CHARACTERIZING AND SIMULATING A ROTATIONAL GRAZING STRATEGY
Marie-Josée Cros
Roger Martin-Clouairevisibility
…
description
4 pages
link
1 file
Sign up for access to the world's latest research
checkGet notified about relevant papers
checkSave papers to use in your research
checkJoin the discussion with peers
checkTrack your impact
Abstract
The management of dairy production systems based on rotational grazing and maize-concentrate feeding requires a complex lookahead decision making procedure to elaborate a seasonal strategy that is sufficiently robust to uncertainty. In this paper, the problem is characterized rather precisely as finding for the late winter to early summer period an appropriate combination of commitments concerning the fields allocated to grazing, those kept just in case of need, the time profile of maizeconcentrate distribution, the nitrogen fertilization policy and the field rotation policy. A simulation tool designed to evaluate tentative man-made strategies under given climate scenarios is described.
Related papers
A biophysical dairy farm model to evaluate rotational grazing management strategies
Agronomie, 2003
This paper describes a biophysical dairy farm model developed as part of the SEPATOU system that simulates pasture-dominated feeding strategies in a dairy cow enterprise. It can reproduce the effects of various technical management options applied to a set of grazing fields on a daily basis over a period of several months. The model is made up of three submodels that deal with the soil (availability of water), sward and animal (cow intake and milk yield) components. It includes as driving variables the main factors that farmers can control (nitrogen fertilizer rate, defoliation frequency and intensity, composition of cows' diet) in order to attain some objectives such as the intended grazing herbage contribution in the cows' diet or the amount of milk produced per cow or per ha. The intended purpose of the SEPATOU system has led to several original developments. In order to simulate various defoliation regimes, growth and senescence processes are dissociated as there is no longer synchronization between them, and the effect of grazing intensity is expressed in terms of the ratio between the herbage mass after and before grazing. The animal intake submodel that can deal with mixed feeding combines two approaches, one based on energy requirements and the other on the relationship between herbage mass, herbage digestibility and intake. The model of herbage digestibility is based on herbage mass and takes into account its variation down the sward profile; this latter aspect plays a key role in the plant-animal interaction. The model has been validated using typical farm cases in Brittany. It provides realistic estimates of the state variables involved in the processes, such as herbage mass and daily milk yield, and it credibly predicts the timing of key events (e.g. date of turnout to grass, end of first grazing cycle).
Simulating management strategies: the rotational grazing example
Agricultural Systems, 2004
This paper argues in favour of a simulation approach relying on an explicit and rigorous modelling of the management strategy that underlies the farmerÕs decision-making behaviour. A strategy is defined as a roadmap of intended technical tasks over a management period. It is tied to an overall objective and specifies what to do depending on the situations encountered. An essential feature is its flexibility, enabling it to cope with stochastic fluctuations of the environment. In order to evaluate the worth of a strategy, the advocated approach relies on a simulation tool with which the effects of applying the strategy are evaluated under different hypothetical weather conditions. The example of a rotational grazing dairy production system is used to illustrate the modelling and simulation of a management strategy covering a production season. SEPATOU, the simulator built for this application domain, is designed to be used by extension services and farming system scientists. It provides a framework for virtual experimentation that can help to enhance decision-making capabilities of primary producers and find new or more profitable management strategies.
