SlideShare a Scribd company logo
1 of 55
Download to read offline
Fitness for the impatient
Fran¸cois Rousset
May 2013
Fran¸cois Rousset Fitness for the impatient May 2013 1 / 33
The message
To understand the forces operating in social evolution, particularly in
spatially structured populations:
Fran¸cois Rousset Fitness for the impatient May 2013 2 / 33
The message
To understand the forces operating in social evolution, particularly in
spatially structured populations:
Relatedness concepts under localized dispersal
Fran¸cois Rousset Fitness for the impatient May 2013 2 / 33
The message
To understand the forces operating in social evolution, particularly in
spatially structured populations:
Relatedness concepts under localized dispersal
Stability of kin recognition polymorphisms
Relationship between inclusive fitness and evolutionary stability
Fran¸cois Rousset Fitness for the impatient May 2013 2 / 33
The message
To understand the forces operating in social evolution, particularly in
spatially structured populations:
Relatedness concepts under localized dispersal
Stability of kin recognition polymorphisms
Relationship between inclusive fitness and evolutionary stability
A fundamental impatience: reduce these problems to trivial building blocks
Fran¸cois Rousset Fitness for the impatient May 2013 2 / 33
The message
To understand the forces operating in social evolution, particularly in
spatially structured populations:
Relatedness concepts under localized dispersal
Stability of kin recognition polymorphisms
Relationship between inclusive fitness and evolutionary stability
A fundamental impatience: reduce these problems to trivial building blocks
Draw connections to other methods such as diffusion theory, multilocus
methods
Fran¸cois Rousset Fitness for the impatient May 2013 2 / 33
Continuous evolutionary stability for the very impatient
Two alleles, a and A, inducing phenotypes za and zA
Fran¸cois Rousset Fitness for the impatient May 2013 3 / 33
Continuous evolutionary stability for the very impatient
Two alleles, a and A, inducing phenotypes za and zA
FitnessofAallele
Phenotypes
za, zA
1
a
z1 zm z2
Fran¸cois Rousset Fitness for the impatient May 2013 3 / 33
Continuous evolutionary stability for the very impatient
Two alleles, a and A, inducing phenotypes za and zA
FitnessofAallele
Phenotypes
za, zA
1
a
z1 zm z2
Fran¸cois Rousset Fitness for the impatient May 2013 3 / 33
Continuous evolutionary stability for the very impatient
Classification of different cases:
Continuously stable
strategy CSS
Branching point
Evolutionarily stable
noninvasible
Invasible
Unattainable
Convergence stable
attainable
Fran¸cois Rousset Fitness for the impatient May 2013 4 / 33
Eshel 1983,1996; Christiansen, 1991; Abrams et al., 1993
Implicit assumptions
Resident phenotype z, mutant z + δ
∆p ∼ δstuff + δ2
blob (1)
assumed to be essentially of the form
∆p ∼ δp(1 − p)S(z) + δ2
p(1 − p)(1 − 2p)blob(z) (2)
where S(z) and blob(z) are of constant sign wrt p.
Fran¸cois Rousset Fitness for the impatient May 2013 5 / 33
Implicit assumptions
Resident phenotype z, mutant z + δ
∆p ∼ δstuff + δ2
blob (1)
assumed to be essentially of the form
∆p ∼ δp(1 − p)S(z) + δ2
p(1 − p)(1 − 2p)blob(z) (2)
where S(z) and blob(z) are of constant sign wrt p.
Actually S(z) is independent from p in many models (strong claim!).
Why?
Fran¸cois Rousset Fitness for the impatient May 2013 5 / 33
A study of frequency (p) dependence
The first-order term
A conceptual device
Fitness
The minimal algorithm
Two views of the Prisoner’s dilemma
Many views of inclusive fitness
The more general logic: dominance, kin recognition
Kin recognition
Glimpses of second-order results
Continuous evolutionary stability
Kin recognition
Fran¸cois Rousset Fitness for the impatient May 2013 6 / 33
A conceptual device
p =
parents i
Ai Xi
X indicator variable for an allele;
A frequency of copies of parental genes
Consider the function f giving (conditional) probabilities
E[p | . . .] =
i
ai (. . .)Xi
where ai is the probability that a descendant copy originates from parental
copy i.
Fran¸cois Rousset Fitness for the impatient May 2013 7 / 33
A conceptual device
p =
gene copies g
Ag Xg
X indicator variable for an allele;
A frequency of copies of parental genes
Consider the function f giving (conditional) probabilities
E[p | . . .] =
i
ai (. . .)Xi
where ai is the probability that a descendant copy originates from parental
copy i.
Fran¸cois Rousset Fitness for the impatient May 2013 7 / 33
Trivial example
Individuals share a resource in total amount R, their fecundity being equal
to the amount of resource they consume
Their share of resource is proportional to the value of some trait z:
sharei =
zi
i zi
= ai , fecundityi = R
zi
i zi
Fran¸cois Rousset Fitness for the impatient May 2013 8 / 33
Trivial example
Individuals share a resource in total amount R, their fecundity being equal
to the amount of resource they consume
Their share of resource is proportional to the value of some trait z:
sharei =
zi
i zi
= ai , fecundityi = R
zi
i zi
Conflict between individual and group
fecundityi = (R − ¯z)
zi
i zi
but ai is still zi
i zi
.
Fran¸cois Rousset Fitness for the impatient May 2013 8 / 33
Fitness
In terms of the number of adult offspring, or“fitness”W
p =
gene copies g Xg Wg
g Wg
= mean(Xg Wg )
Define fitness functions so that
E[p |p] =
1
Ntot g
Xg w(zg (p))
In a demographically stable population Ntota = w (e.g., w = zi
¯z ).
A (minimal: ¯W = 1) Price equation
p = ¯X = ¯W ¯X + Cov(wg , Xg ),
E[∆p|z] = Cov(wg (z), Xg ).
Fran¸cois Rousset Fitness for the impatient May 2013 9 / 33
Minimal algorithm
Express fitness in terms of indicator variables
e.g., phenotype zi = z + Xi δ
Differentiate with respect to some measure of strength of selection
To first order in δ,
wf(zf, zp) ∼ linear combination of (X, X)
Collect products of indicator variables
p ∼ mean(wi Hi ) =
linear combination of means of (H2
, HH) =
p + δ
∂w
∂zf
mean(H2
− HH)
Take expectations E.g., a large population without (spatial) structure
p = p + δ
∂w
∂zf
(p − p2
)
Fran¸cois Rousset Fitness for the impatient May 2013 10 / 33
Two views of the Prisoner’s dilemma
(“dilemma”: T > R, P > S, R > P i.e. T > R > P > S)
Fran¸cois Rousset Fitness for the impatient May 2013 11 / 33
Two views of the Prisoner’s dilemma
Phenotype (z): probability of cooperating in a prisoner’s dilemma
Fecundityf ∝ 1 + Rzfz◦ + Szf(1 − z◦) + T(1 − zf)z◦ + P(1 − zf)(1 − z◦)
No spatial structure:
wf =
Fecundityf
Mean fecundity
phenotype zi = zres + Xi δ
∆p = Cov(wi , Xi ) = δpq[S−P+(z+pδ)(R−S+P−T)]+O[δ3
, (R, S, T, P)2
]
View 1: R, T, S, P are given ecological constraints; evolution of z ⇒
expansion in δ. ∆p ∼ δpq[S − P + z(R − S + P − T)]
View 2: expansion in R, T, S, P, not in δ.
Fran¸cois Rousset Fitness for the impatient May 2013 12 / 33
Population structure
Color code: focal, neighbor, population
wf ∼ linear combination of (H, H, H)
so that
mean(H)t+1 ∼ mean(wi Hi ) =
p + linear combination of means of (H2
, HH, HH) =
p + δ
∂w
∂zf
(H2
− HH) +
∂w
∂zn
(HH − HH)
Traditional population genetic argument:
E[H|H, p] = FSTH + (1 − FST)p
for“relatedness”FST independent of p.
Fran¸cois Rousset Fitness for the impatient May 2013 13 / 33
Genealogical interpretation: island model
past
. . . . . .
p p2
Lineages from distinct demes can be considered as draws of independent
genes copies, each A with frequency p.
Probability that first event is coalescence: FST
Such coalescences are recent if migration“not too small”; p then considered
constant if selection is“weak”and total population size is“large”.
Fran¸cois Rousset Fitness for the impatient May 2013 14 / 33
Inclusive fitness under weak selection
Traditional argument about relatedness“r”(or FST) :
E[H|H, p] = rH + (1 − r)p
for“relatedness”r independent of p. Hence
E[HH − HH|p] = p(r + (1 − r)p) − p2
= rpq
hence (with E[HH − HH|p] = pq)
δ
∂w
∂zf
(H2
− HH) +
∂w
∂zn
(HH − HH) = pq δ
∂w
∂zf
+
∂w
∂zn
r
inclusive fitness −c + rb
.
Selection gradient is independent of p.
Fran¸cois Rousset Fitness for the impatient May 2013 15 / 33
Genealogical interpretation: “stepping stone”
past
. . . . . .
Lineages from distinct demes cannot be considered as draws of
independent genes copies, each A with frequency p
Fran¸cois Rousset Fitness for the impatient May 2013 16 / 33
Frequency-dependence in the stepping-stone model
(Circular stepping-stone model with 200 demes of 10 haploid individuals,
dispersal rate 0.2, and a two allele model with mutation rate 10−5)
Fran¸cois Rousset Fitness for the impatient May 2013 17 / 33
Frequency-dependence in the stepping-stone model
(Circular stepping-stone model with 200 demes of 10 haploid individuals,
dispersal rate 0.2, and a two allele model with mutation rate 10−5)
Fran¸cois Rousset Fitness for the impatient May 2013 17 / 33
Weak selection under localized dispersal
Asymptotic results for large number of demes:
(Infinite) island model
∆p ∼ δpqsIF = δpq(1 − FST)φ
FST ≡
E(XX) − E(XX)
E(XX) − E(XX)
sIF: scaled inclusive fitness effect, selection gradient
φ: p-independent localized selection gradient
Fran¸cois Rousset Fitness for the impatient May 2013 18 / 33
Weak selection under localized dispersal
Asymptotic results for large number of demes:
(Infinite) island model
∆p ∼ δpqsIF = δpq(1 − FST)φ
Localized dispersal
∆p ∼ δpqsIF(p) = δpq(1 − FST(p))φ
FST(p) same as FST but for conditional probabilities
FST ≡
E(XX) − E(XX)
E(XX) − E(XX)
FST(p) ≡
E(XX|p) − E(XX|p)
E(XX|p) − E(XX|p)
sIF: scaled inclusive fitness effect, selection gradient
φ: p-independent localized selection gradient
Fran¸cois Rousset Fitness for the impatient May 2013 18 / 33
What does that mean ?
Fran¸cois Rousset Fitness for the impatient May 2013 19 / 33
The localized selection gradient φ and local FST’s
The finite population meaning and computation of φ
φ ≡
∂π
∂δ
= −
k=f
∂w
∂zk
Tk
T0
Tk
T0
= lim
µ→0
E[p − XXk]
E[p − XX0]
= lim
µ→0
1
1 − FSTk
Localized interactions: short distances (k) only. Everything
understandable as the result of local interactions (global p doesn’t
matter!).
Expressed in terms of local population structure parameters relatively
easy to estimate using genetic markers, and with genealogical
interpretation(s)
Genealogical: FSTk(p) = FSTk at neutral genetic markers
