it depends on what the authors of the original problem would understand to be the "volume" of those very ions...
i am always uncomfortable with those values, as they more often than not are derived form ion mobility experiments in solution, i.e. with a complete hull of water around them and thatlike
those data might be useful, if I considered
HSAB (which deals with ions in solution) as the topic of the hour [which to me seems to be the background of your thread ?] , but could be completely meaningless in other experiments
so. maybe the relative volumes mentioned are valid with respect to HSAB-discussions
in this case, the relative diameters of the ions (here: calculated as "ball shaped") might have been recalculated from their respective λ-values
with respect to a completely different approach, i.e discussing bond length, you would result in a completely different picture (i.e.: you were right): carbonate and nitrate being isoelectronic and isosymmetric (sp
2 at the central atom, 3 identical ligands due to mesomeric effects) , we could take a look at the bondlength N-O and C-O , respectively
and, surprise, surprise: N-O in nitrate is reported as 127,3 pm , C-O in carbonate as 131 pm
so, from this analysis, carbonate in fact should be larger than nitrate
as I said: it depends
regards
Ingo