A Cox nonnested test is conducted using a fast double bootstrap (FDB) methodto select among three competing functional forms (linear response plateau,quadratic, and Mitscherlich-Baule) to model forage yield response tonitrogen applied with swine effluent. The quadratic is rejected in favor ofone of the other functional forms in all cases. The FDB pvalues differed slightly from the single bootstrap pvalues. Buffalograss was slightly more profitable than bermudagrass and hasthe ability to use almost as much nitrogen as bermudagrass.
