Ancient mitochondrial genomes unveil the origins and evolutionary history of New Zealand's enigmatic takahe and moho.

Many avian species endemic to Aotearoa New Zealand were driven to extinction or reduced to relict populations following successive waves of human arrival, due to hunting, habitat destruction and the introduction of mammalian predators. Among the affected species were the large flightless South Island takahe (Porphyrio hochstetteri) and the moho (North Island takahe; P. mantelli), with the latter rendered extinct and the former reduced to a single relictual population. Little is known about the evolutionary history of these species prior to their decline and/or extinction. Here we sequenced mitochondrial genomes from takahe and moho subfossils (12 takahe and 4 moho) and retrieved comparable sequence data from takahe museum skins (n = 5) and contemporary individuals (n = 17) to examine the phylogeny and recent evolutionary history of these species. Our analyses suggest that prehistoric takahe populations lacked deep phylogeographic structure, in contrast to moho, which exhibited significant spatial genetic structure, albeit based on limited sample sizes (n = 4). Temporal genetic comparisons show that takahe have lost much of their mitochondrial genetic diversity, likely due to a sudden demographic decline soon after human arrival (~750 years ago). Time-calibrated phylogenetic analyses strongly support a sister species relationship between takahe and moho, suggesting these flightless taxa diverged around 1.5 million years ago, following a single colonisation of New Zealand by a flighted Porphyrio ancestor approximately 4 million years ago. This study highlights the utility of palaeogenetic approaches for informing the conservation and systematic understanding of endangered species whose ranges have been severely restricted by anthropogenic impacts.

Tissue-specific patterns of regulatory changes underlying gene expression differences among Ficedula flycatchers and their naturally occurring F1 hybrids.

Changes in interacting cis- and trans-regulatory elements are important candidates for Dobzhansky-Muller hybrid incompatibilities and may contribute to hybrid dysfunction by giving rise to misexpression in hybrids. To gain insight into the molecular mechanisms and determinants of gene expression evolution in natural populations, we analyzed the transcriptome from multiple tissues of two recently diverged Ficedula flycatcher species and their naturally occurring F1 hybrids. Differential gene expression analysis revealed that the extent of differentiation between species and the set of differentially expressed genes varied across tissues. Common to all tissues, a higher proportion of Z-linked genes than autosomal genes showed differential expression, providing evidence for a fast-Z effect. We further found clear signatures of hybrid misexpression in brain, heart, kidney, and liver. However, while testis showed the highest divergence of gene expression among tissues, it showed no clear signature of misexpression in F1 hybrids, even though these hybrids were found to be sterile. It is therefore unlikely that incompatibilities between cis-trans regulatory changes explain the observed sterility. Instead, we found evidence that cis-regulatory changes play a significant role in the evolution of gene expression in testis, which illustrates the tissue-specific nature of cis-regulatory evolution bypassing constraints associated with pleiotropic effects of genes.

Correlated evolution in an ectomycorrhizal host-symbiont system.

Mechanisms of diversification in fungi are relatively poorly known. Many ectomycorrhizal symbionts show preference for particular host genera or families, so host-symbiont selection may be an important driver of fungal diversification in ectomycorrhizal systems. However, whether ectomycorrhizal hosts and symbionts show correlated evolutionary patterns remains untested, and it is unknown whether fungal specialisation also occurs in systems dominated by hosts from the same genus. We use metabarcoding of ectomycorrhizal fungi collected with hyphal ingrowth bags from Nothofagus forests across southern New Zealand to investigate host-symbiont specialisation and correlated evolution. We examine how ectomycorrhizal communities differ between host species and look for patterns of host-symbiont cophylogeny. We found substantial differences in ectomycorrhizal communities associated with different host taxa, particularly between hosts from different subgenera (Lophozonia and Fuscospora), but also between more closely related hosts. Twenty-four per cent of fungal taxa tested showed affiliations to particular hosts, and tests for cophylogeny revealed significant correlations between host relatedness and the fungal phylogeny that extended to substantial evolutionary depth. These results provide new evidence of correlated evolution in ectomycorrhizal systems, indicating that preferences among closely related host species may represent an important evolutionary driver for local lineage diversification in ectomycorrhizal fungi.

Paternal hypoxia exposure primes offspring for increased hypoxia resistance.

