RESEARCH ARTICLE
Habi a E ec s on he B eeding Pe o mance
o Th ee Fo es -Dwelling Hawks
Heidi Bjö klund
1,2
*, Ja i Valkama
1
, E kki Tomppo
3
, Toni Laaksonen
4
1The Zoology Uni , Finnish Museum o Na u al His o y Luomus, Uni e si y o Helsinki, Helsinki, Finland,
2Depa men o Biosciences, Uni e si y o Helsinki, Helsinki, Finland, 3Na u al Resou ces Ins i u e Finland,
Van aa, Finland, 4Depa men o Biology, Uni e si y o Tu ku, Tu ku, Finland
*heidi.bjo klund@helsinki. i
Abs ac
Habi a loss causes popula ion declines, bu he mechanisms a e a ely known. In he Eu o-
pean Bo eal Zone, loss o old o es due o in ensi e o es y is suspec ed o cause declines
in o es -dwelling ap o s by educing hei b eeding pe o mance. We s udied he bo eal
b eeding habi a and habi a -associa ed b eeding pe o mance o he no he n goshawk
(Accipi e gen ilis), common buzza d (Bu eo bu eo) and Eu opean honey buzza d (Pe nis
api o us). We combined long- e m Finnish bi d-o -p ey da a wi h mul i-sou ce na ional o -
es in en o y da a a a ious dis ances (100–4000 m) a ound he hawk nes s. We ound ha
b eeding success o he goshawk was bes explained by he habi a wi hin a 2000-m adius
a ound he nes s; b eeding was mo e success ul wi h inc easing p opo ions o old sp uce
o es and wa e , and dec easing p opo ions o young hinning o es . None o he habi a
a iables a ec ed signi ican ly he b eeding success o he common buzza d o he honey
buzza d, o he b ood size o any o he species. The amoun o old sp uce o es dec eased
bo h a ound goshawk and common buzza d nes s and h oughou sou he n Finland in
1992–2010. In con as , he a ea o young o es inc eased in sou he n Finland bu no
a ound hawk nes s. We emphasize he impo ance o s udying habi a s a se e al spa ial
and empo al scales o de e mine he ele an species-speci ic scale and o de ec en i on-
men al changes. Fu he e o is needed o econcile he socioeconomic and ecological
unc ions o o es s and habi a equi emen s o old o es specialis s.
In oduc ion
Un a ou able habi a changes a e he main h ea s o h ea ened species wo ldwide [1].
Ad e se an h opogenic habi a changes include habi a loss, de e io a ion and agmen a ion,
which can a ec ep oduc i e success and su i al o species [2]. In addi ion o hese di ec
impac s, habi a change can cause cascading e ec s among o be ween ophic le els h ough
in e speci ic in e ac ions [2,3]. Fo ins ance, inc easing in e speci ic compe i ion o high-qual-
i y habi a s may o ce a subdominan compe i o in o in e io habi a s [4]. Human-caused
en i onmen al change has hus he po en ial o a ec species in di e en ways.
PLOS ONE | DOI:10.1371/jou nal.pone.0137877 Sep embe 30, 2015 1/19
a11111
OPEN ACCESS
Ci a ion: Bjö klund H, Valkama J, Tomppo E,
Laaksonen T (2015) Habi a E ec s on he B eeding
Pe o mance o Th ee Fo es -Dwelling Hawks. PLoS
ONE 10(9): e0137877. doi:10.1371/jou nal.
pone.0137877
Edi o : Se gio A Lambe ucci, INIBIOMA
(Uni e sidad Nacional del Comahue-CONICET),
ARGENTINA
Recei ed: No embe 5, 2014
Accep ed: Augus 23, 2015
Published: Sep embe 30, 2015
Copy igh : © 2015 Bjö klund 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.
Da a A ailabili y S a emen : Due o e hical and
legal es ic ions conce ning endange ed o sensi i e
species in he s udy, da a a e a ailable upon eques .
All in e es ed esea che s may con ac D . Heidi
Bjö klund a heidi.bjo klund@helsinki. i o access o
da a.
Funding: This s udy was unded by he Kone
Founda ion (www.koneensaa io. i/en), he Doc o al
P og amme in Wildli e Biology Resea ch Luo a
(www.helsinki. i/luo a) and he Finnish Museum o
Na u al His o y Luomus (h p://www.luomus. i/en),
unds o HB. Adminis a ion o he Finnish Common
Changes in bo eal o es s caused by in ensi ica ion o o es y p ac ices since he 1960s a e
one o he g ea es ecen an h opogenic en i onmen al changes in No he n Eu ope [5].
Regene a ion cu ings, es ablishmen o new o es and o es managemen ha e led o o es
agmen a ion, dec eases in a eas o old-g ow h o es , small openings and o es i es, and
changes in o es age s uc u e and ee species composi ion [5–9]. S uc u al changes in o es s
esul ing om in ensi ied o es managemen (e.g. e en-aged s ands, ewe la ge ees wi h
hick b anches and emo al o decaying snags o mal o med ees) ha e a ec ed he quali y o
bo eal o es s as habi a s and si es o ep oduc ion o many axa [10–12].
The consequences o changes in o es s can be pa icula ly d ama ic o o es -dwelling
bi ds o p ey [13], because esou ces ( ood, nes si es) a e usually spa se o op ap o s [14]
and u he mo e, habi a change e ec s can escala e in ood webs. Ou s udy species, he no h-
e n goshawk (Accipi e gen ilis, he ea e goshawk) wi h ci cumbo eal dis ibu ion, and he
common buzza d (Bu eo bu eo) and Eu opean honey buzza d (Pe nis api o us, he ea e
honey buzza d) wi h Eu asian dis ibu ions, a e capable o inhabi ing di e se habi a s, includ-
ing coni e ous, deciduous and mixed o es s [15–22]. Ma u e o es and No way sp uce (Picea
abies) seem impo an o hese species [13,23–25]. Due o a he simila habi a equi emen s,
he species can compe e o e i o ies and nes si es in hei sha ed b eeding ange [26–29],
al hough he goshawk is dominan as i can ake o e he nes s o he o he wo species o e en
p eda e hem [16,26,30].
Popula ions o he goshawk ha e declined in No he n Eu ope [31] and in pa s o No h
Ame ica [20,25,32]. The common buzza d and he honey buzza d ha e o e all s able popula-
ions excep o hei long- e m declines in No he n Eu ope and a decline in he honey buz-
za d in Wes e n Eu ope [31,33–35]. These declines can pa ly be due o in ensi ied o es
managemen and i s consequences o p ey a ailabili y [25,36–40]. The common buzza d and
honey buzza d popula ions ha e dec eased s eeply in Finland and hese species (bu no he
goshawk) a e lis ed as ulne able in he Finnish Red Lis [40]. The goshawk popula ion has
only sligh ly declined in Finland, which has aised conce ns ha he dominan goshawk could
be eplacing he buzza ds om p ime nes si es [26,39].
Despi e he as a ay o b eeding habi a s udies on hese ap o s, mos o hem a e om
he Tempe a e Zone. Only a ew s udies ha e been accomplished in he Bo eo-nemo al Zone
[21,23] and only one on he goshawk in he coni e ous o es -domina ed Bo eal (Taiga) Zone
in Eu ope [41]. This de iciency is s iking, aking in o accoun ha hese hawks ha e wide-
sp ead dis ibu ions h oughou he la ge Eu asian Bo eal Zone [22]. Mo eo e , declining pop-
ula ion ends o he hawks in No he n Eu ope ha e aised conce ns o he s a e o he o es s
in hei b eeding g ounds [36,40].
