Assortative mating by an obliquely transmitted local cultural trait promotes divergence
Justin Yeh University of North Carolina at Chapel Hill Do learned birdsong dialects lead to speciation? • Molecular similarity might correlate with dialect similarity (Baker 1975; Baker & Thompson 1982; Zink & Barrowclough 1984) • Genetic cline overlap with cultural cline, but at lower steepness, and the birdsong does not predict genotype (Kenyon et al. 2016) • Bird lineage with song learning has more species than that without (Lachlan & Servedio 2004)
Justin Yeh 2 correlation ⇏ causation no correlation ⇏ no causation
Justin Yeh 3 Oblique learning: from unrelated individual of the previous generation • Juvenile songbirds learn from adult neighbors • Obliquely transmitted traits are immune to sexual selection (Yeh & Servedio 2015) • If any juvenile can learn from any adult, there should not be correlation between gene and culture
Justin Yeh 4 Spatial structure creates gene- culture association • Learning is spatially restricted • Cultural traits can be locally adapted – High-freq vocal signals in cities (Luther & Derryberry 2012) – Tonal languages in humid environment (Everett et al. 2014)
Justin Yeh 5 Hypothesis
Local genes & Local song & local local oblique reproduction learning
Justin Yeh 6 Hypothesis
Local genes & Local song & local local oblique reproduction learning
Gene-culture association
Justin Yeh 7 Hypothesis
Local genes & Local song & local local oblique reproduction learning
Divergent selection Gene-culture & sexual selection association on song
Genetic divergence
Justin Yeh 8 Hypothesis
Local genes & Local song & local local oblique reproduction learning
Divergent selection Gene-culture & sexual selection association on song
Genetic divergence
Justin Yeh 9 Population genetic model
m
Justin Yeh 10 Population genetic model
m A B a b
Justin Yeh 11 Population genetic model
m A B a b
s H sH
Justin Yeh 12 Population genetic model
C1 C2 m A B a b
s H sH
Justin Yeh 13 Population genetic model
sD , α sD , α
C1 C2 m A B a b
s H sH
Justin Yeh 14 Population genetic model
sD , α sD , α
C1 C2 m A B a b
s H sH
Justin Yeh 15 Population genetic model
sD , α, μ sD , α , μ
N C1 C2 m N A B a b
s H sH
Justin Yeh 16 Simulation conditions
• All genes and cultural traits start diverged – Contact zone • 5000 generations
Justin Yeh 17 Both sexual and divergent selection promote LD
Local gene Local song
Gene-culture Selection on association song (α, sD)
Genetic divergence (LD)
Justin Yeh 18 Both sexual and divergent selection promote LD
m=0.01, sH=0.09, μ=0, no drift
Local gene Local song LD s Gene-culture Selection on D association song (α, sD)
0.01 0.1 1 10 100 Female preference strength (α) Genetic divergence (LD) Note: Intermediate α maximizes sexual selection in phenotype-matching models (Servedio 2011)
Justin Yeh 19 Drift and mutation just make things messy Drift & mutation
Local gene Local song
Gene-culture Selection on association song (sD, α)
Genetic divergence (LD)
Justin Yeh 20 Drift and mutation just make things messy
-3 Drift & mutation m=0.01; sH=0.1; α=1; μ=10 ; N=1000
Local gene Local song LD
Gene-culture Selection on sD association song (sD, α)
Cultural trait divergence (calculated like Nei’s genetic distance) Genetic divergence (LD)
Justin Yeh 21 Mate choice and viability give better evidence than gene-culture correlation • Correlation can be caused by geography • Correlation can be difficult to observe – Oblique learning temporarily and locally removes gene-culture association – Stochastic process may lead to high cultural variation within patches
Justin Yeh 22 Why should cultural evolution care about speciation? • group formation, ideological divide, radicalization, etc. • Reduced genetic exchange <--> reduced memetic exchange • Genetic divergence <--> cultural divergence
Justin Yeh 23 Acknowledgement
• Maria Servedio • Marc Feldman • Courtney Fitzpatrick • National Science Foundation (USA)
Justin Yeh 24 Life cycle
Drift
Oblique learning with Migration cultural mutation
Viability Sexual selection selection & reproduction
Justin Yeh 25 Drift
Migration Oblique learning w/ cultural mutation
Sexual selection & Viability selection reproduction
• D-M incompatibility for A and B (sH)
• Divergent selection on C (sD)
Genotype Culture Patch 1 Patch 2
AB & ab C1, C3… 1 1-sD
Ab & aB C1, C3… 1-sH (1-sH)(1-sD)
AB & ab C2, C4… 1-sD 1
Ab & aB C2, C4… (1-sH)(1-sD) 1-sH
Justin Yeh 26 Drift
Migration Oblique learning w/ cultural mutation
Viability Sexual selection & selection reproduction • Polygyny (sexual selection on males) • Preference (α) to mate with individuals of the same cultural trait (phenotype-matching) ♂ ♀ C1 C2 C3 …
C1 1+α 1 1 1
C2 1 1+α 1 1
C3 1 1 1+α 1 … 1 1 1 … 27 Drift Oblique learning w/ Migration cultural mutation
Viability Sexual selection selection & reproduction • From the previous generation in the same patch
f(ABC1)juvenile=f(AB)juvenile((1- μ) f(C1)adult + μ f(C2)adult)
• Gene-culture association becomes 0 within each patch!
Justin Yeh 28 Life cycle order & number of possible cultural traits
-3 m=0.01; sH=0.1; α=1; μ=10 ; N=1000
m=0.01; sH=0.1; sD=0-0.5; α=1 # of C alleles
LD
sD
Cultural trait divergence
Drift
Viability Oblique learning selection
Migration Sexual selection
29 sD selection
Local gene Local song Local gene Local song migration migration
Gene-culture Gene-culture α, s α association D association selection
Genetic divergence (LD) Genetic divergence (LD)
Justin Yeh 30 Scientific Empirical Proof-of-concept method research models
Observe nature and build hypothesis
Choose Design protocol Gather evidence assumptions and gather data and build model
Analyze evidence Analyze data Analyze model
Evaluate hypothesis and propose new directions
Servedio et al. (2014) Not Just a Theory—The Utility of Mathematical Models in Evolutionary Biology. PLoS Biol 12(12): e1002017