Au ho Que ies
Jou nal: P oceedings o he Royal Socie y B
Manusc ip : RSPB20200649
As he publishing schedule is s ic , please no e ha his migh be he only s age a which you a e able o ho oughly
e iew you pape .
Please pay special a en ion o au ho names, a ilia ions and con ac de ails, and igu es, ables and hei cap ions.
The co esponding au ho mus p o ide an ORCID ID i hey ha en’ done so al eady. I you o you co-au ho s ha e
an ORCID ID please supply his wi h you co ec ions. Mo e in o ma ion abou ORCID can be ound a h p://o cid.
o g/.
No changes can be made a e publica ion.
Q1 A Da a accessibili y s a emen has been added o you pape ; please check ha his is co ec .
Q2 Manda o y end sec ion has been added o you pape ; please check i ha is co ec .
Q3 A unding s a emen has been added o you pape ; please check ha his is co ec .
Q4 Please supply he page ange in e e ence [13].
Q5 Please p o ide publishe loca ion [ci y/s a e/coun y] de ails in e e ences [14,16].
Q6 While he online e sion o igu e 1 will be in colou , we ha e been ins uc ed o p in he igu e in black and
whi e. Please no e ha i you ha e explici ly e e ed o colou in he cap ion his may a ec he legibili y o he
igu e in p in .
SQ1 You supplemen a y ma e ial will be published online alongside you a icle and on s. igsha e.com exac ly as
he ile(s) a e p o ided. The e o e, please could you ei he con i m ha you supplemen a y ma e ial is co ec ,
o –i you ha e any changes o make o hese iles –email hese along wi h you p oo co ec ions o he
jou nal inbox. You ESM iles a e lis ed he e o you con enience:
Supplemen a y ma e ial_p oc_ oy_NO ack.docx
SQ2 You pape has exceeded he ee page ex en and will a ac page cha ges.
oyalsocie ypublishing.o g/jou nal/ spb
Resea ch
Ci e his a icle: Mag ach A, La a C, Luna
UM, Díaz-In an e S, Pa ke I. 2020
Communi y-le el eo ganiza ions ollowing
mig a o y pollina o dynamics along a
la i udinal g adien . P oc. R. Soc. B 20200649.
h p://dx.doi.o g/10.1098/ spb.2020.0649
Recei ed: 22 Ma ch 2020
Accep ed: 9 June 2020
Subjec Ca ego y:
Ecology
Subjec A eas:
ecology, plan science, en i onmen al science
Keywo ds:
mig a o y species, u ous hummingbi d,
niche complemen a i y, ui se , pollina ion,
plan –pollina o in e ac ion
Au ho o co espondence:
Ainhoa Mag ach
e-mail: ainhoa.mag ach@bc3 esea ch.o g
Elec onic supplemen a y ma e ial is a ailable
online a s. igsha e.com.
Communi y-le el eo ganiza ions
ollowing mig a o y pollina o dynamics
along a la i udinal g adien
Ainhoa Mag ach1,2, Ca los La a3, Ubaldo Má quez Luna3, Se gio Díaz-In an e3
and Ing id Pa ke 4
1
Basque Cen e o Clima e Change-BC3, Edi . Sede 1, 1°, Pa que Cien í ico UPV-EHU, Ba io Sa iena s/n,
48940 Leioa, Spain
2
IKERBASQUE, Basque Founda ion o Science, Ma ía Díaz de Ha o 3, 48013 Bilbao, Spain
3
Cen o de In es igación en Ciencias Biológicas, Uni e sidad Au ónoma de Tlaxcala. Km 10.5 Au opis a
Tlaxcala-San Ma ín Texmelucan, San Felipe Ix acuix la, Tlaxcala 9012, Mexico
4
Depa men o Ecology and E olu iona y Biology, Uni e si y o Cali o nia San a C uz, CA 95064, USA
AM, 0000-0003-2155-7556
P edic ing how communi ies e-a ange in esponse o changes in species
composi ion emains a key challenge in ecology. Mig a o y species, which
en e and lea e communi ies ac oss la i udinal g adien s, o e us a unique
oppo uni y o e alua e communi y and species-le el esponses o a shi
in communi y composi ion. We ocused on a mig a o y hummingbi d and
he communi ies ha hos i along a la i udinal and species di e si y g adi-
en . Ou esul s show highe niche o e lap in mo e di e se communi ies,
allowing esiden species o compensa e o he loss o he mig an in p o id-
ing pollina ion se ices. Con as ingly, in less di e se communi ies, he
mig an beha es as a specialis , monopolizing abundan esou ces. In i s
absence, i s ole is no ully co e ed by esiden species, esul ing in a
dec ease in he ui se o he mig an ’s p e e ed plan species. These esul s
help us unde s and he po en ial impac s o biodi e si y loss and ha e
impo an implica ions o communi y pe sis ence gi en expec ed changes
in he mig a o y beha iou s o some species.
1. Backg ound
Ecosys ems a e su e ing om he p essu es o on-going global change, including
clima e change [1,2] and habi a loss [3]. One o he main consequences o ecosys-
em dis u bance is he local ex inc ion o species, ye we ha e li le unde s anding
o he consequences o hese ex i pa ions o ecological in e ac ions, communi y
dynamics and ecosys em unc ions. To p edic how ecosys ems, which a e na u-
ally dynamic, will eac o hese p essu es, we i s need o unde s and how
communi ies eac o he na u al dynamics ha lead o changes in hei compo-
si ion o species, wi h special emphasis on changes in species in e ac ions and
he abili y o he communi y o e-a ange i sel and main ain i s unc ioning.
Up o now, unde s anding communi y-le el ea angemen s ollowing
changes in species composi ion has p o ed elusi e, gi en he g ea complexi y
o ecological sys ems, which ea u e high le els o species di e si y, in e ac ions
ac oss species and en i onmen al a iabili y. The use o ne wo k analyses o ep-
esen some o he bio ic in e ac ions has allowed us o add ess pa o his
complexi y [4,5]. Howe e , many ne wo k s udies ha e used empo ally and
spa ially agg ega ed da a o obse ed in e ac ions ep esen ing a snapsho o a
communi y [6–8]. Agg ega ing da a omi impo an in o ma ion ega ding he
dynamic na u e o ecological in e ac ions [6], and in pa icula conce ning species
unc ional oles, which can change due o compe i ion o esou ces [9], he
p esence o pa asi es and pa hogens [10] o changes in species composi ion [11].
