Repeated signatures of balancing selection in small and large populations of guppies ( Poecilia reticulata )

Repeated signatures of balancing selection in small and large populations of guppies (Poecilia reticulata) | bioRxiv /* */ /* */ <!-- <!-- /*! * yepnope1.5.4 * (c) WTFPL, GPLv2 */ (function(a,b,c){function d(a){return"[object Function]"==o.call(a)}function e(a){return"string"==typeof a}function f(){}function g(a){return!a||"loaded"==a||"complete"==a||"uninitialized"==a}function h(){var a=p.shift();q=1,a?a.t?m(function(){("c"==a.t?B.injectCss:B.injectJs)(a.s,0,a.a,a.x,a.e,1)},0):(a(),h()):q=0}function i(a,c,d,e,f,i,j){function k(b){if(!o&&g(l.readyState)&&(u.r=o=1,!q&&h(),l.onload=l.onreadystatechange=null,b)){"img"!=a&&m(function(){t.removeChild(l)},50);for(var d in y[c])y[c].hasOwnProperty(d)&&y[c][d].onload()}}var j=j||B.errorTimeout,l=b.createElement(a),o=0,r=0,u={t:d,s:c,e:f,a:i,x:j};1===y[c]&&(r=1,y[c]=[]),"object"==a?l.data=c:(l.src=c,l.type=a),l.width=l.height="0",l.onerror=l.onload=l.onreadystatechange=function(){k.call(this,r)},p.splice(e,0,u),"img"!=a&&(r||2===y[c]?(t.insertBefore(l,s?null:n),m(k,j)):y[c].push(l))}function j(a,b,c,d,f){return q=0,b=b||"j",e(a)?i("c"==b?v:u,a,b,this.i++,c,d,f):(p.splice(this.i++,0,a),1==p.length&&h()),this}function k(){var a=B;return a.loader={load:j,i:0},a}var l=b.documentElement,m=a.setTimeout,n=b.getElementsByTagName("script")[0],o={}.toString,p=[],q=0,r="MozAppearance"in l.style,s=r&&!!b.createRange().compareNode,t=s?l:n.parentNode,l=a.opera&&"[object Opera]"==o.call(a.opera),l=!!b.attachEvent&&!l,u=r?"object":l?"script":"img",v=l?"script":u,w=Array.isArray||function(a){return"[object Array]"==o.call(a)},x=[],y={},z={timeout:function(a,b){return b.length&&(a.timeout=b[0]),a}},A,B;B=function(a){function b(a){var a=a.split("!"),b=x.length,c=a.pop(),d=a.length,c={url:c,origUrl:c,prefixes:a},e,f,g;for(f=0;f<d;f++)g=a[f].split("="),(e=z[g.shift()])&&(c=e(c,g));for(f=0;f<b;f++)c=x[f](c);return c}function g(a,e,f,g,h){var i=b(a),j=i.autoCallback;i.url.split(".").pop().split("?").shift(),i.bypass||(e&&(e=d(e)?e:e[a]||e[g]||e[a.split("/").pop().split("?")[0]]),i.instead?i.instead(a,e,f,g,h):(y[i.url]?i.noexec=!0:y[i.url]=1,f.load(i.url,i.forceCSS||!i.forceJS&&"css"==i.url.split(".").pop().split("?").shift()?"c":c,i.noexec,i.attrs,i.timeout),(d(e)||d(j))&&f.load(function(){k(),e&&e(i.origUrl,h,g),j&&j(i.origUrl,h,g),y[i.url]=2})))}function h(a,b){function c(a,c){if(a){if(e(a))c||(j=function(){var a=[].slice.call(arguments);k.apply(this,a),l()}),g(a,j,b,0,h);else if(Object(a)===a)for(n in m=function(){var b=0,c;for(c in a)a.hasOwnProperty(c)&&b++;return b}(),a)a.hasOwnProperty(n)&&(!c&&!--m&&(d(j)?j=function(){var a=[].slice.call(arguments);k.apply(this,a),l()}:j[n]=function(a){return function(){var b=[].slice.call(arguments);a&&a.apply(this,b),l()}}(k[n])),g(a[n],j,b,n,h))}else!c&&l()}var h=!!a.test,i=a.load||a.both,j=a.callback||f,k=j,l=a.complete||f,m,n;c(h?a.yep:a.nope,!!i),i&&c(i)}var i,j,l=this.yepnope.loader;if(e(a))g(a,0,l,0);else if(w(a))for(i=0;i (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0];var j=d.createElement(s);var dl=l!='dataLayer'?'&l='+l:'';j.src='//www.googletagmanager.com/gtm.js?id='+i+dl;j.type='text/javascript';j.async=true;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-M677548'); Skip to main content Home About Submit ALERTS / RSS Search for this keyword Advanced Search New Results Repeated signatures of balancing selection in small and large populations of guppies ( Poecilia reticulata ) View ORCID Profile Josephine R Paris , View ORCID Profile James R Whiting , View ORCID Profile Joan Ferrer Obiol , View ORCID Profile Kimberly A Hughes , View ORCID Profile Bonnie A Fraser doi: https://doi.org/10.1101/2025.06.12.659363 Josephine R Paris 1 Department of Biosciences, University of Exeter , Exeter, UK Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Josephine R Paris For correspondence: parisjosephine{at}gmail.com James R Whiting 1 Department of Biosciences, University of Exeter , Exeter, UK Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for James R Whiting Joan Ferrer Obiol 2 Department of Biology, Colorado State University , Fort Collins, CO, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Joan Ferrer Obiol Kimberly A Hughes 3 Department of Biological Science, Florida State University , Tallahassee, FL, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Kimberly A Hughes Bonnie A Fraser 1 Department of Biosciences, University of Exeter , Exeter, UK Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Bonnie A Fraser Abstract Full Text Info/History Metrics Supplementary material Preview PDF Abstract Balancing selection is a powerful evolutionary force that maintains adaptive genetic and phenotypic diversity. Although methods to detect the footprints of balancing selection in genomic data have advanced, we still lack a clear understanding of how repeatable these signatures appear in wild populations, and how this repeatability is shaped by demographic history and existing genetic variation. The Trinidadian guppy ( Poecilia reticulata ) provides an ideal model to test the repeatability of balancing selection in the wild as there is strong evidence that negative frequency-dependent selection (NFDS) maintains colour polymorphism. Analysing whole-genome sequencing data from 11 guppy populations ( n = 195) with contrasting demographic contexts, we apply scans of balancing selection to explore which genomic regions show evidence of repeatability. We find that populations with small N e show less genetic repeatability but still exhibit population-specific regions of elevated diversity, implicating independent balancing selection or other evolutionary mechanisms. Despite this, we identify 23 regions with repeated signatures of balancing selection, including a region on LG22 containing genes involved in colour, vision, mate choice, and social behaviour. Investigating the repeatability of balancing selection in small and large populations improves our knowledge of how demographic factors interact with selective processes to shape natural variation. 1. Introduction The persistence of high polymorphism in ecologically important traits poses compelling questions in evolutionary biology. Among the mechanisms capable of maintaining balanced polymorphisms, negative frequency-dependent selection (NFDS) - whereby the fitness of a given allele increases as it becomes rarer in the population - is theoretically well-supported ( 1 – 4 ) and empirical examples of its prevalence continue to grow (see this issue). Recent advances in statistical approaches for identifying intermediate-frequency variants have facilitated the development of genome-wide scans for balancing selection ( 5 , 6 ). However, little is known about our ability to detect the signatures of balancing selection in natural populations that differ in demographic history and genetic variation. Such investigations would enhance our understanding of both the strength and constraints of NFDS, as well as its interplay with demography in shaping natural variation. The effectiveness of balancing selection in maintaining polymorphism is dependent on population size ( N e ) ( 7 , 8 ). When N e is high, even weak balancing selection can maintain allelic diversity over extended periods of time ( N e s≫1). In contrast, in small or recently bottlenecked populations, drift is expected to dominate ( N e s≪1), making it difficult for selection to preserve intermediate-frequency alleles. Despite these expectations, a number of empirical studies have identified balanced polymorphisms when N e is severely reduced, particularly at immunity-related loci (e.g. ( 9 – 13 ), suggesting that strong balancing selection can, in some cases, overcome the effects of drift ( 14 ). However, intermediate allele frequencies can also be produced by incomplete selective sweeps and recent population bottlenecks ( 6 , 15 – 17 ), making it difficult to distinguish true cases of balancing selection at individual loci. Population structure and admixture have also been shown to mimic the signatures of balancing selection, leading to false positives ( 18 , 19 ). One promising way to overcome these limitations is to search for repeated signatures of balancing selection across independent populations. The repeated maintenance of polymorphisms provides stronger evidence for selection than patterns observed within any single population, as drift and neutral mutations are unlikely to affect the same regions of the genome ( 20 ). This is especially true in populations with small N e , where the influence of drift is expected to be strong; in such cases, examining multiple populations can help distinguish true signals of selection from the background noise of neutral demographic processes. Yet despite its potential, often only single populations or species are examined for balancing selection (but see ( 21 , 22 ). Identifying genomic signatures of balancing selection across both small and large populations is therefore key to understanding how demographic factors interact with selective processes to shape natural variation. Natural populations of the Trinidadian guppy ( Poecilia reticulata ) provide an ideal system to investigate the repeatability of genomic signatures of balancing selection. At the phenotypic level there is strong, repeatable evidence for NFDS. Male guppies have highly variable and heritable colour patterns, with levels of genetic variance that exceed expectations under recurrent mutation and purifying selection alone, implicating balancing selection ( 23 ). NFDS, specifically, has been demonstrated for colour pattern frequencies in natural populations. In a