RESEARCH ARTICLE
Quan i ying in si u pheno ypic a iabili y in he
hyd aulic p ope ies o ou ee species
ac oss hei dis ibu ion ange in Eu ope
N. Gonza
´lez-Muñoz
1
*, F. S e ck
2
, J. M. To es-Ruiz
1
, G. Pe i
3
, H. Cocha d
4
, G. on A x
5,6
,
A. Lin unen
7
, M. C. Caldei a
8
, G. Capde ille
1
, P. Copini
2,9
, R. Gebaue
10
, L. G o
¨nlund
7
,
T. Ho
¨l a
¨
7
, R. Lobo-do-Vale
8
, M. Pel oniemi
11
, A. S i ih
12
, J. U ban
10
, S. Delzon
1
1BIOGECO, INRA, Uni e si e
´de Bo deaux, Pessac, F ance, 2Fo es Ecology and Fo es Managemen
G oup, Wageningen Uni e si y & Resea ch, Wageningen, The Ne he lands, 3Uni e si àdegli S udi di
Pado a, Dep. TeSAF, Legna o (PD), I aly, 4PIAF, INRA, Uni e si e
´Cle mon -Au e gne, Cle mon -Fe and,
F ance, 5Swiss Fede al Ins i u e o Fo es , Snow and Landscape Resea ch WSL, Bi mensdo , Swi ze land,
6Clima ic Change and Clima e Impac s, Ins i u e o En i onmen al Sciences, Gene a, Swi ze land,
7Depa men o Fo es Sciences, Uni e si y o Helsinki, Helsinki, Finland, 8Fo es Resea ch Cen e, School
o Ag icul u e, Uni e si y o Lisbon, Tapada da Ajuda, Lisboa, Po ugal, 9Wageningen En i onmen al
Resea ch (Al e a), Wageningen, The Ne he lands, 10 Depa men o Fo es Bo any, Dend ology and
Geobiocoenology, Mendel Uni e si y, Zemědělska
´3, B no, Czech Republic, 11 Na u al Resou ces Ins i u e
Finland (Luke), La oka anonkaa i 9, Helsinki, Finland, 12 Swiss Fede al Ins i u e o Technology ETH,
Planning o Landscape and U ban Sys ems, Zu ich, Swi ze land
*[email protected]
Abs ac
Many s udies ha e epo ed ha hyd aulic p ope ies a y conside ably be ween ee spe-
cies, bu li le is known abou hei in aspeci ic a ia ion and, he e o e, hei capaci y o
adap o a wa me and d ie clima e. He e, we quan i y pheno ypic di e gence and clinal a -
ia ion o embolism esis ance, hyd aulic conduc i i y and b anch g ow h, in ou ee spe-
cies, wo angiospe ms (Be ula pendula,Populus emula) and wo coni e s (Picea abies,
Pinus syl es is), ac oss hei la i udinal dis ibu ion in Eu ope. G ow h and hyd aulic e i-
ciency a ied widely wi hin species and be ween popula ions. The a iabili y o embolism
esis ance was in gene al weake han ha o g ow h and hyd aulic e iciency, and e y low
o all species bu Populus emula. In addi ion, no and weak suppo o a sa e y s. e i-
ciency ade-o was obse ed o he angiospe m and coni e species, espec i ely. The lim-
i ed a iabili y o embolism esis ance obse ed he e o all species excep Populus emula,
sugges s ha o es popula ions will unlikely be able o adap hyd aulically o d ie condi ions
h ough he e olu ion o embolism esis ance.
In oduc ion
Massi e o es mo ali y e en s due o d ough s ess and ising empe a u es ha e been
obse ed a he global and egional scales [1–5]. Conside ing ha clima e change models p e-
dic u he inc eases in mean empe a u e and in he equency and se e i y o ex eme
d ough e en s [6], mo e nega i e impac s on ee su i al a e expec ed [7]. In his con ex ,
PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0196075 May 1, 2018 1 / 17
a1111111111
a1111111111
a1111111111
a1111111111
a1111111111
OPEN ACCESS
Ci a ion: Gonza
´lez-Muñoz N, S e ck F, To es-Ruiz
JM, Pe i G, Cocha d H, on A x G, e al. (2018)
Quan i ying in si u pheno ypic a iabili y in he
hyd aulic p ope ies o ou ee species ac oss
hei dis ibu ion ange in Eu ope. PLoS ONE 13(5):
e0196075. h ps://doi.o g/10.1371/jou nal.
pone.0196075
Edi o : Be hold Heinze, Aus ian Fede al Resea ch
Cen e o Fo es s BFW, AUSTRIA
Recei ed: July 5, 2017
Accep ed: Ap il 5, 2018
Published: May 1, 2018
Copy igh : ©2018 Gonza
´lez-Muñoz 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 : All da a unde lying
he s udy a e a ailable a he D yad Digi al
Reposi o y (h ps://doi.o g/10.5061/d yad.
b2pg468).
Funding: This a icle is based on wo k om COST
Ac ion FP1106 STReESS, suppo ed by COST
(Eu opean Coope a ion in Science and
Technology). This s udy was unded in pa by he
“In es men s o he Fu u e” p og amme (g an no.
