Die Quali y Limi s Summe G ow h o Field Vole
Popula ions
K is ian M. Fo bes
1,2
, Pe e S ua
1,3
, Tapio Mappes
2
, Ka ine S. Hose
4
, Heikki Hen onen
5
, O so Hui u
1
*
1Suonenjoki Resea ch Uni , Finnish Fo es Resea ch Ins i u e, Suonenjoki, Finland, 2Depa men o Biological and En i onmen al Science, Uni e si y o Jy a
¨skyla
¨,
Jy a
¨skyla
¨, Finland, 3Depa men o Bo any and Zoology, Masa yk Uni e si y, B no, Czech Republic, 4Sec ion o Ecology, Depa men o Biology, Uni e si y o Tu ku,
Tu ku, Finland, 5Van aa Resea ch Uni , Finnish Fo es Resea ch Ins i u e, Van aa, Finland
Abs ac
Ma ked a ia ion occu s in bo h seasonal and mul iannual popula ion densi y peaks o no he n Eu opean small mammal
species, including oles. The a ailabili y o die a y p o eins is a key ac o limi ing he popula ion g ow h o he bi o e
species. The objec i e o his s udy is o in es iga e he deg ee o which p o ein a ailabili y in luences he g ow h o
inc easing ole popula ions. We hypo hesise ha he summe g ow h o oli o ous ole popula ions is posi i ely associa ed
wi h die a y p o ein a ailabili y. A ield expe imen was conduc ed o e a summe ep oduc i e pe iod in 18 ege a ed
enclosu es. Popula ions o ield oles (Mic o us ag es is) we e andomised amongs h ee ea men g oups: 1) ood
supplemen a ion wi h ad libi um high p o ein (30% d y weigh ) pelle s, 2) ood supplemen a ion wi h ad libi um low p o ein
(1% d y weigh ; bo h supplemen ed oods had equi alen ene gy con en ) pelle s, and 3) con ol (no ood
supplemen a ion), n = 6 pe ea men . Vole densi y, su i al, demog aphic a ibu es and condi ion indica o s we e
moni o ed wi h li e- apping and blood sampling. Highes inal ole densi ies we e a ained in popula ions ha ecei ed
high p o ein supplemen a ion and lowes in low p o ein popula ions. Con ol popula ions displayed in e media e densi ies.
The su i al a e o oles was simila in all ea men g oups. The p opo ion o emales, and o hose ha we e p egnan o
lac a ing, was highes in he high p o ein supplemen ed popula ions. This sugges s ha a ia ion in ep oduc i e, a he
han su i al a es o oles, accoun ed o densi y di e ences be ween he ea men g oups. We ound no clea associa ion
be ween popula ion demog aphy and indi idual physiological condi ion. Ou esul s demons a e ha die a y p o ein
a ailabili y limi s ole popula ion g ow h du ing he summe g owing season. This sugges s ha he nu i ional quali y o
o age may be an unde es ima ed sou ce o in e annual a ia ion in he densi y and g ow h a es o widely luc ua ing
popula ions o he bi o ous small mammals.
Ci a ion: Fo bes KM, S ua P, Mappes T, Hose KS, Hen onen H, e al. (2014) Die Quali y Limi s Summe G ow h o Field Vole Popula ions. PLoS ONE 9(3):
e91113. doi:10.1371/jou nal.pone.0091113
Edi o : Gab iele So ci, CNRS, Uni e si e
´de Bou gogne, F ance
Recei ed No embe 4, 2013; Accep ed Feb ua y 7, 2014; Published Ma ch 12, 2014
Copy igh : ß2014 Fo bes 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: The expe imen was unded by he Academy o Finland (www.aka. i/eng: g an 133495 o OH, g an 132190 o TM). 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: Tapio Mappes is a PLOS ONE edi o ial boa d membe . This does no al e he au ho s’ adhe ence o PLOS ONE edi o ial policies and
c i e ia.
* E-mail: o so.hui u@me la. i
In oduc ion
Popula ions o no he n small mammals a e enowned o hei
high-ampli ude densi y cycles, wi h peaks e e y 3–5 yea s [1–5].
Al hough delayed densi y-dependen p eda ion is o en conside ed
he p inciple d i e o cyclic dynamics [6–10], egula o y p ocesses
a e likely o be mul i ac o ial and geog aphically a iable [1,11].
Hence, consensus on causal ac o s behind cyclici y has no been
eached despi e se e al decades o esea ch [1,10–14].
Bo eal ole cycles ypically in ol e wo successi e yea s o
a iable bu posi i e popula ion g ow h in summe and nega i e o
ze o popula ion g ow h in win e [2,15–16]. The peak o a
mul iannual cycle is usually a ained in la e summe o au umn o
he la e inc ease yea , a e which win e ood deple ion ini ia es
a popula ion decline [17–18]. The yea ollowing peak densi y is
cha ac e ized by a summe decline, when popula ions ypically
dec ease in size om sp ing o au umn [1–2,15].
The g ow h a e o ole popula ions a ies p o oundly be ween
yea s, including yea s ep esen ing he same cycle phase. The
o e all ampli ude o mul i-annual cycles (i.e. he di e ence
be ween maximum and minimum densi ies) also a ies ma kedly
wi hin and be ween si es [1–2,8,19]. Cycle ampli ude is gene ally
g ea e in coole and mo e con inen al a eas han in empe a e,
mild coas al a eas, whe e densi y a ia ions a e p edominan ly
seasonal [1,20]. These di e ences ha e adi ionally been a ib-
u ed o win e se e i y and amoun o snow all, which a e
nega i ely associa ed wi h he s abilising e ec o gene alis
p eda o s on oles and hei specialis p eda o s [1,8].
Recen epo s ha e documen ed a widesp ead collapse o small
oden popula ion cycles [21], o en a ibu ed o changing win e
clima e [22–24]. Ko pela e al. [19] p esen ed e idence o
challenge his associa ion, and ins ead, using ex ensi e ime-se ies
ole moni o ing and clima ic da a om Finland, highligh ed a
connec ion be ween wea he condi ions du ing sp ing and
summe , and ole popula ion g ow h. The la e ela ionship is
po en ially media ed by a ia ion in o age quali y (e.g. [25–26]).
