Temporal genetic structure in a poecilogonous polychaete: the interplay of developmental mode and environmental stochasticity

This is an electronic reprint of the original article. This reprint may differ from the original in pagination and typographic detail. Author(s): Title: Year: Version: Please cite the original version: All material supplied via JYX is protected by copyright and other intellectual property rights, and duplication or sale of all or part of any of the repository collections is not permitted, except that material may be duplicated by you for your research use or educational purposes in electronic or print form. You must obtain permission for any other use. Electronic or print copies may not be offered, whether for sale or otherwise to anyone who is not an authorised user. Temporal genetic structure in a poecilogonous polychaete: the interplay of developmental mode and environmental stochasticity Kesäniemi, Jenni; Mustonen, Marina; Boström, Christoffer; Hansen, Benni W; Knott, Emily Kesäniemi, J., Mustonen, M., Boström, C., Hansen, B. W., & Knott, E. (2014). Temporal genetic structure in a poecilogonous polychaete: the interplay of developmental mode and environmental stochasticity. BMC Evolutionary Biology, 14(12). https://doi.org/10.1186/1471-2148-14-12 2014 RESEARCH ARTICLE Open Access Temporal genetic structure in a poecilogonous polychaete: the interplay of developmental mode and environmental stochasticity Jenni E Kesäniemi 1* , Marina Mustonen 1 , Christoffer Boström 2 , Benni W Hansen 3 and K Emily Knott 1 Abstract Background: Temporal variation in the genetic structure of populations can be caused by multiple factors, including natural selection, stochastic environmental variation, migration, or genetic drift. In benthic marine species, the developmental mode of larvae may indicate a possibility for temporal genetic variation: species with dispersive planktonic larvae are expected to be more likely to show temporal genetic variation than species with benthic or brooded non-dispersive larvae, due to differences in larval mortality and dispersal ability. We examined temporal genetic structure in populations of Pygospio elegans, a poecilogonous polychaete with within-species variation in developmental mode. P. elegans produces either planktonic, benthic, or intermediate larvae, varying both among and within populations, providing a within-species test of the generality of a relationship between temporal genetic variation and larval developmental mode. Results: In contrast to our expectations, our microsatellite analyses of P. elegans revealed temporal genetic stability in the UK population with planktonic larvae, whereas there was variation indicative of drift in temporal samples of the populations from the Baltic Sea, which have predominantly benthic and intermediate larvae. We also detected temporal variation in relatedness within these populations. A large temporal shift in genetic structure was detected in a population from the Netherlands, having multiple developmental modes. This shift could have been caused by local extiction due to extreme environmental conditions and (re)colonization by planktonic larvae from neighboring populations. Conclusions: In our study of P. elegans, temporal genetic variation appears to be due to not only larval developmental mode, but also the stochastic environment of adults. Large temporal genetic shifts may be more likely in marine intertidal habitats (e.g. North Sea and Wadden Sea) which are more prone to environmental stochasticity than the sub-tidal Baltic habitats. Sub-tidal and/or brackish (less saline) habitats may support smaller P. elegans populations and these may be more susceptible to the effects of random genetic drift. Moreover, higher frequencies of asexual reproduction and the benthic larval developmental mode in these populations leads to higher relatedness and contributes to drift. Our results indicate that a general relationship between larval developmental mode and temporal genetic variation may not exist. Keywords: Pygospio elegans, Poecilogony, Population genetics, Developmental mode, Genetic drift, Temporal, Sweepstakes reproductive success, Full-sibs * Correspondence: [email protected] 1 Department of Biological and Environmental Science, University of Jyväskylä, P.O. Box 35, Jyväskylä FI-40014, Finland Full list of author information is available at the end of the article © 2014 Kesäniemi et al.; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Kesäniemi et al. BMC Evolutionary Biology 2014, 14:12 http://www.biomedcentral.com/1471-2148/14/12 Background Several factors can affect the temporal stability of population genetic structure in the unpredictable marine environment. Unstable habitat can make a population more vulnerable to local extinctions and recolonizations [1,2]. Stochastic factors, such as variation in oceanic currents, can affect the movement of pelagic individuals and indirectly genetic patterns [3-5], or they can have a more direct effect, e.g. via high larval mortality. Especially in species with dispersive adults or larvae, temporal genetic differentiation could be caused by recruits migrating from different genetic sources at different times [6-10]. Sweepstakes reproductive success, extreme variation in the reproductive success of individuals, in which only a limited number of individuals contribute to the next generation, affects temporal population genetic structure in some species [11-14]. In small or fragmented populations, the enhanced effects of genetic drift also can lead to significant changes in temporal genetic structure [15-17]. Nevertheless, in most population genetic studies of marine invertebrates, only spatial patterns of genetic structure are examined without acknowledging the possibility for temporal genetic variation. In many cases, the developmental mode of larvae can be used to generalize the dispersal potential of marine species (see [18] for review). Consequently, it may also indicate when spatial differentiation between populations and temporal variation within populations is likely [13,19]. Populations of species developing via dispersive pelagic larvae may be particularly prone to temporal genetic variation since such larvae are known to face high mortality in the plankton [20-22] and populations will receive recruits from potentially many different sources. On the other hand, species with direct development (those lacking a larval stage) and species with benthic, brooded, or encapsulated larvae that do not disperse long distances may be less prone to temporal genetic variation since such larvae are thought to be more protected from predation [20,23,24] and to show higher local recruitment. Lee and Boulding [13] studied temporal genetic variation in four closely related Littorina gastropod species with different larval developmental modes. They found the expected pattern of significant temporal genetic variation in two species with planktotrophic (pelagic) larvae, whereas the species with direct-developing offspring were temporally stable. Even