These changes in species composi ion ha e been p ima ily assessed h ough
s udies ocusing on species ex inc ions o in asions. Some o hem ha e used
© 2020 The Au ho (s) Published by he Royal Socie y. All igh s ese ed.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
ARTICLE IN PRESS
RSPB20200649—17/6/20—15:48–Copy Edi ed by: No Men ioned
expe imen al se -ups o explo e communi y le el dynamics
ollowing species ex inc ions [11–13]. Fo example, B osi &
B iggs [11] empo a ily emo ed he mos abundan bumble-
bee species and analysed how he es o he pollina o
communi y esponded. They ound ha in manipula ed
si es, lo al ideli y dec eased, wi h consequences o plan
ep oduc i e success, and also ha he loss o a single polli-
na o species changed pollina ion ne wo k s uc u e [12].
One un esol ed ques ion is whe he communi ies ha bou ing
di e en di e si y le els migh espond di e en ly. In o he
wo ds, does biodi e si y ha e a modula ing ole in hese
esponses o changing species composi ion? We need o
emo e he same species om di e en backg ound commu-
ni ies o see how esponses di e ac oss a di e si y g adien .
Gi en he complexi y o hese expe imen s, i is easy o see
why such s udies a e no commonly done.
Mig a o y species and he communi ies ha ha bou hem
ep esen an ideal na u al expe imen in which he same
species empo a ily lea es ecosys ems ac oss la i udinal and
di e si y g adien s. We su eyed communi ies o plan s and
he hummingbi ds ha pollina e hei lowe s along a la i udi-
nal g adien om cen al Mexico o sou he n Alaska. We
ocused on communi ies ha ha bou he mig a o y u ous
hummingbi d Selaspho us u us and e alua ed he o aging
niches and he unc ional ole o hummingbi ds in he p esence
and he absence o he mig a o y species. We u he explo ed
he consequences o he emo al o his species o plan ep o-
duc i e success. Ou s udy aims o answe h ee main
ques ions: (1) How does he ole o a single pollina o species
change h oughou i s dis ibu ion ange, (2) how do he
oles o he es o he hummingbi d pollina o s in he commu-
ni y change in esponse o a empo a y species loss and
(3) wha a e he consequences o he unc ions hey pe o m.
Gi en i s widesp ead dis ibu ion, we expec ha he
mig a o y hummingbi d will be a gene alis species eeding
on a di e se se o esou ces ac oss all he ange. Fu he , we
expec esiden species o change hei eeding p e e ences
when he mig a o y species is p esen o absen . In pa icula ,
we expec hem o specialize on a small se o esou ces in
he p esence o he mig a o y compe i o and hen o expand
hei die in i s absence. Finally, we expec he e o be changes
on ep oduc i e success pa icula ly o hose species o plan s
on which esiden species specialized du ing he pe iod in
which he mig a o y species was p esen , bu ha will po en-
ially ecei e he e ospeci ic pollen du ing he pe iod in which
he mig an is absen as a consequence o an expansion in he
die s o esiden hummingbi ds ollowing [11].
2. Me hods
(a) S udy species and si es
Selaspho us u ous (Gmelin, 1788), he u ous hummingbi d,
unde akes sp ing and au umn mig a ions be ween win e ing
a eas in Cen al Mexico and b eeding a eas in sou he n Alaska
and Canada [14]. We e alua ed whe he he unc ional ole o
his species changes wi hin h ee communi ies along i s
mig a o y ou e: wes e n Mexico, cen al Cali o nia and sou he n
Alaska, USA. In addi ion, a wo si es (Cali o nia and Mexico),
we we e able o compa e ime pe iods when he mig a o y
species was p esen and absen o e alua e how he loss o one
species changes he unc ional oles o he emaining species,
wi h implica ions o plan ep oduc i e success.
Las Joyas Biological S a ion is a 1245-ha na u al ese e
loca ed in he Sie a de Manan lan wi hin he s a es o Jalisco
and Colima in Mexico (19°350N; 1040°160W, a 1952 m.a.s.l).
Las Joyas expe iences a mean annual empe a u e o 14.6°C
and mean annual p ecipi a ion o 1610 mm, concen a ed
be ween June and Oc obe . The a ea hos s a complex mosaic o
ege a ion, including pine-oak o es s, coni e ous and cloud
o es s [15].
Landels-Hill Big C eek Rese e (36°40000 N 121°350000 W,
0–1,067 m.a.s.l) is a 1752-ha na u al ese e in he San a Lucia
Moun ains along he Big Su coas in Cali o nia. Big C eek Rese e
expe iences a mean annual empe a u e o 12.8°C and mean
annual p ecipi a ion o 620 and 1020 mm a coas and uppe
peaks, espec i ely, concen a ed be ween No embe and
Ma ch. Vege a ion includes an he e ogenous mix u e o coas al
sc ub, edwood o es s, coas al g asslands, oak woodlands and
pine-oak o es s [16].
In he a ea o P ince William Sound, Alaska, we conduc ed
su eys a ei he side o he sound, a he locali ies o Valdez
and Sewa d (61°07’50.9900 N−146°20’53.9900 W and 60°06’15.3400
N−149°26’36.6000 W and 30 and 0 m.a.s.l. espec i ely). A e age
p ecipi a ion in his a ea is 1520 mm while mean annual emp-
e a u e is 5°C. Vege a ion includes a mix u e o closed
needlelea o es o moun ain helmock, Si ka sp uce, ed and
yellow ceda and open hea h and bog a eas.
(b) Hummingbi d and plan su eys
Su eys we e ca ied ou du ing he yea 2019. A each si e, we
selec ed six 1-km ansec s sepa a ed by a leas 500 m om
each o he . Each ansec was su eyed se en imes pe pe iod
du ing wo pe iods: when he mig an species was p esen
(Janua y in Mexico, Ma ch in Cali o nia and June in Alaska)
and again a e he mig an depa ed (Ap il in Mexico and
Ap il–May in Cali o nia). Be ween h ee and ou ansec s
we e su eyed pe day. The ime a which ansec s we e ca ied
ou was andomized e e y day. No su eys we e done in Alaska
du ing he absence pe iod as no esiden hummingbi ds occu in
his a ea. Along each ansec , we eco ded all he eeding in e -
ac ions obse ed be ween e e y species o hummingbi d p esen
in he s udy a ea and e e y lowe ing plan species. In addi ion,
we eco ded lowe a ailabili y o he di e en hummingbi d-
isi ed plan s once o each pe iod as he o al sum o lowe s
obse ed pe species.
