Received: 7 December 2020 | Revised: 16 December 2020 | Accepted: 17 December 2020 DOI: 10.1002/ece3.7160 ORIGINAL RESEARCH Genetic response to human-induced habitat changes in the marine environment: A century of evolution of European sprat in Landvikvannet, Norway María Quintela1 | Àlex Richter-Boix2 | Dorte Bekkevold3 | Cecilie Kvamme1 | Florian Berg1 | Eeva Jansson1 | Geir Dahle1 | François Besnier1 | 4 1,5 Richard D. M. Nash | Kevin A. Glover 1 Institute of Marine Research, Bergen, Norway Abstract 2 CREAF, Campus de Bellaterra, Autonomous Habitat changes represent one of the five most pervasive threats to biodiversity. University of Barcelona, Barcelona, Spain However, anthropogenic activities also have the capacity to create novel niche 3 DTU-Aqua National Institute of Aquatic Resources, Technical University of Denmark, spaces to which species respond differently. In 1880, one such habitat alterations Silkeborg, Denmark occurred in Landvikvannet, a freshwater lake on the Norwegian coast of Skagerrak, 4 Centre for Environment, Fisheries and which became brackish after being artificially connected to the sea. This lake is now Aquaculture Science (Cefas), Lowestoft, UK 5 home to the European sprat, a pelagic marine fish that managed to develop a self- Institute of Biology, University of Bergen, Bergen, Norway recruiting population in barely few decades. Landvikvannet sprat proved to be ge- netically isolated from the three main populations described for this species; that Correspondence María Quintela, Institute of Marine is, Norwegian fjords, Baltic Sea, and the combination of North Sea, Kattegat, and Research, P.O. 1870, Nordnes, N-5817 Skagerrak. This distinctness was depicted by an accuracy self-assignment of 89% Bergen, Norway. Email: maria.quintela.sanchez@hi.no and a highly significant FST between the lake sprat and each of the remaining samples (average of ≈0.105). The correlation between genetic and environmental variation Funding information European Maritime and Fisheries Fund, indicated that salinity could be an important environmental driver of selection (3.3% Grant/Award Number: 33113-B-16-065; of the 91 SNPs showed strong associations). Likewise, Isolation by Environment Norwegian Department of Trade and Fisheries; Research Council of Norway, was detected for salinity, although not for temperature, in samples not adhering to Grant/Award Number: 299554 an Isolation by Distance pattern. Neighbor-joining tree analysis suggested that the source of the lake sprat is in the Norwegian fjords, rather than in the Baltic Sea de- spite a similar salinity profile. Strongly drifted allele frequencies and lower genetic di- versity in Landvikvannet compared with the Norwegian fjords concur with a founder effect potentially associated with local adaptation to low salinity. Genetic differen- tiation (FST ) between marine and brackish sprat is larger in the comparison Norway- Landvikvannet than in Norway-Baltic, which suggests that the observed divergence was achieved in Landvikvannet in some 65 generations, that is, 132 years, rather than gradually over thousands of years (the age of the Baltic Sea), thus highlighting the pace at which human-driven evolution can happen. María Quintela and Àlex Richter-Boix contributed equally to this work. This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. © 2021 The Authors. Ecology and Evolution published by John Wiley & Sons Ltd. Ecology and Evolution. 2021;11:1691–1718.  www.ecolevol.org | 1691 1692 | QUINTELA et al. KEYWORDS evolutionary response, habitat alteration, local adaptation, SNP, sprat 1 | I NTRO D U C TI O N Humans are fundamentally changing connections within and among ecosystems over a wide range of spatial scales and habitat Humans have dramatically impacted the Earth's surface and pro- types, hence modifying the levels of connectivity (Crook et al., 2015). moted striking ecosystem and biodiversity alterations over the Such changes can pose direct threats to communities, but may also course of the last two centuries, hence becoming an evolution- create novel environments that influence the evolutionary trajecto- ary force of extraordinary influence (Albuquerque et al., 2018; ries of populations and species (Allendorf et al., 2012), and can alter Ceballos et al., 2015; Hooper et al., 2012). Human activities gen- the phenotypic landscapes of species by decreasing or increasing erate major pressures on habitats and organisms and are asso- genetic diversity (Figure 1) (Hendry et al., 2017). Many examples ciated with evolutionary changes that can occur within tens of of contemporary evolution are associated with colonization events, years, a phenomenon known as “contemporary evolution” (Besnier species introductions, or invasions (Colautti & Lau, 2015; Johnston et al., 2014; Otto, 2018; Pelletier & Coltman, 2018; Stockwell & Selander, 1964; Reznick & Ghalambor, 2001). Populations colo- et al., 2003). Human-driven evolution can happen at a pace and nizing new environmental conditions can be exposed to novel se- extent that is significantly higher than that of natural causes (Bull lective forces that lead to adaptive divergence and differentiation & Maron, 2016; Hendry et al., 2008; Palumbi, 2001; Therkildsen from the original population (Björklund & Gustafsson, 2015; Hendry et al., 2019). Anthropogenic activities have altered and created et al., 2002). novel niche spaces and species' responses to ecosystem alter- The construction of navigation canals is an example of hu- ations vary from avoidance to adaptation, including exploitation man-facilitated connectivity between two previously isolated eco- (Bull & Maron, 2016). systems (Galil et al., 2007). Canals can link marine and freshwater F I G U R E 1 Hypothetical adaptive landscapes showing mean population fitness (color contours) and its genetic consequences (a). Black circles show potential distribution of phenotype/genotype. The starting adaptive landscape has three fitness peaks that are each occupied by its own genetic and adapted population (Coastal Norway, North Sea, and Baltic Sea), where the blue circle of each population depicts the environment (dark blue for marine environment, light blue for brackish environment). When humans' actions created the connection of Landvikvannet with the sea, added a new peak (brackish environment) to the original sprat adaptive landscape. Two plausible scenarios are possible following the creation of the new ecological niche. (b) The neighboring population colonizes the new habitat, but the selection is not enough for the new population to differentiate itself from the ancestral one. (c) Selection in the new environment is strong enough so that the population of the new habitat differs from the ancestral population. In that case, a phenomenon of parallel adaptation to brackish environments can occur QUINTELA et al. | 1693 bodies allowing aquatic organisms to disperse to new areas and To test for local adaptation and parallel evolution, we first char- eventually colonize novel environments (Crook et al., 2015). acterized Landvik sprat with a suite of recently developed SNP One such connectivity change took place in 1880, when the lake markers and investigated the origin and connectivity of the lake Landvikvannet (henceforth denoted as Landvik for abbreviation), population using a set of 42 geographically explicit samples, most on the southern Norwegian Skagerrak coast, was artificially con- of which were described in Quintela et al. (2020). Secondly, we in- nected to the adjacent marine fjord (Strandfjorden, Grimstad, vestigated whether loci putatively under selection could be identi- Norway) by a 3 km long narrow canal (Reddal Canal). The con- fied across these samples. Correlation between outlier loci and two struction of the canal, built to transport logs down to the dock- environmental variables, salinity and temperature, was examined to yards by the sea as well as to drain the lake to increase the surface test the potential role of selection in population divergence, and the of arable land, lowered the water level in the lake by 3 m, turning possibility to identify genetic signals of parallel evolutionary change the lake brackish as saltwater inflows over the tidal cycle while between Landvik and the Baltic Sea populations with respect to the there is a continuous flux of freshwater from streams into the lake marine populations. (Kanalkontoret, 1883). This human alteration drove changes in species assemblages, facilitating the colonization of marine spe- cies like the Atlantic herring (Clupea harengus) (Linnaeus, 1758) 2 | M ATE R I A L S A N D M E TH O DS and European sprat, Sprattus sprattus (Linnaeus, 1758). Although it is unsure when these marine species colonized Landvik, the first 2.1 | Sampling and environmental data sprat sample taken by the Institute of Marine Research dates back to 1999. Three samples of sprat from Landvik, comprising a total of 300 in- The European sprat is a small pelagic fish that is widely distrib- dividuals, were collected in 2012, 2015, and 2019, respectively. In uted from northern Norway to Morocco, the Baltic Sea, the north- addition, to compare among local populations in the area, a further ern Mediterranean basins, and the Black Sea (Debes et al., 2008). 