Given the evidence of significant local ancestry-based assortative mating that we observed for a number of traits, we evaluated whether there were particular ancestry components that were most relevant to mate choice.
For significant polygenic phenotype gene sets of interest, we computed the observed versus expected ancestry homozygosity for each ancestry separately across all genes in the set Figure 5. In Colombia, Peru, and Puerto Rico, assortative mating for this polygenic phenotype is driven by an excess of African homozygosity, whereas in Mexico there is a lack of African homozygosity. Patterns of assortative mating for this trait in Mexico and Peru are driven mainly by European ancestry, whereas Colombia and Puerto Rico show an excess of African ancestry homozygosity for this same trait.
Figure 5. Individual examples of ancestry-based assortative and disassortative mating. Results of meta-analysis of dis assortative mating on polygenic phenotypes along with their ancestry drivers are shown for A an anthropometric trait: Height, B a neurological trait: Schizophrenia, and C the immune-related HLA class I and II genes. Mexico and Peru, on the other hand, have some evidence for assortative mating for the HLA loci; Mexico has the highest estimates of ancestry homozygosity at HLA loci for any of the four populations, and Peru has an excess of European and Native American ancestry homozygosity and a deficit of African homozygosity for these genes.
Similar results for two additional anthropometric phenotypes are shown in Supplementary Figure S11 : body mass index and facial development. These phenotypes show assortative mating in all four populations, with varying components of ancestral homozygosity driving the relationships. When these results are considered together, African ancestry consistently shows the strongest effect on driving assortative and disassortative mating in admixed Latin American populations Figure 5 and Supplementary Figure S We further evaluated the extent to which specific ancestry components may drive assortative mating patterns among admixed individuals by evaluating the variance of the three continental ancestry components among individuals within each Latin American population.
Assortative mating is known to increase population variance for traits that are involved in mate choice; thus, the ancestry components that drive assortative mating in a given population are expected to show higher overall variance among individual genomes. African ancestry fractions show the highest variation among individuals for all four populations Figure 6 , consistent with the results seen for the five specific cases of assortative mating evaluated in Figure 5 and Supplementary Figure S Figure 6.
Inter-individual ancestry variance for the four admixed Latin American populations analyzed here. Variance among individuals for the African blue , European orange , and Native American red ancestry fractions within each population are shown.
Assortative mating is a nearly universal human behavior, and scientists have long been fascinated by the subject Vandenburg, ; Buss, Studies of assortative mating in humans have most often entailed direct measurements of traits — such as physical stature, education, and ethnicity — followed by correlation of trait values between partners.
Decades of such studies have revealed numerous, widely varying traits that are implicated in mate choice and assortative mating. Studies of this kind typically make no assumptions regarding, nor have any knowledge of, the genetic heredity of the traits under consideration. Moreover, the extent to which the expression of these traits varies among human population groups has largely been ignored.
The first attempts to evaluate the genetic contributions to assortative mating entailed twin studies, whereby the similarity of mate choice for dizygotic versus monozygotic twins were compared Lykken and Tellegen, While twin studies did uncover a genetic contribution to variance in human mate choice, they often yielded widely inconsistent results. More recent studies of assortative mating, powered by advances in human genomics, have begun to explore the genetic architecture underlying the human traits that form the basis of mate choice in more detail Domingue et al.
In addition, recent genomic analyses have underscored the extent to which human genetic ancestry influences assortative mating Risch et al.
However, until this time, these two strands of inquiry have not been brought together. The approach that we developed for this study allowed us to directly assess the connection between local genetic ancestry — i. Previous studies on human mate choice have demonstrated pronounced sex differences in mate preference; for example, females value earning capacity more in potential mates, whereas males value reproductive capacity, as inferred from youth and physical attractiveness Buss, ; Geary, It should be noted that the genome-based approach that we employed here does not allow us to consider sex differences in mate preference since we are essentially observing the effects of assortative mating on the offspring of mate pairs, by comparing ancestry homozygosity levels in the genomes of all individuals, rather than directly observing mate choice in couples.
Our approach relies on the well-established principle that assortative mating results in an excess of genetic homozygosity Sebro et al. However, we do not analyze homozygosity of specific genetic variants per se , as is normally done; rather we evaluate excess homozygosity, or the lack thereof, for ancestry-specific haplotypes Figure 2B. By merging this approach with data on the genetic architecture of polygenic human phenotypes, we were able to uncover specific traits that inform ancestry-based assortative mating.
This is because, when individuals exercise mate choice decisions based on ancestry, they must do so using phenotypic cues that are ancestry-associated. In other words, ancestry-based assortative mating is, by definition, predicated upon traits that vary in expression among human population groups Supplementary Figure S An obvious example of this is skin color Hancock et al.
