While working through Adam Johnson's excellent Boost your Git DX, I decided to make a repo to help solidify how the different merging strategies work. I created and modified some trivial files through three different PRs, merging each using one of the strategies.

Here is a summary of my findings:

Commits from the feature branch appear sequentially in the main branch's commit history, plus an extra Merge commit that shows all the changes together.

Only the squash merge commit appears in the main branch's commit history (no individual commits from the feature branch); however, by default, it will include each individual commit message in the details when you click on it.

Only the individual commits from the feature branch appear in the main branch's commit history (no merge commit); the original commit time is also lost, as each commit is dated to the time of rebase.

You can check the commit history to see how these work in detail, but it will probably be more tangible for you if you create your own repo and try this out yourself.