Fran¸cois Rousset Fitness for the impatient May 2013 20 / 33
Fitness costs and benefits
Fran¸cois Rousset Fitness for the impatient May 2013 21 / 33
Fitness costs and benefits
Effects B on neighbors’ fecundity and −C on focal’s fecundity
rb − c = rB − C? In general No!
Fran¸cois Rousset Fitness for the impatient May 2013 21 / 33
Fitness costs and benefits
Effects B on neighbors’ fecundity and −C on focal’s fecundity
rb − c = rB − C? In general No!
Local competition ⇒“inclusive fitness”= −C (Taylor, 1992).
Actually ∆p ∼ −pq(1 − FST)C in island model.
Fran¸cois Rousset Fitness for the impatient May 2013 21 / 33
Fitness costs and benefits
Effects B on neighbors’ fecundity and −C on focal’s fecundity
rb − c = rB − C? In general No!
Local competition ⇒“inclusive fitness”= −C (Taylor, 1992).
Actually ∆p ∼ −pq(1 − FST)C in island model.
How to obtain rb − c ∝ rB − C?
Fran¸cois Rousset Fitness for the impatient May 2013 21 / 33
Fitness costs and benefits
Effects B on neighbors’ fecundity and −C on focal’s fecundity
rb − c = rB − C? In general No!
Local competition ⇒“inclusive fitness”= −C (Taylor, 1992).
Actually ∆p ∼ −pq(1 − FST)C in island model.
How to obtain rb − c ∝ rB − C?
Hamilton (1975): “groups break up completely and re-form in each
generation”“young animals take off to form a migrant pool”... [one type]
assort[s] positively with its own type in settling from the migrant pool(...)
to such a degree that the correlation of two separate randomly selected
members (...) is F”
⇒ ∆p ∼ (RB − C)pq
Fran¸cois Rousset Fitness for the impatient May 2013 21 / 33
Fitness costs and benefits
Effects B on neighbors’ fecundity and −C on focal’s fecundity
rb − c = rB − C? In general No!
Local competition ⇒“inclusive fitness”= −C (Taylor, 1992).
Actually ∆p ∼ −pq(1 − FST)C in island model.
How to obtain rb − c ∝ rB − C?
Version with spatially restricted dispersal
Individuals disperse as a group and compete as a group against other
groups for access to whole group breeding spots. The winners of such
group contests can then occupy whole demes
⇒ ∆p ∼ (1 − F)(RB − C)pq
(Gardner & West 2006; Lehmann et al. 2006).
Fran¸cois Rousset Fitness for the impatient May 2013 21 / 33
More general logic: example of dominance
Phenotype zi = z + δ[2h(Xi1 + Xi2)/2 + (1 − 2h)Xi1Xi2]
To first order,
wf(zf, zp) ∼ linear combination of (X1, X2, X1X2, ...)
so that
E[p ] = mean(wi Xi ) =
linear combination of means of (X2
1, X1X2, X1X, X1X1X2, . . . , X1X1X2) =
Triplets generally lead to frequency-dependence, although there are
intriguing exceptions:
Helping among diploid full sibs and among haplodiploid sisters, partial sib
mating (α)
∆p ∼ δpq
∂w
∂zf
+
∂w
∂zn
r f (h, α, /p)
Fran¸cois Rousset Fitness for the impatient May 2013 22 / 33
Three gene lineages... island model
past
. . . . . .
p p2 p3
Lineages from distinct demes can be considered as draws of independent
genes copies, each A with frequency p.
Probability that first event is coalescence: FST
Fran¸cois Rousset Fitness for the impatient May 2013 23 / 33
More general logic: example of kin recognition
Demes of N individuals; two loci; indicator variables R, H for alleles
Conditional helping:
fecundityf = 1 +
1
N − 1
neighbours k
[RfRk + (1 − Rf)(1 − Rk)](−CHf + BHk)
(two recognition alleles case)
wf =
(1 − d)ff
(1 − d)fdeme + dfothers
+ d
ff
fothers
(dispersal probability d; regulation after dispersal)
mean(H)t+1 =
mean(wi Hi )
mean(wi )
= mean(wi Hi ) over individuals i.
(strict regulation)
Fran¸cois Rousset Fitness for the impatient May 2013 24 / 33
Expectedly...
To first order in C and B,
wf ∼ linear combination of (H, H, RH, RH, RRH, RH, RRH,
H, RH, RRH, RH)
so that
mean(H)t+1 ∼ mean(wi Hi ) =
linear combination of expectations of (H2
, HH, RH2
, RH2
, RRH2
,
RHH, RRHH, HH, RHH, RRHH, RHH)
RH RH R RH H
Fran¸cois Rousset Fitness for the impatient May 2013 25 / 33
Meaning
An act of helping always involves the configuration
HR R
−C B
Fran¸cois Rousset Fitness for the impatient May 2013 26 / 33
Meaning
An act of helping always involves the configuration
HR R
−C B
RH RH
describes the probability that the receiver bears the helping allele;
can be computed in terms of the probability of joint coalescence within the
deme
Fran¸cois Rousset Fitness for the impatient May 2013 26 / 33
Meaning
An act of helping always involves the configuration
HR R
−C B
R RH H
describes the probability that a third individual bears the helping allele.
The fitness of this individual is reduced in proportion to (B-C), that is, in
proportion to the increase in fitness of the pair of interacting individuals;
can be computed in terms of the probability of joint coalescence within the
deme
Fran¸cois Rousset Fitness for the impatient May 2013 26 / 33
Kin recognition: results
∆pH ∼ − C (1 − F) (1 − 2pqR) pqH
+ −C F + B φ − (1 − m)2
(B − C)
1
N
(F + φ) + 1 −
1
N
γ
∆pR ∼pHpqR(1 − 2pR)
B − C
N
[Z(N, m) < 0]
Polymorphism lost at the recognition locus.
Fran¸cois Rousset Fitness for the impatient May 2013 27 / 33
Synthesis and developments
Many results follow mechanically from a trivial description of
allele-frequency changes in terms of fitness functions:
* Write a properly defined fitness function in terms of individual
behaviour
* Express behaviour in term of genotypes (indicator variables)
* Expand wi Xi to appropriate order, and take expectations of products
of indicator variables (special case: “direct fitness”method, Taylor &
Frank 1996)
Fran¸cois Rousset Fitness for the impatient May 2013 28 / 33
Synthesis and developments
Many results follow mechanically from a trivial description of
allele-frequency changes in terms of fitness functions:
* Write a properly defined fitness function in terms of individual
behaviour
* Express behaviour in term of genotypes (indicator variables)
* Expand wi Xi to appropriate order, and take expectations of products
of indicator variables (special case: “direct fitness”method, Taylor &
Frank 1996)
Deterministic“multi”locus models, with a recombination step:
∆(mean(any thing X)) = ∆sel(mean(X)) + ∆recomb(mean(X)).
Multilocus models often in terms of“centered”associations that are 0
in expectation in neutral models, e.g. E[(R − pR)(H − pH)].
Fran¸cois Rousset Fitness for the impatient May 2013 28 / 33
Synthesis and developments
Many results follow mechanically from a trivial description of
allele-frequency changes in terms of fitness functions:
* Write a properly defined fitness function in terms of individual
behaviour
* Express behaviour in term of genotypes (indicator variables)
* Expand wi Xi to appropriate order, and take expectations of products
of indicator variables (special case: “direct fitness”method, Taylor &
Frank 1996)
Deterministic“multi”locus models, with a recombination step:
Diffusion methods (first order) e.g. approximation for fixation
probability
π ∼
1 − e−2Ntotφpini
1 − e−2Ntotφ
for φ taken in an infinite-deme limit.
Fran¸cois Rousset Fitness for the impatient May 2013 28 / 33
Synthesis and developments
Many results follow mechanically from a trivial description of
allele-frequency changes in terms of fitness functions:
* Write a properly defined fitness function in terms of individual
behaviour
* Express behaviour in term of genotypes (indicator variables)
* Expand wi Xi to appropriate order, and take expectations of products
of indicator variables (special case: “direct fitness”method, Taylor &
Frank 1996)
Deterministic“multi”locus models, with a recombination step:
Diffusion methods (first order)
Diffusion with p-dependence (in first order, e.g. dominance)
Fran¸cois Rousset Fitness for the impatient May 2013 28 / 33
Synthesis and developments
Many results follow mechanically from a trivial description of
allele-frequency changes in terms of fitness functions:
* Write a properly defined fitness function in terms of individual
behaviour
* Express behaviour in term of genotypes (indicator variables)
* Expand wi Xi to appropriate order, and take expectations of products
of indicator variables (special case: “direct fitness”method, Taylor &
Frank 1996)
Deterministic“multi”locus models, with a recombination step:
Diffusion methods (first order)
Diffusion with p-dependence (in first order, e.g. dominance)
Evolutionary stability
Fran¸cois Rousset Fitness for the impatient May 2013 28 / 33
Continuous evolutionary stability in the island model
Second-order computation for any p feasible.
Simpler computation for extreme p, and connexion to“number of
successful emigrants”approach (Chesson 1984; Metz and Gyllenberg 2001)
B = ∂zf,zf
w +2F∂zf,zn
w +K∂zn,zn
w +4F(N −1) K∂zn
wp + F∂fwp wp∂zn
w
where wp: local offspring
In contrast to formula proposed by Day and Taylor (1998):
(1) three-genes coefficient K;
(2) products of identity coefficients, product of derivatives: joint effect of
first-order change in number of offspring and first-order change in parental
population structure.
Fran¸cois Rousset Fitness for the impatient May 2013 29 / 33
Kin recognition: searching for stable polymorphisms
Second-order computation involves 30 associations, for up to 7 gene copies
in 5 individuals: R RH/R RH/R
Search for conditions for stable polymorphism: low migration and low
recombination
Convergent orbits plus drift:
Fran¸cois Rousset Fitness for the impatient May 2013 30 / 33
“Conclusions”
First order: Simple formalism provides connections between different
modelling approaches (inclusive fitness, adaptive dynamics, diffusion,
coalescence). Relatively simple pattern of frequency-dependence.
Second order: Essentially the same formalism previously used for
multilocus models; Algebraically (and algorithmically) messy in
spatially structured populations; still allows an analysis of the forces
acting on a trait.
Fran¸cois Rousset Fitness for the impatient May 2013 31 / 33
More general coalescent interpretation of relatedness
Fran¸cois Rousset Fitness for the impatient May 2013 32 / 33
Quasi equilibrium
Quasi linkage equilibrium
D(t + 1) = (1 − r)D(t) + O(δ) ⇒ ˆD =
O(s)
r
understood as
O(s)
r
=
O(s)
1 − (1 − r)
= (1 − r)O(s) + (1 − r)2
O(s) + (1 − r)3
O(s) . . .
Quasi equilibrium
D(t + 1) − D◦
= λ(D(t) − D◦
) + O(δ) ⇒ D(t + 1) − D◦ =
O(s)
1 − λ
Actually
D(t + 1) − D◦
= A(D(t) − D◦
) + O(δ) ⇒ D − D◦ = (I − A)−1
O(s)
Fran¸cois Rousset Fitness for the impatient May 2013 33 / 33

