Differential introgression between two Iberian Podarcis lizards

Abstract

Genomic analysis have demonstrated that many hybrid zones are semipermeable boundaries where introgression can be highly variable, with some alleles introgressing and recombining into genomic background of the other species while others are less able to cross the species boundary. This way we can infer which fraction of the genome is counter selected in the other species and which fraction does not participate in reproductive isolation. We investigated this phenomenon across a contact zone where two Iberian lizards, Podarcis bocagei and P. carbonelli, meet and hybridize. We employed a genomic RAD sequencing approach to evaluate how species boundaries are maintained in syntopy using a wider geographical sampling to elucidate the genomic signature of past introgression. Geographic and genomic cline analyses revealed variable patterns introgression among loci. Most alleles do not cross the species boundaries and the hybrid zone is mainly bimodal, suggesting well-established barriers to gene exchange. However, several loci were able to introgress, mainly from P. bocagei into P. carbonelli, and have reached populations away from the contact zone. Mapping those loci onto the Podarcis muralis genome should enable us to test whether they correspond to a particular genomic region and potentially link with traits related to reproductive isolation.

Full text

Fig. 6 – Species distribution
map with HZAR[9] geographic
cline represented. ADMIXTURE
assignments for both SNP
dataset were used as hybrid
index (HI) leading to similar
results. Top bar represents HI.
00.2 0.6 1
0.8
0.4
Cline
width:
0.2-3.2Km!
0.08–3.2Km introgression
tail to south
Cline center!
12–14Km !
Almost no introgression
tail to north
20!
162!
23!
18!
26!
32!
26!
P.#bocagei#
P.#carbonelli#
23!
!
!
!
!
!
!
!
!
Introgression!
Genomic cline parameters
α
: cline center (probability of ancestry); and
β
: introgression rate.
Pb ancestry (
α
>>0) for 1031 loci (44.8%); Pc ancestry (PbA,
α
<<0) for
1148 loci (49.9%); 101 loci (4.4%) with no specific ancestry (
α
≈0).
960 loci (41.7%) shown restricted introgression (
β
>>0), 619 loci
(26.9%) with increased introgression (
β
<<0) and 721 loci (31.3%)
followed the neutral expectations (
β
≈0).
No outlier loci were detected for both
α
and
β
.
Main Conclusions
Are results similar to previous studies?
Yes: Bimodal HZ; No: Structure detected in HZ.
Is hybridization introgressive? Yes.
Similar introgression across genome? No; Significant
nr of loci involved in reproductive isolation.
Direction of introgression? Mostly Pb è Pc.
Kinds of selection present? Selection against heterozygote genotypes.
Ongoing Work
Estimate geographic cline for each loci to compare with genomic clines.
Background Information
A narrow hybrid zone (HZ) is known between Podarcis bocagei (Pb) and
P. carbonelli (Pc).
Previous studies in this HZ show low hybridization, limited
introgression for nearby populations, evidence for a bimodal hybrid
zone [1,2] and no intermediate morphology [1,3,4].
Few loci were used (<22).
Objectives
Are results similar to previous studies?
Is hybridization introgressive?
Similar introgression across genome?
Which is the direction of introgression?
Which kinds of selection are present?
Sampling and SNP data
Double digestion RAD library construction
with 330 individuals.
Complete dataset with 6905 loci after
demultiplexing and filtering.
“Disgnostic” dataset with 2300 loci with frequency
>0.8 in one parental and <0.2 in other.
Structure of the contact zone
1 CIBIO-InBio, Centro de Investigação em Biodiversidade e Recursos Genéticos – Laboratório Associado,
Universidade do Porto, Vairão, Portugal
2 Departamento de Biologia, Faculdade de Ciências da Universidade do Porto, Porto, Portugal
3 Department of Ecology and Evolution, University of Lausanne, Lausanne, Switzerland
4 Biology Department, University of California Riverside, California, USA
5 CEFE UMR 5175, CNRS – University. Montpellier – University Paul-Valery Montpellier – EPHE, Montpellier,
France
Literature cited
[1]Pinho C, Kaliontzopoulou A, Carretero MA, Harris DJ, Ferrand N (2009) Genetic admixture between the
