We can phrase anything in a negative light. Food which fuels 100% of Denmark’s population causes only 33% of Denmark’s emissions. This is a wonderful achievement.
Wild-caught fish are likely not included in that statisic, but vegetables surely would be. Growing vegetables is the quintessential example of farming.
I'm not sure if farmed fish would be counted, since that is not traditional agriculture.
But fishing still burns a lot of diesel. If this statistics don't take in account Denmark emissions from activities in the sea or from commerce fleet out of Denmark, they are probably misleading to wrong results.
And if we take in account that Greenland acts as a buffer in this sense (they don't have probably a lot of cows, so the emissions effect should be diluted). This perfect "one third" statistical value seems just a raw assumption, or just invented for filling a report hiding the lack of data.
If we want to bet for a future, we have to be extra careful to not just repeat slogans, dogmas or cite incomplete studies, even if they say what we want to hear. This only will delay the necessary measures
