SEPATOU : simulator of rotational grazing management in a dairy production system

Presentation
Modeling
Software
Demonstration
Documentation

 

Presentation


Context

Market, environmental and production economics concerns together with the promotion of sustainable agriculture are urging for a reconsideration of feeding systems of herbivores.  In France maize silage has taken an increasing importance in the feeding of dairy cows in the past decades. The current heavy use of maize silage can be explained by economic influences such as grants for maize silage. Another reason stems from the management difficulties associated with grazing. It is particularly hard to determine in advance appropriate rates for nitrogen fertilizer, amounts of hay or silage, the area to be allocated per cow and rotation policy that will meet the needs given specific constraints. These difficulties may have contributed to the decline in the use of grazing for feeding cows. The trend is changing however. Indeed, in many areas grazing can to some extent replace feeding of maize silage, thereby reducing pollution risks, at least during the period of rapid grass growth. Moreover, grazed grass is less expensive than maize silage and presents a more attractive image of dairy produce to consumers.

Dairy farmers need to find a compromise between a high grazing efficiency and low N losses on the one hand, and constraints which exist on farm resources or economic targets on the other. The feeding management problem is a difficult one. It may change from one year to another because the stock of maize silage available may differ and the size and characteristics of the herd may also vary. Moreover due to climate variability the herbage growth fluctuates largely over years and within a season. The underlying control problem involves a multivariable optimization with both direct immediate effects (e.g. cow intake) and indirect delayed effects on the seasonal scale (availability and quality of the pasture for subsequent grazing) or annual scale (feed shortage or excessive silage surplus). An appropriate quantity/quality trade-off of the available herbage should be maintained throughout the period under consideration. The herbage growth rate can be controlled to some extent by appropriate nitrogen fertilization or by grazing pressure. Too much offered herbage per cow can be as big a problem as too little. It has been shown that in order to have herbage of good quality, grazing should be intense and regular. The turnout time and the timing of rotations must be carefully chosen to match the state of the pasture. At some period, poor quality or an excess of herbage can be corrected by harvesting some pastures as hay or silage. In addition, the optimum production (per cow or per unit area) from any milk production system will depend on the constraints in which the farmers operate (milk quota per ha for example) as well as available resources at the farm enterprise level (land availability, labour force). One of the roles of research is to create tools enabling to infer system malfunctions from observations, assess the soundness of technical choices, and help formulate new management practices adapted to specific production systems and objectives. The SEPATOU simulator is one such a tool that allows virtual experimentation of the implementation of feeding management strategies in a dairy farm from mid-winter to mid-summer.
 
 


Problem approach and goal

The approach underlying SEPATOU puts emphasis on the modeling and simulation of the interactive dynamics between the farmer’s decision behavior, the biophysical system (field, feed stocks, cows) and the climatic factors. A key aspect of agricultural management practices is that successful farmers tend to anticipate situations by relying on flexible plans that span the full production period, taking into account contingencies and various constraints such as those on production resources (e.g. maize silage stock available for complementary feeding). At the core of the system lies a modeling formalism in which the management strategy that underlies a dairy farmer’s decision-making behavior can be expressed. A management strategy is defined as a set of planned tasks adaptable to fluctuation of weather.  It specifies how to make determine appropriate daily operational choices (e.g. diet composition of conserved feed, field to graze)  successively throughout the production period depending on the situation encountered and how to adjust the task plans when necessitated by the occurrence of particular events.

SEPATOU simulates the day-to-day dynamics of the farmer's decision process and the response biophysical system for which models of grass growth, animal consumption and milk production are used. SEPATOU provides the means to evaluate and compare tentative strategies by simulating their application throughout the production season under one or several climatic scenarios that are either taken from climate records or generated by a climate generator. The simulation outputs that might be used for evaluation include:

Variations over years can also be examined.

SEPATOU has been designed for use by grazing experts as a means of training extension agents (or dairy farmers) in the operational management of rotational grazing systems and help them test and therefore discover innovative management strategies. By providing the opportunity to formulate alternative management behaviors and dairy farm configurations SEPATOU allows virtual experimentation that can serve as a tangible basis for discussion of issues and as a way of making understandable the complexity of interactions between pasture, animal and management. This capability makes it valuable for disseminating knowledge and improving  management practices empirically. SEPATOU does not aim to support on-farm decision-making for any particular farm but rather to provide a comprehensive view of the impact that management decisions might have on typical production systems considered over a given season and under different weather scenarios. No attempt is made to match very closely any existing system, which would require an extremely intensive modeling and data collection effort; the modeled production systems are artificial representative examples of real cases that are slightly simplified but realistic as regards management.

Project participants

The conceptual and software developments of SEPATOU have been realized by:


The simulator has been evaluated in two Fench regions : Brittany and Aveyron (Ségala). In both cases, different types of production system configurations (farm and management strategy) have been formalized and simulated.
This phase has benefited from contributions of:


Last modifications: December 2002