Long-term effective population sizes, temporal stability of genetic composition and potential for local adaptation in anadromous brown trout (Salmo trutta) populations

We examined the long-term temporal (1910s to 1990s) genetic variation at eight microsatellite DNA loci in brown trout (Salmo trutta L) collected from five anadromous populations in Denmark to assess the long-term stability of genetic composition and to estimate effective population sizes (Ne). Contemporary and historical samples consisted of tissue and archived scales, respectively. Pairwise thetaST estimates, a hierarchical analysis of molecular variance (amova) and multidimensional scaling analysis of pairwise genetic distances between samples revealed much closer genetic relationships among temporal samples from the same populations than among samples from different populations. Estimates of Ne, using a likelihood-based implementation of the temporal method, revealed Ne >or= 500 in two of three populations for which we have historical data. A third population in a small (3 km) river showed Ne >or= 300. Assuming a stepping-stone model of gene flow we considered the relative roles of gene flow, random genetic drift and selection to assess the possibilities for local adaptation. The requirements for local adaptation were fulfilled, but only adaptations resulting from strong selection were expected to occur at the level of individual populations. Adaptations resulting from weak selection were more likely to occur on a regional basis, i.e. encompassing several populations. Ne appears to have declined recently in at least one of the studied populations, and the documented recent declines of many other anadromous brown trout populations may affect the persistence of local adaptation.     

The problem of sampling families rather than populations: relatedness among individuals in samples of juvenile brown trout Salmo trutta L.

In species exhibiting a nonrandom distribution of closely related individuals, sampling of a few families may lead to biased estimates of allele frequencies in populations. This problem was studied in two brown trout populations, based on analysis of mtDNA and microsatellites. In both samples mtDNA haplotype frequencies differed significantly between age classes, and in one sample 17 out of 18 individuals less than 1 year of age shared one particular mtDNA haplotype. Estimates of relatedness showed that these individuals most likely represented only three full-sib families. Older trout exhibiting the same haplotypes generally were not closely related.     

Dynamic micro-geographic and temporal genetic diversity in vertebrates: the case of lake-spawning populations of brown trout (Salmo trutta)

Conservation of species should be based on knowledge of effective population sizes and understanding of how breeding tactics and selection of recruitment habitats lead to genetic structuring. In the stream‐spawning and genetically diverse brown trout, spawning and rearing areas may be restricted source habitats. Spatio–temporal genetic variability patterns were studied in brown trout occupying three lakes characterized by restricted stream habitat but high recruitment levels. This suggested non‐typical lake‐spawning, potentially representing additional spatio–temporal genetic variation in continuous habitats. Three years of sampling documented presence of young‐of‐the‐year cohorts in littoral lake areas with groundwater inflow, confirming lake‐spawning trout in all three lakes. Nine microsatellite markers assayed across 901 young‐of‐the‐year individuals indicated overall substantial genetic differentiation in space and time. Nested gene diversity analyses revealed highly significant (≤P = 0.002) differentiation on all hierarchical levels, represented by regional lakes (F_LT = 0.281), stream vs. lake habitat within regional lakes (F_HL = 0.045), sample site within habitats (F_SH = 0.010), and cohorts within sample sites (F_CS = 0.016). Genetic structuring was, however, different among lakes. It was more pronounced in a natural lake, which exhibited temporally stable structuring both between two lake‐spawning populations and between lake‐ and stream spawners. Hence, it is demonstrated that lake‐spawning brown trout form genetically distinct populations and may significantly contribute to genetic diversity. In another lake, differentiation was substantial between stream‐ and lake‐spawning populations but not within habitat. In the third lake, there was less apparent spatial or temporal genetic structuring. Calculation of effective population sizes suggested small spawning populations in general, both within streams and lakes, and indicates that the presence of lake‐spawning populations tended to reduce genetic drift in the total (meta‐) population of the lake.     