ANR-10-EQPX-16, XYLOFOREST) om he F ench
assessmen s o he p ope ies associa ed wi h d ough esis ance in ees, and o he capaci y o
species o deal wi h en i onmen al changes, may help us o an icipa e he impac o clima e
change on o es ee species.
Hyd aulic ailu e due o xylem embolism is now conside ed one o he main causes o
d ough -induced ee mo ali y [8–10]. When soil wa e po en ial d ops due o wa e sho age,
he ension o he xylem wa e column inc eases, p omo ing he o ma ion o embolisms ha
educe he hyd aulic unc ioning o he plan [11,12]. In cases o p olonged d ough , soil wa e
po en ial con inues o all, igge ing he sp ead o embolisms h oughou he xylem condui
ne wo k, leading o he hyd aulic dys unc ion o he plan ascula sys em and, inally, o le hal
damage o he plan [13,14,15]. The e o e, de e mining he esis ance o embolism o he spe-
cies is c ucial o e alua ing he consequences ha he expec ed inc ease in d ough e en e-
quency can ha e on a gi en popula ion, o es o biome. P
50
is he xylem p essu e a which
50% o conduc i i y is los due o embolism o ma ion, and i is widely used o assess plan
hyd aulic sa e y o embolism. Xylem-speci ic hyd aulic conduc i i y (K
S
), i.e. he a e o wa e
anspo h ough a gi en a ea o sapwood pe uni p essu e di e ence and pe uni leng h, is
commonly used o assess hyd aulic e iciency [12]. Ac oss species, li e a u e shows a weak co -
ela ion be ween hyd aulic sa e y and hyd aulic e iciency, bu he absence o species display-
ing bo h high hyd aulic e iciency and sa e y sugges s a possible sa e y-e iciency ade-o [16].
In coni e s, P
50
and K
s
a e only weakly co ela ed, as embolism esis ance is d i en mos ly by
he o us-ape u e o e lap in pi pai s [17–19], whe eas xylem hyd aulic e iciency is no in lu-
enced by his pi ai . By con as , in angiospe ms, bo h P
50
and K
s
a e associa ed wi h pi
memb ane s uc u e [20–22] and hickness [23], as well as wi h he pe o a ion s uc u e
[24,25].
Di e ences in esis ance o embolism, i.e. in P
50
, ac oss species ha e been widely epo ed
[17,19,26,27,28]. Howe e , less a en ion has been paid o wi hin-species pheno ypic a ia ion
in his hyd aulic p ope y. Pheno ypic a iabili y esul s om a combina ion o gene ic a ia-
ion (di e ences in geno ype among di e en indi iduals wi hin he popula ion and be ween
popula ions) and pheno ypic plas ici y (geno ype p ope y o ende di e en pheno ypes in
di e en en i onmen s [29]), and de ines he capaci y o popula ions o succeed unde chang-
ing en i onmen al condi ions [30,31]. Low le els o pheno ypic a iabili y ac oss la ge spa ial
scales may indica e a low po en ial o species o adap o ongoing clima e change. Con a y o
o he plan unc ional p ope ies (see o ins ance [32] o lea phenology, [33] o lea unc-
ional ai s), p e ious wo ks on hyd aulic p ope ies show ha pheno ypic di e ences wi hin
species a e by a lowe han hose ound ac oss species [34–38], and hese di e ences a e e en
smalle in gymnospe ms han in angiospe ms [39]. Fo ins ance, Lamy e al. 2011, 2014 ound
nei he pheno ypic a iabili y in si u and no gene ic di e en ia ion be ween ma i ime pine
popula ions and sugges ed uni o m selec ion a he han gene ic d i , o P
50
. Howe e ,
whe he he pheno ypic a ia ion o hyd aulic p ope ies a ies ac oss species dis ibu ion
anges emains la gely unexplo ed. Fu he mo e, s udies assessing he ex en o which pheno-
ypic a iabili y in hyd aulic p ope ies is lowe han ha o o he key species ai s o e la ge
scales a e also lacking.
The main aim o his s udy was o e alua e pheno ypic a iabili y in he unc ional hyd au-
lic sa e y and e iciency (P
50
and K
S
) o ou Eu opean ee species ( wo coni e s and wo angio-
spe ms) along a la i udinal g adien co e ing mos o hei dis ibu ion ange. We assessed he
capaci y o he species o adap o changing en i onmen al condi ions by explo ing he links
be ween hyd aulic p ope ies and la i ude and clima e. We also e alua ed he pheno ypic a i-
abili y o b anch g ow h in ees om he same popula ions o assess he ex en o which
hyd aulic p ope ies we e conse ed ela i e o o he key ai s. Finally, we assessed he sa e y-
e iciency ade-o a in aspeci ic le el. We hypo hesized 1) a weake pheno ypic a ia ion o
Pheno ypic a iabili y in hyd aulic ai s
PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0196075 May 1, 2018 2 / 17
Na ional Agency o Resea ch and he Clus e o
Excellence COTE (ANR-10-LABX-45, wi hin he
DEFI p ojec ) o SD. NGM was suppo ed by he
Ag eenskills+ Fellowship P og amme, which has
ecei ed unding om he EU’s Se en h
F amewo k P og amme unde g an ag eemen
No. FP7-26719 (Ag eenskills+ con ac ). G A was
suppo ed by a g an om he Swiss S a e
Sec e a ia o Educa ion, Resea ch and Inno a ion
SERI (SBFI C14.0104). RLdV was unded by a
pos doc o al ellowship om he Po uguese
Fundac¸ão pa a a Ciência e a Tecnologia (FCT;
SFRH/BPD/86938/2012). RG and JU we e
suppo ed by he Minis y o Educa ion, Spo s and
You h o he Czech Republic (COST-LD13017). MP
was suppo ed by EU Li e P og amme (LIFE12
ENV/FI/000409). Fo es Resea ch Cen e (CEF,
School o Ag icul u e, Uni e si y o Lisbon) is a
esea ch uni unded by FCT (UID/AGR/00239/
2013).