series of independent experiments, males with colour patterns that were manipulated to be locally rare had a significant reproductive advantage ( 24 – 26 ) and significant survival advantage compared to common patterns ( 27 , 28 ). It is likely that female mate choice is the primary driver of maintaining colour polymorphism; indeed, female preference for rare colour patterns has been shown to be widespread across multiple populations ( 29 , 30 ). In addition to strong evidence for NFDS, population structure and demography is well documented in the Northern Trinidad guppy system ( 31 – 34 ). Here, the northern mountain range isolates populations among rivers and waterfalls, resulting in upstream populations founded by small numbers, likely from their downstream counterparts ( 35 , 36 ). Finally, experimentally introduced guppy populations, where downstream ‘high-predation’ guppies are transplanted to upstream guppy, and predator-free, locations ( 37 – 40 ), gives a unique opportunity to examine populations with known ages and sizes of founding bottleneck events ( 41 – 43 ). Taken together, the guppy system offers a powerful comparative framework, integrating natural populations and experimental introductions, with contrasting demographic histories and well-characterised ecological drivers and phenotypic responses to NFDS. Here, we perform genome-wide scans across 11 guppy populations to test for the repeatability of balancing selection. We selected populations with empirically documented ecological drivers of NFDS - specifically, female preference for rare male phenotypes ( 29 ), which also span a range of contrasting demographic histories ( 32 , 34 ). First, we quantify population structure, demography, divergence, and genetic diversity to categorise populations into small and large demographic groups. We then examine regions of elevated genetic diversity found exclusively in small populations and assess their repeatability using change-point detection methods. Next, we apply newly developed model-based approaches ( 44 – 47 ) to scan for balancing selection, to identify whether, and where, we find repeatable signatures across populations. Finally, we examine candidate loci that are consistently found to be under balancing selection to assess their potential links to phenotypic traits targeted by NFDS. 2. Methods (a) Study sites and sample collection Guppy samples (n = 195) were collected from 11 sites (across three drainages) in the Northern mountain ranges of Trinidad ( figure 1a ), ranging from 10 to 20 samples per site (table S1). For the Guanapo (Caroni drainage), Aripo (Caroni drainage) and Marianne (Northern drainage) rivers, guppies were sampled both downstream and upstream of waterfalls. Downstream (D) populations are typified by high-predation pressure ( 48 ) and are considered to be the source of naturally-colonised upstream populations. Naturally-colonised upstream populations (U) are hypothesised to have been founded by a small number of individuals, and are generally typified by lower genetic diversity and low predation pressure ( i.e ., “low-predation” populations) ( 32 , 33 ). We also sampled experimentally-introduced populations from the El Cedro (Caroni drainage) and Turure (Oropouche drainage) rivers and their downstream counterparts. In 1981, ~100 guppies were introduced from a downstream El Cedro site into a guppy-predator-free upstream site in the same tributary ( 37 , 38 ). In 1956, ~200 guppies were introduced from a downstream Guanapo site into an upstream site in the Turure river ( 40 ), where they subsequently naturally colonised a previously occupied downstream Turure site ( 49 , 50 ). All samples were stored in either 95% EtOH at −20° C or RNeasy at 4° C prior to DNA extraction. Download figure Open in new tab Figure 1. Overview of the guppy study system and analysis of population structure and divergence. ( a ) Location of the 11 study sites in the Northern Mountain ranges of Trinidad. Inset shows the location (black box) of the main map of the island of Trinidad. Populations are colour-coded by river: Guanapo (purple); El Cedro (orange); Aripo (green); Turure (blue); Marianne (red); Paria (pink). Downstream (D) populations are denoted with a circle. Upstream (U) populations are denoted by a triangle. ( b ) Principal Components Analysis (PCA) showing PC1 (~20%) and PC2 (~14%). ( c ) PCA showing PC1 (~20%) and PC3 (~8%). ( d ) Heatmap of pairwise F ST calculated between all 11 populations, where purple represents lower F ST and yellow represents higher F ST. ( e ) Pairwise co-ancestry matrix between individuals, with light yellow indicating low co-ancestry, and darker yellows, oranges and reds to pink indicating progressively higher co-ancestry. ( f ) Dated divergence tree highlighting the relationships and divergence between populations. Guppy picture © by Erica Robertson. (b) DNA extraction, whole-genome sequencing, and variant calling Total genomic DNA was extracted using the Qiagen DNeasy Blood and Tissue kit (QIAGEN, Germany), including an on-column RNAse step. Low Input Transposase Enabled (LITE) DNA libraries were prepared at The Earlham Institute, UK. Libraries were sequenced on an Illumina HiSeq 4000 (150bp paired-end) with a target sequencing coverage of 10x per sample. Sequencing statistics can be found in table S2. Variant calling followed previous approaches ( 33 , 41 ). Briefly, reads were cleaned using cutadapt ( 51 ), aligned to the guppy genome (GCA_904066995.1; ( 52 ) using bwa mem v0.7.17 ( 53 ), and sorted and indexed using samtools v1.17 ( 54 ). Quality control and coverage of samples were checked using samtools stat, flagstat and idxstats . Variant calling was performed using GATK4 v4.1.8.1 ( 55 ). Variants were recalibrated using a truth-set generated from high-coverage, PCR-free sequencing data ( 52 ). Variants were filtered using quality control filters (QD 60.0 | MQ 13.0), retaining only biallelic SNPs, at a minimum depth of 4 and a maximum depth of 200. Population-specific VCF files were created and filtered at either a maximum missing value of 50% or 80% using vcftools ( 56 ) and then re-combined. For analyses requiring phased variants, phasing was performed using Beagle v2 ( 57 ), followed by shapeit v2 ( 58 ). To incorporate recently documented genome rearrangements (on LG12 and LG20) ( 33 , 59 , 60 ), output file coordinates were lifted over using custom scripts. Further details on the specific variant filtering performed for each analysis can be found in table S3. For divergence time estimates and for polarising ancestral alleles, we also made use of data for two other species: Poecilia wingei (sister) and Poecilia picta (outgroup). P. wingei WGS data (n=6) were accessed via the European Nucleotide Archive (PRJEB26489) ( 61 ). For P. picta , we performed WGS on ten laboratory-reared individuals (5 males and 5 females). Data were sequenced on an Illumina HiSeq 2500 (150bp paired-end; table S2). (c) Population structure, genetic diversity, and demographic history To assess population structure and shared ancestry, we performed a Principal Components Analysis (PCA) in Plink v.19 ( 62 ), estimated mean pairwise F ST in PopGenome v2.75 ( 63 ), and used fineStructure v4.1.1 ( 64 ) to build a co-ancestry matrix based on haplotype relationships. For divergence dating, we used a multispecies coalescent (MSC) approach implemented in SNAPPER ( 65 ). We followed the recommendations of Stange et al. ( 66 ) by constraining the root of the tree (divergence between P. reticulata and P. wingei ) to follow a normal distribution with a mean of 3.41 mya and a standard deviation of 0.329 ( 33 ) (supplementary methods S1). Relative cross-coalescence rates between upstream-downstream pairs and contemporary N e for each population were estimated using ancestral recombination graphs (ARGs) in Relate v1.1.6 ( 67 , 68 ). Analysis was performed using a starting effective population size of 1×10 4 , a mutation rate of 2.90 × 10 −9 per site per generation ( 69 ), and smoothed recombination maps for each chromosome ( 70 ). The ancestral allele was assigned as the one more common in the sister ( P. wingei ) and outgroup ( P. picta ). To convert the ARGs into estimates of population size, we used the script EstimatePopulationSize.sh using a generation time of 7.9 months (0.66 years) ( 71 ). Nucleotide diversity (π) and Tajima’s D were calculated in 50 Kb non-overlapping windows in PopGenome v2.75 ( 63 ). Observed heterozygosity ( H o ) was estimated per site per population in vcftools v0.1.16 ( 56 ) using the --hardy parameter and was summarised as a mean in 50 Kb windows. Change point detection (CPD) was performed on the Tajima’s D estimates per population using the changepoint package ( 72 ) in R, using the cpt.mean function with the BinSeg method, an SIC penalty and a Q of 100. Repeated CPD regions with elevated Tajima’s D were investigated further using lostruct ( 73 ) by performing local PCA in windows of 100 SNPs, and by calculating LD using the –r2 square option in Plink, plotting the results with LDheatmap ( 74 ). (d) Scanning the genome for balancing selection Scans for balancing selection were conducted using BalLeRMix+ ( 44 , 45 ), which implements a series of composite-likelihood ratio (CLR) statistics (known as “ B statistics”) by modelling local deviations in the allele frequency spectrum as it is locally perturbed by selection. The CLR test includes the flanking region surrounding the site of interest, avoiding issues of window size selection. We chose BalLeRMix+ over other scans of balancing selection because it has been shown to be robust under different demographic scenarios (population expansions and bottlenecks), different