Transcript Conflict & cooperation

The evolution of conflict
and cooperation
Lecture in the population biology
and population genetics seminar
series
Tom Wenseleers, 2001
Major transitions in evolution
Szathmary &
Maynard Smith
Gene
Genome
Prokaryotes
Eukaryotes
Genomes
Unicellularity
Genome-alliances
(diploidy, sex)
Multicellularity
Individual organisms
Societies
Why cooperate?
Two approaches :
- game theory: cooperation is only a good
strategy when it has mutual or delayed
benefits (false altruism)
- kin selection: cooperation also possible
when it has personal costs, but only when
interactants are genetically related (true
altruism)
Delayed benefits:
helpers at the nest
Young individuals stay at home and
help their parents raise more
offspring, rather than breeding
themselves. Many birds (including
these fairy wrens) do this.
Or, in group-living animals
sometimes help raise other group
members’ offspring (ostriches, some
primates).
Altruism? Helpers may gain useful experience in raising their own
offspring; or they may have hopes of inheriting a valuable breeding
territory.
Mutual (synergistic) benefits
Wolves hunt in
packs and then
share their prey.
Is this altruism?
All the wolves get a benefit from pack hunting: they can bring
down larger prey. Cooperation is increasing each wolf’s own
fitness.
Reciprocal altruism
(Trivers)
E.g. blood sharing in vampire
bats.
Sharing blood does have costs to
the donor.
But they may hope to get
something back: the next night,
they might miss out and the
neighbor they fed will feed them
in return (future benefits).
For reciprocal altruism, you need
individual recognition and enough
memory of past encounters to
eliminate freeloaders.
Game theory
Von Neumann & Morgenstern 1944
Theory of Games and Economic Behaviour
Game theory
•
Optimal (rational) behaviour in conflict
situations?
•
“players”
•
may each choose a “strategy”
•
(genes, individuals, groups)
Each pairs of strategies is associated with a
“payoff”
Prisoner’s dilemma
“PAYOFFS”
player 1
defect
cooperate
defect
cooperate
player 2
cooperate
cooperate
defect
defect
consequence
for player 1
B
B-C
0
-C
DOVE
Hawk-dove game
Maynard Smith
& Price 1973
DOVE
HAWK
0
-B
B
-C
Hawk-dove game
- SYNERGY
Fitness player 1 w1 =B.z1-B.z2-C.z1.z2
z1 en z2= phenotypes of players 1 & 2
(hawk=1, dove=0)
Advantage of playing hawk depends on what the other
player does: benefit = B-C.z2
At equilibrium B-C.z2=0, and the ESS is to play hawk
with a probability of z*=B/C
Hawk-dove game
Limited degree of cooperation is in this case not
altruistic!
It avoids mutual destruction!
Cf. “Mutually Assured Destruction” (MAD) in
cold war
True altruism
Definition :
Reproductive altruism: An individual
behaves in such a way as to enhance the
reproduction of another individual, at a
cost to its own fitness.
Paradox: how can natural selection ever
favour such behaviour (Darwin) ?
True altruism
Sterile workers in social
insects: give up all
reproduction for the benefit
of their mother queen.
How can such behaviour be
selected?
Mutual or delayed benefits can’t
account for this one: sterile workers
never get to produce any daughters.
Group selection
(Wynne-Edwards)
W-E proposed that individuals in group-living
species might altruistically restrict their
reproduction to avoid overpopulation and starvation.
The behavior would be favored because groups
containing such individuals would survive, while
groups without them would starve and go extinct.
In general, an altruist that promotes reproduction of
its groupmates might be favored.
But there’s a problem with group selection.
A
Group with altruists, busily outcompeting all
the other groups.
A
Z Z Z
A Z A
The selfish individuals in the group are
getting the benefit but paying no cost. In
the next generation they’ve increased
within the group.
Z
Z
A
Z
Z Z
Z
A
Z
Z
Z Z Z
Z
Z
Z
And now altruists are
extinct even though
they’ve helped the
group.
An important distinction
DNA
DNA/Gene: the
“Replicator” that
actually gets copied
in reproduction.
Organism
Organism: the “Vehicle”, a
machine built by the DNA
to do the copying.
