Range-wide genetic homogeneity in the California sea mussel ( Mytilus californianus ): a comparison of allozymes, nuclear DNA markers, and mitochondrial DNA sequences

Molecular Ecology (2008) 17, 4222–4232

doi: 10.1111/j.1365-294X.2008.03905.x

Range-wide genetic homogeneity in the California sea mussel (Mytilus californianus): a comparison of allozymes, nuclear DNA markers, and mitochondrial DNA sequences Blackwell Publishing Ltd

J A S O N A . A D D I S O N , B R I A N S . O RT ,* K AT H RY N A . M E S A and G R A N T H . P O G S O N Department of Ecology and Evolutionary Biology, University of California, Santa Cruz, CA 95064, USA

Abstract We tested for genetic differentiation among six populations of California sea mussels (Mytilus californianus) sampled across 4000 km of its geographical range by comparing patterns of variation at four independent types of genetic markers: allozymes, single-copy nuclear DNA markers, and DNA sequences from the male and female mitochondrial genomes. Despite our extensive sampling and genotyping efforts, we detected no significant differences among localities and no signal of isolation by distance suggesting that M. californianus is genetically homogeneous throughout its range. This concordance differs from similar studies on other mytilids, especially in the role of postsettlement selection generating differences between exposed coastal and estuarine habitats. To assess if this homogeneity was due to M. californianus not inhabiting estuarine environments, we reviewed studies comparing allozymes with other classes of nuclear DNA markers. Although both types of markers gave broadly consistent results, there was a bias favouring studies in which allozymes were more divergent than DNA markers (nine to three) and a disproportionate number of these cases involved marine taxa (seven). Furthermore, allozymes were significantly more heterogeneous than DNA markers in three of the four studies that sampled coastal and estuarine habitats. We conclude that the genetic uniformity exhibited by M. californianus may result from a combination of extensive gene flow and the lack of exposure to strong selective gradients across its range. Keywords: allozymes, female mitochondrial DNA, gene flow, male mitochondrial DNA, natural selection, population structure, scnDNA markers Received 4 June 2008; revision received 16 July 2008; accepted 17 July 2008

Most marine species with long-lived planktonic larvae exhibit low levels of population structuring consistent with high levels of ongoing gene flow (Waples 1998; Bohonak 1999; Grosberg & Cunningham 2001). However, an increasing number of reports have found that time spent in the plankton is not necessarily a good correlate with dispersal distance (e.g. Jones et al. 1999; Wirth & Bernatchez 2001; Taylor & Hellberg 2003; Veliz et al. 2006). Although studies detecting small but significant genetic differences among populations of ‘high gene flow’ marine species have forced a re-appraisal Correspondence: Jason A. Addison, Department of Biology, University of New Brunswick, PO Box 4400, Fredericton, NB, Canada E3B 5A3. E-mail: [email protected] *Present address: Department of Environmental Science, Policy and Management, University of California, Berkeley, CA 94720, USA

of the openness of marine systems and the relative importance of local recruitment, the biological significance of weak genetic structure remains unclear (Hedrick 1999). In the northeast Pacific, population genetic studies on near-shore marine taxa have revealed a variety of different patterns. Some studies have detected small but significant differences among populations but no clear geographical trend (e.g. Edmands et al. 1996; Moberg & Burton 2000; Miller et al. 2006), while others have found declines in genetic diversity with increasing latitude consistent with post-Pleistocene dispersal out of southern refugia (e.g. Edmands 2001; Hickerson & Ross 2001; Marko 2004; Hickerson & Cunningham 2005). Many taxa exhibit significant intraspecific phylogeographical structure that sometimes corresponds to well-known biogeographical boundaries such as Point Conception (see reviews by Burton 1998; © 2008 The Authors Journal compilation © 2008 Blackwell Publishing Ltd

G E N E T I C H O M O G E N E I T Y I N C A L I F O R N I A S E A M U S S E L S 4223 Dawson 2001), while others exhibit isolation by distance (e.g. Hellberg 1994) or signs of secondary contact and limited introgression (e.g. Marko 1998; Sotka et al. 2004). At the other extreme are species such as mussels (Mytilus spp.) that, on the basis of allozymes, exhibit no significant genetic structure along the entire western coast of North America (Levinton & Koehn 1976; Levinton & Suchanek 1978; Sarver & Foltz 1993). In situations where allozymes have suggested genetic uniformity, it is difficult to assess whether subtle population structure has been missed (due to low resolution) or if the patterns of variation have been influenced by natural selection. One way to examine both of these issues is to undertake comparisons between the allozymes and other presumably neutral DNA markers. Simulations have shown that this approach has considerable potential to identify loci experiencing either balancing or diversifying selection (McDonald 1994) and a growing number of studies have observed significant discrepancies between protein and DNA markers (e.g. Karl & Avise 1992; Pogson et al. 1995; Lemaire et al. 2000; Dufresne et al. 2002; see, however, McDonald et al. 1996). In other cases, highly variable microsatellites have revealed subtle genetic structure among populations that were indistinguishable for allozymes (e.g. Shaw et al. 1999; Jørgensen et al. 2005), although the biological significance of this limited divergence (FST = 0.006 and ΦCT = 0.0042, respectively) is not clear (see Waples 1998). However, concordance between allozymes and DNA markers appears to be the rule rather than the exception and, when discrepancies are observed, they are usually caused by the aberrant behaviour of one or two loci (see review by Allendorf & Seeb 2000). Here we re-evaluate earlier reports on the California sea mussel, Mytilus californianus, that observed no significant differences between populations sampled from southern California and southeast Alaska at two allozyme loci (Lap and Pgi) (Levinton & Koehn 1976; Levinton & Suchanek 1978). To assess whether the apparent absence of population structure was caused by the low mutation rates of the allozymes, we compared the protein-coding loci to partial sequences from four genes (nad2, cox1, atp6, and nad5) from the male (M) and female (F) mitochondrial DNA (mtDNA) molecules whose haplotype diversities approach 100% (Ort & Pogson 2007). To address the potential impact of natural selection, we compared six allozyme loci to an equal number of single-copy nuclear polymorphisms (scnDNA markers) developed from randomly selected anonymous complementary DNA (cDNA) clones. Comparisons between the allozymes and DNA markers allowed us to determine the importance of selection in shaping geographical patterns of variation at broad geographical scales and to assess if subtle genetic structure went undetected due to limited sampling of protein-coding loci with low mutation rates. © 2008 The Authors Journal compilation © 2008 Blackwell Publishing Ltd

Materials and methods Samples Samples of 150 adult mussels (3–10 cm shell height) were collected in 1999 or 2000 from mussel beds in the middle and southern parts of the species’ geographical range (Punta Baja, Mexico; Arroyo Hondo, California; Terrace Point, California; and Strawberry Hill, Oregon) as detailed in Engel (2004) and Ort & Pogson (2007) (Fig. 1). In the spring of 2000, two additional samples of 150 individuals were obtained from the northern part of the species’ range (Bamfield Marine Station, Canada, 48°50.5′N, 125°08.5′W and Elfin Cove, Alaska, 58°11.6′N, 136°21.2′W). For the allozyme analyses, mantle and posterior adductor muscle tissues were dissected, stored at –80 °C, and scored as described in Engel (2004). The F mtDNA sequences and the scnDNA markers were scored from total DNA isolated from ethanol-preserved gill tissue using the method of Pogson et al. (1995). Sequences from the male mtDNA genome were obtained from mitochondrially enriched total DNA isolated from male gonad tissue as described in Ort & Pogson (2007).