if a relationship between temporal stability of genetic structure and developmental mode exists, demographic differences between species and/or species-specific behaviours affecting larval recruitment could mask the general correlative patterns. Therefore, more data are needed before evaluating the generality of a relationship between temporal genetic variation and larval developmental mode. Poecilogonous species provide a means to examine this possible relationship without the influence of speciesspecific factors. Poecilogony refers to developmental mode polymorphism, in which there are multiple larval developmental modes within a single species [25,26]. Poecilogony is a rare phenomenon, known in some spionid polychaetes (e.g. [27,28]) and sacoglossan sea slugs [29]. In different poecilogonous species, individual females produce larvae developing via different developmental modes either simultaneously or seasonally, or the different modes are seen among multiple females either within or among populations (reviewed in [30]). Such variety implies that poecilogony could have arisen through different mechanisms in different species. These could include different genetic backgrounds, different environmental cues triggering the production of different larval types (developmental plasticity or bet-hedging), maternal effects, or a combination of these mechanisms (reviewed in [26,30]). For example, in Alderia willowi, poecilogony is based on reliable environmental cues, and females change the developmental mode of their larvae seasonally [29]. A polymorphic strategy may be favored in unpredictable habitats with spatial and temporal heterogeneity. In this case, poecilogony might be best described as a bet-hedging strategy, in which different phenotypes are produced in response to the unpredictability of the environment in an effort to maximize mean long term fitness [31,32]. Although poecilogony has not been proven to be a bet-hedging strategy (but see [30]), poecilogonous species are commonly found in intertidal habitats [25,27,28,33], which are characterized by rapid environmental fluctuations. Our study species, Pygospio elegans, is a poecilogonous spionid polychaete [34,35] commonly found in a variety of sub-tidal and intertidal habitats [34,36,37]. This cold adapted species is widely distributed in the Northern hemisphere and has wide salinity tolerance [38]. P. elegans has been described as opportunistic [34,39] and can reach high densities, especially in nutrient rich intertidal mud and sand flats [40,41]. However in some regions (i.e. the northern Baltic Sea), relatively low densities are observed [37,42,43]. Across its distribution, populations are often described as patchy, and worm densities in populations have been observed to fluctuate over time ([34,40], pers. obs. JEK and KEK). Adult P. elegans worms inhabit sand tubes and are relatively sedentary. After fertilization via direct transfer of spermatophores from males to females, the females lay their embryos inside egg capsules within the maternal tube (up to 34 capsules in an egg string, [36]). Larval developmental mode varies and is related to the number of embryos laid per capsule [34,36,42,44]. When the number of embryos is large (>20/capsule), the larvae have a short brooding period and a long pelagic period. Larvae with this planktonic developmental mode actively swim and feed in the plankton. In laboratory experiments, the pelagic period of these larvae was 1– Kesäniemi et al. BMC Evolutionary Biology 2014, 14:12 Page 2 of 16 http://www.biomedcentral.com/1471-2148/14/12 2 months, depending on the temperature [38]. When the number of embryos laid per capsule is small (1-2/ capsule), the larvae are brooded throughout their development within the egg capsules, feeding on nutritional nurse eggs (adelphophagy) provided at egg-laying by the mother. Larvae with this benthic developmental mode lack a pelagic stage and build their own sand tube soon after release from the capsules. An intermediate developmental mode with an intermediate brooding period and a short pelagic period is also known [35,44]. In addition, P. elegans can reproduce asexually by fragmentation [45], but asexual reproduction has not been observed in all populations [42]. The developmental mode polymorphism observed in P. elegans makes the species particularly suited for examining the relationship between developmental mode and temporal population genetic structure. In some populations multiple larval types have been observed simultaneously or seasonally [35,36,42], but many populations of P. elegans are known to produce only one larval type (e.g.[34,40,46]). In Europe, a broad scale pattern in developmental mode and environment can be seen: the planktonic developmental mode is more common in populations from marine intertidal sand and mud flats of the North Sea, whereas populations with longer brooding (intermediate and benthic modes) are commonly found in the (estuarine) subtidal habitats of the Baltic Sea [42]. However, variation in developmental mode in this species does not appear to be a plasticresponsetovariablesalinityortemperature[46]. Nevertheless, environmental characteristics could serve as cues which trigger the levels of polymorphism in different populations. In the Baltic Sea, P. elegans is often seen associated with Zostera marina sea grass [42,43]. At local scales, these vegetated areas can have more stabilized sediment and show higher species richness and abundance of individuals due to the stabilizing effect of plant roots, reduced water movement and lower predation risk (e.g. [37,47-49]). Intertidal mud and sand flats however, are unstable habitats, e.g. due to disturbance from tidal flow, currents and human impact, which can lead to desiccation stress, sediment transportation and fluctuations in salinity, temperature and oxygen availability [39,50,51]. Since the planktonic, benthic and intermediate larval developmental modes of P. elegans differ in the number of larvae produced per female, the duration of brooding time within capsules and the duration of the planktonic period, as well as in the size of larvae at release from the capsules and the presence or absence of larval swimming setae, the larvae developing via these different modes are expected to differ in their dispersal potential and their susceptibility to predation. As a result, developmental mode of larvae may affect the population genetic structure of P. elegans both spatially [42] and temporally. Although somewhat higher population genetic connectivity is found among P. elegans populations with planktonic larvae, overall a pattern of significant spatial population genetic structure and low connectivity has been found among European P. elegans populations [42]. Moreover, asexual reproduction or mixed strategies of reproduction could affect population genetic structure [42,52,53]. In this study, we examined temporal genetic structure in P. elegans populations and its relationship to developmental mode. Following previous observations [13], we expected to find greater