strengths and ages of selection, and window-size selection and recombination rate biases ( 44 ). Amongst the five B statistics, the B 2 statistics, which model the allele frequency spectrum whilst incorporating substitution information, have been shown to be the most powerful in detecting balancing selection ( 44 ). We used BalLeRMix+ to calculate B 2 statistics using both the derived allele frequencies (DAF; supplementary methods S2) and the minor allele frequencies (MAF), inferred for each population. Prior to calculating allele frequencies, the VCF file was filtered to account for potential artefacts which might lead to false positives, including mappability, repeat-rich regions, and high/low coverage regions (supplementary methods S3). We compared the B 2 DAF and B 2 MAF statistics due to some uncertainty in our identification of the derived allele in some of the populations (figure S1 & figure S2). BalLeRMix+ was run using a constant recombination rate of 2.12e-06, estimated from smoothed linkage maps ( 70 ). While this constant recombination rate is likely not reflective of the entire genome, BalLeRMix+ is robust to errors in recombination rate ( 44 ). To be able to compare the BalLeRMix+ estimates with our other estimates of population genetic statistics, we summarised the B 2 statistic in 50 Kb windows, taking the largest CLR (which is appropriate given that the test accounts for flanking regions) and corresponding log 10 ŝ score. We excluded windows that had fewer than five variable sites. To ensure robustness of our inferences from BalLeRMix+, we also independently calculated the β statistic using BetaScan2 ( 46 , 47 ). Concordant results between the two approaches were used to increase confidence in candidate loci. (e) Identifying repeatable signatures of balancing selection To identify genomic regions with repeatable signals of balancing selection, we first examined the overlap of windows in the top 1% of CLR scores across populations, but found limited consistency (see Results). While informative, such an approach is conservative, as it only considers the most extreme outliers and moreover, depends on choosing an arbitrary cutoff for significance. To overcome these limitations, we used PicMin ( 75 ), a method that evaluates whether multiple populations show a coordinated shift toward extreme values. By leveraging order statistics, PicMin offers a robust and sensitive framework for detecting repeatable signals of selection without relying on predefined significance thresholds. We used 10 million replicates for the null distribution. Significant repeatable windows were examined if they occurred in ≥ 9 populations (config_est parameter) at a −log 10 (FDR) > 4. Candidate windows were then examined further by quantifying the CLR statistic and log 10 ŝ estimates (where ŝ > 0 indicates balancing selection and ŝ < 0 implies positive selection). To explore the genes in candidate regions, we performed a new annotation of the guppy genome using BRAKER3 ( 76 ) (supplementary methods S4). 3. Results (a) Genetic structure and divergence reveals strong drainage effects and traces of introduction history Assessment of population structure, co-ancestry, and divergence among the 195 guppies sampled from 11 sites across Northern Trinidad revealed strong genetic clustering that reflects the river drainages and the known introduction history of the experimental populations ( figure 1 ; figure S3). PC1 (19.6%) separated populations from the Northern drainage populations (MD, MU, PU) from those in the Caroni drainage, while PC2 (13.7%) distinguished the Aripo populations (AD, AU), and PC3 (8.3%) further separated MD from its upstream counterparts (MU, PU), which clustered more closely together ( figure 1b, 1c ). Pairwise F ST estimates among natural downstream-upstream pairs ranged from low (0.08 between AD and AU), modest (0.32 between GD and GU) to high (0.56 between MD and MU), indicating varying degrees of differentiation due to the time since upstream colonisation and subsequent gene flow ( figure 1d ). Experimentally-introduced populations reflected their sources to varying extents: EU was moderately diverged from ED (F ST = 0.10), while TU was genetically proximate to both its source (GD; F ST = 0.07) and its downstream counterpart (TD; F ST = 0.04), indicative of introduction history and post-introduction gene flow, respectively. Coancestry analyses showed a clear clustering by drainage, strong similarity between MU and PU, and a very low shared co-ancestry of GU with all other populations, even with its putative source (GD) ( figure 1e ). Divergence time estimates inferred deep splits between the Northern and Caroni drainages (~2.2 mya), and more recent divergences among introduced and source populations. However, the latter were older than known introduction times, most likely due to biases from the MSC model assumption of no post-divergence gene flow ( figure 1f ). However, relative cross-coalescence rates (RCCR) (which unlike divergence dating, incorporate gene flow) showed expected shallow splits between introduction pairs (e.g. ED–EU, TD–TU), more recent divergence between GD–GU, and deeper divergence for Northern drainage pairs (MD–MU, MD–PU) (figure S4). Together, these results point to strong drainage-driven structure and a mosaic of both shared and independent demographic histories across the 11 guppy populations. (b) Variation in Ne and genetic diversity across upstream and downstream guppy populations Genome-wide estimates of diversity demonstrated that guppies occupying downstream (D) sites, generally harbour more genetic variation compared to upstream (U) sites ( table 1 ). Four of the six upstream sites showed much lower genetic diversity (GU, MU, PU, EU), with two showing extremely reduced diversity (GU and MU; figure 2 ). The upstream populations, AU and TU, did not show as much reduced diversity, likely due to higher amounts of connectivity with their downstream counterparts (figure S4). Estimates of N e further highlighted the low diversity of these four upstream populations, in particular GU and MU ( table 1 ; figure S5). Of the experimentally-introduced sites, TU showed similar amounts of diversity and a similar N e relative to its source population (GD), but EU had a lower diversity and N e compared to its source (ED), and ED had the lowest diversity and N e compared to other downstream populations. Together, these analyses allowed us to classify the populations into small populations: GU, MU, PU and EU; and large populations: GD, AD, AU, TD, TU, MD, ED. Download figure Open in new tab Figure 2. Heterozygosity estimated across the 23 chromosomes of the guppy genome for the 11 studied populations. Plots represent H o estimated in 50 Kb windows across the 23 chromosomes of the guppy genome. Plots on the right summarise the density distributions. Downstream populations (denoted by circles) generally show higher levels of genetic diversity than upstream populations (denoted with triangles). Rivers are colour-coded: Guanapo (purple); El Cedro (orange); Aripo (green); Turure (blue); Marianne (red); Paria (pink). View this table: View inline View popup Table 1. Summary of effective population size ( N e ) and metrics of genetic diversity calculated for each of the 11 guppy populations. (c) Distinctive diversity patterns and shared Tajima’s D outliers in small populations We found that several small populations exhibited a characteristic “sawtooth” pattern of diversity, marked by distinct islands of elevated diversity ( figure 2 ). Furthermore, the distribution of Tajima’s D values revealed greater variance in small populations compared to large ones, characterised by a distinctly bimodal pattern. Distributions in small populations were skewed toward both highly negative values ( i.e. , an excess of rare alleles), and highly positive values ( i.e. , an excess of intermediate-frequency alleles; figure S6). To explore this further, we performed change point detection (CPD) on Tajima’s D (table S4), finding that the small populations GU, EU, and to a lesser extent, MU, have more and longer segments of consistently elevated Tajima’s D . Such patterns were not identified in any other populations. To investigate shared regions of elevated diversity in these three small populations, we examined potential overlaps in CPD regions. While numerous pairwise overlaps were detected when Tajima’s D = 2-3, only one region was shared across all three populations (LG18: 25.05 - 25.65 Mb). In the more extreme Tajima’s D CPD regions (>3), we identified a shared region between MU and GU on LG3 (7.65 - 7.80 Mb), and another between EU and MU on LG22 (12.25 - 13.35 Mb). Further investigation of these three regions showed that they did not share a common recombination landscape (inferred using local PCA; figure S7), but showed elevated LD in nearly all populations (figure S8). (d) Balancing selection signatures show limited repeatability across populations Using the B 2 statistic, we found signatures of selection across the 11 populations using both the derived allele frequency (DAF) spectra (table S5) and the minor allele frequency (MAF) spectra (table S6). Outlier windows (defined as ≥ 99% of the B 2 CLR distribution) differed among populations, with no obvious regions showing high repeatability (figure S9 & S10). The large majority of outlier windows identified as being under strong balancing selection using the MAF spectra were also identified as outliers using the DAF spectra (table S7). We therefore used the results obtained via modelling of the DAF spectra for the remainder of our analyses. The majority of outlier windows in each population showed evidence of