Compe ing in e es s: The au ho s ha e decla ed
ha no compe ing in e es s exis .
hyd aulic sa e y han o hyd aulic e iciency and b anch g ow h, gi en he highly conse ed
e olu iona y na u e o P
50
[37,38]; 2) a pheno ypic cline -a g adual change o a pheno ypic
cha ac e in a species o e a geog aphical a ea- in bo h hyd aulic sa e y and e iciency and 3) a
weak sa e y-e iciency ade-o wi hin species. This s udy p o ides o he i s ime a mul i-
species assessmen o in e and in a-speci ic pheno ypic a iabili y in unc ional hyd aulic
p ope ies along a la ge la i udinal g adien . Ou esul s will help o cha ac e ize he adap i e
capaci ies o Eu opean o es s, which will ha e o ace d ie and wa me clima ic condi ions in
he u u e.
Ma e ials and me hods
S udy species and popula ions
We ocused on ou widely dis ibu ed Eu opean species, wi h di e en wa e - anspo s uc-
u es, om di use po ous wi h scala i o m pe o a ion pla es (Be ula pendula Ro h) o wi h
simple pe o a ion pla es (Populus emula L.) o so wood ( wo acheid-bea ing species, Picea
abies (L.) Ka s and Pinus syl es is L.). Fo each species, ou o six popula ions we e selec ed
ac oss hei dis ibu ion ange (see he dis ibu ion ange o he species and he loca ion o he
popula ions in S1 Fig). The mean annual empe a u e and o al annual ain all ac oss selec ed
popula ions anged om -1.8 o 9.5˚C and 538 o 1739 mm, espec i ely (S1 Table). We also
selec ed wo di e en si es a ew kilome es apa , o each popula ion.
Clima ic da a
Da a o mean annual empe a u e (MAT) and o al annual p ecipi a ion (MAP) we e
ob ained om Wo ldClim o iginal 30-s da a (h p://www.wo ldclim.o g/bioclim) [40] down-
scaled o 100-m esolu ion based on a high- esolu ion digi al ele a ion model (DEM) and
mo ing window eg ession echnique [41] o all bu he I alian (IT) and Swiss popula ions
(SW-LOE and SW-PFY). The MAP da a o he IT (I aly), SW-LOE (Swi ze land-Loe schen-
al) and SW-PFY (Swi ze land- P ynwald) popula ions we e ob ained om nea by wea he s a-
ions a San Vi o di Cado e (Cen e o Alpine En i onmen S udies) and Sie e (www.
me eoswiss.ch), espec i ely, due o conside able a ia ions in opog aphy. The a idi y index
(AI) was calcula ed as MAP/PET ( o al annual p ecipi a ion/annual po en ial e apo anspi a-
ion). PET was ex ac ed om he Global A idi y and PET Da abase (h p://www.cgia -csi.
o g). We a e aged he mean empe a u es (T_Sum) o a idi y indices (AI_Sum) o June, July
and Augus o ob ain mean alues o he summe (see S1 Table o he clima ic condi ions o
he popula ions s udied).
Xylem ulne abili y o embolism
We collec ed b anches om i e o 11 heal hy ma u e ees pe popula ion in he ea ly mo n-
ing du ing he we season (sp ing 2015). One o wo b anches wi h h ee o i e unc ional
ings we e sampled a mid-c own and sou h o ien ed. Samples had a s anda d leng h o 45 cm.
T anspi a ion losses we e p e en ed by emo ing he lea es o needles immedia ely a e sam-
pling and w apping he b anches in mois pape o keep hem humid and cool (3˚C) un il he
measu emen o embolism esis ance (wi hin h ee weeks o sampling). The ba k was emo ed
om coni e b anches o p e en esin o ill he ca i on ese oi s (see below, [17]), and all
b anches we e ecu wi h a azo blade, unde wa e , o a s anda d leng h o 0.27 m. Fo each
angiospe m species, 10 samples pe species we e used o es he open essel a e ac [42] by
injec ing ai a 2 ba s a one end and no open essels we e de ec ed o any o hem.
