section/column annotations should probably go in slot usage

[turbomam]

If we start building schemas representing multiple DH interfaces (soil, sediment, water, etc.) we might want different section assignment and column ordering for each interface. Each interface corresponds to one class in the schema, like the current 'soil_biosample'. In that case, the annotations shouldn't be made on the slots themselves, but rather on slot usage for each interface class.

[pkalita-lbl]

This is done