In addi ion, in o de o e alua e he impac o species
emo al on plan ep oduc i e success, we u he eco ded se -
e al measu es o plan ep oduc i e success o a subse o plan
species ha we e lowe ing in bo h pe iods: ou in Mexico (Fuch-
sia encliand a; Onag aceae, Lobelia laxi lo a; Lobeliaceae, Rubus
adeno ichos; Rosaceae and Sal ia iodan ha; Lamiaceae), and
h ee in Cali o nia (Ribes sanguineum,Ribes menziesii; G ossula ia-
ceae and Cas illeja a inis; O obanchaceae). These plan s we e
selec ed because hey we e p e e ed esou ces by hummingbi ds
in he a ea and because hey had wo lowe ing pe iods, one
when he mig a o y species was p esen and one in i s absence.
In Mexico, we coun ed he o al numbe o seeds pe ui and
measu ed ui leng h and wid h in 30 ui s om i e indi id-
uals pe plan species. In Cali o nia, we eco ded ui se o
10 indi iduals pe plan species (only ou indi iduals in he
case o Ribes menziesii). Fo wo o he species (Ribes sanguineum
and Ribes menziesii), gi en he la ge numbe o lowe s p oduced,
we ma ked h ee b anches wi hin each o he 10 indi iduals pe
species and coun ed he numbe o lowe s and h ee weeks la e
he numbe o ui s p oduced a each o he wo pe iods
(mig an p esen and absen ). Fo he hi d species, Cas illeja
a inis, we eco ded ui se o he whole plan . In addi ion,
we measu ed ui leng h, wid h and we weigh o a subse
oyalsocie ypublishing.o g/jou nal/ spb P oc. R. Soc. B 20200649
2
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
ARTICLE IN PRESS
RSPB20200649—17/6/20—15:48–Copy Edi ed by: No Men ioned
o 10 ui s pe b anch o each o he wo Ribes species and o all
ui s o C. a inis indi iduals.
(c) Species unc ional oles and in e ac ion ne wo ks
To de e mine species unc ional oles, we ex ac ed a se ies o
me ics om plan –hummingbi d in e ac ion ne wo ks. Fo
each ansec wi hin each pe iod, we cons uc ed a weigh ed
bipa i e in e ac ion ne wo k [17] by pooling he da a o he
se en ounds o sampling. We hen calcula ed a se ies o ele an
me ics o he whole hummingbi d communi y and o each o
he hummingbi d species in pa icula .
(d) Communi y-le el me ics
A he communi y le el, we ocused on wo me ics ha p o ide
pa allel in o ma ion on niche occupancy o he hummingbi d
guild. Fi s , we calcula ed niche o e lap using Ho n’s index
[18], which es ima es he simila i y in in e ac ion pa ne s
be ween hummingbi d species. Second, we calcula ed unc ional
complemen a i y. This me ic, no co ela ed wi h he p e ious
one (elec onic supplemen a y ma e ial, able S1), p esen s
addi ional in o ma ion on he di e si y o oles o niches occu-
pied by hummingbi d species. I is calcula ed as he o al
b anch leng h o a unc ional dend og am based on he Eucli-
dean dis ance be ween hummingbi ds in plan assemblages
isi ed [19,20].
(e) Species-le el me ics
A he species le el, we ocused on me ics ha p o ide in o -
ma ion on he unc ional ole o each hummingbi d species
wi hin he communi y. In pa icula , we calcula ed no malized
deg ee, species-le el specializa ion (d0), and s eng h o each
species a each ansec and pe iod. No malized deg ee gi es
an idea o he di e si y o plan species isi ed by each hum-
mingbi d species. I is calcula ed by di iding a hummingbi d
species’deg ee by he o al numbe o plan species. In his
case, we used he o al numbe o lowe ing plan species
ound wi hin ou independen lowe a ailabili y su eys as
he denomina o . d0p o ides in o ma ion on he le el o special-
iza ion o each hummingbi d species based on disc imina ion
om a andom selec ion o pa ne s [21]. In calcula ing his
me ic, we included abundance da a as he lowe a ailabili y
o each plan species om ou independen su eys. S eng h
p o ides complemen a y in o ma ion on he dependence o
plan species on a pa icula hummingbi d species. I is calcu-
la ed as he sum o he dependencies on ha species o he
plan species isi ed by a ocal hummingbi d [22]. All ne wo k
me ics we e calcula ed using package bipa i e [17].
( ) Da a analyses
Fi s , we e alua ed sampling comple eness wi hin ou s udy by
es ima ing he asymp o ic numbe o plan and pollina o species
p esen , as well as plan –pollina o links [23]. This is a non-pa a-
me ic es ima o o species ichness o abundance da a which
includes non-de ec ed species (o links), allowing us o calcula e
he p opo ion o species (o links) de ec ed wi h ou o iginal
sampling da a. We used Chao 1 asymp o ic species ichness es i-
ma o s [23] and es ima ed he ichness o pollina o s, plan s and
plan –pollina o links accumula ed as sampling e o inc eased
up o 100% sampling co e age using package iNEXT [24]
wi hin he R en i onmen [25].
Then, o answe ou i s ques ion ela ed o how he ole o a
single species changes h oughou i s dis ibu ion ange, we e al-
ua ed how he mig a o y species’no malized deg ee changed
along i s mig a o y pa hway by i ing a gene al linea model
including si e as he explana o y a iable.
In addi ion, we ocused on changes in he ole o mig a o y
species h ough he indi ec in e ac ions i engages in by using
mo i analyses [26]. To his end, we compa ed he mo i ole
signa u es o S. u us in Mexico and Cali o nia (Alaska was no
included in his analysis as he e is only one hummingbi d species
he e). Addi ionally, we compa ed he signa u es o S. u us o
hose o wo o he abundan species, Hylocha is leuco is in
Mexico and Calyp e anna in Cali o nia. Mo i ole signa u es we e
calcula ed using package bmo i [27]. Mo i equencies we e no -
malized by di iding he posi ion coun s o each node by he o al
numbe o imes ha node appea s in any o he posi ions [27].
