou o
mo e han 200 woodpecke species (see [5]). One o hese is he
h ee- oed woodpecke (Picoides idac ylus) which is ypically socially
monogamous [6], al hough occasional cases o simul aneous social
polyand y ha e been eco ded [5], [7]. Howe e , mo e accu a e
pa en age analysis ools o he h ee- oed woodpecke s would be
needed o add ess ques ions ela ing o he gene ic bene i s o ma e
choice, inb eeding a oidance and he ac ual b eeding sys em in his
species (c . [8]). Likewise, addi ional da a om non-passe ine bi ds
will be also use ul in unde s anding he e olu iona y signi icance and
li e-his o y co ela es o p omiscui y in bi ds. In compa ison wi h he
mul ilocus DNA inge p in ing analyses p e iously conduc ed in
h ee- oed woodpecke s [5], applica ion o high- esolu ion mic o-
sa elli es would ep esen a mo e e icien and s aigh o wa d
echnique o pa en age assignmen and kinship analyses [9].
The main pu pose o he p esen s udy was o es ima e he
p e alence o EPP and QP in he h ee- oed woodpecke . Since he
males o his species alloca e signi ican ly mo e ime o e i o y
de ence, ca i y exca a ion and eeding o he young han emales [5],
we p edic ed ha his should esul in low equency o EPP and
possible occu ence o QP. The s udy was conduc ed in a popula ion
b eeding in sou he n Finland, which has been s udied since he la e
1980s and has been a subjec o a i e-yea (2003–2007) in ensi e
popula ion s udy (e.g. [10], [11]). To his end, we applied a se o 10
polymo phic mic osa elli e loci de eloped o he species [12]. In
addi ion, gi en he s a is ical limi a ions acing mos pa en age s udies
(e.g. [13]), we u he applied an app oach which was implemen ed
in p og am CERVUS and has been p o en o be wi h high
con idence in pa en age assignmen in an open ma ing sys em [14].
Me hods
E hics S a emen
The me hods we e app o ed by he ins i u ion ha coo dina es
inging ac i i y in Finland (Finnish Museum o Na u al His o y),
based on he egula ion by he Minis y o he En i onmen (No.
17/5713/2002).
PLoS ONE | www.plosone.o g 1 No embe 2009 | Volume 4 | Issue 11 | e7895
S udy Species
The h ee- oed woodpecke (Picoides idac ylus) is a ca i y nes ing
habi a specialis inhabi ing coni e ous aiga o es s in he no h,
and high ele a ion alpine coni e ous o es a he sou he n edge o
he bo eal zone [5]. They exhibi nes -si e ideli y o e yea s [7],
ma e gua ding [15], long du a ion o ca i y exca a ion, bi-
pa en al ca e, and in pa icula , a e y high deg ee o pa e nal
ca e due o exclusi e incuba ion/b ooding a nigh by he males
[5]. The a e age b eeding densi y a ies a lo depending on he
incidence o i e acciden s [16] and he quali y o o es landscape,
habi a s and spa ial scale conside ed, and has been es ima ed o
a y be ween 0.1 and 1.5 e i o ies/km
2
in an in ensi e s udied
a ea o 150 km
2
in sou he n Finland [10], [11].
S udy Si e and Popula ion
The s udy was conduc ed in he E o a ea (ca.61u119N,
25u069E) in sou he n Finland in an a ea o 150 km
2
. The s udy
a ea has been desc ibed in mo e de ail in [10], [11].
The e i o y numbe s and nes si es o he h ee- oed
woodpecke popula ion inhabi ing he E o a ea ha e been s udied
since 1987 [11]. An in ensi e s udy on b eeding biology, including
indi idual ma king o bi ds wi h colou ings, was s a ed in 2004.
Du ing yea s 2004–2007 e i o ies and nes s we e sea ched in he
s udy a ea by using me hods desc ibed in [11]. Ten o 25 nes s
wi h nes lings we e ound annually. Fo each nes possible, he
adul s we e apped and nes lings we e pulled ou om he nes
ca i ies wi h a special ool (so ongs) ha is in gene al use in he
woodpecke inging p ojec s in Finland. The adul s we e apped
using mis ne s o a ne designed o ca ching bi ds coming ou
om he nes ca i y. Each bi d was agged wi h an indi idual
combina ion o colou ings, measu ed, aged, sexed (adul s only)
using he mo phological c i e ia gi en in [6], and body ea he
samples we e collec ed om he bi ds. Two o i e ea he s we e
plucked wi h a pai o weeze s om he en al body ea he ac s
in case any ea he s did no all away du ing he handling o he
bi ds. No ad e se e ec s on bi ds we e obse ed du ing o a e
he ca ching, inging, measu ing and ea he emo al. Finally he
nes lings we e pu back in he nes and adul s we e eleased. Due
o cha ac e is ics o ees, ca i ies o si es, all ound nes s could no
be sampled comple ely. In addi ion, some nes s we e ound oo
la e in he cou se o he b eeding season he e o e pulling he la ge
nes lings ou om he ca i y was no longe sa e o he indi iduals.