balancing selection (ŝ > 0). Only the introduced large TU population showed a large fraction of windows showing positive selection (where ŝ is −3), and noticeably, none of the small populations (GU, EU, MU, PU) had any windows with a signal of positive selection (table S5). This is expected as B 2 models the global SFS to the local SFS and many of the small populations had a large amount of fixed variation ( figure 2 ). B 2 windows showing the strongest signals of balancing selection (≥ 99% and ŝ = 9) exhibited limited, but some overlap with candidate windows (≥ 99%) identified by BetaScan (table S8). These B 2 windows were mostly located in the right-hand tails of the Tajima’s D and π distributions (figure S11). However, repeatability in these windows across populations was largely inconsistent, with no outlier windows shared in eight or more populations, only one window found in seven populations, and just seven windows found in five populations (table S9). To further explore potential repeatability in signatures of balancing selection among our populations, we used PicMin, hypothesising that this would increase our power to detect candidate windows under balancing selection. Using PicMin, we identified 43 windows with a −log 10 (FDR) > 4 supported by ≥ 9 populations (N=16 windows in 9 populations; N=18 windows in 10 populations; N=9 windows in 11 populations; table S10). Of these, 23 windows (found across 13 chromosomes) had an ŝ>0 in all 11 populations and therefore show repeatability only for balancing selection ( figure 3 ). Download figure Open in new tab Figure 3. Heatmap showing the 23 candidate windows and their patterns of repeatability across the 11 guppy populations. Windows are numerically ordered by chromosome and window on the y-axis and the populations are ordered along the x-axis. Squares are coloured by their signature of balancing selection (log 10 ŝ) as indicated in the legend. White circles within each square denote significant windows ( i.e. , top 1% of B 2 scores) for any given population. (e) Balancing selection candidates We investigated the 23 candidate regions with the strongest evidence for repeated balancing selection as identified using PicMin, by integrating population-level B 2 scores (≥ 99%), β statistics (>90%), and functional gene annotations (table S11). Several regions emerged as particularly interesting due to their high statistical support and the presence of genes with functional annotations potentially related to NFDS in guppies (see Discussion). For example, a region on LG2 (~33.6 Mb) was among the top 1% of β scores in all 11 populations and among the top 1% of B 2 scores in seven populations. On LG10 (~16.5 Mb), two adjacent windows showed strong evidence of balancing selection, with the first window in the top 1% of both B 2 and β scores across seven populations. A region on LG16 (~34.6 Mb) was in the top 1% of B 2 scores in five populations and in the top 5% of β scores in six populations. On LG18 (~26.6 Mb), we identified a region containing seven genes which scored in the top 1% of B 2 scores in four populations and in the top 1% of β scores in seven populations. Finally, a window on LG22 (~13.10 Mb) containing five genes ( figure 4 ), was in the top 1% of B 2 scores in five populations and the top 5% of β scores in six populations. Notably, two of these regions, on LG18 and LG22, also overlapped with CPD segments showing elevated Tajima’s D in the small populations, further supporting their role as repeated targets of balancing selection. Download figure Open in new tab Figure 4 Evidence for repeated balancing selection across 11 guppy populations on LG22 (13.03 - 13.1 Mb). B 2 scores across the genomic region on chromosome 22 surrounding four genes, from left to right: iars2 ( isoleucyl-tRNA synthetase 2, mitochondrial ; in pink); nlgn3a ( neuroligin 3a , in brown); qrsl1 ( glutaminyl-tRNA amidotransferase subunit ; in green); rtn4ip1 ( reticulon 4 interacting protein 1 ; in blue); slc45a2 ( solute carrier family 45 member 2 ; in orange). Population B 2 scores are coloured by their signature of selection (log 10 ŝ) as indicated in the legend. We explored the candidate region on LG22 in more detail as it was among the top 1% of B 2 scores for five populations, including all four of the small populations, where log 10 ŝ was also high: GU ŝ = 9, EU ŝ = 9, MU ŝ = 9, PU ŝ = 6. The LG22 window also overlapped with the region identified in the Tajima’s D CPD analysis (LG22: 12.25 - 13.35 Mb) and contained genes relevant to the guppy NFDS phenotype (see Discussion; figure 4 ). With the exception of the introduced and small EU population, all populations showed elevated B₂ over the nlgn3a gene. Signals across other genes in the window varied by population: GD, ED, AD and TU maintained high B₂ scores across both qrsl1 and rtn4ip1 , while EU exhibited even higher B₂ over these genes compared to nlgn3a and other populations. Increased B₂ scores over slc45a2 were also observed in GD, ED, EU, AU, TD, TU and MD. Examination of per-SNP β scores within the region revealed a pronounced peak in all populations (in the top 1% in eight of them), again with the exception of EU (figure S12). LD patterns in the candidate region revealed strong and extensive LD in the small populations (GU, EU, MU, PU) consistent with reduced recombination or long-range haplotype maintenance (figure S13). In contrast, larger populations exhibited more fragmented LD, but a region of elevated LD was apparent surrounding the coordinates encompassing the gene nlgn3a . 4. Discussion Despite its potential to maintain stable polymorphisms, the genomic signatures of NFDS, and balancing selection more broadly, remain understudied in natural populations, limiting our understanding of the strength and constraints of NFDS in shaping natural variation. We analysed 11 guppy populations spanning a range of demographic contexts to investigate the repeatability of balancing selection. Populations exhibiting reduced diversity and low N e were found to harbour localised regions of elevated diversity, but the repeatability of these regions across populations was low. Using new statistical methods to scan the genome for balancing selection, we initially found limited overlap in the top 1% of outlier regions across all populations. However, by taking advantage of the repeatability of NFDS in this system, we applied an order-based statistical framework, identifying 23 genomic regions (across 13 chromosomes) with consistent signatures of selection. Further investigation of these regions revealed genes with putative links to the targets of NFDS in guppies, such as those involved in colour, vision, mate choice, and social behaviour. Together, these results suggest that, despite demographic heterogeneity, certain genomic regions are recurrently targeted by balancing selection, reflecting shared ecological drivers of NFDS, conserved adaptive functions, and evolutionary constraint on where in the genome adaptive variation can persist. Our results provide new insights into how balancing selection operates in small, bottlenecked populations, where the effects of drift are expected to overwhelm selection. Rather than relying solely on an upstream-downstream classification, we used multiple metrics of genetic diversity and effective population size ( N e ) to classify the 11 populations, revealing that not all upstream populations are equally small. For example, AU and TU were found to retain diversity levels more comparable to their downstream counterparts, likely due to higher gene flow ( 34 ). In contrast, the small populations (GU, MU, PU, and EU) exhibited a distinctive “sawtooth” pattern of diversity, resembling patterns observed previously in other small populations ( 11 , 77 ), attributed to both recent inbreeding and balancing selection. Indeed, the bimodal distribution of Tajima’s D , with both strongly negative and positive values, suggests that the observed diversity peaks reflect the influence of multiple, potentially interacting evolutionary processes operating in small populations. The guppy system offers a unique opportunity to compare these islands of elevated diversity with independent signals of balancing selection, and to assess their repeatability across populations, providing insights into the mechanisms that maintain genetic diversity in small populations. Recombination, and how it interacts with selection and drift, may be playing an important role in maintaining diversity. In small populations, genetic drift can generate LD which indirectly favours modifier alleles that increase recombination, thereby enhancing the probability that beneficial combinations of alleles arise and persist ( 78 ). Empirical evidence supports the idea that recombination rate variation can modulate the efficiency of selection in small populations ( 79 ), and inbreeding alone can cause the indirect selection of higher recombination rates ( 80 ). Moreover, regions with elevated recombination may retain more diversity even under strong drift, potentially helping to preserve intermediate-frequency alleles ( 81 ). In our small populations, elevated LD in candidate regions with consistently high Tajima’s D , along with the absence of shared reduced recombination patterns, suggest that recombination, and its interplay with drift and selection, may be a key, and often overlooked factor influencing how diversity is maintained in small populations. Whilst we detected regions of elevated Tajima’s D in the small populations, they exhibited the lowest repeatability in balancing selection signals. This raises an important question: is balancing selection genuinely rarer in small populations, or simply