Agronomy Journal 94:337-345 (2002)
© 2002 American Society of Agronomy
MODELING
Timothy Yield and Nutritive Value by the CATIMO Model
I. Growth and Nitrogen
Helge Bonesmoa and
Gilles Bélanger*,b
a Norwegian Crop Res. Inst., Kvithamar Res. Cent., NO-7500 Stjordal, Norway
b Soils and Crops Res. and Dev. Cent., Agric. and Agri-Food Can., 2560 Hochelaga Blvd., Sainte-Foy, QC, Canada G1V 2J3
* Corresponding author (belangergf{at}em.agr.ca)
Received for publication December 4, 2000.
 |
ABSTRACT
|
|---|
Mechanistic simulation models can assist in developing recommendations to optimize yield and nutritive value and in understanding the complex interaction among plant growth, nutritive value, and environmental conditions. In this paper, we present the growth and N concentration modules of an integrated model [CATIMO (Canadian Timothy Model)] of timothy (Phleum pratense L.) primary growth and nutritive value. This growth model features radiation interception and use efficiency, leaf and stem growth, leaf senescence, and a N function based on the critical N concentration of whole plants. Model parameters were calibrated to key model attributes: leaf area index (LAI); forage N concentration; and leaf, stem, and forage dry matter (DM) yields. Calibration measurements were taken weekly on timothy primary growth in four different years at one location (Fredericton, NB, Canada). Overall, the model satisfactorily fitted the measured values with root mean square errors of 32.8, 42.0, and 65.9 g m-2 leaf, stem, and forage DM yield, respectively. The model tended to underestimate stem DM yield at the end of the primary growth cycle, overestimate forage N concentration under nonlimiting N conditions, and underestimate N concentration under limiting N conditions. The model satisfactorily fitted LAI in 3 of 4 yr. Summary statistics of the calibration indicate a successful description of growth and development of the essential plant components required for modeling digestibility.
Abbreviations: DM, dry matter • LAI, leaf area index • PAR, photosynthetically active radiation • RGRL, relative growth rate of leaf area • RMSE, root mean square error of estimation • RUE, radiation use efficiency • SLAn, specific leaf area of new leaves • TSUM, temperature sum
|
INTRODUCTION
|
|---|
FORAGE NUTRITIVE VALUE OF TIMOTHY decreases with advancing phenological development, but dry matter (DM) yield increases. An acceptable compromise between the decreasing nutritive value and increasing DM yield is to harvest timothy at the early heading stage (Berg et al., 1996). Traditionally, this has been a well-accepted practice in eastern Canada and other timothy-growing areas. However, due to increasing demand for more resource-efficient production, this general recommendation should be replaced by site-specific guidelines that enable grass farmers to adapt harvesting strategies to their own individual situations.
Mechanistic computer simulation models, based on daily weather data and initial soil conditions, can assist in developing site-specific recommendations to optimize harvest yield and nutritive value. Furthermore, mechanistic models help in our understanding of the complex interaction among plant growth, plant nutritive value, and environmental conditions. The level of detail required depends on the objectives and the data available to construct and run the model. It is desirable to build a model one or two hierarchical levels below that required for predictions; e.g., crop models should incorporate process models from the plant and organ levels (Whisler et al., 1986). Such growth models, based on radiation interception and radiation use efficiency (RUE) (Monteith, 1977), have been successfully developed for wheat (Triticum aestivum L.) (CERES; Ritchie and Otter, 1985), soybean [Glycine max (L.) Merr.] (Sinclair, 1986), potato (Solanum tuberosum L.) (LINTUL; Spitters and Schapendonk, 1990), ryegrass (Lolium perenne L.) (LINGRA; Schapendonk et al., 1998), and wheat and corn (Zea mays L.) (STICS; Brisson et al., 1998). These models are able to balance the complexity of various modules and require a limited amount of input data to run.
For forage nutritive value, statistical models are the norm, usually formulated without reference to yield (Fick et al., 1994). Gustavsson et al. (1995), however, formulated an integrated model of timothy growth and nutritive value but without linkages between crop growth and nutritive value through the plant components.
