Molecular Ecology (2009)
doi: 10.1111/j.1365-294X.2009.04101.x
Dynamic micro-geographic and temporal genetic diversity
in vertebrates: the case of lake-spawning populations of
brown trout (Salmo trutta)
Blackwell Publishing Ltd
J A N H E G G E N E S ,*§ K N U T H . R Ø E D ,† P E R E R I K J O R D E ‡ and Å G E B R A B R A N D *
*Freshwater Ecology and Inland Fisheries Laboratory, Natural History Museum, University of Oslo, Box 1172 Blindern, N-0318 Oslo,
Norway, †Norwegian School of Veterinary Science, Box 8146, Dep. N-0033 Oslo, Norway, ‡Centre for Ecological and Evolutionary
Synthesis (CEES), Department of Biology, University of Oslo, Box 1066 Blindern, N-0316 Oslo, Norway
Abstract
Conservation of species should be based on knowledge of effective population sizes and
understanding of how breeding tactics and selection of recruitment habitats lead to genetic
structuring. In the stream-spawning and genetically diverse brown trout, spawning and
rearing areas may be restricted source habitats. Spatio–temporal genetic variability patterns
were studied in brown trout occupying three lakes characterized by restricted stream
habitat but high recruitment levels. This suggested non-typical lake-spawning, potentially
representing additional spatio–temporal genetic variation in continuous habitats. Three
years of sampling documented presence of young-of-the-year cohorts in littoral lake areas
with groundwater inflow, confirming lake-spawning trout in all three lakes. Nine microsatellite markers assayed across 901 young-of-the-year individuals indicated overall
substantial genetic differentiation in space and time. Nested gene diversity analyses revealed
highly significant (≤ P = 0.002) differentiation on all hierarchical levels, represented by
regional lakes (FLT = 0.281), stream vs. lake habitat within regional lakes (FHL = 0.045), sample
site within habitats (FSH = 0.010), and cohorts within sample sites (FCS = 0.016). Genetic
structuring was, however, different among lakes. It was more pronounced in a natural lake,
which exhibited temporally stable structuring both between two lake-spawning populations
and between lake- and stream spawners. Hence, it is demonstrated that lake-spawning
brown trout form genetically distinct populations and may significantly contribute to
genetic diversity. In another lake, differentiation was substantial between stream- and
lake-spawning populations but not within habitat. In the third lake, there was less apparent
spatial or temporal genetic structuring. Calculation of effective population sizes suggested
small spawning populations in general, both within streams and lakes, and indicates that
the presence of lake-spawning populations tended to reduce genetic drift in the total (meta-)
population of the lake.
Keywords: brown trout, cohorts, conservation, effective population size, genetic diversity, lake
spawning, microsatellites, Salmo trutta
Received 10 December 2008; accepted 10 December 2008
Introduction
Reproductive strategies may determine genetic population
structure and diversity in animals and plants (Charlesworth
Correspondence: Jan Heggenes, Fax: +47 35 95 27 02;
E-mail: jan.heggenes@hit.no
§Present address: Department of Environmental Sciences,
Telemark University College, Halvard Eikas Plass, N-3800 Bø i
Telemark, Norway.
© 2009 Blackwell Publishing Ltd
& Wright 2001; Goodisman & Hahn 2005; Friesen et al.
2007). Recruitment areas are key habitats (Pulliam 1988),
influencing effective population size and therefore the
amount of genetic variability transferred to the next
generation (Chesser 1991; Allendorf & Waples 1996). Fish
exhibit a particularly wide range of breeding strategies
(e.g. Helfman et al. 1997; Skubic et al. 2004). Fertilisation
and egg development is usually external, and successful
breeding will to a large extent be dictated by environmental
2 J. HEGGENES ET AL.
Fig. 1 Map of southern Norway with location of Lake Jølstervannet (= J), Lake Røldalsvannet (= R), and Lake Oldenvannet
(= O). Arrows indicate inflowing and outflowing rivers. Sampling
sites are marked with first letter of lake name, river (R) or lake (L)
site, and number.
conditions at breeding sites and recruitment areas. Selection
and competition for suitable spawning areas may therefore
be intense and characterize species and their reproduction
(Largadièr et al. 2001; Fleming & Reynolds 2004). Such
evolutionary and ecological mechanisms have been invoked
to explain breeding tactics and species differentiation in
spawning-site selection in salmonids (Crisp & Carling 1989;
Fleming & Reynolds 2004).
Although they may inhabit marine, lake and stream
habitats as adults, most salmon and trout species migrate
to stream ecosystems for spawning (e.g. Northcote 1992;
Hendry & Stearns 2004). Within streams, these species do
not utilize the entire area but select restricted habitats for
spawning and rearing (e.g. Crisp & Carling 1989; Heggenes
et al. 1999). The widely distributed and stocked brown
trout (Salmo trutta), an ecologically polytypic species,
typically spawns in running water to ensure sufficient
oxygenation of its developing eggs (Jones & Ball 1954;
Crisp & Carling 1989). Lake-spawning has been reported in
some instances for this species (e.g. Sømme 1941; Jensen
1963; Barlaup et al. 1994) and may be a breeding strategy
adapted to reduce competition for limited stream-spawning
habitats. Genetic differentiation is a priori often considered to be a sign of adaptation that enhances survival
and reproduction in that particular local environment
(e.g. Carvalho 1993; Adkison 1995) but may well be a
non-adaptive consequence of genetic drift in small
populations (Adkison 1995; Hansen et al. 2002; Østergaard
et al. 2003).
Several Norwegian lakes with natural brown trout
populations have surprisingly high recruitment rates
considering the limited stream breeding and recruitment
habitats (e.g. Brabrand 1998; Haugen 1998). Most of these
lakes are situated in previously glaciated areas with littoral
gravel spawning areas. Influx of groundwater associated
with quarternary geological structures and the occurrence
of young-of-the-year has been documented in several lakes
(Brabrand et al. 2002, and own data). If lake spawning is an
alternative recruitment strategy that is stable over time, it
may represent additional, and until now unknown, local
genetic variation. If so, it may represent important conservation concerns. We hypothesize that potential lake
spawning and recruitment will be reflected in increased
local genetic diversity, with genetically distinct populations
located in stable lake-spawning areas. Such microgeographic
differentiation is, however, likely to imply small effective
population sizes, and the issue of temporal stability across
cohorts, i.e. whether there is a persistent genetic structure
or not, is of major conservation relevance. Our objectives
were: (i) to identify the number and structure of local
lake-spawning natural brown trout populations in three
lakes with limited stream breeding and recruitment areas;
and (ii) to test if this was stable over time by comparing
cohorts and estimate the genetically effective population
sizes.
Materials and methods
Study areas
The study sites are all lakes formed by glacial activity (fjord
lakes) located in western Norway (Fig. 1), generally with
steep shorelines and high maximum depths. The lakes lack
anadromous salmonids and no stocking has been undertaken in them. The natural Lake Oldenvannet (Fig. 1; 37 m
a.s.l., surface area 8.4 km2, maximum depth 90 m) has two
major inlet rivers that are both glacial, summer-cold and
with high turbidity. Stream-spawning occurs only in the
lower part (c. 70 m) of the Rustøygrova River (Fig. 1), and
lake spawning is unknown. In addition to brown trout,
eel (Anguilla anguilla L.), Arctic charr (Salvelinus alpinus L.)
and three-spined stickleback (Gasterosteus aculeatus L.)
are present in the lake. A second lake is the regulated Lake
Jølstervannet (Fig. 1; 205.5–207 m a.s.l., maximum surface
area 39 km2; maximum depth 233 m), which also has
glacier-fed inlet rivers with low summer temperatures.
