I could write a small book about this but will try to stick to a few pertinent points.
Sticking to the Dales for now it seems perfectly obvious that fishes in caves don't get there by swimming upstream.
Neither do they get there as eggs. These are far too delicate and would not survive the journey. Neither, probably, would they obtain sufficient food to grow and mature into the sizes of fish we see. So we look to Occam's razor and suggest that in all probability they either swim downstream, or, as likely or more likely, they are moved d/s against their will in floods. And now we arrive at the crux of the problem, and its most interesting aspect. The fishes in the caves come from populations in the upland becks that drain into the caves. We see them in the caves but they are not in fact of much biological interest (I will mention some aspects that are in a separate post), as they are unwilling strays doomed to starve.
[Interestingly the Ingleborough Cave specimen mention in the video had been feeding but Avernus and Pluto are atypical environments in Yorkshire caves with low current with sediment dropout and lots of Gammarus and stones fly nymphs etc. Also occurs in sumps of course but there are fewer divers than cavers so less often reported. Active Dales vadose streamways are not high in trout food and are too fast.]
The upland beck populations are very significant but, as far as I know, have never been studied. The obvious questions are:
What is their origin (post glacial colonisation but by what mechanism, or placed there as a food supply by man).
How do they maintains their populations against severe and constant downstream losses that cannot be made good with upstream migration?
There is a well known polymorphism in an enzyme, lactate dehydrogenase, which has two forms, Ldh5(100) and Ldh5(90) with the former ancestral to the latter and with the former common in post-glacial times and the latter coming from more recent colonisation from the south. If the first of the two possibillities above then the 100 form should be present in these becks. Of course we don't these days need to rely on enzyme polymorphism as we could just look at any number of DNA markers like mitochondrial COI. But none of this has ever been done and we have no idea.
It would be relatively simple to get to the bottom of this story with some DNA work but I don't have the resources, or the energy, to do it myself.
Graham Proudlove