Mitochondrial DNA sequence variation and phylogeography among Salmo trutta L. (Greek brown trout) populations

To investigate the phylogenetic relationships and geographical structure among brown trout S. trutta L. populations from the South Adriatic-Ionian and Aegean sea basins, mitochondrial DNA nucleotide sequence comparisons were used. A 310-base-pair (bp) segment of the control region (D-loop), and an additional 280-bp segment of the cytochrome b gene were sequenced from representatives of 13 brown trout populations. Phylogenetic analyses, conducted after combining the data presented with published data from other Eurasian brown trout, revealed four major phylogenetic groups, three of which were found widely distributed within the southern Balkan region. The phylogeographical patterns revealed by mtDNA represent one of the few cases where phylogenetic discontinuity in a gene tree exists without obvious geographical localization within a species' range and has most likely resulted from the differentiation of the major mtDNA clades during Messinian or early Pleistocene times. Finally, the genetic relationships among the populations suggested by mtDNA were generally not in accordance with either allozyme or morphological data.     

Spatio-temporal genetic variability in sea trout (Salmo trutta) populations from north-western Spain

1. Genetic variation at five microsatellite loci was investigated in six sea trout ( Salmo trutta ) populations to describe their spatio-temporal genetic variation in north-western Spain. We observed significant genetic variation between river basins, and isolation by distance with restricted gene flow between neighbouring rivers, which suggests an important homing behaviour.
2. Despite these populations suffering a serious demographic decline during 1998, we did not detect any reduction in their genetic variation, suggesting a reasonably high effective population size and temporal stability.
3. Genetic differences among rivers should be taken into account in future management activities. Given the high genetic variability and the temporal stability observed, we believe that no supportive breeding programmes are presently needed in these populations.     

Initial feeding in migratory brown trout (Salmo trutta L.) alevins

Initial feeding of brown trout was investigated under laboratory conditions. Fifty per cent feeding occurred when yolk constituted approximately 31% of total alevin dry weight, and feeding rate was positively correlated to developmental stage. The possible ecological implications of initial feeding are discussed.     

Genetic detection of sex-specific dispersal in historical and contemporary populations of anadromous brown trout Salmo trutta

The study of sex-biased dispersal has attracted considerable attention in birds and mammals, but less in other taxa, including fishes. We analysed sex-specific dispersal in historical (1910s and 1950s) and contemporary (1990s) samples of anadromous brown trout Salmo trutta. We tested the hypothesis that dispersal is unbiased using information from microsatellite DNA and applying an assignment index for 11 temporally and spatially separated samples. Our results are most consistent with brown trout dispersal being male biased, and provide no evidence of female bias. We found no evidence that dispersal patterns changed over time, indicating that stocking with hatchery strains did not affect sex-specific dispersal.     

Genetic structure and demographic history of brown trout (Salmo trutta) populations from the southern Balkans

1. The present study was designed to characterize the genetic structure of brown trout ( Salmo trutta ) populations from the southern Balkans and to assess the spread of non-native strains and their introgression into native trout gene pools. We analysed polymorphism at nine microsatellite loci in seven supposedly non-admixed and three stocked brown trout populations.
2. The analyses confirmed the absence of immigration and extraordinarily strong genetic differentiation among the seven non-introgressed populations in parallel with low levels of intrapopulation genetic variability. In contrast, analyses of the stocked populations revealed that the genetic integrity of the local populations had been substantially changed, and the populations must be characterized as hybrid swarms. The pattern of population differentiation observed at microsatellites contrasted to that depicted previously by mtDNA variation. However, the close relationships between populations from the Danube and Axios river systems proposed solely by microsatellites could be explained by palaeogeographic events.
3. Our research showed that most of the populations examined represent unique gene pools, whose existence is critically compromised. Therefore, appropriate management and conservation strategies should be developed urgently in order to protect the subspecific biodiversity and to reverse currently negative trends.     

Microsatellite and mitochondrial DNA polymorphism reveals life-history dependent interbreeding between hatchery and wild brown trout (Salmo trutta L.)

