I am writing my thesis protocol and I use many technical definitions, some of them not related to my career.
If I copy paste for those definitions, should I include the reference where I took it from? Wouldn't that be too many references?
This definition for example.
I'd ask your advisor. In linear algebra we don't source the definition of common terms like "orthogonal" or "symmetric tridiagonal matrix". The example you gave does not look the original definition of "water stress", and doesn't reference an original definition, so I presume it is a common term in water management.
A reference should be to the original author that invented the term or first used the term in a publication; references should not be to another anonymous document that is not academically published and just uses or defines the term.
The point is to give credit to inventors (or originators) where credit is due; and your current reference does not do that. If you are already supplying the definition, you can attribute it inline, to "Common term of Water Management" or something like that. Again, consult your thesis advisor, this is their job.
So there may be no possible reference for "Water Stress".
That said, numerous references don't hurt in a thesis, unless you intend to publish it in a journal or conference. In the latter case, references do count toward the page count, which is usually strictly limited in such venues.
So you would likely have to trim your thesis, but that is common, even dividing it into multiple academic papers.
Again, consult your advisor, don't rely on the advice of random Internet dudes.
For a thesis it is always a good idea to be clear where all your information comes from. Your thesis will be examined and clarity makes that examination simpler and shows your work.
How you reference the glossary source would depend upon how you are indicating all references. You could simply use the same reference number against all terms taken from the same place.