BACKGROUND: In a time of rapid environmental change, understanding how the challenges experienced by one generation can influence the fitness of future generations is critically needed. Using tolerance assays and transcriptomic and methylome approaches, we use zebrafish as a model to investigate cross-generational acclimation to hypoxia. RESULTS: We show that short-term paternal exposure to hypoxia endows offspring with greater tolerance to acute hypoxia. We detected two hemoglobin genes that are significantly upregulated by more than 6-fold in the offspring of hypoxia exposed males. Moreover, the offspring which maintained equilibrium the longest showed greatest upregulation in hemoglobin expression. We did not detect differential methylation at any of the differentially expressed genes, suggesting that other epigenetic mechanisms are responsible for alterations in gene expression. CONCLUSIONS: Overall, our findings suggest that an epigenetic memory of past hypoxia exposure is maintained and that this environmentally induced information is transferred to subsequent generations, pre-acclimating progeny to cope with hypoxic conditions.

Concordant phylogeographic responses to large-scale coastal disturbance in intertidal macroalgae and their epibiota.

Major ecological disturbance events can provide opportunities to assess multispecies responses to upheaval. In particular, catastrophic disturbances that regionally extirpate habitat-forming species can potentially influence the genetic diversity of large numbers of codistributed taxa. However, due to the rarity of such disturbance events over ecological timeframes, the genetic dynamics of multispecies recolonization processes have remained little understood. Here, we use single nucleotide polymorphism (SNP) data from multiple coastal species to track the dynamics of cocolonization events in response to ancient earthquake disturbance in southern New Zealand. Specifically, we use a comparative phylogeographic approach to understand the extent to which epifauna (with varying ecological associations with their macroalgal hosts) share comparable spatial and temporal recolonization patterns. Our study reveals concordant disturbance-related phylogeographic breaks in two intertidal macroalgal species along with two associated epibiotic species (a chiton and an isopod). By contrast, two codistributed species, one of which is an epibiotic amphipod and the other a subtidal macroalga, show few, if any, genetic effects of palaeoseismic coastal uplift. Phylogeographic model selection reveals similar post-uplift recolonization routes for the epibiotic chiton and isopod and their macroalgal hosts. Additionally, codemographic analyses support synchronous population expansions of these four phylogeographically similar taxa. Our findings indicate that coastal paleoseismic activity has driven concordant impacts on multiple codistributed species, with concerted recolonization events probably facilitated by macroalgal rafting. These results highlight that high-resolution comparative genomic data can help reconstruct concerted multispecies responses to recent ecological disturbance.

Genomics Reveals Widespread Ecological Speciation in Flightless Insects.

Recent genomic analyses have highlighted parallel divergence in response to ecological gradients, but the extent to which altitude can underpin such repeated speciation remains unclear. Wing reduction and flight loss have apparently evolved repeatedly in montane insect assemblages and have been suggested as important drivers of hexapod diversification. We test this hypothesis using genomic analyses of a widespread wing-polymorphic stonefly species complex in New Zealand. We identified over 50,000 polymorphic genetic markers generated across almost 200 Zelandoperla fenestrata stonefly specimens using a newly generated plecopteran reference genome, to reveal widespread parallel speciation between sympatric full-winged and wing-reduced ecotypes. Rather than the existence of a single, widespread, flightless taxon (Zelandoperla pennulata), evolutionary genomic data reveal that wing-reduced upland lineages have speciated repeatedly and independently from full-winged Z. fenestrata. This repeated evolution of reproductive isolation between local ecotype pairs that lack mitochondrial DNA differentiation suggests that ecological speciation has evolved recently. A cluster of outlier single-nucleotide polymorphisms detected in independently wing-reduced lineages, tightly linked in an approximately 85 kb genomic region that includes the developmental "supergene" doublesex, suggests that this "island of divergence" may play a key role in rapid ecological speciation. [Ecological speciation; genome assembly; genomic island of differentiation; genotyping-by-sequencing; incipient species; plecoptera; wing reduction.].

Receding ice drove parallel expansions in Southern Ocean penguins.

