Within the entire gradient, 37 out of 57 alleles were private to a single population Table 1. We studied the MHC and microsatellite genetic variation in R. Four main results can be derived from our analyses. First, the postglacial migration history has resulted in two major clusters currently present in northern Germany and the Scandinavian Peninsula: a northern and a southern. Second, within population genetic variation is higher in the southern as compared to the northern cluster for all the studied genetic markers.
Third, there are indications that selection is likely weaker and drift stronger in the northern cluster.
Fourth, these forces combined have led to a complex pattern of differentiation along the gradient where some loci are more divergent among populations than predicted from drift expectations due to diversifying selection, while other loci are more uniform among populations due to stabilizing selection.
We will discuss each of these conclusions in detail below. Previous phylogeographical studies of R. Accordingly, R. Under this scenario the two lineages met in northern Sweden forming a contact zone, which exact location remains to be formally identified. Our results are in accordance with the previous studies. However, our study suggests that the contact zone lies further south i.
Our results show lower genetic variation for both neutral microsatellites and the MHC II exon 2 locus in populations from the north of Sweden as compared to southern populations. These results are in concordance with studies on different taxa, which have found lower genetic variability in northern Europe as compared to central European populations [ 13 , 16 ], and this is usually interpreted as a consequence of northern Sweden being the last area in Europe to be recolonized after the last glaciation events.
This is also what is predicted by the central marginal hypothesis of genetic variation [ 7 , 8 ]. However, in the case of R. This leads to a prediction that these two localities are the least diverse along each of the respective colonization routes.
The depletion in genetic variation observed in the north suggests lower adaptive potential in response to climate change in northern populations [ 69 ]. The reduced variation in MHC exon II in the northern cluster is in line with the earlier results on two amphibian species [ 52 , 53 , 54 ].
The high levels of expected heterozygosity, the large number of rare alleles - numbers exceeding or being in line with what has previously been reported for refugial populations in unglaciated areas of Europe [ 53 ] — as well as the heterozygosity excess for the MHC exon II in some southern populations might be explained by two hypotheses on pathogen mediate-selection mechanisms PMS. First, the heterozygote advantage hypothesis assumes that heterozygous individuals are favored because they can recognize a broader range of pathogens [ 72 , 73 ].
However, published evidence confirming a MHC-specific heterozygote advantage is limited [ 38 , 74 , 75 ]. Second, the rare allele advantage hypothesis assumes that uncommon alleles in the populations are likely to offer more protection than common alleles and thus confer a selective advantage [ 76 ]. In our data, we found rare alleles in almost all the populations over the entire gradient.
These rare alleles could be a potential source for defense against pathogens in these populations. With our data we cannot distinguish among these two hypotheses and they are not mutually exclusive [ 35 , 38 , 39 , 77 , 78 ]. Further investigation regarding allele frequency distributions and parasite infection are needed to understand which mechanisms are responsible for maintaining genetic variation in relation to parasite resistance.
Outlier analyses suggested that the MHC II exon 2 and RCO loci have been affected by different evolutionary processes in the northern and southern cluster. We saw signs of diversifying selection only in the southern populations while all markers seem to be mainly influenced by drift processes in northern populations.
However, this pattern could be enforced by the fact that we studied fewer populations in the northern cluster. A study in Scandinavian common frog R. In our study, R. Therefore, drift processes are likely to be more important at high latitudes due to a high degree of population fragmentation and low effective population sizes even though we cannot find a clear difference concerning effective sizes among southern and northern populations, Table 2. We found that RCO was under selection in the southern cluster, possibly suggesting selection on development rate along the southern part of the gradient, as found previously at local and broader geographical scales in northern European anurans e.
While we did not find significant diversifying selection in the northern cluster, F ST was still high along the northern part of the gradient. It would be very interesting to link variation in this locus to phenotypic variation in development rate along the latitudinal gradient. When analyses were made on all the populations we found evidence of diversifying selection on two MHC II exon 2 and RCO and signs of stabilizing selection on five loci.