S a is ical compa isons we e done using pe mu a ional mul i-
a ia e analysis o a iance (PERMANOVA) using B ay–Cu is
as he dissimila i y dis ance. Visual compa isons we e done by
means o non-me ic mul idimensional scaling plo s.
To answe ou second ques ion, ela ed o how he oles o
he o he hummingbi d species in he communi y shi when
he mig a o y species is emo ed, we an analyses a he whole
communi y and a he species le el. A he communi y-le el,
we e alua ed whe he niche o e lap and unc ional complemen-
a i y a ied ac oss pe iods by unning gene al linea mixed
models (GLMMs) wi h pe iod, plan species ichness and lo al
esou ce a ailabili y as explana o y a iables, and ansec
nes ed wi hin si e (Cali o nia and Mexico) as a andom ac o .
A he species le el, we e alua ed whe he no malized deg ee,
d0, and s eng h a ied ac oss pe iods by unning GLMMs ha
included pe iod and i s in e ac ion wi h hummingbi d species
as explana o y a iables. T ansec nes ed wi hin si e was
included as a andom ac o . Fo hese species-le el analyses,
we emo ed he mig a o y species o ocus on how he oles o
he es o he species change.
Finally, o answe ou las ques ion ela ed o how changes in
communi y s uc u e and species unc ional oles migh a ec
plan ep oduc i e success, we an sepa a e GLMMs o he
wo si es (Mexico and Cali o nia) as we we e able o collec
di e en measu es o ep oduc i e success. In he case o
Mexico, esponse a iables we e ui leng h, ui weigh and
he numbe o seeds pe ui , scaled p io o analysis (i.e. we
sub ac ed column means and di ided by s anda d de ia ion)
o allow meaning ul compa isons ac oss species wi h con as ing
li e his o ies. Fo Cali o nia, esponse a iables we e ui se ,
ui leng h, wid h and weigh . In bo h cases, models included
pe iod (mig a o y species p esen o absen ) and i s in e ac ion
wi h plan species as well as he plan ’s no malized deg ee, a
measu e o he di e si y o pollina o s isi ing i , as explana o y
a iables. Plan indi idual was included as a andom ac o . We
used a no mal dis ibu ion o i all models excep in he case o
ui se whe e da a we e i ed o a binomial dis ibu ion. All
s a is ical analyses we e done using R [25] and all GLMMs
we e i ed using package lme4 [28].
3. Resul s
An analysis o he comple eness o ou sampling e ealed ha
wi h ou su ey we we e able o cap u e e y high le els o
hummingbi d and plan species di e si y a all si es anging
om 88 o 100% in hummingbi d species, 56–100% in plan
species and 55–100% in plan –hummingbi d in e ac ions
(elec onic supplemen a y ma e ial, igu e S1).
As demons a ed by ou analysis o o aging niches o he
mig a o y hummingbi d, he ole o he mig a o y species
S. u us changed subs an ially h oughou i s dis ibu ion
ange. In pa icula , no malized deg ee, a measu e o he
numbe o plan species isi ed di ided by all possible plan
species, is much smalle in Cali o nia han in ei he Alaska o
Mexico ( igu es 1and 2a). Howe e , he di e ence is only
oyalsocie ypublishing.o g/jou nal/ spb P oc. R. Soc. B 20200649
3
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
ARTICLE IN PRESS
RSPB20200649—17/6/20—15:48–Copy Edi ed by: No Men ioned
signi ican be ween Cali o nia and Alaska (z- a io = 2.64,
p< 0.01, ma ginal R
2
= 0.32), because in e - ansec a iabili y
is high in Mexico. Despi e he e being a simila numbe o
a ailable plan esou ces in Cali o nia and Alaska ( h ee o
ou species), in Cali o nia he mig a o y species eeds almos
exclusi ely on one esou ce, Ribes sanguineum.
Ou mo i app oach analyses o species oles shows ha
S. u us has e y di e en oles a wo o he loca ions sampled
(elec onic supplemen a y ma e ial, igu e S2A, F= 6.43,
p< 0.001). While he species occupies mos ly specialis oles
in Cali o nia, in Mexico i engages in in e ac ions in ol ing
mo e han one plan species (elec onic supplemen a y
ma e ial, igu e S2A). When compa ing he ole o he mig an
species wi h ha o wo abundan species a each o he si es,
Hylocha is leuco is in Mexico and Calyp e anna in Cali o nia
(see Resul s sec ion), ou esul s show a ce ain o e lap
be ween S u us and each o he wo species when hey co-
occu (F= 7.12, p< 0.001), while he o e lap o hei mo i
signa u es is smalle in he absence o he mig a o y species.
In ela ion o he esponse o he hummingbi d commu-
ni y o species emo al, we ocused ou analyses on wo
di e en le els: a he le el o he whole communi y and a
he le el o each o he hummingbi d species. A he commu-
ni y le el, when he mig a o y species lea es he a ea he
niche o e lap be ween di e en esiden species dec eases
( igu e 2b, ma ginal R
2
= 0.31). This e ec is g ea es in
Cali o nia ( able 1a). In addi ion, we ind ha unc ional com-
plemen a i y, i.e. niche complemen a i y o he hummingbi d
species, dec eases in Mexico when he mig an lea es ( able 1
and igu e 2c, ma ginal R
2
= 0.37).
A he species le el, all h ee a iables e alua ed change
be ween pe iods be o e and a e he emo al o he mig a o y
species. Howe e , esiden hummingbi d species a y s ongly
in he magni ude and di ec ion o hese changes ( igu e 3;
(a)
(b)(c)
(d)(e)
mig a o y species p esen mig a o y species absen
Figu e 1. The s uc u e o in e ac ion ne wo ks be ween plan s and hummingbi ds a h ee si es ac oss a la i udinal g adien , (a) Alaska, (b,c) Cali o nia and (d,e)
Mexico. (a), (b) and (d) show pe iods when he mig a o y species is p esen , and (c) and (e) show pe iods when i is absen . Uppe polygons ep esen humming-
bi d species, ligh blue: esiden species, da k blue: mig a o y species. Flagged in ed a e he wo species ha show la ges changes be ween he wo pe iods,
Calyp e anna in Cali o nia and Hylocha is leuco is in Mexico. Lowe yellow polygons ep esen plan species. (Online e sion in colou Q6.)