79 (immature juvenile) individuals were collected in 2019 in two Three geographically distinct genetic groups have been described Norwegian fjords in the vicinity of Landvik (Tvedestrandsfjord and with nuclear markers: (a) Norwegian fjords, (b) Baltic Sea, and (c) Sørfjord) from a beach seine survey. These five samples were ana- a wide-ranging component spanning the North Sea, Kattegat– lyzed and compared with genotype data from 40 reference samples Skagerrak in north to the Celtic Sea, and Bay of Biscay in south of sprat, 2,425 individuals in total, collected from a range of locations (Glover et al., 2011; Limborg, Hanel, et al., 2012; Limborg et al., 2009; in the Atlantic and the Baltic Sea, as well as in the Adriatic and Black Quintela et al., 2020). Furthermore, mitochondrial control region re- Seas, representing southern outgroups (Table 1, Figure 2). Genetic vealed two additional demes in the Mediterranean Sea, Gulf of Lyon, structure in the 40 reference samples was determined in Quintela and Adriatic Sea (Debes et al., 2008). Differences found in candidate et al. (2020), showing three highly distinct and relatively homoge- loci for divergent selection between the fresh- to brackish water nous groups: (a) Norwegian fjords; (b) Baltic Sea; and (c) Northeast Baltic Sea and fully marine populations suggest that local adaptation Atlantic including the North Sea, Kattegat–Skagerrak, Celtic Sea and to low salinity is likely (Quintela et al., 2020), as has been shown in Bay of Biscay. Evidence of genetic admixture and possibly physical other Clupeid species such as the Atlantic herring in the Baltic Sea mixing was detected in the transition zone between the North and (Guo et al., 2016; Limborg, Helyar, et al., 2012), and the European an- Baltic seas, but not elsewhere. chovy (Engraulis encrasicolus Linnaeus, 1758) in the Adriatic (Ruggeri Spawners and embryos have been identified as the most tem- et al., 2016) and Tyrrhenian Seas (Catanese et al., 2017). The colo- perature-sensitive stages in the life cycle of fish (Dahlke et al., 2020). nization of Landvik's brackish waters might have been possible due Data about temperature and salinity corresponding to the average to the sprat's standing genetic variation allowing adaptation to a summer values for the period 2005–2012 were retrieved from NOAA range of salinities, as conditions in Landvik partly resemble those in database (National Oceanic and Atmospheric Administration). The the Baltic Sea, the largest brackish water body in the world (Florian depth at which measurements were chosen was 10 m for being rele- Berg, 2018). vant both for spawners and embryos (Table 1). The relatively recent colonization of Landvik by sprat provides an opportunity to study a contemporary evolution process, testing whether the creation of this new environment has promoted ge- 2.2 | DNA isolation and genotyping netic differentiation from standing variation through ecological ad- aptation. This happens when barriers to gene flow evolve between DNA was extracted from fin clips stored in ethanol using the Qiagen populations due to divergent selection, with niche adaptation and DNeasy 96 Blood & Tissue Kit in 96-well plates, each of which con- competition as driving mechanisms (Bolnick, 2004; Schluter, 2000). tained two or more negative controls. All 45 samples were geno- Landvik's salinity is similar to that of parts of the Baltic Sea, which typed with the 91 SNPs for which protocols are described in their thus allows the use of it as a replicate model to study parallel evolu- entirety in Quintela et al. (2020). In addition, a subset of 15 of the tion and the role of the environment in ecologically driven speciation 45 samples was genotyped with eight microsatellite markers (see (Bailey et al., 2017; Bolnick et al., 2018). Table A1), as described in Glover et al. (2011). The main aim of the TA B L E 1 Summary statistics of sprat genotyped at 91 SNP: Sampling site within geographic regions, time of collection (year/month), coordinates, number of individuals; observed 1694 heterozygosity, Ho (mean ± SE); unbiased expected heterozygosity, uHe (mean ± SE); inbreeding coefficient, FIS (mean ± SE); number of deviations from Hardy–Weinberg equilibrium (HWE) at | α = 0.05; number of deviations from Linkage Disequilibrium (LD) at α = 0.05 both before and after (B) Bonferroni correction HWE Region No Site Year Month Code Latitude Longitude No ind Ho uHe FIS (B) LD (B) Temp Salin Norwegian fjords 1 Holandsfjord 2008 12 HOL 66,705 13,562 31 0.278 ± 0.018 0.282 ± 0.016 0.003 ± 0.021 7 (0) 125 (1) 10.15 32.98 2 Melfjord 2008 12 MEL 66,610 13,686 79 0.286 ± 0.016 0.281 ± 0.015 −0.021 ± 0.013 7 (0) 193 (2) 12.06 32.62 3 Finneidfjord 2008 12 FIN 66,268 13,942 75 0.276 ± 0.015 0.286 ± 0.015 0.016 ± 0.015 8 (0) 178 (2) 12.17 33.05 4 Stjørdalsfjord 2008 12 TRH 63,807 11,041 80 0.281 ± 0.015 0.286 ± 0.015 0.005 ± 0.013 5 (1) 193 (4) 12.21 32.20 5 Nordfjord 2015 12 NOR1 61,959 6,429 39 0.275 ± 0.017 0.286 ± 0.016 0.024 ± 0.020 7 (2) 141 (2) 12.27 32.98 6 Nordfjord 2001 5 NOR2 61,860 6,004 74 0.284 ± 0.017 0.286 ± 0.016 0.008 ± 0.014 7 (0) 196 (3) 12.19 32.36 7 Nordfjord 2015 12 NOR3 61,811 6,111 49 0.264 ± 0.016 0.283 ± 0.016 0.053 ± 0.019 7 (3) 153 (2) 12.19 32.36 8 Sognefjord 2008 11 SOG1 61,489 7,589 47 0.264 ± 0.016 0.278 ± 0.016 0.024 ± 0.018 10 (1) 155 (2) 12.47 31.99 9 Sognefjord 2015 12 SOG2 61,485 7,597 116 0.266 ± 0.014 0.283 ± 0.015 0.047 ± 0.014 16 (2) 202 (4) 12.47 31.99 10 Hardangerfjord 2015 12 HAR1 60,225 6,058 100 0.269 ± 0.014 0.283 ± 0.015 0.032 ± 0.010 7 (0) 203 (3) 12.87 33.09 11 Hardangerfjord 2008 11 HAR2 60,000 5,935 77 0.269 ± 0.015 0.284 ± 0.016 0.035 ± 0.016 12 (2) 172 (3) 13.78 31.74 12 Hardangerfjord 2008 11 HAR3 60,410 6,701 46 0.286 ± 0.016 0.284 ± 0.015 −0.018 ± 0.017 5 (0) 160 (2) 12.87 31.64 13 Hardangerfjord 2008 11 HAR4 60,236 6,589 99 0.278 ± 0.015 0.282 ± 0.015 0.004 ± 0.013 9 (3) 188 (2) 12.87 33.09 14 Lysefjord 2008 11 LYS 58,987 6,221 100 0.273 ± 0.015 0.286 ± 0.016 0.032 ± 0.013 7 (1) 175 (4) 11.07 31.75 15 Tvedestrandsfjord 2019 9 TVE 58,589 8,982 37 0.251 ± 0.018 0.267 ± 0.018 0.033 ± 0.022 11 (1) 129 (1) 13.68 30.17 16 Sørfjord 2019 9 SORF 58,730 9,079 42 0.263 ± 0.019 0.264 ± 0.017 −0.007 ± 0.018 5 (1) 142 (1) 13.68 31.44 17 Oslofjord 2007 9 OSL 59,881 10,665 89 0.269 ± 0.016 0.281 ± 0.016 0.034 ± 0.013 7 (1) 207 (2) 13.68 31.00 Landvik 18 Landvikvannet 2012 5 LAND12 58,317 8,491 27 0.239 ± 0.022 0.231 ± 0.019 −0.039 ± 0.020 3 (0) 112 (1) 14.86 16.00 19 Landvikvannet 2015 3 LAND15 58,317 8,491 205 0.225 ± 0.017 0.234 ± 0.018 0.025 ± 0.012 11 (4) 186 (11) 14.86 16.00 20 Landvikvannet 2019 9 LAND19 58,317 8,491 68 0.251 ± 0.017 0.264 ± 0.017 0.031 ± 0.018 11 (1) 194 (1) 14.86 16.00 North 21 North Sea 2018 7 NS1 56,049 7,723 57 0.214 ± 0.018 0.231 ± 0.018 0.082 ± 0.024 5 (1) 155 (1) 14.33 32.94 Sea,Kattegat– Skagerrak 22 North Sea 2015 5 NS2 57,130 4,526 77 0.231 ± 0.018 0.238 ± 0.018 0.008 ± 0.013 6 (1) 136 (1) 15.55 34.44 23 North Sea 2008 ? NS3 54,307 1,842 93 0.233 ± 0.018 0.246 ± 0.018 0.029 ± 0.014 5 (1) 138 (1) 14.11 34.61 24 North Sea 2005 5 NS4 54,277 7,802 59 0.224 ± 0.019 0.235 ± 0.019 0.042 ± 0.018 16 (3) 127 (1) 16.07 31.43 25 North Sea 2016 8 NS5 53,417 3,833 40 0.229 ± 0.019 0.234 ± 0.019 0.027 ± 0.020 3 (1) 127 (1) 16.65 34.48 26 North Sea 2016 8 NS6 53,449 2,858 38 0.231 ± 0.018 0.241 ± 0.019 0.014 ± 0.017 9 (2) 100 (1) 14.34 34.11 27 English Channel 2009 6 EC 51,265 1,959 50 0.218 ± 0.018 0.228 ± 0.019 0.019 ± 0.016 6 (1) 109 (1) 15.67 34.38 28 Bay of Biscay 2008 8 BoB 47,192 −1,318 57 0.214 ± 0.018 0.234 ± 0.019 0.085 ± 0.023 4 (0) 111 (1) 14.52 35.04 QUINTELA et al. (Continues) QUINTELA et al. TA B L E 1 (Continued) HWE Region No Site Year Month Code Latitude Longitude No ind Ho uHe FIS (B) LD (B) Temp Salin 29 Celtic Sea 2009 10 CEL 52,814 −9,856 79 0.242 ± 0.018 0.245 ± 0.019 −0.007 ± 0.013 5 (1) 121 (1) 13.58 34.16 30 Kattegat 2018 6 SK1 58,011 11,156 58 0.231 ± 0.019 0.239 ± 0.019 0.017 ± 0.016 3 (1) 114 (1) 15.79 30.65 31 Kattegat 2006 3 SK2 57,734 10,809 38 0.227 ± 0.020 0.227 ± 0.019 −0.004 ± 0.019 9 (2) 81 (1) 15.26 32.37 32 Kattegat 2018 9 SK3 57,718 11,015 38 0.211 ± 0.018 0.239 ± 0.019 0.107 ± 0.026 9 (2) 140 (1) 15.33 31.06 33 Kattegat 2018 7 SK4 57,134 11,854 41 0.218 ± 0.018 0.230 ± 0.019 0.021 ± 0.018 7 (2) 100 (1) 16.32 30.65 34 Kattegat 2018 7 SK5 57,021 11,744 73 0.228 ± 0.019 0.241 ± 0.019 0.044 ± 0.018 4 (1) 125 (1) 15.62 22.65 35 Uddevalla fjord 2008 5 UV 58,250 11,428 59 0.232 ± 0.017 0.244 ± 0.018 0.023 ± 0.016 11 (5) 133 (1) 16.05 29.10 36 Great Belt 2006 3 GB 55,684 10,437 47 0.251 ± 0.018 0.254 ± 0.017 0.005 ± 0.016 7 (1) 137 (1) 14.56 23.64 37 Øresund 2006 3 ØS 55,767 12,731 46 0.263 ± 0.019 0.257 ± 0.018 −0.025 ± 0.017 12 (4) 131 (1) 14.83 17.38 Baltic Sea 38 Arkona Basin 2006 5 AB 55,147 13,845 59 0.232 ± 0.018 0.237 ± 0.019 0.006 ± 0.015 5 (1) 127 (1) 12.81 7.96 39 Bornholm Basin N 2006 3 BBN 55,571 16,408 39 0.224 ± 0.020 0.226 ± 0.019 −0.016 ± 0.016 4 (0) 103 (1) 14.16 7.37 40 Bornholm Basin S 2006 3 BBS 54,933 15,690 43 0.238 ± 0.020 0.238 ± 0.019 −0.014 ± 0.017 6 (0) 88 (1) 17.32 7.75 41 Gdansk Deep 2006 3 GD 54,751 18,993 56 0.227 ± 0.019 0.238 ± 0.019 0.036 ± 0.019 6 (2) 121 (1) 16.39 7.54 42 Gotland Basin 2006 5 GOTB 58,406 20,526 55 0.225 ± 0.019 0.232 ± 0.019 0.007 ± 0.015 5 (0) 130 (1) 13.56 6.94 43 Baltic Sea, 2006 3 GOT 57,815 19,509 56 0.240 ± 0.022 0.223 ± 0.019 −0.053 ± 0.017 8 (0) 110 (1) 15.57 6.94 Gotland Out 44 Adriatic Sea 2005 12 ASA 45,360 13,340 45 0.179 ± 0.018 0.195 ± 0.020 0.050 ± 0.021 8 (3) 75 (0) na na 45 Black Sea 2008 12 BS 41,087 40,027 21 0.193 ± 0.020 0.208 ± 0.020 0.036 ± 0.027 9 (0) 47 (0) na na Note: Samples from ripe individuals are depicted in boldface type in the year/month columns. Temperature (°C) and salinity were measured at 10 m depth (summer values averaged from 2005–2012). | 1695 1696 | QUINTELA et al. F I G U R E 2 Map of the sampling sites as well as detailed view of Landvikvannet. Codes and associated full names of sampling locations can be found in Table 1. The colors depict the genetic clusters according to STRUCTURE. Landvikvannet samples are coded as 18–20 second set was to estimate genetic diversity through allelic richness, conducted using the software ParallelStructure (Besnier & Glover, and hence, results derived of the microsatellite data will be mainly 2013), was used to identify genetic groups under a model assuming presented in Appendix 1. admixture and correlated allele frequencies without using popula- tion information as a prior. Ten runs with a burn-in period consist- ing of 100,000 replications and a run length of 1,000,000 MCMC 2.3 | Statistical analysis iterations were performed for K = 1 to K = 7 clusters. To deter- mine the number of genetic groups, STRUCTURE output was ana- All statistical analyses were performed separately for SNPs and mi- lyzed using two approaches: (a) the ad hoc summary statistic ΔK of crosatellites. The observed (Ho) and unbiased expected heterozygo- Evanno et al. (2005), and (b) the four statistics (MedMed, MedMean, sity (uHe) as well as the inbreeding coefficient (FIS) were computed MaxMed, and MaxMean) both implemented in StructureSelector for each sample with GenAlEx v6.1 (Peakall & Smouse, 2006). The (Li & Liu, 2018). The ten runs for the selected Ks were then aver- genotype frequency of each locus and its direction (heterozygote aged with CLUMPP v.1.1.1 (Jakobsson & Rosenberg, 2007) using deficit or excess) was compared with Hardy–Weinberg expectations the FullSearch algorithm and the G′ pairwise matrix similarity