Fo he bi o es, o age quali y is o a conside able deg ee
de e mined by ni ogen con en , which is o en a p ima y limi ing
ac o o he g ow h o popula ions (ni ogen limi a ion hypo hesis
[27–30]). Ni ogen le els in plan s a y in esponse o a ange o
PLOS ONE | www.plosone.o g 1 Ma ch 2014 | Volume 9 | Issue 3 | e91113
bio ic and abio ic ac o s, such as wea he , leading o bo h spa ial
and empo al a ia ion in i s a ailabili y o he bi o es [31–32].
Fo example, Cole and Ba zli [33] iden i ied ha di e en
ege a ion ypes al e ed he densi y, ep oduc i e pe o mance
and su i al o wild p ai ie ole popula ions, and concluded ha
highly nu i ional o age can ele a e peak popula ion densi ies.
Addi ionally, a midsumme cessa ion o b eeding, o en occu ing
in cyclic oli o ous oles du ing he heigh o he summe g owing
season ( e med a ‘midsumme c isis’ o oles), is hypo hesized o
esul om nu i ional changes in plan s du ing hei ep oduc i e
phenology [15].
The physiological heal h s a e o indi iduals may a y be o e
ansla ing in o changes in popula ion demog aphy. Fo example,
popula ions o small mammals a e cha ac e ised in decline yea s
by small indi idual body size, as well as low ep oduc i e ou pu
and adul su i al ( he Chi y e ec [34–36]). Haema ological
indices, e.g., le els o albumin, haema oc i and immunoglobulins,
can also e lec he quali y o indi idual die a y in ake [37–40].
The objec i e o ou s udy is o e alua e he ex en o which
p o ein a ailabili y limi s he densi y and popula ion g ow h o
small mammals du ing no he n Eu opean summe , a ime o
seemingly supe abundan ood esou ces. We hypo hesise ha
p o ein supplemen a ion will ha e posi i e e ec s, p oxima ely on
he physiological condi ion o oles and ul ima ely on popula ion
g ow h, as compa ed o non-supplemen ed popula ions. Speci i-
cally, we p edic ha he posi i e esponse will be mo e
p onounced in popula ions ha ecei e high p o ein ood han
in hose ha ecei e supplemen al ood wi h equi alen le els o
ene gy bu low p o ein. As model species we use he oli o ous
ield ole (Mic o us ag es is), he mos widely dis ibu ed o
luc ua ing small oden s h oughou Fennoscandia, and o en
conside ed he d i e o popula ion cycles in no he n Eu ope
[41].
Ma e ials and Me hods
E hics s a emen
The expe imen was conduc ed on p i a e land nea he own o
Suonenjoki in Cen al Finland [la 62u45.6729, lon 27u6.0159;
ETRS89 geog aphic coo dina es (,WGS84)]. Pe missions o
ca ying ou expe imen s a his loca ion we e ob ained om he
land owne , whose con ac in o ma ion is a ailable om he
au ho s upon eques . The s udy did no in ol e endange ed o
p o ec ed species. The expe imen was app o ed by he Finnish
Animal E hics Council (pe mi ESAVI/1437/04.10.03/2011).
Field echnicians we e ained p io o he expe imen and ook
all possibly p ecau ions o minimise animal s ess du ing apping
and sampling.
Enclosu es and expe imen al design
The expe imen was conduc ed in 18 adjoining ield enclosu es
(20625 m each) wi h na u al meadow ege a ion, domina ed by
he g asses Phleum p a ense and Deschampsia caespi osa. Enclosu es
we e cons uc ed o shee me al ising app oxima ely one me e
abo e g ound and ex ending 50 cm unde g ound. The eby, ole
mo emen be ween enclosu es was p e en ed and access by
mammalian p eda o s o oles (mus elids) es ic ed. A ian
p eda o s had access o he enclosu es bu we e e y a ely
obse ed in he a ea du ing he expe imen . Each enclosu e
con ained eigh shee -me al shel e boxes (40640650 cm, wi h
wo en ance hole a he base) app oxima ely 10 m apa , in a
36263 con igu a ion. An Ugglan Special li e ap (G ahnab,
Sweden) was placed in each shel e box.
Enclosu es we e andomly alloca ed o one o h ee ea men
g oups: 1) ad libi um high-p o ein (30 pe cen d y weigh c ude
p o ein) ood supplemen a ion, 2) ad libi um low-p o ein (1 pe cen )
ood supplemen a ion, o 3) con ol (no ood supplemen a ion).
The ene gy con en o he wo p o ein ea men s (30% and 1%)
was unchanged a 3500 kcal/kg. Food supplemen a ion was
supplied h ough speci ically o mula ed pelle s (Al omin, Lage,
Ge many) a ailable om a wi e mesh eede placed in each shel e
box.
A he beginning o June 2011, six ield oles ( h ee males, h ee
emales) we e in oduced o each enclosu e. The i s apping
occasion was conduc ed wo weeks la e o ob ain baseline
abundance es ima es ep esen a i e o es ablished indi iduals. A
o al o wo male and h ee emale oles we e in oduced o ou
enclosu es (2 high p o ein and 2 low p o ein) o eplace oles ha
had appa en ly died be ween in oduc ion and baseline apping.
Food supplemen a ion began immedia ely ollowing baseline
apping on June 16, 2011, and con inued un il immedia ely p io
o he inal apping in mid-Sep embe (13
h
) 2011. The
expe imen hus encompassed he p ima y ep oduc i e pe iod
o ield ole popula ions in cen al Finland [15].