This lake was the first locality in Norway where extensive
lake-shore spawning in brown trout was documented
(Sægrov 1990). Other fish species present are three-spined
stickleback, and since the late 1980s also European minnow
© 2009 Blackwell Publishing Ltd
G E N E T I C D I V E R S I T Y I N L A K E - S PAW N I N G B R O W N T R O U T 3
(Phoxinus phoxinus L.). The third lake is the regulated Lake
Røldalsvannet (Fig. 1; 363–380 m a.s.l., maximum surface
area 9 km2, maximum depth 90 m). This lake has one major
tributary (Røldalselva River, Fig. 1) that provides limited
spawning and recruitment areas for brown trout because of
extremely variable regulated and low residual water flows.
Brown trout is the only fish species in the system. The three
lakes are widely separated geographically and are not
connected (cf. inset in Fig. 1).
Field sampling
In all three lakes, potential lake-spawning and recruitment
sites (Fig. 1) were located on the basis of geological maps
and field surveys, and subsequently confirmed by
groundwater inflow measurements and collection of alevins,
larvae newly emerged from spawning substrata from late
June to mid-July (for details, see Brabrand et al. 2002).
Spawning by brown trout in littoral areas was verified
by surface scuba-diving, and in Røldalsvannet and
Jølstervannet also by bottom gill netting (Sægrov 1990).
The lake-spawning sites were all found along the shore on
gravel bottom substrata, 20–100 m in length, and extending
from the shoreline to approximately 1–10 m depth, and
down to 20 m in Røldalsvannet. Additional sites in inflow
rivers (Fig. 1) were 70–150 m in length and covered the
width of the river. In Oldenvannet, fish were sampled from
two sites in the lake (OL1–2) and from one site in the main
inlet spawning river, Rustøygrova (OR1; Fig. 1); in
Jølstervannet, fish were sampled from three sites in the
lake (JL1–3) and one in each of the two inlet rivers
Årdalselva and Helgheimselva (JR1–2; Fig. 1); and in
Røldalsvannet, fish were sampled from three sites in the
lake (RL1–3) and two from sites in the inlet river (RR1–2;
Fig. 1).
All sites (Fig. 1) were sampled by electrofishing in
September with a back-pack unit (pulsed DC back packer,
current output 1500 V, pulse length 1.8 ms at 70 Hz).
Capture of young-of-the-year brown trout in the lakes was
taken to indicate lake recruitment, and approximately 30
individuals were collected for microsatellite DNA analysis.
To ensure that samples represented locally lake-spawned
and -hatched trout, only young-of-the-year fish were used
for analysis. Genetic differentiation in young-of-the-year
trout may be more variable than for older fish (Carlsson &
Carlsson 2002). At site JR1 and JR2, additional samples of
spawners (JR1ad and JR2ad) were analysed to supplement a
small sample of young-of-the-year.
To test if the observed lake recruitment represented a
stable pattern over time, the same sites were electrofished
and sampled in three consecutive years (2000, 2001 and
2002, JR2ad in 2001–2002 and JR1ad in 2001 only). All trout were
measured to the nearest mm, weighed and immediately
preserved in 96% ethanol.
© 2009 Blackwell Publishing Ltd
Microsatellite DNA
Approximately 100–200 mg of the caudal fin tissue was
used to extract DNA by a modification of the salt-extraction
procedure of Miller et al. (1988) and adopted for brown
trout by Heggenes et al. (2002). Genetic variation was
analysed in nine unlinked DNA microsatellites (Appendix
I; Heggenes et al. 2002). The forward primers were endlabelled with fluorescence and the polymerase chain
reactions (PCRs) were performed on a GeneAmp PCR
System 9700 (Applied Biosystems) thermal cycler in 10 μl
reaction mixtures containing 20–40 ng of genomic template
DNA, 2 pmol of each primer, 50 mm KCl, 1.5 mm MgCl,
10 mm Tris-HCl, 0.2 mm dNTP and 0.5 U AmpliTaq
(Applied Biosystems). Thermocycling parameters after
denaturation at 94 °C for 5 min were: 30 cycles of: 95 °C for
1 min, 54 °C annealing temperature for 30 s, extension at
72 °C for 1 min. The last polymerization step was extended
to 10 min. The PCR products were electrophorezed using
an ABI Prism 310 Genetic-Analyzer for fluorescent labelled
products.
Data analysis
Standard descriptive statistics of microsatellite diversity,
including expected heterozygosity (HE), observed heterozygosity (HO), number of alleles (a) and average number
of alleles per locus, were compiled using tfpga version 1.3
(Miller 1997). Differences in heterozygosity estimates and
number of alleles per locus across loci and populations
were tested with non-parametric Kruskal–Wallis 1-way
anova of ranks (Sokal & Rohlf 1995).
The test by Guo & Thompson (1992) for conformity to
Hardy–Weinberg equilibrium (HWE): (i) at each locus–
population combination; (ii) globally across loci within
populations; and (iii) across populations within loci was
implemented in genepop version 3.2 (Raymond & Rousset
1995, Markov chain method with default values). To
guard against inflated Type-I-error rates in multiple
comparisons, all critical significance levels for simultaneous tests were evaluated using the Bonferroni correction
(Rice 1989).
Genetic differentiation among samples from different
locations and/or years was quantified using F ST as estimated by θ (Weir & Cockerham 1984), and the 95% confidence intervals (CIs) were obtained using fstat (Goudet,
1995, 2002, 10 000 permutations). Microsatellite differentiation
as expressed by F-statistics was partitioned in a hierarchical
fashion using hierfstat (Goudet 2005) over all levels,
namely: (i) individuals within cohorts (i.e. year classes); (ii)
cohorts within sample site; (iii) sample sites within habitat
(referring to stream and lake spawning areas); (iv) habitat
within regional lakes; and (v) regional lakes. Randomization
tests for the effect on population structure independently
4 J. HEGGENES ET AL.
of each hierarchical level were implemented in hierfstat
with 5000 permutations.
Genetic distances among population pairs were estimated with Cavalli-Sforza (C-S) chord distance calculated
in the phylip version 3.63 software package (Felsenstein
1993), and used to build an unrooted neighbour-joining
tree to visualize the genetic relationships among sample
sites and temporal replicates (year classes).
Microsatellite allele frequencies were tested for evidence
of recent population bottlenecks by the bottleneck
software (Cornuet & Luikart 1996). A bottleneck reduces
both HE and the number of alleles. Alleles are, however, lost
faster than HE and the population will, for a short time, not
be at drift-mutation equilibrium. A Two-Phased Model
of Mutation (TPM) was used with variance = 30.00 and
proportion of stepwise mutation model = 90.00% and
1000 replications. The TPM test assumes that the populations
are near mutation-drift equilibrium.
Effective population sizes (Ne) were estimated separately
for each site (population) using temporal data as follows.
First, the amount of temporal shift in allele frequency at the
nine loci (or rather eight, as locus Br22 was fixed at some
sites; Appendix I) were estimated between consecutive
cohorts (annual samples) using the estimator Fs of Jorde &
Ryman (2007, equation 9 therein). Because the brown trout
displays overlapping generations, the allele-frequency
shift thus estimated will deviate from the corresponding
genetic change in the total population over the same time
interval and from the change over a full generation (the
basis for defining Ne). These deviations arise because each
annual cohort represents only a part of the generation's
reproduction and depends on the generation length as well
as on how the various age classes participate in reproduction.
