Coevolution in Multispecific Interactions among Free-Living Species

hystericalcoolΚινητά – Ασύρματες Τεχνολογίες

10 Δεκ 2013 (πριν από 3 χρόνια και 6 μήνες)

81 εμφανίσεις

Evo

Edu

Outreach

(2010)

3:4
0

46





DOI

10.1007/s12052
-
009
-
0197
-
1


ORIGINAL

SCIENTIFIC

A
R
TICLE




Coevol
u
tion

in

Multispecific

Interacti
o
ns

among

F
r
ee
-
Living

Species



Pedro

Jordano








Published online:

29

December 2009

#

Springer

Science+Business

Media,

LLC

2009



Abst
r
act

Ec
o
log
ical

int
e
ractions among

spec
i
es

are

the
backbone

o
f

biodiversit
y
.

Inte
r
actions take

a

tremendous
vari
e
ty

of for
m
s

in

natu
r
e

and

have

perva
s
ive

conseq
u
ences
for

the

populati
o
n

d
y
namics

and

evolu
t
ion

of

spec
i
es.

A
persiste
n
t challenge

in

ev
o
lutionary bio
l
ogy

has

been

to
under
s
tand

how

coe
v
olution

has

produc
e
d

complex

webs

of
interact
i
ng

spec
i
es,

whe
r
e

a

large

numb
e
r

of

spec
i
es

interact
thro
u
gh

mu
t
ual

depende
n
ces

(e.g.,

mutual
i
sms) or

influen
-

ces

(e.g.,

preda
t
o
r

prey

interactio
n
s

in

food

web
s
).

Recent
work

on

mega
d
iverse

species

asse
m
blages in

ecolo
g
ical
com
m
unities has

uncove
r
ed

int
e
resting

repea
t
ed

patterns
that

emerge

in

these

c
omplex

net
w
orks of

multispeci
e
s
interact
i
ons.

They

include

the

prese
n
ce

of

a

core

of

super
-

genera
l
ists, proper

patterns

of

interac
tion

(that

resemble
nest
e
d

chinese

boxes
)
,

and

mu
l
tiple

modules

that

act

as

the
basic

blocks

of

the

comp
l
ex net
w
ork. The

structure

of
multi
s
pecies

int
e
ractions

rese
m
bles

o
ther

comp
l
ex

networks
an
d

i
s

centra
l

t
o

und
e
rs
t
an
din
g
i
ts

e
v
olut
i
o
n

an
d

the
conseq
u
en
ces

of

species

los
s
es

for

the

p
ersisten
c
e

o
f

the
whole

network.

Th
e
se

pat
t
erns

sugge
s
t

both

prec
i
se

ways
o
n

ho
w

coe
v
olut
i
o
n
goe
s

o
n

bey
o
n
d

s
i
m
p
le

p
a
i
rwise
interact
i
ons

and

scales

up

to

whole

com
m
unities.


Keywor
d
s

Co
e
volution

.

Multi
s
pecies

interact
i
ons

.

Mut
u
alism

.

Comp
l
ex

networks

.

P
o
llination

.

S
e
ed

disp
e
rsal

Introduction


Comp
l
exity

is

a

basic

property

of

the

biological

world.

The
funda
m
e
nta
l
deta
i
l
s

t
ha
t

s
urroun
d

th
e

na
t
ura
l

live
s

of
anim
a
ls

and

plants

fasc
i
nate

natu
r
alists but

often

limit

our
un
d
e
rstandi
n
g
o
f

th
e

basi
c

p
ri
n
ciple
s

tha
t

driv
e

a
nd
genera
t
e their

diversi
t
y
. What

are

the

main

patterns

of
interact
i
ons among

spec
i
es

in

mega
d
iverse

assembla
g
es,
suc
h

a
s

th
e

mutualisti
c

i
nteractio
n
s
amon
g

v
ertebrate
frug
i
v
o
res that

disp
e
rse

see
d
s

and

pl
an
t
s

that

pro
d
uce
fleshy

fru
i
ts

in

a trop
i
cal

fore
s
t?

Th
e
se

int
e
ractions

can
t
a
k
e

o
n

a

f
orm
i
d
a
bl
e
di
v
er
s
it
y

(Fig
.

1
a
)
,

generating
comp
l
ex

patterns

of

mutual

depende
n
ce

among

animals
and

plan
t
s that

are

more

than

the

sum

of

the

pairw
i
se
interact
i
ons.

The

complex

deta
i
ls

of

the

intim
a
te

associa
-

tions

bet
w
een

plants

and

ani
m
als we
r
e

already

recogn
i
zed
by

e
a
rly

b
otanists,

like

Christian

Kon
r
ad

Sprenge
l
,

who
s
e
seminal

book


D
i
scovery

o
f

the

sec
r
et

of Natu
r
e

in

the
st
r
u
c
tu
r
e

a
n
d

fe
r
ti
l
i
z
ati
o
n
of

f
lo
w
e
r
s


(
S
p
r
e
n
g
e
l

1
7
9
3
)
prese
n
ted

ample

evide
n
ce

for

the

importa
n
ce of

c
ross
-

fertili
z
ation in

plan
t
s.

His

work

was

sem
i
nal

for

Darwi
n

s
e
x
p
e
r
i
m
e
n
t
al

a
p
p
r
o
a
c
h
es

w
i
th

o
r
c
h
i
d
s

a
n
d

t
h
e

r
e
a
l
i
z
a
t
i
o
n

of
t
h
e


e
n
t
a
n
g
l
e
d

b
a
n
k


of

e
c
o
l
o
g
i
c
al

r
e
l
a
t
i
o
n
s
h
i
p
s

a
m
o
n
g

p
l
a
n
ts
a
n
d

a
n
i
m
a
l
s. T
h
is

h
a
s

b
e
en

a
n
d

c
o
n
t
i
n
u
e
s

to

be

(
T
h
o
m
p
s
o
n

2
0
0
6
)

o
n
e

of

t
h
e

m
o
st

c
h
a
l
l
e
n
g
i
n
g i
s
s
u
es

in

e
v
o
l
u
t
i
o
n
a
ry
b
i
o
l
o
g
y:

h
o
w

s
p
e
c
i
e
s

c
o
e
v
o
l
ve

w
h
en

i
n
t
e
g
r
a
t
ed in

c
o
m
p
l
e
x
w
e
bs

of

m
u
t
u
a
l
i
s
t
i
c,

a
n
t
a
g
o
n
i
s
t
i
c
,

c
o
m
p
e
t
i
t
i
v
e,

or

p
a
r
a
s
i
t
i
c
i
n
t
e
r
a
c
t
i
o
n
s
.

W
e

s
h
oul
d
n

t


b
e

s
u
rpr
is
ed

t
h
at

em
p
ir
i
cal

n
atu
r
ali
s
ts




rem
a
in

skeptical

about

the new

ins
i
ghts

that network

theo
r
y

P
.

Jordano

(
*
)

Integrative

Ecology

Group

(IEG),

Estación

Biológica

de

Doñana,

CSIC,

C/Americo

V
espucio

s/n,

Isla

de

La

Cartuja,

41092

Sevilla,

Spai
n

e
-
mail:

jordano@ebd.csic.es

and

the

formal

analysis of

complexi
t
y

bring

to

the

study

of
biodi
v
ersity

(
W
eitz

et

al.

2
0
07
).

First,

it

might

be

difficult

to
inco
r
porate research

a
gendas

that

inc
l
ude

a

v
ast

amou
n
t

o
f
multi
d
iscipl
i
nary

approa
c
hes

that

are

far

away

from

our
exper
t
ise.