Vole moni o ing and sampling
Abundance moni o ing and ole blood sampling was conduc ed
e e y ou h week o a o al o ou apping occasions. On each
occasion, aps we e bai ed wi h oa s and checked consecu i ely a
7 am, 2 pm and 9 pm, o a o al o 8–9 imes o e h ee days. An
elec onic PIT- ag (EID Aal en BV, Aal en, Ne he lands) was
subcu aneously injec ed in o e e y ole upon i s cap u e and he
unique iden i ica ion numbe eco ded a each encoun e . Voles
we e placed in o en ila ed bucke s and aken o an on-si e ield
labo a o y whe e hei sex and ep oduc i e s a us (males:
subadul , ma u e; emales: subadul , ma u e, p egnan and/o
lac a ing) was de e mined h ough ex e nal examina ion. Body
mass and head wid h we e measu ed ( o nea es 0.1 g and
0.1 mm, espec i ely), and app oxima ely 150 ml o blood was
collec ed om he e o-o bi al sinus wi h hepa inized capilla y
ubes. Blood was no collec ed om ju enile indi iduals weighing
unde 15 g. Voles we e hen eleased in o he same enclosu e as
cap u ed, excep on he inal apping occasion when oles we e
emo ed om enclosu es. Upon encoun e ing an indi idual ha
had al eady been sampled o blood du ing he apping occasion,
he ole was immedia ely eleased a he poin o cap u e a e
eco ding i s iden i ica ion numbe , sex, ep oduc i e s a us and
weigh .
Vole abundance (he ea e densi y) was es ima ed sepa a ely o
each enclosu e and apping occasion (18 enclosu es 64
occasions = 72 popula ion densi y es ima es) using he p og am
CAPTURE [42]. M
h
models (which inco po a e he e ogenei y in
cap u e a es) wi h he jackkni e es ima o we e employed o
apping occasions one o h ee. Th oughou his pe iod, ou
enclosu es expe ienced one apping occasion in which no
indi idual was ecap u ed a e hei ini ial cap u e. In hese
cases, densi y was es ima ed wi h emo al (M
bh
) models (Pollock
and O o’s es ima o [43]). Du ing he inal apping occasion,
oles we e emo ed om enclosu es upon i s cap u e and densi y
was es ima ed wi h emo al models. Ra ely, oles we e ound dead
in aps o died du ing sampling (app oxima ely 3% o cap u es).
These indi iduals we e excluded om he densi y es ima ion
models, bu added o he inal es ima e [42]. A popula ion g ow h
a e was calcula ed o each apping in e al based on he
o mula, R
= ln(N
-1
/N
), whe e N
is he popula ion densi y a
ime [18,44].
Die Quali y Limi s Summe Vole Popula ions
PLOS ONE | www.plosone.o g 2 Ma ch 2014 | Volume 9 | Issue 3 | e91113
Vole su i al a e was calcula ed sepa a ely o each enclosu e
and apping in e al using p og am MARK 7.0 [45]. Since
su i al es ima es pa ially depend on ecap u e a e, Akaike’s
in o ma ion c i e ion (AIC) -based model selec ion was employed
[46] o compa e ecap u e a e models including enclosu e,
apping occasion, hei pe mu a ions o only he in e cep . Due o
a small di e ence in AIC alues be ween he wo mos
pa simonious models (DAIC ,2), inal su i al es ima es we e
ob ained using a weigh ed model a e aging p ocedu e, aking
model selec ion unce ain y in o accoun [46].
Condi ion indices
Body condi ion index was exp essed as he s uden ized esiduals
o a andom coe icien s eg ession model o indi idual body mass
on head wid h [47]. Iden i y o he head wid h measu e was
en e ed as a andom ac o in he model o adjus o po en ial
indi idual a ia ion in head wid h measu emen s. Only ma u e
males we e included in he analysis o condi ion index o a oid
con oundmen by ju eniles and ep oducing emales.
Vole blood was cen i uged a 12 000 g o i e minu es, and
haema oc i exp essed as he pe cen age o packed ed blood cells
in o al olume. Blood plasma was hen sepa a ed and ozen
(,220uC) be o e enzyme-linked immunoso ben assays (he ea e ,
ELISA).
To al IgG an ibody i es we e measu ed acco ding o he
ollowing p o ocol. Solid an i-mouse conjuga e was pushed
h ough a 0.22 mm sy inge il e and dissol ed in 0.135 M NaCl.
Pla e wells we e hen coa ed wi h 50 ml o an i-mouse IgG (M-
8642, Sigma, lo 060M6082) solu ion (1 mg/ml) and incuba ed o
a minimum o 12 hou s a +4uC. Wells we e emp ied, and masked
wi h 100 ml 1% bo ine se um albumin in phospha e-bu e ed
saline (BSA-PBS) and incuba ed o 60 minu es a oom
empe a u e. Wells we e hen emp ied, washed and pa d ied.
50 ml o plasma sample (dilu ed a 1:40000 wi h BSA) was added
o duplica e wells. A s anda d was p epa ed by combining 2 ml
om each sample o e all apping occasions. Duplica e s anda d
concen a ions o 200, 150, 100, 50, 25, 10, 5 and 0 we e un on
each pla e simul aneously wi h samples. Pla es con aining samples
and s anda ds we e hen incuba ed o 3 hou s a oom
empe a u e. Following incuba ion, solu ions we e emo ed and
he wells washed. 50 ml o alkaline phospha ase conjuga ed an i-
mouse IgG (A-2179, Sigma, lo 31K4852), dilu ed a 1:4000 wi h
BSA-PBS, was added o each well and pla es incuba ed o a
minimum o 12 hou s a +4uC. Following incuba ion, wells we e
washed and pa d ied, and 50 ml o subs a e (1 mg pNPP [P4744,
Sigma, lo 109K6076] o 1 ml DEA bu e ) was added o each.
Pla es we e hen incuba ed in he da k and ead a 405 nm wi h a
The mo Labsys ems Mul iskan Ascen 354 pla e eade a e 15,
30, 45, 60 and 75 minu es. An abso bance app oxima ely mid-way
be ween he s anda d dilu ions is mos desi able. A e compa ing
abso bance le els, 45 minu es o incuba ion was deemed he mos
app op ia e. The mean abso bance o he sample duplica es was
used as he inal measu e. On a e occasions when an anomalous
esul occu ed, he plausible duplica e was used alone.
A comme cially a ailable mouse-albumin ELISA ki (Alpha
Diagnos ics In e na ional, Texas) was used o measu e he
albumin concen a ion o ole plasma as pe manu ac u e ’s
ins uc ions. An an i-mouse albumin-HPR conjuga e was used and
pla es we e ead a 450 nm using he The mo Labsys ems
Mul iskan Ascen 354 pla e eade .
