Hayo M.G. van der Werf

Evaluation of the environmental impact of agriculture at the farm level: a comparison and analysis of 12 indicator-based methods

Abstract An increasing variety of evaluation methods is being proposed to address the question of the environmental impacts of agriculture. This paper compares and analyses 12 indicator-based approaches to assessing environmental impact at the farm level, in order to propose a set of guidelines for the evaluation or development of such methods. These methods take into account a number of environmental objectives (e.g. soil erosion, water quality). A set of indicators is used to quantify the degree to which these objectives are attained. A total of 26 objectives were taken into account by one or several of the methods. A great diversity in breadth of analysis exists: the number of objectives considered per method varies from 2 to 13. Indicator-based methods for environmental evaluation at the farm level should take into account a range of objectives covering both local and global effects. Indicators based on the environmental effects of farmer practices are preferable to indicators based on farmer practices as the link with the objective is direct and the choice of means is left to the farmer. Indicators based on farmer practices cost less in data collection but do not allow an actual evaluation of environmental impact. Indicators allowing expression of impacts both per unit surface and per unit product are preferable. Indicators producing output in the form of values are preferred to indicators producing scores. If possible, science-based threshold values should be defined for indicators. The method should be validated with respect to (a) the appropriateness of its set of objectives relative to its purpose and (b) its indicators.

Assessing the impact of pesticides on the environment

Pesticide use in agriculture can cause undesirable effects on humans and the natural environment. One of the objectives of integrated agriculture is the elimination or reduction of possible sources of environmental pollution such as pesticides. To achieve this objective, farmers need a method to assist them in estimating the environmental impact of pesticide use. This paper addresses a two-part question: what factors should be taken into consideration to assess pesticide environmental impact, and how can impact be quantified? As the environmental impact of a pesticide depends on its dispersion in the environment and on its toxicological properties, the literature on these topics is reviewed to address the first part of the question. To address the second part of the question, six recent approaches to assess the impact of pesticides on the environment are compared regarding choice, transformation and aggregation of input parameters. The use of simulation models to assess environmental impact is discussed.

An operational method for the evaluation of resource use and environmental impacts of dairy farms by life cycle assessment

This paper describes and applies EDEN-E, an operational method for the environmental evaluation of dairy farms based on the life cycle assessment (LCA) conceptual framework. EDEN-E requires a modest amount of data readily available on-farm, and thus can be used to assess a large number of farms at a reasonable cost. EDEN-E estimates farm resource use and pollutant emissions mostly at the farm scale, based on-farm-gate balances, amongst others. Resource use and emissions are interpreted in terms of potential impacts: eutrophication, acidification, climate change, terrestrial toxicity, non-renewable energy use and land occupation. The method distinguishes for each total impact a direct component (impacts on the farm site) and an indirect component (impacts associated with production and supply of inputs used). A group of 47 dairy farms (41 conventional and six organic) was evaluated. Expressed per 1000kg of fat-and-protein-corrected milk, total land occupation was significantly larger for organic than for conventional farms, while total impacts for eutrophication, acidification, climate change, terrestrial toxicity, and non-renewable energy use were not significantly different for the two production modes. When expressed per ha of land occupied all total impacts were significantly larger for conventional than organic farms. This study largely confirms previously published findings concerning the effect of production mode on impacts of dairy farms. However, it strikingly reveals that, for the set of farms examined, the contribution of production mode to overall inter-farm variability of impacts was minor relative to inter-farm variability within each of the two production modes examined. The mapping of impact variability through EDEN-E opens promising perspectives to move towards sustainable farming systems by identifying the structural and management characteristics of the farms presenting the lowest impacts.

Environmental Impact Assessment of Salmonid Feeds Using Life Cycle Assessment (LCA)

Abstract Understanding the environmental burdens associated with aquafeeds is a critical component for assessing and improving the environmental performance of aquaculture. The aim of the study was to assess the environmental impacts associated with feeds for rainbow trout production in France, using Life Cycle Assessment (LCA). The stages assessed are: the extraction of the raw materials, the production and transformation of the primary ingredients used, the manufacturing of the feeds, the use of the feeds at the farm, transport at all stages, and the production and use of energy resources. The assessment revealed that the use of fishery resources (such as biotic resource use) and nutrient emissions at the farm (such as eutrophication potential) contribute most to the potential environmental impacts of salmonid aquafeeds. Improvements in feed composition and management practices seem to be the best ways for improving the environmental profile of aquafeeds.

Variability in environmental impacts of Brazilian soybean according to crop production and transport scenarios.