Second,

we

ne
e
d

to

fully

under
s
tand

how

abstract

Evo

Edu

Outreach

(2010)

3:4
0

46

41










a


b

Animals





A
b
ur
r
ia

jacutinga




Eute
r
pe

edulis




Plants

Ramphastos

vitelli
n
us


Fig.

1

a

A

schematic

representation of

a

mu
tualistic

network

of
interactions

among

plants

producing

fleshy

fruits

and

the

vertebrate
frugivores

that

disperse

their seeds

in

an

Atlantic

rainforest

locality

in
SE

Brazil (redrawn from

Silva

et

al.

2005
).

The

figure

is

a

bipartite
graph,

i.e.,

a

repres
entation

of

the interactions

(link
s
)

occurring

among
the

species

(nodes)

in

one

set

(plants) and

the

species in

the

other

set
(frugivorous animals),

indicating

the

interactions that

occur

in

this
communit
y
.

b

The

basic

building

blocks

of

many

types

of

ecol
ogical
interactions

are

pairwise

relationships

of

mutual

dependence

or

mutual
influence

among

partner

species.

Each

link

in

a
(
highlighted

black

links)

actually

embeds

the

relative

dependence

of

a

given

plant

(e.g.,
palmito

Euterpe

eduli
s
) on

the

dispers
al service

of

the

frugivore
species

(e.g.,

the

Jacutinga

Aburria

jacutinga

or

the

Channel
-
billed
touca
n

Rhamphasto
s

vitell
i
nu
s
;
d
a
r
k

a
r
r
o
w
)

an
d

th
e

reciproc
a
l
dependence

of

the

frugivore

on

the

fruit

food

resource

provided

by
the

plant

(light

ar
r
o
w
).

In

th
is

case,

the

interaction

is

asymmetrical
since

the

jacutinga

depends

heavily

on

palmito

fruits

while

the

bird
has

only

a

minor

contribution

to

the

overall seed

dispersal

of

palmito,
which

is

a

keystone

resource

consumed by

a

diverse

coterie of
frugivores i
n

the

Atlantic

forest

(Galetti

and

Aleixo

1998
)




repre
s
entations

like

network

graphs

can

be

embe
d
ded

with
importan
t
na
t
u
r
a
l

h
istor
y

detai
l
s

o
f

specie
s

an
d

their
i
n
ter
a
ct
i
ons
. T
h
e

a
n
aly
s
i
s

o
f

ecol
o
gica
l

n
e
t
w
ork
s

i
s

a
formal

way

to

visu
a
lize,

explore,

an
d

addre
s
s

the

shared
patterns

that

lie

beyond

the

myriad

int
e
ractions

invo
l
ved

in
mega
d
iverse

multi
s
pecies

assembla
g
es

like

trop
i
cal forests,
coral

reefs

or

soil

mi
c
roorganisms, and

plants.

W
e

have

a
formal

too
l
,

with

a

solid

multi
d
iscipl
i
nary

knowledge

ba
se,
to

dis
s
ect

the

complexi
t
y

of

ecolo
g
ical

systems

by

mov
i
ng
from

the

reduc
t
ionist

an
a
lysis

of

their

comp
o
nent

parts
(e.g.,

pairw
i
se

int
e
ractions) to

the

analysis

of

the
i
r

mac
r
o
-

scop
i
c

prope
r
ties (Basco
m
pte and

Jorda
n
o

200
7
).

This
essay

briefly

revi
e
ws the

principl
e
s of

network

theo
r
y
applie
d

t
o

ecologi
c
a
l
s
ystem
s

an
d

c
onsider
s

th
e

n
e
w
insight
s

gaine
d

abou
t

th
e

coevolu
t
io
n
o
f

megadiverse
asse
m
blages

of

interact
i
ng

species.



Mutual

Ben
e
fits,

Ant
a
gonism,

a
n
d

Who

Eats

Whom

Co
m
plexity


P
r
e
d
a
to
r

p
r
e
y i
n
t
er
a
c
t
io
n
s

re
p
r
e
s
e
nt

an

i
c
o
n
ic

v
i
ew

o
f
ecolo
g
ical webs

and

have

been

a

central

focus

of

resea
r
ch
for

years

(Pi
m
m

et

al.

199
1
).

A net
w
ork

view

of

a

food

web
(Fig.

2a
)

includ
e
s

information about

the

mu
l
tiple

(inter
-

actio
n
s) lin
k
s

among

spec
i
es

(nodes)

in

t
he

web:

who

eats
whom

and

the

relative

magn
i
tude

of

energy

transfers each
int
e
r
a
cti
o
n re
p
rese
n
ts

(D
u
nne

e
t

a
l.

200
2
).

The

re
c
ent
analy
s
is

of

food

webs as

complex

networks

has

highligh
t
ed
basic

general

prin
c
iples

that

influence

their

stabil
i
ty

and

the

poss
i
b
ilities

for

recovery

after

severe

disturban
c
es

like

the
suppr
e
ssion

of

keys
t
one

super

predators

(Jack
s
on

et

al.

200
1
)

or loss

of

habi
t
at

genera
l
ists

that

comp
o
se

the core

of

the

interactio
n
s

(Sri
n
ivas
a
n

et

al.

200
7
).

The

vari
e
ty

of

antag
o
nistic,

preda
t
o
r

pre
y
, or

comp
e
ti
-

tive

e
f
f
ects port
r
ayed

in

food

web

analy
s
is

are

among

the
multi
p
le

types

of

interactions that

occur

in

natural

systems
(Thom
p
son

198
2
)

and

not

neces
s
arily

the

most

im
p
ortant
ones.

Think

for

exam
p
le

about

the

keys
t
one

relevance

of
plan
t

pollinator

and

plan
t

frugivore

interactions

for

tropi
-

cal

forests where

up

to

9
5
%

of

the

trees

and

subcan
o
py
shrubs

need

these

animals

for

effective

pol
l
en

tran
s
fer

and
succe
s
sful

regene
r
ati
o
n

(Bawa

199
0
;

Jorda
n
o

200
0
).

F
o
rest
regene
r
ation

wou
l
d

si
m
ply

col
l
apse

without

the

interven
-

tion

of

animal

mutual
i
sts.

Mut
u
alistic

webs

of

int
e
raction
are

best

depicted

as

bipartite

graphs

(Fig.

1a
,
2b
),

where

the
mutu
a
l

depende
n
ces

o
f

each

pairwise

interact
i
on can

be
repre
s
ented.

The

t
w
o

distinct

sets

of

spec
i
es

(a
ni
m
als

and
plants)

are

linked

through

coevo
l
ved

interact
i
ons

of

mutual
depende
n
ce

(Fig.

1b
)

that

d
epict

the

reciprocal

conseq
u
en
-

ces

of

the
i
r

interact
i
on.

In

the

same

way

as

trad
i
tional

food
webs

portray

the

patterns

of

energy

tran
s
fer in

ecosy
s
tems
(Fig.

2a
),

bipa
r
tite graphs

capture

the

ma
i
n

elements

of
coevo
l
ved

interact
i
ons

in

these

multis
p
ecific

assembla
g
es
of

spec
i
es:

genera
l
ization,

asy
m
metry

of

mutual

depend
-

ences
,

c
ompartments
,
etc
.

(Fig
.

2b
,

c
)
.

Thu
s

th
e

n
et
outco
m
es

of

these

int
e
ractions, as

we
ll

as

the
i
r

overall
complexit
y

i
n

for
m

(topolo
g
y
)

a
n
d

s
tru
c
tur
e
,
c
a
n

be
analyzed
.
Ne
t
wor
k

topolog
y

refer
s

mainl
y

t
o

i
t
s

s
ize
(nu
m
ber

of

nod
e
s)

and

the

form

and

d
ensity

of

links



42

Evo

Edu

Outreach

(2010)

3:4
0

46






















a

b

c
















d

e







Fig.