More Related Content

Viewers also liked

Guidance, Please! Towards a Framework for RDF-based Constraint Languages.
Guidance, Please! Towards a Framework for RDF-based Constraint Languages.Guidance, Please! Towards a Framework for RDF-based Constraint Languages.
Guidance, Please! Towards a Framework for RDF-based Constraint Languages.
Kai Eckert
 
SCAHD - FEP In The Raiser's Edge
SCAHD - FEP In The Raiser's EdgeSCAHD - FEP In The Raiser's Edge
SCAHD - FEP In The Raiser's Edge
Heather Paul
 
Nicolas Loeuille - présentation MEE2013
Nicolas Loeuille - présentation MEE2013Nicolas Loeuille - présentation MEE2013
Nicolas Loeuille - présentation MEE2013
Seminaire MEE
 
Dive & donne
Dive & donneDive & donne
Dive & donne
yaya117
 
Thomas Bataillon - présentation MEE 2013
Thomas Bataillon - présentation MEE 2013Thomas Bataillon - présentation MEE 2013
Thomas Bataillon - présentation MEE 2013
Seminaire MEE
 
François Massol - présentation MEE2013
François Massol - présentation MEE2013François Massol - présentation MEE2013
François Massol - présentation MEE2013
Seminaire MEE
 
Virginie Ravigné - Dynamique adaptative
Virginie Ravigné - Dynamique adaptativeVirginie Ravigné - Dynamique adaptative
Virginie Ravigné - Dynamique adaptative
Seminaire MEE
 
Thomas Lenormand - Génétique des populations
Thomas Lenormand - Génétique des populationsThomas Lenormand - Génétique des populations
Thomas Lenormand - Génétique des populations
Seminaire MEE
 
Nils Poulicard - Relations entre histoire évolutive et capacité d'adaptation ...
Nils Poulicard - Relations entre histoire évolutive et capacité d'adaptation ...Nils Poulicard - Relations entre histoire évolutive et capacité d'adaptation ...
Nils Poulicard - Relations entre histoire évolutive et capacité d'adaptation ...
Seminaire MEE
 
François Blanquart - Evolution of migration in a fluctuating environment
François Blanquart - Evolution of migration in a fluctuating environmentFrançois Blanquart - Evolution of migration in a fluctuating environment
François Blanquart - Evolution of migration in a fluctuating environment
Seminaire MEE
 

Viewers also liked (20)

Metadata Provenance Tutorial at SWIB 13, Part 1
Metadata Provenance Tutorial at SWIB 13, Part 1Metadata Provenance Tutorial at SWIB 13, Part 1
Metadata Provenance Tutorial at SWIB 13, Part 1
 
JudaicaLink: Linked Data from Jewish Encyclopediae
JudaicaLink: Linked Data from Jewish EncyclopediaeJudaicaLink: Linked Data from Jewish Encyclopediae
JudaicaLink: Linked Data from Jewish Encyclopediae
 
Extending DCAM for Metadata Provenance
Extending DCAM for Metadata ProvenanceExtending DCAM for Metadata Provenance
Extending DCAM for Metadata Provenance
 
Guidance, Please! Towards a Framework for RDF-based Constraint Languages.
Guidance, Please! Towards a Framework for RDF-based Constraint Languages.Guidance, Please! Towards a Framework for RDF-based Constraint Languages.
Guidance, Please! Towards a Framework for RDF-based Constraint Languages.
 