B eeding habi a s udies a e o en conduc ed a a he small spa ial scales (a ound nes s, in
es ic ed s udy a eas), and ypically o e a sho ime pe iod (such as a single yea ). I would
be wo hwhile o s udy habi a composi ion also a la ge scales, e.g. a a landscape scale o he
nes ing si e, since hese a e o en biologically mo e meaning ul o species wi h la ge e i o ies
[25,36,42–45]. Addi ionally, i would be impo an o moni o empo al changes in habi a
s uc u e a hese wide scales [25], and o plan managemen s a egies in la ge geog aphic
a eas o he bene i o hawks [25,46,47]. Mode n me hods, such as emo e sensing combined
wi h a geog aphic in o ma ion sys em, can aid in ul illing hese needs.
We s udied he Eu opean Bo eal Zone b eeding habi a and habi a -associa ed b eeding pe -
o mance o he goshawk, common buzza d and honey buzza d. Ou wo k combines wo
unique da ase s: geog aphically wide-scale and long- e m Finnish hawk b eeding da a and he
ou pu da a om a sa elli e image–aided mul i-sou ce na ional o es in en o y (MS-NFI) a
a ious dis ances a ound he nes ees. This allowed us o ex end ou esea ch in se e al
dimensions: om he nes -si e le el o he e i o y and landscape le els, om a local a ea o a
Habi a E ec s on Fo es -Dwelling Hawks
PLOS ONE | DOI:10.1371/jou nal.pone.0137877 Sep embe 30, 2015 2/19
Bi ds-o -P ey Su ey was unded by he Finnish
Minis y o he En i onmen and he Finnish Museum
o Na u al His o y Luomus. The unde s had no ole in
s udy design, da a collec ion (excep he ole o
Luomus as an o ganise o he Common Bi ds-o -
P ey Su ey) 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 .
na ionwide geog aphic a ea and om ew yea s o 19 yea s. Speci ically, we aimed a in es iga -
ing he ollowing ques ions on mul iple spa ial scales: 1) is he b eeding pe o mance o he gos-
hawk, common buzza d and honey buzza d associa ed wi h hei b eeding habi a in No he n
Eu opean bo eal o es s, and 2) wha a e he ypical cha ac e is ics o he bo eal b eeding habi-
a s o he hawks, and ha e any empo al changes in he b eeding habi a s aken place du ing
he s udy pe iod ha would e lec landscape changes. We we e pa icula ly in e es ed in he
ole o old No way sp uce o es , since he impo ance o sp uce and ma u e o es o he
hawks has eme ged om o he s udies, and o es y p ac ices ha e pa icula ly a ec ed such
o es s [48].
Ma e ials and Me hods
S udy a ea
Da a on nes s we e included om he sou he n hal o Finland (land a ea app. 154,000 km
2
)
om an a ea ex ending app oxima ely 640 km om sou h o no h and 440 km om wes o
eas (S1,S2 and S3 Figs). In gene al, he landscape is low-lying (mean heigh 152 m) and mos
o he land a ea is domina ed by o es s ha a e managed (77%). Fo es s in sou he n Finland
a e domina ed by Sco s pine (Pinus syl es is, 56%), No way sp uce (31%) and b oadlea ed
ees (11%, mainly sil e bi ch Be ula pendula and downy bi ch B.pubescens)[49].
S udy species
The goshawk, common buzza d and honey buzza d a e middle-sized, o es -dwelling hawks
[22,50] ha build hei s ick nes s unde he c own laye [51]. Goshawks p eda e mainly o es
g ouse in Finland, bu also o he bi ds and mammals [37]. Mic o us oles, o es g ouse and
ha es a e he main p ey o common buzza ds [52], and honey buzza ds eed mainly on wasps
Vespidae, bu also on ogs and small bi ds [53]. Adul goshawks a e seden a y while he com-
mon buzza d subspecies B.b.bu eo (wes e n Finland) is a sho -dis ance mig an o cen al
Eu ope [54]. The subspecies B.b. ulpinus (eas e n Finland) and he honey buzza ds a e long-
dis ance mig an s o A ica [54].
Nes da a
Nes ca d da a on hawk nes s we e ga he ed by olun ee ap o inge s and en husias s since
1982 as a pa o he Finnish Common Bi ds-o -P ey Su ey coo dina ed by he Finnish
Museum o Na u al His o y Luomus. A nes ca d includes in o ma ion on nes loca ion, nes
ype, b eeding species and nes isi s o he inge [51]. Hawk e i o ies can ha e se e al al e -
na i e nes s which should be aken in o accoun in s a is ical analyses o a oid pseudo eplica-
ion [55], since nes s o he same e i o y ha e habi a s po en ially mo e simila han nes s
om di e en e i o ies. The e o e, each nes was p o ided wi h a e i o y iden i y code
(de ails in S1 Tex ).
We included nes s wi h a e i ied b eeding a emp ha was ensu ed by obse a ions o
eggs, eggshells, chicks o emnan s o chicks. This excludes occupied bu only deco a ed nes s,
and possible b eeding a emp s in which eggs o chicks may ha e disappea ed wi hou ace. In
o al, we had da a on he ollowing numbe s o b eeding a emp s: goshawk 1475 ( om 861
nes s), common buzza d 774 (529) and honey buzza d 166 (126).
Fo b eeding pe o mance analyses, we included b eeding a emp s wi h known esul s (S1,
S2 and S3 Figs) and disca ded nes s ha we e no isi ed a e he incuba ion o ea ly nes ling
pe iod. We quan i ied b eeding success on a bina y scale (success ul, unsuccess ul), and he
b eeding a emp was conside ed success ul when a leas one chick was aised o inging age
Habi a E ec s on Fo es -Dwelling Hawks
PLOS ONE | DOI:10.1371/jou nal.pone.0137877 Sep embe 30, 2015 3/19
(14–28 days old). Fo b ood size analyses, we included only hose nes s wi h an accu a e num-
be o chicks [51].
E hics S a emen
Rap o inge s and en husias s ollowed he guidelines o Finnish Ringing Cen e a he Finnish
Museum o Na u al His o y Luomus. Acco ding o hese guidelines, unnecessa y nes isi s
and nes climbing should be a oided du ing he b eeding season in o de o minimize dis u -
bance. Ringing licences we e issued by The Cen es o Economic De elopmen , T anspo and
he En i onmen . Hawk nes s we e loca ed on di e en land ypes (p i a e, s a e o company
owned lands) which a e accessible acco ding o Finnish public igh o access. Nes s a p o-
ec ed a eas we e examined wi h a speci ic licence issued by The Cen es o Economic De el-
opmen , T anspo and he En i onmen , o by Me sähalli us. No samples o p o ec ed species
we e aken o his s udy.
Mul i-sou ce na ional o es in en o y (MS-NFI) da a
Ou habi a da a we e based on MS-NFIs o he Na u al Resou ces Ins i u e Finland. The
MS-NFI da a is a combina ion o in o ma ion om sa elli e images (Landsa Thema ic Map-
pe , TM), ield plo s o Finnish na ional o es in en o ies (NFIs) and o he geo e e enced digi-
al da a [56,57], u he de ails on he MS-NFI da a in S2 Tex .
We used he MS-NFI da a (he ea e habi a da a) om ou MS-NFI pe iods (S2 Tex ) and
we ma ched he yea o habi a da a (sa elli e image yea ) wi h he yea o b eeding da a o
each nes . Addi ionally, we gene alized he habi a da a o wo p eceding and wo subsequen
b eeding yea s (bu no be o e 1992, see S2 Tex ). Fo ins ance, i he habi a da a a ound a nes
we e om he yea 1999, hese da a we e used o he b eeding yea s 1997–2001 in his nes . As
a esul o he gene aliza ion o habi a da a om ou MS-NFI pe iods, we had b eeding da a
on ou b eeding pe iods ha we he ea e e e o as he i s , second, hi d and ou h b eed-
ing pe iods. The i s b eeding pe iod co e ed b eeding yea s 1992–2004 (median b eeding yea
o all species combined 1998), he second b eeding pe iod 2002–2007 (2004), he hi d b eeding
pe iod 2005–2008 (2008) and he ou h b eeding pe iod 2008–2010 (2009). Since he numbe
o honey buzza d nes s was low in he ou h b eeding pe iod, we combined he hi d and
ou h b eeding pe iods as he las b eeding pe iod (median b eeding yea o all species 2008).
