E. Ladewig et B. Marlander, DETERMINATION OF OPTIMUM N-SUPPLY IN SUGAR-BEET BY DIFFERENT MATHEMATICAL-METHODS, Zeitschrift fur Pflanzenernahrung und Bodenkunde, 160(1), 1997, pp. 85-88
Various mathematical models used to predict optimum N-supply (fertiliz
er-N + N-min) for maximum root yield, sugar yield and white sugar yiel
d were compared by fitting data from 39 field trials in Northwest Germ
any (1980-1992). Quadratic, modified exponential and square root model
s resulted in higher estimates of optimum N-supply than did 2-regressi
ons and plateau type spline models with differences as high as 130 kg
N/ha between models. Residual variance was lower with the latter two m
odels. Data were grouped at 40 kg/ha intervals of N-supply for analysi
s of variance and resulted in the lowest estimates for optimum N-suppl
y. Based on the overall average, the optimum fertilizer-N dressings we
re 80 kg N/ha for root and sugar yield and 70 kg N/ha for white sugar
yield while mineral soil-N in spring was 70 kg N/ha.