Development and validation of a mechanistic whole dairy farm model to evaluate farming strategies under grazing conditions in Uruguay
A mechanistic, dynamic whole farm simulation model was developed to evaluate the effect of farming strategies on the productivity of dairy grazing systems. The model integrates local available information on pasture growth and quality and current knowledge on animal nutrition and metabolism. The pastoral component simulates the pasture rotation structure of the farm, with variable number and size of paddocks, to which the user must assign a pasture type from an available database. Each pasture type is represented by initial herbage mass (HM) and two vectors: monthly dry matter (DM) growth rate values and organic matter digestibility (OMD) values. The model is driven by pasture growth rate (PGR) on a monthly interval step. Several pasture production and management strategies can be defined as a per paddock basis. The cows are defined in terms of their potential for milk production (MPP), body condition score (BCS, scale 1-5), biotype Frame (body weight with BCS of 3), calving date, and contents of fat and protein in milk. These variables are used to characterize the average of up to six groups of adult cows which are defined by the user to represent the current situation of a dairy farm or a theoretical system. Average grazing DM intake (DMI) of each calving group of cows is estimated considering animal factors: Frame, MPP and days in milk (DIM); pasture factors: OMD, pre-grazing HM (pg-HM) and substitution rate (SR) of supplementary feed. The model is based on metabolisable energy (ME) and environmental thermo neutrality is assumed. Total ME intake (MEI) is partitioned among body functions following a defined priority: maintenance, pregnancy, milk production potential and body reserves (BR). One distinct feature of this model is that the approach used implies an active role of BR in defining the partition of MEI. If ME balance for potential milk is not achieved then BR are mobilized at a constant rate (κ) to give an absolute amount which is proportional to the current size of estimated mass of BR, whose initial level is set when inputting the initial BCS. Another feature of this model is that it can manage decisions taken at different system levels (pasture rotation structure, annual DM yield and seasonal distribution, reserves production and supplementation strategies, variables stocking rates, effects of animal size, BCS, milk potential, etc.), to quantitatively assess the impact of these decisions on cows and farm productivity. The model output was initially validated at the "cow biotype level" using published farmlet trials. The relative prediction error (RPE) and concordance correlation coefficient (CCC) were used as measures of fitness; models with values of RPE less than 10% and values of CCC greater than 0.90 were considered to have significant predictive power. Daily milk yield per cow, live weight and BCS change through the lactation were validated using a set of 12 monthly values for each trait, obtained from cows of contrasting body sizes (Heavy and Light).The RPE and CCC were 16% and 0.94 in Heavy, 20 % and 0.87 in Light cows for milk yield; 3 % and 0.72 in Heavy, 2 % and 0.81 in Light cows for live weight; 6 % and 0.18 in Heavy and 9% and-0.47 in Light cows for BCS change. Monthly intake of pasture per ha was validated using another independent set of 12 average monthly values for each of 5 farmlet stocking rates treatments (2.2; 2.7; 3.1; 3.7 and 4.3 cows/ha). RPE and CCC were: 13% and 0.77; 9% and 0.87; 12% and 0.93; 13 % and 0.91; 16 % and 0.88 respectively. The model was responsive to contrasting cow type and farming management. These results show that the model has acceptable predictive power and can be used to better understand actual farming systems and also to evaluate the expected productive impact of some technical changes introduced at the farm level.
Simulating rotational grazing management
Environment International, 2001
Dairy systems predominantly based on rotational grazing are notoriously hard to manage. In order to ensure profitability, this type of production requires quite good organisation, planning, and operating capability on the part of the farmer. A simulation-based decision support system, called SEPATOU, has been developed for this purpose. At the core of the decision support approach lies an explicit and rigorous modelling of the management strategy that underlies a dairy farmer's decision-making behaviour (real or hypothetical). The SEPATOU system is a discrete-event simulator that reproduces the day-to-day dynamics of the farmer's decision process and the response of the controlled 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 different hypothetical weather conditions. The relative worth of a strategy can be assessed by analysing the effects on the biophysical system and their variability across the representative range of possible conditions that is considered. The activities to be managed concern the type and amount of conserved feed, where to fertilise and how much, the choice of fields to harvest, and most importantly, which field to graze next. Typically, SEPATOU is designed to be used by extension services and farming system scientists. It is implemented in C++ and is currently undergoing a validation process with the intended users. D
A methodology for designing and evaluating alternative cropping systems: Application on dairy and arable farms
Ecological Indicators, 2012
To improve the sustainability of agricultural systems of the Lombardia region (northern Italy), a mixed indicator-model-expert approach was used. Starting from the results of a previous assessment of current management (ACT) in dairy and arable farms, alternative management scenarios at field level were designed in order to reduce nitrogen (N) losses whilst maintaining or improving the environmental and economic sustainability at the farming system level. By working with a group of experts supported by a mechanisation model and a cropping system model, two alternative N management scenarios were defined following a step-by-step decision procedure. The first scenario (FERT) is an improvement of the current fertiliser management scheme, applied at the same crops as in ACT and aimed at maintaining the same yields. The second scenario (ROT) is based on changes in crop rotations by introducing new crops to reduce N losses and to maintain economic profitability. The sustainability of the two scenarios was assessed and compared with agro-ecological and economic indicators. The results of FERT, indicate that the application of adequate N management plans tuned to the production target and the promotion of best management practices may help to reduce N surplus and consequently to save fossil energy and to decrease the costs of production. In the ROT scenario, the introduction of alfalfa cultivation reduces N surplus on maize, whereas intensive double cropping systems (two crops harvested in 12 months) increase N surplus and require higher energy consumptions and production costs compared to cultivating a summer crop only. However, in rotational systems more favourable weed population dynamics are expected compared to ACT. Both alternative scenarios were not implemented in practice, but they are realistic and are consistent with results of experiments where management options similar to those introduced in FERT and ROT were tested. This work indicates that the rational integration between scientific tools (indicators and models) and expert knowledge is adequate to deal with complex farming and cropping systems, which require a multidisciplinary approach.