Our overall objective was to develop an integrated model of timothy growth and nutritive value based on the energetic approach of radiation interception and RUE, the plant morphological components, the N requirements defined by a model of N dilution, and easily obtainable input data. In this paper, we describe the growth and N modules of the model and their calibration. The digestibility module of the model is described in a companion paper (Bonesmo and Bélanger, 2002) in this issue of Agronomy Journal.
|
MATERIALS AND METHODS
|
|---|
Field Experiments
Field experiments were conducted on timothy primary growth following winter at the Fredericton Research Centre of Agriculture and Agri-Food Canada (45°55' N lat) in 1991, 1992, 1993, and 1995 (Table 1). Data from three cultivars in 1991, two cultivars in 1992, four N rates in 1993, and one cultivar in 1995 were extracted from those experiments. Cultivars and N rates were replicated four times, and the plot size was at least 10 m2. In each year, plots were sampled weekly to determine DM yield of forage, leaves, and stems. Weekly measurements of leaf area index (LAI) and forage N concentration were also recorded. The N concentration in plant tissue was determined by the sulfuric acid (H2SO4)–hydrogen peroxide (H2O2) digestion method described by Richards (1993). Nitrogen fertilizer was applied in early May of each year. Initial soil inorganic N contents from 0 to 45 cm were 2.49 g m-2 in 1993 and 7.31 g m-2 in 1995; no measurements were taken in 1991 and 1992. Detailed descriptions of the 1991, 1992, and 1993 experiments were reported previously (Bélanger and Richards, 1995, 1997).
Daily global radiation, maximum and minimum air temperatures, precipitation amounts, and hours of bright sunshine were obtained from an automatic climatic station at the experimental site. The daily mean air temperature was calculated as the average of the maximum and minimum air temperatures. Maximum and minimum air temperatures and sunshine hours were used to calculate potential evapotranspiration according to Baier and Robertson (1965). Among the 4 yr of the study, the growth period was warmest in 1991 and coldest in 1993 (Table 2). Precipitation amounts were the least in 1991 and the greatest in 1992.
View this table:
[in this window]
[in a new window]
|
Table 2. Mean values of weather conditions from 1 May to the end of the growth period for the four field experiments used for model calibration.
|
|
Soil moisture was estimated from soil textural composition, soil organic matter, and dry bulk density, using relationships reported by Riley (1996). Water stored in the tension range from 0.01 to 0.1 MPa is regarded as readily plant available in soil layers above 60 cm while water of the high tension storage from 0.1 to 1.5 MPa is assumed to be plant available only above a 30-cm depth (Riley, 1992). Water available to plants estimated for a soil at field capacity and from the surface to a depth of 60 cm was 104 mm for the Riverbank soil and 106 mm for the Fredericton soil. Both soils were classified as Orthic Dystric Brunisol according to Canadian classification and Typic Dystrochrepts according to USDA classification. At the onset of primary growth, the soil was assumed to be at field capacity because of water from snowmelt and very low potential evapotranspiration. Growth was assumed to begin on 1 May in 1991, 1992, and 1993, but on 24 April in 1995; these dates were based on field observations.
Model Description
The model applies to the primary growth of timothy, which starts shortly after snowmelt and ends approximately 450°C-d later when timothy is normally harvested at the early heading stage. Most parameters were estimated by calibration to key model attributes: LAI; forage N concentration; and leaf, stem, and forage DM yield. A realistic calibration range, reflecting the possible variation for each parameter, was specified (Tables 3 and 4). Four parameter estimates were set according to literature values.
View this table:
[in this window]
[in a new window]
|
Table 4. Estimates and calibration ranges of parameters accounting for N deficient conditions in the timothy growth model.
|
|
Radiation Use Efficiency
The model was based on the approach of RUE (Sinclair and Muchow, 1999). This approach offers a relatively simple, mechanistically based description of the key factors that influence the accumulation of DM. The model was designed to simulate timothy growth by assuming a constant potential RUE (RUEpot) (Table 3). The actual RUE [RUEact, g DM MJ-1 photosynthetically active radiation (PAR)] was computed daily using functions that decreased RUEpot due to suboptimal growing conditions:
 | [1] |
where f(PAR) and f(T) are reduction functions that account for suboptimal daily PAR and temperature, respectively, and f(w) and f(N) are reduction functions that account for suboptimal water and N conditions, respectively. The term min is a programming function that selects an entry of lowest numerical value. f(PAR) and f(T) were multiplicative, whereas the minimum-value approach was used for f(w) and f(N).
The reduction at high daily PAR was calculated as follows:
[2] where PAR is the cumulative daily PAR (MJ m-2) and PARc is the threshold PAR above which the linear decline occurs with slope coefficient PARcoeff. The daily PAR was calculated as daily incoming global radiation multiplied by 0.48 (Varlet-Grancher et al., 1982).
