Reproductive isolation between chromosomal races of the house mouse Mus musculus domesticus in a...

Reproductive isolation between chromosomal races of the housemouse Mus musculus domesticus in a parapatric contact arearevealed by an analysis of multiple unlinked loci

P. FRANCHINI, R. CASTIGLIA & E. CAPANNA

Dipartimento di Biologia Animale e dell’Uomo, University of Rome ‘‘La Sapienza’’, Rome, Italy

Introduction

Several authors have stressed the importance of chro-

mosomal rearrangements [e.g. Robertsonian (Rb) fusions

and fissions, translocations and inversions] in speciation

(White, 1978; King, 1993).

In the western house mouse, Mus musculus domesticus,

the repeated occurrence and rapid fixation of Rb fusions

have led to the formation of several chromosomal races,

providing an excellent model to investigate this process

(Sage et al., 1993).

Fusion between different telocentric chromosomes

and, more rarely, whole-arm reciprocal translocations

(WARTs) have produced 97 populations with distinct

combinations of metacentrics, with a diploid number

reduced to 2n = 22 from the standard 2n = 40 typical of

the species (Capanna, 1982; Pialek et al., 2005). The

rearranged races are scattered within the distribution

area of M. m. domesticus, i.e. western Europe and North

Africa.

Since the discovery of the first race in the Swiss-Italian

Alps (Gropp et al., 1969), several studies have been

carried out on the house mouse, with particular attention

Correspondence: Paolo Franchini, Dipartimento di Biologia Animale e

dell’Uomo, University of Rome ‘‘La Sapienza’’, Via A. Borelli 50, 00161

Rome, Italy.

Tel.: +39 06 4991 8022; fax: +39 06 4457516;

e-mail: [email protected]

ª 2 0 0 8 T H E A U T H O R S . J . E V O L . B I O L . 2 1 ( 2 0 0 8 ) 5 0 2 – 5 1 3

502 J O U R N A L C O M P I L A T I O N ª 2 0 0 8 E U R O P E A N S O C I E T Y F O R E V O L U T I O N A R Y B I O L O G Y

Keywords:

chromosomal races;

gene flow;

house mouse;

hybrid zone;

microsatellite;

reproductive isolation.

Abstract

The house mouse, Mus musculus domesticus, exhibits a high level of chromo-

somal polymorphism because of the occurrence and fast fixation of Robert-

sonian fusions between telocentric chromosomes. For this reason, it has been

considered a classical speciation model to analyse the role of the chromosomal

changes in reproductive isolation. In this study, we analysed a parapatric

contact area between two metacentric races in central Italy, the Cittaducale

race (CD: 2n = 22) and the Ancarano race (ACR: 2n = 24), to estimate gene

flow at the boundary. Hybrids between these two races show high levels of

structural heterozygosity and are expected to be highly infertile. A sample of

88 mice from 14 sites was used. The mice were genotyped by means of eight

microsatellite loci mapped in four different autosomal arms. The results show

clear genetic differentiation between the CD and ACR races, as revealed by

differences in allele frequencies, factorial correspondence analysis and indexes

of genetic population (e.g. FST and RST) along the contact zone. The genetic

differentiation between the races was further highlighted by assignation and

clustering analyses, in which all the individuals were correctly assigned by

their genotypes to the source chromosomal race. This result is particularly

interesting in view of the absence of any geographical or ecological barrier in

the parapatric contact zone, which occurs within a village. In these conditions,

the observed genetic separation suggests an absence of gene flow between the

races. The CD–ACR contact area is a rare example of a final stage of speciation

between chromosomal races of rodents because of their chromosomal

incompatibility.

doi: 10.1111/j.1420-9101.2007.01492.x

to the contact zones between Rb races and between Rb

races and surrounding standard populations (Castiglia &

Capanna, 1999a; Britton-Davidian et al., 2000). These

zones represent a perfect model to study ongoing

chromosomal speciation, based on the evidence that

structural heterozygosity plays a significant role in

the reduction of fitness of hybrid individuals (Redi &

Capanna, 1988; Hauffe & Searle, 1998; Castiglia &

Capanna, 2000). In this situation, gene exchange should

be reduced and the barriers to gene flow should depend

on the karyotype compatibility of the races involved.

Furthermore, gene flow can be influenced by chiasma

repatterning, leading to recombination suppression of

pericentromeric zones of hybrids in structural heteroz-

ygosis (Bidau et al., 2001; Castiglia & Capanna, 2002;

Panithanarak et al., 2004).

Many studies have indirectly investigated gene flow

through hybrid fertility. However, with the diffusion of

molecular techniques, mitochondrial and nuclear DNA

are now the best tools to estimate the amount of gene

exchange between taxonomical units. Recent studies

(Dallas et al., 1998; Panithanarak et al., 2004) have

shown that microsatellites, i.e. nuclear markers charac-

terized by a high mutation rate, are an ideal instrument

to genetically compare the Rb races of this house mouse

subspecies, which presumably have appeared in the last

10 000–3000 years (Auffray et al., 1990; Cucchi et al.,

2005).

In this study, we used microsatellite loci to analyse a

well-known contact area between two metacentric races

in central Italy, the Cittaducale race (CD: 2n = 22) and

the Ancarano race (ACR: 2n = 24). Background studies

of this area were based on cytogenetics (Castiglia &

Capanna, 1999b), mtDNA control region analysis

(Castiglia et al., 2002) and behavioural experiments

(Carpineti & Castiglia, 2004). The CD and ACR races,

which share no metacentrics, represent an extreme case

of diversification between karyotypes in Mus: the CD

karyotype is characterized by the Rb chromosomes 1.7,

2.18, 3.8, 4.15, 5.17, 6.13, 9.16, 10.11, 12.14 (where the

nomenclature ‘x.x’ indicates the telocentric chromo-

somes involved in the centric fusion) and the telocentric

19; the ACR karyotype is characterized by the Rb 1.2, 3.9,

4.17, 5.13, 6.16, 8.14, 10.12, 11.15 and the telocentrics 7,

18, 19 (Fig. 1). These karyotypes lead to the formation in

first generation hybrids of a long multivalent ring of 14

metacentrics and a shorter chain of five chromosomes

(Castiglia & Capanna, 1999b).

A preliminary study based on electron microscopic

observation of the testes suggested the sterility of F1

males (Malorni et al., 1982), although the low number of

animals analysed is inadequate to definitively confirm

this event. In fact, the sterility of some hybrid males may

be detected among a larger sample of fertile ones, as

revealed by Chatti et al. (2005) in a hybrid zone in

Tunisia. Furthermore, the female fertility has not yet

been analysed.

Previous cytogenetic characterization of the present

sample revealed the parapatric distribution of the two

karyotypes, except for the occurrence of one hybrid with

an ACR chromosomal background presenting two differ-

ent metacentrics (Castiglia & Capanna, 1999b): one typical

of the CD race (6.13) and one arising by WART (5.16).

