White-blooded Antarctic icefishes are an example of extreme biological specialization both to the chronic cold of the Southern Ocean and to life without hemoglobin. As a result, icefishes display derived physiology that limits them to the cold and highly oxygenated Antarctic waters. Despite these constraints, a single icefish species, the pike icefish Champsocephalus esox, is presently found in temperate South American waters. To study the genetic mechanisms underlying temperate adaptation in this species, we generated chromosome-level genome assemblies of both C. esox and its Antarctic sister species, Champsocephalus gunnari. The C. esox genome is similar in structure and organization to that of Antarctic icefishes. However, we observe evidence of chromosomal rearrangements, some of which coincide with regions of elevated genetic divergence in pike icefish populations. Our results show several key biological pathways under selection, including genes related to the mitochondria, iron transport, and light sensing, highlighting candidates behind temperate adaptation in this species. The C. esox genome also shows antifreeze glycoprotein pseudogenization, likely due to relaxed selection following ancestral escape from Antarctica. While the organization of the canonical antifreeze glycoprotein locus is conserved in both C. esox and C. gunnari, both species show a translocation of antifreeze genes, previously unobserved in notothenioids. Our results present the first genomic characterization of a secondarily temperate notothenioid to date and serve as a basis for understanding the group’s adaptive potential against a rapidly changing Antarctic environment.

Genomes for C. esox and C. gunnari were assembled using flye v2.5. The E. maclovinus genome was assembled with wtdbg2. All genomes were scaffolded with juicer v1.6.2. Annotation generated using BRAKER v2.1.6 and TSEBRA v1.0.1. A more in depth description is available on the README file and in the publication (Rivera-Colón et al. 2022).

All files are gzipped, but are otherwise standard bioinformatic formats (i.e., FASTA for genome assembly and coding/amino acid sequences), GTF for annotation, AGP for scaffolding). See links for a description of the FASTA, and GTF, and AGP file formats.

File format specifications

*.fa

*.agp

*.gtf

*.cds.fa

*.protein.fa

*Does not include the gzipped compression suffix (.gz).