Fig. 1 Mytilus californianus collection sites with the average samples sizes (n) for the nuclear markers (allozyme/scnDNA).

4224 J . A . A D D I S O N E T A L .

Development of scnDNA markers A cDNA library was constructed in the phagemid vector Lambda ZAP II (Stratagene) from mRNA isolated from a mix of adult mantle, gill, and adductor muscle tissues. A total of 115 clones ranging in size from 0.5 to 1.7 kb were partially, or fully, sequenced using an Applied Biosystems Model 373 Automated DNA Sequencer. Sequences were checked against GenBank to eliminate ribosomal and mitochondrial genes. Polymerase chain reaction (PCR) primers were then constructed from a random sample of 24 anonymous clones and tested for their ability to amplify products from genomic DNA. PCR products ranging in size from 0.4 to 1.5 kb were consistently amplified from six primer sets, and these regions were selected for further detailed characterization. Two represented open reading frames (ORF) from anonymous exons (Mca20 and Mca125), one contained several small putative ORFs less than 70 amino acids (Mca146), while three contained a combination of introns and exons whose locations were deduced from the original cDNA sequences (Mca24, Mca28, and Mca39). The identity of genes tagged remains unknown except for clone Mca39, which exhibits a 92% amino acid identity (over 264 amino acids) with the receptor for activated C kinase 1 (RACK1) protein of Mya arenaria characterized by Siah et al. (2007). To identify polymorphisms within each gene region, PCR products from 14 to 18 individuals were sequenced from the Terrace Point population. Point mutations and indels (in introns) were commonly observed in all gene regions and those segregating at frequencies between 0.20 and 0.45 (representing point mutations in Mca20 and Mca125, and small indels in Mca24, Mca28, and Mca39) were targeted for construction of allele-specific PCR primers (Table 1). At each locus, we attempted to use the most common mutations amenable to the design of allele-specific primers. After careful optimization of allele-specific PCR conditions using genotypes identified from direct sequencing, we proceeded to score the population samples using both allele-specific primers paired with the same reverse/ forward primer. PCR products were resolved in 1% agarose gels and visualized using ethidium bromide. To confirm that the scnDNA markers did not belong to multigene families, we used the allele-specific primers to sequence alleles from various genotypes identified in all six populations. In all cases, the allele-specific primers generated the expected pair of haploid DNA sequences from heterozygotes and only one allelic class from homozygotes. Allele-specific primers could not be developed successfully for a sixth locus (Mca146). For this scnDNA marker, a single PCR product was obtained from all individuals and scored for a polymorphic DraI restriction site (identified by direct sequencing) following the manufacturer’s protocol (New England BioLabs). A subset of the highest quality cDNA

Table 1 Allele-specific primers and PCR conditions

Primer set

Sequence (5′–3′)

Mca20F Mca20R1 Mca20R2 Mca24F Mca24R1 Mca24R2 Mca28F1 Mca28F2 Mca28R Mca39R1 Mca39R2 Mca39F Mca125F1 Mca125F2 Mca125R Mca146F Mca146R

TCACGCTTCTAGTGTTCATCTG AGCTTTATTGTCTGGTATAAAA AGCTTTATTGTCTGGTATAAAG ACCATCAGGCTTAACAGGTCTG TAATCACACATGTTATCAAGG ATCACACATGTTATCATTGAC ACAAGCAGACATAGCTGCTAG CAGTAAAGGGGAATGCTAGG AGGTTGGYTTAGAAGCTGTAC ATGTTCACAATCGTGAATGAGT ATGTTCACAATCGTGAATGACA TAATGGGCTGAATCATTGGAC CATCAAGGGGAATTTCCAAAGA CATCAAGGGGAATTTCCAAAGT GGAATCTCATGCGTAGATAG ATCCTGGTGTGCATCAATGTGG AGACGTGCATTGTGTGACTCTCG

Annealing temperature (°C), extension time (min) 56°, 40′′

56°, 45′′

54°, 1′20′′

54°, 45′′

57°, 45′′

55°, 45′′

clones have been deposited in GenBank (accession nos FF339523–FF339585).

DNA sequencing Sequences from the female nad2 gene were obtained from the two northern populations (Bamfield, Canada and Elfin Cove, Alaska) using the procedures described in Ort & Pogson (2007). PCRs were performed using an Idaho Technology A1605 Air Thermo-Cycler in 10 μL sealed glass capillary tubes. After an initial denaturation step of 45 s at 94 °C, the tubes were exposed to 30 cycles of denaturation at 94 °C for 1 s, primer annealing at 54 °C for 1 s, primer extension at 72 °C for 50 s, and a final hold at 72 °C for 1 min. PCR products were gel purified using ZymoSpin columns (Zymo Research) and sequenced off both strands using an ABI PRISM 3100 Genetic Analyser. Sequences were edited and aligned using the SequenceNavigator and AutoAssembler programs (Applied Biosystems) and deposited in GenBank (accession nos DQ916911–DQ916948, DQ919109–DQ919117).

Statistical analyses Our comparative analyses were based on allele frequency data from six allozyme loci (Pgi, Lap, Odh, Idh1, Idh2, and Pgm) described in Engel (2004), allele frequency data from six anonymous scnDNA markers (Mca20, Mca24, Mca28, Mca39, Mca125, and Mca146), and partial sequences from four mitochondrial genes (nad2, nad5, cox1, and atp6) from the F and M molecules (Ort & Pogson 2007). Data for the allozymes, scnDNA markers and female nad2 gene were © 2008 The Authors Journal compilation © 2008 Blackwell Publishing Ltd

