SUPPORTING INFORMATIONS1
Title: Genetic landscape with sharp discontinuities shaped by complex demographic history
in moose (Alcesalces)
Authors: Lovisa Wennerström, Nils Ryman, Jean-Luc Tison, Anna Hasslow, Love Dalén,
and Linda Laikre
Affiliation of first and senior authors (LW, NR, LL): Division of Population Genetics,
Department of Zoology, Stockholm University, SE-106 91 Stockholm, Sweden
E-mail adresses: lovisa.wennerstrom@zoologi.su.se, nils.ryman@popgen.su.se,
linda.laikre@popgen.su.se
Journal: Journal of Mammalogy
Samples and allozyme data incl. statistical power and summary statistics
We used material from a frozen tissue bank collection kept at the Division of Population
Genetics, Department of Zoology, Stockholm University, that includes material from over
20,000 moose sampled all over Sweden during the late 1970s to early 1980s hunting seasons.
From this tissue bank, and from genotypic databases linked to it, a dataset was chosen, and for
subsamples new genetic data were generated as follows. First, we combined collections from
different provinces representing samples from a single year (1980), for which geographic
coordinates and genotypic data at 3 polymorphic allozyme loci were available for each
individual (Pmi, Mdh-2, and Pgi-1, each segregating for 2 alleles; Ryman et al. 1980). This
dataset comprised 20,358 moose that had not been analyzed jointly before, although parts of
the data have been reported previously (Ryman et al. 1977, 1980; Reuterwall 1980, 1981;
Chesseret al. 1982; Charlier et al. 2008). The samples cover most parts of rural Sweden
(except for the most northern mountainous regions; Fig. S1).
The very large number of individuals results in good statistical power for detecting allele
frequency heterogeneity despite the analyses being limited to only 3 allozyme loci. We
divided Sweden into "squares" of 50x50 km, and moose of each square were regarded as a
"subpopulation" resulting in over 100 squares comprising on average 160 moose per square.
We used POWSIM (Ryman and Palm 2006; Ryman et al. 2006) to simulate probabilities of
obtaining statistical significance (chi-square P<0.05) when sampling 100 moose from each of
a series of such squares with different true degrees of divergence (FST) at 3 bi-allelic loci with
similar allele frequencies as in our allozymes. Each simulation was 1,000 runs. Sampling 100
moose from each of 25 squares provides >96 percent probability of detecting a true FST of
0.005. With 100 squares sampled this probability is almost unity, and the probability of
detecting an FST=0.001 is ~70 percent.
1
Table S1.1.Summary statistics for the 3 di-allelic allozyme loci for the full dataset used in the
present study. The measures were obtained after grouping the material – 20,358 individual
moose samples (cf. Fig. S1) - into 50x50 km squares and using squares with n≥50 moose,
resulting in a total of 19,283 moose. The allele frequency range refers to the variant (less
common) allele at each locus.HT is expected heterozygosityfor the total population (all 19,238
individuals included in the analysis). We used FSTAT v. 2 (Goudet 2001) and CHIFISH
(Ryman 2006) for estimation of F-statistics and significance testing. *P<0.05; ***P<0.001
Locus
Allele frequency
range
HT
FST
FIS
FIT
Mdh-2
0.000 - 0.017
0.002
0.005***
-0.005
-0.001
Pgi-1
0.000 - 0.102
0.029
0.031***
-0.002
0.029***
Pmi
0.014 - 0.305
0.252
0.045***
0.015*
0.059***
0.094
0.044***
0.013*
0.057***
Total
Fig. S1. Distribution of 20,358 moose collected during the 1980 hunt and genotyped at 3
allozyme loci; each mark represents an individual moose. Areas void of samples include
regions where the moose is rare or non-existing, including the northern mountainous regions,
the city of Stockholm, large lakes (e.g. Lakes Vänern, and Vättern) the Island of Gotland, and
the farmland of southernmost Sweden.
2
Table S1.2. Number of moose analyzed for 3 allozyme (AZ) loci (20,358 individuals) and 12 microsatellite (MS) loci (1207 individuals). All
individuals genotyped for allozymes have data for all 3 allozyme loci. Among the microsatellites, missing data is present. All individuals have a
unique geographic coordinate. County codes and names as of 1980, map over geographic location of counties to the right. After 1980 the
counties O, P, and R have been merged into a single county.