Iberian endemic lizards Podarcis bocagei and P. carbonelli: evidence for limited natural hybridization and a
bimodal hybrid zone. 󲣱J Zoolog Syst Evol Res, 47(4), 368-377.
[2]Ribeiro, M (2014) A landscape genetics perspective on the spatial dynamics of hybridization between two
species of wall lizards (MSc Thesis), Faculty of Sciences of University of Porto.
[3]Carretero, MA, Sá-Sousa P, Barbosa D, Harris DJ, Pinho C (2002) Sintopía estricta entre Podarcis bocagei
y P. carbonelli. Boletín de la Asociación Herpetológica Española, 13(1-2), 20-24.
[4]Kaliontzopoulou A, Carretero MA, Llorente GA (2007) Multivariate and geometric morphometrics in the
analysis of sexual dimorphism variation in Podarcis lizards. J Morphol, 268(2), 152-165.
[5]Jombart T (2008) adegenet: a R package for the multivariate analysis of genetic markers. Bioinformatics,
24(11), 1403-1405.
[6]Corander J, Marttinen P, Sirén J, Tang J. (2008) Enhanced Bayesian modelling in BAPS software for
learning genetic structures of populations. BMC Bioinformatics, 9(1), 539.
[7]Alexander DH, Novembre J, Lange K (2009) Fast model-based estimation of ancestry in unrelated
individuals. Genome Res, 19(9), 1655-1664.
[8]Gompert Z, Buerkle, CA (2012) bgc: Software for Bayesian estimation of genomic clines. Mol Ecol Res,
12(6), 1168-1176.
[9]Derryberry GE, Maley JM, Brumfield RT (2014) HZAR: hybrid zone analysis using an R software package.
Mol Ecol Res, 14(3), 652-663.
Differen-al!introgression!between!two!Iberian!
Podarcis!lizards!
POR!MAD!
FRA!AGU! ESP! ESM! TOR!
SIL!
Fig. 3 - Results from
individual multilocus
genotype clustering
computed with
ADMIXTURE[7] for the 6905
loci. Similar results were
obtained for both datasets
and with BAPS. Each
individual is represented by
a vertical line proportionally
partitioned into the K
colored segments. Top chart
represents K=2 and bottom
chart K=3. Population
acronyms as in Fig. 1.
Fig. 2 - Principal Component Analysis of 6906 SNP variation in the 330 individuals
calculated with ADEGENET[5] R package. Circles represent individuals from populations
north of the contact zone, triangles correspond to the individuals from the contact zone,
squares identify the individuals from populations south of the contact zone and stars
represent the admixed individuals identified with BAPS[6]. Population acronyms as in
Fig. 1.
Guilherme Caeiro-Dias1,2,5, Alan Brelsford3,4, Pierre André-Crochet5, Mariana Ribeiro1,2, Catarina Pinho1
Fig. 1 – Species distribution in the study area and sampling
location along the north-south transept with number of
samples for each one.
Study of Hybrid Zones
Understand
mechanisms
promoting
Speciation Maintenance of
species boundaries
Differentiation
High-density SNP genotyping
through Restriction Site
Associated DNA (RAD)
sequencing to evaluate
Introgression
Hybridization
Pre-zygotic
barriers to
gene flow
Post-zygotic
barriers to
gene flow
No
Yes
Extensive genomic introgression
No
Yes
Introgression restricted to few loci
Neutral introgression
Directional introgression
Genome-scale
analysis
20 Admixed
individuals
identified
Most variability
was found
between the
two species!
Some variability
between Pc from
Espinho and other Pc
populations!
Acknowledgments
Elena Pérez, Miguel Carretero, Antigoni Kaliontzopoulou
Ana Perera, Neftalí Sillero, Angelica Crottini for help in the field work.
Host institutions Financial support
Fig. 4 – Variable introgression between Pb and Pc based on
α
and
β
computed with
BGC[8]. Each circle represents one of the 2300 loci. Main genomic clines representative
of loci in the corresponding region of the
α
−
β
parameter space are shown in red.
α
<<0;
β
>>0
Pc ancestry
Restricted introgression
Population structure
may overestimate β
α
>>0;
β
<<0
Pb ancestry
High introgression!
α
>>0;
β
≈0
Pb ancestry
Neutral introgression!
α
≈0;
β
≈0
Neutral loci
β
α
3%
8.4%
20%
23.6%
39.6%
23.6%!
3%!
39.6%!
Neutral
loci
20%!
Selection against
heterozygote
genotypes
Directional
introgression!
Barriers to
gene flow!
Fig. 5 – Schematic
representation of
variable introgression
(arrows) and
distribution of each
individual in Espinho
(map on the right;
symbols and colors as
Fig. 2).
8.4%!
1.3%!
1.9%!