G E N E T I C H O M O G E N E I T Y I N C A L I F O R N I A S E A M U S S E L S 4225 obtained from all six populations. The complete set of M and F mtDNA sequences were collected only for the four most southerly populations as detailed in Ort & Pogson (2007). For the allozymes and scnDNA markers the GenePop program (version 3.4) of Raymond & Rousset (1995) was used to obtain expected heterozygosities (HE) and to perform tests for linkage disequilibrium and conformity to Hardy– Weinberg expectations (HWE). For the mtDNA sequence data haplotype diversities (h), numbers of segregating sites (S), nucleotide diversities (π), and Tajima’s D statistic were obtained by the DnaSP program (version 4.10) of Rozas et al. (2003). For the allozymes and scnDNA markers, Fstatistics (FIS, FIT, FST) were estimated using Weir and Cockerham’s method by the gda program (version 1.0) of Lewis & Zaykin (2001) and tested for significance (95% confidence intervals) by bootstrapping across loci (10 000 replications). We tested for pairwise differences between populations using the fstat (version 2.9.3) (Goudet 1995) and GenePop programs for the allozyme and scnDNA markers and the Arlequin (version 3.0) program (Excoffier et al. 2005) for the mtDNA sequences. To correct for the bias of comparing markers with different numbers of alleles (cf. McDonald 1994), all allozyme analyses were repeated after recoding the data as bi-allelic polymorphisms (i.e. homozygotes for the most common allele, heterozygotes for the most common allele, and others). We used the PowSim (version 4.0) program of Ryman & Palm (2006) to determine the statistical power of the allozyme and scnDNA markers to detect genetic differentiation given differences in their levels of allelic diversity and sample sizes. Empirical allele frequencies for the allozymes (treated as multiallelic and bi-allelic, see above) and scnDNA markers were used to define an initial base population that was then divided into six subpopulations each with effective sizes of 10 000 individuals. Simulations were run using default parameter values for dememorizations (1000), batches (100), and iterations per batch (1000) to predefined levels of population divergence (FST = 0.001, 0.0025, 0.005, 0.010, and 0.020) using the mean sample sizes of our study (allozymes n = 150; scnDNA markers n = 65). Each simulation was replicated 1000 times and power was determined as the proportion of simulations that Fisher’s exact test detected significant differences at the 0.05 level. To assess consistency, all simulations were repeated at least three times. We examined the relationships between the inferred levels of gene flow and the geographical distance between populations using the ibdws program (Jensen et al. 2005). Nonparametric Mantel tests were performed to test for nonrandom associations between matrices of genetic distances between all population pairs and matrices of pairwise geographical distances. Genetic distances were computed using Slatkin’s (1993) similarity index M-hat © 2008 The Authors Journal compilation © 2008 Blackwell Publishing Ltd

that was based on allele frequencies for the allozymes and scnDNA loci and Kimura 2-parameter distances for the mitochondrial sequences. Great circle distances were used to calculate the matrix of pairwise geographical separation. The significance of the major axis regression was assessed by 30 000 permutations of the data. Analyses of molecular variance (amovas) were performed using either Tamura–Nei (1993) distances (mtDNA) or conventional F-statistics (allozymes, scnDNA) in Arlequin and tested for significance with 5000 permutations. To assess whether the extremely high haplotype diversities of the F and M mitogenomes acted to constrain the detection of population structure (cf. Hedrick 1999), we sequentially pruned mutations from the female and male data sets (i.e. first removing singletons, then doubletons, etc.) performing amovas and monitoring haplotype diversities at each pruning level. The greater polymorphism present in the M genome necessitated higher levels of pruning than the F genome to reduce haplotype diversities by an arbitrary level of at least 10%.

Literature review Building on an earlier study by Allendorf & Seeb (2000), we reviewed the literature for studies comparing the degree of population structuring exhibited by allozymes and various types of nuclear DNA markers [typically microsatellites, random amplified polymorphic DNAs (RAPD), or amplified fragment length polymorphisms (AFLP)]. We used the modified Lewontin & Krakauer (1973) test developed by Pogson et al. (1995) to test for differences between the mean FST-values exhibited by the allozymes and DNA markers. Tests were undertaken only when the mean FST-values of one, or both, sets of markers were significantly greater than zero. Negative FST-values were set to zero before testing. To avoid situations where significant differences could have resulted from comparing populations sampled at different times and/or locations, we restricted our survey to studies in which both sets of markers were scored in the same individuals. Although two species (Salmo trutta and Oryza rufipogon) were involved in multiple studies (four and two, respectively), we included all six studies in our review because each involved comparisons between allozymes and different types of DNA markers, or compared similar marker types in populations sampled from different geographical locations.

Results Allozymes and scnDNA markers Single locus statistics for the allozymes and scnDNA markers are provided in Table 2. Heterozygosity levels for the allozyme loci varied considerably from a minimum of

4226 J . A . A D D I S O N E T A L . 0.090 at Idh2 to a maximum of 0.744 at Lap. In contrast, the scnDNA markers exhibited a narrower range of HE but the means were similar for both groups (allozymes HE = 0.433; scnDNA markers HE = 0.400). The two northern populations showed no evidence of reduced genetic diversity associated with postglacial population expansion (see Engel 2004). Significant deficits of heterozygotes were detected at four of the six allozyme loci (Odh, Idh1, Idh2, and Pgm) and the exact test of Raymond & Rousset (1995) indicated a global departure from HWE (FIS = 0.081, P < 0.001, Table 2). Similarly, two scnDNA markers (Mca20 and Mca125) exhibited significantly positive FIS values and, overall, the DNA markers also showed significant heterozygote deficiencies (FIS = 0.076, P < 0.001). As expected, collapsing the allozyme data into two-allele polymorphisms caused mean heterozygosity to drop (to 0.349) but did not affect

Table 2 Single locus statistics for the allozymes and scnDNA markers

Locus Allozymes Pgi Lap Odh Idh1 Idh2 Pgm MEANS scnDNA markers Mca20 Mca24 Mca28 Mca39 Mca125 Mca146 MEANS

n

No. of alleles

HE

FIS

FST

888 893 891 892 893 880 889.5

12 8 10 8 6 8 8.7

0.434 0.744 0.474 0.258 0.090 0.595 0.433

–0.010 0.041 0.079* 0.128* 0.291* 0.146* 0.081**

0.0011 –0.0021 0.0011 0.0001 –0.0027 0.0020 0.0002

553 317 343 285 319 472 381.5

2 2 2 2 2 2 2.0

0.298 0.477 0.366 0.499 0.419 0.343 0.400

0.111* 0.093 0.048 0.093 0.127* –0.040 0.076**

–0.0051 –0.0070 0.0047 –0.0095 0.0075 0.0086 –0.0007

*P < 0.05, **P < 0.001.

the expression of significant heterozygote deficits (mean FIS = 0.072, P < 0.001). After Bonferroni corrections for multiple tests (Rice 1989), no significant linkage disequilibrium was detected between any of the allozymes or scnDNA markers. No significant disequilibrium was observed between Odh and Lap, a known linkage group in other Mytilus species (see Beaumont 1994).

Mitochondrial DNA sequences We sequenced 726 bp of the female nad2 gene from 47 individuals collected from two northern populations (Alaska, n = 25; and Canada, n = 22) to compare with sequences previously obtained from the four southern populations (see Ort & Pogson 2007). Although haplotype and nucleotide diversities were slightly lower in the Canadian sample (Table 3), similar levels of polymorphism were observed among populations sampled throughout the species’ entire range. All populations exhibited significantly negative values of Tajima’s D at the nad2 locus, similar to that found at other genes in the F and M mtDNA molecules, reflecting a large excess of singleton and low frequency mutations (see Ort & Pogson 2007).