Analysis of the mtDNA control region clearly showed the

genetic differentiation of the two races, with the near

absence of shared haplotypes (Castiglia et al., 2002).

The current situation suggests reduced gene flow

between the chromosomal races, although sporadic

hybridization events observed prevent us from establish-

ing the precise strength of the barrier to gene exchange.

At present, there are limitations to the estimation of

gene flow by mtDNA analysis and by cytogenetic

comparisons, because of certain features and mecha-

nisms of these methods. The mtDNA-based techniques

only investigate the maternal lineages and the mutation

rate of mitochondrial markers is slow compared with the

extremely rapid raciation process of M. m. domesticus.

Moreover, trying to detect a hybridization event by

cytogenetic analysis alone could lead to underestimation

of the effective gene exchange. In fact, segregation

events occurring during the meiosis of F1 hybrids may

separate the chromosomes involved in the ring and

in the chain, leading to the formation of gametes

bearing the parental karyotypes. These events can

restore parental karyotypes even in the first backcross

generation.

The aim of this study was to obtain an accurate

estimate of the gene flow between the ACR and CD races

through microsatellite analysis. The karyotype incom-

patibility is a good reason to assume a reduction but not

complete absence of gene flow between the races. In

another well-known Rb system, i.e. Sorex araneus, An-

dersson et al. (2004) and Basset et al. (2006) used

microsatellites to show the occurrence of gene flow

despite the high level of structural heterozygosity in the

F1 hybrids. Therefore, the mechanisms supporting gene

flow in such chromosomal hybrid zones must still be

fully clarified.

This case-study deals with a key issue in the chromo-

somal speciation of M. m. domesticus.

Materials and methods

Sampling and molecular methods

Eighty-eight mice were collected between June 1998 and

March 2000 from 14 localities along the Aterno River, in

an Apennine valley in central Italy (Fig. 2). According to

the karyotype analysis performed during the sampling

period, the specimens were assigned to the CD or ACR

chromosomal race (Castiglia et al., 2002). As the two

races exhibit parapatric contact, only CD or ACR spec-

imens were found in each locality. In some analyses, the

individuals of several sample sites were grouped together

Contact area between chromosomal races 503


J O U R N A L C O M P I L A T I O N ª 2 0 0 8 E U R O P E A N S O C I E T Y F O R E V O L U T I O N A R Y B I O L O G Y

to increase the population size for the microsatellite

analysis (Table 1, Fig. 2). Groupings were made on the

basis of proximity and habitat continuity. We refer to

these groups of sites as ‘populations’ throughout the

paper. Genetic analyses were performed on six popula-

tions, four with the ACR karyotype (CAP, CAS, MAR and

PizACR) and two with the CD karyotype (PizCD and

SVT) (Table 1, Fig. 2). The mice were trapped using

Sherman live traps (H.B. Sherman Traps, Inc., Tallahas-

see, FL, USA) and tissue samples were stored in ethanol

(80%) at 4 �C for DNA extraction.

Eight microsatellite loci in four different autosomal

arms were selected from the mouse genome database of

the Whitehead Institute for Biomedical Research (http://

www.genome.wi.mit.edu): D6Mit138, D6Mit26,

D9Mit251, D9Mit52, D12Mit182, D12Mit30, D19Mit6

and D19Mit94. They are named ‘DXMitY’ according to

the nomenclature suggested by Davisson (1996), where

‘X’ indicates the chromosome arm number and Y the

locus number.

Total genomic DNA was extracted from tissue samples

(kidney or muscle) with the DNeasy Tissue Kit (Qiagen,

Valencia, CA, USA). Polymerase chain reaction (PCR)

was performed in a 10 lL final volume containing: 1 lL

template DNA (�25 ng), 1 lL 10x Tris buffer, 0.6 lL

MgCl2 25 mMM, 1 lL dNTP mix 2.5 mMM, 1 lL 5 lMM of each

primer and 0.25 U Taq Polymerase (Promega, Madison,

USA). Amplifications were carried out in a Thermocycler

T1 (Biometra, Gottingen, Germany) using the following

conditions: one cycle of 4 min at 94 �C (initial denatur-

ation); 30 cycles of 1 min at 94 �C (denaturation), 45 s at

55–57 �C (depending on the primer pair annealing

temperature) and 45 s at 72 �C; ending with 10 min at

72 �C.

The PCR products were run on 6.5% denaturing

polyacrylamide gels using a LI-COR DNA 4200 auto-

mated sequencer (LI-COR Biosciences, Cambridge, UK).

Allele sizes were estimated using GENEENEIMAGEMAGEIR v4.05

software (Scanalytics, Fairfax, VA, USA) by comparison

with infrared molecular size standards for LI-COR DNA

analysers.

Statistical analysis

Genotype frequencies were tested for departures from

Hardy–Weinberg (H–W) equilibrium expectations with

Fig. 1 G-banded karyotypes of the CD race (2n = 22) and of the ACR race (2n = 24).

504 P. FRANCHINI ET AL.



Fisher’s exact test (Louis & Dempster, 1987) using GENE-GENE-

POPPOP v3.4 (Raymond & Rousset, 1995). GENEPOPGENEPOP was also

used to calculate linkage disequilibrium between all pairs

of loci according to Goudet et al. (1996). Significance

levels were calculated by the Markov Chain method

(Guo & Thompson, 1992) using 10 000 permutations,

5000 dememorization steps and 500 batches.

The inter-population level of genetic differentiation

was analysed with the F- and R-statistic. FST values,

estimated by theta (h), were calculated according to Weir

& Cockerham (1984) using FSTATFSTAT v2.9.3.2 (Goudet,

2001). RST values, estimated by rho (q), were calculated

according to Slatkin (1995) using RSTCALCRSTCALC v2.2 (Good-

man, 1997). A statistical analysis of the single locus G

values was used to test the significance of FST over all loci

(Reich et al., 1999). For the significance of pairwise FST

values, we applied a test based on the Markov Chain

method using 10 000 permutations, 5000 dememoriza-

tion steps and 500 batches. Significance levels of RST

values were calculated in a bootstrap procedure using

1000 permutations.

The correlation between geographical and genetic

distances was analysed with a Mantel test (‘Isolde’

subroutine, GENEPOP)GENEPOP). The value FST ⁄ (1 ) FST) was com-

pared with the natural logarithm of the distance

(Rousset, 1997).

To describe population differentiation, we performed a

factorial correspondence analysis (FCA) on all genotypes

using GENETIXGENETIX v4.05 (Belkhir et al., 2004). FCA is a

statistical method allowing to condense the information

contained in a contingency data table by representing set

of objects and set of variables, graphically and synthet-

ically, in a lower dimensional space, denominated facto-

rial space (Benzecri, 1973).

Fig. 2 Location of the 14 sample sites along the contact zone. In Table 1, sample sites and karyotype of each site are described. Pooled sites are

grouped into circles (see text for detailed explanation).