Al oge he 26 nes s we e sampled adequa ely o u he analyses.
DNA Ex ac ion, Molecula Sexing and Mic osa elli e
Geno yping
Genomic DNA was isola ed om he body ea he sha s using
he Chelex-based ex ac ion p o ocol (Bio-Rad, Helsinki, Finland)
ollowing he manu ac u e ’s ins uc ions. The sex o all samples
was iden i ied ollowing a simple and uni e sal me hod o
molecula sexing o non- a i e bi ds using PCR ampli ica ion o
he CHD1 gene as de ailed in [17]. In hose cases whe e bi ds
we e sexed on he basis o c own ea he colo a ion (e.g. [6]) in he
ield, he esul s o ield and molecula sex iden i ica ion me hods
ma ched each o he pe ec ly. A o al o 10 polymo phic
mic osa elli es de eloped o he h ee- oed woodpecke [12] we e
included in his in es iga ion (Table 1). The PCR geno yping
p o ocols a e a ailable om [12]. All geno ypes we e double
checked independen ly by wo pe sons.
Mic osa elli e Va ia ion
De ia ions om Ha dy-Weinbe g equilib ium (HWE) o each
locus and om linkage equilib ium be ween all pai s o loci we e
es ed wi h Fishe ’s exac es s based on he app oach o [18] using
GENEPOP e sion 3.4 [19] wi h 100 000 s eps in he Ma ko
chain (100 ba ches wi h 1000 i e a ions). Basic di e si y indices,
including he numbe o alleles, obse ed he e ozygosi y, Nei’s
[20] unbiased es ima es o expec ed he e ozygosi y, wi hin-
popula ion inb eeding coe icien (F
IS
; [21]), and equency o
null alleles we e es ima ed a each locus as well as o e all loci
using GENEPOP. S anda d exclusion p obabili ies o each locus
and o he selec ed loci combined (Table 1) we e es ima ed wi h
he p og am CERVUS 3.0 [14].
The dis ibu ion o geno ypes a he en loci con o med o he
expec a ion o HWE and all he locus pai s we e in linkage
equilib ium (Table 1; P.0.05; da a no shown o he esul s o
es s o linkage disequilib ium). The cumula i e exclusion
p obabili ies o he se o loci used in he pa en age analysis we e
high: 0.9935 o he i s pa en and 0.9998 o he second pa en
(assuming he i s pa en was assigned co ec ly; Table 1).
Pa en age Analysis
We i s checked misma ch dis ibu ions be ween he pu a i e
pa en s and he nes lings. The ac ha mos o sp ing ma ched he
pu a i e mo he o a he exac ly, o misma ched a a single locus,
s ongly sugges s ha mos o he pu a i e pa en s we e ue
gene ic pa en s. O he cases whe e misma ches occu ed, i e
misma ched by mo e han one epea a wo o mo e loci
(Figu e 1).
The pa en age analysis o gene ic da a om he ield-collec ed
samples was u he pe o med using he compu e so wa e
CERVUS e sion 3.0 [14]. By using a likelihood-based app oach
desc ibed in [14], CERVUS calcula es pa en age in e ence
likelihood a ios and gene a es a s a is ic, DLOD, de ined as he
di e ence in posi i e log likelihood a ios (LOD) be ween he op
wo candida e pa en s.
A o al o 10,000 es s, which is hough o be su icien in mos
cases [14], we e used he e. We de ine ‘candida e pa en s’ as adul s
o he popula ion in a speci ic yea . Bo h male and emale h ee-
oed woodpecke s we e assumed o be capable o p oducing
o sp ing in hei second calenda yea (one yea old) and,
Table 1. Summa y s a is ics o he 10 mic osa elli e loci used
in his s udy.