This latter information was calculated from age-specific
survival and birth rates (i.e. the elements of a Leslie
population matrix) and summarized in a single coefficient
(equation 23 in Jorde & Ryman 1995). This coefficient or
‘correction factor’ (C) was applied along with the generation
length (G) to adjust the observed temporal shifts in order to
estimate the ‘per generation’ effective population size
as Ne = C/(G*2 Fs′) (equation 25 in Jorde & Ryman 1995).
Here, the apostrophe in Fs′ indicates that this estimate of
temporal allele-frequency shift has been adjusted for
sampling (equation 13 in Jorde & Ryman 2007), assuming
sampling plan II (Nei & Tajima 1981; Waples 1989). The
pertinent demographic information is not available for the
present brown trout populations. However, this method
has been used for other Scandinavian brown trout populations, including sub-alpine lakes (four populations: Jorde
& Ryman 1996) and stream-resident trout (two populations:
Palm et al. 2003), spanning a wide range of survival- and
birth rates. While the factors C and G differ among populations, their ratio C/G (which is what enters the estimator
for Ne: above) is quite similar and was restricted to the
range 1.43–2.22. We used the average value of G/C = 2.0
when estimating Ne herein. Finally, the standard error (SE)
of the estimated Fs′ was calculated by jackknifing over the
loci, and used to construct 95% CIs for Fs′ assuming that it
has a normal distribution: CI Fs′ = Fs′ ± 1.96 x SE (e.g. Laikre
et al. 1998). These CIs for Fs′ were subsequently used in the
expression for Ne to calculate CI for the estimated effective
sizes. When more than one pair of consecutive cohorts
were available at a site (as it was for all but one site;
Table 3), the average Fs′ and SE (cf. equation 7.1 in Sokal &
Rohlf 1995) were calculated over pairs and used to estimate
Ne and its CI for that site.
Results
Genetic diversity within sites
There was considerable within-site genetic variability
(average number of alleles across loci = 15.3 ± 12.41 SD,
average observed heterozygosity 0.48 ± 0.07 SE, Appendix
I), but no significant differences in the levels of variability
among samples (cohort–site combinations and sites within
lakes) for any of the three lake systems (P ranged from
0.5710 to 0.9967, Kruskal–Wallis tests, for each test).
The genotypic data suggested a Wahlund effect most
likely because of pooled cohorts (see below). Virtually all
locus–site combinations were in HWE when Bonferronicorrected (9 loci × 14 sites = 126 possible tests; Appendix I).
However, global tests across loci within sites revealed
Bonferroni-corrected significant deviations (four of 14 sites,
Appendix I; significant heterozygote deficit, P < 0.0351), as
did the global test across all loci and all sites (P < 0.0001).
Genetic differentiation among sites, habitats and lakes
There was substantial genetic differentiation among
sample sites as expressed by exact tests for population
differentiation (P < 0.0001) and the average FST, over loci
(Table 1; mean for all sample sites = 0.259, 95% CI from
0.181 to 0.350). In particular, differentiation among sampled
regional lakes was high, with pairwise FST values ranging
from 0.203 to 0.434 (Table 1; mean 0.312 ± 0.074 SD, n = 63).
The variation among sites within lakes was lower, but
significant for 18 of the 28 the pairwise tests within lakes,
resolving on average 5% of the total microsatellite DNA
variation (cf. Table 1). Therefore, the analysis suggested
differentiation into several separate spawning populations
within all three lakes. Overall hierarchical F-statistics
indicated both spatial and temporal genetic heterogeneities (Table 2). From the lowest hierarchical level and
upwards, genetic differentiation was significant among
cohorts within sites (FCohort/Site = 0.016, P < 0.001), among
sites within habitats (FSite/Habitat = 0.010, P < 0.001), between
habitats within lakes (FHabitat/Lake = 0.045, P = 0.002), and
© 2009 Blackwell Publishing Ltd
0.229
0.221
0.251
0.227
0.224
0.219
0.370
0.382
0.397
0.251
0.238
0.051
0.059
—
0.266
0.258
0.297
0.270
0.266
0.255
0.408
0.420
0.434
0.288
0.278
0.065
—
*
0.214
0.209
0.245
0.218
0.207
0.203
0.368
0.378
0.396
0.247
0.235
—
*
*
0.266
0.293
0.294
0.251
0.280
0.275
0.082
0.083
0.102
—
ns
*
*
*
0.391
0.421
0.420
0.396
0.419
0.407
0.005
0.006
—
*
*
*
*
*
0.361
0.391
0.390
0.359
0.382
0.376
—
ns
ns
*
*
*
*
*
—
*
*
*
ns
ns
*
*
*
*
*
*
*
*
Jølster R1
Jølster L1
Jølster L2
Jølster L3
Jølster R1ad
Jølster R2ad
Røldal L1
Røldal L2
Røldal L3
Røldal R1
Røldal R2
Olden R1
Olden L1
Olden L2
0.013
—
*
*
ns
*
*
*
*
*
*
*
*
*
0.024
0.014
—
ns
*
*
*
*
*
*
*
*
*
*
0.019
0.017
0.001
—
ns
*
*
*
*
*
*
*
*
*
–0.003
0.010
0.035
0.031
—
ns
*
*
*
*
*
*
*
*
0.002
0.008
0.030
0.024
0.005
—
*
*
*
*
*
*
*
*
0.372
0.403
0.399
0.368
0.398
0.388
0.0001
—
ns
*
*
*
*
*
0.258
0.279
0.282
0.242
0.265
0.265
0.079
0.081
0.096
0.007
—
*
*
*
Olden L2
Olden L1
Olden R1
Røldal R2
Røldal R1
Røldal L3
Røldal L2
Røldal L1
Jølster R2
Jølster R1ad
Jølster L3
Jølster L2
Jølster L1
Jølster R1
Population
Table 1 Genetic differentiation among sampling sites (tentative populations) across nine loci for lake-spawning brown trout in three lakes. Pairwise values of FST are given above the
diagonal and significance of differentiation in allele frequencies below the diagonal. Significant allele-frequency differentiation (Bonferroni corrected, k = 91) is marked with an asterisk.
FST values that were not significantly different are in italics
G E N E T I C D I V E R S I T Y I N L A K E - S PAW N I N G B R O W N T R O U T 5
© 2009 Blackwell Publishing Ltd
Fig. 2 Neighbour-joining unrooted tree from Cavalli–Sforza chord
distance among the studied brown trout populations in Jølstervannet, Røldalsvannet and Oldenvannet. The first digit after the
name indicates sampling site, and the second digit indicates cohort
(0 = 2000, 1 = 2001, 2 = 2002). Numbers at the branch forks
indicate bootstrap support.
among lakes (FLake/Total = 0.281, P < 0.001). However,
population structures varied among lakes (Table 2). In
Oldenvannet, spatial structuring among sites within
habitats was stronger than in the other two lakes as
indicated by the F-values (Table 2), whereas habitat per se
did not contribute much variation. In Røldalsvannet,
however, stream-spawning and lake-spawning sites
appeared to differentiate, although not significantly in this
analysis. In the more powerful pairwise FST analysis
(Table 1), however, differences were significant between
lake- and river-spawning sites in Røldalsvannet, but not
within habitats. In Jølstervannet, there were less, but
partially significant, genetic heterogeneities among sites
within the lake, and also between one river site and the
lake, but not between river sites (Table 1).
The unrooted Cavalli–Sforza chord-distance tree (Fig. 2)
corroborated these results. The lakes were clearly separate
and within-lakes sites clustered together. In Oldenvannet,
temporal replicates clustered separately within sites, and
all sites clustered separately. This was also largely the
case in Jølstervannet, although much less pronounced. In
Røldalsvannet, the lake-spawning and river-spawning sites
were separate (Fig. 2).