Portifolio designer Claudio Lopes
Portifolio designer Claudio LopesPortifolio designer Claudio Lopes
Portifolio designer Claudio Lopes
 
Towards Interoperable Metadata Provenance
Towards Interoperable Metadata ProvenanceTowards Interoperable Metadata Provenance
Towards Interoperable Metadata Provenance
 
Specialising the EDM for Digitised Manuscript (SWIB13)
Specialising the EDM for Digitised Manuscript (SWIB13)Specialising the EDM for Digitised Manuscript (SWIB13)
Specialising the EDM for Digitised Manuscript (SWIB13)
 
Metadata Provenance
Metadata ProvenanceMetadata Provenance
Metadata Provenance
 
The DM2E Data Model and the DM2E Ingestion Infrastructure
The DM2E Data Model and the DM2E Ingestion InfrastructureThe DM2E Data Model and the DM2E Ingestion Infrastructure
The DM2E Data Model and the DM2E Ingestion Infrastructure
 
LOHAI: Providing a baseline for KOS based automatic indexing
LOHAI: Providing a baseline for KOS based automatic indexingLOHAI: Providing a baseline for KOS based automatic indexing
LOHAI: Providing a baseline for KOS based automatic indexing
 
SCAHD - FEP In The Raiser's Edge
SCAHD - FEP In The Raiser's EdgeSCAHD - FEP In The Raiser's Edge
SCAHD - FEP In The Raiser's Edge
 
Nicolas Loeuille - présentation MEE2013
Nicolas Loeuille - présentation MEE2013Nicolas Loeuille - présentation MEE2013
Nicolas Loeuille - présentation MEE2013
 
Dive & donne
Dive & donneDive & donne
Dive & donne
 
Thomas Bataillon - présentation MEE 2013
Thomas Bataillon - présentation MEE 2013Thomas Bataillon - présentation MEE 2013
Thomas Bataillon - présentation MEE 2013
 
François Massol - présentation MEE2013
François Massol - présentation MEE2013François Massol - présentation MEE2013
François Massol - présentation MEE2013
 
Amy keynote x
Amy keynote xAmy keynote x
Amy keynote x
 
Virginie Ravigné - Dynamique adaptative
Virginie Ravigné - Dynamique adaptativeVirginie Ravigné - Dynamique adaptative
Virginie Ravigné - Dynamique adaptative
 
Thomas Lenormand - Génétique des populations
Thomas Lenormand - Génétique des populationsThomas Lenormand - Génétique des populations
Thomas Lenormand - Génétique des populations
 
Nils Poulicard - Relations entre histoire évolutive et capacité d'adaptation ...
Nils Poulicard - Relations entre histoire évolutive et capacité d'adaptation ...Nils Poulicard - Relations entre histoire évolutive et capacité d'adaptation ...
Nils Poulicard - Relations entre histoire évolutive et capacité d'adaptation ...
 
François Blanquart - Evolution of migration in a fluctuating environment
François Blanquart - Evolution of migration in a fluctuating environmentFrançois Blanquart - Evolution of migration in a fluctuating environment
François Blanquart - Evolution of migration in a fluctuating environment
 

Similar to François Rousset - présentation MEE2013

GEOGRAPHY Population Ecology HSC MAHARASHTRA
GEOGRAPHY Population Ecology HSC MAHARASHTRAGEOGRAPHY Population Ecology HSC MAHARASHTRA
GEOGRAPHY Population Ecology HSC MAHARASHTRA
TwinsIT2
 
Genetic markers in characterization2
Genetic markers in characterization2Genetic markers in characterization2
Genetic markers in characterization2
Bruno Mmassy
 

Similar to François Rousset - présentation MEE2013 (20)

Dependent Types and Dynamics of Natural Language
Dependent Types and Dynamics of Natural LanguageDependent Types and Dynamics of Natural Language
Dependent Types and Dynamics of Natural Language
 
HDR Olivier Gimenez
HDR Olivier GimenezHDR Olivier Gimenez
HDR Olivier Gimenez
 
Population.ppt
Population.pptPopulation.ppt
Population.ppt
 
PopulationEcology52.ppt
PopulationEcology52.pptPopulationEcology52.ppt
PopulationEcology52.ppt
 
GEOGRAPHY Population Ecology HSC MAHARASHTRA
GEOGRAPHY Population Ecology HSC MAHARASHTRAGEOGRAPHY Population Ecology HSC MAHARASHTRA
GEOGRAPHY Population Ecology HSC MAHARASHTRA
 
population genetics of gene function (talk)
population genetics of gene function (talk)population genetics of gene function (talk)
population genetics of gene function (talk)
 
Population Genetics and Hardy Weinberg Law for B.Sc. (Ag.)
Population Genetics and Hardy Weinberg Law for B.Sc. (Ag.)Population Genetics and Hardy Weinberg Law for B.Sc. (Ag.)
Population Genetics and Hardy Weinberg Law for B.Sc. (Ag.)
 
An Introduction To Basic Statistics And Probability
An Introduction To Basic Statistics And ProbabilityAn Introduction To Basic Statistics And Probability
An Introduction To Basic Statistics And Probability
 
Approximate Bayesian model choice via random forests
Approximate Bayesian model choice via random forestsApproximate Bayesian model choice via random forests
Approximate Bayesian model choice via random forests
 
Oaxaca-Blinder type Decomposition Methods for Duration Outcomes
Oaxaca-Blinder type Decomposition Methods for Duration OutcomesOaxaca-Blinder type Decomposition Methods for Duration Outcomes
Oaxaca-Blinder type Decomposition Methods for Duration Outcomes
 
RIVER NETWORKS AS ECOLOGICAL CORRIDORS FOR SPECIES POPULATIONS AND WATER-BORN...
RIVER NETWORKS AS ECOLOGICAL CORRIDORS FOR SPECIES POPULATIONS AND WATER-BORN...RIVER NETWORKS AS ECOLOGICAL CORRIDORS FOR SPECIES POPULATIONS AND WATER-BORN...
RIVER NETWORKS AS ECOLOGICAL CORRIDORS FOR SPECIES POPULATIONS AND WATER-BORN...
 
NBBC15, Reyjavik, June 08, 2015
NBBC15, Reyjavik, June 08, 2015NBBC15, Reyjavik, June 08, 2015
NBBC15, Reyjavik, June 08, 2015
 
ESSLLI2016 DTS Lecture Day 5-2: Proof-theoretic Turn
ESSLLI2016 DTS Lecture Day 5-2: Proof-theoretic TurnESSLLI2016 DTS Lecture Day 5-2: Proof-theoretic Turn
ESSLLI2016 DTS Lecture Day 5-2: Proof-theoretic Turn
 
Chapter18
Chapter18Chapter18
Chapter18
 
Generalizing phylogenetics to infer shared evolutionary events
Generalizing phylogenetics to infer shared evolutionary eventsGeneralizing phylogenetics to infer shared evolutionary events
Generalizing phylogenetics to infer shared evolutionary events
 
FOLBUKCFAIZ.pptx
FOLBUKCFAIZ.pptxFOLBUKCFAIZ.pptx
FOLBUKCFAIZ.pptx
 
Hardy weinberg law
Hardy weinberg lawHardy weinberg law
Hardy weinberg law
 
Phd Defence talk
Phd Defence talkPhd Defence talk
Phd Defence talk
 
06 random drift
06 random drift06 random drift
06 random drift
 
Genetic markers in characterization2
Genetic markers in characterization2Genetic markers in characterization2
Genetic markers in characterization2
 

More from Seminaire MEE

Philippe Huneman - présentation MEE2013
Philippe Huneman - présentation MEE2013Philippe Huneman - présentation MEE2013
Philippe Huneman - présentation MEE2013
Seminaire MEE
 
Laurent Lehmann - Evolution of long lasting behaviours
Laurent Lehmann - Evolution of long lasting behavioursLaurent Lehmann - Evolution of long lasting behaviours
Laurent Lehmann - Evolution of long lasting behaviours
Seminaire MEE
 