S a is ical analyses
Random coe icien s eg ession models (PROC MIXED) we e
used o e alua e he indi idual and in e ac i e e ec s o ime (week
o yea as a con inuous a iable) and ea men on ole densi y,
wi h he in e cep and week as andom e ec s. The e ec o
popula ion mean condi ion index on densi y was e alua ed in a
sepa a e model. Fo his, da a we e es ic ed o he inal h ee
apping occasions, and condi ion a he p e ious apping
occasion ( -1) se as an ini ial explana o y a iable, along wi h
ime, ea men and hei in e ac ions. The in e cep and ime
we e again used as andom e ec s.
Due o he posi i e co ela ion be ween densi y and week (P,
0.001), analyses o g ow h a e (R
), su i al, condi ion index, o al
IgG, haema oc i and albumin con en we e ca ied ou wi h
epea ed-measu es mixed ANOVA models (PROC MIXED) wi h
apping occasion as a epea ed ca ego ical a iable. O he ini ial
ixed explana o y a iables we e ea men , densi y and all possible
in e ac ions. Enclosu e and enclosu e 6 apping occasion we e
included as andom ac o s ( o IgG and albumin models he
ELISA pla e numbe was also included as a andom ac o ).
Repea ed co a ia e ype (au o eg essi e, uns uc u ed, compound
symme y o oepli z) selec ion was based on AIC o he ull model.
Model selec ion was he ea e based on a s epwise educ ion
app oach, guided by AIC alues and biological impo ance, using
Kenwa d and Roge es ima ion [48]. Model compa isons we e
made using he maximum likelihood (ML) me hod, and inal
alues ob ained om he mos pa simonious model wi h es ic ed
maximum likelihood (REML). Sexes we e analysed sepa a ely
when possible, and model alidi y was e i ied ia he esidual
dis ibu ion. To assess o delayed e ec s o densi y, he da a we e
es ic ed o he inal wo apping occasions and each esponse
model inco po a ing cu en densi y compa ed o models includ-
ing densi ies o he wo p eceding apping occasions ( -1, -2).
Uns uc u ed epea ed co a ia e ype was employed in hese
educed models.
To acili a e in e p e a ion o a h ee-way in e ac ion be ween
densi y, ea men and apping occasion in he inal su i al
model (Table 1), a mixed model was cons uc ed wi h densi y o
explain su i al. Enclosu e, wi h in e cep , was se as a andom
ac o . Residuals o his model we e hen used as he esponse
a iable in a epea ed ANOVA model in which su i al was
explained by ea men , apping occasion and hei in e ac ion, as
pe he me hods desc ibed abo e.
Gene alized linea mixed models (PROC GLIMMIX), em-
ploying he same me hodology and ixed and andom ac o s, we e
used o e alua e changes in he p opo ion o males (sex a io),
oles weighing less han 20 g (as ep esen a i e o ju enile
ec ui men ), and ep oducing emales om he o al emale
popula ion. As ex e nal signs o ep oduc ion o ju eniles we e no
ye p esen a he onse o he expe imen , baseline da a we e
emo ed om hese models. Gene alized models we e assessed o
o e -dispe sion. Da a we e analysed in SAS e sion 9.3 (SAS
Ins i u e Inc., Ca y, NC).
Resul s
Popula ion size
The e ec o ea men on densi y changed o e ime (Table 1).
Densi ies we e simila among ea men g oups o he ini ial h ee
apping occasions (Figu e 1a). Howe e , by Sep embe densi ies
we e g ea e in high p o ein han low p o ein ea men
popula ions, while con ol g oups displayed an in e media e le el
o densi y, which did no di e om ei he supplemen a ion g oup.
Popula ion densi y was no in luenced by mean condi ion index
(F
condi ion index ( -1)
1, 38
= 2.47, P = 0.124).
Mean popula ion g ow h a es we e p edominan ly posi i e
h oughou he expe imen and a ied be ween ea men g oups
Die Quali y Limi s Summe Vole Popula ions
PLOS ONE | www.plosone.o g 3 Ma ch 2014 | Volume 9 | Issue 3 | e91113
(Table 1, Figu e 1b). Popula ion g ow h a e was nega i ely
associa ed wi h densi y in all ea men g oups (Table 1, Figu e 1c).
Demog aphics and su i al
Su i al a es di e ed wi h densi y, ea men and ime
(Table 1). Su i al was highe in low p o ein popula ions han in
o he g oups om June o July (Figu e 2). F om July o Sep embe ,
all a es s abilized wi h app oxima ely 70% o oles su i ing
be ween apping occasions (Table 1). Nei he ea men
(F
ea men 2, 18
= 0.80, P = 0.47) no apping occasion (F
occasion2, 33
= 0.96, P = 0.39) a ec ed su i al in he densi y-co ec ed model.
The sex a io o he popula ions was una ec ed by he
ea men s (Table 1, Figu e 3a). Meanwhile, he p opo ion o
ep oducing emales, ou o all emales, dec eased wi h inc easing
Table 1. Mos pa simonious model o explain each esponse a iable.
Response Sou ce o a ia ion Num. d Denom. d
FP
Densi y week 1 56 45.35 ,0.0001
ea men 2 51 6.35 0.0034
week 6 ea men 2 56 8.12 0.0008
G ow h a e occasion 2 22 3.08 0.07
ea men 2 11 4.22 0.0445
densi y 1 32 11.12 0.0022
Su i al occasion 2 32 6.55 0.0042
ea men 2 35 4.98 0.0126
densi y 1 36 4.71 0.0368
densi y 6 ea men 2 35 4.60 0.0168
densi y 6occasion 2 34 7.20 0.0025
ea men 6occasion 4 31 1.67 0.18
densi y 6 ea men 6occasion 4 33 2.76 0.0434
P op. males occasion 3 44 2.53 0.07
ea men 2 119 0.41 0.67
densi y 1 119 1.41 0.24
ea men 6occasion 6 119 0.89 0.50
P op. ep oducing emales occasion 2 26 2.94 0.07
ea men 2 51 0.78 0.47
densi y 1 51 7.12 0.0102
ea men 6occasion 4 51 1.04 0.40
P op. ,20 g occasion 2 227 3.06 0.06
ea men 2 79 0.13 0.89
densi y 1 79 7.62 0.0072
densi y 6occasion 2 79 3.36 0.0397
ea men 6occasion 4 79 5.27 0.0008
Final alues we e ob ained wi h REML. Full models con ained ime (week o apping occasion), ea men g oup, densi y, and all hei in e ac ions as ini ial explana o y
a iables. T apping occasion is a ca ego ical a iable. Week deno es he week o yea and is con inuous. Enclosu e and enclosu e 6 ime we e se as andom a iables.
