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

This Article Abstract Figures Only Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Skinner, R. H. Articles by Gilmanov, T. G. Search for Related Content PubMed Articles by Skinner, R. H. Articles by Gilmanov, T. G. Agricola Articles by Skinner, R. H. Articles by Gilmanov, T. G. Related Collections Crop Physiology & Metabolism Other Forage Crops Crop Models

MODELING

Simulating Gross Primary Productivity of Humid-Temperate Pastures

^a USDA-ARS, Pasture Systems and Watershed Management Research Lab., Bldg. 3702 Curtin Rd., University Park, PA 16802
^b INRA, Agrocampus Rennes, UMR 1069, Sol, Agro- et hydro-systèmes, Spatialisation, F-35000 Rennes, France
^c Dep. of Biology and Microbiology, AgH 310, P.O. Box 2207B, South Dakota State Univ., Brookings, SD 57007

^* Corresponding author (howard.skinner{at}ars.usda.gov).

	ABSTRACT

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
Although most pasture growth models simulate many above- and belowground components of the plant community, calibration and validation are usually based only on periodic measurements of aboveground forage yield. This research used daily measurements of gross primary productivity (GPP) to validate the photosynthesis subroutine of the Integrated Farm System Model (IFSM). The model was calibrated for a pasture grazed by beef cattle (Bos taurus) in 2003, then validated with data from 2004 through 2006. Predicted and observed annual yield differed by 14 ± 9%, whereas predicted GPP differed from observed GPP by only 7 ± 3%. Seasonal trends in GPP were also adequately simulated, although a slight overestimation in the spring and early-summer and underestimation in the later half of the year occurred. Overestimation occurred when wintertime temperatures were above freezing or when N availability was high following fertilizer application. Late-season underestimation was related to low soil N availability which resulted from excessive N uptake by plants earlier in the year. Only minor adjustments in model structure were needed to improve simulation of GPP. Most adjustments involved changes in parameter values, many of which are often difficult to find or lacking in the literature. Refinement of models to accurately simulate the seasonal distribution of physiological parameters such as GPP will help ensure that model structures correctly represent the true dynamics of C assimilation and pasture growth.

Abbreviations: IFSM, Integrated Farm System Model • GPP, gross primary productivity • SPUR, Simulation of Production and Utilization of Rangelands Model

	NOTES

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher.

¹ Mention of a specific brand name is for identification purposes only and does not constitute endorsement by the USDA-ARS at the exclusion of other appropriate sources.

Received for publication August 4, 2007.

	INTRODUCTION

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
THE IFSM (Rotz and Coiner, 2005) is a deterministic, process-based model that predicts effects of weather and management on hydrology and soil nutrient dynamics, forage and crop yields, milk or beef production, and farm economics in temperate regions at a whole-farm scale. Future enhancements intend to incorporate a whole-farm C budget into the model framework. The model was recently enhanced to represent the growth and competition of multiple plant species in pastures and their effects on pasture productivity and botanical composition in temperate climates (Corson et al., 2006). This enhanced model incorporated plant, water, and soil components of the Simulation of Production and Utilization of Rangelands Model (SPUR 2.4; Foy et al., 1999). The revised pasture model was calibrated and validated with field data from experimental pastures in central Pennsylvania containing primarily orchardgrass (Dactylis glomerata L.) and white clover (Trifolium repens L.). Although this is a comprehensive model that simulates both above- and below-ground components of the pasture community, calibration and validation were based only on periodic measurements of aboveground forage yield. No information was available on how well the model simulated the physiological processes such as photosynthetic CO₂ uptake (GPP) that underlie yield predictions.

Properly managed agricultural systems have the ability to sequester C in the soil profile (Follett et al., 2001; West and Marland, 2002; Lal, 2004). This ability is becoming of great interest as human-induced increases in atmospheric CO₂ concentration contribute to global climate change (Karl and Trenberth, 2003). Whole-farm models, such as IFSM, provide a means to evaluate the ability of different production systems to optimize plant and animal productivity and economic returns while at the same time delivering ecological goods and services such as increased soil C storage and reduced emissions of other greenhouse gasses. To predict whole-farm C dynamics, a plant growth model must be able to accurately predict photosynthetic inputs and respiratory losses. However, few data sets are available that contain the year-round measurements of canopy photosynthesis and respiration necessary to properly validate such models.

