Tillage - fertilizer changes: Effect on some soil quality attributes underlong-term crop rotations in a thin Black Chernozem

A crop rotation experiment initiated in 1958 on a thin Black Chernozemic cl ay at Indian Head, Saskatchewan, was managed using conventional tillage unt il 1989 and changed to zero-tillage in 1990. We soil sampled in 1987 and 19 97 to determine management effects on selected soil biochemical characteris tics, and the change in some of the more labile soil quality attributes rel ative to the change in soil organic C and total N. Rotations examined were: fallow-wheat (Triticum aestivum L.) (F-W), fallow-wheat-wheat (F-W-W), con tinuous wheat (Cont W), legume green manure (GM)-W-W, and F-W-W-hay (legume -grass)-hay-hay (F-W-W-HH-H). The monoculture cereal rotations were either fertilized with N and P based on soil tests or unfertilized, while the legu me-containing systems were unfertilized. There was also a F-W-W ON + P) tre atment, in which about 20% of the straw was harvested each crop year. With the change to zero-tillage management in 1990 and in anticipation of greate r soil water storage, higher rates of N were added thereafter. This resulte d in an upward trend in stubble-crop yields and a positive yield response o f wheat grown on fallow, where before the change wheat grown on fallow did not respond to fertilizer. The corresponding increase in crop residue produ ction and residue C inputs resulted in all fertilized systems gaining organ ic C and total N in the 0- to 15-cm depth between 1987 and 1997, while the unfertilized systems remained unchanged. Soil organic C and total N, microb ial biomass C (MBC), light fraction organic C and N (LFC and LFN), minerali zable N (N-min) and wet aggregate stability (WAS), generally had positive r esponses to fertilization, to increased cropping frequency, and to the incl usion of legume green manure or legume hay crops in cereal-based rotations. Straw harvesting did not influence grain yields, nor did it influence the soil biochemical characteristics, though it tended to render the soil more prone to erosion. Response to cropping frequency was apparent only in the f ertilized systems, where the more labile soil quality attributes, (i.e., MB C, LFC, LFN, and N-min) were more sensitive than organic C or total N. Howe ver, gains in LFC and MBC in response to fertilizer did not account for a s ignificant fraction of the gain in total organic C. During the period 1987 to 1997, MBC in the 0- to 15-cm depth increased by 40% in absolute value an d by 33% relative to organic C (3.6% of organic C in 1997 vs. 2.7% in 1987) . The same was true for LFN in the fertilized treatments and in the green m anured system (1.46% of total N in 1997 vs. 1.15% in 1987). However, LFC ha rdly changed over this period. Relative to total N, N-min decreased in 1997 compared with 1987, likely due to higher immobilization. Wet aggregate sta bility was generally greater in 1997 compared with 1991, reflecting greater crop residue inputs and less soil disturbance under zero-tillage managemen t.