In common parlance where I live, "data" is almost invariably a singular, with the form "datum" avoided altogether and "piece of data" or "point of data" used in its stead. Some very particular professors of mine have insisted that "data" be treated as a plural, but the majority of science professors I've dealt with simply don't care. Scientists tend to be relatively practical people - as long as your writing is understandable, they're usually fine with however you deal with "data." "The data show" and "the data shows" are usually both acceptable. "Datum" and "criterion," I think, sound a little stiff - which is why "criteria," which generally still takes a plural verb form even today, is often used in contexts where "criterion" would technically be correct. Even some scientists feel uncomfortable using "bacterium," and consequently "bacteria" is also acceptable in the singular ("this bacteria...").
Another word that does this in American English is "media." "The media" usually "reports" (singular). Of course the plural form "mediums" is more common than "media" anyways. The "media" is viewed as a single entity, which thus takes a singular verb. I'm guessing this is largely American, seeing as British people talk about collective entities of people with plural verbs ("his family say...", "England are playing tonight") which sounds wrong to the American ear.
As for "hibernaculum," the Latin plural is indeed "hibernacula," but it's also possible to circumvent this whole issue by saying "hibernaculums." As a relatively advanced student of Latin, I know, of course, how to create all the Latin plurals in science - but I nevertheless say "genuses," "nucleuses," "spectrums," "funguses," "appendixes" (over "genera," "nuclei," "spectra," "fungi," "appendices") - I always speak them, and usually write them (but there are occasionally professors who penalize for such Anglicized forms). What's interesting is that it's much less acceptable to do the same thing with Greek plurals. I still say "analysises" and "thesises" sometimes, but never in writing, and "phenomenons" sounds wrong to me. But as far as Latin goes, it's generally perfectly acceptable not to use the Latin plural at all.