temporal genetic variation in the population with strictly planktonic larvae, whereas temporal genetic stability was expected for populations with predominately the benthic developmental mode. The possibility for greater temporal genetic variation in the planktonic population is expected to stem from the very high mortality rates known for the planktonic larvae of P. elegans [22] and higher gene flow among populations with planktonic larvae [42]. We also expected populations to differ in estimates of effective population size and sibship depending on developmental mode and the prevalence of asexual reproduction, with larger N e and mostly unrelated individuals expected in those populations with primarily the planktonic developmental mode and lacking asexual reproduction. Our analysis of a poecilogonous species aims to clarify the connection between developmental mode and temporal population genetic variation and could shed light on the role of stochastic environmental variation on the evolution of marine invertebrate larvae. Methods Sample collection and DNA analyses P. elegans adults were collected from seven sites in Europe during a period of four years (2008–2011; 2–3 temporal samples/site, Table 1). In a previous analysis, samples from these sites collected in 2010 showed significant genetic structure [42], so we considered each site to host a distinct P. elegans population. Sampling sites differed in terms of salinity (measured with portable refractometer with ATC), worm density and in the larval developmental modes that we observed (Table 1). The five sites sampled in the Baltic Sea were all sub-tidal and had sandy substrates, but differed in depth. In the brackish Finnish archipelago (Northern Baltic Sea), sediment sampling was done at two sites (FIA and FIF) by scuba diving in approximately 3– 5 m deep water. In the Isefjord and Roskilde fjord estuary complex in Denmark (Zealand Island, Southern Baltic Sea), three sites were sampled by shoveling sediment from 50–100 cm deep water. Here, the sites also vary in exposure (DKR, located at the mouth of the estuary complex, is more exposed than DKV and DKH) as well as in salinity (Table 1). Marine intertidal sand and mud flats in the Wadden Sea (NET) and North Sea (UK) were also sampled. In both, sediment sampling was done during low tide when the Kesäniemi et al. BMC Evolutionary Biology 2014, 14:12 Page 3 of 16 http://www.biomedcentral.com/1471-2148/14/12 sediment and worm tubes were partly exposed. In the Netherlands, we sampled a mudflat from the mainland side of Schiermonnikoog Island (site NET) and in the UK, the Drum Sands sand flat (near Edinburgh, Scotland) was sampled (Table 1). Even though we did not measure density of P. elegans during our collection, the difference between the intertidal marine sites and the sub-tidal Baltic Sea sites was striking. In UK and NET, the worm density was high with densely packed P. elegans tubes. In contrast, the distribution of worms was patchy and worm density was noticeably lower in the Baltic Sea. Densities from 150 to 2800 ind. m 2 have been observed in the Northern Baltic Sea previously (pers. obs. CB), whereas 11 000 ind. m 2 is common at the UK site [40,41]. For all populations, collecting was done at the same location each year by the same person/s. At each location, approximately 15 sediment samples were collected, with a minimum distance of 0.4 m and maximum distance of 20 m between any two samples. Fine scale genetic patterns may be likely in species with restricted larval dispersal (e.g. [53]), however it is not expected in P. elegans at this spatial scale (see [54]). The sediment samples were gently sieved with a 1 mm mesh sieve, and the worms’sand tubes were removed with forceps and combined in a sampling bottle with sea water. In the laboratory, the tubes were placed in trays with sea water. After they emerged from their tubes, the worms were examined, sexed and checked for either sexual reproduction (indicated by the presence of gametes, which can be seen through the transparent body wall) or asexual reproduction (indicated by regeneration usually at both ends of the body, or multiple fragmented worm pieces within a tube). Sand tubes were also examined for the presence of egg capsules. When capsules were found, the developmental mode of the larvae was determined (Table 1). At site FIA we have observed only benthic larvae in egg capsules collected in early spring, but at site FIF we have not observed any sexually reproducing worms during our collections. All developmental modes have been observed in the Danish sites and NET, but benthic and intermediate modes predominate in the less saline environment in Denmark, whereas the planktonic developmental mode predominates in NET (Table 1). Only the planktonic developmental mode has been observed in UK [40,42]. Asexual reproduction in P. elegans was observed only in the Baltic Sea populations (Finland and Denmark) and was not frequent. These observations imply that developmental mode is associated with salinity, but because they are based on a limited number of samples taken primarily in spring, we cannot be sure that other developmental modes at these sites have not been overlooked. For example, we have visited the Finnish sites at different times during spring and summer, but have not observed gametes or larvae in the FIF individuals. Adult worms were preserved individually in ethanol until DNA extraction. DNA was extracted using Qiagen chemicals and a Kingfisher magnetic processor (Thermo Fisher Scientific), after which the samples were genotyped using eight highly polymorphic microsatellite markers following protocols described in [54]. Statistical analyses Genetic variation within each sample was estimated by calculating expected and observed heterozygosity and gene diversity with Arlequin v.3.5.1.2 [55]. Also, allelic richness was calculated for each sample using a rarefaction method with HP-RARE [56]. Rarefaction standardizes the estimate to a minimum sample size to allow for comparison of the estimates among samples from the different populations. To investigate variation between temporal samples (within populations), the number of private alleles (alleles only seen in one sample) was caluculated for each population separately. Due to different sample sizes, the number of private alleles can only be compared among the temporal samples within a single population and not between populations. Deviations from Hardy-Weinberg equilibrium were calculated with Arlequin, and p-values were adjusted with Bonferroni correction. FSTAT [57] was used to estimate F IS values for each sample and their Table 1 Sampling information Sea Population Code Years sampled (number of genotyped individuals) Salinity (psu) Worm density Observed developmental modes* Baltic Sea (st) Ängsö FIA 2008(53); 2009(52); 2010(42) 6-8 Low B Fårö FIF 2008(45); 2009(40); 2010(39) 6-8 Low - Vellerup DKV 2008(43); 2009(47); 2010(43) 20-21 Low to medium I, P, B Rörvig DKR 2009(42); 2010(40) 20-22 Medium I, B, P Herslev DKH 2008(24); 2010(42) 14 Low I, B Wadden