When doing these, one crucial step that is often unavoidable is visualizing the region. I've thought about some ways to manipulate limits of integration without visualizing, but I think it would make things much harder. So you should learn to visualize the regions the describe.
You have a bowl shape that is parabolic (paraboloid is the term in this dimension), that is, consider the y=0 plane, then z=x^2 is parabolic (the term for curves). The bowl occupies space between z=0...4. At z=0, x and y must be zero. At the z=4 plane, you have the circle x^2+y^2=4, that is, radius 2. You had the right integrand, √(1+4x^2+4y^2) dxdy.
So to find the region that paramet(e)rizes the bowl via x and y, we need to know which values of x and y will garauntee we hit all points of the bowl exactly once. So that is the projection of the bowl down (and/or up) onto the xy plane. The projection is of course the disk, x^2+y^2≤ 4. The limits would be hard to integrate with, so our look would be much better if we use a change of coordinates, from rectangular to polar coordinate.
So we want to map from (x,y) to (r,θ). Then the integrand would be √(1+r^2), while dxdy becomes rdrdθ. Why don't you try to find out what the limits of integration would be.