The effects of stocking hatchery trout into wild populations were studied in a Danish river, using microsatellite and mitochondrial DNA (mtDNA) markers. Baseline samples were taken from hatchery trout and wild trout assumed to be unaffected by previous stocking. Also, samples were taken from resident and sea trout from a stocked section of the river. Genetic differentiation between the hatchery strain and the local wild population was modest (microsatellite FST = 0.06). Using assignment tests, more than 90% of individuals from the baseline samples were classified correctly. Assignment tests involving samples from the stocked river section suggested that the contribution by hatchery trout was low among sea trout (< 7%), but high (46%) among resident trout. Hybrid index analysis and a high percentage of mtDNA haplotypes specific to indigenous trout observed among resident trout that were assigned to the hatchery strain suggested that interbreeding took place between hatchery and wild trout. The latter result also indicated that male hatchery trout contributed more to interbreeding than females. We suggest that stronger selection acts against stocked hatchery trout that become anadromous compared to hatchery trout that become resident. As most resident trout are males this could also explain why gene flow from hatchery to wild trout appeared to be male biased. The results show that even despite modest differentiation at neutral loci domesticated trout may still perform worse than local populations and it is important to be aware of differential survival and reproductive success both between life-history types and between sexes.     

A comparison of AFLPs and microsatellites to identify the population structure of brown trout (Salmo trutta L.) populations from Hardangervidda, Norway

A total of 495 fish from 11 Hardangervidda lakes were genotyped in order to compare amplified fragment length polymorphisms (AFLP) and microsatellites in terms of their capacity to infer population genetic structure. The 11 microsatellites used in this study gave a greater polymorphism information content and greater gene diversity, with an average of 14.8 alleles per locus, than the six AFLP primer combinations used. However, the AFLPs resulted in 178 polymorphic loci and a 3.1 times larger marker index (effective multiplex ratio multiplied with the gene diversity). Comparable population structuring, for example in terms of distinguishing fish from the different river systems, was obtained with both marker systems. An AFLP and microsatellite multilocus Bayesian assignment test with the structure program divided the fish into six groups largely concurrent with main branches on a population neighbour-joining tree. Yet, the admixture status of individuals is mostly contradictory in the AFLP and the microsatellite analyses. The results are discussed concerning migration between lake populations.     

Allozymic evidence of parapatric differentiation of brown trout (Salmo trutta L.) within an Atlantic river basin of the Iberian Peninsula

The genetic variation of brown trout from Duero, one of the main Atlantic Iberian river basins, was assessed at 34 enzymatic loci in 62 native populations. A strong intrabasin differentiation was detected (G(ST) = 0.46; range D: 0-0.066), mainly attributable to the existence of two divergent groups of populations within Duero: southern and northern groups. This divergence was mainly a consequence of the unequal distribution of *75 and *100 alleles at sMDH-B1,2* isoloci, which were correlated with substantial differences in genetic diversity among regions. The Lower Course region (nearly fixed for the *100 allele) and Pisuerga River (nearly fixed for the *75 allele) showed lower heterozygosities (H approximately 0.8%) in contrast with adjacent areas, which evidenced intermediate frequencies for both alleles and higher heterozygosities (H: 2.2-3.1%). Vicariance appeared as the more probable explanation for the significant positive correlation detected between genetic and geographical distances in Duero Basin. Genetic relationships with adjacent Iberian drainages indicate a close similarity between the southern group and Cantabric trout, whereas the northern group constitutes an ancient form from this basin. This study confirmed complex genetic relationships in brown trout from northwest Iberia, reasserting the existence of clines at several loci and for genetic diversity. The interaction between Cantabric and Duero trout, as well as the location of the limit of the anadromous form around the 42 degrees N parallel, are both required to understand the genetic characteristics of brown trout from this area.     

Genetic structure of brown trout (Salmo trutta, L.) populations from south-western France: data from mitochondrial control region variability

The genetic relationships between several French Atlantic populations of brown trout were analysed using polymerase chain reaction-single-strand conformation polymorphism and sequencing of two parts of the mitochondrial control region. Ten mitochondrial haplotypes were identified, separated by a small number of mutations. The distribution of these haplotypes confirmed our hypothesis of the existence of two genetically well-differentiated groups of populations in this area, which has already been suggested using nuclear markers. Nuclear and mitochondrial markers, however, show different patterns of genetic differentiation: some populations which belong to one group according to allozyme results appear to be part of the other when using mitochondrial DNA. The origin of these different forms is discussed; they seem to belong to the Atlantic grouping. Nevertheless, the hypothesis of an ancestral and a modern form suggested for northern populations does not appear valid in this particular case.     