Table 1 Localities, sample size (in brackets), populations and

chromosomal characteristic of mice from the contact zone.

Capture sites

number Locality Population

Chromosomal

race

1 Marana (4) MAR (4) ACR, 2n = 24

2 Mopolino (2) CAP (9) ACR, 2n = 24

3 Capitignano (7)

4 Cascina (20) CAS (20) ACR, 2n = 24

5 Barete (3) PizACR (33) ACR, 2n = 24

6 Pizzoli 8 (8)

7 Pizzoli 9 (10)

8 Pizzoli 10 (9)

9 Pizzoli 11 (3)

10 Pizzoli 13 (5) PizCD (19) CD, 2n = 22

11 Pizzoli 14 (5)

12 Villa S.Pietro (6)

13 San Lorenzo (3)

14 San Vittorino (3) SVT (3) CD, 2n = 22




Assignment tests and clustering analyses were per-

formed to identify a genetic structure in the dataset. An

assignment test was carried out to assign the individuals

of the entire sample to the different populations and, in a

further level of analysis, to the chromosomal races. The

software GENECLASSENECLASS v2.0 (Piry et al., 2004) was used to

compute the probability that an individual belonged to

each reference group, in this case populations and races

respectively. Each individual was then assigned to its

more probable group. For each group, the software

simulated a large population of multilocus genotypes

according to the allele frequency in the sample. The

analysis was performed using a Bayesian algorithm based

on the maximum likelihood method. For a more precise

computation, the number of simulated individuals was

set at 10 000. The assignment threshold was set at 0.05 to

assign each individual to a group only when the

probability of it belonging to that group was > 5%.

STRUCTURE VTRUCTURE V2.1 (Pritchard et al., 2000; Falush et al.,

2003) was used to infer the presence of genetic clustering

in the sample. The analysis was performed without prior

information about the population structure using the

admixture model used in the software. The model is

based on the evidence that individuals may have mixed

ancestry, so that they inherit some fraction of their

genome from ancestors in population K. The program

assumes K clusters and, with a Bayesian-based algorithm,

all individuals were assigned to produce H–W equilib-

rium and the absence of linkage disequilibrium within

the hypothesized K groups. The optimal values of K were

selected according to the formula suggested by Evanno

et al. (2005): DK = m|L¢(K)| ⁄ s[L(K)]. This procedure

allows one to choose the cluster subdivision that captures

the uppermost level of genetic structure. The final step

was to assign the average proportion of membership (Qi)

of the sampled population to each of the inferred clusters,

setting a threshold Qi > 0.90. When an individual had

Qi < 0.90, it was considered admixed jointly to two or

more clusters. To select an appropriate burning length, so

as to obtain accurate parameter estimates, several runs at

each K (K = 1–6) were performed. The analysis was

conducted with the default option of the admixture

model used by STRUCTURETRUCTURE, with a burn-in chain

length = 1 000 000 and a Markov Chain Monte Carlo

length = 1 000 000.

An accurate estimate of gene flow was obtained through

the use of BAYESAYESASS+SS+ v1.3 (Wilson & Rannala, 2003). The

Bayesian approach implemented in the software allows

computing the asymmetrical migration rates, m, between

the races and among the populations without assuming

H–W equilibrium within them. The program was run with

1 000 000 iterations of the Monte Carlo Markov Chains

(with a length of the burn-in of 1 000 000). To infer

posterior probability distributions of the migration rates,

samples were collected every 2000 iterations.

The detection of hybrids in the sample was also

investigated by the software NEWEWHYBRIDSYBRIDS v1.11.1 (Ander-

son & Thompson, 2002).. As revealed by Vaha & Primmer

(2006), NEWEWHYBRIDSYBRIDS is more indicated than STRUCTURETRUCTURE

to analyse hybridization when both F1 and backcross

hybrids are present in a dataset. NEWEWHYBRIDSYBRIDS computes

the posterior probabilities that each individual of the

sample belongs into different parental ⁄ hybrid categories.

The six genotypic classes hypothesized are constituted by

the two parental types (Pure CD and Pure ACR), F1 and

F2 hybrids, backcrosses of F1 with the first parental (B1)

and backcrosses of F1 with the second parental (B2). The

probabilities of belonging to one of these classes were

evaluated after 100 000 iterations of the Monte Carlo

Markov Chains, without using any individual or allele

frequency prior information. The threshold probability to

assign individuals to one of the six classes was set to

P > 0.90.

A hierarchical differentiation analysis (AMOVAAMOVA) was

performed using Wright’s F-statistics (Wright, 1951) as

implemented in ARLEQUINRLEQUIN v3.0 (Excoffier et al., 2005). In

this analysis, the geographical populations were clustered

by chromosomal races (CD and ACR).

Results

Microsatellite genotyping for the eight loci was successful

for the entire sample (see Appendix A).

The intra-population level of analysis showed that the

observed heterozygosity over all loci was rather high in

all populations, ranging from 0.46 in PizCD to 0.70 in

SVT (Appendix A). The exact test for H–W equilibrium

for each locus in each population showed a significant

deviation from expectations at loci D9Mit52 and

D12Mit30 in PizACR and at locus D19Mit6 in PizCD

(tablewide Bonferroni corrected level of a = 0.0011).

After Bonferroni correction, there was no evidence of

genotypic linkage disequilibrium at any pair of loci (all

P > 0.0003).

The allele frequencies at each of the 14 sample sites

revealed some ‘race-specific’ alleles that characterize the

CD and ACR races (Appendix A). The trends of the five

alleles of locus D12Mit182 (the locus with the highest

genetic variation at the ‘among races’ level: AMOVAAMOVA locus

by locus, data not shown) clearly discriminated between

the two chromosomal races (Fig. 3). Alleles 152 and 154

were exclusive to the ACR race, while alleles 142 and 144

were exclusive to the CD race. Allele 140 was present in

both races, but with a much higher frequency in CD

(24.2% in CD race, 72.7% in ACR race).

The FST and RST values over all geographic populations

were high and significant (FST = 0.187 and RST = 0.210),

indicating strong genetic differentiation among the pop-

ulations. As shown in Table 2, the results for all pairwise

comparisons differed significantly from zero (Bonferroni

correction applied: P = 0.003), except between SVT and

PizCD (considering FST values) and among SVT–PizCD,

CAP–SVT, CAP–PizACR and MAR–CAP (considering RST

values). There was no correlation between the genetic




and geographic distances [regression of FST ⁄ (1 ) FST)

over geographic distance, Mantel test, P > 0.05] for the

total set and for the ACR race.

As described in Materials and methods, a factorial

correspondence analysis of microsatellite genotypes was

performed to investigate population differentiation. The

individual genotypes of the ACR and CD populations

were grouped separately on the left and right side of the

FCA plot (Fig. 4), along the axis that explains most of the

genetic variation (10.4%).

