Supplementary Material METHODS DAPC analysis details The first

Supplementary Material
METHODS
DAPC analysis details
The first step consists of a principal components analysis in which uncorrelated principal components are obtained from the
original multilocus dataset and are then used in a discriminant analysis in the second step. The R (R Development Core Team 2012)
package ‘adegenet’ (Jombart 2008) was used to perform DAPC and obtain posterior probabilities of assignment of individuals to
clusters obtained using K-means clustering. To determine the optimal number of clusters, the Bayesian information criterion (BIC)
was plotted against the number of clusters. The point at which the rate of change in the BIC or the distance between points becomes
trivial is the optimum number of clusters. We used a subset of principal components that captured approximately 95% of the variation
in the original multilocus dataset for the discriminant analysis step of the DAPC.
Dispersal and migration analyses details
FST, mAIc, and relatedness are expected to be smaller, while vAIc is expected to be higher, in the sex with the higher dispersal
rate. AIc values average zero within a population, and individuals with a negative value are more likely to be dispersers than
individuals with positive AIc values (Goudet et al. 2002). Thus, corrected assignment indices (AIc) of each individual were computed
following Goudet et al. (2002), and variance of AIc values between the sexes was also compared using an F-ratio test under the
hypothesis that variance should be higher for the sex displaying higher dispersal (Goudet et al. 2002). Statistical significance of FST,
mAIc, vAIc, and relatedness values were tested with 10,000 permutations. Calculations were performed with both female and male
selected as the philopatric sex, in both one-sided and two-sided tests. We used both types of test despite a one-sided test being
appropriate for this species (rodents tend to display female philopatry with male dispersal (Calhoun 1962)) because the two-sided test
is also appropriate for individuals sampled from a disturbed urban environment and in a small geographic area (Gardner-Santana et al.
2009).
NC estimation details
For every locus, the number of observed genotypes was lower than the number of expected genotypes. The highest expected-toobserved ratio (greater than 1), selected among all the loci, was multiplied by NE to give preliminary NC values. This calculation
connects the observed pattern of variation in reproductive success summarized in NE to a hypothetical pattern of no difference in
reproductive success, which would be equivalent to NC, under the assumption of random mating and non-overlapping generations. We
calculated the distribution in NC values, which might arise from slight differences in reproductive success resulting in the same
genotypic combinations (due to the genotypes of the breeding parents), as they would if there was no difference in reproductive
success, using a Bayesian method (Petit and Valiere 2006) adapted from Gazey and Staley (1986) using an R software script. This
method applies to a sampling scheme with replacement (non-invasive genetic sampling), but was adapted to our sampling scheme
without replacement by carrying out resampling, with 100,000 replications, of a population of size NC, using a sample size equal to the
number of individuals we analyzed. This was done to obtain an estimate of the number of samples from which the number of uniquely
genotyped individuals (i.e., number of individuals analyzed) would have been obtained.Giraudeau F, Apiou F, Amarger V et al. (1999)
Linkage and physical mapping of rat microsatellites derived from minisatellite loci. Mammalian Genome, 10, 405–409.
Jacob HJ, Brown DM, Bunker RK et al. (1995) A genetic linkage map of the laboratory rat, Rattus norvegicus. Nature Genetics, 9, 63–69.
Jombart T (2008) Adegenet: a R package for the multivariate analysis of genetic markers. Bioinformatics, 24, 1403–1405.
R Development Core Team (2012) R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing,
Vienna.TABLES
AND FIGURES
Table S1. Pairwise geographic distance in meters between sites in Salvador, Brazil. On average sites are 2.7 km apart, but most sites
are between 300 m and 1.7 km apart.
Site
PL1(V1) PL2(V2) PL6(V4) PL7(V4) PL8(V2) PL9(V3) SA3(V5) SA4(V6) SJ5(V7)
393
812
640
592
678
1,580
1,655
8,442
PL1(V1)
500
375
370
333
1,617
1,653
8,671
PL2(V2)
317
425
263
1,682
1,598
8,892
PL6(V4)
79
431
1,359
1,330
8,920
PL7(V4)
488
1,313
1,293
8,950
PL8(V2)
1,800
1,774
8,657
PL9(V3)
296
10,073
SA3(V5)
10,124
SA4(V6)
SJ5(V7)
Table S2. Locus information with bold names representing the final loci selected for analyses. Locus name is provided in the first
column followed by the reference, the size of the locus in base pairs (bp; with allele ranges provided when available). Loci referenced
as “non-referenced by Robertson and Gemmell 2004” were used in the 2004 publication without an original source cited. The fourth
columns list the paired locus for multiplex analyses (Multiplex Partner) The The last three columns report the DNA sequence of the
Forward (F-Primer) and reverse (R-primer) primers used for the PCR amplifications and the fluorescent dye used to label each locus
(Label).