Simulating the influence of management decisions on the nutrient balance of dairy farms
Agricultural Systems, 2005
In order to evaluate the influence of management decisions on the nutrient balance of dairy farms a simulation model was developed. Three farm systems have been simulated: zero grazing, winter milk and summer milk. From the simulated farm systems the zero grazing farm has in all scenarios the lowest N-surplus. The winter milk farm system has a higher N-surplus than zero grazing but lower than the summer milk farm system. The results further indicate the positive effects of maize feeding in addition to grazing. More maize in the ration is especially good to lower the N-surplus during the grazing period in the summer. The benefits of more maize in the ration decrease when the fertilizer application rates decrease.
Optimising pasture and grazing management decisions on the Cicerone Project farmlets over variable time horizons
Animal Production Science, 2013
This study addresses the problem of balancing the trade-offs between the need for animal production, profit, and the goal of achieving persistence of desirable species within grazing systems. The bioeconomic framework applied in this study takes into account the impact of climate risk and the management of pastures and grazing rules on the botanical composition of the pasture resource, a factor that impacts on livestock production and economic returns over time. The framework establishes the links between inputs, the state of the pasture resource and outputs, to identify optimal pasture development strategies. The analysis is based on the application of a dynamic pasture resource development simulation model within a seasonal stochastic dynamic programming framework. This enables the derivation of optimum decisions within complex grazing enterprises, over both short-term tactical (such as grazing rest) and long-term strategic (such as pasture renovation) time frames and under climatic uncertainty. The simulation model is parameterised using data and systems from the Cicerone Project farmlet experiment. Results indicate that the strategic decision of pasture renovation should only be considered when pastures are in a severely degraded state, whereas the tactical use of grazing rest or low stocking rates should be considered as the most profitable means of maintaining adequate proportions of desirable species within a pasture sward. The optimal stocking rates identified reflected a pattern which may best be described as a seasonal saving and consumption cycle. The optimal tactical and strategic decisions at different pasture states, based on biomass and species composition, varies both between seasons and in response to the imposed soil fertility regime. Implications of these findings at the whole-farm level are discussed in the context of the Cicerone Project farmlets.
SEPATOU: a Decision Support System for the Management of Rotational Grazing in a Dairy Production
1999
The paper presents the simulator SEPATOU that can reproduce the day-to-day dynamics of two interactive systems: the decision system representing the dairy farmer's management behavior and the biophysical system that encompasses the herbage production, consumption and transformation into milk. The activities to be managed concern the type and amount of conserved feed, where to fertilize and how much, the choice of fields to cut and, most importantly, what field to graze next. Typically, SEPATOU is designed to be used by extension services and farming system scientists. It provides a flexible environment through which learning about a satisfying management strategy of a given dairy production system can take place by iterating simulation and evaluation of tentative ones.
SIMULATION OPTIMIZATION OF GRAZING MANAGEMENT STRATEGIES
2001
In farming systems research, simulation is a common investigation tool that enables to study the dynamic behavior of production systems in response to climatic factors and more or less sophisticated management strategies. SEPATOU is one such a simulator that reproduces the functioning of a rotational grazing dairy system. This paper considers simulation optimization as a means to derive the best values of some parameters involved in a management strategy. We evaluate a stochastic optimization algorithm, the Kiefer-Wolfowitz method, based on a stochastic approximation of the objective function gradient. It appears that this approach is reliable. However the algorithm requires a delicate parametrization that is specific to each application.
