di e si y
A icle
Do Roads Al e he T ophic Beha io o he Mesoca ni o e
Communi y Li ing Close o Them?
Pablo Ruiz-Capillas 1, C is ina Ma a 1,2,* , Bea iz Fe nández 1, Ca los Fe nandes 3,4 and Juan E. Malo 1,2


Ci a ion: Ruiz-Capillas, P.; Ma a, C.;
Fe nández, B.; Fe nandes, C.; Malo,
J.E. Do Roads Al e he T ophic
Beha io o he Mesoca ni o e
Communi y Li ing Close o
Them? Di e si y 2021,13, 173.
h ps://doi.o g/10.3390/d13040173
Academic Edi o : Michael Wink
Recei ed: 22 Feb ua y 2021
Accep ed: 16 Ap il 2021
Published: 18 Ap il 2021
Publishe ’s No e: MDPI s ays neu al
wi h ega d o ju isdic ional claims in
published maps and ins i u ional a il-
ia ions.
Copy igh : © 2021 by he au ho s.
Licensee MDPI, Basel, Swi ze land.
This a icle is an open access a icle
dis ibu ed unde he e ms and
condi ions o he C ea i e Commons
A ibu ion (CC BY) license (h ps://
c ea i ecommons.o g/licenses/by/
4.0/).
1Te es ial Ecology G oup (TEG-UAM), Depa amen o de Ecología, Facul ad de Ciencias, Uni e sidad
Au ónoma de Mad id, 28049 Mad id, Spain; [email p o ec ed] (P.R.-C.);
[email p o ec ed] (B.F.); [email p o ec ed] (J.E.M.)
2
Cen o de In es igación en Biodi e sidad y Cambio Global (CIBC-UAM), Uni e sidad Au ónoma de Mad id,
28049 Mad id, Spain
3CE3C—Cen e o Ecology, E olu ion and En i onmen al Changes, Depa amen o de Biologia Animal,
Faculdade de Ciências, Uni e sidade de Lisboa, 1749-016 Lisboa, Po ugal; [email p o ec ed]
4Faculdade de Psicologia, Uni e sidade de Lisboa, Alameda da Uni e sidade, 1649-013 Lisboa, Po ugal
*Co espondence: [email p o ec ed]; Tel.: +34-91-497-80-11
Abs ac :
Roads ha e impac s on he auna a ising om habi a agmen a ion, oadkill and he
ba ie e ec . Fu he mo e, oads lead species o change hei ac i i y wi h epe cussions on p eda o –
p ey in e ac ions and igge indi ec e ec s ha a e cu en ly unknown. This s udy analyzes
he e ec o a mo o way on he ophic beha io o he e es ial ca ni o e communi y o i s
su oundings. Mon hly sca sampling was conduc ed o e a yea a h ee dis ances om a mo o way
(0–50 m,
500–550 m
and 1000–1050 m). We collec ed 498 sca s, hese o igina ing om ed ox
(39.16%),
ca (24.50%),
s one ma en (24.09%) and badge (12.25%). The ela i e abundance o he
ophic esou ces in hem was es ima ed oge he wi h he ophic di e si y and niche o e lap o he
ca ni o e species. The esul s showed a dis inc e ec o dis ance om he oad on ophic beha io
o ca ni o es, as well as di e ences be ween species and seasons. The sca s nea es he oad had
10–20% mo e biomass o small mammals, equi alen in ela i e e ms o a 21–48% inc ease in small
mammals’ biomass when compa ed wi h sca s collec ed u he om he oad. This inding indica es
changes in p eda o –p ey in e ac ions nea he oad and shows ha he human-gene a ed s uc u al
and unc ional changes o ecosys ems sp ead h oughou ophic ne wo ks.
Keywo ds: eeding ecology; p eda o –p ey in e ac ions; oad e ec s; small mammals
1. In oduc ion
Roads ep esen a massi e en i onmen al pe u ba ion wo ldwide, ha ing a global
ex en o o e 64 million kilome e s [
1
,
2
]. Al hough abou 80% o he Ea h’s e es ial
su ace emains oadless, he e ain has been agmen ed in o some 600,000 agmen s,
o e hal o which a e smalle han one squa e kilome e [
3
]. The spa ial impac o oads
con inues o g ow, and hei global ex en is p ojec ed o inc ease by o e 60% om 2010 o
2050 [
4
]. By 2050, an es ima ed 25 million addi ional kilome e s o new oads will a ec
egions ha cu en ly sus ain excep ional biodi e si y and o e i al ecosys em se ices [
5
].