Virginie Rougeron - Sexuality and clonality in Leishmania
Virginie Rougeron - Sexuality and clonality in LeishmaniaVirginie Rougeron - Sexuality and clonality in Leishmania
Virginie Rougeron - Sexuality and clonality in Leishmania
Seminaire MEE
 
Nicolas Rode - Coévolutions mâle-femelle dans la nature : Mise en évidence pa...
Nicolas Rode - Coévolutions mâle-femelle dans la nature : Mise en évidence pa...Nicolas Rode - Coévolutions mâle-femelle dans la nature : Mise en évidence pa...
Nicolas Rode - Coévolutions mâle-femelle dans la nature : Mise en évidence pa...
Seminaire MEE
 
Ophélie Ronce - Evolution de la dispersion
Ophélie Ronce - Evolution de la dispersionOphélie Ronce - Evolution de la dispersion
Ophélie Ronce - Evolution de la dispersion
Seminaire MEE
 
Michel Morange - La modélisation comme pratique scientifique
Michel Morange - La modélisation comme pratique scientifiqueMichel Morange - La modélisation comme pratique scientifique
Michel Morange - La modélisation comme pratique scientifique
Seminaire MEE
 
Patrice David - Modélisation en évolution et génétique quantitative
Patrice David - Modélisation en évolution et génétique quantitativePatrice David - Modélisation en évolution et génétique quantitative
Patrice David - Modélisation en évolution et génétique quantitative
Seminaire MEE
 
Sylvain Gandon - Epidémiologie évolutive
Sylvain Gandon - Epidémiologie évolutiveSylvain Gandon - Epidémiologie évolutive
Sylvain Gandon - Epidémiologie évolutive
Seminaire MEE
 
Marco Andrello - Incongruency between model-based and genetic-based estimates...
Marco Andrello - Incongruency between model-based and genetic-based estimates...Marco Andrello - Incongruency between model-based and genetic-based estimates...
Marco Andrello - Incongruency between model-based and genetic-based estimates...
Seminaire MEE
 
Laurent Lehmann - The evolution of long lasting behaviours
Laurent Lehmann - The evolution of long lasting behavioursLaurent Lehmann - The evolution of long lasting behaviours
Laurent Lehmann - The evolution of long lasting behaviours
Seminaire MEE
 
Betty Courquin - Etude des adaptations locales chez une espèce menacée
Betty Courquin - Etude des adaptations locales chez une espèce menacéeBetty Courquin - Etude des adaptations locales chez une espèce menacée
Betty Courquin - Etude des adaptations locales chez une espèce menacée
Seminaire MEE
 

More from Seminaire MEE (11)

Philippe Huneman - présentation MEE2013
Philippe Huneman - présentation MEE2013Philippe Huneman - présentation MEE2013
Philippe Huneman - présentation MEE2013
 
Laurent Lehmann - Evolution of long lasting behaviours
Laurent Lehmann - Evolution of long lasting behavioursLaurent Lehmann - Evolution of long lasting behaviours
Laurent Lehmann - Evolution of long lasting behaviours
 
Virginie Rougeron - Sexuality and clonality in Leishmania
Virginie Rougeron - Sexuality and clonality in LeishmaniaVirginie Rougeron - Sexuality and clonality in Leishmania
Virginie Rougeron - Sexuality and clonality in Leishmania
 
Nicolas Rode - Coévolutions mâle-femelle dans la nature : Mise en évidence pa...
Nicolas Rode - Coévolutions mâle-femelle dans la nature : Mise en évidence pa...Nicolas Rode - Coévolutions mâle-femelle dans la nature : Mise en évidence pa...
Nicolas Rode - Coévolutions mâle-femelle dans la nature : Mise en évidence pa...
 
Ophélie Ronce - Evolution de la dispersion
Ophélie Ronce - Evolution de la dispersionOphélie Ronce - Evolution de la dispersion
Ophélie Ronce - Evolution de la dispersion
 
Michel Morange - La modélisation comme pratique scientifique
Michel Morange - La modélisation comme pratique scientifiqueMichel Morange - La modélisation comme pratique scientifique
Michel Morange - La modélisation comme pratique scientifique
 
Patrice David - Modélisation en évolution et génétique quantitative
Patrice David - Modélisation en évolution et génétique quantitativePatrice David - Modélisation en évolution et génétique quantitative
Patrice David - Modélisation en évolution et génétique quantitative
 
Sylvain Gandon - Epidémiologie évolutive
Sylvain Gandon - Epidémiologie évolutiveSylvain Gandon - Epidémiologie évolutive
Sylvain Gandon - Epidémiologie évolutive
 
Marco Andrello - Incongruency between model-based and genetic-based estimates...
Marco Andrello - Incongruency between model-based and genetic-based estimates...Marco Andrello - Incongruency between model-based and genetic-based estimates...
Marco Andrello - Incongruency between model-based and genetic-based estimates...
 
Laurent Lehmann - The evolution of long lasting behaviours
Laurent Lehmann - The evolution of long lasting behavioursLaurent Lehmann - The evolution of long lasting behaviours
Laurent Lehmann - The evolution of long lasting behaviours
 
Betty Courquin - Etude des adaptations locales chez une espèce menacée
Betty Courquin - Etude des adaptations locales chez une espèce menacéeBetty Courquin - Etude des adaptations locales chez une espèce menacée
Betty Courquin - Etude des adaptations locales chez une espèce menacée
 

Recently uploaded

Jual Obat Aborsi Hongkong ( Asli No.1 ) 085657271886 Obat Penggugur Kandungan...
Jual Obat Aborsi Hongkong ( Asli No.1 ) 085657271886 Obat Penggugur Kandungan...Jual Obat Aborsi Hongkong ( Asli No.1 ) 085657271886 Obat Penggugur Kandungan...
Jual Obat Aborsi Hongkong ( Asli No.1 ) 085657271886 Obat Penggugur Kandungan...
ZurliaSoop
 

Recently uploaded (20)

Accessible Digital Futures project (20/03/2024)
Accessible Digital Futures project (20/03/2024)Accessible Digital Futures project (20/03/2024)
Accessible Digital Futures project (20/03/2024)
 
Holdier Curriculum Vitae (April 2024).pdf
Holdier Curriculum Vitae (April 2024).pdfHoldier Curriculum Vitae (April 2024).pdf
Holdier Curriculum Vitae (April 2024).pdf
 
FSB Advising Checklist - Orientation 2024
FSB Advising Checklist - Orientation 2024FSB Advising Checklist - Orientation 2024
FSB Advising Checklist - Orientation 2024
 
Exploring_the_Narrative_Style_of_Amitav_Ghoshs_Gun_Island.pptx
Exploring_the_Narrative_Style_of_Amitav_Ghoshs_Gun_Island.pptxExploring_the_Narrative_Style_of_Amitav_Ghoshs_Gun_Island.pptx
Exploring_the_Narrative_Style_of_Amitav_Ghoshs_Gun_Island.pptx
 
TỔNG ÔN TẬP THI VÀO LỚP 10 MÔN TIẾNG ANH NĂM HỌC 2023 - 2024 CÓ ĐÁP ÁN (NGỮ Â...
TỔNG ÔN TẬP THI VÀO LỚP 10 MÔN TIẾNG ANH NĂM HỌC 2023 - 2024 CÓ ĐÁP ÁN (NGỮ Â...TỔNG ÔN TẬP THI VÀO LỚP 10 MÔN TIẾNG ANH NĂM HỌC 2023 - 2024 CÓ ĐÁP ÁN (NGỮ Â...
TỔNG ÔN TẬP THI VÀO LỚP 10 MÔN TIẾNG ANH NĂM HỌC 2023 - 2024 CÓ ĐÁP ÁN (NGỮ Â...
 