Piscivory by brown trout Salmo trutta L. and Arctic charr Salvelinus alpinus (L.) in Norwegian lakes

Size and frequency of occurrence of prey of brown trout Salmo trutta L. and Arctic charr Salvelinus alpinus (L.) were recorded in 13 Norwegian lakes during 1973–1990. Piscivores usually comprised less than 5% of the total population. Arctic charr were less piscivorous than brown trout. Trout and charr became piscivorous at 13 and 16 cm length, respectively. These size thresholds were similar to those of other facultative piscivorous freshwater fish species. When present, three-spined sticklebacks, Gasterosteus aculeatus (L.), were preferred by all length groups of piscivorous brown trout and Arctic charr. Length of prey increased with increasing predator length, and the mean body length of prey was about 33 and 25% of predator length for trout and charr, respectively. Yearlings of charr were not recorded as prey.     

Spatial and temporal genetic differentiation and effective population size of brown trout (<Emphasis Type="Italic">Salmo trutta</Emphasis>, L.) in small Danish rivers

The spatial and temporal genetic structure of brown trout populations from three small tributaries of Lake Hald, Denmark, was studied using analysis of variation at eight microsatellite loci. From two of the populations temporal samples were available, separated by up to 13 years (3.7 generations). Significant genetic differentiation was observed among all samples, however, hierarchical analysis of molecular variance (AMOVA) showed that differentiation among populations accounted for a non-significant amount of the genetic differentiation, whereas differentiation among temporal samples within populations was highly significant (0.0244, P<0.001). Estimates of effective population size (N _e) using a maximum-likelihood based implementation of the temporal method, yielded small values (N _e ranging from 33 to 79). When a model was applied that allows for migration among populations, N _e estimates were even lower (24–54), and migration rates were suggested to be high (0.13–0.36). All samples displayed a clear signal of a recent bottleneck, probably stemming from a period of unfavourable conditions due to organic pollution in the 1970–1980’s. By comparison to other estimates of N _e in brown trout, Lake Hald trout represent a system of small populations linked by extensive gene flow, whereas other populations in larger rivers exhibit much higher N _e values and experience lower levels of immigration. We suggest that management considerations for systems like Lake Hald brown trout should focus both on a regional scale and at the level of individual populations, as the future persistence of populations depends both on maintaining individual populations and ensuring sufficient migration links among these populations.     

Estimating the long-term effects of stocking domesticated trout into wild brown trout (Salmo trutta) populations: an approach using microsatellite DNA analysis of historical and contemporary samples

Indigenous salmonid fish gene pools are affected by domesticated conspecifics, derived from aquaculture escapes and deliberate releases. Variability was examined at nine microsatellite loci in order to assess the long-term impact of stocking domesticated trout in two brown trout populations. The study was based on analysis of two historical samples (1945-56), represented by old scale collections, and seven contemporary samples (1986-2000). In one population historical and contemporary samples were remarkably genetically similar despite more than a decade of intense stocking. Estimation of admixture proportions showed a small genetic contribution from domesticated trout (approximately 6%), and individual admixture analysis demonstrated a majority of nonadmixed individuals. The expected genetic contribution by domesticated trout was 64%, assessed from the number of stocked trout and assuming equal survival and reproductive performance of wild and domesticated trout. This demonstrates poor performance and low fitness of domesticated trout in the wild. In another population there was a strong genetic contribution from domesticated trout (between 57% and 88% in different samples), both in samples from a broodstock thought to represent the indigenous population and in a sample of wild spawners. Survival of domesticated trout and admixture with indigenous fish in the broodstock and subsequent stocking into the river, combined with a low population size of native trout relative to the number of stocked trout, could explain the observed introgression. Few nonadmixed individuals remained in the introgressed population, and I discuss how individual admixture analysis can be used to identify and conserve nonintrogressed remains of the population.     