It follows that the assortative mating traits that our study uncovered in admixed Latin American populations must be both genetically heritable and variable among African, European and Native American population groups. The traits we found to influence ancestry-based assortative mating vary among the continental population groups that admixed to form modern Latin American populations Supplementary Table S2.
For example, the anthropometric traits found in our study — body mass, height and facial development — are both heritable and known to vary among ancestry groups. This implies that the genetic variants that influence these traits should also vary among these populations. Accordingly, it is readily apparent that mate choice decisions based on these physical features could track local genetic ancestry.
Interpretation of the neurological traits that show evidence of local ancestry-based assortative mating — schizophrenia and educational attainment — is not quite as straightforward. For schizophrenia, it is far more likely that we are analyzing genetic loci associated with a spectrum of personality traits that influence assortative mating, as opposed to mate choice based on full-blown schizophrenia, and indeed personality traits are widely known to impact mate choice decisions Merikangas, ; Hur, ; Kandler et al.
In addition, since schizophrenia prevalence does not vary greatly world-wide Saha et al. While educational attainment outcomes are largely environmentally determined, recent large-scale GWAS studies have uncovered a substantial genetic component to this trait, which is distributed among scores of loci across the genome Rietveld et al. The population distribution of education-associated variants is currently unknown, but our results suggest the possibility of ancestry-variation for some of them.
Indeed, the average allele frequencies for the variants that influence our top four traits of interest — height, BMI, schizophrenia, and educational attainment — show significant variation among ancestry groups Supplementary Figure S Mate choice based on divergent MHC loci, apparently driven by body odor preferences, is the best known example of human disassortative mating Wedekind et al. However, studies of this phenomenon have largely relied on ethnically homogenous cohorts.
In one case where females were asked to select preferred MHC-mediated odors from males of a different ethnic group, they actually preferred odors of males with more similar MHC alleles Jacob et al. Another study showed differences in MHC-dependent mate choice for human populations with distinct ancestry profiles Chaix et al.
Ours is the first study that addresses the role of ancestry in MHC-dependent mate choice in ethnically diverse admixed populations. Unexpectedly, we found very different results for MHC-dependent mate choice among the four Latin American populations that we studied.
Interestingly, disassortative mating for HLA loci in Colombia and Puerto Rico is largely driven by African ancestry, and these two populations have substantially higher levels of African ancestry compared to Mexico and Peru.
The population- and ancestry-specific dynamics of MHC-dependent mate choice revealed here underscore the complexity of this issue. Given the complexity of the results reported here, particularly as they relate to differences among populations, it should also be noted that anomalous patterns of linkage disequilibrium at MHC loci could confound the analysis at this region. Assortative mating alone is not expected to change the frequencies of alleles, or ancestry fractions in the case of our study, within a population.
Assortative mating does, however, change genotype frequencies, resulting in an excess of homozygous genotypes. Accordingly, ancestry-based assortative mating is expected to yield an excess of homozygosity for local ancestry assignments i. By increasing homozygosity in this way, assortative mating also increases the population genetic variance for the traits that influence mate choice. In other words, assortative mating will lead to more extreme, and less intermediate, phenotypes than expected by chance.
This population genetic consequence of assortative mating allowed us to evaluate the extent to which specific continental ancestries drive mate choice decisions in admixed populations, since specific ancestry drivers of assortative mating are expected to have increased variance. We found that the fractions of African ancestry have the highest variance among individuals for all four populations, consistent with the idea that traits that are associated with African ancestry drive most of the local ancestry-based assortative mating seen in this study Figure 6.
It is important to reiterate that previous studies have shown evidence for assortative mating on both genetic ancestry and specific traits; accordingly, ancestry-based population stratification could lead to the appearance of trait-based assortative mating Sebro et al.
For example, assortative mating occurs among European-Americans along a North-South European ancestry cline, which happens to mirror the cline in height along this same axis Sebro et al. This begs the question as to whether similarities in height among European-American couples is due to ancestry or due to the trait itself.
Studies have shown conflicting results regarding this question. On the one hand, genetic ancestry population stratification alone was posited to account for observed patterns of assortative mating in the US Abdellaoui et al. Here, we have tried to tease apart the overall effects of ancestry-based assortative mating versus trait-specific mate choice by permuting random sets of genes and re-computing our AMI test statistic for each of the four Latin American populations analyzed here. This procedure allowed us to control for the background levels of local ancestry homozygosity in these populations, which could have been generated by ancestry-based assortative mating alone.
