Improving the McCall herbage growth model

Abstract

The objective of this study was to improve the McCall herbage growth model. A more mechanistic senescence function was incorporated, replacing an empirical senescence function. The new function is based on leaf lifespan (LLS), measured in thermal time, which is characteristic of a given grass species, and can be measured independently or sourced from the literature. Other complementary changes were also incorporated, the major ones being to incorporate soil water infiltration constraints and to improve seasonal predictions in winter based on greater photosynthetic efficiency of leaves at low light intensity. The existing and new versions of the model were each calibrated against five time series of herbage growth data using the first half of each time series. The second half of each time series was used for validation, along with an independent data‐set. A suite of goodness‐of‐fit indicators was used to evaluate and compare the existing and new versions of the model. The new version more accurately predicted herbage growth following calibration, showing higher concordance correlation coefficients (ρ_c = 0.59, 0.62, 0.81, 0.81 and 0.53 versus 0.85, 0.88, 0.81, 0.88 and 0.80, for existing and new versions respectively). Most goodness‐of‐fit indicators also improved for the independent dataset (e.g. ρ_c = 0.708 versus 0.787, for existing and new versions, respectively). It is concluded that the new improved version of the model can be confidently used as a replacement for the existing version.

Keywords:

• calibration
• leaf lifespan
• pasture
• ryegrass
• senescence
• validation