2

Complex

networks o
f

ecological

interactions

can

vary

in

their
shape,

link

densit
y
,

and

component

structu
r
e

depending

on

the

type of
interaction

they

embed.

While

food

webs

typically

describe

all

the
interactions

occurring

in

a

given

ecosystem

(
a
)

with

multiple

trophic
level
s

(Dunne

et

al.

2002
),

most

plant

animal

interactions

can

be
displayed

as

bipartite

graphs

(
b
)

describing

the

pairwise

pattern

of
mutual

interdependencies

(Jordano

1987
)

among

two

distinct

sets

of
an
i
ma
l
s

(
or
a
nge

nod
e
s
)

and

pl
a
nts

(
yel
l
o
w
). I
n
t
e
r
a
cti
o
n
s

a
m
ong
species

with

a

higher

degree

of

intimac
y
,

such

as

ant
-
plants

show

a
distinct

pattern

of

structure (
c
),

often

with

multiple

distinct

groups
(module
s
)

of

closely

intimate

associations

(Guimarães

et

al.

200
7
).

The

three

types

of

webs

share

a

complex

pat
tern

of

interactions

made
up

of

multiple

simple


building


blocks



or

motifs

(Bascompte

and
Melián

2005
)

that

vary

in

shape

and

frequency

across

these

networks
(
d
,

e).

Motifs

in

food

webs

(
d
)

include

sim
p
le

trop
h
ic cha
i
ns,
omnivor
y
,

apparent

competition,

and

intraguild

predation

(from

left
to

righ
t
);

those

in

bipartite graphs

(e)

include

diffe
r
ent

forms

of
generalization/specialization,

with

more

specialized (
e
,

to
p
)

and

more
generalized

(
e
,

botto
m
)

motifs.

Images

a,

b
,

and

c

produced

with
Food
W
eb3D,

wri
tten

by

R.

J.

W
illiams

and

provided

by

the

Pacific
Ecoinformatics

and

Computational

Ecology

Lab

(
ww
w
.foodwebs.or
g
)


among

them

(re
l
ative

to

the

maximum

possi
b
le),

a
s

well

as
thei
r

d
istr
i
but
i
on
.
T
w
o

n
etwork
s

migh
t

d
i
f
fe
r

i
n

the
intensi
t
y

of

mutual

effects among

species,

y
e
t

share

the
s
a
me

t
o
pol
o
g
y
.

Ne
t
w
o
rk

s
tru
c
ture c
o
n
v
eys

inf
o
rma
t
i
o
n
about

the

iden
t
ity of

the

nodes:

who

eats

whom

and
whether dis
t
inct subse
t
s of

nodes

more

linked

among

them
exist.

Re
c
ent

an
alyses

indi
c
ate

that

mutual
i
stic networks

have
spec
i
fic

signatures in

the
i
r

topo
l
ogy

and

structure

(i.e.,

the
way

spec
i
es

are

interconne
c
ted thro
u
gh

mu
t
ual

depend
-

ences
)

tha
t

c
onfe
r

mor
e

robustnes
s
a
n
d

stabilit
y

than
expec
t
ed

for

rando
m
ly

assembled

intera
ct
i
ons (
B
ascompte
and

Jorda
n
o

200
7
).

Resem
b
ling other

complex

networks
(bot
h

b
i
o
logica
l

an
d

n
on
b
iotic
;
Amara
l

e
t

al
.

2
0
00
),
mutu
a
listic

webs

are

charac
t
eriz
e
d

by

their

hetero
g
eneity
(Fig.

2b
),

w
i
th

a

numb
e
r

of

n
o
des

having

a

high

number

of
int
e
ra
c
tion
s
an
d

a

h
ig
h

n
um
b
e
r

o
f

n
o
de
s

w
it
h

few
interact
i
ons.

That

is,

there

are

a

few

super
-
generalist

species
that

form

a

well
-
c
o
nnect
e
d core

of

the

network

a
n
d

many
other

spec
i
es

with

few

interactions

(Jor
d
ano et

al.

200
3
).

A
rando
m
ly asse
m
bled web

would,

in

cont
r
ast,

have

a

more
even

distri
b
ution

of

interact
i
ons

among

spec
i
es,

as

null

mode
l
s

of

mu
t
ualistic

webs

indicate

(Váz
q
uez

and

Aizen

200
3
).

An

interes
t
ing

p
roperty

of

hete
r
ogeneous

networks
is

that

they

are

very

robust

to

random

dis
t
urbanc
e
s

(loss

of
a

nod
e
)

but

very

sens
i
tive

to

sele
c
tive

losses

of

nodes

at

the
core

(Albert

et

al.

2
0
00
).

Therefo
r
e,

mutual
i
stic

networks

of
interact
i
on organized

around

a

distinct

subset

of

super
genera
l
ists

can

be

reaso
n
ably

robust

to

dis
t
urbanc
e
s not
dire
c
ted

to

this

cent
r
al

b
ackbone

of

their

str
u
cture.

T
h
r
e
e

a
d
d
it
i
o
n
al

si
g
n
a
t
u
r
e
s

o
f

m
u
tu
a
l
i
st
i
c

n
e
t
w
o
r
k
s
have

been

described

as

charac
t
eristic

prope
r
ties,

indep
e
n
-

dent

of

the

type

of

int
e
raction

and

its

geogra
p
hic

setting
(Basco
m
pte

et

al.

200
3
;

Olesen

et

al.

200
7
).

First, t
he
int
e
racti
o
ns
a
re

n
e
sted.

If

we

repre
s
ent

the

int
e
ract
i
on
network (e.g.,

Fig.

2
b
)

as

a

matrix,

with

animal

spec
i
es as
rows,

and plant

species

as colu
m
ns,

we

can

tally

the

spec
i
es
pairs

that

int
e
ract (cell

values

of

the

ma
t
rix

would

be

one)
or

not

(cells

values

as

zero).

F
o
r

instance, a

series of

field
censu
s
es

similar

to

used

for

mon
i
toring

the

spec
i
es

p
resent
in

a

given

area

can

help

us

to

cata
l
og

in

matrix

form

the
mu
t
u
a
l
ist
ic

in
t
e
r
action
s

p
r
e
s
en
t

i
n

a

give
n

c
om
m
unity
(Fig.

3a
).

If

the

spec
i
es

in

rows

and

colu
m
ns

are

sor
t
ed
from

the

most

genera
l
ist

to

the

most

spec
i
alist,

we

can



Evo

Edu

Outreach

(2010)