Pheno ypic a iabili y in hyd aulic ai s
PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0196075 May 1, 2018 3 / 17
Vulne abili y o d ough -induced embolism was de e mined a he Ca iplace (Uni e si y o
Bo deaux, Talence, F ance; h p://syl ain-delzon.com/ca iplace) and INRA-Cle mon -Fe -
and acili ies, wi h he Ca i on echnique [43,44]. Samples we e in il a ed wi h a e e ence
ionic solu ion o 10 mm 25 KCl and 1 mm CaCl
2
in deionized ul apu e wa e . Cen i ugal
o ce was used o gene a e nega i e p essu e in o he xylem and induce ca i a ion. This
me hod allows o measu e xylem conduc ance unde nega i e p essu e using he cus om so -
wa e Ca iso 4.0 (Uni . Bo deaux, Pessac, F ance). Ini ially, he maximum conduc ance o
s em (K
max
, in m
2
MPa
-1
s
-1
) was calcula ed unde low xylem p essu es. The pe cen age loss o
conduc ance (PLC) o he s ems was calcula ed a di e en xylem p essu es (P
i
) om -0.8 o -5
MPa wi h he ollowing equa ion:
PLC ¼100 1 K
Kmax
 
We ob ained one ulne abili y cu e pe ee by measu ing one o wo o he collec ed
b anches. These ulne abili y cu es show he pe cen age loss o xylem conduc ance as a unc-
ion o xylem p essu e [17]. Fo each b anch, he ela ionship be ween PLC and xylem wa e
p essu e was i ed wi h he ollowing sigmoidal equa ion [45]:
PLC ¼100
1þexp S
25 ðPiP50Þ
  
whe e P
50
(MPa) is he xylem p essu e inducing a 50% loss o conduc i i y and S(% MPa
-1
) is
he slope o he ulne abili y cu e a he in lec ion poin . All sigmoidal unc ions we e signi i-
can and i ed wi h he NLIN p ocedu e in SAS ( e sion 9.4 SAS Ins i u e, Ca y, NC, USA).
The xylem-speci ic hyd aulic conduc i i y (K
s
, kg m
1
MPa
-1
s
-1
) was calcula ed by di iding he
hyd aulic conduc i i y measu ed a low speed by he sapwood a ea o he sample.
B anch g ow h measu emen s
We also collec ed one b anch pe ee om h ee o i e ees pe popula ion and si e. We
selec ed s aigh b anches and did no keep any sample wi h eac ion wood o ou measu e-
men s. The allome ic ela ionship be ween b anch adius (mm) and xylem age (numbe o
yea s) was used as a su oga e o ee g ow h, as adial b anch g ow h and ee g ow h pa -
e ns a e highly co ela ed [46]. The b anch su ace a ea and he numbe o ee ings we e sys-
ema ically measu ed a 70 cm om he b anch apex.
S a is ical analyses
We assessed he pheno ypic a iabili y o unc ional hyd aulic p ope ies (P
50
and K
s
) and
b anch g ow h (b anch adius/xylem age) in each species, by es ing he e ec o popula ion
and si e wi h nes ed ANOVAs, in which he popula ion and he si e nes ed in popula ion we e
conside ed ac o s. I s a is ically signi ican di e ences we e obse ed, pos -hoc Tukey es s
we e conduc ed o mul iple compa isons be ween popula ions. Be o e unning he ANOVAs,
we checked ha he da a sa is ied he assump ions o no mali y and homoscedas ici y. As es-
sel size can apidly inc ease wi h b anch size du ing ea ly yea s o ee g ow h, and hen may
ha e a po en ial e ec on hyd aulic conduc i i y, we es ed any po en ial co ela ion be ween
K
s
and b anch diame e . We also calcula ed he in e -popula ion and in aspeci ic coe icien s
o a ia ion (% CV
in e
and CV
sp
, espec i ely). Fo each species, Spea man´s o Pea son´s co -
ela ion coe icien s (depending on he linea i y condi ion) we e calcula ed be ween he a e -
aged by si e P
50
,K
s
and b anch g ow h and he la i ude and he i e clima ic a iables
Pheno ypic a iabili y in hyd aulic ai s
PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0196075 May 1, 2018 4 / 17
men ioned abo e. Finally, we also checked o in aspeci ic sa e y-e iciency ade-o s, wi h
Spea man´s o Pea son´s co ela ion es s.
S a is ical analyses we e pe o med wi h he R p ojec o s a is ical compu ing (R De elop-
men Co e Team, 2016) [47].
Resul s
Pheno ypic a iabili y ac oss species dis ibu ion anges
Xylem ulne abili y cu es ollowed a sigmoid unc ion in all species (Fig 1,S2 Fig), showing
he lack o an open essel a e ac and he accu acy o he esul s ob ained he e. Be ula pendula
and Picea abies showed, espec i ely, he lowes and highes esis ance o embolism o he ou
species e alua ed. The mean P
50
±SE (MPa) was -1.78 ±0.02 o Be ula pendula, -2.45 ±0.08
o Populus emula, -3.16 ±0.03 o Pinus syl es is, and -3.58 ±0.02 o Picea abies (Fig 1).
Di e ences in P
50
be ween popula ions we e obse ed o all species, whe eas he si e
(nes ed in popula ion) had an e ec on P
50
in all species bu Picea abies (Table 1,Fig 2). The
CV in P
50
was low o all species o he han Populus emula. The a iabili y in P
50
o Be ula
pendula,Picea abies and Pinus syl es is anged om 4.15 (CV
in e
o Picea abies) o 10.23%
(CV
sp
o Pinus syl es is), whe eas ha o Populus emula anged om 24.82 (CV
in e
) o
25.07% (CV
sp
) (Table 2). The high a iabili y obse ed o Populus emula was mos ly due o
he popula ion o Finland-Va¨ io¨(FI-VA), which had he leas nega i e P
50
alues o any o he
popula ions s udied (Fig 2).