We find two reasons which might explain why we find a high number of loci under stabilizing selection: 1 microsatellites were developed using known sequences of coding genes See; [ 56 , 57 ], 2 the long gradient with a high global F ST allows for a better the detection of stabilizing selection. When the populations were divided into a northern and southern cluster we found evidence for selection in the southern cluster three cases of stabilizing and two cases of diversifying selection while in the northern cluster we could only find signs of stabilizing selection on one locus.
These patterns could reflect actual differences among regions in the relative importance of drift and selection but we advise caution when interpreting the results of this study.
While drift is predicted to be more important in small and fragmented populations and selection is more important in large connected ones [ 3 ], we cannot entirely rule out the possibility that the detected patterns may be partly due to lower sample sizes and fewer population contrasts in the north. So while in line with predictions from population genetic theory, these results should be deemed as tentative. We suggest that genetic variation among the regions and populations can be explained by complex patterns of selection, drift and the two recolonization routes of Scandinavia since the last glaciation see [ 81 ] for a similar example.
Our results provide an example of a situation where the level of adaptive MHC II exon 2 diversity seems to be correlated with neutral diversity among populations.
This is not always the case as depletion in overall genetic variation may or may not be correlated with the amount of adaptive genetic variation [ 4 ]. These results suggest that the MHC II exon 2 locus is under diversifying selection and are in agreement with previous studies finding more differentiation in MHC than in neutral loci suggesting adaptation to local parasite faunas e.
However, earlier studies on differentiation at MHC loci show substantial heterogeneity. Other studies have found no difference among population differentiation at MHC and neutral markers indicating the dominant role of drift, and yet others find no differentiation at the MHC indicating balancing selection maintaining genetic variation see [ 82 ] for a recent example, summarized by [ 83 ].
There is a need for further studies of the processes shaping within and among population genetic variation in natural populations to further improve our understanding on how genetic variation is geographically portioned and distributed.
It lives in diverse habitats, from forests to pastures, and breeds in semi-permanent to permanent ponds and lakes. The centre of the distribution of Rana arvalis is located in the area of eastern Germany and western Poland see distribution map in Sillero et al. We collected R. The average distance between collection sites in the same region was 20 km range 8 to 50 km. The collection sites were ponds situated in flat terrain dominated by mixed forest, pastures and agricultural land.
At each site we collected ca. We successfully genotyped all individuals at 18 microsatellite loci isolated from different Rana species, and tested another set of ten microsatellites without success Additional file 1 : Table S1. Some of the successfully genotyped loci are situated in or near coding regions, increasing the probability of these markers being under selection.
PCR amplification was performed individually for each microsatellite. Prior to genotyping, the PCR products were diluted in water , or Both forward and reverse primers were modified for Illumina MiSeq sequencing with an individual 8 bp barcode and a sequence of three N to facilitate cluster identification.
Each amplicon was marked with an individual combination of a forward and a reverse barcode for identification. All amplifications were carried out using filter tips in separate pre- and post-PCR rooms, and negative controls were included in all amplifications to avoid contaminations.
PCR products were run and visualized on a 1. To reduce the number of samples for subsequent purification, 3—9 PCR products with similar concentrations were pooled based on estimations from the gel image. These sample pools were run on 1. The final amplicon pooling was done according to the measured concentrations and consisted of equimolecular amounts of each sample.
The jMHC software [ 89 ] was used to remove primer sequences and unique tags, and to generate alignments of all variants per amplicon. In this study, however, the distinction of true alleles and artefacts was greatly simplified, for three reasons: 1 amplifying a single gene will only yield one or two true MHC alleles per amplicon, and the project was designed in a way that both 2 replication rate and 3 per amplicon coverage was markedly high.
In a single locus system, it is not expected that chimera are generated in higher frequencies than the two parental alleles, and hence, no chimeric sequence that could have been taken for a valid MHC alleles could be detected when thoroughly checking the sequences per amplicon by eye.