HMCS Max Bernays Pre-Deployment Brief (May 2024).pptx
HMCS Max Bernays Pre-Deployment Brief (May 2024).pptxHMCS Max Bernays Pre-Deployment Brief (May 2024).pptx
HMCS Max Bernays Pre-Deployment Brief (May 2024).pptx
 
On_Translating_a_Tamil_Poem_by_A_K_Ramanujan.pptx
On_Translating_a_Tamil_Poem_by_A_K_Ramanujan.pptxOn_Translating_a_Tamil_Poem_by_A_K_Ramanujan.pptx
On_Translating_a_Tamil_Poem_by_A_K_Ramanujan.pptx
 
On National Teacher Day, meet the 2024-25 Kenan Fellows
On National Teacher Day, meet the 2024-25 Kenan FellowsOn National Teacher Day, meet the 2024-25 Kenan Fellows
On National Teacher Day, meet the 2024-25 Kenan Fellows
 
REMIFENTANIL: An Ultra short acting opioid.pptx
REMIFENTANIL: An Ultra short acting opioid.pptxREMIFENTANIL: An Ultra short acting opioid.pptx
REMIFENTANIL: An Ultra short acting opioid.pptx
 
2024-NATIONAL-LEARNING-CAMP-AND-OTHER.pptx
2024-NATIONAL-LEARNING-CAMP-AND-OTHER.pptx2024-NATIONAL-LEARNING-CAMP-AND-OTHER.pptx
2024-NATIONAL-LEARNING-CAMP-AND-OTHER.pptx
 
Jual Obat Aborsi Hongkong ( Asli No.1 ) 085657271886 Obat Penggugur Kandungan...
Jual Obat Aborsi Hongkong ( Asli No.1 ) 085657271886 Obat Penggugur Kandungan...Jual Obat Aborsi Hongkong ( Asli No.1 ) 085657271886 Obat Penggugur Kandungan...
Jual Obat Aborsi Hongkong ( Asli No.1 ) 085657271886 Obat Penggugur Kandungan...
 
Understanding Accommodations and Modifications
Understanding  Accommodations and ModificationsUnderstanding  Accommodations and Modifications
Understanding Accommodations and Modifications
 
Jamworks pilot and AI at Jisc (20/03/2024)
Jamworks pilot and AI at Jisc (20/03/2024)Jamworks pilot and AI at Jisc (20/03/2024)
Jamworks pilot and AI at Jisc (20/03/2024)
 
Towards a code of practice for AI in AT.pptx
Towards a code of practice for AI in AT.pptxTowards a code of practice for AI in AT.pptx
Towards a code of practice for AI in AT.pptx
 
Single or Multiple melodic lines structure
Single or Multiple melodic lines structureSingle or Multiple melodic lines structure
Single or Multiple melodic lines structure
 
Application orientated numerical on hev.ppt
Application orientated numerical on hev.pptApplication orientated numerical on hev.ppt
Application orientated numerical on hev.ppt
 
UGC NET Paper 1 Mathematical Reasoning & Aptitude.pdf
UGC NET Paper 1 Mathematical Reasoning & Aptitude.pdfUGC NET Paper 1 Mathematical Reasoning & Aptitude.pdf
UGC NET Paper 1 Mathematical Reasoning & Aptitude.pdf
 
HMCS Vancouver Pre-Deployment Brief - May 2024 (Web Version).pptx
HMCS Vancouver Pre-Deployment Brief - May 2024 (Web Version).pptxHMCS Vancouver Pre-Deployment Brief - May 2024 (Web Version).pptx
HMCS Vancouver Pre-Deployment Brief - May 2024 (Web Version).pptx
 
Micro-Scholarship, What it is, How can it help me.pdf
Micro-Scholarship, What it is, How can it help me.pdfMicro-Scholarship, What it is, How can it help me.pdf
Micro-Scholarship, What it is, How can it help me.pdf
 
How to Add New Custom Addons Path in Odoo 17
How to Add New Custom Addons Path in Odoo 17How to Add New Custom Addons Path in Odoo 17
How to Add New Custom Addons Path in Odoo 17
 