Sea (it) Netherlands NET 2009(48); 2010(46); 2011(42) 28 High to very high P, I, B North Sea (it) Drum sands UK 2009(28); 2010(49) 28 High P Pygospio elegans sampling locations (st = sub-tidal, it = intertidal), population codes, and years sampled, with number of individuals genotyped per sample in parentheses. Salinity measured at each location, qualitative estimates of worm density and observed larval developmental modes are noted, with the most common developmental mode listed first. *Observed larval developmental modes: B = benthic, I = intermediate, P = planktonic. Kesäniemi et al. BMC Evolutionary Biology 2014, 14:12 Page 4 of 16 http://www.biomedcentral.com/1471-2148/14/12 statistical significance as well as to test for linkage disequilibrium between loci within the populations. Frequencies of null alleles were estimated using FreeNA [58]. To examine proportions of genetic variation explained by both spatial and temporal variation, analysis of molecular variance (AMOVA) was performed with two different sample sets using Arlequin (20 000 permutations). First, an analysis was carried out grouping the samples according to sample year (4 groups: 2008, 2009, 2010 and 2011). Second, the samples were grouped according to population (7 groups: temporal samples from the same population combined). To further investigate temporal genetic variation within the different populations, pair-wise F ST values (using Arlequin and 10 000 permutations) and Jost’s [59] pairwise Dest values were calculated (using the R package DEMEtics, 3000 permutations [60]) among the temporal samples within each population (temporal F ST / Dest). Pair-wise comparisons of F ST were also calculated among populations collected in the same year (2008, 2009, 2010). Linearized pair-wise F ST values were analyzed with PCA in GenAlEx v. 6.5 [61] to visualize differences among the samples and populations. After these analyses, we noticed a large temporal genetic shift in our samples from NET, as described in the results. To further investigate this shift, we used STRUCTURE v.2.3.4, [62] and also included some data reported in our previous spatial genetic analysis (populations from France and Netherlands, [42]) in order to identify potential source populations of the NET2011 sample. STRUCTURE was runassuminganadmixturemodel and correlated allele frequencies. A prior with sampling locality information was used, and the analysis consisted of a burn-in of 250,000 generations followed by MCMC sampling of 400,000 generations. One to six possible clusters were tested, each with four replicate runs. The lowest likelihood scores were used to evaluate the number of genetic clusters in the data, but the result of interest was the placement of the NET2011 sample relative to the other samples. STRUCTURE results were visualized with the program Distruct 1.1 [63]. Since the benthic larvae of P. elegans are not expected to disperse after emerging from the maternal sand tube, we expected that populations with predominately benthic developmental modes would also have a large number of related individuals. To test this hypothesis, we did a sibship analysis using COLONY [64] to identify full-sib families within each temporal sample from the different populations. The analysis was run with no information on parental genotypes, and assuming both male and female polygamy as well as possible inbreeding. Following suggestions from Wang [65], the full-likelihood model was run with run length and precision set to medium. Data from seven of the microsatellite loci were used (excluding Pe12, the locus with the highest proportion of suspected null alleles) and genotyping error rate was estimated to be 0.0001 for each locus. Although P. elegans is not expected to live longer than one year in nature [38], its ability to reproduce asexually allows extension of the life-span of some individual genotypes. Consequently, we also analyzed population data sets in which the data from temporal samples were combined. We did not expect to find full-sib families consisting of individuals sampled in different years, but hypothesized that cross-year families would be more common in those populations with asexual reproduction (Baltic Sea). If full-sib pairs or families were found, one individual from each pair was chosen randomly and removed and the descriptive analyses (H O ,H E and F IS ) and F ST calculations were repeated to investigate the effect of related individuals on population genetic patterns. We also estimated mean relatedness (r) values for each sample usingthemaximumlikelihoodmethodinML-Relate[66]. ML-Relate estimates relatedness for each individual pairwise comparison, and the mean of these (excluding selfcomparisons) was calculated for sample mean relatedness. Relatedness estimates were adjusted for the presence of null alleles. If the same site is sampled at two or more time points, a short term effective population size (N e ) can be calculated using temporal methods. These methods are based on changes in allele frequencies between the temporal samples [67]. Most methods used for calculating effective population sizes are based on assumptions of closed populations (no gene flow), discrete generations and random mating [68]. Although our populations are technically not closed, they are genetically differentiated from each other and show high self-recruitment rates [42].Therefore,weexpectedthatthedatawouldbeappropriate for the methods assuming closed populations. We first used the likelihood based method (MLNE) from Wang and Whitlock [69] to estimate N e assuming closed populations (no migration), but then we also estimated N e assuming open populations (migration allowed). In these analyses, we defined potential source populations as all the sites reported in this study, except when NET was the focal population. In this case, samples from other locations in the Netherlands and France, reported previously [42], were used as potential sources of gene flow (see results and discussion for more information). In all of the MLNE analyses, different maximum N e values were tested with no change in the results, and N e max = 10000 was used in the final analyses. Mean N e values over the sampling period were also estimated using the Moment Based Temporal method (MBT, [67]) implemented in NeEstimator [70]. TempoFs [71] was used to estimate genetic drift (observed allele frequency change) between the temporal samples. This method should produce unbiased results even with highly variable microsatellite markers. Mean Fs’(genetic drift corrected for sampling plan) over all loci was calculated Kesäniemi et al. BMC Evolutionary Biology 2014, 14:12 Page 5 of 16 http://www.biomedcentral.com/1471-2148/14/12 for the sample intervals. Sample plan 2 was used, in which samples are collected fatally before reproduction. In all of these temporal methods, the number of generations between temporal samples is required. For the Finnish sites (FIA and FIF), we estimated one generation per year (pers. obs. JEK), whereas for all other populations, two generations per year