The depth distribution of weed seed, their location in relation to soil aggregates, and the size distribution of soil aggregates were determined in fields that had been under continual conventional or reduced tillage for more than 12 years. Soil samples were taken at three depths in the spring and fall of 1982, and soil aggregates were classified into seven size classes by dry, rotary sieving. Subsamples were deflocculated and washed through sieves to recover weed seed. In the spring, 85% of all seed in the reduced-tillage and 28% of those in the conventional-tillage soil were in the 0- to 5-cm-depth layer. Conventional tillage incorporated weed seed uniformly into various soil aggregate classes, whereas with reduced tillage more seed accumulated in the unaggregated fraction of the soil. In the fall, weed seed distribution in relation to soil depth and among soil aggregate classes was similar for both tillage regimes.

No abstract

B. Maximum Likelihood Approximation 103 C. Maximum Likelihood Applied to Rhodes' Model 109 V. MONTE CARLO STUDY 113 A. Methodology 113 B. Results 114 C. Examples 124 VI. GOODNESS OF FIT STUDY 148 VII. DISCUSSION AND CONCLUSIONS 157 A. Moments of the Logistic Process 157 B. Comparison of Linearization Methods 158 C. Maximum Likelihood Approximation 160 D. Fitting the Logistic Curve 160