Climate shifts are key drivers of ecosystem change. Despite the critical importance of Antarctica and the Southern Ocean for global climate, the extent of climate-driven ecological change in this region remains controversial. In particular, the biological effects of changing sea ice conditions are poorly understood. We hypothesize that rapid postglacial reductions in sea ice drove biological shifts across multiple widespread Southern Ocean species. We test for demographic shifts driven by climate events over recent millennia by analyzing population genomic datasets spanning 3 penguin genera (Eudyptes, Pygoscelis, and Aptenodytes). Demographic analyses for multiple species (macaroni/royal, eastern rockhopper, Adélie, gentoo, king, and emperor) currently inhabiting southern coastlines affected by heavy sea ice conditions during the Last Glacial Maximum (LGM) yielded genetic signatures of near-simultaneous population expansions associated with postglacial warming. Populations of the ice-adapted emperor penguin are inferred to have expanded slightly earlier than those of species requiring ice-free terrain. These concerted high-latitude expansion events contrast with relatively stable or declining demographic histories inferred for 4 penguin species (northern rockhopper, western rockhopper, Fiordland crested, and Snares crested) that apparently persisted throughout the LGM in ice-free habitats. Limited genetic structure detected in all ice-affected species across the vast Southern Ocean may reflect both rapid postglacial colonization of subantarctic and Antarctic shores, in addition to recent genetic exchange among populations. Together, these analyses highlight dramatic, ecosystem-wide responses to past Southern Ocean climate change and suggest potential for further shifts as warming continues.

Genomic inference of contemporary effective population size in a large island population of collared flycatchers (Ficedula albicollis).

Due to its central importance to many aspects of evolutionary biology and population genetics, the long-term effective population size (Ne ) has been estimated for numerous species and populations. However, estimating contemporary Ne is difficult and in practice this parameter is often unknown. In principle, contemporary Ne can be estimated using either analyses of temporal changes in allele frequencies, or the extent of linkage disequilibrium (LD) between unlinked markers. We applied these approaches to estimate contemporary Ne of a relatively recently founded island population of collared flycatchers (Ficedula albicollis). We sequenced the genomes of 85 birds sampled in 1993 and 2015, and applied several temporal methods to estimate Ne at a few thousand (4000-7000). The approach based on LD provided higher estimates of Ne (20,000-32,000) and was associated with high variance, often resulting in infinite Ne . We conclude that whole-genome sequencing data offers new possibilities to estimate high (>1000) contemporary Ne , but also note that such estimates remain challenging, in particular for LD-based methods for contemporary Ne estimation.

Genomic signatures of parallel alpine adaptation in recently evolved flightless insects.

Natural selection along elevational gradients has potential to drive predictable adaptations across distinct lineages, but the extent of such repeated evolution remains poorly studied for many widespread alpine taxa. We present parallel genomic analyses of two recently evolved flightless alpine insect lineages to test for molecular signatures of repeated alpine adaptation. Specifically, we compare low-elevation vs. alpine stonefly ecotypes from parallel stream populations in which flightless upland ecotypes have been independently derived. We map 67,922 polymorphic genetic markers, generated across 176 Zelandoperla fenestrata specimens from two independent alpine stream populations in New Zealand's Rock and Pillar Range, to a newly developed plecopteran reference genome. Genome-wide scans revealed 31 regions with outlier single nucleotide polymorphisms (SNPs) differentiating lowland vs. alpine ecotypes in Lug Creek, and 37 regions with outliers differentiating ecotypes in Six Mile Creek. Of these regions, 13% (8/60) yielded outlier SNPs across both within-stream ecotype comparisons, implying comparable genomic shifts contribute to this repeated alpine adaptation. Candidate genes closely linked to repeated outlier regions include several with documented roles in insect wing-development (e.g., dishevelled), suggesting that they may contribute to repeated alpine wing reduction. Additional candidate genes have been shown to influence insect fecundity (e.g., ovo) and lifespan (e.g., Mrp4), implying that they might contribute to life history differentiation between upland and lowland ecotypes. Additional outlier genes have potential roles in the evolution of reproductive isolation among ecotypes (hedgehog and Desaturase 1). These results demonstrate how replicated outlier tests across independent lineages can potentially contribute to the discovery of genes underpinning repeated adaptation.

The genomic footprint of coastal earthquake uplift.

Theory suggests that catastrophic earth-history events can drive rapid biological evolution, but empirical evidence for such processes is scarce. Destructive geological events such as earthquakes can represent large-scale natural experiments for inferring such evolutionary processes. We capitalized on a major prehistoric (800 yr BP) geological uplift event affecting a southern New Zealand coastline to test for the lasting genomic impacts of disturbance. Genome-wide analyses of three co-distributed keystone kelp taxa revealed that post-earthquake recolonization drove the evolution of novel, large-scale intertidal spatial genetic 'sectors' which are tightly linked to geological fault boundaries. Demographic simulations confirmed that, following widespread extirpation, parallel expansions into newly vacant habitats rapidly restructured genome-wide diversity. Interspecific differences in recolonization mode and tempo reflect differing ecological constraints relating to habitat choice and dispersal capacity among taxa. This study highlights the rapid and enduring evolutionary effects of catastrophic ecosystem disturbance and reveals the key role of range expansion in reshaping spatial genetic patterns.