We used this control to parameterize the significance levels for the test statistic that we used to discover trait-specific ancestry-based assortative mating Figure 4B.
In other words, we only find a specific trait to be implicated in ancestry-based assortative mating if the levels of ancestry homozygosity for the genes associated with that trait are significantly higher than the genome-wide background levels. In this sense, we have shown how these specific traits may serve as cues that underlie, to some extent, ancestry influenced mate choice in Latin American populations. A corollary to this conclusion is the fact that ancestry- and trait-based assortative mating cannot be completely disentangled for modern admixed populations.
The confluence of African, European and Native American populations that marked the conquest and colonization of the New World yielded modern Latin American populations that are characterized by three-way genetic admixture Wang et al.
Nevertheless, mate choice in Latin America is far from random Risch et al. Indeed, our results underscore the prevalence of ancestry-based assortative mating in modern Latin American societies. The local ancestry approach that we developed provided new insight into this process by allowing us to hone in on the phenotypic cues that underlie ancestry-based assortative mating.
Our method also illuminates the specific ancestry components that drive assortative mating for different traits and makes predictions regarding traits that should vary among continental population groups. Whole genome sequence data and genotypes were merged, sites common to all datasets were kept, and SNP strand orientation was corrected as needed, using PLINK version 1.
The resulting dataset consisted of 1, individuals from 38 populations with variants characterized for , SNPs.
This phased set of SNP genotypes was used for local ancestry analysis. This pruned set of SNP genotypes was used for global ancestry analysis. Continental African, European, and Native American populations were used as reference populations, and contiguous regions with the same ancestry assignment, i. TRACTS was used to evaluate possible admixture timings across 1, bootstrap attempts, with the most likely series of admixture events chosen to represent each population.
For each diploid genome analyzed here, individual genes can have 0, 1, or 2 ancestry assignments depending on the number of high confidence ancestry-specific haplotypes at that locus. Our assortative mating index AMI, see below can only be computed for genes that have 2 ancestry assignments in any given individual, i. The gene sets for the polygenic phenotypes were collected by directly mapping trait-associated SNPs to genes. For each Latin American population, phenotype gene sets were filtered to only include genes that passed the ancestry genotype threshold, as described previously.
Finally, the polygenic phenotype gene sets were filtered based on size, so that all polygenic phenotypes included two or more genes. LD pruning was done using pairwise r 2 -values between genic SNPs for all pairs of genes in any given set. The final data set contains gene sets for polygenic phenotypes, hierarchically organized into three functional categories, including unique genes haplotypes Supplementary Figure S7.
To assess local ancestry-based assortative mating, we developed the AMI, a log odds ratio test statistic that computes the relative local ancestry homozygosity compared to heterozygosity for any given gene.
Ancestry homozygosity occurs when both genes in a genome have the same local ancestry, whereas ancestry heterozygosity refers to a pair of genes in a genome with different local ancestry assignments. The AMI is calculated as:. The observed values of local ancestry homozygous and heterozygous gene pairs are taken from the gene-to-ancestry mapping data for each gene in each population.
Analytical results in models with dominance are difficult to obtain and hence rare in the literature. In our model, three cases are analytically tractable to some extent, namely random mating and weak or strong dominance, and complete assortment and arbitrary intermediate dominance.
The invasion criterion for modifiers of small effect cannot be used in the case of dominance. Instead, we have to calculate approximations for the leading eigenvalues.
Weak dominance: Let dominance be sufficiently weak to neglect terms of order and higher see Appendix S1. In this case, the leading eigenvalue of the linearized transition matrix is 21 Hence, a modifier can invade if and only if. Although the strength of selection for a modifier is a decreasing function in , the invasion criterion is not affected by weak dominance. Strong dominance: Let and assume that terms of order and higher can be neglected see Appendix S1. The leading eigenvalue is 22 and a modifier increasing assortment can invade if.
Note, that the invasion fitness is again a decreasing function in. In the case of complete dominance, , modifiers for assortative mating are selectively neutral and the leading eigenvalues equals 1. This can easily be generalized to modifiers with arbitrary effect.
The above results suggest that dominance decreases the strength of selection for rare assortment modifiers, but has no effect on the condition for invasion cf. This is of course only true in the deterministic model. In a stochastic version dominance would also decrease the probability of successful invasion. Clearly, 21 and 22 imply that the invasion fitness becomes higher as the frequency-dependent effect of competition increases, ie.
Moreover, for small and 21 becomes approximately for weak dominance, and 22 becomes approximately for strong dominance. In particular, as intuitively expected, modifiers become almost selectively neutral for high levels of dominance.
