At the present time, no one really knows, some speculation.
I can add a sealed O2 probe to measure, or a flow through gas hypo to withdraw gas and send it through an IR analyzer for CO2 etc.
You can also add air and time the bubble till it's gone etc, or N2 gas etc.
It might be just total dissolved gases coming out of a solution, not any one in paricular.
I think the last one is much more eplanatory in the results and in what we know about venturi action on a sealed water/gas interface. Say the water has 30ppm of CO2, 7ppm of O2, 400ppm of N2 etc.......this just equilibrtaed in gas form in there, and some of the gas just builds up as it comes out of solution. Same deal in filters.
CO2 in and of itself alone would dissolve pretty fast, but mixture with much less % likely would not.
I'm not so sure that the 30ppm that came out of solution into the gas phase would rapidly dissolve back, vs the 10000000 ppm.
Regards,
Tom Barr