Why isn't my database cluttered with artifacts from my Spring DBUnit tests?

I am currently testing my Spring repositories and decided to use a MariaDB server instance instead of an in-memory Derby instance because of some complications in a test that involved a database view.

While the tests eventually did succeed without errors and failures, I noticed that I didn't add a @DatabaseTeardown annotation to my test case. So I decided to check my database for unwanted rows leftover from the test and found that my database was just as empty as before the test.

Could someone here explain why this is happening?