We e ie ed habi a da a in ci cles wi h adii o 100 m, 250 m, 500 m, 1000 m, 2000 m and
4000 m a ound each nes (see [41,45,58] o simila adii) ha co esponded o a eas o 3.14
ha, 19.6 ha, 78.5 ha, 314.1 ha, 1256.6 ha and 5026.4 ha, espec i ely. Habi a da a om di e en
scales ensu ed ha we would include all po en ially impo an habi a composi ion scales o
each species.
Each pixel o he aw habi a da a was classi ied in o one o se en biologically ele an habi-
a classes (de ails in S2 Tex ): 1) sp uce-domina ed o es wi h ee s em olume 150 m
3
ha
-
1
(he ea e old sp uce o es ); 2) o he o es wi h ee s em olume 150 m
3
ha
-1
(he ea e
o he old o es , hal o which consis s o pine o es a all scales); 3) young hinning o es wi h
ee s em olume 60 m
3
–<150 m
3
ha
-1
(young g owing s ock a he hinning cu ings s age
[9]); 4) low s ocking o es wi h ee s em olume om ze o ( eeless pea land o logged a ea)
o <60 m
3
ha
-1
(young seedling, seedling o seed ee s and, o pea land wi h low numbe o
ees); 5) wa e ; 6) a able land (o which ce eals 52%, cul i a ed g asslands 29% and allow
a eas 11%; [59]); and 7) buil -up land (se lemen , oad o pea p oduc ion a ea). We use he e
he e ms ‘old sp uce o es ’and ‘o he old o es ’ o b e i y, bu i should be no ed ha he
co ela ion be ween ee age and size is no pe ec [60,61]. By ‘old sp uce o es ’o ‘o he old
o es ’we do no e e exclusi ely o na u al, old-g ow h o es , which a e a e in he s udy a ea
Habi a E ec s on Fo es -Dwelling Hawks
PLOS ONE | DOI:10.1371/jou nal.pone.0137877 Sep embe 30, 2015 4/19
[62]. Howe e , ou limi o ee s em olume 150 m
3
ha
-1
e e s o ad anced o ma u e o -
es , since he mean s ock olume o ad anced hinning s ands is 163 m
3
ha
-1
and ha o ma u e
s ands 207 m
3
ha
-1
in Finland (NFI 11: 2009–2012 [48]). We excluded clouds and a eas wi hou
habi a da a om he ci cula a eas and hen calcula ed he p opo ion o each habi a class in
each a ea.
We used log- a io ans o ma ion o he habi a a iables in he s udy ques ion 1 (me hod
o [63] desc ibed in [64,65]). This was done o pu ge he mu ual co ela ion o he habi a p o-
po ions because he sum o he habi a p opo ions is 1 (see S3 Tex ).
The habi a a iable es ima es wi hin a adius con ain some e o s (e o sou ces lis ed in
[62], p. 91 onwa ds, see also S2 Tex ) ha dec ease when he size o he s udy a ea inc eases.
Es ima ion o hese e o s and hei inco po a ion in subsequen s a is ical analyses is a e y
complex and ambiguous issue [62,66], which we a oided in his s udy. Impac s o e o s a e
po en ially p onounced when e alua ing empo al changes in habi a es ima es (di e en
model-based habi a da a es ima es); in such cases, compa isons can be done o a pa icula
a ea, using NFI ield da a only [62].
S a is ical analyses
1) How is b eeding habi a associa ed wi h b eeding pe o mance? We included da a
om he nes s om all b eeding pe iods o b eeding success and b ood size analyses. We
in es iga ed he in luence o log- a io habi a a iables on b eeding success and b ood size wi h
gene alized linea mixed models (GLMMs) wi h he e i o y iden i y as a andom e ec . We
p e e ed e i o y iden i y ins ead o nes iden i y as a andom e ec , because he landscape
a ound al e na i e nes s o he same e i o y is p esumably simila and he e o e he isk o
pseudo eplica ion would s ill exis a e conside ing nes iden i y as a andom e ec . The
b eeding success GLMMs o he goshawk and honey buzza d, assuming a binomial (wi h logi
link) dis ibu ion, we e hen o he o m ( ollowing [67]):
Yij Binð1;pijÞ
log i ðpijÞ¼aþb1log ðold sp uce o es Þij þb2log ðyoung hinning o es Þij
þb3log ðlow s ocking o es Þij þb4log ðwa e Þij þb5log ða able land Þij
þb6log ðbuil up land Þij þai
aiNð0;s2
aÞ
ð1Þ
whe e Y
ij
is 1 i nes j on e i o y i has a success ul b eeding a emp ; o he wise Y
ij
is 0. Log is
he log- a io o he habi a a iable and is he adius. We assumed he andom in e cep a
i
o
he e i o y iden i y o be no mally dis ibu ed (mean 0, a iance s2
a). We ound ea lie wi h a
longe - e m da ase ha he b eeding success o he common buzza d dec eases owa ds he
no h in Finland [51]. Thus, we added la i ude o he b eeding success model (1) o he com-
mon buzza d.
The b ood size GLMMs assumed a Poisson (log link) dis ibu ion and included he same
explana o y a iables as in he model 1.
Since he app op ia e scale was unknown, we i s in es iga ed, a which scale he habi a
composi ion in luences mos he b eeding pe o mance o each species. The e o e, we i ed
ou GLMMs o bo h b eeding success and b ood size, using habi a a iables wi hin he adii
o 100 m, 500 m, 1000 m and 2000 m. We s anda dized he model a iables o mean = 0 and
s anda d de ia ion (SD) = 0.5 [68]. We compa ed he Akaike in o ma ion c i e ion (AIC) al-
ues o he models wi h habi a da a a di e en scales, and chose o each species he model
Habi a E ec s on Fo es -Dwelling Hawks
PLOS ONE | DOI:10.1371/jou nal.pone.0137877 Sep embe 30, 2015 5/19

wi h he lowes AIC as he model bes explaining b eeding success o b ood size [69]. I se e al
models we e almos equally good (AIC-di e ence o he bes model 2[68,69]), we inspec ed
whe he he pa ame e es ima es o each o hese op 2AIC models ga e he same in o ma ion
han he bes model. We u he educed he bes model o each species i he log- a io habi a
a iables we e highly co ela ed (| | >0.7 [70]). We used he likelihood- a io es o decide
which o he co ela ed a iables could be d opped. In cases whe e bo h a iables could be
d opped, we disca ded he a iable whose emo al led o he lowes AIC. Model i was e alu-
a ed g aphically, and wi h pa ame ic boo s apping [71]. We es ed he esiduals o he bes
models o spa ial au oco ela ion wi h global Mo an’sI[72] and ound no spa ial
au oco ela ion.
When i ing models o small coun s (such as b ood size o 1–5), bu excluding ze o (b ood
size 0) om possible alues may po en ially bias he pa ame e es ima es o he model [67],
while models including ze o (b ood sizes 0–5) could con ain he same in o ma ion al eady cap-
u ed by b eeding success models (since unsuccess ul nes s a e he ones wi h b ood size 0). To
e i y he b ood size model esul s, we i ed ze o- unca ed gene alized linea models (ze o-
unca ed GLMs) wi h uns anda dized log- a io habi a a iables, using he same a iable se s
as in he bes models. The ze o- unca ion app oach adjus s he pa ame e es ima es by aking
in o accoun he exclusion o ze o [67].