The impac o his “gian emb acing us”, as so e med by Fo man [
6
], is no es ic ed o
he su ace ha he in as uc u es physically occupy bu ins ead may ex end o hund eds
o me e s o ei he side (see de ini ion o “ oad e ec zone” [
7
]). Hence, he a ea a ec ed
ecologically co e s a high pe cen age o he o al su ace in many egions [
8
,
9
]. Fo example,
hal o he nonu banized Eu opean e i o y lies less han 1.5 km om he nea es oad o
ailway line, and a qua e is less han 500 m away. In Spain, 99% o he e i o y is less
han 10 km om a oad, ailway line o simila in as uc u e [9].
The scien i ic communi y has made a g ea e o o e ecen decades o iden i y and
analyze he e ec s o anspo in as uc u es on he bio ic and abio ic componen s o
Di e si y 2021,13, 173. h ps://doi.o g/10.3390/d13040173 h ps://www.mdpi.com/jou nal/di e si y
Di e si y 2021,13, 173 2 o 13
he ecosys ems ha hey a e se (see e iews in [
10
–
12
]). Road–wildli e in e ac ions a e
one o he mos s udied aspec s, no only because o hei impac on biodi e si y
[9,13]
bu also because o he isks o use s and he signi ican economic cos s [
14
]. Mos s udies
ha e ocused on he di ec nega i e e ec s o anspo in as uc u es (such as oadkill,
habi a deg ada ion and agmen a ion), while he indi ec e ec s ha may cascade h ough
he aunal communi y emain la gely unknown. Rega ding he la e , an impo an as-
pec is he change ha oads b ing abou in he use o e i o ies, species beha io and
in e speci ic in e ac ions.
Recen s udies ha e shown how ac i i y pa e ns close o oads a e al e ed in di e en
ways among e eb a e species and unc ional g oups [
13
,
15
]. P eda o and p ey species
espond o oads by changing hei ac i i y and abundance pa e ns [
15
,
16
] and al e ing
hei p eda o y beha io o an ip eda o y esponses [
17
,
18
]. The s uc u e and dynamics o
p ey species, such as small mammals and lagomo phs, ha e been shown o change in he
p esence o a oad [
19
–
21
], wi h likely epe cussions on he ac i i y pa e ns, e i o y use
and beha io o hei p eda o s. Indeed, i has also been shown ha he ac i i y pa e ns o
Medi e anean ca ni o es a e simila ly modi ied by oad p esence [
13
,
21
,
22
], esul ing pe -
haps om changes in ophic esou ces caused by he oad, om a ac ion o oadkill [
23
]
o om changes in he in ensi y o local in e speci ic in e ac ions, such as compe i ion o
in aguild p eda ion [16].
Unde s anding changes in e eb a e ophic webs is especially ele an o un a el
unexpec ed consequences o human in as uc u es, such as possible changes in p eda ion
le els nea oads [
24
,
25
] and hei epe cussions on ecosys em s uc u e and dynamics.
The e a e indica ions ha some o he nega i e e ec s o oads may de i e om he ophic
beha io o di e en species and he p eda o –p ey in e ac ions ha a ise. Fo example,
some s udies ha e shown a ela ionship be ween die ype and he p obabili y o becoming
oadkill [
26
,
27
] and how o some ca ni o es being killed on oads is associa ed wi h p ey
a ailabili y in hei icini y [
28
,
29
]. In u n, p eda o –p ey in e ac ions may de e mine he
use and e ec i eness o he wildli e passages ha a e he mos equen mi iga ion measu e
employed a oads [30–32].
The p esen s udy is he i s o analyze he die o he mammalian p eda o guild in
he icini y o a oad, in a case o a guild domina ed by gene alis species. Die analyses
a e a use ul ool o his pu pose because hey can e eal he use o ophic esou ces
by di e en species and he ophic in e ac ions be ween hem. They also allow ophic
a ailabili y and esou ce dis ibu ion in space and ime o be analyzed [
33
]. Gi en ha
wi hin ou s udy a ea small mammals ha e been shown o be mo e abundan nea he
oad [
21
] and ha local ca ni o e ac i i y is also al e ed [
34
], ou s a ing hypo hesis was
ha dis ance om he oad will ha e a signi ican e ec on ca ni o e ophic beha io .
Speci ically, we expec ed ha he local ca ni o e die close o he oad would be en iched in
small mammals, e lec ing he unc ional esponses de i ed om he changes in p eda o
and p ey densi y and ac i i y.
2. Ma e ials and Me hods
2.1. S udy A ea
The s udy ook place along a 5 km s e ch (UTM 30 T 373259 4510571 o 30 T 368053
4507625) o he ou -lane AP-51 mo o way in Á ila p o ince, cen al Spain (Figu e S1).
