Soil Profile a Physical Appearance The surface of the plowed blocks appeared of uniform texture and color following the smoothing operation and rain which occurred between plowing and sampling. The backhoe had no difficulty in making holes which retained vertical structure in this region. The soil appeared to be reasonably well-mixed, though occasional darker (organic) patches or layers could be seen running through lighter coral regions. Such layers occurred from 5 to 40 cm in the "16" blocks, but were less noticeable in the "14" blocks, which appeared well-mixed down to the coral area at 40 em. b. Radionuclide Distribution Results of profile sampling are presented in Table B-9~-3 and average values are graphed in Figure B-9-11. Statistical Analysis The plow experiment area consisted of eight stake locations laid out in a 2 x 4 rectangle at 25 m spacing. Before plowing the surface TRU values (from IMP readings) at these locations ranged from 48.8 to 109 pCi/g, with a mean of 71.5 pCi/g. After plowing the TRU surface values ranged from 12.3 to 4.3 pCi/g, with a mean of 8.2 pCi/g..It was decided that half the area would remain unplowed so that the necessary "control" areas could be available for possible future plant uptake studies. These control plots were irrelevant in analyzing the effect of plowing on redistributing radionuclides in the soil. Each plowed location served as both untreated (before plowing) and treated (after plowing) observations for statistical purposes. Practical limitations on the plowing technique coupled with mechanical difficulties in the plow precluded application of standard randomization methods. As a compromise, the plot was divided into four sets of 2 x 1 rectangles, each containing either two unplowed blocks or two plowed blocks. It was also known from previous experience that adjacent strips should not be treated the same, so that only two possible configurations (first plot plow or first plot control) were available. One of these was chosen at random, resulting in the experimental configuration shownin Figures B-9~7, -8 and -10. Pre-plowing samples were takenin all eight blocks, but post-plowing samples only in the plowed blocks. The post-samples were taken in different locations from the original samples to avoid confounding plow effects with backhoe effects. Profile samples were taken at seven depths (0 to 5 em, 5 to 10 em, 15 to 20 em, 30 em, 40 em, 50 em) in each of four backhoe holes in each treatment block. This resulted in a total of 32 profile sets pre-plowing and 16 sets post-plowing. During site preparation operations, the surface soil was disturbed in some areas. Some similar operations would be necessary in any field plowing application, so this disturbance was considered an integral part of the plowing treatment for statistical purposes. "a There were two primary aspects of interest in the experiment: the effect of plowing on surface TRU contamination, and the possibility that plowing alters the distribution of TRU contamination in the soil profile. IMP surveys at the eight stake locations before and after plowing measured the first effect, and a series of backhoe profile soil samples taken before and after measured the second. Results of Statistical Analysis The surface changes, as measured by the IMP, were analyzed with a two-way analysis of variance (ANOVA). The experiment was handled as a randomized block design with two treatments (before and after plowing) on each of four blocks. The ANOVA results are shownin all - Table B-9~4, mid‘ B-9-19