were assumed [36,40]. Again, since the life span of P. elegans is probably short in nature (worms lived for approximately one year in the laboratory [38]), the adults sampled in consecutive years are assumed to not be from the same cohort. Results Genetic diversity Genetic variation in the temporal samples is summarized in Table 2. Genetic diversity was relatively high in most populations, lowest in Finland (FIF, H E from 0.614 to 0.655) and highest in the UK (H E from 0.797 to 0.799). There were fluctuations in allelic richness and heterozygosity among samples within most populations, with UK, DKR and FIF showing the most consistent values among samples (Table 2). Fluctuations in diversity did not follow any particular pattern. For example, in DKV diversity (H E , gene diversity, allelic richness) was lower in 2010 compared to the previous years, but in DKH diversity was higher in 2010 than in previous years. Most fluctuations in genetic diversity within populations were not of high magnitude, except in DKV, where there was a decline in variation from 2008 to 2010, and in NET, where there was a noticeable increase in variation in the 2011 sample. The fluctuations among samples within a population are indicated best by the numbers of private alleles (alleles observed only in one temporal sample) which were calculated for each population separately Table 2 Genetic diversity and sibship patterns in the temporal samples Sample H O H E GD A R (N = 40) PrivateA R within site F IS FS families (prob >0.98) FIA2008 0.583 0.660 0.600 10.8 1.9 (N = 70) 0.116 1 FIA2009 0.570 0.664 0.653 10.7 1.3 0.140 1 FIA2010 0.688 0.714 0.680 11.5 2.1 0.015 0 All: 10 FIF2008 0.531 0.643 0.570 11.3 1.6 (N = 58) 0.153 0 FIF2009 0.478 0.614 0.558 10.0 1.5 0.202 0 FIF2010 0.567 0.655 0.588 11.0 1.8 0.128 0 All: 1 (no cross-year) DKR2009 0.584 0.696 0.644 10.3 3.7 (N = 56) 0.146 2* DKR2010 0.606 0.698 0.622 10.6 4.2 0.092 1 All: 5* DKV2008 0.650 0.771 0.700 12.7 3.3 (N = 74) 0.126 4 DKV2009 0.704 0.775 0.721 12.2 2.2 0.076 2* DKV2010 0.689 0.729 0.678 10.8 1.4 0.030 1 All: 9* DKH2008 0.596 0.677 0.596 10.0 3.5 (N = 40) 0.099 0 DKH2010 0.565 0.718 0.653 11.5 5.3 0.189 3 All: 4 (no cross-year) NET2009 0.586 0.703 0.616 11.4 2.3 (N = 72) 0.169 1 NET2010 0.627 0.717 0.669 11.4 2.0 0.124 4 NET2011 0.615 0.774 0.696 13.2 4.4 0.182 0 All: 4 (no cross-year) UK2009 0.606 0.799 0.765 13.2 4.4 (N = 52) 0.242 0 UK2010 0.664 0.797 0.738 13.7 5.3 0.165 0 All: 1 (no cross-year) Observed (H O ) and expected (H E ) heterozygosity, gene diversity (GD) and allelic richness (A R ) based on a sample size of 40 for each sample (here, N = number of genes used in the rarefaction method of HP-rare). Private allele richness (PrivateA R ) was calculated for temporal samples of each population separately (here, N [in italics] varies among populations, making the values comparable only among temporal samples within a population). Inbreeding coefficients (F IS , with significant values in italics) were calculated for each temporal sample. Number of full-sib families (FS) estimated within each temporal sample and within the combined dataset for each population (all) are listed. Samples marked with *have families consisting of more than two members (see text). In the combined data sets, cross-year full-sib families were found unless indicated. Kesäniemi et al. BMC Evolutionary Biology 2014, 14:12 Page 6 of 16 http://www.biomedcentral.com/1471-2148/14/12 and, unlike allelic richness, are not comparable among populations but only among temporal samples within a population. Most extreme increases in the number of private alleles were seen in DKH from 2008 to 2010 (3.5 to 5.3), and in NET from the 2009 and 2010 samples (2.0 and 2.3, respectively) to the 2011 sample (4.4, Table 2). Overall, allelic richness was highest in UK, the population with only planktonic larvae, in line with our findings from a previous study of additional populations from a broader geographic scale [42]. Heterozygote deficiencies were seen in most samples and significant deviations from Hardy-Weinberg equilibrium were observed in some loci in most populations as reflected by significant positive F IS values (Table 2). There were temporal changes in F IS values in many populations. As shown in previous analyses [42,54], null alleles were estimated to occur in either low (<0.05) or moderate frequencies (<0.20) in most loci in some of the samples (including locus Pe7 in FIF, DKH and NET, locus Pe17 in DKV, DKR, DKH, NET and UK, locus Pe18 in DKH, NET and UK, locus Pe13 in FIA, FIF, NET and UK, and Pe19 in NET). Loci Pe15 and Pe12 were the most problematic, showing moderate null allele frequencies in almost all temporal samples. However, since the global F ST and the F ST values for each locus (data not shown) were similar when calculated with (ENA) and without estimating a null allele correction (F ST =0.0306, F ST ENA = 0.0308), we feel confident that the presence of null alleles has not affected our calculations of genetic structure. Regardless, Pe12 was eliminated from the data set when estimating the number of full-sib families with COLONY and when estimating the effective population size. Significant linkage disequilibrium was observed in only a few comparisons within some of the Danish populations (Pe7xPe15 in DKH & DKV, Pe7xPe17 in DKV, Pe7xPe13 in DKR, DKV & DKH, and Pe13xPe12 in DKR, Pe7xPe12 in DKV), occurring within some years, but not others. Although asexual reproduction is possible in P. elegans, we observed only two instances of identical multilocus genotypes in our samples: one pair of identical individuals in the DKV2008 sample and one pair of individuals in the DKR2010 sample. One individual from each of these pairs was removed from the COLONY, STRUCTURE and PCA analyses. Temporal and spatial genetic structure AMOVA analyses (Table 3) indicated that the variation among the different populations was significant (3.61% of variation explained), and that this population structure was similar in different years (AMOVA grouping samples by year showed non-significant variation among years). Similarly, global F ST values in different years were not significantly different (2008: 0.034, 2009: 0.046, 2010: 0.029, P = 0.617). Genetic structure between populations was also evident in our analysis of (within year) population pairwise F ST but there was temporal variation in structure among the Baltic Sea populations (Additional file 1). For example, in Denmark, most population pair-wise comparisons within a year were significant, however in 2010, DKV and DKH have non-significant pair-wise F ST .InFinland, the pair-wise F ST values between FIA and FIF are low each year (F ST < 0.01), and significant in 2008 and 2010, but not in 2009 (Additional file 1). When the temporal samples were grouped according to population in AMOVA (Table 3), we found not only significant among population (spatial) variation (2.85% of variation explained), but also low but significant temporal variation within populations (0.85%). To clarify how specific temporal samples differed, we