3:4
0

46

43



























































1

1

1

1

1

0

1

1

1

1

1

0

0

0

1

0

0

1

1

1

1

0

0

1

1

1

1

1

1

1

1

1

1

1

1

1

1

0

0

0

0

1

0

0

0

0

1

0

0

0

1

1

1

0

0

1

0

1

0

0

0

0

1

1

0

1

1

1

1

0

1

0

0

0

0

0

1

1

1

1

1

1

1

1

1

1

0

0

0

0

1

0

0

0

0

0

0

0

0

0

0

1

1

1

1

0

1

1

1

1

1

1

0

0

0

0

0

0

0

0

0

0

0

0

0

1

0

1

0

1

1

0

1

1

0

0

0

0

1

1

1

0

0

0

0

0

0

0

0

0

1

1

1

0

1

0

0

0

1

0

0

0

1

0

0

0

0

0

0

1

0

0

0

0

0

1

1

0

0

1

0

0

0

0

0

0

1

0

1

0

0

0

0

0

1

0

0

1

0

0

1

1

1

0

1

1

0

0

1

0

0

0

1

0

0

0

0

0

0

0

0

0

0

0

0

1

1

0

1

0

0

1

1

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

1

1

0

1

0

0

0

1

0

1

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

1

1

1

1

0

0

0

0

0

0

1

0

0

0

0

0

0

0

0

0

0

0

0

0

0

1

1

1

0

1

0

0

0

0

0

0

0

1

0

0

0

0

0

0

0

0

0

0

0

0

0

1

1

1

0

0

1

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

1

1

0

1

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

1

1

1

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

1

0

1

0

0

0

0

0

0

0

0

0

0

0

1

0

0

0

0

0

0

0

0

0

0

1

1

0

0

0

0

0

0

0

0

0

1

0

0

0

0

0

0

0

0

0

0

0

0

0

1

0

0

0

0

1

0

0

0

0

0

0

0

0

0

0

1

0

0

0

0

0

0

0

0

0

1

0

0

0

1

1

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

1

0

0

0

0

0

0

0

0

0

1

0

0

0

0

0

0

0

0

0

0

0

0

0

1

0

0

0

0

1

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

1

0

0

0

0

1

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

1

1

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

1

1

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

1

0

0

0

0

1

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

1

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

1

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

1

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

1

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

1

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

1

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

1

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0


Anima
l

p
h
yloge
n
y


a

b


P
lant

p
h
yloge
n
y



1

1

0

0

1

1

0

0

0

0

1

1

0

0

1

1







1

1

0

0

0

0

1

1

1

1

0

0

0

0

1

1






1

1

0

0

0

1

0

1

1

0

1

0

0

0

1

1


Fig.

3

P
h
y
l
o
g
e
n
e
t
i
c

p
at
t
e
r
ns

a
r
e

e
m
be
d
d
ed

i
n

p
l
a
nt

a
ni
m
a
l

i
n
t
e
r
a
c
ti
on
n
e
twor
k
s

a
n
d

ca
n

i
n
fl
u
enc
e

the
i
r

co
e
vo
l
ut
i
on
.

Th
e

i
nter
a
cti
o
n

p
attern
ca
n

b
e

d
escr
i
b
e
d

a
s

a

p
re
s
enc
e

absen
c
e

matr
i
x

d
enot
i
n
g

th
e

ob
s
erved
an
d

u
nobserve
d

int
e
raction
s

(a
s

show
n

h
e
re
)

o
r

a

qu
a
nti
t
ati
v
e

matrix
w
i
th

d
a
ta

on

i
n
te
r
ac
t
i
on

s
t
r
e
n
g
t
h

f
or

e
ac
h

o
b
s
e
r
v
e
d

p
a
i
r
w
is
e

i
n
t
e
r
a
c
t
i
o
n
.
H
o
w

e
a
c
h

s
p
ecie
s

i
n
teract
s
c
a
n

b
e

a
f
f
e
cte
d

it
s

e
vo
l
ut
i
on
a
r
y

hi
s
tory
(p
h
en
o
t
y
pi
c

trai
t
s
;

a
)
.

Fo
r

example
,

closel
y

relat
e
d

specie
s

m
i
ght

h
a
v
e

a

s
i
mila
r

p
att
e
r
n

o
f

i
ntera
c
tion
,
s
imp
l
y

becaus
e

o
f

n
iche
conservati
s
m
(Rezend
e

e
t

a
l
.

20
0
7
)
.

B
esides
,

th
e

in
t
eractio
n

p
attern
it
s
el
f

ca
n

b
e

s
u
bj
e
c
t

t
o

t
h
e

e
f
f
ec
t
o
f

b
o
t
h

th
e

a
nima
l

a
n
d

p
lant
p
h
y
l
og
e
net
i
c

h
i
st
o
rie
s

(
b
)
,

wit
h

a
marke
d

tren
d

fo
r

th
e

interact
i
on
s

t
o
matc
h

t
h
e

p
h
yl
o
g
e
n
eti
c

h
i
st
o
r
y

o
f

th
e

tw
o

g
ro
u
p
s

(
b
,

to
p
)
,

on
e

o
f
the
m

(th
e
p
l
ant
s

i
n

thi
s

case
;

b
,

m
i
ddle
)

o
r

n
o
n
e

(b
,

b
o
tt
o
m
)

Jor
d
ano
an
d

B
ascompte
,

i
n

p
rep




defi
n
e

the

nest
e
dness

of

the

ma
t
rix

as

a

devia
t
ion

from

a
random

pattern

and

its

c
loseness

to

a

perfectly

packed

arra
y
.
That

is,

how

c
lose

the

matrix

resembles

perfectly

order
e
d
interact
i
ons such

that

each

spec
i
es

interactions

are

subsets
of

those

with

wh
i
ch

the

more

genera
l
ist spec
i
es

interact
(i.e.,

a

situation

where

all


1


in

the

matrix

would

be
packed

to

the

left

of

the

matr
ix diago
n
al, so

that


0


values
rem
a
in

to

the

righ
t
). The

ma
t
rix

in

Fig.

3a

is

highly

nest
e
d.
T
ake

the

int
e
ractions of,

sa
y
,

the

third

plant

speci
e
s

(third
colu
m
n

from

the

lef
t
); all

of

them

except

one

invo
l
ve
anim
a
l

spec
i
es

with

which

the

more

genera
l
ized

first

and
second

plant

spec
i
es

also

int
e
ract. And

if

we

carry

on

the
comp
a
rison for

more

spec
i
alized

plants

(col
u
mns

to

the
right

o
f

the

matrix), the

trend

is

prese
r
ved

so

that

the

final
pattern

(Fig.

3a
)

is

character
i
stical
l
y
n
ested

with

most
intera
c
ti
o
n
s
m
a
pp
e
d

i
n

th
e

u
p
p
e
r

h
al
f

o
f

th
e

m
a
trix.
Basco
m
pte

et

al.

(
2
003
)

have

sho
w
n

that

plan
t

pol
l
inator
and

plan
t

frug
i
vore

asse
m
blages

often

show

n
ested

patterns

(like

the

one

sho
w
n

in

Fig.
3
a
) with two

key

prope
r
ties:

the
p
r
e
s
e
n
ce

of

a

c
o
re

of

g
e
n
e
r
a
l
i
s
t
s

t
h
at

i
n
t
e
r
a
ct

a
m
o
n
g

t
h
em
a
n
d

a

s
e
t

of

m
o
re

s
p
e
c
i
a
l
i
z
e
d

s
p
e
c
i
e
s

t
h
at

i
n
v
a
r
i
a
b
ly

t
e
n
d
s

to
i
n
t
e
r
a
ct

w
i
th

s
p
e
c
i
e
s

in

t
h
e

c
o
re

(
i
.
e
.,

a
s
y
m
m
e
t
r
i
c

s
p
e
c
i
a
l
i
-

z
at
i
o
n
)
.

R
e
c
e
nt

r
e
s
e
a
r
ch h
a
s

s
h
o
w
n

t
h
at

n
e
s
t
ed

p
a
t
t
e
r
ns
i
n
c
r
e
a
se

r
o
b
u
s
t
n
e
ss

to

t
he

l
o
ss

of

s
p
e
c
i
e
s

a
n
d

i
n
t
e
r
a
c
t
i
o
ns
a
n
d

f
a
v
o
r

i
n
c
r
e
a
s
e
d

d
i
v
e
r
si
ty

in

c
o
m
p
a
r
i
s
on
w
i
t
h

r
a
n
d
o
m
ly
a
s
s
e
m
b
l
e
d

m
u
t
u
a
l
i
s
t
i
c

c
o
m
m
u
n
i
t
i
e
s

(
B
a
s
c
o
m
p
te

et

a
l
.