Genetic differentiation among cohorts
Although the hierarchical analysis revealed a significant
temporal component (FCohort/Site; Table 2) when the three
6 J. HEGGENES ET AL.
Table 2 Genetic differentiation as expressed by nested hierarchical F-statistics over lakes, habitats (river or lake spawning), sample site and
cohort (year–class). P are estimated P-values for effect on population structure of each hierarchical level independently, based on
randomizing (5000 permutations) the unit defined by the next lower level within the next upper level
Hierarchical level
Locality
FLake/Total
(PLake)
FHabitat/Lake
(PHabitat)
FSite/Habitat
(PSite)
FCohort/Site
(PCohort)
FInd/Cohort
All lakes
Lake Jølstervannet
Lake Røldalsvannet
Lake Oldenvannet
0.281
(0.001)
0.045
0.013
0.084
–0.001
(0.002)
(0.063)
(0.096)
(0.332)
0.010
0.001
–0.001
0.053
(0.001)
(0.018)
(0.901)
(0.046)
0.016
0.014
0.017
0.015
(0.001)
(0.001)
(0.001)
(0.001)
0.029
0.064
–0.006
0.037
Table 3 Genetic differentiation among three consecutive juvenile (age young-of-the-year) cohorts (0 = 2000; 1 = 2001; 2 = 2002) within
sampling sites, averaged over nine microsatellite loci. Left panel: pairwise estimates of FST (θ: estimated according to Weir & Cockerham
1984) among all three cohorts. Bold values are significantly > 0 at the 5% level (no correction for multiple tests). Right panel: mean temporal
allele-frequency shifts ( Fs′ : estimated according to Jorde & Ryman 2007) among consecutive cohorts, averaged over cohort pairs 2000–2001,
2001–2002, and the corresponding estimated effective size [Ne, adjusted for overlapping generations according to Jorde & Ryman (1995)
using correction factor C/G = 2.0]. SE is the standard error of the mean Fs′ and was calculated by jackknifing over loci and averaged over
the two cohort pairs. SE was used to calculate 95% confidence intervals (CI) for Ne, assuming that the mean is normal distributed. Negative
estimates of Ne were interpreted and reported as infinite (∞)
Estimates of FST (θ)
Estimated parameters
Cohorts of comparison
Amount of genetic drift
Effective size
Lake
Site
0 vs. 1
0 vs. 2
1 vs. 2
Mean Fs′
(SE)
Ne
(95% CI)
Jølstervannet
L1
L2
R1
R2
L1
L2
L3
R1
R2
L1
L2
R1
0.036
0.006
0.015
–0.000
0.007
0.007
0.005
0.007
0.016
0.029
0.006
0.028
0.016
0.003
–0.004
0.011
0.018
0.031
0.025
0.049
0.010
0.008
0.009
0.007
0.007
0.034
–0.003
–0.004
0.018
0.020
0.057
0.024
0.016
0.001
0.0512
0.0157
0.0303
0.0666
0.0019
0.0032
0.0348
0.0496
0.0637
0.0559
0.0346
0.0048
(0.0234)
(0.0168)
(0.0125)
(0.0359)
(0.0091)
(0.0102)
(0.0225)
(0.0166)
(0.0151)
(0.0357)
(0.0182)
(0.0149)
20
63
33
15
536
316
29
20
16
18
29
209
(10–191)
(21– ∞)
(18–171)
(8– ∞)
(51– ∞)
(43– ∞)
(13– ∞)
(12–59)
(11–29)
(8– ∞)
(14– ∞)
(29– ∞)
Røldalsvannet
Oldenvannet
different year classes of young-of-the-year brown trout
from each site were analyzed separately (37 samples); each
year consistently generated a similar spatial pattern (cf. Fig. 2).
Hence, significant genetic variation across cohorts did not
substantially change the spatial differentiation pattern.
Results differed among sites with respect to the amount
of temporal genetic differences among the three consecutive
young-of-the-year cohorts. In Jølstervannet, allele-frequency
differentiation among sites and cohorts was not significant
for pairwise tests (Table 3). In Røldalsvannet, tests among
three within-lake sites revealed no significant differentiation
(Table 3). For the two river sites, two out of six displayed
significant differences. In Oldenvannet, the separation
among cohorts was stronger than in the other lakes. Of
nine pairwise tests among cohorts within sites, five were
significant (Table 3). The Cavalli–Sforza chord distances
clearly grouped cohorts within sites in Oldenvannet, as
opposed to Røldalsvannet (cf. Fig. 2), where lake and river
cohorts clustered separately.
In summary, the analysis indicated significant subdivision
among lake-spawning brown trout in Oldenvannet.
Substantial genetic variation across cohorts did not change
this pattern. In Jølstervannet and Røldalsvannet, however,
although there appeared to be some differentiation among
lake spawners, inter-cohort variation was of the same
magnitude and the signal did not generate a consistent
pattern across tests. This suggests small effective population
sizes.
The estimates of effective population size for each sample
site indicated small Nes in general (Table 3). Although the
© 2009 Blackwell Publishing Ltd
G E N E T I C D I V E R S I T Y I N L A K E - S PAW N I N G B R O W N T R O U T 7
uncertainty associated with each point estimate was large
as judged by the lower (2.5%) and upper (97.5%) confidence
limits (CL), seven of the 12 estimates were < 40, indicating
substantial genetic drift in most of the brown trout populations studied here. Only two estimates, both referring
to lake-spawning populations in Røldalsvannet (sites L1
and L2), were substantially and significantly [as judged
by the lower (2.5%) CLs] larger than this. The two streamresident populations in this lake system (sites R1 and R2)
both had small estimated effective sizes, with finite and
small (59 and 29, respectively) upper CLs. The estimated
effective sizes in Oldenvannet varied from 18 to 209 and
were all associated with wide CLs, even though several of
the cohort pairs exhibited significant allele-frequency
heterogeneity (with two significant comparisons in each of
the sites, L1 and L2; Table 3). In Jølstervannet, all estimates
of effective size were small or moderately small, and two
sites (L1 and R1) had finite upper CLs. From the R2 site
samples were available only for two cohorts (2001 and
2002), of which the latter included adults and is included
here only for completeness. The estimate obtained for this
site had too wide a CI (ranging from 8 to ∞) to yield much
information on effective size.
In concordance with the above estimates of current
effective sizes, tests for historical population bottlenecks
suggested that brown trout in most of the sample sites (11
of 14 tests) have recently lived through episodes with small
parent numbers (Wilcoxon two-tailed tests for heterozygote
excess or deficiency, 0.0019 < P < 0.0488). The exceptions
were the small sample of adults from the inlet river in Lake
Jølstervannet (sample R1ad; Table 1, P = 0. 1953) and two
of the three sites in Lake Oldenvannet (R1 and L1, Table 1,
P > 0.2499). The third site in this lake was only just about
significant (P = 0.0488).
nations within lakes appear to represent additional and
substantial local genetic variation in these brown trout
populations.
Spatial structuring
The brown trout is a vertebrate with high genetic diversity
and population structuring, i.e. among lakes and rivers
(e.g. Estoup et al. 1998; Poteaux et al. 1999; Caputo et al.
2004), in separate tributaries to lakes (e.g. Crozier &
Ferguson 1986; Hansen et al. 1993; Heggenes et al. 2002)
and in large rivers (e.g. Hindar et al. 1991; Antunes et al.
1999; Bouza et al. 2001). Indeed, genetic differentiation may
occur within streams over relatively short spatial distances
because of barriers or semi-barriers to migration (Estoup
et al. 1998; Carlsson et al. 1999; Heggenes & Røed 2006).