examined the samples from each population separately in pair-wise comparisons of F ST and Dest (Table 4). In these analyses, significant differences were found between temporal samples in some but not all populations, and the F ST and Dest values were generally low. There were no differences in temporal samples from UK and DKH. When significant differences between temporal samples were indicated, these were found between samples taken in subsequent years (2009 vs. 2010 in FIF, NET, DKR and DKV) as well as between samples from longer time periods (2008 vs. 2010 in FIA). Analysis of either F ST or Dest indicated the same patterns of differentiation, except in DKR (Table 4). The largest differences among temporal samples was seen in the NET population, in which the 2011 sample was significantly differentiated from the two previous samples (2009 and 2010), which did not differ from each other (F ST and Dest). Likewise, PCA analysis of linearized F ST showed that in most cases, temporal samples from the same population clustered together (Additional file 2). However, NET2011 did not cluster with other temporal samples from NET and was placed between clusters of the Danish and UK samples, not grouping clearly with either (Additional file 2). We further investigated the differentiation of the NET2011 sample by performing additional pair-wise comparisons of F ST and Dest with data from other nearby populations with planktonic larvae (from Netherlands and France) that had been sampled only in 2010 and were not the focus of this analysis (data from [42]). We found that although the NET2011 sample was significantly differentiated from earlier samples at the same site, it was not differentiated from other Dutch populations (NLH2010 and NLB2010) and French populations (FRS2010, and FRC2010 only in Dest) sampled in 2010 (Table 5). Moreover, we examined pair-wise F ST and Dest between NET2011 and the 2010 samples collected from Denmark and UK (this study). Comparisons with the Danish samples indicated significant genetic differentiation, but with UK, the results were inconsistent (significant F ST , non-significant Dest) (Table 5). STRUCTURE showed clear assignment of NET2011 individuals to a different cluster than the other temporal Kesäniemi et al. BMC Evolutionary Biology 2014, 14:12 Page 7 of 16 http://www.biomedcentral.com/1471-2148/14/12 samples from NET. Also, the analysis showed a shared cluster membership of NET2011 with the other Dutch and French populations sampled in 2010, which were distinct from the UK samples (K = 3 had the best likelihood, Figure 1). Sibship analysis and relatedness Full-sibs were found within almost all temporal samples, but not in FIF and UK (Table 2). In most samples, fullsibs were pairs of individuals, but larger families were found in two Danish populations. In DKR2009 and DKV2009 samples, full-sib families of five and three individuals, respectively, were identified. In NET, full-sib pairs were seen in 2009 (one pair) and 2010 (four pairs), but none were observed in the 2011 sample. When temporal samples within a population were combined to identify cross-year full-sibs, typically the same full-sib families were identified as in the analyses of each temporal sample separately, but also additional full-sib families were seen. However, cross-year full-sibs (full-sib pairs or families consisting of individuals from different temporal samples) were found only in three Baltic Sea populations (FIA, DKR and DKV). In the analyses of DKR and DKV, in addition to full-sib pairs, two larger cross-year full-sib families were identified (DKR: families of five and four individuals, DKV: families of three and four individuals). Even though no full-sib families were identified in the temporal samples in FIF and UK when they were analyzed separately, in each of these populations one full-sib pair was identified when the temporal samples were combined. In both cases, the full-sibs identified were individulas within a single temporal sample (no cross-year full-sibs), indicating the effect of sample size on identifying potentially highly related individuals. Table 3 AMOVA results of spatial and temporal variation in Pygospio elegans Temporal groups (samples grouped by year) Source of variation Sum of squares Variance components Percentage of variation Fixation indices (P-value) Among years 22.3 −0.006 −0.27 F CT =−0.003 (0.816) Among populations within years 132.2 0.085 3.61 F SC = 0.036 (<0.001) Within samples 3436.3 2.273 96.66 F ST = 0.033 (<0.001) Total 3590.8 2.351 Geographical groups (samples grouped by population) Source of variation Sum of squares Variance components Percentage of variation Fixation indices (P-value) Among populations 110.9 0.067 2.85 F CT = 0.029 (<0.001) Among temporal samples within populations 43.6 0.020 0.85 F SC = 0.009 (<0.001) Within samples 3436.3 2.273 96.3 F ST = 0.037 (<0.001) Total 3590.8 2.360 Temporal groups consist of samples from different populations grouped by sampling year (4 groups: 2008, 2009, 2010, 2011) and spatial groups consist of temporal samples grouped according to population (7 groups). Table 4 Temporal pair-wise F ST /Dest values for samples within each population Finland FIA 2008 2009 FIF 2008 2009 2009 0.002/0.018 2009 0.007/0.065 2010 0.008*/0.055* 0.004/0.048 2010 0.004/0.040 0.011*/0.112** Denmark DKR 2009 DKV 2008 2009 2010 0.008*/0.041 2009 0.002/0.006 2010 0.005/0.045 0.007*/0.061** DKH 2008 2010 0.007/0.062 Netherlands NET 2009 2010 UK 2009 2010 0.001/0.024 2010 −0.002/-0.001 2011 0.038***/0.223*** 0.031***/0.149*** *P < 0.05, **P < 0.01, ***P < 0.001. Kesäniemi et al. BMC Evolutionary Biology 2014, 14:12 Page 8 of 16 http://www.biomedcentral.com/1471-2148/14/12 26. Knott KE, McHugh D: Introduction to symposium: poecilogony-a window on larval evolutionary transitions in marine invertebrates. Integ Comp Biol 2012, 52(1):120–127. 27. Levin L: Multiple patterns of development in Streblospio Benedicti Webster (Spionidae) from 3 coasts of North America. Biol Bull 1984, 166(3):494–508. 28. Gibson GD: Variable development in the spionid Boccardia proboscidea (Polychaeta) is linked to nurse egg production and larval trophic mode. Invertebr Biol 1997, 116(3):213–226. 29. Krug PJ: Poecilogony and larval ecology in the gastropod genus Alderia. Am Malacol Bull 2007, 23(1–2):99–111. 30. Collin R: Nontraditional life-history choices: what can “intermediates”tell us about evolutionary transitions between modes of invertebrate development? Integr Comp Biol 2012, 52(1):128–137. 31. Simons AM: Modes of response to environmental change and the elusive empirical evidence for bet hedging. Proc R Soc B, Biological Sciences 2011, 278(1712):1601–1609. 32. Wennersten L, Forsman A: Population-level consequences of polymorphism, plasticity and randomized phenotype switching: a review of predictions. Biol Rev 2012, 87(3):756–767. 33. Levin L: Life-history and dispersal patterns in a dense infaunal polychaete assemblage - community structure and response to disturbance. Ecology 1984, 65(4):1185–1200. 