The assignment test performed by GENECLASSENECLASS showed

a high percentage of correctly assigned individuals in all

populations, ranging between 33.3% for individuals from

SVT (low value possibly explained by the small popula-

tion size) and 100% for individuals from CAS and MAR

(Table 3). The total value of 89.7% of correct assign-

ments, supported by a high quality index value of 88.4%

(an estimate of the test significance that indicates the

cumulative measure of the percentage assignment of

each individual to a group chosen a priori), showed

a level of genetic differentiation constituted by the

populations. Most significantly, the analysis revealed

another level of genetic separation constituted by the

chromosomal races: there were no bad assignments from

a population of one chromosomal race to a population of

the other race. In the light of this evidence, the

assignment test was also performed on the two chromo-

somal races. In this approach, the analysis was applied to

a dataset in which each specimen was assigned to the

source chromosomal race. The strong genetic subdivision

between the two groups of specimens highlighted in the

previous analysis was confirmed. In fact, all individuals

(100%) were correctly assigned to the source race, with a

high quality index value (95.6%).

The STRUCTURETRUCTURE software revealed strong genetic struc-

turing in the dataset. The posterior probability of the

Fig. 3 Allele frequencies at the locus D12Mit182 for each of the 14

sample sites.

Table 2 FST- (below diagonal) and RST-values (above diagonal) for

each pair of populations.

SVT PizCD PizACR CAP CAS MAR

SVT – )0.043 0.201 0.099 0.282 0.305

PizCD 0.107 – 0.258 0.187 0.384 0.419

PizACR 0.183 0.166 – 0.136 0.110 0.075

CAP 0.150 0.154 0.124 – 0.161 0.277

CAS 0.240 0.254 0.156 0.119 – 0.299

MAR 0.225 0.245 0.154 0.215 0.215 –

Significance differences are marked in bold (P < 0.003, corrected

threshold through the Bonferroni procedure).

Axis 1 (10.4%)

Axi

s 2

(6.

7%)

0 1 2–1

0

1

–1

SVTPizCDPizACRCAPCASMAR

CD raceACR race

Fig. 4 Factorial correspondence analysis (FCA) of the individual

mouse genotypes computed using eight microsatellite loci; the graph

shows the first two factorial components, FCI and FCII.

Table 3 Number of individuals in each population assigned to the

source population (in bold) or to other populations.

Population

Assignation populationCorrect

assignations (%)SVT PizCD PizACR CAP CAS MAR

SVT 1 2 – – – – 33.3

PizCD 2 16 – – – – 88.8

PizACR – – 30 2 2 – 88.2

CAP – – 1 8 – – 88.8

CAS – – – – 20 – 100

MAR – – – – – 4 100




data, P(D), increased from K = 1 to K = 5, where it

reached its maximum value and exhibited a plateau,

showing genetic differentiation between most of the

sampling sites. Further confirmation of a structure came

from the asymmetric proportion of the sample assigned

to each population (P „ 1 ⁄ K). Evanno et al. (2005)

formula DK = m|L¢(K)| ⁄ s[L(K)] indicates that splitting of

the samples into two or three groups should represent

the optimal subdivision of the data and that further

clustering could correspond to unjustified oversplitting

(Fig. 5). This means that the information included in the

K > 3 clusters is less important than the information in

the first two. As shown by the bar-plots (Fig. 6), when K

was set at 2, there was a high percentage of assignment of

each individual, Qi (average Qi of all individuals = 0.98),

to one of the two clusters, as well as full correspondence

of the inferred clusters with the chromosomal races. A

threshold probability Qi = 0.90 excluded only one spec-

imen from the ACR cluster, belonging to CAP, even

though the assignment was stronger for the source

chromosomal race (Qi = 0.67) than for the CD one.

When three clusters were hypothesized, one of them still

showed full correspondence with the CD race, while the

other two consisted of several admixed individuals,

showing a weak genetic structure within the ACR race.

Increasing the number of K resulted in a weaker genetic

structure within the ACR race and the absence of

individuals with admixed genotypes between the two

chromosomal races.

The analysis performed by the software BAYESAYESASS+SS+

detected a low migration rate among all the populations,

except for the two ones of the CD race (Table 4), where

the migration rate is 0.13 (from SVT to PizCD). It should

be noted that gene flow is also nearly absent between the

two parapatric populations PizCD and PizACR (m < 0.03

in each direction: PizCDMPizACR).

A specialized approach to detect hybridization through

the algorithms implemented in NEWEWHYBRIDSYBRIDS was used.

The analysis performed by the program showed similar

results to that obtained by STRUCTURETRUCTURE. Eighty-four of the

total eighty-eight individuals were assigned to the source

parental race with a probability (pA) higher than 0.90

(Fig. 7a). As revealed by the box-plots in Fig. 7b, the four

2 3 4 5 6 7 8 9 10 11 12 13 14 150

2

4

6

8

10

12

K

K

Fig. 5 DK calculated as DK = m|L¢(K)| ⁄ s[L(K)]. The modal value of

this distribution is the true K. As shown in the graph, K = 2 and

K = 3 correspond to the uppermost levels of structure.

SVT PizACR CAP CAS MAR

CD Group ACR Group

K = 2

K = 4

K = 3

K = 5

PizCD

SVT PizACR CAP CAS MARPizCD



Fig. 6 Assignment graphs based on the individual genotype built by

STRUCTURETRUCTURE software through the admixture model. Each individual is

represented by a single vertical bar broken in K tinted segments, with

lengths proportional to each of the K inferred clusters. The populations

are shown to identify the individuals, as the clustering analysis infers a

genetic structure from a genotype dataset alone, without inclusion of

the source population in the input file.

Table 4 Mean values of the posterior distributions of m, the

migration rate into each population.

Recipient

population

Source population

SVT PizCD PizACR CAP CAS MAR

SVT 0.746 0.003 0.003 0.007 0.003 0.021

PizCD 0.125 0.981 0.003 0.008 0.003 0.025

PizACR 0.031 0.003 0.961 0.012 0.003 0.024

CAP 0.031 0.003 0.013 0.950 0.003 0.039

CAS 0.028 0.004 0.014 0.010 0.982 0.051

MAR 0.037 0.003 0.004 0.010 0.003 0.837

Source populations (the origin of the migrants) are listed in rows,

while recipient populations (population in which individuals are

migrating) are listed in the columns. Migration rates ‡ 0.100 are in

bold. Standard deviations for all the mean distributions are < 0.050.




‘admixed genotype’ individuals can no considered

hybrids because of the low assignation coefficient to the

hybrids predetermined classes (in each of the four cases

the pA was considerable higher for the belonging parental

race: 0.81, 0.74, 0.71, 0.54). Both NEWEWHYBRIDSYBRIDS and

STRUCTURETRUCTURE identified the same individual corresponding

to the lowest assignment to the ACR race, AT40 (pA =

0.54: NEWEWHYBRIDSYBRIDS; Qi = 0.67: STRUCTURE).TRUCTURE).