sta- (HWE) using the program GENEPOP 7 (Rousset, 2008), as was link- tistic, and graphically displayed using barplots. Furthermore, the age disequilibrium (LD) between pairwise loci. relationships between Landvik and the reference samples were ex- Landvik sprat were compared with the remaining collections using amined using discriminant analysis of principal components, DAPC pairwise FST (Weir & Cockerham, 1984) computed with ARLEQUIN (Jombart et al., 2010) implemented in adegenet (Jombart, 2008), as v.3.5.1.2 (Excoffier et al., 2005). The Bayesian clustering approach well as with the principal coordinates analysis (PCoA) built using implemented in STRUCTURE v. 2.3.4 (Pritchard et al., 2000), and Nei (1978)'s genetic distance between pairs of populations with QUINTELA et al. | 1697 (a) (b) F I G U R E 3 Relationship between Landvikvannet sprat and the reference samples genotyped at 91 SNP loci according to (a) STRUCTURE, (b) DAPC, and (c) PCoA. Plot in d) represents genetic distance measured as pairwise FST/(1 − FST ) between the northernmost site (HOL) and each of the 41 remaining ones versus the corresponding shortest water distance (in km). Analyses in a) and d) were performed without the distant southern outgroups to increase the resolution GenAlEx v6.1 (Peakall & Smouse, 2006). To examine demographic divergent selection (Lewontin & Krakauer, 1973). In BayeScan, sample relationships between geographically explicit samples, the genetic size was set to 10,000 and the thinning interval to 50. Loci with a pos- distance, measured as FST/(1 − FST ), between the northernmost sam- terior probability over 0.99, corresponding to a Bayes Factor > 2 (i.e., ple (HOL) and all other samples (excluding the southern European “decisive selection” (Foll & Gaggiotti, 2006)), were retained as outliers. outgroups) was plotted against the corresponding shortest water- In LOSITAN, a neutral distribution of FST with 1,000,000 iterations way distance, calculated using the path function in GoogleEarth. was simulated, with forced mean FST at a significance level of 0.05 The assignment of individuals to STRUCTURE genetic clusters was under an infinite allele model for SNPs and under a stepwise model conducted with the program GeneClass 2 (Piry et al., 2004) using for microsatellites. To avoid pseudo replication, outlier analyses were the Rannala and Mountain (1997) method of computation. Finally, conducted using a random sample of 300 individuals from each of a neighbor-joining (NJ) tree based upon pairwise Nei's genetic dis- the four genetic clusters identified with STRUCTURE (after excluding tance DA (Nei et al., 1983) for all SNPs was constructed with the southern distant outgroups). Analyses were performed either using software POPTREE2 (Takezaki et al., 2010) using 1,000 bootstraps jointly the four sets of samples or using subsets, as appropriate. and visualized using FigTree 1.4.3 (Rambaut, 2009). Adaptation to local environments often occurs through natural Two analytic approaches, BayeScan (Foll & Gaggiotti, 2008) and selection acting on a large number of loci, each having a weak pheno- LOSITAN (Antao et al., 2008), were combined to detect loci deviat- typic effect. LFMM, “latent factor mixed model” (Frichot et al., 2013), ing from neutral expectations and therefore reflecting either even- was used to assess whether salinity or water temperature could be a tual selective responses or linkage disequilibrium with genes under potential selective pressure driving local adaptation by identifying loci 1698 | QUINTELA et al. (c) (d) FIGURE 3 (Continued) QUINTELA et al. | 1699 F I G U R E 4 Origin of Landvik sprat: Neighbor-joining tree placing Landvik in context with the reference samples genotyped at 91 SNP loci (NJ tree using after removing the loci under positive selection can be found in Figure A3 in Appendix 1). To increase the resolution, analyses were performed after excluding the distant southern outgroups showing unusual associations with these environmental factors com- STRUCTURE outcome as suggested by Frichot et al. (2013). The cor- pared to the genetic background. Thus, the environmental informa- responding z-scores of the ten replicates were combined following the tion used corresponded to the season of the year where fish are at its recommendations described in Frichot and François (2015). First, the most temperature-sensitive stages (Dahlke et al., 2020). This method, genomic inflation factor (λ) was obtained after computing the median which has formerly proved to be efficient for a suite of scenarios of of the squared (combined) z-scores for each K, divided by the median demographic history (Lotterhos & Whitlock, 2015; de Villemereuil of the chi-square distribution with one degree of freedom. Finally, et al., 2014), uses a linear mixed model to test for associations between p-values were adjusted using the genomic inflation factor (λ), and false genetic variation and environmental factors, while controlling for neu- discovery rates were set using the Benjamini and Hochberg (1995) tral genetic structure with (random) latent factors. Ten runs of LFMM algorithm. were conducted using 1,000 sweeps for burn-in and 10,000 additional In addition, the relationship between genetic distance (FST) and sweeps. The number of latent factors was set at K = 4 according to each environmental factor was examined using Mantel (1967) tests 1700 | QUINTELA et al. TA B L E 2 LFMM analysis for salinity and temperature (measured both in summer at 10 m depth) LFMM, log10 (PO) Global candidate loci Pairwise candidate loci (LOSITAN) Norway vs. Norway Landvik Locus Annotation Salinity Temperature LOSITAN BayeScan Landvik vs. Baltic vs. Baltic Ssp248 Protein kinase C epsilon 130.07 4.54 0.976 0.056 0.624 0.993 0.774 Ssp210 116.82 0.51 0.990 2.308 0.972 1 -100 Ssp215 ATP-dependent 51.50 1.28 0.750 -1.190 0.919 0.675 0.702 6-phosphofructokinase liver-like Ssp253 9.35 3.07 0.996 -0.852 0.917 0.821 1 Ssp263 3.10 32.36 1 2.920 0.658 0.500 1 Ssp213 Tensin-2-like 28.52 0.41 1 0.502 0.997 0.500 1 Ssp268 25.08 6.50 1 0.419 0.998 0.765 1 Ssp279 30.16 0.83 0.799 -1.192 1 0.745 0.250 Ssp260 13.95 1.48 0.991 -1.060 0.994 0.493 0.999 Ssp315 11.83 7.58 0.267 -1.020 1 0.500 0.500 Ssp290 2.55 15.14 0.374 -0.987 1 0.050 -100 Ssp272 Neurochondrin transcript 18.18 0.80 0.606 -1.226 0.465 1 0.500 variant X1-3 Ssp226 17.21 34.26 0.720 0.001 0.437 1 0.500 Ssp269 0.12 6.80 0.621 -1.173 0.103 1 0.002 Ssp286 0.05 9.14 0.435 -1.182 0.414 1 0.500 Ssp302 Procollagen 7.87 0.18 0.638 -1.110 0 1 1 C-endopeptidase enhancer 2-like Ssp225 1.77 18.25 0.835 0.040 0.232 1 1 Ssp300 Zinc finger protein 184- 5.88 0.75 0.468 -1.211 0.500 0.648 1 like transcript variant Ssp319 TOG array regulator of 0.82 28.17 0.972 0.454 0.343 0.826 1 axonemal microtubules 1 Ssp222 8.26 25.54 0.500 -0.602 0.269 0.500 -100 Ssp002 8.19 46.96 0.800 -1.033 0.944 0.684 0.830 Ssp264 2.57 46.41 0.884 -0.615 0.281 0.925 0.807 Ssp207 2.10 22.86 0.722 -1.195 0.851 0.383 0.854 Note: Outlier analyses were performed using 300 randomly sampled individuals per genetic cluster instead of geographically explicit samples. Cells shaded in dark grey depict significant associations at LFMM after genomic inflation correction as well as candidates for positive selection according to LOSITAN after FDR correction (P(Simul FST 0.8 to the Norwegian fjord cluster, the tion, LFMM identified three loci associated with salinity and seven hypothesis of them being migrants is plausible (see Figure A1 in with temperature (Table 2), although the strength of the association Appendix 1). The 15 samples genotyped with microsatellites repro- was larger with salinity (Figure 5a,b). Locus Ssp263 was associated duced the patterns of genetic differentiation and clustering found with temperature as well as flagged as an outlier by both procedures, with SNPs and clearly depicted the distinctness of Landvik sample whereas locus Ssp210 was associated with salinity, flagged as outlier (Figures A2a-d in Appendix 1). with BayeScan and marginally with LOSITAN. Locus Ssp248, the one 1704 | QUINTELA et al. showing the strongest association with salinity (log10 (P0) = 130.1), Norwegian fjords, North Sea–Kattegat–Skagerrak, and Baltic Sea was annotated to the vicinity of a predicted protein kinase C epsilon (Quintela et al., 2020). The study also suggests signatures of contem- in the herring genome whereas only one of the loci associated with porary adaptation to brackish habitat in Landvik sprat population, temperature could be annotated, that is, locus Ssp319 to TOG array which represents a potential model system to study parallel evolu- regulator of axonemal microtubules 1 also in the herring genome tion in comparison with the Baltic. (Table 2). The relationships between genetic distance and environ- mental factors revealed that the correlation between the matrices of FST and salinity followed an Isolation by Environment pattern 4.1 | Origin of the Landvik population (R xy = 0.47, p = .0001), conversely to the lack of correlation with temperature (R xy = 0.05, p = .404). The relationship between genetic differentiation and shortest water dis- Furthermore, outlier detection analyses were conducted for the tance revealed that samples from Landvik strongly departed from any geo- same clusters of 300 randomly sampled individuals albeit in a pairwise graphically driven expectation (see Figure 3d), a situation also described fashion through comparisons involving the Norwegian sprat, Landvik, for three spine sticklebacks (Gasterosteus aculeatus Linnaeus, 1758), where and the Baltic Sea (Table 2). Locus Ssp210, showing strong associa- populations inhabiting anthropogenic modified habitats deviated from the tion with salinity, was reported to be a candidate to positive selection general pattern of Isolation by Distance (Scharsack et al., 2012). Another in the comparison between marine and brackish samples in Norway striking characteristic of the Landvik sprat population is the relatively low versus Baltic and marginally in Norway versus Landvik, whereas no genetic diversity displayed by microsatellite markers in terms of Ho, uHe, positive selection was reported between brackish environments. On and allelic richness. The only sample that exhibited a comparably low allelic the other side, four loci (Ssp253, Ssp263, Ssp300, and Ssp319) were richness was collected in Gotland (Baltic Sea), in the brink of Baltic sprat's candidates to directional