harder to detect? While drift undoubtedly reduces the efficacy of selection and may limit the persistence of polymorphisms, it is also possible that stochastic demographic effects, such as inbreeding, or expansions after a bottleneck, could obscure otherwise detectable signals. Despite our documented differences in demographic contexts among the 11 populations, we identified repeated signatures of selection at 23 regions, which included candidate genes related to the NFDS phenotype. These included two copies of a sacsin-like gene, best known for its role in neuronal function ( 82 ). Mutations in this gene cause retinal abnormalities in zebrafish ( 83 ), and it has also been associated with sociality in cichlids ( 84 ). We also detected signatures of balancing selection on a region containing repeated copies of protocadherin genes, which belong to a large family of neural cell-adhesion molecules implicated in synaptic connectivity and neuronal identity ( 85 ). This suggests a potential link between NFDS and variation in neural processing of colour cues, sensory perception, or behavioural responses. Notably, cadherin genes have been previously implicated in adaptation to predation regimes in guppies ( 33 , 41 ). Balancing selection was also evident in a region containing two adgrb1 (BAI1) genes which in knockout mice are associated with spatial memory deficits, altered synaptic plasticity ( 86 ), and differences in social discrimination ( 87 ). One region overlapping with elevated Tajima’s D on LG18 included several compelling genes: ednrba involved in pigmentation and expressed in melanocyte, iridophore, and xanthophore cells in zebrafish ( 88 ), two copies of nlrp12l, a key innate immune regulator ( 89 ); and dclk2 , involved in dendritic and retinal development ( 90 , 91 ). In the LG22 candidate region, we found particularly strong evidence for balancing selection at the nlgn3a ( neuroligin-3 ) gene, a synaptic adhesion molecule linked to social novelty responses in mice ( 92 – 94 ). Moreover, expression of this gene has been specifically linked to mate preference in Poeciliids ( 95 , 96 ), including guppies ( 97 , 98 ). Downstream of nlgn3a , we found variable signatures of balancing selection in qrsl1 , essential to mitochondrial protein synthesis, and rtn4ip1 , important to optic health ( 99 ), which could influence energy expenditure related to mating displays and visual processing critical to mate recognition. Finally, we observed varied signatures of balancing and positive selection on slc45a2 , a key gene involved in melanin synthesis ( 100 ). Future studies could validate the functional relevance of these genes by integrating gene expression analyses, CRISPR-based functional knockouts, or behavioural assays linking genotype to colour variation, visual discrimination, or social responsiveness under controlled ecological and demographic contexts. It is likely that we missed some regions under balancing selection, as the statistical approaches we used are designed to identify excess intermediate-frequency polymorphisms, which are characteristic of long-term balancing selection. Such methods are less sensitive to detecting ancient trans-species polymorphisms ( 101 ) or recent balancing selection, such as those arising during a selective sweep ( 5 , 8 ). Nevertheless, long-term balancing selection is the most plausible model for our hypothesised mode of NFDS, i.e. , that male colour polymorphisms have been maintained across guppy populations through consistent female preference for rare male phenotypes ( 29 , 30 ). However, the experimentally introduced populations have only recently experienced changes in predation selection pressures, and any newly emerging balanced polymorphisms may not yet have had sufficient time to exhibit strong genomic signatures (although see ( 17 )). Such signatures would be particularly interesting to investigate further as recent empirical evidence has shown that fixation of beneficial alleles often follows an initial period of balancing selection ( i.e. , a staggered sweep model) ( 102 ). We employed conservative filters to avoid false positives arising from repetitive, duplicated, or misassembled regions of the genome. While this approach reduces spurious signals ( 103 ), it also likely excluded more complex loci, including those in gene families known to evolve under balancing selection, such as immunity genes. Indeed, MHC genes are known to exist as multiple copies in guppies ( 104 – 106 ). Similarly, we did not detect repeatable balancing selection on male-specific regions of elevated diversity on the Y chromosome (LG12) ( 107 ), nor on the more complex haplotypes previously implicated in male colour variation ( 108 ). Future work using long-read sequencing and pangenome-based approaches will improve the detection of balancing selection in complex regions ( 109 ), especially at loci with high copy number variation and/or extensive structural variation ( 22 , 110 , 111 ). Finally, as with most genome scan approaches, our detection framework is inherently biased toward loci of large effect ( 112 ). Traits like male colouration and mate choice in guppies are likely polygenic, meaning that weaker or more diffuse selection signals could go undetected ( 113 ). Conversely, some regions identified here may represent false positives, since patterns of intermediate allele frequency can also arise from incomplete sweeps, population structure, or admixture ( 18 ). However, our divergence tree showed no evidence for recent admixture across drainages, reducing the likelihood that our most repeatable signals are the result of introgression. Overall, our strategy of comparing populations with contrasting predation-selective histories and demographic contexts strengthens the inference that our candidate regions represent robust targets of balancing selection. We primarily relied on the B 2 statistic using the derived allele frequency (DAF) spectrum ( 44 , 45 ), alongside the rank-based PicMin approach ( 75 ) to identify candidate regions of repeatable balancing selection. To strengthen the evidence for these candidates, we also compared them with the B 2 statistic modelled using the minor allele frequency (MAF) spectra, β statistics ( 46 , 47 ), and standard diversity statistics. We chose to use the B 2 statistic modelled on the DAF spectra because it incorporates both substitutions and polymorphisms, and has been shown to outperform other methods across a wide range of demographic scenarios and recombination rate heterogeneity ( 44 ); the latter being particularly critical as we applied a uniform recombination map. One challenge of DAF-spectra-based approaches is the requirement for reliable inference of the ancestral allele. Despite having two outgroup species, we still observed an excess of fixed derived alleles in our data, likely due to ancestral allele miscalls. Nonetheless, the B 2 statistic modelled using minor allele frequency (MAF) spectra produced highly similar results, with the majority of outlier windows identified under MAF also detected in the DAF results. This reinforces the robustness of our candidate regions, but also highlights the potential of using MAF-based approaches for identifying balancing selection when reliable inference of ancestral alleles can be difficult ( 114 ), especially for non-models. When comparing B 2 to β , we found limited overlap in top-scoring windows across populations; an unexpected result given that both methods aim to detect similar signatures of balancing selection. This discordance did not appear to differ substantially among populations. However, many of our most repeatable candidate regions showed elevated β scores, emphasising the value of integrating multiple populations and applying rank-based approaches (like PicMin; ( 75 )) instead of relying on arbitrary thresholds. Finally, we compared our candidates to traditional measures of excess diversity (Tajima’s D and π). While top B 2 outliers typically ranked highly in these measures, this was not always true. This is likely due to the challenges in modelling the allele frequency spectra in small populations, where the assumptions underlying statistics like Tajima’s D do not always hold ( 115 ). Additionally, reduced linkage in certain populations, such as EU, may have lowered our overall power to detect balancing selection. However, by integrating information from all these complementary approaches, we narrowed our results to a high-confidence set of 23 candidate regions for functional investigation. In summary, our study provides a comparative analysis of genome-wide scans for detecting balancing selection across multiple natural populations, revealing how NFDS may contribute to the maintenance of genetic diversity, even across demographically diverse populations. To our knowledge, this is among the first studies to explore evolutionary repeatability in the context of balancing, as opposed to positive selection, across natural populations. Our findings demonstrate the value of incorporating genome scans for balancing selection to complement phenotypic, ecological, and behavioural studies, offering a broader perspective on how NFDS operates. Future studies expanding genome-wide investigations of NFDS to additional taxa and demographic contexts will be critical to increasing our understanding of how balancing selection operates in maintaining diversity. Data Accessibility All scripts pertaining to data analysis are available on GitHub ( https://github.com/josieparis/guppy_balancing_selection ). Author contributions Josephine R. Paris - Formal analysis, Investigation, Methodology, Visualization, Writing - original draft, Writing - review & editing James R Whiting - Formal analysis, Methodology, Writing - review & editing Joan Ferrer Obiol - Formal analysis, Methodology, Visualization, Writing - review & editing Kimberly A Hughes - Conceptualization, Funding acquisition, Project administration, Writing - review & editing Bonnie A Fraser - Conceptualization, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Writing - original draft, Writing - review & editing Conflict of interest declaration We declare we have no competing interests. Funding This research was funded by the UK Research and Innovation (UKRI) Natural Environment Research Council (NERC, NE/P013074/1) (J.R.P., B.A.F.), EU Research Council grant (GuppyCon 758382) (B.A.F., J.R.W.) and the National Science Foundation of the United States (NSF) ISO-1354775 and DEB-1740466 (K.A.H.). Acknowledgements The authors wish to thank Helen Rodd for the collection of guppies from the Paria and Turure rivers and useful discussions at the conception of this project. Thanks to Mijke van der Zee for useful discussions and suggestions on analyses. Computational infrastructure support was provided by The University of Exeter’s High-Performance Computing (HPC) facility (ISCA). DNA sequencing was performed by the Earlham Institute, UK. Funder Information Declared Natural Environment Research Council, https://ror.org/02b5d8509 , NE/P013074/1 European Research Council, https://ror.org/0472cxd90 , GuppyCon 758382 National Science Foundation , ISO-1354775 National Science Foundation , DEB-1740466 References 1. ↵ Ayala FJ , Campbell CA . Frequency-dependent selection . Annu Rev Ecol Syst . 1974 Nov ; 5 ( 1 ): 115 – 38 . OpenUrl CrossRef 2. Clarke BC . The evolution of genetic diversity . Proc R Soc Lond B Biol Sci . 1979 Sep 21 ; 205 ( 1161 ): 453 – 74 . OpenUrl CrossRef 3. Wright S . The distribution of self-sterility alleles in populations . Genetics . 1939 Jun ; 24 ( 4 ): 538 – 52 . OpenUrl FREE Full Text 4. ↵ Turelli M , Barton NH . Polygenic variation maintained by balancing selection: Pleiotropy, sex-dependent Allelic effects and G × E interactions . Genetics . 2004 Feb 1 ; 166 ( 2 ): 1053 – 79 . OpenUrl Abstract / FREE Full Text 5. ↵ Bitarello BD , Brandt DYC , Meyer D , Andrés AM . Inferring balancing selection from genome-scale data . Genome Biol Evol . 2023 Mar 3 ; 15 ( 3 ). 6. ↵ Fijarczyk A , Babik W . Detecting balancing selection in genomes: limits and prospects . Mol Ecol . 2015 Jul ; 24 ( 14 ): 3529 – 45 . OpenUrl CrossRef 7. ↵ Charlesworth B . Fundamental concepts in genetics: effective population size and patterns of molecular evolution and variation . Nat Rev Genet . 2009 Mar ; 10 ( 3 ): 195 – 205 . OpenUrl CrossRef PubMed Web of Science 8. ↵ Charlesworth D . Balancing selection and its effects on sequences in nearby genome regions . PLoS Genet . 2006 Apr ; 2 ( 4 ): e64 . OpenUrl CrossRef PubMed 9. ↵ Castro-Prieto A , Wachter B , Sommer S . Cheetah paradigm revisited: MHC diversity in the world’s largest free-ranging population . Mol Biol Evol . 2011 Apr ; 28 ( 4 ): 1455 – 68 . OpenUrl CrossRef PubMed Web of Science 10. Aguilar A , Roemer G , Debenham S , Binns M , Garcelon D , Wayne RK . High MHC diversity maintained by balancing selection in an otherwise genetically monomorphic mammal . Proc Natl Acad Sci U S A . 2004 Mar 9 ; 101 ( 10 ): 3490 – 4 . OpenUrl Abstract / FREE Full Text 11. ↵ Benazzo A , Trucchi E , Cahill JA , Delser PM , Mona S , Fumagalli M , et al. Survival and divergence in a small group: The extraordinary genomic history of the endangered Apennine brown bear stragglers . Proc Natl Acad Sci U S A . 2017 Nov 7 ; 114 ( 45 ): E9589 – 97 . OpenUrl Abstract / FREE Full Text 12. Niskanen AK , Kennedy LJ , Ruokonen M , Kojola I , Lohi H , Isomursu M , et al. Balancing selection and heterozygote advantage in major histocompatibility complex loci of the bottlenecked Finnish wolf population . Mol Ecol . 2014 Feb 1 ; 23 ( 4 ): 875 – 89 . OpenUrl CrossRef 13. ↵ Marmesat E , Schmidt K , Saveljev AP , Seryodkin IV , Godoy JA . Retention of functional variation despite extreme genomic erosion: MHC allelic repertoires in the Lynx genus . BMC Evol Biol . 2017 Jul 4 ; 17 ( 1 ): 158 . OpenUrl CrossRef PubMed 14. ↵ Ejsmond MJ , Radwan J . MHC diversity in bottlenecked populations: a simulation model . Conserv Genet . 2011 Feb ; 12 ( 1 ): 129 – 37 . OpenUrl CrossRef 15. ↵ Harris RB , Jensen JD . Considering genomic scans for selection as coalescent model choice . Genome Biol Evol . 2020 Jun 1 ; 12 ( 6 ): 871 – 7 . OpenUrl CrossRef PubMed 16. Jensen JD , Thornton KR , Bustamante CD , Aquadro CF . On the utility of linkage disequilibrium as a statistic for identifying targets of positive selection in nonequilibrium populations . Genetics . 2007 Aug ; 176 ( 4 ): 2371 – 9 . OpenUrl Abstract / FREE Full Text 17. ↵ Isildak U , Stella A , Fumagalli M . Distinguishing between recent balancing selection and incomplete sweep using deep neural networks . Mol Ecol Resour . 2021 Nov ; 21 ( 8 ): 2706 – 18 . OpenUrl CrossRef PubMed 18. ↵ Soni V , Jensen JD . Temporal challenges in detecting balancing selection from population genomic data . G3 (Bethesda) . 2024 Jun 5 ; 14 ( 6 ): jkae069 . OpenUrl CrossRef 19. ↵ Muirhead CA . Consequences of population structure on genes under balancing selection . Evolution . 2001 Aug 1 ; 55 ( 8 ): 1532 – 41 . OpenUrl CrossRef PubMed Web of Science 20. ↵ Fraser BA , Whiting JR . What can be learned by scanning the genome for molecular convergence in wild populations? Ann N Y Acad Sci . 2020 ; 1476 ( 1 ): 23 – 42 . OpenUrl CrossRef PubMed 21. ↵ Koenig D , Hagmann J , Li R , Bemm F , Slotte T , Neuffer B , et al. Long-term balancing selection drives evolution of immunity genes in Capsella . Elife 2019 Feb 26 . 22. ↵ Lee D , Zdraljevic S , Stevens L , Wang Y , Tanny RE , Crombie TA , et al. Balancing selection maintains hyper-divergent haplotypes in Caenorhabditis elegans . Nature Ecology & Evolution . 2021 Apr 5 ; 5 ( 6 ): 794 – 807 . OpenUrl CrossRef PubMed 23. ↵ Hughes KA , Rodd FH , Reznick DN . Genetic and environmental effects on secondary sex traits in guppies ( Poecilia reticulata ) . J Evol Biol . 2005 Jan ; 18 ( 1 ): 35 – 45 . OpenUrl CrossRef PubMed Web of Science 24. ↵ Hughes KA , Houde AE , Price AC , Rodd FH . Mating advantage for rare males in wild guppy populations . Nature . 2013 Nov 7 ; 503 ( 7474 ): 108 – 10 . OpenUrl CrossRef PubMed Web of Science 25. Potter T , Arendt J , Bassar RD , Watson B , Bentzen P , Travis J , et al. Female preference for rare males is maintained by indirect selection in Trinidadian guppies . Science . 2023 Apr 21 ; 380 ( 6642 ): 309 – 12 . OpenUrl CrossRef PubMed 26. ↵ Daniel MJ , Koffinas L , Hughes KA . Habituation underpins preference for mates with novel phenotypes in the guppy . Proc Biol Sci . 2019 May 15 ; 286 ( 1902 ): 20190435 . OpenUrl CrossRef PubMed 27. ↵ Fraser BA , Hughes KA , Tosh DN , Rodd FH . The role of learning by a predator, Rivulus hartii , in the rare-morph survival advantage in guppies . J Evol Biol . 2013 ; 26 ( 12 ): 2597 – 605 . OpenUrl CrossRef PubMed 28. ↵ Olendorf R , Rodd FH , Punzalan D , Houde AE , Hurt C , Reznick DN , et al. Frequency-dependent survival in natural guppy populations . Nature . 2006 Jun 1 ; 441 ( 7093 ): 633 – 6 . OpenUrl CrossRef PubMed Web of Science 29. ↵ Valvo JJ , Rodd FH , Hughes KA . Consistent female preference for rare and unfamiliar male color patterns in wild guppy populations . Behav Ecol . 2019 Nov 8 ; 30 ( 6 ): 1672 – 81 . OpenUrl CrossRef 30. ↵ Valvo JJ , Aponte JD , Daniel MJ , Dwinell K , Rodd H , Houle D , et al. Using Delaunay triangulation to sample whole-specimen color from digital images . Ecol Evol 2021 Aug 20 ;( ece3.7992 ). 31. ↵ Suk HY , Neff BD . Microsatellite genetic differentiation among populations of the Trinidadian guppy . Heredity (Edinb ). 2009 May 1 ; 102 : 425 – 34 . OpenUrl CrossRef PubMed 32. ↵ Willing EM , Bentzen P , van Oosterhout C , Hoffmann M , Cable J , Breden F , et al. Genome-wide single nucleotide polymorphisms reveal population history and adaptive divergence in wild guppies . Mol Ecol . 2010 Mar ; 19 ( 5 ): 968 – 84 . OpenUrl CrossRef PubMed 33. ↵ Whiting JR , Paris JR , van der Zee MJ , Parsons PJ , Weigel D , Fraser BA . Drainage-structuring of ancestral variation and a common functional pathway shape limited genomic convergence in natural high- and low-predation guppies . PLoS Genet . 2021 May 24 ; 17 ( 5 ): e1009566 . OpenUrl CrossRef PubMed 34. ↵ Fraser BA , Künstner A , Reznick DN , Dreyer C , Weigel D . Population genomics of natural and experimental populations of guppies ( Poecilia reticulata ) . Mol Ecol . 2015 Jan ; 24 ( 2 ): 389 – 408 . OpenUrl CrossRef 35. ↵ Crispo E , Bentzen P , Reznick DN , Kinnison MT , Hendry AP . The relative influence of natural selection and geography on gene flow in guppies . Mol Ecol . 2006 Jan 1 ; 15 ( 1 ): 49 – 62 . OpenUrl CrossRef PubMed Web of Science 36. ↵ Barson NJ , Cable J , Van Oosterhout C . Population genetic analysis of microsatellite variation of guppies ( Poecilia reticulata ) in Trinidad and Tobago: evidence for a dynamic source-sink metapopulation structure, founder events and population bottlenecks . J Evol Biol . 2009 Mar ; 22 ( 3 ): 485 – 97 . OpenUrl CrossRef PubMed Web of Science 37. ↵ Reznick DA , Bryga H , Endler JA . Experimentally induced life-history evolution in a natural population . Nature . 1990 Jul ; 346 ( 6282 ): 357 – 9 . OpenUrl CrossRef Web of Science 38. ↵ Reznick DN , Bryga H . Life-history evolution in guppies ( Poecilia reticulata ): 1. Phenotypic and genetic changes in an introduction experiment . Evolution . 1987 Nov ; 41 ( 6 ): 1370 – 85 . OpenUrl CrossRef PubMed Web of Science 39. Endler JA . Natural selection on color patterns in Poecilia reticulata . Evolution . 