Therefore, invasion fitness seems to be a decreasing function of the level of dominance. The decrease in invasion fitness is not linear and 21 even suggest that modifiers might not be able to invade if dominance is intermediate. However, neither 21 nor 22 is a good approximation for intermediate levels of dominance, and conclusions on this case cannot be drawn. Typical, for the quadratic model is the condition. The fitness changes from stabilizing to disruptive as becomes larger than cf. Intermediate dominance and complete assortment: Suppose dominance is intermediate, i.
Furthermore, assume that the population mates completely assortatively. Then, a unique polymorphic equilibrium exists see Appendix S1. Consider an initially rare modifier that decreases the strength of assortment by an arbitrary amount. In Appendix S1 , we show that such a modifier can never invade, as long as the modifier leads to a positive mating probability between the homozygotes at the ecological locus.
Note that invasion of such a modifier would imply that complete assortment could not be achieved by small steps. If the mating probability between homozygotes is zero, a rare modifier decreasing assortment is neutral. Complete dominance and complete assortment: In Appendix S1 , we show that modifiers decreasing assortment by an arbitrary degree are selectively neutral in populations in which dominance and assortment are initially complete.
The same holds for modifiers decreasing dominance by an arbitrary degree. Here, we compare the initial strength of selection for an increased level of assortment with the selection pressure for an increased level of dominance. The strength of selection for a rare dominance modifier in a randomly mating population for the same ecological model is given in [34].
Hence, we can compare the strength of selection for the different modifiers. If the modifier effects go to zero, the selection coefficients for a dominance modifier and an assortment modifier behave differently see Appendix S1. The strength of selection for a dominance modifier decreases faster than the strength of selection for an assortment modifier. This is consistent with previous results [30] that showed that in symmetric cases selection for an increased level of assortment is stronger than selection for an increased level of dominance if both modifiers have infinitesimally small effects.
Here, we consider the complete evolutionary trajectory of the gene-frequency vector and the population size. A newly introduced modifier can either rise to fixation, die out, or can be maintained at intermediate frequency.
Furthermore, the existence of multiple stable equilibria is possible. Consequently, the fate of a modifier may depend on its initial frequency.
First, we consider modifiers of small effect in an initially randomly mating population. The impact of the modifier's effect size is discussed in Section 3.
Figure 3 illustrates the evolutionary outcome for a modifier with effect. Multiple stable equilibria were not detected, except for complete dominance. Thus for , all results apply for the standard, rare-modifier, and frequent-modifier scenario. For there seems to exist a manifold of equilibria, at which both phenotypes are equally frequent. All trajectories converged to a different equilibrium dependent on the initial conditions. For an initially randomly or completely assortatively mating population the invasion fitness of modifiers equals one, i.
In Figures 3 — 6 , the regions with are marked as regions of maintenance. Modifier that are initially at low frequency or high frequency, will neither get lost nor become fixed. Regions of extinction, maintenance, and fixation of a modifier increasing assortment slightly in an initially randomly mating population. We used a grid with stepsize 0. The other parameters are and.
In addition to the color code, different regions are labeled , where and are the selection regimes that apply if the modifier is rare or frequent, respectively. The color code indicates the different evolutionary outcomes. In the extinction regions, the modifier died out in all runs. In the maintenance regions, the modifier coexisted with the wild type in all runs, whereas in the fixation region the modifier was fixed for all runs.
Parameter combinations for which none of the runs equilibrated within generations are indicated as slow run regions. Regions of extinction, maintenance, and fixation of a modifier increasing assortment with different effects in an initially randomly mating population.
The parameters , , , , and are as in Figure 3. The modifier effects are A , B , C , and D. In addition to the color code, different regions are labeled or , where , , and are the selection regimes that apply if the modifier is rare, at intermediate frequency, or frequent, respectively.
In A, there is no dominance and the modifier cannot go to fixation. In B, dominance is intermediate and the modifier goes to fixation if sufficiently frequent. The strength of competition is in both figures. Furthermore, and. Equilibrium frequencies of homozygotes at the ecological locus are indicated by black bars. The equilibrium frequencies of heterozygotes are negligible and not visible. Regions of extinction, maintenance, and fixation of a modifier increasing assortment slightly.
The parameters , , , and are as in Figure 4. The degree of initial assortment is A and B. As seen in Figure 3 , higher levels of assortment are favored according to the regime in almost the whole parameter space. If viability 5 is shaped and positive frequency-dependence is absent or weak , the regime applies: Two niches exist at the boundary of the phenotypic range, and stabilizing sexual selection is too weak to counteract disruptive selection resulting from competition Figure 1A.