Micrometeorological measurements of CO₂ exchange over pastures provide one means of monitoring net C gain or loss on a daily basis. Partitioning of net fluxes into photosynthetic and respiratory components permits further refinement of our understanding of the processes influencing the seasonal and yearly grassland C balance (Gilmanov et al., 2003; Novick et al., 2004; Verburg et al., 2004; Xu and Baldocchi, 2004). Continuous CO₂ flux measurements are obtained using eddy covariance systems that correlate changes in vertical wind velocity with changes in atmospheric CO₂ concentration. Fluxes are commonly measured at a frequency of 10 Hz and averaged over time steps of 20 to 30 min. Such frequent measurements allow the effects of changes in the environment, or in management practices such as fertilization and defoliation, to be rapidly quantified.

This paper describes the use of daily micrometeorological measurements of GPP to evaluate and improve the photosynthesis subroutine of the pasture growth component of the recently revised SPUR/IFSM. Such refinements will help ensure that model structures correctly represent the inter-annual and seasonal dynamics of C acquisition and pasture growth.

	MATERIALS AND METHODS

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
Development, sensitivity analysis and predictive ability of the pasture growth component of IFSM have been previously described (Corson et al., 2006). Briefly, in the photosynthesis subroutine, photosynthetic rate without stress limitations was initially calculated as a function of photosynthetically active radiation and leaf area index. This optimal photosynthetic rate was then adjusted for limiting factors including water stress, temperature stress, and leaf N concentration. A stress multiplier value ranging from 0 to 1 was calculated for each potential limiting factor and the most limiting factor was identified as the one with the smallest multiplier value. The optimal photosynthetic rate was then multiplied by the stress multiplier value to obtain the actual photosynthetic rate. The model was initially parameterized for a rotationally grazed, cool-season grass–legume pasture in central Pennsylvania (Corson et al., 2006).

Field data for testing the photosynthetic outputs of the model were collected from 2003 through 2006 from a pasture at the Penn State University Haller Research Farm located about 10 km northeast of State College, PA. This grass-dominated permanent pasture was cut once for hay in the spring of 2003 then rotationally grazed by beef cattle three to four times per year for the rest of the monitoring period. The pasture had been under grazing management since 1968. It was last reseeded in 1982 and was dominated by a mixture of cool-season grasses including orchardgrass, tall fescue (Festuca arundinacea Schreb.), smooth bromegrass (Bromus inermis Leyss.), and Kentucky bluegrass (Poa pratensis L.).

Pasture-scale CO₂ fluxes were quantified beginning in May 2002 using a Campbell Scientific¹ eddy covariance CO₂ flux system featuring a LI-7500 open path CO₂/H₂O analyzer and CSAT3 3-D sonic anemometer. This system uses micrometeorological techniques to monitor biosphere-atmosphere exchanges of CO₂ by correlating fluctuations in vertical wind velocity with CO₂ density (Dugas et al., 1991). Data were collected continuously at 10 Hz and averaged over 20-min time intervals. Coordinate rotation, frequency response corrections (Moore, 1986), corrections for density effects due to heat and water vapor transfer (Webb et al., 1980), and corrections for internal and external heating of the LI-7500 (Burba et al., 2006) were applied to the raw CO₂ flux data. The open path CO₂/H₂O analyzer and CSAT3 3-D sonic anemometer were placed at a height of 1.75 m above the soil surface in the center of a 7 ha pasture, providing >200 m fetch in the direction of the prevailing winds. Plant canopy height was generally <0.3 m during the growing season, although canopy height approached 1.0 m before the spring cutting in 2003. Canopy height during the dormant season was <0.1 m.