Among amplicons, valid MHC alleles had to be present in at least two amplicons 2-independent-PCR-criterion, [ 91 ] , therefore we replicated all amplicons that revealed unique MHC alleles in the second Miseq run, and ran every individual twice in the second Miseq run from independent PCRs.
In addition, we used the DOC method, not assuming any specific number of loci to identify and estimate the number of alleles Ai per individual. This procedure is based on the break point in sequencing coverage between variants within each individual and avoids choosing a subjective threshold to separate true alleles from artefacts.
In this procedure, variants are sorted top-down by coverage, followed by the calculation of the coverage break point DOC statistic around each variant. The variant with the highest DOC value is assumed to be the last true allele see [ 65 ]. All valid allele sequences were imported and aligned in MEGA 6.
Allele sequences were extensively compared to other sequences for the same locus: giant spiny frog Quasipaa spinosa ; GenBank: KM We removed the intron sequences for further analyses and ended up with a MHC II exon 2 fragment of bps.
We used the programs Lositan [ 94 ] and BayeScan v. In Bayescan, outliers were detected by implementing the multinomial Dirichlet model. We did the division after analyzing our data see below and in accordance with a bidirectional recolonization hypothesis of Scandinavia [ 67 , 68 ].
Outlier analyses were also repeated excluding the German populations from the data set See Additional file 14 : Table S8 and Additional file 15 : Figure S7. Input files were converted for different analyses program formats using the Excel add-in Microsatellite toolkit [ 97 ]. The frequency of null alleles was estimated with two different softwares: FreeNA [ 98 ] and Genepop [ 99 ]. To assess neutral genetic diversity, expected heterozygosity H E , observed heterozygosity H O and allelic richness AR , allele numbers rarified to the smallest sample size were calculated for each locality in FSTAT 2.
Effective population size was estimated based on two different methods: the linkage disequilibrium method and the coancestry method using the software Ne estimator [ ]. To examine population structure and differentiation, global and pair-wise F ST between all the populations [ ] were calculated according to Nei et al.
The Euclidean distance matrix was estimated using the R package Geosphere see Hijmans et al. To test if pair-wise F ST was spatially auto-correlated, Mantel tests were performed in R running the package Adegenet [ ]. We visualized the spatial structure of microsatellite data using discriminant analysis of principal components DAPC implemented in Adegenet [ ].
The optimal number of clusters for the DAPC was chosen based on the lowest Bayesian Information Criterion BIC value for the different clustering solutions, which coincided with a sharp break in the curve of BIC values as a function of k. We investigated the likelihood of various numbers of K 2—5 , following the approach suggested by Evanno et al.
We estimated and visualized the gene flow patterns between all pairs of population samples by using divMigrate-online [ 66 ], which can detect asymmetric gene flow patterns. We included 12 populations, used G ST as implemented in divMigrate-online as the differentiation metric and set the filter threshold to 0. Confidence limits on gene flow estimates were determined by bootstrap replicates.
We tested for Isolation by distance IBD; [ ] , as described above for the microsatellite data. We then computed the restricted major axis regression slopes for the northern and southern population comparisons separately for each comparison with a differentially selected marker.
We hypothesized that if population differentiation was stronger among southern populations, the slope of this regression would be steeper than among northern populations. In summary, populations were shown to be subjected to different selective regimes and combined with different historical demographic patterns affecting the strength of genetic drift, a complex pattern of differentiation have evolved along the gradient.
Some loci are more divergent than expected by drift among populations due to diversifying selection while others are more uniform among populations due to stabilizing selection. Our data show a high number of MHC exon 2 alleles in comparison to other European amphibian species. Both overall and MHC genetic variation are lower at northern latitudes which suggest a high risk of extinction when confronted with emerging pathogens and climate change.