Replicators & Vehicles
(Dawkins)
Replicators that get copied a lot become more
common, replacing those that get copied less:
that’s just what selection is.
Traits in the vehicles are favored by selection if
they help the replicators that code for them get
copied.
In other words, you always need to look at gene
frequency change, not at ecological success.
W.D. Hamilton (1936-2000)
Kin selection
Hamilton’s Rule (1964)
Relatedness to partner
r.B > C
Benefit to partner
Personal cost
This rule predicts when a gene for altruism should be selected.
Prediction: cooperation at high relatedness, conflict at low relatedness.
Inclusive fitness
Hamilton’s rule leads us to the idea of inclusive
fitness :
Fitness is not only based on own reproduction but
also depends on the effects on other individuals,
weighted by relatedness.
Inclusive fitness = direct fitness (own reproduction) +
indirect fitness (reproduction of others) x relatedness
Empirical tests
•
•
Reproductive conflicts in insect societies
- Sex-ratio conflicts
- Conflicts over male production
- Conflict over caste fate
Parent-offspring conflict (Trivers)
Sex-ratio theory
Fisher
Trivers & Hare
Trait that has a + effect on the production of females
(F) and a - effect on the production of males (M) :
EF.rF > EM.rM
(Hamilton’s rule)
EF = mating success of females ~ M
EM = mating success of males ~ F
ESS F:M sex-ratio = F/M* = rF / rM
Social insect colonies
X
C
AB
AC
0,5
0,5
0,75
AC
BC
0,25
A, B
Relatedness coefficients in an ant colony.
BC
Worker
generation
Calculating relatedness
X
C
1
AB
Relatedness
between sisters?
Share genes via
father with a
chance of
1 x 0.5
0,5
AC
BC
AC
BC
Worker
generation
Calculating relatedness
X
C
AB
0.5
Relatedness
between sisters?
Share genes via
mother with a
chance of
0.5 x 0.5
0,5
AC
BC
AC
BC
Worker
generation
Calculating relatedness
X
C
1
AB
0.5
Sisters share
genes via father
OR mother, so
average chance is
1 x 0.5 + 0.5 x 0.5 =
0.75
0,5
AC
BC
0,5
0,75
AC
BC
Worker
generation
Sex-ratio conflicts
Trivers & Hare
•
Mother queen: equally related to sons and
daughters (rF=0.5, rM=0.5)
 Wants to invest equally in both sexes.
•
Workers: 3 x more related to sisters than to
brothers (rF=0.75, rM=0.25)
 Prefer 3:1 F:M sex-ratio
•
Parent-offspring conflict !
Fratricide in
ants
Often have female biased
sex ratios. Indicates that
sex allocation is
controlled by the
workers.
Except in slave-making
ants: slaves have no
genetic stake in the
slave-makers sex-ratio.
Wood ant Formica exsecta : faculatative
sex-ratio biasing. Some colonies with
single mated queen, others with double
mated queen. Workers only eat their
brothers in colonies headed by a single
mated queen. (Sundström)
Conflicts over male
production
•
•
•
Workers can also produce own sons
rw-son=0.5 > rw-brother=0.25  worker reproduction
But: rQ-son=0.5 > rQ-grandson=0.25
 ’queen policing’
At mating frequencies > 2 a worker is less related
to an average worker produced male than to a
brother
’worker policing’ (Ratnieks)
Calculating relatedness
Single mating
Relatedness WQ produced male =
0.25
X
0,5
0,5
0,25
Relatedness WW produced male =
0.375
0,75
0,5
0,375
0,5
 no worker
policing
Calculating relatedness
1
Treble
mating
2
X
Relatedness WQ produced male =
0.25
3
0,5
0,5
1
0,25
2
0,25
3
0,125
Relatedness WW produced male =
(1/3) x 0.375
+ (2/3) x 0.125
= 0.21
0,375
1
2
3
 worker policing
Empirical evidence
Worker reproduction in monandric species
(stingless bees, bumble bees, some wasps).
Worker policing in honey bees
(polyandrous, mating with 10-15 males).
Worker policing in honey
bees (Ratnieks) 
Empirical evidence
Facultative worker policing in
Dolichovespula saxonica :
workers only police in
polyandrous nests.
