I recently saw a study claiming:

Distributed mosquito nets are intended to be used for malaria protection, yet increasing evidence suggests that fishing is a primary use for these nets, providing fresh concerns for already stressed coastal ecosystems.
  —The perverse fisheries consequences of mosquito net malaria prophylaxis in East Africa (Jones and Unsworth, 2019)

Mosquito nets are harmful fishing tools because (a) they're insecticide-treated and (b) with such small holes you catch a lot of immature fish before they've had a chance to reproduce. But how harmful this practice is overall depends on how widespread it is.

The paper only tries to answer the question "is using a mosquito net for fishing a good idea," and to show that it's a "primary use" of distributed bednets they cite The use of mosquito nets in fisheries: A global perspective (Short et al. 2018). This paper, however, doesn't describe the kind of research that could back up their claim.

They ran on online survey which they left open to anyone to respond for two months in 2015. They got 94 responses from people saying they'd seen mosquito net fishing and 36 from people saying they hadn't. This sort of survey can be useful for getting initial qualitative information about whether it's a thing, how it looks, and where to find it. On the other hand, it's not a good tool for estimating frequency: you don't know what sort of penetration the survey got or how much people only responded to the survey when they had something to share.

A study that tried to answer the question of whether distribution of mosquito nets leads to mosquito net fishing would make a sample of waterside communities that had recent net distributions and compare them to a similar sample that had not. But this is not that study, and it's not capable of supporting a "primary use" claim.

Now, mosquito net fishing could still be bad even if only a very small fraction of nets were used that way, due to the indiscriminate way it collects fish regardless of size. It doesn't mean "people aren't sleeping under nets" but could mean "nets might be making communities worse off." But there's not much reason to think this, and this study doesn't give much new information. GiveWell's 2015 blog post in response to an alarmist New York Times article is still very reasonable:

Randomized control trials consistently show large declines in child mortality from distributing nets and trends in malaria mortality and net coverage rates also suggest that mass distribution of mosquito nets has contributed to major declines in the burden of the disease. This evidence comprises one of the most robust cases for impact we've seen. The article makes the case for a possible harm to fish stocks relying on highly limited evidence.

The article does highlight a potential need to experiment with alternative approaches to malaria control in waterside, food-insecure communities that have very low net usage rates. In these areas, people shouldn't have to choose between malaria and hunger. But again, we see this as a likely isolated problem, and a much smaller one than the problem of insufficient nets for preventing malaria.