The a ic olume he e a e ages 8396 ehicles pe day. The chosen s e ch c osses a well-
p ese ed landscape domina ed by Ibe ian holm oak (Que cus ilex subsp. ballo a) woodlands
(c. 40% co e ) and expanses o Medi e anean sc ub chie ly o gian ea he g ass S ipa
gigan ea Lag., gum cis us Cis us ladani e L. and holm oak eg ow h (40%), mixed wi h
nonin ensi e holm oak g azing woodlands (dehesas; 10%) and ce eal c ops (10%). The local
clima e is Medi e anean, wi h a h ee-mon h summe d ough pe iod, a mean annual
empe a u e o 10
◦
C and mean annual p ecipi a ion o 364 mm. The opog aphy is ela i ely
le el, anging in ele a ion be ween 1050 and 1250 m.
Di e si y 2021,13, 173 3 o 13
2.2. Sca Collec ion
Th ee ansec lines we e su eyed a mon hly in e als du ing he cou se o a yea
(Decembe 2009–No embe 2010). The lines we e pa allel o each o he and o he ca iage-
way, a h ee dis ances om he oad ha ypically show di e en ial densi ies o small
mammals [
16
] (0–50 m, 500–550 m, 1000–1050 m). Each line was some 3 km long and
chie ly ollowed di oads, wi h a small p opo ion (c. 10% in o al) c oss-coun y along
ails and pa hs. The o al dis ance sampled du ing he s udy was 127 km.
Two expe ienced obse e s eco ded and iden i ied ca ni o e sca s encoun e ed du -
ing he su eys. Iden i ica ion was based as much on mo phological cha ac e is ics, such as
size and shape, as on o he cha ac e is ics such as smell and placemen [
35
]. Sca s ha could
no be sa ely iden i ied we e excluded om he die s udy. Iden i ica ion was o species
le el, wi h he excep ion o ca s, which we e eco ded as Felis sp. gi en he impossibili y o
dis inguish be ween sca s o he domes ic ca (Felis ca us) and wildca (Felis syl es is).
2.3. Sca Analysis
A o al o 498 sca s we e analyzed ou o 1005 ound du ing he su eys. Those in
poo e condi ion o no sa ely iden i iable we e disca ded. Those selec ed we e a ep e-
sen a i e sample in e ms o species, season and dis ance om he mo o way. The species
analyzed we e he ed ox (Vulpes ulpes), ca (Felis sp.), s one ma en (Ma es oina) and
badge (Meles meles). Seasons we e g ouped in o wo pe iods, Au umn–Win e and Sp ing–
Summe , aking in o accoun he known di e ences in ophic esou ce a ailabili y be ween
hese pe iods [
21
,
36
]. Table 1gi es he o al numbe o sca s analyzed pe species, sea-
son and dis ance om oad.
Table 1.
Numbe o sca s analyzed (N) o de e mine ca ni o e die pe species, season (A–W:
Au umn–Win e ; Sp–Su: Sp ing–Summe ) and dis ance om oad (D): D
1
: 0–50 m; D
2
: 500–550 m;
D3: 1000–1050 m.
Dis ance Red ox Ca S one ma en Badge N
A–W Sp–Su A–W Sp–Su A–W Sp–Su A–W Sp–Su
D173 28 31 8 35 23 6 6 210
D224 14 21 6 15 7 18 5 110
D334 22 38 18 19 21 14 12 178
N 131 64 90 32 69 51 38 23 498
Collec ed sca s we e ea ed wi h insec icide, o p e en possible a ack by cop ophagous
insec s, be o e s o age in pape bags in a d y place. Subsequen labo a o y analysis ollowed
he p ocedu e desc ibed by Reynolds and Aebische [
37
]: sca s we e i s disin eg a ed in a
0.5 mm sie e unde a je o wa e . Mac oscopic componen s (bones, hai s, ea he s, seeds,
a h opods exoskele on agmen s, ui s) we e a e wa d sepa a ed and so ed in o six ca e-
go ies: small mammals, lagomo phs ( abbi and ha e), ui s and seeds, a h opods (insec s
and c ay ish), ep iles and bi ds. Ca ion was excluded as a ood esou ce as we conside i s
consump ion o be inciden al. Mammal emains we e iden i ied om he cu icula su ace
pa e ning and in e nal s uc u e o hai s [
38
]. The d y mass o each o he ood ca ego ies
p esen in each sca (wi) was also eco ded.
In o de o assess he magni ude o possible iden i ica ion e o s, apa om he
o al sample, we collec ed an addi ional sample o 70 esh sca s no included in he die
analyses. Obse e s we e equi ed o assign hese esh sca s o a species i espec i e
o whe he o no he sca posed iden i ica ion p oblems, and hese es samples we e
subsequen ly submi ed o gene ic analysis. The sca s we e p ese ed in plas ic ials
wi h 96% e hanol un il DNA ex ac ion. DNA was ex ac ed using he PSP Spin S ool
DNA Ki (S a ec, Bi ken eld, Ge many), and he samples we e assigned o species using
species-speci ic mi ochond ial DNA ma ke s speci ically designed o he iden i ica ion o
Di e si y 2021,13, 173 4 o 13
nonin asi e samples om Ibe ian ca ni o es [
39
]. The desc ip ion and compa ison o he
esul s o he mo phological and gene ic iden i ica ions a e shown in Table S1. Unde hese
condi ions, designed o es ima e he (po en ial) maximum le el o iden i ica ion e o s
based on mo phology, i was ound ha mo phological iden i ica ion o ox sca s incu ed
a low a e (10%) o misiden i ica ion. Sca s o o he species incu ed highe e o a es,
wi h a bias owa ds ox sca s being iden i ied as coming om hese species. This indica es
ha he analyzed samples o o he species may include a ce ain p opo ion o ox sca s.
