Grazing frequency and ecosystem processes in a northern mixed prairie, USA

The objective of this study was to evaluate for a 6-yr period the effects of a twice-over rotation grazing system (ROT) and a season-long grazing system (SL), and compare these effects with long-term grazing exclosures (NG) in terms of (1) species composition and basal cover, (2) aboveground net primary production (ANPP) and aboveground N uptake (ANPP-N), (3) rates of litter and root decomposition and N release, (4) soil N mineralization and immobilization, (5) aboveground C and N flow, and (6) grazing intensity (GI) and animal performance, The study period included the drought of 1988. No major differences were found in ANPP and ANPP-N among treatments, but there were important seasonal variations. An average of 72% of ANPP and >82% of ANPP-N occurred by mid-June. There were no differences among treatments in terms of decomposition and N release rates from litter and root biomass, or in soil N mineralization. Grazing, however, reduced the amount of C and N immobilized in standing dead and litter and the flow of C and N from standing dead to litter to soil organic matter. The NG and ROT treatment were more similar in this regard when compared to the SL treatment, and their similarities increased after the drought of 1988. There were no consistent differences in GI between the ROT and SL treatments. Before 1988 GI averaged 21% but in 1988 and 1989 GI increased to an average of 49% as a result of the drought and its aftereffects. Cumulative animal performance was similar under both grazing treatments but there were significant seasonal variations. Species composition was more responsive to grazing than were C and N flows. The differences in this case were found between the grazed and NG treatments but not between the two grazing treatments studied. There were no broad patterns of change in total plant basal cover as a result of grazing patterns or drought. Changes in species composition were highly dependent on range site. The most consistent pattern involved Bouteloua gracilis, which had higher relative cover in the grazed treatments than in the NG treatment. Results from this study indicate that in the grasslands of western North Dakota (1) the recommended stocking rate may be too conservative, (2) rotation grazing may allow for higher stocking rates than season-long grazing without a major impact on animal performance, (3) rainfall is more important than grazing or grazing systems in the control of the ecosystem-level variables measured, (4) species composition is affected by drought and grazing (but not by grazing systems) but the responses are highly dependent on range site, and (5) drought and grazing tend to increase the relative composition of warm-season grasses and forbs.