(Foster & Ratnieks)
Conflict over caste fate
Conflict over caste fate
Bourke & Ratnieks 1999
Stingless bees
Colonies are swarm founded and
therefore mainly need workers, just a
few queens.
But : 20% of all females develop as
queens. A clear excess!
Conflict over caste fate
Wenseleers et al. 2002
Explanation: each larva is more related to own
offspring than to sisters’ offspring
 larva prefers to become a queen herself
overproduction of queens if self determination is
possible
Colony doesn’t need so many queens  mother
queen and adult workers are selected to prevent
excess queen production (‘policing’)
Melipona bees
0.25
HIGH
RELATEDNESS
MALES
QUEEN PRODUCED
SOME WORKER PRODUCED
GLZ, p < 0.0006
0.2
PREDICTED
ESS
0.15
0.1
0.05
ua
dr
i fa
sc
.
.f
M
.q
M
.b
M
av
os
a
0
ee
ch
ei
i
Prop. of queens produced
LOW
RELATEDNESS
(data are from months with maximum queen production)
Policing of caste fate
stingless bees
Self determination
20% queen production
honey bees
Social determination
0.005% queen production
In the ’70 Bob Trivers showed that
there are also conflict of interests
in the seemingly solid parentoffspring bond.
Parent-offspring conflict
Each offspring would like to favour itself over its
siblings (r=0.5)
Parent on the other hand would prefer to treat all
offspring equally (equally related).
Offspring are selected to be more selfish than
their parents should be willing to tolerate!
E.g. intra-uterine conflicts
Major transitions in evolution
All the previous also applies at other levels
E.g. conflicts between genes within cells
or
between cells within multicellular organisms
“intragenomic conflicts”
Intragenomic conflicts
Forms of intragenomic conflict :
- between genes on homologous chromosomes over
transmission to gametes (meiotic drive)
- nucleo-cytoplasmic conflicts over optimal sex
allocation
- conflicts between cells over who ends up producing
the gametes
Meiotic drive cf. hawk-dove game
DRIVE
0
-B
B
-C
COOPERATE
COOPERATE
But there are differences
Genes
Organisms
Option to breed
independently
no
usually
Fighting strategy
poisoning
physical aggression
Type of ESS
pure
mixed
Nucleo-cytoplasmatic
conflict
•
•
Nuclear genes
rF=0.5 , rM=0.5
Cytoplasmic genes (mitochondria, some
bacterial symbionts)
rF=1, rM=0
Enhance their own transmission if they manipulate their
host to produce a more female biased sex ratio. Males are
a dead end.
Male killing
Selective killing of males.
Increases the survival of
sisters in the same brood,
who carry copies of the
maternal element.
Works through kin selection,
cf. fratricide.
E.g. Ricketssia, Wolbachia and
Spiroplasma in ladybird beetle
Cytoplasmic
male sterility
(CMS)
In approx. 4% of all
hermaphrodite plants.
Mitochondrial gene that
benefits the female
function by sterilising the
male function.
Nuclear genes are selected
to suppress CMS.
Feminisation
Feminisation of genetic
males.
Presumably works by
suppressing the androgenic
gland.
Occurs in woodlice
(Wolbachia).
Induction of
parthenogenesis
Induction of asexual
reproduction, resulting in
an all-female brood.
Occurs in some parasitoid
wasps (Wolbachia).
“Maternal sex-ratio”
Manipulates her host (Nasonia)
to fertilise more eggs than she is
selected to.
Nasonia is haplodiploid, so
fertilised eggs develop as
females.
Exact nature of “maternal sex
ratio” is as yet unknown.
Evolution of multicellularity
Slime molds
Dilemma cf. caste conflict
EACH LARVA WANTS TO
BECOME A NEW QUEEN
?
EACH CELL WANTS
TO BECOME A SPORE
SPORE
?
SOMA CELL
An experiment
1 clone
HIGH r
DeAngelo et al. 1990
>1 clone
LOW r
Strassmann et al. 2001
Green beard genes: ultimate
selfish genes
Bb
Bb
Bb
Bb
BB
Bb
Gp-9 allozyme locus
Keller & Ross 1998
The End