Population structure No significant differentiation was observed among the complete set of six populations for the allozymes (mean FST = 0.0002), scnDNA markers (mean FST = –0.0007), or the F nad2 gene (ΦST = –0.0058). Identical results were obtained for the allozyme loci collapsed into two-allele polymorphisms (mean FST = –0.0008). We also failed to detect significant genetic structure among the four southern populations using the 2307 bp of combined data from four genes (nad2, nad5, cox1, and atp6) from the M (ΦST = 0.0010) or F (ΦST = –0.0029) mtDNA molecules (Table 3). For the complete allozyme data, pairwise FST-values between samples were uniformly low (ranging from –0.0021–0.0016) and failed to produce consistent results using tests implemented by

Table 3 Single locus statistics for the mtDNA sequence data Population

Gene(s)

N

Size (bp)

h

S

π

Tajima’s D

ΦST

1. Alaska 2. Canada 3. Oregon 4. Terrace Point, CA 5. Arroyo Hondo, CA 6. Punta Baja All populations (1–6) Southern only (3–6) Southern only (3–6)

F nad2 F nad2 F nad2 F nad2 F nad2 F nad2 F nad2 F combined M combined

25 22 24 24 22 24 141 87 85

726 726 726 726 726 726 726 2307 2307

0.993 0.965 0.993 0.993 0.974 0.964 0.979 0.997 0.999

45 29 41 42 40 46 126 269 432

0.0063 0.0050 0.0062 0.0063 0.0057 0.0065 0.0060 0.0045 0.0122

–2.37* –2.15* –2.32* –2.28* –2.46** –2.45** –2.63** –2.76** –2.42**

na na na na na na –0.0058 –0.0029 0.0010

*P < 0.01, **P < 0.001. Data for the combined female (F) and male (M) sequences include partial sequences from the nad2, cox1, atp6, and nad5 genes as described in Ort & Pogson (2007). h, haplotype diversity; S, segregating sites; π, nucleotide diversity; na, not applicable. © 2008 The Authors Journal compilation © 2008 Blackwell Publishing Ltd

G E N E T I C H O M O G E N E I T Y I N C A L I F O R N I A S E A M U S S E L S 4227 Fig. 2 Effect of sequentially pruning mutations from the combined F (open circles) and M (closed circles) mtDNA sequence data on (a) haplotype diversities and (b) ΦST-values.

GenePop and fstat. No significant differences between population pairs were observed for the scnDNA markers or the allozymes coded as two-allele polymorphisms. Analyses of hierarchical population genetic structure with regional groupings based on known biogeographical boundaries (i.e. Point Conception) also revealed no evidence of population structure for the allozymes or scnDNA markers. Extreme levels of haplotype diversity, as observed in Mytilus californianus mtDNA (Ort & Pogson 2007), can obscure detection of population structure (Hedrick 1999). Pruning away the recent mutations has been observed to improve phylogenetic signal when inferring the evolutionary histories of humans (Wills 1996), warblers (Milot et al. 2000), and plant pathogenic fungi (Banke et al. 2004). To examine if high haplotype diversities were constraining the detection of subtle population structure, we modified this approach and sequentially pruned mutations from the combined mtDNA data, re-performing amovas at each pruning level. Pruning mutations from the complete male and female mtDNA data had the desired effect of reducing haplotype diversity (Fig. 2a) but had no impact on ΦST values (that tended to become even more negative at higher pruning levels; Fig. 2b). Hierarchical amovas for the mtDNA sequences were also not significant indicating that subtle population structure was not being obscured by high gene diversities. Power analyses indicated that the full allozyme data had the ability to detect a true FST of 0.0025 with a probability > 99% (Table 4). Collapsing the allozyme loci into two-allele polymorphisms had the predicted effect of reducing statistical power but the effect was weak; the bi-allelic coded allozymes could still identify weak population structure (FST = 0.0050) with a probability exceeding 96%. Compared to the allozymes, the scnDNA markers had reduced statistical power (due to their lower allelic diversities and smaller sample sizes) but could detect true FST values in the range of 0.005–0.0100 with reasonable power. Type I error rates (i.e. the probabilities of identifying significant genetic structure when the true FST was 0) were consistently about 5% in all simulations (Table 4). © 2008 The Authors Journal compilation © 2008 Blackwell Publishing Ltd

Table 4 Estimates of statistical power for detecting true levels of population structure (FST) by the allozyme and scnDNA markers. Power for the allozymes was determined using all alleles (complete data) and after re-coding the data as bi-allelic polymorphisms

True FST

Allozymes complete

Allozymes bi-allelic

scnDNA markers

0.0000 0.0010 0.0025 0.0050 0.0100 0.0200

0.043 0.779 0.998 1.000 1.000 1.000

0.041 0.306 0.729 0.967 0.999 1.000

0.051 0.141 0.328 0.676 0.944 0.999

Mantel tests provided no evidence for nonrandom associations between geographical distance and genetic similarity (M) for the scnDNA markers (r2 = 0.149, P = 0.996), allozymes (r2 = 0.008, P = 0.551), F nad2 (r2 = 0.106, P = 0.102), M combined (r2 = 0.107, P = 0.334), and F combined (r2 = 0.017, P = 0.415) mtDNA sequences.

Discussion It has been established by population genetics theory that the proper evaluation of genetic structure requires the use of multiple independent loci (e.g. Nei & Roychoudhury 1974; Nei 1978; Slatkin 1987). Our study on the California sea mussel, Mytilus californianus, represents one of the most comprehensive attempts to identify population structuring in its use of four independent types of genetic markers: allozymes, scnDNA markers, and DNA sequences from the female and male mtDNA genomes. All four markers failed to uncover any significant differentiation among mussel populations spanning 4000 km of coastline in the northeast Pacific. Given the reasonable power of our study to detect population structure (Table 4), we conclude that the California sea mussel is genetically homogeneous throughout its range. Such homogeneity is striking because, although the high vagility of many species suggests that

4228 J . A . A D D I S O N E T A L . genetic uniformity should be common in nature, in reality panmixia is rare and most species exhibit some level of differentiation among geographical localities (see Avise 2004). Our choice of markers was designed to control for two potential problems faced by allozymes: complications arising from selection and low resolution due to limited polymorphism. To assess the possible impact of natural selection, we developed a set of scnDNA markers from anonymous clones randomly isolated from a cDNA library. Since both the allozymes and scnDNA markers were tightly associated with transcribed regions of the genome, each had similar probabilities of being influenced by natural selection (either directly at the targeted gene region or indirectly via linkage disequilibrium with loci experiencing selection). The similar FIS and FST-values exhibited by the allozymes and scnDNA markers (Table 2) eliminate the possibility that natural selection has influenced any specific locus in an unusual manner. To determine if the limited polymorphism of the allozymes had constrained their ability to detect genetic structure, we examined mtDNA sequence variation at the highly polymorphic F and M mtDNA molecules. No significant genetic structure was observed at the female nad2 gene or at sequences obtained from four genes (nad2, nad5, cox1, and atp6) in the M or F mtDNA molecules. Sequentially pruning mutations from the combined data, an exercise that circumvented the problem of detecting significant structure using highly variable markers (Hedrick 1999), also failed to reveal significant heterogeneity (Fig. 2). Heterozygote deficiencies have been commonly observed at allozyme loci in marine bivalves but have yet to be adequately explained (see Zouros & Foltz 1984; Raymond et al. 1997; Addison & Hart 2005). In the present study, we observed nearly identical deficits of heterozygotes at both the protein and scnDNA markers. The similar magnitude of deficiencies at both marker types provides evidence against hypotheses explicitly directed at the protein level (i.e. viability selection, imprinting, or null alleles) or genotyping errors (Bonin et al. 2004; David et al. 2007). Furthermore, since M. californianus does not hybridize with other congeners in the northeast Pacific (i.e. Mytilus trossulus and the invasive Mytilus galloprovincialis), we can also rule out the possibility that the deficits resulted from introgressive hybridization. Our results suggest that the heterozygote deficiencies commonly reported in bivalves might relate to their breeding structures or, perhaps, to sweepstakes events (i.e. temporal Wahlund effects). The genetic uniformity exhibited by the California sea mussel is undoubtedly the result of high gene flow. There are a number of life-history traits of this species that should contribute to wide-scale dispersal. First, M. californianus has a long planktonic larval phase lasting up to 45 days (Strathman 1985) allowing for substantial movement in the water column before settlement and empirical studies have

