GK,
thanks for posting that article, very interesting. It would suggest that salmon may stop feeding some way offshore. However, there are a lot of variables in play, and we do know that a lot of salmon don't just make a straight line to their destination and often move significant distances up and down the coast before entering a river (leave aside the 10% or so of 'strayers')(
see here for more on navigation). The many Scottish fixed coastal nets operated on that basis, catching stock from a variety of rivers.
Another issue is the speed of the salmon's digestive system, which is as formidable as its catch and 'swallow' mechanism, and clearly evolved to operate in tandem to fuel its explosive growth (you can't increase your body mass twentyfold in 12 months with a leisurely approach to eating or digesting). The 'swallow' mechanism is automatic: as soon as a salmon takes a prey fish onto the serrated 'tongue' it is reflexively passed into the highly elastic primary stomach. This allows a salmon to take several prey fish in a single pass through a school, then turn and repeat as often as availability dictates. The primary stomach of a 10lbs 2SW salmon can take 20+ sandeels. I haven't done dissections on sea-caught salmon, but plenty on bass, which are much slower growing but nonetheless clear their similarly capacious stomach within the space of a tide (i.e. 6 hours), with only particles and sludge remaining. On that basis the fact that the stomach is empty doesn't give is much guidance as to how long before the salmon stopped feeding.
In any event there is the question of food availability. The supply of prey isn't continuous and feeding is episodic and furious. But finding prey isn't random: if it was salmon would rapidly starve as its chance of colliding with a 50 metre blob of sprats in a 1km square of ocean is less than 0.6%. Finding prey requires
a team search effort by a school of salmon, using their sensors (primarily vibration via the lateral line) in concert. Operating in this mode the chance of success rises over 25%, depending on the 'noisiness' of the prey species. So what the research may show is not necessarily that the individual salmon has lost its appetite, but rather that by travelling in a roughly straight line rather than ranging and hunting, the availability of food may have sharply declined.
There are also interesting questions on outward migration destinations and return routes. Not all salmon from a river go to the same place, which would create serious survival and evolutionary risks (see Norwegian research - some Alta fish turn left, others right). The AST's 'Lost at Sea' project appeared hooked on the notion that all Scottish salmon went to the Norwegian Basin: no doubt lots do, because the prevailing currents can lead them that way, but I doubt that it's all. In that regard I find the case of Yorkshire grilse interesting. In recent years catches have been so poor that we can't draw inferences, but in the period 2010-20 we saw enough to make some inferences. First there's the question of size: they're much bigger than their Scottish peers, in the range 3 1/2 - 6lbs. Second, there's condition: grilse are inefficient swimmers (poor mass to cross section ratio) and lose proportionately more weight in migration than salmon. Come September, plump Findhorn grilse are a rarity (I've caught far more there than anywhere else), whereas Ure grilse tend to be in fuller condition. I have no factual explanation for this. However, I sometimes wonder whether some of the Ure grilse never leave the North Sea, a strategy that would offer earlier access to food as a departing smolt and a shorter and easier return trip.