Locus
nN
A
H
O
H
E
F
IS
F
Null
P
(Ex1
)
P
(
Ex2
)
P
H-W
P i13 129 11 0.767 0.768 0.001 20.002 0.372 0.549 0.1761
P i17 126 11 0.817 0.845 0.033 0.027 0.525 0.692 0.5625
P i20 129 3 0.101 0.097 20.036 20.016 0.005 0.049 1
P i22 129 11 0.829 0.806 20.029 20.013 0.448 0.623 0.2304
P i23 123 7 0.715 0.662 20.081 20.045 0.264 0.448 0.1574
P i24 129 7 0.69 0.655 20.054 0.028 0.232 0.392 0.2489
P i30 127 15 0.827 0.864 0.044 0.020 0.57 0.728 0.0346
P i31 125 6 0.864 0.8 20.081 20.044 0.422 0.6 0.4087
P i36 129 11 0.783 0.844 0.073 0.036 0.519 0.687 0.1451
P i38 127 9 0.732 0.782 0.064 0.032 0.426 0.609 0.0493
O e all 127.3 8.7 0.713 0.712 0.0001 – 0.9935 0.9998 –
Numbe o bi ds sc eened (n), numbe o alleles (N
A
), obse ed he e ozygosi y
(H
O
), expec ed he e ozygosi y (H
E
), Wei and Cocke ham’s (1984) wi hin-
popula ion inb eeding coe icien (F
IS
), equency o null alleles (F
Null
), exclusion
p obabili y o he locus o he i s pa en (P
Ex1
), exclusion p obabili y o he
locus o he second pa en wi h he i s assigned (P
Ex2
), and he exac
p obabili y o de ia ion om Ha dy-Weinbe g equilib ium (P
H–W
).
doi:10.1371/jou nal.pone.0007895. 001
Ma ing Sys em o a Woodpecke
PLoS ONE | www.plosone.o g 2 No embe 2009 | Volume 4 | Issue 11 | e7895
he e o e, su i ing males o emales om ea lie coho s we e
included as candida e pa en s o o sp ing bo n in la e yea s. The
numbe o candida e emales was 10, 21, 34, and 42 and he
numbe o candida e males was 13, 32, 48, and 59 o he 2004–
2007 coho s, espec i ely. The sampling o pa en s in he s udy
a ea was no exhaus i e, and i was es ima ed ha ca. 25–40% o
he adul s we e sampled depending on he yea (M. Piha, pe sonal
obse a ion). Thus, a sampling a e o 25% was used o he 2004
coho and 40% o he 2005–2007coho s. The p opo ion o
success ully geno yped loci was on a e age 98.5% as es ima ed
om he gene ic da a (see esul s). A yping e o a e o 1.2% was
inco po a ed in o he simula ion o ma e ni y and pa e ni y
assignmen s. Assignmen s we e ca ied ou a a elaxed le el o
80% and a s ic le el o 95%.
We assigned pa en age unde wo scena ios o s eps. (i) Fo
comple e amilies in which bo h pu a i e pa en s we e sampled
(N= 46 nes lings in 14 b oods), we i s assigned ma e ni y wi h
unknown pa e ni y using he p og am CERVUS. Once a emale
was assigned, we hen a emp ed o assign pa e ni y o ei he he
pu a i e a he o a po en ial b eeding male om he popula ion
wi h known ma e ni y; o he wise, pa e ni y assignmen s we e
implemen ed wi h unknown ma e ni y. We included he pu a i e
pa en s when possible and all po en ial emales/males in he
popula ion as possible candida es o ma e ni y/pa e ni y. (ii) Fo
amilies which he e was sample a ailable o only he pu a i e
a he (N= 34 nes lings in 12 b oods) we again a emp ed o assign
pa e ni y using CERVUS wi h unknown ma e ni y. The e we e
no cases whe e a DNA sample was a ailable o jus he pu a i e
mo he .
The dis ibu ion o he LOD sco es o assigned and excluded
pa en s is shown in Figu e 2. The LOD sco es o unequi ocal
wi hin-pai o sp ing (i.e. assigned o pu a i e pa en s) assigned on
he basis o ma ching eigh o mo e loci a e all posi i e and he
majo i y o hem a e g ea e han h ee (Figu e 2a,b), bu his is
no ue o he i e ex apai o sp ing assigned wi h ex apai
pa en age (Figu e 2c). The Dc i e ion calcula ed o assignmen o
pa en age was be ween 1.44 and 5.14 in di e en yea s o 95%
con idence, and be ween 0 and 2.94 o 80% con idence whe e
one pa en was known (Table 2).
Resul s
Ac oss he yea s, we assigned 56.5% (26/46) o o sp ing o he
pu a i e mo he wi h a high deg ee o con idence (P.95%) in he
14 comple e amilies. Mo eo e , he pu a i e mo he s we e
con i med as he gene ic mo he s o o sp ing o he as majo i y
o cases (93.5%, 43/46) wi h .80% con idence (Table 3). O he
14 b oods, 12 wi h wo o mo e chicks, all o sp ing we e assigned
o he same pu a i e mo he .