Hence, in e speci ic compa isons may be pa ly loaded (and di e ences blu ed) by he
inclusion o inco ec ly iden i ied sca s.
2.4. Da a Analysis
The pe cen age o esh biomass inges ed pe ood esou ce, PBi, was used as a
desc ip i e index o ca ni o e die :
PBi=100 ×CFi×wi/∑n
i(CFi×wi), (1)
whe e wi is he d y mass o emains om a pa icula ood esou ce i,CFi is a co ec ion
ac o and nis he o al numbe o ood esou ce ca ego ies conside ed. The co ec ion
ac o s used we e hose mos o en employed wi hin he li e a u e o each species [
40
–
44
]
and a e desc ibed in Table S2.
The die a y niche b ead h o each species employed Le in’s s anda dized index [
45
]
using he es ima ed biomass p opo ions. The index is gi en as ollows:
B=∑n
ip2
i−1, (2)
whe e pi is he p opo ion in biomass e ms o i(
PBi/
100
)
. The ollowing s anda dized
equa ion was used o ensu e compa abili y wi h o he s udies:
Bs=(B−1)/(n−1). (3)
The index alues ange be ween nea 0 (lowes niche b ead h) and 1 (g ea es niche
b ead h = 1).
Pianka’s index, S[
46
], was used o es ima e ophic niche o e lap, as applied o he
p opo ional in ake o esh biomass. I is calcula ed as ollows:
S=∑pij ×pik/∑p2
ij ×∑p2
ik0.5, (4)
whe e p
ij
is he p opo ion o ood esou ce iin he die o p eda o jand p
ik
is he p o-
po ion o iin ha o p eda o k. Index alues ange be ween 0 ( o al sepa a ion) and
1 ( o al o e lap).
In addi ion, ce ain o he indices ha a e commonly employed in he li e a u e (see
e iew in [
47
]), based on he equency o occu ence (FO) o di e en ood esou ces
and hei ela i e equency o occu ence (RF), we e calcula ed. Such indices end o
o e es ima e he impo ance o small ood i ems (e.g., insec s, [
47
]). Hence, hey we e only
calcula ed o ensu e ha he p esen ed da a (Table S3) can be compa ed wi h hose o o he
s udies bu hey we e no used in he s a is ical analyses.
2.5. S a is ical Analysis
The gene al ca ni o e die pa e ns we e o dina ed using mul idimensional scaling
(MDS) o compa e sca s on he basis o he ec o s o he PBi alues o each. The nonpa a-
me ic MDS was cons uc ed om he ma ix o dissimila i ies compu ed as Manha an
dis ances [
48
]. The numbe o dimensions used o he o dina ion o he 498 obse a ions
was es ablished aking accoun o he K uskal s ess alue, a measu e o how well he
con igu a ion ma ches he da a [
49
]. To de e mine which ood ca ego ies we e con ibu ing
mos o di e ences in ca ni o e die s, Spea man co ela ions we e pe o med be ween he
Di e si y 2021,13, 173 5 o 13
loca ion o obse a ions on he dimensions and he PBi o he ood esou ces used in he
inpu ma ix.
A h ee-way MANOVA was conduc ed o de e mine he e ec o he analyzed ac o s
(species, season and dis ance o mo o way and hei in e ac ions), wi h MDS coo dina es
as esponse a iables, be o e we con i med MANOVA assump ions o mul i a ia e ho-
mogenei y and no mali y. Di e ences be ween g oups we e de e mined by es ing he
pai wise compa isons be ween he di e en ac o le els.
To assis he g aphical in e p e a ion o esul s, a di ec ion ec o in o dina ion planes
o each ood esou ce was de ined om he Spea man co ela ion coe icien s ob ained
be ween he PBi alues and he axis coo dina es [
50
]. In addi ion, o each le el o he
dis ance om he mo o way ac o , he di ec ion ha i would ha e on he o dina ion plane
was ep esen ed in acco dance wi h he ec o de ined by he be a alues ob ained in he
pai wise compa ison MANOVAs.