The reduction function for temperature was:
 | [3] |
where T is the daily mean air temperature (°C) and f(T) is 0 at or below the base temperature for growth and development (Tb) and 1 at the optimum temperature for RUE (To). The term max is a programming function that selects an entry of highest numerical value. The reduction functions that account for suboptimal water and N conditions [f(w) and f(N)] are described by Eq. [13] and Eq. [16], respectively.
Leaf Area Growth and Senescence
The LAI of green leaves was obtained by integrating the daily net result of the leaf area growth rate (GLAI) and senescence rate (DLAI) (Spitters and Schapendonk, 1990). During early growth, the rate of leaf extension is constrained more by temperature than by the supply of assimilates (Gastal et al., 1992). In this early stage, leaf area increases exponentially with time:
where LAI is in m2 leaf m-2 soil, LAIc is critical LAI, and RGRL is the relative growth rate of leaf area, expressed in degree-days (°C-d). Crop N status affects RGRL (Eq. [17]).
Above an LAIc where new leaves do not increase light interception, the phase of exponential leaf area growth changes into a linear growth phase (Goudriaan and Monteith, 1990). In the linear phase, the model calculates leaf area growth using the product of the increase in leaf weight, GLV (g m-2 d-1), and specific leaf area of new leaves, SLAn (m2 leaf g-1 leaf DM):
 | [4b] |
The SLAn was calculated as a function of temperature, based on the assumption that SLAn increases up to an optimum temperature and then decreases (Solhaug, 1991):
 | [5] |
where SLAmin is the minimum SLAn, SLAadd is the maximum increase of SLAn due to temperature (T), and f(T)SLA is a reduction function that accounts for suboptimal T:
 | [6a] |
 | [6b] |
where ToSLA is the optimum temperature of SLAn.
Leaf senescence was assumed to begin when the temperature sum (TSUM) above a base temperature of 0°C, from initial growth, reached 260°C-d (Bélanger, 1996), after which it was assumed to be affected by temperature. The senescence rate of LAI (DLAI, m2 leaf m-2 soil d-1) is described on the basis of a relative death rate due to aging (RDR); DLAI was zero for TSUM < 260°C-d. For TSUM 
260°C-d, DLAI was defined by:
 | [7] |
where RDRmax is the maximum value of RDR and f(T)RDR is the relative linear effect of temperature between 0 for T < 5°C and 1 for T 30°C. A relative death rate due to aging as a function of temperature was used in the potato version of the LINTUL model (Spitters and Schapendonk, 1990).
Crop Growth Rate and Dry Matter Production of Leaves and Stems
The daily crop growth rate (GC, g m-2 DM) was calculated by multiplying actual RUE by the amount of PAR intercepted. The daily values of intercepted PAR were derived by assuming that light interception increases with LAI according to a negative exponential function, with an extinction coefficient (K):
 | [8] |
Dry weights (g m-2 DM) of the green leaves, senescent leaves, and stems were obtained by integrating the respective growth and death rates. The growth rates of leaves (GL) and stems including sheaths (GS) were calculated as:
 | [9] |
 | [10] |
The proportion of GC partitioned to leaves (FL) was related to the phenological development, described by the TSUM:
where TSUMc is the TSUM at the initial decrease in FL, FLmax is the maximum fraction of GC (Eq. [8]) partitioned to leaves, and TSUMh is the TSUM at heading stage.
The senescence rate of leaves in terms of DM weight (DL) was zero for TSUM < 260°C-d. For TSUM 260°C-d, DL was defined using the same relative senescence rate (RDR) that applies to LAI (Eq. [7]) multiplied by the DM weight of the green leaves (WGL):
 | [12] |
The disappearance rate of senescent leaves is assumed to be similar to DL but with a time delay. Hence, the disappearance rate of senescent leaves on a given day is equal to DL before the time delay. The time delay was required to exclude senesced leaves from LAI but keep them on the plants until they fall off for the calculation of forage nutritive value.
Water
The drought stress [f(w)] was calculated as the estimated ratio between actual evapotranspiration (Ea) and potential evapotranspiration (Ep) from plants:
 | [13] |
where Ep (mm d-1) and Ea (mm d-1) were derived from a model that estimates soil water evaporation and plant evapotranspiration separately (Ritchie, 1972); the model was expanded to include a soil water budget (Skjelvåg, 1981). The combined model calculated potential and actual evapotranspiration from plants on the basis of potential evapotranspiration, LAI, and the content of plant readily and weakly available water in the root zone. The upper limit of cumulative evaporation from soil during an uninterrupted Drying Phase 1 in Ritchie's (1972) model was set to 8 mm, and the hydraulic conductivity coefficient determining the soil water evaporation rate in Drying Phase 2 was set at 3.7 mm d-0.5 according to the soil classified as sandy loam (Soil Survey Staff, 1975). In Drying Phase 1, the soil is sufficiently wet for the water to be transported to the surface at a rate at least equal to the evaporation potential. In Phase 2, soil water evaporation depends on the flux of water through the upper layer of the soil.