O he h ee unassigned nes lings, ma e ni y o one could no
be assigned wi h .80% con idence and i was om a b ood o
ou nes lings. Since he a he o he ma e nally unassigned
nes ling was con i med in he la e pa e ni y analysis, addi ional
ma e ni y analysis wi h known a he did no assign i any gene ic
mo he wi h .80% con idence ei he . In bo h analyses wi h
unknown and known a he s, he nes ling could no be assigned o
any ma e ni y, nei he he pu a i e mo he , no o any o he
candida e adul emales in he popula ion. The pu a i e mo he /
o sp ing pai , iden i ied as ha ing mo e han wo geno ype
misma ches, we e cha ac e ized by nega i e LOD sco es. Thus,
he ma e ni y o he nes ling could no be esol ed and he ue
gene ic mo he s we e unlikely o ha e been sampled as all
candida e emales we e excluded a leas on basis o misma ches in
Figu e 2. LOD sco e dis ibu ions om CERVUS analyses o
pa en age in h ee- oed woodpecke s. (a) LOD sco e o he
candida e males ha ha e been unequi ocally assigned as a he s (&)
and o candida e males ha we e he second mos likely candida e
excluded (%)(N= 80); (b) LOD sco e o he candida e emales ha
ha e been unequi ocally assigned as mo he s (&) and o candida e
emales ha we e he second mos likely candida e excluded (%)
(N= 46); (c) LOD sco e o pai male wi h o sp ing assigned o ex apai
a he (&) and pai emale wi h o sp ing assigned o ex apai mo he
(%)(N= 5).
doi:10.1371/jou nal.pone.0007895.g002
Figu e 1. His og am showing he equency dis ibu ion o
misma ches be ween each h ee- oed woodpecke nes ling
and i s pu a i e a he (&) and mo he (%).
doi:10.1371/jou nal.pone.0007895.g001
Ma ing Sys em o a Woodpecke
PLoS ONE | www.plosone.o g 3 No embe 2009 | Volume 4 | Issue 11 | e7895
wo loci. These esul s also do no exclude he possibili y ha he
o sp ing may ha e esul ed om ex apai e iliza ion.
Nex , we a emp ed o assign pa e ni y o he 46 o sp ing om
14 comple e amilies including he o sp ing wi h unassigned
ma e ni y. In he assignmen analyses, he pu a i e a he was no
success ully assigned o wo o 46 o sp ing in wo o 12 b oods,
while he emaining o sp ing could be assigned o a candida e
male wi h .95% con idence (Table 3). Fo he 12 amilies (34
Table 2. C i ical DLOD sco es and ac ual and p edic ed success a e o en mic osa elli e loci used o assign pa en age.
Ma e ni y assignmen Pa e ni y assignmen
95% con idence 80% con idence 95% con idence 80% con idence
Yea DLOD Ra e DLOD Ra e DLOD Ra e DLOD Ra e
2004 1.79 55 (60) 0 100 (100) 1.53 89 (95) 0 100 (100)
2005 1.44 67 (75) 0 100 (100) 1.92 88 (94) 0 100 (100)
2006 2.01 57 (67) 0 100 (100) 2.66 87 (94) 0 100 (100)
2007 2.33 49 (53) 0 75 (100) 5.14 85 (90) 2.94 95 (100)
Calcula ions we e pe o med ac oss he samples, exp essed as pe cen age o o al numbe o indi iduals analysed (p edic ed success a es in pa en heses).
doi:10.1371/jou nal.pone.0007895. 002
Table 3. De ails o pa en age assignmen analysis using CERVUS, including he sampling yea , he nes iden i y, he numbe o
nes lings in he nes (n), he a ailabili y (+) and una ailabili y (2) o social a he s (=) and social mo he s (R), he numbe o
nes lings assigned o he social a he s and social mo he s using CERVUS wi h 80% con idence, he a es o ex a-pai ma e ni y
(QP) and ex a-pai pa e ni y (EPP).