2
0
0
3
;
B
u
r
g
o
s

et

a
l
.

2
0
0
7
;

B
a
st
o
l
la

et

a
l
.

2
0
0
9
).

S
e
cond,

the

w
i
reframe

of

interact
i
ons among

spec
i
es

is
b
u
ilt

o
n

a
s
y
m
m
e
t
ric

and

w
e
ak

re
c
ip
r
o
c
al
d
e
p
e
n
d
e
nc
e
s
(Jor
d
ano

198
7
;

Ba
s
compte

et

al.

200
6
).

Just

realize

that
the

z
ero

to

one

recor
d
s

(pre
s
enc
e

absenc
e
)

for

the

pairw
i
se
interact
i
ons can

take

the

form

of

quant
i
tative

estimates

of
t
h
e

a
c
tu
a
l

st
r
en
g
th

o
f

d
ep
e
n
d
e
n
ce
o
f

e
ach

p
a
rt
n
e
r
.

In

F
i
g
.

1b
,

t
h
e

st
r
e
ngt
h

o
f

m
utua
l

d
ependence
s
b
e
t
w
een
Euterpe

edulis

and

two

of

its

ma
j
or

seed

dispersers

in

the
Atlan
t
ic

rainforest

of

SE

Brazil

vary

signif
i
cantl
y
.

While

the



44

Evo

Edu

Outreach

(2010)

3:4
0

46




palm

is

qui
t
e

dep
e
ndent on

the

toucan

for

succe
s
sful fruit
rem
o
val,

the

fru
i
ts

are

a

minor

part

of

the

toucan

genera
l
ist
diet;

in

cont
r
ast,

the

jacutinga

rel
i
es

e
x
tensively

on

the

palm
fruits,

but due

to low

abundan
c
e

and inf
r
equent

visitation

to
t
h
e

f
r
u
i
t
i
n
g
pa
lm
s
,

it

p
l
a
y
s

a

s
e
co
n
da
ry

r
o
le

a
s

s
e
e
d
disp
e
rse
r
. This

pat
t
ern

is

extrem
e
l
y

common

in

p
lan
t


anim
a
l

mutual
i
sms:

most

int
e
ractions

are

weak;

and the

few
of

them that

are

str
o
nger

tend

to

be

quite asymme
t
ric,

with,
e
.
g.
,

t
h
e

plan
t

d
epe
n
din
g

hea
v
il
y

o
n

a

pollin
a
t
o
r
o
r
frugi
v
ore spec
i
es

but

the

ani
m
al

b
e
ing

a

super
-
generalist
that

relies

only

marginally

on

that

plant.

If

we

sum the

total
depende
n
ces

that

a

plant

or

an

animal

spec
i
es has

in

a
network, we

get

a

mea
s
ure of

the

spec
i
es

stre
n
gth

(the
quanti
t
at
i
v
e
a
na
l
o
g

t
o

th
e

numbe
r

o
f

interaction
s

per
spec
i
es

in

the

zero

to

one

matrix
)
.

Again,

the

distri
b
ution
of

str
e
ngth

values

is

extreme
l
y

skew
e
d

across

mutual
i
stic
spec
i
es

(Basco
m
pte

et

al.

200
6
),

with

only

a

few

genera
l
ists
concen
t
rating most

of

the

depende
n
cies

for

the

pol
l
ination
or

seed

dispersal services of

the

rest

of

the

comm
u
nit
y
.
Central

species

thus

comb
i
ne

a

high

numb
e
r

of

int
e
ractions
and

a

high

value

of

summed

d
ependen
c
es (strengt
h
)

of
part
n
er

spec
i
es

thus

being

pivotal

for

the

func
t
ioning

of

the
network.

Third,

the

networks show

distinct

modu
l
es

or

comp
a
rt
-

ments,

i.e.,

distinct

subsets

of

species

that

interact more
strongl
y

amon
g

themselve
s
tha
n

w
i
t
h

othe
r

modules.
Olesen

et

al.

(
200
7
)

sho
w
ed

that

most

pol
l
ination

networks
ar
e

modula
r
,
wit
h

distinc
t

subset
s

o
f

p
lant

pollinator
group
s
,

such

as

but
t
erfly
-
po
l
linated

plants

or

those

visited
predo
m
inantly by

hummin
g
birds.

Modu
l
es

are

the

basic
blocks

that

structure

these

network
s
, analo
g
ous

to

the
different walls

that

made

up

a

building.

Individual

spec
i
es
can

have

different

roles

in

this

scena
r
io:

while

some

species
only

inte
ract heavi
l
y

with

spec
i
es

of

the
i
r

own

modules,
other

super
-
generalists


glue


toge
t
her

all

the

modules

by
showing

extr
e
mely generalized

int
e
ractions. Th
e
se

inter
-

actio
n
s

tie

toge
t
her

peripheral

spec
i
es

and

can

be

extr
e
mely
importa
n
t

for

ma
i
ntaining

the

c
ohes
i
veness

o
f

the

network.
For

instance,

inva
s
ive

species

can

be

peri
p
heral in

the
network during

the

early

stages

of

inva
s
ion

but

q
u
ickly
incr
e
ase in

strength

and

get

to

the

core

of

the

network,
displaci
n
g

nat
i
ve

species

(Aizen

et

al.

200
8
).

A
l
l

these

ma
i
n

prope
r
ties

of

the

mu
t
ualistic networks
appear

to

be

omn
i
prese
n
t

in

natu
r
e,

indep
e
ndent

of

the

type
of

interact
i
on or

the

specific

ecosy
s
tem

or

com
m
unity

we
stud
y
.

Whi
l
e

enormous

progr
e
ss

has

been

made

in

recent
y
e
a
r
s

t
o

u
n
d
er
s
t
a
n
d
th
e

b
asi
c

p
a
tte
r
n
s

a
n
d

mo
d
e
s

o
f
interact
i
on

(
B
ascompte

and

Jorda
n
o

200
7
),

the

chal
l
enge
to

und
e
rstand

how

these

mega
d
iver
s
e asse
m
blages

co
-

evolve

rem
a
ins,

i.e.,

how

pairwise

int
e
ractions

add

up

to
modu
l
es

of

tig
h
tly

interact
i
ng

species

to

whole

commun
i
-

ties

as

d
iversified

as

those

we

can

docu
m
ent

in

tropical
fore
s
ts.

Coe
v
olution

of

Mult
i
speci
e
s

Interac
t
ions


What

is

the

basic

proce
s
s

asse
m
bling these

megadiver
s
e
n
e
t
w
o
rk
s
?

U
l
ti
m
a
tel
y
,
t
he

r
o
le

o
f

e
a
c
h

s
p
e
ci
e
s

i
n

t
h
e
network

d
e
pends

on

the

numb
e
r

of

interact
i
ons

it

e
stab
-

lishe
s

wit
h

th
e

p
o
t
e
nti
a
l
p
a
rt
n
e
r

species
.

Est
a
blis
h
ed
interact
i
ons

are

thus,

like

the

basic

b
locks

that

form

larger
m
o
d
u
l
e
s
t
h
a
t

i
n

t
u
r
n

m
a
d
e

u
p

t
h
e

wh
o
l
e

c
om
p
l
e
x
arch
i
tecture

of

the

network.

These

basic

blocks

are

c
alled
interact
i
on

motifs

(Milo

et

al.

200
2
),

or

repea
t
ed

patterns

or
forms

of

interact
i
on,

that

occur

in

the

network

(Fig.

2b,

c
).
Dep
e
nding

on

the

relative frequ
e
ncy

of

these

di
f
ferent
motif
s
,

the

overa
l
l

aspect of

the

whole

net
w
ork

can

be

very
different:

more

modular

and

spec
i
alized

if

built

predo
mi
-

nantly

on

speci
a
lized mo
t
ifs

(e.g.,

Fig.