confirmed that Mytilus spp. larvae can disperse up to 20–50 km (e.g. Koehn et al. 1980a; Hilbish & Koehn 1985; McQuaid & Phillips 2000; Gilg & Hilbish 2003; Becker et al. 2007). Second, although M. californianus exhibits peaks of spawning in the spring or fall (Coe & Fox 1942; Gosselin 2004), it spawns at low levels throughout the entire year (Young 1942; Suchanek 1981). This year-round spawning ensures that larvae are released under a wide range of oceanographic conditions thus avoiding limited and/or biased dispersal that may result from short breeding periods (see Hamm & Burton 2000). Third, since M. californianus is extremely long-lived, up to 50 years or more (Gosling 1992), multiple breeding efforts over many years should dampen longer-term heterogeneity in dispersal patterns mediated by either Pacific Decadal Oscillation (PDO) or El Niño Southern Oscillation (ENSO) events. Although the absence of population structuring in M. californianus may be a consequence of extensive gene flow, high dispersal does not necessarily guarantee genetic homogeneity. For example, along the eastern seaboard of North America the blue mussel, Mytilus edulis, exhibits little genetic structure among open coast populations to the north and south of Cape Cod (Koehn et al. 1976, 1984) but exhibits a sharp allele frequency cline at the aminopeptidase1 (Lap) locus at the entrance to Long Island Sound caused by strong postsettlement selection against the Lap94 allele (Koehn et al. 1980b; Hilbish & Koehn 1985). Differences between estuarine and oceanic populations have also been reported at the Lap locus in M. trossulus (McDonald & Siebenaller 1989), M. galloprovincialis (Gardner & Kathiravetpillai 1997; Gardner & Palmer 1998; see, however, Škalamera et al. 1999), Mercenaria mercenaria (Koehn et al. 1980a), and Modiolus modiolus (Koehn & Mitton 1972). If these earlier studies had compared the patterns of geographical variation at allozyme loci to neutral DNA markers, it is possible that differences may have been observed due to the aberrant behaviour of the Lap locus. It is notable that in the northeast Pacific M. californianus is restricted to open coasts while the native M. trossulus and the invasive M. galloprovincialis inhabit both estuarine and coastal habitats. Furthermore, the gradient in water temperatures along the northeast Pacific coast is far less than that along the eastern seaboard of North America (Powers & Shulte 1998; Schoch et al. 2006). These observations suggest that the genetic uniformity exhibited by M. californianus may not be solely a consequence of high gene flow but also to the absence of strong selective gradients experienced across its geographical range. To test this hypothesis, we undertook a literature review of studies comparing the degree of population structuring between allozyme and nuclear DNA markers (usually microsatellites, RAPDs, or AFLPs), extending an earlier report by Allendorf & Seeb (2000). We used the modified Lewontin & Krakauer (1973) test developed by Pogson et al. (1995) to test for © 2008 The Authors Journal compilation © 2008 Blackwell Publishing Ltd

Gomez-Uchida et al. (2003) Larsson et al. (2007) Lemaire et al. (2000) De Innocentiis et al. (2001) Evans et al. (2004) Mamuris et al. (1999) Vadopalas et al. (2004) Pampoulie et al. (2004) Bouza et al. (2002) Dufresne et al. (2002) Exadactylos et al. (1998, 2003) Geertjes et al. (2004) Diaz-Jaimes & Uribe-Alcocer (2003) no no yes no no no yes yes no yes no no no allozymes > AFLPs allozymes = microsatellites allozymes > microsatellites allozymes > microsatellites allozymes = microsatellites allozymes = RAPDs allozymes > microsatellites allozymes = microsatellites allozymes = microsatellites allozymes > microsatellites RAPDs > allozymes RAPDs > allozymes allozymes = RAPDs

© 2008 The Authors Journal compilation © 2008 Blackwell Publishing Ltd

*P < 0.05; **P < 0.01; ***P < 0.001; NS, not significant; NR, not reported.

0.026*** –0.001NS 0.339*** 0.214*** 0.031NR 0.035** 0.0024*** –0.008NS 0.002NS 0.047*** 0.026NS 0.019* 0.048** Cancer setosus Clupea harengus Dicentrarchus labrax Epinephelus marginatus Haliotis midae Mullus surmuletus Panopea abrupta Pomatoschistus minutus Scophthalmus maximus Semibalanus balanoides Solea solea Sparisoma viride Thunnus albacares Hairy edible crab Atlantic herring European sea bass Dusky grouper South African abalone Striped red mullet Geoduck clam Sand goby Turbot Acorn barnacle Dover sole Stoplight parrotfish Yellowfin tuna

0.007NS 0.006*** 0.017*** –0.002NS 0.033** 0.053*** 0.0009* 0.006** 0.002NS 0.007*** 0.193*** 0.044*** 0.030NS

Reference Estuary/lagoon comparisons? Mean divergence DNA markers Allozymes Species name Common name

Mean FST-values

Table 5 The subset of studies (13 of 60) comparing the magnitude of population genetic structure exhibited between allozyme and nuclear DNA markers in marine taxa. Species are listed alphabetically