doi:10.1371/jou nal.pone.0091113. 001
Figu e 1. Size and g ow h o expe imen al ield ole popula ions. (A) densi y (mean 6se), (B) g ow h a e (R ) (leas squa ed mean 6se), (C)
popula ion g ow h a e by densi y.
doi:10.1371/jou nal.pone.0091113.g001
Die Quali y Limi s Summe Vole Popula ions
PLOS ONE | www.plosone.o g 4 Ma ch 2014 | Volume 9 | Issue 3 | e91113
popula ion densi y, ega dless o ea men (Es ima e = 20.0161,
s.e = 0.006, Table 1, Figu e 3b). The p opo ion o ju enile oles
(,20 g) a ied be ween ea men g oups and apping occasions
(Table 1, Figu e 3c), being highes in con ol popula ions in July,
and lowes by Augus . High p o ein popula ions displayed he
g ea es p opo ion o ju eniles in Augus , bu lowes in
Sep embe .
Indica o s o condi ion
No associa ions be ween ea men and body condi ion index
we e iden i ied (Figu e 4a). No we e he e signi ican di e ences
be ween ea men g oups in male haema oc i (Figu e 4b). Male
haema oc i was nega i ely densi y dependen a he beginning o
he expe imen , bu he ela ionship dissipa ed wi h ime
(F
densi y6occasion
3, 62
= 4.35, P = 0.008). Densi y in he p e ious
apping occasion explained male haema oc i om Augus o
Sep embe be e han cu en densi y (DAIC = 2.0). Male
haema oc i hus exhibi ed delayed densi y dependence in high
p o ein popula ions (F
densi y( -1)6 ea men 2,17
= 4.22, P = 0.032). A
nega i e e ec o densi y on male albumin, ha was p esen
a he beginning o he expe imen , elaxed wi h ime
(F
densi y6occasion
3, 86
= 2.62, P = 0.056, Figu e 4c). Cu en densi y
explained male albumin le els be e han pas densi y (DAIC
= 4.7), bu none o he explana o y a iables eached signi icance.
Meanwhile, male IgG was highe in June ( = 2.69, d. = 5,
P = 0.046) and Sep embe ( = 3.55, d. = 12, P = 0.004) han July,
bu did no a y be ween ea men g oups (F
occasion
3, 8
= 5.05,
P = 0.032, Figu e 4d). Densi y in he p e ious apping occasion
was again a be e p edic o o male IgG han cu en densi y
(DAIC = 3.2), bu none o he explana o y a iables eached
signi icance.
Female haema oc i consis en ly inc eased in high p o ein
popula ions du ing he expe imen (Figu e 5a). Howe e , in e -
p e a ion is con ounded by h ee-way in e ac ions wi h bo h
cu en and pas densi y (F
densi y6occasion6 ea men
6, 34
= 33.8,
P = 0.039; F
densi y( -2)6occasion6 ea men
2, 77
= 3.18, P = 0.047).
Meanwhile, no signi ican e ec s on emale albumin we e
iden i ied (Figu e 5b), including densi y wo occasions p io ,
which explained he da a be e han cu en densi y (DAIC
= 6.6). No e ec s o ea men g oup we e iden i ied in emale
o al IgG in he ull model (Figu e 5c). Howe e , a delayed
densi y-dependen dec ease in emale IgG p esen in Augus ,
had disappea ed by Sep embe (F
densi y( -1)6occasion1, 93
= 7.56,
P = 0.007).
Discussion
Consis en wi h ou hypo hesis, he summe g ow h o ole
popula ions was limi ed by he a ailabili y o die a y p o eins.
Food esou ces a e o g ea impo ance o he popula ion
dynamics o he bi o es [49–51], including cyclic small mammals
[17,52–53]. In gene al, he quan i a i e e ec s o esou ces on
e eb a e he bi o e popula ions ha e been ex ensi ely s udied
[54], while he e ec s o ood esou ce quali y emain li le
in es iga ed. Simila ly, in oles he limi ing e ec s o ood on
popula ion demog aphy has mani es ed h ough quan i y, and
p edomina ely only du ing win e [18]. As such, ou expe imen al
esul s o e impo an insigh s in o p ocesses con ibu ing o
a ia ion in he bi o e densi y – namely he a ailabili y o high-
quali y ood du ing he g owing season.
Su i al a es o oles did no di e be ween ea men g oups
du ing he expe imen . The e o e, he obse ed di e ences in
densi y a e la gely a ibu able o inc eased ec ui men h ough
ep oduc ion. This is suppo ed by he endency o high p o ein
popula ions o consis o ew males and many ep oducing emales
as compa ed o he o he ea men g oups. Howe e , inc eased
a es o ep oduc ion we e no e lec ed in he p opo ion o
ju enile oles, which we e lowes in Sep embe when he
popula ion g ow h a e was highes . Desy and Ba zli [12] iden i ied
he same peculia i y which hey a ibu ed o as e g ow h o
ju eniles wi h ood supplemen a ion. In o he wo ds, high quali y
ood enables oles o g ow as e han lowe quali y ood. Indeed,
p o ein supplemen a ion has been ound o accele a e he g ow h
o small oden indi iduals [55]. Fo his eason, i was no
app op ia e o e alua e unc ional g oup di e ences in su i al
and condi ion be ween ju enile and adul oles in he cu en
expe imen . I should be no ed ha he quali y o a ailable ood
esou ces may a ec he demog aphic a es o ole popula ions
di e en ly in win e han in summe – his emains a opic o
u he expe imen a ion.
