My review is attached in PDF format

I waive anonymity. Stuart J.E. Baird Studenec 3 July 2020

Overview

The manuscript is much improved. The shift in emphasis away from introgression comets and toward the decay of hybrid zones is not superficial, and really works for this amazing dataset. The degree to which hybrid zone decay has been understudied is striking. While single-contact examples of decay have been reported (eg Macholan et al 2008*), the large number of replicate contacts in the current study is what is really necessary to move our understanding of patterns of decay forward. The shift in emphasis has, naturally, introduced a new set of typos and some text sections that could be clearer, but the effort necessary to clear these up should be trivial.

*M. Macholan et al., Genetic conflict outweighs heterogametic incompatibility in the mouse hybrid zone? BMC Evolutionary Biology 8, (2008).

Typos and minor issues

L100 “dies’ -> “dyes”
L152 “geographically close from it” -> “geographically close to it”

L202-210 “Each hybrid zone was defined by parent 1 (P1), parent 2 (P2) and central populations (Table 3). Each parental population was classified as either “local” or “distant”, according to observed levels of introgression or geography. A local parental population is defined as a population close to the focal hybrid zone and being potentially subject to recent gene flow from the hybrid zone. A distant parental population is defined as a population distant from the hybrid zone, which has been less influenced by recent gene flow from it but can still have a substantial rate of established introgression due to past contacts. In a few cases a local parental population was not available in our sampling design and in these cases, a distant population was used instead (Table 3).”

Here ‘close’ and ‘distant’ need to be defined. Their natural units would be with respect to the per generation dispersal scale of mussels. A more arbitrary thresholding in eg great circle Kms would also be fine (eg close <200kms; distant >1000kms), but currently there seems to be no definition at all. (See also lines 354-359).

L211 “computation of hybrid indexes”-> ““estimation of hybrid indexes”
L235-6 “Both analyses considers” -> “Both analyses consider”
L262-3 “only the marker 147” -> “only marker 147”
L274 “with Structure with K = 8”-> “with Structure at K = 8”
L281 “Group E contains a small group” -> “Group E contains a small set”
L283 “an admixture, largely between” -> “admixture largely between”
L309-310 “Genetic differences between patches of parental populations close or distant to hybrid zones” -> “Genetic differences between patches of parental populations close to or distant from hybrid zones”
L319 “The unbalance between populations” -> “The discordance between populations”
L327 “done by Fraïsse, Belkhir, et al., 2016).” -> “done by (Fraïsse, Belkhir, et al., 2016).”
L403-404 “an analysis treating as admixed the North-European M. edulis individuals in the Russian populations” -> “an analysis treating the North-European M. edulis individuals as admixed in the Russian populations”
Fig 5 caption “See Table 3 for a list of used populations.” -> “See Table 3 for a list of the populations shown.”
L448-450 “resulting in a higher rate of shared and high frequency private polymorphisms than expected at first sight for well isolated and low Ne species.” Unclear wording: what is a ‘shared and high frequency private polymorphism’?
L466-467 “The biogeography of the complex also allowed us to study introgression of a very particular kind. More specifically, the” -> “The biogeography of the complex also allowed us to study introgression of a very particular kind: The”
L471 “France, are separated” -> “France are separated”
L500 “(2014, July)” -> “(2014)”
L501 “such as in” -> “such as”
L505 “the occasional failures of our cline analyses” Refer back to some examples of these failures to help the reader follow. Currently nothing up to this point in the manuscript has been clearly flagged as a failure except the failure of SNPs to be diagnostic over large samples at L432.
L524 “where” -> “were”
L535 “Mediterranean, is consistent” -> “Mediterranean is consistent”
L541-43 “This is because, given its position enclosed in the mosaic, it is probable that this population only became established after the last post-glacial period.” -> “This is probably because, given its position enclosed in the mosaic, this population only became established after the last post-glacial period.”
L545 “Cherbourg, represent” -> “Cherbourg represent”
L564 “the last glacial maxima” -> “the last glacial maximum”
L566 “in a refugia” ->“in a refugium”
L581 “not used” -> “not taken into account”