G E N E T I C H O M O G E N E I T Y I N C A L I F O R N I A S E A M U S S E L S 4229 differences between the mean FST-values exhibited by the allozymes and DNA markers. To avoid spurious results, we restricted our survey to studies in which both sets of markers were scored in the same individuals. We identified 60 allozyme-DNA marker comparisons involving 55 species (1 amphibian, 10 invertebrates, 2 birds, 2 mammals, 15 fishes, and 25 plants; see Table S1, Supporting Information) and observed 12 cases where significant differences were found between the protein and DNA markers. Overall, three times as many studies observed that allozymes exhibited significantly greater mean FSTvalues than DNA markers than the converse (nine to three, respectively). Although highly variable markers such as microsatellites are expected under some conditions to exhibit lower FST-values than allozymes (Balloux et al. 2000; Balloux & Goudet 2002), this is unlikely to have contributed to this bias because, similar to that reported by Allendorf & Seeb (2000), the greater differentiation of the allozymes was usually caused by the aberrant behaviour of one or two loci [a notable exception being Lemaire et al.’s (2000) study on the Mediterranean sea bass]. We found that a significantly higher proportion of studies reporting differences between marker types involved marine (7 of 13) compared to nonmarine taxa (5 of 47, Fisher’s exact test, P = 0.0021; Table 5). Here, there was an even stronger bias towards allozymes being more differentiated than the DNA markers (five vs. two, respectively). Interestingly, in three of the four studies that included samples from both oceanic and estuarine/lagoon populations, the allozymes exhibited larger mean FST-values than the DNA markers. Although limited in number, these results support the hypothesis that the lack of exposure to strong selective gradients, rather than high gene flow per se, might be a key element underlying the broad-scale genetic homogeneity exhibited by M. californianus. In conclusion, all four types of genetic markers examined were unable to detect any significant differences among populations of the California sea mussel sampled throughout its geographical range. This striking genetic homogeneity may result from extensive gene flow and the absence of strong postsettlement selection in heterogeneous habitats. Our literature review confirmed that comparing the geographical patterns exhibited by allozymes and nuclear DNA markers represents a powerful way to detect natural selection. However, the limited numbers of studies finding discordant patterns (< 25%), and the small number of loci responsible for these discrepancies, suggest that the proportion of loci responding to heterogeneous selection pressures across complex landscapes may be rather low.

Acknowledgements We are grateful to Matt Edwards and Dan Monson for their assistance in collecting mussel samples, and to A. J. Addison and 2

4230 J . A . A D D I S O N E T A L . anonymous reviewers for providing helpful comments on the manuscript. Funding for the study was provided by the Monterey Bay Regional Studies (MBRS) program, the National Science Foundation (OCE-0350443 and DEB-0412976), and by the Partnership for the Interdisciplinary Study of Coastal Oceans (PISCO).

References Addison JA, Hart MW (2005) Spawning, copulation and inbreeding coefficients in marine invertebrates. Biology Letters, 1, 450–453. Allendorf FW, Seeb LW (2000) Concordance of genetic divergence among sockeye salmon populations at allozyme, nuclear DNA, and mitochondrial DNA markers. Evolution, 54, 640–651. Avise JC (2004) Molecular Markers, Natural History and Evolution, 2nd edn. Sinauer & Associates, Sunderland, Massachusetts. Balloux F, Goudet J (2002) Statistical properties of population differentiation estimators under stepwise mutation in a finite island model. Molecular Ecology, 11, 771–783. Balloux F, Brünner H, Lugon-Moulin N, Hausser J, Goudet J (2000) Microsatellites can be misleading: an empirical and simulation study. Evolution, 54, 1414–1422. Banke S, Peschon A, McDonald BA (2004) Phylogenetic analysis of globally distributed Mycosphaerella graminicola populations based on three DNA sequence loci. Fungal Genetics and Biology, 41, 226–238. Beaumont AR (1994) Linkage studies in Mytilus edulis, the mussel. Heredity, 72, 557–562. Becker BJ, Levin LA, Fodrie FJ, McMillan PA (2007) Complex larval connectivity patterns among marine invertebrate populations. Proceedings of the National Academy of Sciences, USA, 104, 3267–3272. Bohonak AJ (1999) Dispersal, gene flow, and population structure. Quarterly Review of Biology, 74, 21–45. Bonin A, Bellemain E, Bronken Eidesen P, Pompanon F, Brochmann C, Taberlet P (2004) How to track and assess genotyping errors in population genetics studies. Molecular Ecology, 13, 3261–3273. Bouza C, Presa P, Castro J, Sánchez L, Martínez P (2002) Allozyme and microsatellite diversity in natural and domestic populations of turbot (Scophthalmus maximus) in comparison with other Pleuronectiformes. Canadian Journal of Fisheries and Aquatic Sciences, 59, 1460–1473. Burton RS (1998) Intraspecific phylogeography across the Point Conception biogeographic boundary. Evolution, 52, 734–745. Coe WR, Fox KL (1942) Biology of the California sea-mussel (Mytilus californianus). I. Influence of temperature, food supply, sex and age on the rate of growth. Journal of Experimental Zoology, 90, 1–30. David P, Pujol B, Viard F, Castella V, Goudet J (2007) Reliable selfing rates estimated from imperfect population genetic data. Molecular Ecology, 16, 2474–2487. Dawson MN (2001) Phylogeography in coastal marine animals: a solution from California? Journal of Biogeography, 28, 723– 736. De Innocentiis S, Sola L, Cataudella S, Bentzen P (2001) Allozyme and microsatellite loci provide discordant estimates of population differentiation in the endangered dusky grouper (Epinephelus marginatus) within the Mediterranean Sea. Molecular Ecology, 10, 2163–2175. Diaz-Jaimes P, Uribe-Alcocer M (2003) Allozyme and RAPD variation in the eastern Pacific yellowfin tuna (Thunnus albacares). Fisheries Bulletin, 101, 769–777.

Dufresne F, Bourget E, Bernatchez L (2002) Differential patterns of spatial divergence in microsatellite and allozyme alleles: further evidence for locus-specific selection in the acorn barnacle, Semibalanus balanoides? Molecular Ecology, 11, 113–123. Edmands S (2001) Phylogeography of the intertidal copepod Tigriopus californicus reveals substantially reduced population differentiation at northern latitudes. Molecular Ecology, 10, 1743–1750. Edmands S, Moberg PE, Burton RS (1996) Allozyme and mitochondrial DNA evidence of population subdivision in the purple sea urchin Strongylocentrotus purpuratus. Marine Biology, 126, 443–450. Engel JD (2004) Population genetic structure of the California sea mussel: influence of the Pleistocene, biogeography, and micro-evolutionary processes. PhD Thesis, University of California, Santa Cruz, California. Evans BS, Sweijd NA, Bowie RCK, Cook PA, Elliott NG (2004) Population genetic structure of the perlemoen Haliotis midae in South Africa: evidence of range expansion and founder events. Marine Ecology Progress Series, 270, 163–172. Exadactylos A, Geffen AJ, Thorpe JP (1998) Population structure of the Dover sole, Solea solea L., in a background of high gene flow. Journal of Sea Research, 40, 117–129. Exadactylos A, Geffen AJ, Panagiotaki Thorpe JP (2003) Population structure of Dover sole Solea solea: RAPD and allozyme data indicate divergence in European stocks. Marine Ecology Progress Series, 246, 253–264. Excoffier L, Laval G, Schneider S (2005) Arlequin ver. 3.0: an integrated software package for population genetics data analysis. Evolutionary Bioinformatics Online, 1, 47–50. Gardner JPA, Kathiravetpillai G (1997) Biochemical genetic variation at a leucine aminopeptidase (LAP) locus in blue (Mytilus galloprovincialis) and greenshell (Perna canaliculus) mussel populations along a salinity gradient. Marine Biology, 128, 619– 625. Gardner JPA, Palmer NL (1998) Size-dependent spatial, and temporal genetic variation at a leucine aminopeptidase (Lap) locus among blue mussel (Mytilus galloprovincialis) populations along a salinity gradient. Marine Biology, 132, 275–281. Geertjes GJ, Postema J, Kamping A, van Delden W, Videler JJ, van de Zande L (2004) Allozymes and RAPDs detect little genetic population substructuring in the Caribbean stoplight parrotfish Sparisoma viride. Marine Ecology Progress Series, 279, 225–235. Gilg MR, Hilbish TJ (2003) The geography of marine larval dispersal: coupling genetics with fine-scale physical oceanography. Ecology, 84, 2989–2998. Gomez-Uchida D, Weetman D, Hauser L, Galleguillos R, Retamal M (2003) Allozyme and AFLP analyses of genetic population structure in the hairy edible crab Cancer setosus from the Chilean coast. Journal of Crustacean Biology, 23, 486–494. Gosling EM (1992) The Mussel Mytilus: Ecology, Physiology, Genetics, and Culture. Elsevier, Amsterdam. Gosselin LA (2004) Localised synchronous spawning of Mytilus californianus Conrad in Barkley Sound, British Columbia, Canada. Journal of Shellfish Research, 23, 529–533. Goudet J (1995) fstat version 1.2: a computer program to calculate F-statistics. Journal of Heredity, 86, 485–486. Grosberg RK, Cunningham CW (2001) Genetic structure in the sea: from populations to communities. In Marine Community Ecology (eds Bertness MD, Gaines SD, Hay ME), pp. 61–84 . Sinauer & Associates, Sunderland, Massachusetts. © 2008 The Authors Journal compilation © 2008 Blackwell Publishing Ltd