2) Cha ac e is ics and changes in he bo eal b eeding habi a ? We included nes s om
he i s and las b eeding pe iods, and quan i ied i s he b eeding habi a p opo ions in he
wo pe iods a all scales ( adii o 100–4000 m). Fo he analyses o changes in he b eeding hab-
i a , we chose he adii 100 m, 1000 m and 2000 m, whe e 100 m ep esen s he nes -si e scale,
1000 m he e i o y co e scale and 2000 m he e i o y on a b oad scale. The habi a p opo -
ions we e a csine squa e- oo - ans o med; his ans o ma ion is commonly used o p opo -
ions [73].
We analysed he di e ences in habi a p opo ions be ween he i s and las b eeding pe i-
ods o each species wi h linea mixed-e ec models, in which he dependen a iable was an
a csine squa e- oo - ans o med habi a p opo ion, and he explana o y a iable was he
b eeding pe iod. We included e i o y iden i y as a andom e ec wi h a andom in e cep . I
necessa y, we allowed a di e en a iance o he wo b eeding pe iods. We adjus ed he
h eshold o a signi ican p- alue wi h a Bon e oni co ec ion.
We con as ed o es habi a changes a ound hawk nes s wi h o e all changes in o es s in
sou he n Finland using esul s o he NFI ield da a o he compa ison [48]. The habi a classi-
ica ion in NFIs sligh ly di e om ha o ou s, bu we used a classi ica ion ha bes ma ched
wi h ou habi a classes. We examined in he NFI da a he changes in he a ea o >60-yea -old
sp uce-domina ed o es (he ea e ‘olde sp uce-domina ed o es ’; his co esponds oughly
o ou old sp uce o es ), o >60-yea -old pine and deciduous o es (‘o he olde o es ’, com-
pa es oughly wi h ou o he old o es ), and o 21–60-yea -old young o es (‘young o es ’,
co esponding app oxima ely o ou young hinning o es ) in sou he n Finland be ween NFI-
9 (1996–2003) and NFI-11 (2009–2012).
Resul s
1) Associa ion o b eeding pe o mance wi h he habi a
B eeding success. The o e all b eeding success was high in ou nes ca d da a. The p o-
po ion o success ul b eeding a emp s was 89.8% o he goshawk (N = 1454 b eeding
a emp s), 90.9% o he common buzza d (N = 762) and 87.0% o he honey buzza d
(N = 161). Mos b eeding ailu es occu ed a he egg-s age (S3 Table). Since nes ca ds may
o e es ima e success ul b eeding a emp s (ea ly ailu es a e missed o nes ca ds a e illed
Habi a E ec s on Fo es -Dwelling Hawks
PLOS ONE | DOI:10.1371/jou nal.pone.0137877 Sep embe 30, 2015 6/19
mo e o en om success ul b eeding a emp s), we p o ide espec i e p opo ions o success ul
b eeding a emp s om he Rap o Ques ionnai e da a o he Finnish Common Bi ds-o -P ey
Su ey in 1986–2014 o a compa ison: 86.7% o he goshawk (N = 20928 b eeding a emp s),
88.7% o he common buzza d (N = 10111), and 80.9% o he honey buzza d (N = 2077,
Finnish Museum o Na u al His o y Luomus).
The b eeding success o he goshawk was bes explained by habi a p opo ions a he 2000
m adius scale. Fo he common buzza d, he bes model was he one wi h habi a p opo ions
wi hin 100 m. The model wi h habi a p opo ions wi hin 100 m was he bes also o he
honey buzza d, bu habi a p opo ions a he 1000 m scale we e almos as good in explaining
honey buzza d b eeding success.
The p opo ions o old sp uce o es (Fig 1A) and wa e wi hin 2000 m we e signi ican ly
and posi i ely associa ed wi h he b eeding success o he goshawk, whe eas he p opo ion o
young hinning o es was signi ican ly and nega i ely associa ed wi h goshawk b eeding suc-
cess (Fig 1B,Table 1). To illus a e he esul s on he biologically in e es ing o iginal habi a
p opo ion scale, we i ed addi ional b eeding success GLMMs in which he only explana o y
a iable was each o he un ans o med habi a a iable p opo ion in u n a he 2000 m adius
scale. Also he e, he p opo ion o old sp uce o es showed a signi ican posi i e and he p o-
po ion o young hinning o es a signi ican nega i e associa ion wi h goshawk b eeding suc-
cess (S4 and S5 Figs). The p opo ion o buil -up land had a signi ican posi i e associa ion
wi h goshawk b eeding success, whe eas he p opo ion o wa e and a able land we e no sig-
ni ican ly associa ed wi h goshawk b eeding success.
None o he habi a p opo ions o he bes models we e signi ican ly associa ed wi h he
b eeding success o he common buzza d o honey buzza d (Table 1). We also inspec ed he
second-bes b eeding success model o he honey buzza d (wi hin 1000 m), bu he in e p e a-
ion was he same; none o he habi a a iables we e signi ican ly associa ed wi h b eeding
success.
B ood size. The a e age b ood size in success ul nes s was 2.89 (SD 0.86, N = 1167) o he
goshawk, 2.26 (0.85, N = 592) o he common buzza d, and 1.79 (0.41, N = 121) o he honey
buzza d. Two b ood size models we e included in he op 2AIC o bo h he goshawk (wi h
habi a p opo ions wi hin 2000 m and 1000 m) and common buzza d (wi hin 2000 m and 500
m) while each o he ou honey buzza d b ood size models had a simila AIC (models wi h he
lowes AIC in S1 Table). We e i ied he pa ame e s o each compe ing model wi hin he op
2AIC, and hey made no change o he in e p e a ion o he a iables.
None o he habi a a iables we e signi ican ly associa ed wi h he b ood size o any o he
hawks in he bes models (S1 Table), o in he compe ing models. The esul s o he ze o-
unca ed GLMs (no shown) we e in gene al simila o hose o he GLMMs, con i ming he
non-signi ican associa ion o he habi a a iables wi h he b ood size.
2) B eeding habi a and habi a changes
Habi a p opo ions a all scales a ound he nes s a e shown in Fig 2A–2C o he i s b eeding
pe iod, and in h ee scales o he i s and las b eeding pe iods (S2 Table).
The p opo ion o old sp uce o es was highes o he goshawk a he nes -si e scale (100
m) and dec eased g adually wi h inc easing adius (Fig 2A,S2 Table, see also S4 Tex ). The
same pa e n eme ged o he honey buzza d, bu in he i s b eeding pe iod only (Fig 2C). In
con as , he p opo ions o wa e and a able land we e low a he goshawk nes -si es and
inc eased wi h inc easing adius (Fig 2A). Young hinning o es was p ominen in all adii
and in bo h b eeding pe iods o all species (Fig 2A–2C,S2 Table). The habi a p opo ions
we e e y simila in di e en adii o he common buzza d (Fig 2B). Howe e , a able land was
Habi a E ec s on Fo es -Dwelling Hawks
PLOS ONE | DOI:10.1371/jou nal.pone.0137877 Sep embe 30, 2015 7/19
a la ge componen in common buzza d nes si es and a e i o y scales (Fig 2B), while a high
p opo ion o wa e was appa en a he e i o y scales a ound honey buzza d nes s (Fig 2C,
S2 Table).
The p opo ion o old sp uce o es declined signi ican ly be ween he i s and las b eeding
pe iods wi hin 1000 m and 2000 m a ound goshawk nes s, and wi hin 1000 m a ound common
buzza d nes s (Fig 3A,S2 Table). A he nes -si e scale (100 m), he e was no signi ican di e -
ence in he p opo ion o old sp uce o es be ween he i s and las b eeding pe iods o any
o he species, al hough o he honey buzza d, he a e age p opo ion o old sp uce o es was
hal ed (Fig 3A). Howe e , he sample size o he honey buzza d was low in he las b eeding
pe iod.
