I’m in a specific branch in a field of biology where every PI basically knows every other PI, and there is quite a bit of competition amongst them.
One of my PIs during my graduate studies told me he got scooped by a colleague who, as he suspects but cannot proof, stalled a paper on the matter and changed his own project to incorporate the ideas he had found in the MS. That’s just anecdotal, of course, but it gives you an idea of the problem.
In other fields in biology you would have other reasons. Someone who works in bionics, for example, might have trouble sharing raw data due to their financial sources in military or industry. I knew people who worked on biological IR sensors, others on surface coatings, others on water absorbing materials. They told me of some weird situations. They do publish - but only aggregated data and stats go into their papers.
The weirdest story so far was that the comitee had to sign a NDA before they could take a PhD candidates viva, and the candidate had to redact his thesis after the viva for the public version, i.e. the one which went to the libraries, including our national library. Thus, his comitee has given him his degree based on a different thesis than what is available to the public. Seriously, so I was told - I have no proof for that, and it sounds absolutely absurd to me. And no, there was nothing even remotely military- or security-related in that thesis.