A complete description of the algorithm for flux partitioning between GPP and ecosystem respiration can be found in Gilmanov et al. (2003). In summary, on each day for which flux, soil temperature (T_soil), and photosynthetically active radiation (PAR) data were available, GPP was derived from a light-response function F = f(PAR) or F = f(PAR,T_soil). Its parameters (initial slope, plateau, curvature, daytime respiration, temperature exponent of daytime respiration, daytime respiration at T_soil = 0, etc.) were estimated using iterative nonlinear regression. Using a 7-d window, dependence of nighttime respiration on soil temperature, R_night = g(T_soil), was identified.

Missing data were gap-filled using: (a) linear interpolation for gaps 2 h; (b) light-response functions for missing daytime data; and (c) temperature-response function for nighttime data. When nighttime temperature response data were not available, average daytime respiration or temperature dependence of the respiration term of the daytime light response function were used. Averaged over the 4 yr, 37% of daytime flux measurements had to be gap-filled for one reason or another. For the rare days for which measurements were completely missing (19 d total over 4 yr), interpolation was performed at the daily time step using either linear interpolation, or nonlinear regression functions of daily gross photosynthesis and daily ecosystem respiration on available predictors (daily PAR, T_soil, soil moisture, etc.). Because of limitations in the models ability to simulate the soil component of ecosystem respiration, this report focuses on GPP, which is strictly a plant physiological process.

Biomass production was determined during the growing season (March through November) by taking weekly rising-plate meter readings calibrated against clipped biomass obtained on a monthly basis. Monthly harvests and plate meter measurements were also made during the winter to provide a complete description of live biomass distribution throughout the year. With the exception of the one cutting for hay in the spring of 2003, all biomass harvests occurred as grazing by beef cattle using management intensive rotational grazing practices. Under this system, pastures were subdivided into approximately 0.5 ha paddocks and each paddock was typically grazed for 3 to 4 d by 10 to 25 cows or cow/calf pairs depending on the amount of available forage and number of available cows. Model structure, however, called for biomass removal on a single day, which was set as the mid-point of each grazing event. The pasture received N fertilizer as urea twice each year at rates of 56 kg N ha^–1 in April and 34 or 45 kg N ha^–1 in August, except in 2006 when the spring and late-summer rates were 67 and 22 kg N ha^–1, respectively. When pastures were grazed, 37% of the consumed biomass was considered to have been returned to the pasture as manure. Manure deposition rate was based on forage digestibility results from a nearby study employing similar forage species and grazing management (Soder et al., 2006).

	RESULTS

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
Initial testing of the default parameters and structure of the model, using 2003 data, resulted in simulated annual forage yield of 331 g dry matter m^–2 compared with measured forage yield of 391 g dry matter m^–2 (a 15% underestimation). Simulated yield at the first of four harvests was 99% of observed yield. Simulated yield then decreased relative to observed yield for each subsequent harvest until the model predicted no measurable forage yield for the final harvest (Table 1 ). Simulated total annual GPP was very close to measured annual GPP, differing by 4%. Even though little difference existed between simulated and observed annual results, modeled GPP was severely overestimated at the beginning of the growing season and greatly underestimated thereafter.

Two parameters relating to seed production and frost kill were adjusted to essentially remove them from the simulation. Because pastures were managed to maintain the sward in a vegetative state, the date seed production begins (CRIT6) was set to Day 365. Also, field studies have suggested that winter mortality is low, and that these cool-season pastures retain some green leaf area and are capable of photosynthetic uptake at subfreezing temperatures approaching the original temperature for frost kill used in the model (Skinner, 2007). Therefore, the temperature for frost kill (CRIT2) was reduced from –6°C to –20°C to ensure that low temperatures did not eliminate all subsequent physiological activity.

Following model adjustments, predictions of total annual GPP increased by 136 g CO₂ m^–2, and the difference between the simulated and observed uptake for the calibration data in 2003 was reduced to 1% (Table 1). Goodness-of-fit calculations between observed and predicted GPP(Gauch et al., 2003) indicated that 68% of the prediction error came from scatter around the 1:1 line, 32% from systemic bias, with no mean bias. Although simulated GPP remained greater than observed GPP during the first growth period and less than observed for most growth periods thereafter, the discrepancy for all harvests was greatly reduced compared with the initial model. Simulated forage yield increased to 390 g dry matter m^–2 in the revised model, which was almost identical to observed yield (Table 1).