CERVUS
Yea Nes iden i y n =R Ma e ni y assigned o RPa e ni y assigned o =QP EPP
2004 E o13 4 ++4/4 4/4 22
E o14 2 +22 2/2 22
E o15 4 ++4/4 4/4 22
E o16 3 ++3/3 3/3 22
E o19 3 ++3/3 3/3 22
E o20 3 +23/3 3/3 22
E o21 4 ++4/4 4/4 22
E o22 2 +22 2/2 22
E o23 3 ++3/3 3/3 22
2005 E o24 4 ++2/4 4/4 2/4
E o25 3 ++3/3 3/3 22
E o27 2 +22 2/2 22
E o28 2 ++2/2 2/2 22
E o34 3 +22 3/3 22
E o37 3 ++3/3 3/3 22
2006 E o40 3 ++3/3 2/3 21/3
E o41 4 +22 4/4 22
E o42 3 +22 3/3 2
E o44 3 ++3/3 3/3 22
E o45 2 +22 2/2 22
2007 E o48 3 +22 3/3 22
E o49 3 ++22/3 21/3
E o51 4 ++3/4 4/4 1/4 2
E o52 4 +22 4/4 22
E o53 3 +22 3/3 22
E o54 3 +22 3/3 22
To al 26 80 26 14 43/46 78/80 3/80 2/80
doi:10.1371/jou nal.pone.0007895. 003
Ma ing Sys em o a Woodpecke
PLoS ONE | www.plosone.o g 4 No embe 2009 | Volume 4 | Issue 11 | e7895
nes lings) om which he pu a i e a he s alone had been sampled,
pu a i e a he s we e assigned o all o sp ing wi h a .80%
con idence. Thus, he pu a i e a he was excluded o wo (2.5%)
o 80 nes lings in wo (7.7%) o 26 b oods. When all o he po en ial
candida e adul males we e es ed agains he wo ex a-pai chicks
wi h known mo he s, he ue gene ic a he o one chick was
de ec ed wi h .95% con idence. Howe e , no male eme ged as a
likely candida e a he o he o he nes ling - none me e en an
80% con idence c i e ion.
O he nes lings, wo we e ound o be unassigned o any
ma e ni y o pa e ni y. Gene ic sampling o adul males and
emales a he s udy a ea was no comple e in he yea s, and we
suspec — by analogy wi h o he published s udies (e.g. [22], [23])
- ha he gene ic pa en s o he wo ex a-pai nes lings we e
esiden , unsampled e i o ial adul s.
Discussion
The main aim o his s udy was o gain insigh in o he ma ing
sys em o he h ee- oed woodpecke s wi h he aid o igo ous
s a is ical analyses ( iz. he ex emely high cumula i e exclusion
p obabili y, dis inc misma ch be ween social pa en s and ex a-
pai o sp ing and he powe ul likelihood-based app oach) o
gene ic da a. The esul s p o ide he i s gene ic e idence o he
co-occu ence o polyand y and polygamy in he h ee- oed
woodpecke . To his end, hey add o ou unde s anding o
b eeding beha iou o non-passe ine bi ds, and o an inc easing
numbe o s udies epo ing occu ence o ex a-pai e iliza ions
in na u al bi d popula ions (see [2], [4]).
Ex apai Pa e ni y o Th ee-Toed Woodpecke s
This is he i s gene ic s udy showing ha EPP and QP occu
wi hin a single woodpecke species. O e all, howe e , his species
is p edominan ly gene ically monogamous. Since ma e swi ching
wi hin a b eeding season has ne e been isually obse ed in his
species (see [5], [7]), EPPs and QPs p edominan ly esul om
ex apai copula ions. Ma e swi ching can, howe e , occu o
example when male o emale dies acciden ally du ing he ea ly
b eeding season.
The equency o ex apai pa e ni y a ies ma kedly wi hin and
be ween species (see [2]). Ou poin es ima e o he p opo ion o
EPP is 2.5% which is much less han he a e age o ca. 11% in
passe ines, bu mo e equen han in some o he gene ically
monogamous species such as he New Zealand saddlebacks
Philes u nus ca uncula us and obins Pe oica aus alis whe e no EPP
has e e been de ec ed [24]. Compa a i e s udies sugges ha
many ac o s such as phylogene ic his o y, b eeding synch ony and
b eeding densi y, demands o pa e nal ca e, he a e o adul
mo ali y as well as he in ensi y o sexual con lic s all in luence he
cos s and bene i s o ex apai copula ions, and he e o e,
con ibu e o he a ia ion in EPC equency among species (see
[2]). In he con ex o his s udy he ques ion becomes: wha migh
keep ex a-pai e iliza ion a es low in h ee- oed woodpecke s as
compa ed o he a e age ex apai pa e ni y a e o ca. 11% o
e.g. passe ines [2]? We p edic ha he g ea es po en ial o he
need o pa e nal ca e hypo hesis will be in explaining he
di e ences in he le el o EPP among he species because he high
male in es men in b ood ca e is essen ial o emale ep oduc i e
success. In h ee- oed woodpecke s, males alloca e signi ican ly
mo e ime o e i o y de ence, ca i y exca a ion and eeding
young han emales [5]. Noc u nal incuba ion and b ooding as
well as nes cons uc ion also cons ain males wi h espec o social
polygamy (e.g. [6]). In addi ion, since he e a e signi ican sex
di e ences in he p o ision o a ious ypes o ca e and he o al
du a ion o di e en componen s o ca e, hese di e ences could
be ano he possible beha iou al explana ion [25] o he low
ex apai pa e ni y obse ed he e.