Fig 1. Xylem ulne abili y cu es o each popula ion o he ou species s udied (Be ulapendula,Populus emula,Picea abies and Pinus
syl es is). The shaded band ep esen s he s anda d de ia ion. CR: Czech Republic; Fi-RU: Finland-Ruo sinkyla¨; NE: The Ne he lands; PO:
Po ugal; Fi-HYY: Finland-Hyy ia¨la¨; Fi-VA: Finland-Va¨ io¨; IT: I aly; SW-LOE: Swi ze land-Loe schen al; SW-PFY: Swi ze land- P ynwald.
h ps://doi.o g/10.1371/jou nal.pone.0196075.g001
Pheno ypic a iabili y in hyd aulic ai s
PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0196075 May 1, 2018 5 / 17

Va iabili y le els we e much highe o K
s
han o P
50
(Table 2,Fig 2). K
s
di e ed signi i-
can ly be ween popula ions, o all species o he han Populus emula (Table 2,Fig 2). Fo his
Table 1. E ec s o popula ion and si e on P
50
(MPa), xylem-speci ic hyd aulic conduc i i y (K
s
, kg m
-1
MPa
-1
s
-1
) and b anch g ow h (BG, es ima ed as b anch
adius/xylem age (mm/yea )) o s udy species, acco ding o nes ed ANOVAs. The F, p- alues and deg ees o eedom a e shown. Pop: popula ion.
Angiospe ms Coni e s
Be ula pendula Populus emula Picea abies Pinus sy es is
d F p d F p d F p d F p
Popula ion 3 9.978 <0.001 3124.885 <0.001 57.971 <0.001 515.284 <0.001
P
50
Si e (Pop) 4 7.894 <0.001 49.905 <0.001 6 1.042 0.405 6 6.505 <0.001
Popula ion 3 34.949 <0.001 2 0.249 0.781 5 31.381 <0.001 54.810 <0.001
K
s
Si e (Pop) 4 1.022 0.405 3 2.809 0.051 6 1.743 0.123 6 2.019 0.070
Popula ion 3 10.957 <0.001 312.999 <0.001 537.755 <0.001 550.702 <0.001
BG Si e (Pop) 4 2.140 0.107 4 1.021 0.416 6 5.539 <0.001 614.674 <0.001
h ps://doi.o g/10.1371/jou nal.pone.0196075. 001
Fig 2. Mean P
50
(MPa), xylem-speci ic hyd aulic conduc i i y (K
s
, kg m
-1
MPa
-1
s
-1
) and b anch g ow h (mm/yea ) pe species, popula ion and si e. The wo si es
a e ep esen ed in di e en colou s (whi e and g ey). The ba s ep esen he nominal ange o da a a ia ion, wi h he uppe and lowe ends showing he uppe qua ile
plus 1.5 imes he in e qua ile ange and he lowe qua ile minus 1.5 imes he in e qua ile ange, espec i ely. Values beyond hese limi s a e plo ed as ci cles. CR:
Czech Republic; PO: Po ugal; NE: The Ne he lands; Fi-RU: Finland-Ruo sinkyla¨; Fi-VA: Finland-Va¨ io¨; Fi-HYY: Finland-Hyy ia¨la¨; IT: I aly; SW-LOW: Swi ze land-
Loe schen al; SW-PFY: Swi ze land- P ywald. Di e en le e s indica e s a is ically signi ican di e ences be ween popula ions.
h ps://doi.o g/10.1371/jou nal.pone.0196075.g002
Pheno ypic a iabili y in hyd aulic ai s
PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0196075 May 1, 2018 6 / 17
species, we could no ob ain absolu e alues o K
s
o he FI-HYY popula ion, due o so wa e
eco ding issues. No di e ences be ween si es we e obse ed o K
s
(Table 1). Populus emula
had he smalles CV
in e
o he ou species s udied (14.16%), bu he la ges CV
sp
(80.04%,
Table 2). Be ula pendula had he la ges CV
in e
, wi h a mean di e ence o up o 2.8 kg m
-1
MPa
-1
s
-1
be ween he popula ions loca ed a he ex eme ends o i s la i udinal dis ibu ion
ange (Table 2,Fig 2). When signi ican , he co ela ions be ween K
s
and b anch diame e
we e weak (Be ula pendula ho = -0.348, p = 0.008; Pinus syl es is ho = 0.267, p = 0.005;
Populus emula ho = 0.090, p = 0.542; Picea abies ho = -0.210, p = 0.055).
Finally, b anch g ow h di e ed be ween popula ions o all species, whe eas si e (nes ed in
popula ion) had a signi ican e ec on b anch g ow h only o coni e s (Table 1,Fig 2). The
pheno ypic a iabili y o b anch g ow h was g ea e han ha o P
50
(Table 2,Fig 2). Fu he -
mo e, he pheno ypic a iabili y o b anch g ow h was g ea e han ha o K
s
in mos cases
(Table 2,Fig 2). Pinus syl es is had he la ges CV
sp
and CV
in e
in b anch g ow h (85.81 and
76.67%, espec i ely), whe eas hese wo coe icien s we e he lowes in Populus emula (43.29
and 37.66%, espec i ely) (Table 2).