G E N E T I C H O M O G E N E I T Y I N C A L I F O R N I A S E A M U S S E L S 4231 Hamm DE, Burton RS (2000) Population genetics of black abalone, Haliotis cracherodii, along the central California coast. Journal of Experimental Marine Biology and Ecology, 254, 235–247. Hedrick PW (1999) Perspective: highly variable loci and their interpretation in evolution and conservation. Evolution, 53, 313– 318. Hellberg ME (1994) Relationships between inferred levels of gene flow and geographic distance in a philopatric coral, Balanophyllia elegans. Evolution, 48, 1829–1854. Hickerson MJ, Cunningham CW (2005) Contrasting quarternary histories in an ecologically divergent sister pair of low-dispersing intertidal fish (Xiphister) revealed by multilocus DNA analysis. Evolution, 59, 344–360. Hickerson MJ, Ross JRP (2001) Post-glacial population history and genetic structure of the northern clingfish (Gobbiesox maeandricus), revealed from mtDNA analysis. Marine Biology, 138, 407–419. Hilbish TJ, Koehn RK (1985) Dominance in physiological phenotypes and fitness at an enzyme locus. Science, 229, 52–54. Jensen JL, Bohonak AJ, Kelley ST (2005) Isolation by distance, web service. BMC Genetics, 6, 13. Jones PG, Milicich MJ, Emslie MJ, Lunow C (1999) Self-recruitment in a coral reef fish population. Nature, 402, 802–804. Jørgensen HBH, Hansen MM, Bekkevold D, Ruzzante DE, Loeschcke V (2005) Marine landscapes and population genetic structure of herring (Clupea harengus L.) in the Baltic Sea. Molecular Ecology, 14, 3219–3234. Karl SA, Avise JC (1992) Balancing selection at allozyme loci in oysters: implications from nuclear RFLPs. Science, 256, 100–102. Koehn RK, Mitton JB (1972) Population genetics of marine pelecypods. I. Ecological heterogeneity and evolutionary strategy at and enzyme locus. American Naturalist, 106, 47–56. Koehn RK, Milkman R, Mitton JB (1976) Population genetics of marine pelecypods. IV. Selection, migration and genetic differentiation in the blue mussel, Mytilus edulis. Evolution, 30, 2–32. Koehn RK, Hall JG, Zera AJ (1980a) Parallel variation of genotypedependent aminopeptidase-1 activity between Mytilus edulis and Mercenaria mercenaria. Marine Biology Letters, 1, 245–263. Koehn RK, Newell RIE, Immerman F (1980b) Maintenance of an aminopeptidase allele frequency cline by natural selection. Proceedings of the National Academy of Sciences, USA, 77, 5385– 5389. Koehn RK, Hall JG, Innes DJ, Zera AJ (1984) Genetic differentiation of Mytilus edulis in eastern North America. Marine Biology, 79, 117–126. Larsson LC, Laikre L, Palm S, André C, Carvalho GR, Ryman N (2007) Concordance of allozyme and microsatellite differentiation in a marine fish, but evidence of selection at a microsatellite locus. Molecular Ecology, 16, 1135–1147. Lemaire C, Allegrucci G, Naciri M, Bahri-Sfar L, Kara H, Bonhomme F (2000) Do discrepancies between microsatellite and allozyme variation reveal differential selection between sea and lagoon in the sea bass (Dicentrarchus labrax)? Molecular Ecology, 9, 457– 467. Levinton JS, Koehn RK (1976) Population genetics of mussels. In: Marine Mussels: Their Ecology and Physiology (eds Bayne BL), pp. 357–384. Cambridge University Press, Cambridge, UK. Levinton JS, Suchanek TH (1978) Geographic variation, niche breadth and genetic differentiation at different geographic scales in the mussels Mytilus californianus and M. edulis. Marine Biology, 49, 363–375. Lewis PO, Zaykin D (2001) Genetic Data Analysis: Computer Program for the Analysis of Allelic Data, version 1.0 (d16c). Free © 2008 The Authors Journal compilation © 2008 Blackwell Publishing Ltd