In aspeci ic a ia ion in he equency o EPP can occu a bo h
a he spa ial (e.g. he house spa ow Passe domes icus, [26]) and
empo al le els (e.g. he ed-winged blackbi d Agelaius phoeniceus,
[27]). Recen s udies o ex apai pa e ni y ound a somewha bu
no signi ican ly (Fishe ’s exac es ; P= 0.37) highe a e o EPP
(3.625.5%) in a Ge man popula ion o h ee- oed woodpecke s
[5], [7]. Al hough ecological ac o s could explain di e en le els
o EPP in h ee- oed woodpecke s obse ed in his and ea lie
s udies [5], [7], some addi ional po en ial explana ions may be
e oked. Fi s ly, he ea lie s udies may ha e lowe s a is ical powe
due o he smalle sample size (n= 55 chicks, 95% CI:
41.79263.21), he lowe - esolu ion molecula ools (mul i-locus
DNA inge p in ing) and s a is ical me hods (exclusion-based
analysis) employed (see [28]). Secondly, oppo uni ies o adop
al e na i e ep oduc i e s a egies may di e be ween popula ions
(see [29]), o ins ance due o habi a di e ences be ween he
Ge man and Finnish popula ions: s udy a ea o he Ge man
popula ion is 60022700 me e s abo e sea le el [5], while he
a e age al i ude o he s udy a ea o he Finnish popula ion is ca.
130 me e s [5], [7], [10]. Thi dly, spa io empo al luc ua ions in
popula ion densi y and esou ces a e likely o induce empo al
a ia ion in EPP equency. Howe e , u he s udies a e needed
o inden i y p oxima e and ul ima e de e minan s o EPP
occu ence in he species.
Quasi-Pa asi ism o Th ee-Toed Woodpecke s
We ound a low (3.8%) equency o ex apai ma e ni y
esul ing om quasi-pa asi ism. This has a ely been epo ed in
he ela ed woodpecke s such as he lesse spo ed woodpecke
Dend ocopos mino [30]. A numbe o explana ions ha e been pu
o wa d o explain he occu ence o QP [31]. The ‘ emale-d i en
QP’ sugges s ha a emale may choose, o assen a e an
app oach, o copula e wi h an ex apai male and goes on o lay
one o mo e o he eggs in his nes . This op ion implies ha
emales selec ‘high quali y’ males o e ilize hei eggs, and ei he
a oid he cos s o pa en al ca e associa ed wi h p o isioning some
young o bene i om he chosen males’ ‘good genes’ o di ec ly
om beha iou al o o he con ibu ions such as e i o y quali y
(e.g. [32]). Ano he one o he main hypo hesis sugges s ha QP in
non-passe ine bi ds is an insu ance mechanism agains he
po en ial de imen al e ec o inb eeding, o mo e simply, males’
own low quali y ma e [31]. Ne e heless, gi en he low le el o QP
in his s udy popula ion i seems unlikely ha QP is an inb eeding
a oidance s a egy as high le els o ex apai copula ions would be
expec ed in such a case (c . [33]). Fu he mo e, ou da a is hin
abou he ac ual ela edness be ween pa ne s, making i di icul
o es he inb eeding a oidance hypo hesis wi h much con idence.
In e es ingly, he a e o QP o he emales in h ee- oed
woodpecke s (3.75%) is a he lowe ange o es ima es epo ed
o many sho ebi ds (e.g. Common sandpipe . Ac i is hypoleucos,
5.7%, [1]) and passe ines (e.g. Sand ma in Ripa ia ipa ia, 2.4%,
[31]). The limi s o QP could a ise om he species cha ac e is ics
such as a high deg ee o male pa en al ca e, long du a ion o ca i y
exca a ion (and hus a na ow ime ame o e iliza ion), long
day- ime incuba ion and b ooding shi s (mo e han 3 hou s, [34]),
and ew e-ma ing oppo uni ies [5], all o which a e likely o
cons ain bo h males and emales in hei abili y o ob ain
addi ional ma es, and also limi hei abili y o seek ex apai
pa ne s. Howe e , he es ima es o bo h EPP and QP ob ained
he e should be conside ed wi h cau ion. One po en ial ca ea is
ha we did no sample unha ched eggs o dead chicks. This a ec s
Ma ing Sys em o a Woodpecke
PLoS ONE | www.plosone.o g 5 No embe 2009 | Volume 4 | Issue 11 | e7895

he es ima es by making hem conse a i e unde he assump ion
ha mo ali y be o e sampling is andom in espec o EPP and
QP. Ano he possible bias in he es ima es comes om he idea
ha nes s wi h QP may be mo e hea ily p eda ed i emales de end
hem less igo ously. I is also wo h no ing ha since he
es ima ed a es o EPP and QP a e jus based on ela i ely ew
nes lings sampled, he con idence limi s o hese es ima es a e
p obably b oad and hence he es ima es a e po en ially imp ecise.