In e es ingly, con ol ea men g oups in ou expe imen
a ained app oxima ely hal he densi ies o high p o ein
Figu e 2. T ea men -wise popula ion su i al a e (mean ±se).
doi:10.1371/jou nal.pone.0091113.g002
Figu e 3. Demog aphic a ibu es o expe imen al popula ions (leas squa ed mean ±se). (A) p opo ion o males in o al popula ion, (B)
p opo ion o ep oducing (p egnan and/o lac a ing) emales in o al emale popula ion, (C) p opo ion o oles ,20 g om o al popula ion.
doi:10.1371/jou nal.pone.0091113.g003
Die Quali y Limi s Summe Vole Popula ions
PLOS ONE | www.plosone.o g 5 Ma ch 2014 | Volume 9 | Issue 3 | e91113

supplemen ed g oups (Fig. 1). The p o ein con en o ou high-
quali y supplemen a ion was 30%, while c ude p o ein le els in
g asses (including Phleum p a ense) a he end o he g owing season
a e abou 10–15% o d y weigh [56]. I is he e o e emp ing o
en e ain he idea ha summe ole densi ies closely e lec he
le els o die a y p o ein a ailable o oles in hei o age.
Con a y o ou p edic ions, nei he body condi ion no
haema ological indices we e clea ly associa ed wi h expe imen al
ea men s o popula ion densi y. Ne e heless, he iden i ica ion
o changes o e ime, and in e ac ions wi h cu en and pas
densi y, highligh he complex in e ac ions and po en ial u ili y o
hese measu es in popula ion ecology esea ch. I should be no ed
ha in e p e a ion o haema ological indices is di icul and se e al
pa ame e s a e usually equi ed o p o ide an adequa e ep esen-
a ion o heal h s a us [57]. Fo example, ele a ed o al IgG could
ep esen high baseline immuni y le els esul ing om good heal h
Figu e 4. Condi ion indices o male oles om expe imen al popula ions (mean ±se). (A) body condi ion index, (B) haema oc i , (C)
plasma albumin, (D) o al IgG an ibody i e .
doi:10.1371/jou nal.pone.0091113.g004
Figu e 5. Condi ion indices o emale oles om expe imen al popula ions (mean ±se). (A) haema oc i , (B) plasma albumin, (C) o al IgG
an ibody i e .
doi:10.1371/jou nal.pone.0091113.g005
Die Quali y Limi s Summe Vole Popula ions
PLOS ONE | www.plosone.o g 6 Ma ch 2014 | Volume 9 | Issue 3 | e91113
o an immune esponse o in ec ion [39]. Simila ly, low albumin
may be a sign o a p o ein de icien die o in ec ion [58]. Since we
employed se e al heal h indica o s wi hou obse ing ea men
e ec s, i appea s ha oles we e able o main ain good
physiological condi ion du ing he b eeding season on na u al
ood esou ces alone ( o con as ing esul s du ing he non-
b eeding season, see [39]). In he con ex o ou ea men s, i
seems plausible ha oles which ecei ed supplemen al p o ein
we e alloca ing i o emos o ep oduc ion, as opposed o
ele a ing hei own physiological condi ion (i.e., income b eeding
[59]).
Ko pela e al. ecen ly highligh ed he associa ion be ween
summe g owing condi ions and he dynamics o ole popula ions
[19]. Howe e , hei s udy did no iden i y p oxima e mechanisms
ac ing du ing summe . Clima e, amongs o he hings, has been
shown o al e ni ogen le els in plan s [31–32], and we ha e
demons a ed he e ha summe p o ein le els a e a plausible
mechanis ic link be ween clima e and ole popula ion demog a-
phy. In u he suppo , a ‘midsumme c isis’, hypo hesized o
esul om a sho age o high-quali y ood due o g aminoid
senescence [15], did no p esen in high p o ein ea men g oups.
Meanwhile, he g ow h a e o low p o ein popula ions was clea ly
educed be ween he inal wo apping in e als wi hou ob ious
changes in su i al a es. Fu he esea ch is none heless needed o
elucida e causali ies be ween clima e and he bi o e die quali y.
Conside able deba e has ocused on ac o s which limi and
egula e cyclic popula ions o small mammals. A common line o
di e en ia ion is be ween in insic ( o example age s uc u e and
ma e nal o ju enile en i onmen : see [60–63]) and ex insic
ac o s ( he en i onmen , including p eda ion: see [6,9]. Howe e ,
ecen ansplan expe imen s ha e p o ided compelling e idence
o suppo an impo an e ec o he immedia e en i onmen on
he li e his o y ai s o oles [64–65] (see howe e [63]). Ou
iden i ica ion o die quali y limi a ion on inc easing ole
popula ions is consis en wi h he la e indings. Speci ically, we
ha e demons a ed ha p o ein a ailabili y limi s he g ow h o
summe ole popula ions. We he e o e conclude ha die quali y,
ul ima ely de e mined by s ochas ic a ia ion in clima e, is likely o
ha e a hi he o unde es ima ed in luence on he popula ion
dynamics o small mammals.
Acknowledgmen s
We hank Jean-Emmanuel Fou nie , S ephen Ryan, Ilkka Taponen and
Anaı
¨s Zimme o hei con ibu ions o ield wo k, Nu ia Blanco and Sami
Ky o¨la¨inen o assis ance wi h ELISA es s, and Te o Klemola o ad ice
ega ding abundance modelling and s a is ical analyses.
Au ho Con ibu ions
Concei ed and designed he expe imen s: KF PS TM HH OH. Pe o med
he expe imen s: KF PS TM OH. Analyzed he da a: KF KH OH.
Con ibu ed eagen s/ma e ials/analysis ools: TM OH. W o e he pape :
KF PS TM KH HH OH.
Re e ences
1. Hansson L, Hen onen H (1985a) G adien s in densi y a ia ions o small
oden s: he impo ance o la i ude and snow co e . Oecologia 67: 394–402.
2. Hansson L, Hen onen H (1985b) Regional di e ences in cyclici y and
ep oduc ion in Cle h ionomys species: a e hey ela ed? Ann Zool Fenici 22:
277–288.
3. No dahl K (1995) Popula ion cycles in no he n small mammals. Biol Re
Camb Philos Soc 70: 621–637.
4. Kallio ER, Begon M, Hen onen H, Koskela E, Mappes T, e al. (2009) Cyclic
han a i us epidemics in humans – p edic ed by oden hos dynamics. Epidemics
1: 101–107.
