Custom Mesh and how does Culling work in Unity

I'm learning Unity right now and I'm trying to port a terrain class I've written a while ago in C++ and Irrlicht.

I get the basics but I have some question on how(if) Unity does any sort of Culling.

So I tried the Occlusion Culling but it seems I need to bake data in the editor and I don't know how and if I can use this in runtime.
Irrlicht has occlusion culling and once you create a mesh the engine will calculate whatever it needs for it on-the-fly (usually takes 1 sec or so).

So one question would be: Do I have to worry about that? Will unity optimize the mesh at runtime and will it take care of this for me.

Is Unity doing viewing frustum culling or do I need to do that by myself.

What I did for Irrlicht is to split my mesh in smaller meshes. Since they are arranged like a grid it isn't hard to arrange them in tree that has a boundary box for each node. It take a few checks to see what node of the tree should be visible and what not. This allows me to do giant terrains and keep the same FPS - only limit is memory.

Here comes my second question: If I arrange objects inside an empty game object and mark that as Invisible, will Unity traverse the child during rendering or not. If it does how do I prevent Unity from doing so.

Sorry if those are generic questions - I hope someone more experienced with Unity may give me some tips.

Here's what my terrain looks in Irrlicht (it's pretty basic) :

 

Yes. If you want runtime occlusion culling you'd have to write your own system for that. I didn't quite understand your second question...are you referring to having the renderer disabled? Any objects with disabled renderers don't need to be culled; being a child or not isn't relevant.

--Eric