My Bleg from last week has generated two different possibilities for why we should be taxing methane emissions from cows. BK Drinkwater and John Small (who have both weighed in with their own posts here and here, respectively) have posited the fertiliser theory: If the fertiliser used on the land grazed by cows contains carbon taken from the earth, then the methane emissions from a steady-state stock of cows will produce a continued increase in the stock of GHGs in the atmosphere (measured in CO2 equivalents).
John suggests that the source of the net carbon would be urea, although it is not clear from a quick reading of the Wikipedia article on it if the CO2 used to create the urea is a net addition of has already been created as a by-product of existing processes. In any event, as John points out, the policy implication would be to tax the use of urea.
The other possibility, suggested by PaulL in the comments is that the concern about methane is that the stock of GHGs in the atmosphere affects the flow of temperature increases, at least until temperature catches up. Paul suggests that this provides a reason for taxing a one-off increase in methane levels every year for a century.
The trouble with this line of reasoning, as I see it, is that the same argument would apply in spades to on-going emissions. That is, if methane is twenty times as powerful a GHG as CO2 for the same amount of carbon, my logic suggests that the tax on methane emissions that are part of a natural cycle (assuming no net additions from fertiliser) should be way less than 20 times the tax on CO2 emissions.