Locus
Source
Size (bp)
D6Wox1
D3Mit13
D1Cebr3
D6Wox2
D3Wox12
D10Mit5
D4Cebr2
D4Wox7
D5Cebr1
D11Cebr1
D12Wox1
D15Rat77
D10Rat20
D20Rat46
D16Rat81
D3Rat183
D3Cebr3
D11Mgh5
D2Mit14
D8Mgh7
D1Wox23
D6Cebr1
D19Wox11
D2Wox27
D17Rat115
D12Rat76
D5Rat33
D8Rat123
D7Rat97
D2Rat185
Heiberg et al. 2006
Jacob et al. 1995
Giraudeau et al. 1999
Heiberg et al. 2006
Heiberg et al. 2006
Jacob et al. 1995
Giraudeau et al. 1999
Heiberg et al. 2006
Giraudeau et al. 1999
Giraudeau et al. 1999
Heiberg et al. 2006
Jacob et al. 1995
Jacob et al. 1995
Jacob et al. 1995
Jacob et al. 1995
non-referenced by Robertson and Gemmell 2004
Giraudeau et al. 1999
Jacob et al. 1995
Jacob et al. 1995
Jacob et al. 1995
Heiberg et al. 2006
Giraudeau et al. 1999
Heiberg et al. 2006
Heiberg et al. 2006
non-referenced by Robertson and Gemmell 2004
Jacob et al. 1995
non-referenced by Robertson and Gemmell 2004
non-referenced by Robertson and Gemmell 2004
non-referenced by Robertson and Gemmell 2004
non-referenced by Robertson and Gemmell 2004
86
98
101
102
125
140
143
146
235
263
401
449
102-120
140-174
148-160
161-173
167
230-250
88-102
191
198
223
230
234
191-235
87-105
111-136
198-208
172-186
170-192
Multiplex
Partner
D8Mgh7
D1Wox23
D6Cebr1
D19Wox11
unpaired
D17Rat115
D7Rat97
D12Rat76
unpaired
D6Wox2
D3Wox12
D10Mit5
D4Cebr2
D5Cebr1
D20Rat46
D12Wox1*
D10Rat20
F-primer
R-primer
Label
GCTTCTCATGAAAAGGAAGG
TCCTCTTAGTAAAATTGCACGC
CTTGGGAGCTGGGAGTGT
CCAGTCCATACTTATCCATCTG
TATAGTAAGTTCGAGGCCGG
TGCTGGGTGAACCAGAGAG
TGTCAAAGAAAGCCAGTAAAAC
GATAGCATAAAATCCCTAGAGGTT
AACCGCCTGTATTTCTATTTC
TCTTGGGGATACACGGACT
GACATTAAGGGGTCTTCCTAAG
CATGTGGGGAAAGCATTACC
AGTGATTGCCATACCTGCCT
AAGTACTGAGTGGGCTGCGT
GAGCCTTAGCACAGTGGCTT
GGTCAATGGTGTTTTGACTGAA
CAGGGAATGCAGAAGATACAG
CAGCTCTAATTCCAGAAAGGTTT
AGACCTGGGACAGGGTCCT
TGAAGAGATTTTACTGGGTAGCTCC
TCTGACCCATACTTGTACTTTGC
TGGTTTGGTTGGGGAGAA
CTACCCACCCATCTATTCATCC
GATAATTGACATGTCCAGTTCC
TGGAGTTGGACTGGGAATTC
TGCCTTTTAAAATGATGTGCA
TGGAGAAAAGAAGAACCTCCA
ACACAGGGGAGCAGCTAGTG
CAAGTTTTCCTCTGCCCAAG
TGTTTCCTTCTGATTGGAGTTG
GGGGTGGTCCTTTTCCTT
TCAGCCCTTCTCCTGTCTA
GAAGGCTGAGGTATGAAGACTG
CATTTAGATAGGTGATAGATTCAG
AGGGGACCAGTGAGACTCAC
CTGCCCTCCAAACCACC
GTTTCCAGCACCCATGTCC
TCGATTTATCTGAAACCATCAC
GCCCAAGTTTGATCCTCAG
GGCCAATCAGTGTAGACAAAT
TATCTTTGCAACGCTGAGG
ACAGAGGGAACCCATCACAG
GAAATGGCCAGGATAAACCA
GGCAAAACACCAATGCCTAT
GGCCCACATGTGCATGTATA
AAGGAGAAGGAAGATGGGGA
GTGGCTTTAGGACTCTGGAG
GAATCGATTGACAGATGTCTGTG
ATGAGAGGTCAAAGCTTCTCA
TGGACCAGGCAAGTTCTCTT
AATTTCTGCCTCTTTTTCTCAG
GTGCTGTCAGGGAAAGATGTA
GTTTCCAGCACCCATGTCC
CTGGCTGATGGTAGGATGAG
TCACTCTTTCTGCAAATTTGCTAT
ATTGGCAATGCACTCATGTG
GTGCCCTCAGACTGAACTC
CTTTGAACAGAGCAGCCTGG
GCTGTCATTCCACTGGGTTT
TGAAACAGAAAAGACAAAGGACA
FAM
FAM
FAM
FAM
FAM
FAM
FAM
FAM
FAM
FAM
FAM
FAM
FAM
FAM
FAM
FAM
HEX
HEX
HEX
HEX
HEX
HEX
HEX
HEX
HEX
HEX
HEX
HEX
HEX
HEX
*removed from dataset for final analyses
Table S3. AMOVA results by sampling year for all rat samples (temporal structure).