An AMOVAAMOVA hierarchical analysis, grouping the ACR

and CD populations, was performed to investigate the

amount of genetic differentiation between the CD and

ACR races (Table 5). The highest proportion (V = 78.9%;

P = 0.000) of the total variation was from the ‘within

populations’ level. A significant percentage of variation

was also shown by the ‘between populations’

(V = 13.8%; P = 0.000) and ‘among races’ levels

(V = 7.3%; P = 0.048), implying genetic structuring in

these two biological units.

Discussion

Our study of the parapatric contact area between the

ACR race (2n = 24) and the CD race (2n = 22) is part of a

survey investigating the influence of chromosomal rear-

rangements in the speciation process. Because of their

high chromosomal polymorphism for Rb fusions, the

western house mouse is an emblematic model to study

speciation related to karyotypic changes in mammals.

Contact zones between different chromosomal races are

ideal scenarios to observe the effects of chromosomal

rearrangements on gene flow and to determine the

degree of reproductive isolation. Most studies have been

based on contact between Rb and standard races of the

western house mouse, while contact between Rb races

has received much less attention (for reviews see Searle,

1993; Pialek et al., 2005).

In the present study, microsatellite analysis showed a

drastic reduction of gene flow between the ACR race and

the CD race. The absence of any shared metacentrics

indicates that the CD–ACR contact zone is likely an

example of secondary contact of chromosomal races that

originated independently from standard populations and

differentiated in allopatry (Castiglia et al., 2002).

As suggested by many authors (e.g. Moritz, 1994), it is

advisable to use different types of genetic markers to

clearly discriminate between short-term and long-term

population differentiation. Microsatellites, characterized

by a high mutation rate, are a good instrument to

determine the ongoing gene flow, so as to estimate the

genetic fragmentation between chromosomal races and

populations in M. m. domesticus.

The intra-population level of analysis showed that only

two loci in PizACR and one locus in PizCD significantly

deviated from H–W equilibrium. These two populations

consisted of four and five sample sites respectively (more

sample sites than other populations showing H–W

equilibrium); these are situations in which a subpopula-

tion structure could occur and could contribute to the

observed reductions of heterozygosity (Hartl & Clark,

B2

B1

F2

F1

Pure ACR

Pure CD

AT20 AT74 AT40

Individuals

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Ass

ignm

ent

prob

abili

ty

AT22

(a)

(b)

Individuals

Ass

ignm

ent

prob

abili

ty

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Pure ACR (P > 0.9) Pure CD (P > 0.9) ACR type (0.5 < P < 0.9) CD type (0.5 < P < 0.9)

AT20

AT74

AT40

AT22

Fig. 7 (a) Assignment probability of the CD and the ACR individ-

uals to the source chromosomal race computed by the program

NEWEWHYBRIDSYBRIDS. The height of the circles (or triangles for the admixed

individuals) is determined by the posterior probability of being Pure

CD (black circles) or Pure ACR (white circles). (b) The assignment

probabilities of the four individuals with admixed genotypes (rep-

resented in the previous graph by triangles with the ID collection

name) are shown. The height of the different patterns in the column

denotes the posterior probability of each mouse belonging to each of

the six different genotype frequency classes [Pure CD, Pure ACR, F1

and F2 hybrids, backcrosses of F1 with the first parental (B1) and

backcrosses of F1 with the second parental (B2)].

Table 5 AMOVAAMOVA hierarchical analysis examining the partitioning of

genetic variance at three hierarchical levels (‘among races’, ‘between

populations within races’ and ‘within populations’).

Degree of

freedom

Variation

(%)

Fissation

indexes

P-value

Among races 1 7.39 FCT = 0.074 0.048

Between populations

within races

4 13.78 FSC = 0.147 0.000

Within populations 170 78.83 FST = 0.211 0.000




1997). Moreover, the absence of linkage disequilibrium

at any pair of loci agreed with the loci mapping; the eight

loci were selected in four different chromosome arms and

the pair of loci in the same arm were mapped in

pericentromeric and telomeric positions, most probably

excluding the occurrence of linkage because of crossing-

over events.

The inter-population level of analysis demonstrated a

high genetic structure over all populations (FST = 0.187

and RST = 0.210), suggesting significant levels of genetic

fragmentation within the study area. However, the result

of the pairwise comparison among all populations

(F- and R-statistic) indicated a higher level of genetic

differentiation between the races than within them,

suggesting that gene flow is more limited inside the

chromosomal races. The F- and R-statistic revealed that

genetic fragmentation within and between races is high.

This result could be explained by two hypotheses. Firstly,

the observed genetic fragmentation can be explained by

the typical population dynamics of the house mouse,

where commensalism plays a fundamental role in lim-

iting dispersal events between different sites. This situa-

tion may enhance the genetic differentiation between

populations, partially hiding the genetic separation

because of the chromosomal races. The high genetic

structuring detected in the present study is comparable to

the levels of genetic structuring found in other studies of

M. m domesticus (FST = 0.15–0.34, values calculated

between 0.5 and 60 km: Dallas et al., 1998; FST = 0.39,

RST = 0.45: Panithanarak et al., 2004). This hypothesis is

supported by microsatellite analyses of noncommensal

small mammals, for example the common shrew Sorex

araneus, which showed weak genetic fragmentation

(FST = 0.05, RST = 0.11: Lugon-Moulin et al., 2000;

FST = 0.018: Andersson et al., 2004), thus confirming

that gene flow in the house mouse is strongly affected by

commensalism. The absence of a correlation between the

genetic and geographical distances can also be explained

by the role of commensalism. In small mammals, it is

intimately related to passive transport via human vehi-

cles (Cucchi et al., 2005), especially in a farming area like

the Apennine valley along the Aterno River, and it could

be the cause of introgression events independent of

house mouse dispersal. Secondly, the high degrees of

polymorphism in microsatellite alleles, leading to high

levels of homoplasy, may cause overestimation of abso-

lute gene flow between populations (underestimating

the real genetic separation between the races). This can

easily occur when a frequency-based model is used, e.g.

the F-statistic (Gaggiotti et al., 1999).

Other analysis, in addition to the F- and R-statistic,

highlighted the genetic differentiation between the ACR

and CD races. FCA plotting (Fig. 4) showed two separate

clusters constituted by the CD and ACR populations

respectively.

The present evidence is also clearly confirmed by the

results of the assignment test performed with the algo-

rithms used in GENECLASS,ENECLASS, by the Bayesian estimates

executed through BAYESAYESASS+SS+ and by the clustering

analysis conducted with STRUCTURETRUCTURE. The assignment

test clearly identified a strong reduction of gene flow

between the chromosomal races, as shown by the 100%

correct assignment of each individual to the source race.

When applied at the population level, the test provided

wrong assignments only between populations belonging

to the same chromosomal races. The absence of incorrect

assignments of individuals of the two races at Pizzoli is

particularly significant. In this village, the two races are

parapatrically distributed without a geographical barrier.