program distributed by the authors via internet from http:// lewis.eeb.uconn.edu/lewishome/software.html. Lewontin RC, Krakauer J (1973) Distribution of gene frequency as a test of the selective neutrality of polymorphisms. Genetics, 74, 175–195. Mamuris Z, Stamatis C, Triantaphyllidis C (1999) Intraspecific genetic variation of striped red mullet (Mullus surmuletus L.) in the Mediterranean Sea assessed by allozyme and random amplified polymorphic DNA (RAPD) analysis. Heredity, 83, 30– 38. Marko PB (1998) Historical allopatry and the biogeography of speciation in the prosobranch snail genus Nucella. Evolution, 52, 757–774. Marko PB (2004) ‘What’s larvae got to do with it?’ Disparate patterns of post-glacial population structure in two benthic marine gastropods with identical dispersal potential. Molecular Ecology, 13, 597–611. McDonald JH (1994) Detecting natural selection by comparing geographic variation in protein and DNA polymorphisms. In: Non-neutral Evolution: Theories and Molecular Data (eds Golding B), pp. 88–100. Chapman & Hall, New York. McDonald JH, Siebenaller JF (1989) Similar geographic variation at the LAP locus in the mussels Mytilus trossulus and M. edulis. Evolution, 43, 228–231. McDonald JH, Verelli BC, Geyer LB (1996) Lack of geographic variation in anonymous nuclear polymorphisms in the American oyster, Crassostrea virginica. Molecular Biology and Evolution, 13, 1114–1118. McQuaid CD, Phillips TE (2000) Limited wind-driven dispersal of intertidal mussel larvae: in situ evidence from the plankton and the spread of the invasive species Mytilus galloprovincialis in South Africa. Marine Ecology Progress Series, 201, 211–220. Miller KM, Supernault KJ, Li S, Withler RE (2006) Population structure in two marine invertebrate species Panopea abrupta and Strongylocentrotus franciscanus targeted for aquaculture enhancement in British Columbia. Journal of Shellfish Research, 25, 33–42. Milot E, Gibbs HL, Hobson KA (2000) Phylogeography and genetic structure of northern populations of the yellow warbler (Dendroica petechia). Molecular Ecology, 9, 667–681. Moberg PE, Burton RS (2000) Genetic heterogeneity among adult and recruit red sea urchins, Strongylocentrotus franciscanus. Marine Biology, 136, 773–784. Nei M (1978) Estimation of average heterozygosity and genetic distance from a small number of individuals. Genetics, 89, 583– 590. Nei M, Roychoudhury AK (1974) Sampling variance of heterozygosity and genetic distance. Genetics, 76, 379–390. Ort BS, Pogson GH (2007) Molecular population genetics of the male and female mtDNA molecules of the California sea mussel, Mytilus californianus. Genetics, 177, 1087–1099. Pampoulie C, Gysels ES, Maes GE et al. (2004) Evidence for finescale genetic structure and estuarine colonization in a potential high gene flow marine goby (Pomatoschistus minutus). Heredity, 92, 434–445. Pogson GH, Mesa KA, Boutilier RG (1995) Genetic population structure and gene flow in the Atlantic cod Gadus morhua: a comparison of allozyme and nuclear RFLP loci. Genetics, 139, 375–385. Powers DA, Shulte PM (1998) Evolutionary adaptations of gene structure and expression in natural populations in relation to a changing environment: a multidisciplinary approach to address

4232 J . A . A D D I S O N E T A L . the million-year saga of a small fish. Journal of Experimental Zoology, 282, 71–94. Raymond M, Rousset F (1995) GenePop (version 1.2): population genetics software for exact tests and ecumenicism. Journal of Heredity, 86, 248–249. Raymond M, Väänto RL, Thomas F, Rousset F, de Meeüs T, Renaud F (1997) Heterozygote deficiency in the Mytilus edulis species complex revisited. Marine Ecology Progress Series, 156, 225–237. Rice WR (1989) Analyzing tables of statistical tests. Evolution, 43, 223–225. Rozas J, Sánchez-DeIBarrio JC, Messeguer X, Rozas R (2003) DnaSP, DNA polymorphism analyses by the coalescent and other methods. Bioinformatics, 19, 2496–2497. Ryman N, Palm S (2006) PowSim: a computer program for assessing statistical power when testing for genetic differentiation. Molecular Ecology Notes, 6, 600–602. Sarver SK, Foltz DW (1993) Genetic population structure of a species’ complex of blue mussels (Mytilus spp.). Marine Biology, 117, 105–112. Schoch GC, Menge BA, Allison G, Kavanaugh M, Thompson SA, Wood SA (2006) Fifteen degrees of separation: latitudinal gradients of rocky intertidal biota along the California Current. Limnology and Oceanography, 51, 2564–2585. Shaw PW, Pierce GJ, Boyle PR (1999) Subtle population structuring within a highly vagile marine invertebrate, the veined squid Loligo forbesi, demonstrated with microsatellite DNA markers. Molecular Ecology, 8, 407–417. Siah A, Masson R, Loup B et al. (2007) Receptor activated C kinase is down-regulated in the male gonad of the marine bivalve mollusc Mya arenaria exposed to tributyltin (TBT). Aquatic Toxicology, 83, 295–305. Škalamera J-P, Renaud F, Raymond M, de Meuss T (1999) No evidence for genetic differentiation of the mussel Mytilus galloprovincialis between lagoons and the seaside. Marine Ecology Progress Series, 178, 251–258. Slatkin M (1987) Gene flow and the geographical structure of natural populations. Science, 236, 787–792. Slatkin M (1993) Isolation by distance in equilibrium and nonequilibrium populations. Evolution, 47, 264–279. Sotka EE, Wares JP, Barth JA, Grosberg RK, Palumbi SR (2004) Strong genetic clines and geographical variation in gene flow in the rocky intertidal barnacle Balanus glandula. Molecular Ecology, 13, 2143–2156. Strathman RR (1985) Feeding and non-feeding larval development and life-history evolution in marine invertebrates. Annual Review of Ecology and Systematics, 16, 339–361. Suchanek TH (1981) The role of disturbance in the evolution of life history strategies in the intertidal mussels Mytilus edulis and Mytilus californianus. Oecologia, 50, 143–152. Tamura K, Nei M (1993) Estimation of the number of nucleotide substitutions in the control region of mitochondrial DNA in

humans and chimpanzees. Molecular Biology and Evolution, 10, 512–526. Taylor MS, Hellberg ME (2003) Genetic evidence for local retention of pelagic larvae in a Caribbean reef fish. Science, 299, 107–109. Vadopalas B, LeClair LL, Bentzen P (2004) Microsatellite and allozyme analyses reveal few genetic differences among spatially distinct aggregations of Geoduck clams (Panopea abrupta, Conrad 1849). Journal of Shellfish Research, 23, 693–706. Veliz D, Duchesne P, Bourget E, Bernatchez L (2006) Genetic evidence for kin aggregation in the intertidal acorn barnacle (Semibalanus balanoides). Molecular Ecology, 15, 4193–4202. Waples RS (1998) Separating the wheat from the chaff: patterns of genetic differentiation in high gene flow species. Journal of Heredity, 89, 438–450. Wills C (1996) Topiary pruning and weighting reinforce an African origin for the human mitochondrial DNA tree. Evolution, 50, 977–989. Wirth T, Bernatchez L (2001) Genetic evidence against panmixia in the European eel. Nature, 409, 1037–1039. Young RT (1942) Spawning season of the California mussel, Mytilus californianus. Ecology, 23, 490–492. Zouros E, Foltz DW (1984) Possible explanations of heterozygote deficiency in bivalve molluscs. Malacologia, 25, 583–591.

Jason Addison works on the population genetics and phylogeography of marine invertebrates, with a special focus on echinoderms. Brian Ort studies the geographic distribution of genetic diversity in marine and island species of invertebrates and plants in relation to their restoration and conservation. Kathryn Mesa works on the population genetics of marine fishes and invertebrates. Grant Pogson’s research focuses on population genetic structure, adaptation, and speciation in the marine environment.

Supporting Information Additional Supporting Information may be found in the online version of this article. Table S1 Summary of studies comparing the magnitude of population structuring between allozymes and nuclear DNA markers. The mean divergence column indicates the concordance between mean FST or RST using the method of Pogson et al. (1995). Species names are listed alphabetically. Please note: Wiley-Blackwell are not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the Molecular Ecology Central Office.

© 2008 The Authors Journal compilation © 2008 Blackwell Publishing Ltd