In addi ion, he e ec o he h ee analyzed ac o s and hei in e ac ions on he PBi
alues o he p incipal ophic esou ces in biomass e ms o he whole sample (small
mammals and lagomo phs) was also analyzed. Fo his pu pose, a gene alized linea model
(GLM) was conduc ed using a csine squa e oo ans o med p opo ions and co ec ing
s a is ical es s o o e dispe sion.
The signi icance le el was se a p= 0.05, and a sequen ial Bon e oni p obabili y
co ec ion [
51
] was applied o mul iple es ing. All analyses we e ca ied ou wi h he
STATISTICA 8.0 s a is ical p og am [52] using he indi idual sca as he sampling uni .
3. Resul s
The analyzed o al o 498 sca s comp ised 39.16% ed ox, 24.50% Felis sp., 24.09%
s one ma en and 12.25% badge sca s. The main desc ip o s o he die o each o hese
a e gi en in Table S3. The die a y analysis shows he communi y in gene al o ha e a
medium-low niche-b ead h ( ange Bs = 0.23–0.56) wi h medium-high ophic niche o e lap
(S = 0.57–0.88). The mos impo an die a y componen o he ca ni o es in biomass e ms
was small mammals (49.34%), ollowed by lagomo phs (28.76%), ui s and seeds (11.44%),
ep iles (4.58%), a h opods (2.96%) and bi ds (2.92%). Howe e , he main ood sou ces
a ied among ca ni o es (Table S3), wi h small mammals and lagomo phs being especially
dominan in ca and ed ox die s. F ui s and seeds we e a he ele an in he die o s one
ma en and o a lesse ex en in ha o badge . A h opods we e also ele an o he la e .
The MDS (Figu e 1) p o ided h ee o dina ion axes (S ess = 0.087). In acco dance
wi h he co ela ion coe icien s ob ained (see Table S4), he i s dimension ep esen ed
an inc ease in he p opo ion o small mammals ela i e o he o he ophic esou ces.
The second dimension dis inguished an inc ease in he p opo ion o lagomo phs, ui s and
ep iles a i s posi i e ex eme agains an inc eased p opo ion o bi ds and a h opods
a he nega i e ex eme. The hi d dimension ep esen ed a p opo iona e inc ease in
lagomo phs and a h opods agains majo consump ion o he o he ood esou ces.
The MANOVA esul s demons a ed he exis ence o signi ican be ween-species
di e ences in die , season and dis ances om he mo o way, as well as a signi ican
in e ac ion be ween he ac o s “species” and “season” (see Table 2). Rega ding he main
ocus o ou s udy, he pai wise compa isons showed signi ican di e ences in die be ween
0 and 1000 m om he oad (p< 0.001). Howe e , di e ences be ween 0 and 500 m (p= 0.040)
we e no signi ican a e applying he Bon e oni co ec ion, and di e ences be ween 500
and 1000 m (p= 0.206) we e no signi ican ei he . The g aphical ep esen a ion o he be a
alues ob ained (Figu e 1) shows ha he zone nea es he oad saw inc eased p eda ion on
small mammals, whe eas consump ion o o he ophic esou ces cha ac e ized he mo e
dis an zones.

Di e si y 2021,13, 173 6 o 13
Di e si y 2021, 13, x 6 o 13
Figu e 1. Th ee-dimensional o dina ion (MDS) o gene al ends in ca ni o e die . The in luence o
ood esou ces in e ms o pe cen age o esh biomass (PB) is ep esen ed by a ows o leng h and
di ec ion ob ained om hei co ela ion wi h he axes. To assis he g aphical in e p e a ion o
esul s, he igu es include do ed ed a ows ep esen ing he di ec ions o he β coe icien s o
MANOVA o he h ee dis ances, wi h a module equi alen o β mul iplied by 10 o cla i y
p esen a ion. PB Abb e ia ions: MAM, small mammals; LAG, lagomo phs; FRU, ui and seeds;
ART, a h opods; BIR, bi ds; REP, ep iles.
The MANOVA esul s demons a ed he exis ence o signi ican be ween-species di -
e ences in die , season and dis ances om he mo o way, as well as a signi ican in e ac-
ion be ween he ac o s “species” and “season” (see Table 2). Rega ding he main ocus
o ou s udy, he pai wise compa isons showed signi ican di e ences in die be ween 0
and 1000 m om he oad (p < 0.001). Howe e , di e ences be ween 0 and 500 m (p =
0.040) we e no signi ican a e applying he Bon e oni co ec ion, and di e ences be-
ween 500 and 1000 m (p = 0.206) we e no signi ican ei he . The g aphical ep esen a ion
o he be a alues ob ained (Figu e 1) shows ha he zone nea es he oad saw inc eased
Figu e 1.