Model outputs were validated using 2004, 2005, and 2006 data. The model was able to simulate annual yield (Fig. 1 ) to within 14 ± 9% of observed values. Simulated yield exceeded observed yield by 99 g m^–2 in 2006, with all of the overestimation occurring during the fourth of five harvests taken that year. Simulated GPP averaged –7 ± 3% of observed GPP (Fig. 2 ), with predicted values differing form observed values by no more than 16% in any given year. Both simulated and observed GPP showed similar seasonal dynamics (Fig. 3 ). Values were low from January through March, increased rapidly to a maximum in May, then gradually decreased throughout the summer and autumn to a low in December that was similar to January rates. As with the calibration year, GPP was slightly overestimated during the first half of the year, then underestimated from June on, with the exception of September, when simulated GPP exceeded observed GPP following fertilizer applications in late August of each year.

View larger version (13K): [in this window] [in a new window]   Fig. 1. Observed vs. simulated yearly forage yield for a cool-season grass pasture in central Pennsylvania. Yearly totals are the sum of four to five harvests per year.

View larger version (14K): [in this window] [in a new window]   Fig. 2. Observed vs. simulated yearly gross primary productivity (GPP) for a cool-season grass pasture in central Pennsylvania.

View larger version (21K): [in this window] [in a new window]   Fig. 3. Monthly distribution of measured and simulated gross primary productivity (GPP) for a cool-season grass pasture in central Pennsylvania. Data are averaged over 2004 through 2006 and error bars indicated variability among years ( ± 1 SE).

View larger version (23K): [in this window] [in a new window]   Fig. 4. Daily measured (symbols) and simulated (lines) gross primary productivity (GPP) from 1 January through 31 May of each year. Horizontal bars indicate days when snow cover was >2 cm. Arrows indicate dates when the pasture was grazed.

View larger version (28K): [in this window] [in a new window]   Fig. 5. Influence of air temperature on observed and simulated daily gross primary productivity (GPP) during the first 3 mo of the year.

	DISCUSSION

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

Aber et al. (1996) validated a generalized model of forest photosynthesis with daily C balance data from the Harvard Forest. Averaged over 4 yr, the model predicted daily GPP to within 2% of observed values. They found no major discrepancies between predicted and observed values, although their model tended to underestimate mid-summer GPP. Wohlfahrt et al. (2001) conducted a similar analysis for a multi-species model of vegetation/atmosphere CO₂ exchange in mountain grasslands. They only monitored net ecosystem exchange (NEE) and did not provide estimates of GPP. However, regressions of daily predicted vs. observed NEE for two sites were highly significant (P < 0.01) with intercepts near zero and slopes of 0.94 to 0.96. Together these results suggest that models can be developed that accurately predict GPP across a range of ecosystems.

The model was also able to provide excellent estimates of total forage yield, differing from observed values by 1% in 2004 then overestimating yield by 14% in 2005 and 30% in 2006. Weather conditions in 2004 were almost identical to the calibration year of 2003, whereas 2005 and 2006 experienced varying degrees of drought stress. Interestingly, periods when GPP was overestimated early in the year (Table 1, Fig. 3) were accompanied by an underestimation of aboveground biomass, while underestimation of GPP later in the year was accompanied by overestimation of yield. Similarly, overestimations of yield in 2005 and 2006 were accompanied by underestimations of GPP in both years.

Comparisons with data from root cores suggested that too much C was being partitioned by the model to roots early in the year, with little root growth occurring after the first harvest (data not shown). This overestimation of root biomass during the spring occurred despite the reduction in maximum root to shoot ratio (PR9) and increases in root respiration (PR24) and root mortality (PR25). Overestimation of root growth reflects the rangeland origin of the SPUR model since rangeland grasses tend to have relatively large root systems. Further refinement of the model to better match seasonal differences in GPP and yield will have to take into consideration partitioning of C within the plant and the complex environmental interactions that affect C uptake and partitioning (Davidson, 1969; Saggar et al., 1997; Stewart and Metherell, 1999).