2e
,

top)

or

more
nest
e
d

and

genera
l
ized

if

nonspe
c
ialized

motifs

are

domi
-

nant

(e.g.,

Fig.

2e
,

bott
o
m).

Data

from

emp
i
rical networks
of

plan
t

pollinator

a
nd

plan
t

frug
i
vore

interact
i
ons

shows
that

local

abundan
c
e

has

perva
s
ive

inf
l
uence on

these
patterns,

but

other

importa
n
t

spec
i
e
s

spec
i
fic

traits

(size,
pheno
l
og
y
,

color)

also

restrict

the

range

of

partn
e
rs

each
spec
i
es

int
e
racts with

(Jor
d
ano

198
7
;

Jorda
n
o

et

al.

200
3
;
Váz
q
uez

e
t

al.

2
0
07
),

dete
r
mi
n
ing the

typ
e

o
f

mo
t
ifs
cont
r
ibuted.

A

future

challe
n
ge

would

be

to

explo
r
e

how
spec
i
e
s

specific

tra
i
ts

mold

the

pattern

of

int
e
raction

and
add

up

to

genera
t
e

these

net
w
ork
-
wi
d
e

patterns.

It

is

far

from

clear

how

coe
v
olved

sele
c
tion

pressures
cont
r
ibute

to

the

emergen
ce of

high
l
y

nest
e
d

pat
t
erns

of
i
n
t
e
r
a
c
t
i
o
n,

g
i
v
e
n

t
h
e

o
m
n
i
p
r
e
s
e
n
ce

of

a
s
y
m
m
e
t
r
y

of

m
u
t
u
al
d
e
p
e
n
d
e
n
c
i
es (
a
n
d
,

p
r
e
s
u
m
a
b
l
y
,

a
s
y
m
m
e
t
r
y

of

p
h
e
n
o
t
y
p
ic
selectio
n

i
n
t
ens
i
ties
;

Jordan
o

1
9
87
;

Bascompt
e

e
t

a
l
.

2
0
0
6
).

It

is

e
x
p
e
c
t
ed t
h
at

t
h
e

s
e
l
e
c
ti
on p
r
e
s
s
u
r
e
s

o
r
i
g
i
n
a
t
i
n
g
f
r
om

p
a
i
r
w
i
se

i
n
t
e
r
a
c
t
i
o
ns s
h
o
u
ld

be

m
o
re

s
y
m
m
e
t
r
i
c

in
a
nt
a
g
o
n
i
s
t
i
c

i
n
t
e
r
a
c
t
i
o
ns

or in

m
u
t
u
a
l
i
s
ms

w
i
th

a

h
i
gh

d
e
g
r
ee
of

i
n
t
i
m
a
cy

a
n
d

s
p
e
c
i
a
l
i
z
a
t
i
on (
l
i
k
e

a
n
t
-
p
l
a
n
t
s
;

F
i
g.

2
c
;
G
u
i
m
a
r
ã
e
s

et

a
l
.

2
0
0
7
),

r
e
s
u
l
t
i
n
g

in

m
o
re

m
o
d
u
l
a
r

n
e
t
w
o
r
ks
w
i
th

d
i
s
t
i
n
ct

g
r
o
u
ps

of

c
o
e
v
o
l
v
i
ng s
p
e
c
i
e
s.

T
h
e

a
s
y
m
m
e
t
r
ic
p
at
t
e
rn of

i
n
t
e
r
a
c
t
i
o
n

t
h
a
t

p
e
r
v
a
d
es

m
u
t
u
a
l
i
s
t
i
c

n
e
t
w
o
r
ks
of

f
r
e
e
-
l
i
v
i
n
g

s
p
e
c
i
e
s

f
a
v
o
rs

t
h
e

d
i
v
e
r
s
i
f
i
c
a
t
i
o
n

a
n
d

g
r
o
w
th
of

t
he

w
e
b

by

a
d
d
i
ng

n
e
w

s
p
e
c
i
e
s t
h
at

l
i
nk

w
i
th

t
h
e

c
o
re

of
sup
e
r
-
g
e
n
e
r
a
lists
.
Fo
r

i
nsta
n
ce
,

r
a
r
e

p
l
a
n
t

s
p
e
cie
s

can
p
r
o
b
a
b
ly

p
e
r
si
st

a
n
d

h
a
ve

a

f
u
n
c
t
i
o
n
al s
e
r
v
i
c
e

of

p
o
l
l
en
t
r
a
n
s
f
er

or

s
e
ed

d
i
s
p
e
r
s
a
l

by

d
e
p
e
n
d
i
n
g

s
t
r
o
n
g
ly

on

g
e
n
e
r
a
l
i
st
a
ni
m
a
l

p
a
r
t
n
e
rs t
h
a
t
,

in

t
u
r
n
,

o
n
ly

m
a
r
g
i
n
a
l
l
y r
e
ly

on

t
h
e
p
l
a
n
t

r
e
s
o
u
r
c
e
s.

W
e

might

expect

a

vari
e
ty

of

influences

of

the

plant
s


an
d

a
n
i
mals


evolutionar
y
histo
r
y

i
n

sh
a
pin
g

networ
k
patterns

(Fig.

3b
).

As

new

spec
i
es
a
dd

up

in

the

network,
the

overa
l
l

leve
l
s

of

pheno
t
ypic

convergence

and

comple
-

ment
a
rity would

increase.

For

instance,

new

frugivores
would

tend

to

c
onverge

(be

more

simi
l
ar

in

morpholo
g
y)
with

preex
i
sting frug
i
vore

species

and

share

codispersed
pl
a
nts;

in

t
u
rn,

s
e
lec
t
i
o
n

pr
e
ssur
e
s

wou
l
d

incr
e
ase

the



Evo

Edu

Outreach

(2010)

3:4
0

46

45




pheno
t
ypic

ma
t
ching

of

ani
m
al

traits

(body

part
s
,

pheno
l
-

ogy)

and

fruit

traits,

inc
r
e
asing

complem
e
ntar
i
ty

between
interact
i
ng

partners.

Convergence

wou
l
d

tend

to

facilitate
the

persiste
n
ce of

a

given

spec
i
es

with
i
n

a

multi
s
pecific
mutu
a
lism;

comp
l
ementar
i
ty

would

tend

to

facili
t
ate

how
the

species

efficiently

uses

the

mutual
i
stic

servic
es

prov
i
ded
by

the

part
n
ers.

The

interaction

pattern

will

then

reflect

the
phylo
g
enies

of

the

two

groups

of

species

(Fig.

3b
,

top)

and
mar
k
edly

deviate

from

the

ch
e
ckerspot pattern

(Fig.

3b
,
bottom)

of

interactions

e
xpected

in

the

absence

of

phylo
-

genet
i
c

sign
a
l.

Sign
i
ficant inf
l
uences

of

the

evolutio
n
ary
history of

only

one

of

the

spec
i
es groups

would

mean

that
the

group

has

driven

the

evolution

of

the

net
w
ork

(Fig.

3b
,
middle):

clos
e
ly

related

plan
t
s

sho
w
ing

a

trend

to

interact
with

similar sets

of

frugi
vore

species,

but

these

animal
spec
i
es

being

n
o
t

phylo
g
enetical
l
y

related.

A

given

plant
would

be

using

the

dispersal

serv
i
ces

of

a

wide

array

of
frugi
v
ore clade
s
,

but

a

g
iven

frug
i
vore

species

would

tend
to

e
x
p
l
o
i
t

a

s
u
b
s
e
t

o
f

p
h
y
l
o
g
e
ne
t
i
c
a
l
ly r
es
t
r
i
c
t
e
d

f
r
u
it
spec
i
es.