1980 ; 34 ( 1 ): 76 – 91 . OpenUrl CrossRef PubMed Web of Science 40. ↵ Magurran AE . Population differentiation without speciation . Philos Trans R Soc Lond B Biol Sci . 1998 Feb 28 ; 353 ( 1366 ): 275 – 86 . OpenUrl CrossRef Web of Science 41. ↵ van der Zee MJ , Whiting JR , Paris JR , Bassar RD , Travis J , Weigel D , et al. Rapid genomic convergent evolution in experimental populations of Trinidadian guppies ( Poecilia reticulata ) . Evol Lett . 2022 Apr ; 6 ( 2 ): 149 – 61 . OpenUrl CrossRef PubMed 42. Fitzpatrick SW , Gerberich JC , Angeloni LM , Bailey LL , Broder ED , Torres-Dowdall J , et al. Gene flow from an adaptively divergent source causes rescue through genetic and demographic factors in two wild populations of Trinidadian guppies . Evol Appl . 2016 Aug 1 ; 9 ( 7 ): 879 – 91 . OpenUrl CrossRef PubMed 43. ↵ Kemp DJ , Reznick DN , Grether GF , Endler JA . Predicting the direction of ornament evolution in Trinidadian guppies ( Poecilia reticulata ) . Proc Biol Sci . 2009 Dec 22 ; 276 ( 1677 ): 4335 – 43 . OpenUrl CrossRef PubMed 44. ↵ Cheng X , DeGiorgio M . Flexible Mixture Model Approaches That Accommodate Footprint Size Variability for Robust Detection of Balancing Selection . Mol Biol Evol . 2020 Oct 4 ; 37 ( 11 ): 3267 – 91 . OpenUrl CrossRef PubMed 45. ↵ Cheng X , DeGiorgio M . BalLeRMix +: mixture model approaches for robust joint identification of both positive selection and long-term balancing selection . Bioinformatics . 2022 Jan 12 ; 38 ( 3 ): 861 – 3 . OpenUrl CrossRef PubMed 46. ↵ Siewert KM , Voight BF . BetaScan2: standardized statistics to detect balancing selection utilizing substitution data . Genome Biol Evol . 2020 Feb 1 ; 12 ( 2 ): 3873 – 7 . OpenUrl CrossRef PubMed 47. ↵ Siewert KM , Voight BF . Detecting long-term balancing selection using allele frequency correlation . Mol Biol Evol . 2017 Nov 1 ; 34 ( 11 ): 2996 – 3005 . OpenUrl CrossRef PubMed 48. ↵ Endler JA , Houde AE . Geographic variation in female preferences for male traits in Poecilia reticulata . Evolution . 1995 Jun ; 49 ( 3 ): 456 – 68 . OpenUrl CrossRef PubMed Web of Science 49. ↵ Shaw PW , Carvalho GR , Seghers BH , Magurran AE . Genetic consequences of an artificial introduction of guppies ( Poecilia reticulata ) in N. Trinidad . Proc Biol Sci . 1992 May 22 ; 248 ( 1322 ): 111 – 6 . OpenUrl CrossRef 50. ↵ Shaw PW , Carvalho GR , Magurran AE , Seghers BH . Population differentiation in Trinidadian guppies ( Poecilia reticulata ): patterns and problems . J Fish Biol . 1991 Dec ; 39 ( sA ): 203 – 9 . OpenUrl CrossRef Web of Science 51. ↵ Martin M . Cutadapt removes adapter sequences from high-throughput sequencing reads . EMBnet.journal . 2011 May 2 ; 17 ( 1 ): 10 – 2 . OpenUrl 52. ↵ Fraser BA , Whiting JR , Paris JR , Weadick CJ , Parsons PJ , Charlesworth D , et al. Improved reference genome uncovers novel sex-linked regions in the guppy ( Poecilia reticulata ) . Genome Biol Evol . 2020 Oct 1 ; 12 ( 10 ): 1789 – 805 . OpenUrl CrossRef PubMed 53. ↵ Li H . Aligning sequence reads, clone sequences and assembly contigs with BWA-MEM arXiv [q-bio.GN]. 2013 . 54. ↵ Li H , Handsaker B , Wysoker A , Fennell T , Ruan J , Homer N , et al. The Sequence Alignment/Map format and SAMtools . Bioinformatics . 2009 Aug 15 ; 25 ( 16 ): 2078 – 9 . OpenUrl CrossRef PubMed Web of Science 55. ↵ McKenna A , Hanna M , Banks E , Sivachenko A , Cibulskis K , Kernytsky A , et al. The Genome Analysis Toolkit: a MapReduce framework for analyzing next-generation DNA sequencing data . Genome Res . 2010 Sep ; 20 ( 9 ): 1297 – 303 . OpenUrl Abstract / FREE Full Text 56. ↵ Danecek P , Auton A , Abecasis G , Albers CA , Banks E , DePristo MA , et al. The variant call format and VCFtools . Bioinformatics . 2011 Aug 1 ; 27 ( 15 ): 2156 – 8 . OpenUrl CrossRef PubMed Web of Science 57. ↵ Browning SR , Browning BL . Rapid and accurate haplotype phasing and missing-data inference for whole-genome association studies by use of localized haplotype clustering . Am J Hum Genet . 2007 Nov ; 81 ( 5 ): 1084 – 97 . OpenUrl CrossRef PubMed Web of Science 58. ↵ Delaneau O , Howie B , Cox AJ , Zagury JF , Marchini J . Haplotype estimation using sequencing reads . Am J Hum Genet . 2013 Oct 3 ; 93 ( 4 ): 687 – 96 . OpenUrl CrossRef PubMed 59. ↵ Charlesworth D , Bergero R , Graham C , Gardner J , Yong L . Locating the sex determining region of linkage group 12 of guppy ( Poecilia reticulata ) . G3: Genes, Genomes, Genetics . 2020 Oct 1 ; 10 ( 10 ): 3639 – 49 . OpenUrl CrossRef 60. ↵ Whiting JR , Paris JR , Parsons PJ , Matthews S , Reynoso Y , Hughes KA , et al. On the genetic architecture of rapidly adapting and convergent life history traits in guppies . Heredity . 2022 March 8 ; 128 : 250 – 260 . OpenUrl CrossRef PubMed 61. ↵ Morris J , Darolti I , Bloch NI , Wright AE , Mank JE . Shared and species-specific patterns of nascent y chromosome evolution in two guppy species . Genes . 2018 May 3 ; 9 ( 5 ). 62. ↵ Purcell S , Neale B , Todd-Brown K , Thomas L , Ferreira MAR , Bender D , et al. PLINK: a tool set for whole-genome association and population-based linkage analyses . Am J Hum Genet . 2007 Sep ; 81 ( 3 ): 559 – 75 . OpenUrl CrossRef PubMed 63. ↵ Pfeifer B , Wittelsbürger U , Ramos-Onsins SE , Lercher MJ . PopGenome: an efficient Swiss army knife for population genomic analyses in R . Mol Biol Evol . 2014 Jul ; 31 ( 7 ): 1929 – 36 . OpenUrl CrossRef PubMed Web of Science 64. ↵ Lawson DJ , Hellenthal G , Myers S , Falush D . Inference of population structure using dense haplotype data . PLoS Genet . 2012 Jan ; 8 ( 1 ): e1002453 . OpenUrl CrossRef PubMed 65. ↵ Stoltz M , Baeumer B , Bouckaert R , Fox C , Hiscott G , Bryant D . Bayesian inference of species trees using diffusion models . Syst Biol . 2021 Jan 1 ; 70 ( 1 ): 145 – 61 . OpenUrl CrossRef PubMed 66. ↵ Stange M , Sánchez-Villagra MR , Salzburger W , Matschiner M . Bayesian divergence-time estimation with genome-wide single-nucleotide polymorphism data of sea catfishes (Ariidae) supports Miocene closure of the Panamanian isthmus . Syst Biol . 2018 Jul 1 ; 67 ( 4 ): 681 – 99 . OpenUrl CrossRef PubMed 67. ↵ Speidel L , Forest M , Shi S , Myers SR . A method for genome-wide genealogy estimation for thousands of samples . Nat Genet . 2019 Sep 2 ; 51 ( 9 ): 1321 – 9 . OpenUrl CrossRef PubMed 68. ↵ Speidel L , Cassidy L , Davies RW , Hellenthal G , Skoglund P , Myers SR . Inferring population histories for ancient genomes using genome-wide genealogies . Mol Biol Evol . 2021 Aug 23 ; 38 ( 9 ): 3497 – 511 . OpenUrl CrossRef PubMed 69. ↵ Burda K , Konczal M . Validation of machine learning approach for direct mutation rate estimation . Mol Ecol Resour . 2023 Nov ; 23 ( 8 ): 1757 – 71 . OpenUrl CrossRef PubMed 70. ↵ Whiting JR , Paris J , Parsons P , Matthews S , Reynoso Y , Hughes K , et al. On the genetic architecture of rapidly adapting and convergent life history traits in guppies . Heredity (Edinb ). 2021 Mar 19 ; 128 ( 4 ): 250 – 60 . OpenUrl 71. ↵ Potter T , Bassar RD , Bentzen P , Ruell EW , Torres-Dowdall J , Handelsman CA , et al. environmental change, if unaccounted, prevents detection of cryptic evolution in a wild population . Am Nat . 2021 Jan ; 197 ( 1 ): 29 – 46 . OpenUrl CrossRef PubMed 72. ↵ Killick R , Eckley I. changepoint: An R package for changepoint analysis . J Stat Softw . 2014 ; 58 ( 3 ): 1 – 19 . OpenUrl CrossRef 73. ↵ Li H , Ralph P . Local PCA shows how the effect of population structure differs along the genome . Genetics . 2019 Jan 1 ; 211 ( 1 ): 289 – 304 . OpenUrl Abstract / FREE Full Text 74. ↵ Shin JH , Blay S , McNeney B , Graham J , Others. LDheatmap: an R function for graphical display of pairwise linkage disequilibria between single nucleotide polymorphisms . J Stat Softw . 2006 ; 16 ( 3 ): 1 – 10 . OpenUrl CrossRef 75. ↵ Booker TR , Yeaman S , Whitlock MC . Using genome scans to identify genes used repeatedly for adaptation . Evolution . 2023 Mar 1 ; 77 ( 3 ): 801 – 11 . OpenUrl CrossRef PubMed 76. ↵ Gabriel L , Brůna T , Hoff KJ , Ebel M , Lomsadze A , Borodovsky M , et al. BRAKER3: Fully automated genome annotation using RNA-seq and protein evidence with GeneMark-ETP, AUGUSTUS, and TSEBRA . Genome Res . 2024 Jun 25 ; 34 ( 5 ): 769 – 77 . OpenUrl Abstract / FREE Full Text 77. ↵ Robinson JA , Räikkönen J , Vucetich LM , Vucetich JA , Peterson RO , Lohmueller KE , et al. Genomic signatures of extensive inbreeding in Isle Royale wolves, a population on the threshold of extinction . Sci Adv . 2019 May ; 5 ( 5 ): eaau0757 . OpenUrl FREE Full Text 78. ↵ Otto SP , Barton NH . Selection for recombination in small populations . Evolution . 2001 Oct 1 ; 55 ( 10 ): 1921 – 31 . OpenUrl CrossRef PubMed Web of Science 79. ↵ Leroy T , Rousselle M , Tilak MK , Caizergues AE , Scornavacca C , Recuerda M , et al. Island songbirds as windows into evolution in small populations . Curr Biol . 2021 Mar 22 ; 31 ( 6 ): 1303 – 10.e4 . OpenUrl CrossRef PubMed 80. ↵ Brazier T , Stetsenko R , Roze D , Glémin S . Mating system and the evolution of recombination rates in seed plants . J Evol Biol . 2025 May 5 ; voaf008 . 81. ↵ Burri R . Evolution: Small populations, low recombination, big trouble? Curr Biol . 2021 Mar 22 ; 31 ( 6 ): R282 – 4 . OpenUrl CrossRef PubMed 82. ↵ Parfitt DA , Michael GJ , Vermeulen EGM , Prodromou NV , Webb TR , Gallo JM , et al. The ataxia protein sacsin is a functional co-chaperone that protects against polyglutamine-expanded ataxin-1 . Hum Mol Genet . 2009 May 1 ; 18 ( 9 ): 1556 – 65 . OpenUrl CrossRef PubMed Web of Science 83. ↵ Naef V , Damiani D , Licitra R , Marchese M , Vecchia SD , Baggiani M , et al. Modeling sacsin depletion in Danio Rerio offers new insight on retinal defects in ARSACS . Neurobiol Dis . 