I do not pretend to be an expert in the Morris Water Maze! But I can see no reason why you should not try this. As for the optimal time interval, I'd suggest that this is an opportunity to learn more about short-term/working memory in rats. You should try different delays, ranging from immediate on up, until you reach an interval long enough that no residual memory remains to assist the rat in its search for a secure base.

Yes, you can. You need to vary the position of the hidden platform across trials. You need to arrange separate blocks of place trials for each of these hidden platform locations. Since it's short-term/working memory assessment, you should consider running these blocks of trials in one session. You should also have probe trials (no hidden platform) in between and/or these blocks of trials, each with a distinct hidden platform location. You can refer to the Method and Discussion of Zhong et al. (2017) in my "Virtual Water Maze Cognitive Aging Project" for hints about how to do this.

Dear Colleagues,

Thanks for your advice and comments!

But still. If I want to evaluate both short-term and long-term memory in one experiment (without relocation of the platform). Should I also have a trial (with platform) between probe trials?

You can assess performance at different retention delays, using a delayed-matching-to-place paradigm that uses variable goal locations.

Please see, for example, da Silva, Bast, Morris, 2014, Learn Memory, for how performance at different retention delays can be compared:

http://learnmem.cshlp.org/content/21/1/28.full.html

If you only use one platform, I'm afraid the interpretation of your tests at different retention delays will be confounded by the different number of acquisition trials that rats will have received at the different retention intervals (if you include additional training trials with platform) or by extinction (if you simply run two probes at two different retention delays).

Dear Dr. Bast,

Thanks for your review!