When

convergence

and

increased

complem
e
ntarity

re
-

main

res
t
ricted

to

distinct

subse
t
s

of

spec
i
es,

then

modu
-

larity

will

incre
a
se

thro
u
gh

a

disp
r
oportiona
t
e

gro
w
th

of
s
p
e
ci
a
l
i
z
e
d m
ot
i
f
s,

c
r
e
a
t
i
n
g

vo
r
t
e
xe
s

o
f

coe
v
o
l
u
t
i
o
na
r
y
change

(
T
hompson

200
5
,

200
6
).

W
e

can

expect

these
tren
d
s

for

h
ighly

intim
a
te

mutual
i
stic assoc
i
ations

such

as
a
n
t
-
p
l
a
n
t
s
,

s
y
m
b
i
o
s
e
s,

an
d

h
i
g
h
-
s
p
e
c
i
f
i
c
i
t
y
a
n
t
a
g
on
i
s
ms
s
u
c
h

a
s

h
ost

parasit
e
interaction
s
.

I
n

contrast
,

super
genera
l
ists are

expec
t
ed

to

evolve

and

coevo
l
ve

w
i
thin
mega
d
iver
s
ified webs

of

int
e
ractions

among

free
-
liv
i
ng
spec
i
e
s

m
a
i
n
l
y

b
y

e
volvin
g

a
b
i
li
t
ie
s
t
o

interac
t

w
i
t
h
multi
p
le,

distinct

groups

of

part
n
ers.

A

charac
t
eristic

pattern
in

nested

netw
o
rks

of

mutual
i
sts

(Olesen

et

al.

200
7
) is

that
the

super
-
g
enerali
s
ts

are

true

hubs

in

the

netw
o
rk,

adding
interact
i
ons

that

c
o
nnect

different

modules.

T
o

some

e
xtent,
the

evolution

of

the

supergeneralist lifestyle

all
o
ws

the
gluing

toge
t
her of

the

dive
r
se

bricks

and

blocks

that

make
up

the

fasc
i
nating

arch
i
tectur
e

of

these

ecological

services
and

the
i
r

biod
i
versit
y
.



Conc
l
uding

Remar
k
s


The

recent

dev
e
lopment

of

network
-
based

too
l
s

applied

to
the

study

of

complex

patterns

of

ecological int
e
ractions
brid
g
es

mu
l
tidiscipl
i
nary

approa
c
hes

from

statistical

me
-

chani
c
s

in

phys
i
cs,

bioco
m
plexit
y
,

ecolo
g
ical

modelin
g
,
and

basic

natu
r
al

histor
y
.

It

is

probab
l
y the

only

approa
c
h
that

can

succe
s
sfully decip
h
er

the

si
m
ple,

general

patterns
that

lie

behind

the

e
x
treme

complexi
t
y

of

interact
i
on

webs
in

ecosy
s
tems.

A

fascina
t
i
ng

aspect

of

these

webs

is

their
simi
l
arity

and analo
g
y

to

other

complex

network
s
,

spann
i
ng
biotic

(e.g.,

gene

regu
l
ation,

cell

metabol
i
c

reactions)

and

abiotic

(e.g.,

the

internet)

scena
r
ios.

The

multidisc
i
plinary
integrat
i
ve

approa
c
h

to

the

study

of

c
omp
l
ex

net
w
orks

can
be

a

key

to

d
e
veloping e
a
rly
-
warning

diagnost
i
c

criteria

to
identify

cri
t
ical

si
t
uations

of

disturban
c
e in

n
atural

areas
wel
l

befor
e

th
e

f
u
n
ctiona
l
a
s
p
ect
s

o
f

k
e
y

ecos
y
s
tem
servic
e
s,

like ani
m
al
-
medi
a
ted

pollina
t
ion

or

seed disp
e
rsal,
r
each

a

no
-
re
t
urn

point

for

their

succe
s
sful

restoration. On
to
p

o
f

thes
e

app
l
ie
d

object
i
v
e
s
,
researc
h

o
n

c
omplex
ecolo
g
ical networks

has

taken

the

first

steps

to

a

fuller
u
n
d
e
rstan
d
in
g
o
f

h
o
w

co
e
v
o
l
u
ti
o
n

driv
e
s

me
g
adi
v
erse
asse
m
blages

of

mu
t
ualistic

spec
i
e
s,

which

are

the

backbone
of

ecosy
s
tems

like

the

tropi
c
al

rain
f
orest.

Und
e
rstanding

comp
l
ex

coevol
v
ing

netw
o
rks

is

impor
-

tant

because

spec
i
es

and

the
i
r

interact
i
ons

do

not

exist

in

an
e
c
olo
g
i
c
a
l
va
c
uum
.

B
y

u
sin
g

ne
w

m
ulti
d
isciplin
a
r
y
approa
c
hes, we

aim

to

better

foreca
s
t

the

risks

of

losing

a
single

species

o
r

collapsi
n
g

a

sing
l
e

int
e
raction to

the
persiste
n
ce

of

the

whole

ec
o
system.

W
e

know

that

these
comp
l
ex

systems

are

more

than

the

sum

o
f

their

part
s
,

so
the

conseq
u
ences

of

losing

one

of

the

parts

may

extend

well
beyond

its

immed
i
ate

inf
l
uence.

Th
u
s,

this

understand
i
ng
will

help

us

to

be

b
e
tter

prepa
r
ed

to

effectively

restore

the
key

functions and

serv
i
ces
n
eeded

to

rebuild

disturbed
ecosy
s
tems.

W
e

n
eed

a

solid

scientific

theory

of

conse
r
va
-

tion

w
i
th

a

kno
w
ledge

a
nd

und
e
rst
a
nding of

c
omp
l
ex
patterns

of

biodivers
i
ty

that

at

first

sight

appear

im
p
ossible
to

handle

a
nd

analy
z
e

(the


e
n
tangled

ban
k

).

Coev
o
lving
network
s

o
f

mu
l
tispecie
s
interaction
s

underpi
n

this
entan
g
led

bank,

and

we

are

just

starting

to

grasp

the

fine
details

of

their

coevo
l
ution.


Acknowledgments

I

appreciate

the

invitation

by

John

N.

Thompson
and

Rodrigo

Medel

to

contribute to

this

special

issue

on

coevolution.
Their

advice

and

comments,

together

with

one

anonymous

referee,
helped

to

improve a

draft

of

this

pape
r
.

Over

the

years,

my

ideas

on
networks

have

benefited

from

discussions

and

collaborations with
Jordi

Bascompte,

Jens

M.

Olesen,

Paulo

R.

Guimarães

J
r
.,

John

N.
Thompson,

Thomas

Lewinsohn,

and

people

at

the Integrative

Ecology
Group

(Sevilla),

especially

Alfredo

V
alido,

Carlos

Melián,

Miguel

A.
Fortuna,

and

Jofre

Carnice
r
. My

work

has

been

supported by

grants
from

the

Spanish

Ministry

of

Science

(MICINN;

CGL2006
-
00373)
and

Junta

de

Andalucía

(P07
-
RNM2824).

This

paper

is

dedicate
d

to
Myriam

Márquez,

for her

birthday

and

the

suggestions

to

design
Figs.

1

and

3
.




References


Aizen

M,

Morales

C,

Morales

J.

Invasive

mutualists

erode

native
pollination

webs.

PLoS

Biol.

2008;6:e31.

Albert R,

Jeong

H,

Barabási

A.

Error

and

attack
tolerance

of

complex

networks.

Nature.

2000;406:37
8

82.

Amaral

L,

Scala

A,

Barthelemy

M,

Stanley

H.