Substantial within-stream genetic variation has been
reported also in the absence of apparent barriers to
movement (Carlsson & Nilson 2000). The present study is,
however, to our knowledge, the first to document genetic
structuring among lake-spawning populations within
continuous lakes. Genetic differentiation among populations
within lakes was lower than among regional lakes, but
hierarchical analysis and pairwise comparisons among
sites within lakes indicated spatially separate populations
within all three lake systems. Significant differentiation
among sites within lakes cannot be explained by the
presence of (semi)barriers, and the plausible mechanism
for such population-genetic structuring in trout appears
to be spawning-site fidelity (Carlsson & Nilsson 2000).
Reproductive isolation and lake-spawning life histories
may evolve rapidly in salmonids (Hendry et al. 2000). Such
distinct reproductive ‘ecotypes’ may represent adaptations
to divergent selective regimes and contribute to increased
ecological and genetic diversity (see below).
Discussion
High anthropogenic exploitation rates, habitat fragmentation, and destruction of breeding or recruitment habitats
have reduced many natural vertebrate populations, for
example in riverine salmonids (e.g. Gende et al. 2002;
Burkhardt-Holm et al. 2005). Conservation strategies should
be based on an understanding of how breeding strategies
and selection of recruitment habitats lead to the structuring
of populations, effective population sizes, and thereby
genetic diversity.
This study documented strong regional population
structuring among lakes, as expected for this species.
However, genetic structuring was also identified among
lake young-of-the-year recruits from different sites within
a lake and between lake and stream recruits within the
same water system. Furthermore, apparent population
structuring was confounded when three consecutive
cohorts were considered. The different site–cohort combi© 2009 Blackwell Publishing Ltd
Effective population sizes
Tests across cohorts indicated significant local temporal
genetic change within lakes and streams. Previous studies
have documented similar inter-cohort genetic differences
at isozyme loci (Jorde & Ryman 1996; Palm et al. 2003) and
mitochondrial DNA (Laikre et al. 1998, 2002), and have
ascribed them to random genetic drift in age-structured
populations of limited effective size. On a larger timescale
(i.e., decades), recent studies have indicated substantial
temporal genetic variation in stream-resident brown
trout, attributed to environmentally induced population
bottlenecks (Østergaard et al. 2003; Jensen et al. 2005).
Despite genetic drift, the genetic composition of brown
trout stream populations appears to be relatively stable
over decades, suggesting that such temporal variation may
not be an immediate conservation concern (Hansen et al.
2002; Palm et al. 2003).
8 J. HEGGENES ET AL.
The present study documents substantial differentiation
among consecutive cohorts and small estimates of effective
population sizes for most study sites. The estimates are
comparable to those previously obtained for brown trout
inhabiting small and unstable streams (Østergaard et al.
2003; Jensen et al. 2005), suggesting that lake- and streamspawning populations may be relatively small and isolated
also when the habitat is continuous and there are no physical
barriers to migration and intermixing. This finding may,
at least in part, be explained by the degree and extent of
distinct lake-spawning areas, and spawning-site fidelity. In
Jølstervannet, natural spawning is scattered along the
littoral zone around most of the lake (Sægrov 1990), and the
potential for migration and intermixing is obvious. In the
regulated Røldalsvannet, natural site-specific lake spawning
could have been disrupted by the change in water levels
since 1969, possibly generating admixtures. In the natural
Oldenvannet, lake spawning is located to distinct small
littoral areas, and young-of-the-year were captured along
spatially separated 20–40-m lakeshore segments, perhaps
facilitating population fragmentation within this lake.
The estimated effective sizes are subject to various
uncertainties and possible errors. Uncertainties due to the
stochastic nature of genetic drift and statistical sampling
errors are accounted for in the SE of the mean Fs′, used
to set CIs for the estimated Ne′s. These rather wide CIs
(cf. Table 3) reflect the restricted annual sample sizes
(n = 14–30), number of loci (8–9 loci) and short sample
intervals (< 1 generation) (cf. Waples 1989). Additional
uncertainties not accounted for in the CIs arise from overlapping generations and the lack of demographic data for
the study populations. The demographic parameters may
differ among populations, and using a common factor
C/G for all 12 sample locations may seem problematic. The
most important characteristic determining C/G is the
extent of mixing of genes across cohorts during reproduction,
i.e. when multiple age classes participate in spawning.
Such appears to be the norm in brown trout, and results
from earlier studies (Jorde & Ryman 1995; Palm et al. 2003)
indicate that populations tend to differ demographically
in ways that do not strongly modify C/G. Hence, our
approach is unlikely to substantially bias the estimates, or
bias them in any particular direction.
Small effective population sizes may result from high
natural egg mortalities and intrabrood correlation in egg
mortalities. Although there may be a high number of fish
participating in spawning in the lakes, the number of
spawners contributing hatching alevins is likely to be low.
Hatching success appears to largely depend on groundwater
upwelling in areas with spawning substrate, which tend to
be very localized (Brabrand et al. 2002). Thus, only a few
redds (‘nests’) may produce hatching alevins. This may
also explain why lake spawning in brown trout appears
to be a less common life-history strategy than stream
spawning. On the other hand, these lake-spawning trout
populations clearly represent alternative life-history
strategies and contribute to an increased local ecological
and genetic diversity. To our knowledge, this type of
research has hitherto attracted little attention (Brabrand
et al. 2002). Populations with effective sizes as small as
those reported here clearly cannot maintain appreciable
levels of genetic variation over time if isolated. Isolated
populations are expected to lose a fraction 1/(2Ne) of its
genetic variability each generation because of random
genetic drift. When Ne is as small as reported here (Table 3)
(say, 20), such drift would erode 90% of the populations
variability within less than 100 generations. Levels of
genetic variability in these brown trout populations are,
however, very similar among populations, despite differences in estimated Ne among them, and similar to those
reported for microsatellites in brown trout elsewhere. By
implication, the populations are unlikely to be completely
isolated from each other. Gene flow from neighbouring
populations, most likely those occupying the same lake
system, probably occurs, perhaps in a metapopulation
structure (e.g. Hanski 2001; Baguette 2004).
Such implied gene flow may bias estimates of effective
size calculated from temporal change in allele frequencies.
However, while only a few individuals per generation are
enough to maintain genetic variability in the recipient
population, such low numbers of immigrants are unlikely
to contribute much to temporal change and should therefore
not bias the estimates to any great extent. Some sample
localities were quite similar genetically, including the three
lake-spawning sites in Lake Røldalsvannet (cf. Table 2),
which implies higher levels of gene flow among them.
High gene flow may stabilize allele frequencies and lead to
an upward biased estimate for the local Ne. In such cases,
the estimated effective size may refer to the combined Ne of
the entire population system (Nei & Tajima 1981).