The p opo ion o o he old o es inc eased a all scales (100 m, 1000 m and 2000 m) o
bo h he goshawk and he common buzza d (Fig 3B,S2 Table). O he signi ican b eeding hab-
i a changes o he goshawk included a decline in he p opo ion o low s ocking o es and an
inc ease in buil -up land wi hin 2000 m a ound he nes s. No signi ican changes we e de ec ed
in he p opo ions o young hinning o es , wa e and a able land be ween he b eeding pe i-
ods o any o he species.
The abo e-men ioned habi a changes we e de ec ed in e i o ies ha he hawks accep ed
o b eeding and ha can hus ep esen mo e sui able en i onmen s o he hawks han on
a e age in he landscape. I was hus in e es ing o compa e he habi a changes a ound nes s
o o es habi a changes h oughou sou he n Finland. A eas o olde sp uce-domina ed o es
and o he olde o es dec eased in sou he n Finland by 24% ( om 19,040 km
2
o 14,471 km
2
)
and 10% ( om 27,626 km
2
o 24,838 km
2
), espec i ely. Ins ead, he a ea o young o es
inc eased by 14% ( om 42,901 km
2
o 48,877 km
2
)[48].
Discussion
We s udied habi a e ec s on he b eeding pe o mance o h ee hawk species in bo eal o es
landscapes. The con inuously declining popula ion ends o hese hawks ha e aised conce ns
Fig 1. P obabili ies o success ul goshawk b eeding based on he bes b eeding success GLMM (2000
m adius). All b eeding a emp s wi h a e i ied b eeding esul we e included om all b eeding pe iods. Thick
line: p edic ed alues, hin lines delinea e 95% o he a ia ion be ween e i o ies in p edic ed alues, do s:
da a poin s; 0 = unsuccess ul, 1 = success ul b eeding a emp in he y-axis. (A) P obabili y o success ul
b eeding along s anda dized log- a io p opo ion o old sp uce o es . (B) P obabili y o success ul b eeding
along s anda dized log- a io p opo ion o young hinning o es . Goshawk b eeding success inc eases wi h
(A) an inc easing p opo ion o old sp uce o es and (B) a dec easing p opo ion o young hinning o es
wi hin 2000 m a ound he nes .
doi:10.1371/jou nal.pone.0137877.g001
Habi a E ec s on Fo es -Dwelling Hawks
PLOS ONE | DOI:10.1371/jou nal.pone.0137877 Sep embe 30, 2015 8/19
abou un a ou able changes in hei b eeding habi a , and some ad e se changes we e de ec ed
in his s udy. We ound ha goshawk b eeding success inc eased wi h inc easing p opo ions
o old sp uce o es and wa e , and dec easing p opo ions o young hinning o es wi hin
2000 m a ound hei nes s. A he same ime, old sp uce o es dec eased a he e i o y scales
a ound goshawk nes s and h oughou sou he n Finland. We ound no signi ican associa ion
wi h he habi a composi ion and he b eeding success o he common buzza d and honey
buzza d.
Habi a -associa ed b eeding pe o mance
We ound a p e e ence o he goshawk o b eeding in old sp uce o es , since hey p edomi-
na ed a goshawk nes si es (Fig 2A), while hei p opo ion was much smalle on he landscape
scale. Mo eo e , a highe p opo ion o old sp uce o es a ound he nes inc eased goshawk
b eeding success. The goshawk’s p e e ence o ma u e s ands has been con i med in many
s udies [13,17,21,24,25,37,44]. The goshawk a ou s la ge o es pa ches [42,74] and hun s
inside he o es o a o es edges [37,75,76]. Fu he mo e, goshawk popula ions a e sensi i e
Table 1. Logi es ima es o he GLMMs
1
ha bes explained he b eeding success o each species. GLMMs included log- a io habi a p opo ions
wi hin he adius o 100 m, 500 m, 1000 m o 2000 m, and he model wi h he lowes AIC was selec ed. Low s ocking o es was emo ed om he bes gos-
hawk model due o high collinea i y. Va iance o he andom a iable ( e i o y iden i y) desc ibes a ia ion among e i o ies.
Species Radius selec ed Va iable Es ima e SE z- alue p (>|z|)
2
Goshawk 2000 m In e cep 2.74 0.13 21.34 <0.001
Old sp uce o es 0.54 0.22 2.47 0.014*
Young hinning o es –0.93 0.27 –3.40 <0.001***
Wa e 0.52 0.23 2.25 0.024*
A able land 0.22 0.25 0.90 0.369
Buil -up land 0.43 0.24 1.79 0.074
andom: Te i o y, σ
2
: 1.25
Common buzza d 100 m In e cep 7.77 0.95 8.21 <0.001
Old sp uce o es 0.01 0.66 0.02 0.988
Young hinning o es 0.63 2.11 0.30 0.767
Low s ocking o es 0.32 0.89 0.35 0.724
Wa e 2.64 2.23 1.19 0.235
A able land –1.55 1.23 –1.26 0.206
Buil -up land 0.13 1.19 0.11 0.912
La i ude –1.71 1.72 –0.99 0.322
andom: Te i o y, σ
2
: 56.75
Honey buzza d 100 m In e cep 11.26 3.90 2.89 0.004
Old sp uce o es 0.37 7.67 0.05 0.961
Young hinning o es 1.46 3.06 0.48 0.632
Low s ocking o es –0.39 3.28 –0.12 0.905
Wa e 5.86 6.31 0.93 0.353
A able land 0.68 11.02 0.06 0.951
Buil -up land –1.75 3.59 –0.49 0.626
andom: Te i o y, σ
2
: 288.95
1
GLMM: gene alized linea mixed model,
2
Significance le els:
*** <0.001,
*<0.05, n.s. 0.05
doi:10.1371/jou nal.pone.0137877. 001
Habi a E ec s on Fo es -Dwelling Hawks
PLOS ONE | DOI:10.1371/jou nal.pone.0137877 Sep embe 30, 2015 9/19
8. Myllyn aus T, Ma ila T. Decline o inc ease? The s anding imbe s ock in Finland, 1800–1997. Ecol
Econ. 2002; 41: 271–288. dx.doi.o g/10.1016/S0921-8009(02)00034-4.
9. Tomppo E, Heikkinen J, Hen onen HM, Ihalainen A, Ka ila M, Mäkelä H, e al. Designing and con-
duc ing a o es in en o y—case: 9 h na ional o es in en o y o Finland. 1s ed. London: Sp inge ;
2011.
10. Hanski I. Ex inc ion deb and species c edi in bo eal o es s: Modelling he consequences o di e en
app oaches o biodi e si y conse a ion. Ann Zool Fennici. 2000; 37: 271–280. A ailable: h p://www.
helsinki. i/~ihanski/AnnZool2000 Hanski.pd .
11. Kuulu ainen T, Kallio E. S uc u al cha ac e is ics and di e si y o na u al, selec i ely cu and man-
aged old Pinus syl es is-domina ed o es s in Kuhmo. In: Heikkilä R, Lindholm T, edi o s. Biodi e si y
and conse a ion o bo eal na u e. P oceedings o he 10 yea s anni e sa y symposium o he Na u e
Rese e F iendship. Van aa: Kainuu Regional En i onmen Cen e; 2003. pp. 89–92.
12. Rou inen S, Kuulu ainen T. T ee diame e dis ibu ions in na u al and managed old Pinus syl es is-
domina ed o es s. Fo Ecol Manage. 2005; 208: 45–61. dx.doi.o g/10.1016/j. o eco.2004.11.021.
