Submitted to: Soil/Water Research, Progress Report
Publication Type: Experiment Station
Publication Acceptance Date: April 20, 2009
Publication Date: July 1, 2009
Citation: Riedell, W.E., Pikul Jr, J.L., Jaradat, A.A., Schumacher, T.E. 2009. Crop Rotation and N Input Effects on Mineral Elements in Soil, Corn, and Grain as Revealed by Discriminant Analysis. p. 1-3. In: L. Ulvestad (ed.) Soil/Water Research 2008 Progress Report #PR 08-23. South Dakota Agricultural Experiment Station. Brookings SD. Interpretive Summary: Crop rotation has been an important component of agricultural systems for centuries. With the advent of synthetic pesticides and fertilizers during the mid-20th century, however, extensive crop rotations were supplanted by intensive monoculture or short rotation cropping in many areas of the US. Concerns and costs associated with these intensely-managed systems include decreased soil organic matter, degraded soil structure, increased soil erosion, increased surface and groundwater contamination, and increased production costs. Under a 4 year corn-soybean-wheat/alfalfa-alfalfa rotation, where corn followed a forage legume, corn grain yield was stable across all N input levels studied. Conversely, corn yield decreased as N input level was reduced under the continuous corn monoculture and a 2 year corn-soybean rotation treatments. Thus, growing corn in extended rotations that include forage legumes may be a more sustainable practice than growing corn in either monoculture or 2 year rotation with soybean.
Technical Abstract: We were interested in identifying soil, plant, and grain mineral nutrient variables that best discriminate N input treatments as well as rotation treatments from one another. Study objectives were to use multivariate discriminant analysis to measure soil, shoot, and grain mineral nutrient responses to corn monoculture (C-C), 2-yr rotation (C-S) with soybean, or 4-yr rotation (C-S-W/A-A) with soybean, wheat, and alfalfa under different N input levels. The first and second canonical discriminant functions (CAN1 and CAN2) extracted from the whole data set for N input treatments accounted for 68 and 32% of total variation, respectively. Discriminant analysis indicated large negative loadings of plant N, grain N, and grain S as well as large positive loading of plant Zn on CAN1. A stronger level of discrimination was observed between crop rotations when compared with the level of discrimination between N-inputs. The C-C and C-S-W/A-A rotations were totally separated from each other along CAN1, which accounted for a large portion (85%) of total variation. The C-S rotation was scattered on both sides of CAN1. However, the C-S rotation was totally separated from the other rotations along CAN2, which accounted for 15% of total variation. Given the effects of N input and rotation treatments (and their interactions) on grain N, it was not surprising that grain N played an important role in discriminating between N inputs and between rotations. Grain N, as well as grain S, had large loading on CAN1 important for discriminating between N input treatments as well as on both CAN 1 and CAN2 important for discriminating between rotation treatments. The similarity in loadings of grain N and grain S may have resulted because grain S responded to N input and rotation treatments in a manner similar to that of grain N. These data confirm a close association between N and S concentration in corn.