ij,
fertilizer is just stuff that completes the mass balance. You're not thinking of C, O, and H as fertilizer just because it's effectively free. There's lots of N that would be effectively free but people are impatient and prissy so they're happy to pay nominal amounts for some more pleasant and manageable form of ammonium/amine.
Things like sulfur and magnesium that are consumed by plants in large quantities are readily/cheaply available from so many sources that i'd ignore any fertilizer supplier whose income relied on supplying those components (i.e. potash suppliers). Today's oil refiner could be tomorrow's fertilizer supplier and they'll do it just to get rid of waste.
And that brings up another point: with the exception of phosphorus, a lot of what is waste in other processes is fertilizer. The top example being poop. Phosphate is the only fertilizer material which is not conveniently available in a concentrated form from a large number of sources.
i'd quibble a little bit with Roy's use of 'efficiency'. Ag biotechs have produced plants that produce a higher ratio of fruit/(non-fruit plant parts) and those plants grow in higher density (plants/acre). The reason for my quibble is that i suspect a higher mass fraction of materials that must be replaced as 'fertilizer' are removed in the fruit. However, since most of that 'lost' mass is cheaply replaced his point is valid.
slight correction on your legumes comment: the nitrogen fixers are usually symbiotic bacteria that grow amongst the plant roots. maybe the next area for ag biotechs will be to modify symbiotic bacteria to more rapidly cycle inorganic materials to their biologically useful forms.
charlie