This indicates the presence of a genetic barrier that

prevents gene exchange between the CD and ACR races.

The clustering analysis also identified two main clus-

ters isolated by a barrier to gene flow, corresponding to

the chromosomal races. However, there were different

patterns of differentiation. In the ACR group, the

‘between populations’ level of genetic fragmentation

was shown by the inferred K > 2 clusters. With K = 3, 4,

5, 6, a group subdivision was detected and there was

partial correspondence between the inferred clusters and

the ACR populations. The fragmentation observed in the

ACR race did not occur in the CD race for K > 2, probably

reflecting the different sample sites.

Because of the consistency of the results, the clustering

analysis used in STRUCTURETRUCTURE could be an instrument to

assign individuals collected in this area to one of the two

chromosomal races without karyotypic characterization.

The Bayesian analysis performed by BAYESAYESASS+SS+ con-

firmed the pattern showed by the assignment test and by

the clustering analysis. Indeed, the program detected a

significant gene flow only between the two CD popula-

tions. The most important result obtained by this

approach was the absence of gene flow between the

two races living in the village of Pizzoli. Considering

the parapatric distribution of these populations, without

the presence of physical barriers, this absence demon-

strated that other kinds of barrier to gene flow (easily

represented by kariotypyc incompatibility) were operat-

ing.

The sporadic hybridization events detected in previous

research were not confirmed in the present study.

Castiglia et al. (2002) described one putative hybrid via

a cytogenetic analysis (an individual with ACR chromo-

somal background presenting two different metacentrics,

one typical of the CD race and one arising by WART) and

another ACR individual with CD-like mtDNA haplotype.

The same mice showed no evidence of hybridization in

the microsatellite loci analysis of their genotypes. In the

first case, the WART could arise directly in a mouse with

an ACR normal karyotype. In the second case, the

discrepancy between the different results could be

explained by the rapid mutation rate of microsatellite

sequences, which tends to hide past hybridization events.

Hence, the genotypes detected in this individual could

have originated from hybrid ancestors subjected to many




generations of interbreeding within the ACR race. This

consideration is supported by the results obtained by the

software NEWEWHYBRIDSYBRIDS, where there was no evidence of

recent hybridization events as revealed by the absence of

F1, F2 and backcrosses.

The reduction of gene flow revealed by the mtDNA

analysis (Castiglia et al., 2002) was underlined by the

unlinked nuclear loci, indicating complete reproductive

isolation of the races. A preliminary behavioural study

suggested that ACR mice are more aggressive than CD

mice (Carpineti & Castiglia, 2004); after the occurrence of

secondary contact, a premating barrier could have been

established between the two races and its uninterrupted

action may be the cause of the increasing genetic isolation

detected by microsatellites, a short-term discriminator.

Capanna et al. (1984) suggested the importance of

aggressiveness between males of two chromosomal

races, Poschiavo (POS: 2n = 26) and Upper Valtellina

(UV: 2n = 24), in the village of Migiondo (Valtellina,

northern Italy) as a likely reason for the absence of

hybridization events. Hauffe & Searle (1992) recorded

the extinction of the less aggressive race (POS) in

Migiondo. This scenario is highly improbable in our

study area because of the lack of geographical isolation.

Thus, there should only be a shift of the contact zone

(detected in the village of Pizzoli), with a geographical

regression of the CD race.

Recent studies considered the possibility of a chromo-

somal-based speciation (reviewed in Rieseberg, 2001)

emphasizing the role of crossover suppression, as a

mechanism reducing gene flow in the rearranged chro-

mosomes. At the contrary, the present work reveals a

case study where, in presence of high level of chromo-

somal heterozigosity, gene flow may be completely

absent, probably because of hybrid dysfunction.

Acknowledgments

This work was supported by MURST 40% and PRIN

funds. Authors thank E. Solano for help in collecting

specimens. A special acknowledgment to Prof. Marco

Corti, promoter of this research programme and great

friend, who prematurely disappeared.

References

Anderson, E.C. & Thompson, E.A. 2002. A model-based method

for identifying species hybrids using multilocus genetic data.

Genetics 160: 1217–1229.

Andersson, A.C., Narain, Y., Tegelstrom, H. & Fredga, K. 2004.

No apparent reduction of gene flow in a hybrid zone

between the West and North European karyotypic groups

of the common shrew, Sorex araneus. Mol. Ecol. 13: 1205–

1215.

Auffray, J.-C., Vanlerberghe, F. & Britton-Davidian, J. 1990. The

house mouse progression in Eurasia: a palaeontological and

archaeozoological approach. Biol. J. Linn. Soc. 41: 13–25.

Basset, P., Yannic, G., Brunner, H. & Hausser, J. 2006. Restricted

gene flow at specific parts of the shrew genome in chromo-

somal hybrid zones. Evolution 60: 1718–1730.

Belkhir, K., Borsa, P., Chikhi, L., Raufaste, N. & Bonhomme, F.

2004. GENETIXGENETIX 4.05, Logiciel Sous WindowsTM Pour la

Genetique Des Populations. Laboratoire Genome, Populations,

Interactions, CNRS UMR 5000, Universite de Montpellier II,

France.

Benzecri, J.P. 1973. L’analyse des Donnees. Tome 1: La Taxinomie;

Tome II: L’analyse des Correspondences. Dunod, Paris.

Bidau, C.J., Gimenez, M.D., Palmer, C.L. & Searle, J.B. 2001.

The effects of Robertsonian fusions on chiasma frequency

and distribution in the house mouse (Mus musculus domes-

ticus) from a hybrid zone in northern Scotland. Heredity 87:

305–313.

Britton-Davidian, J., Catalan, J., Ramalhinho, M.G., Ganem, G.,

Auffray, J.-C., Capela, R., Biscoito, M., Searle, J.B. & Mathias,

M.L. 2000. Rapid chromosomal evolution in island mice.

Nature 403: 158.

Capanna, E. 1982. Robertsonian numerical variation in animal

speciation: Mus musculus, an emblematic model. In: Mechanism of

Speciation (C. Barigozzi, ed.), pp. 155–177. Alan Liss, New York.

Capanna, E., Corti, M., Mainardi, D., Parmigiani, S. & Brain, P.F.

1984. Karyotype and intermale aggression in wild house mice:

ecology and speciation. Behav. Genet. 14: 195–208.

Carpineti, M. & Castiglia, R. 2004. Analysis of behavioural

discrimination mechanisms in a contact zone between two

metacentric races of the house mouse, Mus musculus domesticus,

in central Italy. Rend. Fis. Acc. Lincei 15: 31–41.

Castiglia, R. & Capanna, E. 1999a. Contact zones between

chromosomal races of Mus musculus domesticus. 1. Temporal

analysis of hybrid zone between the CD chromosomal race

(2n = 22) and populations with the standard karyotype.

Heredity 83: 319–326.