Aber et al. (1996) found that simulation of GPP was much more sensitive to parameters related to maximum photosynthetic rate than to those related to light, temperature, or leaf mass. In the default parameter set of the SPUR-IFSM, maximum photosynthetic rate (P_max) was 28.8 µmol m^–2 s^–1, and the model underpredicted annual GPP in 2003 by 180 g CO₂ m^–2. Increasing P_max to 35 µmol m^–2 s^–1 (Table 2) improved predictions of annual GPP to within 44 g CO₂ m^–2 (1%) of the observed amount. Stout (1994) also found it necessary to increase the photosynthetic rate in SPUR to properly simulate growth of grasslands in the northeastern United States. The greatest observed mid-day photosynthetic rates in the current study ranged from a low of 25 µmol m^–2 s^–1 in 2005 to a high of 45 µmol m^–2 s^–1 in 2004 (data not shown). Other studies have reported single-leaf P_max values of 27 µmol m^–2 s^–1 for orchardgrass (Peri et al., 2003) and canopy P_max of about 45 µmol m^–2 s^–1 for well-fertilized tall fescue (Gastal and Belanger, 1993).

Much of the overestimation of GPP in the beginning of the year occurred on relatively warm days in January, February, and March (Fig. 5). Photosynthetic activity in many plant species can occur at subzero temperatures as long as temperatures are greater than the freezing point of water in the leaves (Pisek, 1973). Skinner (2007) showed that in the absence of snow cover, photosynthetic CO₂ uptake occurred on these pastures at daytime temperatures near –5°C. Even after incorporating subfreezing photosynthetic uptake into the model structure, simulated GPP was often much lower than observed GPP at temperatures well below 0°C. Overestimation of wintertime GPP was most common at temperatures near or above freezing (Fig. 5). Part of the overestimation during winter resulted from the lack of inclusion of snow cover in the weather data set, and the subsequent inability of the model to reduce photosynthetic uptake when snow cover limited light availability (Fig. 4). The model also appeared to respond more rapidly than the actual pasture to improved light, temperature, and canopy cover conditions in the early spring. Plant processes do not always respond instantaneously to improving environmental conditions but require time for genetic and physiological responses to be translated into increased photosynthetic rates. Incorporating lag periods into simulated responses during periods of otherwise favorable environmental conditions could improve correlations between simulated and observed uptake rates.

In addition to air temperature, major constraints to CO₂ uptake throughout the year were water availability and leaf N status (Table 3). When averaged over all 3 yr, shoot N content was the most limiting factor for 53% of the days. Although there were relatively few days during May each year when simulated gross photosynthesis greatly overestimated measured values (Fig. 4), all occurred on days when leaf area index was high and shoot N stress was minimal after fertilizer application in the spring produced short-term spikes in soil N availability. Modeled N uptake by plants was primarily controlled by plant demand rather than by supply, and often occurred faster than soil replacement through fertilizer and manure applications or soil mineralization. Because of this, soil in the model tended to become depleted of N during the summer. This lack of N appeared to be the major constraint limiting GPP after the first harvest in 2004, and during the fall and early winter in 2005 and 2006. Moisture stress was the main limiting factor on only 1 d in 2004, but was more important during the summers of 2005 and 2006. Although N limitations were partially improved by the changes to model structure listed in Table 2, additional changes in model structure will be necessary to improve late-season simulations of GPP. Adjusting parameter values in the current model beyond the changes shown in Table 2 tended to have unacceptable effects on yield, suggesting that additional improvement in simulating GPP will require substantial changes to the actual structure of the model.

	CONCLUSIONS

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
Access to daily photosynthesis measurements provided an excellent opportunity to evaluate key components of the SPUR/IFSM pasture growth submodel. Following adjustments to model parameters, annual GPP could be predicted with a high degree of accuracy, as could general seasonal patterns in C uptake. Improvements in components of the model's photosynthetic CO₂ uptake routine will enhance its ability to simulate the whole farm C budget. Total annual yield could be predicted more easily than yield of individual harvests. On a seasonal basis, photosynthetic uptake was not always closely related to forage yield, primarily because of differences in root–shoot partitioning. This disconnect between CO₂ uptake and yield underscores the importance of properly simulating each component of the pasture C budget. Further refinement of the model to accurately simulate the seasonal distribution of additional functions, including respiratory loss and root development, will help ensure that model structures correctly represent all aspects of pasture C dynamics.