5. Rikalainen K, Aspi J, Gala za JA, Koskela E, Mappes T (2012) Main enance o
gene ic di e si y in cyclic popula ions - a longi udinal analysis in Myodes gla eolus.
Ecol E ol 2: 1491–1502.
6. Hanski I, Hansson L, Hen onen H (1991) Specialis p eda o s, gene alis
p eda o s, and he mic o ine oden cycle. J Anim Ecol 60: 353–367.
7. Hanski I, Tu chin P, Ko pima¨ki E, Hen onen H (1993) Popula ion oscilla ions
o bo eal oden s: egula ion by mus elid p eda o s leads o chaos. Na u e 364:
232–234.
8. Hanski I, Hen onen H, Ko pima¨ki E, Oksanen L, Tu chin P (2001) Small-
oden dynamics and p eda ion. Ecology 82: 1505–1520.
9. Ko pima¨ki E, Oksanen L, Oksanen T, Klemola T, No dahl K, e al. (2005)
Vole cycles and p eda ion in empe a e and bo eal zones o Eu ope. J Anim Ecol
74: 1150–1159.
10. Dalk is T, Sibly RM, Topping CJ (2011) How p eda ion and landscape
agmen a ion a ec ole popula ion dynamics. PLoS One 6: e22834.
11. Lambin X, B e agnolle V, Yoccoz NG (1996) Vole popula ion cycles in no he n
and sou he n Eu ope: is he e a need o di e en explana ions o a single
p oblem? J Anim Ecol 75: 340–349.
12. Desy EA, Ba zli GO (1989) E ec s o ood a ailabili y and p eda ion on p ai ie
ole demog aphy: a ield expe imen . Ecology 70: 411–421.
13. And eassen HP, Glo igen P, Re´my A, Ims RA (2013) New iews on how
popula ion-in insic and communi y-ex insic p ocesses in e ac du ing he ole
popula ion cycles. Oikos 122: 507–515.
14. K ebs CJ (2013) Popula ion luc ua ions in oden s. Chicago: The Uni e si y o
Chicago P ess.
15. Myllyma¨ki A (1977) Demog aphic mechanisms in he luc ua ing popula ions o
he ield ole Mic o us ag es is. Oikos 29: 468–493.
16. No dahl K, Ko pima¨ki E (2002) Changes in indi idual quali y du ing a 3-yea
popula ion cycle o oles. Oecologia 130: 239–249.
17. Klemola T, Koi ula M, Ko pima¨ki E, No dahl K (2000) Expe imen al es s o
p eda ion and ood hypo heses o popula ion cycles o oles. P oc R Soc
Lond B Biol Sci 267: 351–356.
18. Hui u O, Koi ula M, Ko pima¨ki E, Klemola T, No dahl K (2003) Win e ood
supply limi s g ow h o no he n ole popula ions in he absence o p eda ion.
Ecology 84: 2108–2118.
19. Ko pela K, Delgado M, Hen onen H, Ko pima¨ki E, Koskela E, e al. (2013)
Nonlinea e ec s o clima e on bo eal oden dynamics: mild win e s do no
nega e high-ampli ude cycles. Glob Chang Biol 19: 697–710.
20. Tkadlec E, S ense h NC (2001) A new geog aphical g adien in ole popula ion
dynamics. P oc R Soc Lond B Biol Sci 268: 1547–1552.
21. Co nulie T, Yoccoz NG, B e agnolle V, B omme JE, Bu e A, e al. (2013)
Eu ope-wide dampening o popula ion cycles in keys one he bi o es. Science
340: 63–66.
22. Ho¨ n eld B, Hipkiss T, Eklund U (2005) Fading ou o ole and p eda o cycles?
P oc R Soc Lond B Biol Sci 272: 2045–2049.
23. Ims RA, Henden JA, Killeng een ST (2008) Collapsing popula ion cycles.
T ends Ecol E ol 23: 79–86.
24. Kaus ud KL, Mys e ud A, S een H, Vik JO, Øs bye E, e al. (2008) Linking
clima e change o lemming cycles. Na u e 456: s93–97.
25. Chapin SF, Johnson DA, McKend ick JD (1980) Seasonal mo emen o
nu ien s in plan s o di e ing g ow h om an Alaskan und a ecosys em:
implica ions o he bi o y. J Ecol 68: 189–209.
26. Jo´nsdo´ i IS, Khi un O, S ens o¨m A (2005) Biomass and nu ien esponses o a
clonal und a sedge o clima e wa ming. Can J Bo 83: 1608–1621.
27. Ma son WJ J (1980) He bi o y in ela ion o plan ni ogen con en . Annu Re
Ecol Sys 11: 119–161.
28. Sche e TA, Lochmille RL, Leslie DM J , Engle DM, Pay on ME (1998)
Examina ion o he ni ogen limi a ion hypo hesis in non-cyclic popula ions o
co on a s (Sigmodon hispidus). J Anim Ecol 67: 705–721.
29. Else JJ, Fagan WF, Denno RF, Dobbe uhl DR, Fola in A, e al. (2000)
Nu i ional cons ain s in e es ial and eshwa e ood webs. Na u e 6812:
578.
30. Whi e TCR (2005) Why does he wo ld s ay g een? Nu i ion and su i al o
plan -ea e s. Collingwood: CSIRO Publishing.
31. Kumme ow J, McMas e GS, K ause DA (1980) Tempe a u e e ec on g ow h
and nu ien con en s in E iopho um agina um unde con olled en i onmen al
condi ions. A c Alp Res 12: 335–341.
32. Laine KM, Hen onen H (1987) Phenolics/ni ogen a ios in he bluebe y
Vaccinium my illus in ela ion o empe a u e and mic o ine densi y in Finnish
Lapland. Oikos 50: 389–395.
33. Cole R, Ba zli GO (1979) Nu i ion and popula ion dynamics o he p ai ie ole,
Mic o us och ogas e , in cen al Illinois. J Anim Ecol 48: 455–470.
34. Chi y D (1952) Mo ali y among oles (Mic o us ag es is) a Lake Vy nwy,
Mon gome yshi e, in 1936-9. Philos T ans R Soc Lond B Biol Sci 236: 505–552.
35. Boons a R, K ebs C (1979) Viabili y o la ge- and small-sized adul s in
luc ua ing ole popula ions. Ecology 60: 567–573.