2025 Feb 1 ; 205 ( 106793 ): 106793 . OpenUrl CrossRef PubMed 84. ↵ Nyman C , Hebert FO , Bessert-Nettelbeck M , Aubin-Horth N , Taborsky B . Transcriptomic signatures of social experience during early development in a highly social cichlid fish . Mol Ecol . 2020 Feb 1 ; 29 ( 3 ): 610 – 23 . OpenUrl CrossRef 85. ↵ Hayashi S , Takeichi M . Emerging roles of protocadherins: from self-avoidance to enhancement of motility . J Cell Sci . 2015 Apr 15 ; 128 ( 8 ): 1455 – 64 . OpenUrl Abstract / FREE Full Text 86. ↵ Zhu D , Li C , Swanson AM , Villalba RM , Guo J , Zhang Z , et al. BAI1 regulates spatial learning and synaptic plasticity in the hippocampus . J Clin Invest . 2015 Apr ; 125 ( 4 ): 1497 – 508 . OpenUrl CrossRef PubMed 87. ↵ Shiu FH , Wong JC , Yamamoto T , Lala T , Purcell RH , Owino S , et al. Mice lacking full length Adgrb1 (Bai1) exhibit social deficits, increased seizure susceptibility, and altered brain development . Exp Neurol . 2022 May 1 ; 351 ( 113994 ): 113994 . OpenUrl CrossRef PubMed 88. ↵ Parichy DM , Mellgren EM , Rawls JF , Lopes SS , Kelsh RN , Johnson SL . Mutational analysis of endothelin receptor b1 (rose) during neural crest and pigment pattern development in the zebrafish Danio rerio . Dev Biol . 2000 Nov 15 ; 227 ( 2 ): 294 – 306 . OpenUrl CrossRef PubMed Web of Science 89. ↵ Tuladhar S , Kanneganti TD . NLRP12 in innate immunity and inflammation . Mol Aspects Med . 2020 Dec ; 76 ( 100887 ): 100887 . OpenUrl CrossRef PubMed 90. ↵ Shin E , Kashiwagi Y , Kuriu T , Iwasaki H , Tanaka T , Koizumi H , et al. Doublecortin-like kinase enhances dendritic remodelling and negatively regulates synapse maturation . Nat Commun . 2013 ; 4 ( 1 ): 1440 . OpenUrl CrossRef PubMed 91. ↵ DeOliveira-Mello L , Vicente I , Gonzalez-Nunez V , Santos-Ledo A , Velasco A , Arévalo R , et al. Doublecortin in the fish visual system, a specific protein of maturing neurons . Biology (Basel ). 2022 Feb 6 ; 11 ( 2 ): 248 . OpenUrl PubMed 92. ↵ Hörnberg H , Pérez-Garci E , Schreiner D , Hatstatt-Burklé L , Magara F , Baudouin S , et al. Rescue of oxytocin response and social behaviour in a mouse model of autism . Nature . 2020 Aug 5 ; 584 ( 7820 ): 252 – 6 . OpenUrl CrossRef PubMed 93. Bariselli S , Hörnberg H , Prévost-Solié C , Musardo S , Hatstatt-Burklé L , Scheiffele P , et al. Role of VTA dopamine neurons and neuroligin 3 in sociability traits related to nonfamiliar conspecific interaction . Nat Commun . 2018 Aug 9 ; 9 ( 1 ): 3173 . OpenUrl CrossRef PubMed 94. ↵ Tabuchi K , Blundell J , Etherton MR , Hammer RE , Liu X , Powell CM , et al. A neuroligin-3 mutation implicated in autism increases inhibitory synaptic transmission in mice . Science . 2007 Oct 5 ; 318 ( 5847 ): 71 – 6 . OpenUrl Abstract / FREE Full Text 95. ↵ Cummings ME , Larkins-Ford J , Reilly CRL , Wong RY , Ramsey M , Hofmann HA . Sexual and social stimuli elicit rapid and contrasting genomic responses . Proc Biol Sci . 2008 Feb 22 ; 275 ( 1633 ): 393 – 402 . OpenUrl CrossRef PubMed Web of Science 96. ↵ Cummings ME . Looking for sexual selection in the female brain . Philos Trans R Soc Lond B Biol Sci . 2012 Aug 19 ; 367 ( 1600 ): 2348 – 56 . OpenUrl CrossRef PubMed 97. ↵ Reyes AS . Spatiotemporal changes in the expression of genes junb and nlgn3 associated with mate choice in the brain of the guppy ( Poecilia reticulata ). [Bogota, Columbia]: University of the Andes; 2022 . 98. ↵ Bloch NI , Corral-López A , Buechel SD , Kotrschal A , Kolm N , Mank JE . Early neurogenomic response associated with variation in guppy female mate preference . Nat Ecol Evol . 2018 Nov 8 ; 2 ( 11 ): 1772 – 81 . OpenUrl CrossRef PubMed 99. ↵ Angebault C , Guichet PO , Talmat-Amar Y , Charif M , Gerber S , Fares-Taie L , et al. Recessive mutations in RTN4IP1 cause isolated and syndromic optic neuropathies . Am J Hum Genet . 2015 Nov 5 ; 97 ( 5 ): 754 – 60 . OpenUrl CrossRef PubMed 100. ↵ Tsetskhladze ZR , Canfield VA , Ang KC , Wentzel SM , Reid KP , Berg AS , et al. Functional assessment of human coding mutations affecting skin pigmentation using zebrafish . PLoS One . 2012 Oct 10 ; 7 ( 10 ): e47398 . OpenUrl CrossRef PubMed 101. ↵ Cheng X , DeGiorgio M . Detection of shared balancing selection in the absence of trans-species polymorphism . Mol Biol Evol . 2019 Jan 1 ; 36 ( 1 ): 177 – 99 . OpenUrl CrossRef PubMed 102. ↵ Pivirotto AM , Platt A , Patel R , Kumar S , Hey J . Analyses of allele age and fitness impact reveal human beneficial alleles to be older than neutral controls . eLife . 2024 . 103. ↵ Bubb KL , Bovee D , Buckley D , Haugen E , Kibukawa M , Paddock M , et al. Scan of human genome reveals no new loci under ancient balancing selection . Genetics . 2006 Aug 1 ; 173 ( 4 ): 2165 – 77 . OpenUrl Abstract / FREE Full Text 104. ↵ Fraser BA , Ramnarine I , Neff BD . Selection at the MHC class IIB locus across guppy ( Poecilia reticulata ) populations . Heredity (Edinb ). 2010 Feb 1 ; 104 ( 2 ): 155 – 67 . OpenUrl CrossRef PubMed 105. Lighten J , van Oosterhout C , Paterson IG , McMullan M , Bentzen P . Ultra-deep Illumina sequencing accurately identifies MHC class IIb alleles and provides evidence for copy number variation in the guppy ( Poecilia reticulata ) . Mol Ecol Resour . 2014 Jul ; 14 ( 4 ): 753 – 67 . OpenUrl CrossRef PubMed 106. ↵ Herdegen-Radwan M , Phillips KP , Babik W , Mohammed RS , Radwan J . Balancing selection versus allele and supertype turnover in MHC class II genes in guppies . Heredity (Edinb ). 2021 Mar ; 126 ( 3 ): 548 – 60 . OpenUrl CrossRef PubMed 107. ↵ Almeida P , Sandkam BA , Morris J , Darolti I , Breden F , Mank JE . Divergence and remarkable diversity of the Y chromosome in guppies . Mol Biol Evol . 2020 Oct 6 ; 38 ( 2 ): 619 – 33 . OpenUrl 108. ↵ Paris JR , Whiting JR , Daniel MJ , Obiol JF , Parsons PJ . A large and diverse autosomal haplotype is associated with sex-linked colour polymorphism in the guppy . 2022 Mar ; 13 : 1233 . OpenUrl 109. ↵ Wang T , Antonacci-Fulton L , Howe K , Lawson HA , Lucas JK , Phillippy A , et al. The Human Pangenome Project: a global resource to map genomic diversity . Nature . 2022 Apr 1 ; 604 : 437 – 46 . OpenUrl CrossRef PubMed 110. ↵ Schield DR , Perry BW , Adams RH , Holding ML , Nikolakis ZL , Gopalan SS , et al. The roles of balancing selection and recombination in the evolution of rattlesnake venom . Nature Ecology & Evolution . 2022 Jul 18 ; 1 – 14 . 111. ↵ Berdan EL , Blanckaert A , Butlin RK , Flatt T , Slotte T , Wielstra B . Mutation accumulation opposes polymorphism: supergenes and the curious case of balanced lethals . Philos Trans R Soc Lond B Biol Sci . 2022 Aug 1 ; 377 ( 1856 ): 20210199 . OpenUrl CrossRef PubMed 112. ↵ Teshima KM , Coop G , Przeworski M . How reliable are empirical genomic scans for selective sweeps? Genome Res . 2006 Jun 1 ; 16 ( 6 ): 702 – 12 . OpenUrl Abstract / FREE Full Text 113. ↵ Wellenreuther M , Hansson B . Detecting polygenic evolution: Problems, pitfalls, and promises . Trends Genet . 2016 Mar ; 32 ( 3 ): 155 – 64 . OpenUrl CrossRef PubMed 114. ↵ Keightley P , Jackson BC . Inferring the probability of the derived vs. The ancestral allelic state at a polymorphic site . Genetics . 2018 Jan 30 ; 209 : 897 – 906 . OpenUrl Abstract / FREE Full Text 115. ↵ Nielsen R . Statistical tests of selective neutrality in the age of genomics . Heredity (Edinb) . 2001 Jun 1 ; 86 ( Pt 6 ): 641 – 7 . OpenUrl CrossRef PubMed View the discussion thread. Back to top Previous Next Posted June 17, 2025. Download PDF Supplementary Material Email Thank you for your interest in spreading the word about bioRxiv. NOTE: Your email address is requested solely to identify you as the sender of this article. Your Email * Your Name * Send To * Enter multiple addresses on separate lines or separate them with commas. You are going to email the following Repeated signatures of balancing selection in small and large populations of guppies (Poecilia reticulata) Message Subject (Your Name) has forwarded a page to you from bioRxiv Message Body (Your Name) thought you would like to see this page from the bioRxiv website. Your Personal Message CAPTCHA This question is for testing whether or not you are a human visitor and to prevent automated spam submissions. Share Repeated signatures of balancing selection in small and large populations of guppies ( Poecilia reticulata ) Josephine R Paris , James R Whiting , Joan Ferrer Obiol , Kimberly A Hughes , Bonnie A Fraser bioRxiv 2025.06.12.659363; doi: https://doi.org/10.1101/2025.06.12.659363 Share This Article: Copy Citation Tools Repeated signatures of balancing selection in small and large populations of guppies ( Poecilia reticulata ) Josephine R Paris , James R Whiting , Joan Ferrer Obiol , Kimberly A Hughes , Bonnie A Fraser bioRxiv 2025.06.12.659363; doi: https://doi.org/10.1101/2025.06.12.659363 Citation Manager Formats BibTeX Bookends EasyBib EndNote (tagged) EndNote 8 (xml) Medlars Mendeley Papers RefWorks Tagged Ref Manager RIS Zotero Tweet Widget Facebook Like Google Plus One Subject Area Evolutionary Biology Subject Areas All Articles Animal Behavior and Cognition (7640) Biochemistry (17706) Bioengineering (13902) Bioinformatics (41978) Biophysics (21465) Cancer Biology (18611) Cell Biology (25528) Clinical Trials (138) Developmental Biology (13387) Ecology (19920) Epidemiology (2067) Evolutionary Biology (24332) Genetics (15615) Genomics (22519) Immunology (17747) Microbiology (40424) Molecular Biology (17194) Neuroscience (88662) Paleontology (667) Pathology (2838) Pharmacology and Toxicology (4827) Physiology (7650) Plant Biology (15160) Scientific Communication and Education (2046) Synthetic Biology (4302) Systems Biology (9826) Zoology (2271)