Source of variation
Variance component
% of variation
p-value
Among groups (sampling times)
Va = -0.02
-0.48
0.66
Among sites within groups (sites)
Vb = 0.43
9.76
0.00
Within sites (individuals)
Vc = 4.01
90.72
0.00
Table S4. Population pairwise Fst values with samples temporally separated. Non-significant values are in bold. *Comparisons
between the same site at two different years.
Site
PL1_2010
(V1)
PL1_2011
(V1)
PL2_2010
(V2)
PL2_2011
(V2)
PL6_2010
(V4)
PL6_2011
(V4)
PL7_2010
(V4)
PL8_2010
(V2)
PL8_2011
(V2)
PL9_2011
(V3)
SA3_2010
(V5)
SA4_2010
(V6)
SJ5_2010
(V7)
PL1_2010
(V1)
PL1_2011
(V1)
PL2_2010
(V2)
PL2_2011
(V2)
PL6_2010
(V4)
PL6_2011
(V4)
PL7_2010
(V4)
PL8_2010
(V2)
PL8_2011
(V2)
PL9_2011
(3)
SA3_2010
(V5)
SA4_2010
(V6)
SJ5_2010
(V7)
0.01*
-
0.06
0.06
-
0.03
0.04
0.02*
-
0.12
0.12
0.17
0.15
-
0.08
0.08
0.12
0.08
0.02*
-
0.06
0.07
0.13
0.08
0.14
0.04
-
0.08
0.06
0.12
0.10
0.13
0.08
0.11
-
0.06
0.10
0.11
0.09
0.16
0.11
0.12
0.13*
-
0.06
0.07
0.11
0.09
0.05
0.05
0.09
0.09
0.10
-
0.06
0.09
0.10
0.10
0.20
0.13
0.14
0.16
0.09
0.12
-
0.09
0.16
0.12
0.12
0.27
0.15
0.14
0.22
0.09
0.16
0.08
-
0.12
0.14
0.17
0.08
0.19
0.12
0.14
0.17
0.14
0.13
0.21
0.28
-
Table S5. Population pairwise Fst values (lower diagonal) with pairwise geographic distance in meters (upper diagonal) between sites.
Non-significant values are in bold.
Site(Valley)
PL1(V1)
PL2(V2)
PL6(V4)
PL7(V4)
PL8_2010(V3)
PL8_2011(V3)
PL9(V3)
SA3(V5)
SA4(V6)
SJ5(V7)
PL1(V1) PL2(V2) PL6(V4) PL7(V4) PL8_2010(V2) PL8_2011(V2) PL9(V3) SA3(V5) SA4(V6) SJ5(V7)
0.09
0.08
0.06
0.07
0.08
0.06
0.07
0.11
0.12
393
0.10
0.10
0.10
0.09
0.09
0.09
0.11
0.11
812
500
0.06
0.09
0.11
0.05
0.14
0.17
0.12
640
375
317
0.11
0.12
0.09
0.14
0.14
0.14
592
370
425
79
0.13
0.09
0.16
0.22
0.17
592
370
425
79
0.10
0.09
0.09
0.14
678
333
263
431
488
488
0.12
0.16
0.13
1,580
1,617
1,682
1,359
1,313
1,313
1,800
0.08
0.21
Figure S1. Plot of linearized pairwise FST [FST/(1- FST)] over pairwise distance (m) between sites in Pau da Lima.
1,655
1,653
1,598
1,330
1,293
1,293
1,774
296
0.28
8,442
8,671
8,892
8,920
8,950
8,950
8,657
10,073
10,124
-
Table S6. Population size (NE) with 95% confidence intervals (CI) and census size (NC) estimates with 95% highest probability
density (HPD) intervals for clusters 1, 2, and 3. Estimates of NE intervals were computed using the linkage disequilibrium method
implemented by LDNE, while NC estimates were calculated with a sequential Bayesian method (Petit and Valiere 2006).
Cluster 1
Cluster 2
Cluster 3
Mean NE
94
34
54
95% CI
75-125
10-40
35-85
Mean NC
897
273
454
95% HPD
357-1608
167-396
233-722