|CORSON, MICHAEL - INRA, RENNES, FRANCE
|Rotz, Clarence - Al
Submitted to: Agricultural Systems
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 9/24/2008
Publication Date: 11/1/2008
Citation: Skinner, R.H., Corson, M.S., Rotz, C.A. 2008. Comparison of Two Pasture Growth Models of Differing Complexity. Agricultural Systems. 99:35-43.
Interpretive Summary: With the increasing interest in sequestering carbon in agricultural soils, models are needed that can predict management and environment effects on the processes controlling carbon gain and loss from pastures. Such models must be complex enough to capture key processes yet simple enough to allow for parameterization with readily available data. Two pasture growth models that shared many common features, but differed in model complexity, were incorporated into the Integrated Farm System Model (IFSM) and compared for their ability to predict photosynthesis, forage yield, and shoot respiration. The main difference between models was the inclusion of roots in the complex model and their absence in the simple model. Simulated shoot respiration was always greater in the simple model but no field data were available to determine which model provided the best estimate of observed respiration rates. On average, little difference existed between models in their ability to predict photosynthesis and yield although large differences for specific years and sites were sometimes observed. Our results suggest that the ability of the complex model to simulate roots was not needed to adequately simulate photosynthesis, respiration and shoot growth on these pasture systems.
Technical Abstract: Two pasture growth models that share many common features but differ in model complexity have been developed for incorporation into the Integrated Farm System Model (IFSM), a whole-farm model that predicts effects of weather and management on hydrology, soil nutrient dynamics, forage and crop yields, milk or beef production, and farm economics. Major differences between models include the explicit representation of roots in the more complex model and their effects on carbon portioning and growth. The overall goal was to develop a model capable of representing growth, competition, and ecosystem carbon fluxes among multiple plant species in pastures while maintaining a relatively simple model structure that minimizes the number of required user inputs. Models were compared against observed yield data for twelve site-years from three experiments in central Pennsylvania, USA. Each model performed best under conditions that were similar to those under which it was developed. The simple model was equal to the complex model for yield predictions and for predictions of gross primary productivity (GPP), despite the fact that the complex model was specifically developed to optimize simulation of GPP. Greatest differences between models were in the simulation of shoot respiration and its effect on carbon partitioning between above and belowground plant tissues. These results highlight the importance of avoiding model over parameterization, thereby matching the complexity of a model’s structure with its main goals.