34. Morgan T, Rogers A, Paterson G, Hawkins L, Sheader M: Evidence for poecilogony in Pygospio elegans (Polychaeta: Spionidae). Mar Ecol Prog Ser 1999, 178:121–132. 35. Kesäniemi JE, Rawson PD, Lindsay SM, Knott KE: Phylogenetic analysis of cryptic speciation in the polychaete Pygospio elegans.Ecol Evol 2012, 2(5):994–1007. 36. Rasmussen E: Systematics and ecology of the Isefjord marine fauna (Denmark). Ophelia 1973, 11:1–495. 37. Boström C, Bonsdorff E: Community structure and spatial variation of benthic invertebrates associated with Zostera marina (L.) beds in the northern Baltic Sea. J Sea Res 1997, 37((1):153–166. 38. Anger K, Anger V, Hagmeier E: Laboratory studies on larval growth of Polydora ligni,Polydora ciliata,andPygospio elegans (Polychaeta, Spionidae). Helgoländer Meeresunters 1986, 40(4):377–395. 39. Desprez M, Rybarczyk H, Wilson J, Ducrotoy J, Sueur F, Olivesi R, Elkaim B: Biological impact of eutrophication in the Bay of Somme and the induction and impact of anoxia. Neth J Sea Res 1992, 30:149–159. 40. Bolam SG: Population structure and reproductive biology of Pygospio elegans (Polychaeta: Spionidae) on an intertidal sandflat, Firth of Forth, Scotland. Invertebr Biol 2004, 123(3):260–268. 41. Bolam SG, Fernandes TF: Dense aggregations of Pygospio elegans (Claparède): effect on macrofaunal community structure and sediments. J Sea Res 2003, 49(3):171–185. 42. Kesäniemi JE, Geuverink E, Knott KE: Polymorphism in developmental mode and its effect on population genetic structure of a spionid polychaete. Pygospio elegans. Integr Comp Biol 2012, 52(1):181–196. 43. Boström C, Bonsdorff E: Zoobenthic community establishment and habitat complexity-the importance of seagrass shoot-density, morphology and physical disturbance for faunal recruitment. Mar Ecol Prog Ser 2000, 205:123–138. 44. Gudmundsson H: Life history patterns of polychaete species of the family Spionidae. J Mar Biol Assoc U K 1985, 65(01):93–111. 45. Rasmussen E: Asexual reproduction in Pygospio elegans Claparede (Polychaeta, Sedentaria). Nature 1953, 171(4365):1161–1162. 46. Anger V: Reproduction in Pygospio elegans (Spionidae) in relation to its geographical origin and to environmental conditions: a preliminary report. Fortschr Zool 1984, 29:45–51. 47. Constable AJ: Ecology of benthic macro‐invertebrates in soft‐sediment environments: a review of progress towards quantitative models and predictions. Aust J Ecol 1999, 24(4):452–476. 48. Lee S, Fong C, Wu R: The effects of seagrass (Zostera japonica) canopy structure on associated fauna: a study using artificial seagrass units and sampling of natural beds. J Exp Mar Biol Ecol 2001, 259(1):23–50. 49. Fredriksen S, De Backer A, Boström C, Christie H: Infauna from Zostera marina L. meadows in Norway. Differences in vegetated and unvegetated areas. Mar Biol Res 2010, 6(2)):189–200. 50. Moran A, Emlet R: Offspring size and performance in variable environments: field studies on a marine snail. Ecology 2001, 82(6):1597–1612. 51. Zajac R: Macrofaunal responses to pit-mound patch dynamics in an intertidal mudflat: local versus patch-type effects. JExpMarBiolEcol 2004, 313(2):297–315. 52. Calderón I, Ortega N, Duran S, Becerro M, Pascual M, Turon X: Finding the relevant scale: clonality and genetic structure in a marine invertebrate (Crambe crambe, Porifera). Mol Ecol 2007, 16(9):1799–1810. 53. Barbosa SS, Klanten SO, Puritz JB, Toonen RJ, Byrne M: Very fine-scale population genetic structure of sympatric asterinid sea stars with benthic and pelagic larvae: influence of mating system and dispersal potential. Biol J Linn Soc 2013, 108(4):821–833. 54. Kesäniemi JE, Boström C, Knott KE: New genetic markers reveal population genetic structure at different spatial scales in the opportunistic polychaete Pygospio elegans.Hydrobiologia 2012, 691(1):213–223. 55. Excoffier L, Lischer HE: Arlequin suite ver 3.5: a new series of programs to perform population genetics analyses under linux and windows. Mol Ecol Resour 2010, 10(3):564–567. 56. Kalinowski S: HP-RARE 1.0: a computer program for performing rarefaction on measures of allelic richness. Mol Ecol Notes 2005, 5(1):187–189. 57. Goudet J: FSTAT (Version 1.2): a computer program to calculate F-statistics. J Hered 1995, 86(6):485–486. 58. Chapuis M, Estoup A: Microsatellite null alleles and estimation of population differentiation. Mol Biol Evol 2007, 24(3):621–631. 59. Jost L: G ST and its relatives do not measure differentiation. Mol Ecol 2008, 17:4015–4026. 60. Gerlach G, Jueterbock A, Kraemer P, Deppermann J, Harmand P: Calculations of population differentiation based on GST and D: forget GST but not all of statistics! Mol Ecol 2010, 19(18):3845–3852. 61. Peakall R, Smouse PE: GenAlEx 6.5: genetic analysis in Excel. Population genetic software for teaching and research-an update. Bioinformatics 2012, 28(19):2537–2539. 62. Pritchard JK, Stephens M, Donnelly P: Inference of population structure using multilocus genotype data. Genetics 2000, 115:945–959. 63. Rosenberg N: DISTRUCT: a program for the graphical display of population structure. Mol Ecol Notes 2004, 4(1):137–138. 64. Jones OR, Wang J: COLONY: a program for parentage and sibship inference from multilocus genotype data. Mol Ecol Resour 2010, 10(3):551–555. 65. Wang J: Computationally efficient sibship and parentage assignment from multilocus marker data. Genetics 2012, 191(1):183–194. 66. Kalinowski ST, Wagner AP, Taper ML: ml-relate: a computer program for maximum likelihood estimation of relatedness and relationship. Molr Ecol Notes 2006, 6((2):576–579. 67. Waples R: A generalized-approach for estimating effective population-size from temporal changes in allele frequency. Genetics 1989, 121(2):379–391. 68. Waples RS: Spatial-temporal stratifications in natural populations and how they affect understanding and estimation of effective population size. Mol Ecol Resour 2010, 10(5):785–796. 69. Wang J, Whitlock M: Estimating effective population size and migration rates from genetic samples over space and time. Genetics 2003, 163(1):429–446. 70. Ovenden JR, Peel D, Street R, Courtney AJ, Hoyle SD, Peel SL, Podlich H: The genetic effective and adult census size of an Australian population of tiger prawns (Penaeus esculentus). Mol Ecol 2007, 16(1):127–138. 71. Jorde PE, Ryman N: Unbiased estimator for genetic drift and effective population size. Genetics 2007, 177(2):927–935. 72. Ostergaard S, Hansen M, Loeschcke V, Nielsen E: Long-term temporal changes of genetic composition in brown trout (Salmo trutta L.) populations inhabiting an unstable environment. Mol Ecol 2003, 12(11):3123–3135. 73. Ng W, Leung FCC, Chak STC, Slingsby G, Williams GA: Temporal genetic variation in populations of the limpet Cellana grata from Hong Kong shores. Mar Biol 2010, 157(2):325–337. 74. Robainas-Barcia A, Lopez G, Hernandez D, García-Machado E: Temporal variation of the population structure and genetic diversity of Farfantepenaeus notialis assessed by allozyme loci. Mol Ecol 2005, 14(10):2933–2942. 75. Barshis D, Sotka E, Kelly R, Sivasundar A, Menge B, Barth J, Palumbi S: Coastal upwelling is linked to temporal genetic variability in the acorn barnacle Balanus glandula.Mar Ecol Prog Ser 2011, 439:139–150. 