Implications for genetic diversity
In order to quantitatively assess the influence of lakespawning trout on the regional total (meta)populations, we
may tentatively pool sample sites from within each lake
system and calculate the amount of genetic drift from
pooled allele frequencies (e.g. Fraser et al. 2007). For the
three lakes, this approach yielded average estimates of
drift ( Fs′ ) of 0.0092 (Jølstervannet, excluding adults), 0.0083
(Røldalsvannet) and 0.0231 (Oldenvannet), corresponding
to effective population sizes of 109, 120 and 43, respectively,
when correcting for overlapping generations. These
tentative estimates of metapopulation effective sizes are
substantially larger than the corresponding sizes for the
river components alone (which become 39, 31 and 209,
respectively, when pooling river sites within lake systems)
for two of the three regional lakes. This result of pooling
© 2009 Blackwell Publishing Ltd
G E N E T I C D I V E R S I T Y I N L A K E - S PAW N I N G B R O W N T R O U T 9
sample sites on estimated drift and effective sizes has
two important implications. First, it confirms Nei &
Tajima’s (1981) theoretical predictions that pooled samples
should yield larger estimates, representing a more
inclusive population unit. This indirectly supports the
notion that our estimates for local populations (Table 3) are
not strongly biased upwards by gene flow, since they are
much lower than the total for the lake. Second, and
importantly, the result implies that lake-spawning trout
have a stabilizing effect on genetic drift and thus contribute
to maintaining genetic variability within the population
system. This stabilizing effect is expected theoretically, as
effective size of a population system should increase with
increasing number of constituent populations, all else
being equal (Wang & Caballero 1999; Fraser et al. 2007). For
metapopulations, even small and isolated patches may
play a role in genetics by maintaining genetic variance
among demes (e.g. Lehmann & Perrin 2006), as observed
in diverse types of organisms (e.g. Lavigne et al. 2001;
Antolin et al. 2006; Zamudio & Wieczorek 2006). For the
third population system (in Oldenvannet), these predictions
were not born out. In this system, however, only one river
site was sampled, and this happened to harbour a relatively
large population. Combined with two lake sites of low
effective sizes (cf. Table 3), pooling resulted in a fairly low
estimate for the total population in this lake system. It is not
clear if this result is an artefact of unbalanced sampling of
populations in this lake, perhaps augmented by the strong
spatial differentiation among the sampled sites.
In conclusion, this study documents that lake-spawning
brown trout, when present, may constitute distinct populations that can be identified genetically. Such populations
may contribute substantially to increased local genetic
diversity in brown trout, even in spatio–temporally stable
and continuous lake environments, and therefore may give
rise to conservation concerns. In particular, lake-spawning
brown trout represent a distinct ecotype from the usual
stream-spawning form and may constitute an important
source for maintaining genetic variability in local populations. This is important because the species is typically
fragmented into effectively small local populations that
cannot, in isolation, maintain appreciable levels of genetic
variability and may be prone to extinction for genetic and
demographic reasons. Hence, the presence of additional, in
this case lake-spawning, populations represents increased
resistance to loss of genetic variability in the total population system and may in addition provide a rescue effect
against population extinction (e.g. Gonzalez et al. 1998;
Lehmann & Perrin 2006).
Acknowledgements
This project has been funded by the Norwegian Electricity
Industry Association (EBL). We thank L. Midthjell for her assistance
© 2009 Blackwell Publishing Ltd
in the fish genotyping and the subject editor and two anonymous
referees for comments that greatly improved the paper.
References
Adkison MD (1995) Population differentiation in Pacific salmon:
local adaptation, genetic drift, or the environment? Canadian
Journal of Fisheries and Aquatic Sciences, 52, 2762–2777.
Allendorf FW, Waples RS (1996) Conservation and genetics of
salmonid fishes. In: Conservation Genetics: Case Histories from
Nature (eds Avise JC, Hamrick JL), pp. 238–280. Chapman &
Hall, New York.
Antolin MF, Savage LT, Eisen RJ (2006) Landscape features
influence genetic structure of black-tailed prairie dogs (Cynomus
ludovicianus). Landscape Ecology, 21, 867–875.
Antunes A, Alexandrino P, Ferrand N (1999) Genetic characterization of Portuguese brown trout (Salmo trutta L.) and comparison
with other European populations. Ecology of Freshwater Fish, 8,
194–200.
Baguette M (2004) The classical metapopulation theory and the
real, natural world: a critical appraisal. Basic and Applied Ecology,
5, 213 –224.
Barlaup BT, Lura H, Sægrov H, Sundt RC (1994) Inter- and
intra-specific variability in female salmonid spawning behaviour.
Canadian Journal of Zoology, 72, 636–642.
Bouza C, Castro J, Sanchez L, Martinez P (2001) Allozymic
evidence of parapatric differentiation of brown trout (Salmo
trutta L.) within an Atlantic river basin of the Iberian Peninsula.
Molecular Ecology, 10, 1455 –1469.
Brabrand Å (1998) Naturlig rekruttering hos ørret i reguleringsmagasiner. In: Fiskesymposiet 1998 (ed. Erlandsen A), pp. 9–16.
Publikasjon Nr. pp. 281–1998. Energiforsyningens Fellesorganisasjon, Oslo.
Brabrand Å, Koestler AG, Borgstrøm R (2002) Lake spawning
of brown trout related to groundwater influx. Journal of Fish
Biology, 60, 751–763.
Burkhardt-Holm P, Giger W, Göttinger H et al. (2005) Where have
all the fish gone? Environmental Science and Technology 39, 441–447.
Caputo V, Giovannotti M, Nisi Cerioni P et al. (2004) Genetic
diversity of brown trout in central Italy. Journal of Fish Biology,
65, 403 –418.
Carlsson J, Carlsson JEL (2002) Micro-scale distribution of brown
trout: an opportunity for kin selection? Ecology of Freshwater Fish,
11, 234 –239.
Carlsson J, Nilsson J (2000) Population genetic structure of brown
trout (Salmo trutta L.) within a northern boreal forest stream.
Hereditas, 132, 173–181.
Carlsson J, Olsen KH J, Øverli Ø, Stabell OB (1999) Microsatellites
reveal fine-scale genetic structure in stream-living brown trout.
Journal of Fish Biology, 55, 1290–1303.
Carvalho GR (1993) Evolutionary aspects of fish distribution:
genetic variability and adaptation. Journal of Fish Biology, 43, 53–
73.
Charlesworth D, Wright SI (2001) Breeding systems and genome
evolution. Current Opinion in Genetics and Development, 11, 685–
690.
Chesser RK (1991) Influence of gene flow and breeding tactics on
gene diversity within populations. Genetics, 129, 573 –583.
Cornuet JM, Luikart G (1996) Description and power analysis of
two tests for detecting recent population bottleneck from allele
frequency data. Genetics, 144, 2001–2014.
10 J . H E G G E N E S E T A L .
Crisp DT, Carling PA (1989) Observations on siting, dimensions
and structure of salmonid redds. Journal of Fish Biology, 34, 119–
134.
Crozier WW, Ferguson A (1986) Electrophoretic examination of
the population structure of brown trout, Salmo trutta L., from the
Lough Neagn catchment; Northern Ireland. Journal of Fish
Biology, 28, 459 –477.
Estoup P, Rousset F, Michalakis Y et al. (1998) Comparative
analysis of microsatellite and allozyme markers: a case study
investigating microgeographic differentiation in brown trout
(Salmo trutta). Molecular Ecology, 7, 339–353.
Felsenstein J (1993) phylip (Phylogeny Inference Package) version
3.5c. Department of Genetics, University of Washington, Seattle,
Washington.
Fleming IA, Reynolds JD (2004) Salmonid breeding systems. In:
Evolution Illuminated — Salmon and Their Relatives (eds Hendry
AP, Stearns SC), pp. 264–294. Oxford University Press, Oxford.
Fraser DJ, Hansen MM, Østergaard S et al. (2007) Comparative
estimation of effective population sizes and temporal gene flow
in two contrasting population systems. Molecular Ecology, 16,
3866 –3889.
Friesen VL, Burg M, McCoy KD (2007) Mechanisms of population
differentiation in seabirds. Molecular Ecology, 16, 1765–1785.
Gende SM, Edwards RT, Willson MF, Wipfli MS (2002) Pacific
salmon in aquatic and terrestrial ecosystems. Bioscience, 52, 917–
928.
Gonzalez A, Lawton LH, Gilbert FS et al. (1998) Metapopulation
dynamics, abundance, and distribution in a microecosystem.
Science, 281, 2045 –2047.