Nitrogen
Nitrogen deficiency reduces grass yield, mainly by limiting photosynthetic capacity and restricting leaf area expansion and, thus, light interception (Bélanger et al., 1992). Consequently, in our model, N deficiency affects those parameters associated with RUE and leaf area expansion (RGRL and SLAn). The crop N status is expressed by the relative N concentration (RNC); that is, the ratio of the actual N concentration (Na) to the critical N concentration (Nc):
 | [14] |
where Nc represents the N concentration required to reach maximum shoot growth. The Nc is related to crop biomass (W) (Bélanger and Gastal, 2000):
 | [15] |
where Ncmax is the maximum value of critical N concentration and NDcoeff is the N dilution coefficient (Table 4). For W < 100 g m-2 DM, Nc was set equal to Ncmax.
The effect of the N status on RUE [f(N)] was calculated according to Bélanger and Richards (1997):
 | [16] |
In the early exponential growth phase, the effect of N deficiency on the leaf area expansion was accounted for by a reduction of RGRL:
 | [17] |
where RGRLmax is the maximum value of RGRL (Eq. [4a]). Based on the examination of field data, we set RGRLmax equal to 0.015 (°C-d)-1. For LAI > LAIc, the effect of N deficiency on the leaf area expansion was accounted for by a reduction of SLAn:
where SLANn is the reduced SLAn.
The actual crop N concentration (Na) was calculated by assuming that a fraction of the absorbed N (FNH) remained in the aboveground biomass. The FNH increases with increasing N status of the crop (Brégard et al., 2000). In our model, FNH was related to RNC:
 | [19] |
where FNHmax is the maximum FNH.
The total N in above- and belowground biomass was obtained by integrating the daily rate of N absorption. The daily rate of N absorption was assumed to be at the minimum of crop demand (Eq. [20]) and soil supply (Eq. [22]). The N demand, including above- and belowground biomass (Ndem, g N m-2), was characterized by the difference between the amount of N required to go from limited conditions (Na) to the maximum limit of N absorption (Nm), described by a N dilution curve of the aboveground biomass (Brisson et al., 1998):
 | [20] |
The Nm was calculated as:
 | [21] |
where Nmmax is the maximum value of Nm.
The potential supply of N to the crop (Nsup, g N m-2) was assessed by:
 | [22] |
where Nsoil is the current soil mineral N content and FNA is an availability factor for soil N. The FNA was assumed to be constant when the soil N content was above a critical value (NSc), which represents the soil N content for maximum N availability. Below NSc, the FNA decreased linearly with the soil N content:
 | [23] |
where FNAmax is the maximum fraction of the available soil mineral N. We computed daily Nsoil by integrating the initial soil mineral N content, rates of N fertilization, N mineralization, and crop N absorption. The daily rate of N mineralization (RM, g N m-2 d-1) was related to the daily mean air temperature by the use of the combined rate constant for all soil processes (Kall) (Kirschbaum, 1995):
 | [24] |
where RMmax is maximum rate of mineralization. The Kall was calculated as:
[25] where T is the daily mean air temperature. Kirschbaum (1995) used soil temperature, but we chose to include air temperature in CATIMO because these data are more generally available.
Parameter Estimation and Model Performance
The calibration of the CATIMO model was based on data from weekly samplings of green leaf and stem DM yield and LAI. For N concentration, however, only the 1993 measurements were used to obtain an appropriate balance between limiting and nonlimiting N conditions. The program package Powersim Solver was used to estimate a set of parameters (Tables 3 and 4) that minimized the root mean square error of estimation (RMSE) over all experiments (Powersim, 1996; Bonesmo, 1999). A linear regression analysis between simulated and measured values was carried out to quantify possible over- or underestimation of the simulations.
|
RESULTS
|
|---|
The growth model parameters were not estimated at the limit of the specified calibration ranges, except for potential RUE and the maximum fraction of daily crop growth allocated to leaves (FLmax) (Table 3). The potential RUE was estimated to be at the highest value (3.0 g DM MJ-1 PAR) of the calibration range, whereas the maximum fraction of daily crop growth allocated to leaves was estimated at the lower value (0.80).
