1) Perhaps yes.
2) Mathemtical contstans are purely abstact - pi only exists in a perfectly euclidian plane. It's an abstraction that works for us in a scertain scale because it matches close enough. In real life, there aren't circles.. they're just close enough. So no - a universe could have completely different maths required to describe it.
3) Math is purely abstract and detached from the universe. It operates on theoretical models that try to model reality. Math has it's own universal truths, but reality does not. The universe came before math - math is a tool invented by man to analyze the universe.
> So no - a universe could have completely different maths required to describe it.
That wasn't the question, though. Pi is the ratio between circumference and diameter of a perfect circle, whether or not you can create a perfect circle in your universe. We can't even create one in ours, and yet pi still "exists" here. Undoubtedly, wiggly-space mathematicians would also have a conception of pi, though it might be as esoteric to them as toroidial universes are to us.
To put that another way: Math has very simple axioms (e.g. set theory), and everything else derives from them. What theories can be proven from those axioms in this universe, can be proven from those axioms in any universe (and a universe that doesn't obey those axioms likely wouldn't "function" as a universe—it wouldn't have any reason to be causal, for example.) The maths describing a particular universe is its physics, which do change from universe to universe—however, those physics are all just different subsets of the set of mathematical physical models—which are, themselves, a subset of the mathematical theories provable from the axioms.