Roe deer on ice: Selection despite limited effective population size through the Pleistocene.

Roe deer (Capreolus spp.) are a little odd. They are one of only a few placental mammals-and the only genus among even-toed ungulates-capable of putting embryonic development "on ice", also known as embryonic diapause (Figure 1). It would seem such an unusual trait is probably the product of natural selection, but a big question is, how does selection for important traits, such as diapause, interact with the historical demography of a species? In a 'From the Cover' article in this issue of Molecular Ecology, de Jong et al. (2020) demonstrate that selection is acting on genes associated with reproductive biology in roe deer, despite heightened genetic drift due to reduced effective population size through the Pleistocene.

SNP analyses reveal a diverse pool of potential colonists to earthquake-uplifted coastlines.

In species that form dense populations, major disturbance events are expected to increase the chance of establishment for immigrant lineages. Real-time tests of the impact of disturbance on patterns of genetic structure are, however, scarce. Central to testing these concepts is determining the pool of potential immigrants dispersing into a disturbed area. In 2016, a 7.8 magnitude earthquake occurred on the South Island of New Zealand. Affecting approximately 100 km of coastline, this quake caused extensive uplift (several metres high), extirpating many intertidal populations, including keystone intertidal kelp species. Following the uplift, we set out to determine the geographic origins of detached kelp specimens which rafted into the disturbed zone. Specifically, we used genotyping-by-sequencing (GBS) approaches to compare beach-cast southern bull-kelp (Durvillaea antarctica and Durvillaea poha) samples to established populations throughout the species' ranges, and thus infer the geographic origins of potential colonists reaching the disturbed coast. Our findings revealed an ongoing supply of diverse lineages dispersing to the newly uplifted coastline, suggesting potential for establishment of "exotic" lineages following disturbance. Furthermore, we found that some drifting individuals of each species came from far-distant regions, some >1,200 km away. These results show that diverse lineages - in many cases from very distant sources - can compete for new space in the wake of an exceptional disturbance event, illustrating the potential of long-distance dispersal as a key mechanism for reassembly of coastal ecosystems. Furthermore, our findings demonstrate that high-resolution genomic baselines can be used to robustly assign the provenance of dispersing individuals.

Biased Inference of Selection Due to GC-Biased Gene Conversion and the Rate of Protein Evolution in Flycatchers When Accounting for It.

The rate of recombination impacts on rates of protein evolution for at least two reasons: it affects the efficacy of selection due to linkage and influences sequence evolution through the process of GC-biased gene conversion (gBGC). We studied how recombination, via gBGC, affects inferences of selection in gene sequences using comparative genomic and population genomic data from the collared flycatcher (Ficedula albicollis). We separately analyzed different mutation categories ("strong"-to-"weak," "weak-to-strong," and GC-conservative changes) and found that gBGC impacts on the distribution of fitness effects of new mutations, and leads to that the rate of adaptive evolution and the proportion of adaptive mutations among nonsynonymous substitutions are underestimated by 22-33%. It also biases inferences of demographic history based on the site frequency spectrum. In light of this impact, we suggest that inferences of selection (and demography) in lineages with pronounced gBGC should be based on GC-conservative changes only. Doing so, we estimate that 10% of nonsynonymous mutations are effectively neutral and that 27% of nonsynonymous substitutions have been fixed by positive selection in the flycatcher lineage. We also find that gene expression level, sex-bias in expression, and the number of protein-protein interactions, but not Hill-Robertson interference (HRI), are strong determinants of selective constraint and rate of adaptation of collared flycatcher genes. This study therefore illustrates the importance of disentangling the effects of different evolutionary forces and genetic factors in interpretation of sequence data, and from that infer the role of natural selection in DNA sequence evolution.

Ecological gradients drive insect wing loss and speciation: The role of the alpine treeline.