Classes

of

small
-
world
networks.

Proc

Natl

Acad

Sci

USA.

2000;97:
11
149

52.

Bas
c
omp
t
e

J,

Jordano

P
.

Pla
n
t

ani
m
al

m
u
tua
l
ist
i
c

netwo
r
ks:

the

architecture

of

biodiversit
y
.

Ann

Rev

Ecol

Evol

Syst.

2007;38:

567

93.



46

Evo

Edu

Outreach

(2010)

3:4
0

46




Bascompte

J,

Melián

C.

Simple

trophic

modules

for

complex

food
webs.

Ecolog
y
.

2005;86:286
8

73.

Bascompte

J,

Jordano

P
,

Melián

C,

Olesen

J.

The

nested

assembly of
plant

animal

mutualistic

network
s.

Proc

Natl

Acad

Sci

USA.

2003;100:9383

7.

Bascompt
e

J
,

J
ordan
o

P
,

Olese
n

J
.

As
y
m
m
etri
c
coe
v
olutio
n
a
ry
networks

facilitate

biodiversity maintenance.

Science.

2006;312:

431

3.

Bastolla

U,

Fortuna

MA,

Pascual
-
García A,

Ferrera

A,

Luque

B,
Bascompte

J.

The
a
rchitecture

of

mutualistic

networks

minimizes
competition

and

increases

biodiversit
y
.

Nature.

2009;458:1018


20.

Bawa

KS.

Plant

pollinator interactions

in

tropical

rain

forests.

Ann

Rev

Ecol

Evol

Syst.

1990;21:399

422.

Burgos

E,

Ceva

H,

Perazzo

RPJ,

Devoto

M,

Medan

D,

Zim
m
ermann

M,

et

al.

Why

nestedness

in

mutualistic

networks?

J

Theor

Biol.

2007;249:30
7

13.

Dunne

J,

W
illiams

R,

Martinez

N.

Food
-
web

structure

and

network
theory:

the

role

of

connectance and

size.

Proc

Natl

Acad

Sci
USA.

2002;99:12917

22.

Gal
etti

M,

Aleixo

A.

Effects

of

palm

heart

harvesting

on

avian

frugivores

in the

Atlantic

rain

forest

of

Brazil.

J

Appl

Ecol.

1998;35:286

93.

Guimarães

Jr

PR,

Rico
-
Gray

V
,

Oliveira

PS,

Izzo

TJ,

dos

Reis

S
F
,
Thompson

JN.

Interaction intimacy

aff
e
ct
s

structure

and

coevo
-

lutionary

dynamics

in

mutualistic

networks.

Curr

Biol.

2007;17:

1797

803.

Jackson

JBC

et

al.

Historical

overfishing

and

the

recent

collapse

of
coastal

ecosystems.

Science.

2001;293:62
9

38.

Jordano

P
.

Patter
n
s

of

mutualistic

interacti
ons in

pollination and

seed
d
i
s
p
e
r
s
al
:

c
o
n
n
e
c
t
a
nc
e,

d
e
p
e
n
d
e
n
ce

a
s
y
m
m
e
t
r
ie
s,

a
n
d

c
o
e
v
o
l
u
t
i
o
n.
A
m

N
a
t.

1
9
8
7
;
1
2
9
:
6
5
7

7
7
.

Jordano

P
.

Fruits

and

frugivor
y
. In:

Fenner

M,

edito
r
.

Seeds:

the
ecology

of

regeneration

in

plant

comm
u
nities.

W
allingford,

UK:
CAB

Inter
national;

2000.

pp.

125

66.

Jordano

P
,

Bascompte J,

Olesen

JM.

Invariant

properties

in

coevolu
-

tio
na
r
y

networ
k
s

o
f

pl
an
t

anima
l

i
nter
a
ctions
.

Eco
l

Lett.

2003;6:6
9

81.

Milo

R,

Shen
-
Orr

S,

Itzkovitz

S,

Kashtan

N,

Chklovskii

D,

Alon

U.

Network

motifs:

simple

building

blocks

of

complex

networks.
Science.

2002;298:82
4

7.

O
l
es
e
n

J,

B
a
s
c
om
p
te

J,

D
u
p
o
nt

Y
,

J
o
rd
a
no

P
.

T
he

m
o
d
u
l
a
r
i
t
y

o
f
p
o
l
l
i
n
a
t
i
on

n
e
t
w
o
r
k
s
.

P
r
oc

N
a
tl

A
c
a
d

S
c
i

U
SA
.

2
0
0
7
;
1
0
4
:
1
9
8
9
1

6.
Pimm

SL,

Lawton

JH,

Cohen

JE.

Food

web

patte
rns

and

their

consequences.

Nature.

1991;350:66
9

74.

Rezende

EL,

Lavabre

JE,

Guimarães

Jr

PR,

Jordano

P
,

Bascompte

J.

Non
-
random

coextinctions

in

phyloge
n
etically

structured

mutu
-

alistic

networks.

Nature.

2007;448:92
5

8.

S
i
l
v
a

W
R,

G
u
i
m
a
r
ã
e
s
J
r

P
R
,

D
o
s

R
e
is

S
F
,

G
u
i
m
a
r
ã
e
s

P
.

I
n
v
e
s
ti
g
a
t
i
ng
f
r
a
g
i
l
i
ty

i
n

p
l
a
n
t

fr
u
g
i
v
o
r
e

n
e
t
w
o
r
k
s
:

a

c
a
se

s
t
u
d
y

of

t
h
e

A
t
la
n
t
ic
f
o
re
st

i
n

B
ra
z
il
.

I
n:

D
e
n
n
is

A
J,

S
c
h
u
p
p

E
W
,

G
r
e
en

R
J
,

W
e
st
c
o
tt

D
A
,
e
d
i
t
o
r
s.

S
ee
d

d
i
s
p
e
r
s
a
l
:

t
h
e
o
r
y

a
nd

i
t
s

a
p
p
li
c
a
t
i
on

i
n

a

c
h
a
n
g
i
ng
w
o
r
l
d
.

W
a
l
li
n
g
f
o
r
d,

U
K
:

C
A
B

I
nt
e
r
n
a
t
i
o
n
a
l
;

2
0
0
5
.

p.

5
6
1

7
8
.

Sprengel

CK.

Das

entdeckte

Geheimniss

der

Natur im

Bau

und

in

der
Berfruchtung

der

Blumen

I.

Berlin,

Germany:

V
ieweg

sen;

1793.
Srini
v
asan

U
T
,

Dunne

JA,

Harte

J,

Martin
e
z

ND.

Resp
o
nse

of
complex

food

webs

to

realistic

extinction

sequences.

Ecolog
y
.

2007;88:671

82.

Thompson

JN.

Interaction

and

coevolution.

New

Y
ork,

US:

John

W
iley

and

Sons;

1982.

Thompson

JN.

The

geographic

mosaic

of

coevolution.

Chicago,

USA:
University of

Chicago

Press;

2005.

Thompson

J.

Mutualistic webs

of

species.

Science.

2006;312:37
2

3.

Vázquez

D
P
,

Aizen

MA.

Null

model

analyses

of

specialization in
plant

pollinator

interactions.

Ecolog
y
.

2003;84:249
3

501.

Vázquez

D
P
,

Melián

CJ,

W
illiams

NM,

Bluethgen

N,

Krasnov

BR,
Poulin

R.

Species

abundance

and

asymmetric

interaction

strength
in

ecological networks.

Oikos.

2007;
1
16:
1
120

7.

W
eitz

JS,

Benfey

PN,

W
ingreen

NS.

Evolution,

interactions, and
biological networks.

PLoS

Biol.

2007;5:e
1
1.