HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Figures Only Full Text Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in ISI Web of Science Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by van Oijen, M. Articles by Caldwell, N. Search for Related Content PubMed Articles by van Oijen, M. Articles by Caldwell, N. Agricola Articles by van Oijen, M. Articles by Caldwell, N. Related Collections Agroclimatology Forage Management Other Forage Crops Data acquisition and assimilation Other Models

Modeling

Process-Based Modeling of Timothy Regrowth

^a CEH-Edinburgh, Bush Estate, Penicuik, EH26 0QB, UK
^b The Norwegian Crop Research Inst., Særheim Research Centre, Postvegen 213, N-4353 Klepp St., Norway

^* Corresponding author (mvano{at}ceh.ac.uk)

Received for publication October 4, 2004. Previously, the literature on the growth of timothy (Phleum pratense L.) in Scandinavia was reviewed and a simple process-based model was developed to explain and predict the response of this species to different environments and management regimes. The model could not be tested in detail, because only biomass data were available at the time. However, recent experimental work has generated a large (n = 633) and uniquely detailed dataset on the growth and underlying physiology of timothy and its response to cutting at different growth stages. The present study aimed to use this dataset to test the model, and to use the model to identify the key physiological and morphological mechanisms that determine the regrowth rate of timothy after cutting. Model testing consisted of comparing simulations and measurements for eight variables: biomass, leaf area index (LAI), tiller and leaf density, rates of leaf appearance and elongation, carbohydrate concentration, and specific leaf area. Although the new data referred to a different cultivar, a different site, and different years from those used in the original model parameterization, the model was still able to account for nearly half the variation in the dataset [r² = 0.468, normalized root mean squared error (RMSE) = 0.631]. This suggested that the key assumptions of the model (i.e., dependence of growth and allocation on the source-sink balance of the plants and a close link between tillering and leaf area dynamics) were plausible. However, the original model was not able to account for the observation that cutting at early heading tended to be followed by a longer lag phase than cutting at anthesis. We identified six mechanisms, not previously incorporated in the model, that improved its behavior: (1, 2) dependence of tillering and leaf appearance rate on carbohydrate concentration, (3, 4) dependence of leaf appearance and leaf elongation rate on plant phenological stage, (5) sprouting of new tillers from decapitated generative tillers, and (6) proportionality of the number of elongating leaves with tiller size. Incorporation of these mechanisms, followed by reparameterization using a Metropolis-Hastings Monte Carlo method, improved performance statistics (r² = 0.521, normalized RMSE = 0.415) and explained the long duration of slow growth after early cutting. These mechanisms may thus be keys to understanding timothy regrowth.

Abbreviations: CV, coefficient of variation • LAI, leaf area index • RMSE, root mean squared error