Alpine ecosystems are frequently characterized by an abundance of wing-reduced insect species, but the drivers of this biodiversity remain poorly understood. Insect wing reduction in these environments has variously been attributed to altitude, temperature, isolation, habitat stability or decreased habitat size. We used fine-scale ecotypic and genomic analyses, along with broad-scale distributional analyses of ecotypes, to unravel the ecological drivers of wing reduction in the wing-dimorphic stonefly Zelandoperla fenestrata complex. Altitudinal transects within populations revealed dramatic wing reduction over very fine spatial scales, tightly linked to the alpine treeline. Broad biogeographical analyses confirm that the treeline has a much stronger effect on these ecotype distributions than altitude per se. Molecular analyses revealed parallel genomic divergence between vestigial-winged (high altitude) and full-winged (low altitude) ecotypes across distinct streams. These data thus highlight the role of the alpine treeline as a key driver of rapid speciation, providing a new model for ecological diversification along exposure gradients.

Genomics detects population structure within and between ocean basins in a circumpolar seabird: The white-chinned petrel.

The Southern Ocean represents a continuous stretch of circumpolar marine habitat, but the potential physical and ecological drivers of evolutionary genetic differentiation across this vast ecosystem remain unclear. We tested for genetic structure across the full circumpolar range of the white-chinned petrel (Procellaria aequinoctialis) to unravel the potential drivers of population differentiation and test alternative population differentiation hypotheses. Following range-wide comprehensive sampling, we applied genomic (genotyping-by-sequencing or GBS; 60,709 loci) and standard mitochondrial-marker approaches (cytochrome b and first domain of control region) to quantify genetic diversity within and among island populations, test for isolation by distance, and quantify the number of genetic clusters using neutral and outlier (non-neutral) loci. Our results supported the multi-region hypothesis, with a range of analyses showing clear three-region genetic population structure, split by ocean basin, within two evolutionary units. The most significant differentiation between these regions confirmed previous work distinguishing New Zealand and nominate subspecies. Although there was little evidence of structure within the island groups of the Indian or Atlantic oceans, a small set of highly-discriminatory outlier loci could assign petrels to ocean basin and potentially to island group, though the latter needs further verification. Genomic data hold the key to revealing substantial regional genetic structure within wide-ranging circumpolar species previously assumed to be panmictic.

Linked selection and recombination rate variation drive the evolution of the genomic landscape of differentiation across the speciation continuum of Ficedula flycatchers.

Speciation is a continuous process during which genetic changes gradually accumulate in the genomes of diverging species. Recent studies have documented highly heterogeneous differentiation landscapes, with distinct regions of elevated differentiation ("differentiation islands") widespread across genomes. However, it remains unclear which processes drive the evolution of differentiation islands; how the differentiation landscape evolves as speciation advances; and ultimately, how differentiation islands are related to speciation. Here, we addressed these questions based on population genetic analyses of 200 resequenced genomes from 10 populations of four Ficedula flycatcher sister species. We show that a heterogeneous differentiation landscape starts emerging among populations within species, and differentiation islands evolve recurrently in the very same genomic regions among independent lineages. Contrary to expectations from models that interpret differentiation islands as genomic regions involved in reproductive isolation that are shielded from gene flow, patterns of sequence divergence (d(xy) and relative node depth) do not support a major role of gene flow in the evolution of the differentiation landscape in these species. Instead, as predicted by models of linked selection, genome-wide variation in diversity and differentiation can be explained by variation in recombination rate and the density of targets for selection. We thus conclude that the heterogeneous landscape of differentiation in Ficedula flycatchers evolves mainly as the result of background selection and selective sweeps in genomic regions of low recombination. Our results emphasize the necessity of incorporating linked selection as a null model to identify genome regions involved in adaptation and speciation.

Sex-biased gene expression, sexual antagonism and levels of genetic diversity in the collared flycatcher (Ficedula albicollis) genome.

Theoretical work suggests that sexual conflict should promote the maintenance of genetic diversity by the opposing directions of selection on males and females. If such conflict is pervasive, it could potentially lead to genomic heterogeneity in levels of genetic diversity an idea that so far has not been empirically tested on a genomewide scale. We used large-scale population genomic and transcriptomic data from the collared flycatcher (Ficedula albicollis) to analyse how sexual conflict, for which we use sex-biased gene expression as a proxy, relates to genetic variability. Here, we demonstrate that the extent of sex-biased gene expression of both male-biased and female-biased genes is significantly correlated with levels of nucleotide diversity in gene sequences and that this correlation extends to diversity levels also in intergenic DNA and introns. We find signatures of balancing selection in sex-biased genes but also note that relaxed purifying selection could potentially explain part of the observed patterns. The finding of significant genetic differentiation between males and females for male-biased (and gonad-specific) genes indicates ongoing sexual conflict and sex-specific viability selection, potentially driven by sexual selection. Our results thus indicate that sexual antagonism could potentially be considered as one viable explanation to the long-standing question in evolutionary biology of how genomes can remain so genetically variable in face of strong natural and sexual selection.

