Yep. If yeast are provided with oxygen, they can convert an equal amount of sugar into *four times* more CO2 (if I recall right), producing no alcohol.
The problem is that air is mostly nitrogen. All that inert nitrogen, which is insoluble in water, has to be purged out of the system. Which, depending on your diffusion scheme, creates bubbles that carry away much of your bonus CO2 uselessly to the top of the tank. Or fill a reactor until it burps out nitrogen and CO2.
And roughly 18x the energy yield from using the Embden Meyerhof Parnas pathway. If the fermentation is being done through the Entner Doudoroff pathway, CO2 production is roughly 1/6 the Krebs Cycle with 1/36th the energy yield. (Due to thermodynamic inefficiency in transmembrane energy transduction, the numbers aren't quite as clean as textbooks state)
Back to the OP's question:
The fizzing effect from shaking is because the solution is "saturated" in CO2 and the kinetic energy you're imparting to the system is just enough to push it over the energy threshold required for a shift in its equilibrium. In other words, you're making CO2 when you shake the bottle.
Unlike O2 (which interacts with water through dipole alignment), CO2 reacts directly with the water: CO2 + H2O --> H2CO3
. After a certain amount of time, the reaction making carbonic acid (H2C03) slows until the concentrations of reactants and products reach equilibrium. At equilibrium, the reaction rate in the forward and reverse directions is balanced: CO2 + H2O <--> H2CO3
When you impart energy to the system (heating, shaking, gorgon death ray, etc), you shift that balance. As a result of the reaction being exothermic, the reaction goes faster in reverse than forward, so that the net reaction produces C02 until the relative concentrations reach a new equilibrium state and the rates go back to being in balance: H2C03 --> H2O + C02.
Edit: If you really want to go down the rabbit hole, here's a more complete explanation: http://www.newton.dep.anl.gov/askasc.../chem99661.htm