Limited Introgression between Rock-Wallabies with Extensive Chromosomal Rearrangements

Abstract

Chromosome rearrangements can result in the rapid evolution of hybrid incompatibilities. Robertsonian fusions, particularly those with monobrachial homology, can drive reproductive isolation amongst recently diverged taxa. The recent radiation of rock-wallabies (genus Petrogale) is an important model to explore the role of Robertsonian fusions in speciation. Here, we pursue that goal using an extensive sampling of populations and genomes of Petrogale from north-eastern Australia. In contrast to previous assessments using mitochondrial DNA or nuclear microsatellite loci, genomic data are able to separate the most closely related species and to resolve their divergence histories. Both phylogenetic and population genetic analyses indicate introgression between two species that differ by a single Robertsonian fusion. Based on the available data, there is also evidence for introgression between two species which share complex chromosomal rearrangements. However, the remaining results show no consistent signature of introgression amongst species pairs and where evident, indicate generally low introgression overall. X-linked loci have elevated divergence compared with autosomal loci indicating a potential role for genic evolution to produce reproductive isolation in concert with chromosome change. Our results highlight the value of genome scale data in evaluating the role of Robertsonian fusions and structural variation in divergence, speciation, and patterns of molecular evolution.

Keywords: Robertsonian fusion; chromosome rearrangement; introgression; marsupial; speciation.

Fig. 1.

Karyotypes of six species of rock-wallaby (Petrogale) from north-eastern Australia which form part of the penicillata complex, including a map of sample locations and distributions for each species. The color of individuals on the map is species specific and linked to the karyotypes. Robertsonian fusions are highlighted by chromosome numbers, inversions are denoted as (i), and polymorphic karyotypes are shown for chromosome 2 and the X chromosome. The SAM (P. sharmani, P. assimilis, P. mareeba) complex is outlined. The outcomes of experimental hybrid crosses are outlined for both males and females: X = infertile; SF = subfertile; ? = unknown. Based on simple versus complex heterozygote formation between species pairs, we have predicted whether we expect to see introgression or not. For simple heterozygotes which form trivalents during meiosis, we expect introgression could be present as chromosomes could segregate properly during meiosis of hybrids. For complex heterozygotes which from chains of four or five chromosomes, we expect to see no introgression.

Fig. 2.

Summary of population genomic (a, b) and phylogenomic (c, d) results based on the DArT and phased exon data sets, respectively. (a) PCoA results for the five penicillata complex taxa examined in this study (not including P. penicillata used as an outgroup). Species are colored individually and the symbols match species identification from (c). The PCoA (left) includes four known F1 hybrids (designated by +) between P. godmani and P. mareeba. Structure results for these same SNPs and individuals (right) are grouped into three clusters, with P. assimilis, P. mareeba, and P. sharmani (SAM) forming one group despite K = 4 clusters supported. (b) DAPC result for the SAM group based on 19 principal components (PCAs) and two discriminant functions (DAs), colored by species (left). Structure results for just the SAM group indicate two populations and are colored by cluster (right). The lines on the right represent the species boundaries. (c) Phylogenetic species tree of the penicillata complex based on the multispecies coalescent analysis in SVDquartets. Species colors and symbols match those in the population analyses (a, b) and the bootstrap support values are located on branches. (d) Network for individuals estimated using a distance-based approach in SplitsTree (Neighbor-Net). The regions of the network where multiple branches connect individuals and groups indicate reticulation.

Fig. 3.

(a) Average pairwise divergence (d_XY) between P. inornata and P. sharmani, P. assimilis, P. mareeba (SAM) species for comparisons between rearranged (R: 5,6,9,10) and nonrearranged (NR: 1,2,7,8) chromosomes, and between X and autosomal (A) loci. Tests for significant differences between the average d_XY were calculated using t-tests and significant values (P < 0.05) are marked with an asterisk (*). (b) Average d_XY within the SAM group only, for R versus NR and X versus A.

Fig. 4.

(a) TreeMix results which support a single introgression event from the ancestor of P. godmani into P. mareeba. The migration rate is intermediate for this introgression event. (b) TreeMix results from a model of two introgression events, building off of the m = 1 model (a) and including an introgression event with high migration from P. assimilis into P. inornata.