Repositioning the forage base for dairy production in a volatile and risky climate
Climate change projections for warmer and possibly drier conditions will impact on the productivity of pasture based dairy systems. Biophysical modelling tools, such as DairyMod and APSIM, provide a means to assess the impact of a range of climatic changes on pasture systems and develop resilient forage systems. The influence of projected climatic changes will vary according to existing climate and amount of change likely at a site. For example, increased pasture production was modelled in a cool temperate environment at Elliott (NW Tasmania) while lower production was modelled in a temperate climate at Ellinbank (SW Victoria) where there was a tendency for higher winter and early spring growth rates but with a shorter spring growing season. Incorporating deeper rooted and heat tolerant plant traits were shown to be effective in moderating the production decline at Ellinbank. In irrigated regions, supplemental autumn and spring irrigation of annual ryegrass based pastures was shown to be a more efficient use of irrigation water than summer irrigation of perennial ryegrass pastures. 'Double' and 'triple' forage cropping systems based on the heat tolerant and water use efficient maize plant were also effective at increasing DM production per ML irrigation applied. Pointbased biophysical modelling of forage systems can highlight the biophysical limitation of the systems but need to be put into a farm system context to fully evaluate the impacts on financial performance. Linking the outputs of biophysical models with farming system tools can help to bridge this gap.
Introduction: decision support for rotational grazing
In agriculture as a production industry it is striking to see how profitability varies from one farm to the other just because of the differences in management skills of the farmers. The substantial technical evolution of agriculture as well the ongoing changes of the socio-economic context have even more sharpened this peculiarity. Consequently higher importance has been given to the abili-ty of making as wise as possible decisions both in terms of strategic choices and day-to-day management practices. Usual management aspects concerning for instance accounting, and marketing have been addressed for long time by economists and software systems developed for such purpose are now commercially available. Much less has been done on providing decision support in optimi-zing or improving the technical efficiency of agricultural production system although this could be of great benefit for particular ones such as a dairy cow farm as considered in this paper.
The farmers' difficulties in technical management are particularly acute in dairy production systems that rely as much as possible on grassland feeding resource which is used through rotational grazing and completed with maize and concentrate feeding in winter time when the herbage mass is still insufficient for grazing. In the late winter to early summer period the system must switch progressively from a fully maize-concentrate feeding to a predominantly or fully herbage-based feeding. Such a delicate transition requires a coherent and robust conditional plan (also called a strategy) in order to keep the milk production at its optimal level despite the uncontrollability of some important factors such as weather. Basically the decision problem consists in finding for the above five-month period an appropriate combination of the following main commitments: the set of fields definitely allocated to grazing, the set of fields set aside to cope with weather deviation and grazed only if necessary, the profile of maize-concentrate distribution over the whole period, the fertilization policy and the field rotation policy. A combination is appropriate if it ensures an optimal production of milk over the whole period for a sufficiently representative range of climatic conditions. This problem is a complex one because it involves a multivariable optimization: one has to decide on the above issues so that a good quantity/quality tradeoff of the available herbage is maintained along the considered period given that the maize distribution profile can only be non-increasing and the grass growth rate is partially controllable by the fertilization but also partially uncontrollable due to the climatic influence. Some agronomists' results, mainly coming from studies on continuous grazing, have shown that in order to have herbage of good quality it is necessary that the grazing intensity be high and regular on rotational periodicity. The problem of strategic management of a herd feeding based predominantly on rotational grazing (henceforth we shall talk simply of rotational grazing management) has to be solved once every year because the stock of maize at the start of the period varies from one year to the other and the size and characteristics of the herd may change too. From an economical point of view, it is important to solve the problem properly because the herbage resource is much cheaper than the maize-concentrate one.
The management of rotational grazing is a difficult problem for farmers because they lack sufficient capabilities of predicting the dynamics of growth and quality of grass and because the problem intrinsically involves a complex planning dimension in which several fields have to be dealt with concurrently given that the elaboration of the resource grass and the process of its use strongly interfere. For all these reasons we have setup a project for constructing a decision support tool that could provide help in constructing a rotational grazing strategy for the period considered. So far the project has concentrated on a preliminary step that consisted in clarifying the understanding of what a strategy might be and how it could be characterized and represented concretely. Simultaneously we have started to develop a software enabling us to simulate the application of such a strategy on a particular configuration of the production system under an assumed climatic scenario.