13. Lõhmus A. Nes - ee and nes -s and cha ac e is ics o o es -dwelling ap o s in eas -cen al Es onia:
Implica ions o o es managemen and conse a ion. P oc Es onian Acad Sci Biol Ecol. 2006; 55:
31–50.
14. New on I. Popula ion ecology o ap o s. 1s ed. Be khams ed: T & A D Poyse ; 1979.
15. del Hoyo J, Ellio A, Sa ga al J, edi o s. Handbook o he Bi ds o he Wo ld. Vol. 2. New Wo ld Vul-
u es o Guinea owl. 1s ed. Ba celona: Lynx Edicions; 1994.
16. Gamau A, Tebb G, Neme h E. Honey buzza d Pe nis api o us nes -si e selec ion in ela ion o habi a
and he dis ibu ion o goshawks Accipi e gen ilis. Ibis. 2013; 155: 258–270. doi: 10.1111/ibi.12023
17. Pen e iani V. Goshawk nes ing habi a in Eu ope and No h Ame ica: A e iew. O nis Fenn. 2002; 79:
149–163.
18. Pen e iani V, Fai e B. B eeding densi y and landscape-le el habi a selec ion o common buzza ds
(Bu eo bu eo) in a moun ain a ea (Ab uzzo Apennines, I aly). J Rap o Res. 1997; 31: 208–212.
19. Pen e iani V, Fai e B. B eeding densi y and nes si e selec ion in a goshawk Accipi e gen ilis popula-
ion o he Cen al Apennines (Ab uzzo, I aly). Bi d S udy. 1997; 44: 136–145. doi: 10.1080/
00063659709461049
20. Reynolds RT, G aham RT, Reise MH, Basse RL, Kennedy PL, Boyce DA J , e al. Managemen ec-
ommenda ions o he no he n goshawk in he Sou hwes e n Uni ed S a es. Gene al Technical
Repo RM-217. 1992: 1–90.
21. Selås V. Nes -si e selec ion by ou sympa ic o es ap o s in sou he n No way. J Rap o Res. 1997;
31: 16–25.
22. C amp S, Simmons KEL. Handbook o he bi ds o Eu ope, he Middle Eas and No h A ica: The
bi ds o he Wes e n Palea c ic, Hawks o bus a ds. Ox o d: Ox o d Uni e si y P ess; 1980.
23. Amco M, Tje nbe g M, Be g Å. Nes si e choice o honey buzza d Pe nis api o us [Bi åkens Pe nis
api o us bopla s al]. O nis S ec. 1994; 4: 145–158.
24. Selås V, S een OF, Johnsen JT. Goshawk b eeding densi ies in ela ion o ma u e o es in sou heas -
e n No way. Fo Ecol Manage. 2008; 256: 446–451. doi: 10.1016/j. o eco.2008.04.047
25. Squi es JR, Kennedy PL. No he n goshawk ecology: An assessmen o cu en knowledge and in o -
ma ion needs o conse a ion and managemen . S ud A ian Biol. 2006; 31: 8–62.
26. Hakka ainen H, Myk ä S, Ku ki S, To nbe g R, Jungell S. Compe i i e in e ac ions among ap o s in
bo eal o es s. Oecologia. 2004; 141: 420–424. doi: 10.1007/s00442-004-1656-6 PMID: 15375689
27. Kos zewa A. A compa a i e s udy o nes -si e occupancy and b eeding pe o mance as indica o s o
nes ing-habi a quali y in h ee Eu opean ap o species. E hology Ecology and E olu ion. 1996; 8:
1–18. doi: 10.1080/08927014.1996.9522931
28. K üge O. Analysis o nes occupancy and nes ep oduc ion in wo sympa ic ap o s: Common buz-
za d Bu eo bu eo and goshawk Accipi e gen ilis. Ecog aphy. 2002; 25: 523–532. doi: 10.1034/j.
1600-0587.2002.250502.x
29. K üge O. In e ac ions be ween common buzza d Bu eo bu eo and goshawk Accipi e gen ilis: T ade-
o s e ealed by a ield expe imen . Oikos. 2002; 96: 441–452. doi: 10.1034/j.1600-0706.2002.
960306.x
30. Kos zewa A. In e speci ic in e e ence compe i ion in h ee Eu opean ap o species. E hology Ecol-
ogy and E olu ion. 1991; 3: 127–143.
31. Bi dLi e In e na ional. Bi ds in Eu ope: Popula ion es ima es, ends and conse a ion s a us. Cam-
b idge: Bi d Li e In e na ional; 2004.
Habi a E ec s on Fo es -Dwelling Hawks
PLOS ONE | DOI:10.1371/jou nal.pone.0137877 Sep embe 30, 2015 16 / 19

32. Doyle FI. Goshawks in Canada: Popula ion esponses o ha es ing and he app op ia eness o using
s anda d bi d moni o ing echniques o assess hei s a us. S ud A ian Biol. 2006; 31: 135–140.
33. Bijlsma RG, Ve meulen M, Heme ik L, Klok C. Demog aphy o Eu opean honey buzza ds Pe nis api-
o us. A dea. 2012; 100: 163–177.
34. Honkala J, Sau ola P, Valkama J. B eeding and popula ion ends o common ap o s and owls in Fin-
land in 2012. The yea book o he Linnu magazine 2012. 2013: 50–61. A ailable: h p://www.luomus.
i/si es/de aul / iles/ iles/linnu _ k2012_050-061_pe olinnu .pd .
35. Tje nbe g M, Ry man H. Su i al and popula ion de elopmen o he honey buzza d Pe nis api o us
in Sweden [Bi åkens Pe nis api o us ö e le nad och bes åndsu eckling i S e ige]. O nis S ec.
1994; 4: 133–139.
36. Widén P. How, and why, is he goshawk (Accipi e gen ilis) a ec ed by mode n o es managemen in
Fennoscandia? J Rap o Res. 1997; 31: 107–113.
37. To nbe g R, Ko pimäki E, Byholm P. Ecology o he no he n goshawk in Fennoscandia. S ud A ian
Biol. 2006; 31: 141–157. A ailable: h p://www.plane o bi ds.com/no he n-goshawk-a icle-2.
38. Valkama J, Vepsäläinen V, Lehikoinen A. The Thi d Finnish B eeding Bi d A las. Finnish Museum o
Na u al His o y and Minis y o En i onmen . 2011. A ailable: h p://a las3.lin ua las. i/english/.
Accessed 19 May 2014.
39. Sau ola P. Moni o ing bi ds o p ey in Finland: A summa y o me hods, ends, and s a is ical powe .
Ambio. 2008; 37: 413–419. dx.doi.o g/10.1579/0044-7447(2008)37[416:MBOPIF]2.0.CO;2. PMID:
18833793
40. Mikkola-Roos M, Tiainen J, Below A, Ha io M, Lehikoinen A, Lehikoinen E, e al. Bi ds. In: Rassi P,
Hy ä inen E, Juslén A, Manne koski I, edi o s. The 2010 Red Lis o Finnish Species. Helsinki:
Ympä is öminis e iö & Suomen ympä is ökeskus; 2010. pp. 320–331. A ailable: h p://www.
ympa is o. i/en-us/Na u e/Species/Th ea ened_species/The_2010_Red_Lis _o _Finnish_species.
41. Byholm P, Nikula A, Ken ä J, Tai almäki J-P. In e ac ions be ween habi a he e ogenei y and ood
a ec ep oduc i e ou pu in a op p eda o . J Anim Ecol. 2007; 76: 392–401. doi: 10.1111/j.1365-
2656.2007.01211.x PMID: 17302847
42. Becke TE, Smi h DG, Bosakowski T. Habi a , ood habi s, and p oduc i i y o no he n goshawks
nes ing in Connec icu . S ud A ian Biol. 2006; 31: 119–125.