Goodisman MAD, Hahn DA (2005) Breeding system, colony
structure, and genetic differentiation in the Camponotus festinates
species complex of carpenter ants. Evolution, 59, 2185–2199.
Goudet J (1995) fstat (vers. 1.2): a computer programme to
calculate F-statistics. Journal of Heredity, 86, 485– 485.
Goudet J (2002) fstat, version 2.9.3. User Manual. http://
www2.unil.ch/izea/softwares/fstat.htm.
Goudet J (2005) Hierfstat, a package for R to compute and test
hierarchical F-statistics. Molecular Ecology Notes 5, 184–186.
Latest version available at http://www.unil.ch/popgen/
softwares/hierfstat.htm.
Guo SW, Thompson EA (1992) Performing the exact test of HardyWeinberg proportions for multiple alleles. Biometrics, 48, 361–372.
Hansen MM, Loeschcke V, Rasmussen G, Simonsen V (1993)
Genetic differentiation among Danish brown trout (Salmo trutta)
populations. Hereditas, 118, 177–185.
Hansen MM, Ruzzante DE, Nielsen EE et al. (2002) Long-term
effective populations sizes, temporal stability of genetic composition and potential for local adaptation in anadromous brown
trout (Salmo trutta) populations. Molecular Ecology, 11, 2523–
2535.
Hanski I (2001) Spatially realistic theory of metapopulation
ecology. Naturwissenschaften, 88, 372–381.
Haugen T (1998) Svarer årlige utsettinger til forventningene?
Aursjøen som eksempel. In: Fiskesymposiet 1998 (ed. Erlandsen
A), pp. 1–5. Publikasjon Nr. 298. Energiforsyningens Fellesorganisasjon, Oslo.
Heggenes J, Røed KH (2006) Do dams increase genetic diversity in
brown trout (Salmo trutta)? Microgeographic differentiation in a
fragmented river. Ecology of Freshwater Fish, 15, 366–375.
Heggenes J, Baglinière JL, Cunjak E (1999) Spatial niche variability
for young Atlantic salmon (Salmo salar) and brown trout (S. trutta)
in heterogeneous streams. Ecology of Freshwater Fish, 8, 1–21.
Heggenes J, Røed KH, Høyheim B, Rosef L (2002) Microsatellite
diversity assessment of brown trout (Salmo trutta) population
structure indicate limited genetic impact of stocking in a
Norwegian alpine lake. Ecology of Freshwater Fish, 11, 93 –100.
Helfman GS, Collette BB, Facey DE (1997) The Diversity of Fishes.
Blackwell Science, Oxford.
Hendry AP, Stearns SC, eds. (2004) Evolution Illuminated — Salmon
and Their Relatives. Oxford University Press, Oxford.
Hendry AP, Wenburg JK, Bentzen P et al. (2000) Rapid evolution of
reproductive isolation in the wild: evidence from introduced
salmon. Science, 290, 516–518.
Hindar K, Jonsson B, Ryman N, Ståhl G (1991) Genetic relationships among landlocked, resident and andromous brown trout,
Salmo trutta L. Heredity, 66, 83–91.
Jensen KW (1963) Brown trout spawning in lakes. Ferskvandsfiskeribladet, 8, 123–126.
Jensen LF, Hansen MM, Carlsson J et al. (2005) Spatial and temporal
genetic differentiation and effective population size of brown
trout (Salmo trutta, L.) in small Danish rivers. Conservation Genetics,
6, 615–621.
Jones JW, Ball JN (1954) The spawning behaviour of the brown
trout and salmon. British Journal of Animal Behaviour, 2, 103 –114.
Jorde PE, Ryman N (1995) Temporal allele frequency change and
estimation of effective size in populations with overlapping
generations. Genetics, 139, 1077–1090.
Jorde PE, Ryman N (1996) Demographic genetics of brown trout
(Salmo trutta) and estimation of effective population size from
temporal change of allele frequencies. Genetics, 143, 1369–1381.
Jorde PE, Ryman N (2007) Unbiased estimator for genetic drift and
effective population size. Genetics, 177, 927–935.
Laikre L, Jorde PE, Ryman N (1998) Temporal change of mitochondrial DNA haplotype frequencies and female effective size in a
brown trout (Salmo trutta) population. Evolution, 52, 910–915.
Laikre L, Järvi T, Johansson L et al. (2002) Spatial and temporal
population structure of sea trout at the Island of Gotland,
Sweden, delineated from mitochondrial DNA. Journal of Fish
Biology, 60, 49 –71.
Largadièr CR, Estoup A, Lecerf F et al. (2001) Microsatellite analysis
of polyandry and spawning site competition in brown trout
(Salmo trutta L.). Genetics Selection Evolution, 33(Suppl 1), 205–
222.
Lavigne C, Reboud X, Lefranc M et al. (2001) Evolution of genetic
diversity in metapopulations: Arabidopsis thaliana as an experimental modell. Genetics Selection Evolution, 33(Suppl. 1), S399–
S423.
Lehmann L, Perrin N (2006) On metapopulation resistance to drift
and extinction. Ecology, 87, 1844–1855.
Miller MP (1997) Tools for Population Genetic Analysis (TFPGA) v. 1.3.
Department of Biological Sciences, Northern Arizona University,
Flagstaff, AZ.
Miller SA, Dykes DD, Polesky HF (1988) A simple salting out
procedure for extracting DNA from human nucleated cells.
Nucleic Acids Research, 16, 1215.
Nei M, Tajima F (1981) Genetic drift and estimation of effective
population size. Genetics, 98, 625–640.
Northcote TG (1992) Migration and residency in stream salmonids
— some ecological considerations and evolutionary consequences. Nordic Journal of Freshwater Research, 67, 5–17.
Østergaard S, Hansen MM, Loeschcke V, Nielsen EE (2003)
Long-term temporal changes of genetic composition in brown
trout (Salmo trutta L.) populations inhabiting an unstable
environment. Molecular Ecology, 12, 3123–3135.
© 2009 Blackwell Publishing Ltd
G E N E T I C D I V E R S I T Y I N L A K E - S PAW N I N G B R O W N T R O U T 11
Palm S, Laikre L, Jorde PE, Ryman N (2003) Effective population
size and temporal genetic change in stream resident brown trout
(Salmo trutta, L.). Conservation Genetics, 4, 249–264.
Poteaux C, Bonhomme F, Berrebi P (1999) Microsatellite polymorphism and genetic impact of restocking in Mediterranean
brown trout (Salmo trutta L.). Heredity, 82, 645–653.
Pulliam HR (1988) Sources, sinks and population regulation.
American Naturalist, 132, 652–661.
Raymond M, Rousset F (1995) genepop (Version1.2): Population
genetics software for exact tests and ecumenism. Journal of
Heredity, 86, 248–249.
Rice WR (1989) Analyzing tables of statistical tests. Evolution, 43,
223–225.
Sægrov H (1990) Er innsjøgyting undervurdert? In: Fiskesymposiet
1990 (ed. Erlandsen A), pp. 99–113. Vassdragsregulantenes
Forening, Oslo.
Skubic E, Taborsky M, McNamara JM, Houston AI (2004) When to
parasitize? A dynamic optimization model of reproductive
strategies in a cooperative breeder. Journal of Theoretical Biology,
227, 487–501.
Sokal RR, Rohlf FJ (1995) Biometry. W.H. Freeman, San Francisco,
CA.
Sømme I (1941) Ørretboka. Aschehoug Forlag, Oslo.
© 2009 Blackwell Publishing Ltd
Wang J, Caballero A (1999) Developments in predicting the
effective size of subdivided populations. Heredity, 82, 212–226.