76. Strathmann R: Hypotheses on the origins of marine larvae. Annu Rev Ecol Syst 1993, 24:89–117. 77. Levin L, Huggett D: Implications of alternative reproductive modes for seasonality and demography in an estuarine polychaete. Ecology 1990, 71(6):2191–2208. Kesäniemi et al. BMC Evolutionary Biology 2014, 14:12 Page 15 of 16 http://www.biomedcentral.com/1471-2148/14/12 78. Beukema JJ, Flach EC, Dekker R, Starink M: A long-term study of the recovery of the macrozoobenthos on large defaunated plots on a tidal flat in the Wadden Sea. JSeaRes1999, 42(3):235–254. 79. Bolam SG, Fernandes TF: Dense aggregations of tube-building polychaetes: response to small-scale disturbances. JExpMarBiolEcol 2002, 269(2):197–222. 80. Pannell J, Charlesworth B: Effects of metapopulation processes on measures of genetic diversity. Phil Trans R Soc B-Biological Sciences 2000, 355(1404):1851–1864. 81. Reiss H, Meybohm K, Kroencke I: Cold winter effects on benthic macrofauna communities in near- and offshore regions of the North Sea. Helgol Mar Res 2006, 60(3):224–238. 82. Neumann H, Ehrich S, Kroencke I: Effects of cold winters and climate on the temporal variability of an epibenthic community in the German Bight. Climate Res 2008, 37(2–3):241–251. 83. Naumov AD: Long-term fluctuations of soft-bottom intertidal community structure affected by ice cover at two small sea bights in the Chupa Inlet (Kandalaksha Bay) of the White Sea. Hydrobiologia 2013, 706(1):159–173. 84. Hanski IA, Gaggiotti OE: Ecology, Genetics and Evolution of Metapopulations. USA: Elsevier Academic Press; 2004. 85. Swearer S, Shima J, Hellberg M, Thorrold S, Jones G, Robertson D, Morgan S, Selkoe K, Ruiz G, Warner R: Evidence of self-recruitment in demersal marine populations. Bull Mar Sci 2002, 70(1):251–271. 86. Warner RR, Cowen RK: Free Local retention of production in marine populations: evidence, mechanisms, and consequences. Bull Mar Sci 2002, 70((1) Suppl):245–249. 87. Als T, Hansen M, Maes G, Castonguay M, Riemann L, Aarestrup K, Munk P, Sparholt H, Hanle R, Bernatchez L: All roads lead to home: panmixia of European eel in the Sargasso Sea. Mol Ecol 2011, 20(7):1333–1346. 88. Iacchei M, Ben-Horin T, Selkoe KA, Bird CE, García-Rodríguez FJ, Toonen RJ: Combined analyses of kinship and FST suggest potential drivers of chaotic genetic patchiness in high gene-flow populations. Mol Ecol 2013, 22(13):3476–3494. 89. Sponaugle S, Cowen R, Shanks A, Morgan S, Leis J, Pineda J, Boehlert G, Kingsford M, Lindeman K, Grimes C, Munro J: Predicting self-recruitment in marine populations: biophysical correlates and mechanisms. Bull Mar Sci 2002, 70(1):341–375. 90. Morgan SG, Fisher JL, Miller SH, McAfee ST, Largier JL: Nearshore larval retention in a region of strong upwelling and recruitment limitation. Ecology 2009, 90(12):3489–3502. 91. Hedgecock D, Launey S, Pudovkin AI, Naciri Y, Lapègue S, Bonhomme F: Small effective number of parents (Nb) inferred for a naturally spawned cohort of juvenile European flat oysters Ostrea edulis. Mar Biol 2007, 150(6):1173–1182. 92. Veliz D, Duchesne P, Bourget E, Bernatchez L: Genetic evidence for kin aggregation in the intertidal acorn barnacle (Semibalanus balanoides). Mol Ecol 2006, 15(13):4193–4202. 93. Bernardi G, Beldade R, Holbrook SJ, Schmitt RJ: Full-sibs in cohorts of newly settled coral reef fishes. PloS one 2012, 7(9):e44953. 94. Broquet T, Viard F, Yearsley JM: Genetic drift and collective dispersal can result in chaotic genetic patchiness. Evolution 2013, 67(6):1660–1675. 95. Yearsley JM, Viard F, Broquet T: Theeffectofcollectivedispersalonthe genetic structure of a subdivided population. Evolution 2013, 63(6):1649–1659. 96. Kanno Y, Vokoun JC, Letcher BH: Sibship reconstruction for inferring mating systems, dispersal and effective population size in headwater brook trout (Salvelinus fontinalis) populations. Conserv Genet 2011, 12(3):619–628. 97. Wang J: Sibship reconstruction from genetic data with typing errors. Genetics 2004, 166(4):1963–1979. 98. Karaket T, Poompuang S: CERVUS vs. COLONY for successful parentage and sibship determinations in freshwater prawn Macrobrachium rosenbergii de Man. Aquaculture 2012, 324:307–311. 99. Waples RS: Evaluating the effect of stage-specific survivorship on the Ne/Nratio. Mol Ecol 2002, 11:1029–1037. 100. Hedgecock D: Does variance in reproductive success limit effective population sizes of marine organisms? In Genetic and Evolution of Aquatic Organisms. Edited by Beaumont AR. London, UK: Chapman & Hall; 1994:122–134. 101. Hauser L, Adcock GJ, Smith PJ, Ramírez JHB, Carvalho GR: Loss of microsatellite diversity and low effective population size in an overexploited population of New Zealand snapper (Pagrus auratus). Proc Natl Acad Sci 2002, 99(18):11742–11747. 102. Turner T, Wares J, Gold J: Genetic effective size is three orders of magnitude smaller than adult census size in an abundant, estuarine-dependent marine fish (Sciaenops ocellatus). Genetics 2002, 162(3):1329–1339. 103. Waples RS, Yokota M: Temporal estimates of effective population size in species with overlapping generations. Genetics 2007, 175(1):219–233. 104. Palstra FP, Ruzzante DE: Genetic estimates of contemporary effective population size: what can they tell us about the importance of genetic stochasticity for wild population persistence? Mol Ecol 2008, 17(15):3428–3447. 105. West-Eberhard M: Developmental Plasticity and Evolution. Oxford: Oxford University Press; 2003. 106. Marshall DJ, Bonduriansky R, Bussiere LF: Offspring size variation within broods as a bet-hedging strategy in unpredictable environments. Ecology 2008, 89(9):2506–2517. 107. Jørgensen AT, Hansen BW, Vismann B, Jacobsen L, Skov C, Berg S, Bekkevold D: High salinity tolerance in eggs and fry of a brackish Esox lucius population. Fish Manage Ecol 2010, 17(6):554–560. 108. Larsen PF, Nielsen EE, Meier K, Olsvik PA, Hansen MM, Loeschcke V: Differences in salinity tolerance and gene expression between two populations of atlantic Cod (Gadus morhua) in response to salinity stress. Biochem Genet 2012, 50(5–6):454–466. 109. Rossi F: Short-term response of deposit-feeders to an increase of the nutritive value of the sediment through seasons in an intertidal mudflat (Western Mediterranean, Italy). J Exp Mar Biol Ecol 2003, 290(1):1–17. doi:10.1186/1471-2148-14-12 Cite this article as: Kesäniemi et al.:Temporal genetic structure in a poecilogonous polychaete: the interplay of developmental mode and environmental stochasticity. BMC Evolutionary Biology 2014 14:12. Submit your next manuscript to BioMed Central and take full advantage of: • Convenient online submission • Thorough peer review • No space constraints or color figure charges • Immediate publication on acceptance • Inclusion in PubMed, CAS, Scopus and Google Scholar • Research which is freely available for redistribution Submit your manuscript at www.biomedcentral.com/submit Kesäniemi et al. BMC Evolutionary Biology 2014, 14:12 Page 16 of 16 http://www.biomedcentral.com/1471-2148/14/12