selection in the comparison between brack- spawning habitat where salinity conditions approach to those impeding ish environments (Landvik vs. Baltic Sea) but not in the comparison's larval survival (Sjöblom & Parmanne, 1980). Landvik, thus, adheres to the marine versus brackish, which could eventually support rejecting the pattern previously described in the Baltic, where the combination of young hypothesis of the origin of the lake sprat being in the Baltic Sea. Major age, extreme conditions, and limited habitat size leads Baltic populations allele frequency per sample for the loci showing the strongest associ- to often have less intraspecific genetic diversity than their counterparts in ation with salinity was assessed in samples from contrasting environ- the open Northeast Atlantic (Johannesson & Andre, 2006). ments (Table 3) and revealed a similar pattern in low salinity waters (i.e., Landvik is also inhabited by a taxonomically close species to Landvik and the Baltic Sea) as opposed to marine waters (Norwegian sprat: the Atlantic herring. Landvik herring are considered as a fjords), which could suggest that Landvik sprat evolved from the self-sustaining and somewhat stationary population, character- Norwegian make-up to adapt to low salinity environments. Conversely, ized by slower growth, smaller length at maturity, lower vertebral no temperature-related pattern for a similar process was obvious. count, shorter life span, higher relative fecundity, and divergent Finally, the major allele frequency of seven of the SNPs (Ssp253, genetic profiles compared to the migratory oceanic herring in other Ssp321, Ssp260, Ssp268, Ssp213, Ssp251, and Ssp236) showed a parts of the Norwegian waters (Eggers, 2013; Eggers et al., 2014; remarkable drop in Landvik compared with the Norwegian fjord Silva et al., 2013). Meristic trait vertebral count is often used as a samples illustrating a change that could have happened in less than population identifier in herring (e.g., Berg et al., 2017; Mosegaard 132 years (see Figure 6 for four of them). & Madsen, 1996; Rosenberg & Palmén, 1981), and the observa- tion that vertebral count in Landvik herring is similar to that in herring populations in the brackish Western Baltic Sea has led to 4 | D I S CU S S I O N the hypothesis that Landvik was colonized by low salinity adapted herring of Western Baltic Sea origin (Berg et al., 2019; Eggers The brackish lake Landvik, created after excavating a 3 km long canal et al., 2014). In addition, factorial crossing experiments performed to the sea in 1880, represents a model system in which to investigate at a range of salinities ranging from 6 to 35 revealed adaptation of the potential for marine organisms to adapt to rapidly emerging new herring populations to their native salinity conditions and also that environments in the marine realm. Here, we showed that European adaption to salinity is transmitted to the offspring within the fol- sprat, a small pelagic marine fish, were able to colonize and develop lowing generation (Berg et al., 2019). In contrast to herring, which a genetically highly distinct population in few decades. The level rely on a benthic spawning habitat for depositing eggs, sprat is of differentiation observed between samples from Landvik and all a pelagic spawner. As such, salinity may exert an even stronger other reference samples of sprat was equivalent to the genetic dif- selection pressure in sprat to avoid neutrally buoyant eggs from ferentiation displayed among the most geographically distant popu- sinking into deeper anoxic water layers, as has for example been lations of sprat. This level of differentiation has been achieved in observed in Atlantic cod, Gadus morhua, adapted to spawning in a maximum of 132 years, as computed from the completion of the brackish waters (Berg et al., 2015; Nissling et al., 1994). In both cod canal until the sampling date of LAND12, which would mean some and herring, local adaptation is implied to be swift and ongoing, 65 generations of sprat. Thus, Landvik adds to the three distinc- and working on standing genetic variation (e.g., Berg et al., 2015; tive genetic clusters formerly described in European sprat, that is, Lamichhaney et al., 2012). QUINTELA et al. | 1705 The origin of Landvik sprat is unknown, but, based on inference have been changing connections between freshwater and ma- from herring, it would be conceivable that the lake could have also rine ecosystems thus facilitating freshwater introductions (Crook been colonized by fish from the Baltic Sea, already adapted to brack- et al., 2015). Drastic differences in salinity, parasites, competi- ish waters, given the parallelism in the environmental conditions. tors, and predators between marine and freshwater environments However, the analysis of Landvik in conjunction with the reference exert divergent selective pressures on the corresponding popula- samples available in this study does not appear to support the hy- tions. Salinity showed strong associations with 3.3% of the loci pothesis of the Baltic Sea as the source, but points toward founders analyzed in the present study. The genetic change experienced from Norwegian fjordic sprat. This is particularly endorsed by the by the Norwegian sprat colonizing Landvik could be attributed lower genetic differentiation between Norwegian sprat and Landvik, to the strong directional selection driven by the low salinity in as well as by the neighbor-joining tree showing that Landvik sprat the lake. Rapid evolutionary changes are predicted in the face of stems from the Norwegian cluster. Taking into consideration that strong selection following habitat shifts or environmental distur- NJ analyses are sensitive to outliers, the tree was recalculated after bances (Burke & Long, 2012; Kopp & Matuszewski, 2014; Losos purging the candidate loci to positive selection detected by LOSITAN et al., 1997; Turcotte et al., 2011), as happened in Landvik when and BayeScan. The new NJ tree confirmed that the node from which the lake was artificially connected to the sea circa 150 years ago. Landvik sprat stem was the Norwegian sample, LYS (see Figure A3 in Similar processes have been documented in other species such as Appendix 1). Furthermore, Landvik sprat displays a suite of features the threespine stickleback, which managed to evolve from oce- that concur with founder effects such as strongly drifted allele fre- anic ancestors to colonize the freshwater ponds that were formed quencies at both microsatellites and SNPs, together with low genetic during uplift caused by the Great Alaska Earthquake in 1964 diversity at microsatellites assessed as allelic richness, Ho and uHe in (Lescak et al., 2015). Adaptation of a newly established resident comparison with the Norwegian samples. Similarly, losses of genetic population to the brackish environment often proceeds very fast, diversity, a signature compatible with historical founder effect, have over the course of several decades (Barrett et al., 2008; Lescak been reported in other fish species in the face of anthropic chal- et al., 2015; Marques et al., 2018). Data on adaptation associated lenges: for example, Mango tilapia Sarotherodon galilaeus (Linnaeus, with salinity have been reported in fish moving from high to low 1758) in the Sea of Galilee (Borovski et al., 2018), American paddle- salinity such as sticklebacks (DeFaveri & Merilä, 2014; McCairns & fish Polyodon spathula (Walbaum in Artedi 1792) stocked in Poland Bernatchez, 2010), Atlantic killifish Fundulus heteroclitus (Linnaeus, from United States (Kaczmarczyk et al., 2012), or the introduced grass 1766) (Whitehead et al., 2011), and alewives Alosa pseudoharengus carp Ctenopharyngodon idella (Valenciennes in Cuvier & Valenciennes, (A. Wilson, 1811) (Velotta et al., 2014). 1844) with respect to its native Chinese ranges (Chen et al., 2012). Outlier loci experiencing adaptive selection based on envi- Likewise, the isolation of the live-bearing fish Caterina allocota Allotoca ronmental conditions have also been described in other Clupeids. catarinae (de Buen, 1942) from another species of the same complex Ruggeri et al. (2016) related population divergence in microsatel- dated ~1900 years ago represents the first evidence of fish species lite outlier loci in relation to salinity, oxygenation, and temperature translocation by a pre-Hispanic culture of Mexico (Corona-Santiago in the European anchovy Engraulis encrasicolus (Linnaeus, 1758) in et al., 2015). Finally, the sample taken in Landvik in 2019 included 21 the Adriatic Sea. For the same species, Catanese et al. (2017) re- individuals (31%) with a genetic profile compatible with being migrants ported that the selective pressure related to river mouths acts on from surrounding Norwegian fjords as denoted by an inferred ances- the same genes in distant areas in the Atlantic Ocean, Tyrrhenian, try to the Norwegian STRUCTURE cluster of q > 0.8 and by being and North Adriatic Sea. These SNP outliers were also associated assigned to this cluster by GeneClass. However, only one individual with salinity variability or involved in a critical stage of fertilization approached this level of ancestry for the Baltic Sea cluster. The fact process. of finding potential migrants coming from the coastal Norway, but The Baltic Sea was formed after the latest ice age, approximately not from the Baltic Sea, would further support the hypothesis of the 10,000–15,000 years ago, although its “ecological age” is circa Norwegian fjords as a source of Landvik sprat. Despite these levels of 8,000 years (Lass & Matthäus, 2008). The combination of young gene flow, outlier loci revealed patterns of population structure that geological age and contrasting environmental conditions to the sur- support postsettlement selection and suggest that strong selective rounding oceans resulted in fast processes of adaptive evolution, forces could be acting and therefore causing local adaptation. which led to species living in the edge of their physiological toler- ance (Ojaveer et al., 2010). The degree of differentiation between marine and brackish sprat is higher in the comparisons Norway 4.2 | Adaptation as a consequence of brackish fjords versus Landvik (mean FST = 0.080, range 0.029–0.117) than in water colonization Norway fjords versus Baltic (mean FST = 0.037, range 0.026–0.047). Hypothesizing that the origin of Landvik sprat is the Norwegian