We de ec ed no case o conspeci ic b ood pa asi ism (CBP) due
o egg dumping in his s udy while a single case o CBP, as a esul
o egg-dumping o QP, has been epo ed om a Ge man
popula ion o h ee- oed woodpecke s [5]. This sugges s ha he
CBP s emming om egg dumping mus be a e in ou s udy
popula ion. O e all, he wo popula ions did no di e signi ican ly
in he equency o ex apai o sp ing (Fishe ’s exac es ; P= 0.41)
o p opo ion o b oods con aining one o mo e ex apai young
(Fishe ’s exac es ; P= 0.61).
In conclusion, ou esul s o gene ic analysis ound he co-
occu ence o low le els o EPP and QP in he h ee- oed
woodpecke s. Al hough al e na i e explana ions may exis o he
obse a ions in ou s udy species (see [2], [4]), ou da a a e
consis en wi h he hypo hesis ha a high deg ee o male pa en al
ca e play an impo an ole in explaining low a es o EPP and QP
ac oss species. The in o ma ion p o ided in his s udy u he
allows us o examine he success o male and emale ma ing
pa e ns, as well as o unde s and he e olu iona y signi icance and
li e-his o y co ela es o p omiscui y in bi ds.
Acknowledgmen s
We hank Toni Nu mi, Johanna Lakka, Ja mo Pii oinen and Ville
Vepsa¨la¨inen o hei in aluable help in collec ing he ield da a. We a e
g a e ul o Ma ika Ka jalainen, Sonja Jaa i and Henna Fab i ius o
echnical assis ance in labo a o y. John Loeh kindly commen ed on an
ea lie e sion o his pape .
Au ho Con ibu ions
Concei ed and designed he expe imen s: MHL. Analyzed he da a: MHL.
W o e he pape : MHL JM. Pe o med he labo a o y wo k: KV.
Conduc ed he ieldwo k: MP TP.
Re e ences
1. Blomq is D, Ande sson M, Ku¨ppe C, Cu hill IC, Kis J, e al. (2002) Gene ic
simila i y be ween ma es and ex a-pai pa en age in h ee species o sho ebi ds.
Na u e 419: 613–615.
2. G i i h SC, Owens IPF, Thuman KA (2002) Ex a pai pa e ni y in bi ds: a
e iew o in e speci ic a ia ion and adap i e unc ion. Mol Ecol 11: 2195–2212.
3. Wes nea DF, S ewa IRK (2003) Ex a-pai pa e ni y in bi ds: Causes,
co ela es, and con lic . Annu Re Ecol E ol Sys 34: 365–396.
4. G i i h SC, Lyon BE, Mon gome ie R (2004) Quasi-pa asi ism in bi ds. Beha
Ecol Sociobiol 56: 191–200.
5. Pechacek P, Michalek KG, Winkle H, Blomq is D (2005) Monogamy wi h
excep ions: social and gene ic ma ing sys em in a bi d species wi h high pa e nal
in es men . Beha iou 142: 1093–1114.
6. Winkle H, Ch is ie DA (2002) Family Picidae (woodpecke s). In: del Hoyo J,
Ellio A, Sa ga al J, eds. Handbook o he Bi ds o he Wo ld. Ba celona: Lynx
Edicions. pp 296–555.
7. Pechacek P (2006) B eeding pe o mance, na al dispe sal, and nes si e ideli y o
he h ee- oed woodpecke in he Ge man Alps. Ann Zool Fennici 43: 165–176.
8. Richa dson DS, Ju y FL, Blaakmee K, Komdue J, Bu ke T (2001) Pa en age
assignmen and ex a-g oup pa e ni y in a coope a i e b eede : he Seychelles
wa ble (Ac ocephalus seychellensis). Mol Ecol 10: 2263–2274.
9. McRae SB, Amos W (1999) Can inces wi hin coope a i e b eeding g oups be
de ec ed using DNA inge p in ing? Beha Ecol Sociobiol 47: 104–107.
10. Pakkala T, Hanski I, Tomppo E (2002) Spa ial ecology o he h ee- oed
woodpecke in managed o es landscapes. Sil a Fenn 36: 279–288.
11. Pakkala T, Kouki J, Tiainen J (2006) Top p eda o and in e e ence compe i ion
modi y he occu ence and b eeding success o a specialis species in a
s uc u ally complex en i onmen . Ann Zool Fennici 43: 137–164.
