Modelling pasture growth rates in a long-term agronomic rotation experiment
by
Sue Welham
Biomathematics Unit, Rothamsted Research Ltd, Harpenden AL5 2JQ, UK
Coauthors: Brian Cullis (NSW Agriculture, Wagga Wagga Agricultural Institute, Wagga Wagga, NSW 2650), Guangdi Li (NSW Agriculture, Wagga Wagga Agricultural Institute, Wagga Wagga, NSW 2650), Mike Kenward (Medical Statistics Unit, London School of Hygiene and Tropical Medicine, London), Robin Thompson (Biomathematics Unit, Rothamsted Research Ltd, Harpenden AL5 2JQ, UK)

Li et al (2001) describe a large long-term agronomic rotation experiment set up to develop and demonstrate a pasture-crop system that is viable on acid soils of the traditional permanent pasture region in south-eastern Australia, with fertility sustained or improved. There were two pasture-crop rotations and two permanent pasture systems, each with and without lime. For the permanent pasture systems and pasture phases of pasture-crop systems, a rotational grazing pattern was imposed with either a long cycle (2.5 week grazing followed by a 5 week spell without grazing) or short cycle (one week grazing followed by a two-week spell), depending on the season. Pasture dry matter measurements were taken on pasture plots just before and after grazing.

This talk describes the use of smoothing splines to estimate relative pasture growth rates on permanent pastures and the pasture phases of the pasture-crop rotations by modelling the before and after grazing time series separately. The analysis breaks down the pattern in each series into three separate components: a (yearly) periodic pattern described by sine/cosine functions, periodic deviations about this pattern, and non-periodic deviations about the pattern. A power model (correlation decreasing exponentially with time) is used within each plot to take account of temporal correlation in the data. The analysis was done partly in Genstat and partly in ASREML.

Once the models have been fitted, comparisons in relative growth rates between different rotations can be expressed as a linear combination of fixed and random effects and calculated using the PREDICT function in ASREML.

Reference

Li GD, Helyar KR, Conyers MK, Cullis BR, Cregan PD, Fisher RP, Castleman LJC, Poile GJ, Evancs CM and Braysher B (2001) Crop responses to lime in long-term pasture-crop rotations in a high rainfall area in south-eastern Australia. Aust Journal of Agricultural Research, 52, 329-341.

Date received: September 13, 2002

Copyright © 2002 by the author(s). The author(s) of this document and the organizers of the conference have granted their consent to include this abstract in Atlas Conferences Inc. Document # cajn-43.