Crop response to N fertilizer (N-f) is influenced by factors such as N
-f management, soil type, crop sequence, and supply of residual and mi
neralized N, but there is no framework to define the best strategy to
account for a given factor in an N-f recommendation. This paper descri
bes a three-component classification system for evaluating the effect
of any factor on yield response to N-f. This system provides (i) a voc
abulary to describe clearly the effect of a factor on N-f recommendati
ons, and (ii) insight on how to adjust N-f recommendations for the eff
ect. Factors that affect yield response to N-f but not to N supply (N-
s) were classified as shift effects (i.e., movement of a response curv
e in the x and/or y direction, with no change in coefficients of curva
ture). Factors that interact with N-f and N-s response were classified
as interaction effects. Nitrogen supply was defined as the sum of abo
veground plant N content of the control (0 N applied) plot, postharves
t fall NO3- in the surface 1.5 m of the control plot, and N-f applied.
Two 2-yr experiments were conducted at Rosemount and Waseca, MN, to c
ompare N-f response of continuous corn (Zea mays L.) with that of firs
t- and then second-year corn following alfalfa (Medicago sativa L.). W
e used the classification system to evaluate effects of crop sequence,
year, and location on corn yield response to N-f. Year and crop seque
nce effects at Rosemount were primarily shift effects, implying that q
uantifying the effect on Ns would be sufficient to account for these e
ffects on N-s recommendations. In contrast, the interaction model pred
ominated at Waseca. Consequently, at this location simple adjustments
of Ns were not sufficient to account for the complexity of crop sequen
ce effects on N-f recommendations. This classification system facilita
tes the organization, evaluation, and communication of the many factor
s that influence crop yield.