fjord Transitions from marine to freshwater habitats constitute dra- sprat invokes the possibility that the genetic changes occurring in the matic shifts between adaptive habitats that have occurred not lake took place in a maximum of 65 generations (<132 years) rather only on macroevolutionary time scales, but also in the recent than gradually over thousands of years. This hypothesis was also put past (Lee & Bell, 1999). During the last two centuries, humans forward in the threespine stickleback, which achieved in 50 years 1706 | QUINTELA et al. similar levels of divergence as populations that had diverged thou- freshwater populations (Bassham et al., 2018). This population sands of years ago (Lescak et al., 2015). Most likely, such rapid adap- structure and history would provide many opportunities for paral- tation to a new environment cannot depend on de novo mutations lel evolution when new freshwater populations were established and must rely primarily on standing genetic variation (Matuszewski from the marine stickleback population (Stern, 2013). In stickle- et al., 2015; Dolph Schluter & Conte, 2009) as it has been demon- backs, that genetic parallelism is seen on finer geographic scales strated in the threespine stickleback populations (Terekhanova (Jones et al., 2012; Nelson & Cresko, 2018) but not globally, a pat- et al., 2019). The high speed of adaptation of Landvik population to tern attributed to founder events and the loss of genetic diversity the brackish environment probably has been possible by freshwater following colonization of the Atlantic (Fang et al., 2020). Landvik tolerance alleles being present in ancestral marine populations. adaptation to brackish waters could have followed a similar pat- tern, where adaptation independent from Baltic populations has been a consequence of demographic forces of the founder event 4.3 | Parallel or convergent evolution under similar of the lake from the Norwegian coast populations. However, our selection pressures? study has strong limitations to disentangle whether it is a case of molecular parallelism or independent adaptation. Future genomic Adaptation to a radically different environment is likely to be geneti- studies may help reveal the evolutionary history of the sprat and cally complex and to involve many loci, as it has been shown for other the molecular mechanisms involved in its different adaptations to species (Terekhanova et al., 2019). Locus Ssp210, showing strong as- brackish environments. Study the genetics of convergence can sociation with salinity, was reported to be a candidate to positive help shed light on fundamental questions in evolutionary biology, selection in the comparison between marine and brackish samples including whether natural selection is constrained and repeat- (in Norway vs. Baltic and marginally in the comparison Norway vs. able or instead characterized by many molecular paths to similar Landvik) but not in the comparison between brackish environments phenotypes. (Lanvik vs. Baltic), which could suggest parallel or convergent evolu- The uniqueness of Landvik sprat suggests that an appropriate tion processes in Landvik and the Baltic Sea diverging from marine management should be considered for this population. A next step sprat. Despite the geographic proximity between Landvik and the using whole-genome sequencing will allow to further explore intra- Norwegian coastal sites, strong genetic divergence is found among clusters standing genetic variation as well as the origin of Landvik those samples, probably due to differences in abiotic parameters population. Parallel evolution in response to similar environmental (salinity) between habitats. pressures strongly suggests evolution by natural selection; however, Parallel evolution under similar selection pressure has been the underlying genetic basis of this process is unclear. Landvik sprat widely observed in populations of the same species, for example, thus provides an excellent opportunity for testing the genomic as- in bacterial experiments (Baym et al., 2016), recurrent adaptations pects of evolutionary repeatability. of pathogens to their hosts (Collins & Didelot, 2018), and marine threespine sticklebacks that have independently colonized many AC K N OW L E D G M E N T S freshwater habitats (Stuart et al., 2017). Baltic Sea and Landvik sprat We would like to thank the technicians and the commercial fisher- populations could well be the results of parallel or convergent evo- men that helped with the data collection. This study was primarily lution (Arendt & Reznick, 2008), as it has been observed in lake and financed through funding from the Norwegian Department of Trade stream sticklebacks (Colosimo et al., 2005; Stuart et al., 2017). They and Fisheries. Funding was also provided by the European Maritime dwell in discrete and divergent habitats, and are derived from an- and Fisheries Fund for the project “Fordeling af makrel, brisling cestral marine populations, increasing the likelihood of them reusing og silde-bestande,” 33113-B-16-065, and the Research Council of similar ancestral genetic variants for adaptation. However, unlike Norway project “CoastRisk” (299554). the case of some stickleback populations that colonized lakes and streams after the last glaciation from the same ancestral population C O N FL I C T O F I N T E R E S T (Bolnick et al., 2018; Therkildsen et al., 2019), the history of Landvik None declared. and the Baltic Sea are different and could lead to a nonparallel evolution. AU T H O R C O N T R I B U T I O N S Despite the limitations of the current set of SNPs markers, María Quintela: Conceptualization (equal); formal analysis (lead); the candidate outliers that distinguish derived brackish popu- methodology (supporting); writing–original draft (lead); writing–re- lations from the ancestral marine populations are not same be- view and editing (lead). Àlex Richter-Boix: Formal analysis (equal); tween Landvik and Baltic Sea populations, thus not showing a writing–original draft (lead); writing–review and editing (equal). pattern of molecular parallelism, contrary of what has been de- Dorte Bekkevold: Conceptualization (equal); data curation (lead); tected in sticklebacks colonizing different lakes (Terekhanova formal analysis (supporting); funding acquisition (supporting); meth- et al., 2014). Marine populations of stickleback harbor, at low fre- odology (supporting); writing–original draft (supporting); writing–re- quencies, alleles that confer adaptation to freshwater (Schluter view and editing (supporting). Cecilie Kvamme: Conceptualization & Conte, 2009), presumably due to the gene flow from coastal (equal); data curation (lead); funding acquisition (lead); project QUINTELA et al. | 1707 administration (lead); writing–original draft (supporting); writing–re- creates sweeping genomic remodeling in stickleback. Genetics, 209(3), 921. https://doi.org/10.1534/genet​ics.117.300610 view and editing (supporting). Florian Berg: Data curation (equal); Baym, M., Lieberman, T. D., Kelsic, E. D., Chait, R., Gross, R., Yelin, I., & writing–original draft (supporting); writing–review and editing Kishony, R. (2016). Spatiotemporal microbial evolution on antibiotic (supporting). Eeva Jansson: Data curation (equal); formal analysis landscapes. Science, 353(6304), 1147. https://doi.org/10.1126/scien​ (supporting); methodology (supporting); writing–original draft (sup- ce.aag0822 porting); writing–review and editing (supporting). Geir Dahle: Data Benjamini, Y., & Hochberg, Y. (1995). Controlling the False Discovery Rate: A practical and powerful approach to multiple testing. Journal curation (equal); methodology (equal); writing–original draft (sup- of the Royal Statistical Society: Series B (Methodological), 57(1), 289– porting); writing–review and editing (supporting). François Besnier: 300. https://doi.org/10.1111/j.2517-6161.1995.tb020​31.x Formal analysis (equal); methodology (equal); writing–original draft Berg, F. (2018). Population structure and dynamics of Atlantic herring. A (supporting); writing–review and editing (supporting). Richard D. case study of herring inhabiting marginal habitats in Landvikvannet and its vicinity. (Degree of Philosophiae Doctor (PhD) PhD thesis). M. Nash: Data curation (supporting); writing–original draft (sup- University of Bergen. porting); writing–review and editing (supporting). Kevin A. Glover: Berg, F., Slotte, A., Andersson, L., & Folkvord, A. (2019). Genetic origin Conceptualization (equal); data curation (supporting); formal analy- and salinity history influence the reproductive success of Atlantic sis (supporting); funding acquisition (equal); project administration herring. Marine Ecology Progress Series, 617–618, 81–94. https://doi. org/10.3354/meps1​2680 (equal); writing–original draft (supporting); writing–review and edit- Berg, F., Slotte, A., Johannessen, A., Kvamme, C., Clausen, L. W., & ing (supporting). Nash, R. D. M. (2017). Comparative biology and population mix- ing among local, coastal and offshore Atlantic herring (Clupea DATA AVA I L A B I L I T Y S TAT E M E N T harengus) in the North Sea, Skagerrak, Kattegat and western Baltic. PLoS One, 12(10), e0187374. https://doi.org/10.1371/journ​ The raw data for SNPs and microsatellites are available in https://doi. al.pone.0187374 org/10.5061/dryad.zs7h4​4j7h. Berg, P. R., Jentoft, S., Star, B., Ring, K. H., Knutsen, H., Lien, S., Jakobsen, K. S., & André, C. (2015). Adaptation to low salin- ORCID ity promotes genomic divergence in Atlantic cod (Gadus morhua L.). Genome Biology and Evolution, 7(6), 1644–1663. https://doi. María Quintela https://orcid.org/0000-0003-4762-2192 org/10.1093/gbe/evv093 Àlex Richter-Boix https://orcid.org/0000-0002-8559-5191 Besnier, F., Kent, M., Skern-Mauritzen, R., Lien, S., Malde, K., Edvardsen, Dorte Bekkevold https://orcid.org/0000-0002-5297-032X R. B., Taylor, S., Ljungfeldt, L. E. R., Nilsen, F., & Glover, K. A. (2014). Florian Berg https://orcid.org/0000-0003-1543-8112 Human-induced evolution caught in action: SNP-array reveals rapid Eeva Jansson https://orcid.org/0000-0002-7840-7201 amphi-atlantic spread of pesticide resistance in the salmon ecotopar- asite Lepeophtheirus salmonis. BMC Genomics, 15(1), 937. https://doi. Geir Dahle https://orcid.org/0000-0002-6070-8073 org/10.1186/1471-2164-15-937 François Besnier https://orcid.org/0000-0002-7186-2121 Björklund, M., & Gustafsson, L. (2015). The