Since LR(0) parser places reduce-moves in the entire row of "Action", having anything more than just the reduce-move in the state having final-item would lead to SR or RR conflict.
So, Options A, B and C are true.
So Option D must be false. But why is that?
Intuitively, there can be a Regular Language that has its grammar such that the state having a final-item would have some additional item.
But to give a counter example:
$S\rightarrow aA$,
$A\rightarrow b|\epsilon$
It has an $\epsilon$ production. Any grammar with $\epsilon$ production can't be LR(0), because in the parsing table, we'd put reduce-
moves in the corresponding row, on which a shift-move can result in an SR conflict.
In contrast, an SLR(1) parser can have multiple reduce-moves, or both shift and reduce moves in the same state.