(a)(b)(c)
p esen absen p esen absen
pe iod
p esen absen p esen absen
pe iod
no malized deg ee
0.5
0.4
0.3
0.2
0.1
0
50
100
unc ional complemen a i y
150
0.25
0.50
0.75
niche o e lap
1.00
Alaska Cali o nia
Cali o nia
Mexico
Mexico Cali o nia Mexico
Figu e 2. Boxplo s showing (a) no malized deg ee o he mig a o y species Selaspho us u us a h ee si es along i s mig a o y pa hway and (b) how pollina o
niche o e lap and (c) pollina o unc ional complemen a i y change ac oss pe iods (mig a o y species p esen –absen ). G ea e alues indica e highe le els o gen-
e aliza ion. (Online e sion in colou .)
oyalsocie ypublishing.o g/jou nal/ spb P oc. R. Soc. B 20200649
4
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
ARTICLE IN PRESS
RSPB20200649—17/6/20—15:48–Copy Edi ed by: No Men ioned
elec onic supplemen a y ma e ial, igu e S3). In he case o
Cali o nia, we see pa icula ly la ge changes in hese a iables
o he species Calyp e anna, which becomes mo e gene alized
(la ge d0, elec onic supplemen a y ma e ial, igu e S4, model
ma ginal R
2
= 0.5) and mo e impo an o he plan species
ha depend on i (g ea e s eng h, model ma ginal R
2
=0.52)
in he absence o he mig a o y species. In Mexico, he species
whose ole changes mos is Hylocha is leuco is, which also di e -
si ies he numbe o plan species i isi s (la ge no malized
deg ee, model ma ginal R
2
= 0.53) and becomes mo e impo -
an o plan species in he communi y (g ea e s eng h) in
he absence o he mig a o y species ( igu e 3).
As o ou hi d ques ion ega ding he e ec o species
emo al on plan ep oduc i e success, he loss o he
mig a o y hummingbi d is associa ed wi h nega i e e ec s
in Cali o nia and posi i e e ec s in Mexico (elec onic sup-
plemen a y ma e ial, igu e S5). In he case o Cali o nia,
Ribes sanguineum shows dec eased ui se a e he loss o
he mig a o y hummingbi d ( igu e 4b, model ma ginal
R
2
= 0.24), and all h ee plan species show sligh ly dec eased
ui weigh (elec onic supplemen a y ma e ial, igu e S8,
model ma ginal R
2
= 0.03). Cas illeja a inis shows a dec ease
in ui leng h (model ma ginal R
2
= 0.64) and wid h (model
ma ginal R
2
= 0.64) wi h loss o he mig an hummingbi d
(elec onic supplemen a y ma e ial, igu es S9 and S10).
By con as , in Mexico he numbe o seeds pe ui
inc eases wi h he loss o he mig a o y species o h ee o
he ou plan species sampled (elec onic supplemen a y
ma e ial, igu e S5; igu e 4a, model ma ginal R
2
= 0.29). In
addi ion, wo o he species show an inc ease in ui leng h
and ui wid h (elec onic supplemen a y ma e ial, igu es
S6 and S7, ma ginal R
2
= 0.87 and 0.93, espec i ely).
4. Discussion
Ou esul s show ha he o aging niches and unc ional oles
o indi idual hummingbi d species a e dynamic and shi
ac oss hei dis ibu ion a eas. In pa icula , in he case o
S. u us i s ole changes om gene alis o highly specialis
in he communi y, and hese niche shi s ha e a di ec e ec
on i s e ec i eness as a pollina o . Mo eo e , ou esul s
p o e ha he empo a y local emo al o a species can
lead o impo an changes in he unc ional oles co e ed
by he emaining species, and ha hese changes ha e an
e ec on he unc ions hey pe o m. He e, we see he g ea es
changes in he niches occupied by one hummingbi d species
in Mexico (H. leuco is) and one in Cali o nia (C. anna). In bo h
cases, he species expand hei niches and become mo e
impo an in he communi y (g ea e s eng h alues and
hus g ea e dependence o plan s on hem) wi h he loss o
he mig an om he communi y. Howe e , hese changes
en ail di e en consequences ac oss he biodi e si y g adien .
In Mexico, he la ge di e si y o hummingbi d species
allows o ano he species (H. leuco is) o ake o e he ole
o he los mig an , main aining and e en imp o ing on he
unc ion pe o med by he la e . Con e sely, in Cali o nia,
whe e hummingbi d species di e si y is lowe , he ole o
he mig a o y species is no ully co e ed in i s absence,
leading o a educed unc ion.
P e ious expe imen al esea ch has sugges ed ha species’
unc ional oles shi in esponse o changes in communi y
composi ion wi h di e ing consequences o he unc ions pe -
o med [11,13]. While B osi & B iggs [11] ound a nega i e
e ec o he loss o an abundan bumblebee o he ep oduc-
i e success o Delphinium ba beyi (Ranunculaceae), Halle
e al.[13] ound ha he exclusion o bumblebees did no com-
p omise he success o Asclepias e icilla e (Apocynaceae). In
his second case, he ole o he los species was aken o e
by ano he species (Polis es wasps). Howe e , hese s udies,
al hough highly aluable, we e bo h ca ied ou a small spa ial
scales and o ela i ely sho pe iods o ime, and ocused on
he ep oduc i e success o jus one species o plan . By con-
as , ou use o mig a o y species as a p oxy o species loss
allows us o e alua e he consequences o whole-landscape
emo als o he same species on na u al communi ies along
la i udinal g adien s and o ocus on he consequences o a
la ge subse o he plan species in he communi y. O
cou se, he con inen al scale o ou app oach and he ac ha
we ha e da a o jus 1 yea also p esen some ca ea s, such
as he con ounding e ec o la i ude and di e si y. By using a
landscape-le el na u al emo al o a species, we a e able o
show how species loss has la ge nega i e e ec s on some
plan species, no e ec on o he s, and in some cases p oduces
e ec s ha a e o e -compensa ed by changes in he oles o
emaining species. Al hough ou sys ems is no exac ly a
eplica o a species ex inc ion, since he communi ies we
s udy ha e e ol ed wi h he p esence o his mig a o y species,
i clea ly shows ha unde s anding he consequences o species
loss o ecosys em pe sis ence equi es o a communi y-le el
app oach ha ocuses on he combined esponses o mul iple
species and ha akes in o accoun he possible beha iou al
changes o he emaining species.