With respect to most of the simulation tools developed in the agricultural domain the one developed in this project is rather novel 1 on its ability to reproduce the continuous interaction between a decision system and a biophysical system. The first infers what should be done on the basis of the chosen strategy, the perceivable state of the biophysical system and observed data about the external environment (the climate). The biophysical systems reproduce the dynamics of the herbage production, animal intake and milk production under the influence of the external environment and the farmer's actions. The usually reported simulation systems (e.g. Topp & Doyle, 1996) only have a biophysical component or have a rather crude decision system that simply applies a pre-specified sequence of actions rather than a strategy (i.e. a conditional plan).
A model of the biophysical system
The dairy production system can be seen as a biophysical system guided by a decision system (discussed in the next section). The biophysical system, which is sketched here (see Duru et al. 1996 for a complete description), represents through empirical laws structured on a daily basis the dynamics of several interactive subsystems dealing with herbage production, cow intake and milk production. The driving variables include, for climate, the average daily temperature and average incident solar radiation and, for management interventions, the nitrogen level, grazing operation, cutting operation and amount of maize-concentrate 2 feeding.
The pasture is divided into a certain number of fields having different sizes but producing the same kind of forage crop. The herbage production submodel comprises four state variables: leaf area index (LAI), nitrogen index (NI), dry matter (DM in g/m 2 ), and average digestibility (also refereed to as quality) of the sward (AD in %). The dynamics of NI is represented by piecewise linear function linking the levels corresponding to fertilization operations. For a given field at a given date, the LAI, DM and AD may vary slightly differently depending on whether the last completed operation was a cutting or a grazing or whether the current day is within a grazing operation. In the first case, the model reproduces the growth and senescence processes. The LAI is initially 0 and depends on temperature and NI, growth (newly created DM) depends on LAI and solar energy and senescence applies to a proportion of dry matter that has reached a certain age expressed in degree-days. The AD variable decreases with age (again expressed in degree-days) rather slowly before 200 degree-days and faster afterwards; between 200 and 600 degree-days a loss of 12% of average digestibility is assumed. After a grazing operation the only change to made concern the LAI which must be computed with respect to residual LAI at the end of the grazing operation and the AD which is obtained by combining the average digestibility of the dry matter created since the end of the grazing operation and the diminishing impact of the average digestibi-lity of the dry matter that remained after the grazing of the field. During a grazing operation the computation of the DM must incorporate the amount grazed by the herd. The LAI is assumed to be proportional to the remaining DM and the AD is kept at the value it has when the grazing operation started. The latter assumption is acceptable because a grazing operation lasts three days at most. The average digestibility describes the field as a whole but the average digestibility of the cow intake depends on the depth of the sward that is grazed; the upper part of the sward is more digestible therefore the average digestibility of animal intake decreases every day during a grazing operation on a given field. The model uses a function describing the variation of digestibility along the herbage dry matter scale, thus enabling us to represent its variation along the depth axis. The physical properties and the spatial distribution of forage within the field are assumed to be homogeneous over each field (even during or after a grazing operation).
The intake process is controlled by three factors: the dry matter availability, the physiological limit on intake and the physical ability of the cows to consume feed. The actual intake on a given day is determined by the most limiting of these factors. It is assumed that the cows do not graze very close to the ground and most specifically do not graze below a dry matter of 120g/m 2 . The herd is represented by a set dairy cows that are assumed identical with respect to production characteristics (genetic type, age, weight, start of lactation,...). Since the herd may partially be fed with maize and concentrate the availability of dry matter is the sum of the maize-concentrate complement and herbage dry matter above the 120g/m 2 threshold. The physiological limit of intake is determined by the daily metabolizable energy requirements required by a cow which includes the part needed for maintenance and the part devoted to milk production. The required energy of maintenance depends only on whether or not the turnout to herbage has already occur-red. The part of energy that corresponds to the maximal potential milk production has its highest value during the first 40 days after calving and then decreases every day at a constant rate. The physical ability of the cow to consume feed is determined by its average digestibility which, for herbage, is computed as the integral of the digestibility variation function between the states describing the dry matter before and after grazing, divided by the difference between these two states. The amount of herbage dry matter intake is the maximum value compatible with the diges-tibility constraint and the energy requirement not fulfilled by the maize-concentrate complement.
