Project Topics
Get Complete Project Material File(s) Now! »
Environmental Impact Assessment Tools
The first group we identified is related to the environmental impact assessment of agro-food systems. These tools have a clear focus on nutrient losses from agricultural land to the natural environment. These approaches often use nutrient emission indicators that we classified into two categories, depending on the way they estimate impacts. These categories are “real nutrient flow indicators” when nutrient flows and production processes are co-located; and “virtual nutrient flow indicators” when nutrient flows and production processes are spatially disconnected.
Real nutrient flow indicators
In this category, we aggregated different approaches aimed at quantifying nutrient losses from agricultural lands to the environment, e.g., to estimate nitrate leaching to water bodies or nitrous oxide (N2O) emissions to the atmosphere. These approaches have a strong biophysical background and are clearly focused on agro-ecosystem processes, whereas they do not generally take economic agents other than farms into account. Most of these indicators are based on simple statistical calculations – such as N2O emissions as a linear or exponential function of nitrogen fertilizer supplied (Philibert et al., 2012) – but some of them are based on more sophisticated models that incorporate several natural and anthropogenic drivers such as soil, climate and farming practices (Brisson et al., 2003). Many of those indicators have been developed at the field scale, but some upscaling has in fact been attempted to address larger systems such as districts and regions. Among them, we highlighted two widespread approaches that appear to be relevant to assess nutrient flows: environmental risk mapping and integrated assessment of agricultural systems.
The environmental risk mapping approach aims to establish pollution risk maps, generally for one specific pollutant. It uses geographical information systems to superimpose several sets of spatial information underlying the environmental risk, e.g., soil, climate, elevation, land-use, farming practices, etc. (Lahr and Kooistra, 2010). Several cases exist in the literature concerning nutrients, including soil phosphorus fertility depletion (Brown et al., 2000), nitrogen flows within landscapes (Theobald et al., 2004) and nitrate leaching risk (Assimakopoulos et al., 2003). This approach is adapted when risk mapping can be accurately assessed by just superimposing information about biophysical context, land-use and management practices. However, it does not account for complex interactions in management practices, nor for the role of agents and their social organization in that pollution.
The integrated assessment of agricultural systems aims to estimate agricultural impacts on the environment within a multi-criteria perspective. While different impacts are considered, nutrient losses to the environment are almost always considered. In contrast with the previous approach, nutrient losses are often considered in a sophisticated way by accounting for multiple farming practices and their potential interactions. Although there is a wide variety of tools that could be used (Bockstaller et al., 2009), most of them bring together a set of agro-environmental indicators. The INDIGO® tool, for example, is a wide-ranging set of indicators capable of evaluating farm sustainability based on farming practices such as crop rotation, fertilization, irrigation and pesticide use (Bockstaller and Girardin, 2006). Ultimately, these indicators can serve as multi-criteria decision-making tools that support farmers’ groups and policy-makers (Sadok et al., 2009). Most of these tools have been designed at the field or farm scale, whereas very few have been designed at the district scale. In addition, this approach is usually based on local or multi-local data, but is rarely spatialised. Payraudeau and van der Werf (2005) provided one of the very few reviews of the application of those tools at different regional scales, including examples at the local scale. They concluded that different indicators of nutrient losses to the environment could be identified, but the clarification of their position in the cause-effect chain related to farming practices is strongly recommended (Figure 5).
Integrated Crop-Livestock System Analysis
Integrated crop-livestock system analysis aims to represent and simulate nutrient flows through material exchanges among complementary farming enterprises. When applied to the district scale, they focus on flows between livestock and arable farms through feedstuff (e.g., cereals and straw) and organic fertilizers (e.g., animal manure and crop residues) (Thornton and Herrero, 2001). Although these approaches focus on organic materials in agricultural contexts, their system representation may be similar to Industrial symbiosis analysis since both represent detailed flows between partners at the local scale. Nevertheless, crop-livestock system analysis often uses more complex approaches such as mechanistic or dynamic models to derive flows among system components. For example, they may include crop models to simulate crop growth response to fertilizer application or livestock models to simulate manure production in response to livestock feed.
Crop-livestock system analyzes have been used, for example, to assess farming system sustainability (Bonaudo et al., 2014), farming diversification (Sulc and Tracy, 2007), and resource use efficiency (Cortez-Arriola et al., 2014). These methods have been widely applied to improve African smallholder crop-livestock systems in terms of soil carbon, nitrogen and phosphorus management (Alvarez et al., 2014; Barrett et al., 2002; Bouwman et al., 2013; Giller et al., 2011; Manlay et al., 2004) and can also be spatialised (Bellassen et al., 2010). Major progress on nutrient analysis and modelling has been made within the NUANCES Project (Nutrient Use in Animal and Cropping systems – Efficiencies and Scales), where not only actual nutrient flows were analyzed, but also dynamically simulated under different scenarios (Rufino et al., 2011, 2007; Tittonell et al., 2010, 2009a; van Wijk et al., 2009). In these examples, complex crop-livestock models were applied to study nutrient flows among African smallholder farmers and the environment. Many crop-livestock models have been applied in developing countries where access to chemical fertilizers is limited and waste management is, instead, a key issue. In most cases, farm typologies are a prerequisite to identify and model the different agents and their material exchange strategies according to their farming systems (Figure 8).