For the nonlimiting N conditions of 1991, 1992, and 1993, the simulated DM yield of leaves and stems showed good agreement with the measurements (Fig. 1)
. Larger discrepancies between simulated and measured leaf and stem DM yields were observed in 1995, particularly for the leaves (Fig. 1). The slope between simulated and measured values was 0.69 for leaf DM yield and 0.80 for stem DM yield, indicating that the leaf and stem DM yields were overestimated early in the primary growth cycle and underestimated at the end of the primary growth cycle (Table 5). The RMSE was greater for stem DM yield than for leaf DM yield (Table 5); this was due to the larger underestimation of stem growth at the end of the simulated growth period (Fig. 1). The simulated DM yield of leaves and stems generally showed good agreement with the measurements under the two levels of N-limiting conditions in 1993 (Fig. 2)
. However, with 7 g N m-2, the model underestimated the stem DM yield at the end of the simulation period by as much as 138 g m-2. With no N applied, the simulated values of both leaf and stem DM yield on the first measurement date were 13 g m-2 higher than the measured values.
View larger version (25K):
[in this window]
[in a new window]
|
Fig. 1. Measured and simulated DM yield of leaves and stems of timothy grown under nonlimiting N conditions in four separate experiments. The data come from three cultivars in 1991, two cultivars in 1992, and one cultivar in 1993 and 1995.
|
|
View larger version (14K):
[in this window]
[in a new window]
|
Fig. 2. Measured and simulated DM yield of leaves and stems of timothy cv. Champ grown under two levels of N limiting conditions in 1993 (N0, 0 g N m-2; N7, 7 g N m-2).
|
|
The additional estimated parameters for limiting N conditions were also within the limits of the calibration range, except for the maximum fraction of absorbed N (FNHmax), which was at the upper limit (0.8) of the range (Table 4). The calibration range for the soil-related parameters [maximum fraction of the available soil mineral N (FNAmax) and critical soil N content for maximum N availability (NSc)] was wide, probably because of the limited information on soil processes.
Under nonlimiting N conditions in 1993, there was good agreement between simulated and measured values of N concentration (Fig. 3) . For the limiting N conditions, however, N concentration was underestimated by as much as 0.0086 g N g-1 DM. The model overestimated N concentration for the other 3 yr, especially in 1991 when the simulated N concentration was as much as 0.0207 g N g-1 DM higher than the measured value. The normalized RMSE of N concentration based on all data was greater than that of forage DM yield (Table 5).
View larger version (20K):
[in this window]
[in a new window]
|
Fig. 3. Measured and simulated N concentration of timothy grown under limiting and nonlimiting N conditions. The model was calibrated to the 1993 measurements.
|
|
With the exception of 1995, there was a good relationship between the simulated and measured LAI and forage DM yield (Fig. 4
; Table 5). As a result of underestimated stem growth, the model also tended to underestimate forage DM yield at the end of the simulated growth period (Table 5). The normalized RMSE of LAI based on all data was greater than that of forage DM yield (Table 5).
View larger version (19K):
[in this window]
[in a new window]
|
Fig. 4. Simulated LAI and DM yield plotted as a function of measured values of timothy grown under nonlimiting and limiting N conditions in four separate experiments.
|
|
|
DISCUSSION
|
|---|
The growth module of the CATIMO model divides forage DM yield into leaves, stems, and senescent material. This feature is crucial for the development of the associated digestibility module. It is well established that the decrease in the leaf/stem ratio and the increase in the amount of senescent material contributes significantly to the decrease of forage digestibility with time (Marten, 1985); it is also well established that the proportion of those plant components are highly related to weather and managerial factors (Buxton and Casler, 1993). The calibration of the model was possible because the data set included weekly samplings, a prerequisite for dynamic modeling, and also because essential forage components were measured to determine digestibility. Furthermore, necessary weather and soil data for the field sites were available.
