MAIZE TEMPORAL YIELD VARIABILITY UNDER LONG-TERM MANURE AND FERTILIZER APPLICATION - FRACTAL ANALYSIS

Long-term experiments offer unique possibilities to study the effects of management practices on crops and soils with time. Characterizing temporal variability of grain yield provides an opportunity to distinguish between the effects of environment vs. management practices on grain yield, an important parameter. A long-term study was established in 1912 in western Nebraska as part of a rotation study. In 1953, each plot was divided into manure (27 Mg ha(-1) annually) and no-manure sections to which fertilizer treatments of 0, 45, 90, 135, 180 kg N ha(-1), and 135 kg N + 80 kg P ha(-1) were applied annually. Grain yield data from 1953 to 1993 for these treatments were used for fractal analysis and determination of fractal dimension (D), which is an indication of pattern of variation. A D value close to 1 indicates dominance of long-term variation (or trend), while a D value close to 2 indicates dominance of short-term (year-to-year) variation. Grain yield increased with increasing N application rate without manure, but no response to fertilizer application was observed where manure had been applied. Fractal dimensions ranged from 1.942 to 1.996, indicating significant domination of short-term variation of grain yield in the past 41 yr in all treatments. There was no significant difference between D values for manure or fertilizer treatments. Soil fertility amendments did not alleviate year-to-year variability observed in the corn (Zea mays L.) yield. Environmental factors like hail, fall freezing, and temperature variation during the growing season had a significant effect on the grain yield over the years. Although management practices can reduce temporal grain yield variability in some crops, variations in environmental factors in this study were great enough to dominate the yearly maize grain yield regardless of soil fertility amendments.