
View larger version (26K):
[in a new window]
|
Fig. 2. Effect of total N rate applied either as ammonium nitrate or broiler litter on crop yield increases above the no-N treatment in experiments at a Tennessee Valley and a Coastal Plain site from 1990 to 2002. Stepwise regression was performed on least squares year x treatment means to arrive at the best-fitted model based on the coefficient of determination. Because the response variable was yield increase above the zero N treatment, all regressions were modeled without an intercept term. Numbers to the right of regression lines refer to the corresponding equation given in Table 4.
|