Soybean production and its supply chain are highly dependent on inputs such as land, fertilizer, fuel, machines, pesticides and electricity. The expansion of this crop in Brazil in recent decades has generated concerns about its environmental impacts. To assess these impacts, two representative chains supplying soybeans to Europe were identified: Center West (CW) and Southern (SO) Brazil. Each supply chain was analyzed using Life Cycle Assessment methodology. We considered different levels of use of chemical and organic fertilizers, pesticides and machinery, different distances for transportation of inputs and different yield levels. Because transportation contributed strongly to environmental impacts, a detailed study was performed to identify the routes used to transport soybeans to seaports. Additionally, we considered different levels of land occupation and land transformation to represent the impact of deforestation in the CW region. Environmental impacts were calculated for 1000 kg of soybean up to and including the delivery to Europe at the seaport in Rotterdam, at 13% humidity. Overall results showed that the impacts are greater for CW than for SO for all impact categories studied, including acidification (7.7 and 5.3 kg SO(2) eq., respectively), climate change (959 and 510 kg CO(2) eq.), cumulative energy demand (12,634 and 6,999 MJ) and terrestrial ecotoxicity (4.9 and 3.1 kg 1,4-DCB eq.), except eutrophication and land occupation. The same trend was observed for the crop-production stage. Efforts to reduce chemical fertilizers and diesel consumption can reduce CO(2) emissions. Although deforestation for crop production has decreased in recent years, the contribution of deforestation to climate change and cumulative energy demand remains significant. In the CW scenario deforestation contributed 29% to climate change and 20% to cumulative energy demand. Results also showed that although there are different transportation options in Brazil, the current predominance of road transport causes severe environmental impacts. In CW, road transport contributed 19% to climate change and 24% to cumulative energy demand, while in SO it contributed 12% and 15% to these impacts, respectively. Improvements in the logistics of transportation, giving priority to rail and river transports over road transport, can contribute significantly to reducing greenhouse gas emissions and decreasing energy use. Future studies involving Brazilian soybeans should take into account the region of origin as different levels of environmental impact are predicted.

Life Cycle Analysis of field production of fibre hemp, the effect of production practices on environmental impacts

SummaryLife Cycle Assessment (LCA) was used to assess the environmental impacts of field production of fibre hemp and seven other crops in France. The production of 1 ha of hemp yielded a eutrophication potential of 20.5 kg PO4-equivalents, a global warming potential of 2330 kg CO2-equivalents, an acidification potential of 9.8 kg SO2-equivalents, a terrestrial ecotoxicity potential of 2.3 kg 1,4-dichlorobenzene-equivalents, an energy use of 11.4 GJ, and a land use of 1.02 ha.year. A comparison of hemp (low impacts), wheat (intermediate impacts) and sugar beet (high impacts) revealed that the crops were similar for the relative contributions of emitted substances and resources used to impacts, and for the relative contribution of processes to impacts. A reduction of the impacts of hemp production should focus on eutrophication, and consider the reduction of climate change, acidification and energy use as secondary objectives. Given this objective, the overall environmental effect of the substitution of mineral fertiliser by pig slurry is negative. The introduction of reduced tillage is of interest, as it decreases energy use, acidification and climate change. Measures leading to a reduction in NO3 leaching are highly interesting, as they strongly decrease eutrophication. Implications for hemp breeding are discussed.

Perception of the environmental impacts of current and alternative modes of pig production by stakeholder groups.

The current industrial pig production model is in crisis, due to its association with environmental pollution, doubtful product quality and lack of animal well-being. In Bretagne (France), a region of intensive pig production, a survey of seven stakeholder groups concerned with pig production was conducted, as part of a research programme dedicated to the assessment of the environmental impact of different modes of pig production. A very large majority of pig producers (93%) and their suppliers (100%) considers pig farms as an asset for the region, whereas a majority of scientists (58%), activists (78%) and consumers (54%) sees it as a handicap. Differences among stakeholder groups are minor with respect to the perceived importance of environmental and social issues. Stakeholders agree on the relative level of responsibility of pig farms with respect to specific problems. For all groups unpleasant odours and water quality come first with respect to responsibility, for most groups soil quality comes second, followed by product safety and air quality. For a future improved mode of pig production, 76% of pig producers and their suppliers prefer to adapt the current model, for all other groups the majority prefers an alternative model. While pig producers and their suppliers prefer a slurry-based housing system, all other groups prefer a straw-based system. Pig producers see the slurry-based system as technically superior and associate the straw-based system with poor working conditions, whereas consumers associate the slurry-based system in the first place with poor water quality and associate the straw-based system with less pollution. These results will be of use in the research programme on the environmental impact of modes of pig production, as they indicate the environmental impacts to be considered and their relative importance. The results will also help in deciding which options should be assessed. It is concluded that the poor image of the current pig production model and its production practices with consumers does not seem compatible with a sustainable demand for pork products.

