That seems to fit pretty well with what Dale Allen Pfeffer says on p. 8 of his '06 book "Eating Fossil Fuels." In a table titled "Total Energy Directly and Indirectly Consumed in U.S. Farms in 2002..." he gives:
27.3% Diesel
8.6% Gasoline
(so about 36% directly from oil products)
29% Fertilizers
8.3 % Pesticides
3% Natural Gas
4.6% LP Gas
(So about 45% from LP Gas, Natural Gas and products heavily dependent on NG imputs)
The rest, 20.7%, is electricity.
I'm guessing that the majority of the fuel used to get the food to market and to table is going to be diesel and gasoline again (and that's not going to be an inconsiderable amount, since the average distance traveled for US food items farm-to-table is something like 1500 miles).
So oil is not insignificant in the production of food in the US, but NG is at least as large a factor.
I guess I just automatically lump oil and NG together in these discussions, since both seem to be peaking at about the same time and both are used for heating as well as food production...
I think the main challenges for food production will actually come from the uncertainties brought about by climate change. See the just-released excellent report out from carbon equity:
climate code red
We had flooding near here that about doubled the all-time record, while many other counties were experiencing record drought. The SW and SE of the US seem to be just drying up completely. The great breadbasket of the plains is set to return far drier norms for the last 10,000 years (Holocene) even before the effects of GW.
The climate's going nuts, and farmers are on the front lines.