Ou esul s p o ide e idence o he ole o biodi e si y as
insu ance agains species loss [29]. As he mig a o y species
disappea s, we ind ha in he mo e di e se communi y
he unc ions i pe o ms a e co e ed by o he species ha
compensa e o he loss and e en imp o e he unc ion o
he los species, hus ensu ing he s abili y o he sys em.
Howe e , in he less di e se communi y, lowe hummingbi d
di e si y p ecludes he unc ion o he mig a o y species om
being co e ed by o he hummingbi d species and leads o a
g ea e han 10% dec ease in he ep oduc i e success o he
mig an ’s p e e ed plan species. I is impo an o no e
ha al hough he species Hylocha is leuco is seems o ake
Table 1. Resul s o GLMMs showing he e ec o pe iod (mig a o y
hummingbi d species p esen o absen ) on esiden communi y (a) niche
o e lap and (b) unc ional complemen a i y. I alic le e s indica e a iables
wi h la ge and significan e ec s.
es ima e s.e. - alue
(a) niche o e lap
(in e cep ) 0.39 0.12 3.34
pe iod −0.18 0.07 −2.51
plan species ichness −0.01 0.01 −0.47
flo al a ailabili y 0.00 0.00 0.37
(b) unc ional complemen a i y
(in e cep ) 12.58 21.21 0.59
pe iod −18.82 8.19 −2.30
plan species ichness 6.75 2.69 2.52
flo al a ailabili y 0.00 0.00 0.08
oyalsocie ypublishing.o g/jou nal/ spb P oc. R. Soc. B 20200649
5
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
ARTICLE IN PRESS
RSPB20200649—17/6/20—15:48–Copy Edi ed by: No Men ioned
o e he ole as pollina o o mos plan species isi ed by
S. u us in i s absence, ou analyses o he mo i signa u es
o bo h species show some dispa i ies sugges ing ha
he esiden species is no ully able o co e he ole o he
mig an . Ne e heless, he unc ional consequences o chan-
ging he indi ec in e ac ions cap u ed by mo i analyses a e
s ill a om being ully unde s ood.
Insec pollina o s, p esen du ing bo h pe iods, may also
be impo an pollina o s in hese sys ems, ye hei ac i i y
is appa en ly no able o compensa e o he loss o his one
species om ou obse a ions. Howe e , including insec
species would ha e allowed us o e alua e he s uc u e o
he whole ne wo k o in e ac ions in ol ing plan species
which could po en ially e eal in e es ing esul s.
We also show how his species’ o aging niche can d ama i-
cally change along i s dis ibu ion ange. In pa icula , we ind
ha S. u us beha es as a mo e gene alis species a i s win e ing
and b eeding a eas, while i becomes a specialis du ing pa o
(a)
(b)
p esen absen
p esen absen
p esen absen p esen absen p esen absen
p esen absen p esen absen p esen absen
pe iod
pe iod
no malized deg ee
s eng h
0.25
0.50
0.75
1.00
1.25
0.25
0.50
0.75
1.00
1.25
0.25
0
2
4
6
0
2
4
6
0
2
4
6
0.50
0.75
1.00
1.25
Cali o nia Cali o nia
Calyp e anna
A his heloisa Colib i halassinus Eugenes ulgens Hylocha is leuco is
Selaspho us sasin Amazilia be yllina Amazilia ioliceps
Mexico Mexico
Mexico MexicoMexico Mexico
Mexico
Selaspho us pla yce cus
Mexico
Lampo nis ame hys inus
Mexico
Selaspho us calliope
Cali o nia Cali o nia
Calyp e anna
A his heloisa Colib i halassinus Eugenes ulgens Hylocha is leuco is
Selaspho us sasin Amazilia be yllina Amazilia ioliceps
Mexico Mexico
Mexico MexicoMexico Mexico
Mexico
Selaspho us pla yce cus
Mexico
Lampo nis ame hys inus
Mexico
Selaspho us calliope
A his heloisa
Colib i halassinus
Eugenes ulgens
Hylocha is leuco is
Calyp e anna
Selaspho us sasin
Amazilia be yllina
Amazilia ioliceps
species
Selaspho us pla yce cus
Lampo nis ame hys inus
Selaspho us calliope
Figu e 3. Boxplo s showing how species-le el me ics (a) no malized deg ee and (b) s eng h change ac oss pe iods (mig a o y species p esen –absen ) o each o
he esiden hummingbi d species in he communi y. (Online e sion in colou .)
oyalsocie ypublishing.o g/jou nal/ spb P oc. R. Soc. B 20200649
6
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
ARTICLE IN PRESS
RSPB20200649—17/6/20—15:48–Copy Edi ed by: No Men ioned
i s mig a o y jou ney. This di e ence has a consequence o he
ole he species occupies in he communi y, which becomes
mo e impo an in he a ea in which i beha es as a specialis
and appa en ly mo e e icien pollina o . This esul has impli-
ca ions o ying o de e mine he esilience o na u al
communi ies o species loss. Ea ly e o s a doing so assumed
ha species loss mean in e ac ions loss, allowing no es uc u -
ing o ewi ing o in e ac ions (e.g. [30]). Mo e ecen e o s
ha e ied o ake in o accoun he abili y o na u al sys ems
o es uc u e h ough species ole changes byallowing ace ain
le el o in e ac ion ewi ing (e.g. [31]). Al hough we a e s ill a
om unde s anding wha d i es changes in species beha iou s
and wha he consequences o hese changes a e, ou s udy
clea ly shows ha in e ac ion ewi ing may be common
and is impo an o bo h sides o bipa i e in e ac ions like
plan –pollina o ne wo ks. The exis ence o in e ac ion ewi ing
migh be mo e common in sys ems like ou s adap ed o annual
mig a ion p ocesses, ye he equency o his phenomenon
ac oss di e en ecosys ems is no ye clea .