Th ee-dimensional o dina ion (MDS) o gene al ends in ca ni o e die . The in luence
o ood esou ces in e ms o pe cen age o esh biomass (PB) is ep esen ed by a ows o leng h
and di ec ion ob ained om hei co ela ion wi h he axes. To assis he g aphical in e p e a ion
o esul s, he igu es include do ed ed a ows ep esen ing he di ec ions o he
β
coe icien s
o MANOVA o he h ee dis ances, wi h a module equi alen o
β
mul iplied by 10 o cla i y
p esen a ion. PB Abb e ia ions: MAM, small mammals; LAG, lagomo phs; FRU, ui and seeds;
ART, a h opods; BIR, bi ds; REP, ep iles.
Di e si y 2021,13, 173 7 o 13
Table 2.
Fac o s de e mining he die o he ca ni o e communi y o he s udy a ea. Resul s o he
MANOVA on he e ec s o species, season and dis ance o he oad, and hei in e ac ions, on he
posi ion on he h ee axes o he mul idimensional scaling ha summa ize sca con en s.
Fac o s Wilks’ λFp
Season 0.816 35.498 0.000
Dis ance 0.956 3.574 0.002
Species 0.868 7.627 0.000
Season ×Dis ance 0.988 0.931 0.471
Season ×Species 0.912 4.903 0.000
Dis ance ×Species 0.968 0.860 0.629
Season ×Dis ance ×Species 0.953 1.283 0.190
Acco dingly, he GLM esul s showed signi ican di e ences associa ed wi h he
dis ance om he mo o way in he pe cen age o small mammal biomass in he sca s
(Table 3).
In compa ison wi h alues ob ained a 1000–1050 m, he sca s om he zone
nea es he oad con ained a g ea e pe cen age o small mammals (
β0–50
= 0.154
±
0.048,
p= 0.001), wi h a sligh ly inc eased pe cen age o small mammal biomass in sca s om
he in e media e zone (
β500–550
= –0.099
±
0.050, p= 0.047). The pe cen age inc eases we e
a ound 10% in he 0–50 m zone ela i e o he 1000–1500 m zone and 20% ela i e o he
500–550 m
zone, ep esen ing in ela i e e ms inc eases o 21% and 48%, espec i ely,
in small mammal con en (Figu e 2). The GLM did no show signi ican di e en ial
esponses o ca ni o es o he ac o dis ance om he mo o way (nonsigni ican Dis ance
x Species in e ac ion in Table 3), bu he esponse is pa en in ca , s one ma en and badge
while absen in ox (Figu e S2). Highly signi ican di e ences in small mammal PB we e
also ound among species and associa ed wi h he species by season in e ac ion. Con a y o
he case o small mammals, no signi ican di e ences we e ound in he pe cen age o
lagomo ph biomass in he sca s o any o he analyzed ac o s (Table 3).
Table 3.
Fac o s de e mining he pe cen age o esh biomass (PB) o he main ophic esou ces
(small mammals and lagomo phs) p esen in he analyzed sca s. Resul s o he GLM using PBs
o small mammals and lagomo phs as esponse a iables. The a iables we e a csine squa e oo
ans o med o analysis.
Fac o s PB Small Mammals PB Lagomo phs
d Wald pWald p
Season 1 2.31 0.128 2.13 0.144
Dis ance 2 10.76 0.004 2.17 0.337
Species 3 23.77 < 0.001 5.12 0.163
Season ×Dis ance 2 1.20 0.550 0.14 0.931
Season ×Species 3 12.49 0.006 1.93 0.588
Dis ance ×Species 6 4.72 0.579 8.89 0.180
Pe iod ×Species ×Dis ance 6 13.06 0.042 3.58 0.733
Di e si y 2021,13, 173 8 o 13
Di e si y 2021, 13, x 8 o 13
Figu e 2. Weigh ed ma ginal means and con idence in e als (CI–95%) o pe cen ages o esh
small mammal biomass (PBmam) in sca s om he h ee dis ance- om- oad bands conside ed. The
a iable was a csine squa e oo ans o med o analysis bu i s g aphical ep esen a ion has been
back- ans o med o acili a e i s in e p e a ion. No e ha he Y axis does no s a a 0%.
4. Discussion
The esul s show ha he oad has a signi ican gene al e ec on ca ni o e ophic
beha io in i s icini y. In ac , hey e eal p o ound and p e iously un epo ed indi ec
e ec s igge ed by anspo in as uc u es in he s uc u al and unc ional cha ac e is-
ics o he ecosys ems wi hin he oad e ec zone.