These results emphasize the importance of latitudinal gradient studies in order to elucidate and understand the evolutionary processes shaping genetic variation among natural populations. Nei M. Molecular evolutionary genetics.
New York City: Columbia university press; Google Scholar. Lynch M, Walsh B. The origins of genome architecture. Sunderland: Sinauer Associates; Kimura M.
The neutral theory of molecular evolution. Cambridge: Cambridge University Press; Book Google Scholar. Hedrick PW. Evolutionary genetics of the major histocompatibility complex. Am Nat. Article Google Scholar. The abundant centre distribution: to what extent is it a biogeographical rule?
Ecol Lett. Conserv Genet. Parsons PA. Extreme environmental change and evolution. Mol Ecol. Wright S. Evolution in Mendelian populations. Dobzhansky T, Dobzhansky TG. Genetics and the origin of species. Mayr E. Systematics and the origin of species, from the viewpoint of a zoologist. Cambridge: Harvard University Press; Hewitt GM. Speciation, hybrid zones and phylogeography—or seeing genes in space and time.
The structure of biodiversity—insights from molecular phylogeography. Front Zool. Hampe A, Petit RJ. Conserving biodiversity under climate change: the rear edge matters. PubMed Article Google Scholar. Evolutionary conservation genetics. Oxford: Oxford University Press; Some genetic consequences of ice ages, and their role in divergence and speciation.
Biol J Linn Soc. Post-glacial re-colonization of European biota. Comparative phylogeography and postglacial colonization routes in Europe. Genomics and conservation genetics. Trends Ecol Evol. Biodiversity assessment using markers for ecologically important traits. Neutral DNA markers fail to detect genetic divergence in an ecologically important trait.
Biol Conserv. Adaptive vs. Landsc Ecol. To what extent do microsatellite markers reflect genome-wide genetic diversity in natural populations? Sommer S. Major histocompatibility complex and mate choice in a monogamous rodent.
Behav Ecol Sociobiol. But how does genetic variation increase or decrease? And what effect do fluctuations in genetic variation have on populations over time? Mating patterns are important.
Random forces lead to genetic drift. If the individuals at either end of the range reconnect and continue mating, the resulting genetic intermixing can contribute to more genetic variation overall. However, if the range becomes wide enough that interbreeding between opposite ends becomes less and less likely, and the different forces acting at either end become more and more pronounced, and the individuals at each end of the population range may eventually become genetically distinct from one another.
Here is an example of migration affecting relative allele frequency:. The overall effect. Here is an example of how a specific genotype is less favorable than another genotype:. Genetic variation in a population is derived from a wide assortment of genes and alleles. The persistence of populations over time through changing environments depends on their capacity to adapt to shifting external conditions.
Sometimes the addition of a new allele to a population makes it more able to survive; sometimes the addition of a new allele to a population makes it less able. Still other times, the addition of a new allele to a population has no effect at all, yet the new allele will persist over generations because its contribution to survival is neutral.
Key Questions How can genetic variation influence evolution? What is an example of genetic drift? Topic rooms within Genetics Close. No topic rooms are there. Browse Visually. Other Topic Rooms Genetics. Student Voices. Creature Cast. Simply Science. Green Screen. Green Science. Bio 2. The Success Code. Why Science Matters. The Beyond. Migration is the movement of individuals or gametes from one population to another and is equivalent to gene flow when the migrants contribute their genes to the gene pool of the new population.
Migration is similar to dispersal, but the two are studied with different methods and on different timescales; dispersal is estimated on a short, ecological timescale, whereas migration is estimated from patterns of genetic variation on a longer, evolutionary timescale.
Population structure is the pattern of genetic variation that occurs within and between subpopulations. Migration can be estimated directly by tracking the movement of specific alleles or multilocus genotypes. Migration can be estimated indirectly by assessing the extent of genetic differentiation among subpopulations. High migration rates result in large genetically homogeneous populations; restricted migration results in genetically differentiated subpopulations that diverge by random genetic drift.
0コメント