Soil samples were taken from compartments 1, 5, 10, 15, and 20 for treatments CMP and CRT, and from compartments 1, 2, 3, 4, 5, 16, 17, 18, 19, and 20 for treatments GCMP and GCRT.
In addition, for treatments GCMP and GCRT, soil samples were taken in grass barrier strips GS 2 and GS 4, and at distances of 0.15 m downslope and 0.15 m upslope from the lower and upper edges of these grass barrier strips (Figure 1).
Mean downslope soil tracer displacement over a two year period beginning July 1994 was 3.3 m for CMP, 1.8 m for CRT, 2.2 m for GCMP, and 1.5 m for GCRT.
The corn crop for the GCMP and GCRT systems was grown in terraces between adjacent contour grass barrier strips (Figure 1).
Figures 2-7 show the observed chemical species concentrations in the surface soil (0-20cm depth) versus compartment, that is, across the landscape of the "open field" plots (CMP and CRT) and individual terrace of management with grass barrier strips (GCMP and GCRT).
The total C concentration distributions across the entire plot in 1996 for treatments CMP and CRT, and for individual compartments in terraces 1 (footslope) and 4 (shoulderslope) for treatments GCMP and GCRT, are shown in Figure 2.
When the contour grass strip was combined with contour moldboard plowing, the slope of the regression equation in terrace 1 of GCMP increased to 0.50 (Figure 2C), indicating a greater change in total C concentration across the terrace.
In 1992, the P concentration was nearly constant across each plot (CMP and CRT) or terrace (GCMP and GCRT), being [less than or equal to] 2 [micro]g [g.sup.-1] soil.
In 1996, the negative regression slopes for GCMP were 6-9 times steeper than for CMP (compare Figure 4A, C, D) whereas the negative regression slopes for GCRT in 1996 were about 3-4 times steeper than CRT (Figure 4B, 4E, 4F).
The general trend for both Cu (Figure 6) and Zn (Figure 7) was for concentration to increase with increase in elevation in the terraces (GCMP and GCRT) and open fields of CMP and CRT.
The pH versus compartment relationship is periodic for GCMP and GCRT (Figure 8C).