your comment is a little hard to follow, but here's a few specific responses or additions.

> The wider ecosystem effects are not really known.

that's not really true. it is pretty well known that there are major detrimental effects of fish farms. of course we are still learning more and more, but we know at least something. salmon farms tend to take out good stuff from the ocean (fish) and return bad stuff (disease, waste, genetic changes, etc.).

> There are species we don't know exist going extinct, almost certainly, because of commercial fishing.

there are species we know about going extinct due in part to commercial fishing. for example, the southern resident orcas have been struggling heavily due to commercial fishing, pollution, poisoning, capture, river damns, salmon farms, and more. there are very clear actionable things that can be done to save them, but none are even close to being acted upon. it is possible the population there has passed a point of no return such that that particular population will go extinct. they have struggled to even birth new orcas, and due to captures 30-50 years ago, they don't have the leadership they are used to. their behavior has already began to showcase how hard it is for them to even find salmon, as the normally tight knit pods have been seen breaking up to find salmon. there was a month in 2018 where almost none of the pod was seen for an entire month by whale watchers and scientists, an event that hasn't happened in decades or ever since they've started being observed.

(I know there are concerns with many commercial fish farming operations today) Couldn't sustainable fish farms, done right, solve this?

Tree farms seem to work, iirc the amount of forested land in the US actually started increasing in the past few years.

It's true that a lot of aquaculture is not great environmentally. There's a spectrum.

Two points though.

(1) Even some of the worst practices are at least relatively localized. For a given tonage of fish, commercial fishing fleets affect far larger areas, with the most productive areas usually most affected. I think practices need to improve a lot, but even with the current below-par methods... Imagine a ranch with terrible environmental practices. They indiscriminately poison predators, overgraze, use bad fertilizers, run effluent into streams... That 1,000 acre ranch is surrounded by 100,000 acres of wilderness (or good tree farms). The ranch is bad, but enough wilderness can dilute its effects.

(2) A lot depends on the species of fish. With land agriculture, we tend to farm a handful of species that lend well to farming. Before farming, people ate different species. Still, agriculture didn't produce bear meat, otter or some other hard-to-raise species to suit the preferences of hunter-gatherers. With aquaculture, we're still often trying to farm otter and bear.

Tilapia & carp (for example) have a long history of aquaculture. It's cheap and efficient and can be done in environmentally sound ways. The problem is that salmon (raised in environmentally unsound oceanic pens) sells for 5X per kg.

To me that means regulation can be sharp. If salmon, cod or tuna are replaced by more agriculturally enclined species... that's a reasonable compromise, IMO. We can have fish, just not any fish.

Fish farms have their own problems especially open-pen ones they use for things like salmon.

http://www.alexandramorton.ca/salmon-confidential-booklet/

Tree farms are also not a forest. There's no real ecosystem there. They're just rows of commercially viable trees planted and left until they're old enough to harvest.

... which provides an alternative to cutting down the forests for wood, which is why forested land is increasing.

This is not the case, at least in Oregon. Our tree farms are generally healthy, left undisturbed for 40 or so years between harvests, and are parceled out so that entire areas aren't harvested at once. All state run timber sales are surveyed to ensure there are no endangered species present, and, are tagged to reasonably protect against runoff issues (can't cut all the way up to a creek for example). The straight row plantations you may have been referring to are generally for wood pulp, AKA toilet paper.

Note: I am not a ecogeek, I just have a brother in law who is a wildlife biologist and works at a firm that does this stuff for the state, and, I regularly recreate (dirt bike, mountain bike, deer hunt) in a few different "working forests" here in oregon. These working forests aren't pristine old growth, but then again I wouldn't dirt bike etc in a virgin forest either.