All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher.

¹ Mention of a specific brand name is for identification purposes only and does not constitute endorsement by the USDA-ARS at the exclusion of other appropriate sources.

	REFERENCES

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 

• Aber, J.D., P.B. Reich, and M.L. Goulden. 1996. Extrapolating leaf CO₂ exchange to the canopy: A generalization model of forest photosynthesis compared with measurements by eddy correlation. Oecologia 106:257–265.[CrossRef][Web of Science]
• Burba, G.G., D.J. Anderson, L. Xu, and D.K. McDermitt. 2006. Correcting apparent off-season CO₂ uptake due to surface heating of an open path gas analyzer: Progress report of an ongoing study. In Proc. Agric. Forest Meteorol. 27th Conf., San Diego, CA. 22–25 May 2006. Am. Meteorol. Soc., Boston, MA.
• Calanca, P., N. Vuichard, C. Campbell, N. Viovy, A. Cozic, J. Fuhrer, and J.-F. Soussana. 2007. Simulating the fluxes of CO₂ and N₂O in European grasslands with the Pasture Simulation Model (PaSim). Agric. Ecosyst. Environ. 121:164–174.
• Corson, M.S., R.H. Skinner, and C.A. Rotz. 2006. Modification of the SPUR rangeland model to simulate species composition and pasture productivity in humid temperate regions. Agric. Syst. 87:169–191.[Web of Science]
• Davidson, R.L. 1969. Effect of root/leaf temperature differentials on root/shoot ratios in some pasture grasses and clover. Ann. Bot. (London) 33:561–569.[Abstract/Free Full Text]
• Dugas, W.A., L.J. Fritschen, L.W. Gay, A.A. Held, A.D. Matthias, D.C. Reicosky, P. Steduto, and J.L. Steiner. 1991. Bowen ratio, eddy correlation, and portable chamber measurements of sensible and latent heat flux over irrigated spring wheat. Agric. For. Meteorol. 56:1–20.
• Follett, R.F., J.M. Kimble, and R. Lal. 2001. The potential of U.S. grazing lands to sequester soil carbon. p. 401–430. In R.F. Follett et al. (ed.) The potential of U.S. grazing lands to sequester carbon and mitigate the greenhouse effect. CRC Press, Boca Raton, FL.
• Foy, J.K., W.R. Teague, and J.D. Hanson. 1999. Evaluation of the upgraded SPUR model (SPUR 2.4). Ecol. Modell. 118:149–165.[Web of Science]
• Gastal, F., and G. Belanger. 1993. The effects of nitrogen fertilization and the growing season on photosynthesis of field-grown tall fescue (Festuca arundinacea Schreb.) canopies. Ann. Bot. (London) 72:401–408.[Abstract/Free Full Text]
• Gauch, H.G., J.T.G. Hwang, and G.W. Fick. 2003. Model evaluation by comparison of model-based predictions and measured values. Agron. J. 95:1442–1446.[Abstract/Free Full Text]
• Gilmanov, T.G., D.A. Johnson, and N.Z. Saliendra. 2003. Growing season CO₂ fluxes in a sagebrush-steppe ecosystem in Idaho: Bowen ratio/energy balance measurements and modeling. Basic Appl. Ecol. 4:167–183.
• Karl, T.R., and K.E. Trenberth. 2003. Modern global climate change. 2003. Science (Washington, DC) 302:1719–1723.[Abstract/Free Full Text]
• Lal, R. 2004. Soil carbon sequestration impacts on global climate change and food security. Science (Washington, DC) 304:1623–1627.[Abstract/Free Full Text]
• Moore, C.J. 1986. Frequency response corrections for eddy correlation systems. Boundary Layer Meteorol. 37:17–35.