François Rousset - présentation MEE2013

  • 1. Fitness for the impatient Fran¸cois Rousset May 2013 Fran¸cois Rousset Fitness for the impatient May 2013 1 / 33
  • 2. The message To understand the forces operating in social evolution, particularly in spatially structured populations: Fran¸cois Rousset Fitness for the impatient May 2013 2 / 33
  • 3. The message To understand the forces operating in social evolution, particularly in spatially structured populations: Relatedness concepts under localized dispersal Fran¸cois Rousset Fitness for the impatient May 2013 2 / 33
  • 4. The message To understand the forces operating in social evolution, particularly in spatially structured populations: Relatedness concepts under localized dispersal Stability of kin recognition polymorphisms Relationship between inclusive fitness and evolutionary stability Fran¸cois Rousset Fitness for the impatient May 2013 2 / 33
  • 5. The message To understand the forces operating in social evolution, particularly in spatially structured populations: Relatedness concepts under localized dispersal Stability of kin recognition polymorphisms Relationship between inclusive fitness and evolutionary stability A fundamental impatience: reduce these problems to trivial building blocks Fran¸cois Rousset Fitness for the impatient May 2013 2 / 33
  • 6. The message To understand the forces operating in social evolution, particularly in spatially structured populations: Relatedness concepts under localized dispersal Stability of kin recognition polymorphisms Relationship between inclusive fitness and evolutionary stability A fundamental impatience: reduce these problems to trivial building blocks Draw connections to other methods such as diffusion theory, multilocus methods Fran¸cois Rousset Fitness for the impatient May 2013 2 / 33
  • 7. Continuous evolutionary stability for the very impatient Two alleles, a and A, inducing phenotypes za and zA Fran¸cois Rousset Fitness for the impatient May 2013 3 / 33
  • 8. Continuous evolutionary stability for the very impatient Two alleles, a and A, inducing phenotypes za and zA FitnessofAallele Phenotypes za, zA 1 a z1 zm z2 Fran¸cois Rousset Fitness for the impatient May 2013 3 / 33
  • 9. Continuous evolutionary stability for the very impatient Two alleles, a and A, inducing phenotypes za and zA FitnessofAallele Phenotypes za, zA 1 a z1 zm z2 Fran¸cois Rousset Fitness for the impatient May 2013 3 / 33
  • 10. Continuous evolutionary stability for the very impatient Classification of different cases: Continuously stable strategy CSS Branching point Evolutionarily stable noninvasible Invasible Unattainable Convergence stable attainable Fran¸cois Rousset Fitness for the impatient May 2013 4 / 33 Eshel 1983,1996; Christiansen, 1991; Abrams et al., 1993
  • 11. Implicit assumptions Resident phenotype z, mutant z + δ ∆p ∼ δstuff + δ2 blob (1) assumed to be essentially of the form ∆p ∼ δp(1 − p)S(z) + δ2 p(1 − p)(1 − 2p)blob(z) (2) where S(z) and blob(z) are of constant sign wrt p. Fran¸cois Rousset Fitness for the impatient May 2013 5 / 33
  • 12. Implicit assumptions Resident phenotype z, mutant z + δ ∆p ∼ δstuff + δ2 blob (1) assumed to be essentially of the form ∆p ∼ δp(1 − p)S(z) + δ2 p(1 − p)(1 − 2p)blob(z) (2) where S(z) and blob(z) are of constant sign wrt p. Actually S(z) is independent from p in many models (strong claim!). Why? Fran¸cois Rousset Fitness for the impatient May 2013 5 / 33
  • 13. A study of frequency (p) dependence The first-order term A conceptual device Fitness The minimal algorithm Two views of the Prisoner’s dilemma Many views of inclusive fitness The more general logic: dominance, kin recognition Kin recognition Glimpses of second-order results Continuous evolutionary stability Kin recognition Fran¸cois Rousset Fitness for the impatient May 2013 6 / 33
  • 14. A conceptual device p = parents i Ai Xi X indicator variable for an allele; A frequency of copies of parental genes Consider the function f giving (conditional) probabilities E[p | . . .] = i ai (. . .)Xi where ai is the probability that a descendant copy originates from parental copy i. Fran¸cois Rousset Fitness for the impatient May 2013 7 / 33
  • 15. A conceptual device p = gene copies g Ag Xg X indicator variable for an allele; A frequency of copies of parental genes Consider the function f giving (conditional) probabilities E[p | . . .] = i ai (. . .)Xi where ai is the probability that a descendant copy originates from parental copy i. Fran¸cois Rousset Fitness for the impatient May 2013 7 / 33
  • 16. Trivial example Individuals share a resource in total amount R, their fecundity being equal to the amount of resource they consume Their share of resource is proportional to the value of some trait z: sharei = zi i zi = ai , fecundityi = R zi i zi Fran¸cois Rousset Fitness for the impatient May 2013 8 / 33
  • 17. Trivial example Individuals share a resource in total amount R, their fecundity being equal to the amount of resource they consume Their share of resource is proportional to the value of some trait z: sharei = zi i zi = ai , fecundityi = R zi i zi Conflict between individual and group fecundityi = (R − ¯z) zi i zi but ai is still zi i zi . Fran¸cois Rousset Fitness for the impatient May 2013 8 / 33
  • 18. Fitness In terms of the number of adult offspring, or“fitness”W p = gene copies g Xg Wg g Wg = mean(Xg Wg ) Define fitness functions so that E[p |p] = 1 Ntot g Xg w(zg (p)) In a demographically stable population Ntota = w (e.g., w = zi ¯z ). A (minimal: ¯W = 1) Price equation p = ¯X = ¯W ¯X + Cov(wg , Xg ), E[∆p|z] = Cov(wg (z), Xg ). Fran¸cois Rousset Fitness for the impatient May 2013 9 / 33
  • 19. Minimal algorithm Express fitness in terms of indicator variables e.g., phenotype zi = z + Xi δ Differentiate with respect to some measure of strength of selection To first order in δ, wf(zf, zp) ∼ linear combination of (X, X) Collect products of indicator variables p ∼ mean(wi Hi ) = linear combination of means of (H2 , HH) = p + δ ∂w ∂zf mean(H2 − HH) Take expectations E.g., a large population without (spatial) structure p = p + δ ∂w ∂zf (p − p2 ) Fran¸cois Rousset Fitness for the impatient May 2013 10 / 33
  • 20. Two views of the Prisoner’s dilemma (“dilemma”: T > R, P > S, R > P i.e. T > R > P > S) Fran¸cois Rousset Fitness for the impatient May 2013 11 / 33
  • 21. Two views of the Prisoner’s dilemma Phenotype (z): probability of cooperating in a prisoner’s dilemma Fecundityf ∝ 1 + Rzfz◦ + Szf(1 − z◦) + T(1 − zf)z◦ + P(1 − zf)(1 − z◦) No spatial structure: wf = Fecundityf Mean fecundity phenotype zi = zres + Xi δ ∆p = Cov(wi , Xi ) = δpq[S−P+(z+pδ)(R−S+P−T)]+O[δ3 , (R, S, T, P)2 ] View 1: R, T, S, P are given ecological constraints; evolution of z ⇒ expansion in δ. ∆p ∼ δpq[S − P + z(R − S + P − T)] View 2: expansion in R, T, S, P, not in δ. Fran¸cois Rousset Fitness for the impatient May 2013 12 / 33
  • 22. Population structure Color code: focal, neighbor, population wf ∼ linear combination of (H, H, H) so that mean(H)t+1 ∼ mean(wi Hi ) = p + linear combination of means of (H2 , HH, HH) = p + δ ∂w ∂zf (H2 − HH) + ∂w ∂zn (HH − HH) Traditional population genetic argument: E[H|H, p] = FSTH + (1 − FST)p for“relatedness”FST independent of p. Fran¸cois Rousset Fitness for the impatient May 2013 13 / 33
  • 23. Genealogical interpretation: island model past . . . . . . p p2 Lineages from distinct demes can be considered as draws of independent genes copies, each A with frequency p. Probability that first event is coalescence: FST Such coalescences are recent if migration“not too small”; p then considered constant if selection is“weak”and total population size is“large”. Fran¸cois Rousset Fitness for the impatient May 2013 14 / 33
  • 24. Inclusive fitness under weak selection Traditional argument about relatedness“r”(or FST) : E[H|H, p] = rH + (1 − r)p for“relatedness”r independent of p. Hence E[HH − HH|p] = p(r + (1 − r)p) − p2 = rpq hence (with E[HH − HH|p] = pq) δ ∂w ∂zf (H2 − HH) + ∂w ∂zn (HH − HH) = pq δ ∂w ∂zf + ∂w ∂zn r inclusive fitness −c + rb . Selection gradient is independent of p. Fran¸cois Rousset Fitness for the impatient May 2013 15 / 33
  • 25. Genealogical interpretation: “stepping stone” past . . . . . . Lineages from distinct demes cannot be considered as draws of independent genes copies, each A with frequency p Fran¸cois Rousset Fitness for the impatient May 2013 16 / 33
  • 26. Frequency-dependence in the stepping-stone model (Circular stepping-stone model with 200 demes of 10 haploid individuals, dispersal rate 0.2, and a two allele model with mutation rate 10−5) Fran¸cois Rousset Fitness for the impatient May 2013 17 / 33
  • 27. Frequency-dependence in the stepping-stone model (Circular stepping-stone model with 200 demes of 10 haploid individuals, dispersal rate 0.2, and a two allele model with mutation rate 10−5) Fran¸cois Rousset Fitness for the impatient May 2013 17 / 33
  • 28. Weak selection under localized dispersal Asymptotic results for large number of demes: (Infinite) island model ∆p ∼ δpqsIF = δpq(1 − FST)φ FST ≡ E(XX) − E(XX) E(XX) − E(XX) sIF: scaled inclusive fitness effect, selection gradient φ: p-independent localized selection gradient Fran¸cois Rousset Fitness for the impatient May 2013 18 / 33
  • 29. Weak selection under localized dispersal Asymptotic results for large number of demes: (Infinite) island model ∆p ∼ δpqsIF = δpq(1 − FST)φ Localized dispersal ∆p ∼ δpqsIF(p) = δpq(1 − FST(p))φ FST(p) same as FST but for conditional probabilities FST ≡ E(XX) − E(XX) E(XX) − E(XX) FST(p) ≡ E(XX|p) − E(XX|p) E(XX|p) − E(XX|p) sIF: scaled inclusive fitness effect, selection gradient φ: p-independent localized selection gradient Fran¸cois Rousset Fitness for the impatient May 2013 18 / 33
  • 30. What does that mean ? Fran¸cois Rousset Fitness for the impatient May 2013 19 / 33
  • 31. The localized selection gradient φ and local FST’s The finite population meaning and computation of φ φ ≡ ∂π ∂δ = − k=f ∂w ∂zk Tk T0 Tk T0 = lim µ→0 E[p − XXk] E[p − XX0] = lim µ→0 1 1 − FSTk Localized interactions: short distances (k) only. Everything understandable as the result of local interactions (global p doesn’t matter!). Expressed in terms of local population structure parameters relatively easy to estimate using genetic markers, and with genealogical interpretation(s) Genealogical: FSTk(p) = FSTk at neutral genetic markers Fran¸cois Rousset Fitness for the impatient May 2013 20 / 33