Therefore, higher levels of assortment are favored in this scenario. Dominance weakens disruptive selection at the ecological locus. Thus, this scenario is not very robust to changes in the degree of dominance. We will see, the region in which this scenario applies decreases with increasing assortment. From Figure 3 it becomes clear that assortment cannot evolve at all only if dominance is almost complete and competition is at least moderately strong.
Then, 5 is shaped and the regime applies: Assortative mating and competition are strong enough to establish a niche in the middle of the phenotypic range. In addition, competition is strong relative to assortative mating, such that the net effect of selection is stabilizing.
Assortative mating may induce disruptive sexual selection in this scenario Figure 1D. However, higher levels of assortment are not favored because heterozygotes have a significantly higher viability than homozygotes. The strength of competition that is necessary to establish a niche in the middle of the phenotypic range depends crucially on the frequency of heterozygotes at the ecological locus. Since we restrict attention to at most moderate competition, i. However, if the degree of dominance increases, heterozygote advantage decreases.
Whether a modifier can also go to fixation depends crucially on competition and dominance. Remember that in the quadratic model without dominance, a modifier with small effect goes to fixation if competition is sufficiently strong, i. This results needs to be modified in the full model with dominance. Since dominance decreases the effect of competition, we expect the threshold value of for fixation to increase with.
In fact, a modifier cannot go to fixation if is small and see in Figure 3. If the modifier is close to fixation, the regime applies and the modifier is consequently maintained at intermediate frequency. Generally speaking, in the regime assortment is moderately strong and competition is comparatively weak. Heterozygotes are sufficiently common that sexual selection acts against their elimination we call this stabilizing sexual selection.
Stabilizing sexual selection outweighs disruptive selection resulting from competition. Hence, a shaped phenotype distribution is optimal and higher levels of assortment are disadvantageous. Competition can be weak , Figure 1E or moderate , Figure 1F in this scenario. Dominance increases the parameter region in which this scenario applies.
In particular, dominance hinders heterozygotes to exploit a niche in the middle of the phenotype range Figure 1F. Small assortment modifiers cannot go to fixation if the degree of dominance exceeds a critical value.
The reason is that disruptive selection is very weak for sufficiently strong dominance. If the strength of assortment increases, selection becomes stabilizing. If , the regime applies for a sufficiently frequent modifier. In both cases, a modifier will spread while rare, but cannot go to fixation. As discussed in Section 3. Assortment reduces the frequency of heterozygotes at the ecological locus. Hence, it increases the viability of individuals in the middle of the phenotypic range. Moreover, assortative mating induces sexual selection, which can be stabilizing or disruptive, depending on the strength of assortment.
Finally, if assortment is very strong, selection at the modifier locus will be very inefficient because the frequency of heterozygotes at the ecological locus is strongly reduced. For a fixed set of parameters, different regimes can apply at different points in time, especially if modifier effects are large. This may result in multiple stable equilibria. An initially rare modifier with large effect can become fixed only if sufficiently strong disruptive sexual selection is established during its sweep.
Figure 4 illustrates the evolutionary outcome of modifiers with different effect sizes. Note that the effect size does not affect the region in which an initially rare modifier is lost. The reasons for loss of modifiers are the same as in the case of small effects.
In contrast, the fixation region depends in a nonlinear and complicated way on the modifier effect and the initial frequency of the modifier. First, consider a modifier with effect Figure 4A. Again, multiple stable equilibria were not detected. The fixation region collapses to a narrow region in the parameter space and. In this region, the regime applies for a rare modifier, whereas the regime applies for a frequent modifier.
In the regime, assortment is sufficiently strong compared with competition to reduce the overall fitness of heterozygotes at the ecological locus to an extent that overall disruptive selection is established. This can be accomplished if competition is weak , see Figure 1B or moderate , Figure 1C.
Then, because their frequency is relatively low, heterozygous males pay higher costs for being rare. Consequently, an increase in assortative mating is favored. However, selection at the modifier locus is weak because of the low frequency of heterozygotes at the ecological locus.
In addition, dominance decreases the difference between phenotypic values of heterozygotes and homozygotes. Therefore, selection for assortment can be very weak in this scenario. If and , heterozygotes at the ecological locus are less fit than homozygotes for a sufficiently rare modifier. If the modifier increases in frequency the regime applies since competition in the middle of the phenotype range is reduced because of dominance and assortment.
Consequently, the modifier cannot become fixed. If , competition is strong enough to establish a niche in the middle of the phenotypic range during the spread of a modifier, i. Figure 2. If modifiers have large effect , disruptive sexual selection is strong for frequent modifiers.