12. Va¨lima¨ki K, Jaa i S, Piha M, Pakkala T, Me ila¨ J (2008) Isola ion and
cha ac e iza ion o 17 polymo phic mic osa elli e loci o he h ee- oed
woodpecke (Picoides idac ylus). Mol Ecol Resou 8: 1152–1154.
13. Sla e J, Ma shall T, Pembe on J (2000) A e ospec i e assessmen o he
accu acy o he pa e ni y in e ence p og am CERVUS. Mol Ecol 9: 801–808.
14. Ma shall TC, Sla e J, K uuk LEB, Pembe on JM (1998) S a is ical con idence
o likelihood-based pa e ni y es ing in e ence in na u al popula ions. Mol Ecol
7: 639–655.
15. Pechacek P, K is in A (2004) Compa a i e die s o adul and young h ee- oed
woodpecke s in a Eu opean alpine o es communi y. J Wildl Manage 68:
683–693.
16. Hoy JS, Hannon SJ (2002) Habi a associa ions o black-backed and h ee- oed
woodpecke s in he bo eal o es o Albe a. Can J Fo Res 32: 1881–1888.
17. F idol sson AK, Elleg en H (1999) A simple and uni e sal me hod o molecula
sexing o non- a i e bi ds. J A ian Biol 30: 116–121.
18. Guo SW, Thompson EA (1992) Pe o ming he exac es o Ha dy-Weinbe g
p opo ion o mul iple alleles. Biome ics 48: 361–372.
19. Raymond M, Rousse F (1995) GENEPOP: A popula ion gene ics so wa e o
exac es s and ecumenicism. J He ed 86: 248–249.
20. Nei M (1987) Molecula E olu iona y Gene ics. New Yo k: Columbia
Uni e si y P ess.
21. Wei BS, Cocke ham CC (1984) Es ima ing F-s a is ics o he analysis o
popula ion s uc u e. E olu ion 38: 1358–1370.
22. Webs e MS, Chuang-Dobbs HC, Holmes RT (2001) Mic osa elli e iden i ica-
ion o ex apai si es in a socially monogamous wa ble . Beha Ecol 12:
439–446.
23. Hill CE, Pos W (2005) Ex a-pai pa e ni y in Seaside Spa ows. J Field O ni ho
76: 119–126.
24. Sab ina ST, Boessenkool S, Jamieson IG (2008) Gene ic monogamy in wo long-
li ed New Zealand passe ines. J A ian Biol 39: 579–583.
25. Benne PM, Owens IPF (2002) E olu iona y Ecology o Bi ds: Li e His o y,
Ma ing Sys ems and Ex inc ion. Ox o d: Ox o d Uni e si y P ess.
26. Owens IPF, Ha ley IR (1998) Sexual dimo phism in bi ds: why a e he e so
many di e en o ms o dimo phism? P oc R Soc Lond B 265: 397–407.
27. Wea he head PJ, Boag PT (1995) Pai and ex a-pai ma ing success ela i e o
male quali y in ed-winged blackbi ds. Beha Ecol Sociobiol 37: 81–91.
28. Jones AG, A den WR (2003) Me hods o pa en age analysis in na u al
popula ions. Mol Ecol 12: 2511–2523.
29. Wes nea DE, She man PW (1997) Densi y and ex a-pai e iliza ions in bi ds:
a compa a i e analysis. Beha Ecol Sociobiol 41: 205–215.
30. Wik ande U, Olsson O, Nilsson SG (2000) Pa en al ca e and social ma ing
sys em in he lesse spo ed woodpecke Dend ocopos mino . J A ian Biol 31:
447–456.
31. Al es MAS, B yan DM (1998) B ood pa asi ism in he sand ma in, Ripa ia
ipa ia: e idence om wo pa asi ic s a egies in a colonial passe ine. Anim Beha
56: 1323–1331.
32. Kempenae s B, Ve heeyen C, Van den B oeck M, Bu ke T, Van B oekho en C,
e al. (1992) Ex a-pai pa e ni y esul s om emale p e e ence o high quali y
males in he blue i . Na u e 357: 494–496.
33. Hughes JM, Ma he PB, Toon A, Ma J, Rowley I, e al. (2003) High le els o
ex a-g oup pa e ni y in a popula ion o Aus alian magpies Gymno hina ibicen:
e idence om mic osa elli e analysis. Mol Ecol 12: 3411–3450.
34. Ruge K (1971) Zu Biologie des D eizehenspech es Picoides idac ylus.
Beobach ungen wa¨h end de B u zei . De O ni hologische Beobach e 68:
256–271.
Ma ing Sys em o a Woodpecke
PLoS ONE | www.plosone.o g 6 No embe 2009 | Volume 4 | Issue 11 | e7895