  • 32. Fitness costs and benefits Fran¸cois Rousset Fitness for the impatient May 2013 21 / 33
  • 33. Fitness costs and benefits Effects B on neighbors’ fecundity and −C on focal’s fecundity rb − c = rB − C? In general No! Fran¸cois Rousset Fitness for the impatient May 2013 21 / 33
  • 34. Fitness costs and benefits Effects B on neighbors’ fecundity and −C on focal’s fecundity rb − c = rB − C? In general No! Local competition ⇒“inclusive fitness”= −C (Taylor, 1992). Actually ∆p ∼ −pq(1 − FST)C in island model. Fran¸cois Rousset Fitness for the impatient May 2013 21 / 33
  • 35. Fitness costs and benefits Effects B on neighbors’ fecundity and −C on focal’s fecundity rb − c = rB − C? In general No! Local competition ⇒“inclusive fitness”= −C (Taylor, 1992). Actually ∆p ∼ −pq(1 − FST)C in island model. How to obtain rb − c ∝ rB − C? Fran¸cois Rousset Fitness for the impatient May 2013 21 / 33
  • 36. Fitness costs and benefits Effects B on neighbors’ fecundity and −C on focal’s fecundity rb − c = rB − C? In general No! Local competition ⇒“inclusive fitness”= −C (Taylor, 1992). Actually ∆p ∼ −pq(1 − FST)C in island model. How to obtain rb − c ∝ rB − C? Hamilton (1975): “groups break up completely and re-form in each generation”“young animals take off to form a migrant pool”... [one type] assort[s] positively with its own type in settling from the migrant pool(...) to such a degree that the correlation of two separate randomly selected members (...) is F” ⇒ ∆p ∼ (RB − C)pq Fran¸cois Rousset Fitness for the impatient May 2013 21 / 33
  • 37. Fitness costs and benefits Effects B on neighbors’ fecundity and −C on focal’s fecundity rb − c = rB − C? In general No! Local competition ⇒“inclusive fitness”= −C (Taylor, 1992). Actually ∆p ∼ −pq(1 − FST)C in island model. How to obtain rb − c ∝ rB − C? Version with spatially restricted dispersal Individuals disperse as a group and compete as a group against other groups for access to whole group breeding spots. The winners of such group contests can then occupy whole demes ⇒ ∆p ∼ (1 − F)(RB − C)pq (Gardner & West 2006; Lehmann et al. 2006). Fran¸cois Rousset Fitness for the impatient May 2013 21 / 33
  • 38. More general logic: example of dominance Phenotype zi = z + δ[2h(Xi1 + Xi2)/2 + (1 − 2h)Xi1Xi2] To first order, wf(zf, zp) ∼ linear combination of (X1, X2, X1X2, ...) so that E[p ] = mean(wi Xi ) = linear combination of means of (X2 1, X1X2, X1X, X1X1X2, . . . , X1X1X2) = Triplets generally lead to frequency-dependence, although there are intriguing exceptions: Helping among diploid full sibs and among haplodiploid sisters, partial sib mating (α) ∆p ∼ δpq ∂w ∂zf + ∂w ∂zn r f (h, α, /p) Fran¸cois Rousset Fitness for the impatient May 2013 22 / 33
  • 39. Three gene lineages... island model past . . . . . . p p2 p3 Lineages from distinct demes can be considered as draws of independent genes copies, each A with frequency p. Probability that first event is coalescence: FST Fran¸cois Rousset Fitness for the impatient May 2013 23 / 33
  • 40. More general logic: example of kin recognition Demes of N individuals; two loci; indicator variables R, H for alleles Conditional helping: fecundityf = 1 + 1 N − 1 neighbours k [RfRk + (1 − Rf)(1 − Rk)](−CHf + BHk) (two recognition alleles case) wf = (1 − d)ff (1 − d)fdeme + dfothers + d ff fothers (dispersal probability d; regulation after dispersal) mean(H)t+1 = mean(wi Hi ) mean(wi ) = mean(wi Hi ) over individuals i. (strict regulation) Fran¸cois Rousset Fitness for the impatient May 2013 24 / 33
  • 41. Expectedly... To first order in C and B, wf ∼ linear combination of (H, H, RH, RH, RRH, RH, RRH, H, RH, RRH, RH) so that mean(H)t+1 ∼ mean(wi Hi ) = linear combination of expectations of (H2 , HH, RH2 , RH2 , RRH2 , RHH, RRHH, HH, RHH, RRHH, RHH) RH RH R RH H Fran¸cois Rousset Fitness for the impatient May 2013 25 / 33
  • 42. Meaning An act of helping always involves the configuration HR R −C B Fran¸cois Rousset Fitness for the impatient May 2013 26 / 33
  • 43. Meaning An act of helping always involves the configuration HR R −C B RH RH describes the probability that the receiver bears the helping allele; can be computed in terms of the probability of joint coalescence within the deme Fran¸cois Rousset Fitness for the impatient May 2013 26 / 33
  • 44. Meaning An act of helping always involves the configuration HR R −C B R RH H describes the probability that a third individual bears the helping allele. The fitness of this individual is reduced in proportion to (B-C), that is, in proportion to the increase in fitness of the pair of interacting individuals; can be computed in terms of the probability of joint coalescence within the deme Fran¸cois Rousset Fitness for the impatient May 2013 26 / 33
  • 45. Kin recognition: results ∆pH ∼ − C (1 − F) (1 − 2pqR) pqH + −C F + B φ − (1 − m)2 (B − C) 1 N (F + φ) + 1 − 1 N γ ∆pR ∼pHpqR(1 − 2pR) B − C N [Z(N, m) < 0] Polymorphism lost at the recognition locus. Fran¸cois Rousset Fitness for the impatient May 2013 27 / 33
  • 46. Synthesis and developments Many results follow mechanically from a trivial description of allele-frequency changes in terms of fitness functions: * Write a properly defined fitness function in terms of individual behaviour * Express behaviour in term of genotypes (indicator variables) * Expand wi Xi to appropriate order, and take expectations of products of indicator variables (special case: “direct fitness”method, Taylor & Frank 1996) Fran¸cois Rousset Fitness for the impatient May 2013 28 / 33
  • 47. Synthesis and developments Many results follow mechanically from a trivial description of allele-frequency changes in terms of fitness functions: * Write a properly defined fitness function in terms of individual behaviour * Express behaviour in term of genotypes (indicator variables) * Expand wi Xi to appropriate order, and take expectations of products of indicator variables (special case: “direct fitness”method, Taylor & Frank 1996) Deterministic“multi”locus models, with a recombination step: ∆(mean(any thing X)) = ∆sel(mean(X)) + ∆recomb(mean(X)). Multilocus models often in terms of“centered”associations that are 0 in expectation in neutral models, e.g. E[(R − pR)(H − pH)]. Fran¸cois Rousset Fitness for the impatient May 2013 28 / 33
  • 48. Synthesis and developments Many results follow mechanically from a trivial description of allele-frequency changes in terms of fitness functions: * Write a properly defined fitness function in terms of individual behaviour * Express behaviour in term of genotypes (indicator variables) * Expand wi Xi to appropriate order, and take expectations of products of indicator variables (special case: “direct fitness”method, Taylor & Frank 1996) Deterministic“multi”locus models, with a recombination step: Diffusion methods (first order) e.g. approximation for fixation probability π ∼ 1 − e−2Ntotφpini 1 − e−2Ntotφ for φ taken in an infinite-deme limit. Fran¸cois Rousset Fitness for the impatient May 2013 28 / 33
  • 49. Synthesis and developments Many results follow mechanically from a trivial description of allele-frequency changes in terms of fitness functions: * Write a properly defined fitness function in terms of individual behaviour * Express behaviour in term of genotypes (indicator variables) * Expand wi Xi to appropriate order, and take expectations of products of indicator variables (special case: “direct fitness”method, Taylor & Frank 1996) Deterministic“multi”locus models, with a recombination step: Diffusion methods (first order) Diffusion with p-dependence (in first order, e.g. dominance) Fran¸cois Rousset Fitness for the impatient May 2013 28 / 33
  • 50. Synthesis and developments Many results follow mechanically from a trivial description of allele-frequency changes in terms of fitness functions: * Write a properly defined fitness function in terms of individual behaviour * Express behaviour in term of genotypes (indicator variables) * Expand wi Xi to appropriate order, and take expectations of products of indicator variables (special case: “direct fitness”method, Taylor & Frank 1996) Deterministic“multi”locus models, with a recombination step: Diffusion methods (first order) Diffusion with p-dependence (in first order, e.g. dominance) Evolutionary stability Fran¸cois Rousset Fitness for the impatient May 2013 28 / 33
  • 51. Continuous evolutionary stability in the island model Second-order computation for any p feasible. Simpler computation for extreme p, and connexion to“number of successful emigrants”approach (Chesson 1984; Metz and Gyllenberg 2001) B = ∂zf,zf w +2F∂zf,zn w +K∂zn,zn w +4F(N −1) K∂zn wp + F∂fwp wp∂zn w where wp: local offspring In contrast to formula proposed by Day and Taylor (1998): (1) three-genes coefficient K; (2) products of identity coefficients, product of derivatives: joint effect of first-order change in number of offspring and first-order change in parental population structure. Fran¸cois Rousset Fitness for the impatient May 2013 29 / 33
  • 52. Kin recognition: searching for stable polymorphisms Second-order computation involves 30 associations, for up to 7 gene copies in 5 individuals: R RH/R RH/R Search for conditions for stable polymorphism: low migration and low recombination Convergent orbits plus drift: Fran¸cois Rousset Fitness for the impatient May 2013 30 / 33
  • 53. “Conclusions” First order: Simple formalism provides connections between different modelling approaches (inclusive fitness, adaptive dynamics, diffusion, coalescence). Relatively simple pattern of frequency-dependence. Second order: Essentially the same formalism previously used for multilocus models; Algebraically (and algorithmically) messy in spatially structured populations; still allows an analysis of the forces acting on a trait. Fran¸cois Rousset Fitness for the impatient May 2013 31 / 33
  • 54. More general coalescent interpretation of relatedness Fran¸cois Rousset Fitness for the impatient May 2013 32 / 33
  • 55. Quasi equilibrium Quasi linkage equilibrium D(t + 1) = (1 − r)D(t) + O(δ) ⇒ ˆD = O(s) r understood as O(s) r = O(s) 1 − (1 − r) = (1 − r)O(s) + (1 − r)2 O(s) + (1 − r)3 O(s) . . . Quasi equilibrium D(t + 1) − D◦ = λ(D(t) − D◦ ) + O(δ) ⇒ D(t + 1) − D◦ = O(s) 1 − λ Actually D(t + 1) − D◦ = A(D(t) − D◦ ) + O(δ) ⇒ D − D◦ = (I − A)−1 O(s) Fran¸cois Rousset Fitness for the impatient May 2013 33 / 33