In recent years, the rise of non-model pop genomics paired with the development of robust and powerful tests of gene flow [both genome-wide (e.g. f3 tests, D-stats, and rolloff [read here] and treemix) and across in genomic regions (e.g. hapmix, finestructure)] has allowed us to explore the extent of gene flow in in diverged species pairs. While results-to-date likely suffer from biases in ascertainment (i.e. we’re more prone to look for gene flow in cases where it’s likely), and publication (no gene flow might not be a catchy title or result), the weight of evidence suggests that we shouldn’t be shocked by gene flow between recently diverged species/populations . At first blush, evidence for gene flow between species may lead us to think that ‘speciation with gene flow’ [e.g. Feder et. al.)] is common; however, there is a difference between speciation with gene flow vs speciation and gene flow.
The prior describes the case where speciation occurs even with ongoing gene exchange between diverging species-to-be, the later describes minor exchange of genes between good species after speciation is complete. Like the contrast between ‘genomic islands of speciation’ and ‘genomic islands and speciation’, concerns about ‘speciation with gene flow’ and ‘speciation and gene flow’ reflect the extreme caution and strong evidence needed to go from pattern to process in population genomic analyses. Interested readers should check out a recent exciting twitter discussion between Graham Coop, Matt Hahn, Mohamed Noor and Josh Schraiber (below). The take home message is that, in population genomic analyses, we must be careful not to let our excitement about a cool potential process to get ahead of what we can infer from a pattern.
@3rdreviewer Nope. It always confuses me why people think this. I think most popgen analysis done says almost nothing about this.
— Graham Coop (@Graham_Coop) August 18, 2014
@jgschraiber @3rdreviewer the RI loci could easily evolve rapidly in brief lulls of gene flow between even closely connected populations.
— Graham Coop (@Graham_Coop) August 18, 2014
@3rdreviewer @Graham_Coop I think it’s easy to use SWGF and mean something more mundane on accident, perhaps this is commmon
— Joshua G. Schraiber (@jgschraiber) August 18, 2014
.@Graham_Coop Suggested guidelines for when people can claim this? I would exclude any gene flow via secondary contact–maybe too harsh?
— Matthew Hahn (@3rdreviewer) August 18, 2014
@Graham_Coop @jgschraiber One issue is that you don’t want to throw out reinforcement with the SWGF bathwater. Tough call.
— Matthew Hahn (@3rdreviewer) August 18, 2014
@Graham_Coop @3rdreviewer @jgschraiber Related is whether speciation was MOSTLY done, and later bit of gene flow happens- not SWGF, IMHO.
— Mohamed Noor (@mafnoor) August 18, 2014
@mafnoor @Graham_Coop @jgschraiber Agreed. But now we’re back to the problem of knowing when GF happened and how much RI there was. 50.1%?
— Matthew Hahn (@3rdreviewer) August 18, 2014
Discussion
No comments yet.