Global change impac s a e pa icula ly p essing in he
case o mig a o y species, which a e o ced o shi hei
mig a o y beha iou s in esponse o changes in he sui abili y
o hei b eeding and win e ing habi a s [32]. Much global
change esea ch has been de o ed o s udying he changes
o he mig a o y ou es, depa u e o a i al da es [33]o
mig a o y species, ye less a en ion has been gi en o he
unc ional impac s ha hese changes could ha e wi hin
he na u al communi ies ha suppo hem [34]. Indeed,
mig a o y species anspo nu ien s and ene gy as well as
o he o ganisms (e.g. seeds, mollusks, pa asi es and pa hogens,
pollen g ains) be ween dis an loca ions [34], hus coupling
ecological communi ies h oughou hei mig a o y ou es.
S udies ocusing on he in e ac ions be ween mig an and esi-
den species ha e shown ha mig an s can al e ood web
opologies, and he s uc u e and dynamics o na u al commu-
ni ies [34]. Mig an s hus ha e he po en ial o a ec ecosys em
unc ioning ac oss he di e en esiden communi ies hey
connec in hei jou neys, and unde s anding hei impac s
equi es o in eg a i e s udies linking biogeog aphy o
communi y ecology among o he disciplines. In he case o
hummingbi ds in pa icula , mig an species a e key playe s
ha inc ease plan –hummingbi d ne wo k cohesi eness by
in e ac ing wi h a di e se se o plan species [35].
5. Conclusion
Recognizing he dynamic aspec s o na u al communi ies, as
well as he po en ial o eo ganiza ion o hei in e ac ion ne -
wo ks, will inc ease ou abili y o p edic esponses o bo h
na u al and an h opogenic dis u bances—including majo d i-
e s o biodi e si y loss, such as land-use and clima e change.
Clima e change is expec ed o ha e signi ican impac s o bio-
di e si y, including phenological shi s and la i udinal shi s in
he dis ibu ion o many species as hey ack hei clima e
niches, as well as changes o mig a ion ou es o e en a cessa-
ion o mig a o y beha iou s. As shown by ou s udy, such
changes in he dis ibu ion o mig a o y species o high
(a)
(b)
p esen absen
Fuchsia encliand a
Lobelia laxi lo a
Rubus adeno ichos
Sal ia iodan ha
Cas illeja a inis
species
species
Ribes menziesii
Ribes sanguineum
pe iod
p esen absen
pe iod
0.25
0
0.50
ui se seed numbe
0.75
1.00
–2
–1
0
1
2
3
Figu e 4. Boxplo s showing he e ec o pe iod o sampling (mig a o y species p esen –absen ) on (a) he scaled numbe o seeds pe ui in ou Mexican plan s
and (b) ui se in h ee Cali o nia plan s. (Online e sion in colou .)
oyalsocie ypublishing.o g/jou nal/ spb P oc. R. Soc. B 20200649
7
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
ARTICLE IN PRESS
RSPB20200649—17/6/20—15:48–Copy Edi ed by: No Men ioned
conse a ion alue wi h key oles in he communi ies could
ha e indi ec nega i e impac s on many o he species, includ-
ing plan s and o he pollina o s, gi en he in e dependencies
o species wi hin na u al ecosys ems.
Da a accessibili y. This a icle has no addi ional da aQ1 .
Au ho s’con ibu ions. A.M. de eloped he idea and secu ed unding;
A.M., C.L., U.M.L. and S.D.I. collec ed ield da a; A.M. an all he
analyses, all au ho s con ibu ed o w i ing he manusc ip .
Compe ing in e es s. We decla e we ha e no compe ing in e es sQ2 .
Funding. Wo k p oduced wi h he suppo o a 2018 Leona do G an
o Resea che s and Cul u al C ea o s, BBVA Founda ion. The Foun-
da ion akes no esponsibili y o he opinions, s a emen s and
con en s o his p ojec , which a e en i ely he esponsibili y o i s
au ho s. AM ecei ed unding om an Ike basque Resea ch Fellow-
ship. Special hanks o Ma k Readdie, Feynne A ias o suppo
du ing ieldwo k a Landels Hill Big C eek Rese e Q3.
Acknowledgemen s. We also hank Vic o Lau eano Güi ón o he acili-
ies and he pe mi s p o ided o wo k a Las Joyas ield s a ion. We
hank Ignasi Ba omeus o commen ing on a p e ious e sion o his
manusc ip .
Re e ences
1. G imm NB e al. 2013 The impac s o clima e
change on ecosys em s uc u e and unc ion. F on .
Ecol. En i on. 11, 474–482. (doi:10.1890/120282)
2. Ripple WJ, Wol C, Newsome TM, Ba na d P,
Moomaw WR. 2019 Wo ld scien is s’wa ning o a
clima e eme gency. BioScience 1, e1400103. (doi:10.
1093/biosci/biz088)
3. Foley JA e al. 2005 Global consequences o land
use. Science 309, 570–574. (doi:10.1126/science.
1111772)
4. Bascomp e J. 2007 Ne wo ks in ecology. Basic Appl.
Ecol. 8, 485–490. (doi:10.1016/j.baae.2007.06.003)
5. Memmo J, God ay H, Gauld I. 1994 The s uc u e
o a opical hos -pa asi oid communi y. J. Anim.
Ecol. 63, 521–540. (doi:10.2307/5219)
6. Poiso T, S ou e DB, G a el D. 2015 Beyond
species: why ecological in e ac ion ne wo ks a y
h ough space and ime. Oikos 124, 243–251.
(doi:10.1111/oik.01719)
7. Olesen JM, Bascomp e J, Elbe ling H, Jo dano P.
2008 Tempo al dynamics in a pollina ion ne wo k.
Ecology 89, 1573–1582. (doi:10.1890/07-0451.1)
8. Pe anidou T, Kallimanis AS, Tzanopoulos J, Sga delis
SP, Pan is JD. 2008 Long- e m obse a ion o a
pollina ion ne wo k: luc ua ion in species and
in e ac ions, ela i e in a iance o ne wo k s uc u e
and implica ions o es ima es o specializa ion. Ecol.
Le . 11, 564–575. (doi:10.1111/j.1461-0248.2008.
01170.x)
9. Mag ach A, González-Va o J, Boi ie M, Vilà M,
Ba omeus I. 2017 Honeybee spillo e eshu les
pollina o die s and a ec s plan ep oduc i e
success. Na . Ecol. E ol. 1, 1299–1307. (doi:10.1038/
s41559-017-0249-9)
10. Gomez-Mo acho T, Heeb T, Liho eau M. 2017 E ec s
o pa asi es and pa hogens on bee cogni ion. Ecol.
