Yield dynamics in long-term field experience (Lyubertsy experimental field of NIUIF). Message 2. Productivity of variants and field repetitions

Crop yields of 4-field fruit-shifting crop rotation for variants and field repetitions of long-term (42 years) field experience on sod-podzolic sandy loam soil (Lyubertsy experimental field) are analyzed. The effect of various forms of potash fertilizers was studied. Changes in time in the proportions of the influence of experience factors (variants and field repetitions) with a cycle of ≈17 years and the corresponding annual average yields with an average cycle of 8–9 years were noted. These differences in cycle length are unexpected and require further reflection. The presence of these strong changes over time in the proportions of the factors’ influence indicated the instability of their influence. During the study of the aftereffect of K-fertilizers (1976–1985), the influence of experimental factors was insignificant in most years. But it should be noted that in this case, the share of the influence of field repetitions exceeded the share of the influence of the experimental variant. According to the average yield differences of the coupled time samples, the variants were arranged in the following order: Control < NP < NPK, with NP + K_kai being the most productive of the NPK variants. For field replications, repetition IV showed the lowest yield. Due to the strong changes over time, the long-term yield trends were insignificant. Qualitatively, they were negative for the Control, NP and NP + N_sk variants and positive for the remaining variants. For field replications, the time trend of yield was negative for the first three and positive for the fourth. This indicated a tendency for soils with different initial fertility to converge over time. The errors of average long-term yields for variants and field replications are estimated in a direct way and (partially) through the corresponding plots. It is shown that these estimates can vary greatly in two ways. It is also shown that the results of comparing variants and field repetitions may differ by long-term average and by long-term average differences.