Cropping frequency effects on yield of grain, straw, plant N, N balance and annual production of spring wheat in the semiarid prairie

Producers in the semiarid Canadian prairies practice frequent summer fallow to conserve water and reduce the risk of crop failure, but this practice Promotes soil degradation. In contrast, annual cropping enhances soil quality but results in greater economic risk. We need to know what is the most suitable cropping frequency for this region. In 1985, based on results of the first 18 yr of a long-term crop rotation experiment being conducted on a medium-textured, Orthic Brown Chernozem at Swift Current, Saskatchewan, we modified the experiment to allow comparison of four cropping frequencies over the period 1985-2002. These were fallow-spring wheat (Triticum aestivum L.) (F-W), F-W-W, F-W-W-W-W-W, and continuous wheat (Cont W). All systems received recommended rates of N and P fertilizer. Growing season precipitation during the 1985-2002 period was 10% above average so that grain yields were also above average for this region. We assessed yields of grain, straw and N in aboveground plant parts, N concentration in grain and straw, harvest index, nitrogen harvest index and water use efficiency, and the average annual production (on a rotation basis) of grain, straw and aboveground N yield, and related these characteristics to water deficit. Water deficit was the main factor responsible for annual variations in the characteristics assessed. Of these characteristics, cropping frequency influenced only the average annual productivity factors; it rarely influenced the characteristics when they were assessed on a rotation-phase basis. Assuming cropping frequency (x) for the 2-yr, 3-yr, 6-yr and Cont W rotations is 50%, 67%, 83%, and 100%, respectively, then average annual yields (y) in kg ha(-1) yr(-1) were related to frequency as follows: (1) for grain yield y = 768 + 10.7 x, r(2) = 0.99; (2) for straw yield y = 1159 + 18.9 x, r(2) = 0.99; and (3) for aboveground plant N yield y = 23.7 + 0.36 x, r(2) = 0.99. The regression between annual grain production and cropping frequency at Swift Current for 1985 to 2002 had a much higher slope than the relationship for the same experiment in the much drier 1967 to 1984 period. Further, the equations for Swift Current in 1985-2002 were similar to the relationships obtained for systems in the subhumid Black Chernozems, indicating such relationships are greatly influenced by weather conditions. These relationships may be useful for conducting economic analyses and for modeling N balance. We conducted an apparent N balance analysis which indicated that all four rotations have lost N from the system in inverse proportion to the frequency of cropping. However, changes in total N in the 0- to 0.3-m depth of soil suggest that they have not lost N and that Cont W gained N over the course of the study. We are not able to explain this apparent anomaly and recommend further research on this question.