The CATIMO growth model is based on accepted and validated models using RUE (Monteith, 1977), LAI (Spitters and Schapendonk, 1990), plant N (Brisson et al., 1998; Bélanger and Gastal, 2000), and plant-available soil moisture (Ritchie, 1972; Skjelvåg, 1981). The parameters are formulated a priori to be biologically defined and related to essential processes of crop growth and development. In addition, the calibration procedure resulted in physiologically reasonable parameter values (Tables 3 and 4). The number of parameters fitted by the calibration procedure is large compared with the number of measurements. The model parameters, however, are distinctly different from coefficients of traditional multiple regression analyses because they have a biological significance. Overall, there is good agreement between measured and simulated values of key model attributes (Table 5). The leaf and stem DM yields were calculated satisfactorily by the model. The RMSE for forage DM yield in our study (66 g m-2 DM) (Table 5) is similar to the 52 g m-2 DM reported by Gustavsson et al. (1995). Those researchers, however, did not calibrate their timothy growth model to LAI and leaf and stem DM yields. The relatively large RMSE for LAI (1.28 m2 leaf m-2 soil) was primarily due to the large difference between simulated and measured values in 1995 (Fig. 4). When excluding 1995, the RMSE for LAI was 0.72 m2 leaf m-2 soil.
The RUE was estimated to 3.0 g DM MJ-1 PAR, which was the upper limit of the calibration range. Assuming that plant DM has an energy content of 17.5 kJ g-1 (Monteith, 1977), this represents an efficiency of conversion of PAR into DM of 5.1%, which is similar to that found in three other timothy cultivars (6.5, 5.2, and 3.9%) grown under near-optimal growing conditions (Sheehy and Cooper, 1973). The maximum fraction allocated to leaves (FLmax) was estimated to 0.8, which was the lower limit of the calibration range. This might seem low, but the stem fraction in the model included leaf sheaths.
The summary statistics of the calibration indicate a successful description of growth, development, and N concentration although some features of the model need to be clarified. Firstly, the CATIMO model tends to overestimate N concentration for nonlimiting conditions and underestimate N concentration for limiting N conditions. The daily simulations of crop N requirements and soil N supply were calibrated to only the 1993 data. The crop demand follows a well-established approach, but the modeling of the soil N supply may, at best, be regarded as a simplification of a mechanistic model. This simplified approach to the soil N supply is justified in our study because our main focus was on crop growth and nutritive value. However, a more detailed but simple N model, such as the Azodyne model developed for northwestern European conditions (Jeuffroy and Recous, 1998), might calculate the soil N supply more precisely. The RMSE of 0.0089 g N g-1 DM for N concentration (Table 5) was greater than the 0.0025 g N g-1 DM reported by Gustavsson et al. (1995). They attributed the large errors under N-limiting conditions to difficulties in estimating the soil N supply.
The model was not thoroughly calibrated to limiting conditions for water because the included soil moisture model component indicated that there was no water stress, except for a few days in 1991 and 1992. Drought stress has a major influence on RUE. Muchow (1985) reported decreased RUE under water-limited conditions; thus, the effect of water stress was included in the calculation of actual RUE (Eq. [1]). Water stress might also influence light interception through the maximum RGRL and the SLAn.
The range of the model applicability might be improved by including additional functional relationships. The date of initial growth was an input in the present model. Attempts to calculate growth initiation on the basis of the first occurrence of 5 d with a diurnal mean air temperature of 5°C on the prerequisite of snowless ground, as in Bonesmo (1999), was unsuccessful. Reserves of N and carbohydrates at growth initiation have profound effects on the regrowth of perennial grasses (Volenec et al., 1996). The underestimation of leaf growth and the large difference between simulated and measured LAI in 1995 (Fig. 1) might be explained, in part, by substantially greater root N and carbohydrate reserves that resulted from favorable conditions in the previous fall and winter. Daylength also affects growth rates (Deinum et al., 1981). The data used for calibrating the model were all from the same site and latitude so that the effect of phenological development was accounted for by only the TSUM-dependent parameters. Daylength would have to be included if the model was to be used to calculate growth at other latitudes.
The CATIMO model successfully describes the growth and development of the essential plant components that are required for modeling digestibility. The calibration of the digestibility module of the model is described in a companion paper in this issue of Agronomy Journal (Bonesmo and Bélanger, 2002). Work is underway to validate the model using a broader independent data set from a wider selection of sites in eastern Canada.
|
NOTES
|
|---|
Contrib. no. 714, Agric. and Agri-Food Can.
|
REFERENCES
|
|---|
- Baier, W., and G.W. Robertson. 1965. Estimation of latent evaporation from simple weather observations. Can. J. Plant Sci. 45: 276–284.
- Bélanger, G. 1996. Morphogenetic and structural characteristics of field-grown timothy cultivars differing in maturity. Can. J. Plant Sci. 76:277–282.
- Bélanger, G., and F. Gastal. 2000. Nitrogen utilization by forage grasses. Can. J. Plant Sci. 80:11–20.