Therefore, initially frequent modifiers go to fixation in a wide parameter range. These parameter ranges are hatched in Figures 4B—D. However, fixation was only observed if the modifier is initially at very high frequency, i. Since we are primarily interested in the build up of reproductive isolation, we restrict attention to the standard and the rare-modifier scenario for the rest of the section.
If the modifier effect is moderately strong ; Figure 4B , the fixation region increases compared to the case. In particular, a broader range of values for permits fixation of the modifier. The reason is that the for small , and for moderately large regimes are less likely to occur during the spread of modifiers with sufficiently large effect. The range for in which modifiers become fixed also increases compared to the case.
Weak disruptive selection is sufficient for invasion. This occurs if is large. If a modifier increases in frequency, strong disruptive sexual selection will be established and the modifier will go to fixation. Interestingly, intermediate dominance is most favorable for fixation of a modifier. If the level of assortment increases in a part of the population, a niche in the middle of the phenotypic range may be established. Dominance impedes heterozygotes to exploit such a niche cf.
Figure 1C. This means that the regime can be easier established if dominance is moderately strong. If dominance is strong, the mating success of homozygotes and heterozygotes at the ecological locus is almost identical. If competition is sufficiently strong, the regime applies as the modifier rises in frequency. Consequently, an initially rare modifier does not become fixed if the degree of dominance is high and competition is at least moderately strong.
This explains why intermediate dominance maximizes the size of the fixation region. Next, we consider modifiers that lead to almost complete reproductive isolation if fixed. Figures 4C and D illustrate the fate of modifiers with effects and , respectively. Quite surprisingly, the positive effect of dominance on the fixation of modifiers is most pronounced if modifiers have large effects. Strong assortment, which is quickly established if modifiers have large effect, leads to extremely strong disruptive sexual selection.
If , dominance is necessary for fixation of the modifier. In the absence of dominance and if , the reduced mating success of heterozygotes is compensated by the emergence of a niche in the middle of the phenotype range as the modifier becomes sufficiently frequent Figure 5A. Consequently, an initially rare modifier will not spread to fixation.
The presence of dominance does not change the strength of sexual selection unless it is sufficiently strong Figure 5B. Dominance has almost no effect on the strength of disruptive sexual selection as long as the phenotypic value of heterozygotes at the ecological locus stays in this valley. In contrast, if dominance increases, the viability of heterozygotes decreases strongly Figure 5B.
This explains why the optimal degree of dominance increases with increasing modifier effect. Figure 6 illustrates the evolutionary outcome of modifiers with small effect for various initial degrees of assortment. Multiple stable equilibria were not detected. Even a small amount of initial assortment leads to a substantial change of the region in which modifiers are maintained.
The maintenance region shrinks with increasing initial assortment and approaches its minimum at see Figure 6A. If assortment is weak , sexual selection is stabilizing. Thus, the region decreases with increasing assortment.
If competition is weak , stabilizing sexual selection outweighs disruptive selection at the ecological locus and the regime applies. Furthermore, dominance decreases the effect of competition. Therefore, if competition is weak the region is established for weaker assortment. For strong competition , a niche in the middle of the phenotype spectrum can be established if the frequency of heterozygotes is reduced.
Thus, the region is replaced by the region if initial assortment increases. The fixation and maintenance regions increase with increasing initial degree of assortment if.
Then, disruptive sexual selection can be established as long as. Dominance slightly decreases the region in which a modifier is maintained or goes to fixation. However, the effect of dominance is less pronounced compared with the case of weak initial assortment. For moderately strong initial assortment, evolution can be very slow such that slow runs are observed.
For strong assortment only slow runs are observed data not shown. This is consistent with our results about invasion fitness. We conclude that establishment of high levels of assortment via a series of invasion and fixation of modifiers with small effect seems unlikely. The build-up of reproductive isolation via allele substitutions of initially rare modifiers with small effects faces several problems.
Positive frequency-dependence due to an intermediate level of assortment can lead to overall stabilizing selection because it outweighs disruptive selection resulting from competition. On the other hand, for weak or moderate assortment, and sufficiently strong competition, a niche in the middle of the phenotype range appears if heterozygotes become sufficiently rare.
Finally, for high levels of assortment, a severely reduced frequency of heterozygotes can neutralize selection at the modifier locus. Our approach allows us to construct sequences of invasion and fixation of modifiers with different effects. If we consider only initially rare modifiers with small effect, we obtain an estimate for the degree of assortment that can evolve by small steps. Figure 7A shows that only low levels of assortment can evolve, except for a small region of moderate competition and very weak dominance.
