When Isolation Does Not Lead to Genetic Purging
My commentary on a beluga whale study showing how long-term isolation and genetic drift can raise mutation load without leaving a clear signature of genetic purging.
Orton and colleagues compared whole genomes from four eastern Canadian beluga populations, focusing on the endangered Saint Lawrence Estuary population. Relative to the other populations, these belugas had lower nucleotide diversity, more and longer runs of homozygosity, and more homozygous variants predicted to be deleterious. Demographic modelling suggests that the population has been isolated for roughly 48,000–50,000 years.
What I find most important is what the authors did not find: a mutation-impact-specific signal of genetic purging. Small, isolated populations are sometimes expected to expose recessive deleterious alleles to selection and gradually remove them. Here, sustained small effective population size appears instead to have allowed genetic drift to outpace purifying selection. Persistence alone, therefore, is not evidence that a population has successfully shed its genetic load.
The comparison of eight deceased neonates with nine adults adds a provocative, but preliminary, layer. Neonates were not more inbred overall, yet they were more homozygous at sites carrying variants predicted to be deleterious. Four exploratory gene modules were enriched for cellular responses to inorganic substances, raising a possible connection between genetic susceptibility and pollution. That interpretation needs caution: the sample was small, variant effects were predicted computationally, and contaminant exposure was not measured in the individuals.
Why This Matters for My Sable Island Horse Project
My work links genomic pedigrees and runs of homozygosity in Sable Island horses with long-term records of survival, reproduction, morphology, parasites, microbiome and environment. The beluga study reinforces why the length and distribution of runs of homozygosity matter: long runs can reflect recent mating between close relatives, whereas shorter runs may retain the older imprint of genetic drift and historical inbreeding. A single genome-wide inbreeding value can miss that history.
It also sharpens the questions I can ask with the Sable Island data. Rather than assuming that an isolated population has purged harmful variation—or treating a genomic annotation as proof of a fitness effect—I can test whether homozygosity in particular genomic regions predicts observed differences in health, reproduction or survival. The beluga paper offers a useful comparative framework, but not a result that can be transferred directly between species.