- Bélanger, G., F. Gastal, and G. Lemaire. 1992. Growth analysis of a tall fescue sward fertilized with different rates of nitrogen. Crop Sci. 32:1371–1376.[Abstract/Free Full Text]
- Bélanger, G., and J.E. Richards. 1995. Growth analysis of timothy cultivars differing in maturity. Can. J. Plant Sci. 75:643–648.
- Bélanger, G., and J.E. Richards. 1997. Growth analysis of timothy grown with varying N nutrition. Can. J. Plant Sci. 77:373–380.
- Berg, C.C., A.R. McElroy, and H.T. Kunelius. 1996. Timothy. p. 643–664. In L.E. Moser et al. (ed.) Cool-season forage grasses. Agron. Monogr. 34. ASA, CSSA, and SSSA, Madison, WI.
- Bonesmo H., and G. Bélanger. 2002. Timothy yield and nutritive value by the CATIMO model: II. Digestibility and fiber. Agron. J. 94:345–350 (this issue).[Abstract/Free Full Text]
- Bonesmo, H. 1999. Modeling spring growth of timothy and meadow fescue by an expolinear growth equation. Acta Agric. Scand., Sect. B. 49:216–224.
- Brégard, A., G. Bélanger, and R. Michaud. 2000. Nitrogen use efficiency and morphological characteristics of timothy populations selected for low and high forage N concentrations. Crop Sci. 40: 422–429.[Abstract/Free Full Text]
- Brisson, N., B. Mary, D. Ripoche., M.H. Jeuffroy, F. Ruget, B. Nicoullaud, P. Gate, F. Devienne-Baret, R. Antonioletti, C. Durr, G. Richard, N. Beaudoin, S. Rescous, X. Tayot, D. Plenet, P. Cellier, J.-M. Machet, J.M. Meynard, and R. Delécolle. 1998. STICS: A generic model for simulation of crops and their water and nitrogen balances: I. Theory and parameterization applied to wheat and corn. Agronomie 18:311–346.
- Buxton, D.R., and M.D. Casler. 1993. Environmental and genetic effects on cell wall composition and digestibility. p. 685–714. In H.G. Jung et al. (ed.) Forage cell wall structure and digestibility. ASA, CSSA, and SSSA, Madison, WI.
- Deinum, B., J. Beyer, P.H. Nordfeldt, A. Kornher, O. Ostgard, and G. Bogaert. 1981. Quality of herbage at different latitudes. Neth. J. Agric. Sci. 29:141–150.
- Environment Canada. 1993. Canadian climate normals, 1961–1990. Minister of Supply and Serv. Can., Ottawa, ON.
- Fick, G.W., P.W. Wilkens, and J.H. Cherney. 1994. Modeling forage quality changes in the growing crop. p. 757–795. In G.C. Fahey, Jr. (ed.) Forage quality, evaluation, and utilization. ASA, CSSA, and SSSA, Madison, WI.
- Gastal, F., G. Bélanger, and G. Lemaire. 1992. A model of the leaf extension rate of tall fescue in response to nitrogen and temperature. Ann. Bot. 70:437–442.[Abstract/Free Full Text]
- Goudriaan, J., and J.L. Monteith. 1990. A mathematical function for crop growth based on light interception and leaf area expansion. Ann. Bot. 66:695–701.[Abstract/Free Full Text]
- Goudriaan, J., and H.H. van Laar. 1994. Modeling potential crop growth processes. Kluwer Academic Publ., Dordrecht, the Netherlands.
- Gustavsson, A.-M., J.F. Angus, and B.W.R. Torrsell. 1995. An integrated model for growth and nutritional value of timothy. Agric. Syst. 47:73–92.
- Jeuffroy, M.-H., and S. Recous. 1998. Azodyn: A simple model simulating the nitrogen deficiency for decision support in wheat fertilization. Eur. J. Agron. 10:129–144.
- Kirschbaum, M.U.F. 1995. The temperature dependence of soil organic matter decomposition, and the effect of global warming on soil organic C storage. Soil Biol. Biochem. 27:753–760.
- Landström, S. 1990. Influence of soil frost and air temperature on spring growth of timothy in northern Sweden. Swed. J. Agric. Res. 20:147–152.
- Marten, G.C. 1985. Factors influencing feeding value and effective utilization of forages for animal production. p. 89–97. In Proc. Int. Grassl. Congr., 15th, Kyoto, Japan. 24–31 Aug. 1985. Jpn. Soc. of Grassl. Sci., Kyoto, Japan.