Adaptation and use of a fuzzy expert system to assess the environmental effect of pesticides applied to field crops.

Abstract Centre d’Etude pour un Developpement Agricole Plus Autonome (CEDAPA) is a group of farmers in Brittany (western France) which proposes a mainly pasture-based milk and meat production system defined by a set of prohibitions and obligations. CEDAPA farmers claim that their production system causes less environmental damage than the current systems used by most other farmers. With respect to pesticide effects this claim was examined by using an expert system (Ipest), which reflects the authors’ perception of the potential environmental effect of the application of a pesticide to a field crop. Ipest requires three types of input variables: pesticide properties, site-specific conditions and characteristics of the pesticide application. It yields four output variables: one reflecting the rate of application of the pesticide, the other three reflecting the risk for three major environmental compartments (groundwater, surface water and air). In order to fit the hydrogeological conditions of Brittany and to decrease dependence on expert judgement, the Ipest system was modified and enhanced. An existing method for the estimation of runoff risk from the field was included. This new version of the expert system, Ipest-B, was used to compare the environmental effect of pesticide use on winter wheat, silage maize and fodder beet, which overall was largely similar. However, within each of the three crops, large differences occurred from pesticide characteristics, number of treatments and environmental conditions. This type of analysis helps to identify and recommend the least harmful treatment programmes for each crop. The Ipest system was shown to be flexible, and adaptable to the particularities of Brittany. An adaptation of the system to other regional conditions should be possible.

Development of a soil compaction indicator in life cycle assessment

PurposeIntegrating soil quality impacts in life cycle assessment (LCA) requires a global approach to assess impacts on soil quality that can be adapted to individual soil and climate contexts. We have developed a framework for quantifying indicators of impact on soil quality, valid for all soil and climate conditions, and considering both on-site and off-site agricultural soils. Herein, we present one of the framework’s impact indicators, which has not yet been quantified in detail in LCA studies: soil compaction.Material and methodsThe method includes guidelines and tools for estimating midpoint compaction impacts in topsoil and subsoil as a loss of soil pore volume (in cubic metre per functional unit). The life cycle inventory (LCI) and life cycle impact assessment are based on simulation modelling, using models simple enough for use by non-experts, general enough to be parameterised with available data at a global scale and already validated. Data must be as site specific and accurate as possible, but if measured data are missing, the method has a standardised framework of rules and recommendations for estimating or finding them. The main model used, COMPSOIL, predicts compaction due to agricultural traffic. Results are illustrated using a case study involving several crops in different soil and climate conditions: a representative pig feed produced in Brittany, France.Results and discussionPredicted compaction impacts result from the combination of site-specific soil, climate and management characteristics. The data necessary to the LCI are readily available from free soil and climate databases and research online. Results are consistent with compaction observed in the field. Within a soil type, predictions are most sensitive to initial bulk density and soil water content.ConclusionsThe method lays the foundation for possible improvement by refining estimates of initial soil conditions or adding models that are simple and robust enough to increase the method’s capacity and accuracy. The soil compaction indicator can be used in LCAs of bio-based materials and of waste management stages that consider composting. The framework includes other operational indicators (i.e. water erosion, soil organic matter change) to assess impact on soil quality. They complement other impact categories, providing increased ability to identify “impact swapping”.

Sensitivity Analysis of Environmental Process Modeling in a Life Cycle Context: A Case Study of Hemp Crop Production

The aim of this article is to develop a methodological approach allowing to assess the influence of parameters of one or more elementary processes in the foreground system, on the outcomes of a life cycle assessment (LCA) study. From this perspective, the method must be able to: (1) include foreground process modeling in order to avoid the assumption of proportionality between inventory data and reference flows; (2) quantify influences of foreground processes’ parameters (and, possibly, interactions between parameters); and (3) identify trends (either increasing or decreasing) for each parameter on each indicator in order to determine the most favorable direction for parametric variation. These objectives can be reached by combining foreground system modeling, a set of two different sensitivity analysis methods (each one providing different and complementary information), and LCA. The proposed method is applied to a case study of hemp‐based insulation materials for buildings. The present study will focus on the agricultural stage as a foreground system and as a first step encompassing the entire life cycle. A set of technological recommendations were identified for hemp farmers in order to reduce the crops environmental impacts (from –11% to –89% according to the considered impact category). One of the main limitations of the approach is the need for a detailed model of the foreground process. Further, the method is, at present, rather time‐consuming. However, it offers long‐term advantages given that the higher level of model detail adds robustness to the LCA results.