Age
County
Code
County
Name
AB
AC
BD
C
D
E
F
G
H
K
LM
N
O
P
R
S
T
U
W
X
Y
Z
Stockholm
Västerbotten
Norrbotten
Uppsala
Södermanland
Östergötland
Jönköping
Kronoberg
Kalmar
Blekinge
Skåne
Halland
Gbg och Bohus
Älvsborg
Skaraborg
Värmland
Örebro
Västmanland
Dalarna
Gävleborg
Västernorrland
Jämtland
Total
Number of
individuals
AZ
MS
Adults
AZ
MS
Sex
Calves
AZ
MS
Unkn
AZ
MS
AZ
F
MS
AZ
MS
Unkn
AZ
MS
14
56
23
17
19
19
24
38
17
8
8
11
9
21
17
46
15
10
18
21
18
19
128
837
752
314
231
421
569
459
216
90
147
183
192
655
312
1409
578
387
935
1387
846
1051
15
108
39
24
12
35
27
52
14
10
20
10
24
42
32
46
33
18
31
47
39
38
11
57
42
19
10
16
31
46
19
5
11
12
18
34
17
84
31
18
72
78
40
75
1
4
2
1
1
448
12099
716
746
43
210
1338
1094
544
429
758
1011
843
387
156
214
330
318
1100
506
2525
943
670
1560
2385
1322
1715
30
168
64
42
32
54
52
94
32
18
29
21
35
66
50
97
48
30
51
74
58
62
138
950
848
348
251
475
668
494
234
92
121
210
207
767
322
1764
629
442
1116
1484
982
1198
20
123
46
30
20
35
33
63
17
11
15
13
23
44
31
59
42
17
39
42
46
36
61
289
158
170
153
255
315
296
138
57
85
106
92
273
158
646
269
184
360
810
262
395
8
34
10
9
11
17
17
24
15
7
12
6
10
18
17
34
5
12
9
31
6
22
11
99
88
26
25
28
28
53
15
7
8
14
19
60
26
115
45
44
84
91
78
122
2
11
8
3
1
2
2
7
2
2
2
4
2
4
1
1
3
1
6
4
71
444
300
211
188
321
411
338
152
61
56
135
108
411
177
1032
334
265
553
920
436
589
20358
1207
13740
805
5532
334
1086
68
7513
3
M
1
4
1
1
2
3
1
5
2
2
6
1
5
Table S1.3.Expected heterozygosity (He) and summary F statistics for 3 allozyme and 12 microsatellite loci for 1,207 Swedish moose from the
two major subpopulations detected in this study - the northern and the southern one, including the transition zone. We identified the approximate
border between these 2 main clusters by drawing an arbitrary line in the center of the transition zone (Supporting information 5), and refer
individuals north of this line to the northern subpopulation and individuals south of it to the southern one.Map of the geographic location of
samples and subpopulations to the right. *P < 0.05, ** P< 0.01, *** P< 0.001.
No. of
alleles
No. of
individuals
Freq. of most
common allele
Allelic
richness
He
Mdh-2
2
1201
0.999
1.255
0.001
Pgi-1
2
1206
0.979
2.000
Pmi
2
1203
0.862
2.000
All allozymes
6
1207
0.947
BM757
11
1076
BM848
6
CSSM39
Locus
RST
FST
FIT
FIS
-
0.001
0.001
0.040
-
0.042
***
0.071
0.239
-
0.016
***
0.115
**
0.101
**
1.752
0.093
-
0.018
***
0.110
*
0.094
*
0.272
9.569
0.815
0.002
0.064
***
0.095
***
0.034
1055
0.380
5.496
0.721
0.027
0.226
***
0.285
***
0.077
5
1071
0.589
5.272
0.504
0.009
0.129
***
0.148
*
0.023
CSSM003
6
1178
0.661
4.211
0.551
0.004
0.027
***
0.090
*
0.064
*
IDGA8
8
1089
0.285
7.995
0.810
0.048
0.089
***
0.161
***
0.079
***
IDGVA29
11
1028
0.428
9.406
0.746
0.211
0.010
***
0.134
*
0.039
*
MAF46
11
1200
0.484
8.453
0.700
0.034
0.046
***
0.090
***
0.046
*
McM130
4
1186
0.624
3.258
0.539
0.001
0.016
***
0.090
***
0.075
**
McM58
11
1195
0.318
9.385
0.767
0.035
0.209
***
0.226
***
0.022
*
McM64
7
1073
0.362
5.774
0.740
0.066
0.056
***
0.069
*
0.014
OarCP21
2
1074
0.688
2.000
0.430
0.004
0.004
*
0.043
Allozymes
-
-0.001
-
0.030
Microsatellites
***
0.039
RT30
9
1202
0.397
7.699
0.720
0.024
0.104
***
0.160
***
0.062
*
All microsatellites
91
1207
0.457
6.543
0.670
0.070
0.096
***
0.139
***
0.048
***
All loci
97
1207
0.555
5.585
0.555
-
0.093
***
0.138
***
0.050
***
4
Literature cited (not provided in the paper)
RYMAN, N. and S. PALM. 2006. POWSIM: a computer program for assessing statistical power
when testing for genetic differentiation. Molecular Ecology Notes 3:600-602.
RYMAN, N., S. PALM, C. ANDRE, G: R. CARVALHO, T. G. DAHLGREN, P. E. JORDE, L.
LAIKREET AL. 2006. Power for detecting genetic divergence: differences between statistical
methods and marker loci. Molecular Ecology 15:2031-2045.
5