Pheno ypic clines wi h clima e and la i udinal g adien s
Populus emula p esen ed s ong signi ican clines in P
50
, as i e ou o he six clima ic a i-
ables s udied he e showed signi ican co ela ions wi h P
50
(Table 3,Fig 3). P
50
alues o his
species we e posi i ely co ela ed wi h la i ude and a idi y index (AI and AI_Sum), bu nega-
i ely co ela ed wi h MAT and T_sum (Table 3,Fig 3). The e was also a s a is ically signi ican
nega i e co ela ion be ween P
50
and T_Sum in Be ula pendula (Table 3). By con as , no sig-
ni ican clines in P
50
we e obse ed o coni e s (Table 3,Fig 3).
K
s
was less s ongly ela ed o clima e han P
50
. The K
s
/clima e co ela ion was s a is ically
signi ican only o Be ula pendula, wi h lowe K
s
alues a si es wi h highe MAT alues
(Table 3,Fig 4).
We ound s eepe clines o b anch g ow h han o hyd aulic p ope ies, wi h all species
showing a leas one s a is ically signi ican co ela ion be ween b anch g ow h and la i ude/
clima e a iables (Table 3,Fig 5). La i ude was co ela ed wi h b anch g ow h only in Populus
emula, o which he lowes b anch g ow h alues we e ob ained o he no he nmos popu-
la ion (Table 3,Fig 5). In gene al, when s a is ically signi ican , b anch g ow h was posi i ely
co ela ed wi h MAT and T_sum, and nega i ely co ela ed wi h AI, AI_sum and MAP
(Table 3,Fig 5).
Sa e y-e iciency ade-o
A he in aspeci ic le el, we ound s a is ically signi ican bu weak posi i e co ela ions
be ween P
50
and K
s
o he coni e s s udied, wi h he mos ulne able indi iduals ha ing he
Table 2. In aspeci ic (CV
sp
) and in e -popula ion (CV
in e
) coe icien o a iabili y (%) o he xylem p essu e inducing a 50% loss o conduc ance (P
50
, MPa),
xylem-speci ic hyd aulic conduc i i y (K
s
, kg m
-1
MPa
-1
s
-1
) and b anch g ow h (BG, es ima ed as b anch adius/xylem age (mm/yea )) o each s udy species.
CV
sp
CV
in e
Species P
50
K
s
BG P
50
K
s
BG
Be ula pendula 9.67 53.47 58.94 5.56 50.28 47.05
Populus emula 25.07 80.04 43.29 24.82 14.16 37.66
Picea abies 6.57 49.07 57.49 4.15 42.68 53.48
Pinus syl es is 10.23 48.41 85.81 6.45 23.80 76.67
h ps://doi.o g/10.1371/jou nal.pone.0196075. 002
Pheno ypic a iabili y in hyd aulic ai s
PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0196075 May 1, 2018 7 / 17
la ges hyd aulic conduc i i ies (Table 4,S3 Fig). No signi ican co ela ion was ound be ween
P
50
and K
s
o ei he o he angiospe ms s udied (Table 4).
Discussion
We assessed he pheno ypic a iabili y o hyd aulic sa e y and e iciency ai s (P
50
and K
s
,
espec i ely) and b anch g ow h in ou ee species ac oss a long la i udinal g adien co e ing
mos o hei dis ibu ion ange in Eu ope. P
50
displayed lowe pheno ypic a iabili y han K
s
and b anch g ow h, consis en wi h ou ini ial hypo hesis. The low a iabili y o P
50
ac oss
popula ions has been ela ed o uni o m e olu iona y selec ion o canaliza ion [37,38]. Indeed,
hese s udies p o ided e idence o na u al selec ion ac ing on his ai . This uni o m selec ion
ep oduces ai conse a ism and e en ually leads o s asis [37]. In con as , i has been sug-
ges ed ha K
s
a iabili y is ela ed o he in e ac ion be ween geno ype and en i onmen [35].
K
s
may also a y signi ican ly wi h sampling posi ion along he b anch axis [48,49], al hough
we ied o o e come his limi a ion h ough he use o sys ema ic sample p epa a ion p oce-
du es. We also expec ed b anch g ow h o be mo e a iable han P
50
, because b anch g ow h is
s ongly in luenced by mul iple in e ac ing ac o s, such as he a ailabili y o nu ien s, ligh ,
wa e and empe a u e [46,50,51,52,53] and bio ic in e ac ions [54]. The limi ed embolism
esis ance a iabili y obse ed he e, in all species o he han Populus emula, sugges s ha o -
es popula ions o he s udied species will po en ially ind di icul ies o cope wi h a wa me
and d ie condi ions by inc easing hei embolism esis ance. Howe e , conside ing he di e -
ences in clima ic anges be ween he s udied species, we ha e o be cau ious when in e p e ing
hese pa e ns. Fu he s udies in es iga ing la ge p ecipi a ion g adien s and/o ma ginal
Table 3. Co ela ion coe icien s (Pea son o Spea man) and p- alues o he ela ionships be ween he mean xylem p essu e inducing a 50% loss o conduc ance
(P
50
, MPa), xylem-speci ic hyd aulic conduc i i y (K
s
, kg m
-1
MPa
-1
s
-1
) and b anch g ow h (BG, es ima ed as b anch adius/xylem age (mm/yea )) and he clima ic
a iables o each sampling si e.