Furthermore, assortment does not evolve above a moderate level. Evolutionary stable degrees of assortment that can evolve via allele substitutions of initially rare modifiers if modifiers have small positive effect A , or various positive or negative effects B. The numbers in the differently shaded regions indicate the maximum degree of assortment that can evolve starting from random mating.
Measurement of the human allele frequency spectrum demonstrates greater genetic drift in East Asians than in Europeans. Purcell, S. Am J Hum Genet , 81 Robinson, M. Genetic evidence of assortative mating in humans. Larkin, E. Clinical and translational science 3 , — Tishler, P. Incidence of sleep-disordered breathing in an urban adult population: The relative importance of risk factors in the development of sleep-disordered breathing.
Musunuru, K. Candidate gene association resource CARe : design, methods, and proof of concept. Combined admixture mapping and association analysis identifies a novel blood pressure genetic locus on 5p contributions from the CARe consortium.
The American Journal of Human Genetics 88 , 76—82 Thornton, T. Estimating kinship in admixed populations. Curie-Cohen, M. Estimates of inbreeding in a natural population: a comparison of sampling properties. Download references. You can also search for this author in PubMed Google Scholar. Correspondence to Xiaofeng Zhu. Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Reprints and Permissions. Li, X. Height associated variants demonstrate assortative mating in human populations. Sci Rep 7, Download citation. Received : 02 May Accepted : 03 November Published : 16 November Anyone you share the following link with will be able to read this content:.
Sorry, a shareable link is not currently available for this article. Provided by the Springer Nature SharedIt content-sharing initiative. Nature Genetics Nature Human Behaviour By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate. Advanced search. Skip to main content Thank you for visiting nature. Download PDF. Subjects Biomarkers Population genetics.
Abstract Understanding human mating patterns, which can affect population genetic structure, is important for correctly modeling populations and performing genetic association studies. Introduction Human mate choice is relevant to a wide range of scientific disciplines, including biology, sociology, population genetics, evolutionary biology, and psychology 1 , 2 , 3 , 4 , 5. Full size table. Table 2 Comparison of inbreeding coefficient estimated from height associated variants with randomly sampled frequency matched variants: single locus analysis.
Figure 1. Full size image. Figure 2. Figure 3. Table 3 Comparison of inbreeding coefficient estimated from height associated variants and lipids associated variants with randomly sampled frequency matched variants: multiple loci analysis. Figure 4. Discussion In this study, we examined assortative mating for height, using both phenotype and genotype data. Quality Controls All data quality controls QCs were performed for each cohort separately, and only autosomal loci were used.
CARe The authors wish to acknowledge the support of the National Heart, Lung, and Blood Institute and the contributions of the research institutions, study investigators, field staff and study participants in creating this resource for biomedical research.
References 1. Article PubMed Google Scholar 2. Article PubMed Google Scholar 3. PubMed Google Scholar Article Google Scholar Article PubMed Google Scholar Google Scholar View author publications. Ethics declarations Competing Interests The authors declare that they have no competing interests. Additional information Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Electronic supplementary material. Supplementary Figures. About this article. Cite this article Li, X. Copy to clipboard. Further reading Maternal and fetal genetic effects on birth weight and their relevance to cardio-metabolic risk factors Nicole M. Warrington , Robin N. Beaumont , Momoko Horikoshi , Felix R. Robertson , N. Schnurr , Mohammad H. Zafarmand , Jonathan P. Bradfield , Niels Grarup , Marjolein N. Wang , Christian T. Joshi , Jodie N. Richmond , Ana Espinosa , Sheila J.
Barton , Hazel M. Inskip , John W. Lunetta , Joanne M. Murabito , Caroline L. Muglia , Jani Heikkinen , Camilla S. Morgen , Antoine H. Ring , Amanda J. Bennett , Kyle J. Appel , Cilius E. Fonvig , Caecilie Trier , Catharina E. Hougaard , Josep M. Mercader , Allan Linneberg , Katharina E. Schraut , Penelope A. Lind , Sarah E. Medland , Beverley M. Shields , Bridget A. Vinding , Sara M.
Willems , Mustafa Atalay , Bo L. Tuke , Hanieh Yaghootkar , Katherine S. Ruth , Samuel E. Lakka , Cornelia M. QTL associated with the proportion of time males court H. Dashed line represents LOD significance threshold i. Dashes along the x-axis indicate position of genetic markers SNPs. For each simulated effect size, the distribution of all simulated effects blue and those which would be significant in our analysis i. Blue points represent f d values for kb windows. We thank the Smithsonian Tropical Research Institute for support and the Ministerio del Ambiente for permission to collect butterflies in Panama.