The milk production is simply calculated on the basis of energy available for milk production that comes from the overall energy intake minus the amount required for maintenance.
Characterizing and simulating a strategy
In many production system (e.g. cash cropping) it is rather easy, even for farmers, to formulate the strategy of technical management that they intend to follow. This is usually not the case for rotational grazing partly because the management problem is much more complex and also because it has not been studied extensively by agronomers.
Basically a management strategy for an agricultural production system is a temporal structure that specifies, according to global objectives and constraints, a coherent sequence of states and actions that ensures the reach of the objective under a reasonable hypothesis concerning the uncontrollable variables (i.e. the climate): this could be called a nominal plan. A strategy must also incorporate the necessary elements to adapt the nominal plan in case a significant deviation with respect to the normally anticipated evolution of the production system has occurred.
On this basis we have crafted a framework for expressing a rotational grazing strategy. Essentially the late-winter to early-summer period has been divided into a sequence of four episodes corresponding to end-of-winter (before turnout to grass), early-spring (before the deadline for the first grazing on a field), full-spring (before the fields cut for silage become usable for grazing) and remaining sub-period. Within each episode we have identified a specific set of activities to consider and the constraints on these activities (e.g. priority among conflicting activities, precedence). An activity expresses under what conditions and how, depending of the current situation, a particular type of action (e.g. maize feeding) should be performed. In order to insure sufficient robustness of a strategy, the decision-making components of each episode includes rules of adaptation of the activities and constraints. These rules are used to monitor the important deviations of the biophysical system (due to the climate) and specify the required modifications on the originally intended course of actions. For instance, the grazing activity in the third episode specifies for any day the set of candidate fields to be considered for grazing for the rest of the episode and the criteria to chose one among those presently qualified for a grazing operation. The set of candidate fields is augmented (resp. reduced) by one field among those set aside as buffer fields whenever a shortage (resp. a excess) of herbage is anticipated. The way of selecting the field to graze on a particular day may be defined by a hierarchy of criteria specifying, for example, that the preference goes to field grazed the previous day, then (if the preceding criterion is not sufficiently discriminatory) to a field not initially belonging to the buffer set and finally to the one ranking first on dry matter quantity.
The daily application of such a strategy amounts to find out what is the current episode, check if some adaptations are necessary, trigger the activities that are applicable and, finally, execute them in the right order with instances corresponding to the current context. For example, at the end of the second episode a cutting activity is planned to make some field grazable again in the next one ; it applies to all fields that have not been grazed at least once (including those set aside as buffer) and that will never be otherwise due to the poor quality of herbage on them.
The process of supporting the strategy elaboration task by using a simulation-based approach that works as follows. A man-made strategy is provided and then simulations of its application in various climatic conditions are run. If the outputs meet acceptably the objectives with sufficient certainty (i.e. with a certainty above the lower bound specified by the farmer as its risk attitude), then an appropriate (sufficiently robust) strategy has been found. Otherwise a different one should be formulated and the simulation process restarted until a satisfactory solution is reached.
Conclusion
The simulator briefly described here is still under development; the software design relies on an object-oriented representation with the C++ programming language. Clearly the most interesting aspect of such a tool is in the coupling of a decision process and a biophysical system, allowing us to perform virtual experimentation of the functioning of the dairy production system without oversimplifying its decision-making component that in reality plays a key role. At the conceptual level some fundamental strutural aspects underlying the concept of strategy have been identified but further work is required in order to define a general language supporting a rigorous and intelligible representation of strategies.
The complexity of the production system precludes yet the possibility of elaborating a strategy by an automatic configuration program, however, we plan to study the possibility of providing more help than what the present simulator can do. An idea would be to perform a generation on the basis of a cruder model, simple enough to lend itself to a kind of reversal process enabling the determination of a strategy on the basis of desired features, possible futures and results of previous runs of the simulator presented in this paper.