Be ula pendula Populus emula Picea abies Pinus syl es is
Co . p Co . p Co . p Co . p
P
50
La i ude 0.167 0.692 0.750 0.032 -0.568 0.054 0.193 0.547
MAT -0.547 0.161 -0.770 0.025 0.420 0.174 0.056 0.862
MAP 0.539 0.168 -0.214 0.610 0.288 0.364 -0.466 0.127
AI 0.460 0.251 0.886 0.003 -0.112 0.728 -0.462 0.130
T_Sum -0.793 0.019 -0.909 0.002 0.341 0.278 -0.120 0.711
AI_Sum -0.289 0.487 0.934 0.001 0.098 0.761 -0.578 0.049
K
s
La i ude 0.228 0.586 -0.166 0.753 0.001 1.000 -0.122 0.704
MAT -0.886 0.003 0.308 0.553 -0.147 0.649 0.408 0.187
MAP -0.119 0.779 0.086 0.872 -0.414 0.181 0.276 0.384
AI 0.231 0.582 -0.206 0.695 -0.239 0.455 -0.279 0.379
T_Sum -0.428 0.290 0.292 0.575 0.082 0.799 0.225 0.481
AI_Sum -0.180 0.670 -0.439 0.383 -0.028 0.931 0.019 0.952
BG La i ude 0.497 0.210 -0.894 0.003 -0.386 0.215 -0.224 0.484
MAT -0.423 0.296 0.876 0.004 0.755 0.004 0.627 0.029
MAP -0.786 0.021 0.452 0.260 0.133 0.679 0.027 0.934
AI -0.669 0.069 -0.904 0.002 -0.779 0.003 -0.593 0.042
T_Sum 0.364 0.375 0.909 0.002 0.870 0.002 0.394 0.205
AI_Sum 0.527 0.179 -0.848 0.008 -0.394 0.205 -0.387 0.214
S a is ically signi ican co ela ions a e highligh ed in bold. MAT: mean annual empe a u e (˚C); MAP: o al annual p ecipi a ion (mm); AI: a idi y index (MAP/PET
o po en ial e apo anspi a ion). T_Sum (˚C) and AI_Sum: a e aged mean empe a u e and a idi y indices, espec i ely, o June, July and Augus .
h ps://doi.o g/10.1371/jou nal.pone.0196075. 003
Pheno ypic a iabili y in hyd aulic ai s
PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0196075 May 1, 2018 8 / 17
popula ions a e needed. Indeed, a ecen s udy showed ha ma ginal popula ions o beech sig-
ni ican ly di e ed in embolism esis ance [55] while co e popula ions exhibi ed simila P
50
al-
ues [36]. Ou esul s also show ha wi hin species pheno ypic a iabili y in K
s
and g ow h a e
la ge, and in gene al la ge han ha o P
50
, sugges ing ha in a-popula ion a iabili y should
no be neglec ed in u he s udies a local scales.
Fig 3. Mean P
50
(MPa) pe popula ion plo ed agains la i ude (3.a, decimal deg ees) and he clima ic a iables o each sampled popula ion and si e: 3.b. mean annual
empe a u e (MAT, ˚C); 3.c. o al annual p ecipi a ion; (MAP, mm); 3.d. AI: a idi y index (MAP/PET o po en ial e apo anspi a ion).
h ps://doi.o g/10.1371/jou nal.pone.0196075.g003
Pheno ypic a iabili y in hyd aulic ai s
PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0196075 May 1, 2018 9 / 17
39. Ande egg WRL (2015) Spa ial and empo al a ia ion in plan hyd aulic ai s and hei ele ance o cli-
ma e change impac s on ege a ion. New Phy ol 205(3):1008–1014. PMID: 25729797
40. Hijmans RJ, Came on SE, Pa a JL, Jones PG, Ja is A (2005) Ve y high esolu ion in e pola ed clima e
su aces o global land a eas. In J Clim 25:1965–1978.
41. Zimme mann NE, Robe s DW (2001) Final Repo o he MLP clima e and biophysical mapping p ojec ,
Bi mensdo .
42. To es-Ruiz JM, Cocha d H, Choa B, Jansen S, Lo
´pez R, Toma
´s
ˇko a
´I, e al. (2017) Xylem esis ance
o embolism: p esen ing a simple diagnos ic es o he open essel a e ac . New Phy ol 215:489–
499. h ps://doi.o g/10.1111/nph.14589 PMID: 28467616
43. Cocha d H (2002) A echnique o measu ing xylem hyd aulic conduc ance unde high nega i e p es-
su es. Plan , Cell and En i onmen 25:815–819.
44. Cocha d H, Damou G, Bode C, Tha wa I, Poi ie M, Ameglio T (2005) E alua ion o a new cen i uge
echnique o apid gene a ion o xylem ulne abili y cu es. Physiol Plan a um 124:410–418.
45. Pammen e NW, Vande Willigen C (1998) A ma hema ical and s a is ical analysis o he cu es illus-
a ing ulne abili y o xylem o ca i a ion. T ee Physiol 18:589–59. PMID: 12651346
46. Weiski el AR, Magui e DA, Monse ud RA (2007) Response o b anch g ow h and mo ali y o sil icul-
u al ea men s in coas al Douglas- i plan a ions: Implica ions o p edic ing ee g ow h. Fo Ecol Man-
age 251:182–194.