- Monteith, J.L. 1977. Climate and the efficiency of crop production in Britain. Philos. Trans. R. Soc. London, Ser. B. 281:277–294.
- Muchow, R.C. 1985. An analysis of the effects of water deficits on grain legumes grown in a semiarid tropical environment in terms of radiation interception and its efficiency of use. Field Crops Res. 11:309–323.
- Powersim. 1996. Powersim 2.5, reference manual. Powersim Corp., Herndon, VA.
- Richards, J. 1993. Chemical characterization of plant tissue. p. 115–139. In M.R. Carter (ed.) Soil sampling and methods of analysis. Lewis Publ., Boca Raton, FL.
- Riley, H. 1992. Assessment of simple drought indices on the growth of timothy grass (Phleum pratense L.). Norw. J. Agric. Sci. 6:333–348.
- Riley, H. 1996. Estimation of physical properties of cultivated soils in southeast Norway from readily available soil information. Norw. J. Agric. Sci. (Suppl. 25).
- Ritchie, J.T. 1972. Model for predicting evaporation from a row crop with incomplete cover. Water Resour. Res. 8:1204–1213.
- Ritchie, J.T., and S. Otter. 1985. Description and performance of CERES-Wheat: A user oriented wheat yield model. ARS 38. USDA-ARS, Beltsville, MD.
- Schapendonk, A.H.C.M., W. Stol, D.W.G. van Kraalingen, and B.A.M. Bouman.1998. LINGRA, a sink/source model to simulate grassland productivity in Europe. Eur. J. Agron. 9:87–100.
- Sheehy, J.E., and J.P. Cooper. 1973. Light interception, photosynthetic activity and crop growth rate in canopies of six temperate grasses. J. Appl. Ecol. 10:239–250.
- Sinclair, T.R. 1986. Water and nitrogen limitations in soybean grain production: I. Model development. Field Crops Res. 15:125–141.
- Sinclair, T.R., and R.C. Muchow. 1999. Radiation use efficiency. Adv. Agron. 65:215–265.
- Skjelvåg, A.O. 1981. Experimental and statistical methods of plant experiments used in an agroclimatic investigation in Aust-Agder, Norway. Acta Agric. Scand. 31:343–357.
- Soil Survey Staff. 1975. Soil taxonomy. A basic system of soil classification for making and interpreting soil surveys. Agric. Handb. 436. U.S. Gov. Print. Office, Washington, DC.
- Solhaug, K.A. 1991. Influence of photoperiod and temperature on dry matter production and chlorophyll content in temperate grasses. Norw. J. Agric. Sci. 5:365–383.
- Spitters, C.J.T., and A.H.C.M. Schapendonk. 1990. Evaluation of breeding strategies for drought tolerance in potato by means of crop growth simulation. Plant Soil 123:193–203.
- Varlet-Grancher, C., R. Bonhomme, M. Chartier, and P. Artis. 1982. Efficience de la conversion de l'énergie solaire par un couvert végétal. Oecol. Plant. 3:3–26.
- Volenec, J.J., A. Ourry, and B.C. Joern. 1996. A role for nitrogen reserves in forage regrowth and stress tolerance. Physiol. Plant. 97: 185–193.
- Whisler, F.D., B. Acock, D.N. Baker, R.E. Fye, H.F. Hodges, J.R. Lambert, H.E. Lemmon, J.M. McKinion, and V.R. Reddy. 1986. Crop simulation models in agronomic systems. Adv. Agron. 40: 141–208.
This article has been cited by other articles:
|
|
|
|
 
M. van Oijen, M. Hoglind, H. M. Hanslin, and N. Caldwell
Process-Based Modeling of Timothy Regrowth
Agron. J.,
August 17, 2005;
97(5):
1295 - 1303.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
H. Bonesmo, G. Belanger, E. Charmley, R. Drapeau, D. B. McKenzie, R. Michaud, and G. F. Tremblay
Timothy Yield and Nutritive Value by the CATIMO Model: III. Validation for Eastern Canada
Agron. J.,
January 1, 2005;
97(1):
32 - 40.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
H. Bonesmo and G. Belanger
Timothy Yield and Nutritive Value by the CATIMO Model: II. Digestibility and Fiber
Agron. J.,
March 1, 2002;
94(2):
345 - 350.
[Abstract]
[Full Text]
[PDF]
|
|
|
TOP
ABSTRACT
INTRODUCTION