Waples RS (1989) A generalized approach for estimating effective
population size from temporal changes in allele frequency.
Genetics, 121, 379–391.
Weir BS, Cockerham CC (1984) Estimating F-statistics for the
analysis of population structure. Evolution, 38, 1358–1370.
Zamudio KR, Wieczorek AM (2006) Fine-scale genetic structure
and dispersal among spotted salamander (Ambystoma maculatum)
breeding populations. Molecular Ecology, 16, 257–274.
Jan Heggenes’ research interests are in conservation biology and
genetics, in particular how habitat and fragmentation influence
population structure and regulation in freshwater fish. The major
interests of Knut H. Røed are in the use of molecular markers for
understanding patterns and processes of population divergence
and social organization, and their significance to conservation. Per
Erik Jorde is a population geneticist working on temporal genetic
change and spatial genetic structure in various organisms. Åge
Brabrand focuses his research on fish communities, brown trout
ecology and spawning in ground water sites in lenthic environments.
12 J . H E G G E N E S E T A L .
Appendix I Sample sites, sample sizes (n) and diversity indices [number of alleles a, expected heterozygosity (HE), and observed
heterozygosity (HO)], at nine loci found within and among lake-spawning populations of brown trout in three lakes. Data for years/cohorts
are combined. Significant departures from Hardy–Wwinberg equilibrium (heterozygote deficit; Bonferroni-corrected, k = 126) are marked
with asterisks. Uncorrected significant departures are marked with hyphens.
BR13
BR22
Str58
Str15
BR25
Population
/Locus
n
A
HE
HO
n
A
HE
HO
n
a
HE
HO
n
a
HE
HO
n
A
HE
HO
Jølster L1
Jølster L2
Jølster L3
Jølster R1
Jølster R1ad
Jølster R2ad
Jølster all
Røldal L1
Røldal L2
Røldal L3
Røldal R1
Røldal R2
Røldal all
Olden L1
Olden L2
Olden R1
Olden all
58
79
19
65
13
48
282
71
79
70
69
69
358
60
67
70
197
14
13
10
8
6
12
16
7
7
7
11
11
16
11
13
6
14
0.80
0.78
0.78
0.77
0.69
0.72
0.77
0.42
0.38
0.39
0.78
0.67
0.56
0.66
0.76
0.80
0.78
0.69′
0.80
0.68
0.62′
0.77
0.71
0.71
0.42
0.37′
0.41
0.74
0.61
0.51
0.58
0.61′
0.69′
0.63
65
80
19
69
13
50
296
68
79
71
65
72
355
70
73
70
213
2
1
1
1
1
1
2
2
1
1
2
2
2
3
4
2
4
0.15
na
na
na
na
na
0.00
0.02
na
na
0.24
0.25
0.11
0.54
0.68
0.47
0.60
0.15
na
na
na
na
na
0.00
0.02
na
na
0.28
0.26
0.11
0.54
0.59′
0.46
0.53
62
80
19
66
13
50
290
70
84
71
70
74
369
69
73
71
213
7
8
7
9
6
9
15
10
10
9
14
13
18
8
8
10
10
0.59
0.47
0.62
0.56
0.57
0.57
0.55
0.69
0.70
0.60
0.62
0.57
0.65
0.81
0.72
0.72
0.77
0.48′
0.41
0.63′
0.42
0.46
0.50
0.46
0.67
0.73
0.52
0.61
0.57
0.63
0.78
0.77
0.83
0.79
66
79
19
70
13
46
293
72
84
73
62
76
367
71
72
71
214
5
5
4
5
5
6
6
3
3
4
5
5
6
3
3
3
3
0.62
0.51
0.46
0.66
0.73
0.64
0.61
0.61
0.62
0.55
0.42
0.54
0.60
0.31
0.30
0.30
0.30
0.58
0.53
0.37
0.59′
0.62
0.59
0.56
0.69
0.63
0.53
0.49
0.58
0.57
0.31
0.29
0.32
0.31
63
72
19
61
12
50
277
70
83
72
68
71
364
71
71
65
207
20
18
14
14
11
15
30
14
15
14
15
18
22
18
21
17
28
0.91
0.86
0.90
0.85
0.89
0.83
0.89
0.80
0.82
0.82
0.74
0.81
0.81
0.88
0.88
0.90
0.91
0.91′
0.83
1.00
0.77
0.75
0.82
0.84
0.84
0.74′
0.86′
0.75
0.76
0.79
0.82*
0.87
0.77
0.82
BR14
Str60
Str12
BR7
Average
Population
/Locus
n
A
HE
HO
n
A
HE
HO
n
A
HE
HO
n
A
HE
HO
n
a
HE
HO
Jølster L1
Jølster L2
Jølster L3
Jølster R1
Jølster R1ad
Jølster R2ad
Jølster all
Røldal L1
Røldal L2
Røldal L3
Røldal R1
Røldal R2
Røldal all
Olden L1
Olden L2
Olden R1
Olden all
60
77
18
67
13
45
280
69
83
69
66
72
359
55
70
65
190
3
4
3
3
2
3
5
2
3
2
3
2
4
2
2
2
2
0.13
0.26
0.48
0.28
0.21
0.24
0.25
0.27
0.18
0.21
0.49
0.50
0.37
0.04
0.04
0.26
0.12
0.10
0.25
0.28
0.24
0.08
0.22
0.20
0.26
0.18
0.17
0.38
0.47
0.29
0.04
0.04
0.31
0.13
67
79
19
69
13
46
293
72
86
74
69
76
377
72
73
72
217
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
0.50
0.45
0.48
0.48
0.52
0.51
0.49
0.07
0.01
0.01
0.12
0.14
0.07
0.19
0.39
0.27
0.29
0.36
0.37
0.53
0.42′
0.69
0.67′
0.45
0.07
0.01
0.01
0.07′
0.12
0.06
0.15
0.34
0.29
0.26
60
80
19
70
13
49
291
70
82
72
70
72
366
68
73
69
210
6
5
3
7
3
9
12
5
8
7
12
11
14
12
13
10
16
0.10
0.12
0.15
0.32
0.34
0.25
0.20
0.49
0.56
0.56
0.78
0.82
0.67
0.85
0.84
0.82
0.86
0.07
0.10′
0.16
0.27′
0.39
0.25
0.18
0.50
0.52
0.61′
0.79
0.81
0.64
0.82
0.77
0.75
0.78
67
75
17
70
13
48
290
72
87
70
69
75
373
72
75
71
218
7
7
5
9
5
7
9
8
6
7
8
9
10
7
6
4
8
0.67
0.66
0.64
0.68
0.60
0.74
0.67
0.52
0.53
0.43
0.64
0.72
0.58
0.51
0.74
0.72
0.71
0.58
0.61
0.65
0.67
0.69
0.88
0.67
0.50
0.52
0.44
0.77
0.76
0.60
0.44
0.73
0.75
0.64
63.2
77.9
18.7
67.4
12.9
48.0
288
70.4
83
71.3
67.6
73.0
365.3
67.6
71.9
69.3
208.8
7.3
7.0
5.4
6.4
4.6
7.1
6.3
5.9
6.1
5.9
8.0
8.1
6.8
7.3
8.0
6.2
7.2
0.48
0.46
0.49
0.51
0.50
0.50
0.49
0.43
0.42
0.40
0.54
0.56
0.49
0.53
0.59
0.58
0.50
0.42*
0.43
0.50
0.44*
0.49
0.52′
0.45*
0.44
0.41′
0.40′
0.53′
0.55
0.46*
0.50*
0.56*
0.57
0.55*
© 2009 Blackwell Publishing Ltd