Entities, state variables and scales
FAN was developed for application to local or sub-regional case studies (e.g., ≈1,000 km2). Although many agent-based models have been developed for applications to specific geographical contexts, FAN can be adapted for its use in a variety of rural and agricultural case studies where core input data is available. Key requirements include data on land use, livestock numbers and feed rations, as well as crop yields and fertilization rates. Here, we conduct sensitivity analysis by applying FAN to an agricultural case study for a sub-region of southwestern France with about 835 farms (Ribéracois district in the department of Dordogne, ~1,000 km2; see Section 3.1). Although we carried out surveys of major farm partners such as feed and food collectors and food industries in order to guide the development of FAN, the scale and number of farms is large enough to make unfeasible gather surveys of farmers individually. For this reason, FAN is able to generate synthetic populations of agents based on general data collected in public statistics providing land-use and livestock in farms.
In FAN, points and vectors are used to represent agents and flows, rather than as raster grid cells that are common in agent-based models (e.g., Grimm et al., 2005; Rebaudo et al., 2011). As a result, agent attributes, such as the land use of each farm, are stored as attributes. The model is therefore able to take into account the geographic location of the agents (represented as points), flows between agents (represented as vectors), and to distribute these points according to specific addresses or randomly. Simulation of larger and smaller areas is also possible with FAN, as well as the use of different numbers of agents.
The agent population is composed of eight main types of agents, as represented by the agent class diagram in Figure 11. They include agricultural production (farms, feed and forage collectors), food industries (milk & cheese industries, slaughterhouses, and fruits & vegetables industries), waste managers (anaerobic digesters, wastewater treatment plants), and fertilizer wholesalers. We additionally use an intermediary conceptual agent (‘Partner’, white box in Figure 11) as a modelling tool to allow these agents to simultaneously demand and supply materials on each round of exchanges. Although farms are classified into eight different functional groups according to their characteristics (in terms of area, land-use and livestock numbers), all farm agents are capable of a set of defined actions while their annual production is based on their individual agent attributes. For instance, during a simulation, all farms conduct the action of animal feeding at the same time (regardless of the farm type), but farms that do not have animals remain unchanged by this action.
Process overview and scheduling
Since the core purpose of FAN is to simulate local-scale material flows in agro-food networks, here we present how the related processes are scheduled and what rules govern the types and magnitude of these flows. The steps of the model represent different production and exchange activities during a simulation cycle, equivalent to a year (Figure 13).
The simulation cycle begins with the production of fertilizing materials such as manure, digestates and sewage by farms, anaerobic digesters and wastewater treatment plants. These fertilizing materials are then applied to agricultural soils according to each farm’s nutrient demand. Chemical fertilizers can also be used by farms. Excess manure may be saved for bioenergy production by anaerobic digesters. Subsequently, each farm’s crop production is calculated as a function of fertilizing material inputs to soils through a simple linear yield-response model. Animal feed requirements are estimated according to species-specific feed demand, and crop products are exchanged in order to meet these animal requirements. Once feed and forage requirements are satisfied, livestock production is calculated.
After total crop and animal production has been computed, fruits, vegetables, and animal products are exchanged with local food industries, where they are processed, generating processed food and food wastes. Finally, once food wastes have been exchanged with livestock farms and anaerobic digesters, the latter results in bioenergy production. Note that global markets can create competition with local materials (e.g., imported chemical fertilizers can compete with local manures for fertilizing soils) or can compensate local production deficits (e.g., in feedstuff and forage to meet animal requirements).
Table of contents :
Chapter 1 Abstract
1.1. Introduction
1.2. Environmental Impact Assessment Tools
1.3. Stock and Flow Analysis Methods
1.4. Agent-Based Models
1.5. Discussion
1.6. Conclusion and perspectives
Chapter 2 Abstract
2.1. Introduction
2.2. FAN Model Overview
2.3. Process overview and scheduling
2.4. Design Concepts
2.5. Submodels
2.6. Input Data and Initialization
2.7. Comments on FAN processes
2.8. FAN Model Exploration
2.9. Conclusion
Chapter 3 Abstract
3.1. Introduction
3.2. Materials and Methods
3.3. Results
3.4. Discussion and Conclusion
Discussion and Conclusion
Our approach in reference to the state of the art
Scientific fields and terminologies
Case study issues, scale and data collection
Building a material flows agent-based model
The FAN model discussion
Scenarios simulation and assessment
FAN application and improvement perspectives
Conclusion
References