Castiglia, R. & Capanna, E. 1999b. Whole-arm reciprocal

translocation (WART) in a feral population of mice. Chromo-

some Res. 7: 493–495.

Castiglia, R. & Capanna, E. 2000. Contact zone between chromo-

somal races of Mus musculus domesticus. 2. Fertility and segrega-

tion in laboratory-reared and wild mice heterozygous for

multiple Robertsonian rearrangements. Heredity 85: 147–156.

Castiglia, R. & Capanna, E. 2002. Chiasma repatterning across a

chromosomal hybrid zone between chromosomal races of Mus

musculus domesticus. Genetica 114: 35–40.

Castiglia, R., Annesi, F. & Capanna, E. 2002. Contact zones

between chromosomal races of Mus musculus domesticus. 3.

Molecular and chromosomal evidence of restricted gene flow

between the CD race (2n = 22) and the ACR race (2n = 24).

Heredity 89: 219–224.

Chatti, N., Britton-Davidian, J., Catalan, J., Auffray, J. & Said, K.

2005. Reproductive trait divergence and hybrid fertility

patterns between chromosomal races of the house mouse in

Tunisia: analysis of wild and laboratory-bred males and

females. Biol. J. Linn. Soc. 84: 407–416.

Cucchi, T., Vigne, J.D. & Auffray, J.-C. 2005. First occurrence of

the house mouse (Mus musculus domesticus Schwarz & Schwarz,

1943) in the Western Mediterranean: a zooarchaeological

revision of sub-fossil house mouse occurrences. Biol. J. Linn.

Soc. 84: 429–445.

Dallas, J.F., Bonhomme, F., Boursot, P., Britton-Davidian, J. &

Bauchau, V. 1998. Population genetic structure in a Robert-




sonian race of house mice: evidence from microsatellite

polymorphism. Heredity 80: 70–77.

Davisson, M.T. 1996. Rules and guidelines for gene nomencla-

ture. In: Genetic Variants and Strains of the Laboratory Mouse

(M. F. Lyon, S. Rastan & S. D. M. Brown, eds), pp. 1–16.

Oxford University Press, Oxford, UK.

Evanno, G., Regnaut, S. & Goudet, J. 2005. Detecting the

number of clusters of individuals using the software STRUC-TRUC-

TURETURE: a simulation study. Mol. Ecol. 14: 2611–2620.

Excoffier, L., Laval, G. & Schneider, S. 2005. ARLEQUINRLEQUIN ver. 3.0:

an integrated software package for population genetics data

analysis. Evol. Bioinform. Online 1: 47–50.

Falush, D., Stephens, M. & Pritchard, J.K. 2003. Inference of

population structure using multilocus genotype data: linked

loci and correlated allele frequencies. Genetics 164: 1567–

1587.

Gaggiotti, O.E., Lange, O., Rassmann, K. & Gliddon, C. 1999. A

comparison of two indirect methods for estimating average

levels of gene flow using microsatellite data. Mol. Ecol. 8:

1513–1520.

Goodman, S.J. 1997. RST CALCRST CALC: a collection of computer

programs for calculating unbiased estimates of genetic differ-

entiation and determining their significance for microsatellite

data. Mol. Ecol. 6: 881–885.

Goudet, J. 2001. FSTATFSTAT, a Program to Estimate and Test Gene

Diversities and Fixation Indices (Version 2.9.3). Institut

d’Ecologie, Batiment de Biologie, Universite de Lausanne,

Switzerland.

Goudet, J., Raymond, M., De Meeus, T. & Rousset, F. 1996.

Testing differentiation in diploid populations. Genetics 144:

1933–1940.

Gropp, A., Tettenborn, U. & Von Lehmann, E. 1969. Chromos-

omenvariation von Robertson’schen typus bei der tabakmaus,

Mus poschiavinus, und ihren hybriden mit der laboratoriusm-

aus. Cytogenetics 9: 9–23.

Guo, S.W. & Thompson, E.A. 1992. Performing the exact test of

Hardy–Weinberg proportions for multiple alleles. Biometrics 48:

361–372.

Hartl, D.L. & Clark, A.G. 1997. Principles of Population Genetics.

Sinauer, Sunderland, MA.

Hauffe, H.C. & Searle, J.B. 1992. A disappearing speciation

event? Nature 357: 26.

Hauffe, H.C. & Searle, J.B. 1998. Chromosomal heterozygosity

and fertility in house mice (Mus musculus domesticus) from

Northern Italy. Genetics 150: 1143–1154.

King, M. 1993. Species Evolution, the Role of Chromosome Change.

Cambridge University Press, Cambridge.

Louis, E.J. & Dempster, E.R. 1987. An exact test for Hardy–

Weinberg and multiple alleles. Biometrics 43: 805–811.

Lugon-Moulin, N., Balloux, F. & Hausser, J. 2000. Genetic

differentiation of common shrew Sorex araneus populations

among different alpine valleys revealed by microsatellites. Acta

Theriol. 45: 103–117.

Malorni, W., Capanna, E., Cristaldi, M. & De Martino, E. 1982.

Changes of seminiferous epithelium in hybrids of mice

carrying Robertsonian karyotype. Arch. Androl. 9: 333–341.

Moritz, C. 1994. Defining ‘Evolutionarily Significant Units’ for

conservation. Trends Ecol. Evol. 9: 373–375.

Panithanarak, T., Hauffe, H.C., Dallas, J.F., Glover, A., Ward,

R.G. & Searle, J.B. 2004. Linkage-dependent gene flow in a

house mouse chromosomal hybrid zone. Evolution 58: 184–

192.

Pialek, J., Hauffe, H.C. & Searle, J.B. 2005. Chromosomal

variation in the house mouse. Biol. J. Linn. Soc. 84: 535–563.

Piry, S., Alapetite, A., Cornuet, J.M., Paetkau, D., Baudouin, L.

& Estoup, A. 2004. GENECLASSENECLASS 2: a software for genetic

assignment and first-generation migrant detection. J. Hered.

95: 536–539.

Pritchard, J.K., Stephens, M. & Donnelly, P.J. 2000. Inference of

population structure using multilocus genotype data. Genetics

155: 945–959.

Raymond, M. & Rousset, F. 1995. GENEPOPGENEPOP (version 1.2):

population genetics software for exact tests and ecumenicism.

J. Hered. 86: 248–249.

Redi, C.A. & Capanna, E. 1988. Robertsonian heterozygotes in

the house mouse and the fate of their germ cells. In: The

Cytogenetics of Mammalian Autosomal Rearrangements (A. Daniel,

ed.), pp. 315–359. Alan Liss, New York.

Reich, D.E., Feldman, M.W. & Goldstein, D.B. 1999. Statistical

properties of two tests that use multilocus data sets to detect

population expansions. Mol. Biol. Evol. 16: 453–466.

Rieseberg, L.H. 2001. Chromosomal rearrangements and speci-

ation. Trends Ecol. Evol. 16: 351–358.