43. Ca oll C, Rod iguez RL, McCa hy C, Paulin KM. Resou ce selec ion unc ion models as ools o
egional conse a ion planning o no he n goshawk in U ah. S ud A ian Biol. 2006; 31: 288–298.
44. Finn SP, Ma zlu JM, Va land DE. E ec s o landscape and local habi a a ibu es on no he n gos-
hawk si e occupancy in wes e n Washing on. Fo Sci. 2002; 48: 427–436.
45. Ama A, A oyo B, Meek E, Redpa h S, Riley H. In luence o habi a on b eeding pe o mance o hen
ha ie s Ci cus cyaneus in O kney. Ibis. 2008; 150: 400–404. doi: 10.1111/j.1474-919X.2007.00765.x
46. Ha gis CD, McCa hy C, Pe lo RD. Home anges and habi a s o no he n goshawks in eas e n Cali-
o nia. S ud A ian Biol. 1994; 16: 66–74.
47. G aham RT, Reynolds RT, Reise MH, Basse RL, Boyse DA. Sus aining o es habi a o he no h-
e n goshawk: A ques ion o scale. S ud A ian Biol. 1994; 16: 12–17.
48. Fo es s a is ics se ice Me in o. Finnish Fo es Resea ch Ins i u e. 2013. A ailable: h p://www.me la.
i/me in o/ mi/index.h m. Accessed 14 Oc obe 2013.
49. Pel ola A, Ihalainen A. Me sä a a (Fo es esou ces). In: Yli alo E, edi o . Me sä ilas ollinen uosiki ja
2012 (Finnish S a is ical Yea book o Fo es y). Van aa: Finnish Fo es Resea ch Ins i u e; 2012. pp.
37–78 (English summa y: pp. 431–432). A ailable: h p://www.me la. i/me in o/ ilas o/julkaisu / sk/
2012/. Accessed 3 No embe 2014.
50. Kenwa d R. The goshawk. London: T & AD Poyse ; 2006.
51. Bjö klund H, Valkama J, Sau ola P, Laaksonen T. E alua ion o a i icial nes s as a conse a ion ool
o h ee o es -dwelling ap o s. Anim Conse . 2013; 16: 546–555. doi: 10.1111/ac .12028
52. Rei V, To nbe g R, Jungell S, Ko pimäki E. Die a ia ion o common buzza ds in Finland suppo s
he al e na i e p ey hypo hesis. Ecog aphy. 2001; 24: 267–274. www.js o .o g/s able/3683704.
53. I ämies J, Mikkola H. The die o honey buzza ds Pe nis api o us in Finland. O nis Fenn. 1972; 49:
7–10.
54. Sau ola P, Valkama J, Velmala W. The Finnish bi d inging a las. Vol. I. Helsinki: Finnish Museum o
Na u al His o y Luomus and Minis y o En i onmen ; 2013.
55. Hu lbe SH. Pseudo eplica ion and he design o ecological ield expe imen s. Ecol Monog . 1984;
54: 187–211. doi: 10.2307/1942661
Habi a E ec s on Fo es -Dwelling Hawks
PLOS ONE | DOI:10.1371/jou nal.pone.0137877 Sep embe 30, 2015 17 / 19
56. Tomppo E, Haakana M, Ka ila M, Mäkisa a K, Pe äsaa i J. The mul i-sou ce na ional o es in en o y
o Finland—me hods and esul s 2005. Wo king pape s o he Finnish Fo es Resea ch Ins i u e.
2009; 111: 1–277. A ailable: h p://www.me la. i/julkaisu /wo kingpape s/2009/mwp111.pd .
57. Tomppo E, Ka ila M, Mäkisa a K, Pe äsaa i J. The mul i-sou ce na ional o es in en o y o Finland—
me hods and esul s 2009. Wo king pape s o he Finnish Fo es Resea ch Ins i u e. 2013; 273: 1–
216.
58. Hakka ainen H, Myk ä S, Ku ki S, To nbe g R, Jungell S, Nikula A. Long- e m change in e i o y occu-
pancy pa e n o goshawks (Accipi e gen ilis). Ecoscience. 2004; 11: 399–403.
59. Tike In o ma ion Cen e o he Minis y o Ag icul u e and Fo es y. U ilized ag icul u al a ea. 2014.
A ailable: h p://www.maa alous ilas o . i/en/u ilised-ag icul u al-a ea. Accessed 12 Ma ch 2015.
60. Kuulu ainen T, Mäki J, Ka jalainen L, Leh onen H. T ee age dis ibu ions in old-g ow h o es si es in
Vienansalo wilde ness, eas e n Fennoscandia. Sil a Fennica. 2002; 36: 169–184.
61. Tomppo E, Hen onen H, Tuomainen T. Val akunnan me sien 8. in en oinnin mene elmä ja ulokse
me säkeskuksi ain Pohjois-Suomessa 1992–94 sekä ulokse E elä-Suomessa 1986–92 ja koko
maassa 1986–94. Me sä ie een aikakauski ja. 2001; 1B: 99–248.
62. Tomppo E, Haakana M, Ka ila M, Pe äsaa i J. Mul i-sou ce na ional o es in en o y: Me hods and
applica ions. New Yo k: Sp inge ; 2008.
63. Ai chison J. The s a is ical analysis o composi ional da a. London: Chapman and Hall; 1986.
64. Aebische NJ, Robe son PA, Kenwa d RE. Composi ion analysis o habi a use om animal adio-
acking da a. Ecology. 1993; 74: 1313–1325. doi: 10.2307/1940062
65. Hakka ainen H, Myk ä S, Ku ki S, Ko pimäki E, Nikula A, Koi unen V. Habi a composi ion as a de e -
minan o ep oduc i e success o Tengmalm's owls unde luc ua ing ood condi ions. Oikos. 2003;
100: 162–171. doi: 10.1034/j.1600-0706.2003.11906.x
66. McRobe s RE, Cohen WB, Næsse E, S ehman SV, Tomppo EO. Using emo ely sensed da a o
cons uc and assess o es a ibu e maps and ela ed spa ial p oduc s. Scand J Fo Res. 2010; 25:
340–367. doi: 10.1080/02827581.2010.497496
67. Zuu AF, Ieno EN, Walke NJ, Sa elie AA, Smi h GM. Mixed e ec s models and ex ensions in ecol-
ogy wi h R. New Yo k: Sp inge ; 2009.
68. G uebe CE, Nakagawa S, Laws RJ, Jamieson IG. Mul imodel in e ence in ecology and e olu ion:
Challenges and solu ions. J E ol Biol. 2011; 24: 699–711. doi: 10.1111/j.1420-9101.2010.02210.x
PMID: 21272107
69. Bu nham K, Ande son D. Model selec ion and mul imodel in e ence. A p ac ical in o ma ion- heo e ic
app oach. 2 ed. New Yo k: Sp inge -Ve lag; 2002.
70. Do mann CF, Eli h J, Bache S, Buchmann C, Ca l G, Ca é G, e al. Collinea i y: A e iew o me hods
o deal wi h i and a simula ion s udy e alua ing hei pe o mance. Ecog aphy. 2013; 36: 027–046.
doi: 10.1111/j.1600-0587.2012.07348.x
71. Can y A, Ripley B. Boo : Boo s ap R (S-Plus) Func ions. R package e sion 1.3–15; 2015.
72. Do mann CF, McPhe son JM, A aújo MB, Bi and R, Bollige J, Ca l G, e al. Me hods o accoun o
spa ial au oco ela ion in he analysis o species dis ibu ional da a: A e iew. Ecog aphy. 2007; 30:
609–628. doi: 10.1111/j.2007.0906-7590.05171.x
73. McDonald JH. Handbook o biological s a is ics. 2 ed. Bal imo e, Ma yland: Spa ky House Publish-
ing; 2008.