stability of the G-matrix: The Kevin A. Glover https://orcid.org/0000-0002-7541-1299 role of spatial heterogeneity. Evolution, 69(7), 1953–1958. https:// doi.org/10.1111/evo.12703 Bolnick, D. I. (2004). Can intraspecific competition drive dis- REFERENCES ruptive selection? An experimental test in natural popula- Albuquerque, U. P., Gonçalves, P. H. S., Ferreira Júnior, W. S., Chaves, tions of sticklebacks. Evolution, 58(3), 608–618. https://doi. L. S., Oliveira, R. C. D. S., Silva, T. L. L. D., Santos, G. C. D., & Araújo, org/10.1111/j.0014-3820.2004.tb016​83.x E. D. L. (2018). Humans as niche constructors: Revisiting the con- Bolnick, D. I., Barrett, R. D. H., Oke, K. B., Rennison, D. J., & Stuart, Y. E. cept of chronic anthropogenic disturbances in ecology. Perspectives (2018). (Non)Parallel evolution. Annual Review of Ecology, Evolution, in Ecology and Conservation, 16(1), 1–11. https://doi.org/10.1016/j. and Systematics, 49(1), 303–330. https://doi.org/10.1146/annur​ev- pecon.2017.08.006 ecols​ys-11061​7-062240 Allendorf, F. W., Luikart, G., & Aitkent, S. N. (2012). Conservation and the Borovski, T., Tadmor-Levi, R., Shapiro, J., Rubinstein, G., Agyakwah, S. genetics of populations (2nd ed.). Wiley-Blackwell. K., Hulata, G., & David, L. (2018). Historical and recent reductions Antao, T., Lopes, A., Lopes, R., Beja-Pereira, A., & Luikart, G. (2008). in genetic variation of the Sarotherodon galilaeus population in the LOSITAN: A workbench to detect molecular adaptation based on a Sea of Galilee. Conservation Genetics, 19(6), 1323–1333. https://doi. FST-outlier method. BMC Bioinformatics, 9(1), 323. org/10.1007/s1059​2-018-1102-7 Arendt, J., & Reznick, D. (2008). Convergence and parallelism recon- Bull, J. W., & Maron, M. (2016). How humans drive speciation as well sidered: What have we learned about the genetics of adaptation? as extinction. Proceedings of the Royal Society B: Biological Sciences, Trends in Ecology & Evolution, 23(1), 26–32. https://doi.org/10.1016/j. 283(1833), 20160600. https://doi.org/10.1098/rspb.2016.0600 tree.2007.09.011 Burke, M. K., & Long, A. D. (2012). What paths do advantageous alleles take Bailey, S. F., Blanquart, F., Bataillon, T., & Kassen, R. (2017). What drives during short-term evolutionary change? Molecular Ecology, 21(20), parallel evolution?: How population size and mutational variation 4913–4916. https://doi.org/10.1111/j.1365-294X.2012.05745.x contribute to repeated evolution. BioEssays, 39(1), 1–9. https://doi. Catanese, G., Watteaux, R., Montes, I., Barra, M., Rumolo, P., Borme, D., org/10.1002/bies.20160​0176 Buongiorno Nardelli, B., Botte, V., Mazzocchi, M. G., Genovese, S., Barrett, R. D. H., Rogers, S. M., & Schluter, D. (2008). Natural selection Di Capua, I., Iriondo, M., Estonba, A., Ruggeri, P., Tirelli, V., Caputo- on a major armor gene in threespine stickleback. Science, 322(5899), Barucchi, V., Basilone, G., Bonanno, A., Iudicone, D., & Procaccini, G. 255. https://doi.org/10.1126/scien​ce.1159978 (2017). Insights on the drivers of genetic divergence in the European Bassham, S., Catchen, J., Lescak, E., von Hippel, F. A., & Cresko, W. A. anchovy. Scientific Reports, 7(1), 4180. https://doi.org/10.1038/ (2018). Repeated selection of alternatively adapted haplotypes s4159​8-017-03926​-z 1708 | QUINTELA et al. Ceballos, G., Ehrlich, P. R., Barnosky, A. D., García, A., Pringle, R. M., & Ecology & Evolution, 23(7), 347–351. https://doi.org/10.1016/j. Palmer, T. M. (2015). Accelerated modern human-induced species tree.2008.04.004 losses: Entering the sixth mass extinction. Science Advances, 1(5), Fang, B., Kemppainen, P., Momigliano, P., Feng, X., & Merilä, J. (2020). e1400253. https://doi.org/10.1126/sciadv.1400253 On the causes of geographically heterogeneous parallel evolution in Chen, Q., Wang, C., Lu, G., Zhao, J., Chapman, D. C., Zsigmond, J., & sticklebacks. Nature Ecology & Evolution, 4(8), 1105–1115. https://doi. Li, S. (2012). Microsatellite genetic diversity and differentiation org/10.1038/s4155​9-020-1222-6 of native and introduced grass carp populations in three conti- Foll, M., & Gaggiotti, O. (2006). Identifying the environmental factors nents. Genetica, 140(4), 115–123. https://doi.org/10.1007/s1070​ that determine the genetic structure of populations. Genetics, 174(2), 9-012-9663-8 875–891. https://doi.org/10.1534/genet​ics.106.059451 Colautti, R. I., & Lau, J. A. (2015). Contemporary evolution during inva- Foll, M., & Gaggiotti, O. (2008). A genome-scan method to identify se- sion: Evidence for differentiation, natural selection, and local adap- lected loci appropriate for both dominant and codominant mark- tation. Molecular Ecology, 24(9), 1999–2017. https://doi.org/10.1111/ ers: A Bayesian perspective. Genetics, 180(2), 977–993. https://doi. mec.13162 org/10.1534/genet​ics.108.092221 Collins, C., & Didelot, X. (2018). A phylogenetic method to perform ge- Frichot, E., & François, O. (2015). LEA: An R package for landscape and nome-wide association studies in microbes that accounts for popula- ecological association studies. Methods in Ecology and Evolution, 6(8), tion structure and recombination. PLOS Computational Biology, 14(2), 925–929. https://doi.org/10.1111/2041-210X.12382 e1005958. https://doi.org/10.1371/journ​al.pcbi.1005958 Frichot, E., Schoville, S. D., Bouchard, G., & François, O. (2013). Testing Colosimo, P. F., Hosemann, K. E., Balabhadra, S., Villarreal, G., Dickson, for associations between loci and environmental gradients using M., Grimwood, J., & Kingsley, D. M. (2005). Widespread parallel evo- latent factor mixed models. Molecular Biology and Evolution, 30(7), lution in sticklebacks by repeated fixation of ectodysplasin alleles. 1687–1699. https://doi.org/10.1093/molbe​v/mst063 Science, 307(5717), 1928. https://doi.org/10.1126/scien​ce.1107239 Gagnaire, P.-A., Broquet, T., Aurelle, D., Viard, F., Souissi, A., Bonhomme, Corona-Santiago, D. K., Doadrio, I., & Domínguez-Domínguez, O. (2015). F., Arnaud-Haond, S., & Bierne, N. (2015). Using neutral, selected, Evolutionary history of the live-bearing endemic Allotoca diazi spe- and hitchhiker loci to assess connectivity of marine populations in cies complex (Actinopterygii, Goodeinae): Evidence of founder ef- the genomic era. Evolutionary Applications, 8(8), 769–786. https://doi. fect events in the Mexican Pre-Hispanic period. PLoS One, 10(5), org/10.1111/eva.12288 e0124138. https://doi.org/10.1371/journ​al.pone.0124138 Galil, B. S., Nehring, S., & Panov, V. (2007). Waterways as invasion high- Crook, D. A., Lowe, W. H., Allendorf, F. W., Erős, T., Finn, D. S., Gillanders, ways – Impact of climate change and globalization. In W. Nentwig B. M., Hadwen, W. L., Harrod, C., Hermoso, V., Jennings, S., Kilada, (Ed.), Biological Invasions (pp. 59–74). Springer, Berlin Heidelberg. R. W., Nagelkerken, I., Hansen, M. M., Page, T. J., Riginos, C., Fry, B., Glover, K. A., Skaala, O., Limborg, M., Kvamme, C., & Torstensen, E. & Hughes, J. M. (2015). Human effects on ecological connectivity (2011). Microsatellite DNA reveals population genetic differen- in aquatic ecosystems: Integrating scientific approaches to support tiation among sprat (Sprattus sprattus) sampled throughout the management and mitigation. Science of the Total Environment, 534, Northeast Atlantic, including Norwegian fjords. Ices Journal of 52–64. https://doi.org/10.1016/j.scito​tenv.2015.04.034 Marine Science, 68(10), 2145–2151. https://doi.org/10.1093/icesj​ Dahlke, F. T., Wohlrab, S., Butzin, M., & Pörtner, H.-O. (2020). Thermal ms/fsr153 bottlenecks in the life cycle define climate vulnerability of fish. Guo, B., Li, Z., & Merilä, J. (2016). Population genomic evidence for Science, 369(6499), 65–70. https://doi.org/10.1126/scien​ce.aaz3658 adaptive differentiation in the Baltic Sea herring. Molecular Ecology, de Villemereuil, P., Frichot, É., Bazin, É., François, O., & Gaggiotti, O. E. 25(12), 2833–2852. https://doi.org/10.1111/mec.13657 (2014). Genome scan methods against more complex models: When Hendry, A. P., Farrugia, T. J., & Kinnison, M. T. (2008). Human influences on and how much should we trust them? Molecular Ecology, 23(8), 2006– rates of phenotypic change in wild animal populations. Molecular Ecology, 2019. https://doi.org/10.1111/mec.12705 17(1), 20–29. https://doi.org/10.1111/j.1365-294X.2007.03428.x Debes, P. V., Zachos, F. E., & Hanel, R. (2008). Mitochondrial phylogeogra- Hendry, A. P., Gotanda, K. M., & Svensson, E. I. (2017). Human influ- phy of the European sprat (Sprattus sprattus L., Clupeidae) reveals iso- ences on evolution, and the ecological and societal consequences. lated climatically vulnerable populations in the Mediterranean Sea and Philosophical Transactions of the Royal Society B: Biological Sciences, range expansion in the northeast Atlantic. Molecular Ecology, 17(17), 372(1712), 20160028. https://doi.org/10.1098/rstb.2016.0028 3873–3888. https://doi.org/10.1111/j.1365-294X.2008.03872.x Hendry, A. P., Taylor, E. B., & McPhail, J. D. (2002). Adaptive divergence DeFaveri, J., & Merilä, J. (2014). Local adaptation to salinity in the three- and the balance between selection and gene flow: Lake and stream spined stickleback? Journal of Evolutionary Biology, 27(2), 290–302. stickleback in the Misty system. Evolution, 56(6), 1199–1216. https:// https://doi.org/10.1111/jeb.12289 doi.org/10.1111/j.0014-3820.2002.tb014​32.x Eggers, F. (2013). Metapopulation dynamics in Atlantic herring (Clupea har- Hooper, D. U., Adair, E. C., Cardinale, B. J., Byrnes, J. E. K., Hungate, B. A., engus L.) along the coast of the southern Norway and in the local area of Matulich, K. L., Gonzalez, A., Duffy, J. E., Gamfeldt, L., & O’Connor, Landvikvannet. (MSc thesis). University of Bergen. M. I. (2012). A global synthesis reveals biodiversity loss as a major Eggers, F., Slotte, A., Libungan, L. A., Johannessen, A., Kvamme, C., driver of ecosystem change. Nature, 486(7401), 105–108. https://doi. Moland, E., Olsen, E. M., & Nash, R. D. M. (2014). Seasonal dynamics org/10.1038/natur​e11118 of Atlantic herring (Clupea harengus L.) populations spawning in the Jakobsson, M., & Rosenberg, N. A. (2007). CLUMPP: A cluster matching vicinity of marginal habitats. PLoS One, 9(11), e111985. https://doi. and permutation program for dealing with label switching and mul- org/10.1371/journ​al.pone.0111985 timodality in analysis of population structure. Bioinformatics, 