The h ee analyzed ac o s (species, season and dis ance om oad) had signi ican
e ec s on he die o he ca ni o e communi y, bu he e ec o p oximi y o he oad is
he mos no el and in e es ing esul o his s udy. The in e speci ic and phenological di -
e ences ha e been desc ibed in ea lie s udies and co espond mainly wi h he biological
and beha io al cha ac e is ics o each species and wi h seasonal luc ua ions in esou ce
a ailabili y [53–56]. The p esen indings now add ha dis ance om a oad also signi i-
can ly in luences he ophic beha io o he ca ni o e communi y, leading o an inc eased
consump ion o small mammals close o he in as uc u e. This gene al pa e n needs o
be conside ed wi h some cau ion due o wo poin s. Fi s , e en hough he die s o all
species ma ch hose e ealed by p e ious s udies and p ecau ions we e aken o p e en
iden i ica ion e o , he gene ic analyses e ealed some deg ee o con amina ion by ox
sca s o he samples ob ained om he o he species (Table S1, see also [57–59]). Secondly,
high in aspeci ic a iabili y p ecluded he de ec ion o signi ican di e ences among spe-
cies in he esponse, bu as depic ed in Figu e S2, ox sca s keep a cons an p opo ion o
small mammal biomass along he analyzed dis ance g adien while ca , s one ma en and
badge i he gene al pa e n.
Ca ni o e die s nea he oad, as e ealed by sca analyses, show enhanced consump-
ion o small mammals. Ea lie s udies ha e shown ha oads al e he abundance o p ey
species, wi h zones nea es oads cha ac e ized by abundan small mammal popula ions
[16,19,21,60]. In ou s udy si e, small mammals in he a ea nea es he oad a e no only
mo e abundan bu also show a mode a ion o hei seasonal and in e annual popula ion
cycles [21]. Fu he mo e, mice nea oads may expe ience physiological s ess ha makes
hem mo e ulne able o p eda ion [61]. The oadside icini y may he e o e o e an im-
po an p ey sou ce o ca ni o es, especially a imes when ophic esou ces in he wide
en i onmen a e sca ce. Fu he mo e, small mammals a e a key die a y componen in ou
25
30
35
40
45
50
55
60
65
70
75
0-50m 500-550m 1000-1050m
PBmam (%)
Dis ance o mo o way (m)
Figu e 2.
Weigh ed ma ginal means and con idence in e als (CI–95%) o pe cen ages o esh small
mammal biomass (PBmam) in sca s om he h ee dis ance- om- oad bands conside ed. The a iable
was a csine squa e oo ans o med o analysis bu i s g aphical ep esen a ion has been back-
ans o med o acili a e i s in e p e a ion. No e ha he Y axis does no s a a 0%.
4. Discussion
The esul s show ha he oad has a signi ican gene al e ec on ca ni o e ophic
beha io in i s icini y. In ac , hey e eal p o ound and p e iously un epo ed indi ec
e ec s igge ed by anspo in as uc u es in he s uc u al and unc ional cha ac e is ics
o he ecosys ems wi hin he oad e ec zone.
The h ee analyzed ac o s (species, season and dis ance om oad) had signi ican
e ec s on he die o he ca ni o e communi y, bu he e ec o p oximi y o he oad is
he mos no el and in e es ing esul o his s udy. The in e speci ic and phenological
di e ences ha e been desc ibed in ea lie s udies and co espond mainly wi h he bio-
logical and beha io al cha ac e is ics o each species and wi h seasonal luc ua ions in
esou ce a ailabili y [
53
–
56
]. The p esen indings now add ha dis ance om a oad also
signi ican ly in luences he ophic beha io o he ca ni o e communi y, leading o an
inc eased consump ion o small mammals close o he in as uc u e. This gene al pa e n
needs o be conside ed wi h some cau ion due o wo poin s. Fi s , e en hough he die s
o all species ma ch hose e ealed by p e ious s udies and p ecau ions we e aken o
p e en iden i ica ion e o , he gene ic analyses e ealed some deg ee o con amina ion
by ox sca s o he samples ob ained om he o he species (Table S1, see also [
57
–
59
]).
Secondly, high in aspeci ic a iabili y p ecluded he de ec ion o signi ican di e ences
among species in he esponse, bu as depic ed in Figu e S2, ox sca s keep a cons an
p opo ion o small mammal biomass along he analyzed dis ance g adien while ca ,
s one ma en and badge i he gene al pa e n.
Ca ni o e die s nea he oad, as e ealed by sca analyses, show enhanced consump-
ion o small mammals. Ea lie s udies ha e shown ha oads al e he abundance o
p ey species, wi h zones nea es oads cha ac e ized by abundan small mammal popu-
la ions [
16
,
19
,
21
,
60
]. In ou s udy si e, small mammals in he a ea nea es he oad a e
no only mo e abundan bu also show a mode a ion o hei seasonal and in e annual
popula ion cycles [
21
]. Fu he mo e, mice nea oads may expe ience physiological s ess
ha makes hem mo e ulne able o p eda ion [
61
]. The oadside icini y may he e o e
o e an impo an p ey sou ce o ca ni o es, especially a imes when ophic esou ces
in he wide en i onmen a e sca ce. Fu he mo e, small mammals a e a key die a y
componen in ou s udy a ea because al e na i e p ey such as lagomo phs occu only a
low densi ies (unpublished da a), such ha any educ ion in small mammal consump ion
is compensa ed o by small inc eases in all o he a ailable ood ypes in a eas mo e dis an
Di e si y 2021,13, 173 9 o 13
om he oad. The esul s hus sugges ha he p eda o communi y eac s e y locally o
he ophic o e associa ed wi h he oad bo h by al e ing ophic beha io in esponse o
small mammal abundance nea by and by al e ing o aging pa e ns ( his s udy) and use o
space (see below, [34]).