74. Woodb idge B, De ich PJ. Te i o y occupancy and habi a pa ch size o no he n goshawks in he
sou he n cascades o Cali o nia. S ud A ian Biol. 1994; 16: 83–87.
75. Blox on TD J P ey abundance, space use, demog aphy, and o aging habi a o no he n goshawks
in wes e n Washing on. M.Sc. Thesis, he Uni e si y o Washing on. 2002. A ailable: h p://www.
u al ech.o g/pubs/ heses/blox on/blox on_ms_ hesis.pd .
76. Widén P. The hun ing habi a s o goshawks Accipi e gen ilis in bo eal o es s o cen al Sweden. Ibis.
1989; 131: 205–231. doi: 10.1111/j.1474-919X.1989. b02763.x
77. Ku ki S, Nikula A, Helle P, Lindén H. Landscape agmen a ion and o es composi ion e ec s on
g ouse b eeding success in bo eal o es s. Ecology. 2000; 81: 1985–1997.
78. To nbe g R, Lindén A, Byholm P, Ran a E, Valkama J, Helle P, e al. Coupling in goshawk and g ouse
popula ion dynamics in Finland. Oecologia. 2013; 171: 863–872. doi: 10.1007/s00442-012-2448-z
PMID: 22961371
79. Byholm P, Kekkonen M. Food egula es ep oduc ion di e en ly in di e en habi a s: Expe imen al e i-
dence in he Goshawk. Ecology. 2008; 89: 1696–1702. PMID: 18589533
Habi a E ec s on Fo es -Dwelling Hawks
PLOS ONE | DOI:10.1371/jou nal.pone.0137877 Sep embe 30, 2015 18 / 19
80. López-López P, Ga cía-Ripollés C, U ios V. Popula ion size, b eeding pe o mance and e i o y qual-
i y o Bonelli's eagle Hie aae us ascia us in eas e n Spain. Bi d S udy. 2007; 54: 335–342. doi: 10.
1080/00063650709461493
81. To nbe g R, Mönkkönen M, Ki elä SM. Landscape and season e ec s on he die o he goshawk.
Ibis. 2009; 151: 396–400. doi: 10.1111/j.1474-919X.2009.00910.x
82. Coeu dassie M, Riols R, Deco s A, Mionne A, Da id F, Quin aine T, e al. Unin en ional wildli e poi-
soning and p oposals o sus ainable managemen o oden s. Conse Biol. 2014; 28: 315–321. doi:
10.1111/cobi.12230 PMID: 24405288
83. Jonke RM, Chaka o N, K üge O. Clima e change and habi a he e ogenei y d i e a popula ion
inc ease in common buzza ds Bu eo bu eo h ough e ec s on su i al. Ibis. 2014; 156: 97–106. doi:
10.1111/ibi.12124
84. Lehikoinen A, Byholm P, Ran a E, Sau ola P, Valkama J, Ko pimäki E, e al. Rep oduc ion o he com-
mon buzza d a i s no he n ange ma gin unde clima ic change. Oikos. 2009; 118: 829–836. doi: 10.
1111/j.1600-0706.2008.17440.x
85. K üge O. Dissec ing common buzza d li espan and li e ime ep oduc i e success: The ela i e impo -
ance o ood, compe i ion, wea he , habi a and indi idual a ibu es. Oecologia. 2002; 133: 474–482.
A ailable: h p://www.js o .o g/s able/4223446.
86. K üge O. The impo ance o compe i ion, ood, habi a , wea he and pheno ype o he ep oduc ion
o buzza d Bu eo bu eo. Bi d S udy. 2004; 51: 125–132. doi: 10.1080/00063650409461344
87. Agos ini N, Panuccio M, Mellone U, Lucia G, Wilson S, Ash on-Boo h J. Do mig a ion coun s e lec
popula ion ends? A case s udy o he honey buzza d Pe nis api o us. A deola. 2007; 54: 339–344.
88. Bijlsma RG. Wha is he p eda ion isk o Eu opean honey buzza ds Pe nis api o us in Du ch o es s
inhabi ed by ood-s essed no he n goshawks Accipi e gen ilis? De Takkeling. 2004; 12: 185–197.
89. Bijlsma RG. Li e-his o y ai s o honey buzza ds (Pe nis api o us) in A ica. Vogelwa e. 2002; 41:
240–248.
90. S andbe g R, Hake M, Klaassen RHG, Ale s am T. Mo emen s o imma u e Eu opean honey buz-
za ds Pe nis api o us in opical A ica. A dea. 2012; 100: 157–162. doi: 10.5253/078.100.0207
91. Chaka o N, K üge O. Mesop eda o elease by an eme gen supe p eda o : A na u al expe imen o
p eda ion in a h ee le el guild. PLOS ONE. 2010; 5: e15229. doi: 10.1371/jou nal.pone.0015229
PMID: 21151912
92. López-López P, Ga cía-Ripollés C, U ios V. Food p edic abili y de e mines space use o endange ed
ul u es: Implica ions o managemen o supplemen a y eeding. Ecol Appl. 2014; 24: 938–949. doi:
10.1890/13-2000.1 PMID: 25154088
93. Hakka ainen H, Ko pimäki E, Laaksonen T, Nikula A, Suo sa P. Su i al o male Tengmalm's owls
inc eases wi h co e o old o es in hei e i o y. Oecologia. 2008; 155: 479–486. doi: 10.1007/
s00442-007-0929-2 PMID: 18080142
94. Kaila A, Ihalainen A. Fo es esou ces. In: Pel ola A, edi o . Finnish S a is ical Yea book o Fo es y
2014. Van aa: Finnish Fo es Resea ch Ins i u e; 2014. pp. 33–78. A ailable: h p://www.me la. i/
julkaisu /me sa ilas ollinen sk/ ilas o sk-sisal o.h m.
95. Bosakowski T, Speise R. Mac ohabi a selec ion by nes ing no he n goshawks: Implica ions o
managing eas e n o es s. S ud A ian Biol. 1994; 16: 46–49.
96. Huh a E, Aho T, Jän i A, Suo sa P, Kui unen M, Nikula A, e al. Fo es agmen a ion inc eases nes
p eda ion in he Eu asian eec eepe . Conse Biol. 2004; 18: 148–155. doi: 10.1111/j.1523-1739.
2004.00270.x
97. Laaksonen T, Hakka ainen H, Ko pimäki E. Li e ime ep oduc ion o a o es -dwelling owl inc eases
wi h age and a ea o o es s. P oc R Soc B Biol Sci. 2004; 271: S461–S464. doi: 10.1098/ sbl.2004.
0221
98. Mönkkönen M, Reunanen P, Ko iaho JS, Juu inen A, Tikkanen O-P, Kouki J. Cos -e ec i e s a egies
o conse e bo eal o es biodi e si y and long- e m landscape-le el main enance o habi a s. Eu J
Fo es Res. 2011; 130: 717–727. doi: 10.1007/s10342-010-0461-5
99. Myk ä S, Ku ki S, Nikula A. The spacing o ma u e o es habi a in ela ion o species-speci ic scales
in managed bo eal o es s in NE Finland. Ann Zool Fenn. 2000; 37: 79–91. A ailable: h p://www.sekj.
o g/PDF/anz 37/anz 37-079p.pd .
100. Reynolds RT, G aham RT, Boyce DA J . An ecosys em-based conse a ion s a egy o he no he n
goshawk. S ud A ian Biol. 2006; 31: 299–311.
Habi a E ec s on Fo es -Dwelling Hawks
PLOS ONE | DOI:10.1371/jou nal.pone.0137877 Sep embe 30, 2015 19 / 19