Modeling the Biomass and Harvest Index Dynamics of Timothy

Abstract

Residual stubble, following harvest or the grazing of perennial grasses, is an important element of the C cycle in agricultural systems that has not been considered in most growth models. Th e stubble fraction from aboveground biomass can be quantifi ed by predicting the harvest index (HI) of forage grasses, i.e., the amount harvested or grazed as a proportion of the aboveground biomass. A HI simulation module, using new functions to describe shoot apex height and plant weight density, was developed and integrated into the timothy (Phleum pratense L.) growth model CATIMO. Th e model was calibrated and validated using data from two timothy cultivars and diff erent N fertilizer rates from fi ve experiments conducted in eastern Canada. Th e root mean square errors (RMSEs) between simulated and measured values for HI (0.11), shoot apex height (7 cm), stubble biomass (25 g dry matter m -2 ), aboveground biomass (68 g dry matter m -2 ), and harvestable biomass (76 g dry matter m -2 ) were acceptable with normalized RMSEs (13-41%) compared with the coeffi cients of variation of measured data (13-30%). Th e HI also has implications for determining forage nutritive value. Th is improved CATIMO model provides a framework to explore options for optimizing yield and nutritive value and to quantify the stubble biomass of timothy within the context of C sequestration.

Timothy Yield and Nutritive Value under Climate Change in Canada

“…We compared the crop attributes (outputs) simulated by the timothy‐specific grass model CATIMO (Bonesmo and Bélanger, 2002a, 2002b; Bonesmo et al, 2005; Jing et al, 2011, 2012) with baseline conditions for the period 1961 to 1990 and four future climate series for the period 2040 to 2069.…”

“…The CATIMO model has been evaluated previously with experimental data from eastern and western Canada and northern Europe (Bonesmo and Bélanger, 2002a, 2002b; Bonesmo et al, 2005; Jing et al, 2011, 2012), and is able to differentiate the diversity of timothy performance across different agro‐ecological regions with different climate conditions.…”

Simulating Timothy Growth and Nutritive Value with Observed and Synthetic Weather Data

“…In crop model evaluations, the NRMSE between simulated and observed attributes is generally deemed acceptable when it is in the range from 10 to 30% (Jamieson et al, 1991), which was the case in our previous studies (Bonesmo and Bélanger, 2002a; Bonesmo et al, 2005; Jing et al, 2011, 2012a). In this study, the low NRMSE values (≤4.5%) illustrate the high similarity between crop attributes that are simulated from synthetic and observed weather data.…”

“…The low temperature at Normandin might explain it having the lowest NDF concentrations. In the model, DM yield and nutritive value are calculated by a number of complex processes that are impacted by weather conditions (Bonesmo and Bélanger, 2002a, 2002b; Jing et al, 2011, 2012a). Similar variations for DM yield and nutritive value, simulated from synthetic and observed weather data, provide some confidence for applying synthetic weather data to simulate timothy growth and nutritive value within the context of assessing the impact of climate change.…”

“…The CATIMO model has been evaluated for biomass production and nutritive value with independent experimental data for several sites in eastern and western Canada, Finland, and Norway (Bonesmo and Bélanger, 2002a, 2002b; Bonesmo et al, 2005; Jing et al, 2011, 2012a, 2012b). The simulated dynamics of DM yield, N uptake, NDF concentration, and dNDF were close to the observed values, with NRMSE for each crop attribute of spring growth and summer regrowth that were acceptable and in line with the NRMSE standard defined by Jamieson et al (1991).…”