We are also grateful to Edinburgh Genomics for sequencing support. Abstract The evolution of new species is made easier when traits under divergent ecological selection are also mating cues.
Author summary Many closely related animal species remain separate not because they fail to produce viable offspring but because they do not mate in the first place.
Introduction During ecological speciation, reproductive isolation evolves as a result of divergent natural selection [ 1 ].
Download: PPT. Fig 1. Divergence in warning pattern cue and corresponding preference in sympatric Heliconius butterflies. Results We studied male mating preference among first-generation hybrid F1 and backcross hybrid families between H. Three loci contribute to species differences in preference behavior As reported previously [ 25 ], F1 males have a strong preference for the red H. Fig 3. Genetic and physical positions of behavioral QTL and the warning pattern loci and localized levels of admixture f d.
Table 1. Individual and combined QTLs for differences in relative courtship time. Admixture is reduced at the preference—color pattern locus on Chromosome 18 To consider the effects of major color pattern cue and preference loci on localized gene flow across the genome, we used the summary statistic f d to quantify admixture between H.
Different preference QTLs affect different aspects of behavior The male preference QTLs we have identified may influence differences in male attraction toward red H. Discussion Here, we reveal a genetic architecture that will strengthen genetic associations i. Butterfly collection, rearing, and crossing design All butterfly rearing, genetic crosses, and behavioral experiments were conducted at the Smithsonian Tropical Research Institute in Panama between August and August Behavioral assays We measured male attraction to H.
Genotyping and linkage map construction Genotyping and construction of linkage maps has been described elsewhere [ 39 ]. Data analysis All QTL analyses were performed on backcross-to- cydno hybrid males in which the preference behaviors segregate.
Simulations We used simulations to estimate potential inflation of measured effect sizes due to the Beavis effect. Admixture analysis We investigated heterogeneity in admixture across the genome between H. Dryad Underlying data can be found in the online Dryad repository: doi. Supporting information. S1 Fig. S2 Fig. Simulations suggest QTL effect sizes are not greatly overestimated. S3 Fig. Localized levels of admixture f d across all 21 chromosomes. S4 Fig.
Proportion of trial time in which backcross-to- cydno males courted females of either type. S5 Fig. Decline in LD of linked and unlinked loci under an assumption of random mating. S1 Table. Raw behavioral data. S2 Table. References 1. Nosil P. Ecological Speciation. Oxford University Press; Gavrilets S.
Fitness Landscapes and the Origin of Species. Princeton University Press; A framework for comparing processes of speciation in the presence of gene flow. Mol Ecol. Coupling, Reinforcement, and Speciation. Am Nat. Trends Ecol Evol Amst.
Thibert-Plante X, Gavrilets S. Evolution of mate choice and the so-called magic traits in ecological speciation. Ecol Lett. The genetics of adaptive shape shift in stickleback: pleiotropy and effect size.
Genetics of ecological divergence during speciation. A single gene affects both ecological divergence and mate choice in Drosophila. Diversification of complex butterfly wing patterns by repeated regulatory evolution of a Wnt ligand.
The gene cortex controls mimicry and crypsis in butterflies and moths. Curr Biol. Lewontin RC, Kojima K. The Evolutionary Dynamics of Complex Polymorphisms.
View Article Google Scholar Ecological explanations for incomplete speciation. Trends Ecol Evol. Annual Review of Ecology, Evolution, and Systematics. Felsenstein J. Skepticism towards Santa Rosalia, or why are there so few kinds of animals? Estrada C, Jiggins CD. Patterns of pollen feeding and habitat preference among Heliconius species. Ecological Entomology.
Reproductive isolation caused by colour pattern mimicry. Disruptive sexual selection against hybrids contributes to speciation between Heliconius cydno and Heliconius melpomene. Proc Biol Sci. Speciation by hybridization in Heliconius butterflies. Disruptive ecological selection on a mating cue. Pervasive genetic associations between traits causing reproductive isolation in Heliconius butterflies.
The diversification of Heliconius butterflies: what have we learned in years? J Evol Biol. Southcott L, Kronforst MR. Female mate choice is a reproductive isolating barrier in Heliconius butterflies. Jiggins CD. Ecological Speciation in Mimetic Butterflies. Sinauer; Brain composition in Heliconius butterflies, posteclosion growth and experience-dependent neuropil plasticity. J Comp Neurol. Mimicry: developmental genes that contribute to speciation.
Evol Dev. PLoS Biol. Complex modular architecture around a simple toolkit of wing pattern genes. Nat Ecol Evol.
0コメント