Salmonids have a remarkable ability to form sympatric morphs after postglacial colonization of freshwater lakes. These morphs often differ in morphology, feeding, and spawning behaviour. Here, we explored the genetics of morph differentiation by establishing a high-quality, annotated reference genome for the Arctic charr and using this for population genomic analysis of morphs from two Norwegian and two Icelandic lakes. The four lakes represent the spectrum of genetic differentiation between morphs from one lake with no genetic differentiation between morphs, implying phenotypic plasticity, to two lakes with locus-specific genetic differentiation, implying incomplete reproductive isolation, and one lake with strong genome-wide divergence consistent with complete reproductive isolation. As many as 12 putative inversions ranging from 0.45 to 3.25 Mbp in size segregated among the four morphs present in one lake, Thingvallavatn, and these contributed significantly to the genetic differentiation among morphs. None of the putative inversions was found in any of the other lakes, but there were cases of partial haplotype sharing in similar morph contrasts in other lakes. The results are consistent with a highly polygenic basis of morph differentiation with limited genetic parallelism between lakes. The results support a model where morph differentiation is first established through phenotypic plasticity, leading to niche expansion and separation. This is followed by gradual development of reproductive isolation, locus-specific differentiation, and eventually complete reproductive isolation and genome-wide divergence. A major explanation for salmonids' ability to diversify into multiple sympatric morphs is likely their genome complexity from ancient whole genome duplication, which enhances evolvability.

Biodiversity resilience relies on genetic diversity, which sustains the persistence and evolutionary potential of organisms in dynamic ecosystems. Genomics is a powerful tool for estimating genome-wide genetic diversity, offering precise and accurate estimates of the status and trajectory of genetic diversity within species and populations. However, the widespread integration of genomic information into biodiversity conservation and management efforts faces challenges due to a lack of standardised genome-wide data generation methods and applications. The heterogeneity of approaches can make it difficult to consistently interpret the results and clearly communicate key information to stakeholders such as practitioners and decision-makers. To begin to address these challenges, the European Reference Genome Atlas (ERGA) promotes the standardisation of methodologies for high-quality reference genome sequencing and analysis as part of the global network of the Earth BioGenome Project. ERGA is also proactively developing best practices to engage stakeholders in biodiversity genomics research, starting with examining case studies and conducting mapping efforts to familiarise researchers with pathways to effective engagement. An emerging theme is the researchers’ experience of variable perceptions amongst stakeholders of the value and utility of reference genomes and genomics data in biodiversity conservation and management. Addressing this issue calls for consensus on standardised genome-wide data generation methods and applications that will help to deliver the highest standards for accuracy, interpretability, and comparability. We believe converging on consensus methods standardisation is essential for fostering the stakeholder trust and confidence required to successfully promote widespread adoption of genome-wide genetic diversity assessments in biodiversity conservation and management.

The Asian honeybee, Apis cerana, is an ecologically and economically important pollinator. Mapping its genetic variation is key to understanding population-level health, histories, and potential capacities to respond to environmental changes. However, most efforts to date were focused on single nucleotide polymorphisms (SNPs) based on a single reference genome, thereby ignoring larger-scale genomic variation. We employed long-read sequencing technologies to generate a chromosome-scale reference genome for the ancestral group of A. cerana. Integrating this with 525 resequencing datasets, we constructed the first pan-genome of A. cerana, encompassing almost the entire gene content. We found that 31.32% of genes in the pan-genome were variably present across populations, providing a broad gene pool for environmental adaptation. We identified and characterized structural variations (SVs) and found that they were not closely linked with SNP distributions, however, the formation of SVs was closely associated with transposable elements. Furthermore, phylogenetic analysis using SVs revealed a novel A. cerana ecological group not recoverable from the SNP data. Performing environmental association analysis identified a total of 44 SVs likely to be associated with environmental adaptation. Verification and analysis of one of these, a 330 bp deletion in the Atpalpha gene, indicated that this SV may promote the cold adaptation of A. cerana by altering gene expression. Taken together, our study demonstrates the feasibility and utility of applying pan-genome approaches to map and explore genetic feature variations of honeybee populations, and in particular to examine the role of SVs in the evolution and environmental adaptation of A. cerana.