23(14), Evanno, G., Regnaut, S., & Goudet, J. (2005). Detecting the number 1801–1806. https://doi.org/10.1093/bioin​forma​tics/btm233 of clusters of individuals using the software STRUCTURE: A sim- Johannesson, K., & Andre, C. (2006). Life on the margin: Genetic isolation ulation study. Molecular Ecology, 14(8), 2611–2620. https://doi. and diversity loss in a peripheral marine ecosystem, the Baltic Sea. org/10.1111/j.1365-294X.2005.02553.x Molecular Ecology, 15(8), 2013–2029. Excoffier, L., Laval, G., & Schneider, S. (2005). Arlequin ver. 3.0: An in- Johnston, R. F., & Selander, R. K. (1964). House sparrows: Rapid evolution tegrated software package for population genetics data analysis. of races in North America. Science, 144(3618), 548–550. Evolutionary Bioinformatics Online, 1, 47–50. Jombart, T. (2008). adegenet: A R package for the multivariate analy- Excoffier, L., & Ray, N. (2008). Surfing during population expan- sis of genetic markers. Bioinformatics, 24, 1403–1405. https://doi. sions promotes genetic revolutions and structuration. Trends in org/10.1093/bioin​forma​tics/btn129 QUINTELA et al. | 1709 Jombart, T., Devillard, S., & Balloux, F. (2010). Discriminant analysis of Lotterhos, K. E., & Whitlock, M. C. (2015). The relative power of genome principal components: A new method for the analysis of geneti- scans to detect local adaptation depends on sampling design and cally structured populations. BMC Genetics, 11(1), 94. https://doi. statistical method. Molecular Ecology, 24(5), 1031–1046. https://doi. org/10.1186/1471-2156-11-94 org/10.1111/mec.13100 Jones, F. C., Grabherr, M. G., Chan, Y. F., Russell, P., Mauceli, E., Johnson, Mantel, N. (1967). The detection of disease of clustering and a general- J., Swofford, R., Pirun, M., Zody, M. C., White, S., Birney, E., Searle, ized regression approach. Cancer Research, 27(2), 209–220. S., Schmutz, J., Grimwood, J., Dickson, M. C., Myers, R. M., Miller, C. Marques, D. A., Jones, F. C., Di Palma, F., Kingsley, D. M., & Reimchen, T. T., Summers, B. R., Knecht, A. K., … Kingsley, D. M. (2012). The ge- E. (2018). Experimental evidence for rapid genomic adaptation to a nomic basis of adaptive evolution in threespine sticklebacks. Nature, new niche in an adaptive radiation. Nature Ecology & Evolution, 2(7), 484(7392), 55–61. https://doi.org/10.1038/natur​e10944 1128–1138. https://doi.org/10.1038/s4155​9-018-0581-8 Kaczmarczyk, D., Luczynski, M., & Brzuzan, P. (2012). Genetic variation Matuszewski, S., Hermisson, J., & Kopp, M. (2015). Catch me if you can: in three paddlefish (Polyodon spathula Walbaum) stocks based on Adaptation from standing genetic variation to a moving phenotypic microsatellite DNA analysis. Czech Journal of Animal Science, 57(8), optimum. Genetics, 200(4), 1255–1274. https://doi.org/10.1534/ 345–352. genet​ics.115.178574 Kanalkontoret, (1883). Kanalvæsenets Historie. VIII. Vasdrag vestenfor McCairns, R. J. S., & Bernatchez, L. (2010). Adaptive divergence be- Skienselven til og med Sireaaen (in Norwegian). In (pp. 546). Retrieved tween freshwater and marine stickebacks: Insights into the role from https://www.nb.no/items/​8696f​235eb​bbdf6​c 0d50​132cc​ of phenotypic plasticity from an integrated analysis of candi- 356b38a date gene expression. Evolution, 64(4), 1029–1047. https://doi. Kopp, M., & Matuszewski, S. (2014). Rapid evolution of quantitative org/10.1111/j.1558-5646.2009.00886.x traits: Theoretical perspectives. Evolutionary Applications, 7(1), 169– Mosegaard, H., & Madsen, K. (1996). Discrimination of mixed Herring 191. https://doi.org/10.1111/eva.12127 stocks in the North Sea using vertebral counts and otolith microstructure. Lamichhaney, S., Barrio, A. M., Rafati, N., Sundstrom, G., Rubin, C.-J., Nei, M. (1978). Estimation of average heterozygosity and genetic dis- Gilbert, E. R., Berglund, J., Wetterbom, A., Laikre, L., Webster, M. T., tance from a small number of individuals. Genetics, 89(3), 583–590. Grabherr, M., Ryman, N., & Andersson, L. (2012). Population-scale Nei, M., Tajima, F., & Tateno, Y. (1983). Accuracy of estimated phyloge- sequencing reveals genetic differentiation due to local adaptation netic trees from molecular data. II. Gene frequency data. Journal of in Atlantic herring. Proceedings of the National Academy of Sciences, Molecular Evolution, 19(2), 153–170. https://doi.org/10.1007/BF023​ 109(47), 19345. https://doi.org/10.1073/pnas.12161​28109 00753 Lass, H. U., & Matthäus, W. (2008). General Oceanography of the Baltic Nelson, T. C., & Cresko, W. A. (2018). Ancient genomic variation under- Sea. In R. Feistel, G. Nausch, & N. Wasmund (Eds.), State and evolution lies repeated ecological adaptation in young stickleback populations. of the Baltic Sea, 1952–2005 (pp. 5–43). John Wiley & Sons Inc. Evolution Letters, 2(1), 9–21. https://doi.org/10.1002/evl3.37 Lee, C. E., & Bell, M. A. (1999). Causes and consequences of recent fresh- Nissling, A., Solemdal, P., Svensson, M., & Westin, L. (1994). Survival, ac- water invasions by saltwater animals. Trends in Ecology & Evolution, tivity and feeding ability of Baltic cod (Gadus morhua) yolk-sac larvae 14(7), 284–288. https://doi.org/10.1016/S0169​-5347(99)01596​-7 at different salinities. Journal of Fish Biology, 45(3), 435–445. https:// Lescak, E. A., Bassham, S. L., Catchen, J., Gelmond, O., Sherbick, M. L., doi.org/10.1111/j.1095-8649.1994.tb013​26.x von Hippel, F. A., & Cresko, W. A. (2015). Evolution of stickleback in Ojaveer, H., Jaanus, A., MacKenzie, B. R., Martin, G., Olenin, S., 50 years on earthquake-uplifted islands. Proceedings of the National Radziejewska, T., Telesh, I., Zettler, M. L., & Zaiko, A. (2010). Status Academy of Sciences, 112(52), E7204. https://doi.org/10.1073/ of biodiversity in the Baltic Sea. PLoS One, 5(9), e12467. https://doi. pnas.15120​20112 org/10.1371/journ​al.pone.0012467 Lewontin, R. C., & Krakauer, J. (1973). Distribution of gene frequency Otto, S. P. (2018). Adaptation, speciation and extinction in the as a test of the theory of the selective neutrality of polymorphisms. Anthropocene. Proceedings of the Royal Society B: Biological Sciences, Genetics, 74(1), 175. 285(1891), 20182047. https://doi.org/10.1098/rspb.2018.2047 Li, Y.-L., & Liu, J.-X. (2018). StructureSelector: A web-based software to Palumbi, S. R. (2001). Humans as the world's greatest evolutionary select and visualize the optimal number of clusters using multiple force. Science, 293(5536), 1786–1790. https://doi.org/10.1126/scien​ methods. Molecular Ecology Resources, 18(1), 176–177. https://doi. ce.293.5536.1786 org/10.1111/1755-0998.12719 Peakall, R., & Smouse, P. E. (2006). GenAlEx 6: Genetic analysis in Excel. Limborg, M. T., Hanel, R., Debes, P. V., Ring, A. K., Andre, C., Tsigenopoulos, Population genetic software for teaching and research. Molecular C. S., & Bekkevold, D. (2012). Imprints from genetic drift and muta- Ecology Notes, 6(1), 288–295. tion imply relative divergence times across marine transition zones in Pelletier, F., & Coltman, D. W. (2018). Will human influences on evo- a pan-European small pelagic fish (Sprattus sprattus). Heredity, 109(2), lutionary dynamics in the wild pervade the Anthropocene? BMC 96–107. https://doi.org/10.1038/hdy.2012.18 Biology, 16(1), 7. https://doi.org/10.1186/s1291​5-017-0476-1 Limborg, M. T., Helyar, S. J., De Bruyn, M., Taylor, M. I., Nielsen, E. E., Piry, S., Alapetite, A., Cornuet, J.-M., Paetkau, D., Baudouin, L., & Estoup, Ogden, R. O. B., Carvalho, G. R., FPT Consortium, & Bekkevold, A. (2004). GENECLASS2: A software for genetic assignment and D. (2012). Environmental selection on transcriptome-derived first-generation migrant detection. Journal of Heredity, 95(6), 536– SNPs in a high gene flow marine fish, the Atlantic herring (Clupea 539. https://doi.org/10.1093/jhere​d/esh074 harengus). Molecular Ecology, 21(15), 3686–3703. https://doi. Pritchard, J. K., Stephens, M., & Donnelly, P. (2000). Inference of pop- org/10.1111/j.1365-294X.2012.05639.x ulation structure using multilocus genotype data. Genetics, 155(2), Limborg, M. T., Pedersen, J. S., Hemmer-Hansen, J., Tomkiewicz, J., & 945–959. Bekkevold, D. (2009). Genetic population structure of European Quintela, M., Kvamme, C., Bekkevold, D., Nash, R. D. M., Jansson, E., sprat Sprattus sprattus: Differentiation across a steep environmen- Sørvik, A. G., Taggart, J. B., Skaala, Ø., Dahle, G., & Glover, K. A. tal gradient in a small pelagic fish. Marine Ecology Progress Series, 379, (2020). Genetic analysis redraws the management boundaries for the 213–224. https://doi.org/10.3354/meps0​7889 European sprat. Evolutionary Applications, 13(8), 1906–1922. https:// Losos, J. B., Warheitt, K. I., & Schoener, T. W. (1997). Adaptive differen- doi.org/10.1111/eva.12942 tiation following experimental island colonization in Anolis lizards. Rambaut, A. (2009). FigTree 1.4.3. Retrieved from http://tree.bio.ed.ac. Nature, 387(6628), 70–73. https://doi.org/10.1038/387070a0 uk/softw​are/figtr​ee/ 1710 | QUINTELA et al. Rannala, B., & Mountain, J. L. (1997). Detecting immigration by using mul- Contrasting effects of environment and genetics generate a con- tilocus genotypes. Proceedings of the National Academy of Sciences, tinuum of parallel evolution. Nature Ecology & Evolution, 1(6), 0158. 94(17), 9197–9201. https://doi.org/10.1038/s4155​9-017-0158 Reznick, D. N., & Ghalambor, C. K. (2001). The population ecology of Takezaki, N., Nei, M., & Tamura, K. (2010). POPTREE2: Software for con- contemporary adaptations: What empirical studies reveal about the structing population trees from allele frequency data and comput- conditions that promote adaptive evolution. Genetica, 112(1), 183– ing other population statistics with Windows-interface. Molecular 198. https://doi.org/10.1023/A:10133​52109042 Biology and Evolution, 27, 747–752. https://doi.org/10.1093/molbe​v/ Rosenberg, M. S., & Anderson, C. D. (2011). PASSaGE: Pattern msp312 analysis, spatial statistics and geographic exegesis. Version 2. Terekhanova, N. V., Barmintseva, A. E., Kondrashov, A. S., Bazykin, G. Methods in Ecology and Evolution, 2(3), 229–232. https://doi. A., & Mugue, N. S. (2019). Architecture of parallel adaptation in ten org/10.1111/j.2041-210X.2010.00081.x lacustrine threespine stickleback populations from the White Sea Rosenberg, R., & Palmén, L.