En omol. 42,51–64. (doi:10.1111/een.12434)
11. B osi BJ, B iggs HM. 2013 Single pollina o species
losses educe lo al ideli y and plan ep oduc i e
unc ion. P oc. Na l Acad. Sci. US.A 110, 13 044–
13 048. (doi:10.1073/pnas.1307438110)
12. B osi BJ, Niezgoda K, B iggs HM. 2017 Expe imen al
species emo als impac he a chi ec u e o
pollina ion ne wo ks. Biol. Le . 13, 20170243.
(doi:10.1098/ sbl.2017.0243)
13. Halle AC, Mi chell RJ, Chambe lain ER, Ka on JD.
2017 Pollina ion success ollowing loss o a equen
pollina o : he ole o compensa o y isi a ion by
o he e ec i e pollina o s. AoB PLANTS 9.(doi:10.
1093/aobpla/plx020)Q4
14. Johnsga d PA. 1983 The hummingbi ds o No h
Ame ica. Smi hsonian Ins i u ion.Q5
15. A izmendi MC, Dominguez CA, Di zo R. 1996 The
ole o an a ian nec a obbe and o hummingbi d
pollina o s in he ep oduc ion o wo plan species.
Func . Ecol. 10, 119–127. (doi:10.2307/2390270)
16. Bick o d C, Rich P. 1979 Vege a ion and lo a o he
Landels-Hill big c eek ese e, cali o nia. Uni e si y o
Cali o niaQ5 San a C uz.
17. Do mann CF, F und J, Blu hgen N, G ube B. 2009
Indices, g aphs and null models: analyzing bipa i e
ecological ne wo ks. Open Ecol. J. 2,7–24. (doi:10.
2174/1874213000902010007)
18. Ho n HS. 1966 Measu emen o ‘o e lap’in
compa a i e ecological s udies. Am. Na . 100,
419–424. (doi:10.1086/282436)
19. De o o M, Bailey S, C aze P, Memmo J. 2012
Unde s anding and planning ecological es o a ion
o plan –pollina o ne wo ks. Ecol. Le . 15,
319–328. (doi:10.1111/j.1461-0248.2012.01740.x)
20. Pe chey OL, Gas on KJ. 2007 Dend og ams and
measu ing unc ional di e si y. Oikos 116,
1422–1426. (doi:10.1111/j.0030-1299.2007.
15894.x)
21. Blü hgen N, Menzel F, Blü hgen N. 2006 Measu ing
specializa ion in species in e ac ion ne wo ks. BMC
Ecol. 6,1–12. (doi:10.1186/1472-6785-6-9)
22. Bascomp e J, Jo dano P, Olesen JM. 2006
Asymme ic coe olu iona y ne wo ks acili a e
biodi e si y main enance. Science 312, 431–433.
(doi:10.1126/science.1123412)
23. Chao A, Colwell RK, Lin C, Go elli NJ. 2009 Su icien
sampling o asymp o ic minimum species ichness
es ima o s. Ecology 90, 1125–1133. (doi:10.1890/
07-2147.1)
24. Hsieh TC, Ma KH, Chao A.2016 iNEXT: iN e pola ion
and EXT apola ion o species di e si y. package
e sion 2.0.8.
25. R De elopmen Co e Team, R. 2011 R: A language
and en i onmen o s a is ical compu ing.R
Founda ion o S a is ical Compu ing. See h p://
www. -p ojec .o g.
26. Simmons BI, Ci will AR, Bake NJ, Wauchope HS,
Dicks LV, S ou e DB, Su he land WJ. 2018 Mo i s in
bipa i e ecological ne wo ks: unco e ing indi ec
in e ac ions. Oikos. 128, 154–170. (doi:10.1111/oik.
05670))
27. Simmons BI, Swee ing M, Dicks LV, Su he land WJ,
Di Clemen e R. 2018 R. bmo i : A package o
coun ing mo i s in bipa i e ne wo ks. bioRxi
(doi:10.). (doi:10.1101/302356)
28. Ba es D, Mächle M, Bolke B, Walke S. 2015 Fi ing
Linea Mixed-E ec s Models Using lme4. J. S a .
So w. 67,1–48. (doi:10.18637/jss. 067.i01)
29. Lo eau M, Mouque N, Gonzalez A. 2003
Biodi e si y as spa ial insu ance in he e ogeneous
landscapes. P oc. Na l Acad. Sci. USA 100, 12765.
(doi:10.1073/pnas.2235465100)
30. Memmo J, Wase NM, P ice MV. 2004 Tole ance o
pollina ion ne wo ks o species ex inc ions.
P oc. R. Soc. Lond. B 271, 2605–2611. (doi:10.1098/
spb.2004.2909)
31. Kaise -Bunbu y CN, Mu S, Memmo J, Mülle CB,
Ca lisch A. 2010 The obus ness o pollina ion
ne wo ks o he loss o species and in e ac ions: a
quan i a i e app oach inco po a ing pollina o
beha iou . Ecol. Le . 13, 442–452. (doi:10.1111/j.
1461-0248.2009.01437.x)
32. Wilco e DS, Wikelski M. 2008 Going, going, gone: Is
animal mig a ion disappea ing. PLoS Biol. 6, e188.
(doi:10.1371/jou nal.pbio.0060188)
33. Howa d C, S ephens PA, Tobias JA, Shea d C,
Bu cha SHM, Willis SG. 2018 Fligh ange, uel
load and he impac o clima e change on he
jou neys o mig an bi ds. P oc. R. Soc. B 285,
20172329. (doi:10.1098/ spb.2017.2329)
34. Baue S, Hoye BJ. 2014 Mig a o y animals couple
biodi e si y and ecosys em unc ioning wo ldwide.
Science 344, 1242552. (doi:10.1126/science.
1242552)
35. Ma ín González AM, O nelas JF, Dalsgaa d B,
Má quez-Luna U, La a C. 2018 Me a-ne wo ks o he
s udy o biogeog aphical ai s in ecological ne wo ks:
he Mexican hummingbi d-plan assemblage. Sci.
Na . 105, 54. (doi:10.1007/s00114-018-1578-5)
oyalsocie ypublishing.o g/jou nal/ spb P oc. R. Soc. B 20200649
8
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
ARTICLE IN PRESS
RSPB20200649—17/6/20—15:48–Copy Edi ed by: No Men ioned
