Whence the odd-even staggering in nuclear binding?

We explore the systematics of odd-even mass staggering with a view to identifying the physical mechanisms responsible. The BCS pairing and mean field contributions have A- and number parity dependencies which can help disentangle the different contributions. This motivates the two-term parametrization c_1 + c_2/A as a theoretically based alternative to the inverse power form traditionally used to fit odd-even mass differences. Assuming that the A-dependence of the BCS pairing is weak, we find that mean-field contributions are dominant below mass number A~40 while BCS pairing dominates in heavier nuclei.