-E. (1981). Composition of herring stocks in area. Genome Biology and Evolution, 11(9), 2605–2618. https://doi. the Skagerrak-Kattegat and the relations of these stocks with those org/10.1093/gbe/evz175 of the North Sea and adjacent waters. Fisheries Research, 1, 83–104. Terekhanova, N. V., Logacheva, M. D., Penin, A. A., Neretina, T. V., https://doi.org/10.1016/0165-7836(81)90012​-6 Barmintseva, A. E., Bazykin, G. A., & Mugue, N. S. (2014). Fast evo- Rousset, F. (1997). Genetic differentiation and estimation of gene lution from precast bricks: Genomics of young freshwater popula- flow from F-statistics under isolation by distance. Genetics, 145(4), tions of threespine stickleback Gasterostus aculeatus. PLOS Genetics, 1219–1228. 10(10), e1004696. https://doi.org/10.1371/journ​al.pgen.1004696 Rousset, F. (2008). GENEPOP'007: A complete re-implemen- Therkildsen, N. O., Wilder, A. P., Conover, D. O., Munch, S. B., Baumann, tation of the genepop software for Windows and Linux. H., & Palumbi, S. R. (2019). Contrasting genomic shifts underlie par- Molecular Ecology Resources, 8(1), 103–106. https://doi. allel phenotypic evolution in response to fishing. Science, 365(6452), org/10.1111/j.1471-8286.2007.01931.x 487. https://doi.org/10.1126/scien​ce.aaw7271 Ruggeri, P., Splendiani, A., Occhipinti, G., Fioravanti, T., Santojanni, A., Turcotte, M. M., Reznick, D. N., & Hare, J. D. (2011). The impact of rapid Leonori, I., De Felice, A., Arneri, E., Procaccini, G., Catanese, G., evolution on population dynamics in the wild: Experimental test Tičina, V., Bonanno, A., Nisi Cerioni, P., Giovannotti, M., Grant, W. of eco-evolutionary dynamics. Ecology Letters, 14(11), 1084–1092. S., & Caputo Barucchi, V. (2016). Biocomplexity in populations of https://doi.org/10.1111/j.1461-0248.2011.01676.x European anchovy in the Adriatic Sea. PLoS One, 11(4), e0153061. Velotta, J. P., McCormick, S. D., O’Neill, R. J., & Schultz, E. T. (2014). https://doi.org/10.1371/journ​al.pone.0153061 Relaxed selection causes microevolution of seawater osmoregula- Scharsack, J. P., Schweyen, H., Schmidt, A. M., Dittmar, J., Reusch, T. B., & tion and gene expression in landlocked alewives. Oecologia, 175(4), Kurtz, J. (2012). Population genetic dynamics of three-spined stick- 1081–1092. https://doi.org/10.1007/s0044​2-014-2961-3 lebacks (Gasterosteus aculeatus) in anthropogenic altered habitats. Wang, I. J., & Bradburd, G. S. (2014). Isolation by environment. Molecular Ecology and Evolution, 2(6), 1122–1143. https://doi.org/10.1002/ Ecology, 23(23), 5649–5662. https://doi.org/10.1111/mec.12938 ece3.232 Weir, B. S., & Cockerham, C. C. (1984). Estimating F-statistics for the Schluter, D. (2000). The ecology of adaptive radiation. Oxford University analysis of population structure. Evolution, 38(6), 1358–1370. Press. Whitehead, A., Roach, J. L., Zhang, S., & Galvez, F. (2011). Genomic Schluter, D., & Conte, G. L. (2009). Genetics and ecological speciation. mechanisms of evolved physiological plasticity in killifish distributed Proceedings of the National Academy of Sciences, 106(Supplement 1), along an environmental salinity gradient. Proceedings of the National 9955. https://doi.org/10.1073/pnas.09012​6 4106 Academy of Sciences, 108(15), 6193. https://doi.org/10.1073/ Silva, F. F. G., Slotte, A., Johannessen, A., Kennedy, J., & Kjesbu, O. S. pnas.10175​42108 (2013). Strategies for partition between body growth and repro- Wright, S. (1943). Isolation by distance. Genetics, 28(2), 114–138. ductive investment in migratory and stationary populations of spring-spawning Atlantic herring (Clupea harengus L.). Fisheries Research, 138, 71–79. https://doi.org/10.1016/j.fishr​es.2012.07.013 S U P P O R T I N G I N FO R M AT I O N Sjöblom, V., & Parmanne, R. (1980). Sprat off the coast of Finland in 1977 Additional supporting information may be found online in the and 1978. Annals of Biology, 35, 200–203. Slatkin, M. (1993). Isolation by Distance in equilibrium and non-equi- Supporting Information section. librium populations. Evolution, 47(1), 264–279. https://doi. org/10.2307/2410134 Stern, D. L. (2013). The genetic causes of convergent evolution. Nature How to cite this article: Quintela M, Richter-Boix À, Bekkevold Reviews Genetics, 14(11), 751–764. https://doi.org/10.1038/nrg3483 D, et al. Genetic response to human-induced habitat changes Stockwell, C. A., Hendry, A. P., & Kinnison, M. T. (2003). Contemporary in the marine environment: A century of evolution of European evolution meets conservation biology. Trends in Ecology & Evolution, 18(2), 94–101. https://doi.org/10.1016/S0169​-5347(02)00044​-7 sprat in Landvikvannet, Norway. Ecol Evol. 2021;11:1691–1718. Stuart, Y. E., Veen, T., Weber, J. N., Hanson, D., Ravinet, M., Lohman, B. https://doi.org/10.1002/ece3.7160 K., Thompson, C. J., Tasneem, T., Doggett, A., Izen, R., Ahmed, N., Barrett, R. D. H., Hendry, A. P., Peichel, C. L., & Bolnick, D. I. (2017). QUINTELA et al. | 1711 APPENDIX 1 TA B L E A 1 Sprat genotyped at 8 microsatellites Dev Sample N No alleles Ar Ho uHe FIS Dev HWE (B) LD (B) HOL 31 130 16.3 0.88 ± 0.03 0.87 ± 0.03 −0.04 ± 0.04 2 (0) 1 (0) MEL 79 189 17.2 0.80 ± 0.05 0.88 ± 0.03 0.09 ± 0.04 5 (4) 2 (1) FIN 79 185 16.7 0.80 ± 0.05 0.88 ± 0.03 0.09 ± 0.04 4 (3) 1 (1) TRH 80 193 17.3 0.80 ± 0.05 0.88 ± 0.03 0.09 ± 0.04 3 (0) 1 (1) NOR 73 182 16.7 0.81 ± 0.05 0.87 ± 0.03 0.07 ± 0.04 7 (5) 0 (0) SOG 49 150 16.3 0.81 ± 0.05 0.89 ± 0.02 0.08 ± 0.05 2 (1) 3 (2) HAR1 79 173 16.2 0.78 ± 0.04 0.87 ± 0.02 0.10 ± 0.04 4 (4) 0 (0) HAR2 38 129 15.4 0.79 ± 0.02 0.85 ± 0.03 0.05 ± 0.03 2 (1) 1 (1) HAR3 87 193 16.9 0.81 ± 0.03 0.87 ± 0.03 0.06 ± 0.03 4 (4) 0 (0) LYS 99 195 16.5 0.80 ± 0.05 0.87 ± 0.02 0.08 ± 0.04 5 (4) 1 (1) OSL 87 190 16.9 0.81 ± 0.04 0.87 ± 0.03 0.07 ± 0.03 4 (3) 0 (0) LAND15 97 182 12.2 0.63 ± 0.08 0.70 ± 0.08 0.08 ± 0.06 5 (5) 0 (0) NSEA 88 207 18.3 0.79 ± 0.05 0.89 ± 0.03 0.11 ± 0.04 6 (4) 0 (0) CEL 74 193 17.9 0.83 ± 0.06 0.88 ± 0.03 0.06 ± 0.04 2 (1) 1 (1) GOT 43 121 13.5 0.84 ± 0.04 0.82 ± 0.04 −0.04 ± 0.03 3 (0) 1 (0) Note: Summary statistics per sampling site: Number of individuals; number of alleles; allelic richness (Ar); observed heterozygosity, Ho (mean ± SE); unbiased expected heterozygosity, uHe (mean ± SE); inbreeding coefficient, FIS (mean ± SE); number of deviations from Hardy–Weinberg equilibrium (HWE) at α = 0.05; number of deviations from Linkage Disequilibrium (LD) at α = 0.05 both before and after (B) Bonferroni correction. 1712 | TA B L E A 2 Sprat genotyped at 8 microsatellites HOL MEL FIN TRH NOR SOG HAR1 HAR2 HAR3 LYS OSL LAND15 NSEA CEL GOT HOL * 0.040 0.009 0.577 0.021 0.047 0.396 0.817 0.020 0.005 0.001 0.000 0.000 0.000 0.000 MEL 0.005 * 0.226 0.482 0.426 0.126 0.443 0.155 0.353 0.000 0.002 0.000 0.000 0.000 0.000 FIN 0.007 0.001 * 0.561 0.470 0.538 0.527 0.462 0.219 0.000 0.012 0.000 0.000 0.000 0.000 TRH 0.000 0.000 0.000 * 0.288 0.190 0.569 0.794 0.352 0.000 0.107 0.000 0.000 0.000 0.000 NOR 0.005 0.000 0.000 0.001 * 0.114 0.482 0.096 0.411 0.006 0.001 0.000 0.000 0.000 0.000 SOG 0.005 0.003 0.000 0.002 0.003 * 0.640 0.778 0.051 0.044 0.007 0.000 0.001 0.000 0.000 HAR1 0.001 0.000 0.000 0.000 0.000 0.000 * 0.997 0.331 0.462 0.065 0.000 0.000 0.000 0.000 HAR2 0.000 0.003 0.000 0.000 0.003 0.000 0.000 * 0.539 0.127 0.144 0.000 0.000 0.000 0.000 HAR3 0.005 0.001 0.001 0.001 0.000 0.003 0.001 0.000 * 0.000 0.006 0.000 0.000 0.000 0.000 LYS 0.008 0.007 0.006 0.007 0.004 0.003 0.000 0.003 0.005 * 0.000 0.000 0.000 0.000 0.000 OSL 0.011 0.005 0.004 0.002 0.006 0.006 0.002 0.002 0.004 0.011 * 0.000 0.000 0.000 0.000 LAND15 0.087 0.103 0.102 0.100 0.107 0.101 0.086 0.080 0.095 0.079 0.105 * 0.000 0.000 0.000 NSEA 0.016 0.007 0.007 0.008 0.010 0.008 0.011 0.014 0.009 0.015 0.010 0.117 * 0.163 0.000 CEL 0.015 0.008 0.008 0.007 0.010 0.012 0.011 0.019 0.009 0.017 0.011 0.114 0.002 * 0.000 GOT 0.038 0.025 0.020 0.025 0.020 0.026 0.023 0.032 0.023 0.024 0.019 0.137 0.027 0.030 * Note: Heatmap of pairwise FST below diagonal and the corresponding p-values after 10,000 permutations above diagonal. Cells in boldface type depict values significantly different from zero and the ones highlighted in pink depict values that retained significance after sequential Bonferroni correction. QUINTELA et al. QUINTELA et al. | 1713 TA B L E A 3 Sprat genotyped at 91 SNPs: Outcome of STRUCTURESelector and Evanno test after ten STRUCTURE runs from K1 to K7 STRUCTURESelector Evanno test K Runs MedMed MedMean MaxMed MaxMean Mean LnP(K) Stdev LnP(K) Ln′(K) |Ln″(K)| ΔK 1 10 1 1 1 1 −198719.85 0.05 NA NA NA 2 10 2 2 2 2 −195809.77 2.74 2,910.08 731.26 267.02 3 10 3 3 3 3 −193630.95 16.86 2,178.82 1,170.00 69.41 4 10 4 4 4 4 −192622.13 27.26 1,008.82 29.74 1.09 5 10 5 3 5 4 −191643.05 193.65 979.08 445.99 2.30 6 10 3 3 3 3 −191109.96 43.81 533.09 182.68 4.17 7 10 3 3 3 3 −190759.55 24.42 350.41 NA NA MedMedK MedMeaK MaxMedK MaxMeaK 5 4 5 4 (Continues) 1714 | QUINTELA et al. TA B L E A 3 (Continued) TA B L E A 4 Assignment analysis for individuals genotyped at SNPs to the different genetic clusters NorthSea, Correct self-assign Norway Landvik Katt-Skag Baltic Outgr (%) Norway 987 21 135 37 0 83.6 Landvik 23 268 7 2 0 89.3 NorthSea, 70 3 808 69 0 85.1 Kattegat–Skagerrak Baltic 10 0 18 280 0 90.9 Outgroups 0 0 0 0 66 100.0 Note: Values in the diagonal gray cells depict the correct self-assignment to cluster (in numbers). The last column shows the same values but in percentage. Overall correct self-assignment was 86%. QUINTELA et al. | 1715 1.0 0.8 0.6 0.4 0.2 0.0 F I G U R E A 1 Extract of STRUCTURE barplot corresponding to LAND19 with individuals ordered by increasing ancestry to the Norwegian cluster (green). The rest of the clusters are depicted by the following colors: purple (Landvikvannet), blue (Baltic Sea) and orange (North Sea– Kattegat–Skagerrak) 1716 | QUINTELA et al. (a) (b) F I G U R E A 2 Relationship between Landvikvannet sprat and the 14 reference samples genotyped at microsatellites according to (a) STRUCTURE analysis, (b) DAPC, and (c) PCoA. The plot in (d) represents genetic distance measured as pairwise FST/(1 − FST ) between the northernmost site (HOL) and each of the 14 remaining ones versus the corresponding shortest water distance (in km) QUINTELA et al. | 1717 (c) (d) FIGURE A2 (Continued) 1718 | QUINTELA et al. F I G U R E A 3 Origin of Landvikvannet sprat: Neighbor-joining tree placing Landvikvannet in context with the 40 reference samples after removing the SNP loci candidates to directional selection flagged by LOSITAN and BayeScan. Names of sampling sites by numbers can be retrieved in Table 1