36. Bu he SJ, Lambin X, Tel e S, Douglas A, Beldomenico P, e al. (2010)
Indi idual g ow h a es in na u al ield ole, Mic o us ag es is, popula ions
exhibi ing cyclic popula ion dynamics. Oecologia 162: 653–661.
Die Quali y Limi s Summe Vole Popula ions
PLOS ONE | www.plosone.o g 7 Ma ch 2014 | Volume 9 | Issue 3 | e91113
37. Mi a A, Ma hias ML (1994) Seasonal e ec s on he hema ology and blood
plasma p o eins o wo species o mice Mus musculus domes icus and M. sp e us
( oden ia: mu idea) om Po ugal. Hys ix 5: 63–72.
38. Lochmille RL, Dee enbe g C (2000) T ade-o s in e olu iona y immuni y: jus
wha is he cos o immuni y? Oikos 88: 87–98.
39. Hui u O, Jokinen I, Ko pima¨ki E, Koskela E, Mappes T (2007) Phase
dependence in win e physiological condi ion o cyclic oles. Oikos 116: 565–
577.
40. Schoech SJ (2009) Food supplemen a ion expe imen s: a ool o e eal
mechanisms ha media e iming o ep oduc ion. In eg Comp Biol 49: 480–
492.
41. Hen onen H (1987) The impac o spacing beha iou in mic o ine oden s on
he dynamics o leas weasels Mus ela ni alis – a hypo hesis. Oikos 50:366–370.
42. O is DL, Bu nham KP, Whi e GC, Ande son DR (1978) S a is ical in e ence
om cap u e da a on closed animal popula ions. Wildli e Monog aphs 62: 1–
135.
43. Pollock KH, O o MC (1983) Robus es ima ion o popula ion size in closed
animal popula ions om cap u e- ecap u e expe imen s. Biome ics 39: 1035–
1049.
44. Sibly RM, Hone J (2002) Popula ion g ow h a e and i s de e minan s: an
o e iew. Philos T ans R Soc Lond B Biol Sci 357: 1153–1170.
45. Whi e GC, Bu nham KP (1999) P og am MARK: su i al es ima ion om
popula ions o ma ked animals. Bi d S udy 46: 120–138.
46. Bu nham KP, Ande son DR (2002) Model selec ion and in e e ence: a p ac ical
in o ma ion- heo e ic app oach (2nd edi ion). New Yo k: Sp inge .
47. Schul e-Hos edde AI, Milla JS, Hickling GJ (2001) E alua ing body condi ion
in small mammals. Can J Zool 79: 1021–1029.
48. Li ell RC, Milliken GA, S oup WW, Wol inge RD (2006) SAS sys em o
mixed models. SAS Ins i u e, Ca y.
49. Lack D (1954) The na u al egula ion o animal numbe s. London: Ox o d
Uni e si y P ess.
50. Whi e TCR (1978) The impo ance o a ela i e sho age o ood in animal
ecology. Oecologia 33: 71–86.
51. Sinclai ARE, K ebs CJ (2002) Complex nume ical esponses o op-down and
bo om-up p ocesses in e eb a e popula ions. Phil T ans R Soc Lond B 357:
1221–1231.
52. Hansson L (1979) Food as a limi ing ac o o small oden numbe s. Oecologia
37: 297–314.
53. Koskela E, Jonsson P, Ha ikainen T, Mappes T (1998) Limi a ion o
ep oduc i e success by ood a ailabili y and li e size in he bank ole,
Cle h ionomys gla eolus. P oc R Soc Lond B Biol Sci 265: 1129–1134.
54. Bou in S (1990) Food supplemen a ion expe imen s wi h e es ial e eb a es;
pa e ns, p oblems and he u u e. Can J Zool 68: 203–220.
55. Came on GN, Eshelman BD (1996) G ow h and ep oduc ion o hispid co on
a s (Sigmodon hispidus) in esponse o na u ally occu ing le els o die a y p o ein.
J Mammal 77: 220–231.
56. Suleiman A, Okine EK, Goonewa dene LA, Day PA, Ya emcio B, e al. (1999)
Yield and eeding o p ai ie g asses in eas -cen al Albe a. J Range Manage 52:
75–82.
57. Beldomenico PM, Tel e S, Gebe S, Lukomski L, Benne M, e al. (2008) The
dynamics o heal h in wild ield ole popula ions: a haema ological pe spec i e.
J Anim Ecol 77: 984–997.
58. F ommel TO, Seed JR, Sechelski J (1988) Changes in albumin le els in blood
and u ine o Mic o us mon anus ch onically in ec ed wi h T ypanosoma b ucei
gambiense. J Pa asi ol 74: 957–962.
59. Jo¨nsson KI (1997) Capi al and income b eeding as al e na i e ac ics o esou ce
use in ep oduc ion. Oikos 78: 57–66.
60. Boons a R (1994) Popula ion cycles in mic o ines: he senescence hypo hesis.
E o Ecol 8: 196–219.
61. Inchaus i I, Ginzbu g LR (1998) Small mammal cycles o no he n Eu ope:
pa e ns and e idence o a ma e nal e ec hypo hesis. J Anim Ecol 67: 180–194.
62. Mosseau TA, Fox CW (1998) The adap i e signi icance o ma e nal e ec s.
T ends Ecol E ol 13: 403–407.
63. Helle H, Koskela E, Mappes T (2012) Li e in a ying en i onmen s:
expe imen al e idence o delayed e ec s o ju enile en i onmen on adul li e
his o y. J Anim Ecol 81: 573–582.
64. E gon T, Lambin X, S ense h NC (2001) Li e-his o y ai s o oles in a
luc ua ing popula ion espond o he immedia e en i onmen . Na u e 411:
1043–1045.
65. Speakman JR, E gon T, Ca anagh R, Reid K, Scan lebu y DM, e al. (2003)
Res ing and daily ene gy expendi u es o ee-li ing ield oles a e posi i ely
co ela ed bu e lec ex insic a he han in insic e ec s. PNAS 100: 14057–
14062.
Die Quali y Limi s Summe Vole Popula ions
PLOS ONE | www.plosone.o g 8 Ma ch 2014 | Volume 9 | Issue 3 | e91113