Rousset, F. 1997. Genetic differentiation and estimation of gene

flow from F-statistics under isolation by distance. Genetics 145:

1219–1228.

Sage, R.D., Atchley, W.R. & Capanna, E. 1993. House mice as

models in systematic biology. Syst. Biol. 42: 523–561.

Searle, J.B. 1993. Chromosomal hybrid zones in eutherian

mammals. In: Hybrid Zones and the Evolutionary Process (R. G.

Harrison, ed.), pp. 309–353. Oxford University Press, New York.

Slatkin, M. 1995. A measure of population subdivision based on

microsatellite allele frequencies. Genetics 139: 457–462.

Vaha, J.P. & Primmer, C.R. 2006. Efficiency of model-based

Bayesian methods for detecting hybrid individuals under

different hybridization scenarios and with different numbers

of loci. Mol. Ecol. 15: 63–72.

Weir, B.S. & Cockerham, C.C. 1984. Estimating F-statistics for

the analysis of population structure. Evolution 38: 1358–

1370.

White, M.J.D. 1978. Modes of Speciation. W.H. Freeman & Co.,

San Francisco.

Wilson, G.A. & Rannala, B. 2003. Bayesian inference of recent

migration rates using multilocus genotypes. Genetics 163:

1177–1191.

Wright, S. 1951. The genetical structure of populations. Ann.

Eugen. 15: 313–354.

Received 16 July 2007; accepted 4 December 2007

Appendix A

Allele frequencies, observed (Ho) and expected (He)

heterozygosities, number of alleles (n) for the eight

microsatellite loci for each population. Sample size is put

in brackets. Private alleles and their corresponding

frequencies are in bold. Observed (Ho-overall) and ex-

pected (He-overall) heterozygosities for the six populations

calculated over all loci, total number of private alleles

(NPA), proportion of private alleles per locus (PPA) and

mean number of alleles per locus (MNA) for each

population are given at the bottom.




Locus Allele

Population*

SVT

(3)

PizCD

(19)

PizACR

(33)

CAP

(9)

CAS

(20)

MAR

(4)

D6Mit138 111 – 0.053 – – – –

115 0.167 0.368 0.455 0.056 0.225 –

119 – 0.105 0.015 0.444 – –

123 0.500 0.263 0.076 0.167 0.175 0.125

127 – 0.105 0.121 0.056 0.575 0.375

131 0.333 0.105 0.136 0.167 0.025 –

135 – – 0.152 – – 0.375

139 – – 0.045 0.111 – 0.125

He 0.733 0.779 0.740 0.771 0.602 0.785

Ho 0.666 0.631 0.697 0.666 0.550 1.000

n 3 6 7 6 4 4

D6Mit26 190 0.333 0.553 0.500 0.500 0.875 0.250

192 0.333 0.132 0.409 0.500 0.050 0.500

194 0.333 0.316 – – – –

198 – – 0.091 – 0.075 0.250

He 0.8000 0.593 0.583 0.529 0.232 0.714

Ho 0.666 0.368 0.606 0.777 0.250 0.750

n 3 3 3 2 3 3

D9Mit251 129 – – 0.045 – – –

131 0.167 0.132 – – – –

133 – – 0.061 – 0.050 0.125

135 0.500 0.789 0.652 0.667 0.125 0.875

137 0.167 0.026 0.030 – 0.225 –

139 0.167 0.053 0.212 0.333 0.600 –

He 0.8000 0.365 0.531 0.470 0.585 0.250

Ho 0.666 0.315 0.545 0.444 0.550 0.250

n 4 4 5 2 4 2

D9Mit52 170 – – 0.212 0.111 0.300 –

172 – 0.053 – – – –

174 – 0.026 – – – –

176 0.167 0.026 – – – –

178 0.167 0.158 0.152 0.444 0.350 –

180 0.667 0.395 0.045 0.222 0.350 1.000

182 – 0.316 0.167 – – –

184 – 0.026 0.424 0.222 – –

He 0.600 0.734 0.733 0.7320 0.6821 0.000

Ho 0.666 0.631 0.575 0.666 0.600 0.000

n 3 7 5 4 3 1

D12Mit182 140 0.167 0.816 0.121 0.833 0.200 0.125

142 0.667 0.158 – – – –

144 0.167 0.026 – – – –

152 – – 0.212 – 0.325 –

154 – – 0.667 0.167 0.475 0.875

He 0.600 0.3172 0.503 0.294 0.644 0.250

Ho 0.666 0.105 0.484 0.333 0.850 0.250

n 3 3 3 2 3 2

D12Mit30 105 0.333 0.026 – – – –

107 0.333 0.474 – – – –

109 – 0.289 0.515 0.111 – –

111 0.333 0.211 0.061 0.167 0.550 0.625

113 – – 0.167 0.389 0.450 –

115 – – 0.091 0.167 – 0.375

117 – – 0.015 0.167 – –

127 – – 0.152 – – –

He 0.8000 0.664 0.682 0.797 0.507 0.535

Ho 1.0000 0.631 0.363 0.888 0.200 0.750

n 3 4 6 5 2 2

Appendix A (Continued)

D19Mit94 81 – – 0.303 0.222 0.125 0.250

89 0.333 – – – – –

91 – – – – – 0.125

93 – 0.289 – 0.056 – 0.125

95 – 0.105 0.439 0.333 0.625 0.375

97 – – 0.091 – 0.025 –

99 – – 0.015 – 0.225 –

101 – – – 0.222 – 0.125

103 – – 0.045 0.167 – –

105 0.500 0.237 – – – –

107 – 0.079 – – – –

109 0.167 0.289 0.106 – – –

He 0.733 0.779 0.704 0.803 0.556 0.857

Ho 0.333 0.526 0.6471 0.636 0.700 0.750

n 3 5 6 5 4 5

D19Mit6 93 – 0.026 0.136 – – –

95 0.333 0.605 0.364 0.222 0.275 0.250

97 0.167 0.105 0.318 0.389 0.400 0.625

99 0.167 0.053 0.076 0.056 0.050 –

101 0.333 0.158 0.106 0.056 0.275 0.125

103 – – – 0.278 – –

111 – 0.053 – – – –

He 0.866 0.607 0.742 0.758 0.703 0.607

Ho 1.000 0.368 0.727 0.777 0.550 0.500

n 4 6 5 5 4 3

He-overall 0.741 0.605 0.652 0.644 0.564 0.500

Ho-overall 0.708 0.447 0.579 0.652 0.531 0.531

NPA 1 4 2 0 0 1

PPA 0.12 0.50 0.25 0.00 0.00 0.12

MNA 3.25 4.75 5.00 3.88 3.38 2.75

*See Table 1 for population details.




Reproductive isolation between chromosomal races of the house mouse Mus musculus domesticus in a...

Documents

Transcript of Reproductive isolation between chromosomal races of the house mouse Mus musculus domesticus in a...