Covariation in levels of nucleotide diversity in homologous regions of the avian genome long after completion of lineage sorting.

Closely related species may show similar levels of genetic diversity in homologous regions of the genome owing to shared ancestral variation still segregating in the extant species. However, after completion of lineage sorting, such covariation is not necessarily expected. On the other hand, if the processes that govern genetic diversity are conserved, diversity may potentially covary even among distantly related species. We mapped regions of conserved synteny between the genomes of two divergent bird species-collared flycatcher and hooded crow-and identified more than 600 Mb of homologous regions (66% of the genome). From analyses of whole-genome resequencing data in large population samples of both species we found nucleotide diversity in 200 kb windows to be well correlated (Spearman's ρ = 0.407). The correlation remained highly similar after excluding coding sequences. To explain this covariation, we suggest that a stable avian karyotype and a conserved landscape of recombination rate variation render the diversity-reducing effects of linked selection similar in divergent bird lineages. Principal component regression analysis of several potential explanatory variables driving heterogeneity in flycatcher diversity levels revealed the strongest effects from recombination rate variation and density of coding sequence targets for selection, consistent with linked selection. It is also possible that a stable karyotype is associated with a conserved genomic mutation environment contributing to covariation in diversity levels between lineages. Our observations imply that genetic diversity is to some extent predictable.

Divergence in gene expression within and between two closely related flycatcher species.

Relatively little is known about the character of gene expression evolution as species diverge. It is for instance unclear if gene expression generally evolves in a clock-like manner (by stabilizing selection or neutral evolution) or if there are frequent episodes of directional selection. To gain insights into the evolutionary divergence of gene expression, we sequenced and compared the transcriptomes of multiple organs from population samples of collared (Ficedula albicollis) and pied flycatchers (F. hypoleuca), two species which diverged less than one million years ago. Ordination analysis separated samples by organ rather than by species. Organs differed in their degrees of expression variance within species and expression divergence between species. Variance was negatively correlated with expression breadth and protein interactivity, suggesting that pleiotropic constraints reduce gene expression variance within species. Variance was correlated with between-species divergence, consistent with a pattern expected from stabilizing selection and neutral evolution. Using an expression PST approach, we identified genes differentially expressed between species and found 16 genes uniquely expressed in one of the species. For one of these, DPP7, uniquely expressed in collared flycatcher, the absence of expression in pied flycatcher could be associated with a ≈20-kb deletion including 11 of 13 exons. This study of a young vertebrate speciation model system expands our knowledge of how gene expression evolves as natural populations become reproductively isolated.

Genomics reveals repeated landlocking of diadromous fish on an isolated island.

Landlocking of diadromous fish in freshwater systems can have significant genomic consequences. For instance, the loss of the migratory life stage can dramatically reduce gene flow across populations, leading to increased genetic structuring and stronger effects of local adaptation. These genomic consequences have been well-studied in some mainland systems, but the evolutionary impacts of landlocking in island ecosystems are largely unknown. In this study, we used a genotyping-by-sequencing (GBS) approach to examine the evolutionary history of landlocking in common smelt (Retropinna retropinna) on Chatham Island, a small isolated oceanic island 800 kilometres east of mainland New Zealand. We examined the relationship between Chatham Island and mainland smelt and used coalescent analyses to test the number and timing of landlocking events on Chatham Island. Our genomic analysis, based on 21,135 SNPs across 169 individuals, revealed that the Chatham Island smelt was genomically distinct from the mainland New Zealand fish, consistent with a single ancestral colonisation event of Chatham Island in the Pleistocene. Significant genetic structure was also evident within the Chatham Island smelt, with a diadromous Chatham Island smelt group, along with three geographically structured landlocked groups. Coalescent demographic analysis supported three independent landlocking events, with this loss of diadromy significantly pre-dating human colonisation. Our results illustrate how landlocking of diadromous fish can occur repeatedly across a narrow spatial scale, and highlight a unique system to study the genomic basis of repeated adaptation.