47. R De elopmen Co e Team (2016) R: A Language and En i onmen o S a is ical Compu ing. R Foun-
da ion o S a is ical Compu ing, Vienna.
48. Yang S, Ty ee MT (1993) Hyd aulic esis ance in he shoo s o Ace saccha um and i s in luence on lea
wa e po en ial and anspi a ion. T ee Physiol 12:231–242. PMID: 14969914
49. Pe i G, An odillo T, Mencuccini M (2008) Tape ing o xylem condui s and hyd aulic limi a ions in syca-
mo e (Ace pseudopla anus) ees. New Phy ol 177(3):653–64. h ps://doi.o g/10.1111/j.1469-8137.
2007.02291.x PMID: 18069964
50. Makinen H, Sa anpaa P, Linde S (2001) E ec o nu ien op imiza ion on b anch cha ac e is ics in
Picea abies. Scand J Fo es Res 16:354–362.
51. Makinen H (2002) E ec o s and densi y on he b anch de elopmen o sil e bi ch (Be ula pendula
Ro h.) in cen al Finland. T ees S uc Func 16:346–353.
52. Ha ield JL, P uege JH (2015) Tempe a u e ex emes: E ec on plan g ow h and de elopmen .
Wea he Clima e Ex emes 10(A):4–10.
53. Lipiec J, Doussan C, Nosalewicz A, Kond acka K (2013) E ec o d ough and hea s esses on plan
g ow h and yield: a e iew. In Ag ophys 27(4):463–477.
54. Makinen H (1999) E ec o s and densi y on adial g ow h o b anches o Sco s pine in sou he n and
cen al Finland. Can J Fo es Res 29:1216–1224.
55. S ojnićS, Suchocka M, Beni o-Ga zo
´n M, To es-Ruiz JM, Cocha d H, Bol e A, e al. (2017) Va ia ion in
xylem ulne abili y o embolism in Eu opean beech om geog aphically ma ginal popula ions. T ee
Physiol in p ess.
56. Mencuccini M, Coms ock J (1997) Vulne abili y o ca i a ion in popula ions o wo dese species,
Hymenoclea salsola and Amb osia dumosa, om di e en clima ic egions. J Exp Bo any 48:1323–
1334.
57. Vande Willigen C, Pammen e NW (1998) Rela ionship be ween g ow h and xylem hyd aulic cha ac e -
is ics o clones o Eucalyp us spp. a con as ing si es. T ee Physiol 18: 595–600. PMID: 12651347
58. Schuld B, Knu zen F, Delzon S, Jansen S, Mu¨lle -Haubold H, Bu le R, Clough Y, Leuschne C (2016)
How adap able is he hyd aulic sys em o Eu opean beech in he ace o clima e change- ela ed p ecipi-
a ion educ ion? New Phy ol 210(2):443–458. h ps://doi.o g/10.1111/nph.13798 PMID: 26720626
59. Mahe ali H, De Lucia EH (2000) Xylem conduc i i y and ulne abili y o ca i a ion o ponde osa pine
g owing in con as ing clima es. T ee Physiol 20:859–867. PMID: 11303576
60. Ma ı
´nez-Vilal a J, Piñol J (2002) D ough -induced mo ali y and hyd aulic a chi ec u e in pine popula-
ions o he NE Ibe ian Peninsula. Fo Ecol Manage 161: 247–256.
61. Co nwell WK, Bhaska R, Sack L, Co dell D, Lunch CK (2007) Adjus men o s uc u e and unc ion o
Hawaiian Me oside os polymo pha a high s. low p ecipi a ion. Func Ecol 21:1063–1071.
62. Hajek P, Ku jak D, on Wu¨hlisch G, Delzon S, Schuld B (2016) In aspeci ic a ia ion in wood ana omi-
cal, hyd aulic, and olia ai s in en Eu opean beech p o enances di e ing in g ow h yield. F on Plan
Sci 7:791. h ps://doi.o g/10.3389/ pls.2016.00791 PMID: 27379112
63. Ahmad HB, Lens F, Capde ille G, Bu le R, Lama que LJ, Delzon S (2017) In aspeci ic a ia ion in
embolism esis ance and s em ana omy ac oss ou sun lowe (Helian hus annuus L.) accessions. Phy-
siol Plan in p ess.
Pheno ypic a iabili y in hyd aulic ai s
PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0196075 May 1, 2018 16 / 17

64. Lexe C, Fay M, Joseph J, Nica MS, Heinze B (2005) Ba ie o gene low be ween wo ecologically
di e gen Populus species, P. alba (whi e popla ) and P. emula (Eu opean aspen): he ole o ecology
and li e his o y in gene in og ession. Mol Ecol 14: 1045–1057. h ps://doi.o g/10.1111/j.1365-294X.
2005.02469.x PMID: 15773935
65. Lexe C, Joseph J, an Loo M, P enne G, Heinze B, Chase MW, e al. (2009) The use o digi al image-
based mo phome ics o s udy he pheno ypic mosaic in axa wi h po ous genomes. Taxon 58:349–
364.
66. Delzon S, Sa o e M, Bu le R, Dewa R, Lous au D (2004) Hyd aulic esponses o heigh g ow h in
ma i ime pine ees. Plan Cell En i on 27(9):1077–1087.
Pheno ypic a iabili y in hyd aulic ai s
PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0196075 May 1, 2018 17 / 17