On the face of it, the ontology of caves is just like that of holes. But there do seem to be some significant differences. A hole is, as it were, pure nonbeing, albeit of a sort that depends on its walls for its existence. But a cave includes the hole and the walls. When we talk of cave walls, we think of them as parts of the cave. But what is the wall? Is it just a veneer of the rock? It seems to me that if you could take a cave and cut it out of the rock, keeping a thin veneer of rock, and put it in space (so the veneer wouldn't collapse under its own weight), the cave would survive. So only a veneer is essential. Is only a veneer of rock part of the cave? No—more is. For the physical properties of the cave—say, the material causes of stalactites—probably include more than a veneer. Besides, if only a veneer of wall is included in the cave, then only a veneer of stalactites and stalagmites would be included. But that's absurd. So, more than a veneer of wall is part of the cave. But how much? This is, surely, vague.
So caves seem to be more ontologically solid than holes, and it seems that we can answer a bunch of questions about them. But like sand sculptures, they depend ontologically on nonbeing—if you fill in the hole (or cover the sand sculpture with sand) with solid rock, it seems the cave perishes. This ontological dependence shows that caves (and sand sculptures) are not substances. If they are at all.