Phylogeographical lineages in brown trout (Salmo trutta): investigating microgeographical differentiation between native populations from Northern Spain

SUMMARY 1. The large microgeographical differentiation revealed by allozyme studies in brown trout ( Salmo trutta) populations is one of the most striking features of this species. Additionally, allozymes showed great genetic differences between Atlantic and Mediterranean populations on a macrogeographical scale.
2. This study was carried out in order to assess whether the great differences observed between Atlantic and Mediterranean populations persisted where the two are geographically close (the 'microgeographical scale'). Sixteen populations of brown trout, S. trutta , were screened for genetic variation at 25 allozyme loci. The sampling sites, which occupied a relatively small geographical area, were distributed across Cantabrian (Atlantic) and Mediterranean drainages in Northern Spain.
3. The neighbour-joining tree, inferred from Nei's genetic distance, showed that brown trout populations clustered into two different groups. These groups corresponded to the Cantabrian and the Mediterranean groups of populations, although no clear geographical pattern emerged within each of the groups. This geographical pattern is basically caused by significant differences in the frequency distribution of the CK-A1 * locus, with a higher frequency of * 115 in Cantabrian samples (0.586 ± 0.091) while allele * 100 was more frequent in Mediterranean samples (0.931 ± 0.038). In addition, this study revealed alleles exclusive to the Mediterranean and Cantabrian populations, agreeing with previous findings.
4. Genetic differentiation between Cantabrian and Mediterranean regions (14.19%) was similar to that estimated in Spain at a larger scale (13%), showing that most of the differences between the regions can be observed even in a small geographical area.     

Morphological variability among seven populations of brown trout, Salmo trutta L., in Greece

Morphological variation in five meristic and 21 morphometric characters was examined in the brown trout, Salmo trutta L., from seven distinct streams and rivers from N and NW Greece. Variation within each sample, estimated as the multivariate generalization of the coefficient of variation, did not vary substantially among the populations. The phenetic relationships of these populations were examined using the stepwise discriminant analysis. The taxonomic implications of these results are discussed.     

High level of population genetic structuring in lake-run brown trout, Salmo trutta, of the Inari Basin, northern Finland

Rivers draining into (Lake) Inarijärvi, northern Finland, sustain a number of lake‐run brown trout, Salmo trutta, populations but, as with most lake‐run S. trutta systems, the level of population genetic structuring among populations is unknown. To address this and to assist fish stock management in the region, the population genetic structure of S. trutta collected from 28 sampling sites in rivers flowing into Inarijärvi was studied using 13 microsatellite loci. Populations were clustered into three separate groups, largely corresponding to geographic regions, with between‐region F_ST values ranging from 0·11 to 0·16. The significant differentiation observed between most populations within each region also implies that individual populations should be recognized as separate management units and actions to improve, and subsequently maintain, conditions for natural spawning should be prioritized. The results of this study further indicate that the trout from each of these regions may have different biological characteristics, such as local‐lake feeding behaviour among the western populations and strong isolation among the northern stocks. As a consequence, further research is warranted to better understand the level of ecological uniqueness of lake‐run S. trutta populations.     

Spatio-temporal genetic structuring of brown trout (Salmo trutta L.) populations within the River Luga, northwest Russia

The brown trout populations of the Baltic Sea region have been drastically affected by various human activities during the past century. Due to their propensity to home to their natal site to spawn and their tendency to evolve local adaptations, populations may be genetically differentiated in water systems where no physical barriers preventing interbreeding exist. Consequently, identification of management units, a prerequisite for appropriate conservation and management planning, cannot necessarily be deduced from the physical properties of the habitat. In this study, microsatellite markers were employed to assess the spatio-temporal genetic structuring of inter-connected brown trout populations from a river-system in Northwest Russia. Populations were found to be genetically differentiated from each other (global F _ST 0.06) and the genetic structuring within the river to follow an isolation by distance pattern. Indications of temporal stability were found in some populations, however others appeared to be temporally unstable suggesting differences in the demographic forces affecting the populations. Based on the observed isolation by distance pattern of genetic differentiation, preserving several breeding sites spaced evenly throughout the river-system would appear to be more appropriate than focussing conservation effort on any single stretch of the river. Electronic supplementary material  The online version of this article (doi:) contains supplementary material, which is available to authorized users.     

New biochemical genetic markers in the brown trout (Salmo trutta L.)

Three 'new' polymorphic loci in the brown trout (Salmo trutta L.) was detected using starch gel electrophoresis. Evidence for simple Mendelian inheritance at all three loci are presented.