• Novick, K.A., P.C. Stoy, G.G. Katul, D.S. Ellsworth, M.B.S. Siqueira, J. Juang, and R. Oren. 2004. Carbon dioxide and water vapor exchange in a warm temperate grassland. Oecologia 138:259–274.[CrossRef][Web of Science][Medline]
• Peri, P.L., D.J. Moot, and D.L. McNeil. 2003. A canopy photosynthesis model to predict the dry matter production of cocksfoot pastures under varying temperature, nitrogen and water regimes. Grass Forage Sci. 58:416–430.
• Pisek, A. 1973. Effect of temperature on metabolic processes. 1. Photosynthesis. p. 102–127. In H. Precht et al. (ed.). Temperature and life. Springer-Verlag, New York.
• Riedo, M., A. Grub, M. Rosset, and J. Fuhrer. 1998. A Pasture Simulation Model for dry matter production, and fluxes of carbon, nitrogen, water and energy. Ecol. Modell. 105:141–183.[Web of Science]
• Rotz, C.A., and C.U. Coiner. 2005. The Integrated Farm System Model Reference Manual. Available at http://www.ars.usda.gov/Main/docs.htm?docid=8519 (verified 5 Mar. 2008).
• Saggar, S., C. Hedley, and A.D. Mackay. 1997. Partitioning and translocation of photosynthetically fixed ¹⁴C in grazed hill pastures. Biol. Fertil. Soils 25:152–158.
• Skinner, R.H. 2007. Winter carbon dioxide fluxes in humid-temperate pastures. Agric. For. Meteorol. 144:32–43.
• Soder, K.J., M.A. Sanderson, J.L. Stack, and L.D. Muller. 2006. Intake and performance of lactating cows grazing diverse forage mixtures. J. Dairy Sci. 89:2158–2167.[Abstract/Free Full Text]
• Stewart, D.P.C., and A.K. Metherell. 1999. Carbon (¹³C) uptake and allocation in pasture plants following field pulse-labeling. Plant Soil 210:61–73.[CrossRef][Web of Science]
• Stout, W.L. 1994. Evaluation of the SPUR model for grasslands of the northeastern United States. Agron. J. 86:1001–1005.[Abstract/Free Full Text]
• Thornley, J.H.M., and M.G.R. Cannell. 1997. Temperate grassland responses to climate change: An analysis using the Hurley Pasture Model. Ann. Bot. (London) 80:205–221.[Abstract/Free Full Text]
• Verburg, P.S., J.A. Arnone, III, D. Obrist, D.E. Schorran, R.D. Evans, D. Leroux-Swarthout, D.W. Johnson, Y. Luo, and J.S. Coleman. 2004. Net ecosystem carbon exchange in two experimental grassland ecosystems. Glob. Change Biol. 10:498–508.
• Webb, E.K., G.I. Pearman, and R. Leuning. 1980. Correction of flux measurements for density effect due to heat and water vapour transfer. Q. J. R. Meteorol. Soc. 106:85–100.[CrossRef]
• West, T.O., and G. Marland. 2002. A synthesis of carbon sequestration, carbon emissions, and net carbon flux in agriculture: Comparing tillage practices in the United States. Agric. Ecosyst. Environ. 91:217–232.[CrossRef]
• Wohlfahrt, G., M. Bahn, U. Tappeiner, and A. Cernusca. 2001. A multi-component, multi-species model of vegetation-atmosphere CO₂ and energy exchange for mountain grasslands. Agric. For. Meteorol. 106:261–287.
• Xu, L., and D.D. Baldocchi. 2004. Seasonal variation in carbon dioxide exchange over a Mediterranean annual grassland in California. Agric. For. Meteorol. 123:79–96.

This Article Abstract Figures Only Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Skinner, R. H. Articles by Gilmanov, T. G. Search for Related Content PubMed Articles by Skinner, R. H. Articles by Gilmanov, T. G. Agricola Articles by Skinner, R. H. Articles by Gilmanov, T. G. Related Collections Crop Physiology & Metabolism Other Forage Crops Crop Models