Thanks to the OP for the interesting research finding.
So the finding is that hatchery fish have basically the same genome as wild fish, but the level of gene expression has been altered in hundreds of genes as a result of hatchery rearing, and that diferential gene expression occurs quickly (within one generation) and is passed down as an inheritable trait (epigenetics).
In this case, the investigators show that genes involved in metabolism, wound healing and immune response are up-regulated presumably as a result of the high-density hatchery rearing where injury, attacks from other fish, and infection are far more prevalent than in the wild. It would seem like a pretty nifty adaptation, but there is no such thing as a free lunch, so as a result of those changes, they lose some reproductive fitness in other undetermined ways and ultimately only reproduce at 85% the success rate of wild fish in the wild.
One might be tempted to conclude that these hatchery fish have now become weaker or inferior to wild fish but they cite another study which concludes hatchery fish have almost twice the reproductive success as wild fish when spawned in captivity. So really, these fish have simply responded to new selective pressure placed on them at the time (hatchery rearing) and adapted to the new environment, though the speed at which it occurred is pretty amazing. Of course a hatchery is 'unnatural' compared to a wild stream so if these adaptions ends up reducing fitness in an environment they they need to survive in later, then that's not desirable. If these changes end up being slower to revert than they took to appear and last through multiple generations, then the implications are serious.
My question is (and I didn't read all the reports) is if they know how long it takes to undo the changes. If I'm reading it right, it isn't that the genetic code is different, just different items activated. What I would be curious seeing if it a hatchery fish then spawns with a wild fish in the wild, do those activated codes transfer to the offspring or does growing up in the wild undo that? Seems like some of these activations could be because of how the fish grew up and wouldn't necessarily effect offspring long term.
This is a good question as the authors dont claim that these changes are permanent or will not somehow otherwise change again over time. These fish have incredible adaptability so a return to natural selective pressures may result in hatchery fish eventually regaining 'wild' physiology again, and maybe relatively quickly (pure speculation on my part). However the authors do make his cautionary hypothesis: "As subsequent generations of domestication accrue, we speculate that the regulatory changes to expression become codified with gradual and more targeted shifts in allele frequencies". Hatcheries use wild fish for brood stock (someone correct me if I'm wrong), so I imagine there have been concerns about multi-generation effects of hatchery only lineage for some time.
Relating to something more similar, take a child raised in a third world nation, extremely poor, and a child raised wealthy in BC. Very different needs and way to grow up. If both moved to the middle of Alberta and lived similar lives after age 20, they would still be different at 80. However, even without any contact, these two people's kids, both growing up in the same town in the middle of Alberta are going to be much more similar than their parents. If both those live and die in the same town and have kids, that generation is going to be closer to each other than they are to their grandparents.
Actually this is essentially what they did. To use your analogy, they took
salmon steelhead raised in a third world nation (hatchery) and
salmon steelhead raised in wealthy BC (wild stream) that both moved to the ocean (rural Alberta) as teens then lived similar lives from that point on. When they returned to spawn, they found that when hatchery fish were bred with other hatchery fish, their progeny's gene expression was notably different than when wild fish were bred with other wild fish, but that differential expression was not present when hatchery fish were bred with wild fish and vice versa. As to what happens when second generation hatchery fish breed with each other and if any of the upregulated genes start to revert to the 'wild' state seems to be anyone's guess at this point.
It's an intriguing finding and hopefully such interesting research will continue and maybe might lead to changes that can be made in hatchery rearing practices to improve survivability and sustainability.
correction: the study used steelhead, not salmon