Rega ding he use o space, he esul s sugges ha some indi idual ca ni o es make
in ensi e use o he a ea nea es he oad. The s udy design does no allow knowing
he mo emen pa e ns o indi iduals o knowing which indi iduals de eca ed a any
poin o he s udy a ea. Howe e , gi en ha he mean gu e en ion ime o mid-sized
ca ni o es is 7–25 h [62–64], we e indi iduals o ange sys ema ically h ough a eas mo e
and less dis an om he oad wi hin hei home anges (o 300–500 ha, acco ding o
species, [
65
]) hei sca s would include a mix u e o ma e ial inges ed h oughou hei
e i o ies. In such ci cums ances, he con en s o hei sca s would no show any clea
spa ial pa e n (as seems o be he case o he ox) and would no e lec he highe small
mammal abundance ound nea es he oad (seen in he o he h ee species). The simples
explana ion o he inc eased small mammal con en in he sca s nea es he mo o way
is ha indi iduals ha de eca e he e a e hose ha make disp opo iona e use o his
zone o hun ing, becoming wha may be e med “ oad e ge specialis s”. Ea lie s udies
ha e shown ha ca ni o es mos o en ma k whe e hei main p ey li es and hus whe e
hey spend mos ime hun ing [
66
]. The home anges o e ge-specialized indi iduals
may pa ly o e lap wi h hose o indi iduals occupying a eas mo e dis an om he oad,
esul ing in a ansi ional a ea (de ec ed in he 500 m samples) whe e sca s e eal a mixed
die a y con en in e media e be ween hose o he wo ex eme loca ions. Fu he s udies
based on he die o adio- agged indi iduals and/o he gene ic indi idualiza ion o
ca ni o e sca s will be needed o cla i y his poin .
Few adio acking s udies ha e analyzed he use o space by ca ni o es nea oads,
and hese ha e o en gi en con adic o y esul s. Some indica e ha some indi iduals
o s one ma en may no a oid oads, whe e habi a is a ailable, and may e en show an
inc eased likelihood o including hem in hei home anges [
67
]. Ascensão e al. [
22
] also
ound ha one s one ma en included he icini y o a oad in i s home ange and ha
i mainly mo ed pa allel o he mo o way when in i s p oximi y. Such use o he mo o -
way may be as much de e mined by i s a ic olume as by he a ailabili y o oadside
habi a [
67
]. The in ensi e use by ca ni o es o he oadside zone in ou s udy a ea [
34
]
may e lec he a ailabili y o sui able habi a , he abundance o ophic esou ces and he
ela i ely low a ic olume (8,396 ehicles pe day, SE 233), which educes he likelihood
o a oidance beha io o any gene al nega i e e ec on he p eda o communi y (p eda o
elease e ec , [
15
,
25
,
68
]). I may also be ha ju eniles o o he subo dina e ca ni o e indi-
iduals a e displaced owa ds oads by dominan s, as implied by he nonb eeding s a us o
coyo es (Canis la ans) and bobca s (Lynx u us) whose home anges include mo o ways [
69
].
Finally, ou esul s also illus a e po en ial in e speci ic in e ac ions igge ed wi hin
he ca ni o e communi y. They indica e ha he oadside communi y comp ises gene -
alis mesoca ni o es showing high ophic o e lap and na ow ophic niche b ead hs.
P eda ion heo y p edic s a con e gence o die s when esou ces a e mo e abundan and
ophic niche seg ega ion o educe compe i ion when esou ce a ailabili y a ies spa ially
and/o empo ally [
70
–
72
]. The obse ed die a y con e gence among oadside ca ni o es
indica es an abundance o ophic esou ces and illus a es he possible ole o in e e ence
in e ac ions in shaping hei ac i i y pa e ns [
73
,
74
]. The exis ence o some deg ee o
seg ega ion in he ac i i y pa e ns o oadside ca ni o es [
16
,
34
] may e lec a oidance
beha io s be ween in e speci ic pai s o indi iduals wi h high die a y o e lap [73,74].
5. Conclusions
The esul s ob ained show ha oad p esence has indi ec e ec s on ca ni o es ia
hei ophic beha io . This adds o changes in he use o space by he di e en species
and po en ially also o a ia ions in he in ensi y o in e ac ions be ween hem. Such a
combina ion o e ec s is especially impo an gi en ha (i) ecosys ems a ec ed by oads